Genetics 1
The Cucurbits: Cucumber, Melon,
Squash and Pumpkin
H.C. Wien
Department of Fruit and Vegetable Science, Cornell
University, 134A Plant
Science Building, Ithaca, New York 14853-5908, USA
© CAB INTERNATIONAL 1997. The Physiology o f Vegetable
Crops
The species in the family Cucurbitaceae that have been
used as vegetables have enriched and diversified the
diets of humankind for many centuries. For example,
cucumber (Cucumis sativus) is consumed as an
ingredient in salad, or in the pickled form. The fresh
fruits of muskmelon or cantaloupe (Cucumis melo) and
watermelon (Citrullus lanatus) are eaten as desserts,
while the fruit flesh of the squashes and pumpkins is
generally consumed after boiling or baking. The
immature fruit of summer squash or marrows (Cucurbita
pepo), and the mature fruit of all the principal
squash and pumpkin species (C. argyrosperma, C.
moschata, C. maxima, and C. pepo) are harvested and
used as food. Shoot tips of the cucurbit vegetables
are eaten as a cooked vegetable in Southern and
Eastern Africa. Seeds of watermelon and some squash
and pumpkin species are roasted and eaten as snacks,
or ground as an ingredient of sauces. Several
cultivars of C. pepo with hull-less seeds have been
developed that facilitate the food uses of the seeds.
With an oil and protein content of 46 and 34%,
respectively, these could be exploited as alternative
oil-bearing crops (Whitaker and Davis, 1962). In North
America, pumpkins and gourds (Cucurbita pepo) are
frequently grown solely for the ornamental value of
the fruits, in harvest displays and as part of the
Halloween celebration.
The major cultivated cucurbits can be classified into
new world and old world species with regard to their
origin. Cucumbers are thought to have been first
cultivated in India, where their use has been recorded
as long as 3000 years ago (Whitaker and Bemis, 1976).
Cucumis melo is thought to have arisen in the central
part of Africa, and have spread rapidly into Asia,
where many cultivars have since been selected. The
watermelon originated in the dry parts of southern
Africa, and also has South Asia as a second centre of
diversity. The new world species include all the
cultivated Cucurbita. Although the four most
economically important species may have had a common
ancestor in one location in the Americas, such
evidence is difficult to find. Each of these species
has been selected separately, so that interbreeding of
different species is for the most part difficult to
achieve. From archaeological records, the Cucurbita
species are amongst the most ancient of cultivated
crops in the Americas. Indeed, squash was one of the
principal components of the diet of the ancient Mayan
civilizations, together with beans and maize, dating
back as far as 10,000 years (Whitaker and Bemis,
1976).
At the present time, the cucurbit vegetables are
cultivated in all major regions of the world (Table
9.1) (FAO, 1994). Watermelon leads the production
figures both in terms of tonnage and land area (1.8
million ha). Production in Asia comprises half or more
of the total area devoted to each cucurbit vegetable
worldwide.
Our knowledge of the physiology of the cucurbits has
not grown in proportion to their importance in
production. Most effort has been devoted to cucumber,
followed by Cucumis melo, and watermelon in third
place. As in the case of Capsicuni (Chapter 7),
detailed understanding of the growth and productivity
of cucumber has come recently from the need to
optimize production in greenhouse environments. In
addition, intensive investigations into the hormonal
control of flower sex expression during the 1960s and
1970s facilitated the production of hybrid cultivars
of cucumber and squash. More recently, interest in
developing melon cultivars of better fruit quality has
spurred research efforts in fruit carbohydrate
metabolism. Altogether, the state of our physiological
knowledge of the cucurbit crops can be described as
rather uneven, with detailed understanding in some
areas, and others virtually unexplored. The latter is
particularly true of watermelon, squash and pumpkins,
perhaps due to the fact that their sprawling plant
habit make them difficult experimental subjects in
greenhouse, growth chamber and field.
Table 9.1. Production of the major cucurbit vegetables
in the world in 1993 (production: 1000 metric tons)
(FAO, 1994).
Region Watermelon Cucumber Melons Pumpkins and
squash
Asia 15,746 12,761 7726 3552
Europe 2447 2563 2367 1184
NC America 2069 971 1696 428
Africa 2036 400 878 926
South America 994 59 233 768
World 27,063 18,326 12,976 8019
The
GERMINATION AND SEEDLING GROWTH
The germination of cucurbit vegetable seeds requires
relatively warm temperatures (Lorenz and Maynard,
1980), and takes place within 3 or 4 days at 25-30°C.
For cucumber, the lower limit of germination has been
shown to be 11.5°C (Simon et al., 1976), while
germination of muskmelon and watermelon was low at
16°C (Nelson and Sharpies, 1980). Summer squash
(Cucurbita pepo) germination showed a lower threshold
between 5 and 10°C, and optimum germination between 30
and 35°C (NeSmith and Bridges, 1992). Considerable
effort has been expended on improving the capability
of cucumber seeds to germinate at low temperatures.
Soaking the seed in acetone solutions of fusicoccin
(0.5 mmol), or GA4+7 (a mixture of gibberellins 4 and
7) (1 mmol), was the most effective in stimulating
germination at 12°C (Nelson and Sharpies, 1980).
