MECHANISMS OF ADAPTATION IN ORYZA AND ARABIDOPSIS
By
Michael A. Grillo

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Plant Biology – Doctor of Philosophy
Ecology, Evolutionary Biology & Behavior – Dual Major
2013

i

ABSTRACT
MECHANISMS OF ADAPTATION IN ORYZA AND ARABIDOPSIS
By
Michael A. Grillo

Adaptation by natural selection is a driving force in evolution. Understanding the
ecological and genetic basis of adaptation to contrasting environments has been a major focus of
biology since the origins of evolutionary thought. While significant gains have been made in our
understanding of this process, the genetic basis of adaptation in particular remains poorly
understood. To uncover the mechanisms of adaptation we ultimately need to identify the genes
and mutations underlying adaptive traits and experimentally demonstrate their effect on fitness.
During the past century there have been noteworthy theoretical advances in
understanding the genetic basis of evolution. The earliest theory dates back to Fisher’s
geometric model which demonstrates that infinitesimally small mutations have a 50% chance of
being beneficial, whereas this probability quickly diminishes with mutations of increasing effect
size. Thus, Fisher concluded that mutations of small effect are the raw material for adaptation; a
viewpoint that aligned well with the long-standing ‘micromutational’ view of adaptation.
Several decades later Kimura reexamined this work and pointed out that while small mutations
are more likely to be beneficial, they are also highly prone to stochaistic loss by drift. Taking
both factors into account, Kimura suggested that intermediate size mutations are most likely to
contribute to adaptation. More recently, Orr revisited this work and identified that while
Kimura’s prediction is true for a particular step in adaptation it does not explain the distribution
of mutations throughout an adaptive walk. Using analytical theory and computer simulations,
ii

Orr has demonstrated that the effect sizes of beneficial mutations should follow an exponential
distribution, with fewer mutations of large effect fixed early on and mutations of decreasing sizes
fixed subsequently as populations approach the fitness optimum.
Although understanding the genetic basis of adaptation has been a long standing goal of
evolutionary biology, data on the subject has for the most part been limiting. The advent of
Quantitative Trait Loci (QTL) mapping, in particular, has provided a means for obtaining
rigorous such data. QTL mapping has been primarily useful for characterizing the genetic
architecture of phenotypic variation; that is estimating the number, effect size, and location of
loci controlling trait variation. The most noteworthy outcome of this line of research has been
the regular identification of large effect QTL. It should be noted that much of this research has
been conducted in model systems or domesticated plants and animals, although studies in natural
systems are beginning to accumulate.
Here I present a dissertation aimed at understanding the mechanisms of adaptation in two
wild rice species and locally adapted populations of Arabidopsis thaliana. First, I assess the
genetic architecture of adaptation in the wild progenitors of cultivated rice, by identifying QTL
for a number of putative adaptive traits. Through this work flowering time was revealed as a key
adaptation for habitat preference between these species. In the next chapter I attempt to elucidate
the genetic basis of a major flowering time QTL through fine mapping. I continue my
examination of flowering time genetics by examining the genetic basis of flowering time
differentiation between locally adapted populations of Arabidopsis thaliana. Finally, I conduct a
thorough study of comparative floral biology to identify key traits that control mating system
divergence between the wild rice relatives. This work sets the stage for future efforts to
understand the genetic basis of mating system evolution.

iii

This dissertation is dedicated to my Grandfather, Arthur Fredrick Visnaw.
He passed away in 2008 during the early stages of my graduate school career.
I attribute my passion for biology to him.

iv

ACKNOWLEDGMENTS
I am grateful to my graduate committee, Dr. Douglas Schemske, Dr. Jeff Conner, Dr.
Shinhan Shui, and Dr. Steve van Nocker, as well as Dr. Tao Sang. In particular I am
appreciative of Dr. Doug Schemske for his guidance and support. I would like to thank several
undergraduate students for their support, Claire Lorts, Martha McCoy, and Emily Kramer. I also
need to thank my family and friends for support.

v

TABLE OF CONTENTS
LIST OF TABLES……………………………………………………………………………...vii
LIST OF FIGURES……………………………………………………………………………viii
CHAPTER 1: Genetic architecture for the adaptive origin of annual wild rice,
Oryza nivara....................................................................................................................................1
Abstract………………………………………………………………………...…………2
Introduction……………………………………………………………………..………..2
Materials and Methods…………………….………………………………………….…5
Results………………………………………………………………………………….....8
Discussion…………………………………………………………………………….…16
Acknowledgements……………………………………………………………………..28
LITERATURE CITED………………………………………………………………...29
CHAPTER 2: Fine mapping of a major flowering time QTL in Oryza………………….....37
Introduction…………………………………………………………………………..…38
Materials and Methods……….………………………………………………………...43
Results………………………………………………………………………………...…46
Discussion…………………………………………………………………………...…..49
LITERATURE CITED ………………………………………………………...……...58
CHAPTER 3: Genetic architecture of flowering time differentiation between locally
adapted populations of Arabidopsis thaliana…………………………………………...……..65
Abstract……………………………………………………………………………...…..66
Introduction…………………………………………………………………………..…66
Materials and Methods……………………………….……………………………..….71
Results……………………………………………………………………………..….....76
Discussion……………………………………………………………………..….…......79
Acknowledgements……………………………………………………………………..90
LITERATURE CITED ………………………………………………………....…......91
CHAPTER 4: Functional traits underlying mating system differentiation between the wild
relatives of rice……………………………………………………………………………...…100
Introduction………………………………………………………………………....…101
Materials and Methods……………………………….…………………………….…107
Results………………………….……………………………………………………....111
Discussion…………………………………………………………………………...…116
LITERATURE CITED ………………………………………………………………130
vi

LIST OF TABLES
Table 1.1. Effect of QTL alleles of O. nivara found in the combined analysis……...…...…10
Table 1.2. Epistatic interactions between QTL………………………………………….…...15
Table 2.1. Flowering time of Near-isogenic lines (NILs)……………………………….……47
Table 2.2. Genes in the fine mapped region……………………………………………….…48
Table 2.3. Polymorphisms in the FTL1 coding region…………………………………....…52
Table 3.1. Results of quantitative trait locus (QTL) analysis of flowering time……….…..75
Table 3.2. Polymorphisms of candidate gene coding regions between the Italy and Sweden
Arabidopsis thaliana mapping parents……………………………….…………………...…..79
Table 4.1. Potential mating system traits and their putative function…………….………111
Table 4.2. Trait means between O. rufipogon and O. nivara……………………………….117
Table 4.3. Trait means for additional populations of each species………………...………122

vii

LIST OF FIGURES
Figure 1.1. Phenotypic divergence between O. nivara and O. rufipogon……...………….....13
Figure 1.2. Distribution of QTL Effect Size……………………………………………….….26
Figure 2.1. Lifecycle of Oryza species………………………………………………….......….50
Figure 2.2. Fine Mapping Results…………………………………………………………......52
Figure 2.3. Days to flowering in the recombinant mapping population……………………56
Figure 3.1. Proportion of plant flowering…………………………………………..….….….77
Figure 3.2. Days to flowering in the F2 plants………………………………………...…...…82
Figure 3.3. Days to flowering in the Vernalization treatment……………………..……..…87
Figure 4.1. Functional traits of Oryza species…………………………………………….....114
Figure 4.2. Pollen viability time course……………………………………………….…..…118
Figure 4.3. Stigma receptivity time course……………………………………………….…119
Figure 4.4. Autogamy and geitonogamy………………………………………………….…125
Figure 4.5. Flower caricatures of O. nivara and O. rufipogon……………………………..127

viii

CHAPTER 1
Genetic architecture for the adaptive origin of annual wild rice, Oryza nivara

The following chapter was published in Evolution.
Grillo, M.A, C. Li, A.M. Fowlkes, T.M. Briggeman, A. Zhou, D.W. Schemske, and T.
Sang. 2009. Genetic architecture for the adaptive origin of annual wild rice, Oryza
nivara. Evolution 63: 870-883.
1

Abstract
The wild progenitors of cultivated rice, Oryza nivara and Oryza rufipogon, provide an
experimental system for characterizing the genetic basis of adaptation. The evolution of annual
O. nivara from a perennial ancestor resembling its sister species, O. rufipogon, was associated
with an ecological shift from persistently wet to seasonally dry habitats. Here we report a
quantitative trait locus (QTL) analysis of phenotypic differentiation in life history, mating
system, and flowering time between O. nivara and O. rufipogon. The exponential distribution of
effect sizes of QTL fits the prediction of a recently proposed population genetic model of
adaptation. More than 80% of QTL alleles of O. nivara acted in the same direction of
phenotypic evolution, suggesting that they were fixed under directional selection. The loss of
photoperiod sensitivity, which might be essential to the survival of the ancestral populations of
O. nivara in the new environment, was controlled by QTL of relatively large effect. Mating
system evolution from cross- to self-fertilization through the modification of panicle and floral
morphology was controlled by QTL of small-to-moderate effect. The lack of segregation of the
recessive annual habit in the F2 mapping populations suggested that the evolution of annual from
perennial life form had a complex genetic basis. The study captured the genetic architecture for
the adaptive origin of O. nivara and provides a foundation for rigorous experimental tests of
population genetic theories of adaptation.

