Content uploaded by John R. Duval
Author content
All content in this area was uploaded by John R. Duval on Sep 21, 2018
Content may be subject to copyright.
HORTSCIENCE 36(1):60–61. 2001.
Received for publication 10 Dec. 1999. Accepted
for publication 26 June 2000. A contribution of the
Univ. of Georgia Agricultural Experiment Stations,
Georgia Station, Griffin. This research was sup-
ported by state and Hatch Act funds allocated to the
Georgia Agricultural Experiment Stations. The cost
of publishing this paper was defrayed in part by the
payment of page charges. Under postal regulations,
this paper therefore must be hereby marked adver-
tisement solely to indicate this fact.
1Professor. E-mail address: snesmit@gaes.griffin.
peachnet.edu
2Assistant Professor.
Fruit Set of Triploid Watermelons
as a Function of Distance from a
Diploid Pollinizer
D. Scott NeSmith1
Department of Horticulture, Georgia Station, Griffin, GA 30223-1797
John R. Duval2
University of Florida, GCREC–Dover, 13138 Lewis Gallagher Road, Dover,
FL 33257
Additional index words. seedless watermelon, pollination, pollinizer ratio, Citrullus lanatus
Abstract. During 1998 and 1999, ‘Genesis’ triploid watermelons [Citrullus lanatus (Thunb.)
Matsum. & Nak.] were grown in large blocks with a single row of the diploid ‘Ferarri’
planted as a pollinizer in the middle. A once-over harvest each year was made in harvest
lanes 0, 1.5, 3.0, 4.5, 6.0, 7.5, and 9.0 m perpendicular distances from the pollinizer row.
Individual fruit were weighed and counted. Data from both years indicated a similar
distribution of triploid fruit with respect to distance from the pollinizer row. The greatest
number of triploid fruit per unit land area was in the harvest row 3.0 m from the pollinizer
row. When distance from the pollinizer row was 6.0 m or greater, triploid fruit numbers
diminished substantially. Yield estimates made each year using the fruit density data
suggested that a 1 pollinizer : 4 triploid ratio gave the maximum total triploid fruit yield
per hectare for 1.5-m row spacings. These results should prove useful in designing field
planting strategies to optimize triploid watermelon production.
watermelons should provide a good triploid
crop. Maynard and Elmstrom (1992) advised
that every third row of a triploid watermelon
field should be planted with a pollinizer. Par-
sons et al. (1992) recommended that one-third
of the plants in a seedless watermelon produc-
tion field be a standard (diploid) cultivar that is
dissimilar in appearance from the triploid
cultivar (for ease of identification). For the
most part, these recommendations have been
based more on experience than on research
findings. The objective of this research was to
determine pollinizer frequency requirements
for optimizing triploid watermelon produc-
tion per unit land area.
Materials and Methods
In 1998 and 1999, ‘Genesis’ (Shamrock
Seed Co., Salinas, Calif.) triploid watermel-
ons were grown in a field with ‘Ferrari’ (Sham-
rock Seed Co.) as a pollinizer at the Georgia
Experiment Station in Griffin. The soil was a
Cecil sandy clay loam (clayey, kaolinitic, ther-
mic Typic Hapludult). The design was a single
factor (distance), replicated experiment, which
consisted of a single row of the pollinizer
(‘Ferrari’) with six rows of the triploid (‘Gen-
esis’) planted on either side. Four large blocks
or replications of the planting design were
established each year. Row width used was 1.5
m, in-row spacing of plants was 1.2 m, and
individual rows were 24.4 m long. A buffer
distance was left between blocks so that the
closest distance to a pollinizer plant for any
triploid would be in its own block. All plants
were established using 5-week-old transplants.
Field planting dates were 20 May 1998 and 18
May 1999. Fertilization, pesticide applica-
tion, and irrigation practices recommended by
the Univ. of Georgia Cooperative Extension
Service were used in growing the crop (Mizelle,
1988). Pollination relied solely on natural pol-
linators; no supplemental bees were used.
Once-over harvests of ‘Genesis’ fruit
occurred on 4 Aug. 1998 and 10 Aug. 1999.
Each 1.5-m wide row was harvested sepa-
rately. All fruit >10 cm in diameter were
harvested and weighed. The harvest strategy
provided 1.5 × 24.4-m harvest lanes that had
centers 1.5, 3.0, 4.5, 6.0, 7.5, and 9.0 m per-
pendicular distances from the center of the
pollinizer row. The triploid watermelons that
were in the pollinizer row (due to vine growth
Triploid, or “seedless”, watermelons are
becoming increasingly popular among con-
sumers. Marr and Gast (1991) indicated that
consumers were willing to pay 50% more for
seedless watermelons than for traditional
seeded watermelons. Karst (1990) suggested
that there is a potential for seedless watermel-
ons to gain up to 50% of the market share.
Because of this enthusiasm of consumers for
seedless watermelons, growers are also be-
coming more interested in producing the spe-
cialty crop. Many traditional seeded water-
melon cultural practices can be utilized in
growing triploid watermelons, but there are a
few key differences. A major difference is that
triploid watermelons are sparse pollen pro-
ducers; therefore, a source of pollen (tradi-
tional diploid watermelon cultivar) must be
planted with triploids (Maynard and Elmstrom,
1992).
