Table 1 Comparison of mean 9 SD values of FC, FA0 and FA
populations
a
Population FA0
FC FA1
6.37 a Leaf DW g LDW
7.31 b 7.19b
9 2.49 9 2.02
9 2.41 32.8 a
Stem DW g SDW 42.5 b
39.9 b 9 10.9
9 15.6 9 13.9
97.1 a 88.2 a
96.5 b Stem height cm SH
9 5.4 9 11.0
9 12.2 1.54 a
1.00 a 1.48 a
Visible stem cm 9 4.71
9 6.73 VissS
9 6.37 43.4 a
49.8 b Tot seed weight g
48.9 b TSW
9 19.3 9 24.1
9 26.6 Number of seeds NS
2212 a 2216 a
2218 a 9 879
9 812 9 1068
19.25 a 22.05 b
Avg seed weight mg 21.73 b
9 3.42 9 4.72
9 5.54 ASW
SH:SDW ratio 2.99 a
2.71 a 2.75 a
9 1.52 9 1.03
9 1.16 55.4 b
NS:SDW ratio 56.2 b
68.7 a 9 25.3
9 19.3 9 26.4
6.86 a TSW:LDW ratio
6.78 a 6.74 a
9 2.42 9 2.87
9 3.24 112
Population size 123
296
a
A same letter following two values in the same row indicates that they are not distinguishable by a t-test at
PB0.05.
SV
L
C = 2[CL-C
FC
] CL + C
FC
where CL is the mean value of the character C for the line L, and C
FC
is the corresponding value calculated for the whole population of control
plants see Table 1. The absolute value of SV
L
C [abbreviated as ASV
L
C] quantifies the deviation of the line L from the FC population for expres-
sion of the character C, and the sign negative or positive indicates direction of this discrepancy.
For each character C, the difference in ASV between FA1 and their corresponding FA0 line
abbreviated as DASV
A1,A0
C was calculated as follows:
DASV
A1,A0
C =
1mn
n i = 1
m k = 1
ASVFA1
k i
C − ASVFA0
i
C
n
where n is the number of lineages considered, and m is the number of FA1 lines related to each FA0
line n = 3 and m = 2 for classes I and II in the current experiment.
3. Results
3
.
1
. The FC, FA and FA
1
populations Field-grown progeny of control and adapted
plants FC and FA0 plants respectively, see Fig. 1 were analyzed at the end of their life cycle.
Stem height, weight and total seed weight signifi- cantly increased in FA0 plants, but the number of
seeds and the length of the visible stem were not distinguishable from that of the FC plants P B
0.05 by a t-test Table 1. Relative to vegetative development, stem height and total seed weight
respectively SH:SDW and TSW:LDW ratios of the FC and FA0 plants were similar, while a
decrease in relative fertility NS:SDW ratio was observed in FA0 plants Table 1. Progeny of two
consecutive generations of salt-adapted plants FA1 plants, see Fig. 1 was very similar P B
0.05, t-tests to the FA0 population for most of the characters analyzed Table 1.
defined as the leaf DW LDW. Stem height SH was defined as the length between the first adven-
titious root and the top of the spike. The length of visible stem VisS was defined as the distance
between the ligule of the last leaf and the first spikelet branch. This distance may be negative, as
when the first spikelets developed below the ligule of the last leaf. The seeds were weighed and
defined as the total seed weight TSW. The aver- age seed weight ASW was determined after
weighing three groups of ten seeds. The number of seeds was calculated as the ratio between total
seed weight TSW and average seed weight ASW.
A standardized value SV was calculated for each character C and for each line L sepa-
rately, as follows:
Correlations between characters were analyzed within each population. A highly significant rela-
tionship was constantly observed between shoot DW and seed number Table 2. Nevertheless, the
slope of the linear regression differed for FA0 as compared with FC population, but also for FA1
as compared with FA0 Table 2. This suggests that developmental regulations differed in the
FA0 and FA1 populations. The FA1 plants issued from two successive treatments for salt-adaptation
Fig. 1. Enhancement of modifications was ex- pected for these plants, but this is not the case.
The slope calculated for the FA1 population is generally intermediate between that of the FC and
FA0 populations Table 2, suggesting that the development reverted towards the ‘standard val-
ues’ in the FA1 population.
3
.
