Performance of Cucumis melo L. Hybrids and their Heterosis for Fruit Quality in Three-way Reciprocal Crosses

A. Ching and Liyang Wu

Alternative Crops Research Center, Northwest Missouri State University, Maryville, MO 64468

Additional index words. Chamie, CH, fruit weight, fruit shape, total soluble solids

Abstract. In an effort to improve fruit quality characteristics in a Korean melon, Cucumis melo var. Chamie (CH), an inbred genetic line was crossed with five hybrid linesCrenshaw hybrid (CR), Sweet Thing hybrid (ST), Limelight hybrid (LI), Sweet Dream hybrid (SD), and Venus hybrid (VE)in three-way reciprocal crosses. In 1996, F1 progenies were planted in the field and established on a rectangular lattice completely randomized block design with four replications. Heterosis for fruit weight, fruit shape, and total soluble solids (TSS) for the three-way reciprocal crosses were estimated as defined by Rai (1979) and Legg and Souther (1968). Out of 10 F1 hybrid progenies, the overall performance of six F1 hybrids showed favorable heterosis over the mid-parent values ranging from 0.35% to 60.71% for fruit weight. Favorable significant heterosis over the mid-parent value was also found for fruit shape in some of the F1 progenies. The highest TSS was found in the F1 progeny from the ST x CH cross and the lowest from the F1 progeny of the reciprocal cross CH x ST. The overall performance of F1 CR x CH and F1 ST x CH showed favorable significant heterosis over the midparent values. Highly significant heterosis was found on the F1 ST x CH cross as compared to a standard cultivar, Honeydew (HD). The average performance of parental lines, their F1 progenies and the average heterosis for all fruit quality traits showed that only the fruit shape index had favorable and significant heterosis. Three-way cross F1 hybrid progenies tended to give superior shape but smaller size and less heterotic values when compared to their parental values.

 

Northwest Missouri melon producers have been successful in their produc-tion and have often requested information on new specialty melons. 'Chamie', a white-fleshed Korean melon (Cucumis melo L. ), was introduced as a potential alternative crop in 1993. Three years of collected data on this melon have indicated poor fruit quality traits. Fruit size and shape tended to be small oval to pear-shaped, and low brix was shown to be constant.

In muskmelon, fruit weight varies from a few grams to several kilograms. Ma et al. (1995) stated that larger melons are preferred because of a higher market value. Parthasarathy (1978) reported that fruit weight character showed a very high genetic coefficient of variation and heritability, which are of value in obtaining quick genetic improvement. Other desirable features of muskmelons include round to slightly oval shape fruit within the weight range of 0. 68 to 1. 36 kg (Lippert and Legg, 1972a). Lumsden (1914) reported that

round shape was dominant over oval shapes in muskmelon, however, Bains and Kang (1963) indicated that cylindrical fruit shapes were dominant over round, and crosses between these two shapes resulted in oblong fruit shapes. Foster (1968) and Nath and Dutta (1971) observed monogenic control of fruit shape without complete dominance. Ma (1995) reported that the fruit shape index in F1 hybrids was greater than the midparent value or higher than the best parental combination of round or elongated fruit shape. Ma (1995) and Ramasamy et al. (1977) indicated that melon shape inheritance is quantitative and polygenically controlled.

The most important characteristic determining fruit quality of melon is the total soluble solids (TSS). TSS are found to be intermediate in F1 hybrids when compared to their parental lines (Nath and Dutta, 1971; Abadia et al., 1985). Dyutin and Prosvirnin (1977) observed no differences in the TSS in hybrids of muskmelon and their recip

Cucurbitaceae '98


rocal crosses. However, some studies have reported that F1 hybrids produced high sugar content in the mesocarp (Knott and Lorenz, 1950; Mohr, 1960). In other studies, Lippert and Legg (1972) found that the mean square of cultivar vs. crosses, which is a measure of the importance of average heterosis, was significant for the percent TSS.

