- Email: [email protected]
Reanalysis of Vernalization Data of Wheat and Carrot
Annals of Botany 84 : 615–619, 1999 Article No. anbo.1999.0956, available online at http:\\www.idealibrary.com on
Reanalysis of Vernalization Data of Wheat and Carrot W E I K A I Y A N and L. A. H U N T Department of Plant Agriculture, Crop Science Diision, Uniersity of Guelph, Guelph, Ontario, Canada N1G 2W1 Received : 16 April 1999
Returned for revision : 20 May 1999
Accepted : 9 July 1999
Vernalization is an important determinant of the growth, development, and yield of biennial and perennial crops. Accurate simulation of its response to temperature is thus an important component of successful crop systems modelling. Vernalization has a low optimum temperature compared to other temperature responses of plants, and thus may be difficult to treat using expressions that are appropriate for other plant processes. This paper examines the application of a simple equation that has been used for other processes. It reads as l Vmax
0
TmaxkT
10 1 T
Topt Tmax−Topt
, where is the daily rate of vernalization progress at temperature T, Topt and TmaxkTopt Topt Tmax are the optimum and maximum temperatures for vernalization, respectively, and Vmax is the maximum daily rate of vernalization (the inverse of the minimum number of days required to complete vernalization), which occurs at Topt. The model was applied to published vernalization data for wheat and carrot. The fits to data were good (adjusted R# for wheat of 0n94, for carrot 0n98), with estimated Topt and Tmax being 5n7p0n5 and 21n3p1n4 mC, respectively, for wheat ‘ Norin 27 ’ and 6n6p0n2 and 14n1p0n3 mC for carrot ‘ Chantenay Red Cored ’. The estimated parameters, in particular the high Tmax for wheat, were close to those reported using different analytical approaches. It was suggested that the function would be useful for summarizing vernalization data, and that its use would avoid the abrupt changes that are inevitable when different linear relationships are used for part of the overall response. It was also suggested the high Tmax should be taken into account when interpreting data obtained with wheat grown under warm conditions. # 1999 Annals of Botany Company Key words : Plant, vernalization, temperature response, modelling, wheat (Triticum aestium L.), carrot (Daucus carota L.).
INTRODUCTION Vernalization response is a major determinant of the growth and development of biennial and perennial plants. It is generally agreed that the vernalization process has a low optimum temperature, but contrasting opinions exist as to the optimum temperature, the effective range of temperature, and the vernalization requirement for a given species or genotype. These contrasting opinions reflect a number of aspects. The first is the invisible nature of vernalization itself (Lysenko, from Whyte, 1948). The process can only be evaluated in terms of its after-effects, and thus speculation and imagination are required when trying to comprehend the existing evidence. The second is the lack of systematic experimentation, an aspect that reflects the cost of comprehensive experimental investigations of vernalization. The third is the lack of analytical models that can be applied to the existing data. Based on present understanding, a comprehensive investigation into vernalization would involve treating plant material (preferably germinated seeds) at a series of temperatures for a series of durations before moving them into a fully inductive environment (long photoperiods for a long day species plus normal to high temperatures). Chujo E-mail wyan!uoguelph.ca ; thunt!plant.uoguelph.ca
0305-7364\99\110615j05 $30.00\0
(1966) reported such an experiment with the wheat (Triticum aestium L.) cultivar Norin 27 over 30 years ago. However, its value was not realized until recently. Atherton et al. (1990) conducted a similar experiment with carrot (Daucus carota L.) and estimated the optimum, minimum and maximum temperatures for vernalization using a bilinear model. Applying the same methodology, Craigon et al. (1995) reanalysed the data of Chujo (1966). Brooking (1996) also reanalysed the data of Chujo (1966), approaching the issue from a so-called developmental perspective. Robertson et al. (1996) conducted research similar to that reported by Chujo (1966) using two wheat cultivars, and analysed the final leaf number data following the methodology of Brooking (1996). We recently reported the application of an expression that summarizes the temperature response of plant growth and development using only cardinal temperatures as parameters. This expression was successfully used to summarize growth and development data of maize, wheat, barley, beans, sorghum and lambsquarters (Yan and Hunt, 1999). We hypothesize that the expression should also be effective in summarizing the vernalization response, since vernalization is but a kind of temperature response with low optimum temperatures. In this paper we report on the application of the previously used expression to summarize data dealing with the effects of vernalization on wheat (Chujo, 1966) and carrot (Atherton et al., 1990). # 1999 Annals of Botany Company
616
Yan and Hunt—Temperature Response of Vernalization MATERIALS AND METHODS
Data sources Wheat (Chujo, 1966). Germinated seeds of winter wheat (Triticum aestium L. ‘ Norin 27 ’) were treated at 1, 4, 8, 11, 15 and 18 mC for 20, 30, 40, 50 and 60 d before they were transplanted in a glasshouse (temperatures unreported but assumed to be 18 mC) with a 20 h photoperiod. The final leaf number and time from end of treatment to final leaf appearance were recorded (Fig. 1). Carrot (Atherton et al., 1990). Ten-week-old seedlings of carrot (Daucus carota L. ‘ Chantenay Red Cored ’) were treated at temperatures of k1, 3, 5, 7, 10, 12 and 16 mC for 9, 12 and 15 weeks. Treated plants were grown at 16 mC and 16 h daylength for 2 weeks before being transferred to a glasshouse at 18 mC and 16 h daylength. Days from the end of treatment to internode appearance and inflorescence appearance, the percentage of flowered plants, and the total leaf number below the inflorescence were reported ; data dealing with time to internode appearance (Fig. 1) are used here. Temperature response function The temperature response function was the same as that used previously by Yan and Hunt (1999). This function describes the rate (r) of development (or growth) at any temperature T as the maximum rate (Rmax) multiplied by a function of temperature, f(T ) : r l Rmax f(T )
(1)
Alternatively, if the rate is presented as relative to the maximum rate, (1 A) r\Rmax l f(T ) The expression of f(T ) contains only the minimum (assumed to be 0 mC and therefore omitted from the formula), optimum (Topt) and maximum (Tmax) temperatures.
