Pollen Competition in Melon and Effects on Seed and Fruit Vigor

Cucurbit Genetics Cooperative Report 14:63-65 (article 24) 1991

Carl D. Clayberg
Department of Horticulture, Kansas State University, Manhattan, KS 665006

Recent research in Cucurbita has indicated that pollen competition can improve seed quality (1). In the course of a breeding program to develop bush melons using the si-1 (short internode) gene (5), I encountered difficulties in seed germination of some lines, similar to that reported for the compact gene in cucumber (2). I decided to see whether seed germination could be improved by pollen parent choice. Since one of the poorly germinating lines was also white-fleshed, the experiment was designed so that I could also observe the proportion of seed set by each pollen parent, as well as the resulting seed vigor of the respective offspring.

The female parent line chosen was 87C57, and F₄ progeny from a single plant selection in the previous generation and originally derived from a cross of an advanced bush crenshaw breeding line to a commercial crenshaw cultivar. The germination of 87C57 was 22%, and the line was homozygous for si-1 and wf (white flesh). Three pollen parents were used: (1) 87C57 for self pollen” (2) ‘Charentais Improved’, si-1+/ si1+ wf+/ wf+ and 95% germination (from Stokes Seeds, Buffalo, NY); and (3) 87C54, also an F4 line in our same breeding program and from the same cross as 87C57, except that a sib selectionof the advanced bush crenshaw breeding line had been used. 87C54 also had reduced germination (43%). It was used because it was si-1 / si-1 wf+/ wf+ and its larger fruit size was expected to expressed in the F₁. ‘Charentais Improved’ was chosen because its European origin was likely to provide greater genetic differences from the female parent than American si+wf+ cultivars.

Following the suggestion of den Nijs (4), pollination partitioning was accomplished by treating each of the three stigmatic lobes per flower of the female parent differently. All pollinations were done in the field. Female flowers were emasculated in the bud stage on the day previous to pollinations. Pollinations were accomplished by peeling off the perianth of a male flower and using the remaining portion as a brush to apply pollen, transferring as much as possible. Female flowers were protectedbefore and after pollination with gelatin capsules, and benzyladenine was applied after pollination (3).

Three randomly selected plants each of ‘Chartentais Improved’ and 87C54 were chosen at the start of the experiment and used throughout. Each polination with either of these 2 parents involved the use of 3 freshly opened flowers, 1 from each of the 3 selected plants. The pollen from all 3 flowers was transferred to 1 stigmatic lobe. for the self pollinations, 3 freshly opened male flowers from the specific female parental plant were used per stigmatic lobe. In case of mixed pollinations, therefore, each female flower received the pollen from 9 flowers.

At the start of the experiment, 87C57 comprised about 30 plants, and the reduced vigor of this family means that plants could produce only 1-2 fruits each. Therefore, plants were randomly chosen to be used for each of the 5 pollination schemes indicated in Table 1, and 2 fruits were set on each plant, both of the same pollination scheme. Four plants were used for each scheme, but F₁ and F₂ generations were only grown from those plants having sufficient viable seed to conduct the experiment, since some of the plants set almost no seed or the seeds were completely inviable.

The F₁ seeds were planted in a greenhouse maintained at a minimal night temperature of 21C and later transplanted to the field. One typical open-pollinated fruit per plant was weighed, and all seeds were saved for weighing and germination tests. F2 germination tests were conducted in a growth chamber set to a constant 32.2C and 16 hr daylength. Nearly all seeds germinated in 1 week and none after 2 weeks.

Segregation date for the progeny from mixed pollinations are given in Table 2. 57-20-1 and 57-20-2 are different fruits on the same plant; a contingency test indicated that they were segregating in the same ratio (p+.03), so thge results were pooled and tested against those for plant 26. The contingency test here gave p, so pollen transmission rate differed significantly in these 2 female parents, with pollen of genotype si-1 wf+ being more frequently effective on plant 20, whereas si-1+wf+ pollen was more effective on plant 26. Although these data are limited, because only mature plant characters were used, the results indicate unequal gametic transmission for the genes concerned and that maternal parent genotype, even of full sibs within a line, altered transmission frequency. The similarity of segregation ratios for fruits 1 and 2 of plant 20 suggests that these are real differences and not merely the result of small sample size.

As shown in Table 1, pollen genotype influenced offspring vigor, as measured by seed germination and fruit weight. Data for F₁ fruit weight are averages for 5 random selected plants of that phenotype, 1 fruit per plant. The F₂ germination data are averages for 1 sample of 100 seeds per fruit for each of the 5 plants.

