Plant Soil (2009) 316:25–34 DOI 10.1007/s11104-008-9755-5

REGULAR ARTICLE

Genotypic differences in root hydraulic conductance of rice (Oryza sativa L.) in response to water regimes Naoki Matsuo & Kiyoshi Ozawa & Toshihiro Mochizuki

Received: 29 May 2008 / Accepted: 12 August 2008 / Published online: 9 September 2008 # Springer Science + Business Media B.V. 2008

Abstract To determine water uptake by rice in watersaving culture, we examined root hydraulic conductance (L0), plant growth, and root anatomy of three rice genotypes (Oryza sativa L. ssp. indica cv. Beodien, traditional upland; ssp. japonica cv. Sensho, traditional upland; ssp. japonica cv. Koshihikari, improved lowland) under three water regimes: water-saturated (hydroponic), well-irrigated aerobic (control), and water-saving aerobic in soil. In hydroponic culture, although shoot dry weight (SDW) and root number were the largest in Sensho, root L0 was the highest in Koshihikari. There was no significant

relationship between root L0 and SDW in hydroponics, so root L0 might not limit shoot growth under flooding. Root L0 was much less in soil than in hydroponics, and that of Koshihikari was the lowest, especially in water-saving conditions. Root L0 was highly correlated with SDW under water-saving conditions but not in the control, so root L0 limits shoot growth under repeated water stress. Root anatomy was less affected by water regime than root L0 and is genetically controlled. Thus, root L0 may be more affected by water channels than by root anatomy.

Responsible Editor: Hans Lambers.

Keywords Hydroponic culture . Lowland rice . Root hydraulic conductance . Soil culture . Upland rice . Water-saving culture

N. Matsuo (*) Graduate School of Bioresource and Bioenvironmental Sciences, Kyushu University, 111 Harumachi, Kasuya-cho, Kasuya-gun, Fukuoka 811-2307, Japan e-mail: [email protected] K. Ozawa Tropical Agriculture Research Front, Japan International Research Center for Agricultural Sciences, 1091-1, Maezato-Kawarabaru, Ishigaki, Okinawa 907-0002, Japan T. Mochizuki Faculty of Agriculture, Kyushu University, 111 Harumachi, Kasuya-cho, Kasuya-gun, Fukuoka 811-2307, Japan

Introduction In light of the current environmental problems caused by the use of irrigation in agriculture, improvements in water use efficiency, water productivity, and the sustainability of water consumption have become priority research areas. The scarcity of water is also threatening the sustainable production of rice in Asia, notably northern China and parts of Southeast Asia, where rice is the most important cereal crop (Bouman and Tuong 2001). A new development in watersaving technologies is the concept of aerobic rice (Bouman 2001; Bouman et al. 2005). In aerobic rice

26

systems, fields remain unsaturated throughout the growing season, similar to wheat and maize cultivation. Therefore, the soil in aerobic rice systems undergoes repeated cycles of wet and dry conditions. Matsuo et al. (2007a) demonstrated that when upland cultivars were used, more than 70% of irrigation water could be saved compared with flooded paddy conditions, without growth or yield penalties, whereas the growth and yield of lowland cultivars decreased greatly under the same systems. Although many studies have reported on drought resistance in rice, there has been little study of the physiological and morphological traits of rice plants under aerobic rice systems with repeated water stress. Trillo and Fernández (2005) assessed the response of wheat plants to repetitive water stress in pot experiments. Both root hydraulic conductivity (Lpr) and shoot growth were strongly reduced as water stress became severe. Root Lpr of maize plants grown in hydroponic culture was inhibited by water stress induced by polyethylene glycol (Mu et al. 2006). Root Lpr and /or root hydraulic conductance (L0) is one of the major parameters reflecting root water uptake ability and is closely correlated with plant water relations under both normal and stressed conditions. However, in almost all studies of root Lpr of cereals, it was measured while the roots were held in nutrient solution or water. In the field, however, cereals are grown in soil, but few studies have examined root L0 of plants in soil. Therefore, it is unclear whether reported values of root L0 reflect water uptake ability under actual field conditions. In the case of rice, there have been some reports of root Lpr in hydroponic culture (Miyamoto et al. 2001; Ranathunge et al. 2003, 2004, 2005), but little difference was detected between lowland and upland genotypes. Furthermore, there are no studies of the water uptake ability or root L0 of rice plants subjected to repeated water stress. The objectives of this study were to analyze genotypic differences in root L0, plant growth, and root anatomy in the response to different water regimes. We measured root L0 in hydroponic culture by using a pressure chamber method commonly used for root L0 determination, and in soil to reflect actual field conditions. We discuss the relationship between plant growth and root L0 and whether root L0 could be a promising trait for breeding rice cultivars adapted to aerobic rice systems.

