| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Plant Environment Laboratory, Univ. of Reading, Cutbush Lane, Shinfield, Reading RG2 9AF, UK
* Corresponding author (p.q.craufurd{at}rdg.ac.uk).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
3 h after dawn). Exposure to high temperature (centered on the time of peak anthesis) for 6 h reduced spikelet fertility more than exposure for 2 h, and fertility was lower at 38°C than at 35°C. Genotypic ranking for spikelet fertility at 35 and 38°C was highly correlated in both 2003 and 2004. Fertility was also highly correlated across years, suggesting a consistent and reproducible response of spikelet fertility to temperature. The check cultivar N22 was the most heat tolerant genotype (64–86% fertility at 38°C) and cultivars Azucena and Moroberekan the most susceptible (<8%).
Abbreviations: BST, British Summer Time • DAS, days after sowing • IRRI, International Rice Research Institute • OR, odds ratio • RH, relative humidity • VPD, vapor pressure deficit
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Anthesis is the most sensitive stage of rice to high temperatures (Yoshida et al., 1981). The heat sensitive processes of anthesis—anther dehiscence, pollination, pollen germination, and to a lesser extent pollen tube growth—are completed within 45 min of the opening of a rice spikelet (Ekanayake et al., 1989) and fertilization is completed within 1.5 to 4 h (Cho, 1956). Spikelet tissue temperature of 33.7°C for an hour at anthesis was sufficient to induce spikelet sterility (Jagadish et al., 2007). In contrast, spikelet fertility is not affected by high temperatures an hour before or after anthesis, even at 38 and 41°C (Yoshida et al., 1981). Similar effects of short episodes of high temperature during flowering on fertility have been reported in peanut (Arachis hypogaea L.; Prasad et al., 1999, 2000), wheat (Triticum aestivum L.; Wheeler et al., 1996) and sorghum [Sorghum bicolor (L.) Moench; Prasad et al., 2006b]. Rice typically antheses late morning with peak anthesis occurring between 1000 and 1200 h. Therefore, it is essential that screening or phenotyping for heat tolerance at anthesis excludes or minimizes the possibility of avoidance or escape by early or late anthesing spikelets.
Rice genotypes have been screened for tolerance to high temperature during flowering (Satake and Yoshida, 1978; Matsui et al., 2001; Matsui and Omasa, 2002; Prasad et al., 2006a) at temperatures up to 41°C and for durations ranging from 2 h to the whole crop cycle. These studies identified a few heat tolerant genotypes capable of setting seed at high temperature during flowering in both indica and japonica spp. However, these studies did not explicitly separate tolerance from avoidance. The most tolerant cultivar found to date, ‘N22’, is agronomically poor and has not been used in breeding. Furthermore, none of these studies shed light on the inheritance or genetic control of this important trait, which at least in some dicots is controlled by a few genes (Hall, 1992).
Heat tolerance and heat avoidance of high temperature at anthesis are both potentially useful traits for breeding programs for hotter rice growing environments today and more especially in the future (International Rice Research Institute, 2007). Two experiments were conducted at optimum (30°C daytime) and high (35 and 38°C) temperature with parents of some prominent mapping populations (i) to determine whether there are differences in the daily flowering pattern and hence avoidance, and (ii) to identify rice genotypes having true heat tolerance during anthesis
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Genotypes
Eighteen and eight rice genotypes were sown in 2003 and 2004, respectively (Table 1
). These genotypes were all parents of mapping populations, except for N22, a heat-tolerant check, and included genotypes of O. sativa (mostly indicas), O. glaberrima, and interspecifics of O. sativa x O. glaberrima.
|
Greenhouse Conditions
Plants were grown in a naturally lit greenhouse with day and light-proof night compartments. Pots were placed on automated mobile trolleys (2.85 by 0.96 m) that were drawn out and into night compartments (3.57 by 1.74 m) at 0800 h (i.e., dawn) and 1900 h British Summer Time (BST), respectively, giving a constant photoperiod of 11 h. In both years, temperature in the greenhouse was maintained at 30°/24°C day/night through a combination of heating and venting during the day and heating at night. Fans in both the day and night compartments continuously circulated air. Aspirated air temperature was measured at canopy height every 30 min using copper constantan thermocouples and recorded using data loggers (Delta T Devices, Cambridge, UK).
