Agronomy Journal Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kobata, T.
Right arrow Articles by Uemuki, N.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Kobata, T.
Right arrow Articles by Uemuki, N.
Agricola
Right arrow Articles by Kobata, T.
Right arrow Articles by Uemuki, N.
Related Collections
Right arrow Rice
Right arrow Seed Production
Published in Agron. J. 96:406-414 (2004).
© American Society of Agronomy
677 S. Segoe Rd., Madison, WI 53711 USA

SEED PRODUCTION

High Temperatures during the Grain-Filling Period Do Not Reduce the Potential Grain Dry Matter Increase of Rice

Tohru Kobata*,a and Naoya Uemukib

a Faculty of Life and Environ. Sci., Shimane Univ., 1060 Nisikawatu-cho, Matsue 690-8504, Japan
b Satake Co. 2-30 Saijo Nishihon-machi, Higashihiroshima, Hiroshima 739-8602, Japan

* Corresponding author (kobata{at}life.shimane-u.ac.jp).

Received for publication December 2, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
High temperatures during the grain-filling period (GFP) of rice (Oryza sativa L.) increase the grain dry matter increase rate (GIR), but this increase in GIR is insufficient to completely compensate for the concomitant reduced GFP, and as a result, grain yield decreases. The shortfall in GIR as temperatures increase has been believed to signify a reduction of the potential GIR as a sink capacity. However, we suspect that lack of assimilate supply to the grain, rather than the decreased potential GIR, lowers the GIR and causes reduced grain weight. Our objective was to determine if the grain weight could reach full potential under higher temperature conditions if assimilate supply during the GFP was sufficient to sustain the increased GIR. Rice was grown at three locations in western Japan over 3 yr. At one location, plots were covered with plastic film during the GFP to increase temperature. Spikelet filling percentages (F%) at maturity varied between 70 and 90% when mean temperatures ranged between 23 and 29°C during the GFP. When plots were thinned to half density during GFP, all F% were approximated by a single logistic equation based on accumulated temperature, with a ceiling of 90%. Hence, thinning can overcome the lower F%. These results suggest that potential GIR in rice is not reduced by high temperatures during the GFP. Yield reductions commonly associated with such conditions are likely due to the failure of assimilate supply to the grain to meet the requirements of the accelerated GIR.

Abbreviations: AT, accumulated temperature • DRW, dry matter increase rate of whole plant • F%, spikelet filling percentage(s) • GFP, grain-filling period • GIR, grain dry matter increase rate • PG, potential grain dry weight • PGIR, potential grain dry matter increase rate


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
IN DIVERSE CROPS, grain yield is affected by air temperature during the GFP (Yoshida, 1981; Evans, 1996; Egli 1998). In most cases, higher temperatures during GFP increase the GIR but shorten the duration of the GFP. This results in reduction of mean grain weight or in the percentage of ripened grains. This effect has been noted in rice (Sasaki, 1935; Nagato and Ebata, 1965; Sato and Inaba, 1976; Yoshida, 1981; Tashiro and Wardlaw, 1989), wheat (Triticum aestivum L.) (Wardlaw et al., 1980; Bhullar and Jenner, 1985; Tashiro and Wardlaw, 1989; Tashiro and Wardlaw, 1990; Guedira and Paulsen, 2002), barley (Hordeum vulgare L.) (Wallwork et al., 1998), maize (Zea mays L.) (Jones et al., 1981), and soybean [Glycine max (L.) Merr.] (Egli and Wardlaw, 1980; Gibson and Mullins, 1996). Lower temperature reduces the GIR, extends duration of the GFP, and delays grain maturation although moderate cool temperatures sometimes benefit grain yield (Yoshida, 1981; Nishiyama, 1985; Egli, 1998). The temperature range that affects the GIR and GFP duration differs between plant species (Egli, 1998; Tashiro and Wardlaw, 1989).

When grain growth was restricted by high temperature, the main reason for termination of GIR in rice (Inaba and Sato, 1976), wheat (Hawker and Jenner, 1993), and barley (Wallwork et al., 1998) was thought to result from disappearance of enzyme activity relating to starch synthesis of the grains. These results suggest that potential GIR (PGIR) in grains is controlled by decreased metabolic activity under high-temperature conditions, and hence assimilate supply to the grain is only part of the cause of low grain weight.