Similar growth regulator treatments had no significant
effect on rate or final percentage emergence in six
field plantings into cool soils, however (Staub et
al., 1987). Imbibing or pregerminating cucumber seeds
in water at 32°C before planting them in soil at 15°C
also did not shorten the time of emergence, although
plant stand was improved significantly (Staub et al.,
1986). These results indicate that several steps of
the germination and emergence process are limited by
low temperatures. Simon et al. (1976) conjectured that
the low temperatures may cause a denaturation of
proteins in the germinating seedling, resulting in
more severe damage the longer the exposure.
Considerable genetic differences exist in cucurbits in
capacity to germinate at low temperatures (Schulte and
Grote, 1974; Nerson et al., 1982; Nienhuis et al.,
1983). After four cycles of selection for germination
at 15°C, Nienhuis et al. (1983) improved low
temperature germination from 32 to 94%. This increase
was also correlated with improved germination at
higher temperatures. In muskmelon, Nerson et al.
(1982) showed 'bird's nest' cultivars developed in
Iran to have significantly better cold temperature
germination than standard viny and dwarf cultivars
developed in the USA. We do not know yet what
physiological processes differ among lines of either
species contrasting in cold temperature germination.
The low germination rate of cucumber at 15°C may be
partly due to seed dormancy (Nienhuis et al., 1983).
Freshly harvested seed failed to germinate at this
temperature, and remained dormant until it had been
stored for 84 days. Watts (1938) had also encountered
this phenomenon in 'Black Diamond' cucumber, and was
able to overcome the inhibition by removing the
seedcoat, or germinating the seed at 30°C. The
presence of a germination inhibitor in the testa was
further supported by the finding that soaking cucumber
seed in acetone significantly improved cold
temperature germination (Nelson and Sharpies, 1980).
These workers also showed that watermelon seed
germination at 16°C could be improved by washing the
seed for 2 h in water. While dormancy may play a role
in the reduced germination of fresh cucumber seed at
low temperatures, it must be emphasized that poor
germination under cool soil conditions is also a
problem with cucurbit seed that has passed the
apparent rest period.
Another approach to overcoming the growth-inhibiting
effects of cold soils is to graft cucumber onto
rootstocks of species less susceptible to cold root
temperatures (den Nijs, 1980a). This approach, now
widely practised in Japan and Korea, where about 80%
of greenhouse cucumbers are grafted, allows cucumber
and other cucurbits to be grown without soil heating.
Another benefit from the practice is the avoidance of
problems of soil-borne diseases, to which the
rootstock is not susceptible, and the improvement in
fruit quality that comes with the use of certain
rootstocks (Lee, 1994). More particularly with regard
to root growth in cold soils, Tachibana (1982, 1987)
found that Cucurbita ficifolia, one of the most common
rootstocks used for cucumber, maintains active growth,
water and nutrient uptake at 12 and 14°C (Fig. 9.1).
At these temperatures, these processes are sharply
reduced
120
110
100
°_.., 90 .v-~. _
80 0 v Fig-leaf gourd
3ucu r
ucu
0 70 .:. Cuc/gourd
.
60
50
40
10 12 1416 18 20 22 24 26 28 30 32
120
B
100 p.---.____
80 ! . .v.
;` v
c
-17
60
0 U d
40
aD J
20
0
10 12 1416 18 20 22 24 26 28 30 32
Temperature (°C)
Fig. 9.1. Root weight (A) and leaf phosphorus content
(B) of fig-leaf gourd, cucumber or cucumber grafted on
a fig-leaf gourd rootstock, in relation to root
temperature. Values are percentages relative to the
values of each treatment at 20°C (Tachibana, 1982).
in cucumber. Differences in susceptibility to low soil
temperatures between two cucumber cultivars were
particularly well related to differences in phosphorus
uptake at those temperatures (Fig. 9.1) (Tachibana,
1982, 1987). This may suggest that P uptake is one of
the key processes inhibited by low root temperatures.
Genetic variation in tolerance to low temperature in
cucumber has been found both in the seed germination
stage (Nienhuis et al., 1983), as well as in the adult
vegetative and fruiting stages of the plant (den Nils,
1980b). Good progress has been made in selecting
greenhouse cucumber lines that grow vigorously and
produce acceptable yield at 20/15°C (day/night),
temperatures at which current commercial lines are
much reduced in growth. The improved lines maintained
a high leaf area ratio at low temperature during the
vegetative stage, and showed no delay in fruit
development after flowering. It should thus be
possible through breeding and selection to develop new
cultivars that will have significantly lower energy
requirements in greenhouse production. Incorporation
of these genes into field-grown cultivars may widen
the area of adaptation of the crop.
FLOWER DIFFERENTIATION
Before describing the development of flowers in the
cucurbit vegetables, it is necessary to understand the
range of sexual types that commonly occur in these
crops (Table 9.2). Most common is the monoecious
flowering habit, in which male and female flowers are
found on the same plant. In cucumber, many cultivars
with gynoecious flowering have been developed.
