Volume-based pollen size analysis: an advanced method to assess somatic and gametophytic ploidy in flowering plants

TECHNICAL ADVANCE

Volume-based pollen size analysis: an advanced method to assesssomatic and gametophytic ploidy in flowering plants

Nico De Storme • Linda Zamariola •

Martin Mau • Timothy F. Sharbel •

Danny Geelen

Received: 19 October 2012 / Accepted: 31 December 2012

� Springer-Verlag Berlin Heidelberg 2013

Abstract Pollen size is often used as a biological

parameter to estimate the ploidy and viability of mature

pollen grains. In general, pollen size quantification is per-

formed one- or two-dimensionally using image-based

diameter measurements. As these approaches are elaborate

and time consuming, alternative approaches that enable a

quick, reliable analysis of pollen size are highly relevant

for plant research. In this study, we present the volume-

based particle size analysis technique as an alternative

method to characterize mature pollen. Based on a com-

parative assay using different plant species (including

tomato, oilseed rape, kiwifruit, clover, among others), we

found that volume-based pollen size measurements are not

biased by the pollen shape or position and substantially

reduce non-biological variation, allowing a more accurate

determination of the actual pollen size. As such, volume-

based particle size techniques have a strong discriminative

power in detecting pollen size differences caused by

alterations in the gametophytic ploidy level and therefore

allow for a quick and reliable estimation of the somatic

ploidy level. Based on observations in Arabidopsis thali-

ana gametophytic mutants and differentially reproducing

Boechera polyantha lines, we additionally found that vol-

ume-based pollen size analysis provides quantitative and

qualitative data about alterations in male sporogenesis,

including aneuploid and diploid gamete formation.

Volume-based pollen size analysis therefore not only pro-

vides a quick and easy methodology to determine the

somatic ploidy level of flowering plants, but can also be

used to determine the mode of reproduction and to quantify

the level of diplogamete formation.

Keywords Pollen size � Volume-based particle size

analysis � Ploidy determination � 2n Gametes �Microsporogenesis � Arabidopsis thaliana � Boechera

polyantha

Introduction

Due to high prevalence of polyploidy, the development of

reliable and simple techniques that allow for a rapid

determination of the somatic or gametophytic ploidy level

is of great importance for several aspects of plant research.

Indeed, in plants, variations in somatic ploidy level are not

only observed in natural communities (Petit et al. 1999), but

also appear under artificial conditions (Cheng and Korban

2011) or in mutant populations (d’Erfurth et al. 2009).

Besides differences in somatic ploidy, reproductive organs

may also display variations in their gametophytic ploidy

level. In normal sexual reproduction, meiosis typically

halves the somatic diploid chromosome number and gen-

erates haploid gametes. However, in plants with an altered

reproductive system (e.g., apomixis), aneuploid lines and

meiotic or gametophytic mutants, meiosis produces spores

Communicated by Scott Russell.

Electronic supplementary material The online version of thisarticle (doi:10.1007/s00497-012-0209-0) contains supplementarymaterial, which is available to authorized users.

N. De Storme � L. Zamariola � D. Geelen (&)

Department of Plant Production, Faculty of Bioscience

Engineering, University of Ghent, Coupure Links 653,

9000 Ghent, Belgium

e-mail: [email protected]

M. Mau � T. F. Sharbel

Leibniz Institut fur Pflanzengenetik und

Kulturpflanzenforschung, Corrensstraße 3, 06466 Stadt Seeland,

OT Gatersleben, Germany

123

Plant Reprod

DOI 10.1007/s00497-012-0209-0

with a variable genomic DNA content (Kantama et al. 2007;

Liu and Qu 2008; Henry et al. 2005; Consiglio et al. 2004).

Moreover, under certain conditions (e.g., specific meiotic

mutants or environmental stress), meiosis is converted into

a mitotic-like non-reductional cell division and conse-

quently generates gametes that contain the somatic chro-

mosome number (d’Erfurth et al. 2009, 2010; De Storme

and Geelen 2011; Pecrix et al. 2011). These so-called

unreduced or 2n gametes have an important evolutionary

and agricultural relevance (Brownfield and Kohler 2011;

Chan 2010; Ramanna and Jacobsen 2003).

Several techniques have been developed that allow for the

direct determination of the somatic and/or gametophytic

ploidy level in plants, including DNA flow cytometry

(Dolezel and Bartos 2005; Johnston et al. 1999), DNA

photocytometry (Hardie et al. 2002; Vilhar et al. 2001;

Praca-Fontes et al. 2011), and chromosome cytology

(Ahloowal 1965; Gamage and Schmidt 2009). However,

despite the high accuracy in determining the plant’s ploidy

level, all these methods require elaborate, expensive, and

time-consuming sample preparation, making them less

useful for high throughput analyses. A common alternative

practice is to estimate the plant’s ploidy level by means of

morphological features, such as stomata density (Tan and

Dunn 1973; Beck et al. 2003; Przywara et al. 1988), trichome

branching (Hulskamp 2004), and pollen size (Altmann et al.

1994; Zlesak 2009; Katsiotis and Forsberg 1995; Bamberg

and Hanneman 1991). Out of these three biological param-

eters, only pollen size provides additional data on the male

gametophytic ploidy distribution and the associated stability

of the reproductive program (Zlesak 2009; Bretagnolle and

Thompson 1995). In plant research, pollen size is therefore

often considered an attractive parameter to monitor the mode

of male reproduction and to assess the fertility and/or ploidy

of pollen at anthesis (Kelly et al. 2002).

Pollen size measurements are generally performed one-

or two-dimensionally by means of image-based approa-

ches, in which either the diameter or the transsectional

particle area is registered (Zlesak 2009; Altmann et al.

1994; Jones and Reed 2007). This methodology has three

major disadvantages. At first, the associated staining pro-

cedure and slide preparation protocol often change the

morphometric features of the pollen grains and alters the

size and morphology of the mature pollen, leading to

aberrations in size determination. Secondly, as pollen of

many plant species show a specific, non-spherical body

shape with the additional presence of pores and exine

ornaments (Shaheen et al. 2009), one- and two-dimensional

pollen size measurements generally give a false and/or

simplified representation of the real pollen grain size. And

thirdly, image-based size approaches often show inaccu-

racies in particle size assessment, mostly caused by tech-

nical issues such as out-of-focus or incorrect image

analysis settings (e.g., inclusion of debris) (Keyvani and

Strom 2013).

In the search for an alternative method enabling easy

and accurate pollen size analysis in a high throughput

manner, we assessed the applicability of a three-dimen-

sional volume-based particle size analysis technique; for

example, the Multisizer III Coulter Counter. In this

approach, particles of interest are suspended in a weak,

isotonic electrolyte (Isoton II) and guided through an

aperture separating two electrodes between which an

electric current is set. When a particle passes through the

aperture, it causes a temporal increase in impedance, which

is registered as a voltage pulse (Shekunov et al. 2007). The

amplitude of this pulse is proportional to the particle’s

body volume (Coulter principle) and thus allows for a

volume-based representation of the particles’ diameter

(McTainsh et al. 1997; Miller and Lines 1988). Based on

its high accuracy and high throughput measurement

capacity, volume-based size analysis is already widely used

in many research fields for the characterization of both

organic and non-organic particles (Avdeef et al. 2009;

Walker et al. 1974; Hirsch and Gallian 1968; Walstra and

Oortwijn 1969). However, in pollen biology, this method

has not been adopted yet.

In this study, we report the use of volume-based particle

size analysis as an alternative technique to characterize

mature pollen grains. Using different plant species, we test

whether volume-based pollen size is an appropriate

parameter to determine the plant’s pollen size distribution

and additionally monitor if volume-based pollen size

analysis can be used to assess the somatic ploidy level.

Moreover, based on observations in gametophytic mutants

and apomictically reproducing plant species, we monitor to

which extent this technique can be used to characterize

alterations in male gamete formation (e.g., 2n pollen) and

to determine the mode of reproduction in flowering plants.

