Introduction

The sexual expression of zucchini squash (Cucurbita pepo L.) ‘Caserta’ is monoecious, with orange, large, and showy flowers, occurring in separate points on the same plant, and isolated in the leaf axils (^{Filgueira, 2008}). Female flowers, which have an elongated ovary, are usually less numerous than male flowers (^{Filgueira, 2008}).

Zucchini squash has two or three sexual phases, starting with production of male flowers exclusively, then male and female flower production occurs alternatively, and finally female flowers are produced (^{Manzano et al., 2010a}; ^{Manzano, Martínez, Megías, Garrido, & Jamilena 2013}; ^{Martínez et al., 2014}; ^{Peñaranda, Payán, Garrido, Gómez, & Jamilena, 2007}).

Genetics, environment and hormonal factors influence sexual expression in Cucurbitaceae (^{Bhowmick & Jha, 2015}; ^{Campbell, Luo, & Mercer, 2013}; ^{Manzano, Martínez, García, Megías, & Jamilena, 2014}). Low temperatures inhibit development of male flowers and stimulates female ones in zucchini squash; high temperatures support partial, or complete, conversion of female flowers into bisexual flowers and stimulates production of male flowers (^{Peñaranda et al., 2007}).

Ethylene is the principal hormone regulating sexual expression in the Cucurbitaceae (^{Manzano et al., 2011}; ^{Papadopoulou, Little, Hammar, & Grumet, 2005}; ^{Yamasaki, Fujii, & Takahashi, 2005}). Exogenous ethylene application can modify the proportion of male/female flowers (^{Manzano et al., 2010a}; ^{Payán et al., 2006}; ^{Wien, 2006}). In melon (Cucumis melo L.), cucumber (Cucumis sativus L.) and zucchini squash, ethylene reduces the initial development of male flowers and decreases the male/female flower ratio, which is not true for watermelon (Citrullus lanatus (Thunb.) Matsum & Nakai) (^{Manzano et al., 2014}).

To increase yield and facilitate breeding programs, some studies suggest using ethylene-based growth regulators to achieve sex reversal in some cucurbits. The female floral buds require a minimal amount of ethylene to complete development and maturation, without occurrence of premature abortion (^{Manzano, Martínez, Gómez, Garrido, & Jamilena, 2010b}; ^{Manzano et al., 2011}; ^{Manzano et al., 2013}). Ethephon (Ethrel^®), when applied to plants, leads to immediate release of ethylene in plant tissues, causing suppression of male flowers, which, after aborted, are replaced by female flowers that develop on secondary buds (^{Nascimento, Pinheiro, & Freitas, 2007}). This induction of ginoicism, through use of ethephon, can be a promising technology, compared to other regulators, due to its positive effects and ease of application (^{Nascimento, Innecco, Medeiros-Filho, & Oliveira, 2005}). However, the intensity of femininity caused by ethephon may depend on the dose, number of applications, and environmental conditions (^{Friedlander, Atsmon, & Gal, 1997}).

The northern region of Brazil is a large producer of cucurbits, such as watermelon and melon. Other cucurbits, among them zucchini, have been gaining importance. The available cultivars are not adapted to the conditions of the northern region, such as the state of Tocantins that has a warm climate. To produce more efficiently, adapted cultivars should be developed or management techniques should be improved. The objective of this study was to verify the influence of doses and number of applications of ethephon on the sexual reversal of zucchini flowers.

Material and methods

The experiment was carried out from November 2013 to January 2014 in red latosol soil (^{Empresa Brasileira de Pesquisa Agropecuaria [Embrapa], 2013}), at the Vegetable Crops Sector of the Universidade Federal do Tocantins, Gurupi Campus, located at 11° 43’ 45’’ South latitude and 49° 04’ 07’’ West longitude, at 280 masl. The regional climate is humid with moderate water deficit, classified as Aw (^{Köppen, 1948}).

The experiment was established under a randomized complete block design with three replicates and a 4 x 4 factorial arrangement, plus a control (natural flowering). The first factor was ethephon dose (125, 250, 375 and 500 mg·L^-1), using the commercial product Ethrel^®, and the second was the number of applications (1, 2, 3 and 4).

The first ethephon application took place 10 days after seedling emergence (dae), and the rest every seven days. The cultivar Caserta was used in direct sowing, for which two seeds were placed per planting hole and a thinning was carried out when the plants had two true leaves, leaving one plant per hole. Raised beds were 0.7 m wide and 0.2 m high. Plant spacing was 0.5 m, in a triangular arrangement. Each plot consisted of five plants, and the three central plants were used for data collection.

Fertilization was with 150 g·m^-1 of a synthetic 5N-25P-15K and 20 g·m^-1 of micronutrient supply (FTE). Topdressing was at 15 dae, using 5 g of formulated 5N-25P-15K per plant. Sprinkler irrigation and all cultural practices were performed according to crop needs.

The doses of the different treatments were prepared from a stock solution, which was diluted to obtain the different concentrations. For all treatments, 2 L were prepared for application, and 1 % mineral oil was added to each one.

Six counts were conducted in male and female flowers at a three-day interval on six occasions. The first count was performed at 22 dae, and the last at 36 dae. Characteristics evaluated were number of male and female flowers, male/female flower ratio, fruit set rate and yield (Mt·ha^-1). Fruit set rate was calculated based on total number of female flowers converted into fruit.

The data were submitted to analysis of variance using the statistical program SISVAR (^{Ferreira, 2011}). As the number of female and male flowers did not meet ANOVA assumptions (independence, homoscedasticity and normal distribution), they were expressed by mean with standard errors. For the rest of the variables, regression models were produced to best explain variation in observed data.

