Pollen–Pistil Interactions in Autochthonous Balkan Sweet Cherry Cultivars—The Impact of Genotype and Flowering Temperature

Abstract

The efficacy of sweet cherry production is highly dependent on the regularity of flowering events and genetic-determined relations between female sporophyte and male gametophyte, which became even more important with higher flowering temperatures caused by climate change. Special attention is paid to the genetic diversity that provides essential sources of potential temperature-tolerance genes. Our study aimed at the genetic and reproductive characterization of Balkan cherry cultivars of autochthonous origin (‘Canetova’, ‘G-2’, ‘Dolga Šiška’ and ‘Ohridska Crna’), and six potential pollenizers. To identify S-haplotypes, the polymerase chain reaction (PCR) method was used to detect the S-ribonuclease (S-RNase) and S-haplotype-specific F-box protein (SFB) alleles, combined with fragment analysis and S-RNase sequencing. Pollination experiments were performed at three Balkan localities over two flowering seasons, and the fluorescence microscopy method was used to assess the cultivars’ male/female reproductive behaviour. A novel S-RNase allele S₄₀ was identified in ‘Ohridska Crna’ for the first time. ‘Ohridska Crna’ also demonstrated the best adaptability to higher temperatures regarding primary ovule longevity. This feature makes it desirable from the aspect of breeding new cultivars that can withstand the impacts of climate change. The findings on male-female relations and their temperature dependence open up the possibility for yield prediction and smart horticultural decisions that can be made to guide cherry production.

1. Introduction

Sweet cherry (Prunus avium L.) is among the most popular temperate fruit crops, due to the attractiveness of its fruits, their nutritive value and beneficial health effects, and suitability for fresh consumption and processing. Despite the number of public and private breeding programmes carried out in many countries and the available commercial cultivars that are continuously being released, cherry fruit production is still based on a small number of genotypes, including landraces [1]. Molecular diversity studies assessed through polymorphisms of microsatellites and self-incompatibility locus (S-locus) revealed a reduction in genetic diversity from wild populations, via landraces, to modern bred sweet cherry cultivars, demonstrating significant genome-wide losses of variations and making the genetic base of modern breeding programmes quite narrow [2,3]. Diversity becomes even more important in the context of adaptation to changes in environmental conditions induced by global warming, especially for perennial fruit crops, which require more than a decade for the release of a new cultivar [4].

A key trait influencing the cultivation and breeding of sweet cherries is self-incompatibility, which prevents fertilization by genetically related cultivars or self-pollination and, thus, promotes continuous genetic exchange within and between populations [5]. The mechanism of gametophytic self-incompatibility (GSI) in sweet cherry is controlled by two multi-allelic-linked genes at the S-locus–S-RNase and SFB, which are expressed in the style and the pollen, respectively [6,7]. To describe the complexity of this locus, the term ‘S-haplotype’ is used, while ‘S-allele’ refers to the polymorphic variants of both genes within the S-haplotype [8]. The structure of the two genes is well-characterized: S-RNase contains five conserved regions, one highly variable region and two introns [9,10], while SFB comprises one F-box region, two variable regions, two hypervariable regions and one intron in the 5’ untranslated region (5′UTR) [11,12]. According to Schuster et al. [13], 23 functional (S₁–S₇, S₉–S₁₀, S₁₂–S₁₄, S₁₆–S₁₉, S₂₁–S₂₂, S₂₄, S₂₇, S₃₀, S₃₇ and S_38/7m) and three non-functional (S_3′, S_4′ and S_5′; pollen-part mutants) S-haplotypes have been identified in cultivated sweet cherries among 34 reported in Prunus avium to our knowledge [9,14,15,16,17,18,19,20,21]. In the numbering S₁ to S_38/7m alleles, S₈, S₁₁, and S₁₅ were excluded after confirming their identity with S₃, S₇, and S₅, respectively [9,14,22], while the alleles S₂₆, S₃₃, and S₃₅–S₃₆ were identified in sour cherries [23]. Diploid sweet cherries are usually self-incompatible (SI), and certain genotype pairs exhibit cross-incompatibility (CI), either reciprocally or unilaterally, requiring compatible pollen for a successful fruit set. To date, 72 incompatibility groups (IGs), as well as one ‘0’ group (unique S-genotypes) and one ‘SC’ group, have been reported in sweet cherries [13,24]. In modern orchard management, it is well known that a significant portion of the sweet cherry area must be allocated to genetically compatible cultivars, known as pollenizers, so that factors such as their flowering and harvesting and their distance from the main cultivar must be carefully considered to ensure successful commercial production. Although self-compatible (SC) sweet cherries represent one of the greatest achievements of breeding efforts, planting two or three cross-compatible pollenizers remains essential to ensure high yields [25,26].

Besides the GSI, investigations on other interactions at the pollen–pistil level, in particular in their relation to flowering temperatures, are crucial for understanding adaptation to different environmental conditions. Sweet cherry is better adapted to the conditions of somewhat colder climates [27], therefore, its fruit production should be considered in light of the increasing frequency of flowering seasons with higher temperatures in growing regions in Europe [4,28], including the Balkan region [29]. Climate parameters observed in Serbia show a trend of temperature increase, with projections (period 1961–2100) over 2.5 °C for mean temperature (‘stabilization’ scenario) or over 5 °C (‘constant-increase’ scenario), which affects the production of fruits and grapes [30,31,32]. Rising spring temperatures due to global warming have led to earlier flowering dates for cherry cultivars in numerous production areas [33,34], increasing their risk of frost damage [35]. On the other hand, warm temperature stress accelerates ovule senescence, which as a consequence leads to female gametophyte degeneration and ovule abortion [36], which is a genotype-dependent reaction in sweet cherry [37]. Flowering temperature is also a key factor affecting stigmatic receptivity [38] and pollen tube kinetics [27]. Genotypes exhibiting tolerance are valuable genetic reservoirs for potential temperature-tolerance genes [36]. Therefore, the identification, characterization and maintenance of these genetic resources are essential for expanding the genetic pool for current and future programmes. Nowadays, breeders are paying more attention to the collections of genetic resources, held by numerous countries [39].

Although remarkable efforts have been made to conserve cherry genetic resources up to now, it remains essential to coordinate both genotypic and phenotypic characterization to identify new sources of diversity for commercial growing of the genotypes as ‘final products’ of natural selection or as male/female parents in planned hybridization. As for the former Yugoslav countries, numerous data related to genotype/phenotype characterization and conservation of sweet cherry landraces, originating from Serbia [40,41,42], Croatia [43,44], North Macedonia [45,46,47,48], and Bosnia and Herzegovina [49,50], have been reported.

Within this research, we used molecular and reproductive biology tools to identify S-haplotypes and characterize the reproductive behaviour of four autochthonous sweet cherry cultivars originating from Serbia and North Macedonia. These traits were analyzed in the context of the specific relationships between the female sporophytes of autochthonous cultivars and commercial cultivars as pollen donors, as well as the impact of flowering temperature on the quality of these relationships, which jointly result in efficient/not efficient ending of the fertilization process.

2. Materials and Methods

2.1. Plant Material and Experiment Design

Four autochthonous sweet cherry cultivars were used in the study: (i) ‘Canetova’ (Figure 1a)—selected by the Faculty of Agriculture, Belgrade (AGRIF) as a spontaneous seedling in the Belgrade region, and released in 2014 as an early-ripening cultivar (2nd week of cherry ripening—WCR), with large fruits (9 g) of dark-red skin and flesh; the aroma is very pleasant and the taste is sweet, due to favourable soluble solid/total acidity ratio [41]; (ii) ‘G-2’ (Synonym ‘Đuti’; Figure 1b)—selected by AGRIF in the Belgrade region, as a very early-ripening (1st WCR) spontaneous seedling; it is a candidate-cultivar in Serbia, with satisfactory fruit size, deep-red skin and flesh colour; soluble solid/total acidity ratio is high, due to satisfactory soluble solid content and relatively low total acidity, which makes this cultivar interesting from the aspect of the consumer acceptance, regardless of very early ripening time [40]; (iii) ‘Dolga Šiška’ (Figure 1c)—old late-ripening local cultivar (5th WCR) grown in North Macedonia, but also in Bulgaria, northern Greece and Serbia; highly appreciated due to very large (12–13 g) wide-hearted fruits of exceptional aroma and taste, long stalk, pronounced firmness, dark-red skin and flesh; shows productive traits on the same level or better than many of European sweet cherries of the same ripening season [45]; and (iv) ‘Ohridska Crna’ (Figure 1d)—old, very late ripening local cultivar (end of 6th WCR), grown in North Macedonia (southern-west parts in vicinity of Ohrid lake), Greece, Bulgaria, and Serbia; characterized by large, elongated heart-shaped fruits of a blackish skin, dark-red flesh, purple juice, and long stalk [46]; the fruits are characterized by a pleasant aroma and recognizable full, bitter-sweet taste.

The choice of pollenizers was based on the data on perennial overlap during the full-flowering phenophase with the autochthonous genotypes, and their commercial value: ‘Burlat’, ‘Lapins’, and ‘Rita’—for early flowering ‘Canetova’ and ‘G-2’ [34,40,41,51,52]; ‘Kordia’, ‘Summit’, and ‘Sunburst’—for late flowering ‘Dolga Šiška’ and ‘Ohridska Crna’ [45,46,52].

Field experiments were conducted during the flowering seasons of 2022 and 2023 at (i) cherry gene-bank collection of AGRIF at the experimental site ‘Radmilovac’, Belgrade, Serbia (44°45′ N; 20°35′ E; 130 m altitude) for ‘G-2’ and its pollenizers; (ii) commercial cherry orchard in Ohrid, North Macedonia (41°07′ N; 20°48′ E; 695 m altitude) for ‘Ohridska Crna’, ‘Dolga Šiška’, and their pollenizers; and during the flowering seasons of 2023 and 2024 at the commercial cherry orchard in Ljubić locality, Čačak, Serbia (43°54′ N; 20°18′ E; 242 m altitude) for ‘Canetova’ and its pollenizers. The climate is temperate continental (Radmilovac, Ljubić) or Mediterranean-like (Ohrid). All three areas are typical for fruit growing in the Balkan Peninsula.