Heterosis is responsible for the superior performance of a hybrid individual when compared with their parental lines. Falconer (1981) has indicated that heterosis can be expressed when the parental lines of a hybrid contain different alleles at a locus and that there are different levels of dominance among these alleles. Fehr (1987) stated that according to the dominance hypothesis, heterosis is caused by complete or partial dominance. However, in the overdominance hypothesis, the value of the heterozygote is considered superior to the value of either homozygote. This was also supported by Moll and Stuber (1974).

In the present research study, known commercial hybrid melons with known phenotypic performance were used for crossing with 'Chamie' an inbred Korean melon in a reciprocal manner. The specific objective was to determine the performance and the heterotic values for some fruit quality traits.

Materials and methods

During Summer 1995, seeds from five commercial muskmelon hybrid cultivars and an inbred line were planted in 12. 0-cm-diameter plastic pots containing a soil media of 1 part soil (silty clay) : 1 part sand by volume. The pots were placed at random in a greenhouse with 28 °C day and 25 °C night temperatures. The pots were watered daily to achieve 80% soil moisture level. As the plants grew, trellises were set up for the plants to climb. Seven grams of 10­10­10 fertilizer were applied to each pot when the plant reached a length of 30 cm.

All mature pistillate flower buds that were likely to open the following day were bagged to prevent free cross-pollination. Staminate flower buds which were likely to open the following day were detached from the plants to prevent self-pollination. These preparations were carried out

between 4:00 and 6:00 pm daily.

The day after the buds opened, pollen from the male parent source was transferred onto the stigma of the female parent by shedding the pollen from the anther of the staminate flower. After the female parent flower was pollinated, the flower was rebagged and tagged with the details of the cross.

The bags were removed when the fruit reached 3 to 4 cm in diameter. The list of parental lines and their reciprocal crosses are presented in Table 1. The Korean melon 'Chamie' was used as a source of pollen to pollinate the other five commercial cultivars: Crenshaw hybrid (CR), Sweet Thing hybrid (ST), Limelight hybrid (LI), Sweet Dream hybrid (SD), and Venus hybrid (VE). Conversely, pollen from the five commercial hybrid cultivars was used to pollinate 'Chamie' (CH). This process completed the three-way crosses for this study.

When the fruit were at the mature stage, the melons were cut open, and the seeds were harvested and rinsed. They were dried at room temperature for two days and then stored for the 1996 growing season.

Seeds of all parents and the F1 progenies were planted during the first week of May 1996 in 4 cm diameter round-plugged plastic containers using Metro-mix as a soil medium. The containers were placed in a greenhouse at 28 °C day and 25 °C

Table 1. Reciprocal cross index and cultivar symbols (female x male).

Parent or crossz

P1 Chamie (CH)

P2 Crenshaw hybrid (CR)

P3 Sweet Thing hybrid (ST)

P4 Limelight hybrid (LI)

P5 Sweet Dream hybrid (SD)

P6 Venus hybrid (VE)

C1 Chamie x Crenshaw hybrids

C2 Chamie x Sweet Thing hybrid

C3 Chamie x Limelight hybrid

C4 Chamie x Sweet Dream hybrid

C5 Chamie x Venus hybrid

C6 Venus hybrid x Chamie

C7 Crenshaw hybrid x Chamie

C8 Sweet Thing hybrid x Chamie

C9 Limelight hybrid x Chamie

C10 Sweet Dream hybrid x Chamie

zAll crosses are designated with the letter C and a number. All parents are designated with the letter P and a number.

Cucurbitaceae '98


night temperatures. The seedlings were watered daily and fertilized at a rate of 30 kg·ha­1 with a soluble 15­15­15 Peters fertilizer. Once the transplants developed the first two true leaves, they were given a final application of the 15­15­15 soluble Peters fertilizer before field transplantation. In the third week of May 1996, transplants of the parent lines and the F1 progenies were grown in a field at the Northwest Missouri State University Experimental Farm in Maryville, on a Sharpsburg silty clay soil (fine, montmorillonitic, mesic Typic Argiudolls).

