Flowering, Sex Expression, and Fruiting of Pumpkin (Cucurbita sp.) Cultivars under Various Temperatures in Greenhouse and Distant Field Trials

Field production of decorative pumpkins (Cucurbita pepo L.) in New York occasionally results in markedly delayed fruit production in spite of normal vine growth. These episodes of fruitlessness appear to be associated with periods of high temperatures. To determine the link between temperature and pumpkin flowering and fruiting, a series of multilocational field trials and confirmatory greenhouse experiments were carried out. The field trials were conducted in the summer seasons of 1996 and 1997 in Ithaca and Albany, N.Y.; Queenstown, Md.; and Bradenton, Fla.; and in Ithaca and Bradenton in 1998. Mean growing season temperatures were 20, 21, 24 and 28 °C, respectively, at the four locations in 1996 and 1997. Delay in fruit formation was indicated by the main stem node number at which the first fruit developed. In Ithaca and Albany, the six cultivars formed their first fruit at node 17, but fruit production shifted to node 24 at Queenstown, and to node 26 or more at Bradenton. The prolonged delay in fruiting at the warmest site resulted in a 74% decrease in total yield of the C. pepo cultivars in 1996 and 1997, compared to Ithaca and Queenstown. In contrast, the yields and yield components of the C. maxima cultivar Prizewinner were similar at all four sites. Greenhouse trials in which ʻHowdenʼand ʻBaby Bearʼwere grown at 32/27, 25/20, and 20/15 °C confirmed that high temperatures delay formation and anthesis of female flowers. This and other published work indicates that there are genetic differences in susceptibility to high temperature flower delay that could be exploited to improve pumpkin performance. The commercial production of pumpkins in North America has traditionally been concentrated in the central and northern regions of the United States, and in southern Canada (Peirce, 1987). As market demand for decorative pumpkins during October and November has increased, more growers in southern states are attempting to produce the crop locally. Production problems of heavy disease and insect pressure, and “lack of adaptation” have resulted in low yields and discouraged local production in the southernmost states (Elmstrom et al., 1988). In cooler growing areas, such as upstate New York, female flower buds sometimes turn yellow and then necrotic before anthesis after occasional hot spells (Wien and Riggs, personal observation). Although susceptible cultivars may not have any female flowers reach anthesis for more than a week in those conditions, there is no quantitative information on the effect of either shortor long-term high temperatures on pumpkin fruit production under field conditions. Failure of cucurbit plants to form female flowers has been documented most extensively in cucumbers, where environmental conditions can influence whether a male or female flower is formed from a common primordium (Cantliffe, 1981; Goffinet, 1990). In general, moderate temperature and high light intensity favored development of pistillate flowers in monoecious cucumbers. High temperatures, especially if combined with low light conditions, reduced female flower numbers (Cantliffe, 1981). Other researchers have noted a delay in female flower production when C. pepo plants are grown at higher temperatures. Nitsch et al. (1952) found that ̒ Acornʼsquash grown under controlled environmental conditions at 30 °C mean temperature and 16-h photoperiod produced up to 160 nodes without a single female flower. Reducing the temperature, particularly at night, allowed female flowers to form at nodes 20 to 30 on the main stem. Similarly, NeSmith et al. (1994) observed that some zucchini cultivars produced fewer pistillate flowers when planted in midsummer in Georgia, when temperatures averaged 26 °C. The objectives of the present study were to observe the temperature response of pumpkin flower sex expression and to determine the effect of different growing season temperatures on flowering and fruit formation of several pumpkin cultivars in a series of greenhouse and multilocational field trials. Materials and Methods Four-location fi eld trial. Six cultivars of pumpkin were grown at Ithaca and Albany, N.Y.; Queenstown, Md.; and Bradenton, Fla., which provided a range of growing season temperatures. Ithaca, Queenstown, and Bradenton were the sites in 1996; all four