Contribution to the knowledge of genome size evolution in edible blueberries (genus Vaccinium )

Abstract

BACKGROUND:

Vaccinium is one of the largest genera (ca. 500 species) of Ericaceae, well known for its edible and ornamental uses. Although there is certain karyological knowledge, information about genome size (GS) is scarce in the genus.

OBJECTIVE:

The main goal is providing GS data for several Vaccinium species with prevalence in Europe and Western Asia and analysing global GS variation in the genus, considering available data and phylogenetic context.

METHODS:

New GS assessments were obtained by flow cytometry and chromosome counts were verified. Phylogenetic analyses (using nuclear ITS, and chloroplastic matK and ndhF) were performed by Bayesian inference and reconstruction of ancestral GS by maximum parsimony.

RESULTS:

We obtained GS data for five Vaccinium species (13 populations). Three species are reported for the first time. Values (2C) ranged between 1.16–1.47 pg at the diploid (2n = 24) and between 3.13–3.16 pg at the tetraploid (2n = 48) levels. The five species here investigated have been placed and analysed in a reconstructed phylogenetic background (including 68 taxa).

CONCLUSIONS:

GS values of Vaccinium can be considered “very small”. The preliminary reconstruction of ancestral GS would point to a reduction in Vaccinium, although more data is needed to establish global GS evolutionary trend in the genus.

Keywords

Berries C-value chromosome counts flow cytometry nuclear DNA amount nuclear DNA content

1 Introduction

Vaccinium L. is one of the major genera in family Ericaceae, with ca. 500 described species [1] although there are only 223 accepted species names (The Plant List, accessed 31st May 2019). All species are perennial, exhibit both self and cross-pollination and they have in common a small and pulpy berry as a fruit which in many cases is edible [2]. Most Vaccinium prefer cool regions of the Northern hemisphere, being broad in Europe, Asia and North America and absent in New Zealand, Australia and most of Africa; yet some species are also present in tropical regions such as Madagascar or Hawaii [3]. Fruits from several species of genus Vaccinium, commonly known as blueberries, enter into the category of the so-called berry fruits, which represent an increasingly consumed crop all over the world. Reasons for this recent boost of interest are the numerous health properties that are related to them. These berries are rich in anti-oxidant compounds, can alter the lipid metabolism, being beneficial for dietary purposes, and are rich in potassium and fibre [4]. They also have beneficial effects on cardiovascular and urinary diseases, and can improve brain function and cognitive ability [5, 6]. Besides, many Vaccinium have ornamental value [7].

The major type of cultivated blueberry (varieties of Vaccinium corymbosum) was developed in the last century [8] and several other Vaccinium have only recently been domesticated [9 –11], in which interspecific hybridization has been an effective breeding strategy [12]. There are conservation projects involving wild Vaccinium to preserve traits potentially useful for breeding purposes, such as disease resistance, winter hardiness, low chilling requirement, adaptation to high pH soils, early ripening, late bloom or cold hardiness among others [8, 13]. However, there are still many knowledge gaps regarding the evolutionary and systematic diversity in this genus, which is essential information for conservation programmes as well as the development of new cultivars [e.g. 14, 15]. Molecular phylogenetic analyses have revealed that Vaccinium constitutes a polyphyletic group, with other closely related genera from tribe Vaccinieae (e.g. Agapetes, Cavendishia, Paphia, Notoptora) embedded within Vaccinium [3, 16]. In addition, the relationships among the major clades of blueberries have not been well resolved in previous phylogenetic treatments of the genus. One possible cause is the extent of hybridization, already noted as common in early cytological investigations [17]. This has led authors to advocate for the recognition of several clades or smaller groups rather than a large Vaccinium redundant to tribe Vaccinieae [3] and to propose sinking Agapetes in Vaccinium [18].

From the karyological point of view, Vaccinium also shows a certain level of evolutionary complexity. Species of this genus have many ploidy levels derived from a single base chromosome number, x = 12, ranging in nature from the diploid to the hexaploid. Auto- and allopolyploids are also common. Interspecific triploids, pentaploids, octoploids and nonaploids have been obtained for domestication purposes [19]. The most extensively studied species of Vaccinium are those from section Cyanococcus (including different ploidy levels) and section Oxycoccus (only diploids) [20]. Other sections such as Myrtillus and Vitis-idaea (only diploids), section Hemimyrtillus, including V. arctostaphylos (only tetraploids), and section Vaccinium, including V. uliginosum (different ploidy levels) have also been largely investigated [7].

Chromosome number and genome size (GS) are basic cytological characters, normally constant within a species at the same ploidy level [21], although there are exceptions, such as the presence of B-chromosomes [22] or genuine intraspecific GS variation [23, 24]. Chromosome numbers are relatively well documented for Vaccinium, as recorded by the Chromosome Counts database [25], where there is data for more than 100 species of the genus. However, GS is only known for a few of its species [12 , 26–28]. The nuclear DNA content is a trait correlated with many biological characters such as seed and chloroplast size (positively correlated in most cases [24]), and related to life cycle or environmental factors [29]. In general, plants with smaller GS tend to show faster growth rates and complete faster their life cycle [21, 30]. From an ecophysiological perspective, GS has been correlated to environmental tolerance towards high or low temperatures [31] or soil nutrients [32]. Other studies have even found a relationship between nuclear DNA amount and response to global warming in which plants with smaller genome sizes would be favoured [33]. In general, these associations of GS with plant growth and yield have great potential interest for conservation and breeding purposes on crop species such as Vaccinium. This work aims to extend GS information on this genus, which will be ultimately useful for breeding purposes and to analyse its variation considering the phylogenetic context of the studied species.

2 Materials and methods

2.1 Species and sample collection

We obtained samples from 13 Vaccinium populations representing five species (V. corymbosum, V. arctostaphylos, V. myrtillus, V. uliginosum, and V. vitis-idaea) from five different subgenera/sections (Cyanococcus, Hemimyrtillus, Myrtillus, Vaccinium and Vitis-idaea, respectively). An indication of the collection place as well as data from herbarium vouchers can be found in Table 1. For chromosome counts, young leaf buds were collected during spring and summer and instantly stored in ice and kept overnight. For flow cytometry assays, we sampled fresh young leaves from five individuals per population and stored them at 4°C until they were processed at the laboratory. Leaf tissue employed for DNA sequencing was sampled from the field and immediately stored in silica-gel.

Table 1
Provenance and voucher data of the studied populations

Species Location, collectors, collection date and voucher number

V. arctostaphylos Kackar Mountains, Rize National Park, Turkey. Sedat Serçe and Nusrat Sultana. 29th June 2018. BC-974075.

V. corymbosum† Cutivated at the Niğde Ömer Halisdemir University greenhouse (origin of the cultivar line: USA), Turkey. Sedat Serçe and Nusrat Sultana. 21st June 2018. BC-974080.

