Scientia Horticulturae 189 (2015) 168–174

Contents lists available at ScienceDirect

Scientia Horticulturae
journal homepage:www.elsevier.com/locate/scihorti

Inheritance of fruit cracking resistance of melon (Cucumis melo L.) fitting E-

0 genetic model using major gene plus polygene inheritance analysis

a,b a a a a
Zhenyu Qi , Junxing Li , Muhammad Ammar Raza , Xiaoxia Zou , Liwen Cao , Linli
a a,∗
Rao , Liping Chen
a
Institute of Vegetable Science, Zhejiang University, Key Laboratory of Horticultural Plants Growth, Development and Biotechnology, Agricultural Ministry of China,
Hangzhou 310058, China
b
Agricultural Experiment Station, Zhejiang University, Hangzhou 310058, China

article info abstract

Article history: Melon (Cucumis melo L.), an important cash crop, is cultivated worldwide. Fruit cracking frequently occurs during ripening,
Received 29 January 2015 resulting in economic loss. Thus, increasing fruit cracking resistance is crucial in melon breeding. However, until recently,
Received in revised form 16 March 2015 few reports have described standard evaluation and identification methods for fruit cracking resistance or the genetic analysis
Accepted 1 April 2015 Available online 21 of this trait, hindering the development of crack-resistant melons. In the present study, a mixed major gene plus polygene
April 2015 inheritance model was used to analyse fruit cracking resistance in six generations (P 1 , P2 , F1 , B1 , B2 and F2 ) of melon,
using the cracking-susceptible variety ‘MOIN-10’ and the cracking-resistant variety ‘RE-33’ as parents. The results showed
Keywords: that the genetic model E-0, incorporating two additive–dominance–epitasis major genes plus an additive–dominance–epitasis
Melon polygene, is the best-fitting genetic model for fruit cracking resistance. The heritabilities of the major genes in the B 1 , B2 and
Fruit cracking resistance
Major gene plus polygene inheritance F2 populations for fruit cracking resistance were 46.24%, 59.19% and 57.98%, respectively, and the heritabilities of
Genetic analysis polygenes for this trait were 21.26%, 2.02% and 9.74%, respectively. Genetic variances in B 1 , B2 and F2 accounted for
67.51%, 61.19% and 67.73% of phenotypic variance, respectively, showing that fruit cracking resistance was controlled
through genetic factors and environmental influence. Thus, in melon breeding, selection for cracking resistance in early
generation melon is desirable.

© 2015 Elsevier B.V. All rights reserved.

1. Introduction resulting from increased fruit turgor pressure (Kamimura et al., 1972; Li et al.,
2007), associated with environmental conditions (Gibert et al., 2007; Dürdane
Melon (Cucumis melo L.) is an important cucurbit crop, culti-vated et al., 2011) and cultivation meas-ures (Liu et al., 2010; Lichter et al., 2002).
worldwide with a high diversity of varieties (Pitrat et al., 2000; Lin, 2012). To some extent, the cultivation and physiological measures to alleviate the
During the ripening period, melon fruits are prone to splitting and cracking, severity of cracking reduce production losses resulting from fruit cracking
severely reducing market value, causing economic loss and negatively (Shimizu, 2005). However, it is difficult to solve this problem funda-mentally.
influencing the development of the melon industry (Wu et al., 2008; Wessel- Breeding varieties with resistance to cracking has been proposed as an
Beaver, 2011). Therefore, studies concerning the reduction of fruit cracking effective solution (Reynard, 1960), while genetic analysis of fruit cracking
have important significance. resistance is a prerequisite for the cultiva-tion of resistant varieties. Resistance
to fruit cracking in tomatoes is a quantitative trait that is controlled through
Fruit cracking during fruit development has previously been described in multiple genes and is strongly affected by environmental factors (Young,
several crops and fruits (Peet, 1992; Lane et al., 2000; Li et al., 2010; Lal et 1960; Prashar and Lambeth, 1960; Avdeev, 1980; Scott and Barten, 1992;
al., 2011; Khanal et al., 2011; Peng et al., 2012; Chen et al., 2012a,b ). It has Emmons and Scott, 1998; Wahyuni, 2014). In watermelon, resis-tance to fruit
been reported that fruit cracking reflects an imbalance between exocarp and cracking is controlled through a few major genes and modified through
flesh growth polygenes (Jiang et al., 2009).
However, in melon, the genetic
causes of fruit cracking resistance remain unknown.
∗
Corresponding author. Tel.: +86 571 88982006. E-mail
address: [email protected] (L. Chen).

http://dx.doi.org/10.1016/j.scienta.2015.04.004 0304-
4238/© 2015 Elsevier B.V. All rights reserved.
 Z. Qi et al. / Scientia Horticulturae 189 (2015) 168–174 169

