Background

The genus Daucus L. is a member of Apiaceae, a large, complex, and cosmopolitan family of approximately 466 genera and 3820 species that are especially diverse in temperate regions of Eurasia and North America [1]. Although the Apiaceae family is well defined morphologically by a wide range of distinctive characteristics, allowing its constituent taxa to be unambiguously assigned to the family, taxonomic divisions within the family have been extensively discussed [2]. The cultivated carrot (D. carota L. subsp. sativus Hoffm.) is economically and nutritionally the most significant member of the genus, providing a major source of vitamin A precursors (α- and β-carotene) in the human diet [3]. Based on a morpho-anatomical study by Sáenz Laín [4], Daucus has traditionally comprised 20 species and has been divided into five sections: Daucus L., Anisactis DC., Platyspermum DC., Chrysodaucus Thell, and Meoides Lange. This classification was further extended by Rubatzky et al. [5], who listed 25 species. Recently, a number of molecular analyses involving plastid genes (rbcL, matK), plastid introns (rpl16, rps16, rpoC1), ribosomal internal transcribed spacer (ITS) sequences, chloroplast and mitochondrial DNA restriction sites, nuclear orthologs, and nuclear single nucleotide polymorphisms (SNPs) have been performed to clarify the phylogenetic relationships among Daucus species and their close relatives in the subfamily Apioideae [6,7,8,9,10,11,12,13,14,15,16,17,18,19]. These studies have resulted in the division of the genus into two subclades: Daucus I and Daucus II. Daucus I comprises D. carota subsp. carota L. (the wild ancestor of the cultivated carrot) with all D. carota subspecies, as well as several Daucus species of Mediterranean origin and some species traditionally placed in other genera, whereas Daucus II includes the remaining members of the genus. Following a recent taxonomic revision of Daucus by Banasiak et al. [16], the genus has been enlarged to include another 18 species from 9 other genera and now contains about 40 species.

The rich genetic diversity of wild species of Daucus could be utilized in carrot breeding programs; hence, it is crucial to better understand the genome structure and relationships among these species. Genome size, defined as the amount of DNA in the holoploid genome of an organism [20], is a fundamental biological characteristic, which can provide insight into the evolutionary background of a species. Knowledge of genome size is important in elucidating the taxonomic relationships among species and tracing evolutionary changes. It may also help resolve conflicting hypotheses concerning the origin of polyploids and serve as additional quality control for species identification in germplasm collections. It can also be useful in a variety of other scientific disciplines, including systematics, phytogeography, phylogeny, and genome sequencing projects, since their scale and cost depend on this parameter [21,22,23,24,25].

For the estimation of nuclear DNA content, flow cytometry has become the predominant method of choice, as it is fast, accurate, and relatively inexpensive [26]. According to the Plant DNA C-value database [27], nuclear DNA amounts have been estimated for 57 members of Apiaceae using flow cytometry; however, it should be noted that this database does not include all estimates.

Cultivated carrot has a relatively small genome of approximately 473 Mb per haploid genome [28, 29]. In the genus Daucus, nuclear DNA content estimates by flow cytometry have been reported for several wild species and subspecies, as well as cultivated carrots, revealing great variation in the 2C DNA amount (0.847–3.019 pg) [30,31,32,33].

Pollen of each plant species can be described by a variety of characteristics, including size, shape, aperture number and features, and exine ornamentation; thus, the study of pollen morphology is especially important for plant identification and taxonomic research, and can help to determine the relationships among taxa at various taxonomic levels. Palynological characteristics also play a significant role in fields such as phylogeny, paleobotany, archeology, and criminology [34,35,36,37]. Although, according to PalDat [38] (the largest database for palynological data), many members of different genera in the family Apiaceae have been subjected to palynological investigation, a comprehensive study on pollen of the genus Daucus is still lacking.

Since the systematics of Daucus remains under debate, revisions with the use of additional data are necessary to better understand the relationships within Daucus species. Therefore, the aims of this research were (1) to estimate the nuclear DNA content in 13 Daucus taxa (14 accessions) and four closely related non-Daucus species, frequently used in previous phylogenetic and cytotaxonomic studies of Daucus [15, 17, 18, 39], by flow cytometry; (2) to investigate the pollen morphology of these taxa by light and scanning electron microscopy (SEM); (3) to determine their pollen nucleus status; (4) to evaluate the taxonomic value of these cytogenomic and palynological data; and (5) to explore whether any correlations exist between genome size and pollen features.

