Study species

Daphne arbuscula is a long-lived, evergreen shrub that grows to a height of 10–30 cm, has coriaceous leaves, and bears bisexual, strongly fragrant flowers ranging in colour from pink to purple, rarely even white (S1 Fig in S1 File) [18–20, 40]. The species is diploid, with a chromosome count of 2n = 2x = 18 [35, 39, 41]. Fruits are one-seeded hard dry or fleshy drupes in brownish to yellow-green colours, often enclosed within marcescent hypanthium structures (S1 Fig in S1 File) [19, 38].

Sampling and design of the study

Four localities were selected to represent the entire range of ecological conditions within the whole distribution range of D. arbuscula in Muránska planina Mts (Fig 1) [19]. Moreover, the presence of clonal growth in D. arbuscula [19] made it difficult to determine the exact number of genetically distinct individuals. To reduce bias in the final results and the interpretation of flower and fruit production, only individuals spatially separated by smaller rock formations or vegetation were considered, minimizing the chances of clone inclusion. Regarding population size, Poludnica is the largest locality, with over 1400 shrubs, followed by Veľká Stožka, with approximately 1100 plants, then Malá Stožka with around 500 shrubs, and lastly, Šiance, the smallest locality with about 115 shrubs. The precise estimation of shrub numbers is challenging due to the highly inaccessible rocky terrain, making precise counting essentially impossible. The study sites, Poludnica and Šiance, predominantly face south and southeast and are situated at lower altitudes (approximately 800 m and 950 m above sea level). They represent a niche characterized by drier and warmer climatic conditions and are hence referred to as ’warm-microclimate’ localities. The Šiance (SI; Fig 1; S1 Fig in S1 File) population is limited to small calcareous rock bodies, with most individuals located in open, sunny rocks (SI41–SI46). A few individuals (SI47–SI50) occur in shaded rock crevices beneath a mixed forest canopy on northern-oriented rocks. In contrast, the other pair of populations, Malá Stožka and Veľká Stožka, represent a niche with colder and more humid climatic conditions, predominantly facing north, and are located at higher altitudes (approximately 1150 and 1275 m above sea level). These localities are hence referred to as ’cold-microclimate’ sites.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Distribution map of studied populations of Daphne arbuscula in the area Muránska planina Mts in the Western Carpathians and distances among the studied populations.

Population codes are as follows: MS–Malá Stožka, PO–Poludnica, SI– Šiance, VS–Veľká Stožka.

https://doi.org/10.1371/journal.pone.0300819.g001

Ten mature individuals of approximately the same size were selected from each population and marked with a unique metal identifier for subsequent identification purposes (S1 Table in S1 File). All the examined populations were of sufficient size to ensure that the research did not negatively impact their fitness and viability. Population size was estimated using data from the monitoring of D. arbuscula conducted by the Muránska planina National Park Administration. Authorization for field research and the collection of D. arbuscula samples, designated by permit number 5408/2017-6.3, was granted by the Ministry of Environment of the Slovak Republic on July 12, 2017. This permission facilitated the collection of all analyzed materials from the designated study locations throughout the entire investigative period, spanning from 2017 to 2019.

Flower and fruit production

The fruits typically ripen within 4–5 weeks; however, this process is highly individual-specific and influenced by climatic conditions. Subsequently, the fruits disperse rapidly and haphazardly from shrubs within a few days. Therefore, after the majority of the flowering shoots on the studied individuals had been in the full-bloom stage for at least a few days, securing pollination, they were isolated using nylon bags to ensure precise analysis of the fruit set.

For each of the ten selected individuals in all populations, fifteen flowering shoots with one inflorescence were randomly chosen and isolated using nylon bags with a mesh size of approximately 0.5 mm (S1 Fig in S1 File). A total of 150 shoots per population were isolated during each season. The flowering period of this species is relatively short, spanning two to three weeks in May and/or June. Populations from cold-microclimate localities often exhibit delayed flowering, sometimes up to one month compared to populations from warm-microclimate sites [20, 23, 36]. The isolation process therefore took place in May and June, depending on local conditions (S1 Fig in S1 File). The number of flowers in each inflorescence was recorded, and the fruits were collected between June and July, once they had ripened and started dropping from the shoots. The number of fruits formed per branch was recorded. The study was conducted over two growing seasons, from April to July in both 2018 and 2019.