The experiment was conducted by employing a rectangular lattice complete block design with four replications. The experimental plots measured 8.23 ¥ 1.53 m with an alley of 1.23 m between

plots. The distance between plots was 2.5 m. Additional irrigation was given whenever the first 8 cm of top soil showed signs of dryness. Sevin-80 was sprayed on the crown of each plant at a rate of 2.40 g·L­1 every 10 days (or after a rainfall) to control cucumber beetles that were present in the field. Plants were checked daily for powdery mildew infection throughout the growing and production cycle.

All mature, full-slip fruit were harvested during August 1996 as they ripened from the middle five plants in each plot. Data were collected by weighing the fruit using a top-load electronic balance. Data were also collected for two fruit quality traits as follows: Shape index(fruit length was divided by fruit width to give a shape index);

Table 2. Mean performance of F1 hybrids and their heterosis in percentage for fruit weight.

Heterosis over

Cross (female x male) Mean (g) Midparent Better parent Honeydewz

CH x CR (overall) 699.2 ± 166.9 ­59.28** ­76.69** ­63.20**

Elongated 735.3 ± 199.3 ­57.18** ­75.49** ­61.30**

Oval 696.9 ±168.3 ­59.41** ­76.77** ­63.32**

Round 672.6 ± 125.7 ­60.83** ­77.58** ­64.60**

CR x CH (overall) 1723.0 ± 1120.0 0.35 ­42.57** ­9.32

Elongated 1747.0 ± 881.0 1.75 ­41.77** ­8.05

Oval 1060.7 ± 390.7 ­38.22** ­64.64** ­44.17**

Round 4320.0 ± 1746.0 151.60** 44.00** 127.37**

CH x ST (overall) 1663.4 ±389.8 47.99** ­8.30 ­12.45*

ST x CH (overall) 1063.6 ± 334.9 ­5.37 ­41.37** ­44.02**

Elongated 1097.0 ± 296.3 ­2.40 ­39.53** ­42.26**

Oval 1017.0 ± 344.5 ­9.52 ­43.94 ­46.47**

Round 1182.9 ± 361.6 5.24 ­34.79 ­37.74**

CH x LI (overall) 1217.2 ± 411.0 ­28.40** ­58.99** ­35.94**

Elongated 1150.7 ± 308.9 ­32.31** ­61.23** ­39.44**

Oval 1222.4 ± 436.0 ­28.09** ­58.81** ­35.66**

Round 1661.0 ± 554.0 ­2.29 ­44.04** ­12.58*

LI x CH (overall) 2400.0 ± 572.9 41.18** ­19.14* 26.32**

Oval 2361.2 ± 604.3 38.89** ­20.44** 24.27**

Round 2447.0 ± 556.0 43.94** ­17.55* 28.79**

CH x SD (overall) 1048.80 ± 316.7 ­17.87* ­50.53** ­44.80**

SD x CH (overall) 2052.3 ± 508.9 60.71** ­3.19 8.02 Oval 1968.0 ± 517.9 54.11** ­7.17 3.58 Round 2101.2 ± 502.4 64.54** ­0.89 10.59*

CH x VE (overall) 1097.2 ± 299.9 2.45 ­35.80** ­42.25**

Elongated 1118.0 ± 296.0 4.39 ­34.58** ­41.16** Oval 1058.3 ± 308.4 ­1.19 ­38.07** ­44.30**

VE x CH (overall) 1130.4 ± 422.9 5.55 ­33.86** ­40.51**

Oval 900.8 ± 190.4 ­15.89* ­47.29** ­52.59**

Round 1265.2 ± 462.9 18.13* ­25.97** ­33.41**

CH=433 CR=3000 ST=1814 LI = 2968

SD=2120 VE=1709 HD=1900

zStandard cultivar.

*,**Significant at P = 0.05 or 0.01, respectively.

Cucurbitaceae '98


total soluble solidspercent of total soluble solids was determined by refractometry on tissue removed from the center of the axis of the fruit.

The determination of fruit shapes was accomplished using the shape index as follows: Shape index >1.6 = elongated; >1.2 < 1.6 = oblong; <1.2 = round.

Measurement of heterosis: the performance of a F1 progeny (resulting from a cross) relative to its parents can be expressed in two ways, as indicated by Fehr (1987) and Rai (1979). The standard heterosis was computed using the method of Rai (1979), and in this study [standard cultivar = Honeydew (HD)] the average heterosis was determined by the method of Legg and Souther (1968).