Plant Soil (2009) 316:25–34

Materials and methods Plant materials We grew three rice (Oryza sativa L.) genotypes: an indica traditional upland cultivar, Beodien; a japonica traditional upland cultivar, Sensho; and a japonica improved lowland cultivar, Koshihikari. We selected them on the basis of our previous study (Matsuo et al. 2007b) of drought resistance at flowering. Beodien is drought-resistant, Sensho is mildly drought-resistant, and Koshihikari is drought-susceptible. The first two grew well under aerobic rice systems but the last did not (Matsuo et al. 2007a). Experimental design We examined root L0, plant growth, and root anatomy in hydroponic and soil culture. Experiments were conducted in a greenhouse at Japan International Research Center for Agricultural Science, Okinawa, Japan in 2007. Since circadian rhythm has a large effect on root L0 (Carvajal et al. 1996), we used fluorescent lamps (∼2.5 μmol s−1 m−2) during night time to minimize the effect. Hydroponic culture Seeds were pre-soaked at room temperature for 2 days and then sown in seed beds on July 1. At 21 days after sowing (DAS), seedlings were transferred to a hydroponic culture system consisting of containers (60×45×15 cm) with 30 L of nutrient solution, each of which accommodated six plants (two plants per genotype in each). The containers were buried in soil to minimize the effect of root temperature on L0 (Radin 1990). The solution contained the following components (mg L−1): N, 150; P, 52; K, 340; Mg, 36; Mn, 1.5; B, 0.59; Ca, 160; Fe, 3.8; Cu, 0.038; Zn, 0.11; Mo, 0.039 (pH 5.5–6.0). It was completely renewed every 7 days and aeration was not provided. At 72 DAS, four plants with similar plant height (PH) and tiller number (TN) of each genotype were selected for root L0 measurements and were transferred to the laboratory, where they were grown in cylindrical PVC pots (height, 35 cm; ∅, 15 cm) filled with nutrient solution. The plants were left at rest for 2 days to acclimatize at 25°C under fluorescent lamps (∼8.5 μmol s−1 m−2).

Plant Soil (2009) 316:25–34

Soil culture Five pre-germinated seeds were sown on May 24 in the same type of cylindrical PVC pot as used for hydroponic culture, filled with sandy clay loam (clay, 19.7%; silt, 11.0%; sand, 69.3%) at a bulk density of 1.42 Mg m−3 to within 5 cm of the rim. Before sowing, a chemical fertilizer (14% N, 14% P, 14% K) was mixed with the soil at a rate of 2.5 g per pot, and the pot contents were irrigated with 1,650 mL tap water. Volumetric water contents of the soil at field capacity (remaining moisture 24 h after irrigation) and the permanent wilting point (−1.5 MPa) were 28.3% and 6.5%, respectively. The permanent wilting point of the soil was determined with a Dewpoint Potentiameter (WP4; Decagon Devices, Pullman, WA, USA). Thirty-two pots were arranged in a 22-cm grid in a large container (90×180×30 cm), which was buried in the soil and filled with tap water for the same reason as above. At 18 DAS, seedlings were thinned to one per pot, and two water regimes were commenced, control and water-saving. In the control, three monitor pots of each genotype were weighed every 3 to 5 days (depending on weather and plant size) on an electric balance, and the average amount of evapotranspired water was supplied to all pots with respect to each genotype (∼1,000 mL). In the watersaving treatment, each pot was weighed when leaf rolling did not recover; the amount of evapotranspired water was recorded, and the same amount of water was supplied. The recovery was observed after dusk. The frequency of irrigation depended on individual pots. In the water-saving treatment, one pot of each genotype was designated as a monitor plant, and the soil was collected from the surface to 30 cm depth with a gauge auger set (30 mm ∅, DIK-104A, Daiki Rika Kogyo, Saitama, Japan) just before each irrigation. The sample was divided into three depths (0–10, 10–20, 20–30 cm) and, after drying at 105°C for 24 h, volumetric soil water content was determined. The same amount of soil was replaced. The pots were arranged in a randomized complete block design with 16 replications for each genotype by water regime. In total, 96 pots and three containers were used in soil culture. At 87 DAS, four plants with similar PH and TN were selected for root L0 measurements, and pots were transferred to the laboratory, where the pots were irrigated to saturation to enable the plants to regain

27

turgor. Plants were acclimatized as in hydroponic culture. During acclimation period, pots were kept saturated with tap water so that the equilibrium of ionic composition and pH of the soil was achieved and air inside the soil was replaced by tap water. Measurements Root L0 measurements were carried out as described by Miyamoto et al. (2001), with some modifications, from 74 to 75 DAS for hydroponic culture and from 89 to 91 DAS for soil culture. During measurements of soil-grown plants, the soil was saturated with tap water. Secondary tillers were cut off with a razor blade at approximately 30 mm from the base. Each pot was placed in a pressure chamber, where the main tiller was threaded through a rubber stopper sealed with silicone sealant (Fig. 1). Cotton wool was attached to the cut surface and covered with plastic film to prevent evaporation. The air in the chamber was pressurized, and the pressure was continuously monitored with a pressure transducer and adjusted (±1.0 kPa). In advance, we had determined the relationship between pressure and exuded xylem sap. The relationship was nonlinear in the range between 0 and 30 kPa and linear thereafter (data not shown). Therefore, we selected 30 and 70 kPa for determination of exuded xylem sap. At each pressure, when the flow rate had stabilized (∼5 min), exuded xylem sap (W, g) was collected at the cut surface for 15 min on the cotton wool that was preweighed first. The cotton wool was weighed on an electric balance with a sensitivity of 0.1 mg. The root system was then washed carefully with tap water and dried thoroughly on paper towels, root number (RN) was counted, and

