...

Evaluation of plant and fruit traits in recombinant inbred lines of

by user

on
Category: Documents
1

views

Report

Comments

Transcript

Evaluation of plant and fruit traits in recombinant inbred lines of
Cien. Inv. Agr. 33(2): 111-118. 2006
www.rcia.puc.cl
research paper
Evaluation of plant and fruit traits in recombinant inbred lines of
tomato obtained from a cross between Lycopersicon esculentum
and L. pimpinellifolium
Gustavo R. Rodríguez1,3, Guillermo R. Pratta1, Roxana Zorzoli2
and Liliana A. Picardi2
1
Consejo Nacional de Investigaciones Científicas y Técnicas
Consejo de Investigaciones de la Universidad Nacional de Rosario.
3
Facultad de Ciencias Agrarias, Universidad Nacional de Rosario. Campo Experimental J.F. Villarino.
CC Nº 14 (S2125ZAA) Zavalla, Argentina.
2
Abstract
G.R. Rodríguez, G.R. Pratta, R. Zorzoli and L.A. Picardi. 2006. Evaluation of plant
and fruit traits in recombinant inbred lines of tomato obtained from a cross between
Lycopersicon esculentum and L. pimpinellifolium. Cien. Inv. Agr. (In English) 33(2):111118. Seventeen recombinant inbred lines of tomato (Lycopersicon esculentum) were obtained
from an interspecific cross between tomato cv. ‘Caimanta’ and the accession LA722 of L.
pimpinellifolium. Plant and fruit traits were evaluated in these lines as well as in the parental
genotypes that served as controls. Significant differences were found among parental genotypes
and recombinant inbred lines for plant traits (internodes length between third and fourth node,
number of flowers per inflorescence, stem perimeter at the basal, middle, and apical part) and
fruit traits (soluble solid content, pH, acidity, diameter, height, shape, weight, and shelf life).
Significant differences were also found among recombinant inbred lines, in spite of the fact that
many of them were similar to the wild parent. Fifteen lines had longer shelf life than LA722 that
was the parent with the longest shelf life. A multivariate analysis of canonical correlation with
plant and fruit traits and a cluster analysis were performed to classify the lines and the parental
genotypes according to their performance for these traits. The first canonical correlation
explained 73% of the total variance (p<0.01). According to the canonical coefficients those
plants with the biggest stem perimeter and the lowest number of flowers per cluster produced
the heaviest and the biggest fruits but they had low acidity. In the cluster analysis, shelf life was
an important discriminatory trait for these lines and the parental genotypes. ‘Caimanta’ was
the only genotype located in one group. The other clusters had genotypes with similar values
for stem perimeter, fruit size, fruit weight, acidity and fruit shelf life. It was possible to obtain
new genotypes that have recombinant traits from the parental genotypes and some of the new
genotypes have an even longer shelf life than the wild parent. In conclusion, some of these new
genotypes would be a new source of variability useful in tomato breeding programs.
Key words: Fruit quality, fruit shelf life, Lycopersicon esculentum, L. pimpinellifolium.
Introduction
The development of new tomato cultivars
(Lycopersicon esculentum Mill.) have intended to
Received 08 Feb 2005; Accepted 09 March 2006.
1
Corresponding author: [email protected]
improve productivity, quality and adaptation to
different production conditions. Sometimes, this
is difficult to achieve due to reduce availability
of genetic resources (Warnock, 1991). Close
related wild species of Lycopersicon, with
fertile crossings with cultivated tomato species,
have been very valuable genetic sources for
the development of new cultivars (Hermsen,
1984).
112
CIENCIA E INVESTIGACION AGRARIA
Fruit quality is one the most important traits in
a breeding program. Quality involves several
traits such as size, weight, shape, soluble solid
content, acidity, texture, and shelf life (SL).
Other important traits, indirectly affecting fruit
quality, are flowers per inflorescence, internode
length, and stem perimeters at the apical,
medium and basal portion of the stems. The
stem perimeters affect the support of branches,
racemes, and the anchorage capacity of the
plant.
Within the genetic background of cultivated
tomato species, some spontaneous mutants
altering the maturation process have been
identified (Tigchelaar, 1986). At homozygote
state, fruits of these mutants present an
extended shelf life, but they negatively affect
other commercially valuable traits (Nguyen et
al., 1991). They produce undesirable pleiotropic
effects on color, pH, flavor, and aroma, even
at heterozygote state (Buescher et al., 1976).