Results

Nuclear DNA content

The 2C values for Daucus taxa ranged from 0.999 (D. carota subsp. sativus [DH]) to 3.228 pg (D. littoralis), giving an overall variation of about 3.2-fold (Table 1, Fig. 1). Among the outgroup species, the 2C values varied from 1.775 (C. platycarpos) to 2.882 pg (T. arvensis). In both groups, the accessions differed in nuclear DNA content (p < 0.001).

Table 1 Nuclear DNA content of Daucus taxa and outgroup species
Full size table
Fig. 1
figure 1

Exemplary histograms of the relative nuclear DNA content of (a) Daucus aureus, (b) D. conchitae, (c) D. glochidiatus, (d) D. guttatus, (e) Caucalis platycarpos, and (f) Torilis arvensis. Asterisks indicate the internal standards: (a–b, d–e) Petunia hybrida ‘P × Pc6’ (2C = 2.85 pg), (c, f) Solanum lycopersicum ‘Stupicke’ (2C = 1.96 pg)

Full size image

The taxa belonging to the Daucus I subclade exhibited lower genome size compared to the taxa from the Daucus II subclade, except for D. muricatus (Daucus I), whose genome size was much higher than the other accessions of this subclade and even higher than some species of Daucus II subclade (Table 1). The monoploid genome size (1Cx) of all Daucus accessions with 18 chromosomes was similar (about 0.5 pg); however, for the species with higher chromosome number it varied from 0.510 to 1.614 pg and did not relate to n. Also, in Daucus taxa there was no correlation between the nuclear DNA content and chromosome number (r = 0.521, p = 0.06).

Means in columns with the same letter were not significantly different at p < 0.001. A one-way ANOVA and Tukey’s HSD test were conducted separately for Daucus taxa and the outgroup species.

Pollen viability and palynological characteristics

The means of pollen viability between Daucus taxa differed significantly (p < 0.001; Table 2). Despite these differences, the pollen viability rate was relatively high (above 70%), except for D. pusillus, whose pollen exhibited lower viability. Similarly, pollen of outgroup species expressed high viability after Alexander’s staining, and no significant differences were found between compared accessions.

Table 2 Palynological characteristics and pollen viability of Daucus taxa and outgroup species
Full size table

The means of the pollen morphometric characteristics (the lengths of the polar and equatorial axes) differed significantly (p < 0.001; Table 2) within both Daucus taxa and the outgroup species. Following the nomenclature for pollen size by Halbritter et al. [36], which classifies pollen as very small (< 10 µm), small (10–25 µm), medium (26–50 µm), large (51–100 µm), or very large (> 100 µm), our observations revealed that the vast majority of Daucus taxa (~ 79%) had medium-sized pollen grains; only three taxa (~ 21%) had small pollen grains. The mean length of the polar axis (P) varied from 21.19 (D. glochidiatus) to 40.38 µm (D. aureus). In all outgroup species, the pollen grains were classified as medium in size, of which the pollen of Orlaya daucorlaya showed the highest P parameter (49.86 µm). The average equatorial diameter ranged from 13.20 to 20.21 µm in Daucus taxa and from 12.06 to 23.54 µm in the outgroup species. Based on the mean ratio of the polar axis to equatorial diameter (P/E), the pollen grains of Daucus taxa were classified as prolate (P/E = 1.33–2.00), except for D. aureus and D. rouyi, whose pollen were perprolate (P/E > 2.00) in shape, whereas all the outgroup species had perprolate-shaped pollen.

Means in columns with the same letter were not significantly different at p < 0.001. A one-way ANOVA and Tukey’s HSD test were conducted separately for Daucus taxa and the outgroup species.

The SEM analysis performed on pollen samples of 11 taxa (12 accessions) revealed that the pollen grains were tricolporate with narrow colpori (Fig. 2). Considering the pollen outline in the polar view, the examined taxa had triangular pollen grains (see Fig. 2b, m–n, q). The exine ornamentation was striate (elongated ornamentation elements separated by parallelly arranged grooves) in D. conchitae (Fig. 2b) and D. rouyi (Fig. 2p); rugulate (elongated and irregularly arranged ornamentation elements) in D. carota subsp. capillifolius (Fig. 2d) and D. involucratus (Fig. 2l); perforate in D. guttatus (Fig. 2j), O. daucoides (Fig. 2t), and O. daucorlaya (Fig. 2v), or the ornamentation pattern was mixed, i.e., striate-perforate in D. carota subsp. sativus (‘Dolanka’) (Fig. 2h), D. littoralis (Fig. 2n), and D. sahariensis (Fig. 2r); striate-rugulate in D. carota subsp. sativus (DH) (Fig. 2f); or rugulate-perforate in T. arvensis (Fig. 2x).

