Inheritance of Orange Flesh Color in Watermelon

Cucurbit Genetics Cooperative Report 12:59-63 (article 26) 1989

Warren R. Henderson
Department of Horticultural Science, North Carolina State University, Raleigh, NC 27695-7609

A variety of flesh colors are present in watermelon: red, orange, yellow, and white. Porter (3) reported yellow flesh from ‘Golden Honey’ was recessive, y to red flesh, Y . Poole (2) showed that yellow flesh, which he termed golden yellow, was also from ‘Golden Honey’ as well as from ‘Yellow Flesh Ice Cream’, was controlled by a single recessive gene, y. Poole demonstrated that Canary yellow from ‘Honey Cream’ was controlled by a single dominant gene, C, to pink flesh, c from ‘Dove’. Shimotsuma (5) found that two pairs of genes with epistasis controlled white, yellow and red flesh color derived from 1) a cultivated form of Citrullus lanatus; 2) a bitter, wild type of C. lanatus; and 3) a non-bitter wild C. lanatus. White flesh was controlled by a single dominant gene, Wf, to yellow and red flesh; thus Wf = B and Wf = bb both gave white flesh; yellow flesh was dominant to red flesh and was expressed only when wfwf was present, thus yellow = wfwf B_, and red flesh was the double homozygous recessive,wfwf bb. An F₂ phenotype segregation ratio would thus be 12 white: 3 yellow: 1 red.

In the present investigation the inheritance of orange flesh derived form ‘Tendersweet Orange Flesh’ (‘Tendersweet OF’) was studied in crosses with red flesh from ‘Dixielee’ and ‘Sweet Princess’, and yellow flesh from ‘Golden Honey’. Following the flesh color symbols referred to earlier and those given by Robinson et. al. (4) and by Henderson et. al. (1) y will be used in this study to designate yellow flesh from ‘Golden Honey’ and Y red flesh color.

The F₁ of ‘Dixielee’ x ‘Tendersweet OF’ was red, indicating dominance of red flesh to orange flesh. The F₂ and BC₁ data support a single gene hypothesis whereby orange flesh was recessive to red flesh (Table 1). Chi-square values are all non-significant indicating a good fit to the single recessive gene hypothesis for orange flesh to red flesh. Also an orange fleshed F2 selection segregated 3:1 for red to orange flesh as would be expected in 2/3 of the red selections.

A test for heterogeneity for Chi-square goodness of fit (Table 2) showed that each family segregated in a similar direction and was similar to the pooled value in both the F₂ and backcross generations. Thus, reliability can be placed in the pooled data e.g. a deficiency in a character of one family was not cancelled by a surplus in another family.

In the test for allelism (Table 3) orange flesh was dominant to yellow flesh in the cross ‘Tendersweet OF’ x ‘Golden Honey’. Further, red flesh was dominant to yellow flesh in the cross ‘Golden Midget’ (red flesh) x ‘Golden Honey’ (yellow flesh). It is tempting to hypothesize a multiple allelic system as is shown in Table 1 and 2 whereby yy = yellow flesh, yoyo or yoy = orange flesh and Y_ = red flesh. However, a dihybrid system with epistasis has not been ruled out and awaits the F₂ and backcross data. Tentatively the symbol yo is given to orange flesh which is recessive to red flesh (Y) but dominant to yellow flesh (y).

Table 1. Segregation and Chi-square goodness of fit test for watermelon flesh color in the cross ‘Tendersweet Orange Flesh’ (orange) x ‘Dixielee’ (red).

			Flesh color Offspring (no. plants)		Expected ratio
Generation	Parents	Hypothesized parental genotype(s)z	Red	Orange	Red:Orange	Chi-Square	Probability
P1	Dixielee	YY	8	0	1:0	–	–
P2	Tendersweet OFy	y0y0	3	0	1:0	–	–
F1	Dixielee x Tendersweet OF	Yy0	12	0	1:0	–	–
F2	Dixielee x Tendersweet OF F1(self)	Yy0 (self)	21	11	3:1	1.50	.50-.75
BC1	F1 x Dixielee	Yy0 x YY	46	0	1:0	0	1.00
BC2	F1 x Tendersweet OF	Yy0 + y0y0	47	65	1:1	2.89	.05-.10
F3	F2 – red selection (self)	Y_ (self)	11	3	1:0 or 3:1	0.10	.75-.90
F3	F2 – orange selection (self)	yoyo (self)	0	10	0:1	0	1.00

^z Tentative flesh color genotypes: Y_=red – dominant to orange and yellow; yoyo or yoy=orange – recessive to red, dominant to yellow; yy=yellow – recessive to both red and orange
^y Tendersweet OF=’Tendersweet Orange Flesh’

Table 2. Heterogeneity test for Chi-square goodness of fit test for F₂ and backcross generations for watermelon flesh color in the cross, ‘Dixielee’ x ‘Tendersweet Orange Flesh’.

Generation	df	Chi-square	Probability
F2
Sum of two chi-squares	2	1.66	.25 – .50
Pooled	1	1.50	.10 – .25
Heterogeneity	1	0.16	.50 – .75
Backcross (F1 x Tendersweet OF)
Sum of four chi-squares	4	6.73	.10 – .25
Pooled	1	2.89	.05 – .10
Heterogeneity	3	3.84	.25 – .50

Table 3. Allelism tests for watermelon flesh color.^z

		Flesh Color (number of plants)
Cross	Family	Red (Y_)	Orange (yoyo or yoy)	Yellow (yy)
Tendersweet OFy (yoyo) x Golden Honey (yy)	1	0	17	0
	2	0	4	0
	Total	0	21	0
Tendersweet OF (yoyo) x Golden Midget (YY)	1	16	0	0
	2	13	0	0
	Total	29	0	0
Golden Honey (yy) x Golden Midget (YY)	1	12	0	0
Tendersweet OF (yoyo) x Sweet Princess (YY)	1	19	0	0
Golden Honey (yy) x Sweet Princess (YY)	1	9	0	0
Golden Midget (yy) x Sweet Princess (YY)	1	14	0	0

^z Yellow derived from ‘Golden Honey’, orange from ‘Tendersweet Orange Flesh’ and red flesh color from ‘Golden Midget’ and ‘Sweet Princess’.
^y Tendersweet OF = ‘Tendersweet Orange Flesh’

^x Tentative gene symbols:

• Y = red
• y^o= orange (recessive to Y dominant to y)
• y = yellow (recessive to Y and y^o)

Literature Cited

1. Henderson, W. R., M. Pitrat, R.W. Robinson, T.C. Wehner and T.W. Whitaker. 1987. Gene list for watermelon. Cucurbit Genetics Coop. Rpt. 10:106-110.
2. Poole, C.F. 1944. Genetics of cultivated cucurbits. J. Heredity 35: 122-128.
3. Porter, D.R. 1937. Inheritance of certain fruit and seed characters in watermelons. Hilgardia 10:489-568.
4. Robinson, R.W., H.W. Munger, T.W. Whitaker and G.W. Bohn. 1976. Genes of the cucurbitaceae. HortScience 11:554-568.
5. Shimotsuma, M. 1963. Cytogenetical studies in the genus Citrullus VI. Inheritance of several characters in watermelon. Jap. J. Breeding 13:2235-240.