V. myrtillus (1) Canillo (1), Vall d’Incles, Andorra. Joan Vallès and Teresa Garnatje. 11th July 2018. BC-974082.

V. myrtillus (2) Canillo (2), Vall d’Incles, Andorra. Joan Vallès and Teresa Garnatje. 11th July 2018. BC-974081.

V. myrtillus (3) Devět Skal, Křižánky, Czech Republic. Sònia Garcia, Aleš Kovařík and Daniel Vitales. 29th May 2018. BC-974079.

V. myrtillus (4) Velke Darko, Škrdlovice, Czech Republic. Sònia Garcia, Aleš Kovařík and Daniel Vitales. 29th May 2018. BC-974078.

V. myrtillus (5) Rostejn, Trestice, Czech Republic. Sònia Garcia, Aleš Kovařík and Daniel Vitales. 24th May 2018. BC-974077.

V. myrtillus (6) Mangartska jama, Triglav National Park, Slovenia. Teresa Garnatje and Oriane Hidalgo. 24th July 2018. BC-974071.

V. myrtillus (7) Kackar Mountains, Rize National Park, Turkey. Sedat Serçe and Nusrat Sultana. 29th June 2018. BC-974074.

V. uliginosum (1) Canillo, Vall d’Incles, Andorra. Albert Ruzafa 8th July 2018. BCN-E-367.

V. uliginosum (2) Kackar Mountains, Rize National Park, Turkey. Sedat Serçe and Nusrat Sultana. 29th June 2018. BC-974076.

V. vitis-idaea (1) Novohradske Hory, Pohorská Ves, Czech Republic. Eva Bártová. 3rd June 2018. BC-974072.

V. vitis-idaea (2) Velke Darko, Czech Republic. Sònia Garcia, Aleš Kovařík and Daniel Vitales. 29th May 2018. BC-974073.

Species	Location, collectors, collection date and voucher number
V. arctostaphylos	Kackar Mountains, Rize National Park, Turkey. Sedat Serçe and Nusrat Sultana. 29th June 2018. BC-974075.
V. corymbosum†	Cutivated at the Niğde Ömer Halisdemir University greenhouse (origin of the cultivar line: USA), Turkey. Sedat Serçe and Nusrat Sultana. 21st June 2018. BC-974080.
V. myrtillus (1)	Canillo (1), Vall d’Incles, Andorra. Joan Vallès and Teresa Garnatje. 11th July 2018. BC-974082.
V. myrtillus (2)	Canillo (2), Vall d’Incles, Andorra. Joan Vallès and Teresa Garnatje. 11th July 2018. BC-974081.
V. myrtillus (3)	Devět Skal, Křižánky, Czech Republic. Sònia Garcia, Aleš Kovařík and Daniel Vitales. 29th May 2018. BC-974079.
V. myrtillus (4)	Velke Darko, Škrdlovice, Czech Republic. Sònia Garcia, Aleš Kovařík and Daniel Vitales. 29th May 2018. BC-974078.
V. myrtillus (5)	Rostejn, Trestice, Czech Republic. Sònia Garcia, Aleš Kovařík and Daniel Vitales. 24th May 2018. BC-974077.
V. myrtillus (6)	Mangartska jama, Triglav National Park, Slovenia. Teresa Garnatje and Oriane Hidalgo. 24th July 2018. BC-974071.
V. myrtillus (7)	Kackar Mountains, Rize National Park, Turkey. Sedat Serçe and Nusrat Sultana. 29th June 2018. BC-974074.
V. uliginosum (1)	Canillo, Vall d’Incles, Andorra. Albert Ruzafa 8th July 2018. BCN-E-367.
V. uliginosum (2)	Kackar Mountains, Rize National Park, Turkey. Sedat Serçe and Nusrat Sultana. 29th June 2018. BC-974076.
V. vitis-idaea (1)	Novohradske Hory, Pohorská Ves, Czech Republic. Eva Bártová. 3rd June 2018. BC-974072.
V. vitis-idaea (2)	Velke Darko, Czech Republic. Sònia Garcia, Aleš Kovařík and Daniel Vitales. 29th May 2018. BC-974073.

^†cultivar ‘Jubilee’.

2.2 Chromosome counts

Chromosomes were prepared according to [34] with some modifications. Fixation of leaf buds was performed in methanol/glacial acetic acid (3 : 1) with 1% PVP. The fixative solution was changed several rounds to ensure complete removal of all plant pigments and finally it was preserved at –20°C in the same solution until slide preparation. After removing the outer leaf coat under the microscope, the inner white meristematic portion of a single leaf bud was washed in citrate buffer (4 mM citric acid and 6 mM sodium citrate, pH 4.5) several times to remove the existing fixative and digested overnight at room temperature with the enzyme mixture PINE containing 2% (w/v) cellulase Aspergillus niger (Sigma C-1184), 4% (w/v) cellulase Onozuka R10 (Sigma 16419), 2% (w/v) cytohelicase from Helix pomatia (Serva C-8274), 0.5% (w/v) pectolyase from A. japonicus and 5% (w/v) pectinase from A. niger in citrate buffer. The material was digested until it became soft, followed by two cycles of centrifugation and washing in citrate buffer and one centrifugation step in fixative at 4°C at 4500 rpm. The final wash was done also in fixative at 4°C, increased to 5000 rpm. The pellet of nuclei was dissolved in 500 μl of fixative and mixed by slow pipetting. For each slide, 13 μl of mixture were spread, instantly washed in fixative three times and air dried. The air-dried slides were stained with 1% aceto-orcein solution and checked several rounds to ensure perfect staining, which is about 30 minutes. After proper staining, the slide was washed with 70% ethanol to remove extra dye and mounted with Canada balsam solution. A total of 10 well-spread cells, containing metaphase chromosomes, were counted for each species. Slides were analysed using a Leica-Epifluorescence microscope (LEICA, DMIL-LED) equipped with LAS software (Leica) and LEICA high resolution digital camera. Images of well-spread metaphase chromosomes were captured at 100× magnification and analysed. After image acquisition, the final picture was processed through Adobe Photoshop 7 software to balance brightness and contrast, and only using functions that affect the image as a whole.