0
Topt Tmax−Topt
10 1
T kT f(T ) l max TmaxkTopt
T Topt
(2)
To describe the daily vernalization progress at any applied temperature T, eqn (1) is written as : l Vmax f(T )
(3)
or, when Vmax is known, \Vmax l f(T )
(3 A)
where Vmax denotes the maximum daily rate of vernalization, which occurs only when the plant is grown at Topt for vernalization. Therefore, if the plant is treated at several temperatures and if the rate of vernalization, , at these temperatures is quantified appropriately, the temperature response of vernalization can be summarized by fitting eqn (3). Equation 3 A can be used only when Vmax is known. Data analysis The rate of vernalization of wheat at each temperature was first calculated by taking the reciprocal of the number
of treatment days required for the plant to reach final leaf appearance in 41 d after termination of treatment. The required periods of treatment, obtained through visual interpolation, ranged from 59 d at 16 mC to 29 d at 7 mC (Fig. 1 A). Forty-one days to final leaf appearance was chosen both to avoid having to extrapolate beyond the range of values encompassed in the data set, and to minimize the error of interpolation while still sampling the whole range of treatment temperatures. The parameters of the temperature response function of vernalization (Vmax, Topt and Tmax) were then calculated by fitting eqn (3) to the vernalization rate data. Once the parameters were obtained, indications of the vernalization status at the end of each treatment were calculated by multiplying the daily vernalization progress for each temperature by the appropriate number of days of treatment. The values thus obtained are referred to as degree of vernalization (DOV). The plant is considered as fully vernalized when DOV 1n0. Alternatively, the status of vernalization at the end of each treatment can be calculated by multiplying the daily relative vernalization progress (\Vmax) for each temperature by the appropriate number of days of treatment. The values so obtained are referred to as standard vernalization days (SVD). The rate of vernalization progress for carrots at each treatment temperature could not be obtained in the same way as wheat for several reasons. First, there were only three different durations of treatment ; second, 100 % flowering occurred only for the 12 and 15 week treatments at 5 and 7 mC so that the 9 week duration could not be used to determine the rate of vernalization at each treatment temperature ; and third, the response curve for the 15 week treatments appeared to be abnormal, possibly reflecting observational errors at 10 or 12 mC (Fig. 1 B). Consequently, an after-effect approach was adopted. In this approach, the rate of development towards internode appearance measured from the end of treatment ( p) was used when fitting the following function : p l Pmax f(T )
(4)
where Pmax is the maximum rate of post-treatment development. The relative rate of post-treatment development can then be calculated through : p\Pmax l f(T )
(4 A)
The fitted Topt and Tmax using eqn (4) were regarded as reflecting the temperature response of vernalization, based on the assumption that the relative rate of post-treatment development after being treated at a temperature is an indicator of the effectiveness of this temperature for vernalization, i.e. \Vmax l p\Pmax. Temperature parameters were determined separately for each of the three durations of treatment. Since this approach could not estimate the maximum rate of vernalization (Vmax), the DOV could not be determined. But the standard vernalization days (SVD) could be calculated by multiplying the daily relative vernalization progress for each temperature (\Vmax, assumed to be equal to p\Pmax) by the appropriate number of days of treatment.
617
Yan and Hunt—Temperature Response of Vernalization B 95
20 d 30 d 40 d 50 d 60 d
*
*
* 1
* *
12 wk
85
15 wk
80 75 70 65 60 55 50 45 40 35 30 25
*
4 7 10 13 16 Treatment temperature (°C)
9 wk
90 Days to internode appearance after treatment
Days to final leaf apperance after treatment
A 131 126 121 116 111 106 101 96 91 86 81 76 71 66 61 56 51 46 41 36 31 26
19
20
2
5 8 11 Treatment temperature (°C)
14
F. 1. Days to a specific developmental stage after various vernalization treatments for wheat ‘ Norin 27 ’ (A) and carrot ‘ Chantenay Red Cored ’ (B). Data from Chujo (1966) and Atherton et al. (1990).
