AU2007201918B2 - Watermelon with improved processing qualities - Google Patents

Description

S&F Ref: 714708D1 AUSTRALIA PATENTS ACT 1990 COMPLETE SPECIFICATION FOR A STANDARD PATENT Name and Address Seminis Vegetable Seeds, Inc., of 2700 Camino Del Sol, of Applicant : Oxnard, California, 93030, United States of America Actual Inventor(s): Benito Juarez Fred McCuistion Greg Tolla Address for Service: Spruson & Ferguson St Martins Tower Level 35 31 Market Street Sydney NSW 2000 (CCN 3710000177) Invention Title: Watermelon with improved processing qualities The following statement is a full description of this invention, including the best method of performing it known to me/us: 5845c(773906 I) I WATERMELON WITH IMPROVED PROCESSING QUALITIES BACKGROUND OF THE INVENTION Field of the Invention [00011 The field of the present invention is watermelon breeding and the 5 genetic improvement of watermelon. More specifically, this application is related to diploid, tetraploid and triploid watermelon seeds and plants for the production of watermelon fruit that (i) have ultra firm flesh and/or liquid-retaining flesh and (ii) are sweet at maturity. Description of Related Art 10 [00021 Watermelon (Citrullus lanatus) is an important commercial member of the Cucurbitaceae family that includes many different varieties. The fruit of these varieties differ in coloring, sweetness, and other traits. For example, watermelon fruit of different varieties display a wide range of coloring on the outside rind. In addition, color in the edible tissue varies from different shades of red to yellow. Watermelon fruit also 15 vary in sweetness, which can be estimated by measuring total soluble solids, or brix, using a refractometer. Because sweetness is especially important to consumers, the U.S. Department of Agriculture has set fruit quality standards based on brix levels (United States Standards for Grades of Watermelon, U. S. Department of Agriculture (1978)). According to these standards, edible parts of the fruit having not less than 8 brix are 20 deemed to be "Good", while edible parts of the fruit having not less than 10 brix are deemed to be "Very Good." [00031 Consumers also have the choice of either seeded or seedless watermelon varieties. Unlike the flesh coloring, which is caused by varying genetic loci, the distinction between seeded and seedless varieties is usually caused by human 25 intervention of making crosses that vary ploidy levels. Similar to humans, watermelons are natural diploids with chromosomes arranged in pairs. Many plants, including watermelons, can undergo a duplication of their entire set of chromosomes and exist as tetraploids. While it is uncommon for watermelons to produce spontaneous tetraploids, this process can be routinely produced in the laboratory using cell biology techniques. A 30 tetraploid parent may then be crossed with a diploid parent to produce triploid seeds, which, in turn, generate plants with seedless fruits. In particular, seed formation in the fruit of triploid plants aborts because of the ploidy level differences, resulting in seedless fruits. Many commercial varieties are triploid and seedless.

2 [00041 Fruits of plants of different ploidy also vary in flesh firmness. Diploid lines typically have the lowest fruit flesh firmness levels. For reasons that are unclear, the process of changing a diploid line to a tetraploid line correlates with firmer fruit flesh. In other words, tetraploid lines usually have firmer fruit flesh than diploids. Triploids, being 5 a cross between a tetraploid and a diploid, typically have an intermediate level of flesh firmness. [0005] In addition to consumer preferences as to coloring, sweetness and seeds, there is increasing consumer demand in the fresh produce business for products that combine quality and convenience. Examples of products that meet these criteria are 10 bagged baby carrots, broccoli and cauliflower and bagged leafy crops, such as lettuce and spinach. [00061 Similarly, there is demand for mature cut fruits, like watermelon, melon, pineapple, papaya and kiwi. A growing segment of watermelon retail sales are cut fruits that are either displayed in large pieces with the rinds attached, or are cut into 15 smaller pieces, without the rind, and offered to the consumers in plastic food containers. The industry term for these products is "minimally processed." By 1998, Perkins-Veazie et al. ((1998) Hortscience 33:605) estimated that 10% of the retail watermelon market was minimally processed. 100071 The advantage of such cut fruit displays is that the consumer can 20 visually inspect the quality of the fruit and, in particular, judge whether the fruit is mature and, thus, ready to consume. Often, immature fruits will not be uniform in pigmentation, and overripe fruit will display signs of decay. Moreover, these products offer convenience to the consumer. 100081 The disadvantage to the produce retailer in presenting minimally 25 processed watermelon products is that cut fruits have a short shelf life. Studies indicate that minimally processed products have a shelf life of 2 to 3 days maximum (ibidem; Wehner et al. In: Watermelons: Characteristics, Production and Marketing. Maynard, editor. ASHS Press, Alexandria VA 2001.) 100091 Watermelon fruits currently available typically undergo rapid quality 30 deterioration after being cut. Cutting the fruit causes decay, which is observed as a softening of the fruit texture. Deterioration is also manifested as liquid leakage; in some varieties, the flesh of a fresh cut watermelon fruit quickly becomes unattractive to the consumer. The rapid deterioration of cut watermelon fruit places both time and space constraints on the retailer. Because cut fruits have a short shelf life, the retailer typically 3 performs the processing on the retail site. In addition, the retailer has to monitor the products often to ensure that deteriorating products are discarded. [00101 Unlike the sweetness standards established by the U.S. Department of Agriculture, there are no industry standards to describe the firmness of the edible portions 5 of watermelon fruits. Therefore, there are a wide range of descriptors in use, from "firm" and "crisp" (Erma Zaden catalog descriptors for varieties Gil 104 and Erma 12) to "very firm flesh" (Zhang et al. in USPTO application numbers 20040060085 and 20030217394 and Seminis watermelon catalog for the variety Cooperstown). Seminis has described cultivars Fenway, Royal Star and Sentinel as having "excellent crispness," "firm flesh" 10 and "crisp juicy flesh," respectively. In addition, Rogers Seed Company advertises the Tri-X Brand 626 as "exceptionally firm" and the Tri-X Brand 313 as having "firm texture" and "crispness." 100111 While advertising terminology used to describe watermelon fruit flesh firmness is quite variable, scientific reports, using quantitative measurements, show that 15 typical commercial germplasm have had substantially lower flesh firmness than the watermelon fruit of this invention. For example, Roberts et al. (2004 Report from: Watermelon Research and Development Working Group. 24 Annual Meeting, Tulsa, OK) measured flesh firmness in a wide range of germplasm, using a penetrometer to measure the amount of force resisted. The data were reported in Newtons, an 20 International System of Measurements term. For purposes of comparison with the penetrometer measurements reported herein, Roberts' data was converted to pounds force (lbf), using the following formula: I lbf= 4.448 Newtons. Roberts reports a range of watermelon flesh firmness between approximately 1.4 (6.2 Newtons) to 3.4 lbf (15.1 Newtons). One of the lines analyzed is Rogers Seed Company line Tri-X Brand 313. As 25 noted above, Rogers Seed Company advertises this line as having "firm" flesh. Roberts et al. measured the flesh firmness in Tri-X Brand 313 as 10.84 Newtons, which converts to approximately 2.4 lbf. The flesh firmness of Tri-X Brand 313 was also tested, using a penetrometer from QA Supplies in Norfolk, Virginia (Model FT01 1) with a probe diameter of 8 mm. Using this methodology, Tri-X Brand 313 has a flesh firmness reading 30 of 1.4 lbf (6.2 Newtons; Table 1). Because Roberts does not report the size of the penetrometer probe used, the present data cannot be directly compared to Roberts'. At least for Tri-X Brand 313, the approximately 77% higher reading measured by Roberts et al. compared with the protocol described herein may be the result of different methodology, and, in particular, the use of differently sized penetrometer probes.

