Heading (flowering) date is one of the most important adaptive characteristics of plants, which has a major impact on grain yield in crop species. In rice, photoperiod and temperature are two crucial exogenous signals that control heading date. Extensive research on photoperiod effects has identified two crucial flowering pathways. However, very little is known about temperature effects. Field observations indicate that low temperature delays the flowering, but the molecular mechanism underlying this process has not yet been identified. The main objective of this study was to dissect the genetic basis of rice flowering regulation by temperature, apart from the photoperiod, which has been extensively researched.
We conducted phenotypic and genotypic analysis on two Oryza sativa L. indica varieties, Zhenshan 97 and Zhongzao 18, and 168 recombinant inbred lines derived from them, grown in two different seasons in two years. Tests under different photoperiod and temperature conditions in growth chambers showed that heading date (HD) of both parents was accelerated at high temperatures but delayed at low temperatures irrespective of photoperiod. The averaged effective cumulative temperatures (ECTs) necessary for Zhenshan 97 and Zhongzao 18 were fluctuated around 1110 and 1260 °C, respectively, either in both field and growth chamber experiments or long day and short day conditions, indicating these values are parents' threshold ECTs, which could be the prerequisite before heading.
Get your grade
or your money back
using our Essay Writing Service!
A major ECT QTL, qEHD10, also with a large effect on HD, was identified in the interval between markers RM1375 and RM3229 on chromosome 10, which could explain 42.2% to 57.0% of ECT variation and 39.8% to 59.4% of HD variation in four environments. Over all these novel findings would improve the knowledge of temperature effects on rice flowering.
Flowering time is one of the most fundamental and important adaptive characteristics of plants, which transit from the vegetative to reproductive stage mediated by endogenous and exogenous signals. Endogenous signals include gibberellin and ascorbic acid, which play a crucial role in flowering. Endogenous signals are not largely influenced by environment . The two major exogenous signals that mainly influence this transition are photoperiod and temperature . Based on the photoperiod requirements for flowering, plants are divided into short-day plants, long-day plants, and day-neutral plants , . Because rice is a short day plant, heading date (HD) is accelerated under short-day (SD) conditions while it is delayed under long-day (LD) conditions.
Photoperiodic controlled pathways have been well described using mutants and natural varieties in Arabidopsis and rice. Flowering in Arabidopsis is stimulated under LD conditions and activates the FLOWERING LOCUS T (FT) , which conceals a mobile florigen . FT expression is activated by CONSTANS (CO), which consists of a zinc finger domain and a CCT domain  and is in turn regulated by GIGANTEA (GI), a part of the circadian clock .
Mean while many genes responsible for photoperiodic flowering in rice, such as Hd1, Ehd1, EHD2, SE5, Ghd7, and Hd3a have been identified. Hd1, an orthologue of CO in Arabidopsis, has dual functions in controlling rice heading, serving as a promoter of heading under SD conditions and a repressor under LD conditions. Plants that lack a functional Hd1 gene flower later under SD conditions compared with wild-type plants, which indicate the crucial role of Hd1 in promoting flowering under SD conditions , . Another QTL for flowering, Early heading date 1 (Ehd1), which encodes a B-type response regulator  promotes flowering under both SD and LD conditions but to a greater extent under SD conditions. Under SD conditions, Ehd1 promotes expression of Hd3a and a related FT-like gene, rice FT1 . Ehd1 is thought to act independently of Hd1, and its expression is up-regulated by the MADS box gene OsMADS51 , which is itself regulated by OsGI. Therefore, OsMADS51 might be acting as an intermediate between OsG1 and Ehd1 to promote the expression of Ehd1 with a diurnal rhythm under SD conditions. In addition, OsGI might regulate diurnal expression of Hd1, similar to the regulation of CO by GI in Arabidopsis . This shows the important role of OsGI, like GI in Arabidopsis, in controlling the two independent photoperiodic pathways. Recently Xue et al.  reported that Ghd7 encodes a zinc finger and a CCT domain, expressed predominantly during the light period of long days, and delays flowering under LD conditions by repressing Hd3a. In addition, studies have also clarified that Oryza sativa INDETERMINATE1 (OsId1, RId1, or Ehd2) is also required for expression of Ehd1 and Hd3a for the promotion of flowering under SD conditions -. However, there are no homologues of Ghd7, OsID1, Ehd1, or OsMADS51 in Arabidopsis.
