History, Important Varieties

by Todd C. Wehner
Department of Horticultural Science
North Carolina State University
Raleigh, NC 27695-7609

Watermelon varieties have been described in the vegetable variety lists maintained by the American Society for Horticultural Science. Also, a complete set of descriptions for all vegetable crops from lists 1 through 25 has been collected into a book to be produced by ASHS Press. Seeds are available for many of the open pollinated and inbred varieties on the list, but there are a significant number of varieties that are no longer available. Watermelon breeders should obtain and evaluate a sample of the varieties available to become familiar with the diversity of germplasm. It is also useful to observe the improvement in horticultural traits that has been made in varieties developed over time.

A breeding program usually is started by intercrossing the best varieties currently available, or by crossing the best varieties with accessions having one or more useful traits missing from the elite varieties. Thus, in the beginning a watermelon breeder will need to obtain seeds of the best varieties, a set of varieties developed at different times in the past, a set of accessions from germplasm repositories, and lines with useful or interesting gene mutants.

A survey of popular varieties in the ten major watermelon-producing states in the United States by D.N. Maynard in 2000 indicated that popular varieties for commercial production were almost all hybrids, with few open-pollinated varieties being used commercially. Popular diploid (seeded) open-pollinated varieties (‘Allsweet’, ‘Black Diamond’, ‘Calsweet’, ‘Crimson Sweet’, ‘Jubilee II’, and ‘Legacy’) were grown mostly in one state each, suggesting regional adaptation or local demand. Hybrids generally were grown in several states, suggesting they have wider adaptation. The ‘Allsweet’ type, generally considered to be of high quality, was represented by more than half of the listed varieties (three of the open-pollinated and 11 of the hybrids). The most popular diploid (seeded) varieties were ‘Sangria’ and ‘Royal Sweet’ (seven states), ‘Fiesta’ (six states), and ‘Mardi Gras’ and ‘Regency’ (five states). For triploid (seedless) varieties, almost half of the varieties were ‘Tri-X-313’ type. The most popular triploid varieties were ‘Tri-X-313’ (ten states), ‘Summer Sweet 5244’ (nine states), ‘Millionaire’ (eight states), ‘Genesis’ (five states), and ‘Tri-X-Shadow’ (four states).

In order to develop improved varieties for an industry in a particular region of the world, the watermelon breeder will need to have seeds of varieties, breeding lines, populations, plant introduction accessions, and gene mutants that express the traits of interest at a high level. The breeder should identify a source that has the highest level of expression. That would be true whether the trait is quantitatively inherited (fruit yield, earliness, size, sweetness) or qualitatively inherited (dwarfness, anthracnose resistance, flesh color). If there is a choice of accession for a particular trait (for example, white flesh), it is better to use an adapted accession with the best genetic background. Thus, ‘Cream of Saskatchewan’ would be a better choice to use in the development of white flesh varieties for use in the United States, than a wild-type, white-fleshed citron having large vines, late maturity, hard flesh, bitter flavor, large green seeds, and seed dormancy.

Watermelons as large as 137 kg have been reported, but most cultivars produce fruit weighing 1 to 15 kg. Fruit sizes can be micro (<1.4 kg), mini (1.4-3.7 kg), icebox (3.8-6.1 kg), small, sometimes called pee-wee (6.2-8.3 kg), medium (8.4-11.1 kg), large (11.2-14.5 kg), and giant (>14.5 kg). The ice box cultivar ‘Sugar Baby’, released by M. Hardin in 1956, produces small, dark green, spherical fruit. It is popular with home gardeners, and is widely grown in Asia. Also, there are cultivars that will mature in production areas having short seasons.

Fusarium wilt resistance, obtained from Citrullus amarus (citron), was first bred into ‘Conqueror’. Resistant cultivars with improved quality were subsequently developed and became more valuable commercially. ‘Klondike R-7’, a Fusarium resistant cultivar introduced by the University of California at Davis in 1937, was popular in that state for many years.

‘Charleston Gray’, which produces fruit with elongate shape, gray (light green) rind, light red flesh, crisp texture, and sweet (10° brix) flavourful taste, was released in 1954 by the USDA. It was a leading cultivar in the USA for decades. From Kansas State University (C. V. Hall), several cultivars were important because of their high-quality fruit, along with resistance to Fusarium wilt and anthracnose: round-fruited ‘Crimson Sweet’ (released in 1963), elongate-fruited ‘Allsweet’ (released in 1973), and small-fruited ‘Petite Sweet’ (released in 1970). From the University of Florida (J. Crall and G. Elmstrom), several cultivars were important because of their high-quality fruit, along with resistance to Fusarium wilt and anthracnose: round-fruited ‘Dixielee’ (released in 1979), elongate-fruited ‘Jubilee’ (released in 1963), and small-fruited ‘Mickylee’ (released in 1986).

Important foundational cultivars of watermelon are: Charleston Gray, Klondike, Miles, and Peacock.

  • An important cross made by Charles V. Hall at Kansas State University, Manhattan Kansas was: (Miles x Peacock) x Klondike -> Allsweet, Crimson Sweet, and Calsweet
  • The parents of that cross were developed as follows:
    (((Africa 8 x Iowa Belle) x Garrison) x Garrison) x ((Hawkesbury x Leesburg) x Garrison) -> Charleston Gray
  • unknown selection in 1908 -> Klondike
  • Dixie Queen x Klondike R-7 -> Miles
  • selection from Klondike -> Peacock

Another important set of foundational cultivars of watermelon are: Africa 8, Iowa Belle, Garrison, Hawkesbury, and Leesburg.

