Mol Gen Genet (1991) 229:57-66 0026892591002784 © Springer-Verlag 1991

A genetic and physiological analysis of late flowering mutants in Arabidopsis thaliana M. Koornneef, C.J. Hanhart, and J.H. van der Veen Department of Genetics, Wageningen Agricultural University, Dreijenlaan 2, NL-6703 HA Wageningen, The Netherlands Received April 3, 1991

Summary. Monogenic mutants of the early ecotype Landsberg erecta were selected on the basis of late flowering under long day (LD) conditions after treatment with ethyl methanesulphonate or irradiation. In addition to later flowering the number of rosette and cauline leaves is proportionally higher in all mutants, although the correlation coefficient between the two parameters is not the same for all genotypes. Forty-two independently induced mutants were found to represent mutations at 11 loci. The mutations were either recessive, intermediate (co locus) or almost completely dominant ~wa locus). The loci are located at distinct positions on four of the five Arabidopsis chromosomes. Recombinants carrying mutations at different loci flower later than or as late as the later parental mutant. This distinction led to the assignment ofeight of the loci to three epistatic groups. In wild type, vernalization promotes flowering to a small extent. For mutants at the loci fca, fve, fy and fpa, vernalization has a large effect both under LD and short day (SD) conditions, whereas co, gi, fd and fwa mutants are almost completely insensitive to this treatment. SD induces later flowering except for mutants at the co and gi loci, which flower with the same number of leaves under LD and SD conditions. This differential response of the mutants to environmental factors and their subdivision into epistatic groups is discussed in relation to a causal model for floral initiation in Arabidop-

sis thaliana. Key words: A r a b i d o p s i s - Flowering - Vernalization Photoperiodism - Linkage map

Introduction

Arabidopsis thaliana is a plant with separated vegetative and reproductive phases. The transformation of the vegetative shoot meristem into a reproductive inflorescence Offprint requests to: M. Koornneef

meristem becomes apparent by changes in the structure (Vaughan 1955) and in the pattern of mitotic activity (Besnard-Wibaut 1977) of the shoot apex. At the macroscopic level this change becomes visible with the appearance of a main stem, rapidly bolting from the rosette. This elongated stem carries several cauline leaves or bracts below the first flower. Inflorescences, which are open racemes with typical crucifer flowers, subsequently also develop from axillary buds in some of the rosette leaves and in most cauline leaves. The phenomenology of flowering in Arabidopsis has been described by several authors (Laibach 1951; Vaughan 1955; Besnard-Wibaut 1977; Irish and Sussex 1990; Smyth et al. 1990). The number of leaves (leaf number, LN) and, correlated with it, the time of floral initiation, depends strongly on environmental factors and genotype (NappZinn 1969, 1985). Flowering is promoted by long day (LD) photoperiods and low temperatures (vernalization) but neither requirement is absolute and probably with very few exceptions all genotypes will eventually flower without these treatments. Genetic differences in flowering time (FT) conspicuously contribute to variation within and between natural populations (Laibach 1951 ; Jones 1971 ; Napp-Zinn 1985). This variation in FT has in some cases been related to differential sensitivity to vernalization (e.g. Jones 1971) and daylength (e.g. Laibach 1951), reflecting local adaptation. Most ecotypes that are used in molecular and genetic experiments, such as Columbia and Landsberg erecta, flower within 3-4 weeks after sowing, with 6-8 rosette leaves and 2-3 cauline leaves, when normal growth conditions and LD are provided. These early genotypes show only a small response to vernalization. However, some ecotypes, especially those from Northern latitudes, such as Stockholm and S6derland, flower very late and are very responsive to vernalization. Lateness of ecotypes is mostly dominant and differs from early ecotypes by a limited number of both dominant and recessive genes (Napp-Zinn 1957; Karlovska 1974). A cross between the medium early ecotypes Dijon and Limburg-2, each containing a dominant gene for late