Results

The mean values of fruit weight for parents and F1 hybrids and the mean performance of F1 hybrids and their heterosis are given in Table 2. The data show that fruit weight was highly variable for all F1 hybrids and their parental lines. The inbred line CH produced the smallest fruit, weighing 430 g, while CR produced the largest fruit, weighing 3,000 g. The F1 hybrids when using CH as the female parent produced small melons. Out of ten hybrids, the overall performance of six hybrids showed heterosis over midparent values ranging from 0.35% to 60. 71%. The F1 hybrids LI x CH, CH x ST and SD x CH have shown significant heterosis. Only round shape melons from the

Table 3. Mean performance of F1 hybrids and their heterosis in percentage for fruit shape index.

Heterosis over

Cross (female x male) Mean (g) Midparent Better parent Honeydewz

CH x CR (overall) 1.41 ± 0.23 ­11.93* ­21.71** 18.42*

Elongated 1.76 ± 0.13 10.11* ­2.13 48.04**

Oval 1.39 ± 0.12 ­13.17* ­22.82** 16.75*

Round 1.14 ± 0.05 ­28.83** ­36.73** 4.30

CR x CH (overall) 1.79 ± 0.37 11.62* ­0.78 50.08**

Elongated 1.99 ± 0.25 24.61** 10.77* 67.55**

Oval 1.45 ± 0.08 ­9.44 ­19.51* 21.76**

Round 1.11 ± 0.05 ­30.53** ­38.24** ­6.59

CH x ST (overall) 1.28 ± 0.10 ­14.55* ­28.79** 7.71

ST x CH (overall 1.51 ± 0.28 0.73 ­16.06* 26.97**

Elongated 1.84 ± 0.18 22.48** 2.07 54.39**

Oval 1.42 ± 0.10 ­5.43 ­21.19** 19.20*

Round 1.09 ± 0.07 ­27.31** ­39.42** ­8.37

CH x LI (overall) 1.56 ± 0.23 0.54 ­13.43* 30.95**

Elongated 1.80 ± 0.15 16.10* ­0.02* 51.23**

Oval 1.45 ± 0.10 ­6.30 ­19.31* 22.05**

Round 1.03 ± 0.05 ­33.39** ­42.64** ­13.24*

LI x CH (overall) 1.28 ± 0.16 ­17.24* ­28.73** 7.80

Oval 1.39 ± 0.12 ­10.26* ­22.73** 16.88* Round 1.11 ± 0.06 ­28.15** ­38.13** ­6.42

CH x SD (overall) 1.59 ± 0.14 5.80 ­11.83 33.36**

SD x CH (overall) 1.10 ± 0.14 ­22.23** ­35.19** ­1.97

Oval 1.31 ± 0.12 ­12.46* ­27.05** 10.34*

Round 1.08 ± 0.07 ­27.89** ­39.91** ­9.11

CH x VE (overall) 1.71 ± 0.21 17.59* ­5.28 43.28**

Elongated 1.83 ± 0.14 26.38** 1.81 53.99**

Oval 1.47 ± 0.08 1.14 ­18.53* 23.24**

VE x CH (overall) 1.21 ± 0.19 ­16.62* ­32.83** 1.60

Oval 1.37 ± 0.19 ­5.25 ­23.67** 15.45* Round 1.11 ± 0.13 ­23.26** ­38.18** ­6.50

CH=1.8 CR=1.4 ST=1.2 LI = 1.3

SD=1.2 VE=1.1 HD=1.19

zStandard cultivar.

*,**Significant at P = 0.05 or 0.01, respectively.

Cucurbitaceae '98


cross CR x CH expressed significant heterobeltiosis.

Muskmelon fruit shape was highly variable. Because of the three-way crosses, most of the F1 hybrid progenies showed variations in fruit shape, from elongated to round. The mean performance of F1 hybrids and their heterosis is shown in Table 3. The overall performance of the progeny from the cross CR x CH produced elongated fruit shape. Sweet Dream hybrid x Chamie had the most symmetrical melons; furthermore, some F1 hybrids showed greater extended shape values than their midparent values, indicating significant heterosis. On the other hand, some F1 hybrids had a more round shape indicating a negative heterosis and a negative nonsignificant heterobeltiosis.