2.3 Flow cytometry

Five plants per population were studied processing two samples per individual. The internal standards chosen were Petunia hybrida ‘PxPc6’ (2C = 2.85 pg) and Pisum sativum ‘Express Long’ (2C = 8.37 pg) in order to cover the possible ploidy differences with good linearity during the experiment [35]. Nuclei of both internal standard and Vaccinium samples were isolated by co-chopping the leaf tissues together in 600 μl of LB01 lysis buffer [36], supplemented with ribonuclease A (RNase A, Boehringer, Meylan, France). The fluorescence measurements were performed in a Gallios flow cytometer (Coulter Corporation; Hialeah, Fla, USA). The instrument was set up with the standard configuration; excitation of the sample was performed using a blue 488 nm laser. Forward scatter (FSC), side scatter (SSC), and red (620/30 nm) fluorescence (peak and area signals) for propidium iodide (PI) were acquired. Removal of aggregates and background noise was performed by gating red fluorescence area versus red fluorescence peak signals and FSC vs red florescence signal, respectively. Acquisition was stopped automatically at 8,000 nuclei. The lysed cell suspension was filtered through 70 μm nylon mesh to remove cell debris. Samples were incubated in ice until measurement and stained with 36 μl PI solution (1 mg/ml, Sigma-Aldrich, Alcobendas, Madrid) to a final concentration of 60 μg/ml. Flow cytometry measurements were taken twice for each sample and a minimum of 2000 particles per peak were acquired to get a consistent result. Genome sizes (total DNA or 2C-value) of the species were calculated according to the peak position ratios and following the formula: (sample peak mode/internal standard peak mode) * GS (2C) of the standard [37].

2.4 DNA extraction, PCR and sequencing strategy

Either the E.Z.N.A.^® Plant DNA Kit (Omega Bio-tek, Inc., Norcross, Georgia, USA) or the CTAB method [38] were used to extract genomic DNA from leaf materials, depending on their quality or available amount. The quality of each extraction was checked with Qubit Fluorometric Quantification (ThermoFisher Scientific, Waltham, Massachusetts, USA). Polymerase chain reaction (PCR) was performed using the T100TM Thermal Cycler (BioRad Laboratories, Hercules, California, USA) in a 25 μL volume following the procedures explained in [24]. Three regions were amplified in this study: one nuclear (Internal Transcribed Spacer of the nuclear ribosomal DNA, ITS) and two chloroplastic (matK and ndhF). The ITS region was amplified by PCR using ITS1f and ITS4r primers [39]. The PCR profile used for amplification was 94°C 3 min; 30×(94°C 20 s; 55°C 1 min; 72°C 1 min); 72°C 10 min [24]. In some cases, the pair of primers 1406f and ITS4r [40] were used when the former did not work. The sequencing primer used for this region was the ITS4r. The matK region was amplified by PCR using 1848f and 710r primers [41]. The ndhF region was amplified by PCR using two primer pairs; for the first segment we used the 90F (5^′ CGT ATC TGG GCT TTT CTA AGT G 3^′) and 912R (5^′ GAG CAA GTG CTA AAG TAG CTC CTA A 3^′) (both specifically designed), and for the second segment we used the primers 803F and 1318R [42]. The PCR profile used for matK and ndhF amplification was described in [43]. Sequencing was performed with the Big Dye Terminator Cycle Sequencing v3.1 (PE Biosystems, Foster City, California, USA) at the Serveis Científics i Tecnològics (Universitat de Barcelona) on an ABI PRISM 3700 DNA analyzer (PE Biosystems, Foster City, California, USA). Due to the length of the fragments amplified in both regions, we used the reverse and forward primers of each segment for the sequencing reaction. GenBank accession numbers for these sequences are MN134402-MN134402 (ITS), MN150130-MN150135 (ndhF) and MN150136-MN150141 (matK).

2.5 Phylogenetic and statistical analyses (including reconstruction of ancestral GS)

The chloroplast (matK and ndhF) and nuclear (ITS) datasets were analysed separately since loci from different genomic compartments may undergo independent evolution, potentially generating incorrect phylogenetic inferences based on concatenated datasets [44]. The cpDNA dataset contained 68 taxa with a sequence length of 2820 characters (1–1543 matK and 1544–2820 ndhF). The nrDNA dataset contained 68 taxa with a sequence length of 694 characters. The two matrices were aligned using MAFFT [45] and finally adjusted manually on Geneious Prime v2019.1.3. Non-sequenced fragments were scored as missing data (N). In total, the matrices included DNA data from 65 (nrDNA) and 64 (cpDNA) taxa, where five of them have been sequenced for the first time in this study and the remaining were obtained from Genbank (Table S1). The two datasets were analysed using Bayesian Inference (BI), previously fitting the best evolutionary model with MEGA X [46], independently for the nuclear and chloroplast matrices. For both datasets, the best evolutionary model fitted was the GTR + G + I. For the BI analyses, the program MrBayes v3.2.1 [47] was used to run two independent Markov chains Monte Carlo (MCMC) for 50,000,000 generations for each dataset, with tree sampling every 1,000 generations. The average standard deviation of the split frequencies was checked to be less than 0.01 and the potential scale reduction factor was near 1.0 for all parameters. The first 25% of the trees were discarded as ‘burn-in’ and the posterior probabilities were estimated constructing the 50% majority rule consensus tree, using as outgroup the Vaccinieae tribe species Gaultheria procumbens, Leucothoë fontanesiana, Zenobia pulverulenta and Andromeda polifolia.

To analyse the evolution of GS in Vaccinium, we performed the ancestral character reconstruction of GS values using unordered maximum parsimony implemented for continuous characters in Mesquite v.3.6 software [48]. First, we selected the species with available genome size and sequencing data for the molecular markers (cpDNA and nrDNA) mentioned above, either generated on this work or obtained from online repositories (i.e. Genbank) as well as published papers (Tables 2 and 3). The trees used as the input files for ancestral character reconstructions were generated by BI using only the species with available GS and sequence data (15 taxa, with Gaultheria procumbens set as outgroup), fitting the best evolutionary model with MEGA X. Finally, using these reduced phylogenetic reconstructions based on nrDNA and cpDNA, ancestral state inference was calculated for holoploid (2C) GS.

Table 2
Chromosome number and genome size of studied Vaccinium species. Standard deviation was calculated from the measurement of five different individuals, in most cases, each measured twice