4 Although the results reported herein were obtained using an 8 mm probe, another commonly used penetrometer has a diameter of Il mm, which would account for the different readings, as penetrometer area is approximately 73% higher for an 11 mm probe as compared to an 8 mm probe. 5 [0012] Schultheis and Thompson (2004 Report from: Watermelon Research and Development Working Group. 24 th Annual Meeting, Tulsa, OK) also survey watermelon fruit flesh firmness. Although these authors use a different model penetrometer than that used in the experiments described herein, they use a very similarly sized probe with a diameter of 5/16" or about 8 mm. Schultheis and Thompson report 10 that line Tri-X 313 had flesh firmness readings between 1.4 and 1.7 (6.2 and 7.6 Newtons), which are similar to the measurements shown in Table 1. In this report, however, the authors describe these firmness data in units of pounds/square inch. It is suspected, however, that the units provided in the Schultheis and Thompson report should be in pounds force, as a reading of 1.4 pounds/square inch, using a 5/16" (about 8 mm) 15 probe, is only 0.15 pounds force (0.67 Newtons). 100131 Maynard and Sidoti (2003 GCREC Research Report BRA-2003; Univ. Florida, Gulf Coast Research and Education Center, Bradenton, FL) report an additional survey of fruit flesh firmness of commercial watermelon lines. In this study, the authors use a different model penetrometer than that used in the method described 20 herein, with a larger sized probe having a diameter of 7/16" or about 11 mm. Their firmness data range from 1.8 to 3.0 pounds/square inch. As with the Schultheis and Thompson report, it is believed that these authors are using the incorrect units in their firmness readings. Assuming that these data are actually in pound force units, they compare well with the results obtained using the methodology described herein. For 25 example, Maynard and Sidoti's firmness measurement of line Tri-X 313 was 2.6. If one adjusts this figure to correct for the approximate 2 times difference in probe area, the new figure is 1.35 lbf (6.01 Newtons), which is nearly identical to measurement reported herein for this same line, (Table 1). On the other hand, if one assumes that the data are correctly reported in lb/square inch, the figure of 2.6 lb/square inch based on a 7/16" 30 probe would be reading of 0.39 lbf (1.73 Newtons). The Tri-X 313 line should have a much higher firmness reading than 0.39 lbf, providing further evidence of inconsistency in how such units have been reported in the prior art. [0014] Leskovar et al. ((2004) J. Horticultural Science and Biotechnology 79: 75-81) also report watermelon fruit firmness. Although this manuscript uses a different 5 measurement protocol, the authors describe in detail their methods, allowing the data to be converted for comparison with the data described herein. After converting to the same units, the range of germplasm analyzed had fruit firmness between 0.9 lbf and 1.5 lbf (4.0 and 6.7 Newtons). 5 [00151 Although measurements of the prior art can be confusing, there is clarity that commercial watermelon lines produced prior to this invention have fruit firmness that is well below 3 lbf (13.3 Newtons). In addition, as shown in Example 4, the fruit of such commercial watermelon lines, once cut, undergo significant liquid leakage. The present invention, therefore, addresses the need in the marketplace for watermelon 10 lines that produce fruits that have a longer shelf life when processed. Specifically, the watermelon of this invention have (i) ultra firm flesh, which avoids the problem of cut fruit becoming overly soft, and/or (ii) liquid-retaining flesh, which delays deterioration of cut fruit by liquid leakage. In addition, these fruits have quality characteristics desired by the consumer, such as sweetness and attractiveness, and offer the retailer both flexibility 15 as to where fruit processing occurs and additional shelf life once fruit is processed. SUMMARY OF THE INVENTION [00161 Herein disclosed are unique watermelon inbred lines and hybrid varieties that produce fruit having ultra firm edible flesh at maturity that resists at least 3.0 Pounds force (lbf - 13.3 Newtons) (measurement techniques defined herein). In 20 addition to the novel ultra firm flesh phenotype, these fruits meet market requirements for sweetness, having not less than 6 brix for the edible tissue (measurement techniques defined herein). Some watermelons disclosed herein are diploid and tetraploid inbred lines that produce sweet tasting ultra firm flesh at maturity that resists at least 3.5 lbf (15.6 Newtons). Any diploid or tetraploid inbred line having ultra firm flesh as disclosed 25 herein can transmit this ultra firm flesh phenotype to a hybrid. Also disclosed are triploid lines that produce sweet tasting ultra firm flesh and/or liquid-retaining flesh at maturity that resists at least 3.5 lbf (15.6 Newtons), though lines that produce sweet tasting ultra firm flesh at maturity that resists at least 4, 5, 6 and even 8 lbf (17.8, 22.2, 26.7 and even 35.6 Newtons) are also contemplated by this invention. Also disclosed are watermelon 30 plants that produce fruit having few to no seeds that produce sweet tasting ultra firm flesh and/or liquid-retaining flesh at maturity that resists at least 3.5 lbf (15.6 Newtons). [0017] According to an aspect, the present invention provides a watermelon plant that produces fruit having few to no seeds, said fruit having ultra firm flesh and soluble solids of at least about 6 brix wherein the ultra firm flesh resists pressure of at 6 least about 3.5 lbf. A plurality of watermelon plants grown in a field are also provided by the invention. 100181 In addition to having ultra firm flesh at maturity, the watermelons of the present invention are capable of developing uniformly pigmented fruit flesh (red, 5 yellow, or orange). In addition, at maturity, fruits from these inbred lines and hybrids will meet or exceed industry standards for sweetness, being at least good (not less than about 8 brix) and preferably very good (not less than about 10 brix). Particularly preferred watermelon fruit of the present invention exhibit a sweetness of greater than 11.5 brix. [00191 Also herein disclosed is a method for producing hybrid watermelon 10 seed comprising crossing an inbred watermelon plant with a second watermelon plant and harvesting resultant hybrid watermelon seed, as well as a hybrid watermelon plant produced by growing the resultant hybrid watermelon seed. 100201 Also herein disclosed is a method for producing the ultra firm watermelon plant comprising the steps of crossing a watermelon variety having a level of 15 sweetness that at least meets industry standards with a low sweetness watermelon variety having ultra firm flesh; performing at least one backcross with the variety having a level of sweetness that at least meet industry standards, and performing one or more cycles of self-pollination of products of the backcross (or recurrent backcross) having the combined traits of ultra firm flesh and sweetness that at least meets industry standards. The method 20 may utilize as a watermelon having ultra firm flesh the watermelon plant of USDA Collection No. P1296341. [00211 Watermelon fruit and watermelon flesh derived from the ultra-firm watermelon are also contemplated. Preferred are watermelon plants producing a fruit weighing at least about 1.5 kg, more preferably producing a fruit weighing at least about 25 3.0 kg. In a further preferred embodiment the watermelon plant produces a fruit weighing at least about 4.5 kg, and in a still further preferred embodiment the plant produces a fruit weighing at least about 6.0 kg. [00221 Also herein disclosed is a watermelon plant having the soluble solids and flesh firmness traits of a plant produced from seed deposited as Accession No. 30 NCIMB 41230, made on July 1, 2004, as well as seed, pollen, ovule and other vegetative tissue derived from the plant, or a watermelon plant regenerated from such tissue. [00231 Also herein disclosed is a watermelon plant with liquid-retaining flesh. 100241 Thus, according to another aspect of the present invention, there is 7 provided a watermelon plant that produces fruit having few to no seeds, said fruit having liquid-retaining flesh and soluble solids of at least about 6 brix. As explained in detail below, this liquid-retaining trait corresponds to the amount of weight that cut watermelon fruit flesh loses over time. Preferred are watermelon plants wherein cut flesh from the 5 watermelon fruit loses less than about three and one-half percent of its weight after three days storage at 40 centigrade. More preferred are such watermelon plants where the cut flesh loses less than about three percent weight after three days storage at 40 centigrade. A still further preferred watermelon plant is provided where the cut flesh loses less than about two percent weight after three days storage at 4" centigrade. In another preferred 10 embodiment, the watermelon plant has cut flesh that loses less than about one and one half percent weight after three days storage at 40 centigrade. This liquid-retaining trait extends the shelf life of processed watermelon fruit. 100251 Watermelon fruit and watermelon flesh derived from the liquid retaining watermelon are also contemplated. 15 [00261 Also herein disclosed is a good ultra firm flesh watermelon diploid inbred line that produces sweet tasting mature fruit. Also herein disclosed is a triploid hybrid, created using as at least one parental line that is either an ultra firm flesh diploid inbred line or an ultra firm flesh tetraploid inbred line that produces good standard sweet tasting mature fruit with ultra firm flesh. The mature watermelon fruit produced in the 20 diploid, tetraploid, or triploid plants disclosed herein may develop full red flesh color and be sweet tasting, with good brix levels. 100271 Other mature watermelon fruits disclosed herein may develop full yellow flesh color and good sweetness in combination with ultra firm flesh. Yet other mature watermelon fruits of this invention may develop full orange color and good 25 sweetness in combination with ultra firm flesh. In a preferred embodiment of the watermelon plants of the present invention, the watermelon flesh from fruits of these plants stays ultra firm after being minimally processed (fresh cut fruit). This ultra firm feature extends the shelf life of the processed fruit. 100281 Also herein disclosed is a novel method of producing diploid and 30 tetraploid watermelon lines and triploid watermelon hybrids that produce sweet tasting mature fruit with ultra firm flesh (resists pressure of at least 4.0 lbf (17.8 Newtons); not less than 8 brix). [00291 One step in this method involves crossing a known watermelon variety or line with a watermelon line of this invention having ultra firm flesh at maturity.