  • An important cross made by James Crall at Florida Agr. Expt. Sta., Leesburg was: Africa 8, Iowa Belle, Garrison, Hawkesbury, Leesburg -> Jubilee
  • The parents of that cross were developed as follows:
    selection from accession collected in South Africa sent by Wagner to Layton -> Africa 8
  • Conqueror x unknown male parent (probably Kleckley Sweet) -> Iowa Belle
  • unknown selection -> Garrison
  • selection from a field of Gray Monarch with dark rather than light colored seed -> Hawkesbury
  • selection from Kleckley Sweet -> Leesburg
  • selection from Boss x Arkansas Traveler -> Kleckley Sweet

‘Charleston Gray’ and ‘Crimson Sweet’ are popular in African and Asian countries. Local selections in tropical West Africa include ‘Accra’, ‘Anokye’ and ‘Volta’. ‘Arka Jyoti’ and ‘Tarmuj’ are Indian cultivars, and ‘Zhongyu No. 1’ is a Chinese landrace.

Seedless watermelon cultivars were developed as a result of the discovery by Kihara (1951) that triploid watermelons are mostly seedless. Triploid hybrids are produced by crossing two inbred lines: a tetraploid female parent with a diploid male parent (the reciprocal cross is unsuccessful). Among the first successful cultivars (6 kg size) were ‘Tri-X-313’ from O. J. Eigsti (American SunMelon, part of Syngenta), ‘AC 5244’ from W. Barham (Barham Seeds, part of Seminis). Later, mini seedless (3 kg size) cultivars were developed. One of the first successful ones was ‘Petite Perfection’ from X. Zhang (Syngenta Seeds).

Tetraploids are produced by the application of colchicine or oryzalin, and can be maintained by self- or sib-pollination. They are used as maternal parents in crosses with diploids. The resulting triploid F1 hybrids, because of their odd chromosome number, are female and male sterile. Germinating pollen grains stimulate enlargement of the triploid ovaries. The fruit are not always entirely seedless, but may have small, empty seed coats and occasional hard seeds. Few tetraploids have been successful in producing triploid hybrids in the marketplace, either in the small or mini fruit sizes.

Because seedless cultivars are male sterile, growers must plant a pollenzier (diploid cultivar) in order to get a high rate of fruit set. Usually, one row of the diploid is planted to for every three or four rows of triploid in the production field. The diploid and triploid cultivars should be distinguishable, such as one having round fruit with narrow stripes (seedless) and the other elongate fruit with wide stripes (seeded), so that they can be separated at harvest. An adequate bee population (two active hives per hectare) is needed for successful fruit set and to minimize the percentage of cull fruit. A second method for pollenizer use is the plant the diploid in the triploid rows, with one diploid plant for every four triploid plants.

Seed production of triploid cultivars is expensive and seeds germinate poorly if conditions are less than optimum. Therefore, growers often use transplants rather than direct seeding the field. Some growers remove the seed coats before planting in order to improve germination. Production of triploid seeds and transplants, along with the need for pollenizers makes seedless cultivars significantly more expensive than seeded cultivars. However, they have become popular in North America, Europe and Japan.

In addition to seedlessness and disease resistance, breeders have selected for earliness, high yield, improved flesh characteristics such as greater sugar content, and thick or thin, but tough and flexible rind, the latter to reduce damage during shipment. Dwarf vined (bush type) cultivars have been developed and are useful for home gardeners having limited space, but the commercially important cultivars are tall (vining type). In Africa and the middle east, triploid cultivars have not yet replaced diploids in the marketplace.

Commonly used genes for cultivar development are tough rind (EE), wide stripes (gWgW), elongate (OO), oval (Oo) or round (oo) fruit shape, scarlet red (YScrYScr) or coral red (YCrlYCrl) flesh color, black (DD RR TT WW) seed color, and short seed size (ll ss). There are also genes for disease reistance, as follows. Anthracnose resistance is from Ar-1 for race 1 and Ar-2-1 for race 2 resistance. Alleles db and Fo-1 provide resistance to gummy stem blight and race 1 of Fusarium wilt, respectively. However, gummy stem blight resistance appears to be due to more than just a single gene. Susceptibility to powdery mildew is governed by pm. Resistance to Papaya ringspot virus-watermelon strain controlled by a single recessive gene, prv. A moderate level of resistance to Zucchini yellow mosaic virus was conferred by a single recessive gene zym-FL. A high level of resistance to Zucchini yellow mosaic virus-Florida strain was controlled by a single recessive gene, zym-FL-2; not the same as zym-FL. Resistance to the China strain of Zucchini yellow mosaic virus was controlled by a single recessive gene zym-CH. Finally, a single dominant allele, Zym, confers resistance to zucchini yellow mosaic virus

In China, there is a large market for confectionery (edible seeded) type, and plant breeders have selected for high yields of large seeds. They are also seeking to improve resistance in their landraces to Fusarium wilt, anthracnose, and gummy stem blight.