Homozygous gynoecious lines usually produce only
female flowers, but in the heterozygous form, male
flowers can frequently be found, especially under
environmental conditions favouring male flower
expression (see below). A range of other sexual types
have been developed in cucumber, such as the
hermaphroditic (perfect flowers), androecious (male
flowers) and the andromonoecious (male and perfect
flowers), but these have not
Table 9.2. Floral morphologies of cucurbit vegetables.
Species Flowering type
Cucumis sativus Monoecious, most common
Gynoecious
Hermaphroditic
Andromonoecious
Cucumis melo Andromonoecious, most common
Monoecious
Cucurbita pepo Monoecious
Citrullus lanatus Monoecious, most common
Andromonoecious
become important in cucumber production. In muskmelon,
the predominant flowering form in commercial cultivars
is andromonoecious, although some monoecious,
gynoecious, androecious and hermaphroditic forms also
exist (Rudich, 1990). The monoecious sexual type is
most prevalent in Cucurbita spp. and in watermelon,
but cultivars differ in ratio of male to female
flowers, and in the lowest node to bear female
flowers.
Flowers are formed early in the seedling stage in
cucumber and melon, and can be found on low main stem
nodes. Morphological studies have found that male,
female and perfect flowers arise from anatomically
similar 'presexual' primordia, from which subsequently
the male, female or both parts develop fully (Atsmon
and Galun, 1960; Goffinet, 1990) (Fig. 9.2). The
regulation of sexual development is under genetic,
environmental and hormonal control, and has been
intensively studied for the last 40 years. As with
other aspects of cucurbit physiology, cucumber has
been investigated most, with the other species getting
much less attention.
The sex type of flowers in cucumber and in the species
of cucurbits having monoecious flowering habit shows a
distinct ontogenetic pattern (Nitsch et al.,1952;
Shifriss and Galun,1956). The basal main stem nodes
are generally male, and these become interspersed with
increasing number of nodes bearing female flowers
(Fig. 9.3). At the upper part of the main stem, the
plant may eventually form a zone of nodes bearing only
female flowers.
GENETIC FACTORS IN SEX EXPRESSION
Detailed analyses of the genes responsible for sex
expression in the cucurbits have been summarized by
Frankel and Galun (1977), and by Lower and Edwards
(1986). Even without knowing the details of the
genetics of sex
60
0
0
aD
0
h
3
0 J
30
20
10
Female flower zone
Male flower zone
12 14 16 18 20 22 24 26 28 3032
Mean temperature (°C)
Fig. 9.3. Influence of mean temperature on the
transition from male to female flowering nodes on
'Acorn' squash (Cucurbita pepo) (Nitsch et al., 1952).
n
expression, it is useful to realize that cultivars
termed 'gynoecious' can be either homozygous or
heterozygous for that trait, and these differences can
have a profound effect on the flowering pattern of the
plant. Homozygous gynoecious lines of cucumber produce
male flowers at very low frequency, and in many
situations will only produce female flowers (Lower et
al., 1983). In contrast, heterozygous gynoecious
lines, especially those developed from crosses between
gynoecious and monoecious lines, produce male flowers
on a significant proportion of their nodes (Cantliffe,
1981; Lower et al., 1983; Lower and Edwards, 1986).
Cultivar differences can also exist for frequency of
male flowering nodes among heterozygous gynoecious
lines. Many cucumber hybrids used in field production
are heterozygous with regard to the gynoecious
character, and are termed 'predominantly female' by
the seed trade. These cultivars show considerable
influence of the environment on sex expression, as
shown below (Cantliffe, 1981; Nienhuis et al., 1984).
ENVIRONMENTAL EFFECTS ON SEX EXPRESSION
Influence of the environment on sex expression can be
most clearly shown by focusing on the zone of
transition between male and female nodes, and
measuring the node number at which the first female
flower is borne, or the ratio of male to female
flowers in a constant number of nodes. Since
monoecious plants tend to bear increasing number of
female flowers with time, it is important to compare
treatments over the same period of time. In assessing
the effect of temperature, or other factors affecting
growth rate, the number of nodes on which the effect
is counted should be kept constant (Matsuo, 1968).
This helps avoid some of the errors in interpretation
found in some of the early literature (Tiedjens,
1928).
At least three environmental factors have an important
influence on sex expression in cucumber and Cucurbita
pepo: temperature, light energy and photoperiod. Under
cool conditions, the production of female flowers is
favoured (Fig. 9.3). Mean temperatures are most
important, but night temperatures also play a
significant role, with warm nights leading to
increased male flower production at a given mean
temperature, compared to warm days (Nitsch et al.,
1952). The temperature influence may occur during
flower primordia differentiation, as in cucumber, or
during the development of the flower to anthesis. In
Cucurbita pepo, low temperature may inhibit male
flower development after differentiation, leading to
precocious female flowering (see section on precocious
female flowering under Physiological Disorders below)
(Rylski and Aloni, 1990; H.C. Wien, Ithaca, NY, 1990,
unpublished data). NeSmith et al. (1994) also found
that the number of male and female flowers reaching
anthesis could be influenced by temperature in a
cultivar-dependent manner.
High light conditions generally favour female flower
production, while shading, or low incident radiation
delays the onset of female flowering (Ito and Saito,
1960; Kooistra, 1967; Cantliffe, 1981). The
combination of low light and high temperatures
prevalent in the fall growing period in the southern
US may explain the predominance of male flowers in the
cucumber crop at that time (Cantliffe,1981).