When rice plants are subjected to water stress or shading during the early GFP but are subsequently well irrigated or exposed to abundant radiation, final grain dry weights reach those of nonstressed controls (Kobata and Takami, 1981; Kobata et al., 2000; Kobata and Sugawara, 2000). These results showed that shortage of available assimilate during the early GFP in water-stressed or shaded conditions does not determine potential grain weight. Thus, potential patterns of grain growth rate are fairly stable under stress conditions during GFP. Based on previous studies, it is doubtful that high temperature alone defines PGIR and determines the final grain weight. The effect of assimilate supply to grains in determining grain weight at high temperature may be discounted in most cases. Past experiments (Sasaki, 1935; Nagato and Ebata, 1965; Yoshida, 1981; Tashiro and Wardlaw, 1989) contained no positive trials in which available assimilate was increased under the high-temperature condition. An experiment in which available assimilate was increased by rachis thinning or CO2 enrichment under high-temperature treatment yielded few positive effects in terms of grain dry matter increase (Sato and Inaba, 1976). In that case, it is not certain that assimilate supply during the GFP increased sufficiently to meet the requirements of increased GIR because it was not clearly shown if thinning or CO2 enrichment increased assimilate supply to attached grains during the principal GFP. It can thus be hypothesized that potential grain weight is not inevitably reduced by increased temperature during the GFP and that assimilate supply to the grain mainly determines grain weight. Restriction of movement of other grain components such as soluble protein or N to rice grains should not limit the GIR under high-temperature conditions because increased concentrations of such components in the straw is caused by termination of GIR in the early GFP (Sato and Inaba, 1973).

It is important to clarify the effects of high temperature on grain filling in rice and on the physiological processes affecting grain filling. High summer temperatures have been observed in Japan, especially during the past decade. West Japan has been particularly affected. At times, hot summer conditions have seriously damaged both rice quality and production (Terashima et al., 2001). Furthermore, future negative impacts on diverse crop production in high- and middle-latitude areas due to climate warming from the greenhouse effect are of concern (Angus, 1990; Evans, 1996).

Our objectives were to test what factor is the primary influence on PGIR and the duration of GFP in field-grown rice under different temperature conditions and to determine if the shortage of assimilate supply to grain under high temperatures during GFP reduces the GIR.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Plant Materials
Rice cultivar Koshihikari was grown in paddy fields in three locations. These were at the Shimane University Matsue experimental farm (35° N, 133° E; 17 m above sea level), Shimane Prefecture experimental farm at Akana (35° N, 132° E; 450 m above sea level), and Kyoto University Takatuki experimental farm in Osaka prefecture (34° N, 135° E; 10 m above sea level). The soil was silty clay loam (Typic Epiaquept) at Matsue and Takatuki and black silty loam (Aquandic Epiaquept) at Akana. Temperature conditions during the rice growth season differ substantially between the experimental sites. Mean temperatures from transplantation to flower initiation and from flower initiation to full heading date were highest at Takatuki (23.1 and 29.1°C, respectively) in 2001 and lowest (17.4 and 23.4°C, respectively) at Akana in 1999. Temperatures in Matsue were intermediate (20.0 to 20.7°C and 24.8 to 26.6°C) in 1999 and 2000. Seeds were sown in seedbeds (60 by 30 by 3 cm) containing seedling soil on 12 Apr. 1999 and 10 Apr. 2000 (Matsue), 9 Apr. 1999 (Akana), and 10 May 2001 (Takatuki), and seedlings were grown in a non–temperature-controlled greenhouse. Four-leaf stage seedlings were transplanted to paddy fields at the farms on 22 May 1999 and 10 May 2000 (Matsue), 30 Apr. 1999 (Akana), and 9 June 2001 (Takatuki). Seedlings at Matsue and Akana were planted in rows 0.30 m apart at a spacing of 0.15 m, whereas those at Takatuki were planted in rows 0.33 m apart at 0.18-m spacing. Fertilizer N [as (NH4)2SO4)], P (as CaHPO4), and K (as KCl) were applied at all three sites. Rates applied were 4 g N m–2, 8 g P m–2, and 4 g K m–2 at Matsue; 4 g N m–2, 10 g P m–2, and 5 g K m–2 at Akana; and 3 g N m–2, 8 g P m–2, and 3.5 g K m–2 at Takatuki. An additional 4, 2, and 3 g N m–2 and 4, 2.3, and 3.75 g K m–2 were applied 3 wk before heading at Matsue, Akana, and Takatuki, respectively. In the Akana experiment, an additional 2 g N m–2 and 2.3 g K m–2 were applied at the booting stage. Plant densities and amounts of fertilizer were determined according to conventional procedure at each location. The plants at Matsue, Akana, and Takatuki were grown in 10.5- by 13.5-, 10.0- by 30-, and 25- by 20-m field plots, respectively. Flooded conditions were maintained throughout the growing season.