Thus, their use as parental has been limited in
the search of plant material producing fruits
with long shelf life.
Zorzoli et al. (1998) demonstrated that fruits of
wild tomato (L. pimpinellifolium) have a longer
shelf life than fruits of commercial tomato
cultivars. Nevertheless, this was lower than
shelf life of homozygote genotypes for nor (non
ripening) mutant and rin (ripening inhibitor)
mutant of L. esculentum. Furthermore, it was
found that the unfavorable pleiotropic effects of
nor and rin mutants, associated to fruit quality
traits, decreased because of the contribution of
the wild genotypes (Pratta et al., 2000).
The objective of this work was to characterize
recombinant inbred lines of tomato obtained by
selection from an interspecific crossing between
L. esculentum and L. pimpinellifolium, after
six generations of self-fertilization, in order to
verify the magnitude of the genetic variability
obtained for plant and fruit traits.
Materials and methods
Recombinants
In this experiment, seventeen recombinant
inbred lines of tomato (F2:6), obtained through
an agonistic-divergent selection of shelf life
and fruit weight (Zorzoli et al., 2001) were
studied. The F2 of an interspecific crossing
between L. esculentum ‘Caimanta’ (Estación
Experimental Agropecuaria de INTA Cerrillos,
Salta, Argentina) and the LA722 line of L.
pimpinellifolium, (Tomato Genetic Resources
Center, Department of Vegetable Crops,
University of California, Davis, USA) was
used as base generation. Parentals (16 plants)
and these seventeen recombinant lines (136
plants) were transplanted to field conditions
(Estación Experimental, Facultad de Ciencias
Agrarias, UNR) (33º 02’ lat. South and 60º 53’
lat. West) in a complete randomized design,
including eight plants per each genotype as
average. Recombinant lines, obtained from
the F2 generation, were self-fertilized to the F6
generation. Selection process was done based
on the genealogical method (Hallauer, 1981).
Agronomic practices recommended for tomato
in the area were performed (Ferratto et al., 1997).
Evaluation
Length of internodes (LI) between the third
and fourth node; stem perimeter, determined
at the base (BP), middle (MP), and apical
(AP) portion of the stems, and flowers per
inflorescence, considering the mean of the first
three inflorescences, were evaluated 60 days
after transplant.
Fruits were characterized for height (Hf),
diameter (Df), shape (Sf, Hf/Df), weight (Wf)
and SL. An 18 fruit sample per plant (2,778
fruits), harvested 45 days post-anthesis, was
evaluated. Shelf life was visually determined as
days between harvest and the beginning of the
fruit softening. With this purpose, fruits were
randomly arranged in a shelf at 25 ± 3°C after
harvest (Schuelter et al., 2002).
The homogenized juice, obtained from
pericarps of six to ten randomly taken fruits per
each genotype, was characterized for soluble
solids (SS, using a manual refractometer type
EMA), pH and titratable acidity (TA). Titratable
acidity (grams of citric acid per 100 g of juice)
was determined by the volume of 0.1 N NaOH
needed to neutralize to pH 8.1, 10 g of juice in
100 ml of distilled water (Nguyen et al., 1991).
Four replications per trait and genotype were
used.
VOL 33 N˚2 MAY - AUGUST 2006
Statistical Analysis
Data was subjected to normal distribution
tests (Shapiro and Wilk, 1965) and analysis
of variance (ANOVA). Means were compared
according to Fisher least significant difference
test (Snedecor and Cochran, 1980). Phenotypic
correlations (rF) among variables were used to
estimate degree of association between traits
(Sokal and Rohlf, 1969). Multivariate analysis
of canonic correlations was made to establish
phenotypic associations among plant and fruit
traits. A cluster multivariate analysis was also
performed to classify recombinant lines and
parental genotypes based on 13 traits evaluated.
Cluster analysis was performed according
to Ward (1963), using Euclidean distance to
establish genotypes proximity.
Results
All traits were normally distributed, including
TA that was transformed as logarithmic function.
Mean values for plant and fruit traits, obtained
for each genotype, were summarized in Tables 1
and 2, respectively. Highly significant (p < 0. 01)
differences between parental were obtained
for all traits. Among recombinant lines, highly
113
significant (p < 0.01) differences were obtained
for FL. Recombinant lines were also significant
different in fruit quality traits.
The results on phenotypic correlations among
traits are presented in Table 3. Based on canonic
correlations, 24.1 was the first eigenvalue
between plant and fruit traits. The first canonic
correlation was highly significant (p=0.0028),
and it explained 73% of the total variation.