The genotypes are grafted on wild cherry (Prunus avium L.) seedlings and represented by nine trees each, allowing all experiments to accord to a statistically randomized block design. The orchards were in full fertility stage (10–12 years old) and maintained with standard cultural practises in conditions without irrigation.

2.2. S-Allele Identification

2.2.1. Extraction of Genomic DNA

Total genomic DNA from each of the four sweet cherry cultivars of autochthonous origin and six commercially important cultivars used as pollenizers were isolated from fresh young leaves collected in the spring of 2022. The leaf samples were frozen in liquid nitrogen, ground with a Mixer Mill MM 400 (Retsch GmbH, Haan, Germany), and genomic DNA was then extracted using the CTAB method [53], with some modifications of the extraction buffer [54]. Dried DNA pellets were dissolved in 50 μL of TE buffer (10 mM Tris pH 8.0 and 1 mM EDTA) containing RNase A (10 μg/mL) (Invitrogen, Groningen, The Netherlands) and incubated at 37 °C for 30 min to digest RNA. The DNA samples were stored at −20 °C until the PCR reactions were performed.

2.2.2. PCR Amplification of S-RNase and SFB Alleles

To nominate the most likely S-RNase alleles in each sweet cherry cultivar, consensus primers were used to amplify the first PaConsI-F/PaConsI-R (corresponding regions for the signal peptide and mature protein) and the second intron PaConsII-F/PaConsII-R (corresponding to the C2 and C5 conserved regions) of this gene [9]. The PCR methodology for both primer pairs closely followed the procedures described by Sonnenveld et al. [9]. In addition, Pru-C2 and PCE-R consensus primers (corresponding to the C2 and C3 conserved regions) were used to determine the S-RNase alleles in ‘Ohridska Crna’ [55,56]. The PCR reaction and amplification conditions for this primer pair were adjusted according to Sebolt et al. [23] as follows: approximately 100 ng of genomic DNA was used in a 25 μL reaction containing 1× PCR reaction buffer, 3 mM MgCl₂, 200 μM dNTPs, 0.8 μM of each primer, and 0.625 U Taq DNA polymerase (Qiagen GmbH, Hilden, Germany). PCR cycling conditions in the Mastercycler^® nexus gradient (Eppendorf AG, Hamburg, Germany) were as follows: 94 °C for 2.5 min, followed by 35 cycles of 94 °C for 30 s, 61 °C for 30 s and 72 °C for 75 s, with a final 7 min extension step at 72 °C.

Primer pairs for alleles S₁- through S₇-, S₉-, S₁₀-, and S₁₂-RNases, along with the PCR reactions, were reported in the studies by [9,14]. The PCR cycling conditions in a Mastercycler^® nexus gradient (Eppendorf AG, Hamburg, Germany) were slightly modified as follows: 94 °C for 2 min, followed by 35 cycles of 94 °C for 30 s, a specific annealing temperature for 30 s, and 68 °C for 1 min, with a final 5 min extension step at 72 °C. The annealing temperatures for these alleles were provided in the study by [57], with 62 °C used for S₁₂-RNase. Additionally, for ‘Ohridska Crna’, S-RNase-specific PCR reactions were performed for alleles S₁₃, S₁₄, S₁₆ through S₁₉, S_21/25, S₃₄, and S₃₇, applying annealing temperatures for certain alleles [9,21,22], combined with some variations implemented in our study.

To identify the allelic constitution of the SFB gene, the following allele-specific primer pairs were used: SFB₁ through SFB₆ [58], SFB₉ and SFB₁₂ [23], and SFB_4′ (distinguishing the functional variant SFB₄ from the non-functional variant SFB_4′, which has a 4 bp deletion) [16]. Compared to Zhu et al. [16], the PCR reaction and conditions for amplifying the SFB_4′ allele using BFP200 and BFP201 primers in ‘Lapins’, ‘Ohridska Crna’, and ‘Sunburst’ were modified based on Sebolt et al. [23] and reported by Marić et al. [42]. For the amplification of the other SFB alleles, we also followed the PCR protocols [23], with modifications to the annealing temperatures for certain alleles: 63 °C for SFB₁, 55 °C for SFB₃, 64 °C for SFB₆, and 60 °C for SFB₉.

The PCR products generated using the consensus and the allele-specific primers were separated on 2% and 1.5% agarose gels, respectively, with the Biometra Horizon 11.14 system (Analytik Jena GmbH, Jena, Germany) for approximately 3–4 h at 70 V/cm. A 1 Kb plus DNA ladder (Invitrogen, Groningen, The Netherlands) was used for fragment sizing. DNA bands were visualized through ethidium bromide (0.5 µg/mL) staining and observed under ultraviolet light (UV) using the BIO-PRINT-1500/26M imaging system (Vilber Lourmat, Collégien, France).

2.2.3. Fragment Analyses of S-Locus Genes

The fluorescently labelled forward primer PaConsI-F (Yakima Yellow), in combination with the reverse primer PaConsI-R2 [59], was used to amplify the first intron of the S-RNase, whereas the primer set F-BOX5′A (6-FAM) and F-BOX intronR [12] was used for amplification of the intron located in the 5′UTR of the SFB gene in the studied autochthonous cultivars and pollenizers. The PCR reactions and cycling conditions for the S-RNase and SFB intron regions were performed according to the protocols of Sonneveld et al. [59] and Cachi and Wünsch [60], respectively. Fluorescently tagged PCR products for allele sizing of both S-genes were prepared as follows: 2 µL of PCR product, 0.25 µL of 500 LIZ dye size standard (Thermo Fisher Scientific, Waltham, MA, USA) and 17.75 µL of Hi-Di formamide (Thermo Fisher Scientific, Waltham, MA, USA). The PCR fragments were sized using the Applied Biosystems SeqStudio™ Genetic Analyzer (Thermo Fisher Scientific, Waltham, MA, USA) with SeqStudio™ Data Collection Software 1.2.4 and GeneMapper 6.1 (Thermo Fisher Scientific, Waltham, MA, USA).

2.2.4. Sequencing of S-RNase Fragment

The second intron region of S-RNases in ‘Ohridska Crna’ was amplified using Pru-C2 and PCE-R consensus primers in a total volume of 250 µL (five reactions of 50 µL each). For the elution of the S_x allele of ‘Ohridska Crna’, approximately 220 µL of the PCR product was loaded onto a 2.2% agarose gel. The fragment, approximately 440–450 bp in size and corresponding to the S_x-RNase, was extracted and purified from the gel using the GeneJET Gel Extraction kit (Thermo Fisher Scientific, Waltham, MA, USA). The S_x-RNase was re-amplified and purified to remove residual PCR primers using the GeneJET PCR Purification kit (Thermo Fisher Scientific, Waltham, MA, USA). It was then sequenced using the Applied Biosystems™ BigDye™ Terminator v3.1 Cycle Sequencing Kit on the SeqStudio Genetic Analyzer (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA, USA). The boundaries of the second intron and the coding regions of S-RNase were predicted by pairwise alignment with similar sequences in the nucleotide collection of the National Centre for Biotechnology Information (NCBI; https://www.ncbi.nlm.nih.gov/) (accessed on 31 March 2024). Nucleotide sequences and their deduced amino acid sequences were compared using the Clustal W algorithm of the MegAlign program (DNAStar).

2.3. Pollination Experiment

2.3.1. Air Temperature

The air temperature was continuously monitored before and during the full flowering phenophase at all experimental sites. The average mean, maximum, and minimum daily temperatures of full flowering (the date of full flowering beginning at ‘zero-date’, BBCH stage 65 [61], and ten days, thereafter), as well as for the period before full flowering (ten days up to the ‘zero-date’) were calculated per season for each autochthonous cultivar.

2.3.2. Pollination Procedure

Two-year-old branches of each autochthonous cultivar were chosen, with a synchronized population of 80–100 flowers at the late balloon stage (BBCH 60–61). In this manner, about 2000 flowers per cultivar were selected, emasculated, and protected with paper bags. The branches at the same flowering stage were also chosen for the open-pollination variant (500 flowers per cultivar, without emasculation and bagging). Simultaneously, the anthers of autochthonous cultivars and pollenizers were collected from the flowers at the late balloon stage and allowed to dehisce for 24–48 h at 20 °C. Pollination of emasculated flowers was performed at the beginning of full flowering (BBCH 65) when stigmatic secretion was evident. Approximately equal amounts and abundance of pollen were ensured with two touches of stigma [62]. Pistils of each cultivar were self-pollinated and cross-pollinated with three pollenizers (four combinations per pollinated cultivar; 16 combinations of hand-pollination in total). Together with the open-pollination variant, 20 combinations and 10,000 flowers per flowering season were processed. Pollinated flowers were re-isolated using protective bags, which were permanently removed three weeks after the pollination.