High-Temperature Treatments
High-temperature treatment was applied to 2.1- by 2.8-m areas of the field plots at Matsue on the full heading date. Steel frames 1.5 m high were placed in the rice field. These were covered with high-transparency polyester sheets designed for greenhouse roofing (Six-Light, Taiyo-Kogyo Co., Tokyo). The polyester sheets covered the tops, and thin polyethylene sheets were loosely hung on the sides. Natural ventilation was permitted through a small hole 0.01 m in diameter in the center of each roof and under the bottom edge of the side sheets. No artificial ventilation or air circulation was made. The plastic canopy increased the interior mean temperature 1 to 4°C above the outside temperature. Carbon dioxide concentrations monitored with an infrared gas analyzer decreased by a maximum of 6.6% at noon on fine days compared with ambient air. A photon flux meter set under the films showed that the plastic canopy decreased photon flux density by a maximum of 14%.

Thinning Treatments
Half the plots in all locations and years were unthinned, and the remaining plots were thinned to every other plant to reduce the plant density by half for the duration of the GFP. The GIR in rice can reach maximum rate due to increased assimilation under improved radiant conditions if plant density is halved by thinning (Kobata and Moriwaki, 1990; Kobata et al., 2000).

Measurements and Analysis
Six neighboring hills were harvested from each replication every 9 to 11 d from the full heading date until 40 d after the full heading date. The full heading date is defined as the date at which 90% of tillers bear seed heads. The plants were divided into straw and spikelet, dried in an oven at 80°C for 48 h, and weighed. Plant dry weights were determined for three parts: grain, straw (leaf plus stem), and the whole plant (grain plus straw). Fifty days after the full heading date, 10 plants from each replication in the unthinned treatment were harvested and measured for their yield and yield components. Filled grains were selected by specific gravity method. To separate unfilled grains from filled grains, all spikelets were immersed in (NH4)2SO4 solution (1.06 x 103 kg m–3), and floated spikelets were removed. Selected spikelets were counted, oven-dried, and weighed. The husks were removed and oven-dried grain weights (brown rice) determined. Mean single-husk weight was estimated by dividing the spikelet number into weight differences between the spikelets and grain. Husk weight per hill was calculated by multiplying the mean husk weight by the number of spikelets. Grain weight was estimated by subtracting the husk weight from the spikelet weight. Changes and differences between fertile and sterile grain in the husk weight during the GFP were ignored (Kobata et al., 2000). Grain filling was measured by percentage of ripened grain and spikelet-filling percentage. Percentage of ripened grain is the ratio of the number of filled to total spikelet numbers where the number of filled grain is that selected by the gravitational method. Thus, incompletely filled grains are omitted from the percentage of ripened grain. Spikelet-filling percentage is a measure of grain growth capacity (Murayama, 1982; Tsukaguchi et al., 1996; Horie et al., 1997) and is defined as the ratio of the observed grain dry weight (G) to the potential grain dry weight (PG)

[1]
where PG is calculated by multiplying the spikelet number by the dry weight of a single fully ripened grain at harvest, as derived from seed selection by specific gravity (Murayama, 1982; Tsukaguchi et al., 1996). Even under favorable field conditions, some spikelets cannot ripen due to infertility at flowering, so F% is generally less than 100 (Matsushima, 1959; Murayama, 1982).

