Abstract
The genus Magnolia includes over 250 species that range in ploidy level from diploid to hexaploid. Although there is basic information on ploidy levels of various species, sampling has been limited and little information on specific cultivars and hybrids is available. The objective of this research was to determine relative genome sizes and relationships to ploidy levels among a diverse collection of species, hybrids, and cultivars using flow cytometry. Nuclei were extracted, stained with 4′, 6-diamidino-2-phenylindole (DAPI), and analyzed using a flow cytometer. Relative genome sizes were determined using Pisum sativum as the reference genome. Genome size was calibrated with ploidy level for species with documented chromosome numbers. Relative genome size for a given ploidy level varied significantly among most taxonomic sections indicating these groups have undergone considerable genomic divergence. These data also indicate it is desirable to calibrate ploidy level with relative genome size for each section separately. Within a section, relative 2C genome sizes, for a given ploidy level, had narrow ranges and could be used to clearly distinguish between euploid levels. Genome size estimates, determined with DAPI or propidium iodide fluorochromes, varied (by 0% to 14%) as a function of species and base pair (bp) composition. Both methods were suitable for determining euploid level. Base pair composition of representative Magnolia species ranged from 61.6% to 63.91% AT. Genome sizes and ploidy levels are presented for a broad range of species and hybrids within genus Magnolia. This information also provides further insight into reproductive biology, substantiation of numerous hybrids and induced polyploids, and comparison of methods for determining genome size that will help facilitate the development of improved hybrids in the future.
Polyploidy has been an important process in the evolution of plants that can contribute to reproductive isolation, novel gene expression, and ultimately divergence and speciation (Adams and Wendel, 2005; Comai, 2005; Hegarty and Hiscock, 2008; Soltis and Burleigh, 2009; Soltis et al., 2003). Polyploidy is also an important factor in plant breeding because it can influence reproductive compatibility, fertility, and phenotypic traits (Chen and Ni, 2006; Jones and Ranney, 2009; Ranney, 2006; Soltis et al., 2004). In some cases, the artificial induction of polyploidy in Magnolia also can enhance ornamental characteristics, including thicker leaves and larger flowers with thicker petals that persist longer (Kehr, 1985). As such, accurate and specific knowledge of ploidy levels of species and cultivars is important information for magnolia breeders.
The genus Magnolia comprises more than 250 species belonging to various sections within three subgenera (Figlar and Nooteboom, 2004). Although basic information on chromosome counts and ploidy levels of different Magnolia species have been compiled (Callaway, 1994; Chen et al., 2000), sampling has been limited and little is known about ploidy levels of specific hybrids and cultivars. The base chromosome number for Magnolia is 1n = 1x = 19. However, different subgenera contain species with a variety of ploidy levels ranging from 2n = 2x = 38 to 2n = 6x = 114. Crosses between species with varying ploidy levels may yield hybrids with nonstandard chromosome numbers that can result in reduced fertility or sterility. Because of these constraints, Magnolia breeders have attempted to induce new polyploids to overcome these limitations, yet most of these putative polyploids have never been confirmed. The range in ploidy levels within this genus also provides an opportunity to indirectly substantiate hybridity when parents differ in ploidy levels.
Because many Magnolia species are polyploids with high chromosome numbers, traditional cytology based on light microscopic examination is a difficult and time-consuming process. Flow cytometry has proved to be an efficient means of estimating genome size and associated ploidy level (Doležel et al., 2007; Jones et al., 2007). Therefore, the objectives of this study were to determine the genome sizes and relationships to ploidy levels of a diverse collection of species, hybrids, and cultivars of Magnolia to 1) develop an extensive database of ploidy levels for use by magnolia breeders; 2) determine the ploidy levels of plants that were chemically treated to artificially induce polyploidy; 3) confirm hybridity of interploid and interspecific (when parents vary substantially in genome size) crosses; and 4) compare estimates of genome size using DAPI (AT preferential) or propidium iodide (PI) (intercalating) fluorochrome stains and estimate bp composition for representative taxa from 10 taxonomic sections.
Materials and Methods
Relative genome size and ploidy level determination.