The effect of photoperiod on sex expression appears to
be less striking than the factors of temperature and
light conditions in most cultivars (Matsuo, 1968;
Cantliffe, 1981). Short photoperiods tend to favour
the production of female flowers. It is difficult
under field conditions to avoid an interaction of
photoperiod and light energy levels, with short
photoperiods coinciding with periods of reduced light
conditions. In that situation, it appears that light
energy plays the more important role, as in the
example cited by Cantliffe (1981) in the preceding
paragraph. If photoperiod is controlled by keeping the
amount of light energy constant and extending
daylength with low intensity light, some cucumber
cultivars are day neutral with regard to sex
expression, and others show delayed female flowering
ilong days (Matsuo, 1968).
The sex expression of cucumbers in the field generally
follows the predictions from controlled environments.
For instance, Shifriss and Galun(1956) found that
planting in the relatively cooler, and shorter day
length conditions of April in Israel resulted in
plants with a higher proportion of female flowers than
when the same cultivars were planted in July. Where
conditions appear to conflict, temperature tends to be
more important. Novak (1972) found that 'Tablegreen'
monoecious cucumber produced female flowers earlier
under the cool, long-day conditions of Ithaca, New
York summers than in the hot, short-day conditions of
the Philippines i(Table 9.3). NeSmith et al. (1994)
demonstrated that planting date tends to affect female
flower number more markedly for some summer squash
cultivars than for others.
Nitrogen status of the plant has also been shown to
influence sex expression in cucumber, with high
nitrogen fertilizer levels delaying the production of
female flowers (Ito and Saito, 1960). High plant
population density, or close spacing within rows also
increases male flowering, and may operate through
reduced light levels available to individual plants in
crowded stands (Lower et al.,1983; Nienhuis et
al.,1984).
~c
Table 9.3. The influence of growing conditions on
node to first female flower, and node
number of start of female flower zone in `Tablegreen'
cucumber, grown in the field in Ithaca,
NY, and Los Banos, Philippines. Mean temperatures of
21 and 27°C, and daylengths of 16 and
;-:, 11.5 h prevailed during the growing seasons at
Ithaca and Los Banos, respectively (Novak,
1972).
Location Node to first female flower Node of start of
female zone
Ithaca, USA 8.3 ± 4.0 13.4 ± 4.1
~;jf~ Los Banos, Philippines 16.6 ± 1.8 31.7 ± 7.4
po
' ~I
In summary, the environmental conditions which
encourage the buildup of carbohydrates and which
reduce the amount of vegetative growth tend to favour
female flower expression. Conditions which foster stem
extension and reduce carbohydrate build-up, such as
high temperatures, low light conditions, high nitrogen
levels and close spacings, also increase the tendency
for male flower production in the cucurbit vegetables.
GROWTH HORMONES
Plant growth regulating compounds play a key role in
determination of flower gender in the cucurbits.
Evidence for their importance comes both from studies
in which endogenous levels of the growth hormones are
related to sex expression, and from the effect of
applying the compounds exogenously to the plant. Aside
from helping us understand the mechanism of flower sex
development, use of growth regulators to manipulate
sex expression has facilitated the development of
hybrid cucurbit cultivars, in ways that will be
illustrated below.
Gibberellin
The production of male flowers on nodes of cucumber
that would normally produce female flowers can be
brought about by the application of gibberellic acid
(GA3), and even more effectively, by GA4+7 (Mitchell
and Wittwer, 1962; Atsmon and Tabbak, 1979). Similar
results have been obtained with cucumber, Cucurbita,
and muskmelon (Splittstoesser, 1970; Rudich et al.,
1972a; Atsmon and Tabbak, 1979).
Higher levels of endogenous gibberellins have been
found in monoecious and andromonoecious cucumbers than
gynoecious cultivars, in parallel to their tendency to
produce male flowers (Hemphill et al., 1972).
Environmental conditions that favoured male flower
development, such as high temperature and long
daylengfh also increased the amount of gibberellins
detected in the apical regions (Saito and Ito, 1963).
In muskmelon, levels of endogenous gibberellins were
less well correlated with male tendency in one study
(Hemphill et al., 1972). In another experiment,
however, the increased femaleness of andromonoecious
melons brought about by applications of the growth
inhibitor SADH (succinic acid -2,2-dimethylhydrazide,
or daminozide) was matched by lower gibberellin
content (Rudich et al., 1972a).
Ethylene
The capacity of ethylene to stimulate female flower
development in the
cucurbits first became known when researchers applied
the ethylene-pro-
ducing chemical ethephon (2-chloroethylphosphonic
acid) to cucumbers,
(Robinson et al., 1969). When applied to seedling
monoecious cucumber
plants, the chemical eliminated male flowers on the
lower nodes, and increased the number of female
flowers. Ethephon stimulated the formation of female
flowers on monoecious muskmelon, without much change
in male flower numbers, while on andromonoecious and
hermaphroditic plants there was an increase in perfect
and suppression of male flowers (Karchi, 1970). On
monoecious Cucurbita pepo, the chemical suppressed
male and increased female flowers. Ethephon
application has thus become useful to ensure that the
inbred line used as the female parent in hybrid
combinations does not develop male flowers, and is
used for this purpose in both cucumber and squash
breeding (Shannon and Robinson, 1979; Lower and
Edwards, 1986).