Introduction
Darwin considered adaptation by natural selection a primary driving force for the origin
of new forms and species. The following century witnessed noteworthy theoretical advances in
elucidating the genetic basis of adaptation. The earliest theory traces back to Fisher’s (1930)

2

geometric model predicting that favorable mutations fixed during adaptation should have small
phenotypic effect, because mutations of large effect were more likely to generate negative,
pleiotropic effects. Kimura later suggested that mutations of small effect could suffer from
stochastic loss and concluded that mutations of intermediate sizes should be most frequently
fixed. More recently, Orr (1998) proposed that the effect sizes of beneficial mutations should
follow an exponential distribution, with fewer mutations of large effect fixed early on and
mutations of decreasing sizes fixed subsequently as populations approach the fitness optimum.
As molecular genetic markers and techniques became increasingly accessible to evolutionary
biologists, empirical evidence for the genetic basis of adaptation, revealed largely by quantitative
trait locus (QTL) mapping, has accumulated over the past two decades (Mackay, 2001a; Mackay,
2001b; Mauricio, 2001; Barton & Keightley, 2002; Rieseberg et al., 2002; Rieseberg et al., 2003;
Orr, 2005; Slate, 2005; Mitchell-Olds & Schmitt, 2006). Perhaps the most remarkable progress
has been made in elucidating the genetic basis of crop domestication under artificial selection
(Doebley et al., 2006). Genes and in some cases mutations responsible for major domestication
transitions have been identified through molecular cloning of QTL in a number of important
crops, such as maize, rice, wheat, barley, and tomato (Doebley et al., 1997; Frary et al., 2000;
Wang et al., 2005; Konishi et al., 2006; Li, CB et al., 2006a; Simons et al., 2006; Komatsuda et
al., 2007). Similar insights were recently gained in natural systems such as sticklebacks and
beach mice, where genes or mutations underlying adaptive traits were identified (Colosimo et al.,
2005; Hoekstra et al., 2006). These successful evolutionary genetic analyses have benefited
from a study system that consisted of an ancestor and its recent derivative with phenotypic traits
of clear adaptive importance and evolutionary direction. With these attributes, the wild
progenitors of rice, Oryza nivara and O. rufipogon, offer an excellent natural system for studying

3

the mechanisms of adaptive evolution. They are recently diverged sister species that have
undergone marked adaptive differentiation to distinct habitats in tropical and subtropical Asia
(Vaughan, 1994; Ge et al., 1999; Vaughan et al., 2008). Phylogenetic analysis has indicated that
the annual species O. nivara was derived from a perennial ancestor resembling O. rufipogon
(Zhu & Ge, 2005). A QTL analysis of phenotypic divergence between the two species thus
should estimate the overall genetic architecture for the evolution of O. nivara from the O.
rufipogon-like ancestor. Oryza nivara has differentiated from O. rufipogon in a number of traits
that are considered of general importance for plant adaptation, including life history, mating
system, and flowering time. Inhabiting deep-water swamps, O. rufipogon is perennial,
outcrossing, and photoperiod sensitive. Adapted to the seasonally dry habitat, O. nivara became
annual, predominately self-fertilized, and photoperiod insensitive. The derivation of an annual
from a perennial life history and selfing from outcrossing mating system in O. nivara
exemplifies not only the widely documented trends of life history and mating system evolution,
but also the tendency of evolutionary association between annual habit and self-fertilization in
flowering plants (Stebbins, 1974; Barrett et al., 1996). The correlation between life history and
habitat stability in these Oryza species, that is annual in seasonal environment and perennial in
aseasonal environment, also represents a frequently observed phenomenon in plant adaptation
(Stebbins, 1974). The ability to control flowering time by responding to photoperiod is another
important adaptive mechanism in plants. It allows for the optimization of timing of pollination
and seed dispersal and consequently maximizes reproductive success. Moreover, divergent
flowering time can also promote assortative mating and act as a prezygotic barrier to gene flow
(Kirkpatrick, 2000; Weis & Kossler, 2004), which may have served as a primary mechanism of
reproductive isolation between the two Oryza species (Kuroda et al., 2005). The adaptive

4

importance of these phenotypic differentiations between O. nivara and O. rufipogon was
supported by a series of common garden experiments (Morishima et al., 1961; Oka &
Morishim.H, 1967; Oka, 1976), in which the two species were taxonomically treated as the
annual and perennial strains of O. perennis. Here we report a QTL analysis of important
adaptive traits involving life history, mating system, and photoperiod sensitivity of the two wild
Oryza species. The availability of rice genome sequences has allowed the development of a
high-quality linkage map with microsatellite markers evenly covering the 12 chromosomes. The
results illustrate the genetic footprints of the adaptive origin of the annual species, O. nivara.
The estimate of the number and magnitude of effect of QTL underlying the adaptive evolution
provides empirical data for evaluating the predictions of the population genetic models of
adaptation. The estimate of the dominant effect and direction of effect of QTL offers
opportunities to address questions concerning the genetic consequences of the interplay between
mating systems, standing genetic variation, and natural selection during adaptive evolution.

Materials and Methods
Oryza nivara and O. rufipogon are distributed from southeastern Asia to India. Based on
our greenhouse observation of both species, two individuals representative of typical O.
rufipogon (IRGC 80506 from India) and O. nivara (IRGC 80470 from India) were chosen as
mapping parents (Fig. 1.1). Because the outcrossing parent O. rufipogon is highly heterozygous,
we generated two F2 mapping populations (A, B) by self-fertilizing two F1 plants that were
divergent in O. rufipogon alleles. Each mapping population consisted of 192 individuals. Both
populations were grown in the same size pots under uniform environmental conditions and
natural day length in a greenhouse at Michigan State University. The F2 seeds were germinated

5

in April 2004, and the mapping populations were grown for 2 years. Self-fertilized seeds of each
parent were germinated at the same time, and grown under the same conditions as the F2
populations (N = 15, O. nivara; N = 30, O. rufipogon).
Simple sequence repeats (SSRs), or microsatellites, were designed based on previous
publications and rice genome sequences (Temnykh et al., 2001; McCouch et al., 2002; Li, CB et
al., 2006a). SSRs were amplified by polymerase chain reaction (PCR) from DNA isolated using
FastDNA kits (Q-BIOgene, Carlsbad, CA), and recorded on agarose gels stained with ethidium
bromide. A total of 116 polymorphic markers with confirmed linkage using JoinMap 3.0 (Van
Ooijen & Voorips, 2001) were chosen to construct the genetic map. The markers evenly covered
the 12 chromosomes, with an average density of approximately 13 cM. Segregation distortion
was tested using JoinMap 3.0.
Flowering time was determined as days from germination to flowering. The seeds of the
F2 mapping populations and the selfed progenies of the parents germinated between April 30 and
May 6, 2004.
Traits associated with mating system include panicle shape, panicle exsertion, anther
length, and the timing of anther dehiscence (Fig. 1.1). Panicle shape was measured by the angles
between primary branches and rachis. Panicle exsertion was measured by the distance from the
flag leaf node to the first primary branch node. Each trait was measured twice when
approximately 25% and 75% of flowers had opened on a panicle. The average of these two
values was used to estimate panicle shape and exsertion around the peak flowering time. Three
panicles were measured for a plant. Anther length was measured from three flowers that were
about to open. In O. nivara, anthers become dehiscent immediately after the flower opens,
whereas anther dehiscence is delayed from several minutes to more than half an hour in O.

6

rufipogon. This trait, however, was difficult to evaluate in the F2 populations because the timing
of anther dehiscence was influenced by environmental factors, such as light, temperature, and
humidity. As a result, we did not include it in the QTL study.
To determine whether F2 plants are annual like O. nivara, the longevity of the F2s were
evaluated while the mapping populations were kept for 2 years. Other life-history traits
involving resource allocation are tiller length, number of flowers (spikelets) per panicle, grain
weight, and flowering duration. Tiller length was measured as the mean of the length of the
three longest tillers. The number of spikelets per panicle was calculated as the mean of the five
largest panicles. Grain weight was calculated as the mean of at least 30 grains. Flowering
duration was measured as the proportion of flowering tillers in February and May, 2005 and
scored as: 0% (0), 1–24% (1), 25–49% (2), 50–74% (3), 75–100% (4).
QTL were identified using composite interval mapping (CIM) implemented in the
software package QTL Cartographer 2.5. A forward and backward stepwise regression
procedure was used to identify cofactor for CIM.A10-cM scan window was employed, and the
likelihood-ratio statistic was computed every 2 cM. Experiment-wide significance (P < 0.05)
thresholds for QTL detection were determined with 1000 permutations. LOD values and R2
were determined based on likelihood-ratio tests under a hypothesis allowing both additive and
dominance effects (H3:H0). Multiple interval mapping (MIM) was also conducted (Kao et al.,
1999). A Bayesian information criterion (BIC) model selection with the penalty function c(n) =
ln(n) was used to perform MIM analysis at 1 cM steps. CIM and MIM analyses yielded very
similar results with essentially the same number, chromosomal locations, and effect size of QTL
identified. The main difference was that the LOD values of MIM were lower. Because the
permutation test is still under development for MIM (Z. B. Zeng, per. comm.), we could not

7

compare the significance levels associated with the LOD values resulting from the two methods.
Thus the results of CIM analysis were reported. Epistasis between QTL was tested using twoway analysis of variance (ANOVA) (Westerbergh & Doebley, 2004). The markers closest to the
LOD peaks of QTL were tested for interaction. To test QTL clustering, we compared the
distribution of the nearest neighbor QTL distances to a random distribution according to Protas et
al. (2008).
Based on the genetic mapping and rice genome sequences, we identified a candidate
gene, Hd1, for one of the flowering time QTL, DF6. The entire coding region of Hd1 was PCR
amplified from O. rufipogon using primers, AFH 5’
CACGAAGCTTATGAATTATAATTTTGGTGGC and ARM 5’
CGTGACGCGTTCAGAACCATGGAACAGTAC. The region was cloned into TA plasmids
and fully sequenced. The verified sequence was then moved into a GA plasmid and driven by a
35S promoter. The plasmid was introduced into Agrobacterium tumedaciens strain LBA4404,
which was then used to infect calli induced from germinating seeds of the O. nivara mapping
parent. RNA was isolated from leaves of transgenic plants using the RNeasy Plant Mini Kit
(Qiagen, Valencia, CA). The expression of the introduced O. rufipogon Hd1 coding region was
confirmed by RT-PCR and sequencing. Transgenic plants that expressed the O. rufipogon Hd1
were grown under long-day conditions (>14-h day time) for phenotypic evaluation.