Rubber band Main stem

Plastic film Cotton Rubber stopper Cut tillers Pressure chamber

Pot

Pressurized air

Air outlet

Fig. 1 A schematic diagram of the pressure chamber used in this study

28

Plant Soil (2009) 316:25–34

root fresh weight (RFW, g) was determined. The flow rate Jv (g g−1 RFW s−1) was calculated as:

Plant growth

Jv ¼ W =ðRFW
sÞ: Root L0 (g g as:

−1

RFW kPa

Results

−1

−1

s ) was then calculated

L0 ¼ Jv =ΔP; where ΔP is the pressure difference between 30 and 70 kPa. The root system was separated into three sections (0–10 cm, 10–20 cm, and below 20 cm). Twenty nodal roots in the uppermost soil layer were sampled at 95–100 mm from the root base (= 5 mm long) and fixed with FAA (5% formalin, 5% acetic acid, 90% alcohol) for anatomical observation. Ten medium-size nodal roots of 20 samples were selected, and transverse cross-sections (100–120 μm thick) were cut on a plant microtome (MTH-1, Nippon Medical & Chemical Instruments Co. Ltd., Osaka, Japan), and then stained with safranin (0.5% w/v). The sections were viewed under a light microscope (Optiphot, Nikon Co. Ltd., Tokyo, Japan) and photographed with a digital camera (Coolpix-950, Nikon). The photographs were analyzed with Image J v. 1.38 software (http://rsb.info.nih.gov/ij/) to determine xylem vessel number (XN), cross-sectional area (CA), stele area (SA), total xylem vessel area (TXA), and the ratio of SA to CA. The rest of the root system and the shoots were oven-dried at 70°C for at least 3 days to determine root dry weight (RDW) in each section and shoot dry weight (SDW). Statistical analyses Because the plant growth duration and nutrient status in hydroponic culture were different from those in soil culture, we used different statistical tests in Unistat v. 5.6 software (Unistat Ltd., London, UK). For hydroponic culture, we used the LSD test (P<0.05) to detect genotypic differences in plant growth parameters, root L0 and root anatomical traits. For soil culture, we analyzed the effects of genotype and water regime on these parameters by a two-way ANOVA to determine which factors influenced them, then the LSD test (P<0.05) to determine genotypic differences in each treatment.

Growth parameters are summarized in Table 1. In hydroponic culture, there were significant differences among genotypes in SDW, RN, and RDW. The SDW and RN of Sensho were higher than those of the other two genotypes, and the RFW of Sensho was higher than that of Koshihikari. In soil culture, the total amount of evapotranspired water during the growth period was 10.7 kg in Beodien, 12.3 kg in Sensho, and 12.4 kg in Koshihikari under control conditions, and 7.8, 6.5, and 5.6 kg under water-saving conditions (data not shown). The ratio of the latter to the former was 72.9% in Beodien, 52.8% in Sensho, and 45.2% in Koshihikari. Among the three shoot growth parameters measured, the SDW was the most affected by the water regime in soil culture: Koshihikari was the most susceptible to water deficit, and Beodien was the least (ratio of water-saving SDW to control SDW was 31.7% and 63.6%, respectively). Water regime had larger effects on RDW than on RN. The RDW of Koshihikari was the most affected by water regime, and was significantly lower than those of the other two genotypes under water-saving conditions (RDW ratio: 64.1% in Beodien, 46.4% in Sensho, and 21.5% in Koshihikari). Two-way ANOVA indicated a significant main effect of both water regime and genotype on all parameters. It also revealed a genotype by water regime interaction in all traits except in PH. In hydroponic culture, root distribution was the greatest at 0–10 cm in all genotypes (Table 2). Sensho and Koshihikari had almost no roots below 20 cm, and the root distribution below 10 cm was significantly lower than that in Beodien. In both control and water-saving conditions, Koshihikari had a significantly higher proportion of roots at 0–10 cm than the other genotypes and lower below 20 cm, indicating its shallower root system in soil culture. This caused that more soil water was remained at 20–30 cm depth in Koshihikari just before irrigation than in the other two genotypes (data not shown). Root hydraulic conductance In hydroponic culture, the root L0 value of Koshihikari was approximately twice that of the other two genotypes, and significantly higher than that of

