A fruit tree in full bloom is a remarkable sight - and a deeply frustrating one if the blossom falls without setting any fruit. Pollination failure is the most common reason for abundant flowering and absent harvest, and it is almost always preventable once you understand what the tree actually needs. This article covers the biology of fruit tree pollination from first principles: how flowers work, why most varieties cannot pollinate themselves, how compatibility is determined genetically, what pollinators are doing and when they can't do it, and how to plan a planting that reliably sets a full crop.
This is part of a four-article series on fruit tree growing. The other articles cover general husbandry and care, pruning techniques and the science behind them, and grafting. If you want to assess whether specific varieties are compatible with each other before you plant, the Fruit Tree Planner is designed for exactly that.
The Flower: Anatomy of a Pollination
To understand pollination you need to start with the flower. A typical fruit tree flower - an apple blossom is a good model - consists of five white or pink petals, five sepals beneath them, a central pistil, and a ring of stamens surrounding it. The stamens are the male structures: each consists of a slender filament topped by an anther, which is the pollen-producing organ. The pistil is the female structure: it consists of a stigma (the receptive surface at the top), a style (the stalk connecting it to the base), and the ovary at the bottom, which contains the ovules that will become seeds if fertilisation is successful.
Pollen is not a single cell. It is a two-celled structure - the male gametophyte - that contains the genetic material needed for fertilisation. When a pollen grain lands on the stigma of a receptive flower, it germinates: it produces a pollen tube that grows down through the style toward the ovary. This tube is the delivery mechanism. The generative cell inside the pollen grain divides to produce two sperm cells, which travel down the tube. One sperm fertilises the egg cell in the ovule to produce the embryo (the seed); the other fuses with the central cell of the ovule to produce the endosperm, which is the seed's food reserve. This double fertilisation is characteristic of flowering plants and is what makes a fertilised ovule develop into a viable seed.
Fruit development is triggered by the production of hormones, particularly auxins, in the developing seeds. A fruit with no viable seeds produces little auxin and typically drops prematurely, before it has sized up. This is why poor pollination results not just in no fruit, but in small, early-dropping fruit: the hormonal signal to the tree to sustain fruit development is weak or absent. In apple, each fruitlet contains up to ten ovules; a fruitlet that has only one or two fertilised ovules will be smaller, less symmetric, and more likely to drop than one with seven or more.
Self-Incompatibility: Why Most Varieties Cannot Pollinate Themselves
Most fruit tree varieties are self-incompatible: they cannot set fruit using their own pollen or the pollen of genetically similar varieties. This is not a deficiency - it is an evolved mechanism to prevent inbreeding and maintain genetic diversity. The mechanism by which incompatibility works at the molecular level is one of the better-understood systems in plant genetics.
In apples, pears, plums, and cherries, self-incompatibility is controlled by a single chromosomal region called the S-locus. Every individual plant carries two alleles at the S-locus, called S-alleles, one inherited from each parent. Each S-allele encodes two proteins that work together: a protein expressed in the stigma (the receptor) and a protein expressed in the pollen (the ligand). When pollen lands on a stigma, the stigma recognises the pollen's S-allele. If that S-allele matches either of the stigma's own S-alleles, the stigma rejects the pollen: it produces compounds that block the growth of the pollen tube in the style, usually within the first fraction of an inch of growth. The pollen grain germinates but the tube never reaches the ovary. No fertilisation occurs.
This system is called gametophytic self-incompatibility (GSI), because incompatibility is determined by the haploid genome of the pollen grain (the gametophyte) rather than the diploid genome of the whole plant. The practical consequence is that two varieties sharing even one S-allele are mutually incompatible as pollinators: each rejects the other's pollen just as it would reject its own. For cross-pollination to succeed, the two varieties must carry entirely different S-allele pairs.
Sweet cherries illustrate this particularly clearly. They share the same GSI system, and for most of the twentieth century the compatibility relationships between varieties were understood mainly through trial, error, and observation in orchards. The underlying S-alleles were identified and mapped from the 1990s onward, and this work revealed why certain historically popular combinations - Williams Bon Chretien paired with Conference pear, or Stella as a near-universal cherry donor - worked so well, and why others consistently failed despite appearing to bloom at the same time.