The observed means showed variations ranging from 4.6% to 15% TSS. The highest mean of 15% is from the cross ST x CH and was observed only when ST was used as the female parent.

However, when CH was used as the female parent, the observed mean for TSS was 4.6% (Table 4). The studies of the TSS showed that ST has the highest sugar content while LI has the lowest sugar content. The highest TSS of all F1 hybrid fruit was produced by ST x CH, while the lowest TSS of hybrid fruit was produced by CH x ST. The overall performance of CR x CH and ST x CH showed significant heterosis over midparent value, as shown in Table 4. Furthermore, F1 progenies of the crosses ST x CH and CH x VE showed nonsignificant heterobeltiosis. Standard heterosis was only found in the ST x CH. When comparing the heterosis values with a standard cultivar, in this case honeydew (HD), highly significant heterosis was found in the F1 ST x CH.

The average performance of the parents, their F1 hybrids and their average heterosis for all quality traits are given in Table 5. The results showed

Table 4. Mean performance of F1 hybrids and their heterosis in percentage for total soluble solids (°Brix).

Heterosis over

Cross (female x male) Mean (g) Midparent Better parent Honeydewz

CH x CR (overall) 9.6 ± 1.1 ­11.93* ­21.95** ­23.2**

Elongated 10.0 ± 1.5 ­8.26 ­18.70* ­20.0**

Oval 9.2 ± 1.0 ­15.60* ­25.20** ­26.4**

Round 9.8 ± 1.0 ­10.09* ­20.33** ­21.6**

CR x CH (overall) 12. ± 1.4 12.84* 0.00 ­1.6

Elongated 12.5 ± 1.4 14.68* 1.63 0.0

Oval 12.1 ± 1.5 11.01* ­1.63 ­3.2

Round 11.6 ± 0.5 6.42 ­5.69 ­7.2

CH x ST (overall) 4.6 ± 0.9 ­65.67** ­68.28** ­63.2**

ST x CH (overall) 15.0 ± 1.2 11.94* 3.45 20.0**

Elongated 11.2 ± 1.2 6.16 ­8.94 ­10.4*

Oval 9.2 ± 0.9 ­12.80* ­25.20** ­26.4**

Round 9.0 ± 1.5 ­14.69* ­26.83** ­28.0**

LI x CH (overall) 8.0 ± 1.9 ­24.17** ­34.96** ­36.0**

Oval 8.0 ± 2.2 ­24.17** ­34.96** ­36.0**

Round 7.9 ± 1.7 ­25.12** ­35.77** ­36.8**

CH x SD (overall) 13.0 ± 1.7 1.17 ­2.99 4.0

SD x CH (overall) 10.7 ± 2.1 ­16.73* ­20.15** ­14.4*

Oval 10.3 ± 2.3 ­19.84* ­23.13** ­17.6*

Round 11.2 ± 1.9 ­12.84* ­16.42* ­10.4*

CH x VE (overall) 12.5 ± 1.7 2.88 1.63 0.0

Elongated 12.6 ± 1.5 3.70 2.44 0.8 oval 12.2 ± 1.2 0.41 ­0.81 ­2.4

VE x CH (overall) 11.6 ± 2.0 ­4.53 ­5.69 ­7.2

Oval 12.3 ± 1.5 1.23 0.00 ­1.6

Round 11.0 ± 2.2 ­9.47 ­10.57 ­12.0*

CH=12.3 CR=9.5 ST=14.5 LI=8.8

SD=13.4 VE=12.0 HD=12.5

zStandard cultivar.

*,**Significant at P = 0.05 or 0.01, respectively.

Cucurbitaceae '98


Table 5. Average heterosis for three quantitative characters.