Over 300 accessions were sampled from various sources that included 62 species, 125 hybrids, and 16 induced polyploids representing taxa from each subgenus of Magnolia as well as both species of Liriodendron, the only other genus in family Magnoliaceae per Figlar and Nooteboom (2004). Nuclei from newly expanded leaf or tepal tissue were extracted, stained with DAPI, and then analyzed (minimum of 2500 nuclei per sample) using a flow cytometer (PA-I; Partec, Münster, Germany) to determine relative holoploid 2C DNA content following the methods of Jones et al. (2007). Genome sizes were determined by comparing mean relative fluorescence of each sample with an internal standard, Pisum sativum ‘Ctirad’, with a known genome size of 8.76 pg (Greilhuber et al., 2007). Because tetraploid Magnolia taxa have similar genome sizes to P. sativum ‘Ctirad’, Magnolia virginiana ‘Jim Wilson’ [NCSU 2004-24 (3.92 pg)] was used as a secondary standard. Absolute genome size for the secondary standard was calculated as the mean of 10 separate subsamples determined with P. sativum ‘Ctirad’ as an internal standard and PI as the fluorochrome stain (see procedure subsequently in “Comparison of fluorochromes and estimate of base pair composition”). Holoploid, 2C DNA contents were calculated as: 2C = DNA content of standard × (mean fluorescence value of sample ÷ mean fluorescence value of the standard).
The relationship between ploidy levels and genome sizes was determined for plants with documented chromosome numbers (Chen et al., 2000). Mean 1Cx monoploid genome size (i.e., DNA content of the non-replicated base set of chromosomes with x = 19) was calculated as 2C genome size ÷ ploidy level to assess variability in base genome size. A minimum of two subsamples was tested to derive a mean relative genome size for each accession. Data for species were subjected to analysis of variance and means separation using the Waller procedure (Proc GLM, SAS Version 9.1; SAS Institute, Cary, NC). Ploidy levels for hybrid taxa and suspected aneuploid hybrids were derived in the following manner: ploidy level = mean 2C genome size ÷ weighted average 1Cx genome size of the reported parental species.
Comparison of fluorochromes and estimate of base pair composition.
Ten species were sampled that included taxa from each subgenus of Magnolia. Nuclei were extracted, stained, and analyzed as described previously using a minimum of 3000 nuclei per sample. Sample preparation was similar to methods described for DAPI with the exception that the staining solution consisted of 2 mL staining buffer, 6 μL RNase A, and 12 μL PI (CyStain PI absolute P; Partec) and the samples were maintained at 4 °C for 1 h before flow cytometry analysis using a 488-nm laser for excitation (PA-II; Partec). The experimental design was a split-plot design with fluorochrome (DAPI versus PI) as the whole plot and species as the subplot. Samples were collected and analyzed over time in complete blocks. Data were subjected to analysis of variance and mean separation using Fisher's least significant difference specifically calculated for comparing two whole plot (fluorochrome) factors for a given subplot (species). Base pair composition was calculated following the equation: AT% = AT% for internal standard × [(fluorescence internal standard, DAPI/fluorescence sample, DAPI) ÷ (fluorescence internal standard, PI/fluorescence sample, PI)](1/binding length) (Godelle et al., 1993), where AT% of the internal standard, Pisum sativum = 61.50% and binding length of DAPI ≈3.5 bp (Meister and Barrow, 2007).
Cytology.
Actively growing root tips of container-grown seedlings of putative octaploid M. cylindrica were collected at midday and placed in the mitotic inhibitor 8-hydroxyquinoline for 2 h at 5 °C in dark conditions. They were then transferred to a fixative solution of three parts 95% ethanol:one part glacial acetic acid (v/v) for 24 h while remaining at 5 °C in dark conditions. Tissue was excised from just behind the root tip and placed in 12 N HCl for 10 s. Squashes were prepared with a small amount of this tissue and a drop of modified Fuelgen stain on a slide with a coverslip.
Results and Discussion
Relative genome size and ploidy level among species.
Relative genome sizes and ploidy levels were determined for 175 accessions, representing 62 species of Magnoliaceae and arranged by taxonomic sections following Figlar and Nooteboom (2004) (Tables 1 and 2). Base, 1Cx genome size varied significantly among plants sampled from different taxonomic sections indicating these groups have undergone considerable genome size divergence (Table 1). This variation indicates it is necessary to calibrate ploidy level with genome size for each section to estimate ploidy level from genome size in Magnolia. However, within a section, genome sizes for a given ploidy level had sufficiently narrow ranges that they could be used to clearly determine ploidy levels. Diploidy was prevalent throughout taxonomic sections, but variation in ploidy level occurred among species within several sections. Section Magnolia in subgenus Magnolia had both diploid and hexaploid members, whereas section Yulania in subgenus Yulania was represented by diploid, tetraploid, and hexaploid species. The two species tested in section Gynopodium, subgenus Gynopodium, were both hexaploid.
