Treatments.

Thirteen pollenizer cultigens were evaluated: ‘Companion’, ‘Jenny’, ‘High Set 11’, ‘Mickylee’, ‘Minipol’, ‘Pinnacle’, ‘Summer Flavor 800’ (‘SF800’), ‘Super Pollenizer 1’ (‘SP1’), ‘Sidekick’, ‘TP91’, ‘TPS92’, ‘WC5108-1216’, and ‘WH6818’. Triploid watermelon cultigens, ‘Petite Perfection’ and ‘Tri-X-313’, were included for comparison of vegetative and floral habit.

Cultural practices.

Seeds were sown into 3.8 cm × 3.8-cm cell trays on 20 Apr. each year. Plants were grown in the greenhouse for 2 weeks and moved to a coldframe for 2 weeks of conditioning for field establishment. Seedlings were hand planted into beds covered with black polyethylene mulch at the Central Crops Research Station, Clayton, NC, into a Norfolk loamy sand (fine-loamy, kaolinitic, thermic typic Kandiudults) soil on 24 May 2005 and 17 May 2006. As beds were formed and just before mulch was laid, the soil was fumigated with 1,3-dichloropropene and chloropicrin (Telone C-35: Dow AgroSciences, Calgary, Alberta, Canada) and broadcast-fertilized (0N–0P–50K at 149.1 kg·ha⁻¹ and 16N–0P–0K at 213.0 kg·ha⁻¹) according to soil test recommendations. In-row spacing was 1.3 m and between-row spacing was 3.1 m. Plant spacing around plants was maximized to facilitate individual plant measurements. Four watermelon plants of each cultigen were included in each plot. Cultigens were arranged in a randomized complete block design with five replications. Pesticides and fertigation were applied based on North Carolina cultural recommendations (Sanders, 2004, 2005).

Vegetative development.

Vine and internode lengths were measured at 4 and 5 weeks after transplanting (WAT). The longest vine on each plant was measured to the nearest 0.5 cm from the crown to the tip of the vine. The internode measurement to the nearest 0.5 cm was taken from the internode between the third and fourth fully expanded leaves from the tip of the longest vine. These measurements were not collected after 5 WAT because moving the vines for data collection after fruit set increased the risk of fruit detachment from vines.

Flower production.

Flowers were counted weekly from 3 to 9 WAT between 0730 hr and 1200 hr on individual plants. Flowers that were open only on the day of counting were included in total weekly counts.

Fruit development.

Cumulative fruit set was recorded during the same period as anthesis. Fruit set was determined and counted when the petals had dried and fallen off and the fruit had enlarged. Mature fruit were harvested, counted, and weighed. Rind thickness and soluble solids were evaluated to determine if pollenizer fruit could be marketed as a seeded watermelon. Rind thickness was measured on five cut fruit from each plot and measured from the outside of the fruit to where the interior of the fruit started to turn pink or cream in color. Rind thickness measurements were taken on both sides of the fruit at the equatorial position of the fruit. Soluble solids were measured with a handheld refractometer.

Data were analyzed using GLM and means were separated using Fisher's protected least significant difference (P < 0.05) (SAS Institute, 2002). Years were significantly different; thus, years are presented separately.

Vegetative development.

Vine lengths over both years fit into two distinct vegetative categories (dwarf and standard) (Table 1). The difference between the two groups was noticeable at 4 WAT (data not shown); however, the differences became more distinct at 5 WAT. Dwarf cultigens included in these studies were ‘Companion’, ‘Sidekick’, ‘TP91’, ‘TPS92’, and ‘WC5108-1216’. All other pollenizer cultigens had standard-length vines. The triploid watermelon cultigens had similar vine lengths to standard-length pollenizers. A dwarf cultivar of watermelon has shorter internodes caused by fewer cells in the internodes compared with a standard watermelon vine (Liu and Loy, 1972).

Table 1.Watermelon seed source and length of watermelon vine measured from the crown to the tip of the longest branch/plant/day for different pollenizer and triploid cultigens.

View Fullscreen

In both years, dwarf length cultigens had shorter internodes than cultigens we categorized as having a standard vine length (data not shown). Although not measured directly, the amount of branching varied between cultigens. ‘High Set 11’, ‘Sidekick’, and ‘SP1’ appeared to have substantial multibranching.