According to canonic coefficients, plants with
higher stem perimeter (BP: 0.94; MP: 0.87
and AP: 0.76) and low number of flowers
per inflorescence (-0.61), produce fruits with
higher weight (0.95) and higher size (Hf: 0.92
and Df: 0.92) but, low in acidity (-0.61). The
second canonic correlation was not significant
(p=0.09).
According to multivariate analysis, genotypes
were clustered as shown in Figure 1 when
all traits were used, including parental and
seventeen recombinant lines. In the last
level, there was a group formed only by
cultivated species and other group with the
rest of the genotypes. At this level of grouping,
Table 1. Means values and standard error for plant traits obtained for seventeen recombinant inbred lines obtained from an
interespecific cross between Lycopersicon esculentum ‘Caimanta’ and line LA722 of L. pimpinellifolium.
Genotypes
Caimanta
LA722
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
F
LSD (0.05)
IL
FL
3.3±0.1
2.3±0.1
2.9±0.2
3.1±0.2
2.8± 0.1
2.5± 0.1
3.1± 0.2
3.2± 0.1
3.2± 0.2
2.5± 0.1
2.7± 0.2
3.1± 0.2
2.5± 0.1
2.8± 0.2
3.2± 0.2
2.8± 0.2
3.4± 0.1
2.8± 0.2
2.5± 0.1
38.8**2 0.4 5.2± 0.5
12.1± 0.8
6.1± 0.4
7.1± 0.6
6.5± 0.3
6.4± 0.6
7.0± 0.5
10.0± 0.9
7.1± 0.5
9.9± 0.7
11.4± 0.6
9.5±0.5
7.9± 0.2
8.2± 0.7
10.5± 1.2
7.5± 0.6
6.4± 1.0
9.4±0.6
8.5± 0.6
55.4**2
1.8
Traits1
BP
MP
AP
4.6± 0.2
2.6± 0.2
3.3± 0.1
3.0± 0.1
3.0±0.1
3.0± 0.1
3.3± 0.1
3.0± 0.2
2.9± 0.2
2.8± 0.1
3.1±0.1
3.0± 0.1
2.7± 0.1
2.8± 0.1
3.2± 0.1
2.9± 0.1
3.2± 0.1
3.4± 0.1
3.2± 0.1
64.3**2
0.3
4.5± 0.2
2.1± 0.2
3.3± 0.1
3.2± 0.1
3.1± 0.1
3.0± 0.1
3.0± 0.1
2.8± 0.2
2.7± 0.2
2.7± 0.1
3.1± 0.1
3.2± 0.1
2.5± 0.1
2.9± 0.2
3.2± 0.2
3.0± 0.1
3.3± 0.1
3.6± 0.1
3.3± 0.1
179.2**2
0.4
3.5± 0.1
1.6± 0.1
2.8± 0.1
2.8± 0.1
2.6± 0.1
2.5± 0.1
2.8± 0.2
2.2± 0.1
2.1± 0.1
2.3± 0.1
2.3± 0.1
2.6± 0.1
2.2± 0.1
2.7± 0.1
2.5± 0.1
2.6± 0.1
2.9± 0.1
2.8± 0.1
2.7± 0.1
161.5**2
0.3
LE: internodes length between third and fourth node (cm); FL: number of flowers per inflorescence; BP: stem perimeter at the basal part
(cm); MP: stem perimeter at the middle part (cm); AP: stem perimeter at the apical part (cm).
2
**=p ≤ 0.01
1
CIENCIA E INVESTIGACION AGRARIA
114
Table 2. Means Values and standard error for fruit traits obtained for seventeen recombinant inbred lines obtained from an
interespecific cross between Lycopersicon esculentum ‘Caimanta’ and the line LA722 of L. pimpinellifolium.