Summary of means and ranges for 2C, holoploid genome size (pg), and 1Cx monoploid genome size (pg) of Magnolia species grouped by section and ploidy level.
Relative genome size and estimated ploidy level for a diverse collection of Magnoliaceae representing 62 species.
Ploidy levels of species were generally consistent with past reports (Chen et al., 2000; Treseder, 1978; Xia et al., 2008) with some new additions and clarifications. Samples from wild-collected M. cylindrica (Bartlett 193, Holden 96-111A, Holden 96-115B, and MGA 216/Holden 87-86-93) were found to be tetraploid, having relative 2C genome sizes ranging from 8.82 to 9.11 (Table 2), in agreement with Xia et al. (2008) but not with prior reports (Treseder, 1978) that indicated M. cylindrica was diploid. Earlier reports may have varied as a result of lack of confirmed, wild-collected accessions in gardens of Europe and North America as stated by Callaway (1994). Chromosome counts have not been published for M. zenii, a species recently introduced into cultivation. The three accessions of M. zenii (MGA 440/Arnold 1545-80-B, Chollipo Form, and ‘Pink Parchment’) tested here were diploid with a mean relative genome size of 4.16 pg. Magnolia biondii has been reported to be tetraploid (Xia et al., 2008), although we found two M. biondii accessions (MGA 027 and Bartlett 2002-056) to be diploid with a mean relative genome size of 4.11 pg. In our study, no natural variation in ploidy level was found among accessions within a given species.
Relative genome size and ploidy level among hybrids.
Genome sizes and ploidy levels were determined for a broad range of reported interspecific, intra- and interploid hybrids (Table 3). In certain cases, analysis of genome size helped to substantiate or refute the authenticity of the hybrids. For example, the intersectional, intraploid hybrid Magnolia ‘Katie-O’ (NCSU 2004-012, MGA 307) had a mean 2C genome size of 4.30 pg, intermediate between the reported parents of M. insignis (2C = 4.94 pg) × M. virginiana (2C = 3.72 pg), supporting hybridity. Additional interspecific, intraploid hybrids strongly supported by genome size analysis include M. yuyuanensis × M. virginiana, (NCSU 2009-131), M. virginiana ‘Havener’ × M. insignis Red Form, 111/7, (McCracken), and [(M. tripetala × M. obovata) × M. tripetala] ‘Silk Road’ × M. insignis (MGA). Flow cytometry did not typically allow for distinguishing interspecific hybrids within a given section and ploidy level as a result of conserved genome sizes within sections. Taxa including M. ×kewensis, M. ×loebneri, M. ×brooklynensis, and M. ×veitchii fall into this category.
Relative genome size and estimated ploidy level for interspecific hybrids of Magnolia arranged by reported parentage ploidy levels.
Evidence for successful hybridization between plants of different ploidy levels was apparent based on analysis of genome sizes. In many cases, interploid hybrids were substantiated. These include the following within subgenus Magnolia: [M. grandiflora (6x) × M. virginiana (2x)] ‘Maryland’ (MGA 077, McCracken) with an intermediate genome size of 7.49 pg, and also a seedling of ‘Maryland’ (MGA 325), which was likely open-pollinated by M. grandiflora that had a genome size of 9.00 pg, consistent with a pentaploid derived from a tetraploid by hexaploid cross. An unnamed plant at the U.S. National Arboretum (USNA 2) with morphological similarity to Magnolia ‘Maryland’ was found to have a genome size of 5.62 pg, consistent with a triploid, suggesting a M. grandiflora (6x) × M. virginiana (2x) backcrossed to M. virginiana. An intermediate tetraploid condition was determined for M. insignis (2x) × M. grandiflora ‘Kay Parris’ (6x) (NCSU H2010-026-001), which had an 8.50 pg relative genome size.
Within subgenus Yulania, confirmed interploid hybrids were numerous. Verification of hybridity was readily confirmed for the USNA's Kosar/de Vos hybrids. M. liliiflora (4x) × M. stellata (2x) had genome sizes ranging from 6.28 to 6.69 pg, consistent with triploids. Numerous putative pentaploid hybrid cultivars, derived from crosses of (6x × 4x) species or hybrids, were also verified. These hybrid cultivars include: Alexandrina, Angelica, Apollo, Blushing Belle, Butterflies, Elizabeth, Galaxy, Gold Finch, and Spectrum with 2C genome sizes ranging from 10.11 to 11.02 pg.