Vegetative development of the pollenizer may affect quantity and quality of triploid fruit. Dwarf types are theoretically less suited for the separate row planting method because staminate flowers produced by pollenizers may be a considerable distance from the triploid plants, which may hinder pollen transfer (Motsenbocker and Arancibia, 2002). Planting pollenizers in the same row as triploid plants minimize space between triploid and pollen source plants. Thus, the dwarf cultigens would likely be better suited for in-row planting to minimize the competition with triploid plants. In contrast, the use of standard-length cultigens may be better for the separate row planting method compared with the dwarf vine types. The separate row planting method maximizes the amount of space between the triploid and pollenizer plants. The larger space likely would result in less competition for nutrients, water, and light between the pollenizer and the triploid plants. Neppl et al. (2003) reported differences in vine vigor and growth among watermelon cultivars. The more vigorous growth of ‘Mickylee’ showed a trend to decrease fruit size and number compared with the smaller habit of ‘SP1’, a dedicated pollenizer (Freeman et al., 2007a). A vigorous vining pollenizer may act much like a weed and could reduce the yield and quality of the crop. Weed competition can reduce the yield and quality of many crops (Buker et al., 2003; Monks and Schultheis, 1998); however, a pollenizer is typically not given this consideration.

Triploid plants may cover the shorter internodes of the dwarf pollenizers minimizing visibility of flowers to bees. Standard-length cultigens with longer internodes may more effectively compete with the triploid plants.

Staminate flower production, 2005.

Tables 2 through 4 were used to construct Figures 1A and 1B. The figure is included with the tables to illustrate different concepts. To minimize confusion and permit visual resolution in the figure, selected pollenizer cultigens were chosen to represent the different types of pollenizers. The earliest peak production of staminate flowers was at 5 and 6 WAT for ‘Companion’ (Table 2; Fig. 1A). Although this was the earliest peak time for staminate flowers relative to other cultigens, ‘Companion’ did not produce the most staminate flowers during those weeks. Compared with ‘Companion’, peak staminate flower production was later for a number of cultigens at 7 WAT. Two cultigens, which produced the most staminate flowers per plant per day at 6 and 7 WAT, were ‘SP1’ and ‘Sidekick’. ‘SP1’ produced the most staminate flowers on a given day at 6 and 7 WAT. ‘Sidekick’ produced the highest number of staminate flowers at 4 and 9 WAT relative to the other cultigens. ‘Sidekick’ and ‘SP1’ produced a similar number of staminate flowers per day and were the leading producers of staminate flowers among other cultigens at 5 and 8 WAT. ‘TP91’ and ‘TPS92’, dwarf cultigens, produced the fewest staminate flowers during the entire experiment. The triploid cultigens produced a similar number of staminate flowers as the pollenizer cultigens with the lowest staminate flower production.

Table 2.Total number of open staminate flowers/plant/day for different pollenizer and triploid watermelon cultigens in 2005.

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

Staminate flower production, 2006.

Staminate flower production peaked later in the 2006 experiment compared with 2005 (Tables 2–3; Fig. 1A–B). This delay was likely as a result of cooler temperatures at the time of transplanting. Daily peak production of staminate flowers for ‘Companion’ was 6 and 7 WAT (Table 3; Fig. 1B). Peak production of staminate flowers for the other cultivars was 8 and 9 WAT (Table 3). Staminate flower production after the peak production period was not examined because flower counts were stopped at 9 WAT because fruit set had occurred on the triploid plants and we used concurrent data acquisition times in the previous year's study. ‘SP1’ and ‘Sidekick’ produced the most staminate flowers per day 8 and 9 WAT. The third highest-producing cultigen of all the cultigens examined was ‘High Set 11’ and was comparable to ‘SP1’ or ‘Sidekick’ at 6 and 9 WAT. The dwarf vine cultigens, ‘Companion’, ‘TP91’, ‘TPS92’, and ‘WC5108-1216’, typically produced fewer staminate flowers per day than the standard vine types at 7, 8, and 9 WAT, whereas at 4, 5, and 6 WAT, these cultigens produced a similar number of staminate flowers compared with the standard vine types. ‘TP91’ and ‘TPS92’ produced very few staminate flowers per plant and never exceeded five per plant on a given day over the 6-week period that flowers were counted. The triploid cultigens produced the same number of staminate flowers per day as many standard-length pollenizers, which had an intermediate staminate flower response.