Always on Time
Marked to Standard
Temperature is another important environmental factor involved in flowering. In Arabidopsis, the temperature-regulated vernalization pathway is essential for flowering. Several studies have shown that variation at one or both of two loci, FLOWERING LOCUS C (FLC) and FRI, can account for a large portion of the winter annual habit in Arabidopsis . FLC, which encodes a MADS box transcription factor, represses the floral promoter SUPPRESSOR OF OVER EXPRESSION OF CONSTANS 1 and also binds to the sites within the FT gene to repress transcription and suppress the long-day flowering response . In addition, the vernalization mediated repression of FLC requires VERNALIZATION INSENSITIVE 3 , VERNALIZATION 2, and VERNALIZATION 1, plant-specific DNA-binding proteins . Lee et al.  demonstrated that SHORT VEGETATIVE PHASE is essential for delayed flowering under cool conditions using svp mutants. FT expression is elevated at high growth temperatures , whereas little acceleration is seen in ft mutants, showing that that FT is involved in this phenomenon. Balasubramanian et al.  reported that warmer growth conditions (27.8°C versus 23.8°C) accelerated flowering under SD conditions, when CO is less active, and proposed that there might be another mechanism that induces FT in response to high temperatures.
In rice, however, very little is known about temperature effects because this species does not undergo a vernalization period. Field observations indicate that low temperature delays the flowering, but the molecular mechanism underlying this process has not yet been identified. No homologue of Arabidopsis FLC controlled by temperature signals has been observed in rice. Compared to the extensive work on photoperiod, little research has been conducted to elucidate potential temperature effects on rice heading.
In this study, we tested a recombinant inbred line (RIL) population and its two parents: (1) to measure how they respond to different temperatures and photoperiods under both natural and artificial conditions; (2) to evaluate the temperature effects on the heading of RIL population across two seasons in two years under natural field conditions, by calculating daily temperatures and day lengths; and (3) to identify the temperature-driven HD QTLs using effective cumulative temperature (ECT) as a trait.
Heading dates of parents under different photoperiod and temperature conditions
ZS97 and ZZ18 plants were grown in growth chambers with different combinations of photoperiod and temperature treatments. In high-temperature treatments, HDs for ZS97 were 64 days and 65 days under SD and LD conditions, respectively, and those for ZZ18 were 73 days and 75 days under the same conditions (Figure 1; Table 1), indicating that both the parents are insensitive to photoperiod. The HD of ZS97 was 9 days earlier than that of ZZ18 under SD conditions and 10 days earlier under LD conditions. In the low-temperature treatments, HDs for ZS97 were 86 days and 88 days under SD and LD conditions, respectively, and those for ZZ18 were 96 days and 97 days under the same conditions (Figure 1; Table 1). The HD of ZS97 was 10 days earlier than that of ZZ18 under SD conditions and 9 days earlier under LD conditions.
The SD promotion rates of ZS97 and ZZ18 in the high-temperature treatments were 1.56% and 2.66%, respectively, and 2.27% and 1.03% in the low-temperature treatments (Table 1). The high-temperature promotion rates of ZS97 and ZZ18 under SD conditions were 25.6% and 23.9%, respectively, and 26.1% and 22.6% under LD conditions (Table 1). Temperature had significant effects on ECT and HD, no significant photoperiod effects and interactions between temperature and photo-period was detected using two-way analysis of variance (Table 2)
Photoperiod and temperature fluctuations in field experiments
In experiments 1 and 3, the day length ranged from 13.5 to 14.5 h during the period from sowing to flowering (18 May to 14 August), closely matching the LD conditions. In experiments 2 and 4 (late growing seasons), day length gradually decreased from 13.5 to 11.5 h from sowing to flowering (5 July to 28 September), which tended to be SD conditions.