Plant Soil (2009) 316:25–34

29

Table 1 Plant height (PH), tiller number (TN), shoot dry weight (SDW), root number (RN), and root dry weight (RDW) of each genotype in three water regimes (mean±SE, n=4)

Hydroponic culture Beodien Sensho Koshihikari Soil culture Control Beodien Sensho Koshihikari Water-saving treatment Beodien Sensho Koshihikari ANOVA Genotype Water regime G×W

PH (cm)

TN (plant−1)

SDW (g plant−1)

RN (plant−1)

RDW (g)

81±4 aa 87±4 a 82±1 a

8.5±1.6 a 13.5±2.4 a 7.7±0.9 a

5.8±1.0 b 13.5±2.1 a 6.7±0.5 b

134±18 b 269±27 a 181±12 b

0.65±0.10 ab 0.85±0.12 a 0.40±0.12 b

106±7 a 100±4 ab 87±2 b

13.5±0.3 b 12.8±0.5 b 17.5±0.3 a

16.4±0.4 b 22.5±1.3 a 23.0±0.6 a

123±9 c 242±7 b 283±5 a

2.62±0.22 b 3.45±0.25 a 2.63±0.16 b

90±1 a 92±2 a 83±2 b

8.0±0.4 a 9.3±0.5 a 7.5±1.0 a

10.1±0.3 a 9.6±0.6 a 7.3±0.7 b

99±11 b 160±22 a 146±9 ab

1.68±0.36 a 1.60±0.17 a 0.78±0.09 b

7.3** 10.0** 1.5 ns

5.7** 192.5*** 17.7***

7.8** 393.3*** 22.3***

44.1*** 68.6*** 11.4***

9.9** 105.5*** 4.1*

Factors influencing these traits are also expressed as F-values for soil culture ns Not significant *P<0.05; **P<0.01; ***P<0.001 a

Means within a column followed by the same letter are not significantly different at P<0.05 by LSD test

saving to control L0 (136%), and Koshihikari had the lowest (34.8%). Two-way ANOVA of root L 0 indicated a significant (P < 0.01) main effect of genotype and a genotype by water regime interaction but not of water regime. No significant correlation was detected between root L0 and SDW in hydroponic culture (r=−0.53)

Sensho (Table 3). In soil, Beodien had the highest root L0 under both conditions, followed by Sensho. Under control conditions, a significant difference was detected between Beodien and Koshihikari. Under water-saving conditions, the root L0 of Beodien was significantly higher than those of the other two genotypes. Beodien had the highest ratio of water-

Table 2 Root dry weight (RDW) and its distribution at three depths of each genotype in three water regimes (mean±SE, n=4) RDW (mg plant−1) 0–10 cm Hydroponic culture Beodien 435±95 ba Sensho 795±114 a Koshihikari 338±57 b Soil culture Control Beodien 809±43 b Sensho 1,157±56 a Koshihikari 1,095±44 a Water-saving treatment Beodien 641±101 ab Sensho 671±70 a Koshihikari 421±54 b a

Distribution (%) 10–20 cm

>20 cm

0–10 cm

10–20 cm

>20 cm

166±21 a 88±17 b 66±7 b

51±8 a 1±0 b 0b

65.4±4.2 b 89.1±1.9 a 82.6±2.8 a

26.4±3.1 a 10.8±1.8 b 17.4±2.8 b

8.2±1.4 a 0.2±0.1 b 0b

716±54 b 965±48 a 754±15 b

1,047±136 ab 1,324±166 a 783±135 b

31.2±1.3 b 33.9±1.8 b 41.9±1.8 a

29.2±0.7 a 28.1±0.7 a 29.0±2.1 a

39.6±1.9 a 38.0±2.2 a 29.1±3.6 b

544±46 a 488±69 a 268±21 b

495±40 a 440±47 a 88±31 b

37.6±1.7 b 42.1±2.2 b 54.2±2.6 a

32.6±1.2 a 30.1±1.2 a 35.2±2.7 a

29.8±1.6 a 27.7±2.4 a 10.5±2.9 b

Means within a column followed by the same letter are not significantly different at P<0.05 by LSD test

30

Plant Soil (2009) 316:25–34

Table 3 Root hydraulic conductance (L0) of each genotype in three water regimes (mean±SE, n=4)

Root anatomy

L0 (×10−8 g g−1 RFW kPa−1 s−1) 26.4±7.8 aba 21.1±4.0 b 44.4±7.4 a

2.28±0.31 a 1.78±0.30 ab 1.38±0.23 b 3.11±0.52 a 1.21±0.29 b 0.48±0.23 b

Discussion

15.0** 0.6 ns 3.9*

Some detailed results of rice root Lpr in hydroponic culture have been reported (Miyamoto et al. 2001; Ranathunge et al. 2003, 2004, 2005). Miyamoto et al. (2001) grew young plants (31–40 days old) of a lowland cultivar, IR 64, and an upland cultivar, Azucena, in hydroponic culture, but found no significant differences in total or individual root Lpr between the genotypes. In the present study, however, the root L0 of lowland Koshihikari in hydroponic culture was twice higher than those of upland cultivars, indicating a difference in root L0 among genotypes at an advanced stage (74–75 DAS) (Table 3). We calculated root L0 on RFW basis and

Factors influencing these traits are also expressed as F-values for soil culture ns Not significant *P<0.05; **P<0.01 a

Means within a column followed by the same letter are not significantly different at P<0.05 by LSD test

and control soil culture (r=−0.48) (Fig. 2). On the other hand, root L0 under water-saving conditions was positively correlated with SDW (r=0.86***) and the total amount of evapotranspired water (r=0.75**).