Previous studies have identified fruit polymorphism, which includes two distinct fruit morphotypes that differ in colour and mesocarp thickness and are dispersed at approximately the same time [23, 38]. However, it remains uncertain whether these morphotypes represent true fruit polymorphism or different ontogenetic stages of the fruit ripening process. Despite seeds from the ’small-fruit’ morphotype appearing to have lower viability compared to those from the ’big-fruit’ morphotype, their viability still exceeds a threshold of 50% [38]. This means that both morphotypes might contribute significantly to the sexual reproduction of this species, and therefore we decided not to differentiate the collected fruits by morphotype in this study.

Generalized linear mixed models (GLMM) [42] were utilized to investigate temporal and spatial differences in flower and fruit production of D. arbuscula. The models incorporated random effects of repeatedly measured individuals and fixed effects of time (years 2018 and 2019), population (Šiance, Malá Stožka, Veľká Stožka, and Poludnica), and their interaction. Initially, the number of flowers was fitted using a GLMM with a Poisson error distribution and a logarithmic link function. However, the model residuals displayed considerable underdispersion (φ = 0.28), which can lead to overestimated standard errors and misleading statistical inference [43]. To address this issue and minimize the type II error rate, we refitted the data using a flexible Conway-Maxwell-Poisson distribution, allowing us to model a wide range of dispersion levels [44]. No further violation of the model assumptions was detected in diagnostic plots of residuals.

The probability of fructification was calculated as a proportion of flowers that developed fruits out of all flowers recorded on a given plant. Initially, we modelled the data as binomial proportions with a logit link function. However, the binomial GLMM showed significant overdispersion (φ = 1.33), potentially leading to inflated type I error rates [45]. As a result, we refitted the data using a model with beta-binomial distribution [46] and a complementary log-log link function, efficiently handling the overdispersion. Likelihood ratio tests were employed to assess the significance of the model terms [42]. Whenever a significant overall effect was identified (α = 5%), we conducted a series of post hoc tests with Tukey-adjusted probabilities to compare the estimated marginal means [47]. Marginal pseudo-determination coefficients were calculated for each GLMM to quantify the proportion of the total variation explained by the random and fixed effects (R2c) and fixed effects only (R2m) [48].

The analyses were performed in R v. 4.2.0 [49] using the libraries DHARMa [50], emmeans [51], ggplot2 [52], glmmTMB [53] and performance [54].

Breeding system study

To test for autogamy, nylon bags with a smaller mesh size of less than 0.2 mm were utilized to prevent pollinators from transferring pollen from other inflorescences or plants. A preliminary study (S2 File) showed that a mesh size of 0.5 mm was not effective at preventing pollination by thrips (Thysanoptera), known to be important pollinators of D. arbuscula [17, 24]. During the 2018 growing season, three inflorescences with unopened buds were isolated from each individual in all populations, totalling 120 shoots. Furthermore, to determine compatibility (SC vs SI), assisted pollination studies were conducted on a subset of plants (3–5 shrubs) at the Poludnica and Šiance localities. After the flowers opened, the nylon bags were removed from the isolated inflorescences, and the flowers were manually pollinated between flowers within the same inflorescence using a delicate, hand-made brush. Following pollination, the flowers were enclosed in nylon bags to prevent pollen transfer between flowers or from other plants via animal vectors. Bags from all studies were collected once the fruits ripened.

Genetic analyses

We conducted a comprehensive genetic study on a species that lacked previous genetic data. Thus, to obtain the first insight into the genetic variation of the studied species, we employed traditional molecular markers, specifically the ITS1-5.8S-ITS2 region of ribosomal DNA (referred to as ITS) and the rpl32-trnL^(UAG) region of chloroplast DNA (S3 File). For a more detailed investigation of population-level variability, we utilized double-digest RAD sequencing (dd RADseq; all 40 individuals).

Genomic DNA was extracted from young dried leaves of all sampled individuals using the Invisorb^® Spin Plant Mini Kit (Invitek Molecular GmbH, Berlin, Germany) following the manufacturer’s instructions, supplemented with PVP and 6 μL of RNase A. The extracted DNA was then purified using AMPure XP (Beckman Coulter Inc., Brea, California, USA).