It was essential to establish identification methods and eval-uation In the present study, muskmelon varieties RE-33, with high resistance to
standards for fruit cracking resistance prior to genetic analysis. Currently, cracking, and MOIN-10, with cracking susceptibility, were used as the parents
identification methods and evaluation stan-dards for resistance have only been in a cross. Six generations, P1 , P2 , F1 , B1 , B2 and F2 , were obtained. The
reported for pepper (Johnson and Knavel, 1990), watermelon (Sugiyama et identification methods and evaluation standards for resistance were
al., 1998; Jiang et al., 2010), tomato and apple (White and Whatley, 1955; established according to the classi-fication of fruit cracking based on the
Voisey and Lyall, 1965; Kamimura, 1977 ), and these standards and methods number, length and depth of the cracks. The genetic analysis and identification
might not be suitable for melon, as the texture and structure of melon fruit are of a genetic model were performed to determine the pattern of gene action of
different from the fruits of those crops. melon fruit cracking resistance. Moreover, the major genetic effects and
genetic force were estimated. These results might provide a basis for further
QTL analysis and practical marker-assisted selec-tion breeding for fruit
Several methods have been used for the genetic analysis of quan-titative cracking resistance in melon.
traits (Zalapa et al., 2008; Fita et al., 2008; Zhang et al., 2007), but these
methods only estimate the overall genetic effect and not the effects of
individual loci. In recent years, an analysis method using a major gene plus
polygene mixed genetic model has been developed for plants (Wang and Gai, 2. Materials and methods
1997, 1998; Gai et al., 2003). In contrast with previous methods, this
technique detects and identifies major genes and polygenes and estimates 2.1. Materials
genetic parameters. This gene model has been widely used in various crops,
such as rice (Wang et al., 1999), cotton (Yuan et al., 2002), rape (Zhang et al., A cross was made between muskmelon RE-33 (C. melo ssp. melo var.
2006), sorghum (Guan et al., 2012), Chinese cabbage (Zhuo et al., 2006), cantalupensis Naudin), an inbred line with high resistance to cracking, as the
tomato (Li et al., 2006), pepper (Chen et al., 2012a,b), mustard (Zhang et al., female parent (P1 ) and MOIN-10, an inbred line with cracking susceptibility,
2014), wheat (Ren et al., 2008) and cabbage (Su et al., 2012), for the genetic
as the male parent (P2 ; Fig. 1). MOIN-10, a succulent, crisp muskmelon
analysis of important agro-nomic traits. The results derived from this method
variety, with excellent and unique flavour, derived from a cross between MO-
were consistent with those from quantitative trait loci (QTL) analysis (Wang 21 (C. melo ssp. agrestis var. momordica Roxburgh) and IN-56 (C. melo ssp.
et al., 2008). This method has been demonstrated as a cost-effective tech- melo var. inodorus Jacquin), is susceptible and apt to cracking during the
nique for the study of complex quantitative traits. However, there are no ripening period in either spring or autumn cultivation. MO-21 and IN-56 were
reports concerning the application of the major gene plus polygene mixed supplied from the Vegetable Research Institute of Zhejiang University, and
genetic model for the genetic analysis of melon, and this method was used to RE-33 was obtained from the Mitsuo Seed Com-pany of Zhejiang. F 1 plants
analyse fruit cracking resistance in melon. were generated after a cross conducted in the spring of 2013 in a plastic
greenhouse at Huajiachi Cam-pus, Zhejiang University, and F 2 , B1 and B2
seedlings were obtained

Fig. 1. Two parents of melon and the grade of fruit cracking. Note: (a) fruit of ‘MOIN-10’; (b) fruit of F1 hybrid in “‘RE-33’ × ‘MOIN-10”’; (c): fruit of ‘RE-33’; (d) fruit cracking grade 0; (e) fruit
cracking grade 1; (f) fruit cracking grade 2; (g) fruit cracking grade 3; (h) fruit cracking grade 4 and (i) fruit cracking grade 5.
170 Z. Qi et al. / Scientia Horticulturae 189 (2015) 168–174