Value

Character Parental Hybrid Average

Fruit weight (g) 2007.33 1409.51 ­29.78**

Fruit shape (index) 1.33 1.45 9.02**

Total soluble solids (%) 11.75 10.73 ­8.68*

*,**Significant at P = 0.05 or 0.01, respectively.

produce greater fruit length when crossed with the CR hybrid line, a significant heterosis was obtained for the elongated F1 hybrid fruit. However, no significant heterosis was observed when CH was crossed with ST, LI, and VE.

An inference concerning the relative magnitude of gene actions could not be made from the analyzed data for fruit shape index (Table 3). This is probably due to the distribution of genes between the parental lines having positive and negative effects which diminish the effects of each other (Ghaderi and Lower, 1979).

Total soluble solids is the best indication of fruit quality and probably the most important quality character with respect to consumer acceptance (Kalb and Davis, 1984). From the reciprocal crosses involving ST x CH, significant heterosis values over the midparent values were obtained for TSS; however, highly significant negative heterosis values over the midparent values were obtained for the cross CH x ST (Table 4). Positive heterotic effects were found in the elongated and oval shapes but not in the round shapes, indicating that CH (elongated shape) and ST (round shape) have better combinatory effect on TSS when CH is the male parent. For some crosses, the TSS values tended to always be intermediate in F1 fruit as compared to their parents (Nath and Dutta, 1971; Abadia et al., 1985). This appears to be true for the crosses CH x LI, CH x SD, CH x VE, and CH seems likely to be the female parent. The overall data obtained in the present study indicated that the TSS trait showed a wide range of variation.

Moll and Stuber (1994) stated that heterozygous genotypes could exhibit heterobeltiosis for desired genetic traits, and that this condition capitalized on the wide variable gene pool. Heterosis represents superior performance of the F1 hybrid progeny when compared to its parental lines; however, the level of expression for heterosis is highly variable. On the other hand, the average heterosis given by a three-way hybrid relies on the frequency of loci that keep a dominant allele, and this is in the function of the genotype of the single cross and the third parent. However, according to Fehr (1987), the frequency of loci with a dominant allele generally tends to be less in a three-way cross than in a single cross. Therefore, the average

that only fruit shape index had a significant heterosis. Fruit weight showed highly significant negative heterosis. The three-way F1 hybrid crosses produced an improved shape index and smaller values when compared to their parental values over fruit weight (Table 5).

Discussion

Analysis of the data for fruit weight showed that from all the crosses, including the reciprocal ones, highly significant negative heterosis values were obtained when CH was used as the female parent (Table 2). 'Chamie' possess recessive genes for fruit length and fruit diameter, and perhaps the presence of homozygous recessive loci tend to produce these lower heterotic values, as indicated by Fehr (1987). Similar data were reported with cantaloupes and muskmelons by Sambandam and Chelliah (1972). However, there is a possibility that all F1 hybrids from crosses having CH as the female parent had inherited maternally the small size character. The overall heterosis on the standard variety honeydew (HD) was shown to be highly significant (P = 0. 01) for the cross LI x CH (Table 2) and showed a better combination for this trait when CH was used as the male parent than the rest of the crosses. The heterosis over the midparent values was shown to be highly significant (P = 0. 01) for the cross SD x CH on fruit weight (Table 2). This may indicate additive gene action for the average fruit weight.

Fruit shape is defined as the ratio of fruit length to fruit diameter; however, gene action of fruit length and gene action of fruit are independent of each other. In general, muskmelons display a wide variation in shape which may be due to the nature of the genes controlling fruit length and fruit diameter (Ganesan and Sanbandam, 1979). Because of the genetic tendency of CH to

Cucurbitaceae '98


Robert MacLehose and Co. Ltd., Glasgow; Longman, New York. p. 254­263.

Fehr, W.R. 1987. Heterosis, p. 115­119. In: Principles of cultivar development. vol. 1. Theory and technique. Macmillan.

Foster, R.E. 1968. F1 hybrid muskmelon. V. Monoecious and male sterility in commercial seed production. J. Hered. 59:205­207.

Ganesan, J. and C.N. Sambandam. 1979. Inheritance of certain qualitative characters in muskmelon (Cucumis melo L. ). Annamalai Univ. Agr. Res. Ann. 9:41­44.