2.3.3. Microscopic Observation of Pollen Performance In Vivo and Ovule Fluorescence

A total of 30 pistils of each treatment were fixed on the 3rd, 6th, and 10th day after ‘zero-date’ in FPA (70% ethanol, propionic acid, and formaldehyde, 90:5:5 percentages by volume). The aniline blue staining was used [63,64]. The styles were separated from the ovaries, opened along the suture, and covered with a cover glass (squash preparations). The ovaries were dissected along the suture; integuments of the primary ovules were cut with a razor blade longitudinally and tangentially, to enable better observation of pollen-tube penetration in the micropyle and nucellus [65]. The pistils were observed under UV light on the Olympus BX61 microscope (Olympus, Tokyo, Japan), and analyzed by AnalySIS software Version 3.1, using Multiple Image Analysis. Pollen tubes were counted in the style (upper third, middle third, and the base) and the ovary, at a magnification of 200× and 100×, respectively, and their number was presented as the average of three fixation samples. The pollen tube growth rate was presented as the percentage of pistils with the longest pollen tube penetrating to their particular parts. The total of pistils with penetration into the nucellus on the 10th day after ‘zero-date’ was taken as the fertilization percentage. The ‘unusual’ pollen tube growth rate in the ovary (wandering, branching, and curling up) has been recorded before their entrance into the ovule. Primary ovule fluorescence (from the chalazal region to the entire ovule) [37] was monitored in cross-pollination and open-pollination variants. The unusual pollen behaviour and ovule fluorescence rates are presented as average for all fixation terms.

2.3.4. Fruit Set

The fruit set was recorded at the beginning of ripening (BBCH 87), as the percentage of fruits per total number of pollinated flowers remaining after the final fixation.

2.3.5. Statistical Analysis

The data were statistically analyzed using a two-factor analysis of variance (ANOVA). The significance of differences among mean values was determined by LSD multiple range tests at p ≤ 0.05. Correlations among the parameters were determined by correlation regression analysis and Pearson’s correlation coefficients, through two matrices—one including the reproductive parameters for cross-pollination, and the other for the open-pollination variant. Statistical analyses were performed using the SPSS statistical software package, Version 8.0 for Windows.

3. Results

3.1. Identification of S-Haplotypes in Autochthonous Cultivars and Pollenizers

The cultivars and a promising selection (candidate cultivar) of autochthonous origin, as well as the pollenizers, were genotyped according to the indicated methods.

In the PCR reactions with the aforementioned S-RNase consensus primers, two products were obtained for each autochthonous cultivar and pollenizer, except for ‘Ohridska Crna’, where the primers PaConsII-F and PaConsII-R amplified only a fragment of approximately 1060 bp, corresponding to the S₄-RNase allele (Table S1; Figure S1b).

The length of the PCR products amplified with primers PaConsI-F and PaConsI-R ranged from approximately 300 bp (S₃-RNase allele) to around 420 bp (S₂-/S₇-/S₉-/S₁₀-/S₁₂-/S₂₂-RNase alleles), 450 bp (S₁-/S₅-RNase alleles), and up to about 520 bp (S₄-/S₆-RNase alleles) (Table S1; Figures S1a and S2a). The PCR product lengths for the second intron amplified using the primers PaConsII-F and PaConsII-R in the assessed sweet cherry samples ranged from approximately 570 bp to 2300 bp (Table S1; Figures S1b and S2b). This primer pair facilitated the identification of sweet cherries carrying S₄-RNase (~1060 bp), S₅-RNase (~2160 bp + ~1650 bp, although both bands of this allele were not strong), S₆-RNase (~570 bp), and S₁₂-RNase (~1770 bp + ~1500 bp) alleles. However, distinguishing between S₁- and S₃-RNase (~880 bp), as well as between S₂-, S₇-, and S₂₂-RNase (~2300 bp), required additional analysis (Table S1).

The application of allele-specific primers (Table S1) confirmed the presence of S₁-RNase in ‘Lapins’ and ‘Summit’; S₂-RNase in ‘G-2’ and ‘Summit’; S₃-RNase in ‘Burlat’, ‘Dolga Šiška’, ‘G-2’, ‘Kordia’ and ‘Sunburst’; S₄-RNase in ‘Lapins’, ‘Ohridska Crna’ and ‘Sunburst’; S₅-RNase in ‘Canetova’ and ‘Rita’; S₆-RNase in ‘Canetova’ and ‘Kordia’; S₉-RNase in ‘Burlat’; and S₁₂-RNase in ‘Dolga Šiška’. The lengths of the allele-specific PCR products (Table S1) were consistent with previously reported data [9,14]. Furthermore, to reveal the second allele in ‘Ohridska Crna’ (designated as S_x), no amplification was obtained using primers specific to S₁₃-, S₁₄-, S₁₆- through S₁₉-, S_21/25-, S₃₄-, and S₃₇-RNase.

SFB genotyping using specific primers for SFB₁ (fragment of 768 bp) through SFB₆ (fragment of 781 bp), SFB₉ (fragment of 250 bp), and SFB₁₂ (fragment of 292 bp) confirmed the results previously obtained from the identification of S-RNase alleles in autochthonous cultivars and pollenizers. Additionally, a 453 bp fragment corresponding to SFB_4′ was identified in ‘Lapins’ and ‘Sunburst’. The absence of amplification of this fragment in ‘Ohridska Crna’ verified that this autochthonous cultivar has a functional SFB₄.

To reveal the second allele in ‘Ohridska Crna’ and to confirm the presence of the aforementioned alleles in the assessed cultivars, fragment analyses were performed using fluorescently labelled forward primers to amplify both the first intron of S-RNase and the SFB intron. Ten S-RNase fragments were generated, ranging from 236 bp to 454 bp (Table S2a,b). For SFB, eight fragments ranging from 177 bp to 204 bp were identified, except for SFB₉ in ‘Burlat’ (Table S2a,b). The lengths of the obtained fragments matched to those previously reported [60], using the ABI PRISM 3130xl genetic analyzer (Applied Biosystems), with a difference of −1 bp for S₆-, S₉-, and S₁₂-RNases, and +1 bp for SFB₃ and SFB₅. In addition, the data obtained in our study showed a difference of +2 bp compared to the data generated using the ABI PRISM 310 genetic analyzer (Applied Biosystems) [12,60]; except for S₉-RNase, where the difference was +1 bp, and +3 bp for S₄- and S₆-RNases and SFB₅. The use of different instruments (ABI PRISM 310 and ABI PRISM 3130xl genetic analyzers) resulted in length discrepancies among various S-RNase and SFB alleles in sweet cherries [60]. Some SFB alleles have very similar lengths, making their differentiation difficult due to small differences in length. In such cases, as well as when SFB fragments are identical in length (e.g., 191 bp for SFB₁ and SFB₄), S-RNase amplification is used to confirm haplotype differentiation. This approach allowed the identification of the S₂₂-haplotype (424 bp for S₂₂-RNase and 177 bp for SFB₂₂) in the pollenizer ‘Rita’ (Table S2a,b). In the case of ‘Ohridska Crna’ (S₄S_x), fragments of 454 bp and 191 bp corresponded to the S₄-RNase and SFB₄ alleles, respectively, while fragments of 384 bp and 179 bp were associated with the S_x-RNase and SFB_x alleles (Table S2a,b). The S_x-haplotype displayed S-RNase and SFB intron products, which are unique among the Prunus avium S-haplotypes studied to date.

Based on the methods used, the S-haplotypes were identified in autochthonous cultivars and pollenizers, enabling their assignment to the corresponding IGs and the ‘SC’ sweet cherry group (

Table 1. S-haplotypes and IGs of the analyzed sweet cherry material, including SC cultivars.

Cultivar	S-Haplotype	Incompatibility Group
‘Canetova’	S5S6	XV
‘Dolga Šiška’	S3S12	XXII
‘G-2’	S2S3	IV
‘Ohridska Crna’	S4S40	0 a
Pollenizer
‘Burlat’	S3S9	XVI
‘Kordia’	S3S6	VI
‘Lapins’	S1S4’	SC b
‘Rita’	S5S22	0
‘Summit’	S1S2	I
‘Sunburst’	S3S4’	SC

^a 0: Incompatibility group ‘0’ (group of universal donors). ^b SC: Self-compatible cultivar.

), following the reported classification [24].

3.2. Sequencing and Characterization of the Novel S₄₀-RNase

Fragment analysis revealed two peaks corresponding to the S₄- and S_x-haplotypes for both the S-RNase and the SFB genes in ‘Ohridska Crna’. Since the fragment lengths of the S_x-haplotype did not match any known S-haplotype in sweet cherry, and no amplification was obtained with any of the available S-RNase allele-specific primers, this indicated the presence of a possible undetermined allele. Furthermore, the consensus primer pair PaConsII-F and PaConsII-R, used primarily for S-genotyping, identified only a single fragment of approximately 1060 bp, corresponding to the S₄-RNase allele, while the additional use of the consensus primers Pru-C2 and PCE-R yielded two bands: the first, approximately 830 bp, corresponding to the S₄-RNase, and the second, approximately 440–450 bp, possibly associated with the S_x-RNase. Therefore, the fragment obtained from the second intron region of the S_x-RNase consensus PCR was sequenced. The partial sequence of the S_x-RNase consists of 75 bp of the second exon and 92 bp of the third exon, interrupted by 270 bp of the second intron. A new S-RNase sequence has been submitted to the NCBI GenBank database and named S₄₀ (Accession number: PQ753275), following the first published system [9] and later continued [18,20,21], and our unpublished data deposited at NCBI. Among the 30 deposited sweet cherry reference sequences of the corresponding region of the S-RNase gene, the length of the second intron ranges from 82 bp (S₂₇-RNase) to 1983 bp (S₇-RNase). The S₄₀-RNase is the only allele with a second intron length of 270 bp. The deduced amino acid sequence of S₄₀-RNase exhibited similarity with published sequences ranging from 54.3% (S₂₀-RNase) to 82.6% (S₁-RNase), as shown in Figure 2. Based on all the above results, we identified the S-haplotype of ‘Ohridska Crna’ as S₄S₄₀ (Table 1), which represents a unique S-genotype to date.

3.3. Flowering Temperatures

In the period before full flowering, the average mean air temperatures were relatively uniform throughout the flowering seasons for ‘Canetova’ and ‘Dolga Šiška’. The differences were more expressed for ‘G-2’ and ‘Ohridska Crna’, with temperatures being higher in the first flowering season (

Table 2. Air temperatures before and during the full flowering of autochthonous sweet cherry cultivars.