Table 3.Total number of open staminate flowers/plant/day for different pollenizer and triploid watermelon cultigens in 2006.

View Fullscreen

Pistillate flower production, 2005 and 2006.

In 2005, the peak production of pistillate flowers for the triploid cultigens was at 6 WAT for ‘Petite Perfection’ and 5 WAT for ‘Tri-X-313’ (Fig. 1). In 2006, the peak production of pistillate flowers per day for the triploid cultivars, Petite Perfection and Tri-X-313, was 6 and 9 WAT, respectively.

Vine habit has an effect on the number of staminate and pistillate flowers that are produced. In cucumbers, little leaf types produce more branching than normal leaf and additional branching of the little leaf type cucumber results in higher flower and fruit production (Bowers et al., 1981; Goode et al., 1989; Schultheis et al., 1998). Watermelon flowers develop in the nodes (Rubatzky and Yamaguchi, 1997) and the obvious additional branching observed on ‘Sidekick’ and ‘SP1’ likely created more locations for flower development.

Staminate flower production on pollenizers should overlap with the production of pistillate flowers on triploid plants to maximize the pollination and set of triploid fruit. Maynard and Elmstrom (1992) discussed that earlier or later flowering pollenizers affected the timing of triploid fruit harvest. Several cultigens had consistent, prolific staminate flower production through the season (i.e., ‘SP-1’, ‘Sidekick’, and ‘High Set 11’) (Fig. 1). This continuous production might extend triploid fruit production and expand the length of harvest time. ‘Companion’ produced staminate flowers earlier in the season, which may lead to earlier triploid fruit set and harvest. In 2005, pistillate flower production of the triploids overlapped best with ‘Companion’ staminate flower production. ‘Petite Perfection’ had peak production of pistillate flowers at the same time as the peak production of staminate flowers on ‘WC5108-1216’. In 2006, peak pistillate flower production per day of the triploid fruit overlapped with the peak production of staminate flowers on all pollenizer cultigens, except ‘Companion’. Staminate flower production does not necessarily need to overlap with triploid pistillate flower production, because there likely is an adequate amount of viable pollen being produced by pollenizers during some of its off-peak times. However, for the best chance of triploid fruit set and quality, peak pollenizer staminate flower production should coincide with triploid pistillate flower production.

Fruit development, 2005.

Most cultigens had set one fruit per plant 5 WAT (data not shown). Beginning 5 WAT, ‘Sidekick’ and ‘SP1’ continued to set one or more fruit throughout the 5-week period (Table 4). ‘High Set 11’ averaged 0.5 or more fruit per week between 6 and 9 WAT. ‘Companion’ fruit set was complete 6 WAT with approximately two fruits per plant. By 9 WAT, a total of eight fruit and 13 fruit were set per ‘SP1’ and ‘Sidekick’ plants, respectively. Other cultigens set between two and five fruits per plant between 5 and 9 WAT. Most cultigens set two or three fruit per plant by 6 WAT and one to two additional fruits were set in the subsequent 3 weeks. ‘Petite Perfection’ set most of its fruit 6 and 9 WAT and ‘Tri-X-313’ set most of its fruit 6 and 8 WAT.

Table 4.Cumulative number of set watermelon fruit/plant/day for different pollenizer and triploid watermelon cultigens.

View Fullscreen

Fruit development, 2006.

Delay in staminate flowers caused a corresponding delay in fruit set compared with 2005 with most cultigens not beginning to set fruit until 6 WAT (Table 4). ‘Jenny’, ‘SF800’, and ‘WC5108-1216’ set less than one fruit per plant until 9 WAT. ‘SP1’ and ‘Sidekick’ set the most fruit by 9 WAT. ‘Companion’, ‘TP91’, and ‘TPS92’ initially set one fruit per plant 6 WAT with limited fruit set thereafter. The other cultigens had their first fruit set by 6 or 7 WAT and then a second by 8 to 9 WAT. Concentrated fruit set for the triploid cultigens occurred by 9 WAT.

Fruit quality, 2005 and 2006.