Self-Fertile Varieties
Some varieties are self-compatible: they can fertilise their own flowers and set fruit without a cross-pollinator. Self-compatibility usually arises from a mutation in one of the S-allele proteins that prevents the stigma from recognising and rejecting self pollen. The mutation is typically in the pollen S-allele rather than the stigma side, which means a self-compatible variety can still be pollinated by other compatible varieties and can act as a donor for them - but it doesn't need to.
Self-fertile apples include James Grieve, Discovery, Greensleeves, and Scrumptious. Among pears, Conference is notably self-fertile and can set a meaningful crop even in isolation, though its fruit is more reliably sized and shaped when cross-pollinated. Most European plums are at least partially self-fertile; Victoria, Czar, and Marjorie's Seedling are reliable self-setters. Morello cherry (an acid cherry) is self-fertile. Most sweet cherries historically were not, though modern breeding has produced self-fertile sweet cherries including Stella, Sunburst, and Lapins.
It is worth noting that "self-fertile" does not always mean "does just as well alone as with a partner." Many nominally self-fertile varieties set noticeably heavier, more consistent crops when a compatible cross-pollinator is nearby. Conference pear and Cox's Orange Pippin apple are both partially self-fertile, but both are significantly more productive with a compatible partner in range. Self-fertility should be understood as a floor - a minimum performance - rather than an optimum.
Pollination Groups and Flowering Time
For cross-pollination to occur, two varieties must be in flower at the same time. Nurseries and fruit research stations have grouped varieties into flowering time categories - typically numbered 1 (earliest) through 6 or 7 (latest) for apples - based on when they reach peak bloom under average conditions. Varieties in the same group, and usually those in adjacent groups, have sufficient overlap in their flowering window to pollinate each other reliably.
Varieties two or more groups apart may flower without any overlap at all, particularly in a season where early varieties are advanced by warm weather and late varieties are delayed by a cold snap. Planting only early and late varieties without anything mid-season is a common mistake that guarantees poor pollination.
The groups are a guide, not an absolute rule. Exact bloom timing shifts with temperature, elevation, aspect, and year-to-year weather variation. A tree in a warm, south-facing location may bloom a week earlier than the same variety on a north-facing slope. What the groups capture is the stable relative order: a Group 2 apple will always bloom before a Group 4 apple in the same conditions, even if the absolute dates shift. For practical purposes, choosing varieties within the same group or adjacent groups is reliable; varieties two groups apart carry meaningful risk of inadequate overlap.
Pears, plums, cherries, and other species each have their own parallel grouping systems with the same logic. Many nursery catalogues list the group number alongside each variety, making it straightforward to select compatible combinations.
Triploids: The Special Case
Most fruit trees are diploid: they carry two sets of chromosomes (one from each parent), and their cells divide normally during the formation of pollen and ovules, producing viable haploid gametes. Some apple varieties, however, are triploid: they carry three sets of chromosomes, which arose when a diploid and tetraploid (four-set) parent crossed in breeding. The three chromosome sets cannot pair up correctly during meiosis (the cell division that produces pollen), so the pollen grains produced are chromosomally abnormal and non-viable.
This has a critical practical consequence: a triploid apple cannot be used as a pollinator for any other variety. Its pollen does not fertilise. The triploid tree itself can receive pollen and set fruit normally - it just cannot give any in return. This means that planting a triploid with only one other variety leaves both trees without a functional pollinator: the triploid provides no pollen to its companion, and the companion provides pollen to the triploid but has nothing else to receive pollen from.
The solution is to plant a triploid with at least two compatible diploid varieties, so that the two diploids can pollinate each other as well as the triploid. The triploid benefits from both; the two diploids are each other's partners; and all three trees carry a useful crop.
Well-known triploid apples include Bramley's Seedling (the classic English cooking apple), Blenheim Orange, Jupiter, Jonagold, and Gravenstein. These are all highly regarded varieties, which is why it matters to know about their pollen sterility before planting. A lone Bramley, however magnificent its eventual size, will be deeply unproductive without appropriate partners.