Results
In this study, we evaluated morphological and physiological traits potentially important
for the adaptation of O. nivara to a new habitat (Fig. 1.1). These include flowering time,
continuation of flowering, tiller length, the number of flowers (spikelets) per panicle, grain

8

weight, panicle and floral morphology related to mating system differentiation, and longevity
(see Discussion for details). For these phenotypic traits, the variation in the mapping populations
as well as the mean and standard deviation in the parents is illustrated in Figure 1.1. Two traits
associated with reproduction allocation, number of spikelets per panicle, and grain weight, show
clearly transgressive segregation. As a short-day plant, the mapping parent of O. rufipogon did
not flower when it was kept under a long-day condition (> 14-h day time). Under the natural
day-length condition in the greenhouse, O. rufipogon began to flower in mid-October (∼11-h
day time), which corresponds to the day length experienced during early November in its native
distributional range in India.
The O. nivara mapping parent was photoperiod insensitive and flowered after three to
four months of vegetative growth under any day-length conditions. When grown together with
the mapping populations in May of 2004, the self-fertilized progeny of the O. nivara parent
flowered between the 9th and 16th of August, and the progeny of the O. rufipogon parent
flowered between October 6th and November 7th. Oryza nivara flowered continuously
throughout the growing season. By March 2005, all tillers of O. nivara had flowered and died,
behaving as a true annual plant even with the unlimited supply of water in the greenhouse (Fig.
1.1). All of the mature tillers produced by O. rufipogon progeny completed flowering by
February 2005. The remaining younger tillers did not flower until October 2005, the next shortday cycle. The maintenance of vegetative tillers during the current growing season allowed
continuous growth into the next year, contributing to the perenniality of O. rufipogon. All O.
rufipogon and F1 plants have survived for multiple years, indicating that the annual life history is
recessive.

9

The F2 populations began to flower on July 31st, and the last F2 plant flowered on 15
November 2004. All but three F2 plants survived for 2 years in the greenhouse. The three
individuals died during the first year probably as a result of hybrid weakness, showing short
tillers and abnormal flowers with thin and aborted anthers. Thus, of the total 384 F2 individuals,
there was no typical annual plant.
The O. rufipogon mapping parent was heterozygous (showing length polymorphism) for
66 of 116 or 59% SSR markers. The observation of such a high level of heterozygosity in the O.
rufipogon genome indicated that the mapping parent was from a highly outcrossing population.
In contrast, no SSR polymorphism was found in the O. nivara parent. The two F1 plants
used for developing the F2 mapping populations, A and B, had different O. rufipogon alleles at
38 of the 66 heterozygous SSR loci. Test for segregation distortion showed that significant
deviation from the expected segregation ratio of 1:2:1 (P < 0.01) occurred at 15 SSR markers in
mapping population A and 14 SSR markers in population B. Of these markers, eight showed
segregation distortion in both populations, which covered two blocks on chromosomes 2 and 3.

Table 1.1. Effect of QTL alleles of O. nivara found in the combined analysis.
Effect of QTL alleles of O. nivara found in the combined analysis. R2, percent phenotypic
variance explained by QTL; a, additive effect; d, dominance effect. Modes of gene action (GA):
d/a <−0.75, recessive (R); −0.75 ≤d/a ≤−0.25, partially recessive (r); −0.25
<d/a <0.25, additive (A); 0.25 ≤d/a ≤0.75, partially dominant (d); and d/a > 0.75, dominant (D).
If the additive effect of an allele acted in the opposite direction of phenotypic variation, a “−”
was added to the abbreviation for gene action (e.g., −r for TL7.2). LOD threshold determined by
permutation for each trait in each is given in the parentheses following the trait name.

10

Table 1.1 (cont’d)
trait

QTL

population A

population B

LOD

R2

a

d/a

DF1

28.6

44.0

-21.3

0.23

A

DF6

9.0

9.3

-15.3

-0.64

DF9

11.4

14.1

-17.6

Continued
flowering

CF1

19.4

36.2

CF3

5.6

A:3.5, B: 3.6

CF10
TL1.1

A: 3.5, B:3.5

Tiller length

R2

a

d/a

GA

16.7

24.8

-12.6

0.64

d

r

11.0

13.3

-14.7

-0.71

r

-0.27

r

14.6

21.1

-15.6

0.13

A

1.2

-0.15

A

16.8

34.0

0.63

0.70

d

8.7

0.25

1.6

D

3.9

8.4

0.41

0.97

D

4.9

9.3

0.36

0.89

D

8.5

16.5

0.55

0.84

D

17.1

27.3

-12.1

-0.06

A

11.1

20.6

-8.7

0.33

d

TL1.2

Days to flowering

GA LOD

3.9

6.9

-6.3

-0.76

R

TL3

3.7

5.8

-6.0

-0.53

r

TL4

3.8

7.7

-7.5

-0.40

r

TL6

4.8

9.3

-7.8

-0.46

r

7.8

11.1

-4.3

0.74

d

TL7.1

11.9

25.1

-5.8

1.0

D

10.1

17.3

-4.5

1.3

D

TL7.2

5.7

10.4

-9.0

-0.07

A

PE1.1

4.1

9.8

0.9

0.58

-d
4.4

9.6

-0.84

0.06

A

PE7

4.5

9.4

1.1

-0.06

-A

PS3

5.0

11.5

-3.6

0.31

d

PS6.1

A: 3.5, B:3.5

4.1

9.6

-4.4

-0.45

r

PS6.2

3.9

10.1

-4.1

0.11

A

5.7

12.0

-0.20

0.02

A

3.9

7.9

-0.18

-0.44

r

Panicle exsertion

PE1.2

A: 3.5, B: 3.5

PE6

Panicle shape
A: 3.6, B: 3.4

AL2.1

3.9

6.0

7.9

13.8

-1.3

-0.27

-0.65

-0.48

r

r

Anther length

AL2.2

A:3.6, B: 3.6

AL5

4.3

9.0

-0.22

-0.36

r

4.1

7.1

-0.11

1.00

D

AL6

4.5

9.2

-0.16

-0.37

r

4.1

8.5

-0.18

-0.56

r

Spikelet number

SN1

7.1

13.8

-10.7

0.01

-A

15.0

23.3

-14.1

0.15

-A

A: 3.3, B:3.5

SN4

4.2

6.0

-8.2

-0.56

-r

11

A total of 22 and 28 QTL were identified from populations A and B, respectively. The
QTL detection was based on LOD thresholds determined by permutation for each trait, ranging
from 3.3 to 3.6.
Fifteen QTL had overlapping 1-LOD support intervals between the mapping populations,
suggesting that they may be the same QTL detected independently from the two mapping
populations. These included all of the QTL with effect sizes larger than 15% in at least one of
the populations. QTL detected in only one population invariably had phenotypic effects lower
than 15% in both mapping populations.
Of the 15 shared QTL, nine are found nearest to SSR markers with the same O. rufipogon
allele in both populations and six are nearest to SSR markers with different O. rufipogon alleles
between populations A and B. Of the seven QTL detected only in population A, four are found
nearest to SSR markers with the same O. rufipogon allele in both populations and three are
nearest to SSR markers with different O. rufipogon alleles between populations. The chi-square
test indicates that there is no significant association between the detection of QTL unique to
population A and the presence of the nearest O. rufipogon alleles unique to the population (χ2 =
0.016, df = 1, P = 0.90). No significant association is found for population B either (χ2 = 0.080,
df = 1, P = 0.78), where five of 15 shared QTL and five of 13 unique QTL are found nearest to
SSR markers with different O. rufipogon alleles between populations.
The results indicate that the allelic variation of the outcrossing parent does not seem to
have a significant effect on QTL detection suggesting that O. rufipogon alleles showing SSR
length polymorphism are functionally similar as far as interspecific phenotypic variation is
concerned. Thus we combined the data of the two populations by treating two O. rufipogon
alleles with length polymorphism at a locus as the same allele in the combined analysis.

12

The combined analysis detected a total of 30 QTL with effect sizes ranging from 2.9% to
36.5% (Table 1.1). The distribution of effect sizes (R2) of these QTL is illustrated in Figure 1.2.
Of these QTL, 25 or 83.3% have additive effects in the same direction as the phenotypic
evolution from O. rufipogon to O. nivara. The modes of action of the O. nivara alleles are
broken down as: nine additive, 11 recessive or partially recessive, and 10 dominant or partially
dominant.
Significant epistasis was detected between six pairs of QTL affecting four traits (Table
1.2). To test QTL clustering, 30 points were placed randomly on the linkage map. An average of
three of 19 nearest neighbor distances were within 5 cM, compared to seven of 20 distances for
the QTL dataset (Fisher’s exact test, P = 0.96). This suggests that there is no significant
clustering of QTL.

Figure 1.1. Phenotypic divergence between O. nivara and O. rufipogon.
For interpretation of the references to color in this and all other figures, the reader is referred to
the electronic version of this dissertation. Phenotypic divergence between O. nivara and O.
rufipogon. (A) Oryza rufipogon mapping parent. (B) Oryza nivara mapping parent. (C) Open
and exserted panicle of O. rufipogon. (D) Compact and partially inserted panicle of O. nivara.
(E) An open flower of O. rufipogon, with large, indehiscent, pendant anthers. (F) An open flower
of O. nivara, with smaller, immediately dehiscent, upright anthers. (G) Oryza rufipogon at the
end of the first growing season, with dried tillers that had previously flowered and green tillers
that would flower in the next season. (H) Oryza nivara at the end of the growing season, when
all tillers had flowered and died.

13

Figure 1.1 (cont’d).

14

Table 1.2. Epistatic interactions between QTL.
Epistatic interactions between QTL revealed through two-way ANOVA with markers closest to
the LOD peaks of QTL. ∗ P<0.05, ∗∗ P<0.01, ∗∗∗ P<0.001.

