Cucurbit Genetics Cooperative
Other Crop Genetics Cooperatives
Home About Membership Reports Gene Lists Conferences Links Search NCSU
Cucurbit Genetics Cooperative Report 15:13-16 (article 6) 1992

Correlation Between Parents and F1 Progeny in Earliness Heterosis and the Estimation of Traits Limits of Parents

Hongwen Cui, Yongtao Qi, Jianhui Liu and Zhongbo Ren

Department of Horticulture, Northwestern Agricultural University, Yangling, Shaanxi 172100, P.R. China

Cucumber (Cucumis sativus L.) is one of the important summer vegetables in China. There is an urgent requirement for early cultivars of cucumber for use in plastic tunnel culture in the spring.

Early maturity is an important trait and is composed of many components. It has been suggested that early maturity of cucumber is related to position of first pistillate flower, growth rate of the fruits, percentage of pistillate flowers. and number of pistillate flowers per plant. The objective of this research was to study the relationship of those traits and early yield of F1 hybrids, and to determine an optimum range for each trait to help plant breeders in the development of early maturing cucumbers.

Methods: Field work was conducted at the Vegetable Station, Northwest Agricultural University, China in 1988 and 1989. Four inbred lines were chosen as female parents: 'Yue 82', 'Jing 4-3-1'. 'Xinong 58-5', '7742'. Four inbreds were chosen as male parents: 'Changchun-1', 'Zhangqui M', and 'Black 235'. An incomplete diallel cross was used to generate 16 F1 . 'Changchun Mici' was used as the check. Soil for the experiment was covered by plastic film in the spring. A randomized complete block design was used in the experiment with 3 replications. There were 30 plants in each plot, with 0.67 m between rows and 0.23 m between plants in rows.

Ten plants in each plot were sampled for observation and measurement. The 14 traits observed were: yield per plant in early stage, harvested fruit numbers per plant in early stage, fruit set percentage in early stage, frequency of pistillage flowers/node on main stem in early stage, number of pistillage flowers on main stem in early stage, daily weight increment in grams of commercial fruits, total number of branches, number of fruiting branches, number of non-fruiting branch stem in early stage, leaf area of per plant in early stage, yield of a plant in early stage, number of leaves of a plant at flowering time, days from sowing to first staminate flowering three heterosis were described as follows:

  • Midparent heterosis = MP = (F1 - Midparent)/Midparent;
  • High parent heterosis - HP - (F1- High parent)/High parent;
  • Performance relative to check = CP = (F1 - Check)/Check

Correlations and regressions were performed with the ANALYST program.

Results: Correlation between heterosis of F1 progenies and midparent, and the diversity of two parents are the most direct features of parent combination. Correlations between such features and heterosis of F1 progeny were available both in the size and direction (Table 1). MP and HP were comparable, but the correlation of parents and F1 on CP showed large differences in size and direction.

Early maturity contributes to early yield. MP for early yield of F1 progeny was significantly negatively correlated with midparent (r = -0.831**) and positively correlated with diversity values of the two parents (r = 0.347). In that case, MP depended on the maturity of the parents (the later they mature, the easier it is to have large MP), and the diversity of the two parents for earliness.

HP for early yield was negatively correlated with diversity values of two parents (r = 0.473*), but not correlated with the mean values of the two parents (r = -0.174). In order to obtain F1 progenies with HP, parents with small diversity should be used.

We consider CP more important for cucumber breeding than MP. Correlation between CP for early yield of F1 progeny and the mean values of the two parents was significantly positive (r = 0.693.**), so parents with early maturity should be chosen. Also, the correlation between CP for early yield of F1 progeny and diversity of the two parents was not significant (r = -0.286), so parent diversity should be small.

Early yield was affected by many secondary traits or components. Analysis of the relations of those components and the F1 seemed very important to the breeding of earliness in cucumber. Results indicated that CP for number of fruits harvested per plant in early stage was similar to that of early yield. Number of pistillate flowers on the main stem, and pistillate flower density or frequency of pistillate flowers/node were also the same except that MP and CP seem correlated with diversity of the two parents in those two traits. The more total branches the parents had, the more difficult it was for the F1 to exceed the midparent value for branch number. Correlation between MP or HP for number of branches in the F1 and the mean value of the two parents were significantly negative (r = -0.628** and r = -0.676**, respectively. Therefore, more diverse parents had F1 progeny with less MP. CP was different in tat the forming of fruiting branches at early stage was almost uncorrelated with the features of the two parents. On the other hand, CP for total number of branches and number of non-fruiting branches at early stage in F1 showed significantly positive correlation with mean values of parents (r = 0.670** and r = 0.529**, respectively).

In order to obtain F1 progenies with few non-fruiting branches, parents with fewer branches should be chosen. The more diverse the parents were in flowering time and pistillate flowering time, the more likely it was that F1 progenies would be early flowering with a low node of pistillate flower than the midparents. CP for the time of pistillate and staminate flowering, the node of first pistillate flower and the leaf number at flowering time showed significantly positive correlation with midparents (r = 0.828**, r = 0.856**, r = 0.724**, respectively). Therefore, in order to get negative CP for those four traits and improve earliness of F1 progenies, the parents with the features of early flowering, lower node of first pistillate flower and fewer leaves at flowering time, and the parents with less diversity in these traits should be used.