Fruit interior was examined to determine the acceptability of pollenizer cultigens for potential sales of seeded watermelon. In 2005 and 2006, the pollenizer cultigens with a rind thickness greater than 10 mm were ‘Companion’, ‘Mickylee’, ‘SF800’, ‘TP91’, ‘TPS92’, and ‘WC5108-1216’ (Table 5). Furthermore, all the cultigens except ‘Companion’, ‘TP91’, ‘TPS92’, and ‘WC5108-1216’ had the same rind width as ‘Petite Perfection’. In 2006, ‘High Set 11’, ‘Jenny’, ‘Pinnacle’, and ‘WH6818’ had the same rind thickness as ‘Petite Perfection’. The cultigens with the thinnest rinds were ‘High Set 11’, ‘SP1’, and ‘Sidekick’.

Table 5.Average mass, rind thickness, and soluble solids of different pollenizer and triploid watermelon cultigen fruit in 2005 and 2006.

View Fullscreen

‘SF800’ and ‘Tri-X-313’ have a thick rind and are shipped and marketed in bins. Based on rind thickness alone, ‘Companion’, ‘Mickylee’, ‘TP91’, ‘TPS92’, and ‘WC5108-1216’ all meet the criteria for shipment and marketing in bins. ‘Petite Perfection’ is a mini watermelon cultivar that has a thinner rind than the standard-sized triploid fruit such as ‘Tri-X-313’, ‘High Set 11’, ‘SP1’, and ‘Sidekick’ are cultigens with a thinner rind that easily collapses if stepped on in the field and are less likely to result in injury to workers.

For watermelon fruit to be marketable, watermelon should have a minimum °Brix (soluble solids) of 10.0 (USDA, Agricultural Marketing Service, 2006). The only cultigens to have soluble solids at this level in 2005 were ‘Companion’, ‘Pinnnacle’, ‘SF800’, and ‘WC5108-1216’ (Table 5). In 2006, additional cultivars included Jenny, Mickylee, Minipol, and TPS92. Cultigens with the lightest red flesh color were ‘Mickylee’, ‘Sidekick’, ‘TP91’, and ‘TPS92’. ‘SP1’ has a white flesh color that turns cream as it ages. ‘High Set 11’ segregated red, pink, white, and orange-colored flesh. ‘TP91’, ‘TPS92’, and ‘WC5108-1216’ had thick rinds characteristic of marketable fruits. However, they had a °Brix, which was generally less than 10.0 and therefore would not be suited for market.

This research was not focused on pollenizer cultigens that were suitable for market. However, we would be remiss if we did not at least consider their potential because they are an economic consideration should the producer want to consider their suitability for market. There is still demand in many markets for seeded watermelon. Some of the new dedicated pollenizer cultivars that meet the fruit quality criteria for market are ‘Pinnacle’ and ‘Minipol’. ‘Jenny’ has good fruit quality and is currently sold in Europe but not in the United States. Our conclusions are based on fruit with acceptable red flesh color and sugars. ‘Pinnacle’ and ‘Jenny’, however, have thinner rinds and may require special packaging. ‘Mickylee’ and ‘SF800’ are commonly grown for commercial sale but also used as pollenizers in North and South America.

When selecting a pollenizer, vegetative, floral, and fruit characteristics should be considered. Factors that could have the most impact on the yield of triploid watermelon production are the quantity and timing of peak production of staminate flowers and vine growth of the pollenizer. Although a cultigen may have higher quantities of staminate flowers per plant compared with other cultigens, the higher number of staminate flowers may not coincide with a high quantity of pollen (Stanghellini and Schultheis, 2005). Freeman et al. (2008) found no variability in pollen viability of diploid watermelon pollenizers. Thus, pollen production and not viability most likely has greater influence on triploid fruit set. In subsequent studies, pollen production of these cultigens should be quantified.

Motsenbocker and Arancibia (2002) and Sanders et al. (1999) showed reducing in-row spacing between watermelon plants cause fruit size and number of fruit per vine to decline. However, these two research studies did not examine dedicated pollenizer cultigens. When spacing cucumber plants, reducing the in-row spacing had the same effect on little leaf cucumber with a multibranching and regular leaf cucumber (Bowers et al., 1981). Thus, the multibranching habit of ‘SP1’ and ‘Sidekick’ with more side branching may have limited effect on the production of triploid fruit when pollenizers are interplanted in the same row as transplants. Dwarf plant structure has been used in other crops such as apple, peach, and nectarine to maximize planting density (Vizzoto and Costa, 1997; Weber, 2001). The cultigens used in this study with a dwarf vine structure could be planted at higher quantities or interplanted with triploid watermelons to maximize the number of staminate flowers in the field with potentially little effect on the yield of the triploid fruit.