Ghaderi, A. and R.L. Lower. 1979. Analysis of generation means for yield in six crosses of cucumber. J. Amer. Soc. Hort. Sci. 104:562­572.

Kalb, II, I.J. and D.W. Davis. 1984. Evaluation of combining ability, heterosis and genetic variance for fruit quality characteristics in bush muskmelon. J. Amer. Soc. Hort. Sci. 109:411­415.

Knott, J.E. and O.N. Lorenz. 1950. Vegetable production. Adv. Agron. 2:134.

Lippert, L.F. and P.D. Legg. 1972. Appearance and quality characters in muskmelon fruit evaluated by a ten cultivar diallel cross. J. Amer. Soc. Hort. Sci. 97:84­87.

Lumsden, D. 1914. Mendelism of melons. Bul. N.H. Agr. Expt. Sta. 172:58.

Ma, Z.H. 1995. Heredity of main economic characteristics of the Cucumis melo. L. International symposium on cultivar improvement of horticultural crops. part I: Vegetable crops. Acta Hort. 402:57­61.

Ma, D.W., Z.H. Guo, C.H. Zhang, S.Z. Gae, and M. Wang. 1995. A study on chromosome number and karyotype of melons (Cucumis melo. L. ). International symposium on cultivar improvement of horticultural crops. Part I: vegetable crops. Acta Hort. 402:61­65.

Moll, R.H. and C.W. Stuber. 1974. Quantitative genetics empirical results relevant to plant breeding. Adv. Agron. 26:277­313.

Nath, P. and O.P. Dutta. 1971. Hybridization among muskmelon, snap melon and long melon. Indian J. Hort. 28(2):123­129.

Parthasarathy, V.A. 1978. Studies on certain melon like forms of India with special reference to their compatability. PhD thesis, Annamalai Univ., Tamil Nadu, India.

Rai, B. 1979. Heterosis breeding. Agro. Biol. Publ. New Delhi, India. p. 183.

Ramasamy, B., V.S. Seshadri, and J.C. Sharma. 1977. Inheritance of some fruit characters in muskmelon. Scientia Hort. 6(2):107­120.

Sambandam, C.N. and S. Chelliah. 1972. Scheme for the evaluation of cantaloupes and muskmelons varieties for resistance to the fruit fly (Dacus cucurbitae C. ). USDA PL480. Res. Proj. Rpt. Annamalai Univ. p. 67.

heterosis in a three-way cross will be less than in a single cross, and loci with dominant alleles will also be less, due to the presence of homozygous recessive loci in some of the progenies.

From the analyzed data, no significant heterosis over the best parent value was detected for any fruit quality trait. This further indicates the highly homozygous condition of the inbred line CH and the high frequency of recessive genes that may be found in all of the hybrid parental lines.

Highly significant favorable heterosis based upon the midparent comparison for fruit weight and fruit shape (Tables 2 and 3) indicated the presence of important nonadditive genetic effects that controlled the expression of these traits. These observations may justify the improvement of CH. Any melon breeder wishing to improve populations derived from the genetic materials utilized in this study must develop selection procedures that can effectively use both additive and nonadditive genetic variance. However, it appears that single crosses with CH would be more favorable than the three-way cross performed in this study.

Conclusion

From the results of this melon breeding research study, two conclusions can be drawn. One, heterosis above the midparent value was observed for all fruit quality traits studied, and no heterosis was found to be above the high parent value. Two, the results of this study pointed to the variability that existed and the importance of additive gene effects for the expression of fruit quality traits in melons, using hybrid lines and an inbred line.

Literature cited

Abadia, J., G.M.L. Gomez, J. Caurtero, and F. Nuez. 1985. Inheritance melon fruit characters. Cucurbit Genet. Coop. Rpt. No. 8:34­35.

Bains, M.S. and U.S. Kang. 1963. Inheritance of some flower and fruit characters in muskmelon. Indian J. Genet. 23(1):101­106.

Dyutin, K.E. and V.J. Prosvirrin. 1977. Genetic control of dry soluble substances content character in watermelon and melon. Cytol. Genet. 6:504­508.

Falconer, D.S. 1981. Introduction to quantitative genetics.

Cucurbitaceae '98