Pollination, fertilisation and fruit-set

Updated April 2018

High quality flowers on the female parent variety should be pollinated at the optimal time and under suitable environmental conditions. The likelihood of achieving success is increased by pollinating more than one (emasculated) flower per cluster.

Flower quality

Detailed guidance on how to increase flower quality and pollination rates in commercial apple orchards can be found in the ‘Apple Best Practice Guide’. Much of this is relevant to improving the success of hand pollination when making crosses.

  • Flowers formed on spurs are generally of higher quality than those on axillary shoots.
  • Flowers on spurs growing on two- or three-year-old wood often produce better quality flowers than spurs formed on older wood.
  • Flowers are initiated during the previous summer, and those on axillary shoots are initiated much later than flowers on spurs and short terminals. Consequently, they are less developed at the onset of dormancy in the late autumn, and are usually of poorer quality in the next spring.
  • Flower bud quality is usually improved by light, as opposed to severe, winter pruning. Summer pruning can improve flower bud quality on spurs by removing shading.
  • The first flowers to open on a variety are often those with the best setting and fruit growth potential.

Effective Pollination Period (EPP)

The EPP is a commonly used measurement of ‘flower quality’, defined as the number of days after the flower opens during which it can receive pollen and still set fruit. This period covers the time taken for the pollen to germinate, the pollen tube to grow and the time during which the ovule remains viable. EPPs are determined by measuring the success rates of pollinating flowers on the day they open and on each of the subsequent days. According to the ‘Apple Best Practice Guide’:

  • Good quality apple flowers should have EPPs of 3-5 five days under orchard conditions in the UK. 
  • Flowers only able to set fruits when pollinated on the day of opening have low EPPs and are classed as poor quality.
  • Spur or terminal flowers usually have longer EPPs than those formed as axillaries on one-year-old wood. The short EPPs of the latter are probably due to their shortened flower development. The risk of poor fruit set on axillary flowers is therefore higher than on spur or terminal flowers.
  • EPPs differ between varieties, and also vary from year to year.

The factors controlling EPPs are not fully understood but the ‘Apple Best Practice Guide’ advises that adopting best practice will induce stronger flowers, more viable pollen, better pollination conditions and a higher success rate at fertilisation. 

Optimal time of pollination

If you don’t know the EPPs of a variety, the default action is to pollinate flowers on the day, or day after, they open. According to MAFF (1973), recently opened flowers (i.e. one or two days from opening) ‘if pollinated by good compatible pollen, are likely to be more effective than older flowers in giving a reliable set of fruit’. Similarly, the ‘Apple Best Practice Guide’ states that highest pollination rates are usually achieved on the day the flowers open.


The condition of the stigma is often mentioned when discussing pollination. The best time to pollinate is traditionally considered to be when the surface of the stigma is ‘sticky’. Summarising from the ‘Apple Best Practice Guide’, the surface of the stigma consists of tiny papillae. These are turgid when the flower first opens (anthesis) and a surface secretion develops (stickiness). At this point the flower is considered to be receptive to pollen. The papillae dry out and collapse within two days. However, pollen has been shown to germinate on stigmas as late as ten days after flower opening (Braun and Stösser, 1985).  This suggests that ‘stickiness’ is not a prerequisite for successful pollination and fertilization. However, in general it is likely that the longer hand pollination is delayed once a flower is opened the lower the likelihood of success. 

Optimal conditions for fertilisation

The process of fertilisation comprises germination of the pollen grain on the stigma, the growth of the pollen tube down the style, its union with the ovule, and the subsequent fertilisation of the egg (♀) by one of the two sperm nuclei (♂) to form the embryo.The other sperm nucleus unites with the two polar nuclei in the ovule to form the triploid endosperm tissue. The embryo and endosperm both develop within the seed (pip). The fleshy, edible part of the developing fruit is actually the swelling of maternal ovary and receptable tissue. It is worth noting that this part only contains and expresses genes from the maternal parent.