Table 1 The computer program was obtained from Dr. Zhang Yuan-
Grades and characteristics of fruit cracking in melons. ming, (Soybean Research Center of Nanjing Agricultural University)
Grade Number of Number of cracks Number of cracks (http://jpkc.njau.edu.cn/swtj/Upfiles/2007718202726314.zip).
cracks greater than l/2 of deeper than 1/2 of
circumference fruit flesh thickness
3. Results
Grade 0 0 0 0
Grade 1 [1,+∞) 0 0
Grade 2 [1,+∞) 1 0 3.1. Establishment of identification method and evaluation
Grade 3 [1,+∞) 1 1 standard for resistance in melon
Grade 4 [2,+∞) [2, +∞) 1
Grade 5 [2,+∞) [2, +∞) [2, +∞)
The grade of fruit cracking was classified into six levels (0–5), while the
Note: The circumference was calculated as the average value of two vertical meas-ures, and degree of cracking was associated with the number, length and depth of the
cracks smaller than l/4 of the circumference of the fruit skin were neglected. The fruit flesh cracks. Particularly, the depth of the cracks corresponded well with the degree
thickness was calculated as the average value of the thick-ness at the bottom and middle. All
of cracking (Table 1, Fig. 1). From grade 0 to 5, we observed an increase in
numbers expressed as intervals in the table are integers.
the num-ber, length and depth of the cracks (Fig. 1); these characteristics
corresponded to the grades of fruit cracking listed in Table 1. Furthermore, the
grade of fruit cracking was based on the three different characteristics listed in
through self-fertilization and backcrossing in the autumn of the same year. Table 1. The first characteristic indicated the total number of cracks (only
crack lengths greater than 1/4 of the circumference were included). we
calculated the circumference from the average value of two vertical measures
2.2. Investigation of traits and calculation of resistance index of fruit of the circumference of the fruit skin. The second characteristic indicated the
cracking number of cracks longer than l/2 of the circum-ference. The third
characteristic indicated the number of cracks deeper than l/2 of the fruit flesh
On March 5, 2014, P1 , P2 , F1 , B1 , B2 and F2 were sown, and the thickness, calculated from the average value of the fruit flesh thickness at the
seedlings were transplanted 3 weeks later. A randomized complete block bottom and middle.
design with three replicates was used for the non-segregating generations (P 1 ,
P2 and F1 ). The numbers of P1 , P2 , F1 , B1 , B2 and F2 seedlings were 36,
42, 42, 172, 183 and 229, respectively, with separation at 40 cm within and 90 By classifying fruit cracking in melons, we established a spe-cific
cm between the rows (30,000 plants/ha). The general management levels, identification method and evaluation standard for cracking resistance. The
water, fertilizer and ventilation were uniform, and drip irrigation was fruit cracking resistance index for comparing melon genotypes was calculated
employed. Stan-dard cultivation practices, vertical culture, single-vine
using the following formula:
pruning and 28th leaf topping were conducted during different growth
periods. Pollination was performed in the morning from the ninth to the 12th Resistance index of fruit cracking (RIFC)
nodes, and precisely two fruits were set on each plant.
100 (quantity of a grade × the grade) 100%.
High water saturation of soil was maintained at the matura-tion period = − (number of fruits scored × 5) ×
(approximately 32 days after pollination) to induce fruit cracking. Cracking
was scored every morning when the fruits were beginning to crack, including
the number, length and depth of the cracks, bottom and middle fruit flesh
3.2. Distributions of fruit cracking resistance score in six
thickness and vertical cir-cumference in two areas using Mm tape. Each
populations
cracked fruit was individually examined and classified based on the grade of
fruit cracking (Fig. 1; Table 1). Fruit without cracks received a grade of 0 at
Average RIFC values were compared among the six populations. The
60 days after pollination. The resistance index of fruit cracking (RIFC) was
calculated based on the grade of fruit cracking. average RIFC values of parents P1 and P2 were 92.22 and 16.42, respectively
(Table 2). However, the average RIFC value of F1 was 38.81, a value less than
half of the summed RIFC values of the two parents and closer to the RIFC
value of the male parent MOIN-10, indicating that fruit cracking susceptibility
2.3. Genetic analysis
was partially dominant to resistance in this cross combination. The average
SPSS (version 13.0) was used to analyse the six generations of melon. The RIFC values of the B1 , B2 and F2 populations were 73.37, 30.32 and 51.31,
mixed major gene plus polygene inheritance model was used to analyse fruit respectively, and the variance of the three segregating genera-tions was large,