Species Section Location 2n PL GS (2C pg)±SD 1Cx Internal standard

V. arctostaphylos Hemimyrtillus Turkey 48^* 4x 3.13±0.07 0.78 Pisum sativum

V. corymbosum† Cyanococcus Turkey 48^* 4x 3.17±0.18 0.79 Pisum sativum

V. myrtillus (1) Myrtillus Andorra 24 2x 1.28±0.13 0.64 Petunia hybrida

V. myrtillus (2) Myrtillus Andorra 24 2x 1.16±0.11 0.58 Petunia hybrida

V. myrtillus (3) Myrtillus Czech Republic 24 2x 1.26±0.04 0.63 Petunia hybrida

V. myrtillus (4) Myrtillus Czech Republic 24 2x 1.24±0.06 0.62 Petunia hybrida

V. myrtillus (5) Myrtillus Czech Republic 24 2x 1.20±0.03 0.60 Petunia hybrida

V. myrtillus (6) Myrtillus Slovenia 24 2x 1.30±0.03 0.65 Petunia hybrida

V. myrtillus (7) Myrtillus Turkey 24^* 2x 1.27±0.06 0.64 Petunia hybrida

V. uliginosum (1) Vaccinium Andorra 24 2x 1.47±0.11 0.74 Petunia hybrida

V. uliginosum (2) Vaccinium Turkey 24^* 2x 1.34±0.02 0.67 Petunia hybrida

V. vitis-idaea (1) Vitis-idaea Czech Republic 24 2x 1.28±0.03 0.64 Petunia hybrida

V. vitis-idaea (2) Vitis-idaea Czech Republic 24 2x 1.24±0.03 0.62 Petunia hybrida

Species	Section	Location	2n	PL	GS (2C pg)±SD	1Cx	Internal standard
V. arctostaphylos	Hemimyrtillus	Turkey	48^*	4x	3.13±0.07	0.78	Pisum sativum
V. corymbosum†	Cyanococcus	Turkey	48^*	4x	3.17±0.18	0.79	Pisum sativum
V. myrtillus (1)	Myrtillus	Andorra	24	2x	1.28±0.13	0.64	Petunia hybrida
V. myrtillus (2)	Myrtillus	Andorra	24	2x	1.16±0.11	0.58	Petunia hybrida
V. myrtillus (3)	Myrtillus	Czech Republic	24	2x	1.26±0.04	0.63	Petunia hybrida
V. myrtillus (4)	Myrtillus	Czech Republic	24	2x	1.24±0.06	0.62	Petunia hybrida
V. myrtillus (5)	Myrtillus	Czech Republic	24	2x	1.20±0.03	0.60	Petunia hybrida
V. myrtillus (6)	Myrtillus	Slovenia	24	2x	1.30±0.03	0.65	Petunia hybrida
V. myrtillus (7)	Myrtillus	Turkey	24^*	2x	1.27±0.06	0.64	Petunia hybrida
V. uliginosum (1)	Vaccinium	Andorra	24	2x	1.47±0.11	0.74	Petunia hybrida
V. uliginosum (2)	Vaccinium	Turkey	24^*	2x	1.34±0.02	0.67	Petunia hybrida
V. vitis-idaea (1)	Vitis-idaea	Czech Republic	24	2x	1.28±0.03	0.64	Petunia hybrida
V. vitis-idaea (2)	Vitis-idaea	Czech Republic	24	2x	1.24±0.03	0.62	Petunia hybrida

^†cultivar ‘Jubilee’. ^*chromosome counts obtained in the present study; the other counts obtained through the Chromosome Counts Database [25]

Table 3

Genome size value published so far for different taxa of the family Ericaceae in different individual studies. pg = Picogram, Mbp = Mega base pairs, FC = Flow cytometry, PI = Propidium Iodide, DAPI = 4^′,6-diamidino-2-phenylindole, NGS = data coming from a next generation sequencing project. Superscript numbers after species names indicate the source publication in the literature list

Taxa	2n	Ploidy level	Estimation method	2C (pg)	1Cx (pg)
Genus Vaccinium
Vaccinium boreale²⁶	24	2	FC:PI	1.18	0.59
V. corymbosum²⁶	24	2	FC:PI	1.33	0.67
V. darrowi²⁶	24	2	FC:PI	1.31	0.66
V. elliottii²⁶	24	2	FC:PI	1.26	0.63
V. myrtilloides²⁶	24	2	FC:PI	1.26	0.63
V. pallidum²⁶	24	2	FC:PI	1.21	0.61
V. tenellum²⁶	24	2	FC:PI	1.3	0.65
V. macrocarpon²⁸	24	2	NGS	1.16	0.58
V. arboreum¹²	24	2	FC:PI	1.02	0.51
V. crassifolium¹²	24	2	FC:PI	1.1	0.55
V. darrowii¹²	24	2	FC:PI	1.09	0.55
V. elliottii¹²	24	2	FC:PI	1.05	0.53
V. pallidum¹²	24	2	FC:PI	1.1	0.55
V. stamineum¹²	24	2	FC:PI	1.01	0.51
V. angustifolium¹²	48	4	FC:PI	2.02	0.51
V. arboreum¹²	48	4	FC:PI	1.98	0.50
V. corymbosum¹²	48	4	FC:PI	2.11	0.53
V. myrsinites¹²	48	4	FC:PI	2.11	0.53
V. pallidum¹²	48	4	FC:PI	2.1	0.53
V. pensylvanicum¹²	48	4	FC:PI	1.95	0.49
V. virgatum¹²	72	6	FC:PI	3.8	0.63
V. corymbosum ‘Jubilee’¹²	48	4	FC:PI	2.1	0.53
V. myrtilloides²⁷			FC:PI	1.29
V. angustifolium²⁷			FC:PI	2.07
Genus Arctostaphylos
A. uva-ursi⁵⁷			FC:PI	2.49
Genus Empetrum
E. hermaphroditum⁵⁸	52	4	FC:PI	2.56	0.64
E. nigrum⁵⁸	26	2	FC:PI	1.29	0.65
Genus Erica
E. multiflora⁵⁹			FC:PI	0.95
Genus Chimaphila
Chimaphila umbellata²⁷			FC:PI	18.2
Genus Gaultheria
Gaultheria procumbens²⁷			FC:PI	2.5 to 3.8
G. baccata²⁷			FC:PI	1.1 to 1.3
Genus Monotropa
Monotropa uniflora²⁷			FC:PI	59.8
Genus Pyrola
Pyrola elliptica²⁷			FC:PI	9.6 to 9.3
Genus Rhododendron
Rhododendron delavayi⁶⁰			NGS	1.42
R. catawbiense⁵⁴	2	FC:DAPI	1.44	0.72
R. fortunei⁵⁴	2	FC:DAPI	1.55	0.78
R. maximum⁵⁴	2	FC:DAPI	1.44 to 1.53	0.72
R. ponticum⁵⁴	2	FC:DAPI	1.46	0.73
R. sinogrande⁵⁴	2	FC:DAPI	1.54	0.77
R. edgeworthii⁵⁴	2	FC:DAPI	1.75	0.88
R. augustinii⁵⁴	4	FC:DAPI	3.1	0.78
R. maddenii⁵⁴	6	FC:DAPI	4.39 to 4.47	0.73
R. maddenii⁵⁴	8	FC:DAPI	5.42 to 5.97	0.68
R. alabamense⁵⁴	2	FC:DAPI	1.66	0.83
R. arborescens⁵⁴	2	FC:DAPI	1.65	0.83
R. austrinum⁵⁴	2	FC:DAPI	1.59 to 1.64	0.80
R. canescens⁵⁴	2	FC:DAPI	1.61 to 1.72	0.81
R. cumberlandense⁵⁴	2	FC:DAPI	1.63	0.82
R. eastmanii⁵⁴	2	FC:DAPI	1.58 to 1.60	0.79
R. flammeum⁵⁴	2	FC:DAPI	1.68 to 1.72	0.84
R. occidentale⁵⁴	2	FC:DAPI	1.51	0.76
R. triploides⁵⁴	2	FC:DAPI	2.3 to 2.52	1.15
R. atlanticum⁵⁴	4	FC:DAPI	3.01 to 3.33	0.75
R. austrinum⁵⁴	4	FC:DAPI	3.11 to 3.88	0.78
R. stenopetalum⁵⁴	2	FC:DAPI	1.27	0.64