Hybrids arising from parents with odd ploidy levels (5x or aneuploids) were prevalent and had highly variable genome sizes. Magnolia ×soulangeana, a pentaploid hybrid between M. denudata (6x) and M. liliiflora (4x), exhibits fertility in initial F1 hybrids and subsequent generations (McDaniel, 1968) and when used as parents gave rise to apparent aneuploid progeny ranging from ≈4.6 to ≈8.5x based on genome size. Fertility among M. ×soulangeana cultivars has been examined previously and it was found that pollen viability generally increased with increasing ploidy level above 5x (Santamour, 1970). Relative 2C genome sizes determined here support cytological findings by Santamour (1970) that the cultivars Lennei and Grace McDade are septaploids or higher. Other taxa in Table 3 of approximate septaploid genome size include Magnolia ‘Andre Leroy’ (Milliken), Magnolia ‘Manchu Fan’ (Bartlett 2003-593), Magnolia ‘Sunsation’ (SCC), and Magnolia ‘Todd Gresham’ (Bartlett 2002-641). Each of these hybrids has a parental combination that theoretically could yield 7x offspring. No triploid hybrids were found to be parents of any hybrid surveyed in this study indicating triploids may typically not be fertile.
In a number of cases, interploid hybridization was not validated. Two accessions of Magnolia ‘Sweet Summer’ [11.53 pg (McCracken, MGA 327)], a reported M. virginiana (2x) × M. grandiflora (6x) hybrid, and Magnolia ‘Monland’ [11.29 pg (SCBG)], a reported M. grandiflora (6x) ×virginiana (2x) hybrid (Langford, 1994), both had genome sizes consistent with a subgenus Magnolia hexaploid.
Unreduced gametes can lead to higher than expected genome sizes or ploidy levels in Magnolia hybrids (McDaniel, 1968; Santamour, 1970). In subgenus Yulania, the relative genome size of M. acuminata (4x) × M. stellata (2x) ‘Gold Star’ (NCSU 2004-063) was determined to be 8.22 pg, consistent with the genome size of a tetraploid. This suggests this cultivar is the result of pollination from an unintended source or the product of an unreduced gamete from M. stellata. The hybrids ‘Miranja’ and ‘Sunsation’ may also have resulted from stray pollination or unreduced gametes from at least one parent.
Determination of relative genome size and ploidy level among artificially induced polyploids.
Attempts to develop artificially induced polyploids of Magnolia have met with varying degrees of success. M. stellata and M. cylindrica seedlings treated with colchicine at the Holden Arboretum (C. Tubesing, personal communication) were determined to be tetraploid and octoploid, respectively (Table 4). Magnolia kobus ‘Norman Gould’ [7.79 pg (USNA 59598-H)] was also confirmed to be tetraploid. Additonally, a M. grandiflora ‘Little Gem’ treated with colchicine at Head-Lee Nursery (R. Head, personal communication) was determined to be a 6x - 12x cytochimera. The plant was reported to be treated over 10 years ago and has stabilized as a cytochimera with ≈55% of the leaf tissue comprised of 12x cells. Phenotypic characteristics such as thickened foliage and increased width to length ratio of foliage (Kehr, 1985) were suggestive of polyploidy in M. sieboldii ‘Colossus’, a reported hexaploid. However, samples of M. sieboldii ‘Colossus’ from multiple sources had genome sizes (2C = 4.35 pg to 4.62 pg) consistent with a diploid. Hybrids with Magnolia ‘Colossus’, including M. sieboldii ‘Colossus’ × M. grandiflora ‘Bracken's Brown Beauty’ (McCracken), M. sieboldii ‘Colossus’ × M. grandiflora ‘Kay Parris’ (KP 2008-001), and M. sieboldii ‘Colossus’× Magnolia ‘Sweet Summer’ (MGA 280) (Table 3), all had relative genome sizes consistent with a tetraploid, further confirming the ploidy level of the diploid and hexaploid parents. Other reported induced polyploids that were not confirmed include M. stellata ‘Two Stones’ and M. acuminata ‘Patriot’. Seedlings SCC-2009-004 and SCC-2009-005, derived from open-pollinated octoploid M. cylindrica at the Holden Arboretum, were determined to be ≈7x based on genome sizes of 14.92 to 15.21 pg. This supports the assertion of Charles Tubesing (personal communication) that the octoploids probably outcrossed with other magnolias with lower ploidy levels from their collections. A chromosome count of one of these seedlings, SCC 2009-004, identified ≈133 chromosomes (Fig. 1), in close agreement with genome size data.
