Cucurbit Genetics Cooperative Report 7:12-13 (article 6) 1984

Improving the Fertility of Tetraploid Cucumbers

A.P.M. den Nijs and D.L. Visser

Institute for Horticultural Plant Breeding (IVT), P.O.B. 16, 6700 AA Wageningen, The Netherlands

In our breeding program to introduce resistance genes from wild Cucumis species into C. sativus, we have attempted to overcome crossing barriers by using tetraploid cucumbers as the maternal parent in crosses with C. melo L. and C. zeyheri Sond. Interspecific pollinations, in all possible combinations, between diploid and tetraploid lines of C. sativus and C. melo yielded occasional fruits containing ovules that were sometimes enlarged but without embryos (1). Occasional embryo development up to the globular stage was obtained in diploid and tetraploid cucumbers pollinated by melon (2).

We assembled a small collection of tetraploid cucumber lines for a study of their crossability with other Cucumis species (Table 1). However, fruit set after pollinations by C. melo was disappointingly low, and self pollinations yielded only rarely fruits as well. Seed set upon selfing averaged 5 to 20 seeds per fruit (s/f) disclosing the low fertility of the tetraploids. This was also observed elsewhere (3). We decided to improve the fertility of the tetraploids before further crossability studies were carried out.

Table 1. Designation, creator and source of tetraploid breeding lines.
Code	Designation/derived from	Doubled by	Source
BDR	cv. Butchers Disease Resisting	Grimbly	GCRI, Littlehampton, UK
C16	4x/1976	Mackiewicz	Inst. of Gent., Reguly, Poland
Eng	English slicer	Grimbly	via Mackiewicz
RT	naturally occurring 4x	-	Arkansas Agr. Exp. Stn, USA
Addis	cv Addis	Lower	North Carolina S.U., USA
0918	PI 220860, from Korea	IVT	NCRPIS, Ames, Iowa, USA
1811	WJR 3147, from India	IVT	Vavilov Inst. P1. Ind., USSR

Crosses were made between the tetraploids listed in Table 1 in a half diallel to alleviate the inbreeding depression which could be responsible for the low fertility. Many crosses failed and average fruit set was under 25%. Seed set was also very low, but 'C16', 'Addis' and 'Gbn 1811' averaged about 20 seeds per fruit (s/f) after self pollination and outcrossing. Four out of six F₁ hybrids sown in 1982 germinated, and these were self pollinated and crossed with each other. Fruit set on the hybrid plants was over 90%, i.e. the same as on diploid material. Average seed set ranged from about 30 s/f for '1811 x Eng' and 'Addis x C16', to about 120 s/f for 'C16 x BDR' and 'RT x C16'. The outcrosses between the inbred tetraploid lines therefore significantly improved fertility.

One of the best hybrids, 'C16 x BDR', was studied in more detail in 1983, along with its parents, the F₂ and an outcross with 'RT x C16'. Some results are in Table 2. Poor germination resulted in only 2 plants of 'BDR', which like those of 'C16' lagged behind in plant development in comparison with the hybrid populations. Percentages of stainable pollen (means of at least 6 preparations per plant, at three dates) were uniformly high, which was also true for the fruit set. The very high fruit set on 'C16' and its reasonable seed set surprised us in view of earlier experience, although 'C16' is one of the more fertile lines. The environment must have been conducive to seed set in 1983. The very low seed set on 'BDR' and the dramatic increase to about 90 s/f on the F₁ hybrid agreed with earlier data. The average seed set on the F₂ plants was somewhat depressed which may reflect the again increased level of inbreeding. The combination with the unrelated tetraploid 'RT' appeared to have a positive effect on seed set.

Table 2. Fertility of inbred and hybrid tetraploid breeding lines.
Population	Number of plants	Percentage stain-pollen	Percentage fruit set after selfing	Total number of fruits	Seeds per fruit
P1:C16	9	73	94	16	57
P2:BDR	2	68	75	6	24
F1:Cl6 x BDR	10	72	78	30	88
F2:(Cl6 x BDR) x self	25	74	88	64	74
F1:(Cl6 x BDR) x (RT x C16)	50	70	88	118	95

Although the seed set of these hybrid tetraploids is not yet up to the diploid level, we expect that the improved fertility will enable us to more accurately compare tetraploid and diploid cucumbers with respect to the crossability with melons and interesting wild Cucumis species.

Literature cited:

1. Deakin, J.R., G.W. Bohn and T.W. Whitaker. 1971. Interspecific hybridization in Cucumis. Econ. Bot. 25:195-211.
2. Niemirowicz-Szczytt, K. and B. Kubicki. 1979. Cross fertilization between cultivated species of genera Cucumis L. and Cucurbita L. Genet. Pol. 20:117-124.
3. Smith, O.S. and R.L. Lower. 1973. Effects of induced polyploidy in cucumbers. J. Amer. Soc. Hort. Sci. 98:118-120.