Temperature was also recorded daily, and the temperature curves show fluctuations in both seasons in the two years (Figure 2). In experiment 1, an average of 25°C with a minimum of 20°C and a maximum of 30°C was recorded. Whereas in experiment 3, an average of almost 27°C (higher than in experiment 1) with a minimum of 18°C and a maximum of 32°C was recorded. In experiments 2 and 4 (July to September), the minimum temperature recorded was 22°C and the maximum was 33°C, with an average of 28°C. Thus, the average temperatures from sowing to heading in experiments 2 and 4 were higher than those in experiments 1 and 3.
Days to heading
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
In experiments 1 and 3, average HDs of ZS97 and ZZ18 were 64 and 73 days and 62 and 72 days, respectively; HDs of the RIL population ranged from 59.8 to 87.1 days and 57.3 to 85 days, respectively. In experiments 2 and 4, average HDs of ZS97 and ZZ18 were 60 and 71 days and 60 and 70 days, respectively; HDs ranged from 54.7 to 80 days and 55 to 81.8 days for the RIL population (Table 3).
Effective cumulative temperature
ECTs were 1109°C for ZS97 and 1266°C for ZZ18, with a range of 1025.9 to 1534.7°C in the population, and 1140°C and 1292°C, with a range of 1051.1 to 1429.1°C, in experiments 1 and 2, respectively. In experiments 3 and 4, ECTs for ZS97 and ZZ18 were 1102°C and 1235°C, with a range of 1045.3 to 1502.5°C in the population, and 1130°C and 1261°C, with a range of 1070.2 to 1437.4°C (Table 3).
Polymorphism between parents
A total of 908 SSR markers distributed over the 12 chromosomes were used to identify polymorphism between the parents. Of them, 75 polymorphic markers were observed on 11 chromosomes, except on chromosome 8. The polymorphic ratio of 8.25% indicated that both parents were genetically closely related. These markers were used for the genotypic analysis of the whole population.
The genetic linkage map was difficult to develop for chromosomes having fewer polymorphic markers. Single marker analysis was performed for QTL detection using all chromosomes. However, the marker orders used to create the linkage map were assumed based on the published rice genome RM marker orders , . One million bases on a rice chromosome is approximately equivalent to 4 cM ; using this relationship, the published physical distances between markers (http://www.gramene.org) were used to estimate approximate genetic distances on chromosome 7. Genetic linkage maps were developed for chromosomes 10, 11, and 12, with eight or more polymorphic markers. Interval mapping method was conducted for QTL detection with these three chromosomes.
Location of QTLs
The skewness and kurtosis values (Table 3) of HD and ECT in the RIL population were all less than 1.0 in absolute value, suggesting that both these traits approximately fit normal distributions and the data for the traits are suitable for QTL mapping.
QTLs for HD and ECT were commonly identified in the four experiments (Tables 4 and 5) and termed qEHD (ECT-driven HD QTL). According to single marker analysis, QTLs were detected on chromosomes 7, 10, 11, and 12. qEHD7 was identified near marker RM473 on chromosome 7 in experiment 1, and qEHD11 was near marker RM5349 on chromosome 11 in experiment 1. The QTL qEHD12 near marker RM6837 on chromosome 12 was detected only in experiment 2, whereas qEHD10 was commonly detected near marker RM1375 on chromosome 10 (P < 0.001 in all the four experiments).
Using the interval mapping method, only one QTL was detected in each experiment on chromosome 10. Their 1 - LOD confidence intervals mostly overlapped, indicating there was one common QTL, qEHD10, detected in all four experiments (Figure 3). This QTL explained HD variation of 39.8% to 59.4%, with LOD values of more than 17 (Table 5).