Results and discussion

In the control treatment (without ethephon application), there were around twice as many male flowers, which usually occurs with natural flowering (Figure 1). Ethephon application changed the pattern of flowering, since for some treatments the number of male and female flowers were equal when the lowest dose was applied (Figure 1), while with the highest concentration of ethephon (500 mg·L^-1) more female flowers were produced in all applications. The results obtained agree with those reported by other authors in zucchini (^{Iozi, Rodrigues, Goto, & Ono, 1999}; ^{Yongan, Bingkui, Enhui, & Zunlian, 2002}), in C. moschata (^{Cardoso, da Silva, & Vecchia, 1998}), in Tetsukabuto hybrid squash (^{Nascimento et al., 2007}) and in melon (^{Girek et al., 2013}). ^{Daryono, Prasetya, Sumarlina, Sartika, and Subiastuti (2018)} confirmed that, from 75 ppm, ethephon can increase female and hermaphrodite flowers, inhibit male flower numbers, and increase flower diameter in some melon cultivars.

Figure 1 Number of male and female flowers in zucchini squash (Cucurbita pepo L.) ‘Caserta’ under different doses of ethephon and number of applications.  

In general, the ethephon dose affected the intensity of flower sex change. As the ethephon concentration increased, the greater the number of female flowers, and the lower the number of male flowers. A single application was enough to change the pattern of flowering, and the maximum sexual reversal occurred with two or three applications. ^{Manzano et al. (2011)} report that a single application would be enough to influence sexual expression because ethephon reduces the first phase of development of male flowers. However, ^{Peixoto, Casali, Almeida, and Seraphin (1989)} found that for pumpkin, C. pepo group Baianinha, at least two applications at 400 mg·L^-1 are required to reverse the sex of plants.

Although the 125 mg·L^-1 dose influenced zucchini sexual expression, mainly with four applications, the effect was not enough to induce the appearance of more female than male flowers. ^{Yongan et al. (2002)}, when applying ethephon to zucchini plants at the cotyledon stage, found that 50 mg·L^-1 was the most effective concentration for sexual reversal, and 100 mg·L^-1 caused injury or death of the plants, which was not the case in this work.

From the 250 mg·L^-1 dose upward, more female than male flowers were obtained, except for the 250 mg·L^-1 treatment with two applications and the 375 mg·L^-1 treatment with one application. The ethephon dose that resulted in the highest feminizing effect was 500 mg·L^-1 regardless of the number of applications. No dose provided a complete sexual reversal, i.e. plants with only female flowers, which agrees with the findings reported by ^{Shannoon and Robinson (1979)} and ^{Cardoso et al. (1998)} in zucchini.

^{Cardoso et al. (1998)} also verified that ethephon was not able to induce complete sex reversal, so they proposed using the doses and numbers of applications that resulted in a smaller number of male flowers in a plant that would be a female parent in F₁ hybrid seed production. This would help reduce the manpower required to manually remove the male flowers before anthesis, which would facilitate zucchini hybrid breeding.

As the ethephon concentration increased, there was a reduction in the male/female flower ratio (Figure 2). Ethephon application promoted a linear decrease in this trait in all treatments (dose and number of applications), which was due to an increased number of female flowers and a decreased number of male flowers, or the combined effect of these two elements.

Figure 2 Male/female flower ratio in zucchini squash (Cucurbita pepo L.) ‘Caserta’ as affected by dose and number of ethephon applications. Data transformed to √(x+1). 

The 500 mg·L^-1 dose, with three or four applications led to a lower male/female flower ratio. The effect was less marked when ethephon was applied once; however, a similar behavior to other treatments occurred, indicating that ethephon affects sex regulation in zucchini.

Under field conditions the percent of fruit setting would not be total due to environmental factors that interfere with pollination, plant health and phenotypic expression. Comparing the ethephon application at doses of 125 and 250 mg·L^-1, it was observed that ethephon affected the zucchini fruit setting even when the lowest concentration was applied (Figure 3).

Figure 3 Rate of fruit set in zucchini squash (Cucurbita pepo L.) ‘Caserta’ as affected by ethephon dose and number of applications. 

With increasing ethephon dose and number of applications, at higher concentrations, ethephon exerts no direct influence on conversion of female flowers in fruit. In addition, because the effect of ethephon is linked to its feminizing power, the amount of pollen and the likelihood of efficient pollination is reduced, as the male/female flower ratio is lower than normal, as well as the number of male flowers.

In general, ethephon application adversely affected zucchini yield. For one, three, and four applications quadratic behavior was observed, and a linear comportment for two applications occurred (Figure 4). Yield was related to number of fruit per plant, which was negatively affected by fruit set. Ethephon application is only an agronomically interesting alternative with four applications, since under this condition higher yields were obtained; therefore, if the objective is to increase the number of fruit, four applications can be recommended.

Figure 4 Yield of zucchini squash (Cucurbita pepo L.) ‘Caserta’ as affected by ethephon dose and number of applications. Data transformed to √(x+1). 

With a decrease in the male/female flower ratio, higher yields would be expected due to a greater number of female flowers; however, the yield of the different treatments was lower than that of the control. This may be related to the fruit set rate, which in turn may be related to an insufficient number of male flowers to pollinate all the flowers, or to the influence of ethephon on flower viability.

Conclusions

Use of ethephon influences sexual expression of zucchini, but does not promote a complete sex reversal. Ethephon application can be a useful tool for zucchini squash breeding as it reduces the need for removing male flowers in parents used as the mother line in directed crosses or seed production.