Cultivar	Flowering Season	Period	Average Mean Temperature (°C)	Average Max. Temperature (°C)	Average Min. Temperature (°C)
‘Canetova’	I	BFF a	10.23	18.34	2.67
		FF b	6.28	11.29	2.01
	II	BFF	10.26	18.74	1.91
		FF	15.98	26.39	6.90
‘G-2’	I	BFF	12.00	17.00	6.80
		FF	11.00	16.60	5.20
	II	BFF	10.80	17.60	4.40
		FF	7.10	11.40	3.30
‘Dolga Šiška’	I	BFF	10.59	18.55	3.70
		FF	12.55	23.10	5.30
	II	BFF	10.34	18.74	2.49
		FF	13.49	20.17	8.03
‘Ohridska Crna’	I	BFF	12.74	23.48	4.41
		FF	13.09	22.87	5.83
	II	BFF	10.69	17.22	3.00
		FF	12.99	20.55	6.37

^a BFF: Before full flowering. ^b FF: Full flowering.

). As for the average maximum temperature, the biggest seasonal difference was observed for ‘Ohridska Crna’ (6.3 °C). Maximum temperatures reached more than 25 °C at least one day before full flowering in ‘Canetova’ (both seasons; Figure S3), and in ‘Dolga Šiška’ and ‘Ohridska Crna’ (first season; Figures S4 and S5). During the same period, minimum temperatures dropped to 0 °C in the second flowering season (‘G-2’, ‘Dolga Šiška’), or both seasons (‘Ohridska Crna’); for ‘Canetova’, they even fell below 0 °C, in both flowering seasons (Figures S3–S6).

During the full flowering, average mean air temperatures were higher in the first season for ‘G-2’ and ‘Ohridska Crna’, i.e., in the second season for ‘Canetova’ and ‘Dolga Šiška’ (Table 2). The biggest differences were observed for ‘Canetova’ (9.7 °C; 15.1 °C, and 4.9 °C for average mean, maximum, and minimum temperature, respectively). Maximum air temperatures were higher than 25 °C even in seven days during the full flowering of ‘Canetova’ (second flowering season; Figure S3). They were also higher than 25 °C in some of the full flowering days for ‘Dolga Šiška’ and ‘Ohridska Crna’ (both seasons; Figures S5 and S6). Minimum daily full flowering temperatures fell below 0 °C for ‘Canetova’ (four days in the first season; Figure S3) and for ‘G-2’ (two days in the second season; Figure S6).

3.4. Pollen Tube Growth Efficiency

3.4.1. Pollen Tube Number

The highest values of pollen tube number were observed in the upper part of the styles, followed by a drastic reduction in the lower parts of the pistil in all treatments (

Table 3. Pollen tube growth efficiency in the pistils of ‘Canetova’ under different pollination modes.

Pollination Variant (A)		STU s	STM t	BS u	OVR v	FP w (%)	FS x (%)
‘Canetova’ × ‘Burlat’		33.78 ± 8.55 a	19.98 ± 3.21 a	10.94 ± 1.12 a	2.76 ± 1.20 b	51.51 ± 16.63 a	11.05 ± 1.91 d
‘Canetova’ × ‘Lapins’		30.87 ± 9.36 b	18.12 ± 1.65 b	8.38 ± 0.45 c	2.34 ± 0.18 c	54.61 ± 16.04 a	35.80 ± 13.09 a
‘Canetova’ × ‘Rita’		30.51 ± 9.86 b	19.62 ± 2.85 ab	9.92 ± 1.57 b	3.18 ± 0.48 a	53.57 ± 3.97 a	31.67 ± 16.56 b
‘Canetova’ SP y		24.69 ± 13.30 c	3.48 ± 3.19 d	0.67 ± 0.48 e	0.15 ± 0.02 e	0.00 ± 0.00 c	0.00 ± 0.00 e
‘Canetova’ OP z		23.13 ± 12.31 c	13.11 ± 9.82 c	4.56 ± 2.55 d	1.15 ± 1.09 d	16.65 ± 18.27 b	24.04 ± 4.43 c
Flowering season (B)
I		33.85 ± 11.39 a	16.93 ± 7.80 a	6.45 ± 4.25 b	2.01 ± 1.67 a	31.33 ± 27.92 b	14.54 ± 9.29 b
II		23.34 ± 7.49 b	14.89 ± 7.81 a	7.34 ± 4.02 a	1.82 ± 0.93 b	39.21 ± 24.52 a	26.50 ± 19.54 a
A × B
‘Canetova’ × ‘Burlat’	I	41.39 ± 0.70 a	22.05 ± 0.34 a	11.79 ± 0.67 a	3.84 ± 0.33 a	66.67 ± 0.96 b	10.69 ± 0.97 e
	II	26.17 ± 2.89 d	17.92 ± 3.59 b	10.08 ± 0.72 b	1.69 ± 0.03 d	36.36 ± 1.08 f	11.41 ± 2.65 e
‘Canetova’ × ‘Lapins’	I	39.40 ± 0.91 ab	19.36 ± 0.24 b	8.41 ± 0.43 c	2.43 ± 0.01 bc	40.00 ± 1.32 e	24.35 ± 5.71 bc
	II	22.34 ± 0.62 d	16.87 ± 1.45 b	8.34 ± 0.57 c	2.25 ± 0.25 c	69.23 ± 0.87 a	47.27 ± 1.20 a
‘Canetova’ × ‘Rita’	I	39.55 ± 1.10 ab	22.16 ± 0.78 a	8.69 ± 0.70 c	3.56 ± 0.38 a	50.00 ± 1.04 d	16.86 ± 4.46 d
	II	21.47 ± 1.77 d	17.07 ± 0.43 b	11.17 ± 1.04 ab	2.81 ± 0.10 b	57.14 ± 0.37 c	46.48 ± 2.70 a
‘Canetova’ SP	I	36.83 ± 0.82 bc	6.39 ± 0.42 c	1.07 ± 0.17 f	0.11 ± 0.19 e	0.00 ± 0.00 h	0.00 ± 0.00 f
	II	12.53 ± 0.15 e	0.58 ± 0.14 e	0.27 ± 0.25 g	0.20 ± 0.26 e	0.00 ± 0.00 h	0.00 ± 0.00 f
‘Canetova’ OP	I	12.08 ± 0.67 e	4.16 ± 0.07 d	2.30 ± 0.50 e	0.17 ± 0.30 e	0.00 ± 0.00 h	20.74 ± 0.97 cd
	II	34.19 ± 3.46 c	22.06 ± 0.91 a	6.81 ± 0.86 d	2.14 ± 0.12 c	33.33 ± 1.23 g	27.35 ± 3.92 b
ANOVA
A		*	*	*	*	*	*
B		*	ns	*	*	*	*
A × B		*	*	*	*	*	*

^s STU: Upper third of the style. ^t STM: Middle third of the style. ^u BS: Base of the style. ^v OVR: Ovary. ^w FP: Fertilization percentage. ^x FS: Fruit set. ^y SP: Self-pollination. ^z OP: Open pollination. */ns: Indicates significance at p ≤ 0.05, or absence of significance, respectively, according to the LSD test. Mean values followed by the different lower-case letters in the column represent significant differences.

,

Table 4. Pollen tube growth efficiency in the pistils of ‘G-2’ under different pollination modes.

Pollination Variant (A)		STU s	STM t	BS u	OVR v	FP w (%)	FS x (%)
‘G-2’ × ‘Burlat’		14.01 ± 1.07 d	8.21 ± 1.09 e	4.22 ± 0.53 d	1.93 ± 0.48 b	41.46 ± 16.23 c	8.93 ± 8.34 b
‘G-2’ × ‘Lapins’		38.80 ± 16.09 c	23.97 ± 12.18 b	8.68 ± 2.11 b	3.51 ± 0.73 a	40.20 ± 2.44 c	11.97 ± 7.57 b
‘G-2’ × ‘Rita’		40.04 ± 15.46 c	30.17 ± 10.48 a	9.71 ± 0.99 a	3.16 ± 0.36 a	53.12 ± 3.66 a	8.12 ± 4.64 b
‘G-2’ SP y		42.35 ± 13.13 b	11.80 ± 1.59 d	2.05 ± 0.32 e	0.33 ± 0.37 c	3.35 ± 3.65 d	0.27 ± 0.42 c
‘G-2’ OP z		47.42 ± 7.33 a	19.06 ± 4.20 c	6.19 ± 1.06 c	1.94 ± 0.76 b	44.16 ± 10.09 b	22.44 ± 15.47 a
Flowering season (B)
I		40.44 ± 17.92 a	19.75 ± 9.8 a	6.33 ± 3.22 a	2.20 ± 1.07 a	36.71 ± 20.42 a	11.95 ± 7.06 a
II		32.80 ± 14.04 b	17.53 ± 11.70 b	6.01 ± 3.01 a	2.57 ± 1.26 a	36.20 ± 18.08 a	8.74 ± 13.91 a
A × B
‘G-2’ × ‘Burlat’	I	13.72 ± 0.99 f	7.47 ± 1.01 g	4.15 ± 0.76 d	2.28 ± 0.38 d	56.25 ± 0.67 a	16.43 ± 1.13 bc
	II	15.30 ± 0.17 f	8.94 ± 0.63 fg	4.25 ± 0.33 d	1.58 ± 0.27 e	26.67 ± 1.33 g	1.43 ± 1.96 e
‘G-2’ × ‘Lapins’	I	54.24 ± 0.94 a	35.04 ± 1.75 b	10.54 ± 0.70 a	2.85 ± 0.09 bc	42.31 ± 1.07 d	17.51 ± 4.95 b
	II	23.34 ± 0.32 e	12.92 ± 0.93 e	6.83 ± 0.58 c	4.17 ± 0.11 a	38.09 ± 0.64 e	6.04 ± 3.62 bcd
‘G-2’ × ‘Rita’	I	26.02 ± 2.48 d	20.67 ± 1.38 c	8.84 ± 0.32 b	3.06 ± 0.41 bc	50.00 ± 1.99 c	11.41 ± 3.39 bcd
	II	54.07 ± 1.11 a	39.67 ± 1.44 a	10.59 ± 0.29 a	3.25 ± 0.38 b	56.25 ± 0.54 a	4.82 ± 3.14 cde
‘G-2’ SP	I	54.23 ± 2.32 a	12.99 ± 1.44 ef	2.15 ± 0.46 e	0.33 ± 0.57 f	0.00 ± 0.00 i	0.54 ± 0.00 e
	II	30.54 ± 1.04 c	10.63 ± 0.22 f	1.95 ± 0.05 e	0.30 ± 0.12 f	6.67 ± 0.15 h	0.00 ± 0.00 e
‘G-2’ OP	I	53.98 ± 2.67 a	22.58 ± 2.43 c	5.99 ± 1.26 c	2.59 ± 0.41 cd	35.00 ± 1,57 f	13.46 ± 5.38 bc
	II	40.86 ± 0.36 b	15.53 ± 0.98 d	6.39 ± 1.06 c	1.28 ± 0.07 e	53.33 ± 0.65 b	31.42 ± 18.10 a
ANOVA
A		*	*	*	*	*	*
B		*	*	ns	ns	ns	ns
A × B		*	*	*	*	*	*