Genotypes
Caimanta
722
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
F
LSD (0.05)
SS
pH
AT
4.2± 0.1
6.9± 0.0
4.4± 0.2
4.9± 0.1
4.8± 0.2
5.5± 0.3
5.0± 0.0
5.2± 0.4
4.8± 0.0
5.8± 0.2
3.7± 0.8
5.3± 0.2
5.5± 0.2
5.8± 0.1
5.0± 0.5
4.6± 0.1
5.1± 0.4
4.7± 0.1
4.9± 0.2
2187.0** 2
0.8
5.4± 0.1
5.1± 0.0
5.3± 0.0
5.0± 0.1
5.1± 0.1
5.1± 0.1
5.2± 0.0
4.8± 0.1
5.2± 0.0
5.0± 0.0
5.2± 0.1
5.2± 0.1
4.9± 0.1
5.0± 0.1
5.3± 0.0
5.3± 0.1
5.4± 0.1
5.0± 0.1
5.3± 0.2
11.7**
0.2
0.3± 0.0
1.1± 0.1
0.5± 0.0
0.5± 0.0
0.5± 0.0
0.5± 0.1
0.7± 0.1
1.1± 0.2
0.8± 0.1
1.4± 0.1
1.1± 0.4
1.2± 0.1
0.9± 0.2
1.0± 0.2
0.7± 0.0
0.5± 0.0
0.5± 0.1
0.7± 0.1
0.5± 0.1
97.0**
0.4
Traits1 (standard error) 2
D
H
SH
W
SL
7.2± 0.2
1.0± 0.0
3.8± 0.0
3.5± 0.1
3.5± 0.1
3.2± 0.0
2.4± 0.0
1.5± 0.0
1.8± 0.0
1.4± 0.0
1.5± 0.0
1.8± 0.0
1.8± 0.0
1.8± 0.0
1.7± 0.0
3.8± 0.1
3.3± 0.1
3.2± 0.0
3.1± 0.0
4870.7**
0.1
5.4± 0.1
0.9± 0.0
3.2± 0.0
2.8± 0.0
2.8± 0.0
2.9± 0.0
2.1± 0.0
1.3± 0.0
1.6± 0.0
1.3± 0.0
1.3± 0.0
1.6± 0.0
1.5± 0.0
1.6± 0.0
1.6± 0.0
3.0± 0.0
2.7± 0.0
2.7± 0.0
3.3± 0.0
4730.5**
0.1
0.8± 0.0
0.9± 0.0
0.9± 0.0
0.8± 0.0
0.8± 0.0
0.9± 0.0
0.9± 0.0
0.9± 0.0
0.9± 0.0
0.9± 0.0
0.9± 0.0
0.9± 0.0
0.9± 0.0
0.9± 0.0
0.9± 0.0
0.8± 0.0
0.8± 0.0
0.9± 0.0
1.1± 0.0
48.2**
0.0
98.5± 9.9
0.9± 0.0
27.2± 0.7
19.3± 0.7
20.3± 0.7
18.8± 0.5
8.3± 0.4
2.0± 0.1
4.1± 0.2
2.0± 0.1
2.1± 0.1
4.0± 0.1
3.3± 0.1
3.2± 0.1
3.0± 0.1
25.8± 0.8
19.2± 0.8
17.0± 0.6
18.2± 0.5
479.8**
1.9
9.7± 0.9
15.7± 0.4
21.2± 0.6
21.9± 0.8
25.0± 0.9
16.6± 0.5
17.3± 0.7
31.3± 2.0
21.5± 1.0
28.2± 1.7
15.0± 0.5
21.4± 0.8
18.7± 1.2
20.4± 0.9
19.6± 1.0
17.9± 0.5
11.3± 0.7
15.4± 0.5
23.1± 0.7
22.2**
2.8
1
SS: Soluble solid content (°Brix); pH: hydrogen potential; AT: acidity (g of citric acid/100 g of homogenized fruit juice); D: diameter (cm),
H: height (cm); SH: shape (D/H ratio); W: Weight (g) and SL: Shelf life.
2
**=p ≤ 0.01.
discriminating traits were BP (F=59.6, p<0.001),
MP (F=18.1, p=<0.001), AP (F=8.1, p=0.011),
Df (F=23.3, p<0.001), Hf (F=16.4, p<0.001) and
Wf (F=238.2, p=0.00). Clustering genotypes
in five groups demonstrated that SL was the
discriminating trait (Table 4). Two sub-groups,
presented in Figure 1, included genotypes with
long SL, characterized by high stem perimeter,
fruits with high size and weight and low acidity
(Group 1A). Similar genotypes with shorter SL
were clustered in Group 1B. A second grouping
also separated genotypes with long (Group
2C) from short (Group 2D) SL but, these were
genotypes with lower stem perimeters, smaller
fruits, lower weight but with high acidity.
Discussion
The recombinant inbred lines had intermediate
vegetative development relative to their
parental genotypes. It was interesting that when
the number of flowers per inflorescence, an
important component in tomato production, was
analyzed; thirteen out of seventeen genotypes
had mean values higher than seven flowers per
inflorescence. Furthermore, nine and thirteen
recombinant lines were similar to wild type.