RESULTS The estimated Topt and Tmax for the vernalization of winter wheat ‘ Norin 27 ’ were 5n7 and 21n3 mC, respectively ; equivalent values for carrot ranged between 6n4 and 6n1 for Topt and 14n1 and 15n2 for Tmax (Table 1). The adjusted R# values were 0n936 and 0n978 for wheat and the 9-week carrot data respectively, but lower for the 12- and 15-week carrot data, presumably because the temperature treatments were few. Despite the large difference in the adjusted R# fit for the different durations of treatment, the estimated Topt and Tmax for carrot vernalization were close (Table 1). The estimated Vmax for wheat was 0n0346p0n0014, and therefore the vernalization requirement (l 1\Vmax) was around 29 d. The derived response curves for wheat and carrots (based on averaged parameters) are shown in Fig. 2. This figure
emphasizes that the range of vernalizing temperatures is much broader for the studied cultivar of wheat than for carrot. However, the fit to the data appears equally good for both species. The rate of development after termination of the vernalization treatments was related to the number of standard vernalization days that had accumulated prior to the end of each treatment period for both wheat and carrot (Fig. 3). For wheat there was a strong positive relationship between the rate of development after treatment and the calculated SVD, when SVD was less than about 30 d (the estimated vernalization requirement), but not when SVD was greater than 30 d. This presumably reflects the fact that 30 standard vernalization days satisfied the vernalization requirement of the cultivar, with additional days of treatment having an effect solely on post-vernalization
T 1. Estimated optimum and maximum temperatures for ernalization of wheat cultiar ‘ Norin 27 ’ and carrot cultiar ‘ Chantenay Red Cored ’
Wheat Carrot 9 week treatment 12 week treatment 15 week treatment Average
Vmax (1\d)
Tmax (mC)
Topt (mC)
Adjusted R#
n
0n0346p0n0014
21n3p1n37
5n7p0n49
0n936
6
— — — —
14n1p0n27 15n2p1n72 14n4p2n29 14n6
6n8p0n16 6n5p0n81 6n4p1n37 6n6
0n978 0n588 0n224
5 5 5
Data from Chujo (1966) and Atherton et al. (1990).
618
Yan and Hunt—Temperature Response of Vernalization
Relative rate of vernalization
1.2
DISCUSSION
1.0 0.8 0.6 0.4 0.2 0.0
0
3
6 9 12 15 18 Treatment temperature (°C)
21
24
Wheat: simulated Wheat: measured Carrot: simulated Carrot: measured F. 2. Derived temperature response curves for the vernalization of wheat ‘ Norin 27 ’ and carrot ‘ Chantenay Red Cored ’. Data from Chujo (1966) and Atherton et al. (1990).
development per se. There are two obvious outliers in Fig. 3A—treatments at 1 and 4 mC for 20 d. These outliers presumably reflect the common observation that the vernalization effect of low temperatures is small or lacking if provided for only a short period of time (e.g. Wang et al., 1995). For carrot, there was a strong linear relation with SVD over the whole range of SVD, indicating that the vernalization requirement of carrot is equal to or longer than 15 weeks, the longest duration applied in the experiment. The single outlier in Fig. 3 B (rate of development to internode appearance) is from the 15-week treatment at 10 mC, suggesting that this observation is questionable.
Although the phenomenon of vernalization was discovered early this century, much of it still remains a mystery, mainly due to its invisible nature. For wheat, the vernalization response has been noted during many phases of development, from as early as shortly after the embryo is formed to as late as when plants have developed many leaves (for a recent review see Wallace and Yan, 1998 : 147–206). Although a vernalization treatment generally causes smaller final leaf number and fewer days to a reproductive stage, the visibility of the treatment effect is highly dependent on the post-treatment conditions (photoperiod and temperature). Vernalization is really a process leading to an invisible physiological state that becomes manifest in the subsequent rate of reproductive development, which is faster in longer daylengths with long day species and at high temperatures. Due to this invisibility, neither the progress nor the completion of vernalization can be equated with any morphological stage. The plant’s vernalization status at any time during growth and development can be equated, however, with a somewhat abstract measure of cumulative vernalization days or degree of vernalization. The vernalization temperature function presented in this paper provided a means for calculating such a measure, provided that the two parameters, i.e. the optimum and maximum temperatures for vernalization, are known. The temperature function also provides a means for estimating the minimum number of days required for the completion of vernalization. Such a minimum duration, which has to be termed the vernalization requirement, is an essential parameter for crop simulation. We have demonstrated two approaches to obtaining the parameters of a vernalization temperature response function. The first is direct use of data from temperature and duration treatments that produce the same rate of
A
0.055
B
0.035 Rate of development to internode appearance after treatment (1/d)
Rate of development to final leaf appearance after treatment (1/d)
0.040
0.030 0.025 0.020 0.015 0.010
0.045
0.035
0.025
0.015
0.005 0.000
0
10 20 30 40 50 60 70 Standard vernalization days at end of treatment
0.005 20
60 80 100 120 40 Standard vernalization days at end of treatment
F. 3. Rate of development after vernalization treatment as a function of the number of standard vernalization days at the end of treatment for wheat ‘ Norin 27 ’ (A) and carrot ‘ Chantenay Red Cored ’ (B). Data from Chujo (1966) and Atherton et al. (1990).