suggesting that there were wide variations in these populations.
cracking resistance in the six genera-tions (P1 , P2 , F1 , B1 , B2 and F2 ) of
melon. The parameters of various generations and component distributions
were estimated using a maximum likelihood function, with an iterated
expectation and conditional maximization algorithm (IECM). The optimal
model was selected using the Akaike information criterion (AIC), and a test of Table 2
goodness-of-fit for the best genetic model was performed. The first- and Frequency distribution of the index of fruit resistance to crack in six population from “‘RE-33’ ×
second-order genetic parameters of the optimal genetic model were estimated ‘MOIN-10”’.
using the least-squares method according to Wang (1996). The heritabilities Generation Index of fruit cracking resistance Mean index
of the major genes and polygenes were estimated as: of fruit
0 10 20 30 40 50 60 70 80 90 100 cracking resistance

2 2 2 P1 1 2 5 8 20 92.22
Major gene heritability: h mg = mg / p ; polygene heritability:
P2 13 5 10 12 2 16.42
2 2 2
h pg = pg / p . F1 1 1 14 16 7 2 1 38.81
2 2 2 B1 1 2 1 4 13 17 26 33 32 26 17 73.37
p : phenotypic variance; mg : major-gene variance; pg : polygenic B2 27 38 56 17 15 8 7 0 2 1 1 30.32
2 2
variance; h mg : major gene heritability; h pg : polygene heritability. F2 1 3 18 27 31 24 35 20 10 1 2
51.31
 Z. Qi et al. / Scientia Horticulturae 189 (2015) 168–174 171

Fig. 2. Frequency distribution of the fruit cracking resistance traits in B1 , B2 and F2 populations. Note: a, b and c correspond to B1 , B2 , and F2 populations of “‘RE-33’ × ‘MOIN-10”’.

The frequency distributions of RIFC in the three segregating gen-erations

F2 , B1 and B2 were obtained (Fig. 2). The results based on Table 2 and Fig. 2
showed a clear multimodal distribution in the F2 population and skewed
distributions with clear quantitative genetic characteristics in the B 1 and B2
Table 3
populations, corresponding to the genetic characteristics of a mixed major The estimation of max-likelihood-value and AIC value of the different genetic models.
gene plus polygene model.

Model code Implication of model Max-likelihood-value AIC

A-1 1 MG-AD −3059.1184 6126.2368
3.3. Selection of a best-fitting inheritance model and tests of
A-2 1 MG-A −3062.0930 6130.1860
goodness-of-fit for the best genetic model A-3 1 MG-EAD −3192.9744 6391.9487
A-4 1 MG-AEND −3136.7068 6279.4136
Fruit cracking resistance scores for the six generations were analysed B-1 2MG-AD1 −3034.7869 6089.5737
using the main gene plus polygene mixed genetic model. Next, the maximum B-2 2MG-AD −3045.1477 6102.2954
B-3 2MG-A −3165.3037 6338.6074
likelihood function and AIC values were cal-culated, comprising 24 types of
B-4 2MG-EA −3062.8682 6131.7363
genetic models classified into five categories: one major gene (A), two major B-5 2MG-AED −3186.0449 6380.0898
genes (B), polygene (C), one major gene plus polygene (D) and two major B-6 2MG-EEAD −3186.0449 6378.0898
genes plus poly-gene (E; Table 3). The model with the minimum AIC value C-0 PG-ADI −3036.5518 6093.1035
was identified as the optimal model. Model was E-0 had the lowest estimated C-1 PG-AD −3047.1143 6108.2285
D-0 MX1-AD-ADI −3035.6179 6095.2358
AIC value, followed by E-1 and E-3. D-1 MX1-AD-AD −3036.8823 6091.7646
D-2 MX1-A-AD −3036.8850 6089.7700
Tests of goodness-of-fit were performed for the candidate mod-els (E-0, D-3 MX1-EAD-AD −3038.8323 6093.6646
2 D-4 MX1-AEND-AD −3036.0703 6088.1406
E-1 and E-3) to identify the optimal model, including equal distribution (U1 , E-0 MX2-ADI-ADI −3011.6255 *
6059.2510
2 2 2 E-1 MX2-ADI-AD −3018.0142 *
U2 and U3 ), Smirnov (nW ) and Kolmogorov test (Dn). The model with 6066.0283
E-2 MX2-AD-AD −3038.3953 6098.7905
the lowest number of values reaching statis-tical significance was selected as *
E-3 MX2-A-AD −3027.4063 6072.8125
the optimal model. The number of values reaching the significance level for E-4 MX2-EAED-AD −3046.3025 6108.6050
the E-0, E-1 and E-3 models were 11, 11 and 16, respectively, for the test of E-5 MX2-AED-AD −3047.0549 6112.1099
30 statistics characters (Table 4). Using the goodness-of-fit test and minimum E-6 MX2-EEAD-AD −3066.7083 6149.4165
AIC value, we determined that E-0 was the optimal model for melon fruit Note: MG: major gene model; MX: mixed major gene and polygene model; PG:
cracking resistance, suggesting that fruit cracking resistance was conferred by polygene model; A: additive effect; D: dominance effect; I: interaction (epistasis);
a two additive–dominance–epitasis major genes plus additive–dominance– N: negative; E: equal. For example, model E-0 = MX2-ADI-ADI is a mixed model with 2 major
epistasis polygene model. genes of additive–dominance–epistasis + an additive–dominance–epistasis effects polygene.