^s STU: Upper third of the style. ^t STM: Middle third of the style. ^u BS: Base of the style. ^v OVR: Ovary. ^w FP: Fertilization percentage. ^x FS: Fruit set. ^y SP: Self-pollination. ^z OP: Open pollination. */ns: Indicates significance at p ≤ 0.05, or absence of significance, respectively, according to the LSD test. Mean values followed by the different lower-case letters in the column represent significant differences.

,

Table 5. Pollen tube growth efficiency in the pistils of ‘Dolga Šiška’ under different pollination modes.

Pollination Variant (A)		STU s	STM t	BS u	OVR v	FP w (%)	FS x (%)
‘Dolga Šiška’ × ‘Kordia’		14.75 ± 4.87 c	6.58 ± 2.82 c	3.81 ± 1.82 b	1.68 ± 0.40 ab	11.73 ± 4.49 c	6.94 ± 3.80 c
‘Dolga Šiška’ × ‘Summit’		23.18 ± 6.70 a	11.59 ± 2.69 a	5.41 ± 1.01 a	1.87 ± 0.48 a	9.71 ± 2.30 b	9.39 ± 3.49 b
‘Dolga Šiška’ × ‘Sunburst’		21.54 ± 4.05 b	7.83 ± 1.11 b	3.84 ± 1.35 b	1.62 ± 0.57 b	27.50 ± 2.89 a	0.19 ± 0.47 d
‘Dolga Šiška’ SP y		15.10 ± 2.36 c	0.00 ± 0.00 e	0.00 ± 0.00 d	0.00 ± 0.00 d	0.00 ± 0.00 d	0.00 ± 0.00 d
‘Dolga Šiška’ OP z		22.39 ± 10.76 ab	3.22 ± 1.21 d	1.04 ± 0.66 c	0.35 ± 0.19 c	11.85 ± 3.9 c	17.43 ± 8.80 a
Flowering season (B)
I		18.97 ± 7.49 b	4.49 ± 3.49 b	1.98 ± 1.75 b	1.02 ± 0.83 b	10.55 ± 8.47 b	4.31 ± 4.11 b
II		19.81 ± 6.79 a	7.19 ± 4.89 a	3.65 ± 2.51 a	1.19 ± 0.92 a	13.77 ± 10.34 a	9.27 ± 9.85 a
A × B
‘Dolga Šiška’ × ‘Kordia’	I	10.32 ± 0.22 g	4.03 ± 0.54 d	2.16 ± 0.18 de	1.33 ± 0.17 bc	7.69 ± 0.78 e	3.64 ± 0.36 c
	II	19.19 ± 0.54 d	9.13 ± 0.14 b	5.46 ± 0.26 ab	2.02 ± 0.08 a	15.78 ± 0.92 c	10.24 ± 1.80 b
‘Dolga Šiška’ × ‘Summit’	I	17.23 ± 2.43 e	9.28 ± 1.35 b	4.67 ± 0.57 c	2.27 ± 0.27 a	11.73 ± 0.71 d	7.98 ± 2.10 b
	II	29.13 ± 0.40 b	13.89 ± 0.60 a	6.15 ± 0.74 a	1.48 ± 0.16 b	7.69 ± 0.94 e	10.79 ± 4.50 b
‘Dolga Šiška’ × ‘Sunburst’	I	17.91 ± 1.13 de	7.00 ± 0.56 c	2.66 ± 0.35 d	1.11 ± 0.05 bc	25.00 ± 0.44 b	0.38 ± 0.67 d
	II	25.17 ± 0.48 c	8.67 ± 0.83 b	5.02 ± 0.48 bc	2.14 ± 0.18 a	30.00 ± 1.42 a	0.00 ± 0.00 d
‘Dolga Šiška’ SP	I	17.20 ± 0.35 e	0.00 ± 0.00 f	0.00 ± 0.00 f	0.00 ± 0.00 e	0.00 ± 0.00 f	0.00 ± 0.00 d
	II	13.00 ± 0.77 f	0.00 ± 0.00 f	0.00 ± 0.00 f	0.00 ± 0.00 e	0.00 ± 0.00 f	0.00 ± 0.00 d
‘Dolga Šiška’ OP	I	32.21 ± 0.25 a	2.17 ± 0.45 e	0.45 ± 0.12 f	0.33 ± 0.15 d	8.33 ± 0.41 e	9.54 ± 0.40 b
	II	12.58 ± 0.34 f	4.28 ± 0.35 d	1.63 ± 0.22 e	0.31 ± 0.25 d	15.38 ± 0.84 c	25.32 ± 2.61 a
ANOVA
A		*	*	*	*	*	*
B		*	*	*	*	*	*
A × B		*	*	*	*	*	*

^s STU: Upper third of the style. ^t STM: Middle third of the style. ^u BS: Base of the style. ^v OVR: Ovary. ^w FP: Fertilization percentage. ^x FS: Fruit set. ^y SP: Self-pollination. ^z OP: Open pollination. *: Indicates significance at p ≤ 0.05 according to the LSD test. Mean values followed by the different lower-case letters in the column represent significant differences.

and

Table 6. Pollen tube growth efficiency in the pistils of ‘Ohridska Crna’ under different pollination modes.

Pollination Variant (A)		STU s	STM t	BS u	OVR v	FP w (%)	FS x (%)
‘Ohridska Crna’ × ‘Kordia’		12.21 ± 4.23 c	8.24 ± 4.73 b	4.59 ± 3.65 b	1.85 ± 0.73 a	43.16 ± 7.59 a	24.73 ± 14.71 a
‘Ohridska Crna’ × ‘Summit’		12.02 ± 6.25 c	6.05 ± 4.63 c	3.19 ± 2.13 c	1.76 ± 0.56 a	28.04 ± 2.21 d	14.19 ± 64.48 c
‘Ohridska Crna’ × ‘Sunburst’		24.32 ± 0.40 a	12.83 ± 4.63 a	5.89 ± 0.58 a	1.91 ± 0.52 a	33.33 ± 0.49 c	18.17 ± 13.02 b
‘Ohridska Crna’ SP y		9.97 ± 10.53 d	4.3 ± 4.76 e	0.96 ± 1.05 e	0.00 ± 0.00 b	0.00 ± 0.00 e	0.17 ± 0.41 d
‘Ohridska Crna’ OP z		22.66 ± 5.09 b	5.23 ± 2.95 d	2.29 ± 1.49 d	1.75 ± 0.99 a	34.75 ± 7.65 b	25.91 ± 14.71 a
Flowering season (B)
I		17.17 ± 8.73 a	6.85 ± 3.72 b	2.76 ± 1.81 b	1.32 ± 0.88 b	27.15 ± 14.70 b	10.48 ± 7.94 b
II		15.31 ± 8.26 b	7.81 ± 5.72 a	4.09 ± 3.15 a	1.59 ± 1.04 a	28.55 ± 18.37 a	22.89 ± 14.01 a
A × B
‘Ohridska Crna’ × ‘Kordia’	I	6.33 ± 0.34 g	1.83 ± 0.26 f	1.28 ± 0.25 f	1.25 ± 0.10 c	30.77 ± 0.74 d	22.98 ± 3.99 cd
	II	17.72 ± 0.16 d	10.26 ± 0.47 c	5.11 ± 0.49 c	2.26 ± 0.08 b	55.56 ± 0.88 a	26.47 ± 2.54 bc
‘Ohridska Crna’ × ‘Summit’	I	8.35 ± 0.97 f	3.95 ± 0.43 e	1.27 ± 0.12 f	1.22 ± 0.06 c	30.00 ± 0.75 d	9.16 ± 1.00 ef
	II	16.04 ± 0.33 e	12.55 ± 0.35 b	7.91 ± 0.31 a	2.50 ± 0.28 ab	26.08 ± 0.28 f	19.22 ± 5.31 d
‘Ohridska Crna’ × ‘Sunburst’	I	24.33 ± 0.58 b	11.94 ± 0.42 b	5.72 ± 0.83 bc	1.46 ± 0.16 c	33.33 ± 0.58 c	6.64 ± 0.91 f
	II	24.31 ± 0.27 b	13.71 ± 0.68 a	6.07 ± 0.23 b	2.35 ± 0.26 b	33.33 ± 0.63 c	29.69 ± 4.94 b
‘Ohridska Crna’ SP	I	19.57 ± 0.94 c	8.66 ± 0.73 d	1.92 ± 0.15 f	0.00 ± 0.00 d	0.00 ± 0.00 g	0.33 ± 0.58 g
	II	0.37 ± 0.05 h	0.00 ± 0.00 g	0.00 ± 0.00 g	0.00 ± 0.00 d	0.00 ± 0.00 g	0.00 ± 0.00 g
‘Ohridska Crna’ OP	I	27.25 ± 0.98 a	7.90 ± 0.14 d	3.36 ± 0.36 d	2.64 ± 0.27 a	41.67 ± 1.04 b	12.92 ± 0.68 e
	II	18.08 ± 0.59 d	2.56 ± 0.59 f	0.95 ± 0.13 e	0.86 ± 0.16 d	27.78 ± 0.86 e	39.07 ± 5.28 a
ANOVA
A		*	*	*	*	*	*
B		*	*	*	*	*	*
A × B		*	*	*	*	*	*

^s STU: Upper third of the style. ^t STM: Middle third of the style. ^u BS: Base of the style. ^v OVR: Ovary. ^w FP: Fertilization percentage. ^x FS: Fruit set. ^y SP: Self-pollination. ^z OP: Open pollination. *: Indicates significance at p ≤ 0.05 according to the LSD test. Mean values followed by the different lower-case letters in the column represent significant differences.