The whole fertilisation process takes up to 10 days; two to four days for the pollen to germinate and the pollen tube to reach the base of the style, followed by six to eight days to cross the pericarp and unite with the ovule. Poor fruit set resulting from short EPPs is thought to be related to ovule longevity and growth of the pollen tube down the style, rather than to pollen germination itself.


Environmental conditions have a large impact on fertilisation. Pollen grains need to take up water (hydrate) in order to germinate. Hence, drying winds at the time of pollination will reduce their viability. However, pollen also loses viability if completely wetted, so rain will significantly reduce germination rates. Germination is also temperature dependent, with optimum temperatures of 15 - 25ºC, although pollen from some varieties can germinate and grow at much lower temperatures. The growth rate of the pollen tube is almost entirely dependent on temperature; if the tube does not grow down the style within 2-4 days fertilisation may not occur.

Levels of fruit-set and seed-set

An apple flower has five stigmas. Each stigma is connected via the stye to a carpel, each of which contains two ovules. It is not normally necessary for all ten ovules to be fertilized and produce seed for satisfactory fruit development, and with some varieties very few seeds appear to be necessary. However, many varieties develop misshapen fruit unless most of the cells of the core contain at least one fully developed seed (MAFF, 1972).


Surprisingly, we haven’t come across many figures for the success rate of hand pollination in terms of fruit-set. Trials with the variety Discovery at Long Ashton Research Station (Stott et al.,1973) showed that 9% of flowers set fruits if pollinated on the day of opening but only 1% set fruits when pollinated two days after opening. In a large scale trial (Brittain,1933) with Cox’s Orange Pippin, Golden Russet and Northern Spy, selfing produced a mean fruit set 1.5%, compared with 7% when cross-pollinated. Keulemans et al. (1994) hand pollinated 1, 3 or 5 stigmas per flower, and either 1 or 3 flowers per cluster. They harvested the resulting fruits when ripe and found that seed set increased with the number of pollinated stigmas on the flower, from 0.48 to 1.38 seeds per pollinated flower. Final fruit-set expressed as a percentage of flowers pollinated also increased from 24% to 39% with increasing numbers of pollinated stigmas.


The effects of pollinating 1 or 3 flowers per cluster were mixed. There was little effect on initial fruit set before the ‘June drop’. However, June drop increased from 5% (1 flower) to 21% (3 flowers). Consequently, final fruit-set decreased from 38% (1 flower) to 26% (3 flowers). On the other hand the average number of seeds/fruit harvested was 2.6 (1 flower) and 2.8 (3 flowers). Interestingly they also observed that the germination of seeds harvested from the fruit was negatively correlated with seed number, probably due to competition between seeds in the fruit.


At face value the results of Keulemans et al. (1994) call into question our practice of pollinating three flowers per cluster, in that a lower proportion of the pollinated flowers set fruit. However, our approach increases the chances of at least one of the three flowers on the cluster being in optimal condition when we pollinate the cluster. This should increase the chance of at least one set fruit from a particular cross. Logistically, we also find it quicker and easier to prepare, protect and hand pollinate three flowers per cluster, compared with one flower on each of three different clusters.  

Success rates at Ystwyth Valley Apple Breeders

Table of apple crossing success rates at YV apple breeders

Fruit-set from crosses made at YV Apple Breeders between 2013 and 2017 is shown in the table opposite. Fruit-set is defined here as the number of apples harvested at maturity. Total seed-set is the number of seeds extracted from harvested apples. Each cross usually comprises pollination of three flowers on a single cluster, although in a few cases less than three flowers are pollinated. The values in this table are based on the pooled results of crosses made between many different varieties each year, but include only ‘defined’ (both parents known) as opposed to ‘open’ (only female parent known) crosses. They show: 