8 The product of such cross is then self-pollinated to create a segregating population. In successive generations, individuals from populations segregating for the ultra firm flesh trait are subjected to successive cycles of selection and breeding and the end result is a new watermelon line that produces sweet tasting mature fruit having ultra firm flesh. 5 [00301 Other objects, features and advantages of this invention will become apparent from the detailed description that follows. It should be understood that the detailed description and examples, while stating preferred embodiments of the invention, are by way of illustration only, as modifications and changes within the scope of the invention will become apparent to those skilled in the art. 10 BRIEF DESCRIPTION OF THE FIGURES [00311 Figure 1 is a histogram that illustrates fruit flesh firmness of the third generation of self-pollinated inbred watermelon plants of the present invention. The arrow indicates the average mature fruit firmness of the recurrent parent lines. The shaded portion of the histogram shows that 43% of these fruits have firmness readings at 15 or above 4 lbf (17.8 Newtons). 100321 Figure 2 is a graph showing weight loss among processed fruit of standard commercial watermelon varieties and the hybrids of the present invention. The weight loss closely approximates liquid leakage from the processed fruit. DETAILED DESCRIPTION OF THE INVENTION 20 [00331 The present invention provides a watermelon plant that produces fruit having few to no seeds with (i) ultra firm flesh and/or liquid-retaining flesh and (ii) sweetness of at least 6 brix. Therefore, the fruit of this invention have improved processing qualities, as, once cut, the fruit remains firm and/or retains its juice considerably longer than the commercial watermelon lines of the prior art. 25 Definitions [00341 As used herein, the term "plant" includes plant cells, plant protoplasts, plant cells of tissue culture from which watermelon plants can be regenerated, plant calli, plant clumps and plant cells that are intact in plants or parts of plants such as pollen, flowers, seed, leaves, stems and the like. 30 [00351 As used herein, "diploid plants" means plants or transplants derived from planting diploid seeds or from micropropagation that have two sets of chromosomes in the somatic cells, or twice the haploid number.