Materials and methods

Plant materials

Diploid and tetraploid Arabidopsis Colombia-0 (Col-0)

wild type accessions were obtained from the Nottingham

Arabidopsis Stock Centre (NASC). Octaploid plants were

generated by somatic genome duplication using colchicine

(De Storme and Geelen 2011), and hexaploid lines were

selected out of the progeny of unstable octaploid lines.

Triploid lines were generated by performing reciprocal

crosses between diploid and tetraploid control lines. The

Arabidopsis mutants atspo11-1-3 (SALK_146172; Col-0),

msh5-1 (SALK_N610240; Col-0), tes-4 (CS9353, Ws-2),

and qrt1-1 (CS8845; Ler-0) were obtained from the

Plant Reprod

123

European Arabidopsis Stock Centre. Arabidopsis seed

germination and plant cultivation were performed accord-

ing to De Storme and Geelen (2011). Tomato control plants

(TR 5-306 background) were kindly provided by Rob Dirks

(Rijk Zwaan Breeding N. V.), and Brassica plants (N90)

were obtained from Bayer CropScience (Ghent). Both

tomato and Brassica plants were grown in a climate

chamber under a 16 h/8 h light/dark rhythm at a tempera-

ture of 21 �C. Dry kiwifruit pollen samples were kindly

provided by Filip Debersaques (University College,

Ghent). All other pollen donor plants, for example, white

clover, poppy, and lesser bindweed, were endogenous

species found in the wild.

Pollen size analysis

Extraction of pollen was performed by putting mature

flowers in Isoton II solution at gentle agitation. After

removing the flower material, the resulting pollen suspen-

sion was centrifuged for 1 min at 13,0009g to isolate the

pollen fraction in the pellet. Finally, this pellet was resus-

pended in 20 ll of isoton II solution. For one- and two-

dimensional image-based pollen size analyses, a small

amount (8–12 ll) of pollen suspension was put on a

microscopy slide and covered with a cover slip.

Analysis of the pollen diameter (random or minor/major

axis length) and transsectional surface area was either

performed manually using the Olympus Cell M measure-

ment toolbar or digitally using a MATLAB-based image

analysis software program (supplemented file). For manual

and software-based assessments, a minimum amount of 50

and 500 pollen grains was monitored, respectively. Anal-

ysis of pollen morphology (e.g., circular or ellipsoid) was

assessed by calculating eccentricity values. The eccentric-

ity of a particle is a value between 0 and 1 that states how

much a conic section deviates from being circular, with 0

representing circular objects and higher values corre-

sponding to more ellipsoid-shaped objects. Eccentricity

values were calculated based on the minor (Min) and major

(Maj) axis length value using the following formula:

Ecc = [1 - Min2/Maj2)]1/2.

For volume-based, three-dimensional pollen size anal-

yses, the pollen suspension was transferred into 10 ml of

isoton II solution and automatically analyzed using the

Multisizer II Coulter Counter (Beckman Coulter). In this

approach, at least 1,000 pollen grains were analyzed per

sample. Graphical output was enhanced by binning the

histogram data in groups of three.

Cytology

Nuclear DNA staining of mature pollen grains was performed

based on the protocol described by Durbarry et al. (2005)

with minor modifications described in De Storme and Geelen

(2011). Analysis of meiotic outcome was performed by

selecting flower buds based on size and shape and squashing

them on a slide in a drop of 4.5 % (w/v) lactopropionic orcein

staining solution. Visualization of meiotic chromosome

behavior in male meiocytes was accomplished using the

spreading technique described by Jones and Heslop-Harrison

(1996) with some minor modifications (De Storme and Geelen

2011).

DNA flow cytometry

Flow cytometric ploidy analysis of somatic tissues in

Arabidopsis, tomato, and Brassica was performed based on

the nuclei extraction method of Galbraith et al. (1983).

Fresh green leaf material was chopped with a sharp razor

blade in 200 ml of Galbraith’s buffer, and the resulting

nuclei suspension was filtered through a 40-mm nylon

mesh. After staining the isolated nuclei with propidium

iodide (final concentration of 10 mM), the corresponding

DNA content was analyzed using flow cytometry (Epics

Altra, Beckman; excitation, 488 nm; signal detection,

575 nm).

Quantification of relative nuclear DNA content in Bo-

echera polyantha pollen was performed as described by

Matzk et al. (2000) using pollen from a diploid, sexual

B. stricta (ES 558.2) as an external control. Pollen grains

from at least 20 mature open flower buds at dehiscence

were harvested by shaking gently for 15 min in 500 ll

distilled water, followed by centrifugation at 13,200 rpm

for 6 min. After decanting the supernatant, the pollen

pellets were resuspended in 100-ll Galbraith’s buffer and

ground with two 6-mm steel balls in 2-ml eppendorf tubes

using a Retsch mixer mill MM 400 for 20 s at 30 Hz.

Extracted nuclei were suspended in a total volume of 1-ml

Galbraith’s buffer. Pollen nuclei were filtered through a

10-lm nylon net filter (Millipore) into a measurement vial.

Tissue-specific nuclear DNA content measurement was

performed on a FACStar (Beckton Dickinson, http://www.

bd.com/) equipped with a DAPI 100 micron Near UV laser.

Approximately 2,000 nuclei per run were collected and

plotted on a linear scale using the FACSDiva Software

from Beckton Dickinson. Data analysis was carried out

using WinMDI v2.9 (WinMDI, The Scripps Research

Institute, http://facs.scripps.edu/software.html).

In vitro regeneration of Brassica napus

Brassica plants were regenerated in vitro following the

protocol described by Deblock et al. (1989) with modifi-

cations. In brief, Brassica napus seeds were sterilized and

sown on K1 medium under the following conditions:

photoperiod of 16 h day/8 h night and temperature of

Plant Reprod

123

20 �C. After 7 days, hypocotyls were cut into 1 cm seg-

ments, and explants were cultured on medium M1 con-

taining MS basal medium with 0.5 g l-1 MES, 0.5 g l-1

PVP10, 40 mg l-1 adenine, 5 mg l-1 silver nitrate, and

2 % (w/v) sucrose, supplemented with hormones (1 mg l-1

BAP, 0.5 mg l-1 NAA). The explants were transferred to

new medium every 2 weeks. After 3–6 weeks, the formed

calli were transferred to M1 medium with 1 mg l-1 BAP,

0.25 mg l-1 NAA, and 1 mg l-1 2,4-D for 1 week and

then subcultured every 2 weeks on M1 with 1 mg l-1 BAP

and 0.1 mg l-1 TIBA until they formed shoots. Small

shoots were transferred to M1 medium (without silver

nitrate) containing 2.5 lg l-1 BAP until they developed

normal looking leaves. Normal looking leaves were

transferred to M2 medium containing 1/2 MS supple-

mented with 0.5 g l-1 MES and 1.5 % (w/v) sucrose

without hormones. Rooted shoots were transferred to

controlled climate chambers (16 h day/8 h night, 20 �C).

Results

Volume-based pollen size analysis reveals dynamic

pollen size alterations upon suspension

For determining pollen grain size using the Multisizer III

Coulter Counter, pollen is suspended in an electrolyte

solution. In general, mature plant spores start to swell when

dissolved in an aqueous solution (Edlund et al. 2004). To

monitor the impact of the suspension protocol on the size

dynamics of mature pollen, we suspended spores from

different plant species in Isoton II and performed a time-

lapse volume-based pollen size analysis.

Strikingly, for all species tested (e.g., kiwifruit, tomato,

lesser bindweed, and oilseed rape), suspended pollen

showed a varying body size upon time, generally display-

ing a progressive shrinkage during the first hour after

which a steady-state body size is established (Fig. 1).