58 flowering, gave an FI heterotic towards late and segregated in the F 2 very late double homozygotes (Van der Veen 1965). Mutants affecting FT are frequently observed in M 2 populations obtained upon mutagenic treatment. Mutant induction experiments in Arabidopsis have mostly been performed with early ecotypes such as Estland (Est), Landsberg erecta (Ler) and Columbia (Col) under LD conditions, so almost exclusively mutants were isolated in which FT is delayed (McKelvie 1962; R6dei 1962; Hussein 1968; Vetrilova 1973). Starting with late flowering mutants, Hussein (1968) has also reported the isolation of mutants which flower earlier than the parental genotype. In all cases reported, a delay in FT correlates highly with an increase in the number of leaves in the rosette and on the main stem, so the later flowering results from an extended vegetative period and not from retarded growth. Genetic analysis, where performed (R6dei 1962; Hussein 1968; Vetrilova 1973), has shown that the mutants are mostly recessive but sometimes (semi-) dominant, depending on the locus. Allelism was tested in only a few cases (e.g. Hussein 1968) and linkage studies were reported by R6dei (1962), Hirono and R6dei (1965) and Koornneef et al. (1983), resulting in map positions for seven FT loci (Koornneef 1990). It can be expected that the characterization of loci affecting the initiation of flowering will ultimately allow the cloning of these genes. The phase change from vegetative to reproductive development, that is the initiation of an inflorescence meristem, precedes the formation of flower primordia which requires other highly specific genes (Bowman et al. 1989). In this report we describe the genetic analysis of 39 late flowering mutants induced by us and 1 isolated by Dr. R. Pruitt in the Ler ecotype and compare these with the mutants described by R6dei (1962). A physiological characterization was performed to classify the different loci in relation to their daylength and vernalization response. Finally a tentative model will be proposed for the role of the respective genes in the initiation of flowering in Arabidopsis.

Materials and methods

Plant material. Mutants flowering later and with more rosette leaves than wild type were selected in the selfed progeny of mutagen-treated Landsberg erecta (wild type) seeds, growing under long day conditions. Late mutants not showing a proportional increase in leaf number were discarded, since in these mutants reduced growth was thought to be the main cause of the delay in flowering time. Mutations were induced by treating seeds either with 10raM ethyl methanesulphonate (EMS) or with different doses of X-rays or fast neutrons as described by Koornneef et al. (1982a). The following mutant lines, carrying X-ray-induced mutations at FT loci: co (constants) in combination with a previously undescribed chlorophyll mutant, gi 2 (gigantea), and /d (luminidependens) described by Redei (1962), were all kindly provided by Dr. G. R6dei, R6dei's co allele, now

in combination with the chlorophyll mutant lu (lutescens), was received from Dr. R61ichova (Brno). Probably the same allele but without the additional mutant phenotype, was obtained from Dr. Komeda (Tokyo), who received it from Dr. R6dei. For physiological experiments, representative mutant alleles at 11 loci were used in the Ler background. These lines did not show any other visible mutations. Genetic characterization. Mutant lines derived from selfed progenies of selected M 2 plants were crossed with the wild type and with a gradually built-up set of tester mutants. Allelism was inferred from the mutant phenotype in the F1 and the absence of wild-type (early with low leaf number) plants in the F2 generation. Linkage analysis was done in F2s derived from crosses of late flowering mutants with lines carrying one or several monogenic recessive morphological markers which had been mapped previously (Koornneef et al. 1983). F3 progeny testing was applied: (a) in cases where F 2 scoring was ambiguous; (b) to confirm the phenotype of rare recombinants; and (c) to find recombinants in repulsion phase F2s where no homozygous recombinants between specific markers were observed. Recombination frequencies were estimated by maximum likelihood procedures as described by Koornneef and Stare (1988). Map positions were calculated from the estimated recombination fractions, using the program "Genmap" written by P. Stare (personal communication), which is based on the procedures described by Jensen and J6rgensen (1975). Recombination data obtained previously (Koornneef et al. 1983) for a number of the FT loci ~ca and ft) and other morphological markers were included in this analysis. The behaviour of recombinants (" double mutants") from mutant x mutant crosses at 8 of the 11 loci was studied as part of student exercises in our department. In general, the distribution of FT and LN in digenically segregating generations is continuous, often without distinct peaks. For F3 analysis, about 100, sometimes more, F3 lines per cross were used, which derived from randomly chosen F2 plants. Line means for FT and LN (rosette) enabled the plotting of scatter diagrams for visual inspection. Homozygosity of the respective F2 parents could be assessed from the within line variances (10-12 plants per line). From "Fz-diallel analyses", separate F2 histograms for FT and (rosette) LN, were constructed for a number of mutant x mutant combinations. The F2 population size was in most cases 200 plants (4 replicates of 50); for co/gi, 350 (7 replicates of 50 plants). Growth conditions and physiological characterization. Seeds were sown in plastic petri dishes on water-saturated filter paper and (except for physiological experiments, see later) incubated for 3-5 days at 2 - 5 ° C to break dormancy. The seeds were subsequently incubated for 2 days at 25 ° C and 16 h light (15 W/m 2) and then transplanted to soil with a small brush when radicle protrusion could be observed. In earlier experiments (mutant isolation, some of the mutant × mutant crosses and