Ambient temperature was measured in an instrument shelter adjacent to the field site at 30-min intervals. Readings were recorded to a data logger and mean daily temperature over 24-h periods subsequently calculated. For the high-temperature treatments, chamber interior temperatures were also measured at 30-min intervals and recorded to data loggers. These were covered to guard against radiative heating effects and set from the edge of the chamber two-thirds of the distance along the diagonal from corner to corner.

The experimental design at each site was a randomized block with a split-plot arrangement (Matsue) or a randomized block (Akana and Takatuki). At Matsue, the main plots comprised two temperature treatments, with two densities as subplots. The treatments at Akana and Takatuki comprised two densities at each site. Each treatment was replicated three times. Density effects were compared using data pooled by site, year, and, in the case of Matsue, by temperature treatment for analysis as a 6 site-year-temperature x 2 density treatment factorial experiment with three replications. At Matsue the main effects of temperature and temperature x density treatment interaction were analyzed over years as a 2 yr x 2 temperature x 2 density treatment factorial experiment with three replications. Significant differences (0.05 level) were calculated from the analysis of variance. The least-squares method was used to fit a logistic curve to F% and accumulated temperature (AT) data, and the decision coefficient was calculated to assess the goodness of fit of the curve.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Grain Yield and Temperature
Mean grain yields in the unthinned and unenclosed plots at Matsue were 599 g m–2 in 1999 and 591 g m–2 in 2000 (Table 1). When the plants were covered by the canopy and grown under high-temperature conditions during the GFP, yields decreased significantly. In 1999, mean yield fell to 412 g m–2 (69% of the ambient temperature plot) and to 428 g m–2 (72%) in 2000. In both years, reductions in yield in enclosed plots were accompanied by significant decrease (–23%) in percentage of ripened grain. Thousand-grain weights decreased by only 3 to 6% relative to the ambient temperature plots. Thus, reduction of percentage of ripened grain during the high-temperature treatment led to the reduction in yield.


View this table:
[in this window]
[in a new window]
  		Table 1. Grain yield, yield components of rice cultivar Koshihikari, and mean temperature during the grain-filling period (GFP) in three different locations. In Matsue, a part of rice crop was covered with a plastic canopy to increase temperature (high-temperature treatment, H).

 
Mean daily maximum temperatures in the high-temperature treatment plots were 35°C (1999) and 33°C (2000), 4 and 2°C greater than the maximum ambient temperatures in the respective years (Fig. 1) . Under ambient conditions, mean daily temperatures between 0 and 10 d after the full heading date at Matsue were 29 and 28°C in 2002 and 2001, respectively. The high-temperature treatments increased the mean temperature by 2.1 and 1.0°C above ambient in each respective year (Table 1). A temperature difference between the ambient and the high-temperature plots of 1 to 2°C was maintained in each 10-d interval from Day 10 until 30 d after the full heading date. The temperature differences between plots decreased with decreasing ambient temperature in 1999 but were maintained over the whole GFP in 2000 when ambient temperature did not fall (Fig. 1). As a result, the high-temperature treatment increased mean temperature during the 30 d of the GFP by 1.2 to 1.5°C. Therefore, the high-temperature treatment increased mean temperature and decreased grain yield significantly due to reduction of percentage of ripened grain.



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 1. Mean daily temperatures after full heading date under ambient and high-temperature (H) treatment conditions at Matsue in 1999 and 2000 and under ambient temperature at Akana (1999) and Takatuki (2001).

 
The grain yield of 669 g m–2 in the unthinned treatments at Akana (1999) was the greatest among all plots (Table 1). The yield at Takatuki in 2001 (574 g m–2) was the lowest under ambient temperature conditions among the three locations even though spikelet number was greatest. Although spikelet number at Akana was intermediate between at Takatuki and at Matsue, percentage of ripened grain and yield were the highest. Thus percentage of ripened grain, rather than spikelet number, strongly contributed to differences of grain yields. Thousand-grain weights were less affected by locations or high-temperature treatments than percentage of ripened grain. Consequently, percentage of ripened grain was the major contributor to differences in grain yield between experimental sites.