22

23

24

25

0 30 60 90 120 150

kiwifruit

tomato

80

85

90

95

0 30 60 90 120 150

Lesser bendweed

28

29

30

31

32

0 30 60 90 120 150

Mea

n vo

l. po

llen

suspension time (h)

25.23 - 36.39 µm

18.70 - 36.39 µm

a

b

c

d

e

f

dia

met

er (

µm

)M

ean

vol.

polle

n d

iam

eter

(µ

m)

Mea

n vo

l. po

llen

dia

met

er (

µm

)

Fig. 1 Time-lapse series of volume-based diameter measurements of

tomato and kiwifruit pollen (a), lesser bindweed pollen (b) and

oilseed rape spores (c) upon suspension in isoton II; volumetric pollen

size diameter distributions of pollen samples harvested from tomato

(d), lesser bindweed (e) and oilseed rape (f) at specific time points

upon isoton II suspension

Plant Reprod

123

For example, in kiwifruit (Actinidia deliciosa), the initial

spore diameter of 24.27 lm quickly decreased to 23.60 lm

in the first hour and then slowly decreased to an equilibrium

of 23.27 lm after 2.5 h of suspension (Fig. 1a). Similarly,

for bindweed spores, the initial diameter of 92.31 lm stea-

dily decreased and reached a size of ±82.5 lm after

100 min of suspension (Fig. 1b, e). For the Brassica, the

average pollen diameter also decreased during the first

30 min of suspension, reaching a final value of ±28.7 lm

(Fig. 1c). However, in contrast to the other pollen samples,

this decrease was not due to a progressive size reduction of

all pollen grains but instead was caused by the accumulation

of shriveled spores (20–25 lm) (Fig. 1f).

From these observations, we conclude that both pollen

type (e.g., plant species) and suspension duration have a

strong impact on pollen size, with spores either showing a

gradual decrease in their body volume or accumulating

smaller-sized, shriveled spores upon Isoton II suspension.

Hence, in further experiments, all pollen size analyses were

performed immediately following suspension.

Volumetric particle size analysis enables a quick

and accurate characterization of pollen size

To check the accuracy of the Multisizer III particle ana-

lyzer in determining the size of mature pollen, we moni-

tored pollen size in a set of flowering plants and compared

the volume-based size distribution with image-based

diameter assessments. For all species analyzed, the vol-

ume-based pollen size assay resulted in a Gaussian distri-

bution with a mean diameter closely corresponding to the

diameter obtained by image-based measurements (Fig. 2,

Table 1 and supplemented Fig. S1). However, based on a

more in-depth comparison of both approaches (using the

diameter ratio: DMi/DMu), distinct variations in mean pol-

len diameter were observed. Image-based assessment

hereby resulted either in an under—(DMi/DMu \ 1) or an

overestimation (DMi/DMu [ 1) of the pollen volume

(Table 1). Since these biases were typically observed in

extreme spherical and ellipsoid-shaped spore types, we

hypothesized that one-dimensional image-based measure-

ments are strongly biased by the pollen’s shape.

To test the impact of pollen morphology on the accuracy of

image-based measurements, we assessed mean spore eccen-

tricity and found a clear correlation with the image-based

diameter bias, as represented by the DMi/DMu microscopic

accuracy parameter (Table 1; Fig. 3). For spores with a high

eccentricity (ellipsoid), the image-based diameter value was

consistently higher than the volume-derived one (e.g. Brassica

napus; Ecc. = 0.51, DMi/DMu = 1.11), whereas for spores

with a low eccentricity (spherical), the one-dimensionally

derived diameter value was substantially lower compared to

the volumetric-derived one (e.g. Lycopersicon esculentum;

Ecc. = 0.22, DMi/DMu = 0.91). Hence, image-based diame-

ter measurements typically over- or underestimate the actual

pollen size, largely depending on the shape and eccentricity of

the pollen grain.

In this experiment, we also observed that the variability

in pollen size using one-dimensional parameters (e.g.,

diameter) is generally higher compared to the variability in

volume-based analyses. This increase in variability is

particularly relevant for strong ellipsoid-shaped particles

(Table 1) and is less conspicuous for more spherical par-

ticles. Volume-based size measurements therefore not only

provide a more accurate representation of the pollen size

distribution, but also substantially reduce non-biological

size variability.

Volume-based pollen size strongly correlates

to the somatic ploidy level in Arabidopsis

Pollen size is often used to determine the plant’s somatic

ploidy level. To assess the accuracy of volume-based

pollen size analysis approach in assessing somatic ploidy,

mature pollen from diploid, tetraploid, and octaploid Ara-

bidopsis plants was analyzed using the Multisizer III

Coulter Counter. For all three ploidy levels, the volume-

derived pollen diameter showed a Gaussian distribution

with a mean diameter of, respectively, 20.31, 25.16, and

31.67 lm (Fig. 4a, b). Although these diameter values

largely corresponded to the image-based values, the mean

major and minor axis length and area-derived diameter

value were substantially higher (Table 2), confirming the

notion that image-based diameter measurements overesti-

mate the actual pollen size in Arabidopsis. More impor-

tantly, in contrast to one- and two-dimensional analyses,

volume-derived diameter assessments strongly reduced

size variability for all three pollen types (Table 2) and

consequently diminished pollen size overlap between the

different ploidy levels (supplemented Fig. S2). Indeed,

both the major and minor axis length and the area-derived

diameter exhibited a higher variability compared to the

volume-based diameter for all ploidy levels, indicating that

volume-derived size analysis substantially reduces non-

biological size variability compared to one- and two-

dimensional approaches. Hence, volume-based pollen size

analysis is the most accurate pollen-based method to

characterize Arabidopsis plants that differ in their somatic

ploidy level.

In contrast to plants with an even ploidy level (29,

49 and 89), which typically produce uniformly sized pol-

len, triploid plants exhibited a more variable pollen size

distribution with a subset of small spores (\20 lm) and a

large subpopulation of larger pollen grains (20–30 lm),

having a mean diameter of 23.21 ± 2.04 lm (39: 29 * 49)

and 23.93 ± 2.19 lm (39: 49 * 29), respectively (Fig. 4c

Plant Reprod

123

and supplemented Table S1). No differences in size distri-

bution were observed between the two reciprocal triploid

types tested. Microscopic analysis revealed that the small

particles correspond to shriveled, malformed spores,

whereas the larger particles represent swollen, viable pollen.

Remarkably, in the swollen pollen grain population, diam-

eter variability ranged from ±20 to ±30 lm and was

therefore much higher compared to plants with an even

ploidy level (supplemented Table S1). This large size vari-

ability most presumably reflects the imbalanced meiotic

Arabidopsis thaliana

Lycopersicon esculentum

Brassica napus

Trifolium repens

20.61 µm

25.16 µm

29.06 µm

39.35 µm

30.57 µm

37.13 µm

30.49 µm31.82 µm

29.23 µm

31.86 µm

19.96 µm

23.15 µm

23.24 µm

23.00 µm

24.56 µm

25.73 µm

a

b

c

d

Fig. 2 Volume-based diameter distribution and representative images of mature pollen harvested from Arabidopsis (a), tomato (b), oilseed rape

(c), and white clover (d); scale bar, 25 lm

Plant Reprod

123

chromosome segregation in triploid meiosis, which typically

leads to spores that contain a variable number of chromo-

somes (supplemented Fig. S3).

In contrast to the variably sized pollen in Arabidopsis

triploids, pollen isolated from hexaploids appeared more

uniform and resulted in a Gaussian volume-derived size dis-

tribution with a mean diameter of 28.92 lm and a variability

of 1.81 lm (Fig. 4c). The uniformity in pollen size reflects the

more balanced chromosome segregation in hexaploid meio-

sis, as in other euploid Arabidopsis meiocytes.

To establish the correlation between pollen volume and

somatic ploidy level in Arabidopsis, we next plotted the

volume-based pollen diameter in function of the somatic

DNA content for all ploidy levels analyzed. As such, we

found that there is a strong linear correlation (R2 = 0.98)

between the somatic ploidy level and the volume-based

pollen diameter (Fig. 5 and supplemented Table S1),

indicating that volume-based pollen size analysis provides

an accurate and straightforward methodology to determine

the somatic ploidy level in Arabidopsis.

Volume-based pollen size analysis allows quick

identification of somatic polyploidy in plants

To check whether volume-based pollen size analysis can be

used as a broad methodology to detect somatic polyploi-

dization events in flower plants, we artificially induced

polyploidy in two plant species (e.g., Lycopersicon escu-

lentum and Brassica napus) and assessed the Multisizer

Coulter Counter technique for its ability to detect polyploid

individuals.