These results were coherent with the results
obtained in a crossing between L. esculentum
and L. pimpinellifolium where the dominance
of the wild type prevailed on quantitative traits
(Weller et al., 1988).
Recombinant lines were different from the line
LA722. However, recombinant phenotypes
were closer to wild type for Df, Hf, and Sh. In
relation to Sh, line 17 had the highest value with
pyriform fruits.
Flavor is mainly determined by sugars and
acidity, and it increases as these factors increase
in tomato fruits (Jones and Scott, 1983). The
highest SS contents obtained for wild type
tomato were coherent to other reports (Rick,
1976; Weller et al., 1988). Although this trait was
not evaluated during the selection process, and
none of the recombinant lines overcame LA722
values, the SS contents increased in relation
to the cultivated parental. It was interesting
the increment on SS contents obtained for
recombinants (lines 8 and 12) relative to wild
type parental. The recombination observed in
VOL 33 N˚2 MAY - AUGUST 2006
115
Table 3. Phenotypic correlations (rF) among evaluated traits in recombinant inbred lines obtained from an interespecific
cross between Lycopersicon esculentum ‘Caimanta’ and the line LA722 of L. pimpinellifolium.
IL 1
BP
MP
AP
FL
H
D
SH
0.21
**2
0.6
*
0.19
*
-0.20
**
0.24
ns
0.30
ns
-0.38
ns
0.29 -0.09 -0.45 0.29 -0.24
ns
ns
*
ns
ns
0.76
***
-0.19
**
0.83
***
0.83
***
-0.32
ns
0.82 -0.41 -0.66 0.56 -0.55
***
ns
***
**
*
BP
0.83
***
MP
0.65
***
AP
-0.18
*
-0.30
***
FL
0.80
***
0.83
***
-0.74
***
H
0.82
***
-0.36
ns
0.82
***
-0.37
ns
0.94
***
0.005
ns
-0.75
***
D
W
SL
SS
pH
AT
0.89 -0.47 -0.57 0.53 -0.48
***
*
**
*
*
0.74 -0.34 -0.56 0.52 -0.62
***
ns
**
*
***
0.47
*
-0.62 0.22
***
ns
-0.31
***
0.95 -0.03 -0.52 0.55 -0.78
***
ns
*
**
***
SH
0.36 -0.38 0.75
ns
ns
**
0.92 0.003 -0.52 0.59 -0.81
***
ns
*
**
***
-0.22 0.11
*** ***
W
0.25 -0.031 0.30
ns
ns
ns
-0.005 -0.45 0.54 -0.62
ns
*
*
***
SH
0.22 -0.57 0.42
ns
**
ns
SS
-0.43 0.50
*** ***
pH
-0.42
***
IL: Internodes length between third and fourth node (cm); BP: Stem perimeter at the basal part (cm); MP: stem perimeter at the middle
part (cm); AP: stem perimeter at the apical part (cm); FL: number of flowers per cluster; H: height (cm); D: diameter (cm); SH: shape (D/H
ratio); W: Weight (g); SL: Shelf life; SS: Soluble solid content (°Brix); pH: Hydrogen potential and AT: acidity (g of citric acid/100 g of
homogenized fruit juice).
2
ns= not significant; * p<0.05; ** p<0.01; ***p<0.001.
1
these lines suggested the importance of the
effect of wild type genotype. Previous works
have demonstrated that SS contents can be
improved using germplasm of related species
characterized by fruits with high SS contents.
Among these work, Poysa (1993) has used L.
cheesmanii and L. chmielewskii to increase the
content of dry material of fruit.
None of the recombinant lines overcame pH
obtained for tomato cv. ‘Caimanta’, used as
control line in this study. Recombinant line
6 had the lowest pH, significantly different
from their parental lines. This suggested that
recombination for this trait occurred during the
selection process. Mean values were the sum
effects of genes present in each progenitor that
negatively affected this trait. In general, acidity
of some recombinant lines was equal to one of
their parental genotypes. The recombinant line 8
was the genotype with highest value for acidity.
Fruit weight, Df and Hf of any recombinant
line was as parental genotypes. Lines 1 and 14
had the highest values and lines 6, 8, and 9 had
the lowest fruit weight. However, recombinant
lines were closer to LA722 used as control.
These results agreed with those of Grandillo
et al. (1999), who postulated that quantitative
inheritance alleles of small fruits would be
semidominant over alleles involved in the
expression of bigger fruits. On the other side,
polygene with dominant effects in wild species
could cause fruit weight reduction (Weller et al.,
1988). From the productive point of view, this
reduction in fruit size could be compensated
with the presence of a higher number of flowers
per inflorescence.