Fig. 2
figure 2

Pollen morphology and exine ornamentation of selected Daucus and related taxa by scanning electron microscopy. (a–b) D. conchitae; (c–d) D. carota subsp. capillifolius; (e–f) D. carota subsp. sativus (DH); (g–h) D. carota subsp. sativus (‘Dolanka’); (i–j) D. guttatus; (k–l) D. involucratus; (m–n) D. littoralis; (o–p) D. rouyi; (q–r) D. sahariensis; (s–t) Orlaya daucoides; (u–v) O. daucorlaya; (w–x) Torilis arvensis. Scale bars: 5 µm (b–f, h, j, l, n–r, t, v, x), 10 µm (a, g, i, k, m, s, u, w)

Full size image

As evidenced by DAPI staining, at the time of shedding, the pollen of the 14 investigated taxa was three-celled, with a large, weakly stained vegetative nucleus and two smaller, strongly stained sperm cells (Fig. 3). The vegetative nucleus was diffused and more or less round, whereas the sperm cells, depending on the stage of pollen grain development, were round or spindle-shaped and usually located near each other. Although DAPI is a DNA-specific dye, the staining clearly showed the bone-shaped or elliptical outline of the pollen grains, with the apertural areas often visible.

Fig. 3
figure 3

DAPI-stained pollen grains of selected Daucus and related taxa. (a–b) D. conchitae; (c) D. carota subsp. capillifolius; (d) D. carota subsp. sativus (DH); (e) D. glochidiatus; (f) D. guttatus; (g) D. involucratus; (h) D. littoralis; (i) D. muricatus; (j) D. pusillus; (k) D. rouyi; (l) D. sahariensis; (m) D. syrticus; (n) Orlaya daucoides; (o) Torilis arvensis. sc, sperm cells; vn, vegetative nucleus. Scale bar = 5 µm

Full size image

Interspecific relationships within the genus Daucus

The UPGMA similarity dendrogram, based on three quantitative parameters (2C DNA content, P, E), divided 14 Daucus taxa into three major clusters at a Euclidean distance of 6.5, with a cophenetic correlation of 0.86 (Fig. 4). The first cluster contained three taxa with the largest pollen grains but clearly different 2C DNA content (the variation was about 3.2-fold). In the second cluster, two taxa with the smallest pollen grains and an ~ 2.1-fold variation in 2C value were grouped together. The third cluster was subdivided into two subclusters, one of which included the 18-chromosome taxa form Daucus I subclade, with very similar pollen size and 2C values, and two taxa from Daucus II subclade with an evidently higher 2C value; the second subcluster comprised two taxa with larger pollen grains and a ~ 1.8-fold variation in 2C DNA content.

Fig. 4
figure 4

UPGMA similarity dendrogram showing relationships among Daucus taxa based on 2C DNA content and pollen characteristics. The 2n chromosome numbers are shown in brackets. P, the mean size of pollen grains; Dol, ‘Dolanka’; the same colors indicate similar values of this parameter

Full size image

According to the Pearson’s correlation analysis, high positive correlations were found between the length of the polar axis (P) and the length of equatorial diameter (E) of pollen grains (r = 0.834, p < 0.001), as well as between P and P/E (r = 0.823, p < 0.001) (Fig. 5). No significant correlations were observed when comparing cytogenetic information with palynological data.

Fig. 5
figure 5

Relationship between (a) polar axis (P) and equatorial diameter (E) of pollen grains and (b) P and P/E ratio among Daucus. Three asterisks (***) indicate statistical significance at p < 0.001; r, Pearson's correlation coefficient; R2, coefficient of determination

Full size image

Discussion

In this study, the nuclear DNA content of 13 taxa (14 accessions) of Daucus and four closely related species was estimated by flow cytometry, thus expanding the knowledge of genome size variation in the family Apiaceae. Of these, flow cytometric data for 11 taxa are reported here for the first time. Among the studied taxa, almost all had very small genomes (2C ≤ 2.8 pg), except for D. littoralis and T. arvensis, whose genome size was categorized as small (2.81–7.00 pg), following the nomenclature of Leitch et al. [41]. It was in agreement with the data included in the Plant DNA C-value database [27], where members of Apiaceae are reported as having mostly very small or small genome sizes; only for a few species genome size exceed 7.00 pg.

