SPRING 2003¹

D. N. Maynard² and B. J. Sidoti³

Gulf Coast Research and Education Center

University of Florida

5007 60^th Street East

Bradenton, FL 34203

Diploid (seeded) watermelons generally weigh from 18 to 35 lb and represent about 40% of the commercial crop grown in Florida. Triploid (seedless) watermelons usually weigh 15 to 22 lb and account for about 60% of the shipment from Florida. The proportion of the Florida crop devoted to triploid production is increasing each year. Icebox watermelons weigh 6 to 12 lb each and are grown on a very small acreage. Personal size or mini watermelons were introduced in 2003; these fruit weigh 3 to 7 lb. each. Florida produced 8.6 million cwt of watermelons of all types from 24,000 harvested acres in 2000-2001, which provided an average yield of 310 cwt/acre. The average price was $5.70/cwt resulting in a crop value of over $42 million which accounted for 2.5% of the gross value of the state’s vegetable crops (Fla. Agr. Stat. 2002).

The concept of triploid (seedless) watermelons was first described in the U.S. literature by Kihara (1951) based on experimentation that began in 1939 in Japan. Seed for planting seedless watermelons results from a cross between a tetraploid female parent, developed by treating diploid lines with colchicine or by other means, and a diploid (normal) male parent. The resulting triploid plants are sterile and do not produce viable seed. However, small, rudimentary seeds develop which are consumed along with the flesh just as immature seeds are eaten in cucumber.

Fruit enlargement in seeded fruit, including watermelon, is enhanced by growth-promoting hormones produced by the developing seed. Growth hormones are lacking in seedless watermelons so those agents must be provided by pollen. Since flowers of triploid plants lack sufficient viable pollen to induce normal fruit set, diploid seeded watermelons are interplanted with triploids to serve as pollenizers. Specific pollenizers have been developed by the seed industry that allow for solid plantings of triploids with the pollenizers interplanted. An adequate bee population is necessary to insure that sufficient transfer of pollen occurs. Seedless fruit (from triploid plants) tend to be triangular shaped and are prone to hollowheart without sufficient pollination.

Although the procedure for production of seedless watermelons has been known for about 50 years and commercial varieties have been available for many years, the interest in and acreage of seedless watermelons has remained small in Florida until quite recently. The deterents to seedless watermelon production have now been largely overcome so production is increasing rapidly.

Specialty vegetables are in high demand and seedless watermelons offer an attractive alternative for discriminating consumers and the food service industry. Seedless watermelons are being actively promoted by marketing organizations and seed companies to stimulate demand. At the same time, new varieties are being developed that are superior to those previously available. Seedless watermelons have been evaluated at this location annually since 1988.

The objective of this trial was to evaluate the performance of triploid watermelon cultigens under west-central Florida conditions.

Seeds of 40 triploid watermelon varieties or experimental hybrids (Table 1) were planted in a peat-lite growing mix in planter flats (1 ¼ x 1 ¼ x 2 ¼ in. cells) on 28 January. The watermelon transplants were grown by a commercial plant grower.

Soil samples from the experimental area obtained before fertilization were analyzed by the University of Florida Extension Soil Testing Laboratory (Hanlon and Devore, 1989): pH = 6.1 (target pH is 6.0) and Mehlich I extractable P = 133 ppm (very high), K = 12 ppm (low), Mg = 52 ppm (high), Ca = 725 ppm (adequate), Zn = 12.3 ppm (adequate), Cu = 5.6 ppm (adequate), and Mn = 6.1 ppm (adequate).

The EauGallie fine sand was prepared in early February by incorporation of 0-0.8-0 lb N-P₂O₅-K₂O per 100 linear bed feet (lbf). Beds were formed and fumigated with methylbromide:chloropicrin, 67:33 at 2.3 lb/100 lbf. Banded fertilizer was applied in shallow grooves on the bed shoulders at 3.1-0-4.3 lb N-P₂O₅-K₂O/100 lbf after the beds were pressed and before the black polyethylene mulch was applied. The total fertilizer applied was equivalent to 150-40-208 lb N-P₂O₅-K₂O/A. The final beds were 32-in. wide and 8-in. high, and were spaced on 9-ft centers with four beds between seepage irrigation/drainage ditches, which were on 41-ft centers.

The transplants were set in holes punched in the polyethylene mulch at 3-ft in-row spacing on 4 March. The replicated plots were 24 ft long and had eight plants each and were repeated three times in a randomized, complete block design. Diploid watermelons that were being evaluated were direct seeded in beds on each side of two triploid watermelon beds on 26 February to serve as diploid pollenizers. Plant stands recorded just before vines grew together showed no significant differences among plots. Weed control in row middles was by cultivation and applications of paraquat. Pesticides were applied as needed for control of gummy stem blight (chlorothalonil, azoxystrobin, mancozeb, and maneb), and lepidopterous larvae and silverleaf whitefly (Bacillus thuringiensis, endosulfan, bifentrin, potassium salts of fatty acids, and spinosad).

Watermelons were harvested twice during the 15-22 May and 3-9 June periods. Marketable (U.S. No.1 or better) fruit according to U.S. Standards for Grades of Watermelons (U.S. Dept. Agr., 1978) were separated from culls and counted and weighed individually. Fruit 12 lbs and larger were assumed to be marketable. Tetraploid fruit, where they occurred, were not included in the marketable category because they are not seedless. Six fruit from each entry at the first harvest were used to determine soluble solids (a measure of sweetness) with a digital, hand-held refractometer, polar and equatorial dimensions, rind thickness, flesh color, and the incidence and severity of hollowheart. Flesh firmness was determined with a hand-held Wagner FT 011 fruit tester using the 7/16 inch plunger. Two samples were taken from each fruit in the area between the seeds and the rind. Where possible, data were subjected to analysis of variance and mean separation was by Duncan’s multiple range test.