The temperature at Akana during the early GFP was much lower (4.2 to 7.2°C) than at the other locations (Fig. 1), and the mean temperature over the whole GFP at this site was also the lowest among all locations, being 3.8 to 4.7°C and 5.1 to 6.2°C lower than that of the Matsue ambient and high-temperature plots, respectively (Table 1). Mean temperature at Takatuki during the GFP was similar to that in Matsue in 1999. The low mean temperature at Akana increased percentage of ripened grain (Table 1). However, high temperature was not the only factor associated with a lower percentage of ripened grain; at Takatuki, the high spikelet number reduced percentage of ripened grain. Low percentage of ripened grain in rice carrying many spikelets is sometimes observed (Matsushima, 1959) because assimilate supply is insufficient to fill all grains. Therefore, high temperature during the GFP is a critical factor in reduction of grain ripening, but the effect of the temperature on grain ripening could be magnified by lower assimilate supply to grains.

The Effect of Thinning on Grain and Whole-Plant Dry Matter
After thinning, whole-plant dry weights in Matsue increased to 118% (1999) and 111% (2000) of the unthinned plots (Fig. 2) . Increases under the high-temperature treatment were similar to the ambient temperature plots (118% in 1999 and 115% in 2000). Grain dry weights in the thinned plots under ambient temperature increased to 115 and 105% of the normal density plots, and the increase under the high-temperature plots was 122 and 114% in 1999 and 2000, respectively (Fig. 2). Thus, thinning was effective in increasing grain weight under high-temperature conditions more than under ambient temperature conditions, and the effect of thinning on the whole-plant dry matter increase was similar in the ambient and high-temperature plots. In the high-temperature plots, whole-plant dry weight of the unthinned treatments was reduced by 5% on average over both years compared with the ambient temperature plants. Light interception by the plastic film and increased maintenance respiration loss due to high temperature (Thornley, 1976; Amthor, 1989) were considered reasons for the decrease. However, it is unknown which factor was the main cause of reduction in whole-plant dry weight.



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 2. Dry matter of the whole plant (parts above ground), grain, and straw per hill during the grain-filling period for normal-density (filled circles and solid lines) and thinned (open circles and dotted lines) plants. Plots are given for ambient and high-temperature (H) conditions at Matsue (1999 and 2000) and for ambient temperature conditions at Akana (1999) and Takatuki (2001). Data are the mean ± standard error of three replicates. Horizontal line in the straw plots indicates the straw dry weight at the full heading date.

 
Straw dry weights in the unthinned plants under ambient temperature conditions decreased after the full heading date and in each year reached a minimum 30 d after the full heading date (Fig. 2). In both years of high-temperature treatment, reduction in straw dry weight was greater than that under ambient conditions. Thinning alleviated reduction of straw dry weight after the full heading date under both ambient and high-temperature conditions, and during the late GFP, straw dry weight was almost the same as at the full heading date. At Akana, thinning increased the whole-plant and grain weights, but decrease in straw dry weight after full heading date was more than at Matsue (Fig. 2). In the case of Takatuki, whole-plant dry weight in both the normal-density and thinned plants was greater than that at the other locations, and straw weights did not decrease after the full heading date (Fig. 2). Therefore, in all cases, thinning increased the whole-plant dry matter so that the grain dry weight increased. Surplus assimilate not used by the grain was partitioned to straw dry weight or alleviated reduction of straw dry weight after the full heading date. These results suggest that in the thinned plants, assimilate was produced above the requirement for grain dry matter increase, even if the plants suffered high-temperature conditions, as in Matsue.

The main effect of thinning should be increased assimilation due to reduction in competition between plants. In rice, the increase of nutrient absorption by thinning is considered an indirect effect resulting from the increase of dry matter production (Takenaga, 1995). Furthermore, most macronutrient absorption by the plants has been finished before the GFP (Ishizuka and Tanaka, 1952; Yoshida, 1981).