30 Beodien Sensho Koshihikari

15

25 SDW (g plant-1)

SDW (g plant-1)

20

10 5

(a) 2

0

6

4

20 15 10

(b)

5

0

0

8

0

15

2

3

4

5

10

12 9 r = 0.86***

6 3 0

1

L0 (×10-8 g g-1 RFW kPa-1 s-1)

L0 (×10-7 g g-1 RFW kPa-1 s-1)

SDW (g plant-1)

Fig. 2 Relationship between root hydraulic conductance (L0) and shoot dry weight (SDW) in hydroponic culture (a), control soil (b), and water-saving soil (c) and the relationship between L0 and total water consumed during growth in water-saving soil (d). Each dot represents individual plant. ***P<0.001; **P<0.01

Total water consumed (kg)

Hydroponic culture Beodien Sensho Koshihikari Soil culture Control Beodien Sensho Koshihikari Water-saving treatment Beodien Sensho Koshihikari ANOVA Genotype Water regime G×W

In hydroponic culture, values of XN, CA, SA, and TXA were the highest in Beodien, followed by Sensho. The SA/CA values of Beodien and Sensho were similar and significantly higher than that of Koshihikari. In soil culture, results were almost identical. Two-way ANOVA revealed a significant (P <0.001) main effect of genotype on all root anatomical traits (Table 4), and a significant effect of water regime on SA and CA. No genotype by water regime interaction was detected in any traits.

(c) 0

1

3

2 -8

-1

4 -1 -1

L0 (×10 g g RFW kPa s )

5

8 6

r = 0.75**

4 2 0

(d) 0

1

2 -8

3 -1

4 -1 -1

L0 (×10 g g RFW kPa s )

5

Plant Soil (2009) 316:25–34

31

Table 4 Xylem vessel number (XN), cross-sectional area (CA), stele area (SA), total xylem vessel area (TXA), and ratio of SA to CA (SA/CA) of each genotype in three water regimes (mean±SE, n=4)

Hydroponic culture Beodien Sensho Koshihikari Soil culture Control Beodien Sensho Koshihikari Water-saving treatment Beodien Sensho Koshihikari ANOVA Genotype Water regime G×W

XN (n)

CA (×10−1 mm2)

SA (×10−2 mm)

TXA (×10−3 mm2)

SA/CA (%)

7.3±0.3 aa 6.8±0.5 a 4.9±0.6 b

6.3±0.1 a 4.3±0.2 b 3.4±0.1 c

6.1±0.2 a 4.1±0.6 b 2.2±0.4 c

14.9±0.3 a 9.6±0.4 b 4.5±1.2 c

9.7±0.4 a 9.5±1.2 a 6.6±0.9 b

7.8±0.2 a 6.6±0.2 b 4.9±0.2 c

7.0±0.2 a 5.5±0.2 b 5.1±0.3 c

7.3±0.3 a 5.5±0.1 b 2.7±0.1 c

13.9±1.0 a 9.0±0.5 b 5.2±.0.2 c

10.4±0.3 a 10.1±0.3 a 5.4±0.2 b

7.8±0.1 a 6.2±0.2 b 4.4±0.1 c

6.3±0.2 a 4.4±0.3 b 3.9±0.2 c

6.9±0.1 a 4.8±0.3 b 2.3±0.0 c

14.1±0.6 a 8.2±1.5 b 4.5±0.1 c

11.0±0.4 a 11.1±0.3 a 5.9±0.2 b

143.5*** 3.1 ns 1.2 ns

38.4*** 21.4*** 0.7 ns

323.8*** 11.9** 0.5 ns

130.2*** 0.6 ns 0.4 ns

207.2*** 8.2* 0.5 ns

Factors influencing these traits are also expressed as F-values for soil culture ns Not significant *P<0.05; **P<0.01; ***P<0.001 a

Means within a column followed by the same letter are not significantly different at P<0.05 by LSD test

RDW of Koshihikari was smaller than those of the other two genotypes (Table 1). Small root system might cause higher root L0 in Koshihikari. However, we found no significant correlations between root L0 and shoot growth in hydroponic culture (Fig. 2). Therefore, we can conclude that root L0 does not limit rice plant growth in water-saturated environments such as flooded paddies. As far as we know, this study is the first to measure the root L0 of rice in soil. Soil culture remarkably reduced root L0 compared with that in hydroponic culture. L0 values in hydroponic culture were 11.6, 11.9 and 32.2 times higher than those in control soil in Beodien, Sensho and Koshihikari, respectively, and 8.5, 17.4 and 92.5 times higher in water-saving treatment, respectively. This difference was probably due to the different media during growth (hydroponic vs. soil culture) and might be caused by two reasons. First, in general, if the plants were grown in hydroponic culture, few lateral roots emerged from crown roots, compared with the root systems of plants grown in soil culture. Judging from the relationship between RN and RDW (Table 1), this difference was clearly observed in this study (though not measured).