Yan and Hunt—Temperature Response of Vernalization development after treatment, as illustrated by the wheat data. The second is the after-effect approach in which treatment temperatures are related to rates of development after termination of treatment, as illustrated by the carrot data. The first approach is ideal but datasets seldom allow its application. Most analyses of vernalization, e.g. Atherton et al. (1990) and Craigon et al. (1995) have thus used the aftereffect approach. Based on the 12-week treatments, Atherton et al. (1990) used a bilinear model to analyse the effect of vernalizing temperature on the rate of development of carrot from the end of treatment to internode appearance or flower initiation. They estimated the optimum, minimum and maximum temperatures for carrot vernalization to be k1, 6n5 and 16 mC, respectively. Craigon et al. (1995) used the same methodology as that of Atherton et al. (1990) to reanalyse the wheat data of Chujo (1966). They used the 30-d duration treatments and estimated minimum, optimum and maximum temperatures for wheat vernalization to be k4n8, 5n2 and 26n6 mC, respectively. These estimations, both for wheat and for carrot, are of the same magnitude as those reported in this paper. However, there are some subtle but important advantages in using the continuous function reported here. First, a bilinear model requires four parameters (two for each linear segment) to describe the temperature response curve while the continuous function used here only requires three (Tmax, Topt and Vmax or Pmax) ; second, the bilinear model involves a sharp change in rate of vernalization at the optimum temperature, while a continuous function describes a smooth curve ; third, estimated minimum temperatures from the bilinear models were lower than 0 mC for both wheat and carrot, while the continuous function is constrained to 0 mC, the latter being supported both by data for carrot (Atherton et al. 1990) and by the wheat data of Robertson et al. (1996). Both of these studies indicated no vernalization effect at k1 mC. According to the analysis presented here, the winter wheat ‘ Norin 27 ’ has a maximum vernalization temperature of 21n3p1n4 mC. This maximum appears high for vernalization, but is less than the value of 26n6 mC calculated from the same data by Craigon et al. (1995). Such high values would allow a winter wheat to develop, albeit with much delay, to
619
flowering even under mean temperatures of 20 mC or higher, particularly when such temperatures fluctuate within a range of p5 mC, as is often the case (e.g. Wang et al., 1995). Recognition that maximum temperature for vernalization is over 20 mC could thus eliminate the need to assume an interchangeability between vernalization and plant age (Wang et al., 1995). Recognition of a high maximum and no age effect would be consistent with common observations that temperatures consistently in excess of 20 mC can completely prevent the plant from progressing to reproduction, such as when winter wheat cultivars are sown in the summer. Thus, summarization of the temperature response curve of vernalization in the manner presented here not only has potential to simplify and improve the quantification and simulation of vernalization and hence crop modelling, but also to clarify some physiological aspects that have caused confusion in the past. LITERATURE CITED Atherton JG, Craigon J, Basher EA. 1990. Flowering and bolting in carrot. I. Juvenility, cardinal temperatures and thermal times for vernalization. Journal of Horticultural Science 65 : 423–429. Brooking IR. 1996. Temperature response of vernalization in wheat : A developmental analysis. Annals of Botany 78 : 507–512. Chujo H. 1966. Differences in vernalization effect in wheat under various temperatures. Proceedings of the Crop Science Society of Japan 35 : 177–186. Craigon J, Atherton JG, Sweet N. 1995. Modelling the effects of vernalization on progress to final leaf appearance in winter wheat. Journal of Agricultural Science, Cambridge 124 : 369–377. Robertson MJ, Brooking IR, Ritchie JT. 1996. Temperature response of vernalization in wheat : Modelling the effect on the final number of mainstream leaves. Annals of Botany 78 : 371–381. Wallace DH, Yan W. 1998. Plant breeding and whole-system crop physiology. Wallingford, Oxon, UK : CAB International. Wang SY, Ward RW, Ritchie JT, Fischer RA, Schulthess U. 1995. Vernalization in wheat. I. A model based on the interchangeability of plant age and vernalization duration. Field Crops Research 41 : 91–100. Whyte RO. 1948. History of research in vernalization. In : Murneek AE, Whyte RO, eds. Vernalization and photoperiodism. Waltham, MA : Chronica Notanica Company, 1–38. Yan W, Hunt LA. 1999. An equation for modelling the temperature response of plants using only the cardinal temperatures. Annals of Botany 84 : 607–614.