; Figure 3a). This reduction was especially pronounced in the self-pollination variant—there were no pollen tubes already in the middle part of the style in ‘Dolga Šiška’ (both flowering seasons; Table 5, Figure 3b) and in ‘Ohridska Crna’ (second season; Table 6). For other cultivars and flowering seasons, the longest pollen tubes ended their growth in the middle third and the base of the style in this pollination variant. Sporadically, pollen tubes were observed even in the ovary (‘Canetova’ SP, ‘G-2’ SP; Table 3 and Table 4). Incompatible pollen tubes, characterized by a broadened tip, or strong fluorescence along their entire length, were noticed in the self-pollination variant (Figure 3c) and the semi-compatible crosses (‘Canetova’ × ‘Rita’; ‘G-2’ × ‘Burlat’; ‘Dolga Šiška’ × ‘Kordia’; ‘Dolga Šiška’ × ‘Sunburst’).

In lower pistil parts, the values of pollen tube numbers were higher in cross-pollination than in the open-pollination variant and, particularly, in the self-pollination variant. Only in ‘G-2’ was the number of pollen tubes in open pollination competitive with those in cross-combinations along all pistil parts (Table 4); in ‘Ohridska Crna’, this was the case in the ovary (Table 6).

Pollen tube number in different pistil parts was influenced by variability factors and their interactions in ‘Dolga Šiška’ and ‘Ohridska Crna’ (Table 5 and Table 6), or pollination variant and variability factors’ interaction in ‘Canetova’ and ‘G-2’ (Table 3 and Table 4). The impact of the flowering season was more significant for later flowering cultivars.

3.4.2. Pollen Tube Growth Rate

In the cross-pollination variant, pollen tubes were observed in the ovary on the third day after pollination (Figures S7 and S8), except for the pollen tubes of ‘Lapins’ in the pistils of ‘G-2’ in the second flowering season (Figure S7b). The longest pollen tubes were located predominantly in the obturator zone (‘Dolga Šiška’), micropyle or nucellus (‘Canetova’, ‘G-2’, ‘Ohridska Crna’). On the sixth day after pollination, the pistils generally had the longest pollen tube in the micropyle or nucellus, while on the tenth day, they were found in the nucellus (Figure 3d) in all cross-pollinations. As for the open-pollination, pollen tube growth dynamics were slightly delayed, especially in the first fixation (‘Canetova’ OP, first season, Figure S7a; ‘Ohridska Crna’ OP, second season, Figure S8b), but by the last fixation terms, the values almost reached those of cross-pollination. The opposite tendencies for fertilization percentage values by flowering seasons had statistically manifested as the significance of the variability factors’ interactions (Table 3, Table 4, Table 5 and Table 6). As for the self-pollination, penetrations into the lower regions of the pistils were noticeable at later fixation terms.

3.4.3. Ovule Vitality and Unusual Pollen Tube Growth

Fluorescence of the primary ovules (Figure 3e), as a sign of their senescence/degeneration, mostly recorded higher values in the cross-pollination than in the open-pollination variant (Figure S9). The emergence of ovule fluorescence was the most pronounced in the ovaries of ‘Dolga Šiška’ (on average, 63.68% and 26.34% for cross-pollination and open-pollination variants, respectively), and the least in ‘G-2’ (27.72%; cross-pollination) and ‘Ohridska Crna’ (22.22%; open-pollination). In addition, it was more evident in the first flowering season; the opposite tendency was observed only in ‘G-2’, pollinated with ‘Burlat’ and ‘Rita’ (Figure S9b).

The rate of the ovaries with the ‘unusual growth’ of pollen tubes (Figure 3f–h) before their entrance into the ovules was highest in ‘Dolga Šiška’, and lowest in ‘Ohridska Crna’ (on average, 11.89% and 0.00% for the cross-pollination variant, respectively). In the open-pollination variant, this rate was 0.0% for all cultivars and seasons, except for ‘Canetova’ OP in the second flowering season (3.17%; Figure S9a).

3.4.4. Fruit Set

Pollination variant, flowering season, and/or their interactions significantly influenced fruit set in all autochthonous cultivars (Table 3, Table 4, Table 5 and Table 6). In the cross-pollination variant, the highest and the lowest values of the fruit set (as an average of three cross-pollinations) were achieved in ‘Ohridska Crna’ (19.03%) and ‘Dolga Šiška’ (5.51%), respectively. The same cultivars had the highest and the lowest values of fruit set in the open-pollination variant (25.91% and 17.43%, respectively). Fruit set was generally better in the second flowering season for ‘Canetova’, ‘Dolga Šiška’, and ‘Ohridska Crna’, whereas the opposite tendency was observed for ‘G-2’ in all cross-combinations. As for the self-pollination variant in ‘G-2’ and ‘Ohridska Crna’, a low fruit set was observed in the first flowering season (0.54% and 0.33%, respectively; Table 4 and Table 6).

3.4.5. Correlations Among the Reproductive Parameters

Correlation matrices obtained for the cross-pollination and open-pollination variants (

Table 7. Pearson’s coefficients of linear correlations between the reproductive parameters.

Parameter	STU a	STM b	BS c	OVR d	OF e	UB f	FP g	FS h
Cross-pollination variant
STU	/
STM	0.94 *	/
BS	0.77 *	0.84 *	/
OVR	0.60 *	0.63 *	0.71 *	/
OF	0.07	−0.14	−0.21	−0.13	/
UB	−0.28	−0.32	−0.39	−0.20	0.46 *	/
FP	0.40	0.51 *	0.59 *	0.58 *	−0.40	−0.37	/
FS	−0.01	0.12	0.33	0.15	−0.46 *	−0.18	0.52 *	/
Open-pollination variant
STU	/
STM	0.79 *	/
BS	0.70	0.95 *	/
OVR	0.67	0.74 *	0.70	/
OF	0.14	−0.16	−0.30	−0.36	/
UB	0.15	0.56	0.52	0.33	−0.52	/
FP	0.61	0.61	0.70	0.73 *	−0.30	0.15	/
FS	−0.31	−0.02	0.08	−0.22	−0.80 *	0.19	0.25	/

^a STU: Pollen tube number in the upper third of the style. ^b STM: Pollen tube number in the middle third of the style. ^c BS: Pollen tube number in the base of the style. ^d OVR: Pollen tube number in the ovary. ^e OF: The rate of ovule fluorescence. ^f UB: The rate of unusual behaviour of pollen tubes. ^g FP: Fertilization percentage. ^h FS: Fruit set. *: Indicates significance at p ≤ 0.05.

), showed positive or negative correlations among some reproductive parameters, ranging from medium-strong to very strong. Medium to very strong positive correlations were obtained among pollen tube numbers in different pistil regions, and these parameters also influenced fertilization percentage. Ovule fluorescence did not affect fertilization percentage but affected fruit set in both cross-pollination and open-pollination variants (r = −0.46; r = −0.80, respectively). In addition, ovule fluorescence positively correlated with the rate of unusual pollen tube growth in the cross-pollination variant (r = 0.46). Correlation-regression analysis also showed that the observed reproductive parameters were not related in the same manner, or strength, in open pollination as in the cross-pollination variant.

4. Discussion

4.1. S-Haplotype Identification

In the present study, we provided data on the identification of S-haplotypes in four sweet cherry cultivars of autochthonous origin and confirmed S-haplotypes in six economically important, globally known cultivars used as pollenizers. In the autochthonous material, seven S-alleles were identified, namely S₂ to S₆, S₁₂, and S₄₀, with S₄₀ being revealed in this study for the first time. As reported in the literature, the number of identified alleles in sweet cherry accessions varied from eight in Croatian and Czech genotypes [66,67], through ten in Turkish and Spanish genotypes [60,68], to as many as 17 alleles identified in Italian accessions [69]. Based on the identified alleles, the autochthonous genotypes were assigned to four IGs out of the 72 IGs and the ‘0’ group [24]. ‘G-2’, ‘Canetova’, and ‘Dolga Šiška’ belong to IGs IV, XV, and XXII, respectively, whilst ‘Ohridska Crna’, with its unique S-genotype, falls into the universal pollen donor group ‘0’, which includes 26 genotypes [24].