Cherry Compatibility: A Complex System Simplified
Sweet cherries had, for most of the twentieth century, one of the most complex and least understood compatibility systems of any commonly grown fruit. Many popular varieties were found, through decades of orchard observation, to be simply incapable of pollinating certain other popular varieties - even when they bloomed simultaneously. The explanation, once the S-allele work was done, was straightforward: those pairs shared S-alleles and rejected each other's pollen.
Compatibility groups for sweet cherries - in which all members share at least one S-allele and cannot pollinate each other - were defined by researchers from the 1940s onward and are still used in nursery catalogues. Within any one compatibility group, all varieties are mutually incompatible. Varieties from different groups may or may not be compatible; it depends on their specific S-allele combinations. Some varieties, historically called "universal donors," were found to successfully pollinate almost everything - Stella, for example, produces pollen that is compatible with virtually all other sweet cherry varieties, and it is also self-fertile, making it one of the most useful cherries a gardener can plant for pollination purposes.
The picture has been simplified considerably by modern breeding. Self-fertile sweet cherry varieties - Stella, Sunburst, Lapins, Sweetheart, Regina in the right conditions - can set adequate crops without any cross-pollinator, which makes them far more practical for gardens where there is space for only one or two trees. Morello and other acid cherries are self-fertile and also compatible as pollinators with many sweet cherry varieties within overlapping bloom periods, though the reverse (sweet cherry pollen on Morello) is less reliable.
Plum and Gage Compatibility
European plums (Prunus domestica) and most gages and damsons are self-fertile to varying degrees. Victoria, the most widely grown English plum, is reliably self-fertile. Czar, Marjorie's Seedling, and Opal are others that set well alone. Greengage varieties are more variable; Cambridge Gage is partially self-fertile, but most other gages set better with a compatible partner.
Japanese plums (Prunus salicina) are generally self-incompatible and need a cross-pollinator from within their own species. European and Japanese plums do not cross-pollinate each other reliably: they flower at different times (Japanese plums typically bloom earlier) and the pollen is not functionally compatible between the two species in most cases.
Where a plum has a specific cross-pollination requirement, the nursery catalogue will usually note it. In practice, for most garden situations involving European varieties, the risk of complete pollination failure is lower than with apples or sweet cherries - but planting two compatible varieties still reliably increases crop size and consistency.
Pears and Quince
Pears follow the same gametophytic self-incompatibility system as apples, with S-allele based incompatibility groups. Conference is the notable exception: a mutation in its pollen S-allele makes it self-compatible, and it can also act as a pollinator for many other varieties. Williams Bon Chretien (Bartlett) and Doyenne du Comice are popular pear varieties that are self-incompatible and have specific compatibility requirements; Comice in particular benefits greatly from a partner, with Conference being the most commonly recommended match.
An additional complication with pears is that some variety pairs are cross-incompatible despite being in the same flowering group. Williams Bon Chretien and Seckle, for instance, are mutually incompatible despite overlapping bloom periods. This is another instance of shared S-alleles. Always check specific compatibility rather than assuming that same-group varieties will work together.
Quince (Cydonia oblonga) is generally self-fertile and does not require a cross-pollinator, though yield often improves with more than one plant present.
The Pollinators: Bees and the Conditions They Require
Pollen does not travel from anther to stigma by itself in the case of fruit trees - it requires a vector, and in the vast majority of practical cases that vector is an insect. Bees are the primary pollinators of orchard crops. Understanding bee behaviour and the conditions that affect it explains a great deal about why pollination succeeds or fails in a given season.
Honey bees are the most familiar, and the most commonly thought of, but they are not necessarily the most effective pollinators of fruit trees. Solitary bees - mason bees (Osmia species), mining bees (Andrena species), and others - visit more flowers per unit time, carry a higher proportion of pollen on their bodies relative to their size, and forage at lower temperatures than honey bees. A single red mason bee (Osmia bicornis) has been estimated to do the pollination work of several dozen honey bees in a comparable period. Bumblebees are also significant early-season pollinators, able to fly at lower temperatures than honey bees, which makes them valuable during the early bloom of plums, damsons, and cherries.