Sequencing analysis of photoperiod-sensitive genes in ZS97 and ZZ18
qEHD10 was located in the neighboring regions of Ehd1 and Ehd2 on chromosome 10. In order to determine their relationships to qEHD10, we sequenced the Ehd1 and Ehd2 alleles of ZS97 and ZZ18. Sequencing analysis of Ehd1 showed a few single-base substitutions in exons 1 and 4 but not in exons 2 and 3 (Figure S1A). In ZS97, SNPs were found at 130, 160, and 224 bp away from ATG, and two SNPs occurred in exon 4. In ZZ18, three SNPs were found at 83, 224, and 226 bp away from ATG, and two SNPs occurred in exon 4. The protein sequences of both parents have 256 amino acids, as in Nipponbare, except for slight modifications in 5 amino acids, showing that the parents carried functional alleles (Supporting Information Fig. S1B). The sequencing analysis of Ehd2 showed no sequence difference between ZS97 and ZZ18 in both promoter and coding region (Figure S2A). However, the analysis revealed an SNP at 26 bp from exon 4 when compared with Nipponbare, whereas the amino acid sequence had no difference among the three cultivars (Figure S2B).
Growth temperature determines heading date
ECTs from sowing to heading for the parents were consistent in whatever conditions they were grown. The ECTs of ZS97 and ZZ18 were fluctuated around 1110 and 1260 °C, in the four field environments respectively. Probably indicated, 1110 and 1260 °C were the threshold ECTs, which the plants must obtain before they begin flowering. Thus HDs for the two parents were largely associated with temperature. For the population, days to heading was less in experiments 2 and 4 compared to experiments 1 and 3 (Table 3), in which the average temperatures during the period from sowing to flowering were lower than in experiments 2 and 4 (Figure 2). However, the mean ECTs of the population were similar among the four experiments, with small differences that were probably due to the minor error in calculations of three-time-point mean temperature, Especially, ECTs of the population had no significant difference between experiments in LD (experiments 1 and 3) and SD trended conditions (experiments 2 and 4). These findings indicate that temperature played a key role in controlling HD in the population.
In order to confirm the role of temperature and photoperiod effects on both parents, we performed experiments in growth chambers under different temperature and light conditions. Two-way analysis of variance showed there was no photoperiod effect and interaction effects on HD and ECT, but only temperature had significant effects on HD and ECT (Table 2). Thus, the photoperiod treatments did not change heading in the parents at the same temperature conditions. However, the high-temperature treatment promoted early heading in any photoperiod conditions, and vice versa.
HD was controlled by threshold ECTs
To investigate the effects of ECT on HD in rice, photoperiod effects and interactions between photoperiod and temperature had better to be excluded. Thus, HD measurement of a population should be made in well controlled conditions such as growth chambers. But, there was not enough well controlled space available to plant the large population with 2 replicates. Hence, in order to maintain the constant photoperiod, we planted the population at the same dates in the two seasons in two years and the daily average temperatures were recorded to measure the ECT, to normalize fluctuating daily temperatures. Moreover in this study, the photoperiod and photoperiod by temperature interaction had no effects on HD of our materials. Thus, our materials were suitable to analyze temperature effect on HD in the field.
In rice, the reported base temperature for time to heading varied with different types of models considered, 8 °C with the nonlinear model and 10.7 °C with the linear model for IR8 , and with the range of temperature analyzed, 8.6 °C in the range 20 to 28 °C for IR36 . Using a nonlinear model, Gao et al.  considered a base temperature of 10 °C for japonica, 12 °C for indica, and 13 °C for hybrid rice. Here we considered 10°C as a base temperature for the two parents and population. ECT was used to estimate the temperature effects on flowering in sunflower , wheat , Arabidopsis , and rice , . It was reported that each photoperiod insensitive genotypes of sunflower and rice has to obtain threshold ECT before flowering , . In fact, in China, to ensure the successful production of hybrid, heading synchronization of parents of three-line hybrid system was made by regulating sowing date of parents with ECT as an indicator , . In this study, the averaged ECTs of both parents, ZS97 and ZZ18 in all the four field experiments was 1115 and 1259 °C which were close to the values of 1113 and 1264 °C in growth chamber, although the day lengths and temperatures in the field and growth chamber was different. This suggested that 1110 and 1260 °C were the threshold ECTs of ZS97 and ZZ18, respectively.