There are few published data on the
pollinizer frequency requirement for optimum
pollination and fruit production of triploid
watermelons. Kihara (1951) suggested that
one diploid pollinizer per four to five triploid
Fig. 1. Fruit number of ‘Genesis’ triploid watermelon as a function of distance from pollinizer row during 1998
and 1999 at Griffin, Ga. The pollinizer was ‘Ferrari’. Each vertical bar represents one 1.5-m-wide row. The
0 distance represents the location of the pollinizer row. Standard errors are depicted by vertical lines.
across rows) were also harvested (0 m from the
pollinizer). Harvest lanes of the specified dis-
tances were designated on each side of the
pollinizer row (east and west) to determine if
there was a “directional” effect on fruit density.
Results and Discussion
Overall, fruit yield for the rows of ‘Genesis’
plants varied depending on their distance from
the pollinizer row (Fig. 1). The patterns of fruit
density across the rows were similar for both
years. The pollinizer row itself (0 m distance)
had some triploid watermelons because of vine
growth from nearby plants. The first triploid
row adjacent to the pollinizer had a substantial
number of ‘Genesis’ watermelons, but the great-
est number of seedless watermelons per unit
area were present in the second row (3.0 m
distance) away from the pollinizer each year.
The number of triploid watermelons dimin-
ished by 37% to 40% from the second to the
third row (4.5 m distance). The remaining rows
(distances of 6.0, 7.5, and 9.0 m) set very few
fruit, suggesting that distances of 6.0 m or
Fig. 2. Estimates of yield (total fruit number/ha) of ‘Genesis’ triploid watermelon in response to different
pollinizer : triploid row ratios during 1998 and 1999. The pollinizer was ‘Ferrari’. Estimates are based
on 1.5-m row widths.
greater from the pollinizer row are too great for
adequate pollination, either because of the dis-
tance from the pollen source, or pollen dilution
and/or lack of bee visits.
Directionally biased pollen flow was not
apparent in the planting blocks of these experi-
ments. Fruit yields were similar for triploid
rows planted equal distances on either side of
the pollinizer row (data not shown). Also,
there were no differences in average indi-
vidual fruit weight or percentage of market-
able fruit (data not shown) from the different
harvest lanes, only in total fruit number per
unit area.
Using the fruit density results from the
1998 and 1999 field experiments, yield pro-
jections for different pollinizer : triploid ratios
were calculated (Fig. 2). Estimates of fruit
yield/ha increased with increasing number of
triploid rows up to a ratio of one pollinizer to
four triploid rows in both years. A trend to-
ward declining yields/ha was apparent as the
number of triploid rows per pollinizer row
increased beyond four. Triploid yield (fruit/
ha) for the current, commonly used ratio of 1:2
(every third row a pollinizer) was 25% less
than that resulting from the 1:4 ratio in 1998
and 1999.
The triploid yield estimates in these experi-
ments were based on 1.5-m row widths. This
would probably be very acceptable for triploid
watermelons, since vine coverage area tends to
be smaller than that of traditional seeded water-
melons. NeSmith (1993) reported that diploid
watermelons grown in 1.5-m-wide rows at in-
row spacings of 0.9 m yielded 29% to 34%
more than did watermelons planted at in-row
spacings of 2.1 m. When deciding on pollinizer
ratio using data from the current study, one
must consider that increasing row width be-
yond 1.5 m may slightly modify calculations of
yields. The current data indicated that when
distance from the pollinizer row approached
6.0 m, triploid fruit number declined sharply.
Therefore, if growers use 1.8-m row widths,
perhaps a ratio of 1:3 would be best.
In summary, triploid watermelons do re-
quire a pollinizer and adequate pollinator insect
activity for successful fruit set. Growers should
realize that exceeding distances of 6.0 m from
the diploid pollinizer will probably reduce
yields. Also, planting too many pollinizers
(i.e., ratios of 1:1 and 1:2) will reduce triploid
yields. The results presented here suggest that
triploid yields per unit land area would be
optimized using a ratio of 1 pollinizer row : 4
triploid rows for 1.5-m row widths.
Literature Cited
Karst, T. 1990. Seedless watermelon sure to grow.
The Grower 23(8):61.
Kihara, H. 1951. Triploid watermelons. Proc. Amer.
Soc. Hort. Sci. 58:217–230.
Marr, C.W. and K.L.B. Gast. 1991. Reactions by
consumers in a farmers’ market to prices for
seedless watermelon and ratings of eating qual-
ity. HortTechnology. 1:105–106.
Maynard, D.N. and G.W. Elmstrom. 1992. Triploid
watermelon production practices and varieties.
Acta Hort. 318:169–173.
Mizelle, W.O., Jr. 1988. Commercial watermelon
production. Coop. Ext. Serv. Publ. B-996. Univ.
of Georgia, Athens.
NeSmith, D.S. 1993. Plant spacing influences
watermelon yield and yield components.
HortScience 28:885–887.
Parsons, J., L. Stein, T. Longbrake, S. Cotner, and J.
Johnson. 1992. Seedless watermelon produc-
tion. Agr. Ext. Serv. Bul. L-2303. Texas A&M
Univ., College Station.