Long SL is a highly appreciated trait for fresh
market tomatoes. The SL values in parental
genotypes were similar to previous reports,
demonstrating that L. piminellifolium and L.
CIENCIA E INVESTIGACION AGRARIA
116
Cai
LA722
9
5
7
11
12
10
13
6
8
1
14
2
18
3
4
16
15
0
Cluster 3
2D
Cluster 2
2C
1B
1A
Cluster 1
50
100
150
200
250
300
Linkage Euclidean Distance
Figure 1. Diagram of clustering for thirteen evaluated traits in the both parent:
Lycopersicon esculentum cv. Caimanta and the accession LA722 de L. pimpinellifolium,
and the 17 RILs. Groups 1, 2 and 3 clustered by weight and Groups 1A, 1B, 2C and 2D
by weight and shelf life.
esculentum var. cerasiforme bear genes that
prolong tomato SL (Zorzoli et al., 1998; Pratta
et al., 2000). Among recombinant lines, it
was noticed that line 6 overcame significantly
LA722. Fourteen lines had a higher SL than
their wild parents. The lowest SL value was
obtained with line 15, and it was similar to
tomato cv. ‘Caimanta’. Some recombinants lines
had higher SL values than LA722, providing
additional evidences in favor of recombination
through this interspecific cross.
The analysis of canonic correlations showed the
importance of the relation between vegetative
growth traits and those determining fruit
production and quality. In this study, plant size
was associated with weight, size and acidity
of the fruits. Similar results were obtained by
Rodríguez et al. (2005) when they evaluated the
F2:1 generation of another crossing between L.
esculentum var. cerasiforme and L. esculentum
(a genotype bearing nor gene). Vallejo Cabrera
et al. (1994) stated that morphological and
vegetative traits in tomato, and other commercial
fruit quality attributes, were important in the
determination of the agronomic aptitude of a
genotype. These traits were associated to final
production. All recombinant lines integrated
a group close to LA722, confirming that
genes found in wild species were dominant on
cultivated germplasm. Bernacchi et al. (1998)
stated that this effect would be due to wild genes
belonging to a linkage group difficult to break
through intercrossing and recombination.
Based on these results, SL could be
independently selected in a breeding program.
Based on canonic correlations SL was not
associated to any other trait. Besides, the
power of discrimination of the SL was present
only at five levels of grouping. Recombinant
lines belonging to these clusters showed
recombination for this trait independently of
the rest of the traits.
In conclusion, differences in plant traits and
fruit quality components were found among
seventeen recombinant inbred lines of tomato,
obtained from an interspecific crossing
between Lycopersicon esculentum ‘Caimanta’
and accession LA722 of L. pimpinellifolium.
Characterization of these recombinant lines
demonstrated a wide genetic variability for
plant and fruit quality traits, possible to obtain
from crossing using L. pimpinellifolium. It was
remarkable that SL segregated independently
and most of these recombinant lines had higher
SL than LA722. These new genotypes, in which
the recombination between wild and cultivated
genes was favorable, constitute a new source of
variability for tomato breeding programs.
VOL 33 N˚2 MAY - AUGUST 2006
117
Table 4. Clusters of recombinant inbred lines obtained from an interespecific cross between Lycopersicon esculentum
‘Caimanta’ and the line LA722 of L. pimpinellifolium. Clusters (1A, 1B, 2C, 2D y 3) based on means values and standard
deviation for each trait, and F values and their significance after grouping in five levels.