Reanalysis of Vernalization Data of Wheat and Carrot W E I K A I Y A N and L. A. H U N T Department of Plant Agriculture, Crop Science Diision, Uniersity of Guelph, Guelph, Ontario, Canada N1G 2W1 Received : 16 April 1999
Returned for revision : 20 May 1999
Accepted : 9 July 1999
Vernalization is an important determinant of the growth, development, and yield of biennial and perennial crops. Accurate simulation of its response to temperature is thus an important component of successful crop systems modelling. Vernalization has a low optimum temperature compared to other temperature responses of plants, and thus may be difficult to treat using expressions that are appropriate for other plant processes. This paper examines the application of a simple equation that has been used for other processes. It reads as l Vmax
0
TmaxkT
10 1 T
Topt Tmax−Topt
, where is the daily rate of vernalization progress at temperature T, Topt and TmaxkTopt Topt Tmax are the optimum and maximum temperatures for vernalization, respectively, and Vmax is the maximum daily rate of vernalization (the inverse of the minimum number of days required to complete vernalization), which occurs at Topt. The model was applied to published vernalization data for wheat and carrot. The fits to data were good (adjusted R# for wheat of 0n94, for carrot 0n98), with estimated Topt and Tmax being 5n7p0n5 and 21n3p1n4 mC, respectively, for wheat ‘ Norin 27 ’ and 6n6p0n2 and 14n1p0n3 mC for carrot ‘ Chantenay Red Cored ’. The estimated parameters, in particular the high Tmax for wheat, were close to those reported using different analytical approaches. It was suggested that the function would be useful for summarizing vernalization data, and that its use would avoid the abrupt changes that are inevitable when different linear relationships are used for part of the overall response. It was also suggested the high Tmax should be taken into account when interpreting data obtained with wheat grown under warm conditions. # 1999 Annals of Botany Company Key words : Plant, vernalization, temperature response, modelling, wheat (Triticum aestium L.), carrot (Daucus carota L.).
INTRODUCTION Vernalization response is a major determinant of the growth and development of biennial and perennial plants. It is generally agreed that the vernalization process has a low optimum temperature, but contrasting opinions exist as to the optimum temperature, the effective range of temperature, and the vernalization requirement for a given species or genotype. These contrasting opinions reflect a number of aspects. The first is the invisible nature of vernalization itself (Lysenko, from Whyte, 1948). The process can only be evaluated in terms of its after-effects, and thus speculation and imagination are required when trying to comprehend the existing evidence. The second is the lack of systematic experimentation, an aspect that reflects the cost of comprehensive experimental investigations of vernalization. The third is the lack of analytical models that can be applied to the existing data. Based on present understanding, a comprehensive investigation into vernalization would involve treating plant material (preferably germinated seeds) at a series of temperatures for a series of durations before moving them into a fully inductive environment (long photoperiods for a long day species plus normal to high temperatures). Chujo E-mail wyan!uoguelph.ca ; thunt!plant.uoguelph.ca
0305-7364\99\110615j05 $30.00\0
(1966) reported such an experiment with the wheat (Triticum aestium L.) cultivar Norin 27 over 30 years ago. However, its value was not realized until recently. Atherton et al. (1990) conducted a similar experiment with carrot (Daucus carota L.) and estimated the optimum, minimum and maximum temperatures for vernalization using a bilinear model. Applying the same methodology, Craigon et al. (1995) reanalysed the data of Chujo (1966). Brooking (1996) also reanalysed the data of Chujo (1966), approaching the issue from a so-called developmental perspective. Robertson et al. (1996) conducted research similar to that reported by Chujo (1966) using two wheat cultivars, and analysed the final leaf number data following the methodology of Brooking (1996). We recently reported the application of an expression that summarizes the temperature response of plant growth and development using only cardinal temperatures as parameters. This expression was successfully used to summarize growth and development data of maize, wheat, barley, beans, sorghum and lambsquarters (Yan and Hunt, 1999). We hypothesize that the expression should also be effective in summarizing the vernalization response, since vernalization is but a kind of temperature response with low optimum temperatures. In this paper we report on the application of the previously used expression to summarize data dealing with the effects of vernalization on wheat (Chujo, 1966) and carrot (Atherton et al., 1990). # 1999 Annals of Botany Company
616
Yan and Hunt—Temperature Response of Vernalization MATERIALS AND METHODS
Data sources Wheat (Chujo, 1966). Germinated seeds of winter wheat (Triticum aestium L. ‘ Norin 27 ’) were treated at 1, 4, 8, 11, 15 and 18 mC for 20, 30, 40, 50 and 60 d before they were transplanted in a glasshouse (temperatures unreported but assumed to be 18 mC) with a 20 h photoperiod. The final leaf number and time from end of treatment to final leaf appearance were recorded (Fig. 1). Carrot (Atherton et al., 1990). Ten-week-old seedlings of carrot (Daucus carota L. ‘ Chantenay Red Cored ’) were treated at temperatures of k1, 3, 5, 7, 10, 12 and 16 mC for 9, 12 and 15 weeks. Treated plants were grown at 16 mC and 16 h daylength for 2 weeks before being transferred to a glasshouse at 18 mC and 16 h daylength. Days from the end of treatment to internode appearance and inflorescence appearance, the percentage of flowered plants, and the total leaf number below the inflorescence were reported ; data dealing with time to internode appearance (Fig. 1) are used here. Temperature response function The temperature response function was the same as that used previously by Yan and Hunt (1999). This function describes the rate (r) of development (or growth) at any temperature T as the maximum rate (Rmax) multiplied by a function of temperature, f(T ) : r l Rmax f(T )
(1)
Alternatively, if the rate is presented as relative to the maximum rate, (1 A) r\Rmax l f(T ) The expression of f(T ) contains only the minimum (assumed to be 0 mC and therefore omitted from the formula), optimum (Topt) and maximum (Tmax) temperatures.