Temperature is the most important environmental variable governing bee activity during bloom. Honey bees generally do not forage below about 50°F; solitary bees and bumblebees can operate down to around 41 to 46°F. In a cold spring, when temperatures struggle to reach double figures during the week that cherries or pears are in bloom, pollination can be severely reduced even when both bee populations and compatible varieties are present. There is nothing to be done about temperature in the field, but understanding it explains years of poor fruit set that might otherwise seem mysterious.
Rain during bloom is also seriously detrimental. Bees do not forage in rain. More critically, rain washes pollen from open anthers before it can be transferred, and wet stigmas can be physically impeded in pollen tube germination. A single week of cold, wet weather during peak bloom can eliminate most of a season's fruit set regardless of how well everything else is managed. Stone fruits - particularly cherries and apricots, which bloom earliest and whose blossom windows are relatively narrow - are most exposed to this risk.
Increasing Pollinator Presence
The three most effective things you can do to support pollinator populations around your fruit trees are: provide nesting habitat, avoid pesticide applications during bloom, and maintain a supply of forage plants before and after fruit tree bloom.
Solitary bees nest in the ground (mining bees) or in holes in wood and hollow stems (mason bees and leafcutter bees). A patch of bare, undisturbed ground in a sunny position provides mining bee habitat. A bee hotel - a box containing tubes of various diameters made from bamboo, paper, or drilled wood - provides nesting sites for mason bees. These are not novelties; they produce measurable increases in local solitary bee populations when sited appropriately.
Avoiding insecticides during bloom is non-negotiable if you want fruit. Any contact insecticide - including those marketed as "organic" or "natural," such as pyrethrin - kills bees that contact treated surfaces. If pest pressure during bloom is severe, use the most targeted and least persistent option available, apply only at dusk when bees are not flying, and never spray open flowers. The loss of a year's pollination from a single misapplied spray during flowering is catastrophic compared to almost any pest damage you might be trying to prevent.
Planting flowers that provide pollen and nectar before and after the fruit tree bloom season supports overwintering bumblebee queens emerging in early spring and builds populations of solitary bees that will be active at bloom time. Willow (Salix), flowering currant (Ribes sanguineum), and pulmonaria are useful early-season sources. Phacelia, borage, and open-centred flowers throughout summer support populations through the season.
Hand Pollination
In situations where natural pollination cannot be relied upon - trees growing under glass or polytunnel covers, trees in bloom during an extended cold or wet period, or isolated trees where no compatible partner is within reach - hand pollination is a practical and effective alternative.
The technique is simple. Using a small, dry, soft-bristled paintbrush or even a cotton bud, collect pollen from the open anthers of a compatible donor variety by gently dabbing the brush against the anthers until a visible yellow dusting of pollen coats the bristles. Transfer this to the stigmas of flowers on the recipient tree, gently touching each stigma in turn. Stigmas are receptive from the time the flower opens until the petals begin to fall, a window of roughly 5 to 8 days under cool conditions. For hand pollination to be worthwhile, collect pollen at peak anther dehiscence (when the anthers are fully open and the pollen is dry and loose), which typically occurs on the second or third day after a flower opens.
Pollen can be collected and stored dry in a small sealed container in the refrigerator for several days, which is useful if the donor and recipient trees don't quite bloom simultaneously - pollen from the earlier-blooming variety can be stored until the later-blooming recipient opens its flowers.
Practical Planting Strategy
With the biology understood, planning a planting that reliably pollinates is straightforward. The key decisions are: which species and varieties, how many, how far apart, and how to account for triploids.
For most species, two compatible diploid varieties in the same or adjacent flowering groups, planted within approximately 165 feet of each other, will provide reliable mutual pollination via natural bee activity. Bees routinely travel much further than this - honey bees forage at distances of half a mile to 2 miles, and pollen can arrive from orchards well beyond your own garden - but the probability of adequate transfer is highest when compatible varieties are close together and the shared bee population is working both trees. If your neighbour grows a compatible apple variety, it may well serve as your tree's pollinator, but you cannot depend on this as your only contingency.