In the tomato experiment, young seven-day-old seed-

lings were treated with 0.2 % (w/v) colchicine (3 h sub-

mersion) to induce somatic polyploidization. Out of 50

plants treated, 22 survived and reached the flowering stage.

In this population, two plants were found to produce a

significant number of larger pollen. Indeed, in contrast to

diploid controls, which typically generate one population

of uniformly sized pollen (±24 lm), the two isolated

colchicine lines produced three differentially sized pollen

types; a subset of small spores (\22 lm), pollen with a

slightly larger diameter (±28 lm), and giant pollen

(±34 lm) (Fig. 6a). Microscopic analysis confirmed the

presence of all three pollen types, and nuclear DNA

staining demonstrated that the giant spores contained an

enlarged vegetative and generative nucleus, indicating for

Table 1 Comparative analysis of pollen diameter distribution using one-dimensional image-based assessments and volume-based analysis in a

set of flowering plants

Species Microscopy (one-dimensional) Multisizer (three-dimensional) Diameter ratio

Name n Eccentricity Diameter (lm) Size range (lm) n Diameter (lm) DMi/DMu

Mean SD Mean SD Min Max Mean SD CV (%)

Convolvulus arvensis 168 0.18 0.08 86.93 2.88 83.18 106.60 4,766 93.28 3.92 4.21 0.93

Lycopersicon esculentum 154 0.22 0.09 21.60 0.99 21.26 26.17 3,056 23.75 0.83 3.49 0.91

Papaver rhoeas 142 0.27 0.12 25.72 1.73 23.59 31.49 2,335 27.62 1.64 5.95 0.93

Trifolium repens 166 0.32 0.11 29.27 1.76 27.16 35.06 4,147 31.48 1.46 4.65 0.93

Actinidia deliciosa 164 0.36 0.13 24.26 1.77 20.21 30.04 5,470 24.53 1.71 6.98 0.99

Arabidopsis thaliana 120 0.43 0.13 21.83 1.83 18.22 23.75 4,232 20.72 0.86 4.15 1.05

Brassica napus 130 0.51 0.16 32.71 3.86 26.82 32.54 4,081 29.58 1.01 3.41 1.11

Lilium longiflorum 152 0.63 0.14 84.38 15.01 61.07 109.50 4,204 80.77 9.49 11.70 1.04

The number of pollen analyzed for each species, and each measurement method are listed under (n). Accuracy of image-based measurements

compared to volume-based diameter assessments (e.g. Multisizer III) are represented by the diameter ratio value, giving the bias of the

microscopy-based diameter (DMi) relative to the Multisizer III-based diameter (DMu). Pollen size variability is presented by standard deviation

values (SD) and the coefficient of variation (CV)

C.a.L.e.

P.r. T.r.

A.d.

A.t.

B.n.

L.l.

0,85

0,90

0,95

1,00

1,05

1,10

1,15

0,1 0,2 0,3 0,4 0,5 0,6 0,7

Mic

r. d

iam

./Vol

um. d

iam

.

Eccentricity

Fig. 3 Correlation between the accuracy of image-based pollen

diameter measurements and the eccentricity of the pollen grains; the

accuracy of the microscopic diameter assessment is presented as

the ratio of the microscopic (DMi) to the volume-based diameter

(DMu) with values lower than 1 representing an underestimation

and values higher than 1 representing an overestimation of the real

pollen size; C.a. Convolvulus arvensis, L.e. Lycopersicon esculentum,

P.r. Papaver rhoeas, T.r. Trifolium repens, A.d. Actinidia deliciosa,

A.t. Arabidopsis thaliana, B.n. Brassica napus, and L.l. Liliumlongiflorum

Plant Reprod

123

an increased genomic DNA content (Fig. 6b). In line with

this, ploidy validation using DNA flow cytometry con-

firmed that the two ‘‘larger pollen-producing’’ plants have a

doubled, for example, tetraploid, genomic ploidy level,

whereas all other plants remained diploid (Fig. 6c).

In the second experiment, about 50 Brassica plants were

regenerated in vitro from callus; a process which occa-

sionally induces spontaneous somatic polyploidization. At

flowering, regenerated Brassica plants were monitored for

the production of larger pollen using the Multisizer III. As

such, we isolated one plant that produced a significant

amount of larger spores. In contrast to control plants which

exhibit a bimodal pollen size distribution with a broad peak

at ±22 lm (shriveled spores) and a narrower peak

at ±30 lm, the selected plant showed a bimodal pollen

size distribution with one peak at ±28 lm and one peak

at ±40 lm (Fig. 6a). Microscopic analysis confirmed the

presence of both pollen types and additionally demon-

strated that smaller pollen corresponded to shriveled,

malformed pollen grains, whereas the larger pollen grains

appeared swollen and viable. Nuclear DNA staining

revealed that the larger pollen grains have a normal nuclear

configuration, but additionally showed that gametophytic

nuclei appeared significantly larger, indicating for an

a

b

c

A.t. 2x

A.t. 4x

A.t. 8x

A.t. 3x (4x*2x)

A.t. 6x

20 µm 20 µm 20 µm

Fig. 4 Volume-based diameter distributions (a) and representative images (b) of pollen isolated from diploid, tetraploid, and octaploid

Arabidopsis plants; c Volume-based diameter distributions of pollen isolated from triploid (female overdose) and hexaploid Arabidopsis plants

Plant Reprod

123

increased DNA content (Fig. 6b). To analyze the somatic

ploidy level of the ‘‘larger pollen-producing’’ Brassica

plant, we performed DNA flow cytometry and found that

this plant was octaploid (Fig. 6c), whereas all other plants

remained tetraploid.

Hence, these findings demonstrate that volume-based

pollen size analysis can be used a quick and reliable

method to detect higher ploidy plants in both directed and

non-directed polyploidization experiments.

Characterization of meiotic and gametophytic defects

through volume-based pollen size analysis

Based on the close correlation between size and ploidy of

mature pollen, we hypothesized that volume-based pollen

size can be used as a reliable parameter to determine the

ploidy level of individual pollen grains and to assess the

stability of the male gamete-producing process. To test

this, we monitored the pollen size distribution of a set of

Arabidopsis mutants that are defective in different stages of

male reproduction; for example, the meiotic mutants

spo11-1-3, msh5-1, and tes-4 and the gametophyte-specific

qrt1-1.

Arabidopsis SPO11 and MSH5 are both required for

meiotic double strand break formation and corresponding

mutants show a high level of univalents at the end of

prophase I, typically leading to unbalanced tetrads and

polyads that contain a variable set of differentially sized,

aneuploid spores. Consistent with this, mature spo11-1-3

pollen exhibit a broad volume-based size distribution,

covering a large range of diameter values, reaching

from *12 to *35 lm (Fig. 7b). Microscopic analysis

confirmed this large variability and additionally revealed

that the small pollen (\20 lm) correspond to shriveled,

aborted pollen, whereas larger ones were swollen and

viable. Interestingly, similar to spo11-1-3, the volume-

based pollen size distribution of msh5-1 also displayed a

subset of small pollen grains (e.g., aborted) and a large

population of variably sized spores (Fig. 7c). However,

contrary to spo11-1-3, msh5-1 pollen appeared less vari-

able in size and displayed only a minor subset of small,

aborted pollen grains.

In contrast to alterations in CO formation, which typi-

cally generate aneuploid spores, defects in meiotic cell

division typically lead to triads, dyads, and/or monads that

contain diploid and/or tetraploid spores. Functional loss of

TES, for example, induces a complete loss of meiotic

cytokinesis and consequently generates large syncytial

microspores that contain all four meiotic nuclei. In line with

their increased ploidy, tes-4 pollen appeared substantially

larger, exhibiting a highly uniform volume-based size dis-

tribution with a mean diameter of 31.23 ± 1.60 lm. As this

value closely corresponds to the diameter of tetraploid

pollen (31.67 ± 1.51 lm) (Fig. 7a, d), these data confirm

that volume-based pollen size analysis constitutes a reliable

technique to determine the gametophytic ploidy level of

mature pollen grains.