0
Topt Tmax−Topt
10 1
T kT f(T ) l max TmaxkTopt
T Topt
(2)
To describe the daily vernalization progress at any applied temperature T, eqn (1) is written as : l Vmax f(T )
(3)
or, when Vmax is known, \Vmax l f(T )
(3 A)
where Vmax denotes the maximum daily rate of vernalization, which occurs only when the plant is grown at Topt for vernalization. Therefore, if the plant is treated at several temperatures and if the rate of vernalization, , at these temperatures is quantified appropriately, the temperature response of vernalization can be summarized by fitting eqn (3). Equation 3 A can be used only when Vmax is known. Data analysis The rate of vernalization of wheat at each temperature was first calculated by taking the reciprocal of the number
of treatment days required for the plant to reach final leaf appearance in 41 d after termination of treatment. The required periods of treatment, obtained through visual interpolation, ranged from 59 d at 16 mC to 29 d at 7 mC (Fig. 1 A). Forty-one days to final leaf appearance was chosen both to avoid having to extrapolate beyond the range of values encompassed in the data set, and to minimize the error of interpolation while still sampling the whole range of treatment temperatures. The parameters of the temperature response function of vernalization (Vmax, Topt and Tmax) were then calculated by fitting eqn (3) to the vernalization rate data. Once the parameters were obtained, indications of the vernalization status at the end of each treatment were calculated by multiplying the daily vernalization progress for each temperature by the appropriate number of days of treatment. The values thus obtained are referred to as degree of vernalization (DOV). The plant is considered as fully vernalized when DOV 1n0. Alternatively, the status of vernalization at the end of each treatment can be calculated by multiplying the daily relative vernalization progress (\Vmax) for each temperature by the appropriate number of days of treatment. The values so obtained are referred to as standard vernalization days (SVD). The rate of vernalization progress for carrots at each treatment temperature could not be obtained in the same way as wheat for several reasons. First, there were only three different durations of treatment ; second, 100 % flowering occurred only for the 12 and 15 week treatments at 5 and 7 mC so that the 9 week duration could not be used to determine the rate of vernalization at each treatment temperature ; and third, the response curve for the 15 week treatments appeared to be abnormal, possibly reflecting observational errors at 10 or 12 mC (Fig. 1 B). Consequently, an after-effect approach was adopted. In this approach, the rate of development towards internode appearance measured from the end of treatment ( p) was used when fitting the following function : p l Pmax f(T )
(4)
where Pmax is the maximum rate of post-treatment development. The relative rate of post-treatment development can then be calculated through : p\Pmax l f(T )
(4 A)
The fitted Topt and Tmax using eqn (4) were regarded as reflecting the temperature response of vernalization, based on the assumption that the relative rate of post-treatment development after being treated at a temperature is an indicator of the effectiveness of this temperature for vernalization, i.e. \Vmax l p\Pmax. Temperature parameters were determined separately for each of the three durations of treatment. Since this approach could not estimate the maximum rate of vernalization (Vmax), the DOV could not be determined. But the standard vernalization days (SVD) could be calculated by multiplying the daily relative vernalization progress for each temperature (\Vmax, assumed to be equal to p\Pmax) by the appropriate number of days of treatment.
617
Yan and Hunt—Temperature Response of Vernalization B 95
20 d 30 d 40 d 50 d 60 d
*
*
* 1
* *
12 wk
85
15 wk
80 75 70 65 60 55 50 45 40 35 30 25
*
4 7 10 13 16 Treatment temperature (°C)
9 wk
90 Days to internode appearance after treatment
Days to final leaf apperance after treatment
A 131 126 121 116 111 106 101 96 91 86 81 76 71 66 61 56 51 46 41 36 31 26
19
20
2
5 8 11 Treatment temperature (°C)
14
F. 1. Days to a specific developmental stage after various vernalization treatments for wheat ‘ Norin 27 ’ (A) and carrot ‘ Chantenay Red Cored ’ (B). Data from Chujo (1966) and Atherton et al. (1990).