Concerning ‘Ohridska Crna’, a discrepancy in the identification of the S-genotype was found between our study and the updated table [24], reporting the S-genotypes of 1700 sweet cherries. In our study, the genotype S₄S₄₀, instead of S₃S₁₂, was determined for ‘Ohridska Crna’. The S₃S₁₂ genotype in our study was identified in another Macedonian cultivar, ‘Dolga Šiška’. Indeed, both cultivars are commonly grown in North Macedonia, especially in the southwestern regions near Ohrid, as well as in several other Balkan countries [46], which is why their names are often confused. Ercisli et al. [66] reported S₃S₁₂ for ‘Ohridska Crna’ collected from the continental part of Croatia, while Radičević et al. [70] stated the same S-allelic constitution for a genotype designated as ‘Ohridska 1’ (but not ‘Ohridska Crna’), which was obtained from Macedonia. Although our previous study on the materials received from Macedonia indicated S₄S_x for ‘Ohridska Crna’ and S₃S₁₂ for ‘Ohridska Dolga Šiška’ [47], as part of the CherrySeRB project we performed S-genotyping on material that was morphologically and pomologically characterized, detecting S₄S₄₀ in ‘Ohridska Crna’ and S₃S₁₂ in ‘Dolga Šiška’. Discrepancies in S-genotypes have also been found in other studies, e.g., in Italian sweet cherry accessions [69]. Therefore, this type of work is very important and represents an opportunity to eliminate errors that occurred during the establishment of collections.

Although the number of autochthonous sweet cherry cultivars examined in our study is quite small, making it difficult to discuss allele frequencies, the most frequent allele in this material was S₃ (25%). Besides the new S₄₀ allele, the presence of the other identified alleles was confirmed in our previous study on the relative occurrence of S-alleles in local Serbian, Macedonian, and Bulgarian sweet cherry genotypes [71]. The S₃ allele was the most frequent one observed in sweet cherry accessions from Croatia (39%) [66], Spain (38%) [60], the Czech Republic (34.4%) [67], Turkey (29.6%) [68], and Italy (25%) [69]. In contrast, the S₁₂ allele was not found in the Spanish germplasm [60] but was frequently observed in genotypes originating from Croatia and Turkey (19% and 7.4%, respectively) [66,68], as well as from Serbia and North Macedonia (13.6%), which could be related to their long shared history [71]. The geographical distribution of certain S-haplotypes could indicate a common origin of genotypes found in nearby regions [60]. There are differences in the occurrence of other alleles, such as S₂ and S₄, which were rare in Italian accessions, each with a frequency of 1% [69]. Similarly, the S₂ allele was found with the same frequency (1%) in western Spain but was not detected in genotypes from the eastern and northern parts of the country [60]. In our previous study, the occurrence of this allele was 15.9% [71], which was similar to the frequency reported in Turkish germplasm (14.8%), while its frequency in Croatian landraces was 8% [66,68]. As for S₄, a higher occurrence of this allele was observed in genotypes from northern Spain, the Czech Republic, and Turkey, with frequencies of 23%, 21.9%, and 13.6%, respectively [60,67,68]. The high frequency of the S₅ allele was observed in Ukrainian germplasm (25.9%), which clearly distinguishes these genotypes from those originating in other parts of Europe [67]. This allele was rare in Italian landraces and Spanish genotypes from the western region of the country (1%) and was absent in genotypes from the northern and eastern regions [60,69]. Our previous study revealed a low frequency of the S₅ allele (2.3%) in Serbian, Macedonian, and Bulgarian genotypes [71], which was slightly lower than its occurrence in Turkish (5%) and Croatian (7%) germplasms [66,68]. The S₆ was one of the most frequent alleles in Spanish and Italian landraces (26% and 19%, respectively) [60,69], while it was also common in Croatian and Turkish genotypes (8% and 11.1%, respectively) [66,68], as well as in genotypes (9.1%) from Serbia, Macedonia, and Bulgaria [71].

The reason why specific alleles are common in certain sweet cherry germplasm while others remain rare is still unclear. However, Cachi and Wünsch [60] and Marchese et al. [69] have suggested that some alleles may be associated with traits that confer adaptation to specific agroecological conditions or may result from founder effects and selection events [72].

4.2. Fertilization Efficacy and Impact of Pollenizer

Fertilization success in sweet cherries depends on the pollenizer’s reproductive behaviour arising from its genotypic specificity, and within this research, we tried to determine the measure of that impact. Considering different regions of pistils, ‘Burlat’ and ‘Rita’ induced the highest number of pollen tubes in the pistils of ‘Canetova’ and ‘G-2’, respectively. As for the late-flowering autochthonous cultivars, the highest number of pollen tubes in the pistils of ‘Dolga Šiška’ and ‘Ohridska Crna’ induced ‘Summit’ and ‘Sunburst’, respectively, and this effect was stably manifested along the different pistil’s parts. It is in line with the previous findings [73,74], according to which ‘Summit’ tends to form a high number of pollen tubes in vivo. On the other hand, ‘Kordia’ induced the lowest/almost the lowest number of pollen tubes, especially in the upper pistil parts of both late flowering cultivars. Additionally, ‘Kordia’ generally tends to form a smaller number of pollen tubes in vivo [74]. Analyzing pollen tube growth parameters in conjunction (Table 3, Table 4, Table 5 and Table 6), it seems that the pollen tube number in the upper pistil did not affect the outcome of fertilization; this initial growth strength below the stigma is related to genotypic characteristics, as well as seasonal factors associated with microsporogenesis. This is consistent even for the self-pollination variant, where the pollen tube number in the upper pistil part was comparable to those in other pollination variants. In ‘G-2’ SP, this parameter had a significantly higher value than in cross-pollinations (Table 4).

Correlation coefficients (Table 7) showed that the pollen tube number in the upper part of the style affected the pollen tube numbers in the lower parts, especially in the cross-pollination variant. On the other hand, no big differences were observed in terms of pollen tube growth dynamics by fixation terms, considering that on the third day after the ‘zero date’ pollen tubes were at least in the obturator zone, and on the tenth day in the nucellus, in almost all cross-combinations and open-pollination variants (Figures S7 and S8). On the third day after pollination, pollen tubes were also observed in the ovary of apricot [75], plum [76], sweet cherry [43,74], and sour cherry [77]. Due to the restricted experimental possibilities of monitoring the competitiveness of different pollenizers in the same pistil, it is hard to explain the relationship between the pollen tube number, and the growth rate. For example, ‘Lapins’ had the smallest number of pollen tubes in the pistils of ‘Canetova’, but an excellent growth rate (Table 3; Figure S7a). The same pollenizer in the pistils of ‘G-2’ induced a relatively high number of pollen tubes, but their growth was somewhat slower, especially in the second flowering season (Table 4; Figure S7b).

Our results showed the relative validity of the ‘parameters of number’ in the middle part and the base of the style, and the ovary; they correlated to fertilization percentage in cross-pollination (medium positive correlations; r = 0.51, 0.59 and 0.58, respectively; Table 7). In the open-pollination variant, a strong positive correlation was shown between the pollen tube number in the ovary and fertilization percentage (r = 0.73; Table 7). This suggests that the number of pollen tubes in different pistil parts, particularly in the upper, was certainly not a decisive factor in determining pollenizer efficiency, but that it assumed a better chance of a successful fertilization outcome. Nevertheless, the reproductive behaviour of pollenizers in our work indicates genotypic specificities and, as will be shown below, these specificities cannot be considered themself, but in the context of the pollinated cultivar, flowering temperatures, and their complex interactions.

4.3. Fertilization Efficacy and Impact of Pollinated Cultivar

In this research, inconsistencies in the behaviour of pollenizers, related to the specific impact of female sporophytes, were also observed. For example, ‘Burlat’, which induced the smallest number of pollen tubes in the pistils of ‘G-2’ (Table 4), simultaneously induced the highest pollen tube number in the pistils of ‘Canetova’ (Table 3). Similar behaviour showed ‘Summit’, which had the highest and the lowest pollen tube number in the pistils of ‘Dolga Šiška’ (Table 5) and ‘Ohridska Crna’ (Table 6), respectively. These findings are in line with those of Hedhly et al. [27], who stated that the same pollenizer can exhibit different behaviour when its pollen tubes grow in style transmitting tissue of different cultivars. Style tissue can be more or less conducive to pollen tubes of different pollenizers, at first through the GSI reaction, which eliminates support for gametophytes with the same genetic constitutions. That was most visible in the self-pollination variant, where style tissue stopped the growth of pollen tubes with the same S-haplotypes, regardless of the pollen quality, its abundance, or initial growth strength. In some cases, the GSI reaction was quite strong already in the upper part of the style (‘Dolga Šiška’; Table 5), and somewhere it allowed penetrations into the lower pistil parts and even the ovary (‘G-2’, ‘Ohridska Crna’; Table 4 and Table 6), with a low fruit set recorded. In sweet cherries, incompatible crosses resulted in a fruit set of less than 3% [78]. Certain sweet cherry S-haplotypes had higher incompatibility-breakdown rates after selfing than others, and breakdowns are also related to the air temperature during the flowering [79]. That was noticeable in cross-pollinations as well; the aforementioned male gametophytes of ‘Burlat’ in the pistils of ‘Canetova’ were better supported (fully compatible) than in pistils of ‘G-2’ (semi-compatible pollination).

Genotypic specificities of pollinated cultivars could also explain the different behaviour of pollenizers in lower pistils’ parts, through the appearance of fluorescence, as a key indicator of the ovule vitality lost. In the cross-pollination variant, this biologically and agronomically important attribute was the most pronounced in ‘Dolga Šiška’ (Figure S9c). Hence, the efficiency of pollen tube growth, expressed through the fertilization percentage and fruit set, was the weakest in its pistils (Table 5). ‘G-2’ and ‘Ohridska Crna’ showed a lower incidence of ovule fluorescence (Figure S9b,d), so the effectiveness of the pollenizers was much better for these cultivars (Table 4 and Table 6). Genotypic specificities in terms of the primary ovule’s function and its loss were also found in plums [80], almonds [81], apricots [82], sweet cherries [37], and sour cherries [77].