To date, the two most extensive cytogenomic studies on the genus Daucus have been conducted by Nowicka et al. [32] and Roxo et al. [33]. Nowicka et al. [32] investigated the nuclear DNA content in the collection of diploid members of Daucus from different parts of the world, whereas Roxo et al. [33] estimated the 2C values for 16 taxa of the subtribe Daucine from the Macaronesian islands. The results of these works, combined with our findings, revealed an over 3.8-fold variation of the nuclear DNA content within Daucus. Some authors suggest that the interspecific variation in DNA content has adaptive significance and correlates with environmental and ecological factors; however, current evidence has been inconclusive so far and does not provide a clear answer as to whether environmental pressure has a relevant impact on plant genome size variation [42,43,44,45,46].

In the present research a low variation (less than 6%) in the 2C DNA content of D. carota accessions was found. Moreover, all 18-chromosome taxa had similar genome size, which supports the close relationship between these taxa.

Regarding the species for which cytogenomic data have previously been reported, our results are in large part congruent with those obtained by Nowicka et al. [32]. Our estimations for four wild species (D. involucratus, D. littoralis, D. muricatus, and D. pusillus) were only slightly higher, with a difference of 4–8%, but such discrepancies may be attributed to the different internal standards that were used by the authors (Brassica napus L. ‘Bor’ and D. carota subsp. sativus ‘Dolanka’). In contrast, the 2C value obtained here for D. guttatus conflicted with that of the authors, who found large differences in DNA content for two D. guttatus accessions. This could result from their taxonomic misclassification because the germplasm of the D. guttatus complex is especially problematic; thus, misidentifications are frequent [47]. Therefore, in this regard, the results are difficult to compare.

The great differences in the 1Cx value among wild Daucus species suggest that in the course of speciation, large-scale chromosomal rearrangements or the accumulation of non-coding repetitive DNA sequences (particularly retrotransposons) occurred in the genus.

In this study, the pollen morphology of 17 taxa (18 accessions) of the family Apiaceae (13 Daucus and 4 non-Daucus taxa) was determined, of which 13 taxa (11 Daucus and 2 non-Daucus) were examined for the first time. In all taxa, tricolporate and prolate–perprolate-shaped pollen were observed, which is a common feature of the pollen grains of Apiaceae [48,49,50,51,52].

Regarding the exine ornamentation observed by SEM, pollen of the examined Daucus taxa was striate, rugulate, perforate, or had a mixed ornamentation pattern; thus, this palynological characteristic may be considered useful for species delimitation within Daucus. Exine ornamentation plays an important role in plant systematics; it can be useful for distinguishing closely related genera or sometimes species in the same genus [53, 54]. We also compared the data included in PalDat for species of different genera belonging to the family Apiaceae that were also analyzed by SEM (50 species representing 31 genera) and verified that the pollen grains of most of these species are rugulate and perforate (or only rugulate/perforate), suggesting strong homogeneity for this trait in Apiaceae. Nonetheless, outside PalDat, some other types of exine ornamentation in Apiaceae have also been reported in the literature, e.g., cerebroid, pertectate, and verrucate [50,51,52, 55].

At the time of shedding, the pollen grains of the studied taxa were three-celled, as revealed by DAPI staining, which is a characteristic feature of pollen in the Apiaceae family [56]. In nature, most flowering plants produce pollen that is arrested at the two-celled stage, containing one vegetative cell and one generative cell, and only around 30% of species shed three-celled pollen (one vegetative cell, two sperm cells) at anthesis [57, 58]. Compared to two-celled pollen, three-celled pollen grains are inherently short-lived [59], and they are also more hydrated [60].

Many examples of correlations between genome size and phenotypic traits at the nuclear, cellular, and tissue levels can be found in the literature. Studies have shown that the amount of DNA is associated with nuclear and cell volume, cell size, cell cycle duration, stomatal cell size, cell density [61,62,63,64,65], seed mass [66], leaf mass per unit area [67], and flowering time [68].