Trait
Days to flowering
Tiller length
Spikelet number
Grain weight

QTL
DF1 x DF6
DF1 x DF9
DF6 x DF9
TL1.1 x TL7.2
SN1 x SN9
GW2 x GW6

Markers
RC1–33×MRG5836
RC1–33×RC9–78
MRG5836×RC9–78
RC1–42×RM248
RC1–33×MRG4840
MRG2439×RM528

F-value
4.016
4.422
2.185
2.812
3.985
3.733

Significance
***
***
*
*
**
**

Of the three flowering time QTL identified in this study, the one with the smallest effect,
DF6, mapped near a candidate gene, Hd1. Hd1 is the rice ortholog of Arabidopsis CONSTANS,
one of the central regulators of photoperiod sensitivity in both Arabidopsis and rice (Cremer &
Coupland, 2003; Hayama et al., 2003). We sequenced the coding region of the gene of the
mapping parents and found that the O. nivara parent had a 55-bp insertion in the first exon of the
gene. This was most likely a loss-of-function mutation. Previous studies showed that rice
cultivars carrying a 2-bp insertion at the beginning of the second exon of Hd1, a loss-of-function
mutation, had weaker photoperiod response than cultivars with the normal gene (Yano et al.,
2000).
To investigate the functional consequence of the loss-of-function mutation of Hd1 in O.
nivara, we introduced the Hd1 coding region of O. rufipogon driven by the 35S promoter into
15

the O. nivara parent. A total of 12 T0 transgenic plants that expressed the O. rufipogon gene
were obtained. All of the transgenic plants flowered when grown under a long-day condition
(day length > 14 h). Seeds of five randomly chosen T0 lines together with those of the O. nivara
parent were germinated and 10 T1 plants of each line were grown to maturity. All of the T1
plants flowered under the long-day condition, and their flowering time overlapped with that of O.
nivara. These results showed that 35S-promoter driving expression of O. rufipogon Hd1 coding
region did not increase the sensitivity of photoperiod response of O. nivara.

Discussion
Because the genome of the outcrossing parent, O. rufipogon, is highly heterozygous, two
F2 mapping populations were used to account for potential allelic effect on QTL detection.
Taking advantage of the self-compatibility of the outcrossing parent and the F1s, we developed
the F2 mapping populations through self-fertilization of separate F1 plants. In either mapping
population we found<13% of SSR markers had experienced segregation distortion. The
predominant Mendelian segregation in each mapping population allowed for straight forward
QTL analyses.
With the two mapping populations, we could evaluate the extent to which different O.
rufipogon alleles might affect the results of the QTL analysis. It is noteworthy that QTL with
effect larger than 15% in at least one population are always detected in both populations in the
separate analyses. The effect sizes of these shared QTL are significantly correlated between
mapping populations (Spearman’s Rho, r = 0.83, P = 0.0002). Furthermore, QTL that were
identified in both A and B populations in the separate analyses were also found in the combined

16

analysis. The use of two mapping populations thus reinforces that these QTL are reliably
identified.
In the separate analyses, the fact that only QTL of small effect lack the counterparts in
the other population seems to suggest that this may be due primarily to limited statistical power
for detecting small QTL in one of the mapping populations each with less than 200 F2
individuals (Beavis 1998). In the combined analysis, we were able to detect QTL with effect
size as small as 2.9%, lower than 3.8% of the smallest effect of QTL found in the separate
analyses. The combined analysis identified four new QTL, TL8, SN2, SN12, and GW10, with
effect ranging 4.0–6.4%. However, nine QTL detected in either A or B population in the
separate analyses were not found in the combined analysis, including two A population QTL
with effect sizes of 7.7% and 9.8%, and seven B-population QTL of 5.8–10.1% effect (Table
1.1).
Apparently, the limited detection power of relatively small populations may explain the
absence of small-effect QTL in one of the populations, but cannot explain the presence of QTL
unique to one of the populations but absence in the combined analysis. One straightforward
explanation for this is that the analysis of small populations could have created false QTL.
However, we cannot rule out the possibility that the ability to detect small effect QTL in the
combined analysis were undermined by subtle functional divergence between different O.
rufipogon alleles. In fact, of nine QTL detected in one of the populations but not in the
combined analysis, four are associated with the homozygous loci of O. rufipogon and five are
associated the heterozygous loci. In comparison, of 10 QTL detected in one of the populations
as well as in the combined analysis, seven and three are associated with homozygous and
heterozygous loci of O. rufipogon, respectively. Thus, this raises the question of whether alleles

17

with minor functional differentiation might be easier to detect in separate analyses than in the
combined analysis in which they are treated as the same functional allele.
Taken together, our analyses show that the detection of QTL of relative large effect is
insensitive to the allelic variation, implying that QTL of large effect identified from two sampled
individuals as the crossing parents may harbor genetic mutations important for the differentiation
between the populations and species that they represent. Analyses of small populations could
either omit QTL of small effect or create false QTL. However, the question of whether QTL
present in one of the populations but not in the combined analysis are all or partially false cannot
be fully addressed with the available data. In any event, QTL identified in both separate and
combined analyses are considered to be reliable, and the results of the combined analysis will be
the focus of further discussion.
Like the majority of Oryza species, O. rufipogon is a short-day plant that flowers near the
end of the monsoon season. For perennials that live in relatively aseasonal habitats and rely
primarily on vegetative reproduction, photoperiod sensitivity allows for the synchronization of
flowering time and perhaps is advantageous also in preventing a prolonged flowering period and
overproduction of seeds. Occurring in seasonal habitats, O. nivara is photoperiod insensitive and
begins to flower three to four months after seed germination. This coincides with the middle of
the monsoon season, allowing O. nivara to complete seed production before the onset of the dry
season. Late flowering, however, is probably deleterious in the seasonally dry habitat because it
does not allow enough time for seed production before desiccation. This study thus provides
genetic evidence for the potential importance of the loss of photoperiod sensitivity in the
adaptive evolution of a plant species.

18

Three QTL, DF1, DF6, and DF9, identified for flowering time have moderate-to-large
effect (Table 1.1). Particularly, DF1 is the largest (R2 = 36.5%) of all QTL found in the
combined analysis. Fixation of the allele of such a large effect could have substantially
weakened photoperiod sensitivity of the O. rufipogon-like ancestor and consequently improved
its fitness by permitting seed production in the seasonally dry habitat.
The complementation test for the candidate gene of DF6, Hd1, has the following
implications for the evolution of photoperiod insensitivity of O. nivara. The failure to rescue
photoperiod sensitivity of O. nivara by overexpressing the O. rufipogon gene, which is at least
partially dominant over the O. nivara allele, is consistent with the finding that there are two other
QTL of larger effect on photoperiod sensitivity. But we cannot rule out the possibility that the
lack of phenotypic effect was due to the 35S promoter driving overexpression, because it was
shown that in certain rice cultivars Hd1 complementation with its endogenous promoter delayed
flowering under long-day conditions (Yano et al., 2000). Nevertheless, it would not be
surprising if Hd1 together with the O. rufipogon alleles at DF1 and DF9 are required to make O.
rufipogon an obligated short-day plant. Interestingly, epistatic interactions have been detected
between any two of the three flowering time QTL (Table 1.2), suggesting that they may function
in the same or related pathways. The number and type of mutations selected at these three loci
that led to the evolution of a day neutral plant from an obligated short-day plant is of great
interest and can be investigated through molecular cloning of DF1, DF6, and DF9 followed by
functional and ecological analyses of their influence on flowering time.
Previous studies of insect-pollinated plants have elucidated the genetic basis for the
adaptive evolution of mating systems (Fenster & Ritland, 1994; Goodwillie, 1999). QTL of
small to moderate effect were identified for floral traits associated with the differentiation of

19

selfing and outcrossing mating systems between Mimulus species (Lin & Ritland, 1997; Fishman
et al., 2002). Similarly, QTL of only moderate effect were found for the evolution of floral
morphology associated with self-fertilization in Leptosiphon (Goodwillie et al., 2006).
However, QTL of relative large effect were found underlying floral morphology leading to
higher outcrossing rates in wild tomatoes (Georgiady et al., 2002).
For wind pollinated, self-compatible Oryza species, outcrossing rates were shown to be
closely correlated with floral morphology (Oka & Morishim.H, 1967). Oryza rufipogon has
larger anthers containing more pollen grains than O. nivara, consistent with the scenario that
higher outcrossing rates are associated with higher male allocation of reproductive resources
(Charlesworth & Charlesworth, 1981). The panicles of O. rufipogon are exserted from the flag
leaves and open in shape, promoting pollen dispersal by wind (Fig. 1.1). In contrast, O. nivara
has compact panicles that are partially enveloped by the flag leaves during pollination, reducing
the chance of cross-pollination.
QTL identified for these panicle and floral traits had small to moderate effects that
explained 4.9–10.5% of the phenotypic variance for two populations combined (Table 1.1). This
adds another example, especially for wind-pollinated plants, to the growing body of empirical
data suggesting that QTL of small to-moderate effect are often responsible for the evolution of
self-fertilization through modification of floral morphology.
The differentiation between annual and perennial life histories is common in flowering
plants and is often associated with traits influencing vegetative propagation and reproductive
allocation. Oryza rufipogon inhabits deep-water swamps that support a rather closed cover of
vegetation all year round. Because of relatively high seedling mortality in this type of habitat,
perennials capable of vegetatively propagating are favored. In seasonally dry habitats in which

20

mature plants cannot survive through the dry season, annuals with high reproductive allocation
are selectively advantageous. Field investigations confirmed that the majority of young tillers of
O. rufipogon grew from underground stems produced in the previous growing season, and all
individuals of O. nivara arose from seeds (Sano et al., 1980).
Previous genetic studies of annual and perennial differentiation analyzed either multiple
traits related to life history or a key trait, such as rhizome formation, responsible primarily for
vegetative propagation or overwinter survival. They found that multiple QTL were involved in
both cases (Hu et al., 2003; Westerbergh & Doebley, 2004; Hall et al., 2006). In this study, we
grew the mapping populations for 2 years to allow the direct evaluation of annual versus
perennial life form. All F2 plants survived during the 2-year period of experiment. There was
no typical annual plant as the O. nivara parent found in either population. Given that the
mapping parents showed the annual and perennial differentiation under the same growing
conditions and the perennial habit is dominant in F1, the lack of segregation of the recessive
annual habit suggests that the differentiation of annual and perennial life history, as evaluated by
longevity, has complex genetic control. It may be determined by multiple traits associated with
resource acquisition and allocation, such as tiller length, spikelet number, grain weight, and
flowering duration. A total of 20 QTL were identified for these traits in the combined analysis
(Table 1.1).
Oryza rufipogon reproduces largely through vegetative propagation by producing
horizontal tillers up to 5-m long, which form new stems and adventitious roots at the nodes (Oka
& Morishima, 1967; Xie et al., 2001). In the greenhouse, O. rufipogon grew longer tillers with
one to two more leaves than O. nivara before they flowered. Oryza nivara produces a larger
number of seeds that weigh more than those of O. rufipogon. Thus, each tiller of O. rufipogon