9 [00361 "Triploid plants" refers to plants or transplants derived from planting triploid seeds or from micropropagation that have three sets of chromosomes in the somatic cells, or three times the haploid number. [00371 "Tetraploid plants" are plants or transplants derived from planting 5 tetraploid seeds or from micropropagation that have four sets of chromosomes in the somatic cells, or four times the haploid number. 100381 The term "firm flesh" refers to the edible flesh of a watermelon for which fruit firmness, as measured using a penetrometer by the methods described in Example 2, is greater than about 1.5 lbf (6.7 Newtons) of pressure but less than or equal to to about 2.0 lbf (8.9 Newtons). Botanically, the edible flesh of a watermelon fruit is placental tissue. [00391 The descriptor "ultra firm flesh " refers to the edible flesh of a watermelon with fruit firmness, as measured using a penetrometer by the methods described in Example 2, measuring not less than 3.0 lbf(13.3 Newtons) of pressure, or 15 with higher firmness than fruit produced by standard known cultivars. Ultra-firm flesh watermelon preferably has fruit firmness of about 3.5 lbf (15.6 Newtons). 100401 The term "very firm flesh" refers to the edible flesh of a watermelon with firmness, as measured using a penetrometer by the methods described in Example 2, greater than about 2.0 pound force (8.9 Newtons) of pressure but less than 3.0 (13.3 20 Newtons). 100411 The term "liquid-retaining flesh" refers to edible flesh of a watermelon which, once cut, loses less than about four percent of its weight after three days storage at 40 centigrade, or retains more liquid, over time, than fruit produced by standard known cultivars. About 95-98% of the weight lost from cut watermelon fruit is estimated to be 25 due to liquid leakage. The majority of the remaining weight loss is from soluble solids, such as sugars and acids. Therefore, liquid loss may be approximated by measuring the percent weight loss of watermelon fruit, once cut, over time. 100421 A "penetrometer" is a device designed to measure force and is used herein to measure fruit firmness. It provides a quick, easy and accurate method to 30 determine fruit flesh and skin firmness. The data reported herein was gathered using a hand-held penetrometer to obtain three to five pressure readings on mature fruit. Specifically, Penetrometer model FTO I (QA Supplies, Norfolk, VA) was used with an 8 millimeter, or approximately 5/16 inch, probe. 100431 "Pounds force", or "lbf', is the unit read by the penetrometer model 10 FTOI 1, and is used herein exclusively to indicate readings made using the 8 millimeter probe, unless otherwise indicated. [00441 Coloration of the rind in watermelons, also referred to as "rind pattern", can vary from light green, often termed gray, to medium green, to very dark 5 green, appearing to be almost black. In addition, the rind may have stripes of various designs which are typical of a variety or type. Therefore, the terms "tiger stripe", "mottle stripe", "dark mottle stripe", and the like, are used to identify various patterns. 100451 As used herein, "length to width ratio (L/W ratio)" means the ratios obtained in any of the possible combinations by taking the average length divided by the 10 average width on the watermelon fruit. The ratios can vary from 1:1.2 to 2.2:1. [00461 The term "population" refers to genetically heterogeneous collection of plants sharing a common parental derivation. [00471 As used herein, the term "variety" or "cultivar" refers to a group of similar plants that, by their genetic pedigrees and performance, can be identified from 15 other varieties within the same species. [0048] "Backcrossing" refers to the process in which a breeder crosses a plant with one of its parent lines. [00491 "Recurrent backcrossing" is a breeding strategy designed to recover the genetic composition of a line by crossing a plant in succession back to one of the 20 parent lines. [00501 The term "soluble solids" refers to the percent of solid material found in edible fruit. As used herein, soluble solids are measured quantitatively with a refractometer as percentage brix. Refractometers often include a sucrose scale, as brix is formally defined as weight percent sucrose. If the only soluble solid present in an 25 aqueous solution is sucrose, the sucrose scale should give the actual percentage sucrose. However, if other soluble solids are present, as is almost always the case, the reading is not equal to the percentage sucrose, but approximates the overall percentage soluble solids in the sample. In short, although brix is technically defined as weight percent sucrose, those of skill in the art recognize that weight percent soluble solids, as obtained 30 with a refractometer, approximates weight percent sucrose and accurately indicates sweetness. Therefore, the higher the percentage soluble solids, as indicated by brix level, the higher the perceived sweetness of the fruit. [00511 The U.S. Department of Agriculture has established watermelon fruit quality standards based on brix levels (United States Standards for Grades of 1 Watermelon, U. S. Department of Agriculture (1978)). According to these standards and as used herein, edible parts of the fruit having not less than 8 brix are referred to as "good", while edible parts of the fruit having not less than 10 brix are referred to as "very good." 5 100521 "Sweetness", as used herein, may be measured quantitatively, as described above, using a refractometer, or qualitatively, by taste. [00531 A "quantitative trait loci", or "QTL" is a chromosomal location that encodes for alleles that affect the expressivity of a continuously distributed phenotype. 100541 "Maturity" refers to maturity of fruit development and indicates the 10 optimal time for harvest. Generally, growers of skill in the art harvest fruit at or substantially near its maximum sweetness and flavor intensity. In watermelon, the maturity comes associated with changes in rind appearance, flesh color and sugar content. [00551 The terms "homozygous" and "homozygosity" are genetic terms. When identical alleles reside at corresponding loci on homologous chromosomes, that 15 locus is called homozygous. Homozygosity typically refers to the degree to which a population is fixed at one or more loci. [00561 A "hybrid" is an offspring of a cross between two genetically unlike individuals. [00571 An "inbred" or "inbred line" is a substantially homozygous individual 20 or variety. [00581 "Introgress" is the process a breeder performs to introduce a new trait, usually from a non-cultivated type, into a cultivated type. Typical Characteristics of Commercial Watermelon Fruit [00591 Successful watermelon production depends on attention to various 25 cultural practices. These include soil management, with special attention to proper fertilization; crop establishment, with appropriate spacing; weed control; the introduction of bees for pollination and, if producing fruit from triploid plants, a suitable pollen source for producing seedless (triploid) watermelon; irrigation; and pest management. Watermelon fruit size and shape, rind color, thickness and toughness, seed size, color and 30 number, flesh color, texture, sugar content and freedom from fruit defects are all important characteristics to be considered in selection of watermelon varieties. Commercial seed companies typically offer the grower an opportunity to observe these criteria in demonstration plots of their varieties, and some agricultural universities perform cultivar analysis data for the local growers (Roberts et al. (2004), Maynard and 12 Sidoti (2003), Schultheis and Thompson (2004) and Leskovar et al. (2004)). [00601 Watermelon crops can be established from seed or from transplants. Transplanting is becoming more common because transplanting can result in an earlier crop compared with a crop produced from direct seeding. When a grower wants to raise a 5 seedless fruited crop, transplanting is preferred. Transplanting helps achieve complete plant stands rapidly, especially where higher seed costs, as with triploid seeds, make direct-seeding risky. [00611 Watermelon is the only economically important cucurbit with pinnatifid (lobed) leaves; all of the other species have whole (nonlobed) leaves. 10 Watermelon growth habit is a trailing vine. The stems are thin, hairy, angular, grooved, and have branched tendrils at each node. The stems are highly branched and up to 30 feet long (Wehner et al. In: Watermelons: Characteristics, Production and Marketing. Maynard, editor. ASHS Press, Alexandria VA 2001). [00621 Watermelon breeders are challenged with anticipating changes in 15 growing conditions, new pathogen pressure, and changing consumer preferences. With these projections, a breeder will attempt to create new cultivars that will fit the developing needs of growers, shippers, retailers, and consumers. Thus the breeder is challenged to combine in a single genotype as many favorable attributes as possible for good growing, distribution, and eating. 20 100631 One important characteristic for the breeder is fruit size. Fruit size is an important consideration because there are different market requirements for particular groups of shippers and consumers. The general categories are: icebox (<12 lb; < about 5.4 kg), small (12-18 lb; about 5.4-8.2 kg), medium (18-24 lb; about 8.2-10.9kg), large (24-32 lb; about 10.9-14.5kg), and giant (>32 lb; > about 14.5kg). Fruit size is inherited 25 in polygenic fashion, with an estimated 25 genes involved. Fruit is distributed from the grower to the retailer by shippers, who focus on particular weight categories, such as 18 24 lb (about 8.2-10.9kg) for seeded and 14-18 lb (about 6.4-8.2kg) for seedless. Although historic consumption has been for these general categories of sizes, there is an increasing trend in the marketplace for a new class of small-fruited watermelon hybrids (with fruit 30 weight between 3-9 lb; about 1.4-4.lkg). [00641 Fruit flesh firmness and liquid retention are other important characteristics. Consumers have varying textural preferences for watermelon fruit, and flesh firmness is a determinant of texture. Additionally, fruit firmness is a critical parameter that determines how long cut fruit will last on the retailer's shelf. Liquid 13 retention is also critical to consumer perception of minimally processed watermelon. Cut fruit shelf life research is usually qualitative, with evaluations on when the fruits become 'slimy' (Perkins-Veazie et al. 1998 HortScience 33: 605). Quantitative evaluations of cut fruit shelf life include measuring the flesh firmness directly, using a penetrometer, or 5 measuring percent weight loss of cut fruit over time in order to approximate liquid leakage, as described in Example 4. [00651 The firmness of various fruit was also determined simply by eating them. Indeed, this was how the watermelons of this invention were first determined to have ultra firm flesh compared to prior art watermelons. In taste tests, standard cultivars 10 of the prior art, such as Seminis' diploid Royal Star line, were determined to have firm flesh, while the following lines were determined to have firm to very firm flesh: Tri-X Brand 626 (Syngenta/Rogers - triploid), Extazy (Hazera - triploid) and Solitaire (Golden Valley - triploid). [00661 Another important internal fruit characteristic is quality, which 15 includes sweetness and attractiveness of fruit and rind color. Wehner et al. ((2001) in: Watermelons: Characteristics, production and marketing. Maynard, editor. ASHS Press. Alexandria, VA) describe these characteristics. Among the most important of these characteristics is sweetness, without a bitter taste, which is measured by brix and by taste. Taste panel data demonstrated a direct correlation of good flavor scores with higher brix 20 levels (Nip et al. (1968) Proc. Amer. Soc. Hort. Sci. 93:547). Brix levels increase as the fruit develops and ripens on the vine. Thus, immature fruits will have unacceptably low sweetness to the consumer; if picked too early, the edible tissue will also not have uniform color. Quantitative recommendations for watermelon fruits have been published. While Wehner et al. suggest brix levels between 10% and 14% brix, the United States 25 Department of Agriculture has established standards, as described in detail in the "Definitions" section, in which sweetness of at least 8 brix is good and sweetness of at least 10 brix is very good. Despite some variation in the recommendation and the standards, there is no dispute that fruit sweetness is a critical characteristic of watermelon fruit. 30 Characteristics of Watermelon Fruit of the Present Invention Fruit Firmness [00671 The flesh of watermelon plant fruits of the present invention is firmer and retains liquid better than the fruit flesh of watermelon cultivars of the prior art. In 14 prior art watermelon fruit, mature edible flesh from diploid genotypes are softer than both triploid and tetraploid genotypes. Fruit firmness variation within a line, irrespective of ploidy level, is insignificant. In general, standard diploid cultivars produce fruits with soft to at best firm flesh (i.e., flesh firmness at maturity from less than 1.0 lbf to about 1.5 5 lbf; about 4.5 to about 6.7 Newtons). Standard tetraploid lines typically produce fruit with firm flesh or very firm flesh (i.e., flesh firmness between 1.5 lbf - about 6.7kg - to less than about 3.0 lbf - about 13.3 Newtons - at maturity). Standard triploid hybrids produce seedless fruit with an intermediate level of flesh firmness at maturity, ranging from about 1.3 lbf to 2.5 lbf (about 5.8 to 11.1 Newtons). Table I shows flesh firmness 10 data from the prior art for commercial hybrids and inbred watermelon lines. 100681 All firmness measurements herein were made using a model FTOI I penetrometer from QA Supplies in Norfolk, Virginia with an 8 millimeter probe diameter. Readings were made and are reported in pounds force, a British Engineering measurement for pressure, which is abbreviated lbf and is converted to Newtons 15 according to the following formula: I lbf= 4.448 Newtons. Subject fruits were cut equatorially, midway between the blossom and stem ends of each fruit. Three to five readings were made per fruit, taking samples from the center of each cut fruit. Reported firmness data is an average of these three to five readings. Table 1. Survey of firmness in typical watermelon cultivars and inbred lines. Average 20 firmness readings are in pound force by methodology described herein. Line Origin Ploidy Firmness (lbf) Tri-X 313 Syngenta/Rogers Triploid 1.4 Millionaire Harris Moran Triploid 1.8 Revolution SunSeeds Triploid 1.7 Majestic Seminis Triploid 1.7 Olympia Seminis Triploid 1.6 Omega Seminis Triploid 1.5 PSI 10-5288-9 Seminis Triploid 2.3 4082 Seminis Tetraploid 2.0 4084 Seminis Tetraploid 1.5 4090 Seminis Tetraploid 1.6 4133 Seminis Tetraploid 2.2 4134 Seminis Tetraploid 2.4 4135 Seminis Tetraploid 2.2 15 Line Origin Ploidy Firmness (Ibf) 4137 Seminis Tetraploid 2.7 4138 Seminis Tetraploid 2.2 47602A Seminis Diploid 1.5 4203 Seminis Diploid 1.4 Cooperstown Seminis Triploid 1.5 Fenway Seminis Triploid 2.1 Sentinel Seminis Diploid 1.4 W-1 128 Seminis Diploid 1.4 W-1 119 Seminis Diploid 1.6 BSI 2532 Seminis Diploid 1.7 BSI 2527 Seminis Diploid 1.3 W-2068 Seminis Diploid 1.1 W-2741 Seminis Diploid 1.3 W-1488 Seminis Diploid 1.7 BSI 2543 Seminis Diploid 1.2 [00691 Compared to prior art watermelon lines, the fruit of the present invention both have ultra firm flesh and are sweet. Table 2 displays flesh firmness and sugar content from watermelon line P1296341, which was used as the source of the novel firm flesh fruit of this invention, and hybrid lines created according to the methods 5 described herein. Sweetness measurements were determined quantitatively, using a refractometer (Leica Microsystems Model AR200, Reichert Inc., Depew, NY), according to manufacturer's instructions. One measurement was taken from each half of an equatorially cut fruit. The data were recorded as an average. {0070] As indicated by comparing the firmness readings in Table 2 to those in 10 Table 1, the flesh of the watermelon fruit of the present invention is considerably more firm than the flesh of the watermelon fruit of the prior art. Specifically, watermelon fruit of the present invention resist pressure of at least about 3.0 lbf (about 13.3 Newtons), preferably at least about 3.5 lbf (about 15.6 Newtons), more preferably at least about 4 lbf (about 17.8 Newtons) and most preferably at least about 5 lbf (about 22.2 Newtons). 15 [0071] In addition, as shown in Table 2, watermelon fruit of the present invention are sweet. Specifically, watermelon fruit of this invention display sweetness of at least about 6 brix, more preferably at least about 8 brix and most preferably at least about 10 brix. Particularly preferred watermelon fruit of the present invention exhibit a sweetness of greater than 11.5 brix.