In Arabidopsis qrt1-1, the four spores in a meiotic tetrad

do not separate and remain attached during the whole

process of spore development (Francis et al. 2006). At

flower anthesis, mature qrt1 spores are typically released in

groups of four (Fig. 7f). Consistent with this, volume-based

size analysis of qrt1-1 pollen resulted in a unimodal size

Table 2 The impact of pollen

size methodology on the mean

diameter and non-biological

size variation of pollen isolated

from diploid, tetraploid, and

octaploid Arabidopsis plants

The number of pollen analyzed

for each ploidy level and each

measurement method are listed

under (n)

Pollen size methodology Pollen diameter (lm)

29 Plant 49 Plant 89 Plant

Mean SD Mean SD Mean SD

Volume-based (3D) 20.31 0.87 25.16 1.34 31.67 1.51

n = 3,859 n = 3,817 n = 952

Minor axis (1D) 21.05 0.94 26.61 2.52 32.08 3.37

Major axis (1D) 24.74 1.51 29.80 3.32 35.14 3.46

Area-based (2D) 22.79 1.07 27.92 3.05 33.29 3.57

n = 597 n = 972 n = 1,200

2x

3x (2x*4x)

3x (4x*2x)

4x

6x

8xy = 1,8037x + 17,717

R² = 0,9751

16

20

24

28

32

36

0 2 4 6 8 10

Pol

len

diam

eter

(µm

)

Somatic ploidy level

Fig. 5 Graphical representation of the linear correlation between

mean pollen volume and somatic ploidy level in Arabidopsis

Plant Reprod

123

distribution, representing particles that are substantially

larger (e.g., ±37 lm) compared to normal haploid spores

(Fig. 7e). Thus, although qrt1-1 tetrads have the same

genomic DNA content as tes-4 spores, their total volume is

substantially higher, indicating that volume-based pollen

size analysis allows for the discrimination of both types of

gametophytic defects.

Easy detection and quantification of diplogamete

formation through volume-based pollen size analysis

In natural plant populations, the analysis of the pollen

ploidy distribution provides valuable information on the

structure and diversity of the type of reproduction. In order

to check whether volume-based pollen size analysis tech-

nique can be used to determine the male gametophytic

ploidy distribution (e.g., diploid and/or aneuploid spores)

under natural conditions, we next analyzed the size distri-

bution of pollen from three differently reproducing Boec-

hera polyantha species and from an endogenous

Ranunculus species (e.g. Ranunculus acris).

To assess whether volume-based pollen size analysis can

be used to differentiate between sexual and apomictic

plants, we here compared the pollen size distribution of

three types of diploid Boechera polyantha: one sexual, one

apomictic, and one facultative apomict. In the sexually

reproducing line, mature pollen shows a unimodal size

distribution with a mean diameter of 12.6 ± 1.2 lm,

reflecting the stable formation of meiotically reduced,

haploid spores (Fig. 8a). In both the facultative and the

obligate apomictic lines, however, mature spores appeared

larger and displayed a mean volumetric pollen diameter of

17.6 ± 2.3 and 15.3 ± 3.0 lm, respectively (Fig. 8b, c),

suggesting an increased genomic DNA content. To assess

gametophytic ploidy, we performed DNA flow cytometry

and found that the pollen in both types of apomicts have a

duplicated (e.g., diploid) gametophytic ploidy level

(Fig. 8). However, in contrast to the high similarity in male

gametophytic ploidy, pollen size distribution appeared

slightly different in both types of apomicts, with spores

from the obligate apomict being more variable in size

compared to pollen from the facultative apomict. From this,

Relative fluorescence

Lycopersicon esculentum Brassica napus

Relative fluorescenceRelative fluorescence Relative fluorescence

a

b

c

Fig. 6 Volume-based size distribution (a) and nuclear DNA staining (b) of mature pollen and flow cytometric somatic ploidy determination

(c) in diploid and a tetraploid tomato and tetraploid and octaploid oilseed rape plant; scale bar, 20 lm

Plant Reprod

123

Wild type

atspo11-1-3

msh5-1

tes-4

qrt1-1

Control 2x atspo11-1-3 msh5-1 tes-4 qrt1-1

a

b

c

d

e

f

Fig. 7 Volume-based pollen

size distribution in Arabidopsis

diploid, tetraploid, and

octaploid control plants (a), the

meiotic mutants atspo11-1 (b),

msh5-1 (c), tes-4 (d), and the

gametogenesis-specific mutant

qrt1-1 (e); The large

subpopulation of smaller

particles in the qrt1-1 pollen

size distribution is due to a

severe trips infection, which

causes the presence of shriveled,

aborted pollen grains;

f Representative images of the

male meiotic tetrad stage in

corresponding control and

mutant plants; scale bar, 10 lm

Plant Reprod

123

we conclude that the size of mature Boechera polyantha

spores is not only influenced by the gametophytic ploidy

level, but may also depend on the underlying mode of

reproduction, or, on the genotypic differences between the

accessions tested. Increased pollen size variability in obli-

gate apomicts could additionally be caused by environ-

mental conditions or through the accumulation of

gametophytic mutations (i.e., Muller’s ratchet).

In a second case study, we analyzed the size distribution

of Ranunculus acris pollen and found that they display a

trimodal size distribution with a small subset of pollen

having a diameter lower than 24 lm, a second population

with a mean diameter of ±28 lm, and a large subset of

giant spores with a mean diameter of ±40 lm (Fig. 9a).

Microscopic observation confirmed the presence of these

three pollen types and demonstrated that the small pollen

grains corresponded to shriveled spores, whereas the other

two pollen types appeared swollen and viable (Fig. 9b).

Nuclear DNA staining demonstrated that both types of

viable spore populations display a typical binuclear con-

figuration with one condensed generative nucleus and one

less condensed vegetative one (Fig. 9c, d). However, in the

giant pollen, nuclei were substantially larger and exhibited

an enhanced fluorescence compared to nuclei in the smal-

ler-sized variants, indicating for an increased gametophytic

ploidy level (presumably diploid).

Based on these findings, we conclude that volumetric

pollen size analysis is not only applicable for sporophytic

and gametophytic ploidy estimations, but also offers a

quick and reliable technique to quantify diplogamete

Relative fluorescence

B. polyantha (ES903) – obligate sex

B. polyantha (ES805 ) – facult . apo

B. polyantha (ES776) – obligate apo

x

2x

2x

a

b

c

Fig. 8 Volume-base pollen size distribution and DNA flow cytom-

etry of mature pollen from sexually reproducing (a), facultative

apomictic (b), and obligate apomictic (c) Boechera polyantha plants;

all male gametophytic ploidy measurements were performed with one

1C pollen external control from the diploid sexual Boechera genotype

(ES 558.2), represented by the dotted profile

Plant Reprod

123

formation and to detect apomictically reproducing species,

at least for those in which 2n spore formation is linked to

the clonal seed phenotype.

Discussion

Volumetric versus image-based one- and two-

dimensional pollen size analysis

In several studies, pollen size has already been demon-

strated to be a useful parameter to determine the somatic

and gametophytic ploidy level in flowering plants (Johansen

and Vonbothmer 1994; Kelly et al. 2002; Altmann et al.

1994; Jacob and Pierret 2000; Kapadia and Gould 1964;

Katsiotis and Forsberg 1995). In all these reports, however,

pollen size analysis was performed on a one- or a two-

dimensional basis, respectively, using the diameter and

transsectional surface area as size parameter (Schols et al.

2002). Here, in this study, we present the three-dimensional

volume-based particle sizing technique as an alternative

method to analyze pollen size and demonstrate that it is an

easy, accurate, and reliable method to quickly characterize

the size distribution of large pollen samples, irrespective of

the shape and the position of the pollen grains.