21

can potentially yield more photosynthetic product but allocate a less amount of it to seed
production. As a consequence, the perennial species stores much of the resource to support
continuous vegetative propagation.
Furthermore, O. rufipogon was able to re-establish a vegetative growth phase after
flowering in each growing season. In contrast, O. nivara was unable to halt flowering once
initiated. Continuous flowering in O. nivara maximizes resource allocation to seed production
until the plant is exhausted and dies. Although the developmental mechanisms underlying this
trait are unknown, the loss of photoperiod sensitivity might have contributed to continuous
flowering in O. nivara. Being photoperiod sensitive, the younger tillers of O. rufipogon were
unable to flower from February to September while the day length was increasing. Taken
together, the annual habit of O. nivara may have evolved as a result of a series of phenotypic
modifications that shifted resource allocation from sustaining continuous vegetative growth to
maximizing seed production.
It is noteworthy that 83.3% of the QTL alleles of O. nivara act in the direction of the
phenotypic evolution from O. rufipogon to O. nivara in the combined analyses. For flowering
time and duration, all six QTL act in the same direction to reduce photoperiod sensitivity and
promote continuous flowering. Similarly, all seven QTL underlying mating system traits,
panicle exsertion, panicle shape, and anther length, act in the same direction to potentially
improve selfing rates. For the remaining three traits, tiller length, spikelet number per panicle,
and grain weight, which are related to resource acquisition and allocation, a total of 17 QTL are
identified and 12 act in the direction of phenotypic evolution of O. nivara.
The result is comparable to that of a QTL analysis of domestication transitions from O.
nivara to an indica rice cultivar, in which the same O. nivara accession was used as the mapping

22

parent (Li, CB et al., 2006a). It was found that 81% of QTL alleles of the indica cultivar acted to
improve the domestication traits, consistent with the notion that the origin of cultivated rice was
driven by strong directional artificial selection that should result in the fixation of favorable
mutations largely in the same direction. Although the number of QTL detected in this study is
too small for most of traits to provide enough power for a sign test, the finding that the
proportion of total QTL acting in the evolutionary direction of O. nivara is even higher than that
found in the case of rice domestication supports the hypothesis that the origin of O. nivara was
driven by directional natural selection moving the ancestral populations toward a new fitness
optimum (Orr, 1998; Rieseberg et al., 2002; Carbone et al., 2005).
Colocalization of QTL for different traits may be an indication of pleiotropic mutations
or mutations of linked genes that form an adaptive gene complex. Of 30 QTL identified in the
combined analysis, 15 or 50% have their 1-LOD supporting intervals overlapped with at least
that of another QTL. But we failed to detect any significant clustering of the QTL. This degree
of colocalization is not nearly as striking as was found for rice domestication where 70% of QTL
for all morphological traits related to grain yield were colocalized and two large QTL clusters
were clearly present on chromosomes 4 and 7 (Li, CB et al., 2006a). During rice domestication,
fixation of pleiotropic mutations or tightly linked mutations under strong artificial selection
could have warranted rapid modifications in plant morphology for higher grain yield. We did
not find similar evidence suggesting that such cohorts of traits were under natural selection
during the adaptation of O. nivara.
Another potentially meaningful comparison between the origin O. nivara and
domestication of rice concerns the dominant effect of QTL alleles. The previous analysis
showed that the indica cultivar fixed 16 recessive and partially recessive alleles and only four

23

partially dominant alleles when domesticated from O. nivara (Li, CB et al., 2006a). In this
study, the combined analysis showed that there was nearly the equal number of dominant and
recessive alleles fixed in O. nivara. The difference may be attributed to the different mating
systems of the ancestral species.
Although understanding the history of rice domestication has been a challenging
evolutionary problem by itself, certain consensus began to emerge from recent molecular genetic
and phylogenetic studies. That is, two major types of rice cultivars, namely indica and japonica,
were domesticated from wild populations with divergent genomic backgrounds, and both O.
nivara and O. rufipogon (sometimes treated as two ecotypes of O. rufipogon in domestication
literature) contributed, probably to different extent, to the domesticated gene pool (Londo et al.,
2006; Sang & Ge, 2007a; Sang & Ge, 2007b; Sweeney & McCouch, 2007). Oryza nivara may
have served as the primary genome donor for the indica cultivar (Li, CB et al., 2006b). Thus, the
progenitor of indica rice could be self-fertilized, whereas O. nivara had an outcrossing O.
rufipogon-like ancestor. In this regard, the more abundant dominant mutations found in O.
nivara than in the indica cultivar could be due to the fact that dominant mutations were more
likely to be fixed in the outcrossing ancestral populations of O. nivara. Nevertheless, a total of
11 recessive alleles found in the combined analysis of O. nivara and O. rufipogon divergence are
still larger in number than 10 dominant alleles. The fixation of relatively abundant recessive
mutations in O. nivara can be explained by several reasons. First, beneficial mutations might
have preexisted in the ancestral populations, in which recessiveness would allow these
previously deleterious alleles to persist at the mutation selection balance in the outcrossing
ancestor. The preexisting recessive mutations could be quickly fixed in response to a sudden
environmental change (Orr & Betancourt, 2001). Second, the increase in selfing rates during the

24

evolution of O. nivara might have facilitated the fixation of at least partially recessive mutations
as they occurred along the adaptive walk (Charlesworth, 1992). Finally, recessive mutations
might have occurred more frequently than dominant ones in the Oryza species as suggested by
the study of rice domestication (Li, CB et al., 2006b).
The effect sizes of the QTL from the combined analysis follow roughly an exponential
distribution (Fig. 1.2). That there are fewer QTL with effect sizes 0–5% than those with 5–10%
effect is most likely because many QTL with small effect of 0–5% remain undetected given the
mapping population size of less than 400 individuals (Beavis, 1998). This result thus fits the
prediction of Orr’s (1998a) population genetic model of adaptation. The similar distribution has
been observed in other cases of plant adaptation under natural or artificial selection (Tanksley,
1993; Bradshaw et al., 1998; Doganlar et al., 2002; Doebley, 2004; Lauter et al., 2004). There
were also examples in which QTL effect did not follow an exponential distribution due to the
lack of large effect QTL (Burke et al., 2002; Fishman et al., 2002; Lexer et al., 2005). This
might be explained by several possible reasons including the fixation of standing mutations of
relative small effect, adaptation following a moving optimum such as a cline of intermediate
habitats, and weak selection (Orr, 2005). The generalization of empirical support for the
theoretical models of adaptation still awaits the accumulation of a larger body of genetic data.

25

Figure 1.2. Distribution of QTL Effect Size.
Distribution of magnitude of effect of QTL identified in the combined analysis of F2
mapping populations.

The consistency of our results with Orr’s model of adaptation is not surprising because
the adaptive evolution of O. nivara does seem to satisfy the basic assumptions of the model.
That is, when a population is distantly off its adaptive optimum due to drastic environmental
changes, an adaptive walk begins and moves the population progressively toward a new
optimum. This is likely the case for the wild rice species. The O. rufipogon-like ancestor of O.
nivara already adapted to the deep-water habitat could have been dispersed to the seasonally dry
26

habitat, or experienced severe drought as the original habitat dried up. Given the abundance of
seasonal habitats in regions with a typical monsoon season, such sharp environmental changes
would have repeatedly occurred and consequently imposed strong directional selection for the
origin of O. nivara.
In this regard, Orr’s model also predicts that initial mutations fixed during adaptation
tend to have large phenotypic effect because they are less likely to overshoot the optimum than
those fixed later when the population gets closer to the optimum. The QTL of the largest effect
detected in this study is DF1, which explains 36.5% of the phenotypic variance for flowering
time. This implies that the loss of photoperiod sensitivity could be an early step for the adaptive
evolution of O. nivara. Early flowering in the seasonally dry habitat is no longer deleterious, as
it would have been for the perennial ancestor living in the deep water habitat. Conversely, the
loss of photoperiod sensitivity could have substantially improved seed production and
consequently the fitness of ancestral populations of O. nivara in the seasonal habitat.
Testing this hypothesis requires molecular cloning of DF1 to determine whether the QTL
indeed harbors a mutation of large effect (Phillips, 2005; Mitchell-Olds et al., 2007). Further,
the fitness effect of the QTL needs to be confirmed in the natural habitats using a near-isogenic
line (Bradshaw & Schemske, 2003) that contains the O. nivara allele of DF1 in the genomic
background of O. rufipogon. Both experiments are feasible especially with the availability of
rice genome sequences. The genomic era of biology should soon witness rigorous experimental
tests of the origin of species under natural selection.

27

Acknowledgments
We are grateful to the anonymous reviewers and T. Juenger for their valuable comments and
suggestions that helped improve the previous versions of the manuscript. The research was
supported by the National Science Foundation of USA.