The spikelet-filling percentage can show whether assimilate supply was sufficient for the grain dry matter increase. On unthinned plots, F% at the ambient temperature in Matsue were 80 and 90% in 1999 and 2000, respectively, and decreased to 70 and 85% under high-temperature conditions in the respective years (Fig. 3) . However, in both years, thinning of both the ambient and high-temperature plots increased their F% to almost 90%. This suggests that even under high-temperature conditions, a high ceiling F% can be attained if assimilate supply is enhanced by thinning. At Akana, there was little difference in F% both at maturity and at 10 d after full heading date between unthinned and thinned plots while F% on thinned plots was higher than on unthinned plots at 20 to 30 d after full heading date (Fig. 3). The lowest value of F% at 10 d after full heading was observed at Akana. Spikelet-filling percentage at maturity in the normal-density plots was lowest at Takatuki where increase in F% by thinning was also the highest among all locations. At both Akana and Takatuki, F% reached 90% when plant density was reduced, matching the figure at Matsue. Maximum F% in all thinned plants is thus very similar (90%) in all cases. The shortfall of 10% from the ideal reflects physiological infertility of the spikelets (Matsushima, 1959). Infertility during the GFP was not greatly increased in the high-temperature experiments. Percentages of infertile spikelets in grains harvested from the high-temperature treatment plots in 2000 reached 7% at most compared with 4% in the ambient temperature plots. The small treatment effect may be because the high-temperature treatment was not imposed until flowering had almost finished (90% of panicles appeared). Also, in many japonica rice cultivars, sterility increases significantly when temperatures exceed 35°C (Matsui et al., 2001).



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 3. Spikelet filling percentages (F%) for normal-density (filled circles and solid lines) and thinned (open circles and dotted lines) plants during the grain-filling period under ambient and high-temperature (H) conditions at Matsue (1999 and 2000) and under ambient temperature conditions at Akana (1999) and Takatuki (2001). See Eq. [1] in the text for F%. Data are the mean ± standard error of three replicates. Horizontal line indicates 90% of F%, which was attained by most of the thinned plants.

 
Relationship between Accumulated Temperature and Spikelet-Filling Percentage
In diverse plant species, the developmental stage or seed-filling rate is highly dependent on temperature (Sinclair 1994; Egli, 1998). When the mean daily temperature is T and the temperature at which GIR stops is Tbase, the AT (°C day) is indicated by

[2]
In Japanese rice, Tbase during the GFP is estimated to range from 7 to 8°C (Ebata, 1990). This is supported by observations in field conditions, which show that grain weight increase can continue for up to 3 wk at temperatures below 10°C, provided the plants do not suffer frost injury (Nishiyama, 1985). We therefore adopted 7°C as Tbase by referring to experimental data for cultivar Sasanishiki during the GFP (Ebata, 1990).

Although F% increased with increasing AT, the rate of increase in unthinned plots differed between sites (Fig. 4) . Increase in F% was approximated by a logistic equation for each data set.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 4. Spikelet filling percentages (F%) in the normal-density and the thinned plants vs. accumulated temperature (AT) after the full heading date. Data are for ambient and high-temperature (H) conditions at Matsue (1999 and 2000) and for ambient conditions at Akana (1999) and Takatuki (2001). For the normal-density plants, respective factors and coefficients for the fitted curves of the form F% = a/[1 + b exp(–cAT)] are a = 80, b = 80, c = 0.01, and r2 = 0.999 (Matsue 1999); 70, 75, 0.02, and 0.988 (Matsue H 1999); 80, 80, 0.02, and 0.996 (Akana 1999); 90, 75, 0.02, and 0.999 (Matsue 2000); 85, 80, 0.02, and 0.998 (Matsue H 2000); and 72, 70, 0.02, and 0.993 (Takatuki 2001).

 

[3]
where a was the F% ceiling for each set. The curves for each data set have very good fit (r2 = 0.996–0.999) (Fig. 4). Values for a clearly differ between all of the unthinned plots, with a range between 70 and 90%. In the 1999 and 2000 high-temperature treatments at Matsue, a was 5 to 10% lower than the values observed under ambient temperature conditions. The lowest a was observed under high-temperature treatment at Matsue (70%) and under ambient temperature condition at Takatuki (72%). When the plant densities were halved by thinning after full heading, the relationship between F% and AT in all data was approximated by a single logistic equation (r2 = 0.983) with a value for a of 90% (Fig. 4). Therefore, differences between the ambient and the high-temperature treatments in unthinned plots at Matsue in 1999 and 2000 were diminished if the plant densities were halved by thinning after full heading. Thinning treatment during the GFP should eliminate the effect of assimilate supply on the response of the F% to temperature. The F% response of thinned Koshihikari to AT was therefore almost stable, irrespective of temperature conditions, year, location, and cultivation procedure. This result is consistent with the hypothesis that the grain weight reaches its potential when the assimilate supply during GFP meets the requirements for realizing the increased GIR, even under different temperature conditions. Hence, the GIR in thinned plants can be considered as the PGIR.