Although the nuclear DNA content did not correlate with the pollen features of Daucus taxa, this parameter can be of great use in distinguishing individual taxa within some groups of taxa with a similar pollen size (Fig. 4), e.g., taxa with small (P < 25 µm) pollen grains (D. glochidiatus, D. involucratus, and D. pusillus) differed in terms of 2C DNA content; the same relation was observed in taxa with P ≈ 30–34 µm (D. conchitae and D. rouyi), as well as in those with the largest pollen grains, i.e., P ≈ 37–41 µm (D. aureus, D. muricatus, and D. littoralis). However, in the group of taxa with pollen grains of P ≈ 26–29 µm, only D. guttatus can be separated from the others based on 2C DNA content. In the case of whole plant morphology, for example, two morphologically similar species, D. conchitae (2C = 2.08 pg) and D. guttatus (2C = 2.37 pg), can be easily distinguished based on their DNA content. On the other hand, some taxa that shared very similar DNA content and pollen size, e.g., D. carota subsp. capillifolius and carrot, were morphologically distinct.

Wild Daucus species may play an essential role in carrot breeding programs, as they could be a valuable potential source of agronomically important genes. Thus, to effectively utilize this germplasm, it is crucial to determine the species boundaries and relationships within Daucus [47, 69]. Considering that the correct identification of species is a prerequisite for further use, the application of supplementary methods for this purpose is essential. Therefore, the assessment of relative nuclear DNA content by flow cytometry can be a good choice for simple, rapid, and low-cost screening of genebank accessions during their identification and maintenance, even at seedling stage [24, 70, 71], which could be further combined with palynological measurements to help reliably identify species, as evidenced in this study.

Conclusions

The present study significantly complements the available information on the nuclear DNA content in the genus Daucus and provides comprehensive knowledge of the pollen morphology of its taxa. These results may be of great importance in elucidating the taxonomic relationships among Daucus species and can help in the correct identification of gene bank accessions. From a broader view, the findings of this work could also be meaningful for the interpretation of the evolutionary trends in the genus. Nonetheless, to better understand the relationships within Daucus in the phylogenetic context, further studies comprising the remaining taxa from the genus, as well as the taxa from different genera that have recently been included in Daucus, are needed.

Methods

Plant material

A total of 18 accessions representing 17 taxa (species or subspecies) were examined in this study, including 12 accessions of wild Daucus taxa, a reference doubled haploid carrot line, one carrot cultivar, and four accessions of closely related non-Daucus species (Table 3). Seeds of all wild accessions were provided by the USDA-ARS North Central Regional Plant Introduction Station (Ames, Iowa, USA), whereas carrot seeds were either purchased commercially or obtained from the collections of the Department of Plant Biology and Biotechnology, University of Agriculture in Krakow (Krakow, Poland).

Table 3 List of Daucus and closely related non-Daucus accessions used in this study
Full size table

The seeds were germinated in soil-filled pots and grown in a growth chamber at 18 °C with a long-day photoperiod of 16/8 h (light/dark) for the first few weeks, then transferred to greenhouse conditions (26/14 °C ± 2 °C, day/night temperature; long-day photoperiod) until flowering. In the case of two cultivated carrot accessions, the plants were first vernalized in a cold chamber at 5 °C for three months, then returned to the greenhouse for flowering.

Flow cytometric measurements of nuclear DNA content

For nuclear DNA content measurements, 8–23 plants per accession, depending on availability, were used. Young leaves were collected from plants grown in a growth chamber and samples for flow cytometric analysis were prepared as previously described [72] using—for nuclei isolation—Galbraith’s buffer [73], supplemented with 1% (w/v) polyvinylpyrrolidone (PVP-10, MW 10,000; Sigma-Aldrich, St. Louis, USA), ribonuclease A (RNase A, 50 µg mL−1; Sigma-Aldrich), and propidium iodide (PI, 50 µg mL−1; Sigma-Aldrich). Solanum lycopersicum L. ‘Stupicke’ (2C = 1.96 pg; [74]) were used as an internal standard for D. glochidiatus, D. littoralis, O. daucorlaya, and T. arvensis, while for the remaining accessions, Petunia hybrida Vilm. ‘P × Pc6’ (2C = 2.85 pg; [75]) was applied. The nuclei suspension was analyzed using a CyFlow SL Green flow cytometer (Partec GmbH, Münster, Germany) equipped with a high-grade solid-state laser (λem = 532 nm), long-pass filter RG 590 E, DM 560 A, and side and forward scatters. The PI fluorescence was measured in 3000–5000 nuclei per sample. For histogram evaluation, FloMax software (Partec GmbH, Münster, Germany) was applied. The coefficient of variation (CV) of the G0/G1 peak of sample species ranged between 2.68 and 5.96%. Nuclear DNA content was calculated using the linear relationship between the ratio of the 2C peak positions sample/standard on a histogram of fluorescence intensities.