59

tively, the previously described loci co and gi (R6dei 1962). Mutants allelic to the Id mutation (R6dei 1962) were not found. This implies that mutations in at least 12 different loci can result in later flowering and an increased LN. The results (Table 1) also suggest that mutations are not randomly distributed over the different loci. Despite the fact that the two traits (FT and LN) show variation within homozygous genotypes growing in a uniform environment, monogenic inheritance can be concluded from segregation analysis of LN in F2 generations derived from mutant x wild type crosses (Fig. 1; Table 2). Similar observations on FT (data not shown) give complete confirmation of the monogenic basis of

linkage analyses) seeds were sown on Perlite and treated as described by Koornneef et al. (1982b). All genetic experiments were performed in an air-conditioned greenhouse. From the middle of September until April additional light was given providing, respectively, a daylength of at least 14 h and a light intensity sufficient to allow normal growth. Day temperature was 22-25 ° C and night temperature 16-19 ° C. To investigate the effect of daylength, plants were grown in a climate room at 24 ° C with 9 h (short day, SD) illumination by fluorescent light tubes (Philips TL 57) at 25 W/m 2 supplemented with 5 W/m 2 incandescent bulbs. For the long day (LD) treatment, the same climate room was used with only half of the light intensity of both the fluorescent and incandescent light for a period of 18 h. All plants were grown in soil. Vernalization response was tested by incubating seeds for 22 days in darkness in a cold room at 3° C and subsequently transplanting the etiolated seedlings into soil. For the first 2 days after transplantation the seedlings were protected from excessive water loss by covering them loosely with a layer of filter paper. The seeds used for these experiments were harvested at least I year beforehand and therefore did not require a cold treatment to break dormancy. FT was recorded as the number of days from the time at which plants were placed at 24°C to the time of opening of the first flower. At the same time the number of leaves on the rosette (excluding cotyledons) and on the main stem were counted.

Table 1. Results o f allelism tests o f independently isolated late flowering m u t a n t s Locus

Ethyl methanesulphonate

fy fpa fve fca fe ft fd fwa fha co(=fg) gi (=fb)

Results Genetic characterization

A collection of 42 independent mutants flowering later than the corresponding wild type could be assigned to 11 different loci on the basis of allelism tests. Mutants at the locifg a n d f b were found to be alleles of, respec-

Table 2. Segregation ratio and recessiveness for leaf n u m b e r o f m u t a t i o n s at specific f o w e r i n g time loci

Locus

30 109

X-rays

Neutrons

Total

1

1 2 2 8 1 3 1 2 3

9 7

ib 2b

10 9

34

6

2

1

1

2

42

Includes one m u t a n t isolated by Dr. R. Pruitt b Includes in each group one m u t a n t described by R6dei (1962)

a

Middle

109 106 101 96 100 95 98 31

fha

co-3 co-4 gi-3

1 2 2 5" 1 3 1 1 2

N u m b e r o f plants with leaf n u m b e r Low

fy fpa fve fca fe ft fd fwa

N u m b e r o f i n d e p e n d e n t m u t a n t s induced by:

nt b 72 nt b

Segregation ratio

)~2

Recessiveness o f the m u t a n t allele"

High 32 30 34 29 33 38 39 107

3:1 3:1 3:1 3:1 3:1 3 :1 3:1 1:3

0.40 0.62 0.01 0.22 0.01 0.90 0.88 0.47

35

1 : 2:1

0.72

m nc s nc nc nc nc Partial d o m i n a n t nc Intermediate

21

3 :1

5.43 °

nc

s, slight; m, m o d e r a t e ; nc, strong but n o t complete b nt, n o t tested, since in this p o p u l a t i o n no classification was possible 0.01 < P < 0 . 0 5

60 50

.*

/~0 30 20

Ika _ L. 1"

10 0 40

m

ft

"*

,o

30 ¸ 20 4r

*

fha

~ 10 13.