Assessment of Effects of Temperature on Grain Dry Matter Increase Rate
A close uniform relationship between the F% and AT in thinned plants during full heading was observed in this study. Because the resulting equation was based on the GIR response at varying temperatures (23 and 35°C) during the GFP (Table 1 and Fig. 1), we can use this relationship to estimate the effect of temperature on the PGIR within this temperature range. The PGIR is given by

[4]
where RPF% is the rate of the potential F% (% °C –1). The RPF% in the thinned plants is obtained from the differential calculus (Milthorpe and Moorby, 1979) of F% (Eq. [3]) by

[5]

The effects of temperature increase or decrease during the GFP were assessed using the above equations. Changes in PGIR with daily temperature increases of 2°C (T + 2) or 4°C (T + 4) and decrease by 3°C (T – 3) were predicted using Eq. [4] and [5]. Daily mean temperature during the GFP (Fig. 1) at Matsue (1999 data) was used as the standard temperature (T + 0), and the spikelet number and mean dry weight of a single full-ripened grain (Table 1) was used as a source for calculation of PG. At T + 2 and T + 4, the maximum rate of PGIR increased to 111 and 120% that of T + 0, and the duration of the GFP was shortened by about 3 and 6 d, respectively (Fig. 5 —upper graph). Under T – 3, the maximum rate of PGIR decreased to 82% of that under T + 0, and the GFP was extended by about 4 d over that in T + 0.



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 5. Calculated potential grain dry matter increase rate (PGIR) after the full heading date and observed dry matter increase rate of the whole plant (DRW, solid line; DRW = 46.99 exp(–0.08D), r2 = 0.965 for Matsue in 1999 where D is the number of days after the full heading date] (upper graph), DRW – PGIR (middle), and cumulative DRW – PGIR (lower), when the mean temperature of each day increases by 2°C (T + 2) and 4°C (T + 4) or decreases by 3°C (T – 3) from the observed temperature (T + 0).

 
Current assimilation during GFP is one of the main sources for grain filling in rice (Matsushima, 1959; Yoshida, 1981; Kobata et al., 2000). The rate of dry matter increase of the whole plant (DRW) during the GFP in T + 0 was estimated based on an assumed current dry matter production rate. This rate was derived from the differential equation of the curvilinear relationship between observed whole-plant dry weight and the number of days after full heading for Matsue in 1999 (Fig. 5—upper graph). The difference between the DRW and PGIR (DRW – PGIR) suggests apparent shortage of the assimilate supply required for PGIR (Fig. 5—middle). The DRW in T + 0, T + 2, T + 4, and T – 3 was greater than the PGIR (DRW – PGIR > 0) within at least 1 wk after the full heading date. Therefore, the surplus of current assimilate over the requirements of the PGIR could be deposited into straw. In T + 2 and T + 4, the DRW – PGIR began to decrease 2 or 3 d earlier than in T + 0 or T – 3 and fell to a negative value. The negative value suggests shortening of the assimilate supply for the PGIR. At 22 or 23 d after the full heading date, the DRW – PGIR in T + 2 and T + 4 begins to increase again and reaches a zero or positive rate. In T + 0 or T – 3, positive rates are not regained until 24 or 30 d after full heading. Cumulative DRW – PGIR after the full heading date (Fig. 5—lower) suggests that the assimilate accumulated during the early GFP in T + 2 and T + 4 is completely consumed earlier than in T + 0 or T – 3. The largest decrease in straw weight (100 g m–2) in T + 4 is equivalent to 20% of the grain yield in T + 0 (Table 1). However, all of assimilate accumulated in straw during the early GFP may not be turned into grain dry matter increase during the middle GFP. Part of the assimilate in the straw may be used for the formation of new tillers, as has been shown for the late GFP (Jiang et al., 1988).