28

LITERATURE CITED

29

LITERATURE CITED

Barrett SCH, Harder LD, Worley AC. 1996. The comparative biology of pollination and
mating in flowering plants. Philosophical Transactions of the Royal Society of London
Series B-Biological Sciences 351(1345): 1271-1280.
Barton NH, Keightley PD. 2002. Understanding quantitative genetic variation. Nature Reviews
Genetics 3(1): 11-21.
Beavis WD 1998. QTL analyses: power, precision, and accuracy. In: Paterson AH ed. Molecular
Dissection of Complex Traits. Boca Raton, FL: CRC Press, 145-162.
Bradshaw HD, Otto KG, Frewen BE, McKay JK, Schemske DW. 1998. Quantitative trait
loci affecting differences in floral morphology between two species of monkeyflower
(Mimulus). Genetics 149(1): 367-382.
Bradshaw HD, Schemske DW. 2003. Allele substitution at a flower colour locus produces a
pollinator shift in monkeyflowers. Nature 426(6963): 176-178.
Burke JM, Tang S, Knapp SJ, Rieseberg LH. 2002. Genetic analysis of sunflower
domestication. Genetics 161(3): 1257-1267.
Carbone MA, Llopart A, DeAngelis M, Coyne JA, Mackay TFC. 2005. Quantitative trait loci
affecting the difference in pigmentation between Drosophila yakuba and D. santomea.
Genetics 171(1): 211-225.
Charlesworth B. 1992. EVOLUTIONARY RATES IN PARTIALLY SELF-FERTILIZING
SPECIES. American Naturalist 140(1): 126-148.
Charlesworth D, Charlesworth B. 1981. ALLOCATION OF RESOURCES TO MALE AND
FEMALE FUNCTIONS IN HERMAPHRODITES. Biological Journal of the Linnean
Society 15(1): 57-74.
Colosimo PF, Hosemann KE, Balabhadra S, Villarreal G, Dickson M, Grimwood J,
Schmutz J, Myers RM, Schluter D, Kingsley DM. 2005. Widespread parallel evolution
30

in sticklebacks by repeated fixation of ectodysplasin alleles. Science 307(5717): 19281933.
Cremer F, Coupland G. 2003. Distinct photoperiodic responses are conferred by the same
genetic pathway in Arabidopsis and in rice. Trends in Plant Science 8(9): 405-407.
Doebley J. 2004. The genetics of maize evolution. Annual Review of Genetics 38: 37-59.
Doebley J, Stec A, Hubbard L. 1997. The evolution of apical dominance in maize. Nature
386(6624): 485-488.
Doebley JF, Gaut BS, Smith BD. 2006. The molecular genetics of crop domestication. Cell
127(7): 1309-1321.
Doganlar S, Frary A, Daunay MC, Lester RN, Tanksley SD. 2002. Conservation of gene
function in the Solanaceae as revealed by comparative mapping of domestication traits in
eggplant. Genetics 161(4): 1713-1726.
Fenster CB, Ritland K. 1994. QUANTITATIVE GENETICS OF MATING SYSTEM
DIVERGENCE IN THE YELLOW MONKEYFLOWER SPECIES COMPLEX.
Heredity 73: 422-435.
Fisher RA. 1930. The genetical theory of natural selection. Chicago, IL: University of Chicago
Press.
Fishman L, Kelly AJ, Willis JH. 2002. Minor quantitative trait loci underlie floral traits
associated with mating system divergence in Mimulus. Evolution 56(11): 2138-2155.
Frary A, Nesbitt TC, Grandillo S, van der Knaap E, Cong B, Liu JP, Meller J, Elber R,
Alpert KB, Tanksley SD. 2000. fw2.2: A quantitative trait locus key to the evolution of
tomato fruit size. Science 289(5476): 85-88.
Ge S, Sang T, Lu BR, Hong DY. 1999. Phylogeny of rice genomes with emphasis on origins of
allotetraploid species. Proceedings of the National Academy of Sciences of the United
States of America 96(25): 14400-14405.

31

Georgiady MS, Whitkus RW, Lord EM. 2002. Genetic analysis of traits distinguishing
outcrossing and self-pollinating forms of currant tomato, Lycopersicon pimpinellifolium
(Jusl.) Mill. Genetics 161(1): 333-344.
Goodwillie C. 1999. Multiple origins of self-compatibility in Linanthus section Leptosiphon
(Polemoniaceae): Phylogenetic evidence from internal-transcribed-spacer sequence data.
Evolution 53(5): 1387-1395.
Goodwillie C, Ritland C, Ritland K. 2006. The genetic basis of floral traits associated with
mating system evolution in Leptosiphon (Polemoniaceae): An analysis of quantitative
trait loci. Evolution 60(3): 491-504.
Hall MC, Basten CJ, Willis JH. 2006. Pleiotropic quantitative trait loci contribute to population
divergence in traits associated with life-history variation in Mimulus guttatus. Genetics
172(3): 1829-1844.
Hayama R, Yokoi S, Tamaki S, Yano M, Shimamoto K. 2003. Adaptation of photoperiodic
control pathways produces short-day flowering in rice. Nature 422(6933): 719-722.
Hoekstra HE, Hirschmann RJ, Bundey RA, Insel PA, Crossland JP. 2006. A single amino
acid mutation contributes to adaptive beach mouse color pattern. Science 313(5783): 101104.
Hu FY, Tao DY, Sacks E, Fu BY, Xu P, Li J, Yang Y, McNally K, Khush GS, Paterson AH,
Li ZK. 2003. Convergent evolution of perenniality in rice and sorghum. Proceedings of
the National Academy of Sciences of the United States of America 100(7): 4050-4054.
Kao CH, Zeng ZB, Teasdale RD. 1999. Multiple interval mapping for quantitative trait loci.
Genetics 152(3): 1203-1216.
Kirkpatrick M. 2000. Reinforcement and divergence under assortative mating. Proceedings of
the Royal Society B-Biological Sciences 267(1453): 1649-1655.
Komatsuda T, Pourkheirandish M, He C, Azhaguvel P, Kanamori H, Perovic D, Stein N,
Graner A, Wicker T, Tagiri A, Lundqvist U, Fujimura T, Matsuoka M, Matsumoto
T, Yano M. 2007. Six-rowed barley originated from a mutation in a homeodomainleucine zipper I-class homeobox gene. Proceedings of the National Academy of Sciences
of the United States of America 104(4): 1424-1429.

32

Konishi S, Izawa T, Lin SY, Ebana K, Fukuta Y, Sasaki T, Yano M. 2006. An SNP caused
loss of seed shattering during rice domestication. Science 312(5778): 1392-1396.
Kuroda Y, Sato YI, Bounphanousay C, Kono Y, Tanaka K. 2005. Gene flow from cultivated
rice (Oryza sativa L.) to wild Oryza species (O-rufipogon Griff. and O-nivara Sharma
and Shastry) on the Vientiane plain of Laos. Euphytica 142(1-2): 75-83.
Lauter N, Gustus C, Westerbergh A, Doebley J. 2004. The inheritance and evolution of leaf
pigmentation and pubescence in teosinte. Genetics 167(4): 1949-1959.
Lexer C, Rosenthal DM, Raymond O, Donovan LA, Rieseberg LH. 2005. Genetics of species
differences in the wild annual sunflowers, Helianthus annuus and H-petiolaris. Genetics
169(4): 2225-2239.
Li CB, Zhou AL, Sang T. 2006a. Genetic analysis of rice domestication syndrome with the
wild annual species, Oryza nivara. New Phytologist 170(1): 185-193.
Li CB, Zhou AL, Sang T. 2006b. Rice domestication by reducing shattering. Science
311(5769): 1936-1939.
Lin JZ, Ritland K. 1997. Quantitative trait loci differentiating the outbreeding Mimulus guttatus
from the inbreeding M-platycalyx. Genetics 146(3): 1115-1121.
Londo JP, Chiang Y-C, Hung K-H, Chiang T-Y, Schaal BA. 2006. Phylogeography of Asian
wild rice, Oryza rufipogon, reveals multiple independent domestications of cultivated
rice, Oryza sativa. Proceedings of the National Academy of Sciences of the United States
of America 103(25): 9578-9583.
Mackay TFC. 2001a. The genetic architecture of quantitative traits. Annual Review of Genetics
35: 303-339.
Mackay TFC. 2001b. Quantitative trait loci in Drosophila. Nature Reviews Genetics 2(1): 1120.
Mauricio R. 2001. Mapping quantitative trait loci in plants: Uses and caveats for evolutionary
biology. Nature Reviews Genetics 2(5): 370-381.

33

McCouch SR, Teytelman L, Xu YB, Lobos KB, Clare K, Walton M, Fu BY, Maghirang R,
Li ZK, Xing YZ, Zhang QF, Kono I, Yano M, Fjellstrom R, DeClerck G, Schneider
D, Cartinhour S, Ware D, Stein L. 2002. Development and mapping of 2240 new SSR
markers for rice (Oryza sativa L.). DNA Research 9(6): 199-207.
Mitchell-Olds T, Schmitt J. 2006. Genetic mechanisms and evolutionary significance of natural
variation in Arabidopsis. Nature 441(7096): 947-952.
Mitchell-Olds T, Willis JH, Goldstein DB. 2007. Which evolutionary processes influence
natural genetic variation for phenotypic traits? Nature Reviews Genetics 8(11): 845-856.
Morishima H, Chang WT, Oka HI. 1961. DIRECTIONS OF DIFFERENTIATION IN
POPULATIONS OF WILD RICE, ORYZA PERENNIS AND O SATIVA F
SPONTANEA. Evolution 15(3): 326-&.
Oka HI. 1976. MORTALITY AND ADAPTIVE MECHANISMS OF ORYZA PERENNIS
STRAINS. Evolution 30(2): 380-392.
Oka HI, Morishim.H. 1967. VARIATIONS IN BREEDING SYSTEMS OF A WILD RICE
ORYZA PERENNIS. Evolution 21(2): 249-&.
Oka HI, Morishima H. 1967. VARIATIONS IN BREEDING SYSTEMS OF A WILD RICE
ORYZA PERENNIS. Evolution 21(2): 249-&.
Orr HA. 1998. The population genetics of adaptation: The distribution of factors fixed during
adaptive evolution. Evolution 52(4): 935-949.
Orr HA. 2005. The genetic theory of adaptation: A brief history. Nature Reviews Genetics 6(2):
119-127.
Orr HA, Betancourt AJ. 2001. Haldane's sieve and adaptation from the standing genetic
variation. Genetics 157(2): 875-884.
Phillips PC. 2005. Testing hypotheses regarding the genetics of adaptation. Genetica 123(1-2):
15-24.