~

0

E z 30

y'-3

CO-W

20 10 0

I

fca

•

co-3

...

2O

10.

5

15

25

35

5

15

25

35

5

15

25

35

Number of leaves Fig. 1. Frequency distributions for leaf number (LN) of F2 generations derived from crosses between late flowering mutants and wild type. Arrows indicate the average values for the wild-type parent, the heterozygote and the mutant parent grown in the same plot.

The horizontal lines represent the parental ranges. Numbers added to gene symbols indicate mutant allele number. Symbols without numbers represent mutant allele 1

the mutations. The slight recessive deficit for the gi mutant (Table 2) is not significant at the 1% level and is in general not a characteristic of this mutant. Most o f the mutants are recessive, but the heterozygous progeny (F1 and F J o f the cross m u t a n t (maternal parent) × wild type are to varying degrees, later than the early wild type. This degree o f recessiveness (Table 2) appears to be locus-specific. The two fwa mutants are more or less dominant, and all co mutants (including the less extreme co-4) give an intermediate heterozygote. N e w estimates of recombination percentages between F T mutants and morphological markers are collected in Table 3. Since we found that gi is allelic to f b and

that R 6 d e i ' s p a m u t a n t (Hirono and R6dei 1965) is allelic to McKelvie's lepida (le) m u t a n t (McKelvie 1962), the accurate linkage data for the g i - c h l - I e region published by Hirono and R6dei (1965) were used to supplement our data for the determination of the m a p position of the 11 F T loci (Fig. 2). Two morphological markers (re, reticulata; chp7, pale green), shown in Fig. 2 were not described in the latest classical m a p (Koornneef 1990), although re was located by R6dei and Hirono (1964) at a similar position on c h r o m o s o m e 2. The 11 F T loci are located on all Arabidopsis chromosomes except c h r o m o s o m e 3, and no clustering can be observed.

61 Table 3. Estimates of recombination percentages between markers and late flowering loci Markers

Method"

Recombination (%)

R2 R3 R2 R3 R2 C2 R2 C2 C2 R2 C2 R2 R2 C2 R2 C2

0.0 + 6.2 14.3 +_5.0 23.0 +_5.8 36.4+_ 8.5 36.5 +_2.7 21.4 +- 1.2b 29.5 _ 3 . 4 b 36.5 +_2.7 12.7 _+1.0b 14.9 +- 3.5b 24.8 +- 1.6 42.8 _+3.3b 50.6 _+5.5 34.2 +_1.0u 36.1 +_2.9b 21.5 ___2.8

R2 R3 R3 R3 R2 R3 R2 C2 C2 R2 C2 C2 R2 R2 C2 C2 C2 R2 R2 R2 C2 R2 R2

0.0 _+5.0 6.7 _+2.2 4.4 +_2.2 19.0_+4.8 31.5 _ 4.4 25.2 +-4.4 35.6 +-4.2 19.5 -/- 1.2b 12.5 +_1.4 22.6 +_5.5 25.5 +_1.28 34.0 +_1.4 38.5 +4.6 18.6 +_3.4 17.7 + 1.7b 32.9 +- 2.4 28.3 +_2.6 0.0 +_5.9 19.7 _ 5.6 26.7 +-4.4 13.7 +_0.9 28.3 + 4.9 16.9 + 5.3

Markers

Chromosome 1

fha/ga4 fha/ga4 fha/dis2 fha/dis2 fha/cer5 ga4/dis2 ga4/chl ga4/cer5 dis2/chl dis2/chl dis2/cer5 dis2/apl dis2/fe chl/apl chl/apl apl/fe

Recombination (%)

C2 R2 R2 C2 R2 C2 R2 C2 R2 C2 C2 R2 C2 R2 C2 R2 C2 R2 C2 R2 C2 R2

32.3 ___1.7 33.9 ___4.9 b 42.8 + 7.2 33.6 + 5.5 43.4 +- 7.0 41.4 + 1.8 b 39.8 + 3.4b 1.2 _+0.4 0.0 +_3.4 b 3.9 + 0.8 9.5 _ 1.1 12.2 _. 5.9 16.3 + 0.5 b 13.8 +_2.5 b 2.1 -t-0.9 0.0 +_5.6 15.2 + 1.3 b 12.2 +_2.2b 16.9 _+2.1 0.0 +_6.5 1.3 + 0.6 0.0 _ 4.0