Growth Cabinets
At anthesis (the first day anthers were observed in any spikelet), plants were transferred to modified Saxcil growth cabinets (internal size 1.4 by 1.4 by 1.5 m). Cabinets were maintained at 380 µmol mol–1 CO2 of air (IRGA, ADC Gas Analysis, Great Amwell, UK). The photosynthetic photon flux density at the base of the cabinet was maintained at 650 µmol m–2 s–1 using cool white fluorescent tubes and incandescent lamps. Photoperiod was 11 h, and lights were switched on at 0800 h. A centrally placed fan circulated heat and air uniformly throughout the cabinet. Screened and aspirated air temperature and relative humidity (RH) were measured using copper constantan thermocouples and recorded using a data logger (Delta T Devices) every 10 s and averaged each 10 min. Temperature treatments varied in both the experiments as detailed below. Vapor pressure deficit (VPD) and RH, controlled by adding moisture to air passing through glycol maintained at a set temperature or by removing the excess humidity by condensation, also varied between experiments.
Experiment 1
In 2003 plants were exposed to daytime air temperatures of 30, 35, and 38°C at anthesis in unreplicated growth cabinets. Vapor pressure deficit was maintained at 1.2 kPa at all temperature regimes; therefore, RH ranged from 70% at 38°C to 85% at 30°C. The day after anthesis was first observed in the greenhouse, 10 replicate plants per genotype per temperature treatment were transferred at 0800 h (dawn) to a control growth cabinet maintained at 30°C. Two hours after dawn (1000 h BST), 10 plants were transferred to growth cabinets at air temperatures of 30° (control), 35°, and 38°C for 2 h and then returned to the greenhouse. Plants were transferred 2 h after dawn, so that they were exposed to the temperature treatments when maximum number of spikelets were open.
To identify anthesing spikelets exposed to high temperature—and hence remove the potentially confounding effects of avoidance associated with early or late anthesis—anthesing spikelets were carefully marked with acrylic paint (Jagadish et al., 2007). First, any anthers from spikelets that anthesed before the transfer to the growth cabinets were carefully removed. On the day of the temperature treatments, spikelets that had opened (i.e., with a visible anther) in the 2 h before the temperature treatments (0800–1000 BST) were marked with blue acrylic paint before transferring from the 30°C cabinet. Those that opened during the 2-h temperature treatments were marked with red paint at the end of the temperature treatment. Finally, spikelets opening after the temperature treatments were also marked with blue paint. Ten to 12 d after the temperature treatments, spikelets marked with blue and red paint were scored separately for fertility or seed set. The average number of spikelets marked during the temperature treatments ranged from 156 (35°C) to 166 (38°C).
Experiment 2
In 2004 plants were again subjected to temperature treatments of 30, 35, and 38°C in growth cabinets using the same transfer and spikelet marking system as the previous year. However, in 2004 spikelets were exposed to high temperature for 6 h instead of 2 h to ensure that >90% anthesing spikelets were exposed to high temperature (Jagadish et al., 2007). Growth cabinets were replicated twice, and there were five replicates per genotype per treatment per growth cabinet. Relative humidity was held constant at 60% during the day to maximize anther dehiscence (Matsui et al., 1999a,b, 2001; Matsui and Omasa, 2002). Therefore, VPD ranged from 1.7 (at 30°C) to 2.3 kPa (at 38°C). Spikelet tissue temperature was measured in ‘IR64’ (indica) and ‘Azucena’ (japonica) using copper constantan thermocouples in spikelets of four plants per cabinet and recorded using a data logger (Delta T Devices) every 10 s and averaged over 10 min for the entire period of high temperature exposure. There was no genotypic difference among the two genotypes; hence, the tissue temperature was assumed to be the same in the other genotypes.
On the day that anthesis was first observed, plants were moved to a control growth cabinet (30°/24°C day/night) to give uniform conditions before the temperature treatments. On the following day, plants were subjected to the temperature treatments for 6 h, starting at 0900 (1 h after dawn) until 1500 h. Open spikelets were marked halfway through and at the end of the treatment period. Plants were transferred back to the control cabinet after the temperature treatment, and from there to the greenhouse the following morning. Ten to 12 d after anthesis, marked spikelets were scored for spikelet fertility. The average number of spikelets marked during the temperature treatments ranged from 246 (38°C) to 274 (30°C). More spikelets were marked in 2004 because the plants were exposed to a longer duration (6 h) of high temperature compared to 2003 (2 h).