Based on a comparative pollen size assay, we demonstrate

that the analysis of pollen grain size through volume-based

approaches is more accurate and confers a substantial

reduction of non-biological size variability compared to one-

and two-dimensional image-based approaches. As such, this

technique provides an enhanced discriminative power in

detecting biologically based differences in pollen size. The

comparative assay also revealed that one-dimensional

diameter assessments are strongly biased by the pollen’s

shape and position in the image plane. Indeed, in some

species, image-based pollen diameter quantifications over-

estimate the real pollen size, whereas in other species the

actual pollen body volume is underestimated. Since image-

based diameter overestimations typically occur in highly

eccentric spores, we presume that these biases are caused by

the overrepresentation of the major axis in image-based

measurements. Indeed, ellipsoid spores generally show an

ellipse-shaped transversal body orientation in slide prepa-

rations, causing image-based approaches to equally integrate

the major and minor axis into the final diameter value. This is

in contrast to the real particle dimensions, in which the

Fig. 9 a Volume-based diameter distribution of pollen harvested from Ranunculus acris; Microscopic brightfield visualization (b) and nuclear

DNA staining (c) of mature pollen isolated from a Ranunculus acris flower at anthesis

Plant Reprod

123

relative contribution of the minor axis in an ellipsoid volume

is two times that of the major axis. Alternatively, volume-

based particle size approaches calculate the diameter

parameter based on the three-dimensional body volume

(Hurley 1970). Indeed, in contrast to image-based particle

analysis, in which maximally two dimensions can be

implicated (e.g., diameter and area), volumetric size mea-

surements typically integrate all three dimensions of the

particle’s volume (Narhi et al. 2009; Barreiros et al. 1996).

As this precludes any bias toward a specific axis, this

methodology results in a more accurate assessment of the

particle’s body size, independent of its shape and position.

For spherical spores, microscopic diameter assessments

were found to underestimate the actual size of the particle.

This bias is most presumably caused by imperfections in

the image focus plane and by improper determination of

the particle boundary lines (Keyvani and Strom 2013).

More importantly, this underestimation could also be

attributed to changes in pollen morphology during the

image capturing procedure. Indeed, using time-lapse anal-

ysis series of volumetric pollen size, we found that mature

spores from several plant species upon isoton II suspension

undergo a gradual decrease in their body volume. As such,

these time-dependent changes in pollen size may confer

substantial alterations in pollen size acquisition, particu-

larly in time-consuming approaches such as image-based

pollen size analysis. In contrast, volume-based particle size

approaches typically allow for a quick analysis of large

amounts of pollen grains (±5,000 particles/min) and

therefore substantially reduce the effect of time-dependent

size alterations in suspended pollen grains (Shekunov et al.

2007).

An additional advantage of volume-based particle size

analysis (e.g., Multisizer Coulter Counter) is that the

sample preparation procedure only involves a simple (re-)

suspension of the pollen sample, whereas image-based

approaches typically require laborious slide preparation

and time-consuming microscopic recording and process-

ing. Moreover, as samples are automatically processed, the

Multisizer III enables a fast assessment of large amounts of

pollen grains and thus provides an excellent platform for

monitoring pollen size in a high throughput manner.

Volume-based pollen size analysis as a tool

to determine the somatic ploidy level of plants

For several plant species, it has been found that the size of

mature pollen grains is positively correlated with the somatic

ploidy level of the donor plant. Indeed, in both the dicoty-

ledonous Arabidopsis (Altmann et al. 1994), Arachis

(Singsit and Oziasakins 1992), cauliflower (Ockendon

1988), potato (Bamberg and Hanneman 1991) and Portulaca

(Mishiba and Mii 2000), and in many monocotyledonous

plant species (e.g. Bouteloua, Bromis, Paspalum and Avena),

a positive correlation between pollen size and somatic

chromosome number has been reported (Katsiotis and

Forsberg 1995; Kapadia and Gould 1964; Tan and Dunn

1973; Johansen and Vonbothmer 1994; Quarin and Hanna

1980).

In this study, we demonstrate that volume-based pollen

size analysis can be used as an alternative to microscopic

pollen sizing techniques in determining the somatic ploidy of

the donor plant. Based on observations in an Arabidopsis

2 9 –4 9 –89 ploidy series, we found that volume-based

pollen size analysis confers less non-biological size vari-

ability and reduces the size overlap between haploid, diploid,

and tetraploid pollen populations compared to one- and two-

dimensional approaches. Hence, for species that show a

positive correlation between the somatic ploidy level and

pollen size, volume-based pollen size analysis has a strong

discriminative power in differentiating pollen produced by

plants with a different ploidy level and thus provides an easy

and reliable method to quickly determine the somatic ploidy

level of plants or plant populations. This was further con-

firmed by the analysis of two other species, namely tomato

and oilseed rape, and is particularly interesting for plant

species that show a small ploidy-dependent pollen size var-

iation, such as perennial ryegrass (Lolium perenne L.)

(Jansen and Dennijs 1993).

Characterization of male gametogenesis through

volume-based pollen size analysis

Pollen size analysis of Arabidopsis recombination-defec-

tive mutants spo11 and msh5 showed a much broader

pollen size distribution compared to wild type plants, most

likely reflecting the large variability in the microspore’s

chromosome number. Indeed, since both spo11 and msh5

have defects in meiotic CO formation (Lu et al. 2008;

Grelon et al. 2001; Higgins et al. 2008), meiotic chromo-

some segregation is disturbed, typically resulting in

unbalanced tetrads and polyads that contain a wide variety

of mostly aneuploid spores (Pradillo et al. 2007). Hence,

the large pollen size variability in both spo11 and msh5

strongly confirms the earlier observed pollen ploidy-size

correlation in Arabidopsis, and additionally demonstrates

that volume-based pollen size can be used to detect defects

in male meiotic chromosome segregation. By comparing

spo11 and msh5, we found a substantial difference in the

overall pollen size distribution, with msh5 producing sig-

nificantly less small (\20 lm) aborted pollen grains com-

pared to spo11-1. This discrepancy is most presumably

caused by a differential severity of the recombination

defect. Indeed, in contrast to spo11-1, which shows a

complete loss of COs (Grelon et al. 2001), msh5 mutants

still display a residual level (*25 %) of chiasmata

Plant Reprod

123

(Lu et al. 2008). As such, alterations in msh5 male meiosis

are not as severe and typically generate unbalanced tetrads

with aneuploid spores that show less variability in their

genomic DNA content. Based on the correlation between

the type and severity of the recombination defect and the

associated pollen size distribution, we postulate that vol-

ume-based pollen size analysis can be used as a primary

cue to detect aberrations in meiotic recombination.

Similar to msh5 and spo11, triploid PMCs also exhibit

alterations in meiotic chromosome segregation, typically

leading to a wide variety of aneuploid and euploid

microspores (Henry et al. 2005).Indeed, due to the inherent

presence of three copies of each chromosome and the

associated interaction of all three copies in meiotic CO

formation (e.g., trivalents instead of bivalents), segregation

of homologous chromosomes in MI is always unstable,

typically leading to a 2-1 segregation ratio (Charles et al.

2010; Loidl 1995). Meiotic progression through meiosis II

consequently generates unbalanced tetrads that contain four

aneuploid spores; two with a reduced ploidy level (\1.59)

and two with an increased ploidy level ([1.59). In line with

this large variability in gametophytic ploidy, the size dis-

tribution of triploid pollen appeared highly variable, similar

as in spo11-1 and msh5.