Maximum temperatures were near normal and minimum temperatures were above average in throughout the growing period. This resulted in an earlier than usual crop. Rainfall was less than normal throughout the period.

Early yields (Table 2) ranged from 49 cwt/acre for HSR 2920 to 357 cwt/acre for RWT 8145. Seventeen other entries had early yields similar to those of RWT 8145. Average fruit weight in the early harvest varied from 13.2 lbs for USS 2121 to 25.6 lbs for WX 28. Soluble solids concentrations ranged from 11.4% for ‘Dillon’ to 13.9% for USS 2330. Accordingly, soluble solids in all entries far exceeded the 10% specified for optional use in the U.S. Standards for Grades of Watermelons to describe very good internal quality (U.S. Dept. Agr., 1978). The incidence of hollowheart ranged from 0 for ‘Dillon’, ‘Gypsy’, HSR 2920, ‘Olympia’, ‘Superseedless 7177', ‘Superseedless 7187', ‘Sweet Delight’, STL 8021WM, ‘Talladega’, ‘Trillion’, Tri-X-Palomar’, Tri-X-313, XP 4510759, USS 2121 ZG-8016, and ZG-8818 to 100% in ‘Sunny’, ‘Omega’, and USS 2231. Hollowheart was especially severe in WX 28, HA 5015, USS 2330, ‘Summer Sweet 5244' and ‘Omega’. Flesh firmness varied from 1.8 lb/in² for ‘Sunny’ to 3.0 lb/in² for ‘Millionaire’, ‘Trillion’ and HSR 2920. Some processor specify a minimum of 2.0 lb/in² for acceptance as fresh cut product.

Total yields (Table 2) ranged from 195 cwt/acre for USS 2121 to 959 cwt/acre for WX 28. Twenty-seven other entries produced yields statistically similar to those of WX 28. Average fruit weight for the entire season varied from 13.7 lbs for USS 2121 to 26.5 lbs WX 28. The number of fruit per plant ranged from 0.9 for USS 2121 to 3.4 for ‘Superseedless 7177'. Cull fruit ranged from 2% for RWT 8145 to 38% for ZG 8801.

The distribution of fruit into weight classes is shown in Table 3. When triploid fruit are packed in cartons, there is an advantage in having a high proportion of fruit in the 15 to 18 lb category. ‘Gypsy’, ‘Imagination’, ‘Millionaire’, SR 8021WM, USS 2231, and ZG 8016. produced over 40% of their fruit in this weight range. Fruit are graded into two or more sizes when they are shipped in bins. Large fruit are useful for food service or as a precut product.

Seedless watermelon variety trials have been conducted at this location each spring season since 1988. The highest yields ranged from 507 cwt/acre in 1996 to 1253 cwt/A in 2002. The highest yield in 2003 was 959 cwt/acre which greatly exceeded the 877 cwt/acre average high during the entire 16-year period.

Variety shape and rind patterns and flesh color, based on observations in this trial, are shown in Table 1. Varieties producing oval to oblong fruit may be more suitable for boxing than varieties producing round melons. Generally, the striped melons are more attractive for the U.S. market than those with dark stripes on a very dark green background, or those with a solid dark green rind. Flesh color was outstanding in ‘Genesis’, RWT 8145, SR 8021WM, SSC 31829, ‘Tri-X-313’ and USS 2231.

Based on results of this and previous trials, triploid hybrids, in alphabetical order, that should perform well in Florida include Cooperstown (for trial), ‘Dillon (for trial), ‘Genesis’, ‘Millionaire’, ‘Olympia’ (for trial), ‘Revolution’, ‘Sugar Shack’ (for trial), Super Seedless 7167, 7177, 7187, ‘Summer Sweet 5244’, ‘Sweet Slice’ (for trial), ‘Tri-X 313’, ‘Tri-X Palomar’, and ‘Tri-X Shadow’. ‘Triton’, a yellow-flesh variety should be evaluated for that niche market. RWT 8145 was an outstanding experimental hybrid in the 2003 trial. Other varieties may perform well on individual farms.

The information contained in this report is a summary of experimental results. No discrimination is intended and no endorsement is implied where trade names are used.

We are grateful to the following firms for their financial and material support of vegetable variety evaluation during 2002-2003. Abbott & Cobb; Agrisales, Inc.; BHN Research; Fafard Inc.; Harris Moran Seed Co.; Hazera Quality Seeds; Hollar Seeds; Sakata Seed America; Shamrock Seed Co.; Seminis Vegetable Seeds, Inc.; Southwestern Vegetable Seed Co.; Sugar Creek Seeds, Inc.; Sunseeds; Syngenta Seeds; U.S. Seedless, LLC; Willhite Seed, Inc.; and Zeraim Gedera Ltd.

Florida Agricultural Statistics Service. 2002. Vegetables, acreage, production, value. Orlando, Fla.

Hanlon, E. A. and J. M. DeVore. 1989. IFAS extension soil testing laboratory chemical procedures and training manual. Fla. Coop. Ext. Circ. 812.

Kihara, H. 1951. Triploid watermelons. Proc. Amer. Soc. Hort. Sci. 58:217-230.

U. S. Dept. Agr. 1978. U.S. standards for grades of watermelons. AMS, Washington, D.C.

1 Florida Agricultural Experiment Station Journal Series No. T.  This report is available on the GCREC website (http:gcrec.ifas.ufl.edu).

2 Professor Emeritus

³ Biological Scientist

Back to Top

Photo Gallery

Watermelon Home Page