Cluster 1A
4, 15, 16
Cluster 1B
Cluster 2C
Cluster 2D
Cluster 3
1, 2, 3,
LA 722, 5, 7, 9,
14, 17
6, 8
10, 11, 12, 13
Caimanta
Internodes length1, cm
2.9± 0.5
2.8± 0.2
2.9± 0.5
2.8± 0.4
3.3± 0.0
Stem perimeter:
Basal part, cm
3.2± 0.2
3.1± 0.2
2.9± 0.1
3.0± 0.2
4.6± 0.0
3.3± 0.3
3.1± 0.1
2.8± 0.0
2.8± 0.4
4.5± 0.0
Middle part, cm
Apical part, cm
2.7± 0.2
2.7± 0.1
2.3± 0.0
2.3± 0.4
3.5± 0.0
Flowers per inflorescence, nº
7.4± 1.7
7.1± 0.9
9.9± 0.1
9.2± 2.0
5.2± 0.0
Fruits:
Height (H), cm
2.8± 0.1
3.0± 0.3
1.3± 0.0
1.5± 0.3
5.4± 0.0
Diameter (D), cm
3.3± 0.1
3.5± 0.3
1.5± 0.0
1.7± 0.4
7.2± 0.0
Shape (D/H)
0.9± 0.0
0.9± 0.1
0.9± 0.0
0.9± 0.0
0.8± 0.0
18.3± 1.2
22.1± 4.1
2.0± 0.0
3.6± 2.2
159.2± 0.0
Weight, g
Shelf life, day
14.4± 2.8
21.8± 2.6
29.8± 2.2
18.7± 2.5
9.7± 0.0
Soluble solid, %
5.1± 0.4
4.7± 0.2
5.5± 0.4
5.2± 0.9
4.2± 0.0
5.2± 0.2
5.± 0.2
4.9± 0.1
5.1± 0.1
5.4± 0.0
pH
Titratable acidity 1, g
0.6± 0.1
0.5± 0.0
1.3± 0.2
0.9± 0.2
0.3± 0.0
1
2
F
0.4 ns2
16.8 **
7.7 **
4.7 *
3.0 ns
88.7 **
65.2 **
1.1 ns
751.0 **
15.9 **
1.2 ns
2.8 ns
14.9 **
Internodes length between third and fourth node. Titratable acidity= g of citric acid/100 g of homogenized fruit juice.
ns= not significant; *=p ≤ 0.05; **=p ≤ 0.01. ±, Standard error.
Resumen
Diecisiete líneas recombinates de tomate
generadas por el cruzamiento interespecífico
entre Lycopersicon esculentum cv. Caimanta y la
línea LA722 de L. pimpinellifolium se evaluaron
en su comportamiento para caracteres de planta
(longitud del tercer y cuarto entre nudo, número
de flores por inflorescencia y perímetro basal,
media y apical del tallo) y de fruto (contenido
en sólidos solubles, pH, acidez, diámetro, altura,
forma, peso y vida en estantería). Estas líneas
fueron significativamente distintas entre ellas
y respecto de los progenitores. Sin embargo,
las líneas recombinantes correspondieron a
fenotipos más afines al genotipo LA722 de
L. pimpinellifolium. Quince de estas líneas
tuvieron mejor vida en estantería que LA722,
progenitor silvestre con mayor vida en
estantería. Se utilizó un análisis multivariado
de correlaciones canónicas entre los caracteres
de planta y fruto y un análisis multivariado
de agrupamientos para caracterizar las líneas
recombinantes, incluyendo a los progenitores.
La primera correlación canónica explicó el 73%
de la variación total (p < 0,01). Los coeficientes
canónicos sugieren que plantas con mayor
perímetro del tallo y bajo número de flores
por inflorescencia producen frutos de mayor
peso y tamaño pero con menor acidez. En el
análisis de agrupamiento la vida en estantería
fue un importante carácter para discriminar las
líneas recombinantes y los progenitores cuando
se consideraron cinco niveles. En un grupo se
observó solamente al cv. Caimanta, mientras
que en los demás grupos se combinaron
genotipos con valores afines para los caracteres
perímetro del tallo, tamaño, peso, acidez y
vida en estantería de los frutos. Durante
el proceso de obtención de estas líneas se
recombinaron características de ambos
progenitores obteniéndose genotipos con
valores superiores al progenitor silvestre tal
como fue en la vida en estantería. En conclusión,
algunos de estos nuevos genotipos, constituyen
una nueva fuente de variabilidad, útil en
programas de mejoramiento de tomate.
Palabras clave: Calidad de fruto, Lycopersicon
esculentum, L. pimpinellifolium, vida en
estantería.
References
Bernacchi, D., T. Beck-Bunn, D. Emmatly, Y. Eshed,
S. Inai, J. Lopez, V. Petiard, H. Sayama, J.
Uhlig, D. Zamir, and S.D. Tanksley. 1998.
II. Evaluation of near isogenic lines carrying
single-donor introgressions for desirable
wild QTL-alleles derived from Lycopersicon
hirsutum and L. pimpinellifolium. Theorical
and Applied Genetics 97:170-180.
Buescher, R.W., W.A. Sistrunk, E.C. Tigchelaar, and
118
CIENCIA E INVESTIGACION AGRARIA
T.J. Ng. 1976. Softening, pectolytic activity
and storage life of rin and nor tomato hybrids.