16 Table 2. Firmness and sugar content of inbred and hybrid lines developed from the invention described herein and the P1296341 source. Firmness readings are in pound force and sugar content is reported as %Brix. Both measurement methods are described herein. Line Origin Ploidy Firmness (lbf) Sugar content (Brix) P1296341 USDA collection Diploid 13.5 1.6 7132 This invention Triploid 4.7 10.2 7133 This invention Triploid 6.2 11.7 4201 This invention Diploid 8.0 9.7 4203 This invention Diploid 7.8 10.8 4204 This invention Diploid 6.5 9.7 4207 This invention Diploid 6.5 10 Liquid Retention 5 [00721 The fruit of the present invention also retain liquid better than the fruit of the prior art. Example 4 describes a study that demonstrates this liquid-retaining trait. The study compares liquid leakage rates of cut fruit from watermelon of this invention and of the prior art when stored at 40 centigrade. The results of this study are illustrated in Figure 2. The study measures percent weight loss over time of cut fruit. This 10 measurement approximates liquid loss, as 95-98% of the weight loss is due to liquid leakage. The remaining weight loss is due to leakage of other components of the fruit, such as soluble solids and acids. The primary conclusion from these data is that watermelon fruit of the present invention lose less than about four percent weight after three days storage at 4' centigrade. Preferably, the fruit of the present invention lose less 15 than about three and one-half percent weight after three days storage at 4' centigrade, more preferably less than about three percent weight, even more preferably less than about two percent weight, and most preferably less than about one and one-half percent weight. 100731 In addition to having liquid-retaining flesh, the fruit of the present 20 invention are sweet. Specifically, these watermelon fruit display sweetness at least about 6 brix, more preferably at least about 8 brix and most preferably at least about 10 brix. Particularly preferred watermelon fruit of the present invention exhibit a sweetness of greater than 11.5 brix. Other traits 25 100741 Watermelon plants of this invention may be seeded or seedless.

17 Methods for obtaining diploid, triploid and tetraploid plants are well known in the art. Specifically, methods for obtaining diploid and triploid watermelon plants and seed of the present invention are described in detail below. Tetraploid plants of the present invention may be easily obtained by those of ordinary skill in the art using known cell biology 5 techniques and the diploid plants described below. [00751 Using standard crossing techniques, those of skill in the art may obtain watermelon fruit of the present invention with desirable traits besides those described above, as the ultra firm flesh and liquid-retaining flesh traits are dominantly inherited. For example, breeders may easily obtain watermelons of the present invention that are of 10 a particular size or have a particular flesh color or rind pattern. Breeding Techniques - Inbred and Hybrid Lines 100761 Watermelon lines of the present invention were developed in the United States (Georgia, Florida and California), Mexico and Guatemala beginning in the year 2000. Furthermore, watermelon lines were grown for field performance and 15 evaluation of adaptation in Florida, Georgia and California beginning in the year 2003. Additionally, diploid and triploid watermelon hybrids made with lines that produce watermelons having ultra firm flesh and/or liquid-retaining flesh at maturity were evaluated in field conditions in Florida, California and Mexico in 2003 and 2004. Specific crosses and firmness and quality evaluations of resultant fruits are described in 20 detail in the "Examples" section. (00771 For most breeding objectives, commercial breeders work with germplasm that is often referred to as the 'cultivated type'. This germplasm is easier to breed with because it generally performs well when evaluated for horticultural performance. The performance advantage the cultivated types provide is sometimes 25 offset by a lack of allelic diversity. This is the tradeoff a breeder accepts when working with cultivated germplasm - better overall performance, but a lack of allelic diversity. Breeders generally accept this tradeoff because progress is faster when working with cultivated material than when breeding with genetically diverse sources. [00781 In contrast, when a breeder makes either wide intra-specific crosses, 30 or inter-specific crosses, a converse tradeoff occurs. In these examples, a breeder typically crosses cultivated germplasm with a non-cultivated type. In these crosses, the breeder can gain access to novel alleles from the non-cultivated type but has to overcome the genetic drag associated with the donor parent. Because of the difficulty with this breeding strategy, this approach often fails because of fertility or fecundity problems.

18 The difficulty with this breeding approach extends to many crops, and is exemplified with an important disease resistant phenotype that was first described in tomato in 1944 (Smith, Proc. Am. Soc. Hort. Sci. 44:413-416). In this cross, a nematode disease resistance was transferred from L. peruvianum (PI128657) into a cultivated tomato. 5 Despite intensive breeding, it was not until the mid-1970s before breeders could overcome the genetic drag and release successful lines carrying this trait. Indeed, even today, tomato breeders deliver this disease resistance gene to a hybrid variety from only one parent. This allows the remaining genetic drag to be masked. The inventiveness of succeeding in this breeding approach has been recognized by the USPTO (US Patents 10 6,414,226, 6,096,944, 5,866,764, and 6,639,132). [00791 In watermelon, the plant introduction (PI) accessions are typically lines that produce small fruits with firm white flesh and very poor taste (even bitter). Even though these lines have such poor horticultural qualities, some watermelon breeders, like some other crop breeders, attempt to breed with these PI lines because they 15 potentially contain novel alleles. To date, the most commonly attempted breeding objective for use of the PI line series is to introgress new disease resistance genes. The process of introgressing novel resistance genes from the PI lines into acceptable commercial types is a long and often arduous process. This process can be difficult because the trait may be polygenic, have low heritability, have linkage drag or some 20 combination of the three. [00801 This breeding project began with a wide cross between cultivated watermelons and PI No. 296341, which was obtained from the USDA collection at the Regional Plant Introduction Station in Griffin, Georgia. This accession has been available to watermelon breeders since its deposit into the U.S. Plant Introduction system 25 in 1964. [00811 The original intent of the project, however, was not to make watermelon fruit with firm flesh and/or liquid-retaining flesh. Rather, the original intent of the project was to introgress a resistance to Fusarium wilt, specifically to Fusarium oxysporumf sp. niveum race 2, referred to herein as FON race 2. Although no 30 commercial watermelons currently contain resistance to FON race 2, the possibility of using P1296341 as a source of resistance has been known for many years (Netzer (1989) Plant Disease 73:518; Martyn and Netzer (1991) HortScience 26:429-432; Wehner et al. ((2001) in: Watermelons: Characteristics, production and marketing. Maynard, editor. ASHS Press. Alexandria, VA). That there are no watermelon commercial lines for sale 19 with FON race 2 resistance introgressed from P1296341, despite these reports as long as 15 years ago, underscores the difficulty of introgressing traits from wide crosses and creating commercially successful inbreds and hybrids. 100821 In addition to being resistant to FON race 2, P1296341 is characterized 5 by having very small round fruits between 4 and 6 inches in diameter and weighing between I and 2.6 pounds (between about 0.45 and 1.2kg). Fruit flesh is white and very firm, with low soluble solids content (Table 2). Organoleptic evaluations of these fruits range from no perception of sweetness to bitter. 100831 As described in the "Examples" section below, inbred watermelon 10 plants of the present invention may be obtained by crossing a watermelon with the ultra firm flesh trait and/or liquid-retaining flesh trait (ultra firm parent) with a non-ultra firm flesh watermelon with other desirable quality characteristics, including sweetness (recurrent parent). The ultra firm parent may be plant introduction accession number 296341. 100841 Those of skill in the art will be able to introgress the ultra firm flesh 15 trait and/or the liquid-retaining trait into the recurrent parent by conducting various recurrent backcrosses, selecting for the (i) ultra firm flesh and/or liquid-retaining flesh trait and (ii) the sweetness trait, and finally self-pollinating selected plants of the recurrent backcrosses to create inbred watermelon lines with the above traits. One possible method for accomplishing such introgression is described in the "Examples" section below. 20 [00851 Inbred line 3347 was generated, which generates sweet ultra firm fruit according to the present invention, using the methods described above and in the "Examples" section. See, especially, Example 5. Inbred line 3347 has been deposited with NCIMB and accorded Accession No. NC[MB 41230. Details of the deposit follow the "Examples" section. 25 [00861 Using known methods, breeders may obtain diploid, triploid and tetraploid inbred lines of watermelon having fruit with the (i) ultra firm flesh and/or liquid-retaining flesh trait and (ii) sweetness trait. 100871 In addition, because the ultra firm flesh and liquid-retaining traits of the present invention are dominantly inherited, breeders may obtain hybrids using the 30 watermelons of this invention. Hybrids may be either diploid or triploid. Specifically, breeders crossed inbred watermelon plants with the above desired flesh traits and sweetness traits to either diploid or tetraploid non-ultra firm flesh cultivars to create, respectively, diploid and triploid watermelon plants with fruit having the ultra firm flesh and/or liquid retaining flesh trait and sweetness trait. The non-ultra firm flesh parent used in creating a 20 hybrid may also be used to obtain sweet ultra firm flesh and/or liquid-retaining flesh watermelon with other desirable traits, such as a particular size and/or color. 100881 Those skilled in the art recognize that there are several breeding methods used for the introgression of new traits into commercial germplasm, including 5 mass selection, pedigree selection, recurrent selection and backcrossing. By way of example, and by no means limiting, the introgression of ultra firm flesh watermelon fruit at maturity, with high brix levels is described below. Examples Example I - Generation of Fl Lines and Backcrosses 0 o[00891 In the summer of 2000, four first filial (Fl) generation lines were created by crossing 4 Seminis inbred lines as females to P1296341. The four diploid inbred lines used were W-2388, W-1 128, W-1 119 and W-1488. Line W-2388 is elongated in shape with a length to width (L/W) ratio of 1.8 to 2.2 : 1. The rind color and pattern is of medium green background with wide darker stripes. This shape and rind 15 pattern phenotype is known to those skilled in the art as an "elongated dark mottle stripe" watermelon fruit. The fruit shape of Line W- 1128 is round oval with L/W ratio of 1.0 1.2: 1 and rind color is of light to medium green background and narrow darker green stripes. This phenotype is known to those skilled in the art as "round-oval with narrow (or tiger) stripes" watermelon fruit. Fruit shape of Line W-1 119 is oval to high round 20 with L/W ratio of 1.1-1.3 : 1. Rind color is medium green background with wide darker green stripes. This phenotype is known to those skilled in the art as "round-oval dark mottle stripe" watermelon fruit. Fruit of Line W-1488 is of round shape with L/W ratio of 1.0 to 1.1 : 1. Rind color is light green with some faint mottle/ net pattern in the background. This phenotype is known to those skilled in the art as "round gray (or light 25 green)" watermelon fruit. These four lines provide an array of phenotypic diversity amongst the cultivated types. [0090] In the fall of 2000, the respective FIs were used as females to backcross to the above four inbreds, creating the backcross I (BC I) generation. [00911 The BCI generation plants were grown in the spring of 2001, and 30 selections were made based on overall vigor. It was difficult to take the alleles from the PI line into the cultivated types because many of the BCI and even BC2 plants died. Variation in vine vigor was observed that was associated with survivability. Vine vigor was assumed to be associated with general vigor, and perhaps with pathogen resistance.