The difference of the number of lateral roots between hydroponic and soil culture caused the difference of final total root length between the two. Root hydraulic resistance should increase as root length increases. These results might cause the large difference of root L0 between hydroponic and soil culture. Second, soilroot resistance was existed in soil culture, but not in hydroponic culture. Soil-root resistance increases as soil dries (Tuzet et al. 2003). This resistance might consist of one of the effects of water stress and caused differences of root L0 between hydroponic and soil culture. The other factor might be media differences during measurements. Although media differences during measurements were not compared in the present study, the media in soil culture was essentially water, because pots used in soil culture had been saturated before root L0 measurements and the soil used in this study contained about 70% of sand. Therefore, the growing media could not have imposed high resistance. From these aspects, the effects of media difference on measured root L0 could be negligible. Therefore, our results could clearly reveal a genotypic difference in root L0 of rice plants in the response to water regime. Lowland Koshihikari had

32

the highest root L0 in hydroponic culture and the lowest in soil culture, which was greatly reduced by repeated water stress. In contrast, chronic water stress caused the smallest growth reduction in Beodien, the root L0 of which was increased by repeated water stress. These results indicate that root L0 is an important trait necessary to maximum plant growth under recurrent water stresses conditions. Under water-saving conditions in soil, genotypes showed no significant differences in soil moisture content at depths of between 0 and 20 cm just before irrigation, but Koshihikari had significantly more soil water below 20 cm than the other two genotypes (data not shown). This difference was probably due to the fact that Koshihikari had very few roots below 20 cm (88 mg and 10.5%, compared with the control: 783 mg and 29.1%; Table 2). The responses of shoot and root growth to the repeated water stress during growth differed markedly among the genotypes used. Under water-saving conditions, root L0 was well correlated with SDW and the amount of total evapotranspired water (Fig. 2), suggesting that root L0 is a limiting factor for shoot growth under repetitive water stress conditions. These two variables (SDW and total evapotranspired water) depended not only on root L0 but also on the hydraulic conductance of root-soil system, which can be much bigger and can change a lot, depending on the soil water conditions (Tuzet et al. 2003). The hydraulic resistance of root-soil systems should increase as soil dries. In the present study, however, root L0 was measured keeping root systems in saturated soil. Thus, the effects of soil drying process on root L0 could not be estimated. However, we clearly revealed that the effects of drought history on root L0 were different among cultivars and root L0 was highly correlated with SDW and total evapotranspired water, even though root L0 was measured keeping root systems in saturated soil. Repeated water stress might damage the root systems of Koshihikari, resulting in a small root L0, which in turn caused the low water uptake. Moreover, the reduction in root L0 might not meet the water demand from shoots (i.e., transpiration) and cause poor shoot and root growth by reducing photosynthesis. Water regime had no effect on XN and TXA (Table 4). This result indicates that these traits are stable to the environment and genetically controlled. SA was highly correlated with CA in each genotype (data not shown). Kondo et al. (2000) reported that

Plant Soil (2009) 316:25–34

root diameter was highly correlated with stele diameter, and both were larger in upland cultivars than in lowland cultivars. Our results are consistent with those. Two-way ANOVA revealed a significant main effect of water regime on root L0 and of genotype on root anatomical traits. Moreover, there was a genotype by water regime interaction in L0 but not in root anatomy. These results indicate that factors other than root anatomy influenced root L0. Ranathunge et al. (2004) reported that exodermal apoplastic barriers such as the Casparian bands and suberin lamellae are permeable to water, and that there is substantial apoplastic transport of water across the outer part of the root (i.e., from rhizodermis to cortical cell layer) in rice, even in the presence of the exodermal apoplastic barriers. Ranathunge et al. (2003) also reported that a comparison of detailed measurements of overall root Lpr with that of the outer part of the roots demonstrated that the endodermis/stele made the largest contribution to hydraulic resistance. Water transport within and between plant tissues uses both apoplastic and symplastic routes; therefore, a crucial proportion of the water molecules have to cross a number of cellular membranes (Quigley et al. 2002). Transport via the apoplastic route is driven by physical forces and is regulated mostly by differences in water potential between the soil, plant, and atmosphere (Munns 1993; Maurel 1997). According to the composite transport model (Steudle and Frensch 1996), this route is coupled with the transpiration rate. The fact that the total amount of evapotranspiration was lower in Koshihikari under water-saving conditions might be due to reduced water transport through the apoplastic route, which dominates water passage in the outer part of roots. However, the hydraulic resistance of the outer part of roots is much lower than that between endodermis and stele (as described above), where the symplastic route dominates the water transport. Therefore, the lower transpiration in Koshihikari due to lower root L0 in water-saving conditions might be caused by reduced water transport through the symplastic route. It is well known that the symplastic route is regulated by a family of water channel proteins called aquaporins (Amodeo et al. 1999; Johanson et al. 2001). Water channel activity can be affected by different factors such as temperature, salinity, nutrient deprivation, drought, diurnal rhythm, heavy metals, and oxygen content (Azaizeh and Steudle 1991; Henzler