Interpopulation, intrapopulation, and seasonal variability in flower and fruit production

Flower production of D. arbuscula significantly varied among populations, although the differences were contingent upon the year of observation (time × population: χ2 = 41.8, df = 3, p < 0.0001). In 2018, the number of flowers produced across populations was relatively consistent (Fig 2A). However, the following year, a decline in flower production was observed across all populations, with the most significant reduction occurring in Malá Stožka, which diverged significantly from the flower number of Poludnica (t = 3.5, df = 1166, p = 0.0029) and Veľká Stožka population (t = 4.2, df = 1166, p = 0.0002). Although the overall model was highly significant (χ2 = 508, df = 8, p < 0.0001), the flower counts data were fitted relatively poorly (R2c = 0.09, R2m = 0.04).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Seasonal variability in sexual reproduction of the four studied populations of Daphne arbuscula.

(A) the number of flowers in the inflorescence; (B) the probability of fructification. The estimates based on Generalized Linear Mixed Models (GLMM) are represented by circles, with their 95% confidence intervals displayed as error bars.

https://doi.org/10.1371/journal.pone.0300819.g002

Fruit development took between 5 and 7 weeks after anthesis, depending on microclimatic conditions. The fruit production was low, ranging from 8.97% to 48.49% across populations and both growing seasons (Table 1; Fig 2B). The populations showed significant differences in fruit production between the years (time × population: χ2 = 258, df = 3, p < 0.0001). In 2018, the fruit set was significantly higher in the warm-microclimate locality Šiance than in the cold-microclimate populations of Malá Stožka (t = 4.54, df = 1166, p < 0.0001) and Veľká Stožka (t = 5.61, df = 1166, p < 0.0001; Fig 2B). Poludnica population also showed a significantly higher fruit set than Veľká Stožka (t = 2.95, df = 1166, p = 0.017). However, in 2019, the opposite was true; the warm-microclimate population of Poludnica showed significantly lower fruit production than Šiance (t = 3.67, df = 1166, p = 0.0014) and the cold-microclimate populations of Malá Stožka (t = 5.35, df = 1166, p < 0.0001) and Veľká Stožka (t = 4.20, df = 1166, p = 0.0002). In Poludnica and Šiance fruit production was approximately equal in both years, and the temporal pattern seems to match flower counts. In cold climate populations, considerable variability was observed in fruit production between the two years, which was less affected by flower production. The beta-binomial GLMM was highly significant (χ2 = 508, df = 8, p < 0.0001) and had good performance (R2c = 0.53, R2m = 0.30).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Reproduction characteristics of the studied populations of Daphne arbuscula.

https://doi.org/10.1371/journal.pone.0300819.t001

The success rate of sexual reproduction at the individual level fluctuated more dramatically, ranging from 0% to 83% (S1 Table in S1 File). A few individuals did not produce any fruits during the study period (e.g., PO16). In contrast, only a few individuals from populations Šiance, Malá Stožka, and a single individual from Veľká Stožka exhibited fruit production in excess of 50% (S1 Table in S1 File). During both seasons (S1 Table in S1 File), certain individuals from locality Šiance, specifically SI47–SI49, exhibited remarkably high rates of fruit production (60–83%).

Breeding system study

During the test for autogamy, no fruits developed from 120 inflorescences. Fruit production was not observed in any of the isolated inflorescences, even in those with assisted self-pollination to determine compatibility (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Breeding system testing in studied populations of Daphne arbuscula.