HortScience 11:603-605.
Ferratto, J.A., A. Longo, R. Grasso, M.C. Mondino,
y A.C. Fernández. 1997. Ajuste del diagnóstico
agronómico del Proyecto Hortícola Rosario.
Publicación Miscelánea Argentina 32:1-48.
Grandillo, S., H.M. Ku, and S.D. Tanksley. 1999.
Identifying the loci responsible for natural
variation in fruit and shape in tomato. Theorical
and Applied Genetics 99: 978-987.
Hallauer, A.R. 1981. Selection and breeding
methods. p. 3-55. In: K.J. Frey (ed.). Plant
Breeding II. The Iowa State University Press,
Ames, Iowa, USA.
Hermsen, J.G. 1984. Some fundamental
considerations on interspecific hybridization.
Iowa State Journal Research 58: 461-474.
Jones, R.A., and S.J. Scott. 1983. Improvement of
tomato flavor by genetically increasing sugar
and acid contents. Euphytica 32:845.
Nguyen, V.Q., W.J. Ashcroft, K.H. Jones, and
W.B. Glasson. 1991. Evaluation of F1 hybrids
incorporating the rin (ripening inhibitor) gene
to improve the storage life and fruit quality
of fresh market tomatoes (Lycopersicon
esculentum Mill.). Australian Journal of
Experimental Agriculture 31:407-413.
Poysa, V. 1993. Use of Lycopersicon cheesmanii and
L. chmielewskii to increase dry matter content
of tomato fruit. Canadian Journal of Plant
Science 73:273-279.
Pratta, G., R. Zorzoli y L.A. Picardi. 2000.
Interacciones genéticas entre germoplasma
silvestre y cultivado de Lycopersicon spp. con
efectos sobre la calidad del fruto de tomate.
Plant Genetics Resources Newsletter 124:7-12.
Rick, C.M. 1976. Natural variability in wild species
of Lycopersicon and its bearing on tomato
breeding. Genética Agraria 30:249-259.
Rodríguez, G., G. Pratta, R. Zorzoli y L.A. Picardi.
2005. Caracterización de la generación
segregante de un híbrido de tomate con
genes nor y silvestres. Pesquisa Agropecuaria
Brasileira 40: 41-46.
Schuelter, A.R., F.L. Finger, V.W.D. Casali, S.H.
Brommonschenkel, and W.C. Otoni. 2002.
Inheritance and genetic linkage analysis of a
firm-ripening tomato mutant. Plant Breeding
121:338-342.
Shapiro, S.S., and M.B. Wilk. 1965. An analysis of
variance test for normality (complete samples).
Biometrika 52:591-611.
Snedecor, G.W., and W.G. Cochran. 1980. Statistical
Methods. 7th ed. The Iowa State University
Press, Ames, Iowa, USA. 626 pp.
Sokal, R.R., and F.S. Rohlf. 1969. Biometry. W.H.
Freeman and Company. San Francisco. USA.
253 pp.
Tigchelaar, E.C. 1986. Tomato breeding. p. 135-170.
In: M. J. Basset (ed.). Breeding Vegetable Crops.
AVI Publishing Company, Inc., Westport, CT,
USA.
Vallejo Cabrera, F.A., J.H. Pava, J.A Vargas y
A.P.A. Arango. 1994. Caracterización morfoagronómica de especies y variedades botánicas
del género Lycopersicon. Acta Agronómica
44:37-50.
Ward, J.H. 1963. Hierarchical grouping to optimize
an objective function. Journal of American
Statistical Association 69:236-244.
Warnock, S.J. 1991. Natural habitats of Lycopersicon
species. HortScience 26:466-471.
Weller, J.I., M. Soller, and T. Brody. 1988. Linkage
analysis of quantitative traits in an interspecific
cross of tomato (Lycopersicon esculentum x
Lycopersicon pimpinellifolium) by means of
genetic markers. Genetics 118:329-339.
Zorzoli, R., G. Pratta y L.A. Picardi. 1998. Efecto
de los mutantes nor y rin y de genes de origen
silvestres sobre la calidad postcosecha de los
frutos de tomate. Mendeliana 13:12-19.
Zorzoli, R., G. Pratta y L.A. Picardi. 2001. Líneas
seleccionadas para peso y vida en estantería
de los frutos a partir de un cruzamiento
interespecífico en Lycopersicon. Basic and
Applied Genetics 14:118 (Abstract).
Fly UP