locations were used in 1997; and the trial was repeated in Ithaca and Bradenton in 1998. Cucurbita pepo cultivars were ̒ Howdenʼ, an open-pollinated, large-fruited (normally 10 kg), viney cultivar; ʻWizardʼ and ̒ Appalachianʼ, two hybrid lines with compact growth habit and moderate fruit size (6–8 kg); ʻRocketʼ, an early-fruiting, viney hybrid with medium-large fruit; and ʻBaby Bearʼ, a small-fruited (1 kg) hybrid with extensive vines. The large-fruited (20–30 kg) ʻPrizewinnerʼ hybrid (C. maxima) was also included in the trials. Seeds for all the trials were from commercial sources, and were from the same seed lots for trials at all locations. Seeds for all Ithaca trials were started in Todd Planter polystyrene trays in a greenhouse at Cornell set at 25/20o C day/night temperature, and transplanted after ≈3 weeks. The other plantings were direct-seeded in the field. Sowing and harvest dates were timed to have marketable fruit by mid-October at each location (Table 1). The 12-plant plots consisted of single rows 7.3 m long and 2.7 m apart, with plants 0.61 m apart in the row. The between-row spacing was 3 m in the Queenstown plantings. There were two replications in 1996, and four in 1997 and 1998. Levels of applied nitrogen were at or below state guidelines in each planting (Table 2). Local cultural practice recommendations for control of weeds, insects, and diseases were used at each site, and were adequate to assure normal crop growth in each location. Rainfall was supplemented by sprinkler irrigation at all locations except Bradenton, where seepage irrigation maintained the watertable of the beds. Temperature data were obtained from weather stations in the vicinity of the experimental sites. HORTSCIENCE 39(2):239–242. 2004. Received for publication 26 Sept. 2002. Accepted for publication 17 Jan. 2003. Apr04HortScience.indb 239 3/25/04 10:59:58 AM HORTSCIENCE, VOL. 39(2), APRIL 2004 240 CROP PRODUCTION Table 1. Dates of sowing and harvest for six pumpkin cultivars grown in four locations from 1996 to 1998. Dates of sowing Dates of harvest Location 1996 1997 1998 1996 1997 1998 Ithaca, N.Y. 24 May 19 May 24 May 21 Sept. 29 Sept. 23 Sept. Albany, N.Y. --10 June ----24 Sept. --Queenstown, Md. 6 June 6 June --23 Sept. 22 Sept. --Bradenton, Fla. 16 July 28 July 28 July 3 Oct. 23 Oct. 16 Oct. When the first fruits had reached a diameter of at least 10 cm on the large-fruited cultivars, the location of the first-formed fruits on the main stem was noted on six plants randomly chosen within each plot. Yields were obtained by harvesting all the plants, tracing vines into adjacent plots if necessary, and weighing and counting all fruits. The harvest sample included unmarketable fruits, but not immature fruits that were green and less than typical marketable size for the cultivar. Statistical analysis of the data was conducted with MSTAT software (Michigan State Univ., East Lansing). Data were analyzed separately for each year with locations and cultivars as factors. Greenhouse experiment. To determine the effect of temperature on development of female flower primordia under more controlled conditions than are available in the field, ʻHowdenʼ and ʻBaby Bearʼ were grown in three greenhouse compartments. Seeds for the temperature experiments were germinated at 32/27 °C (day/night) in moist paper towels and normal, vigorous seedlings selected after 3 d and transplanted into 23-cm-diameter (7-L volume) pots filled with peat-vermiculite artificial soil mix. Ten pots per treatment were placed in greenhouse compartments set at 32/27, 25/20, and 20/15 °C. Pots were spaced 100 cm × 46 cm on the bench and vines trained to the bench middles. Because of differences in vine vigor, ʻHowdenʼ and ʻBaby Bearʼ were placed in adjacent blocks on the same bench. Natural light was supplemented by light from metal halide lamps for a 12-h duration during the day, to provide at least 325–400 μmol·cm·s irradiance. Location of male and female flower buds on the main stem was noted at weekly intervals when first visible. Plants continued to grow and develop flowers due to lack of pollination and fruit set in the greenhouse (El Keblawy and Lovett-Doust, 1996). The experiment was conducted twice, beginning in Nov. 2001 and Feb. 2002, in natural photoperiods that ranged from 11.5 to 15 h. Since the data obtained were similar in both iterations, means and standard errors were calculated on the pooled 20 plants per treatment.