https://doi.org/10.1371/journal.pone.0300819.t002

Genetic diversity

The genomic RADseq analyses consistently supported the presence of shallow population structure, indicating genetic distinctiveness among all four populations. PCA and ADMIXTURE analysis at K = 2 and K = 3 showed that Poludnica and Veľká Stožka populations are genetically closer, with separation only occurring at K = 4 (Fig 3A and 3B). Interestingly, even at K = 4, a minor cluster revealed their genetic similarity. Sample VS29 is likely in an outlier position in Splitstree analysis because it has genes from two different gene pools (most likely populations Poludnica and Veľká Stožka). In contrast, Malá Stožka was identified as the most divergent population among all the populations analyzed. Both ADMIXTURE analysis and the phylogenomic network generated by the neighbour-net algorithm suggested limited gene flow between the analyzed populations (Fig 3A and 3C). The four populations exhibited different numbers of private alleles, ranging from 1323 (Šiance– 3.4% of total alleles) to 1958 (Poludnica– 5.1% of total alleles; S2 Table in S1 File). No private fixed alleles were detected in the population. Further population-genetic parameters showed only negligible differences (S2 Table in S1 File). The low levels of between-population divergence were further supported by the pairwise population differentiation values (F_ST), which ranged from 0.0035 to 0.0079 (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Population genetic analyses based on 38,368 single nucleotide polymorphisms and 39 individuals from four populations of Daphne arbuscula.

(A) genetic clustering depicted by ADMIXTURE analysis at K = 2–4. Each individual is represented by a vertical bar, with colour proportionally indicating the segment assignment; (B) Principal component analysis (PCA) analysis. (C) Neighbour-Net as recovered by Splitstree. Population codes are as follows: MS–Malá Stožka, PO–Poludnica, SI– Šiance, VS–Veľká Stožka.

https://doi.org/10.1371/journal.pone.0300819.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Pairwise population F_ST estimates inferred from single nucleotide polymorphism data.

https://doi.org/10.1371/journal.pone.0300819.t003

Seasonal and population fluctuations in D. arbuscula fruit formation

Endemic species with restricted distribution ranges and small population sizes often exhibit decreased ability of sexual reproduction, especially when compared to their more widespread congeners [3, 4, 8–11, 13, 23, 75]. This may also limit their dispersal abilities and threaten their survival, in particular in the case of various disturbances. Reproduction effort can be influenced by a whole range of internal and external factors, e.g. breeding system, pollinator activity or climatic conditions.

Our study revealed that overall sexual reproduction in endemic D. arbuscula also exhibited relatively low fruit production, not exceeding 50% at the population level. We presented evidence that climatic conditions during the flowering and fruiting periods exert a great influence on fruit production, surpassing factors such as genetic variability, location, and ecology of the studied populations [cf. 76, 77]. Seasonal climatic conditions were also identified as a key determinant of sexual reproduction success in high-mountain cushion saxicolous members of the genus Saxifraga [78, 79]. In general, it has been demonstrated that unfavourable climatic conditions during the reproduction stage have the potential to impact plant reproduction, with species exhibiting diverse responses [28, 31, 80].

Daphne arbuscula flowers from late April to June, coinciding with substantial weather variability in Central European mountainous regions [81]. In 2018, drier and warmer weather in April and May, followed by colder and wetter conditions in June, resulted in significantly higher fruit production in the earlier-flowering warm-microclimate populations (20.3% in Poludnica; 42.82% in Šiance; Table 1) compared to the later-flowering cold-microclimate counterparts (10.53% in Malá Stožka; 8.97% in Veľká Stožka; Table 1). Conversely, in 2019, colder and wetter weather in April and May led to a remarkable decrease in fruit formation in warm-microclimate populations (10.9% in Poludnica; 28.93% in Šiance; Table 1), while drier and warmer conditions in June enhanced fruiting in cold-microclimate populations (48.49% in Malá Stožka; 35.58% in Veľká Stožka; Table 1). These findings indicate that cold and wet weather during the reproductive phase diminishes fruit production. However, it appears that fruit production in populations residing in warmer environments demonstrated greater stability across both seasons, suggesting that they were less affected by climatic fluctuations compared to populations with a cold microclimate (Fig 2B). This indicates that the sexual reproduction behaviour of populations residing in cold microclimates at north-oriented sites shows an increased vulnerability to between-year oscillations in a seasonal climate, which could exacerbate their susceptibility to ongoing trends in global climatic changes. Nevertheless, it is reasonable to approach our findings with caution due to some limitations in our study design. The lack of replication for each location, representing distinct climate types and population sizes, renders the design insufficiently robust, potentially strengthening the investigation’s results. Adding more replications, however, presented considerable challenges, given the difficulty or inaccessibility of the majority of populations of this endemic shrub located on rocky cliffs with low population sizes. Similarly, the study’s limitation to two seasons may be considered a constraint in drawing definitive conclusions. While we acknowledge that extending the study period to 10 or 15 years would provide sufficient data to discern clear trends in the impact of seasonal climatic conditions on fruit production, such an extended duration is evidently unfeasible for the majority of scientific projects today. Nevertheless, our observations across two seasons, encompassing climatically disparate years with divergent patterns during the reproductive phase of this species (Fig 2B and S1 File), highlight the influence of seasonal climate conditions on the fruit production of the studied species, even at the microscale level. Despite these constraints, we believe that these limitations have not unduly biased our outcomes or, particularly, the interpretation of our data.