R2 R2 C2 R2 R2 C2 R2 R2 C2

11.6 _+3.48 20.0+_ 5.1 5.6+_ 1.7 14.0 + 7.2 14.1 +_7.2 2.4 +_0.5 b 9.1 +_1.28 15.0 _ 2.8 b 5.8 +__0 . 4 8

Chromosome 4

bp/cer2 bp/cer2 bp/ga5 bp/fwa bp/fd bp/ap2 bp/ap2 cer2/ga5 cer2/ga5 cer2/fwa cer2/fd cer2/fd cer2/ap2 cer2/ap2 ga5/fwa ga5/fd ga5/ap2 ga5/ap2 fwa/ap2 fwa/ap2 fd/ap2 fd/ap2

Chromosome 2

fve/cp2 fve/cp2 fve/hy3 fve/py fve/as fve/as fve/cer8 hy3/py cp2/er cp2/re cp2/as cp2/cer8 cp2/fpa er/re er/as er/cer8 er/fpa re~as re/cer8 re/fpa as/cer8 as/fpa cer8/fpa

Method"

Chromosome 5

fy/msl fy/ttg chp7/co chp7/msl chp7/ttg lu/co co/msl co/ttg msl/ttg

" C2, R2, results of F 2 generations in coupling (C) or repulsion (R) phase b Results pooled with data published by Koornneef et al. (1983)

Recombinants from crosses between late flowering mutants In s e g r e g a t i n g p o p u l a t i o n s d e r i v e d f r o m crosses b e t w e e n m u t a n t s at d i f f e r e n t loci, the r e c o m b i n a n t d o u b l e m u t a n t s m a y o r m a y n o t s h o w t r a n s g r e s s i o n p a s t the l a t e r p a r e n t . I f so, the genes c o n c e r n e d act o n i n d e p e n d e n t p a t h w a y s l e a d i n g to p r o d u c t s affecting f l o w e r i n g time. I f not, t h e y act o n the s a m e p a t h w a y o r o n t w o p a t h w a y s c o n v e r g i n g into one. I n the case o f leakiness o f t w o m u t a n t s o f d i f f e r e n t genes in o n e p a t h w a y , t r a n s g r e s s i o n c a n also be expected. Such i n f o r m a t i o n is p e r t i n e n t to the c o n s t r u c t i o n o f c a u s a l m o d e l s . E x a m p l e s o f the s e g r e g a t i o n for F T in two F2 p o p u l a tions, o n e s h o w i n g t r a n s g r e s s i o n a n d the o t h e r n o t , are

s h o w n in Fig. 3. T h e c o n c l u s i o n s f r o m these digenic segr e g a t i o n a n a l y s e s are p r e s e n t e d in Table 4 a n d c a n be s u m m a r i z e d as t b l l o w s : 1. M u t a n t s at the loci co, gi a n d f t r e p r e s e n t m u t a t i o n s in o n e p a t h w a y . N o t a single p l a n t l a t e r t h a n the p a r e n t s c o u l d be o b s e r v e d a m o n g 350 F2 p l a n t s i n v o l v i n g different alleles o f co a n d gi. 2. fca a n d f y m u t a n t s affect genes in a s e c o n d p a t h w a y . 3. T h e loci fve, fwa a n d p r o b a b l y fe r e p r e s e n t a t h i r d p a t h w a y . It s h o u l d b e n o t e d t h a t crosses w i t h J h a a n d fpa h a v e n o t y e t b e e n a n a l y s e d in this c o n t e x t a n d t h a t crosses w i t h f d p r o v i d e d insufficient evidenc. A m o n g the seven crosses giving t r a n s g r e s s a n t F3 lines, we o b s e r v e d 43 t r a n s g r e s s a n t s in a t o t a l o f 892 F3 lines ( e x p e c t e d 56; 2 2 = 3 . 1 ; 0 . 1 0 > P > 0 . 0 5 ) .