21 [0092] The respective BCl lines, derived from the original four inbreds were crossed as females as follows: 1. [[W-1128 x PI296341]F1 x W-1128](this is the W-1128 BC1) X W - 1128 2. [[W-1119 x PI296341]F1 x W-1119](this is the W-1119 BC1) X W - 1119 5 3. [[W-1488 x P1296341]FI x W-1488](this is the W-1488 BCl) X W - 1488 4. [[W-2388 x P1296341]F1 x W-2388](this is the W-2388 BCl) X W - 2068 5. [[W-2388 x P1296341]FI x W-2388](this is the W-2388 BCl) X BSI - 2543 6. [[W-2388 x P1296341]F1 x W-2388](this is the W-2388 BCl) X BSI - 2527 100931 In these six crosses, the first three were recurrent parent backcrosses. 10 Cross number four was to line W-2068, which is very similar to original parent W-2388. Crosses five and six were to new inbreds. The recurrent backcross program aims to add one or more new traits from the donor parent (in this case, P1296431), while retaining the phenotype of the recurrent parent. However, watermelon breeding is a dynamic process, so it is not uncommon to change the recurrent parent as newer inbred lines are being 15 developed concurrently. Crosses four through six, therefore, were not technically creating the BC2 generation. For clarity in describing the generations, these crosses will be referred to as the BC2* generation. 100941 The BC2 and BC2* generation were grown in the summer of 2001. As with the BCI generation, selection for vine vigor was made. Females thus selected 20 were used to create the BC3 and BC3* generation. 1. W-1128 BC2 X W-1128 = BC3 2. W-1119BC2 XW-2741= BC3* 3. W-1488 BC2 X W-1488= BC3 4. W-2068 BC2* X W-2068 = BC3* 25 5. BSI-2543 BC2* X BSI-2543 = BC3* 6. BSI-2527 BC2* X BSI-2527 = BC3* In the fall of 2001, the BC3 and BC3* lines were grown, and selection was again applied for vine vigor. Selected plants were then crossed to create the BC4 and BC4* generations, respectively. 30 7. W-1128 BC3 X W-1128 = BC4 8. W-2741 BC3* X W-2741 = BC4* 22 9. W-1488 BC3 X W-1488 = BC4 10. W-2068 BC3* X W-2068 = BC4* 11. BSI-2543 BC3* X BSI-2543 = BC4* 12. BSI-2527 BC3* X BSI-2527 = BC4* 5 [00951 In addition to selecting for vine vigor, examination of BC3 and BC3* fruit, which contained the BC4 and BC4* generation seed, produced an unexpected finding. Although the BC3 generation still performs poorly when evaluated by current horticultural characteristics, the fruits were examined for quality characteristics. Although most fruit had poor quality, breeder observations as to a small number of fruit 10 included the following: "good fruit color, sweet taste and ultra firm flesh - like an apple." The unexpected finding was that both ultra firm flesh and sweet tasting flesh could be created. The possibility of creating a sweet tasting flesh, combined with ultra firm flesh for the cut fruit segment of the marketplace resulted in a bifurcation of breeding objectives. A new project was initiated with the goal of creating ultra firm flesh 15 watermelon fruits with sweet taste. The original goal of introgressing FON race 2 resistance remained as a separate, and, temporarily, secondary goal. Example 2: Self-Pollinations of Plants Bearing Ultra Firm Flesh Fruit and Early Flesh Firmness Data 100961 In the spring of 2002, the BC4/BC4* generation was grown and 20 evaluated qualitatively for sweet taste, fruit flesh firmness, and horticultural characteristics. Based on these evaluation criteria, plants were selected to create the next generation. Instead of creating another backcross generation, however, each selection from the lines being developed in parallel was self pollinated. The crossing produced the BC4SI / BC4*SI generation. 25 100971 In the summer of 2002, the BC4Sl/BC4*Sl generation was grown and evaluated qualitatively for sweet taste, fruit flesh firmness, and horticultural characteristics. Based on these evaluation criteria, plants were selected to create the next generation. Self pollination of the selected plants created the BC4S2/BC4*S2 generation. 100981 In the fall of 2002, the BC4S2/BC4S*2 generation was grown and 30 evaluated qualitatively for sweet taste, fruit flesh firmness, and horticultural characteristics. Based on these evaluation criteria, plants were selected to create the next generation. Self pollination of the selected plants created the BC4S3/BC4S*3 generation. [00991 In the spring of 2003 the BC4S3/BC4S*3 generation was grown and 23 evaluated qualitatively for sweet taste, fruit flesh firmness, and horticultural characteristics. Based on these evaluation criteria, plants were selected to create the next generation. Self pollination of the selected plants created the BC4S4/BC4S*4 generation. [001001 For the BC4S3 fruit, both qualitative and quantitative data were 5 obtained for flesh firmness. Specifically, ninety three fruits from individual BC4S3 plants were evaluated for firmness with a penetrometer (model FTO I1 with an 8 millimeter probe, QA Supplies, Norfolk, VA). The FTO1 1 penetrometer has a gauge that reads PF, which is an improper abbreviation for pound force. Pound force is a British Engineering measurement scale for pressure, and is properly abbreviated lbf. The 10 conversion from the British measurement system to the International System of Units (SI) is 1 lbf = 4.448 Newtons. For all flesh firmness measurements using a penetrometer, mature fruits were detached from the plant and cut in an equatorial direction. For orientation, fruits have a stem end and a blossom end. Equatorial slicing means that the fruits are halved such that each half has the blossom end or stem end the farthest distance 15 from the cut site. Samples were taken from the center of the cut fruit. For diploid fruits, sampling occurred inside the seeded ring. Although triploid fruits have few to no seeds, sampling occurred within the same core area of the split fruit. Each half was sampled with the penetrometer, with a total of three to five readings per fruit. Firmness data are reported as an average of the three to five readings. 20 1001011 Even after several generations attempting to fix the firm flesh genotype combined with acceptable horticultural characteristics, including sweetness, Figure 1 shows that significant fruit flesh firmness variation still existed in these samples. Although the data in Figure 1 indicate significant variation, it was clear that improvements to fruit firmness had been made. The arrow shows the average firmness rating of the recurrent 25 parents. Even at this early generation in product development, approximately 43% of the fruits have firmness measurements of not less than 4 lbf (about 17.8 Newtons). [001021 Some phenotypes are determined by the genotype at one locus. These simple traits, like those studied by Mendel, fall in discontinuous categories such as green or yellow seeds. Most variation observed in nature, however, is continuous, like yield in 30 field corn, or human blood pressure. Figure 1 shows a continuous-type pattern of firm flesh variation, similar to a normal distribution. Unlike simply inherited traits, continuous variation can be the result of polygenic inheritance. Loci that affect continuous variation are referred to as quantitative trait loci, or QTLs. Variation in the phenotype of a quantitative trait is the result of the allelic composition at the QTLs and an environmental 24 effect. Several potential causes for the variation were identified: (1) the fruit firmness trait may be controlled by several to many QTLs; (2) the fruit firmness trait may be caused by one or a few genes, but have a low heritability; and (3) the trait may be both polygenic and have low heritability. Those skilled in the art recognize that the 5 marketplace requires product uniformity. Thus, the utility of the invention is higher for those traits with high heritability that are not greatly affected by the environment. The heritability of a trait is the proportion of the phenotypic variation attributed to the genetic variance. This ratio varies between 0 and 1. Thus, a trait with heritability near 1.0 is not greatly affected by the environment. Because the fruit firmness variation shown in Figure 10 1 did not explain the cause of the variation, further experiments were conducted, as described in the examples below, to determine the cause of the variation. Example 3: Generation of Diploid Hybrids with Ultra Firm Flesh Trait 1001031 In the fall of 2002, in addition to the self pollinations, crosses with selected BC4S2/BC4S*2 generation plants were made to other commercial inbreds that 15 do not contain the ultra firm flesh phenotype. These crosses were made to test to what extent the ultra firm flesh trait would be dominantly inherited in a hybrid combination. Those skilled in the art will recognize the importance of establishing how well traits developed in inbred lines function in a hybrid combination. [001041 In the spring of 2003, these test hybrids were evaluated in Florida and 20 California. Although many hybrid combinations were tested in these trials, most of these data are not shown. Instead, data from four top performing hybrids across two trialing locations are shown in Tables 3 and 4. Hybrids were evaluated by a number of criteria, including the rind color pattern. For these hybrids, all had a mottled stripe pattern, designated MS. Also evaluated were fruit length and width, rind thickness, flesh color, 25 firmness and sweetness levels. 100105] When determining sweetness levels quantitatively, a refractometer was used to measure brix levels. Specifically, brix levels were measured with a digital, hand-held refractometer (Leica Microsystems model AR200, Reichert Inc., Depew, NY) according to manufacturer's instructions. Brix levels were determined after the 30 penetrometer firmness readings, by squeezing a sampled fruit until drops of liquid fell into the well of the refractometer. One brix measurement was taken from each half of a cut fruit, and the data were recorded as an average. [001061 Tables 3 and 4 show that the test hybrids do exhibit small variation between the test sites. Taken together, however, the data show that these top performing 25 hybrid combinations performed uniformly in the two locations. In particular, these hybrids consistently had ultra firm flesh, as measured by pound force of pressure and very good soluble solids, as measured by percentage brix. 1001071 Fruit flesh firmness data across the two locations provided insight into 5 the genetics of the trait, answering questions as to heritability posed by the data shown in Figure 1. First, these data show that the ultra firm flesh trait can be delivered into an Fl hybrid from a single parent. In other words, genetic loci selected in the method described herein affect fruit firmness in a dominant manner. This is a critical fact for the design of breeding strategies. Moreover, consistency in the firmness measurements across several 10 hybrids in the two locations show that the ultra firm flesh alleles selected in the method described herein have a high heritability. Those skilled in the art recognize the importance of creating commercial lines with highly heritable horticultural traits. Specifically, such cultivars will allow growers to produce a crop with uniform market specifications. 15 Table 3 Test hybrid evaluations: Florida, Spring 2003 Rind Hybrid Rind Length Width Thickness Flesh Firmness Sweetness (cm) (cm) (cm) Color (lbf) (Brix) 4201 MS 23 19.5 1.5 Red 8.0 9.7 4203 MS 25.5 21.5 1.5 Red 7.5 11.3 4204 MS 23 19 1.0 Red 6.0 9.3 4207 MS 25 20 2.0 Red 7.0 9.6 Table 4 Test hybrid evaluations: California, Spring 2003 Rind Hybrid Rind Length Width Thickness Flesh Firmness Sweetness (cm) (cm) (cm) Color (lbi) (Brix) 4201 MS 22.5 17 1.5 Red 8.0 9.7 Test hybrid evaluations: California, Spring 2003 (Cont.) 4203 MS 23 17 1.5 Red 8.0 10.3 4204 MS 25 18 2.0 Red 7.0 10.0 4207 MS 25 17 1.5 Red 6.0 10.3 26 Example 4: Evaluation ofLiquid-Retaining Flesh Characteristics of Ultra Firm Flesh Hybrids [00108] As described herein, studies agree that minimally processed products have a short shelf life of 2 to 3 days maximum (Perkins-Veazie et al. (1998) Hortscience 5 33:605; Wehner et al. in: Watermelons: Characteristics, Production and Marketing. Maynard, editor. ASHS Press, Alexandria VA 2001). Although the maximum shelf life of cut watermelon fruits is only a few days, product quality begins to deteriorate rapidly after being processed. In cut products presented in plastic food containers, the consumer can see this rapid deterioration because liquid will leak out of the cut products and 10 accumulate in the bottom of the container. a 00109] Mature fruits from the 2003 California hybrid trial (Example 3, Table 4) were evaluated for leakage using a liquid retention test as described herein (see Figure 2). This test was performed at 4" centigrade. Fruits from test hybrids 4201, 4204 and 4207 were tested along with standard diploid and triploid hybrid controls. Test hybrids 15 had the ultra firm flesh trait, with firmness readings of 8.0 lbf, 7.0 lbf and 6.0 lbf, respectively (about 35.6, about 31.1 and about 26.7 Newtons respectively; Table 4). Controls had flesh firmness readings of <2.0 lbf (<about 8.9 Newtons) and <2.5 lbf (<about 11.1 Newtons), respectively. To measure liquid loss, the edible portion of the fruits were cut into approximately 1" cubes and weighed. The approximate 1 inch cube 20 (about 16.4cm 3 ) size was chosen because this best approximates the processed product size found in retail outlets. Over a 16 day period, samples were re-weighed, and the liquid loss was estimated by calculating the percent weight loss. 1001101 Figure 2 graphically displays the percent weight loss of these samples over a 16 day period. Multiple samples per line were tested; the triangles, circle and 25 squares represent the mean values at each time point, and the sample standard deviations are shown as bars. Data in figure 2 show large differences in weight losses between the controls having softer fruit flesh and the fruits with the ultra firm flesh trait. The difference between the controls and the test hybrids with the ultra firm flesh phenotype was apparent by the first time point, which was approximately 6 2 hours after the 30 samples were cut. Therefore, although cut product from standard cultivars may have a shelf life of up to 2 to 3 days, product deterioration begins almost immediately after they are cut. These data show that the ultra firm flesh lines developed using the method described herein will resist the rapid liquid leakage now common in cut watermelon fruits. Because these ultra firm flesh fruits will retain liquid once cut, they will last longer 27 in the minimally processed watermelon market. Example 5: Final Self Pollinations and Creation and Evaluation of Triploid Hybrids 1001111 In the summer of 2003, the BC4S4/BC4S*4 generation, the generation of which is described above in Example 2, was grown and evaluated qualitatively for 5 sweet taste, fruit flesh firmness, and horticultural characteristics. Based on these evaluation criteria, plants were selected to create the next generation. Self pollination created the BC4S5/BC4S*5 generation. 1001121 In the fall of 2003, the BC4S5/BC4S*5 generation was grown and evaluated qualitatively for sweet taste, fruit flesh firmness, and horticultural 10 characteristics. Based on these evaluation criteria, plants were selected to create the next generation. Self pollination created the BC4S6/BC4S*6 generation. [001131 In addition, quantitative firmness data were collected from the BC4S5 generation for lines that were qualitatively sweet. Specifically, twenty six lines were tested, and results are shown below in Table 5. Fourteen of these lines had a single fruit 15 tested, and the remaining 12 lines had 2 or 3 fruits tested per line. The range of firmness amongst the twenty six lines ranged from a low of 4.0 lbf (about 17.8 Newtons) to a high of 8.0 lbf (about 35.6 Newtons). For the lines that had multiple samples, 11 of the 12 lines showed no difference in the penetrometer measurements. One line did show a penetrometer measurement difference of 1 lbf (about 4.5 Newtons). These data provide 20 further insight as to questions raised by Figure 1, which showed variation in the ultra firm flesh trait in the BC4S3 generation. In particular, it was unclear in the BC4S3 generation whether the ultra firm flesh trait displayed a low or high heritability. That many lines developed in parallel gave elevated, but different fruit firmness readings suggested that the ultra firm flesh is polygenic in nature. The very low intra-line variation shown in 25 Table 5, together with the test hybrid data shown in Tables 3 and 4 demonstrate that the ultra firm flesh trait has a high heritability. Those skilled in the art recognize the importance of creating commercial lines with highly heritable horticultural traits because such cultivars allow growers to produce a crop with uniform market specifications. Table 5 shows the inbred line evaluations from the BC4S5/BC4S*5 generation. Line - replication no. Firmness 3333 -1 7.0 lbf 3333 -2 8.0 lbf 3334-1 5.0 lbf 28 Line - replication no. Firmness 3334 -2 5.0 lbf 3335 -1 8.0 lbf 3335 -2 8.0 lbf 3336-1 5.0 lbf 3336-2 5.0 lbf 3337-1 5.0 lbf 3339-1 4.0 Ibf 3340-1 4.5 Ibf 3340-2 4.5 Ibf 3340-3 4.5 Ibf 3341-1 5.0 Ibf 3346-1 5.0 lbf 3346-2 5.0 lbf 3347-1 6.0 lbf 3347-2 6.0 Ibf 3347-3 6.0 Ibf 3348-1 6.0 Ibf 3348-2 6.0 lbf 3348-3 6.0 lbf 3349-1 5.0 lbf 3350-1 5.0 lbf 3350-2 5.0 lbf 3350-3 5.0 lbf 3352-1 5.5 lbf 3353-1 6.0 lbf 3355-1 8.0 lbf 3355-2 8.0 lbf 3357-1 5.0 lbf 3357-2 5.0 lbf 3358-1 5.0 Ibf 3358-2 5.0 lbf 3359-1 6.0 lbf 3378-1 7.0 Ibf 3380-1 7.0 lbf 3384-1 7.0 lbf 29 Line - replication no. Firmness 3386-1 7.0 lbf 3387-1 8.0 lbf 3388-1 7.0 lbf 3390-1 6.0 lbf 3390-2 6.0 lbf 3392-1 8.0 lbf 3392-2 8.0 lbf 3394-1 5.5 lbf 3394-2 5.5 lbf 3396-1 7.0 lbf 3396-2 7.0 lbf 3397-1 7.0 lbf 3397-2 7.0 lbf 3398-1 7.5 lbf 3399-1 7.5 lbf 3399-2 7.5 lbf 3400-1 7.5 lbf 3401-1 7.5 lbf 3401-2 7.5 lbf 1577-1 8.0 lbf 1577-2 8.0 lbf 1577-3 8.0 lbf 1577-4 8.0 lbf 1577-5 8.0 lbf 1577-6 8.0 lbf 1577-7 8.0 lbf 1001141 In addition to the self pollinations described above, crosses with selected BC4S4/BC4S*4 generation plants were made to other commercial tetraploid inbreds that do not contain the ultra firmness phenotype. These tetraploid x diploid crosses were made to test to what extent the ultra firm flesh trait would be dominantly 5 inherited in a triploid hybrid combination. As shown in Table 6 below, the ultra firm flesh trait was inherited by the triploid seedless fruit.