34

Protas M, Tabansky I, Conrad M, Gross JB, Vidal O, Tabin CJ, Borowsky R. 2008. Multitrait evolution in a cave fish, Astyanax mexicanus. Evolution & Development 10(2): 196209.
Rieseberg LH, Raymond O, Rosenthal DM, Lai Z, Livingstone K, Nakazato T, Durphy JL,
Schwarzbach AE, Donovan LA, Lexer C. 2003. Major ecological transitions in wild
sunflowers facilitated by hybridization. Science 301(5637): 1211-1216.
Rieseberg LH, Widmer A, Arntz AM, Burke JM. 2002. Directional selection is the primary
cause of phenotypic diversification. Proceedings of the National Academy of Sciences of
the United States of America 99(19): 12242-12245.
Sang T, Ge S. 2007a. Genetics and phylogenetics of rice domestication. Current Opinion in
Genetics & Development 17(6): 533-538.
Sang T, Ge S. 2007b. The puzzle of rice domestication. Journal of Integrative Plant Biology
49(6): 760-768.
Sano Y, Morishima H, Oka HI. 1980. INTERMEDIATE PERENNIAL-ANNUAL
POPULATIONS OF ORYZA-PERENNIS FOUND IN THAILAND AND THEIR
EVOLUTIONARY SIGNIFICANCE. Botanical Magazine-Tokyo 93(1032): 291-305.
Simons KJ, Fellers JP, Trick HN, Zhang ZC, Tai YS, Gill BS, Faris JD. 2006. Molecular
characterization of the major wheat domestication gene Q. Genetics 172(1): 547-555.
Slate J. 2005. Quantitative trait locus mapping in natural populations: progress, caveats and
future directions. Molecular Ecology 14(2): 363-379.
Stebbins GL. 1974. Flowering plants: evolution above the species level. Cambridge, MA: The
Belknap Press of Harvard University Press.
Sweeney M, McCouch S. 2007. The complex history of the domestication of rice. Annals of
Botany 100(5): 951-957.
Tanksley SD. 1993. Mapping polygenes. Annual Review of Genetics 27: 205-233.
Temnykh S, DeClerck G, Lukashova A, Lipovich L, Cartinhour S, McCouch S. 2001.
Computational and experimental analysis of microsatellites in rice (Oryza sativa L.):
35

Frequency, length variation, transposon associations, and genetic marker potential.
Genome Research 11(8): 1441-1452.
Van Ooijen JW, Voorips RE. 2001. Joinmap 3.0: software for the calculation of genetic
linkage maps. Wageningen, the Netherlands: Plant Research International.
Vaughan DA. 1994. The wild relatives of rice: a genetic resources bankbook. Philippines:
International Rice Research Institute.
Vaughan DA, Lu B-R, Tomooka N. 2008. The evolving story of rice evolution. Plant Science
174(4): 394-408.
Wang H, Nussbaum-Wagler T, Li BL, Zhao Q, Vigouroux Y, Faller M, Bomblies K,
Lukens L, Doebley JF. 2005. The origin of the naked grains of maize. Nature
436(7051): 714-719.
Weis AE, Kossler TM. 2004. Genetic variation in flowering time induces phenological
assortative mating: Quantitative genetic methods applied to Brassica rapa. American
Journal of Botany 91(6): 825-836.
Westerbergh A, Doebley J. 2004. Quantitative trait loci controlling phenotypes related to the
perennial versus annual habit in wild relatives of maize. Theoretical and Applied
Genetics 109(7): 1544-1553.
Xie ZW, Lu YQ, Ge S, Hong DY, Li FZ. 2001. Clonality in wild rice (Oryza rufipogon,
Poaceae) and its implications for conservation management. American Journal of Botany
88(6): 1058-1064.
Yano M, Katayose Y, Ashikari M, Yamanouchi U, Monna L, Fuse T, Baba T, Yamamoto
K, Umehara Y, Nagamura Y, Sasaki T. 2000. Hd1, a major photoperiod sensitivity
quantitative trait locus in rice, is closely related to the arabidopsis flowering time gene
CONSTANS. Plant Cell 12(12): 2473-2483.
Zhu QH, Ge S. 2005. Phylogenetic relationships among A-genome species of the genus Oryza
revealed by intron sequences of four nuclear genes. New Phytologist 167(1): 249-265.

36

CHAPTER 2
Fine mapping of a major flowering time QTL in Oryza

37

Introduction
Adaptation is a driving force for the evolution of new forms and species. While
significant gains have been made in our understanding of this process, the genetic basis of
adaptation in particular remains poorly understood (Orr, 2005). To uncover the mechanisms of
adaptation we ultimately need to identify the genes underlying adaptive phenotypes and
experimentally demonstrate the effect of these genes on fitness (Barton & Keightley, 2002; Slate,
2005). Recent advances in genomic technologies now allow for the rigorous examination of the
genetic mechanisms of adaptation in a broad range of organisms.
Throughout the last twenty years advancements in Quantitative Trait Loci (QTL)
mapping have provided a reliable means for identifying regions of the genome associated with
trait variation (Tanksley, 1993; Orr, 2001; Barton & Keightley, 2002; Orr, 2005; Mitchell-Olds
& Schmitt, 2006; Mackay et al., 2009). During this period a plethora of QTL studies have been
carried out, with the majority of this work being conducted in genetic model systems or
domesticated plants and animals; although examples from natural systems exist and continue to
accumulate (Bradshaw et al., 1998; Colosimo et al., 2004; Goodwillie et al., 2006; Hall &
Willis, 2006; Bradshaw et al., 2012; Leinonen et al., 2013). QTL mapping has been primarily
useful for characterizing the genetic architecture of phenotypic variation; that is estimating the
number, effect size, and location of loci controlling trait variation. Perhaps the most noteworthy
outcome of this line of research has been the regular identification of large effect QTL (Orr,
2005). This finding is surprising given that it is counter to the long standing dogma that
mutations of small effect control adaptation (Fisher, 1930). In response, new theory has been
developed in order to understand the role of large effect mutations (Orr, 1998; Orr, 2005).

38

Through these efforts it is now well established that loci of large effect can indeed influence trait
variation (Barton & Keightley, 2002; Orr, 2005; Phillips, 2005; Mitchell-Olds et al., 2007).
QTL mapping has been pivotal in our knowledge of the genetic architecture of
adaptation, by identifying genomic regions associated with trait variation. These regions,
however, can be quite large and may encompass hundreds of genes. Furthermore, it has been
shown that separate mutations within a single gene can have independent effects on phenotypic
differentiation (Rockman, 2012; Linnen et al., 2013). Thus, a comprehensive understanding of
the genetic basis of adaptation requires elucidating the causal genes and mutations that underlie
QTL (Barton & Keightley, 2002; Orr, 2005; Phillips, 2005; Hoekstra & Coyne, 2007; MitchellOlds et al., 2007; Rockman, 2012). Examination of such causal genetic elements can provide
insight into several facets concerning the genetic basis of adaptation, including i.) the types of
genes controlling adaptive variation (e.g. protein coding vs. regulatory genes), ii.) the role of cisregulatory mutations, and iii.) the relative role of new mutations vs. standing genetic variation
(Phillips, 2005; Mackay et al., 2009).
The wild progenitors of cultivated rice, Oryza nivara and Oryza rufipogon, present an
ideal system for studying the mechanisms of adaptation. Oryza nivara and O. rufipogon are
recently diverged sister species that have undergone marked adaptive differentiation to distinct
wetland habitats in the tropics and sub-tropics of Asia (Oka & Chang, 1960; Oka & Chang,
1961; Vaughan, 1994; Zhou et al., 2008; Zheng & Ge, 2010). Oryza rufipogon thrives in
lowland marshes and wetlands with continuous water availability throughout the year; whereas,
O. nivara occurs in ephemeral pools that have standing water throughout the monsoon season
and then desiccate upon the arrival of the dry season (Vaughan, 1994; Kariali et al., 2008; Zhou
et al., 2008; Zheng & Ge, 2010). In addition to their distinct habitat preferences, these sister

39

species differ in their flowering time, mating system and life history. Oryza rufipogon, is
perennial, out-crossing, and flowers late in the year under a short day length. In contrast, O.
nivara is annual, predominately self-fertilized, and flowers earlier under a long day length (Oka
& Chang, 1960; Vaughan, 1994). Phylogenetic analysis has indicated that O. nivara has recently
diverged from a common ancestor resembling O. rufipogon (Barbier et al., 1991; Zhu & Ge,
2005). The direction of evolution in this system is consistent with broader patterns in plant
evolution with annual selfing species being derived from perennial outcrossers (Stebbins, 1974;
Barrett et al., 1996). The life cycle of each species is depicted in Figure 2.1. Germination
occurs at the onset of the rainy season for both species. O. rufipogon is an obligate short day
plant flowering in November and December. After flowering, O. rufipogon persists and survives
throughout the dry season as it inhabits stable wetlands. Conversely, O. nivara flowers after a
period of sufficient vegetative growth under any day length. O. nivara flowers in August and
September, prior to the onset of the dry season. As an annual, O. nivara dies after seed
maturation and seeds remain dormant during the dry season (Fig. 2.1).
Of the traits that differ between these species, flowering time stands out a crucial
adaptation for survival in an O. nivara habitat. That is, O. nivara has adapted to survive in
seasonal habitats by being able to flower and produce seeds prior to desiccation. Kariali et al.
(2008) have monitored populations of these species throughout the growing season in India and
documented that O. nivara habitats dessicate by early October, prior to the O. rufipogon
flowering period. As a strict short day plant, O. rufipogon, would not be able to reproduce prior
to the onset of the dry season in seasonal habitats.
The timing of reproduction is an important stage in the life history of any organism.
Variation in the flowering time of plants is commonly observed in nature, and in many cases it