The central role of ethylene in cucurbit sex
expression has been further strengthened by the
finding that inhibitors of ethylene formation or
action have effects on flower formation that are
opposite to those of ethephon. For instance, treatment
of homozygous gynoecious cucumbers with the ethylene
action inhibitor silver nitrate or the ethylene
synthesis inhibitor aminoethoxyvinylglycine (AVG)
resulted in the formation of male and perfect flowers
(Atsmon and Tabbak, 1979). Silver nitrate and silver
thiosulphate are used by cucumber breeders to produce
inbred, all-female lines (Lower and Edwards, 1976).
Treated gynoecious plants produce some male flowers
and permit selfing, without the detrimental effects of
elongated, brittle stems brought about by gibberellin
application.
Watermelon appeared to be more sensitive to ethephon
applications than the other cucurbits, and the crop's
reaction was opposite to that found for cucumber
(Christopher and Loy, 1982). Exposure to
concentrations of 30 ul 1-1 retarded female flower
formation, but had less effect on male flowers. Use of
silver nitrate or AVG on watermelon resulted in a
suppression of female and perfect flowers, with only a
partial reduction of male flowers (Christopher and
Loy, 1982). These anomalous responses to hormones with
regard to sex expression indicate that further
clarification is needed on how flower gender is
determined in watermelon.
Determinations of endogenous ethylene levels in
cucurbit seedlings have supported the regulatory role
of this chemical in sex expression. In cucumber,
seedling apices of gynoecious lines produced more
ethylene than those of monoecious or androecious lines
(Rudich et al., 1972c, 1976). Ethylene evolution was
higher for female than for male flower buds, and
monoecious lines increased their ethylene evolution at
the time that female flower primordia were developing.
Short daylengths stimulated ethylene evolution in
comparison to long daylengths, in concert with their
influence on female flower formation (Rudich et al.,
19720.
Auxin
It has been known since the early days of plant
hormone research that auxin is involved in sex
expression of cucurbits, but its exact role is still
uncertain (Rudich,1990). Treatment of young cucumber
plants with auxin or synthetic
y
auxins such as naphthalene acetic acid promoted female
flower formation (Ito and Salto, 1960; Salto and Ito,
1963). Culture of potentially male excised flower buds
in auxin-containing medium stimulated ovary formation
(Galun et al., 1963).
The endogenous auxin level was increased in some
experiments by conditions fostering female flower
formation (Rudich et al., 1972b), while the level
decreased in others (Ito and Saito, 1960; Saito and
Ito, 1963). Furthermore, treatment of Cucurbita pepo
with ethephon increased female flower number, but
decreased endogenous auxin activity (Chrominski and
Kopcewicz, 1972). A clear role for auxin is difficult
to determine, in part because higher levels of auxin
cause the liberation of ethylene in tissues. In
addition, ethylene has been shown to inhibit auxin
translocation, and to contribute to the inactivation
of auxin through decarboxylation (Beyer and Morgan,
1969). Finally, the reliable determination of
endogenous auxin levels in plants has been difficult,
and has contributed to the uncertainty of the;role of
this hormone in sex expression of cucurbits.
Abscisic acid (ABA)
There is considerable uncertainty about the role of
ABA in sex expression ofcucurbits (Rudich, 1990). ABA
applications to gynoecious cucumbers increased female
tendency, but treatment of monoecious lines favoured
male flower production (Rudich et al., 1972b;
Friedlander et al., 1977a). In one study,
ABA content of gynoecious cucumber was higher than
that of a monoecious
line (Rudich et al., 1972b); in another, the reverse
was true (Friedlander et al.,
1977b). The ABA content of both types of cucumber were
increased three- to
ten-fold by applications of ethephon. It is thus not
clear if ABA plays a direct
role in sex expression, or has a secondary involvement
in response to the
more important action of the previously mentioned
hormones.
Summari'
Research since the 1960s has identified the
gibberellins as the main growth
hormone group that stimulates male flower development
in Cucumis and
Cucurbita. Little is known about the biosynthetic
pathway of the gibberellins
in the cucurbits, although many researchers have found
GA 4 and GAS to be
more active than GA
3. Ethylene stimulates female flower formation in
these
cucurbits, and suppresses the development of male
flower buds. We lack a comprehensive understanding of
the interaction of these hormones.
Although ethephon applications lead to reduced
gibberellin levels
(Chrominsky and Kopcewicz, 1972; Rudich et al.,
1972b), it is not known if male flowering tendency
enhanced by ethylene inhibitors operates by
increasing gibberellins. The roles of auxin and of ABA
also need further
investigation, to determine if these play primary or
secondary roles in determination of gender of cucurbit
flowers. For this purpose, it may be rewarding to
repeat the tissue culture experiments of presexual
flower buds
conducted by Galun et al. (1963), and manipulate
ethylene and ABA levels as well as auxin and
gibberellins investigated by them.