*
Indicates the lowest AIC values.
172 Z. Qi et al. / Scientia Horticulturae 189 (2015) 168–174

Table 4
Tests for goodness of fit model of fruit cracking resistance traits in different generations.

Model Generation Statistic parameter of fit model

2 2 2 2
U U U n W Dn
1 2 3

E-0 P1 0.441 (0.5068) 0.173(0.6771) 0.819 (0.3654) 0.6374 (<0.05)* 0.3038 (n = 36, CD(0.05) = 0.2332) (<0.05)*
F1 0.045 (0.8328) 0.221(0.6380) 1.133 (0.2871) 0.3583 (>0.05) 0.2211 (n = 42, CD(0.05) = 0.2150) (<0.05)*
P2 0.009 (0.9231) 0.626(0.4289) 7.785 (0.0053)* 0.4552 (>0.05) 0.2307 (n = 42, CD(0.05) = 0.2150) (<0.05)*
B1 0.221 (0.6381) 0.204(0.6512) 0.000 (0.9896) 0.3559 (>0.05) 0.1118 (n = 172, CD(0.05) = 0.1043) (<0.05)*
B2 4.210 (0.0402)* 5.927(0.0149)* 3.210 (0.0732) 1.2231 (<0.05)* 0.2163 (n = 183, CD(0.05) = 0.1011) (<0.05)*
F2 0.035 (0.8508) 0.094(0.7586) 0.251 (0.6164) 0.4000 (>0.05) 0.0966 (n = 229, CD(0.05) = 0.0903) (<0.05)*
E-1 P1 1.043 (0.3071) 0.662(0.4159) 0.492 (0.4830) 0.6995 (<0.05)* 0.3210 (n = 36, CD(0.05) = 0.2332) (<0.05)*
F1 0.639 (0.4239) 1.026(0.3111) 0.911 (0.3398) 0.4222 (>0.05) 0.2563 (n = 42, CD(0.05) = 0.2150) (<0.05)*
P2 0.417 (0.5185) 0.001(0.9811) 6.736 (0.0094)* 0.4488 (>0.05) 0.2482 (n = 42, CD(0.05) = 0.2150) (<0.05)*
B1 0.397 (0.5285) 0.372(0.5419) 0.000 (0.9990) 0.3776 (>0.05) 0.1184 (n = 172, CD(0.05) = 0.1043) (<0.05)*
B2 0.489 (0.4843) 1.506(0.2197) 4.842 (0.0278)* 0.9071 (<0.05)* 0.1874 (n = 183, CD(0.05) = 0.1011) (<0.05)*
F2 1.663 (0.1973) 1.729(0.1886) 0.070 (0.7908) 0.5565 (<0.05)* 0.1292 (n = 229, CD(0.05) = 0.0903) (<0.05)*
*
E-3 P1 1.777 (0.1826) 2.988(0.0839) 3.069 (0.0798) 0.731 (<0.05) 0.3638 (n = 36, CD(0.05) = 0.2332) (<0.05)*
F1 1.821 (0.1772) 2.307(0.1288) 0.722 (0.3954) 0.5433 (<0.05)* 0.2898 (n = 42, CD(0.05) = 0.2150) (<0.05)*
P2 0.351 (0.5535) 1.540(0.2146) 7.127 (0.0076)* 0.4824 (<0.05)* 0.2253 (n = 42, CD(0.05) = 0.2150) (<0.05)*
B1 17.344 (0.0000)* 16.192(0.0001)* 0.001 (0.9730) 2.068 (<0.05)
*
0.2241 (n = 172, CD(0.05) = 0.1043) (<0.05)*
B2 5.987 (0.0144)* 6.570(0.0104)* 0.602 (0.4377) 1.6412 (<0.05)* 0.2610 (n = 183, CD(0.05) = 0.1011) (<0.05)*
F2 0.000 (0.9912) 0.016(0.8991) 0.303 (0.5823) 0.4135 (>0.05) 0.1041 (n = 229, CD(0.05) = 0.0903) (<0.05)*
2 2 2 2 2 2 2
Note: U1 , U2 and U3 are the statistic of Uniformity test; n W is the statistic of Smirnov test; Dn is the statistic of Kolmogorov test. The probabilities are shown in the brackets of U1 , U2 and U3 ;
2
the threshold limit value of n W is 0.461 at 0.05 level.
*
Indicates significance at the 0.05 level.