3 Results

Chromosome counts were performed for all the studied species in order to verify the ploidy level of the accessions, except for V. vitis-idaea because of unavailability of the appropriate tissue. Two different ploidy levels (diploid and tetraploid) were detected. Vaccinium myrtillus and V. uliginosum were diploid with chromosome number 2n = 2x = 24 and V. corymbosum (‘Jubilee’ cultivar) and V. arctostaphylos were tetraploid with chromosome number 2n = 4x = 48 (Table 2 and Fig. 1). Genome size estimations were performed for all the studied materials (Table 1). The average half peak coefficient of variation (HPCV%) was 1.11% and 3.44% for internal standards and studied samples, respectively. For both V. corymbosum and V. arctostaphylos we only estimated the GS of one population, while seven populations were analysed for V. myrtillus (GS ranging from 1.16 pg to 1.30 pg; i.e. 10.77 % of intraspecific variation), two for V. uliginosum (GS ranging from 1.34 pg to 1.47 pg; i.e. 9.70% intraspecific variation) and two for V. vitis-idaea (GS ranging from 1.24 pg to 1.28 pg; i.e. 3.22% intraspecific variation). Average monoploid GS (1Cx) was 0.63 pg for diploids and 0.79 pg for tetraploids. The relationship between GS and ploidy level was analysed considering previously published data in the genus (Table 2). A positive significant relationship between holoploid GS (2C) and ploidy level was found (Pearson’s r = 0.94, DF = 21, p < 0.0001), while the relationship between monoploid GS (1Cx) and ploidy level was non-significant (Pearson’s r = –0.14, DF = 21, p = 0.4983).

Fig.1

Pictures and chromosome number of the studied Vaccinium species. (A-B) V. myrtillus (2n = 2x = 24); (C-D) V. uliginosum (2n = 2x = 24); (E-F) V. arctostaphylos (2n = 4x = 48); (G-H) V. corymbosum (cultivar ‘Jubilee’) (2n = 4x = 48), with detail of the fruit in the lower right corner. Scale bars 10 μm.

In order to place the studied species in a molecular systematic background we have reconstructed the phylogenetic history of the genus including these taxa (i.e. published phylogenetic reconstructions of Vaccinium did not include all the species here studied). Both reconstructions showed congruent results in the position and structure of many phylogenetic groups but there were also certain inconsistences regarding the placement of some species and clades. Both the nrDNA and the cpDNA trees supported the monophyly of tribe Vaccinieae but not that of genus Vaccinium, since species from several other genera (i.e. Cavendishia, Paphia, Notoptora, etc.) appear mixed in the clades containing most Vaccinium species. According to the reconstruction based on cpDNA (Figure S1), all the newly sequenced accessions were placed together with previously sequenced specimens of the same species. However, in the tree based on nrDNA (Figure S2), the new accessions of V. corymbosum ‘Jubilee’ and V. uliginosus we analysed in this study were split in different clades from previously sequenced specimens of the same species. In this nrDNA phylogenetic inference, the specimens of V. uliginosus from Andorra and Turkey were placed in a clade (PP = 0.97) with V. vitis-idaea and V. macrocarpon, while a previously sequenced accession of V. uliginosum (from Canada [49]) appeared in a clade (PP = 0.99) with V. arctostaphylos, V. cylindraceum and V. padifolium. Based on the same nrDNA reconstruction, the accession of V. corymbosum ‘Jubilee’ we sequenced was placed in an early-diverging clade with V. darrowii, V. angustifolium, V. hirsutum and V. myrsinites, whereas the formerly sequenced specimen of this species (coming from a wild population) occupied a poorly supported but phylogenetically distant position from that clade. Regarding the phylogenetic relationships among Vaccinium groups of species, the tree generated from the analysis of ITS sequences (Figure S1) was considerably more resolved than the one coming from chloroplast DNA (Figure S2).

Ancestral character reconstructions of GS values was limited to 16 taxa of the genus, some species being only present in the reconstruction based on nrDNA (i.e. V. darrowii; V. boreale) or in the one based on cpDNA (i.e. V. mirtylloides), depending on sequence data availability. The reduced evolutionary reconstruction based on rDNA sequences showed good resolution for most of the nodes (PP > 0.95; Fig. 2) and the systematic structure was congruent with the ITS tree of Vaccinieae taxa (Fig. S1) as well as with previous phylogenetic treatments of the tribe [3, 16]. In contrast, as occurred in the cpDNA tree of Vaccinieae (Fig. S2), the ancestral GS phylogenetic reconstruction based on cpDNA data (Fig. 2) showed poor resolution in most of the nodes of the tree. The ancestral state inference indicates a GS reduction for Vaccinium with respect to the outgroup (Gaultheria). According to both nrDNA and cpDNA reconstructions, several independent shifts of GS within the genus have occurred, most of them related to changes in ploidy level of the species. The reconstructed ancestral GS inference for the MRCA of the genus ranged between 1.88 to 2.31 pg for 2C values.

Fig.2

Reconstruction of the ancestral GS (2C) values in the studied species of Vaccinium based on nrDNA and cpDNA phylogenetic inferences. Gaultheria procumbens was used as an outgroup. Values above branches indicate posterior probabilities (PP). Sequences obtained for this study indicated by an asterisk. Numbers in superscripts indicate the provenance of populations (as in Table 1).

4 Discussion

The study has confirmed previously known chromosome numbers for all the studied species excepting for V. vitis-idaea, for which material for chromosomal analysis was not available. According to the Chromosome Counts Database [25] the obtained counts agree with the most commonly found numbers for V. corymbosum (in which the tetraploid count 2n = 48 is much more common than the exceptionally found 2n = 24, 36, 54, 56, 60, 88 and 72), V. arctostaphylos (in which only the tetraploid count 2n = 48 has been found up to date) and V. myrtillus (in which, although 2n = 24 is the overall most common, there are exceptional counts of 2n = 20 and 48). In the case of V. uliginosum, we counted 2n = 24 in the accession from Turkey, while both 2n = 24 and 48 are similarly found in previous studies (although rare counts of 2n = 39, 45 and 72 have also been reported). The accessions of V. uliginosum from the Czech Republic and from Andorra (the last one also being diploid according to our GS estimates) are placed together in our nrDNA phylogenetic reconstruction (Fig. S1) (which is overall consistent with those of [3, 16] on the tribe and genus) but in a different clade from the previously sequenced samples obtained from Genbank. Phylogenetic split among diploid and tetraploid individuals of V. uliginosum was already reported by [50], interpreting this pattern as a result of hybridization, polyploidization and homogenization processes during the complex evolutionary history of the species. As mentioned previously, we could not assess chromosome number for any of our populations of V. vitis-idaea, but the GS values we obtained indicate that both analysed populations are most likely diploid. Indeed, for this species, all extant counts are 2n = 24 except a single one reporting a triploid accession (2n = 36) [25]. However, whenever possible, it is recommended to perform chromosome counts on the same populations in which GS will be assessed to ascertain ploidy, rather than inferring it from previously known data, since intraspecific variation can exist at the chromosome level, involving different ploidy levels, the presence of B-chromosomes, aneuploidy or disploidy.