Effect of High Temperature on Flowering Pattern
In 2004 the daily flowering pattern of the eight genotypes was recorded. Three plants of each genotype were exposed to 30 and 38°C from 0900 to 1500 h for three consecutive days. On each day, the number of open spikelets was counted every 30 min. The anthers of opened spikelets were then removed to make subsequent counting easier.
Statistical Analysis
Spikelet fertility was treated as binomial data (Jagadish et al., 2007) and analyzed as logits ([Log10 (p/100 – p)] of percentages of p [0 < p < 100%]) using the generalized linear mixed models procedure of Genstat version 8.1 and expressed as percent odds ratio (OR). The flowering pattern of each genotype at 30° and 38°C was compared using a Kolmogorov–Smirnov test (an empirical distribution function). Spearman rank correlation was used to compare genotype performance across temperatures and years.
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
35°C from 60 to 68 DAS. This coincided with the flowering period of ‘Lemont’ (68 DAS), N22 (60 DAS), and ‘WAB 450-I-B-P38-HB’ (65 DAS). However, spikelet fertility in the control temperature treatment (29.6°C in Table 2
) was apparently not affected, probably because the maxima occurred after most spikelets had anthesed (see below).
|
Effect of High Temperature on Flowering Pattern
The daily pattern of flowering at 29.6°C was similar in seven O. sativa genotypes (Fig. 1
; Table 2), the first spikelets opening 1 to 3 h after dawn (between 0900 and 1100 h BST), the maximum number of spikelets opening between 3 and 4 h after dawn (between 1100 and 1200 h BST), and spikelet opening ending 4 to 6.5 h after dawn. The duration of spikelet opening was mostly 3 to 4 h, although in N22 it was only 1.5 h. In N22 24, and 17% of spikelets opened in the 30-min interval on either side of the peak anthesis with 46% spikelets opening at the peak. In contrast to O. sativa genotypes, maximum spikelet opening occurred 1.5 h after dawn in ‘CG14’, an O. glaberrima (Fig. 1), and spikelets were open immediately after dawn.
|
Spikelet Fertility
Experiment 1
In 2003 there were significant effects of temperature (P < 0.001) and genotype (P < 0.001), but not their interaction (P > 0.56), on spikelet fertility following a 2-h exposure to 29.6, 33.7, and 36.2°C spikelet temperature. Average spikelet fertility was 84.0, 78.9, and 47.1% OR at 29.6°, 33.7°, and 36.2°C, respectively (Table 3
). Spikelet fertility in most genotypes was therefore only significantly reduced at 36.2°C; the critical or threshold tissue temperature was therefore <36.2°C in this experiment.
|
29.6°C in a similar manner.
|
1 SD of N22), tolerant (>1 and 2 SD of N22), susceptible (>2 and 3 SD of N22) and highly susceptible (>3 SD of N22), where SD = 21.1% OR (Table 3).
Experiment 2
In 2004 seven parents of mapping populations along with N22 as check were exposed at anthesis to spikelet temperatures of 29.6, 33.7, and 36.2°C for 6 h. There were highly significant effects of temperature (P < 0.001), genotype (P < 0.001), and their interaction (P < 0.001) on spikelet fertility. No significant differences (P > 0.16) occurred between replicate growth cabinets. The interaction in 2004 was largely caused by the values for Co 39 and IR64 at 29.6°C, which were lower and higher, respectively, than values at 33.7 and 36.2°C, compared to 2003. IR64 also had much lower values of spikelet fertility at 33.7 and 36.2°C in 2004 than in 2003.