Relative genome sizes and estimated ploidy levels of artificially induced polyploid Magnolia species.
Photomicrograph of a root tip cell of Magnolia SCC 2009-004 in early metaphase with ≈133 chromosomes. Maternal parent Magnolia cylindrica (2n = 8x = 152), paternal parent unknown, but likely (2n = 6x = 114), resulting in a plant that is 7x.
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 135, 6; 10.21273/JASHS.135.6.533
Photomicrograph of a root tip cell of Magnolia SCC 2009-004 in early metaphase with ≈133 chromosomes. Maternal parent Magnolia cylindrica (2n = 8x = 152), paternal parent unknown, but likely (2n = 6x = 114), resulting in a plant that is 7x.
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 135, 6; 10.21273/JASHS.135.6.533
Photomicrograph of a root tip cell of Magnolia SCC 2009-004 in early metaphase with ≈133 chromosomes. Maternal parent Magnolia cylindrica (2n = 8x = 152), paternal parent unknown, but likely (2n = 6x = 114), resulting in a plant that is 7x.
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 135, 6; 10.21273/JASHS.135.6.533
Comparison of fluorochromes and estimate of base pair composition.
Comparison of DAPI and PI stains showed there was a significant interaction between fluorochrome stain and species on the estimation of genome size (P ≤ 0.05) (Table 5). For some species (e.g., M. sinica, M. stellata ‘Royal Star’, and M. yuyuanensis), there was no significant difference in genome size estimates between fluorochromes. In other cases, the difference in genome size estimates varied by as much as 0.73 pg or 14% for M. delavayi. This suggests that as bp composition of the sample deviates from the bp composition of the internal standard (in this case P. sativum = 61.50% AT), the estimate of genome sizes between methods diverges. However, for the purpose of determining euploid levels, either method was sufficiently accurate to provide proper classification and the DAPI procedure is faster, less expensive, uses less toxic compounds, and can have lower cv for mean nuclei fluorescence than the PI procedure. Base pair composition of representative Magnolia species ranged from 61.6% to 63.9% AT. Sequences of 8500 bases of cpDNA from seven different regions of 43 different species of Magnolia showed the relative frequency of AT ranging from 62.9% to 63.1% (H. Azuma, personal communication), similar to the range that we determined for the entire nuclear genome based on differential fluorochrome staining.
Comparison of differential staining of fluorochromes and DNA base pair content for selected species from 10 sections of Magnolia.
Implications of relative genome size for systematics and breeding.
The most recent taxonomic revision of Magnolia (Figlar and Nooteboom, 2004) incorporates both morphological and molecular data (Azuma et al., 1999, 2000, 2001; Kim et al., 2001). In some cases, data on relative genome size support these revised taxonomic groupings. For example, establishment of section Macrophylla to include only M. macrophylla and botanical varieties ashei and dealbata is supported by the difference in 1Cx value (Table 2) of this group compared with other North American species (M. fraseri and M. tripetala) with which it was traditionally grouped (Treseder, 1978). However, in other cases, there is inconsistent variation in genome size within some sections (e.g., M. rostrata in section Rhytidospermum) and similarities in genome size among distantly related taxa (Table 2).
For breeders, the revised taxonomy by Figlar and Nooteboom (2004) provides a greater understanding of the relatedness and potential for interspecific hybridizations among closely allied species that is often supported empirically (Table 3). Yet, development of progeny from hybrids, beyond an F1 generation, requires genome/chromosomal compatibility for meiosis to function properly. Thus, it is reasonable to expect that the greater the difference in genome size among parental species, the less likely hybrid progeny will be fertile.
Results from this study provide data on genome sizes and ploidy levels of a broad range of species and hybrids of Magnolia. This information also gives insight into reproductive biology, confirmation of hybrids and induced polyploids, and comparison of methods for determining genome size that will help facilitate the development of improved hybrids in the future.