This work contributes to the first GS data for three species which are relatively common in mostly cold (V. vitis-idaea and V. uliginosum) or temperate (V. arctostaphylos) regions in the Northern hemisphere, the former two including populations in Eurasia and North America. Both at diploid and tetraploid levels our new data extend the range of GS from the upper side, contributing the highest GS for diploid Vaccinium to V. uliginosum (2C = 1.47 pg) and the highest GS for tetraploid Vaccinium to V. corymbosum ‘Jubilee’ (2C = 3.16 pg). Even if we do not consider this cultivar, in which hybridization or other domestication processes could have been involved (see comment on this below), the second highest GS in the genus for a tetraploid species also belongs to our newly assessed V. arctostaphylos (2C = 3.13 pg). Our study also provides data for new populations of the economically important and widely distributed species V. myrtillus. In this case, we measured geographically distant populations from Andorra, Turkey and the Czech Republic, while the only GS assessment to date (2C = 1.17 pg) [51] belonged to an accession from Bosnia and Herzegovina. Our data, ranging from 2C = 1.16–1.28 pg are consistent with previous measurements of V. myrtillus, indicating a narrow range of GS variation irrespective of the geographical distribution of the populations analysed. Low intraspecific variation has also been found for both V. uliginosum and V. vitis-idaea. Only in the case of V. corymbosum ‘Jubilee’ have we found a remarkable degree of variation with respect to previously known GS estimates, our measures being 50.48% larger than the previously assessed [12]. The different GS values could be explained by the fact that this is one of the most popular and widely extended blueberry cultivars – because of its ornamental and food value – in which hybridization processes for breeding purposes are likely involved [8]. Indeed, the incongruent phylogenetic position of V. corymbosum ‘Jubilee’ between cpDNA and nrDNA reconstructions (Figs. S1 and S2) supports the hybridization hypothesis.

The known GS range for Vaccinium species (from 2C = 1.01 to 1.47 pg at the diploid level and from 2C = 1.95 to 3.17 pg at the tetraploid level) is quite narrow and using Leitch et al. [52] categories (so-called Leitch’s criteria) all species would belong to the group of “very small genomes” (2C≤2.8 pg) except our tetraploids V. corymbosum and V. arctostaphylos and the hexaploid V. virgatum (2C = 3.8 pg, [12]) which would be considered just “small genomes” (2C = 2.9–7 pg). It should be noted that, although the known range may not reliably reflect the real GS variation within the genus, the species with known GS belong to nine sections out of the 30 within the genus [1], so although scarce, the extant data could barely represent the probably small GS dominant in the genus and its low diversity within Vaccinium. The reconstruction of the ancestral GS state in the genus (2C = 1.88–2.31 pg), which needs to be taken with caution given the scarcity of GS data for Vaccinium species (genome size estimated for 19 taxa for a genus with ca. 500 species, and only 16 could be used for the reconstruction), also points to “very small” GS according to Leitch’s categories. A very small genome size is considered the ancestral condition, as well as the most common, in flowering plants [52]. Within the genus (considering diploid GS) there have been scattered events leading to an increased GS in certain species (see Fig. 2). Although their GS would still fall within the “very small” category, these mild changes reflect the dynamic nature of GS evolution, even in groups with a narrow range of variation.

A common response to polyploidy is genome downsizing [53] in which whole genome duplication (WGD) is not accompanied by a proportional GS increase. The available data for Vaccinium GS and ploidy levels indicates a slight genome downsizing from the diploid to the tetraploid level (average 1Cx = 0.60 pg for diploids and average 1Cx = 0.56 pg for tetraploids), however, the only hexaploid with available data (1Cx = 0.63 pg) points to a genome upsizing for this ploidy level. Conversely, the available data for the Ericaceae genus Rhododendron points to a clear and global genome downsizing in which 1Cx is gradually reduced across ploidy levels. However, it would be necessary to have more GS data at the genus level in order to properly assess the extent of genome ups and downs in Vaccinium, since such differences would be better assessed between closely related species, and our dataset is quite scanty.

Genome size data availability for family Ericaceae are even more scattered than for the genus Vaccinium. Despite being one of the largest plant families with more than 4000 species and around 125 genera [16], up to our knowledge we only have GS for a few species from genera Arctostaphylos (1/ca. 60), Empetrum (2/3-18), Erica (1/ca. 900), Chimaphila (1/5), Gaultheria (2/ca. 135), Monotropa (1/2), Pyrola (1/ca. 30) and Rhododendron (21/ca. 1000). Table 3 shows the known GS in family Ericaceae together with the source publications, which would represent roughly 1% of Ericaceae with known GS. Apparently, genus Vaccinium is one of the best represented (among the genera with a relatively high number of species), and genus Rhododendron is similarly known. Both genera are more common in temperate or even cold regions of the Northern hemisphere (although both have tropical species). Certainly, the large diversity of both genera, their widespread distribution and their economic interest (either as edible/medicinal for Vaccinium or mostly ornamental for Rhododendron) may have played a role on relative GS data abundance of these taxa. Despite being not too closely related from the phylogenetic point of view [55], both share similar GS ranges, similar base chromosome number (x = 12 in Vaccinium and x = 13 in Rhododendron) and ploidy levels. However, other Ericacaceae genera (i.e. Monotropa, Pyrola and Chimaphila) show contrastingly higher GS values (see Table 3). In the case of Monotropa uniflora, the parasitic nature of this species could be involved in its extremely large genome, as found previously in other parasitic plants (e.g. Viscum album [56]).