FLOWERING AND FRUIT SET
The time of flowering of the cucurbit vegetables is
primarily determined by temperature, and its influence
on plant growth rate. Temperature is also the
principal factor determining the time of anthesis and
duration of opening of individual flowers. Seaton and
Kremer (1938) found that the flowers of Cucurbita had
a minimum temperature for anthesis and anther
dehiscence of around 10°C. Above this level, flowers
would open at dawn, and remain open until about noon.
Under cooler conditions, anthesis and anther
dehiscence was delayed until the following day. As
temperatures increased beyond 30°C, anthesis occurred
earlier, and flowers closed by mid or late morning.
The minimum temperature for opening of cucumber and
watermelon flowers was found to be about 15°C, whereas
muskmelon anthesis temperatures were between 18 and
21°C (Seaton and Kremer, 1938). The duration of flower
opening for cucumber, watermelon and muskmelon is
generally for the entire daylight period of one day.
The receptivity of the female flower, or of the female
portion of perfect flowers of cucumber has been found
to extend from 2 days before to 2 days after anthesis
under growth chamber conditions (Le Deunff et al.,
1993). In the greenhouse, Munger (1988) reported that
manual crosses were successful on the day of anthesis
and the following morning, but under temperate field
conditions, manual pollination success often decreased
to low levels on the afternoon of the anthesis day.
The factors leading to this rapid decline were not
identified.
POLLEN GERMINATION AND POLLEN TUBE GROWTH
When pollen from the same plant or another plant of
the same species is deposited on the stigmatic
surface, pollen germination follows in less than 30
min under normal conditions (Suzuki, 1969; Sedgley and
Buttrose, 1978).
'Cucumber pollen germinates over a wide range of
temperatures, but pollen
tube growth rates may be inhibited at the extremes,
preventing fruitset
(Matlob and Kelly, 1973). As temperature increased
from 10 to 32°C, pollen
tube elongation rate increased in snake melon (Cucumis
melo var. flexuosus),
but cucumber pollen tubes were only stimulated up to
21°C (Matlob and
Kelly, 1973). Elongation rates were also higher for
pollen from plants grown
under high light and moderate temperature conditions
than from plants
produced in low light conditions.
There are also considerable genetic differences in
pollen tube growth rates (Table 9.4). In general,
these are higher for the cucurbits with larger ovary
and final fruit size, and may be related to the size
of pollen grains of
The Cucurbits: Cucumber, Melon, Squash and Pumpkin 359
Table 9.4. Rates of pollen tube growth and pollen
grain diameters of vegetable cucurbit species,
compiled from several authors.
Pollen Pollen tube Time to
Temp. grain growth ovule
Species (°C) diameter, u' (mm h-') penetration (h)
Reference
Cucumber 21 63 - 36 Matlob (1973)
Muskmelon - 53 0.95 24 Suzuki (1969)
C. melo var.
flexuosus 21 36 Matlob (1973)
Watermelon 30 52 6 24 Sedgley (1978)
Watermelon 43 52 2.5 > 7 Sedgley (1978)
Cucurbita - 142 4.6-6 9-11 Suzuki (1969)
(Ikuse, 1956).
these species (Ikuse, 1956). Nevertheless, the tube
growth rates are sufficiently rapid to ensure that the
pollen tubes reach the nearest part of the ovary
within a few hours. Poole and Porter (1933) calculated
that the tips of the pollen tubes of watermelon would
reach the nearest ovules in 3 h. Suzuki (1969) arrived
at a figure of 5 h for muskmelon in similar
calculations. In spite of this, most researchers have
found that fertilization of the ovules takes from 24
to 36 h (Table 9.4), indicating that other factors
than pollen tube growth rate are involved in this
step.
In some cucumber cultivars, the rate of pollen tube
growth may not be rapid enough to allow fertilization
of ovules over the entire length of the ovary (den
Nijs and Miotay, 1991) (Fig. 9.4). At the same time as
the pollen tubes are growing, the ovary is also
elongating, and on some long-fruited cucumber
cultivars, the middle and far end of the fruit may
never be reached by the pollen tubes. As a result, the
fruit enlarges at the blossom end, and relatively few
seeds are formed (Varga and Bruinsma,1990).
The path of the pollen tubes to the ovary and ovules
is primarily along the conducting tissue connecting
the style and ovary, but once the ovary is reached,
pollen tubes will also travel in the cavities between
fruit lobes (Poole and Porter, 1933). The pattern of
distribution of pollen on the stigmatic surface has
little influence in Cucurbita on the distribution of
fertilized seeds in the fruit, indicating that the
pollen tubes can travel laterally in the style or
ovary to some extent (Hayase, 1953; S.W. Cady and H.C.
Wien, 1994, Ithaca, NY, unpublished). In watermelon,
pollen distribution on the stigma does affect seed
location, and ultimately, fruit shape (Mann, 1943).
The necessity of ovule fertilization and seed set for
fruit set in nonparthenocarpic cucumber remains
unclear. Fuller and Leopold (1975) found that style
removal 12 h after pollination allowed 50% of fruits
to set, while ovule penetration required 30-36 h. They
speculated that stimulation of the ovary to continue
growth did not need the pollen tubes to reach the
ovules, but did not identify the nature of the signal.