30 Table 6. Mature fruit flesh firmness and sweetness scores. Firmness was measured as described herein with a penetrometer. Rind Length Width Rind Flesh Firmness TSS Hybrid (cm) (cm) Thickness (cm) Color (lbo (Brix) SVR8034-7131 TS 28 24 1.2 Red 5.0 10.2 SVR8034-7132 MS 26 25 1.0 Red 4.0 9.7 SVR8034-7133 MS 28 26 1.0 Red 5.0 10.5 SVR8034-7134 MS 26 24 1.1 Red 4.5 10.0 1001151 All references cited herein are hereby expressly incorporated herein by reference. 5 DEPOSIT INFORMATION 1001161 A deposit of the Seminis Vegetable Seeds proprietary inbred and hybrid watermelon line 3347 disclosed above and recited in the appended claims has been made with NCIMB Ltd, 23 St. Machar Drive, Aberdeen AB24 3RY, Scotland. The date of the deposit was 1 July 2004. The deposit of 2500 seeds for this variety were taken 10 from the same deposit maintained by Seminis Vegetable Seeds since prior to the filing date of this application. Upon issuance of a patent, all restrictions upon the deposit will be removed, and the deposit is intended to meet all of the requirements of 37 C.F.R. § 1.801-1.809. The NCIMB accession number for inbred line 3347 was deposited as Accession No. NCIMB 41230. 15 [001171 Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity and understanding, it will be obvious that certain changes and modifications may be practiced within the scope of the invention, as limited only by the scope of the appended claims.