In contrast to the mutants with an altered meiotic chro-

mosome segregation, volume-based pollen size analysis of

the meiotic cytokinesis-defective tes-4 resulted in a uni-

formly sized pollen distribution with a mean diameter

of ±31.50 lm, closely corresponding to the size of pollen

isolated from a tetraploid line. TES is a microtubule-asso-

ciated motor kinesin that plays a key role in post-meiotic cell

plate formation (Yang et al. 2003) and loss of TES function

typically generates large syncytial microspores that enclose

the four meiotic daughter nuclei (Hulskamp et al. 1997). In

subsequent microspore development, these co-allocated

nuclei either fuse or stay separated and generate a mix of

tetraploid and multinuclear pollen (Spielman et al. 1997). In

all types of tes spores, however, the total genomic DNA

content is the same and equals the genomic content of mature

tetraploid pollen. Hence, the observation that tes pollen show

a similar size distribution pattern of tetraploid-derived pollen

confirms the strong pollen ploidy-size correlation in Ara-

bidopsis and additionally indicates that volume-based pollen

size analysis can be used as a reliable technique for the quick

detection and accurate quantification of pollen with an

altered ploidy level (aneuploid and polyploid). A similar

detection of 2n male gametes through pollen size analysis

has already been documented in several meiotic non-

reduction mutants and species (De Storme and Geelen 2011;

Pichot and El Maataoui 2000; Ramsey 2007; Ortiz 1997;

Jansen and Dennijs 1993) and under conditions that favor

diploid and/or polyploid pollen formation (De Storme et al.

2012; Pecrix et al. 2011; Mason et al. 2011). Moreover, based

on observations in differently reproducing Boechera plants,

we here demonstrate that volume-based pollen size analysis

not allows a quick assessment of diploid gamete formation

(Kantama et al. 2007), but additionally provides quantitative

and qualitative data on the underlying mode of reproduction

(e.g., sexual or apomictic).

Thus, since the volume-derived pollen size distribution

closely corresponds to the type of meiotic or gametophytic

defect and strongly reflects the ploidy distribution of the

resulting gametes, we conclude that volume-based pollen

size analysis is a useful tool to distinguish between various

types of defects in the stability or progression of the male

reproductive program. Volume-based pollen size analysis

can therefore additionally be used as an indirect technique

to quickly determine the mode of reproduction (e.g., sexual

or apomictic) and the formation of diploid gametes in

natural plant populations.

Acknowledgments We would like to thank Susan Armstrong

(University of Birmingham) for demonstrating the meiotic spreading

protocol. Gratitude to Tim Bekaert (University of Ghent, Civil

Engineering) for helping with the MATLAB-based image analysis

and particle sizing software. Many thanks to Rob Dirks (Rijk Zwaan

Breeding B. V.), Raphael Mercier (INRA, Versailles), Rod Scott

(University of Bath), and Filip Debersaques (University College

Ghent) for providing plant material. This research is supported by an

aspirant fellowship to Nico De Storme and research grant G006709N

offered by the Flemish Funding Agency for Scientific Research

(FWO). Collaborations and travel were supported by the COST action

FA0903.

References

Ahloowal BS (1965) A root tip squash technique for screening

chromosome number in Lolium. Euphytica 14:170–172

Altmann T, Damm B, Frommer WB, Martin T, Morris PC, Schweizer D,

Willmitzer L, Schmidt R (1994) Easy determination of ploidy Level

in Arabidopsis thaliana plants by means of pollen size Measure-

ment. Plant Cell Rep 13:652–656

Avdeef A, Tsinman K, Tsinman O, Sun N, Voloboy D (2009)

Miniaturization of powder dissolution measurement and estima-

tion of particle size. Chem Biodivers 6:1796–1811

Bamberg JB, Hanneman RE (1991) Rapid ploidy screening of tuber-

bearing Solanum (Potato) species through pollen diameter

measurement. Am Potato J 68:279–285

Barreiros FM, Ferreira PJ, Figueiredo MM (1996) Calculating shape

factors from particle sizing data. Part Part Syst Charact

13:368–373

Beck SL, Dunlop RW, Fossey A (2003) Stomatal length and

frequency as a measure of ploidy level in black wattle, Acaciamearnsii (de Wild). Bot J Linn Soc 141:177–181

Bretagnolle F, Thompson JD (1995) Tansley review no-78—Gametes

with the somatic chromosome number—mechanisms of their

formation and role in the evolution of autopolyploid plants. New

Phytol 129:1–22

Brownfield L, Kohler C (2011) Unreduced gamete formation in

plants: mechanisms and prospects. J Exp Bot 62:1659–1668

Chan SWL (2010) Chromosome engineering: power tools for plant

genetics. Trends Biotechnol 28:605–610

Plant Reprod

123

Charles JS, Hamilton ML, Petes TD (2010) Meiotic chromosome

segregation in triploid strains of Saccharomyces cerevisiae.

Genetics 186:537–550

Cheng ZM, Korban SS (2011) In vitro ploidy manipulation in the

genomics era. Plant Cell Tissue Org 104:281–282

Consiglio F, Carputo D, Monti L, Conicella C (2004) Exploitation of

genes affecting meiotic non-reduction and nuclear restitution:

Arabidopsis as a model? Sex Plant Reprod 17:97–105

De Storme N, Geelen D (2011) The Arabidopsis mutant jasonproduces unreduced first division restitution male gametes

through a parallel/fused spindle mechanism in meiosis II. Plant

Physiol 155:1403–1415

De Storme N, Copenhaver GP, Geelen D (2012) Production of diploid

male gametes in Arabidopsis by cold-induced destabilization of

postmeiotic radial microtubule arrays. Plant Physiol 160:1808–1826

Deblock M, Debrouwer D, Tenning P (1989) Transformation of

Brassica napus and Brassica oleracea using Agrobacteriumtumefaciens and the expression of the bar and neo Genes in the

Transgenic Plants. Plant Physiol 91:694–701

d’Erfurth I, Jolivet S, Froger N, Catrice O, Novatchkova M, Mercier

R (2009) Turning meiosis into mitosis. Plos Biol 7(6). doi:

10.1371/journal.pbio.1000124

d’Erfurth I, Cromer L, Jolivet S, Girard C, Horlow C, Sun YJ, To

JPC, Berchowitz LE, Copenhaver GP, Mercier R (2010) The

CYCLIN-A CYCA1;2/TAM Is required for the meiosis I to

meiosis II transition and cooperates with OSD1 for the prophase

to first meiotic division transition. Plos Genet 6(6). doi:

10.1371/journal.pgen.1000989

Dolezel J, Bartos J (2005) Plant DNA flow cytometry and estimation

of nuclear genome size. Ann Bot London 95:99–110

Durbarry A, Vizir I, Twell D (2005) Male germ line development in

Arabidopsis. duo pollen mutants reveal gametophytic regulators

of generative cell cycle progression. Plant Physiol 137:297–307

Edlund AF, Swanson R, Preuss D (2004) Pollen and stigma structure

and function: the role of diversity in pollination. Plant Cell

16:S84–S97

Francis KE, Lam SY, Copenhaver GP (2006) Separation of Arabid-

opsis pollen tetrads is regulated by QUARTET1, a pectin

methylesterase gene. Plant Physiol 142:1004–1013

Galbraith DW, Harkins KR, Maddox JM, Ayres NM, Sharma DP,

Firoozabady E (1983) Rapid flow cytometric analysis of the cell

cycle in intact plant tissues. Science 220:1049–1051

Gamage HK, Schmidt S (2009) Short communication. A robust

method for chromosome quantification and ploidy determination

in woody species. Aust J Bot 57:87–93

Grelon M, Vezon D, Gendrot G, Pelletier G (2001) AtSPO11-1 is

necessary for efficient meiotic recombination in plants. EMBO J

20:589–600

Hardie DC, Gregory TR, Hebert PDN (2002) From pixels to

picograms: a beginners’ guide to genome quantification by

Feulgen image analysis densitometry. J Histochem Cytochem

50:735–749

Henry IM, Dilkes BP, Young K, Watson B, Wu H, Comai L (2005)

Aneuploidy and genetic variation in the Arabidopsis thalianatriploid response. Genetics 170:1979–1988

Higgins JD, Vignard J, Mercier R, Pugh AG, Franklin FCH, Jones GH

(2008) AtMSH5 partners AtMSH4 in the class I meiotic

crossover pathway in Arabidopsis thaliana, but is not required

for synapsis. Plant J 55:28–39

Hirsch J, Gallian E (1968) Methods for the determination of adipose

cell size in man and animals. J Lipid Res 9:110–119

Hulskamp M (2004) Plant trichomes: a model for cell differentiation.