3.4. Estimation of genetic parameters for the best-fitting model dominant and epistatic effects are important for the inheritance of melon fruit
cracking resistance.
The first- and second-order parameters of the E-0 model were estimated The results from second-order parameters showed that the heritabilities of
using the least-squares method (Table 5). The additive effects of the two 2
the major genes (h mg ) from the B1 , B2 and F2 populations were 46.24%,
major genes controlling cracking resistance were 8.1502 and 8.0989, 2
respectively, and the dominance effects were −13.5600 and −0.0462. The 59.19% and 57.98%, respectively, and the heritabilities of the polygenes (h pg
) were 21.26%, 2.02% and 9.74%, respectively. The genetic variances of the
potential ratio (ha/da) of the first major gene was −1.6338, indicating that the
dominant effect was greater than the additive effect. However, the potential B1 , B2 and F2 populations were 278.6305, 211.4700 and 281.4963,
respectively, accounting for 67.51%, 61.19% and 67.73%, respectively, of the
ratio (hb/db) of the second major gene was −0.0057, suggesting a lower domi-
nant effect. Clear interactions and epistatic effects were observed between the phenotypic vari-ance. The environmental variance of B 1 , B2 and F2
two major genes. The additive × additive effect (i) was 1.7779 and the populations was estimated as 134.1059 according to P1 , P2 and F1 , account-
dominance × dominance effect (l) was 26.1169. The additive × dominant ing for 32.49%, 38.81% and 32.27%, respectively, of the phenotypic variance.
interaction effect (jab ) between the two major genes was 5.9039, a value less These results suggested that fruit cracking resistance is controlled through
genetic factors, with a large effect of environ-mental factors, indicating that
than the dominance × additive inter-action effect (jba ), indicating that
selection for fruit cracking resistance in early generations is most efficient.
dominance × additive interaction effects contributed to melon cracking
resistance. Thus, additive,

4. Discussion

In the present study, six generations of genetic populations were obtained

Table 5 from melon parents with resistance and suscep-tibility to cracking. Joint
The estimation of genetic parameters of fit model of fruit cracking resistance traits.
segregation analysis was used for the genetic analysis of fruit cracking
1st order Estimate 2nd order Estimate resistance, and major-gene and polygenic effects were isolated from the total
parameter parameter phenotypic effects,
E-0 B1 B2 F2 contributing to the understanding of the genetic model of crack-ing resistance
8.1502 2
da p 412.7366 345.5761 415.6022 (Gai, 2005). The results showed that melon fruit cracking resistance was a
2
db 8.0989 e 134.1059 134.1059 134.1059
−13.5600 2 quantitative trait fitting the E-0 model, namely two additive–dominance–
ha mg 190.8690 204.5542 241.0071
hb −0.0462 pg
2
87.7615 6.9158 40.4892 epitasis major genes plus an additive–dominance–epistasis polygene model. In
i 1.7779 2
hmg (%) 46.24 59.19 57.98 addition, addi-tive, dominant and epistatic effects were observed between the
j 2
ab 5.9039 hpg (%) 21.26 2.02 9.74 two major genes. To the best of our knowledge, this is the first report of
j −7.5585
ba genetic studies of fruit cracking resistance in melon.
l 26.1669
ha /da −1.6638
−0.0057
Several identification methods have been developed for testing cracking
hb /db
resistance. Kamimura, 1977 reported a simple and accu-
Note: da : additive effect of the first pair major gene; db : additive effect of the second
rate method of vacuum immersion for testing cracking resistance in tomatoes.
pair major gene; ha : dominant effect of the first pair major gene; h b : dominant
effect of the second pair major gene; [d]: additive effective value of polygene; i: However, the tomato is a berry with a soft and juicy flesh and thin pericarp,
additive effect plus additive effect of the two major genes; l: dominant effect plus while the melon is a pepo with a hard flesh and thick pericarp. Therefore, the
dominant effect of the two major genes; jab : additive effect plus dominant effect vacuum immersion method is not suitable for melons. Jiang et al. (2010)
of the two major genes; jba : dominant effect plus additive effect of the two major
genes; ha /da : dominance degree of the first major gene; hb /db : dominance degree established identifica-tion methods and evaluation standards
2 2
of the second major gene;
2
p : phenotypic variance;
2 2
e : environmental variance; for cracking resistance in watermelon, wherein fruit cracking
mg : major-gene variance; pg : polygenic variance; h mg : major gene heritability;
2
h pg : polygene heritability. is primarily classified accord-ing to the number and length of
the cracks, without considering
 Z. Qi et al. / Scientia Horticulturae 189 (2015) 168–174 173