2
. E6idence for two classes of lineages The FA0 and FA1 populations issued from the
pooling of six FA0 and twelve FA1 lines respec- tively. All these 18 lines issued from six A0 ances-
tors defining six lineages see Section 2. The mean value of each character was calculated sepa-
rately for each line Table 3, and was standard- ized SV values as described in Section 2. The
length of visible stem was omitted from the results because of its considerable range of variations
Table 1. For each line, the SV indices calculated for the nine characters studied were pooled be-
cause of the similar ranges of variation. Plotting the corresponding SV from two lines provides an
estimate of their level of similarity. No significant correlation has been observed between the FA0
line and any one of the FA1 lines of the lineage 701 Fig. 2A. In contrast, a significant positive
correlation P B 0.05 was observed between FA0 and FA1 lines of the lineage 715 Fig. 2B. A
similar analysis was performed for the four other lineages. A significant correlation P B 0.05 may
be seen between FA0 and FA1 lines in lineages: 530, 715 and 716, and the standardized values of
the two corresponding FA1 lines also showed significant correlations Table 4. The FA1 lines
do not resemble the corresponding FA0 line in the lineages 611, 701 and 718, although the two corre-
sponding FA1 lines may show a significant corre- lation lineage 611, Table 4. This analysis reveals
that, in spite of initial homogeneity, all the lin- eages are not identical. The high ‘transmissibility’
from A0 ancestor to FA1 lines lineages 530, 715
Table 2 Comparison of linear regressions for FC, FA0 and FA1 populations
a
x axis y axis
Population FC
FA0 FA1
LDW 0.76
TSW r
0.76 0.65
s 7.28
7.41 7.28
− 3.04
c −
3.12 −
4.42 SDW
0.63 NS
0.60 r
0.77 35.52
45.96 s
52.82 386.1
706.0 478.5
c r
− 0.180 ns
0.043 ns −
0.055 ns SH
NS VissS
− 0.46
− 0.31
− 0.62
r NS
− 40.3
− 107.0
− 76.9
s 2257
2333 c
2377 SH
r VissS
0.58 0.59
0.46 s
0.67 0.97
0.88 c
95.2 96.1
87.2 SDW
r SH
− 0.003 ns
0.006 ns −
0.031 ns d.f.
110 121
294
a
Abbreviations. r, regression coefficient; s, slope value; c, constant values of the linear regression. Characters abbreviated a in Table 1. Ns, non significant correlation P\0.05, PB0.05; PB0.005; PB0.001.
Table 3 Mean values of the FA0 and FA1 lines
a
LDW SH
SDW TSW
NS ASW
SH:SDW NS:SDW
TSW:LDW mg
g ratio
g ratio
cm ratio
g FA
lines 45.0
530 106.7
8.26 65.9
2650 24.2 2.56
59.3 7.94
9.72 57.4
106.5 62.8
2779 611
22.6 1.90
49.0 6.45
701 10.58
65.2 91.8
85.8 2736 31.6
1.43 42.3
8.09 6.85
42.4 95.3
715 52.3
2323 22.3 2.39
57.2 7.54
8.42 43.6
92.4 49.6
2364 716
20.9 2.34
58.3 5.88
5.45 24.6
89.8 23.2
1379 718
16.7 4.77
73.8 4.32
FA
1
lines 45.8
104.6 65.9
530
1
2693 7.70
24.2 2.57
58.8 8.56
530
2
7.37 38.9
100.7 53.1
2307 22.7 2.93
57.8 6.78
7.28 40.4
114.1 41.2
2088 611
1
19.5 3.04
52.9 5.68
6.76 34.7
107.1 32.0
1616 611
2
19.3 3.27
45.1 4.35
8.43 41.0
103.9 55.2
2809 701
1
18.8 3.11
72.9 6.78
701
2
6.94 33.7
95.8 32.5
1899 17.0 3.10
52.8 4.24
715
1
7.11 44.2
89.3 62.5
2753 22.3 2.13
62.0 8.70
715
2
8.21 48.2
82.5 63.2
2728 22.9 1.81
58.4 7.74
6.91 50.3
99.6 59.0
2108 716
1
28.4 2.36
41.0 8.38
9.77 61.3
94.0 75.2
2584 716
2
29.6 1.64
45.0 8.05
7.62 54.2
98.1 71.8
2642 718
1
28.1 2.15
48.2 9.19
718
2
5.75 35.9
98.5 1.3
2015 20.1 3.32
57.1 7.23
a
Same abbreviations as in Table 1.
cant correlation has been observed between mean ASV values of FA0 and FA1 populations belong-
ing to Class I but not to Class II Fig. 3A. This result confirms that the deviation from standard-
ized value in development is similar for FA0 and FA1 populations in Class I. In Class II, however,
the ASV were considerably reduced in FA1 as compared to FA0 populations Fig. 3A.