Nat Rev Mol Cell Biol 5:471–480

Hulskamp M, Parekh NS, Grini P, Schneitz K, Zimmermann I, Lolle

SJ, Pruitt RE (1997) The STUD gene is required for male-specific

cytokinesis after telophase II of meiosis in Arabidopsis thaliana.

Dev Biol 187:114–124

Hurley J (1970) Sizing particles with a coulter counter. Biophys J

10:74–79

Jacob Y, Pierret V (2000) Pollen size and ploidy level in the genus

Rosa. Acta Hortic 508:289–292

Jansen RC, Dennijs APM (1993) A statistical mixture model for

estimating the proportion of unreduced pollen grains in perennial

ryegrass (Lolium Perenne L.) via the size of pollen grains.

Euphytica 70:205–215

Johansen B, Vonbothmer R (1994) Pollen size in Hordeum L.—

correlation between size, ploidy level, and breeding system. Sex

Plant Reprod 7:259–263

Johnston JS, Bennett MD, Rayburn AL, Galbraith DW, Price HJ

(1999) Reference standards for determination of DNA content of

plant nuclei. Am J Bot 86:609–613

Jones G, Heslop-Harrison J (eds) (1996) Classical and molecular

cytogenetics of Arabidopsis thaliana. Oxford University Press,

Oxford

Jones KD, Reed SM (2007) Analysis of ploidy level and its effects on

guard cell length, pollen diameter, and fertility in Hydrangeamacrophylla. HortScience 42:483–488

Kantama L, Sharbel TF, Schranz ME, Mitchell-Olds T, de Vries S, de

Jong H (2007) Diploid apomicts of the Boechera holboelliicomplex display large-scale chromosome substitutions and aber-

rant chromosomes. Proc Natl Acad Sci USA 104:14026–14031

Kapadia ZJ, Gould FW (1964) Biosystematic studies in the Boutelouacurtipendula complex. III. Pollen size as related to chromosome

numbers. Am J Bot 51:166–172

Katsiotis A, Forsberg RA (1995) Pollen grain size in four ploidy

levels of genus Avena. Euphytica 83:103–108

Kelly JK, Rasch A, Kalisz S (2002) A method to estimate pollen

viability from pollen size variation. Am J Bot 89:1021–1023

Keyvani A, Strom K (2013) A fully-automated image processing

technique to improve measurement of suspended particles and flocs

by removing out-of-focus objects. Comput Geosci 52:189–198

Liu JJ, Qu LJ (2008) Meiotic and mitotic cell cycle mutants involved

in gametophyte development in Arabidopsis. Mol Plant 1:

564–574

Loidl J (1995) Meiotic chromosome pairing in triploid and tetraploid

Saccharomyces cerevisiae. Genetics 139:1511–1520

Lu XD, Liu XL, An LZ, Zhang W, Sun J, Pei HJ, Meng HY, Fan YL,

Zhang CY (2008) The Arabidopsis MutS homolog AtMSH5 is

required for normal meiosis. Cell Res 18:589–599

Mason AS, Nelson MN, Yan GJ, Cowling WA (2011) Production of

viable male unreduced gametes in Brassica interspecific hybrids

is genotype specific and stimulated by cold temperatures. BMC

Plant Biol 11:103

Matzk F, Meister A, Schubert I (2000) An efficient screen for

reproductive pathways using mature seeds of monocots and

dicots. Plant J 21:97–108

McTainsh GH, Lynch AW, Hales R (1997) Particle size analysis of

aeolian dusts, soils and sediments in very small quantities using a

coulter multisizer. Earth Surf Process Landf 22:1207–1216

Miller BV, Lines RW (1988) Recent advances in particle size

measurements—a critical review. Crit Rev Anal Chem 20:

75–116

Mishiba K, Mii M (2000) Polysomaty analysis in diploid and

tetraploid Portulaca grandiflora. Plant Sci 156:213–219

Narhi LO, Jiang YJ, Cao S, Benedek K, Shnek D (2009) A critical

review of analytical methods for subvisible and visible particles.

Curr Pharm Biotechnol 10:373–381

Ockendon DJ (1988) The ploidy of plants obtained from another

culture of cauliflowers (Brassica oleracea Var Botrytis). Ann

Appl Biol 113:319–325

Plant Reprod

123

Ortiz R (1997) Occurrence and inheritance of 2n pollen in Musa. Ann

Bot 79:449–453

Pecrix Y, Rallo G, Folzer H, Cigna M, Gudin S, Le Bris M (2011)

Polyploidization mechanisms: temperature environment can

induce diploid gamete formation in Rosa sp. J Exp Bot 62:

3587–3597

Petit C, Bretagnolle F, Felber F (1999) Evolutionary consequences of

diploid-polyploid hybrid zones in wild species. Trends Ecol Evol

14:306–311

Pichot C, El Maataoui M (2000) Unreduced diploid nuclei in

Cupressus dupreziana A. Camus pollen. Theor Appl Genet

101:574–579

Praca-Fontes MM, Carvalho CR, Clarindo WR (2011) C-value

reassessment of plant standards: an image cytometry approach.

Plant Cell Rep 30:2303–2312

Pradillo M, Lopez E, Romero C, Sanchez-Moran E, Cunado N,

Santos JL (2007) An analysis of univalent segregation in meiotic

mutants of Arabidopsis thaliana: a possible role for synaptone-

mal complex. Genetics 175:505–511

Przywara L, Pandey KK, Sanders PM (1988) Length of stomata as an

indicator of ploidy level in Actinidia deliciosa. New Zeal J Bot

26:179–182

Quarin CL, Hanna WW (1980) Effect of three ploidy levels on

meiosis and mode of reproduction in Paspalum hexastachyum.

Crop Sci 20:69–75

Ramanna MS, Jacobsen E (2003) Relevance of sexual polyploidiza-

tion for crop improvement—a review. Euphytica 133:3–18

Ramsey J (2007) Unreduced gametes and neopolyploids in natural

populations of Achillea borealis (Asteraceae). Heredity 98:

143–150

Schols P, Dessein S, D’Hondt C, Huysmans S, Smets E (2002)

Carnoy: a new digital measurement tool for palynology. Grana

41:124–126

Shaheen N, Khan MA, Hayat MQ, Yasmin G (2009) Pollen

morphology of 14 species of Abutilon and Hibiscus of the

family Malvaceae (sensu stricto). J Med Plants Res 3:921–929

Shekunov BY, Chattopadhyay P, Tong HHY, Chow AHL (2007)

Particle size analysis in pharmaceutics: principles, methods and

applications. Pharm Res 24:203–227

Singsit C, Oziasakins P (1992) Rapid estimation of ploidy levels in

in vitro regenerated interspecific Arachis hybrids and fertile

triploids. Euphytica 64:183–188

Spielman M, Preuss D, Li FL, Browne WE, Scott RJ, Dickinson HG

(1997) TETRASPORE is required for male meiotic cytokinesis

in Arabidopsis thaliana. Development 124:2645–2657

Tan GY, Dunn GM (1973) Relationship of stomatal length and

frequency and pollen grain diameter to ploidy level in Bromusinermis Leyss. Crop Sci 13:332–334

Vilhar B, Greilhuber J, Koce JD, Temsch EM, Dermastia M (2001)

Plant genome size measurement with DNA image cytometry.

Ann Bot 87:719–728

Walker PH, Woodyer KD, Hutka J (1974) Particle size measurements

by Coulter Counter of very small deposits and low suspended

sediment concentrations in streams. J Sediment Petrol 44:

673–679

Walstra P, Oortwijn H (1969) Estimating globule size distribution of

oil-in-water emulsions by Coulter Counter. J Colloid Interf Sci

29:424–431

Yang CY, Spielman M, Coles JP, Li Y, Ghelani S, Bourdon V, Brown

RC, Lemmon BE, Scott RJ, Dickinson HG (2003) TETRA-

SPORE encodes a kinesin required for male meiotic cytokinesis

in Arabidopsis. Plant J 34:229–240

Zlesak DC (2009) Pollen diameter and guard cell length as predictors

of ploidy in diverse rose cultivars, species, and breeding lines.

Flor Ornam Biotechnol 3:53–70

Plant Reprod

123