Plant Soil (2009) 316:25–34

and Steudle 1995; Carvajal et al. 1996, 1999, 2000; Henzler et al. 1999; Murai-Hatano et al. 2008; North and Nobel 2000; Tournaire-Roux et al. 2003). Roles of aquaporins in responses to irrigation were also reviewed (Vandeleur et al. 2005). However, studies of the relationship between aquaporin activity and water stress in rice are limited, but Lian et al. (2004) reported that the gene encoding OsPIP1;3 or the production of OsPIP1;3, a water channel protein, was up-regulated early in upland rice (up to 4 h) in response to 20% polyethylene glycol treatment, but there was no significant change in lowland rice. In another paper (Lian et al. 2006), they reported that the production of OsPIP1;2, OsPIP2;1, and OsPIP2;5 was also up-regulated in upland rice under the same conditions. The activities of these water channel proteins might be high in the upland rice cultivars used in this study, especially in Beodien. Recently, Sakurai et al. (2005) identified 33 rice aquaporin genes, whose contribution to water uptake ability can now be studied. Further detailed study is needed to clarify the relationship between the expressions of aquaporin genes and root L0. In conclusion, this is the first report to compare the root L0 values of plants between soil and hydroponic culture. Our results clearly reveal large genotypic differences in root L0 in response to water regime, and that root L0 was highly correlated with SDW only under repeated water stress conditions. They identify root L0 as a promising trait for breeding new rice cultivars adapted to aerobic rice systems. Further, water regime had larger effects on root L0 than root anatomy. Therefore, we can conclude that factors other than root anatomy, such as aquaporins, greatly affect root L0. Further study is needed to analyze the relationship between aquaporins and root L0 under repeated water stress conditions.

References Amodeo G, Dorr R, Vallejo A, Atuka M, Parisi M (1999) Radial and axial water transport in sugar beet storage root. J Exp Bot 50:509–516 doi:10.1093/jexbot/50.333.509 Azaizeh H, Steudle E (1991) Effects of salinity on water transport of excised maize (Zea mays L.) roots. Plant Physiol 97:1136–1145 Bouman BAM (2001) Water-efficiency management strategies in rice production. Int Rice Res Notes 26:17–22 Bouman BAM, Tuong TP (2001) Field water management to save water and increase its productivity in irrigated rice.

33 Agr Water Manag 49:11–30 doi:10.1016/S0378-3774(00) 00128-1 Bouman BAM, Peng S, Castaneda AR, Visperas RM (2005) Yield and water use of irrigated tropical aerobic rice systems. Agr Water Manage 74:87–105 doi:10.1016/j. agwat.2004.11.007 Carvajal M, Cooke DT, Clarkson DT (1996) Responses of wheat plants to nutrient deprivation may involve the regulation of water-channel function. Planta 199:372–381 doi:10.1007/BF00195729 Carvajal M, Martinez V, Alcaraz CF (1999) Physiological function of water channels as affected by salinity in roots of paprika pepper. Physiol Plantarum 105:95–101 doi:10.1034/j.1399-3054.1999.105115.x Carvajal M, Cerda A, Martinez V (2000) Does calcium ameliorate the negative effect of NaCl on melon root water transport by regulating aquaporin activity? New Phytol 145:439–447 doi:10.1046/j.1469-8137.2000. 00593.x Henzler T, Steudle E (1995) Reversible closing of water channels in Chara internodes provides evidence for a composite transport model of the plasma membrane. J Exp Bot 46:199–209 doi:10.1093/jxb/46.2.199 Henzler T, Waterhouse RN, Smyth AJ, Carvajal M, Cooke DT, Schaffner AR et al (1999) Diurnal variations in hydraulic conductivity and root pressure can be correlated with the expression of putative aquaporins in the roots of Lotus japonicus. Planta 210:50–60 doi:10.1007/s004250050653 Johanson U, Karlsson M, Johanson I, Gustavsson S, Sjövall S, Fraysse L et al (2001) The complete set of genes encoding major intrinsic proteins in Arabidopsis provides a framework for a new nomenclature for major intrinsic proteins in plants. Plant Physiol 126:1358–1368 doi:10.1104/ pp.126.4.1358 Kondo M, Aguilar A, Abe J, Morita S (2000) Anatomy of nodal roots in tropical upland and lowland rice varieties. Plant Prod Sci 3:437–445 Lian HR, Yu X, Ye Q, Ding X, Kitagawa Y, Kwak SS et al (2004) The role of aquaporin RWC3 in drought avoidance in rice. Plant Cell Physiol 45:481–489 doi:10.1093/pcp/ pch058 Lian HL, Yu X, Lane D, Sin WN, Tang ZC, Su WA (2006) Upland rice and lowland rice exhibited different PIP expression under water deficit and ABA treatment. Cell Res 16:651–660 doi:10.1038/sj.cr.7310068 Matsuo N, Nhan, DQ, Mochizuki T (2007a) Effect of watersaving irrigation on rice yield and its water productivity. Jpn J Crop Sci 76: Extra issue 44–45 (in Japanese) Matsuo N, Nhan DQ, Mochizuki T (2007b) Effect of deep tillage on growth and yield of rice cultivars grown under water deficit. J Fac Agr Kyushu Univ 52:331–336 Maurel C (1997) Aquaporins and water permeability of plant membrane. Annu Rev Plant Physiol Plant Mol Biol 48:399–429 doi:10.1146/annurev.arplant.48.1.399 Miyamoto N, Steudle E, Hirasawa T, Lafitte R (2001) Hydraulic conductivity of rice roots. J Exp Bot 362:1835–1846 doi:10.1093/jexbot/52.362.1835 Mu Z, Zhang S, Zhang L, Liang A, Liang Z (2006) Hydraulic conductivity of whole root system is better than hydraulic conductivity of single root in correlation with the leaf water status of maize. Bot Stud (Taipei, Taiwan) 47:145–151