Ovule fluorescence rates had higher values in cross-pollination than in the open-pollination variant (Figure S9), which can be explained by the effect of emasculation, i.e., the flower injury on the function of female flower elements, which was previously reported [83]. Interestingly, the highest percentage of ovaries with unusual pollen tube growth before their entrance into the micropyle (cross-pollination variant) was also recorded in ‘Dolga Šiška’. The expression of this phenomenon (Figure S9) showed a certain kind of parallelism to ovule fluorescence across genotypes and flowering seasons, which was confirmed by Pearson’s correlation coefficients in the cross-pollination variant (medium strong correlation, r = 0.46; Table 7). According to Herrero [84], the female structures encountered by the growing pollen tube have to attain particular developmental states before they are competent to ‘encourage’ the passage of the pollen tubes.

4.4. Fertilization Efficacy and Impact of Flowering Temperatures

The higher temperatures during the full flowering affected pollen performance in vivo, influencing a lower number of pollen tubes, in particular in the cross-pollination variant (except for ‘Lapins’ pollen tubes in the pistils of ‘G-2’). They also influenced faster pollen tube kinetics up to the third day after ‘zero-date’; after this term, pollen tube growth kinetics were relatively uniform by flowering seasons, independently of pollination variant, pollenizer, or temperature. Our results are in line with Hedhly et al. [85], according to whom a small increase in the mean temperature (1–3 °C) during the first two weeks after sweet cherry anthesis, has a direct effect on reducing the number of pollen tubes reaching the base of the style, which reflects the generally better adaptation of this species to cooler flowering temperatures. The same authors stated that higher temperatures also increase the rate of pollen tube growth, but only in the first few days after pollination. The aforementioned inconsistent behaviour of ‘Lapins’ was also visible in the pistils of ‘Canetova’, but to a lesser extent than in ‘G-2’. It seems that the growth of ‘Lapins’ pollen tubes was not as dependent on temperature conditions as that of other pollenizers. Different behaviour of pollenizers concerning the influence of flowering temperatures was previously shown in sweet cherry [27,74,85], sour cherry [77], peach [86], apricot [87], and plum [88]. As for the open-pollination variant, the inconsistency resulted from the different (unknown) pollenizers, uneven amounts of their pollen, and time of pollination.

The relative independence of ‘Lapins’ concerning different flowering temperatures, makes this cultivar superior among the early flowering pollenizers. This independence is related primarily to the unchanged growth dynamics of its pollen tubes, regardless of their reduced/not reduced number in conditions of higher flowering temperatures. The fact that it is a SC cultivar, with high production of good quality pollen and an early, extended flowering time [52], contributes to this attitude. As for late flowering cultivars that were exposed to higher temperatures, together with pollenizers, rather than earlier flowering, ‘Summit’ and ‘Kordia’ showed the best pollen performance in vivo in the pistils of ‘Dolga Šiška’ and ‘Ohridska Crna’, respectively. In light of previous results [74], this finding is not surprising for ‘Summit’; ‘Kordia’ shows good results as a pollenizer at higher temperatures only when pollinating well-adapted main cultivars. Higher temperatures had a positive effect on the faster growth of pollen tubes, presumably increasing the probability of successful fertilization from a broader range of pollen donors [89].

As for female reproductive behaviour, our results imply the significant impact of full flowering temperatures on the ovule fluorescence and, consequently, the fertilization percentage and fruit set. This is especially true for ‘Dolga Šiška’, which was exposed to temperatures exceeding 25 °C just before and at the beginning of full flowering in the first season (Figure S5); these temperatures had a detrimental effect on its ovule vitality. On the other hand, the behaviour of ‘Ohridska Crna’, despite the exposure to similar flowering temperatures (Table 2; Figure S4), showed better results in terms of ovule fluorescence. As for early flowering cultivars, ‘G-2’ had ovule fluorescence values similar to those of late-ripening ‘Ohridska Crna’, but it should also be kept in mind that its flowers were not exposed to such high temperatures (Figure S6). This tendency was not clearly visible in ‘Canetova’, owing to minimum daily temperatures before and during full flowering that fell below 0 °C during both flowering seasons (Figure S3), which could have injured ovules and affected ovule fluorescence results.

The rate of ovule fluorescence negatively correlated with the fruit set, and the strength of this correlation was higher in open-pollination (strong correlation; r = −0.80) than in the cross-pollination variant (medium-strong correlation; r = −0.46) (Table 7). This incongruity leads to the conclusion that flower emasculation, combined with higher temperatures, unequally affected the different cultivars. The cultivars with a tendency to loose ovule vitality fast (‘Dolga Šiška’), were affected more than cultivars with generally better ovule vitality (‘Ohridska Crna’), and this created differences in correlation matrices patterns. Previous studies related to sweet cherry [90] and other Prunus species [91,92,93,94] stated that the maximum daily temperatures over 25 °C caused significant modification of the flowering phenophase, which resulted in a drastic reduction in fruit set. It has been shown that higher temperatures affect flower development [94,95,96], shorten stigmatic receptivity [38,86,97], and accelerate ovule degeneration [37,80,90]. In addition, the need to accumulate the amount of chilling during winter and overcome dormancy, to finally achieve flowering in spring, is also genotype-dependent in sweet cherry [98], and in other stone fruits [99,100].

The above-mentioned observations imply that pollen–pistil interactions became more complex under the influence of different flowering temperatures. The statistical significance of the interactions of the variability factors shows that their effects should be not considered independently, but as an overlapping effect of several aspects: (i) pollenizers/pollinated genotype and its behaviour at different temperatures; (ii) the influence of pollinated cultivar on pollen performance; (iii) the influence of temperature on male–female relations; (iv) unequal effect of emasculation presence/absence on different female genotypes; and (v) the effect of other factors that were not considered in this work, but surely could contribute to the final fertilization outcome. The last of the summarized aspects refers not only to flowering/pollination events, but also to the numerous, long-lasted events that occurred before flowering (flower bud differentiation and dormancy), and after it (fruit development). All these steps are dependent on factors that can be classified as environmental, and those that lie inside the bud/flower [101]. The air temperature was certainly one of the most important climate factors at all stages, but other environmental factors, including orchard management (conditions without irrigation), horticultural practises, as well as pollinators (bees and others) activity (open pollination variant), could also contribute to the successful flower-to-fruit development, or its failure.

The idea of conducting the research with pollination experiments at several typical Balkan cherry-growing locations arose from multi-year data on altered temperature conditions during the flowering phenophase, predictions of future temperature changes, and challenges and problems in cherry fruit production resulting from these conditions. In that sense, we counted on the occurrence of higher temperatures during the flowering and had at least one such season for each autochthonous cultivar, so they have fully manifested (and pollenizers as well) their plasticity in terms of reproductive traits.

The results obtained should be considered from the several aspects that include future horticultural solutions, arising from the smart agriculture concept. From one point of view, our results contribute to genetic diversity knowledge, which is essential for climate-smart agriculture, providing the raw material for future breeding because it lays the groundwork for breeding climate-resilient crops [102]. Genetic diversity allows breeders to identify genes associated with GSI (as a new S₄₀ allele in our research) and climate-resilient traits, such as heat tolerance in the reproductive sense (‘Ohridska Crna’). The new concept of precision-designed plant breeding is especially suitable for polygenic traits for which it is difficult to design molecular markers showing major effects or for stress-related traits associated with high phenotyping costs to achieve stress-induction conditions [103]. On the other hand, our findings, related to a strong correlation between ovule fluorescence and fruit set in the open pollination variant, follow previous findings for commercial sweet cherry cultivars [37,74]. This opens up the future possibility of yield prediction in commercial plantations, based on the assessment of yield potential, monitoring flowering temperatures, and rapid ovule vitality screening in days after the flowers open. When predicting future excessive or insufficient yield, suitable horticultural measures could be taken (thinning or giving exceptional largeness to a relatively small number of fruits, respectively). This is under a new approach, where plant phenotypic data, environmental data, and abiotic stress data together form the ‘big data for horticultural cultivation’, that make smart decisions possible [104].

5. Conclusions

Sweet cherry cultivars are extremely dependent on climatic conditions for a consistent and sufficient fruit set, and even more so given the impacts of climate change. The diversity investigations that include genotyping and reproductive-phenotyping are of great importance in gathering genotype-environment interaction data, crucial from the aspects of breeding work, orchard design, and management. This type of study implies, even with a relatively small number of autochthonous genotypes, a large volume of fieldwork and sampling at different locations, with all the challenges that such studies entail. Hence, coordinated efforts at the European or wider level are needed to characterize, preserve, and utilize cherry genetic resources. Autochthonous Balkan sweet cherries investigated in this research showed different adaptability to agroecological conditions in which the pollination experiments were conducted, in the sense of female reproductive behaviour. Our findings undoubtedly point to the exceptional reproductive performance of ‘Ohridska Crna’. Owing to the unique S-haplotype (S₄S₄₀) revealed in this work, the transmitting tissue of its style is supportive of the pollen tubes of the pollinisers with the appropriate flowering time. Generally long-lived and vital primary ovules, and the stability of this trait under flowering temperatures higher than 25 °C, make this cultivar promising for future breeding efforts aimed at developing cultivars that can better respond to the challenges of global warming in the reproductive sense, particularly given the severe bottleneck caused by low diversity in cultivated sweet cherries. It would be interesting to test if the demonstrated tolerance of ‘Ohridska Crna’ in relation to higher flowering temperatures also implies tolerance in other climate-change dependent traits important for cherry fruit production (high temperatures during buds differentiation, winter chill requirements). An open question for all stakeholders in cherry science and production is a response to the requests for temperature-tolerance and exceptional quality of cherry fruits, which is difficult to unite within a genotype.