It was noted that HD and ECT were two kinds of parameters to measure heading time. Hence, the QTL for ECT and HD located in the same interval should be one QTL. The QTL effects can be explained in two ways. For example, in experiment 1, the additive effect of qEHD10 means the QTL promoted ZZ18 heading up to 9.6 days and ECT to 170 °C as compared to ZS97 (Table 5). That is to say, in experiment 1, ECT of 170°C took 9.6 days, indicating that ECT of 19°C was required per day.
Genetic background of the parents is helpful for QTL detection
Generally, QTL detection is based on natural allelic differences between parental lines . It is expected that using a two-parent population to map QTLs will detect only limited QTLs, because the genes will present only limited information about polymorphism content. In this study, the polymorphic percentage was less than 10% and we could not identify any polymorphic markers on chromosome 8, which implied that both parents were highly genetically related. With the cloning of key flowering genes, the photoperiodic pathway in rice is well understood. Two independent flowering pathways were shown to be separately regulated by Ehd1 and Hd1, both of which are upstream of Hd3a, an orthologue of Arabidopsis FT . To identify the status of the parental alleles of the three important flowering genes, comparative sequencing was performed between the parents for Hd3a, Ehd1, and Hd1.
The Hd3a protein sequence is the same in both ZS97 and ZZ18. Compared to Nipponbare, it has one amino acid substitution, but the lengths of the protein sequences are the same, meaning this is a functional allele (Figure S4). Sequencing analysis of the Hd1 locus showed that both ZS97 and ZZ18 contain a 36-bp insertion 334 bp away from the ATG site, where Nipponbare has a 36-bp deletion in exon 1 (Figure S3A). ZS97 has one single-base substitution 531 bp away from ATG. In ZS97 and ZZ18 there is a common SNP at the site of 600 bp away from ATG, and ZS97 has another SNP at 531 bp from ATG in exon 1 as well. We speculate that these Hd1 alleles are nonfunctional in the parents because the protein sequence is modified by the addition of 12 amino acids and the variation of 2 amino acids (Figure S3B). Therefore, it is likely that the Hd1-mediated flowering pathways, which are controlled by photoperiod, did not function in the parents and the population. Ehd1, a B-type response regulator, is involved in early flowering under SD conditions and delayed heading under LD conditions independent of hd1 . ZS97 and ZZ18 carried polymorphic Ehd1 alleles (Fig. S1A), but the two parents did not expressed the phenotype of Ehd1, a key gene in rice photoperiod flowering pathway. qEHD10 detected in this study was closely linked to Ehd1. Currently, we cannot rule out the possibility that qEHD10 is not Ehd1, even though the population does not show photoperiod response. Developing near isogenic lines of qEHD10 was under process for fine mapping, in order to provide a clear proof that qEHD10 is different locus from Ehd1. However, the new findings of temperature promoted HD, riches us the knowledge in rice flowering pathway.
Roles of ECT in the rice flowering pathway
In rice the influence of photoperiod on HD has been extensively researched and is well characterized, whereas the effects of temperature remain unclear. In Arabidopsis, however, temperature can regulate flowering time via the vernalization pathway . Vernalization is very important for winter plants and regulates flowering time through the promotion or suppression of FLC expression. But no homologues of FLC have been identified in rice.
HDs of ZS97 and ZZ18 were accelerated at high-temperature treatments under both SD and LD conditions. In contrast, ZS97 and ZZ18 showed no significant difference in HD between SD and LD conditions in spite of high temperature or low temperature. This contradicts the conventional behavior of rice, in which heading was inhibited under LD and promoted under SD conditions. These findings suggest that some temperature signals may exist apart from the photoperiodic signals that regulate HD. Luan et al.  reported that HDs in rice mutant lF1132 were very early at high temperatures under both SD and LD conditions, whereas at low temperatures the HD was delayed significantly under SD but little earlier under LD conditions by the hd1-3 locus on chromosome 6. They reported that HD was delayed at low temperatures due to the suppression of hd3a expression by hd1-3. The genes involved in the photoperiod pathway might be closely linked to the locus of temperature response . Lin et al.  identified the Hd9 locus and proposed a hypothesis that Hd9 is involved in characteristics other than photoperiod sensitivity. Nakagawa et al.  confirmed that Hd9 was involved in thermal response. These results are similar to those of the present study in which qEHD10 was identified near the photoperiod gene Ehd1 and made response to temperature.