As expected, the larger fruit size of teh si-1 wf+ parent was reflected in the larger fruits produced by F₁ plants of that phenotype in the mixed pollinations, as wella s by the control in which that genotype was the only pollen parent, in comparison to the female parent selfed, either in the mixed pollinations of the 2 selfed controls. Total seed weights for the parental and F₁ generations were also obtained, but differences were not as pronounced as for fruit weight. For example, mean seed weight for the F₁ from mixed pollinations was 14.32 gm for si-1 wf+ pollen for 11.68 gm for selfed pollen.

When I examined seed germination, however, some of the results were unexpected. For the control in which si-1+wf+ was the pollen parent, the mean F₂ germination was 85.5%, whereas for the mixed pollination offspring from the same parent, the mean was only 53.0% (Table 1). The higher F₂ values for plant 27 do not seem to be due to plant 27 itself, becaues its own seed germinated so poorly (F₁ mean = 20.0&). Consequently, the effect here appears to be due to the ‘Charentais Improved’ pollen parent, although why this effect was not similarly expressed in the offspring of the same phenotype in the mixed pollinations is not known.

Insufficient seed production did not permit testing of these results with other plants besides 27, nor do I know whether this effect would carry over to the F₃ si-1 segregants from the crosses of 87C57-27 x ‘Charentais Improved’. These results indicate that mixed pollinations by the present method using the genes tested here do not enhance offspring vigor in si-1 melons, as measured by improved seed germination.

Table 1. Seed germinations percentage and fruit weight values (gms) for F₁ and F₂ generations for mixed pollinations and controls.

			Pollen genotype
♀ Parent (si-1 wf)			si-1+wf+		si-1 wf+		si-1 wf
Plant no.	Fruit weightz	F1Germination	F1Fruit weight	F2 Germination	F1Fruit weight	F2Germination	F1Fruit Weight	F2 Germination
Mixed pollinations		♂ ♂ : si-1+wf+-1 lobe, si-1 wf+-1 lobe, si-1 wf = (x)-1 lobe
57-20-1	1878	38	1914	50	3456	57	1533	54
53-20-2	2028	20	1706	50	4264	61	1923	58
57-26-1	2531	55	1651	59	2676	65	1987	56
Mean	2146	37.7	1757	53.0	3465	61.0	1814	56.0
♂ si-1+wf+ (1 lobe)
57-27-1	2268	16	1460	85
57-27-2	1692	24	1352	86
Mean	1980	20.0	1411	85.5
♂ si-1 wf+(1 lobe)
57-21-2	1570	5			3112	50
57-23-2	1765	6			3583	48
Mean	1668	5.5			3348	49.0
♂ (x)(1 lobe)
57-7-1	2041	30					1932	37
57-19-2	1275	19					1451	73
Mean	1658	24.5					1692	55.0
♂ (x) (3 lobes)
57-24-1	1510	11					1642	44
57-24-2	1882	16					1715	65
Mean	1696	13.5					1678	54.5

^z Mean fruit weights for male parents: si-1+wf+, 2268 gm; si-1 wf+, 4128 gm.

Table 2. Phenotypes of F₁ progeny from mixed pollination and their frequencies.

	F1phenotype (= P2 pollen genotype)
	si-1+	wf+	si-1	wf+	si-1	wf
Parent	No.	%	No.	%	No.	%
57-20-1	19	22	50	57	19	22
57-20-1	28	25	69	62	15	13
57-26-1	52	48	37	34	19	18
Expected		33		33		33

Literature Cited

1. Davis, L.E., A.G. Stephenson, and J.A. Winsor. 1987. Pollen competition improves performance and reproductive output of the common zucchini squash under field conditions. J,. Amer. Soc. Hort. Sci. 112:712-716.
2. Edwards, M.D. 1983. Evaluations of genetic control, selection procedures and responses to selection for improved emergence of compact cucumbers. PhD Diss., Univ. of Wisconsin, Madison.
3. Munger, H.M. and D.P. Lane. 1983. An improved method of BA application for the promotion of fruit set in muskmelon. cucurbit Genetics Cooperative 6:51.
4. Nijs, A.P.M. den. 1983. Genetic evidence for substantial lateral growth of pollen in the cucumber ovary. Cucurbit Genetics Cooperative 6:20.
5. Paris, H.S., H. Nerson, and Z. Karchi. 1984. Genetics of internode length in melons. J. Hered. 75:403-409.