the depth of the cracks, which does not entirely reflect the degree of fruit B2 and F2 populations. Moreover, the obvious differences of heri-tability
cracking in melons. The identification methods and evalu-ation standards for among the populations was associated with the recessive effects of the major
cracking resistance used in the present study were thoroughly and carefully gene in resistance to cracking, consistent with previous studies in other crops
established through classifying fruit cracking according to the number, length (Li et al., 2006; Su et al., 2012; Yuan et al., 2002). These results indicate that
and depth of the cracks. melon fruit cracking resis-tance is dominated by main genes and modified
To date, explanations of fruit cracking have primarily focused on the through polygenes. In practical breeding, efficient selection should be
imbalance among ectocarp, mesocarp and endocarp growth resulting from conducted in early generations, reflecting the high heritability of major genes.
increasing fruit turgor pressure. Opera et al. (1996) reported that fruit with The increased genetic variation and heritability of major genes is favourable
cracking resistance consistently displayed a thick and elastic pericarp. A for molecular marker-assisted selection breeding.
tomato variety with cracking resis-tance was obtained through directional
selection for thick pericarp materials. In the present study, a crack-resistant Fruit cracking can emit fragrance and attract feeding animals, and seeds
parent, with a thick and elastic pericarp, was crossed with a susceptible par- can thereby be taken further afield. This trait is favourable for plant evolution
ent to improve the cracking resistance of melon. No cracking was observed
but disadvantageous for crop production and storage. Thus, breeding varieties
prior to the ripening period (32 days after pollination) in the B 1 , B2 and F2 with resistance to cracking will be the most effective solution (Fernández-
segregating populations, and subsequently the fruit of melon began to crack Trujillo et al., 2013). Encouraging results have been achieved in tomato
during the ripening period, suggest-ing that there is no direct relationship (Reynard, 1960) and persimmon (Yamada et al., 2002). However, in melon,
between pericarp scarring during the expansion period and fruit cracking breed-ing varieties with resistance to cracking have only recently been
during the ripen-ing period. The distributions of cracking resistance index in established. In the present study, we conducted basic research con-cerning
the B1 , B2 and F2 segregating populations showed that these individuals resistance to fruit cracking in melon. These results will provide a theoretical
displayed different degrees of resistance to fruit cracking, indicat-ing that reference for the QTL mapping and breeding of melon resistance to fruit
muskmelon can be used as a good breeding material for cracking resistance in cracking.
melon.

In the present study, the F1 generation, with an RIFC value skewed Acknowledgements
towards the susceptible parent, was obtained using the resistant parent RE-33
The author would like to thank Dr. Y. M. Zhang of the College of
hybridized with the susceptible parent MOIN-10. The RIFC of the B 2
Agriculture, Nanjing Agricultural University and Z. Z. Dong of the
generation following hybridization was also skewed towards the susceptible
parent. In addition, the two major genes controlling cracking resistance both Experimental Station, Zhejiang University for assistance and advice. Financial
showed neg-ative dominance effects (−13.56 and −0.0462), suggesting that assistance and research material were obtained from the Mitsuo Seed
cracking susceptibility had a greater dominant effect than resis-tance, Company of Zhejiang, China (NO. 588870-I21301).
consistent with previous breeding experience, in which the F 1 population
showed a tendency towards either one of the par-ents with cracking fruit in
References
breeding. However, the F1 generations of hybrid combinations from the
resistant materials showed different degrees of fruit cracking. This Avdeev, Y.I., 1980. Inheritance of resistance to concentric fruit cracking in tomato.
phenomenon might reflect the influ-ence of environmental factors, consistent Tsitologiyai Genetika 14 (1), 33–37.
with the results of studies of fruit cracking resistance in apple, tomato and Chen, H.H., Li, J.G., Zhang, J., Du, Y., 2012a. Research progress in mechanism of Chinese
jujube (Ziziphus jujuba Mill.) fruit cracking in Xinjiang. Xinjiang Agric. Sci. 6, 014, in
litchi, showing that fruit cracking is associated with mineral nutrition,
Chinese with English abstract.
endogenous hormones and water (Gibert et al., 2007; Shimizu, 2005; Peet and Chen, X.J., Fang, R., Zhou, K.H., Miao, N.S., Huang, C.L., 2012b. Major gene plus poly-gene
Willits, 1995). The additive effects of the two major genes control-ling inheritance analysis of fruit traits in peppe. Acta Bot. Boreali Occident. Sin. 32 (2), 0246–
cracking resistance were 8.1502 and 8.0989, indicating that the trait of 0251, in Chinese with English abstract.
cracking resistance had a good breed value (BV) in melon. From a research Dürdane, Y., Naif, G., Yusuf, Y., Mine, A., Perihan, C., 2011. Effect of different organic
standpoint, the main effect of cracking resistance traits is an interaction fertilizers on yield and fruit quality of indeterminate tomato (Lycopersicon escu-lentum).
Sci. Res. Essays 6 (17), 3623–3628.
complex, in which the additive effect plus dominant effect of the two major Emmons, C.L.W., Scott, J.W., 1998. Diallel analysis of resistance to cuticle cracking in tomato.
genes in combination (jab ) had a negative effect value of −7.5585. In J. Am. Soc. Hortic. Sci. 123 (1), 67–72.
Fernández-Trujillo, J.P., Lester, G.E., Dos-Santos, N., Martínez, J.A., Esteva, J., Jifon, J.L.,
addition, the dominant effect plus dominant effect of the two major genes (l)
Varó, P., 2013. Pre-and postharvest muskmelon fruit cracking: causes and potential
had a positive value of 26.1169, showing that resistance to fruit cracking has a remedies. HortTechnology 23 (3), 266–275.
strong interaction effect. In melon breeding, selection in the F 1 of hybrid Fita, A., Picó, B., Monforte, A.J., Nuez, F., 2008. Genetics of root system architecture
using near-isogenic lines of melon. J. Am. Soc. Hortic. Sci. 133 (3), 448–458.
combinations of resistance to fruit cracking, should make the two main gene Gai, J.Y., 2005. Segregation analysis of genetic system of quantitative traits in plants.
combinations exist at the same time, not only to the additive effect by using Hereditas 27 (1), 130–136, in Chinese with English abstract.
the major gene, also should can make good use of this interaction effect. Gai, J.Y., Zhang, Y.M., Wang, J.K., 2003. Genetic System of Quantitative Traits in Plants, vol. 8.
Science Press, Beijing, pp. 224–260, 94, (in Chinese).
Gibert, C., Chadœuf, J., Vercambre, G., Génard, M., Lescourret, F., 2007. Cuticular cracking on
nectarine fruit surface: spatial distribution and development in relation to irrigation and
thinning. J. Am. Soc. Hortic. Sci. 132 (5), 583–591.
Guan, Y.A., Zhang, H.W., Fan, Q.Q., Yang, Y.B., Qin, L., Wang, H.L., Wang, H.G., 2012.
Genetic analysis of plant height and brix value by using major gene and polygene
In the present study, the heritabilities of the two major genes in the B 1 , B2 inheritance model in a cross between common sorghum and sweet sorghum. J. Nucl. Agric.
and F2 populations were estimated as 46.42%, 59.19% and 57.98%, Sci. 26 (1), 0036–0042, in Chinese with English abstract.
Jiang, H.K., Yuan, X.H., Zhang, Z., Su, X.J., Xu, H., Song, B., Chen, L.Z., An, L.H., Shan,
respectively, whereas the heritabilities of the polygenes were 21.26%, 2.02% Q.W., Zhang, Q.A., 2009. Research on genotype in fruit cracking of watermelon. Acta
and 9.74%, respectively. These results suggest that the heritability of the Agric. Boreali Sin. S1, 106–109, in Chinese with English abstract.
polygenes within the B1 population is greater than that in the B2 and F2 Jiang, H.K., Zhang, Q., Fang, L., Dong, Y.X., Yan, C.S., Wang, M.X., Wang, Y., Yuan, X.H.,
2010. Evaluation on crack resistance of watermelon (Citrullus lanatus). Acta Hortic. 871,
populations, whereas the her-itability of the major genes of B 1 is less than that 223.
Johnson, D., Knavel, D.E., 1990. Inheritance of cracking and scarring in pepper fruit.
of B2 and F2 , suggesting that resistance to cracking is not only controlled by
J. Am. Soc. Hortic. Sci. 115 (1), 172–175.
the two main genes but also shows a greater degree of modification effects of Kamimura, S., Yoshikawa, H., Ito, K., 1972. Studies on fruit cracking in tomatoes. Bull.
the polygenes. In addition, the B1 population had a higher genetic background Hortic. Res. Stat. Min. Agric. For. Ser. C. 7, 73–138.
Kamimura, S., 1977. A method of testing crack resistance in tomatoes. Jpn. Agric.
of the parents’ resistance to cracking than the
Res. Q. 11, 111–114.
174 Z. Qi et al. / Scientia Horticulturae 189 (2015) 168–174