Rice growth (germination to heading) is divided into two separate and distinct growth phases, namely basic vegetative growth phase (BVG) and photoperiod-sensitive phase (PSP) , . Temperature is considered to affect rice heading date by accelerating or retarding BVP , . Poethig  noted the importance of thermal degree days when estimating the temperature in the regulation of phase change and developmental timing in plants. In our studies, low and high temperatures might retard and accelerate BVG period. More over in high temperature, the parents accumulated more degree heat everyday and quickly reached the threshold ECT which led to short HD, and this might be the reason why both parents responded to high temperature in similar manner. In contrast, in low temperature, it took long time to reach the threshold ECT, leading to long HD. Thus qEHD10 function is triggered by ECT threshold point, which ultimately leads to flowering and indicating that there might be a novel pathway regulated by temperature to control rice heading.
Materials and Methods
Plant material and growth chamber conditions for the parents
A population of 168 RILs, produced by seven consecutive generations of single-seed descent from a cross between two elite Oryza sativa L. indica varieties, Zhenshan 97 (ZS97) and Zhongzao 18 (ZZ18), was used in the present study. Controlled photoperiod and temperature experiments were performed on the parents in Conviron PGV 36 type growth chambers (Conviron Ltd., Winnipeg, Canada). ZS97 and ZZ18 were treated with four different growth conditions with two different photoperiods and temperatures: LD, 27.6°C; LD, 22.6°C; SD, 27.6°C; and SD, 22.6°C. The day-length parameters were: LD, 15 h light and 9 h dark; SD, 9 h light and 15 h dark. For both LD and SD conditions, the high-temperature treatment was 27.6°C (average of 25°C for 15 h and 32°C for 9 h) and the low-temperature treatment was 22.6°C (average of 20°C for 15 h and 27°C for 9 h). The seeds were first sown in a natural field, and 2-week-old seedlings were transferred to the growth chambers. A minimum of 10 plants were investigated for the measurement of HD. The high-temperature promotion rate and the SD promotion rate were calculated by the following formulae : high-temperature promotion rate (%) = [(days to heading at low temperature − days to heading at high temperature)/days to heading at low temperature] -100 and SD promotion rate (%) = [(days to heading under LD − days to heading under SD)/days to heading under LD] -100.
Field experimental design
The RIL population together with its parents were sown in two different seasons each year in 2007 and 2009, first on 18 May (hereafter referred to as experiment 1 for May 2007 and experiment 3 for May 2009) and second on 5 July (referred to as experiment 2 for July 2007 and experiment 4 for July 2009), at the experimental farm of Huazhong Agricultural University located in Wuhan (29°58'N, 113°41'E), China. Each field experiment was performed by randomized complete block design with two replicates. Fourteen seedlings of 25-day-old from each line were transplanted into a two-row plot, with a distance of 16.5 cm between plants within a row and 26.4 cm between the rows. Field management, including irrigation, fertilizer application, and pest control, followed essentially the normal agricultural practice.
Measurement of heading date and effective cumulative temperature
HD was recorded as days from sowing to the appearance of the first panicle for each plant. This trait was evaluated for the 10 plants in the middle of two rows of each RIL, excepting the marginal plants. The heading days averaged over the two replications within each season was used as the raw data for analysis.
The highest, lowest, and mean daily temperatures were recorded from the experimental field based on three hourly point calculations . The temperature data recorded for the four experiments were used to calculate the ECT from the day of sowing to heading of each plant in all the RILs among the two replications. Averaged ECT of two replications was used as the raw data for analysis in each season.
