Euphytica 32: 685-689 (1983)

Institute for Horticultural Plant Breeding (IVT), P.O. Box 16, 6700 AA Wageningen, the Netherlands
Received 17 April 1983


To improve fruit- and seed set in roses, the Hybrid Tea-cultivar Sonia was pollinated with the cv. Ilona 0, 1. 2, 3, 4, 5, 6 and 7 times at 24 h intervals. Unpollinated flowers did not yield fruits. The number of achenes per fruit increased up to 3, fruit weight and weight of achenes increased up to 4, and the number of seeds per pollinated flower (PI) increased up to 5 pollinations.

Both between and within pollination treatments, highly significant correlations occurred between fruit weight, number of achenes per fruit, and weight of achenes. Effects of repeated pollination in rose and apple are compared. The effect of a relatively low fertilization level in rose crosses is discussed.


Rose breeders frequently complain about the low seed production in crosses between rose cultivars or, a still lower production when species roses are involved. Even when every precaution has been taken and glasshouse environment seems optimal, the number of seeds obtained per pollinated Hybrid Tea-flower lies only between 5 and 15 (DE VRIES, unpublished data).

To improve both fruit- and seed set, at IVT a number of experiments are being carried out. The present study deals with the effect of repeated pollination on fruit- and seed set in 'Sonia' x 'Ilona' crosses.


Mother plants consisted of 5-year-old 'Sonia' bushes on R. canina 'Inermis', planted in a heated greenhouse.

Flowers were emasculated when the petals had partly opened, just before ripening of the anthers. After removal of the petals and subsequent emasculation, stigmas were pollinated with previously collected and stored (VISSER et al., 1977) 'Ilona' pollen. There were 8 treatments, involving 25 flowers each, in which the flowers were: 0, 1, 2, 3, 4, 5, 6 and 7 times pollinated at 24 h intervals. Pollinations were made with a No 5 artist's brush, and carried out daily between 8.00 and 10.00 h over a 10 day period, from May 1982 onwards. To prevent fruit rot, flowers were dusted with Eupareen M, 2 days after the end of each treatment. At the time of pollination, of 15 'Sonia' flowers, the number of pistils was counted.

Fig. 1. The percentage of fruit set () and the number of achenes per fruit () in 'Sonia', after repeated pollination with 'Ilona',

Minimum temperature in the greenhouse was adjusted at 20 °C, maximum temperatures, owing to solar irradiation, might rise to 35°C. In October and early November, the ripe (orange coloured) fruits were harvested. The individual weight of each fruit and the total weight and number of its fully developed achenes ('seeds') was determined.

For this study, the seed set per pollinated flower, 'pollination index' (PI) is used to express the overall efficiency of a pollination treatment (VISSER & VERHAEGH, 1980a).


The number of pistils in 'Sonia' flowers, each leading to one ovule, varied between 85 and 125 (mean: 103.4 ± 13).

Emasculated, but unpollinated flowers yielded no fruits (Table 1). The number of days to fruit ripening did not differ significantly between pollination treatments. When the number of pollinations increased from 1 to 5, the number of fruits increased from 12 to 23, but pollinating 6 and 7 times did not further improve fruit set. Fruit weight, the number of achenes per fruit, and the weight of the achenes per fruit, increased from 1 through 4 pollinations per treatment. PI increased from 4.1 in 1 pollination, to 15.6 in 5 pollinations; in 6 and 7 pollinations, PI did not further increase.

Table 1. Fruit- and seed set characters after various number of pollinations, made at 24h interval, in 'Sonia' x 'Ilona' crosses (25 flowers per treatment).

Number of
Days of
fruit ripening
of fruits
weight (g)
Number of
per fruit
Weight of achenes
per fruit (g)
0 0
1 148.8 12 4.52 8.6 0.53 4.1
2 148.7 15 5.67 15.8 0.70 9.5
3 149.3 16 6.15 17.2 0.80 11.0
4 149.2 18 7.37 17.3 0.97 12.5
5 150.7 23 6.82 16.9 0.87 15.6
6 152.0 22 7.41 16.7 0.94 14.7
7 149.9 21 6.19 16.8 0.82 14.1

Table 2. Correlations between the fruit weight, the number of achenes per fruit and the weight of achenes per fruit, both
within and between 7 'Sonia' x 'Ilona' crosses, obtained when flowers were 1, 2, ...., 6 or 7 times pollinated at 24h interval.