For triploid varieties, the rule described earlier applies: plant at least two compatible diploid varieties alongside the triploid so that the diploids provide pollen for each other and for the triploid. The triploid itself contributes no pollen to the system.
Crab apples are worth knowing about as pollinator trees for apples. Most crab apples are diploid, carry an enormous number of flowers over a long bloom period, and are broadly compatible (in S-allele terms) with a wide range of apple varieties. A single crab apple can serve as an effective pollinator for multiple apple trees blooming across a range of groups. Ornamental crab apples planted as garden trees - Malus 'John Downie', Malus 'Evereste', and similar - function as pollinators for orchard apples and are a useful solution when space or variety selection is limited. Many commercial orchards include crab apple trees specifically for their pollination value.
If you are extending an existing planting, consider not just compatibility but also whether your existing trees provide adequate timing coverage. A single mid-season variety pollinating only itself leaves early and late bloomers unpollinated in years where bloom timing diverges from the average. Adding an early-blooming and a late-blooming variety to a mid-season planting gives the collection far greater robustness across variable seasons.
The Fruit Tree Planner lets you enter your specific varieties and see how well they are likely to pollinate each other, what flowering groups they fall in, and whether any are triploids that require additional partners. It covers apples, pears, plums, cherries, and a number of other fruit types, and is the practical companion to the theory covered in this article.
Diagnosing Pollination Problems
Heavy blossom followed by very little or no fruit is almost always a pollination problem. The question is which of several possible causes applies in your situation.
The most common cause is the absence of a compatible cross-pollinator. A self-incompatible variety planted alone, or with a variety from the same S-allele incompatibility group, will flower heavily every year and produce almost nothing. If this is your situation, the solution is to add a compatible variety, or to top-work a compatible variety onto an existing branch of the tree by grafting - described in the grafting guide.
The second common cause is bloom timing mismatch. If two varieties are theoretically compatible but bloom weeks apart in practice, they will not pollinate each other regardless of their genetic compatibility. Review the flowering groups of your varieties; if they are more than one group apart, the overlap may be insufficient in early or late seasons.
The third is weather. A year of poor fruit set on trees that have performed well previously, particularly if the spring was cold and wet, is almost certainly weather-driven. There is little to be done except wait for the following season. If cold springs are reliably a problem in your location, prioritise varieties in mid and late flowering groups that are more likely to open after the worst of the cold weather.
The fourth is pollinator absence. In very isolated gardens, on exposed sites with few flowers nearby, or in years when bee populations are depressed for environmental reasons, even well-matched and timely varieties may set poorly. Hand pollination, described above, is the direct intervention. Over the longer term, improving the habitat and forage available to bees in and around the planting will build the local population.
Finally, consider frost damage to open blossom. Fruit tree flowers are frost-tender once they have opened: temperatures below 28°F for more than a few hours will kill open flowers, which turn brown at the centre and fall without setting. If you see this pattern - browning in the centres of flowers immediately after a cold night - the flowers are dead and no amount of pollination will save them. Protection with horticultural fleece overnight during predicted frosts at bloom time is the only immediate countermeasure. Over the long term, site selection above frost drainage channels (as described in the husbandry guide) significantly reduces this risk.
Pollination as a Systems Problem
The practical lesson of pollination biology is that it is a systems problem, not a single-variable one. Compatible genetics, overlapping bloom timing, viable pollinators in adequate numbers, and sufficient weather windows for foraging and pollen transfer all need to come together within a period of days each spring. No single factor guarantees success; no single factor always explains failure.
Good planning - choosing compatible varieties across adjacent flowering groups, avoiding triploids without appropriate partners, and maintaining habitats that support diverse bee populations - creates a system that is resilient to single-factor failures. A cold week may suppress bee activity, but if pollen is abundant and the bloom window is generous, a few good days at either end of the cold period may be enough. A late frost may kill some early flowers, but if the planting includes mid-season varieties still in bud, those will carry the crop. Resilience, built through variety selection and habitat management, is the most reliable insurance a fruit grower can have.