  • The number of crosses producing one or more fruit at harvest time, expressed as a percentage of the total number of crosses made, were: 2013 (63%), 2014 (60%), 2015 (68%), 2016 (49%) and 2017 (33%), giving an average of 55% across the five years.
  • Averaged over the five years, 45% of crosses failed to produce any fruit, 34% produced one fruit, 17% produced two fruit, and only 4% produced three fruit.
  • Final fruit-set expressed as a percentage of the number of flowers pollinated was: 2013 (29%), 2014 (28%), 2015 (36%), 2016 (26%), 2017 (16%), giving an average of 27% across the five years. These values compare with the final fruit-sets of 38% (1 flower) and 26% (3 flowers) reported by Keulemans et al. (1994).
  • The average number of seeds (pips) per fruit harvested over the five years was 6.1. The value expressed on the basis of the total number of flowers pollinated was 1.7 seeds/pollinated flower. This compares with the 1.38 seeds per pollinated flower reported by Keulemans et al. (1994) when all five stigmas on a single flower were pollinated.

The wide variation in the results between years makes it difficult to accurately estimate the number of crosses required to produce a desired number of seeds. The unusually high failure rate (66%) for crosses made in 2017 is due mainly to frost damage of the flowers.

Effects of parental variety on success rates at YV Apple Breeders

Success rate can vary widely with variety, and in some cases also depends on whether the variety is used as male or female parent. We are compiling a long-term data-set on varietal differences, scoring each cross (3 flowers pollinated) as ‘successful’ if at least one fruit survives to final harvest. ‘Failures’ are classed as either ‘failed pollination’, where no fruit is set from the outset, or ‘failed post-pollination,’ where any remaining fruit are lost during or after the June drop. The average success rates across all varieties used in five or more crosses between 2010 and 2017 are 50%, when used as the female parent, and 55%, when used as the male parent.


The figure below compares a sample of 20 varieties used as female parents between 2010 and 2017.  

Graph of success rates of different apple varieties used as female parents in crosses between 2010-2016
  • Most diploid varieties have success rates >50% when used as female parents.
  • Good female parents include Beauty of Bath, Discovery, Ellison's Orange, Lord Lambourne and Spartan.
  • The majority of failures are associated with pollination and/or fertilisation (64%) rather than losses during June drop and later (36%) .
  • Proven triploid (or aneuploid) varieties (e.g. Adam’s Pearmain, Annie Elizabeth. Ashmead’s Kernel) have low success rates as female parents, associated with failed pollination or fertilisation.
  • Sunset and Scarlet Pimpernel have relatively low success rates for diploids, mainly associated with failed pollination or fertilisation.  

 The figure below compares a sample of 20 varieties used as male parents between 2010 and 2017.  

Graph of success rates of 20 apple varieties used as male parents in crosses 2010 to 2016
  • Most varieties have success rates >50% when used as male parents.
  • Good male parents include Discovery, Gwell Na Mil and Honeycrisp.
  • Some triploid (or aneuploid) varieties like Adam’s Pearmain have relatively high success rates as male parents, whilst others like Annie Elizabeth have poor success rates.

Beauty of Bath is a good example of a variety with widely different success rates depending on whether it is used as the female (56%) or male (28%) parent. In contrast, Katy and Ellison's Orange have very similar success rates irrespective of whether they are female or male parents.  


Braun J, and Stösser R. 1985. Narben- und Griffelstruktur ihr Einfluss auf Pollen keimung, -schlauchwachstum und Fruchtansatz beim Apfel. Angew. Bot. 59: 53-65.

Keulemans J, Eyssen R, and Colda G. 1994. Improvement of seed set and seed germination in apple. p. 225-228. In: H. Schmidt and M. Kellerhals (ads.). Progress in

temperate fruit breeding. Kluwer Academic, Dordrecht, Netherlands.

MAFF. 1972. Apples. Ministry of Agriculture, Fisheries and Food, Bulletin 207.

HMSO, London. 205p.

MAFF. 1973. Flowering Periods of Tree and Bush Fruits. Ministry of Agriculture, Fisheries and Food, Technical Bulletin 26. HMSO, London. 76p.

Stott KG, Jefferies CJ, and Jago C. 1973. Pollination and fruit set of the apple Discovery. Report of Long Ashton Research Station for 1972, 31.