The previous discussion relates to the limits for early yield. If mean value of the two parents exceeded the limit value, it would be impossible for F1 progenies to produce positive (or negative, if desired) CP. Regression analysis was used to estimate that limit value (Table 2). F1 progenies would not produce positive MP in each trait if early yield, fruit set percentage on main stem, percentage of pistillate nodes, number of pistillate flowers, daily increment of fruit, and number of fruiting branches at early stage were higher on average than 0.395 kg/plant, 73.38%, 30.021%, 2.954, 16.790 g/day, and 0.135 branch/stem, respectively.

Similarly, F1 would not produce negative MP if the total number of branches per plant, number of leaves per plant at flowering time, days from seed sowing to first staminate flower, days from seed sowing to first pistillate flower for two parents were on average more or higher than 2.9, 7.0, 38.8, 43.8, and 2.5 respectively. Although our conclusions were influenced by materials, samples, experimental regions and seasons, those trends have some general implications for cucumber breeders.

Discussion: Plastic tunnels for cucumber production in early spring is important in China. Thus, breeding of early maturing F1 hybrids is a high priority. Available information on earliness dealt mainly with first pistillate flower position (1, 2). Cui (3) and Lu (4) studied earliness in cucumber, and suggested that earliness was related to the first pistillate flower position, pistillate flower ratio (pistillate flower density), fruit growth rate, and other traits. Sterlenikova (5) suggested that the early maturity of cucumber was related to leaf area index at early stage. Other components of earliness were rarely reported.

In summary, F1 progeny were earlier when there was positive or negative correlation with parent traits, especially when MP and HP depended on the parent mean. Estimation of parent feature limits was important in choosing parents for maximizing the use of heterosis. The suggestion that there should be fewer leaves on parents and F1 hybrids during flowering, was different from the results of Sterlenikova. That may be due to use of different genotypes and environments. Additional research is needed in that area.

Table 1. Correlation between F1 heterosis and parent trait in cucumber.

Midparent heterosis
High-parent heterosis
Performance relative to check
Parent mean
Parent diversity
Parent mean
Parent diversity
Parent mean
Parent diversity
EYP -0.831** 0.347 0.174 -0.473* 0.693** -0.286
EFP -0.461 0.028 -0.308 -0.628** 0.776** -0.183
EFS -0.967** 0.810** -0.524* -0.003 -0.326 0.150
ED -0.529* 0.056 0.091 -0.610** 0.751** -0.027
FFS -0.543 0.045 -0.0102 -0.424 0.737 0.053
DWI -0.886** 0.821** -0.395 0.184 0.193 -0.134
TBS -0.38* -0.685** -0.380 -0.713** 0.670** -0.047
EBS -0.595** -0.604** -0.582* -0.615** -0.083 -0.039
NFBS -0.470 -0.628** -0.478* -0.676** 0.529** -0.080
LAP -0.285 -0.376 -0.168 -0.172 0.613** 0.474*
LSP -0.789** -0.801** -0.824** -0.886** 0.699** 0.368
DMF -0.477** -0.429 -0.551** -0.749** 0.856** 0.369
DFF -0.547** -0.524** -0.653** -0.815** 0.828** 0.424
NFF -0.515** -0.677** -0.700** -0.781** 0.724**  

Z EYP - early yield per plant, EFP - early fruits harvested per plant, EFS - early fruit set percentage of main vine, FD - pistillate flower density of main vine, FFS - pistillate flowers of main vine, DWI - daily weight increment in gram per commercial fruit, TBS - total branches per plant, FBS - fruiting branches per plant, NFBS - non-fruiting branches per plant, LAP - leaf area per plant in early stage, LSP - leaves per plant in flowering time, DMF - days from sowing to staminate flowering, DFF - days from sowing to pistillate flowering, NFF - node of first pistillate flower.

Table 2. Approximate estimation of limits from F1 MP and parental mean values in a regression model.

Regression constant
Regression coefficient
Early yield per plant
Early fruit-set (% on main stem)
Pistillate flower density of main stem
No. pistillate flowers/main stem
Marketable fruit wt. gain (g/day)
Total branches/plant
Fruiting branches/plant 4714.29 -30774.67** 0.153
leaves/plant at flowering
Days from sowing to staminate flowering
Days from sowing to pistillate flowering
Node of first pistillate flower

Literature Cited

  1. Tan, Qimeng. 1961. Breeding and seed production of vegetable. Agronomy Press. Bejing, China.
  2. Yang, Yuchun 1979. The utilization of cucumber F1heterosis. New Agriculture. 12.
  3. Cui, Hongwen and Junjun Deng. 1987. Combining ability of quantitative characteristics and genetic analysis of cucumber inbred lines. Acta Univ. Septerionali Occident Agric. Vol. 15 (3) pp.63.
  4. Lu, shuzhen and Feng Hou. 1980. The researches on F1 hereditary features of cucumber. Tianjing Agric. Sci. 1: 43-47.
  5. Sterlenikova, H.H. 1984. Hybrids heterosis of cucumber breeding (in Russian).
Home About Membership Reports Gene Lists Conferences Links Search NCSU
Department of Horticultural Science Box 7609North Carolina State UniversityRaleigh, NC 27695-7609919-515-5363
Page citation: Wehner, T.C., Cucurbit Genetics Cooperative;
Created by T.C. Wehner and T. Ng, 1 June 2005; design by C.T. Glenn;
send questions to T.C. Wehner; last revised on 6 November, 2009