Parameters Correlation*
    total within crosses between crosses
    (n =121) (n = 7)
Fruit weight number of achenes x fruit-1 0.82** 0.88* 
Fruit weight weight of achenes x fruit-1 0.92** 0.99**
Number of achenes x fruit-1 weight of achenes x fruit-1 0.88** 0.91* 

 * Correlations significant for p = 0.01* or 0.001**, respectively.

Fig. 1 illustrates that the number of achenes per fruit is stabilized 2 pollinations sooner than the fruit set; the parameters increasing up to 3 and 5 times pollinating respectively.

Table 2 shows a number of highly significant correlations, demonstrating that irrespective of the number of pollinations, as fruits were heavier they contained more, and a greater weight of achenes; when the number of achenes per fruit increased, their total weight did so too. Similarly, as pollination treatments resulted in greater fruit weights, they yielded a larger number and a greater weight of achenes; as treatments induced more achenes per fruit, total achene weight increased correspondingly.


Results showed that repeated pollination in the cross 'Sonia' x 'Ilona', improved both fruit set and the number of achenes per fruit. Compared with one single pollination, 5 successive pollinations were most effective, as they about doubled the fruit set (48% vs. 92%) and produced an almost fourfold number of achenes per pollinated flower (4.1 vs. 15.6).

It is realized that the application of repeated pollination may encounter practical objections. However, as mother plants available for breeding are usually scarce, pollinating more-than-once is advisable as it makes the most of the number of flowers available.

Surprisingly little research on the effects of repeated pollination has been carried out. For the genus Rosa no references are available, but in the related genera Malus and Prunus fruit set has been found to improve by 3-4 successive pollinations (KONDRAT'EV et al., 1972). In Pyrus, trees pollinated 2-3, or 5-6 times by honey bees, yielded 1-2, respectively 3-5 times more fruits than those pollinated once only (PANOV & PETKOV, 1975).

In apple and pear, VISSER & VERHAEGH (1980b), VISSER & MARCUCCI (1983) recently showed that in many cultivars a double pollination induced about twice as many seeds per pollinated flower as a single pollination. Owing to a suitable marker pollen, they were able to demonstrate that the second pollen induced 2-3 times as many seeds as the first. They named the first: 'pioneer' pollen, and suggested that, partly at its own cost, the pioneer pollen paved the way for the second one.

Although carried out with the same pollen, present results show that in rose too, the second pollen may almost double the number of achenes (seeds) per fruit (8.6 vs. 15.8). It is unclear as yet whether this increase is due to the pioneer effect of the first pollen, or to a larger number of stigmas that become receptive as the flower ages. In apple where the number of ovules per flower is usually about 10 (CRANDALL, 1919; KRUMBHOLZ, 1935), even double pollination did not induce on average more than 7-8 seeds per fruit (DORSEY, 1921; VISSER & VERHAEGH, 1980b), or about 80% fertilization in the fruits harvested. In case of 'Sonia', where a mean of 103 ovules per flower was counted, maximum number of achenes per fruit was approximately 17.2, i.e. about 17% fertilization in crosses with 'Ilona'. Although 'Sonia' shows different affinity towards the pollens of various cultivars, at most 37.6 achenes per fruit ('Sonia' x 'Hadley'), i.e. 36% fertilization, were observed (DE VRIES, unpublished data). This means that in apple, but to a much larger extent in rose, notwithstanding sufficient pollen and repeated pollination a high percentage of ovules does not develop into seeds.

In the compatible cross 'McIntosh' x 'Delicious' apples, BRYANT (1935) showed that, just after pollen tubes had entered the embryosac and effected fertilization, embryos might already abort. AUBERTIN (1959) found egg-breakdown to occur as soon as 18 h after pollination, embryo degeneration from 11 days after pollination, and seed abortion over the whole period of apple fruit ripening.

In general, embryo abortion in apple and pear may occur owing to poor nutrition of the flower, competition between ovules, chromosomal aberration, or to genetic unbalance between micro­- and megaspores (HOWLETT, 1932; BRYANT, 1935; DOWRICK, 1958). These factors are likely to play a role in rose too. Owing to their large number per flower, competition between ovules for nutrition, may be the main cause for relatively few seeds per fruit.



Seed Size, Growth and Yield