Average spikelet fertility was 87.3, 66.9, and 26.4% OR at 29.6, 33.7, and 36.2°C, respectively (Table 4 ). Values at the two warmer temperatures were therefore lower than in 2003. As in 2003, there was a strong relation between the ranks (r = 0.99, P < 0.001, n = 8; Fig. 2) and absolute value (r2 = 0.71, n = 8) for spikelet fertility at 36.2 and 33.7°C. There was only a weak relationship between the ranking of spikelet fertility at 29.6 and 33.7°C (P = 0.08) and 29.6 and 36.2°C (P = 0.05).
|
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
There was significant variation in the time of day of flowering between O. sativa and O. glaberrima species, with CG14 flowering 1.5 to 2.5 h earlier than O. sativa genotypes (Table 2, Fig. 1). Yoshida et al. (1981) and Prasad et al. (2006a) observed similar differences. Early time of day of flowering should confer an advantage as temperatures are cooler in the early morning; this advantage can readily be exploited by interspecific hybrids (so-called NERICAs—New Rice for Africa; Africa Rice Center, http://www.warda.org/warda/nericas-at-a-glance.asp). Significant variation among O. sativa cultivars. in the time of day of flowering has also been observed. (S. Heuer, International Rice Research Institute [IRRI], personal communication, 2007).
The marking protocol allowed spikelets that anthesed at high temperature to be identified to measure true heat tolerance (Jagadish et al., 2007). We interpret the lower percentage fertility at 36.2°C during 2004 (6-h exposure) compared with 2003 (2-h exposure) to be because >96% spikelets anthese during a 6-h period compared with 37 to 64% during a 2-h period (Jagadish et al., 2007). Therefore, in the longer duration treatment, many more spikelets are exposed to >1 h high temperature—enough to cause sterility. In contrast, with a shorter exposure, spikelets opening at the start and end of the 2-h period will be exposed to high temperature for <1 h and hence may not be sterile (Jagadish et al., 2007). There were also differences in RH and VPD between years, which may have affected fertility, although the lower RH used in 2004 was reported to be optimum for anther dehiscence (Matsui et al., 1999a,b; Matsui and Omasa, 2002).
None of the mapping population parents screened were as tolerant as N22 at high temperature (Tables 3 and 4). There seems no doubt from several different studies at similar day temperatures that N22 is a true heat-tolerant cultivar (Yoshida et al., 1981; Ziska and Manalo, 1996; Prasad et al., 2006a) and should be used as a source of tolerance in breeding programs. Among the other genotypes examined over 2 yr, only Bala (which includes N22 in its background) and Co 39 exhibited moderate heat tolerance. All the other genotypes, including CG14 (O. glaberrima), were susceptible to high temperature. Azucena and Moroberekan were quite sensitive to high temperature, and there is some evidence of a temperature x duration of exposure interaction in Azucena (Jagadish et al., 2007).
The consistent significant relationship of fertility at 33.7 and 36.2°C across years (33.7°C: r = 0.76 and 36.2°C: r = 0.71; both P < 0.05, n = 8) and between 33.7 and 36.2°C (Fig. 2) suggests that the screening protocol is robust in terms of relative cultivar performance. However, one genotype, IR64, exhibited much greater sensitivity in 2004 than 2003. The seed lots of IR64 used in 2003 and 2004 were from multiplications in different seasons at IRRI (wet and dry season 2001), and we assume, given the consistency of the other genotypic responses, that this caused the difference in response. We therefore suggest that 6-h exposure be used for screening to reduce the possibility of escape. The strong correlation between 33.7 and 36.2°C, and the lack of correlation between 29.6°C and warmer temperatures with 6-h exposure, suggests also that the critical temperature was <33.7°C, in agreement with Yoshida et al. (1981) and Nakagawa et al. (2002) for rice, and similar to many other crops (e.g., peanut: Prasad et al., 2001). Yoshida and Nakagawa's data also suggest that it is variation in the critical temperature, which may vary among genotypes by up to 6°C, rather than the response to temperature above the critical temperature, that determines tolerance.
In conclusion, differences in flowering pattern within a day could be used as an avoidance mechanism to ensure good seed set. True heat tolerance of seed set can be reliably measured by marking spikelets exposed to >33.7°C tissue temperature for 6 h. The most heat tolerant cultivars were N22, Bala and Co 39, and these should be exploited in breeding programs and for genomic and proteomic studies.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Received for publication December 27, 2007.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. V. K. Jagadish, R. Muthurajan, R. Oane, T. R. Wheeler, S Heuer, J. Bennett, and P. Q. Craufurd Physiological and proteomic approaches to address heat tolerance during anthesis in rice (Oryza sativa L.) J. Exp. Bot., January 1, 2010; 61(1): 143 - 156. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Q. Craufurd and T. R. Wheeler Climate change and the flowering time of annual crops J. Exp. Bot., July 1, 2009; 60(9): 2529 - 2539. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | 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 | |||