RESULTS The estimated Topt and Tmax for the vernalization of winter wheat ‘ Norin 27 ’ were 5n7 and 21n3 mC, respectively ; equivalent values for carrot ranged between 6n4 and 6n1 for Topt and 14n1 and 15n2 for Tmax (Table 1). The adjusted R# values were 0n936 and 0n978 for wheat and the 9-week carrot data respectively, but lower for the 12- and 15-week carrot data, presumably because the temperature treatments were few. Despite the large difference in the adjusted R# fit for the different durations of treatment, the estimated Topt and Tmax for carrot vernalization were close (Table 1). The estimated Vmax for wheat was 0n0346p0n0014, and therefore the vernalization requirement (l 1\Vmax) was around 29 d. The derived response curves for wheat and carrots (based on averaged parameters) are shown in Fig. 2. This figure
emphasizes that the range of vernalizing temperatures is much broader for the studied cultivar of wheat than for carrot. However, the fit to the data appears equally good for both species. The rate of development after termination of the vernalization treatments was related to the number of standard vernalization days that had accumulated prior to the end of each treatment period for both wheat and carrot (Fig. 3). For wheat there was a strong positive relationship between the rate of development after treatment and the calculated SVD, when SVD was less than about 30 d (the estimated vernalization requirement), but not when SVD was greater than 30 d. This presumably reflects the fact that 30 standard vernalization days satisfied the vernalization requirement of the cultivar, with additional days of treatment having an effect solely on post-vernalization
T 1. Estimated optimum and maximum temperatures for ernalization of wheat cultiar ‘ Norin 27 ’ and carrot cultiar ‘ Chantenay Red Cored ’
Wheat Carrot 9 week treatment 12 week treatment 15 week treatment Average
Vmax (1\d)
Tmax (mC)
Topt (mC)
Adjusted R#
n
0n0346p0n0014
21n3p1n37
5n7p0n49
0n936
6
— — — —
14n1p0n27 15n2p1n72 14n4p2n29 14n6
6n8p0n16 6n5p0n81 6n4p1n37 6n6
0n978 0n588 0n224
5 5 5
Data from Chujo (1966) and Atherton et al. (1990).
618
Yan and Hunt—Temperature Response of Vernalization
Relative rate of vernalization
1.2
DISCUSSION
1.0 0.8 0.6 0.4 0.2 0.0
0
3
6 9 12 15 18 Treatment temperature (°C)
21
24
Wheat: simulated Wheat: measured Carrot: simulated Carrot: measured F. 2. Derived temperature response curves for the vernalization of wheat ‘ Norin 27 ’ and carrot ‘ Chantenay Red Cored ’. Data from Chujo (1966) and Atherton et al. (1990).
development per se. There are two obvious outliers in Fig. 3A—treatments at 1 and 4 mC for 20 d. These outliers presumably reflect the common observation that the vernalization effect of low temperatures is small or lacking if provided for only a short period of time (e.g. Wang et al., 1995). For carrot, there was a strong linear relation with SVD over the whole range of SVD, indicating that the vernalization requirement of carrot is equal to or longer than 15 weeks, the longest duration applied in the experiment. The single outlier in Fig. 3 B (rate of development to internode appearance) is from the 15-week treatment at 10 mC, suggesting that this observation is questionable.
Although the phenomenon of vernalization was discovered early this century, much of it still remains a mystery, mainly due to its invisible nature. For wheat, the vernalization response has been noted during many phases of development, from as early as shortly after the embryo is formed to as late as when plants have developed many leaves (for a recent review see Wallace and Yan, 1998 : 147–206). Although a vernalization treatment generally causes smaller final leaf number and fewer days to a reproductive stage, the visibility of the treatment effect is highly dependent on the post-treatment conditions (photoperiod and temperature). Vernalization is really a process leading to an invisible physiological state that becomes manifest in the subsequent rate of reproductive development, which is faster in longer daylengths with long day species and at high temperatures. Due to this invisibility, neither the progress nor the completion of vernalization can be equated with any morphological stage. The plant’s vernalization status at any time during growth and development can be equated, however, with a somewhat abstract measure of cumulative vernalization days or degree of vernalization. The vernalization temperature function presented in this paper provided a means for calculating such a measure, provided that the two parameters, i.e. the optimum and maximum temperatures for vernalization, are known. The temperature function also provides a means for estimating the minimum number of days required for the completion of vernalization. Such a minimum duration, which has to be termed the vernalization requirement, is an essential parameter for crop simulation. We have demonstrated two approaches to obtaining the parameters of a vernalization temperature response function. The first is direct use of data from temperature and duration treatments that produce the same rate of
A
0.055
B
0.035 Rate of development to internode appearance after treatment (1/d)
Rate of development to final leaf appearance after treatment (1/d)
0.040
0.030 0.025 0.020 0.015 0.010
0.045
0.035
0.025
0.015
0.005 0.000
0
10 20 30 40 50 60 70 Standard vernalization days at end of treatment
0.005 20
60 80 100 120 40 Standard vernalization days at end of treatment
F. 3. Rate of development after vernalization treatment as a function of the number of standard vernalization days at the end of treatment for wheat ‘ Norin 27 ’ (A) and carrot ‘ Chantenay Red Cored ’ (B). Data from Chujo (1966) and Atherton et al. (1990).