Rice straw contains some soluble assimilate reserves at the start of the GFP. This reserve plays a buffering role when the current assimilate supply falls below PGIR (Yoshida, 1981; Takami et al., 1990). Nonstructural carbohydrate contents of improved rice cultivars (including Koshihikari) at the heading period range between 0 and 200 g m–2 (Yoshida, 1981; Tsukaguchi et al., 1996; Horie et al., 1997; Kobata et al., 2000; Kobata and Sugawara, 2000). Not all of this assimilate reserve, however, can be used for grain filling. About 23 to 30% of the reserve is consumed by respiratory loss during GFP, even under favorable growth conditions (Kobata and Takami, 1986). High temperatures may also accelerate respiratory loss of assimilate through increase in maintenance respiration (Thornley 1976; Amthor, 1989). Over short time intervals, net photosynthesis rate scarcely changes below 33°C (Vong and Murata, 1977), but the DRW can be decreased by prolonged high temperatures through increase in maintenance respiration of nonphotosynthetic organs. Furthermore, other factors such as severe competition between plants, low radiation, and low nutrient conditions could increase the impact of high temperature on the GIR through reduction of assimilate supply.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Suppression in the capacity of GIR and shortening of the GFP by high temperatures have previously been thought to reduce grain weight in rice (Sato and Inaba, 1976; Tashiro and Wardlaw, 1989). According to this assumption, the GIR under high-temperature conditions cannot compensate for the shortened GFP that results. However, if plant population densities are reduced, increased GIR can compensate for the shortened GFP caused by high temperatures. The effect on the GIR of the metabolic process relating to the enzyme activity of starch synthesis has been considered to be one of the main reasons for reduction of PGIR under high-temperature conditions in rice (Inaba and Sato, 1976) and other grain crops (Hawker and Jenner, 1993; Wallwork et al., 1998). However, this effect may not be precisely evaluated, unless an adequate assimilate supply to the grain is assured. We conclude that the response of PGIR to high temperatures during the GFP is stable in field-grown rice. Consequently, there is a definite relationship between PGIR and the term of the GFP. Nevertheless, it not known if the stability of the PGIR at higher temperature exists during sudden heat shock or extremely high temperatures, such as are often seen in growth cabinet experiments. It is thought that reduced rice yields associated with high temperatures during the GFP mainly result from shortage of assimilate supply. Potential grain dry matter increase rate is sensitive to temperature and increases as temperature increases. Ultimately, the assimilate supply from current assimilation and from assimilate reserves in the straw cannot meet the increased demand of the PGIR. Our results show that an increase in mean temperature of several degrees during the GFP can have a serious impact on the grain yield of rice grown in a temperate zone. Unless assimilation capacity during the GFP is increased by the use of cultivation technology or until a new cultivar with low response in PGIR at high temperatures is found, grain yields in rice will be reduced as temperatures increase.


   ACKNOWLEDGMENTS
 
Our thanks to the staff of Shimane Prefecture Akana experimental farm and to Dr. Tatuya Inamura and Mr. Hisashi Kgata of the Kyoto University Takatuki experimental farm for their technical assistance and support during the field season. We thank Dr. Barry Roser for reading our manuscript.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Agron. J.Home page
T. Kobata, T. Nagano, and K. Ida
Critical Factors for Grain Filling in Low Grain-Ripening Rice Cultivars
Agron. J., April 11, 2006; 98(3): 536 - 544.
[Abstract] [Full Text] [PDF]

Home page
 ANN BOT (LOND)Home page
S. MORITA, J.-I. YONEMARU, and J.-I. TAKANASHI
Grain Growth and Endosperm Cell Size Under High Night Temperatures in Rice (Oryza sativa L.)
Ann. Bot., March 1, 2005; 95(4): 695 - 701.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kobata, T.
Right arrow Articles by Uemuki, N.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Kobata, T.
Right arrow Articles by Uemuki, N.
Agricola
Right arrow Articles by Kobata, T.
Right arrow Articles by Uemuki, N.
Related Collections
Right arrow Rice
Right arrow Seed Production

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Crop Science Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome