Why sweet cherries fail to set fruit

Sweet cherry in full bloom in an orchard in Pella, northern Greece — Heavy bloom does not mean a crop. Six different things stand between the flower and the cherry, and each one leaves its own trace.

Savvas Pastopoulos, Agronomist MSc, Pastopoulos Agronomics, Neos Mylotopos, Pella, Greece

An orchard that blooms heavily and yields nothing is among the most common and the most misread problems in sweet cherry. The cause is almost never "the bees". It is one of six, and each one leaves a trace that becomes visible once a flower is cut open and the weather log is read. This article gives the figures behind each cause and, at the end, the order in which they can be told apart.

Two combinations planted in Greece that never set fruit Tragana Edessis and Bakirtzeika both carry the genotype S3S4, which places them in the same incompatibility group. Neither fertilises the other, however close together they are planted [1]. The same applies to Van and Regina, both S1S3 [3]. An orchard planted with such a pair will bloom every year and yield nothing, and the cause is neither the weather nor the bees.

9 to 11 dayseffective pollination
period

>30 °Cdays that shorten
ovule life

−2.2 °C10% kill at
full bloom

5 / hahives for good
pollination

Contents

How much fruit set counts as normal
Incompatibility: why the pollen arrives and does not fertilise
The genotypes of the cultivars grown here
How many pollinizers, and where
The effective pollination period
Frost and the dead pistil
Bees, bumblebees and the weather at bloom
Nutrition: what is documented and what is not
Bud quality is decided the previous summer
The diagnostic sequence in the orchard
Sources

1. How much fruit set counts as normal

Before a cause is sought, a measure is needed. Sweet cherry keeps only a small fraction of its flowers, and that is normal: a tree carrying tens of thousands of flowers could not feed as many fruits. The question is not "why did flowers drop" but "why did the tree not keep even the proportion it usually keeps in this orchard".

The comparison is therefore always against the same orchard in other years, and against neighbouring orchards of the same cultivar in the same year. A cultivar that cropped last year and not this year, with the same trees and the same pollinizers, does not have an incompatibility problem, it has a season problem.

2. Incompatibility: why the pollen arrives and does not fertilise

Sweet cherry has gametophytic self-incompatibility. A single genetic locus, the S-locus, encodes a ribonuclease in the style and a protein in the pollen. When the pollen allele matches either of the two alleles of the flower, the ribonuclease degrades the RNA of the pollen tube before it reaches the ovule. Pollination took place normally, fertilisation did not [2].

What an incompatibility group means Two cultivars that share even one of their two alleles are mutually incompatible. Cultivars carrying the same pair of alleles are grouped into numbered groups, and every member of a group is incompatible with every other member of that group. The most recent update of the international catalogue records up to 22 distinct S-alleles and 63 groups across roughly 1,483 cultivars [2].

S4′ and the self-fertile cultivars S4′ is a natural mutation of the S4 allele that has lost its function. A cultivar carrying it can reject neither its own pollen nor the pollen of other cultivars, and is therefore self-fertile and at the same time a universal pollen donor for almost the whole orchard. This is the basis of the self-fertility of Lapins, Stella, Sunburst, Sweetheart, Skeena and Grace Star [3].

Two reservations. First, self-fertile does not mean it gains nothing from a pollinizer: Skeena is described by the Oregon extension service as a less productive self-fertile cultivar and is recommended to be accompanied [5]. Second, self-fertility cancels none of what follows, a self-fertile cultivar freezes, goes without bees and loses its ovules to heat exactly like the rest.

3. The genotypes of the cultivars grown here

The first three rows come from a Greek study that genotyped 21 cultivars grown in Greece and identified two new incompatibility groups not previously described internationally [1]. The rest come from the international compatibility table [3].

Cultivar	Genotype	Group	Self-fertile
Tragana Edessis	S3S4	III	No
Bakirtzeika	S3S4	III	No
Kapsiotika	S2S5	VIII	No
Burlat	S3S9	XVI	No
Van	S1S3	II	No
Regina	S1S3	II	No
Kordia	S3S6	VI	No
Bing	S3S4	III	No
Napoleon	S3S4	III	No
Sam	S2S4	XIII	No
Rainier	S1S4	IX	No
Summit	S1S2	I	No
Canada Giant	not reported	I	No
Lapins	S1S4′	universal donor	Yes
Stella	S3S4′	universal donor	Yes
Sweetheart	S3S4′	universal donor	Yes
Sunburst	S3S4′	universal donor	Yes
Skeena	S1S4′	universal donor	Yes
Grace Star	S4′S9	universal donor	Yes

Sources: 1 2 3

Two traps in the table, and two gaps that are declared

Canada Giant belongs to group I, together with Summit, and is often sold as self-fertile when it is not [3]. Planted next to Summit as a "pollinizer", it contributes nothing.
A nursery catalogue stating "self-fertile" is not evidence. The criterion is the genotype.
For Ferrovia no published genotype was found in an accessible source. It is self-incompatible and needs a pollinizer, but its group is not given here because it could not be confirmed.
For Sweet Aryana no data were found at all, most probably it is too recent for the public databases.

4. How many pollinizers, and where

The schemes in use

10% pollinizers, every third tree in every third row. The basic scheme of the manual of the Pomology Institute of Naoussa [4], which coincides with the recommendation of the Oregon extension service [5].
30% pollinizers, the whole third row. The second scheme of the same manual [4], for situations where pollination is difficult.
One pollinizer every fifth tree in every row, for cultivars of low fertility, with Regina named explicitly [5].

Bloom overlap is checked every year, not once at planting The order of bloom shifts with the winter and with the spring. Two cultivars listed in a nursery catalogue as flowering "at the same time" may in a given year share only a handful of days with open flowers. Overlap is confirmed by observation in the orchard, in the year under examination [3][5]. No extension service was found to publish a maximum distance in metres from the pollinizer; only planting patterns are given, with the general principle that the nearer the better [9].

5. The effective pollination period

Even with the right pollinizer and abundant bees, fertilisation has a deadline. The effective pollination period is defined as the longevity of the ovule minus the time the pollen tube needs to reach it [7]. If the pollen arrives late, the ovule has already aged.

The figures from a Mediterranean climate, and the link with heat

Work on three traditional Mediterranean cultivars gave [6]:

Ovule longevity of 13 to 16 days from anthesis in the more fertile cultivars, but only 8 days in a susceptible cultivar in a hot season.
An effective pollination period of 9 and 11 days in the two better cultivars, and only 4 to 5 days in the susceptible one.
The stigma remained receptive for 12 to 14 days in all three, that is, the limit was set not by the stigma but by the death of the ovule.

And the critical finding: a significant negative correlation was found between ovule longevity and the number of days with a maximum above 30 °C in the interval from bloom to fertilisation. This is the direct link between a hot spring and a failure of fruit set while pollination apparently went well.

Temperature during bloom

Trials on five cultivars at 5, 15 and 25 °C showed high pollen germination, 66 to 75%, at both 15 and 25 °C. At 5 °C germination remained around 50% in some cultivars, but the pollen tube was three to five times shorter, it does not cover the distance to the ovule in time [8]. The Pennsylvania extension service gives the same limits in absolute terms: below 5 °C pollen practically does not germinate, and below 10.6 °C tube growth is extremely slow [9].

There are therefore two different ways to lose the season to temperature, and they are opposites: a cold bloom that does not let the tube advance, and a hot bloom that kills the ovule before the tube arrives.

6. Frost and the dead pistil

Critical temperatures by phenological stage, for an exposure of about half an hour [10]:

Stage	10% kill	90% kill
First swell	−8.3 °C	−15.0 °C
Green tip	−3.9 °C	−10.0 °C
Tight cluster	−3.3 °C	−8.3 °C
Open cluster	−2.8 °C	−6.1 °C
First white	−2.8 °C	−4.4 °C
Full bloom	−2.2 °C	−3.9 °C
Post-bloom	−2.2 °C	−3.9 °C

The most useful tool in this article: cutting the flower open

Frost can kill the pistil alone and leave the flower looking entirely normal, petals, stamens, everything in place. That is exactly what leads to the wrong conclusion of "poor pollination".

The cut is made at an angle, not straight across. A cherry bud holds two to five flowers that mature at different rates, and a flat cut misses some of them [11].
A live pistil is pale green and turgid. A dead one is brown to black, from the oxidation of the phenolics released by the freeze [11].
The discolouration is not instantaneous. Samples are left at room temperature and inspected again the next day before a conclusion is drawn [11].

A sample of 20 to 30 flowers, from several parts of the orchard and not from one tree. Low spots where cold air settles give a different picture from the slopes.

7. Bees, bumblebees and the weather at bloom

The flight limits of the honeybee

Density: five colonies per hectare, with the hives carrying at least four frames of brood early in spring [12].
Temperature: below 12.8 °C there is practically no flight. At 12.2 °C only 21% of the bees forage, at 10.6 °C only 6%. Full activity from 18.3 °C upwards [12].
Wind: optimal foraging below 16 km/h, reduced between 16 and 24, and above 24 km/h foraging stops [12].

Wild bees work when the honeybee does not fly Bumblebees and solitary bees fly at temperatures roughly 3 to 6 °C lower than the honeybee [9]. In a British study on sweet cherry, managed pollinators made 71% of the visits and wild bees only 29%, but the wild ones moved between rows more and contacted the stigma more often per visit, with wild bumblebee queens visiting an average of 19 flowers per minute [13].

No controlled comparison of fruit set between honeybees and bumblebees at low temperature was found for sweet cherry. The conclusion rests on the flight thresholds, not on a direct trial.

8. Nutrition: what is documented and what is not

Boron is not the solution it is advertised to be Boron is essential for pollen tube growth, which is why a foliar application at bloom is widely recommended. The experimental evidence, however, is weak. A two-year trial of zinc plus boron on three cultivars showed a trend towards higher fruit set and yield that was NOT statistically significant [14]. An autumn boron application on two other cultivars raised fruit set in only one of the two blocks and only in one of the two cultivars.

Boron corrects a real boron deficiency, and that is established by analysis. As a general "boost to fruit set" without a documented deficiency, it is not supported. Boron also has a narrow margin between sufficiency and toxicity.

For nitrogen, no trial was found that links the rate quantitatively to a fruit set percentage in sweet cherry. The mechanism, however, is documented generally in fruit trees: excessive nitrogen availability turns the tree towards shoot elongation at the expense of reproductive development and may delay the onset of bearing [19]. In a young orchard growing vigorously and forming no flower buds, this is the first suspicion.

For calcium, all the available literature concerns postharvest fruit quality, firmness and resistance to cracking. No published evidence was found linking calcium to flower quality or to fruit set. It is stated here as an open question and not as a recommendation.

9. Bud quality is decided the previous summer

A flower that opened deformed or weak did not fail in April. It failed the previous summer, when it was differentiated.

Double pistils and double fruit: the heat after harvest A controlled study proved that the cause is high temperature and not water stress [15]. The susceptible window is the differentiation of the pistil primordia, roughly four to six weeks after harvest, and the main trigger is given as temperatures above 37.8 °C from late July to early September [16]. Another service places the practical alarm threshold lower, above 30 °C during differentiation [17]. That range is not carelessness: the two sources measure different things, the absolute trigger and the threshold of raised risk.

The practical consequence: irrigation and shading in July and August are measures for next year's bloom, not only for the current crop.

Reserves and early defoliation After harvest the leaves continue to produce and store carbohydrates, mainly in the roots, and those reserves support next year's bloom, fruit set and early fruit growth [17]. In a controlled trial, one year of manual defoliation 50 days before natural leaf fall did not reduce the following year's yield; two consecutive years of defoliation 60 to 70 days early did reduce it [18].

The tree withstands one bad year but not two in succession. Early leaf loss from cherry leaf spot or from water stress, two seasons running, explains an apparently causeless failure in the third.

10. The diagnostic sequence in the orchard

The causes can be told apart. The order matters, because each step rules out the next.

Step 1, Cut flowers open, within 24 to 48 hours of any cold event

A brown or black pistil inside a flower that otherwise looks normal, across several cultivars and mainly in the low spots → frost. This is diagnostic on its own: it shows the flower was alive and pollinated normally, but the ovary was already dead. A green, turgid pistil → frost is ruled out, and the examination continues.

Step 2, Ovary swelling, one to three weeks after petal fall

No swelling anywhere in a block, while the bees were working and the weather was warm and calm → incompatibility is suspected. The tell-tale sign is that the problem is local, in particular blocks or in a particular cultivar combination, and not across the whole orchard. Cross-check against the table in section 3, and confirm that the pollinizer did in fact bloom together with the main cultivar in that year.

Step 3, The weather log of the bloom week

Maxima mostly below 13 °C, or wind above 24 km/h, or several days of rain → poor pollination weather. Pollen tube growth and insect flight are suppressed at the same time [8][9][12].
Maxima often above 25 to 30 °C → ovule senescence outrunning the pollen tube, that is, a short effective pollination period although the bees worked normally [6].

Step 4, If the flowers themselves were abnormal

Deformed, small, with visible double pistils, with poor petal development → examine the weather of the previous summer, four to eight weeks after the previous harvest. A run of days above 35 to 38 °C in that window is the documented trigger, especially if the previous crop was heavy or the trees lost their leaves early [15][16][17][18].

Step 5, If the flowers wilt, blacken or gum

Brown, collapsed flowers, often with fungal growth → brown rot. Blackened and water-soaked, often after a cold and wet spell → bacterial blossom blast caused by Pseudomonas. Both destroy the flower before it can set, and both species of Monilinia have been confirmed on stone fruit blossoms and fruit in Greece [20]. Identification is by laboratory sample.

Step 6, Only after the above have been ruled out

Excessive vegetative growth, poor spur formation, an unsuitable or aged rootstock, pruning that is too severe or badly timed, planting distances that have closed the orchard in. No controlled trial was found that attributes a specific fruit set percentage to pruning timing or to rootstock choice, so this group of causes remains a diagnosis of exclusion and rests on general physiology rather than on figures.

11. Sources

The numbers in the text refer to the list below.

[1] Ganopoulos, I., Argiriou, A. and Tsaftaris, A. (2010). Determination of self-incompatible genotypes in 21 cultivated sweet cherry cultivars in Greece and implications for orchard cultivation. Journal of Horticultural Science and Biotechnology 85(5):444-448. Source for the genotypes of the Greek cultivars, Tragana Edessis and Bakirtzeika S3S4, Kapsiotika S2S5, and for the identification of two incompatibility groups not previously described internationally.
[2] Schuster, M. (2020). Self-incompatibility (S) genotypes of cultivated sweet cherries, an overview, update 2020. Julius Kühn-Institut. Source for the mechanism of gametophytic self-incompatibility and for the 22 alleles and 63 groups across 1,483 cultivars.
[3] ANFIC Sweet Cherry Compatibility and Bloom Timing Table (2022). Source for the genotypes and groups of the international cultivars, for the S4′ of the self-fertile ones and for the group I placement of Canada Giant.
[4] ELGO DIMITRA, Pomology Institute of Naoussa. Updated manual of sweet cherry cultivation (2013), in Greek. Source for the two pollinizer schemes, 10% and 30%, under Greek conditions.
[5] Sweet Cherry Orchard Establishment in the Pacific Northwest, PNW-642, Oregon State University Extension. Source for the every-third-tree-in-every-third-row scheme, for the denser arrangement in low-fertility cultivars such as Regina, and for the lower productivity of the self-fertile Skeena.
[6] Radunić, M., Jazbec, A. and Goreta Ban, S. (2026). Ovule longevity determines the fruit set and effective pollination period in sweet cherry cultivars under Mediterranean climate conditions. Agronomy 16(14):1382. Source for ovule longevity of 8 to 16 days, for the effective pollination period of 4 to 11 days and for its correlation with days above 30 °C.
[7] Sanzol, J. and Herrero, M. (2001). The effective pollination period in fruit trees. Scientia Horticulturae 90:1-17. Source for the definition of the effective pollination period.
[8] Milatović, D. and Nikolić, D. (2017). The effect of temperature on pollen germination and pollen tube growth in vitro of sweet cherry cultivars. Acta Horticulturae 1161:401-404. Source for pollen germination of 66 to 75% at 15 and 25 °C and for the three to five times shorter pollen tube at 5 °C.
[9] Orchard Pollination: Pollinizers, Pollinators, and Weather, Penn State Extension. Source for the 5 °C and 10.6 °C limits for pollen, and for bumblebees and solitary bees flying 3 to 6 °C lower than the honeybee.
[10] Sweet Cherry Bud Stages and Critical Temperatures, Colorado State University, Western Colorado Research Center. Source for the table of critical temperatures by phenological stage.
[11] How to Assess Cold Damage, Washington State University Tree Fruit. Source for the angled cut, for the discolouration of the pistil and for waiting one day before reading the sample.
[12] Honey Bees, Washington State University Tree Fruit. Source for the five colonies per hectare, for the temperature limits of 12.8 and 18.3 °C and for the wind limits of 16 and 24 km/h.
[13] Mateos-Fierro, Z. et al. (2022). Wild bees are less abundant but show better pollination behaviour for sweet cherry than managed pollinators. Journal of Applied Entomology. Source for the 71% of visits by managed pollinators and for the 19 flowers per minute of wild bumblebee queens.
[14] Usenik, V. and Štampar, F. Effect of foliar application of zinc plus boron on sweet cherry fruit set and yield. Acta Horticulturae 594. Source for the finding that the increase in fruit set from zinc plus boron was not statistically significant.
[15] Beppu, K. and Kataoka, I. (1999). High temperature rather than drought stress is responsible for the occurrence of double pistils in Satohnishiki sweet cherry. Scientia Horticulturae 81:125-134. Source for the finding that the cause of double pistils is temperature and not water stress.
[16] Fruit Doubling and Spurs on Cherry, University of California Integrated Pest Management. Source for the four to six week window after harvest and for the 37.8 °C threshold.
[17] The Next Cherry Crop Starts at Harvest, Washington State University Tree Fruit. Source for the postharvest carbohydrate reserves and for the 30 °C threshold during differentiation.
[18] Early Fall Defoliation in Sweet Cherry, Washington State University Tree Fruit. Source for the finding that one year of early defoliation did not reduce yield while two consecutive years did.
[19] Neilsen, D. et al. (2018). Nitrogen nutrition of fruit trees to reconcile productivity and environmental concerns. Plants 7(1):4. Source for the shift of the tree towards vegetative growth under excess nitrogen and for the delay in the onset of bearing.
[20] Intraspecific diversity of Monilinia fructicola and M. laxa populations from blossoms and fruit of different hosts in Greece. Plant Disease. Source for the presence of both species of Monilinia on stone fruit blossoms in Greece.

Disclaimer

Plant protection products and fertilisers are used only in accordance with the authorisations of the Greek Ministry of Rural Development and Food and with what is stated on their labels: authorised crops, doses, maximum number of applications, re-entry intervals and pre-harvest intervals. Those authorisations are valid in Greece only. They do not apply in any other country, where the national register of that country applies instead.

Any foliar spray on trees carrying fruit risks marking on the fruit and scorch on the foliage. A prior test on a limited number of trees, with the same product and the same dose, followed by a wait of five to seven days, is required before general application. Spraying is done in the late afternoon or early morning, never in a heatwave and never on foliage under water stress.

This is informative and educational material summarising published findings. It is not an individual agronomic recommendation. The cultivar genotypes come from published catalogues and do not replace verification of the planting material that was actually bought. Every intervention is carried out at the user's responsibility, following an on-site assessment by a licensed agronomist. Pastopoulos Agronomics accepts no liability for damage arising from the use of the information in this article.