Yan and Hunt—Temperature Response of Vernalization development after treatment, as illustrated by the wheat data. The second is the after-effect approach in which treatment temperatures are related to rates of development after termination of treatment, as illustrated by the carrot data. The first approach is ideal but datasets seldom allow its application. Most analyses of vernalization, e.g. Atherton et al. (1990) and Craigon et al. (1995) have thus used the aftereffect approach. Based on the 12-week treatments, Atherton et al. (1990) used a bilinear model to analyse the effect of vernalizing temperature on the rate of development of carrot from the end of treatment to internode appearance or flower initiation. They estimated the optimum, minimum and maximum temperatures for carrot vernalization to be k1, 6n5 and 16 mC, respectively. Craigon et al. (1995) used the same methodology as that of Atherton et al. (1990) to reanalyse the wheat data of Chujo (1966). They used the 30-d duration treatments and estimated minimum, optimum and maximum temperatures for wheat vernalization to be k4n8, 5n2 and 26n6 mC, respectively. These estimations, both for wheat and for carrot, are of the same magnitude as those reported in this paper. However, there are some subtle but important advantages in using the continuous function reported here. First, a bilinear model requires four parameters (two for each linear segment) to describe the temperature response curve while the continuous function used here only requires three (Tmax, Topt and Vmax or Pmax) ; second, the bilinear model involves a sharp change in rate of vernalization at the optimum temperature, while a continuous function describes a smooth curve ; third, estimated minimum temperatures from the bilinear models were lower than 0 mC for both wheat and carrot, while the continuous function is constrained to 0 mC, the latter being supported both by data for carrot (Atherton et al. 1990) and by the wheat data of Robertson et al. (1996). Both of these studies indicated no vernalization effect at k1 mC. According to the analysis presented here, the winter wheat ‘ Norin 27 ’ has a maximum vernalization temperature of 21n3p1n4 mC. This maximum appears high for vernalization, but is less than the value of 26n6 mC calculated from the same data by Craigon et al. (1995). Such high values would allow a winter wheat to develop, albeit with much delay, to
619
flowering even under mean temperatures of 20 mC or higher, particularly when such temperatures fluctuate within a range of p5 mC, as is often the case (e.g. Wang et al., 1995). Recognition that maximum temperature for vernalization is over 20 mC could thus eliminate the need to assume an interchangeability between vernalization and plant age (Wang et al., 1995). Recognition of a high maximum and no age effect would be consistent with common observations that temperatures consistently in excess of 20 mC can completely prevent the plant from progressing to reproduction, such as when winter wheat cultivars are sown in the summer. Thus, summarization of the temperature response curve of vernalization in the manner presented here not only has potential to simplify and improve the quantification and simulation of vernalization and hence crop modelling, but also to clarify some physiological aspects that have caused confusion in the past. LITERATURE CITED Atherton JG, Craigon J, Basher EA. 1990. Flowering and bolting in carrot. I. Juvenility, cardinal temperatures and thermal times for vernalization. Journal of Horticultural Science 65 : 423–429. Brooking IR. 1996. Temperature response of vernalization in wheat : A developmental analysis. Annals of Botany 78 : 507–512. Chujo H. 1966. Differences in vernalization effect in wheat under various temperatures. Proceedings of the Crop Science Society of Japan 35 : 177–186. Craigon J, Atherton JG, Sweet N. 1995. Modelling the effects of vernalization on progress to final leaf appearance in winter wheat. Journal of Agricultural Science, Cambridge 124 : 369–377. Robertson MJ, Brooking IR, Ritchie JT. 1996. Temperature response of vernalization in wheat : Modelling the effect on the final number of mainstream leaves. Annals of Botany 78 : 371–381. Wallace DH, Yan W. 1998. Plant breeding and whole-system crop physiology. Wallingford, Oxon, UK : CAB International. Wang SY, Ward RW, Ritchie JT, Fischer RA, Schulthess U. 1995. Vernalization in wheat. I. A model based on the interchangeability of plant age and vernalization duration. Field Crops Research 41 : 91–100. Whyte RO. 1948. History of research in vernalization. In : Murneek AE, Whyte RO, eds. Vernalization and photoperiodism. Waltham, MA : Chronica Notanica Company, 1–38. Yan W, Hunt LA. 1999. An equation for modelling the temperature response of plants using only the cardinal temperatures. Annals of Botany 84 : 607–614.