5 Conclusions

This study provides new GS data for three Vaccinium species and contributes additional GS information for two more, all of them can be considered as “very small”, and an apparent GS reduction has occurred in the genus; altogether GS are known for about 4% of the genus when considering previous data. We also confirm the most common ploidy levels for the assessed species, namely 2x and 4x. Most of the assessed Vaccinium species have economic interest as food and ornamentals or are wild relatives of important crops in this genus such as V. myrtillus, for which this study extends its known GS range of variation, or V. corymbosum, one of the most widely cultivated blueberry species both for edible and ornamental uses. We have also discussed GS variation and ancestral state reconstruction considering the phylogenetic framework of the studied species, although the scarcity of GS data available for the genus limits the extent of its conclusions, and a more detailed study covering a larger taxonomic diversity of the genus would be desirable. Finally, the amount of genomic research data for Vaccinium is steadily increasing [Genome Database for Vaccinium: https://www.vaccinium.org/]. Given that GS is an essential prior information to consider in any study concerning whole genome sequencing, this information will be useful in any genomic approach that may be attempted for these taxa.

Conflict of interest

The authors have no conflict of interest to report.

Footnotes

The supplementary material is available in the electronic version of this article: .

Acknowledgments

This work was supported by the Dirección General de Investigación Científica y Técnica (Spanish Government: CGL2016-75694-P, AEI/FEDER, UE) and the Generalitat de Catalunya (“Ajuts a grups de recerca consolidats” 2017SGR01116). S.G. benefited from a Ramón y Cajal contract (RYC-2014-16608) from the government of Spain and N.S., A.G. and K.I. are thankful to Erasmus+Student traineeship programme of Niğde Ömer Halisdemir University, for providing the fellowship during the visit of Institut Botànic de Barcelona (IBB-CSIC). Eva Bártová, Teresa Garnatje, Oriane Hidalgo, Aleš Kovařík, Albert Ruzafa and Joan Vallès are thanked for providing material of Vaccinium for GS assessments and for technical support. Jaume Comas and Chari González are also thanked for technical support in flow cytometric assessments.

References

Vander Kloet

, Dickinson

. A subgeneric classification of the genus Vaccinium and the metamorphosis of V. section Bracteata Nakai: More terrestrial and less epiphytic in habit, more continental and less insular in distribution. J Plant Res. 2009;122(3):253–68.

Song

and Hancock

. Vaccinium . In: Wild crop relatives: Genomic and breeding resources. Springer, 2011;197–221.

Powell

, Kron

. Hawaiian blueberries and their relatives-a phylogenetic analysis of Vaccinium sections Macropelma, Myrtillus, and Hemimyrtillus (Ericaceae). Syst Bot. 2002;27(4):768–79.

Carbone

, Mourgues

, Perrotta

, Rosati

. Advances in functional research of antioxidants and organoleptic traits in berry crops. BioFactors. 2008;34(1):23–36.

Kalt

, Dufour

. Health functionality of blueberries. HortTechnology. 1997;7(3):216–21.

Szajdek

, Borowska

. Bioactive compounds and health-promoting properties of Berry fruits: A review. Plant Foods Hum Nutr. 2008;63(4):147–53.

Retamales

, Hancock

. Blueberries. CABI. 2012;21.

Ballington

. Collection utilization and preservation of genetic resources in Vaccinium. HortScience. 2001;36(2):213–20.

Brevis

, Bassil

, Ballington

, Hancock

. Impact of wide hybridization on highbush blueberry breeding. J Am Soc Hortic Sci. 2008;133(3):427–37.

10.

Hancock

, Lyrene

, Finn

, Vorsa

, Lobos

. Blueberries and cranberries. In: Hancock

, editor. Temperate fruit crop breeding. Springer; 2008, p. 115–50.

11.

Česonienė

, Daubaras

, Paulauskas

, Žukauskienė

, Zych

. Morphological and genetic diversity of European cranberry (Vaccinium oxycoccos L., Ericaceae) clones in Lithuanian reserves. Acta Soc Bot Pol. 2013;82(3):211.

12.

Sakhanokho

, Rinehart

, Stringer

, Islam-Faridi

, Pounders

. Variation in nuclear DNA content and chromosome numbers in blueberry. Sci Hortic-Amsterdam. 2018;233:108–13.

13.

Ehlenfeldt

, Ballington

. Vaccinium species of section Hemimyrtillus: Their value to cultivated blueberry and approaches to utilization. Botany. 2012;90(5):347–53.

14.

Frankham

, Briscoe

, Ballou

. Introduction to conservation genetics. Cambridge University Press; 2002.

15.

Ngo Ngwe

MFS

, Omokolo

, Joly

. Evolution and phylogenetic diversity of yam species (Dioscorea spp.): Implication for conservation and agricultural practices. PLoS One. 2015;10(12):e0145364.

16.

Kron

, Powell

, Luteyn

. Phylogenetic relationships within the blueberry tribe (Vaccinieae, Ericaceae) based on sequence data from matK and nuclear ribosomal ITS regions, with comments on the placement of Satyria. Am J Bot. 2002;89(2):327–36.

17.

Longley

. Chromosomes in Vaccinium. Science. 1927;66(1719):566–8.

18.

Matuszak

. Evolution of Mountain Plants in the Region of the Qinghai-Tibetan Plateau and Beyond. Verlag Nicht Ermittelbar. 2015.

19.

Vorsa

, Ballington

. Fertility of triploid highbush blueberry. J Amer Soc Hort Sci. 1991;116(2):336–41.

20.

Lobos

, Hancock

. Breeding blueberries for a changing global environment: A review. Front Plant Sci. 2015;6:782.

21.

Bennett

, Leitch

. Genome size evolution in plants. In: Gregory

, editor. The evolution of the genome. Elsevier Academic Press; 2005; p. 89–162.

22.

D’Ambrosio

, Alonso-Lifante

, Barros

, Kovařík

, de Xaxars

, Garcia

. B-chrom: A database on B-chromosomes of plants, animals and fungi. New Phytol. 2017;216(3):635–42.

23.

Olanj

, Garnatje

, Sonboli

, Vallès

, Garcia

. The striking and unexpected cytogenetic diversity of genus Tanacetum L. (Asteraceae): A cytometric and fluorescent in situ hybridisation study of Iranian taxa. BMC Pl Biol. 2015;15(1):174.

24.

Inceer

, Garnatje

, Hayırlıoğlu-Ayaz

, Pascual-Díaz

, Vallès

, Garcia

. A genome size and phylogenetic survey of Mediterranean Tripleurospermum and Matricaria (Anthemideae, Asteraceae). PloS One. 2018;13(10):e0203762.

25.

Rice

, Glick

, Abadi

, Einhorn

, Kopelman

, Salman-Minkov

, Mayzel

, Chay

, Mayrose

. The Chromosome Counts Database (CCDB)–a community resource of plant chromosome numbers. New Phytol. 2015;206(1):19–26.

26.

Costich

, Ortiz

, Meagher

, Bruederle

, Vorsa

. Determination of ploidy level and nuclear DNA content in blueberry by flow cytometry. Theor Appl Genet. 1993;86(8):1001–6.

27.

Bai

, Alverson

, Follansbee

, Waller

. New reports of nuclear DNA content for 407 vascular plant taxa from the United States. Ann Bot. 2012;110(8):1623–9.