Khanal, B.P., Grimm, E., Knoche, M., 2011. Fruit growth, cuticle deposition, water uptake, and Voisey, P.W., Lyall, L.H., 1965. Puncture resistance in relation to tomato fruit cracking.
fruit cracking in jostaberry, gooseberry, and black currant. Sci. Hor-tic. 128 (3), 289–296. Can. J. Plant Sci. 45 (6), 602–603.
Wahyuni, S., 2014. Genetic Analysis of Quantitative and Fruit Cracking Characters in Tomato
Lal, S., Ahmed, N., Mir, J.I., 2011. Effect of different chemicals on fruit cracking in (Lycopersicon esculentum Mill.). Bogor Agricultural University, Bogor, Master Theses.
pomegranate under karewa condition of Kashmir Valley. Indian J. Plant Physiol. 16 (3),
326. Wang, J.K., 1996. Studies on Identification of Major-Polygene Mixed Inheritance of
Lane, W.D., Meheriuk, M., McKenzie, D.L., 2000. Fruit cracking of a susceptible, an Quantitative Traits and Estimation of Genetic Parameters. Department of Plant Breeding
intermediate, and a resistant sweet cherry cultivar. HortScience 35 (2), 239–242. and Biometrics, Nanjing Agricultural University, Nanjing, Doctorate Dissertation; in
Lichter, A., Dvir, O., Fallik, E., Cohen, S., Golan, R., Shemer, Z., Sagi, M., 2002. Cracking of Chinese with English abstract.
cherry tomatoes in solution. Postharvest Biol. Technol. 26 (3), 305–312. Wang, C.E., Gai, J.Y., Fu, S.X., 2008. Inheritance and QTL mapping of tofu and soymilk output
in soybean. Sci. Agric. Sin. 41 (5), 1274–1282, in Chinese with English abstract.
Li, J.S., Shen, H.L., Shi, Z.Q., 2006. Analysis on the major gene and polygene mixed
inheritance of lycopene content in fresh consumptive tomato fruit. Hereditas 28 Wang, J.S., Wang, J.K., Zhu, L.H., Gai, J.Y., 1999. Major-polygene effect analysis of resis-tance
(4), 458–462, in Chinese with English abstract. to bacterial blight (Xanthomonas campestris pv. Oryzae) in Rice. J. Genet. Genomics 27
Li, J.X., Zhu, C.H., Gao, J.Y., Yang, E.C., Yue, J.Q., 2010. Advances in pomelo fruit crack-ing. (1), 34–38, in Chinese with English abstract.
Subtrop. Agric. Res. 6 (3), 185–188, in Chinese with English abstract. Wang, J.K., Gai, J.Y., 1997. Identification of major gene and polygene mixed inheri-tance model
Lin, D.P., 2012. Comments on intraspecific classification of melon. China Cucurbits Veg. (5), and estimation of genetic parameters of a quantitative trait from F 2 progeny. J. Genet.
42–46, in Chinese with English abstract. Genomics 24 (5), 432–440, in Chinese with English abstract.
Liu, T.Z., Fu, Y.L., Jia, Y.L., Zhi, F.J., Zhang, C., Liu, J.K., 2010. Study on effect of bagging on Wang, J.K., Gai, J.Y., 1998. Identification of major gene and polygene mixed inher-itance model
cracking of apricot fruit. Acta Agric. Jiangxi 22 (12), 48–49, in Chinese with English of quantitative traits by using joint analysis of P1 , P2 , F1 , F2 and F2:3 generations. Acta
abstract. Agronomica Sin. 24 (6), 59–651, in Chinese with English abstract.
Li, W.C., Xie, J.H., Chen, J.Y., Zhang, X.M., Gong, D.Q., 2007. Advances on the mech-anism
and preventing measures of cracking fruit. Chin. J. Trop. Agric. 27 (1), 55–60, in Chinese Wessel-Beaver, L., 2011. “Exploding” fruits not unique to watermelon: fruit crack-ing in
with English abstract. Cucurbita moschata. In: Cucurbit Genetics Cooperative Report 33–34., pp. 55–56.
Opera, L.U., Studman, C.I., Banks, N.H., 1996. Fruit skin splitting and cracking. Hortic.
Rev. 19, 217–262. White, D.G., Whatley, B.T., 1955. A method to measure cracked apples and tomatoes. In:
Peet, M.M., Willits, D.H., 1995. Role of excess water in tomato fruit cracking. Proceedings of the American Society of Horticultural Science, pp. 289–290, vol. 65.
HortScience 30 (1), 65–68.
Peet, M.M., 1992. Fruit cracking in tomato. HortTechnology 2 (2), 216–223. Wu, M.Z., Yi, H.P., Feng, J.X., Wang, D.M., Erken, Y.S., Zhang, Y.B., Zhai, W.Q., Wu, H.B., Li,
Peng, X.L., Tian, Q.J., Chen, J.F., Zhang, J.X., Liu, S.B., 2012. Primary experimental report on J.S., 2008. Evaluation and utilization of the Xinjiang melon germplasm. Acta Hortic. 769,
prevention of fruit cracking for Siraitia grosvenorii. Hunan Agric. Sci. 19, 034, in Chinese 129.
with English abstract. Yamada, M., Sato, A., Ukai, Y., 2002. Genetic differences and environmental variations in
Pitrat, M., Hanelt, P., Hammer, K., 2000. Some comments on infraspecific classifi-cation of calyx-end fruit cracking among Japanese persimmon cultivars and selections. HortScience
cultivars of melon. In: VII Eucarpia Meeting on Cucurbit Genetics and Breeding, pp. 29– 37 (1), 164–167.
36, 510. Young, H.W., 1960. Inherited fruit cracking resistance in the tomato. In: Proceedings of the
Prashar, D.P., Lambeth, V.N., 1960. Inheritance of radial fruit cracking in tomatoes. In: Florida State Horticultural Society Held at Tampa, October 25–27 1996, Florida, pp. 53–
Proceedings of the American Society for Horticultural Science, pp. 530–537, 76. 148, vol. 73.
Yuan, Y.L., Zhang, T.Z., Guo, W.Z., Yu, J., Kohel, R.J., 2002. Major-polygene effect anal-ysis of
Ren, L.J., Zhang,.X., Zhou, M.P., Chen, Q.B., Ma, H.X., 2008. Heritability of resistance to super quality fiber properties in upland cotton (G. hirsutum L.). J. Genet. Genomics 29 (9),
wheat spindle streak mosaic virus. J. Triticeae Crops 28 (1), 154–159, in Chinese with 827–834, in Chinese with English abstract.
English abstract. Zalapa, J.E., Staub, J.E., McCreight, J.D., 2008. Variance component analysis of plant
Reynard, G.B., 1960. Breeding tomatoes for resistance to fruit cracking. In: architectural traits and fruit yield in melon. Euphytica 162, 129–143.
Proceedings of Plant Science Seminar, 1960, Camden, NJ, pp. 93–112. Zhang, L.L., Liu, B., Li, J.X., Chen, L.P., 2014. Genetic analysis of swelling traits of tumor-like
Scott, J.W., Barten, J.H.M., 1992. Inheritance of spider track incidence, a tomato fruit disorder. stems in tuber mustard (Brassica juncea Coss. var. tumida Tsen et Lee). J. Nucl. Agric. Sci.
In: Proceedings of the Florida State Horticultural Society, pp. 349–352, 105. 28 (1), 0023–0028, in Chinese with English abstract.
Zhang, S.F., Ma, C.Z., Zhu, J.C., Wang, J.P., Wen, Y.C., Fu, T.D., 2006. Genetic analysis of oil
Shimizu, T., 2005. Effects of GA inhibitor on the fruit cracking of melon (Cucumis melo L.). content in Brassica napus L. using mix model of major gene and polygene. Acta Genet. Sin.
Hortic. Res. (Japan) 4, 89–93. 33 (2), 171–180.
Sugiyama, K., Iwanaga, Y., Kanno, T., 1998. Development of a simple and accurate method of Zhuo, Z.C., Wan, E.M., Zhang, L.G., Zhang, M.K., Hui, M.X., 2006. Major gene plus poly-gene
testing cracking resistance in watermelon. Bull. Natl. Res. Inst. Veg. Ornamental Plants Tea inheritance analysis of bolting trait in heading Chinese cabbage. Acta Bot. Boreali
13, 1–12. Occident. Sin. 29 (5), 0923–0928, in Chinese with English abstract.
Su, Y.B., Liu, Y.M., Fang, Z.Y., Yang, L.M., Zhuang, M., Zhang, Y.Y., Zhang, X.L., Sun, P.T., Zhang, S., Hu, J., Zhang, C.F., Guan, Y.J., Zhang, Y., 2007. Genetic analysis of fruit shape traits
2012. Genetic analysis of head-splitting resistance traits in cabbage. Acta Hortic. Sin. 39 at different maturation stages in sponge gourd. J. Zhejiang Univ. Sci., B 8
(8), 1482–1490, in Chinese with English abstract. (5), 338–344.