Harsh climatic conditions, such as cold, high humidity, and strong winds, can damage gametes, flowers, and young fruits, leading to lower seed quality and fruit set [29, 31, 32]. [37] found that late frosts, prolonged droughts, and strong winds during fruit formation can cause tissue degeneration in D. arbuscula fruits. The degree of fructification was found to depend on the temperature and occurrence of rather sunny and dry weather in the spring also in glacial relict Pulsatilla vernalis [76]. As climate change increases the incidence of extreme climatic events, higher spring temperatures may result in earlier flowering, increasing the risk of exposure to late frosts and subsequent fruit damage [28, 30, 32]. Indeed, late summer frosts exert a more pronounced negative impact on reproductive shoots, consequently affecting sexual reproduction success, compared to vegetative ones in the majority of studied high-mountain plants from the European Alps [82].

Moreover, the flowering period of D. arbuscula shrubs is rather short and typically lasts up to two weeks [20, 36]. Previous field studies on D. arbuscula have identified potential insect pollinators (i.e., from orders Thysanoptera, Coleoptera, Lepidoptera, Hymenoptera–Formicoidea or Apoidea [22]) in the vicinity of blooming shrubs under sunny and warm weather conditions, but contrasting trends may emerge when unfavourable climatic conditions prevail [19, 21, 22]. Thus prolonged rainfall and low temperatures can further diminish insect pollinator activity, reducing the chances of successful pollination [19, 31, 83]. An additional critical factor adversely impacting successful pollination could be the mismatch in plant phenology and pollinator activity resulting from climate change [e.g. 84]. Such a deficiency in pollinators consistently leads to decreased fruit production in other Daphne species as well [85–87].

Nevertheless, rocky habitats exhibit high geomorphological heterogeneity, featuring diverse microsites characterized by distinct microclimates [88–90]. Within D. arbuscula populations, individual shrubs located just a few meters apart can significantly vary in flowering time and reproductive success due to microclimate variations. Microsites with hospitable microclimates might act as a buffer against harsh climatic fluctuations, ensuring that some shrubs experience optimal conditions for sexual reproduction. This was clearly demonstrated in the Šiance population, where shrubs in open, sunny rock crevices (SI41–SI46) flowered 1–3 weeks earlier compared to those on north-oriented, more humid rocks shaded by mixed forests. As a result, shrubs from shaded microhabitats produced substantially more fruit during both seasons (60–83%; S1 Table in S1 File) compared to those from open, sunny places (2.5–59%; S1 Table in S1 File). Similar microsite-dependent sexual reproduction success was observed in the high-alpine species Saxifraga moschata [78]. Plants from early snowmelt sites exhibited a higher abundance of sexual reproduction structures compared to those from later snowmelt sites. Such microsites may thus play a crucial role as long-term reservoirs for sexual reproduction and the maintenance of genetic diversity within the species [91].

The role of additional intrinsic and extrinsic factors in decreasing fruit production of D. arbuscula

The observed reduction in sexual reproduction in D. arbuscula is in line with the phylogenetic context, as other species within the genus Daphne also exhibit insufficient levels of fruit formation below 50% [19, 23, 86, 87, 92–98]. For example, the largely deficient fruit formation was observed also in its two closely related narrow endemic species D. petraea [99] and D. rodriguezii [86]. Nevertheless, there are exceptions, such as the widespread Daphne mezereum, which has been reported to reach up to 72% fruit production [100]. The limitation in fruit production could be linked to the floral structure of the genus, wherein anthers are situated in the floral tube [40], facilitating the dispersal of numerous pollen grains on the stigma of the same flower (i.e., stigma clogging), thus preventing adhesion of pollen from another plant [86, 88, 95, 96]. Consequently, for outcrossing species like D. arbuscula, numerous flowers remain unpollinated, or if pollinated, self-pollinated embryos degenerate during ripening [37, 86, 87, 95, 96]. Despite having high pollen viability (approximately 97%) in outcrossing D. arbuscula, indicating no significant issues with pollen development and fertility [35, 39] this morpho-anatomical pollination issue can lead to a decline in the number of embryos resulting from outcrossing pollination.