62

2 o 3 6

3 ~ve hy3 cp2

14

fho

22

go4

23

PY

gi

32 54

Fe GS

33

17

dis2

60

chl

68 71

le cer5

5

hy2 9

bp

er 29

48

cer8

57

fpo

gll

fy

6

chp7 lu

9 12 27

gi

co

msl ttg

fwa

fe fve

47 48

cer2 gG5

51

fwa

61

fd Gp2

63 76

fca

Table 4. The presence (+) or absence (-) of lransgression in segregating generations derived from crosses between mutants at eight loci

0

21

39

45

4

tt5

fca

fy

co

gi

fi

fwa

- (F2) nt

- (F3)

+ (F3) nt + (F2)

+ (F3) nt + (F3)

nt + (F2) nt

- (F2) - (F2)

+(F2) (1)

+(F2) + (F3)

+(F3) + (F2)

nt nt

fe

fve

fca

nt 1

+ (F2) + (F3) [ nt

+ (F3) / - (F2)

nt, not tested F2, based on F2 plants F3, based on F3 lines (1), no homozygote recombinants due to close linkage In crosses amongfca, co and gi only flowering time (FT) was scored; in the other crosses both FT and leaf number (LN) were scored

ff

91 103

Gpl

125

fe

Fig. 2. The linkage map ofArabidopsis thaliana, indicating the location of the 11 flowering time (FT) loci and the markers used in linkage analysis with these loci. The map positions are based on all classical linkage data available for Arabidopsis (see Koornneef 1990)

20 .m 1:3.

:o-3 xgi-3

IO .13

E

0 20

fcaxg[-3 13_

"5 1 0 - -

I1} .13

E Z

0 18

30

42 54 66 Days to flowering

78

Fig. 3. Frequency distributions for FT of the Fz generations derived from crosses between co-3 and gi-3, showing no transgression and betweenfca and gi-3, which does show transgression. The horizontal lines represent the ranges of the parents and wild type. Each F2 is one replicate (size 50) out of seven and four, respectively

The phenotype of late flowering mutants and the effect of short days and vernalization The correlation between FT and LN has been noted before by many authors and can be explained by assuming a uniform rate of leaf initiation for all genotypes which differ only in the time at which the vegetative

meristem is changed into a reproductive meristem. However, the correlation coefficient between the two parameters is not the same for all genotypes. The data obtained for 2 wild types and 32 different mutants grown simultaneously under identical greenhouse conditions show (Table 5) that most mutants at thefca andfve loci make more leaves in the same period than, for example, mutants at the co and gi loci. This high LN relative to FT reappeared in the late segregants in the F2s derived from the crosses with wild type (data not shown). Another factor, which affects this correlation, is shown by two identical co-1 alleles, where in one case this mutant was combined with the chp7/chp7 genotype. The chlorophyll mutation clearly delayed FT, probably by reducing growth, but did not affect LN. The co mutant (co-l) in the Col background has, compared with the Ler co mutants, a higher LN at the same FT, which can also be observed for the Col wild type itself (Table 5). Apparently, the genotypic background can affect the FT/LN ratio as was previously shown by Hussein (1968). Independently isolated mutant alleles at the same locus were found to differ quantitatively in several cases (co, gi, fca, fve). Radiation-induced mutants, however, which may be more likely candidates for null mutants than EMS-induced alleles, did not, in general, represent the latest alleles. For other loci (e.g. ft, fha, fpa, fwa) the different alleles were rather similar in FT and LN characteristics. Arabidopsis is known to respond to LD and vernalization with a reduction of its FT and LN (Napp-Zinn 1969, 1985). Representative mutants of 11 loci and the wild type Ler, were grown both in LD and SD and a vernalization treatment (seed/seedling treatment for 22 days at 3° C) was applied to half of the plants. For each of the four treatments, FT was plotted against LN. Figure 4 gives the graph for treatment SD, V- (short day, no vernalization). With the exceptions to be mentioned, the points are well fitted by linear regression, which shows again the high correlation between FT and LN both in the rosette and on the main stem. In all four treatmentsfca andfve showed a conspicuous excess

63 Table 5. Leaf number (LN) and flowering

Genotype

time for two wild types and a number of independently isolated mutants at 11 different loci grown under identical long-day greenhouse conditions

Mutagen"

Wildtype-Ler Wildtype-Colb ~-i EMS ~a-1 EMS ~a-2 EMS ~e-1 EMS ~e-2 EMS ~a-2 N ~a-3 X ~a-4 X ~a-5 EMS ~a-1 EMS ~a-6 EMS ~-1 EMS fi-i EMS fi-2 EMS fi-3 EMS fd-1 EMS fwa-2 N fwa-1 EMS ~a-2 EMS ~a-3 X ~a-1 EMS co-4 EMS co-5 EMS co-6 EMS co-7 EMS co-2 EMS co-3 EMS co-1b X co-1/e~7b X gi-4 EMS g#5 X gi-3 EMS g#6 EMS