More recently, Varga and
360 H.C. Wien
140
Ovary length
120
Longest pollen tubes
100
(~ Main front of tubes
_E 80
60
J
40
20
0 1 2 3 4 5 6 7
Cultivars
Fig. 9.4. Length of the ovary, the longest pollen
tubes and the main front of the pollen tubes growing
down the style, for six cucumber cultivars of
differing fruit length (den Nijs and Miotay, 1991).
Bruinsma (1990) demonstrated that style removal after
12 h allowed some ovules to be fertilized and some
fruit to set. They found that cucumber cultivars
incapable of parthenocarpic fruit set had to at least
be induced to form non-viable 'pseudoseeds' (seeds
with seedcoats but lacking embryos and endosperm).
These were induced to form in their experiments by use
of irradiated pollen that was sterile but capable of
pollen tube growth to the ovules. The mechanism by
which true seeds and pseudoseeds stimulate the
continued growth of the ovary is not known, although
it is presumed to be hormonal.
PARTHENOCARPY
Parthenocarpy is fruit set and fruit growth without
the fertilization of ovules. It occurs widely among
the cucurbit vegetables, particularly in cucumber
(Rudich et al., 1977) and in Cucurbita pepo (Rylski,
1974; Robinson, 1993). The tendency to set and develop
seedless fruit is enhanced by cool weather conditions,
and to a lesser extent, by short photoperiods (Rylski,
1974; Rudich ct al., 1977; Dean and Baker, 1983). In
cucumber, parthenocarpic tendency is more strongly
expressed in lines that have a higher proportion of
female flowers. Even on monoecious lines, the tendency
for parthenocarpic fruit production increases with
plant age, as does the femaleness of the plant (Rudich
et al., 1977; Kim et al., 1992a).
The Cucurbits: Cucumber, Melon, Squash and Pumpkin 361
There is considerable genetic variation in
parthenocarpy in cucumber, and the characteristic has
been viewed as a means of overcoming adverse
environmental effects on fruit set or pollinating
insects. To date, parthenocarpic gynoecious cultivars
are common in greenhouse cucumber production, but have
not been successfully developed for field production
(Lower and Edwards, 1986).
The production of seedless fruits in watermelon does
not occur without special measures, since naturally
occurring parthenocarpy has not been found in this
species (Mohr, 1986). Fruit set without seed set can
be brought about by pollinating a self-sterile
triploid watermelon with a diploid pollen parent
(Kihara, 1951).
Commercial production of seedless watermelon has been
hampered by several factors. The abnormally thick
seedcoat of the triploid seeds impedes germination and
makes the use of transplants advisable. In addition,
yields are decreased because one quarter of the field
must be planted to a diploid pollinator cultivar.
Consumer reluctance to accept that the small white
empty seedcoats found in these parthenocarpic fruit
are not seeds has also hampered acceptance of seedless
watermelon cultivars.
HORMONAL REGULATION OF FRUIT SET
Indications of how the retention and growth of fruits
in the cucurbit vegetables might be regulated have
come from studies of endogenous hormones in flowers
and fruits, and from the induction of fruit set by
exogenous hormone applications.
Auxin
Comparisons of parthenocarpic and normal seeded
cucumber lines have generally shown that ovary auxin
content is higher in parthenocarpic ovaries at
anthesis (Kim et al., 1992a,b; Takeno and Ise, 1992).
Pollination increased ovary auxin concentration of the
non-parthenocarpic line, while auxin level in
unpollinated ovaries declined further (Kim et al.,
1992b). The exact site of auxin production in the
developing ovary has not been identified, but may be
in both the pericarp and ovule tissues.
The importance of auxin in cucumber fruit set is
supported by the findings that fruit set induced by
application of exogenous growth regulators such as
benzyl adenine (a synthetic cytokinin), gibberellic
acid, synthetic auxins and auxin transport inhibitors
all increase the endogenous auxin content of the
developing ovary (Beyer and Quebedeaux, 1974; Kim et
al., 1992x; Takeno and Ise, 1992). A number of
investigators have also brought about parthenocarpic
fruit set in cucumber by application of synthetic
auxins to the flower (Elassar et al., 1974x; Watkins
and Cantliffe, 1980; Kim et al., 1992b). Similar
results were obtained with muskmelon (Whitaker and
Prior, 1946; Elassar et al., 1974b), Cucurbita pepo
(Wong, 1941; Y. Zhang and H.C.
362 H.C. Wien
Wien, 1988, Ithaca, NY, unpublished data) and other
cucurbit vegetables (Wong, 1941).
Increasing fruit set of pickling cucumbers by the use
of the auxin transport inhibitor chlorflurenol has
been demonstrated in greenhouse and field experiments
(Cantliffe et al., 1972, 1974; Dean and Baker, 1983;
Ells, 1983). It was thought that simultaneous set of
many fruits could be achieved by using the chemical on
gynoecious cucumber cultivars that had been delayed in
fruit set by lack of male flowers. Unfortunately,
commercial realization of this technique failed
because it was difficult to prevent male flower
formation even on strongly gynoecious genotypes. The
few male flowers that occurred in such plantings were
sufficient to allow early set of a few fruit, which
then inhibited further fruitset on the plants (Ells,
1983).
Cytokinins
Evidence that endogenous cytokinins may be involved in
fruit set of cucumber has been largely indirect.