Claims (44)

1. A watermelon plant that produces fruit having few to no seeds, said fruit having ultra firm flesh and soluble solids of at least about 6 brix wherein the ultra firm flesh resists pressure of at least about 3.5 lbf. 5

2. The watermelon plant of Claim I wherein the ultra firm flesh resists pressure from at least about 4.0 lbf to at least about 5.0 lbf.

3. The watermelon plant of Claim 1, wherein the ultra firm flesh resists pressure of at least about 5.0 lbf.

4. The watermelon plant of Claim 1, wherein the ultra firm flesh resists 10 pressure of at least about 6.0 lbf.

5. The watermelon plant of Claim 1, wherein the ultra firm flesh resists pressure of at least about 8.0 lbf.

6. The watermelon plant of any one of Claims I to 5, wherein said fruit has good soluble solids. 15

7. The watermelon plant of Claim 6, wherein said fruit has very good soluble solids.

8. The watermelon plant of any one of Claims 1 to 7, wherein the plant produces a fruit weighing at least about 1.5 kg.

9. The watermelon plant of Claim 8, wherein the plant produces a fruit 20 weighing at least about 3.0 kg.

10. The watermelon plant of Claim 9, wherein the plant produces a fruit weighing at least about 4.5 kg.

11. The watermelon plant of Claim 10, wherein the plant produces a fruit weighing at least about 6.0 kg. 25

12. The watermelon plant of any one of Claims I to 7, wherein the plant produces a fruit weighing from about 6.4 to about 8.2 kg.

13. The watermelon plant of any one of Claims I to 12, wherein the plant produces fruit having red flesh, yellow flesh or orange flesh.

14. The watermelon plant of Claim 13, wherein the plant produces fruit having 30 red flesh.

15. The watermelon plant of Claim 13, wherein the plant produces fruit having yellow flesh or orange flesh.

16. The watermelon plant of any one of Claims 1 to 15, wherein the plant is triploid. 32

17. A watermelon plant that produces fruit having few to no seeds, said fruit having ultra firm flesh and soluble solids of at least about 6 brix, wherein the ultra firm flesh resists pressure of at least about 3.5 lbf, substantially as hereinbefore described with reference to any one of the examples. 5

18. A plurality of watermelon plants according to any one of Claims I to 17 grown in a field.

19. A plant part or product of a plant of any one of Claims 1 to 17, comprising seed, fruit, pollen, ovule or vegetative tissue.

20. A watermelon plant generated from a plant part or product of Claim 19. 10

21. A watermelon fruit harvested from a plant according to any one of Claims I to 17 and 20.

22. A watermelon plant that produces fruit having few to no seeds, said fruit having liquid-retaining flesh and soluble solids of at least about 6 brix.

23. A watermelon plant that produces fruit having few to no seeds, said fruit 15 having liquid-retaining flesh and soluble solids of at least about 6 brix, substantially as hereinbefore described with reference to any one of the examples.

24. A plant part or product of a plant of Claim 22 or claim 23, comprising seed, fruit, pollen, ovule or vegetative tissue.

25. The fruit of a watermelon plant of claim 22 or claim 23. 20

26. The fruit of Claim 25, wherein said liquid-retaining flesh, when cut, loses about three and one-half percent weight after three days storage at 40 centigrade.

27. The fruit of Claim 26, wherein said liquid-retaining flesh, when cut, loses about three percent weight after three days storage at 40 centigrade.

28. The fruit of Claim 27, wherein said liquid-retaining flesh, when cut, loses 25 about two percent weight after three days storage at 40 centigrade.

29. The fruit of Claim 28, wherein said liquid-retaining flesh, when cut, loses about one and one-half percent weight after three days storage at 40 centigrade.

30. The fruit of any one of Claims 25 to 29, further having ultra firm flesh.

31. The fruit of Claim 30, wherein the ultra firm flesh resists pressure of at 30 least about 3.5 lbf.

32. The fruit of Claim 31, wherein the ultra-firm flesh resists pressure from at least about 4.0 lbf to at least about 5.0 lbf.

33. The fruit of Claim 31, wherein the ultra-firm flesh resists pressure of at least about 5.0 lbf. 33

34. The fruit of Claim 31, wherein the ultra-firm flesh resists pressure of at least about 6.0 lbf.

35. The fruit of Claim 31, wherein the ultra-firm flesh resists pressure of at least about 8.0 lbf. 5

36. The fruit of any one of Claims 25 to 35, having soluble solids of at least about 8 brix.

37. The fruit of Claim 36, having soluble solids of at least about 10 brix.

38. The fruit of Claim 37, having soluble solids of at least about 11.5 brix.

39. A watermelon fruit having few to no seeds, liquid-retaining flesh and Io soluble solids of at least about 6 brix, substantially as hereinbefore described with reference to any one of the examples.

40. A watermelon fruit having few to no seeds, said fruit having ultra firm flesh and soluble solids of at least about 6 brix wherein the ultra firm flesh resists pressure of at least about 3.5 lbf, substantially as hereinbefore described with reference to any one 15 of the examples.

41. The flesh of the fruit according to any one of Claims 25 to 29.

42. The flesh of the fruit according to any one of Claims 30 to 35.

43. The flesh of the fruit according to any one of Claims 36 to 38.

44. The flesh of the fruit according to claim 39 or claim 40. 20 Dated 1 May, 2007 Seminis Vegetable Seeds, Inc. Patent Attorneys for the Applicant/Nominated Person SPRUSON & FERGUSON