Pollen viability, morphology, and nucleus status

Pollen viability was assessed using Alexander’s staining method [76]. Fresh pollen was collected from fully open flowers (from the anthers after dehiscence) of the greenhouse-grown plants onto microscope slides (samples were taken from 1–3 randomly chosen inflorescences per individual, 3–5 plants per accession; except for D. syrticus and D. glochidiatus, for which only one and two plants, respectively, were available), then a drop of Alexander’s stain was applied to each slide and covered with a cover glass. The slides were examined under an Axio Imager.M2 microscope (Carl Zeiss, Göttingen, Germany), and the number of viable (dark red cytoplasm with a green wall) and non-viable (entirely green) pollen grains were counted, with a minimum of 300 pollen grains per slide. The pollen viability was expressed as a percentage of viable pollen.

Pollen size was determined using samples of Alexander-stained pollen that had been used for the viability test. The polar axis (P) and equatorial diameter (E) of the pollen grains were measured from microphotographs captured with a Canon PowerShot G10 digital camera (Canon, Tokyo, Japan) attached to the same microscope as above. At least 100 viable pollen grains per plant (3–5 plants per accession; except for D. syrticus and D. glochidiatus, for which only one and two plants, respectively, were available) were measured. The terminology for pollen size follows that of Halbritter et al. [36]. The pollen of each accession was classified into a shape class based on the ratio of the polar axis to the equatorial diameter (P/E), according to the nomenclature proposed by Erdtman [40].

For SEM analysis, pollen samples from the fully open flowers of 12 accessions were collected into gelatin capsules and stored in an exsiccator until use. Dry pollen grains were mounted on stubs and sputter-coated with gold using a JFC-1100E ion sputter coater (JEOL, Tokyo, Japan). The palynological characteristics (exine ornamentation and aperture number) were examined under a JSM-5410 scanning electron microscope with a wolfram cathode (JEOL, Tokyo, Japan). The terminology for exine ornamentation follows that of Halbritter et al. [36].

To determine the pollen nucleus status (expressed as the number of pollen nuclei in pollen grains after anther dehiscence), pollen samples of 14 taxa (12 Daucus taxa and two outgroup species) were collected in the same way as for the viability test, then mounted in a drop of 4′,6-diamidino-2-phenylindole (DAPI) solution (2.5 µg−1 DAPI, 7.7 mM Tris–HCl, 10 mM spermine tetrahydrochloride, 10 mM NaCl, 2.2% hexylene glycol, and 0.25% Triton™ X-100; mixed in a 1:1 ratio with glycerol), and covered with a cover glass. The slides were examined under the same microscope using the fluorescence mode and an appropriate filter set for DAPI (Zeiss filter set 02: λex = 365 nm, λem > 420 nm). The microphotographs were captured using a BV MV camera (Applied Spectral Imaging, Edingen-Neckarhausen, Germany).

Statistical analyses

For quantitative parameters, means and standard errors of the means were calculated for each accession and subjected to a one-way analysis of variance (ANOVA) followed by Tukey’s honestly significant difference (HSD) test at a significance level of at least p = 0.05 using Statistica v. 13.3 (TIBCO Software Inc., USA). For nuclear DNA content estimation and pollen morphology and viability, the mean of measurements/counts for one plant was considered a single replication. Statistical analyses were conducted separately for the Daucus taxa and the outgroup species.

To determine the relationships among the Daucus taxa, an unweighted pair-group method with arithmetic mean (UPGMA) cluster analysis with Euclidean distance was performed based on nuclear DNA content and pollen morphology (P, E) data using Past v. 3.22 software [77].

To identify relationships among cytogenetic (2C DNA content, somatic chromosome number) and palynological (P, E, P/E) data within the genus Daucus, Pearson’s correlation analysis was carried out using Statistical.