Literature Cited
Adams, K.L. & Wendel, J.F. 2005 Novel patterns of gene expression in polyploidy plants Trends Genet. 21 539 543
Azuma, H., García-Franco, J.G., Rico-Gray, V. & Thien, L.B. 2001 Molecular phylogeny of the Magnoliaceae: The biogeography of tropical and temperate disjunctions Amer. J. Bot. 88 2275 2285
Azuma, H., Thien, L.B. & Kawano, S. 1999 Molecular phylogeny of Magnolia (Magnoliaceae) inferred from cpDNA sequences and evolutionary divergence of floral scents J. Plant Res. 112 291 306
Azuma, H., Thien, L.B. & Kawano, S. 2000 Molecular phylogeny of Magnolia based on chloroplast DNA sequence data and floral scent chemistry Proc. Intl. Symp. Family Magnoliaceae. 219 227
Callaway, D.J. 1994 The world of magnolias Timber Press Portland, OR
Chen, Z., Huang, X., Wang, R. & Chen, S. 2000 Chromosome data of Magnoliaceae Proc. Intl. Symp. Family Magnoliaceae. 192 201
Chen, Z.J. & Ni, Z. 2006 Mechanisms of genomic rearrangements and gene expression changes in plant polyploids Bioessays 28 240 252
Comai, L. 2005 The advantages and disadvantages of being polyploidy Nat. Rev. Genet. 6 836 864
Doležel, J., Greihuber, J. & Suda, J. 2007 Flow cytometry with plant cells: Analysis of genes, chromosomes and genomes Wiley-VCH Weinheim, Germany
Figlar, R.B. & Nooteboom, H.P. 2004 Notes on Magnoliaceae IV Blumea 49 1 14
Godelle, B., Cartier, D., Marie, D., Brown, S.C. & Siljak-Yakovlev, S. 1993 Heterochromatin study demonstrating the non-linearity of fluorometry useful for calculating genomic base composition Cytometry 14 618 626
Greilhuber, J., Temsch, E.M. & Loureiro, J.C.M. 2007 Nuclear DNA content measurement 67 101 Doležel J., Greilhuber J. & Suda J. Flow cytometry with plant cells: Analysis of genes, chromosomes and genomes Wiley-VCH Weinheim, Germany
Hegarty, M.J. & Hiscock, S.J. 2008 Genomic clues to the evolutionary success of polyploidy plants Curr. Biol. 18 R435 R444
Jones, J.R. & Ranney, T.G. 2009 Fertility of neopolyploid Rhododendron and occurrence of unreduced gametes in triploid cultivars J. Amer. Rhododendron Soc. 63 131 135
Jones, J.R., Ranney, T.G., Lynch, N.P. & Krebs, S.L. 2007 Ploidy levels and genome sizes of diverse species, hybrids, and cultivars of Rhododendron L J. Amer. Rhododendron Soc. 61 220 227
Kehr, A.E. 1985 Inducing polyploidy in magnolias J. Amer. Magnolia Soc. 20 6 9
Kim, S., Park, C., Kim, Y. & Suh, Y. 2001 Phylogenetic relationships in family Magnoliaceae inferred from NDHF sequences Amer. J. Bot. 88 717 728
Langford L. 1994 Checklist of cultivated magnolia, revised edition J. Amer. Magnolia Soc. 6178 3 31
McDaniel, J.C. 1968 Magnolia hybrids and selections Proc. Central States For. Tree Improvement Conf. 6 6 12
Meister, A. & Barrow, M. 2007 DNA base composition of plant genomes 177 185 Doležel J., Greilhuber J. & Suda J. Flow cytometry with plant cells: Analysis of genes, chromosomes and genomes Wiley-VCH Weinheim, Germany
Ranney, T.G. 2006 Polyploidy: From evolution to new plant development Proc. Intl. Plant Prop. Soc. 56 137 142
Santamour F.S. Jr 1970 Cytology of magnolia hybrids II. M. ×soulangiana hybrids Morris Arboretum Bul. 21 58 61
Soltis, D.E. & Burleigh, J.G. 2009 Surviving the K-T mass extinction: New perspectives of polyploidization in angiosperms Proc. Natl. Acad. Sci. USA 106 5455 5456
Soltis, D.E., Soltis, P.S., Bennett, M.D. & Leitch, I.J. 2003 Evolution of genome size in the angiosperms Amer. J. Bot. 90 1596 1603
Soltis, D.E., Soltis, P.S. & Tate, J.A. 2004 Advances in the study of polyploidy since plant speciation New Phytol. 161 173 191
Treseder, N.G. 1978 Magnolias Faber and Faber Boston, MA
Xia, N., Liu, Y.H. & Nooteboom, H.P. 2008 Magnoliaceae Flora of China 7 73 75