28.

Georgi

, Herai

, Vidal

, Carazzolle

, Pereira

, Polashock

, Vorsa

. Cranberry microsatellite marker development from assembled next-generation genomic sequence Mol Breeding. 2012;30(1):227–37.

29.

Garcia

, Inceer

, Garnatje

, Valles

. Genome size variation in some representatives of the genus Tripleurospermum. Biol Plantarum. 2005;49(3):381–7.

30.

Gruner

, Hoverter

, Smith

, Knight

. Genome size is a strong predictor of root meristem growth rate. J Bot. 2010.

31.

Knight

, Molinari

, Petrov

. The large genome constraint hypothesis: Evolution, ecology and phenotype. Ann Bot. 2005;95(1):177–90.

32.

Guignard

, Nichols

, Knell

, Macdonald

, Romila

, Trimmer

, Leitch

. Genome size and ploidy influence angiosperm species’ biomass under nitrogen and phosphorus limitation. New Phytol. 2016;210(4):1195–206.

33.

MacGillivray

, Grime

. Genome size predicts frost resistance in British herbaceous plants: Implications for rates of vegetation response to global warming. Funct Ecol. 1995:320–5.

34.

Heitkam

, Petrasch

, Zakrzewski

, Kögler

, Wenke

, Wanke

, Schmidt

. Next-generation sequencing reveals differentially amplified tandem repeats as a major genome component of Northern Europe’s oldest Camellia japonica. Chrom Res. 2015;23(4):791–806.

35.

Doležel

, Greilhuber

. Nuclear genome size: Are we getting closer? Cytom Part A. 2010;77(7):635–42.

36.

Doležel

, Binarová

, Lucretti

. Analysis of nuclear DNA content in plant cells by flow cytometry. Biol Plantarum. 1989;31(2):113–20.

37.

Doležel

, Bartoš

JAN

. Plant DNA flow cytometry and estimation of nuclear genome size. Ann Bot. 2005;95(1):99–110.

38.

Doyle

, Doyle

. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem Bull. 1987;19:11–5.

39.

White

, Bruns

, Lee

SJWT

, Taylor

. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. PCR protocols: A guide to methods and applications. 1990;18(1):315–22.

40.

Nickrent

, Schuette

, Starr

. A molecular phylogeny of Arceuthobium (Viscaceae) based on nuclear ribosomal DNA internal transcribed spacer sequences. Am J Bot. 1994;81(9):1149–60.

41.

Johnson

, Soltis

. matK DNA sequences and phylogenetic reconstruction in Saxifragaceae s. str.. Syst Bot. 1994:143–56.

42.

Olmstead

, Sweere

. Combining data in phylogenetic systematics: An empirical approach using three molecular data sets in the Solanaceae. Syst Biol. 1994;43(4):467–81.

43.

Shaw

, Lickey

, Schilling

, Small

. Comparison of whole chloroplast genome sequences to choose noncoding regions for phylogenetic studies in angiosperms: The tortoise and the hare III. Am J Bot. 2007;94(3):275–88.

44.

Degnan

, Rosenberg

. Gene tree discordance, phylogenetic inference and the multispecies coalescent. Trends Ecol Evol. 2009;24(6):332–40.

45.

Katoh

, Standley

. MAFFT multiple sequence alignment software version Improvements in performance and usability. Mol Biol Evol. 2013;30(4):772–80.

46.

Kumar

, Stecher

, Li

, Knyaz

, Tamura

. MEGA X: Molecular Evolutionary Genetics Analysis across computing platforms. Mol Biol Evol. 2018;35:1547–9.

47.

Ronquist

, Teslenko

, Van Der Mark

, Ayres

, Darling

, Höhna

, Larget

, Liu

, Suchard

, Huelsenbeck

. MrBayes 3.2: Efficient Bayesian phylogenetic inference and model choice across a large model space. Syst Biol. 2012;61(3):539–42.

48.

Maddison

, Maddison

[homepage on the Internet]. Mesquite: A modular system for evolutionary analysis. 2018 [update 2018 Dec 27; cited 2019 Jun 03]. Available from: http://www.mesquiteproject.org/

49.

Saarela

, Sokoloff

, Gillespie

, Consaul

, Bull

. DNA barcoding the Canadian Arctic flora: Core plastid barcodes (rbcL+matK) for 490 vascular plant species. PLoS One. 2013;8(10):e77982.

50.

Eidesen

, Alsos

, Popp

, Stensrud

, Suda

, Brochmann

. Nuclear vs. plastid data: Complex Pleistocene history of a circumpolar key species. Mol Ecol. 2007;16(18):3902–25.

51.

Pustahija

, Brown

, Bogunić

, Bašić

, Muratović

, Ollier

, Hidalgo

, Bourge

, Stevanović

, Siljak-Yakovlev

. Small genomes dominate in plants growing on serpentine soils in West Balkans, an exhaustive study of 8 habitats covering 308 taxa. Plant Soil. 2013;373(1-2):427–53.

52.

Leitch

, Chase

, Bennett

. Phylogenetic analysis of DNA C-values provides evidence for a small ancestral genome size in flowering plants. Ann Bot. 1998;82(suppl_1):85–94.

53.

Leitch

, Bennett

. Genome downsizing in polyploid plants. Biol J Linn Soc. 2004;82(4):651–63.

54.

Jones

, Ranney

, Lynch

, Krebs

. Ploidy levels and relative genome sizes of diverse species, hybrids, and cultivars of Rhododendron. J Amer Rhododendron Soc. 2007;61(4):220–7.

55.

Schwery

, Onstein

, Bouchenak-Khelladi

, Xing

, Carter

, Linder

. As old as the mountains: The radiations of the Ericaceae. New Phytol. 2015;207(2):355–67.

56.

Zonneveld

BJM

. New record holders for maximum genome size in eudicots and monocots. J Bot. 2010.

57.

Siljak-Yakovlev

, Pustahija

, Šolić

, Bogunić

, Muratović

, Bašić

, Catrice

, Brown

. Towards a genome size and chromosome number database of Balkan flora: C-values in 343 taxa with novel values for 242. Adv Sci Lett. 2010;3(2):190–213.

58.

Suda

. New DNA polidy level in Empetrum (Empetraceae) revealed by flow cytometry. Ann Bot Fenn. 2002:133–41.

59.

Pellicer

, Estiarte

, Garcia

, Garnatje

, Peñuelas

, Sardans

& Vallès

. Genome size unaffected by moderate changes in climate and phosphorus availability in Mediterranean plants. Afr J Biotechnol. 2010;9(37):6070–7.

60.

Zhang

, Xu

, Cai

, Ma

, Li

, Xie

, Song

, Peng

, Yan H and

Zou L

. The draft genome assembly of Rhododendron delavayi Franch. var. delavayi. GigaScience. 2017;6(10):1–11.