Rosette LN 6.8±0.1 12.9±0.2 12.1±0.2 14.6±0.4 17.8±0.6 18.5±0.5 33.0±0.3 14.7±0.2 14.9±0.5 18.4±0.6 23.2±0.4 28.2±0.5 31.3±0.8 15.6±0.3 14.8±0.3 15.3±0.3 16.6±0.4 10.8±0.1 13.3±0.3 14.4±0.6 9.7±0.2 10.4±0.2 11.9±0.2 12.6±0.2 18.7±0.4 18.6±0.3 19,8±0.2 19.9±0.3 21.6±0.3 21.7±0.3 22.1±0.5 17.4±0.3 20.2±0.4 23.1±0.3 23.4±0.4

Cauline LN 3.1±0.1 3.5±0.1 4.1±0.1 3.7±0.2 4.4±0.3 5.5±0.3 10.5±0.3 4.7±0.1 4.9±0.2 5.0±0.2 6.2±0.2 8.2±0.2 8.7±0.3 5.4±0.l 5.9±0.2 7.5±0.2 7.6±0.3 4.4±0.1 7.0±0.2 7.0±0.3 4.6±0.1 4.0±0.1 3.3±0.1 4.6±0.1 7.5±0.3 9.5±0.3 11.4±0.3 8.8±0.2 9.6±0.2 8.4±0.2 9.7±0.3 6.2±0.3 8.3±0.3 10.8±0.2 10.9±0.3

Days to flowering 27.9±0.3 30.6±0.3 36.3±0.2 37.7±0.4 40.5±0.7 39.4±0.5 53.3±0.5 34.9±0.4 38.0±0.6 39.5±0.7 45.5±0.4 45.0±0.7 46.3±0.4 38.3±0.7 38.7±0.3 41.0±0.6 43.4±0.4 34.2±0.2 37.9±0.3 39.2±0.5 33.8±0.3 34.4±0.5 35.7±0.3 37.5±0.3 45.8±0.4 48.2±0.3 46.6±0.2 44.0±0.2 47.0±0.3 41.8±0.5 55.2±0.4 41.7±0.6 43.8±0.3 47.9±0.3 49.3±0.4

a EMS, ethyl methanesulphonate; N, fast neutrons; X, X-rays b In Columbia genetic background. All other genotypes are Landsberg erecta The averages and standard errors are based on four replicates of five plants each

70

Q

k--

fca

o~

/

(1) 6O

t

/,m / t

O q___ /

50

/

6°

e,,'"

/

O +-,

~j

/

,,'£2

II..0

Oo

/O

;>~ 4O ~3

t~¥

,, m

b

Q 30

0

fpo

z

//

9i_3 co-3

fd

fve 0~

fy ft fha

WT

i

i

i

10

20

30

Number

of

40

leaves

Fig. 4. The correlation between FT and L N for representative mutant alleles of the different FT loci and wild type grown in a climate

room at short day length (SD). Open symbols represent the number of cauline leaves on the main stem; closed symbols the number of rosette leaves (without the cotyledons)

of rosette leaves. For cauline leaves an excess is shown byft, fwa a n d f e (and p e r h a p s f d ) for the two SD treatments. For the two L D treatments, onlyfwa is in excess. This is also observed in greenhouse experiments with b o t h m u t a n t alleles at this locus. It is interesting that the group of four mutants mentioned above stands apart in combining a good response to SD with little or no response to a vernalization treatment (see later). Regression analysis of F T on L N for the four treatments indicated that the delay in F T due to treatment stress is more p r o n o u n c e d with SD than with V ÷ (data not shown). In view of these non-genetic effects on F T it is obvious that for the present purpose L N is a better p a r a m e t e r for the phase change (lateness) than F T itself. Further, we prefer to use total L N instead of rosette L N as the latter does not account for the differential effect of mutants on stem LN. The effects of the four treatments on L N of the 12 individual genotypes are shown in Fig. 5. For ease of presentation we t o o k the environment o f m u t a n t selection, viz. LD, V - , as a point of reference. F r o m this starting point we speak of the SD response (left vertical arrow) and V ÷ response (middle arrow). Since initial

64 50-

z

~E

30-

+ tao

t iI

20-4

... F

l

2: t'Y

Z

a_ 10 ,