The difference between ASV of the FA0 and the corresponding FA1 lines was estimated sepa-
rately for Class I and Class II lineages. For each character C, the difference in ASV between FA1
and their corresponding FA0 line abbreviated as DASV
A1,A0
C, see Section 2 was calculated and plotted as a function of the mean ASV calculated
for the corresponding FA0 lines Table 5. No relationship was found in Class I lineages, but
most of the DASV indexes were positive. This indicates that development was even more per-
turbed in FA1 than in FA0 plants in Class I Fig. 3B. A significant negative correlation P B 0.05
is observed in Class II lineages Fig. 3B. This and 716 defined a Class I of lineages, while the
low ‘transmissibility’ lineages 611, 701 and 718 defined a Class II of lineages.
3
.
3
. Re6ersion towards normal de6elopment in FA
1
lines Deviation from the standardized values was
analyzed separately for the two classes of lineages. For each character, the absolute value of SV
termed ASV, see Section 2 was averaged for the three FA0 lines belonging the same class. Except
for stem height, the ASV is always higher for FA0 from Class II than those from Class I Table 5.
This indicates that the FA0 plants belonging to Class
II were
the most
modified in
their development.
The mean ASV indexes were calculated sepa- rately for FA1 lines belonging the two classes as
described in Table 5 for FA0 lines. For each character, the mean ASV calculated for FA0 and
FA1 lines were plotted together. A positive signifi-
suggests that the tendency to recover the standard value in FA1 lines negative DASV indexes was
proportional to the deviation observed in the FA0 line belonging to Class II.
3
.
4
. Influence on 6ariability The phenotypic variability within each FA1 line
Table 5 Discrepancy from standard values in FA0 lines from Clases I
and II. The absolute standard value ASV calculated for the three FA0 lines in the same class were averaged for each
character The ASV calculated for each class are averaged as a mean ASV value
a
Character Absolute standard value ASV
Class I Class II
0.203 0.355
Leaf DW Stem DW
0.284 0.497
Stem height 0.104
0.080 0.244
0.542 Tot seed weight
0.262 0.152
Avg seed weight SH:SDW ratio
0.206 0.537
0.294 NS:SDW ratio
0.164 0.226
TSW:LDW ratio 0.131
0.176 Mean ASV value
0.343
a
Abbreviations as in Table 1.
Fig. 2. Relationship between FA0 and FA1 lines from the same lineage. A Lineage 701; B lineage 715. The standard
values SV indexes were calculated from data of Tables 1 and 3 as described in Section 2. Solid circles: FA1
1
line; opened circles: FA1
2
line.
Fig. 3. Relationship between perturbation in FA0 and FA1 plants. A Comparison of standardized values ASV. B
Influence of the perturbation in FA0 on the corrective effect DASV in FA1. Absolute standardized values ASV and
their difference between FA0 and FA1 lines DASV are calculated from data in Table 5 as detailed in Section 2. Open
circles: class I lineages; solid circles: class II lineages. Table 4
Comparison of two lines from the same lineage
a
r coefficient between Lineage
FA0 and FA0 and
FA1
1
and FA1
2
FA1
1
FA1
2
0.981 0.883
0.887 530
0.583 611
0.246 0.902
0.496 701
− 0.014
0.408 0.959
0.901 0.902
715 716
0.655 0.868
0.884 −
0.801 0.233
0.232 718
a
The r coefficient is calculated within the same lineage for standardized values SV calculated from the nine characters
see Table 3. Significant correlation at PB0.05.
was analyzed separately for the two classes. For each one of the FA1 line, the coefficient of variation
abbreviated as CV was calculated for each one of the nine characters. Variability within FA1 lines is
generally greater that variability of the FC popula- tion Table 6. Moreover, significant differences
may be observed in the level of variability between the two classes. Variability is considerably in-
creased in FA1 lines of Class II for the number of seeds or the relative stem height, but the situation
is reversed for average seed weight Table 6. Accordingly, lineages belonging to Classes I and II
cannot be simply distinguished by this analysis. The CV values of the two FA1 lines from the same
lineage were plotted together. No relationship is observed in Class I P \ 0.05, Fig. 4A, while a
significant positive correlation r = 0.50, 25 d.f., P B 0.01 appears in Class II Fig. 4B.
A capacity for developmental reversion was especially observed in Class II lineages Fig. 3,
which also showed a ‘control’ of variability Fig. 4B. The link between these two types of events was
tested. For each FA1 line and for each character, the coefficient of variation was standardized as a
function of the average CV value of the character calculated for the FC population see Table 5.
These standardized values abbreviated as SCV were calculated as described above for mean values
SV, see Section 2. Then, for each character and for each FA1 line, the standardized coefficient of
variation SCV was plotted as a function of the standardized mean value SV. No significant rela-
tionship P \ 0.05 appears for FA1 lines of Class I Fig. 5A, while a negative significant correlation
P B 0.01 is observed for FA1 lines of Class II Fig. 5B.
4. Discussion