Exogenous application of benzyl adenine and other
cytokinins has been shown to increase fruit set, and
stimulate fruit growth (Elassar et al., 1974a; Ogawa
et al., 1990; Shishido et al., 1990; Takeno et al.,
1992). Results indicated that in one case, the fruit
set stimulation was accompanied by an increase in
endogenous auxins (Shishido et al., 1990), but in
another instance, no correlation with endogenous
auxins could be found (Takeno et al., 1992). Takeno et
al. (1992) measured increased ovary cell numbers in
treated flowers, and speculated that the improvement
in fruit set may have been brought about by
cytokinin-stimulated cell division. There is need for
direct measurement of ovary cytokinin levels in
cucumber lines differing in parthenocarpic fruit set
tendency to clarify the role of these growth promoting
compounds in the fruit set process.
Gibberellins
The involvement of gibberellins in fruit set of
cucumber has also been deduced from the result of
exogenous application experiments. Elassar et al.
(1974a) and Ogawa et al. (1989) showed that
application of GA 3 and more effectively, GA4+7,
increased fruit numbers. As with the cytokinins,
detailed measurements of endogenous gibberellins in
the flower and developing fruit are still needed.
POLLINATION
The cucurbit vegetables are among the vegetable crop
species that require insects for pollination. This is
most obvious for the crops that have separate male and
female flowers, such as cucumber, watermelon, squash
and pumpkin. But even in muskmelon (Cucumis melo)
which bears perfect flowers, pollinators are
necessary. In this species, the pollen is sticky and
not
i
Fig. 9.5. Male flowers of Cucurbita pepo, with petals
removed to show the degree of opening of the nectary
slits, from closed (left), to open (right) (Cady,
Glatz and Wien, Ithaca, NY,
unpublished observations).
readily transferred, so that exclusion of bees by
caging plants nearly totally prevented fruit set
(McGregor and Todd, 1952).
It is beyond the scope of this book to delve into the
foraging behaviour
of honey bees and gourd bees, the principal
pollinators of the cucurbit veg-
etables. The physiology of the plant may, however,
influence its attractive-
ness to pollinating insects. For instance, the
domestic bee Apis mell i fera uses
the flowers of cucurbits as both a source of pollen
and of nectar (Free, 1970).
The principal attraction in muskmelon flowers appears
to be nectar.
Plantings of a nectarless mutant had only sporadic
visits by bees, and as a
consequence, poor fruit set and low yields (Bohn and
Mann, 1960; Bohn and
Davis, 1964), even though pollen production by the
mutant was normal.
Similarly, accessibility of the staminal nectaries in
Cucurbita pepo may influ-
ence the frequency of visits by honey bees (S.W. Cady,
R. Glatz and
H.C.Wien,1994, Ithaca, NY, unpublished observations)
(Fig. 9.5).
For successful pollination, both male and female or
perfect flowers must be open on the same day. In
Cucurbita pepo, cool weather conditions early in the
growing season may result in female flowers opening
several days before male flowers, resulting in delayed
fruit set (Rylski and Aloni, 1990) (see section on
physiological disorders below). There are preliminary
indications that low light conditions, and high
temperatures have the opposite effect (R.O. Nyankanga
and H.C. Wien, Ithaca, NY, 1996, unpublished data).
Female flower primordia on some cultivars fail to
develop to anthesis under these conditions, thus
delaying or decreasing fruit production. Little is
364 H.C. Wien
known about the regulatory mechanisms of flower
development in the cucurbit vegetables, but it may
have a similar basis as the hormonal regulation of sex
expression (see section on flower differentiation
above).
FRUIT GROWTH
The development of fruit in the cucurbit vegetables
has been an object of fascination for many years.
Attention has focused on the large-fruited members of
Cucurbita pepo and C. maxima, which can reach
extraordinary size. For instance, contests organized
by pumpkin-growing clubs are currently striving to
break the 1000-pound (454 kg) barrier, and fruit
weights of 375 kg have recently been achieved
(Langevin, 1993) (Fig. 9.6). Such fruits are said to
gain as much as 11 kg per day while growing, but the
physiological attributes needed to bring about such
prodigious size increases have been largely
unexplored.
Fruit growth of smaller-fruited cucurbits was closely
studied by Sinnott more than 50 years ago (Sinnott,
1939, 1945). Ovary growth of Cucurbita pepo before
anthesis was found to consist both of cell division
and cell enlargement activities (Fig. 9.7). By
plotting the diameter of individual cells against
ovary diameter, Sinnott (1939) demonstrated that the
transition from a I
1
Fig. 9.6. Fruit of Cucurbita maxima weighing 377 kg
(829 lb), winner of the 1995 'World m
Pumpkin Federation Weigh-off' for largest squash (H.C.
Wien, Ithaca, NY, unpublished data). fr
__________________________________________________
Do You Yahoo!?
Yahoo! Auctions - buy the things you want at great prices
http://auctions.yahoo.com/
---------------------------------------------------------------------
Pumpkin-growing FAQ: http://www.mallorn.com/lists/pumpkins/search.cgi
To sign-off this list, send email to majordomo@mallorn.com with the
message text UNSUBSCRIBE PUMPKINS