34 Munns R (1993) Physiological processes limiting plant growth in saline soils: some dogmas and hypotheses. Plant Cell Environ 16:15–24 doi:10.1111/j.1365-3040. 1993.tb00840.x Murai-Hatano M, Kuwagata T, Sakurai J, Nonami H, Ahamed A, Nagasuga K et al (2008) Effect of low root temperature on hydraulic conductivity of rice plants and the possible role of aquaporins. Plant Cell Physiol (in press) North GB, Nobel PS (2000) Heterogeneity in water availability alters cellular development and hydraulic conductivity along roots of a desert succulent. Ann Bot (Lond) 85:247– 255 doi:10.1006/anbo.1999.1026 Quigley F, Rosenberg JM, Shachar-Hill Y, Bohnert HJ (2002) From genome to function: The Arabidopsis aquaporins. Genome Biol 3:1–17 Radin JW (1990) Responses of transpiration and Hydraulic conductance to root temperature in nitrogen- and phosphorus-deficient cotton seedlings. Plant Physiol 92:855–857 Ranathunge K, Steudle E, Lafitte R (2003) Control of water uptake by rice (Oryza sativa L.): role of the outer part of the root. Planta 217:193–205 Ranathunge K, Kotula L, Steudle E, Lafitte R (2004) Water permeability and reflection coefficient of the outer part of young rice roots are differently affected by closure of water channels (aquaporins) or blockage of apoplastic pores. J Exp Bot 396:433–447 doi:10.1093/jxb/ erh041

Plant Soil (2009) 316:25–34 Ranathunge K, Steudle E, Lafitte R (2005) Blockage of apoplastic bypass-flow of water in rice roots by insoluble salt precipitates analogous to a Pfeffer cell. Plant Cell Environ 28:121–133 doi:10.1111/j.1365-3040.2004. 01245.x Sakurai J, Ishikawa F, Yamaguchi T, Uemura M, Maeshima M (2005) Identification of 33 rice aquaporin genes and analysis of their expression and function. Plant Cell Physiol 46:1568–1577 doi:10.1093/pcp/pci172 Steudle E, Frensch J (1996) Water transport in plants: role of the apoplast. Plant Soil 187:67–79 doi:10.1007/ BF00011658 Tournaire-Roux C, Sutka M, Javot H, Gout E, Gerbeau P, Luu D et al (2003) Cytosolic pH regulates root water transport during anoxic stress. Nature 425:393–397 doi:10.1038/nature01853 Trillo N, Fernández RJ (2005) Wheat plant hydraulic properties under prolonged experimental drought: Stronger decline in root-system conductance than in leaf area. Plant Soil 277:277–284 doi:10.1007/s11104-005-7493-5 Tuzet A, Perrier A, Leuning R (2003) A coupled model of stomatal conductance, photosynthesis and transpiration. Plant Cell Environ 26:1097–1116 doi:10.1046/j.13653040.2003.01035.x Vandeleur R, Niemietz C, Tilbrook J, Tyerman SD (2005) Roles of aquaporins in root responses to irrigation. Plant Soil 274:141–161 doi:10.1007/s11104-004-8070-z