Another potential explanation for the reduced fruit production in studied species is the limited availability of mating partners within and among populations. Several studies have emphasized the significance of population size and density for the reproductive success of animal-pollinated plants [reviewed by 101]. The scarcity of mating partners is further compounded by self-incompatibility, which has also been demonstrated in D. arbuscula (Table 2) [21, 36, 37], leading to the failure of viable seed formation from self-pollinated flowers [9, 17, 102].

Furthermore, the decreased fruit production could be attributed to potential shifts towards vegetative (clonal) reproduction [19, 37, 103], which might also significantly reduce the number of compatible mating partners [cf. 4, 8, 104–106]. However, the role of clonality in this context can only be speculated upon since exploring clonality falls outside the scope of the present study.

Ultimately, our analyses consistently demonstrated a decline in fruit production across all investigated populations, regardless of population size (cf. Table 1; S1 Table in S1 File). This suggests that population numbers, particularly the number of mating partners within a population, may not play a decisive role in the reduction of the species’ sexual reproduction. Similarly, a lack of correlation between sexual reproduction-associated traits and population size was identified in Allysum montanum from the Jura Mountains in France [107]. Conversely, an opposite trend was detected for various plant species by [15]. Nevertheless, given the limitations in our sampling designs (see Discussion above), we exercise caution in making definitive statements on this matter. Further research is warranted to address this question comprehensively.

Low association between genetic variability at the population level and sexual reproduction in D. arbuscula

Genetic mechanisms may contribute to seed and/or fruit abortion, leading to low fecundity and posing a risk to the long-term survival of plants [e.g., 108–113]. Endemic species, often occurring in small and isolated populations, may be more vulnerable to reduced gene flow within and among populations, potentially experiencing genetic drift or inbreeding depression [10, 11, 15–17, 114, 115].

Daphne arbuscula is considered an ancient tertiary relict species that likely survived Pleistocene glaciations on the south-facing slopes of Muránska planina Mts [19, 20]. Our genetic analyses, providing the initial glimpse into the genetic variation of the studied species, did not unveil any intraspecific diversity among the examined populations using traditional markers (ITS and cpDNA; see S3 File). Instead, they only revealed a relatively shallow genetic structure when employing the more robust RADseq method (Fig 3A–3C). This finding, contrasting with the pronounced genetic structure observed in the Menorcan narrow endemic D. rodriguezii [116] suggests that the fragmentation of D. arbuscula’s distribution range in Muránska planina Mts, or alternatively, colonization from potential microrefugia, occurred more recently in the postglacial period. The genetic differentiation among populations is relatively weak, as indicated by the pairwise population differentiation (F_ST values), a low level of within-population private SNPs (ranging between 3.4% and 5.1% per population), and the absence of private fixed SNPs (see Results section and Tables 1 and 3). Moreover, the stronger genetic differentiation observed in the small populations of Malá Stožka and Šiance, as revealed by ADMIXTURE at K = 2–4 (Fig 3A), may be attributed to genetic drift rather than long-term isolation.

While variations in within-population genetic variability were observed in the studied populations, these differences do not seem to be clearly linked to fruit production. Specifically, the most abundant population, Poludnica, exhibited the highest level of genetic variability, yet reproductive success remained consistently low year after year. In contrast, the population of Malá Stožka, which is spatially the most isolated and the second smallest in size with lower within-population genetic variability, had the highest fruit set during the second year among all studied populations (Table 1; Fig 2B). These findings align with studies that have shown that low genetic diversity does not always result in reduced reproductive success [e.g., 117, 118]. Therefore, it can be concluded that the genetic variability of the analyzed populations does not seem to play a crucial role in the fruit production of this species.