The ECT for each genotype was calculated using the following equation:
where Ke is the ECT; Ti is the mean temperature on the ith day obtained from the average of highest and lowest temperature; and T0 is the developmental zero temperature or base temperature, the minimum temperature needed for plant growth below which development ceases (10°C for rice). ECT for a particular HD was calculated by summing the daily ECTs from the day of sowing to the day of heading. The statistical analysis of the phenotype data was conducted using Microsoft Excel 2003 (Microsoft, Redmond, Washington, USA).
DNA isolation and polymorphic marker screening
DNA was extracted from fresh leaves at the seedling stage from 168 RIL (F7) plants using the cetyltrimethylammonium bromide method  with minor modifications. A total of 908 SSR markers were used to screen the two parents, ZS97 and ZZ18, to identify polymorphic markers across the genome. Markers of the RM-series were designed according to Temnykh et al. ,  and those of the MRG-series according to the rice genome sequences of the Monsanto Company . The SSR assay was conducted by polyacrylamide gel electrophoresis, as described by Wu and Tanksley  with minor modifications.
QTL analysis for HD and ECT was carried out by single marker analysis, and the analysis fits the data to the following simple linear regression model:
y = b0 + b1x + e, (2)
where y is the phenotypic value of a line, b0 is the population mean, b1 is the additive effect of the locus on the trait, e is a residual error term and x is directly related to the genotypic code at the locus being tested for the line considered. The results for b0, b1, and the F statistic for each marker were estimated and indicate whether the marker was linked to a QTL. Linkage analysis was performed by MAPMAKER/EXP software Ver.3.0b with the Haldane function  and interval mapping by the MAPMAKER/QTL ver.1.1b . A minimum LOD value of 3.0 was selected to confirm the presence of a QTL in a given genomic region.
Sequencing photoperiod genes (Ghd7, Hd1, Ehd1, Ehd2, and Hd3a)
Promoter and coding regions of each gene were amplified using LA Taq (Takara) from genomic DNA, and PCR products were purified. These purified PCR fragments were sequenced using the Big Dye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). Data were collected using the ABI Prism 3730 DNA Analyzer (Applied Biosystems) and interpreted using SEQUENCHER 4.6 (Gene Codes Corporation).
The authors kindly thank farm technician Mr. J.B. Wang for his excellent field work.
Figure 1. Heading dates for ZS97 and ZZ18 in different photoperiod and temperature treatments conducted in growth chambers.
Figure 2. Temperature fluctuations at the experimental field in 2007 and 2009.
Figure 3. Linkage map showing QTL positions in relation to those of previously identified QTLs.
Figure 4. Heading dates among the population in the four experiments.
Table 1. The effect of photoperiods, mean temperatures, and effective cumulative temperature on heading date
ECT (°C )b
ECT (°C )b
HD at 27.6°C
HD at 22.6°C
SDPR, short-day promotion rate; ECT, effective cumulative temperature; HD, heading date; HTPR, high-temperature promotion rate
a Delay in HD caused by photoperiod
b Effective cumulative temperature under SD/LD conditions
c Delay in HD caused by temperature
Table 2. Analysis of temperature by photoperiod interaction for HD and ECT in the parents
*PxT photoperiod by temperature interaction
a Heading date
b Effective cumulative temeperature
Table 3. Phenotypic analysis of heading date and effective cumulative temperature in the RIL population of rice across two seasons in two years
Mean ± SD
69.5 ± 6.1
1204.4 ± 112.5
66.6 ± 5.6
1231.7 ± 82.5
72.3 ± 6.2
1247.3 ± 112.6
67.7 ± 5.7
1216.9 ± 77.5
ZS97, Zhenshan 97; ZZ18, Zhongzao 18; HD, heading date; ECT, effective cumulative temperature
Table 4. QTLs identified in four experiments by single marker analysis
a Additive effect, a positive value means ZZ18 allele with increasing effect
Table 5. QTL interval mapping analysis for chromosome 10
a Additive effect, a positive value means ZZ18 allele with increasing effect.
b Variance explained by the QTL