40

has been demonstrated that this variation has a genetic basis (McIntyre & Best, 1978; Marquis,
1988; Van Dijk & Hautekeete, 2007; Wilczek et al., 2009). Additionally, there is substantial
evidence that flowering time is adaptive, as has been revealed through reciprocal transplant
experiments (Hall & Willis, 2006), the observation of latitudinal and altitudinal clines in
flowering time (Lacey, 1988; Kalisz & Wardle, 1994; Olsson & Ågren, 2002), and contemporary
studies of selection (Munguia-Rosas et al., 2011). Plants that do not match their flowering time
to the biotic and abiotic environment can incur severe fitness consequences. For these reasons,
coupled with the relative ease of measuring, flowering time has likely received more attention
than any other adaptive trait.
Furthermore, the molecular genetic pathways underlying the transition from a vegetative
state to flowering have been well characterized (Simpson & Dean, 2002; Amasino, 2010). Many
plants initiate flowering in response to environmental cues. For example, the photoperiod and
vernalization (prolonged exposure to cold temperatures) pathways integrate their respective
environmental signals to promote flowering. Moreover, plants can flower in response to
endogenous signals including the gibberellin and autonomous pathways (Simpson & Dean,
2002). The molecular pathways controlling flowering have predominantly been characterized in
the model plant, Arabidopsis thaliana. In Arabidopsis over 100 genes have been delineated as
part of the flowering time network and many of the core genes are conserved across plant species
(Kardailsky et al., 1999; Kobayashi et al., 1999; Böhlenius et al., 2006; Tamaki et al., 2007; Xia
et al., 2012). While the molecular underpinnings have been well described, only a subset of
these genes have been shown to be associated with natural variation for flowering time
(Stinchcombe et al., 2004; Strange et al., 2011). Although Arabidopsis is considered a genetic
model system, its biology is surely not representative of all angiosperms. Arabidopsis is a long

41

day plant with flowering being promoted by increasing day length as well as vernalization
(simpson and dean). In contrast, rice is a facultative short day plant and does not require
vernalization. In general, rice and Arabidopsis share the same basic genetic machinery for
flowering (Yano et al., 2000; Hayama et al., 2002; Kojima et al., 2002; Hayama et al., 2003)
however, there are key distinctions that provide flowering under short day lengths (Hayama et
al., 2003; Komiya et al., 2009; Takahashi et al., 2009). For example in both rice and
Arabidopsis the major floral regulator CONSTANS (CO), or HEADING DATE1 (Hd1) in rice, is
controlled by the circadian clock with increased expression under long days (Yano et al., 2000;
Izawa et al., 2002). In Arabidopsis the increased expression of CO during long day lengths
promotes FLOWERING LOCUS T (FT) expression, which allows for flowering (Kardailsky et
al., 1999). Conversely, in rice increased expression of Hd1 under long days inhibits expression
of the FT orthologue (Hd3a), and thereby flowering (Hayama et al., 2003). While many
flowering time genes are conserved between rice and Arabidopsis, there are flowering time genes
that they do not share (Izawa, 2007). For example, the MADS-box gene EARLY HEADING
DATE 1 (Ehd1) activates Hd3a independently from Hd1 (Takahashi et al., 2009). Additionally,
this gene has no clear orthologue in Arabidopsis.

With an available genome sequence and

abundant molecular genetic tools, rice has emerged as a model for understanding the basic
framework for flowering in short day plants; however, the genetic basis of flowering time
variation has not been as thoroughly explored as in Arabidopsis.
Previously, I have conducted QTL mapping for a suite of traits that differ between O.
rufipogon and O. nivara (Grillo et al., 2009). Three QTL were identified that control flowering
time differentiation in this system. Here I build off this work and focus on the large effect
flowering time QTL on chromosome 1. This QTL explained 36.5% of the variation in flowering

42

time between the focal species and had the largest effect size of any QTL in this mapping study.
The accurate estimation of QTL effect sizes is quite challenging and can be influenced by a
variety of factors, including the size of the mapping population, the amount of recombination in
a given genomic region, and linkage disequilibrium between mutations within a QTL region
(Mackay et al., 2009; Rockman, 2012). The ability to accurately quantify QTL effect sizes is
critical for our understanding of genetic models of adaptation. Near-isogenic lines (NILs)
provide an empirical means to demonstrate the phenotypic effect of an individual QTL by
introgressing it into a background genome (Bradshaw & Schemske, 2003; McKay et al., 2003).
The goals of the present work are to 1) create a near isogenic line to determine the effect on
flowering time of O. nivara alleles at this QTL, 2) fine map this QTL region to identify the
candidate causal gene, and 3) to determine if a candidate gene differs in coding region sequence.

Materials and Methods
In order to assess the phenotypic effect of individual O. nivara alleles, a near-isogenic
line (NIL) was created through consecutive rounds of backcrossing and marker assisted
selection. To begin this process, an F1 individual was backcrossed to O. rufipogon. Information
on these accessions is provided in Grillo et al. (2009). O. rufipogon was selected as the F1
maternal parent as this mimics the direction of evolution, by which O. nivara alleles arose in an
O. rufipogon like nuclear and cytoplasmic genome. To conduct the cross, O. rufipogon flowers
were manually opened prior to anthesis and were emasculated to prevent selfing. Donor pollen
from O. nivara was collected upon anthesis by shaking anthers over a glass microscope slide or
my fingernail. The slide or fingernail was then brushed against an O. rufipogon stigma. After a
3 month period of dormancy the resulting BC1 seeds were germinated and DNA was extracted

43

from the seedlings with FastDNA kits. These seedlings were scored with a microsatellite marker
(Ch1cM33) which was closely linked to the major flowering time QTL peak (Grillo et al., 2009).
Seedlings were selected that were heterozygous at this locus and were grown to maturity in the
MSU greenhouse. A heterozygous genotype at this locus ensures that the cross was successful
(i.e. not the product of geitonogamous selfing in the O. rufipogon maternal parent). Once the
BC1 plants reached maturity, flowering was induced in growth chambers for another round of
backcrossing to O. rufipogon. This process was continued for 5 total rounds of backcrossing at
which point the background genome was expected be approximately 98.43% homozygous for O.
rufipogon alleles, but be heterozygous at the QTL of interest. At this point a BC5 individual was
self-fertilized to generate NILs that were homozygous for O. rufipogon and O. nivara alleles at
the locus of interest. The BC5 individual was genotyped at 36 microsatellite markers distributed
across the genome to ensure they were homozygous for O. rufipogon alleles. Three markers
were selected on each of the 12 rice chromosomes, including markers that were closely linked to
the QTL peak for the 2 other known flowering time QTL (Grillo et al., 2009). In total it took
approximately 5 years to generate the NILs. The resulting NILs, as well as O. nivara and O.
rufipogon plants, were grown in the greenhouse with supplemental lights to sustain a long day
condition of 16 hours day length. Plants were checked regularly and phenotyped for days to
flower.
After 3 rounds of consecutive backcrossing, a BC2 individual was selected to fine map
the QTL region. In the BC2 generation approximately 87.5% of the background genome is
homozygous for O. rufipogon alleles. This BC2 individual was genotyped with microsatellite
markers at the other 2 QTL on chromosome 6 and 9 to ensure that it was homozygous O.
rufipogon at these loci. The BC2 plant was then grown in the greenhouse and vegetativly cloned

44

extensively, to produce a large number of seeds. The BC2 plants were isolated in a greenhouse
room to prevent outcrossing, and were allowed to flower. Several thousand selfed progeny were
generated through this process that served as a recombinant mapping population.
Microsattelite markers were then designed to flank the QTL region. These markers were
selected by searching the O. sativa reference genome (RGAP7) for simple sequence repeats
(SSRs). Approximately 30 microsatellite markers were generated and markers flanking the QTL
were selected based on their position and ease of deciphering polymorphisms on a 3% agarose
gel. The markers selected encompassed a 9 cM region and were located at approximately 27cM
and 36cM, with the QTL peak being at approximately 31cM.
After a period of dormancy, the selfed BC2 seeds were germinated and screened at the
flanking microsatellites (27 and 36cM). A high throughput DNA extraction protocol developed
by Li et al. (2006) was used to quickly isolate DNA from over 6,000 seedlings. Seedlings were
selected if genetic crossover occurred between the flanking markers. That is, the markers were
heterozygous at one marker and homozygous at the other (i.e. RN-NN). DNA was re-extracted
from recombinant plants using the FastDNA kits, and they were re-genotyped to confirm that
they were true recombinants. Because of the dominant effect of O. rufipogon alleles at this
locus, only O. nivara recombinants (RN-NN) were selected. In O. rufipogon recombinants it
would be difficult to distinguish between the heterozygous and homozygous phenotype. The
recombinant population, along with O. rufipogon, O. nivara, and F1 control plants, were grown
in the greenhouse under a long day condition and phenotyped as described for the NILs.
Single nucleotide polymorphism (SNP) markers were designed to fine map the region.
To identify SNPs, approximately 500bp regions of gene introns were sequenced in the region of
interest. Sequencing primers were designed to amplify the target sequence based on the O.

45

sativa reference genome (RGAP7). Gene regions were PCR amplified using a high-fidelity
polymerase (PfuUltra II Fusion HS DNA Polymerase, Stratagene; La Jolla, California, USA).
Successful amplification was determined on 1.5% agarose gels, and DNA bands were then
purified (Wizard SVGel and PCR Clean-Up, Promega). Sanger sequencing was conducted at the
Michigan State University Genomics Facility. The position of the amplified fragment was
confirmed by a BLAST search of the sequence to the O. sativa genome. The location of the SNP
markers is provided in Figure 2.2.
The coding region of FT-LIKE 1 (FTL1) was sequenced from both O. rufipogon and O.
nivara. Sequencing primers were designed from the FTL1 sequence of O. sativa. Sanger
sequencing was conducted as described previously. The resulting Sequence data was analyzed
with the Staden package Pre-gap4 and Gap4 (http://staden.sourceforge.net/). Polymorphisms
were compared between the focal species in BioEdit
(http://www.mbio.ncsu.edu/bioedit/page2.html). All polymorphisms were confirmed by reamplification and sequencing of the specific region.

Results
Near-isogenic lines (NILs) are useful in demonstrating the phenotypic effect of a single
locus (Bradshaw & Schemske, 2003; McKay et al., 2003). In the present study the NILs that
were homozygous for O. nivara alleles (i.e. NIL-NN), all flowered (n=21); whereas the NILs
with O. rufipogon alleles (NIL-RR) did not flower (n=15; Table 2.1). Similarly the O. nivara
control plants flowered (n=5), whereas O. rufipogon (n=5) did not. While both O. nivara and the
NIL-NN plants flowered, they did so at distinct times. The mean NIL-NN flowering date was

46

38.3 days after the O. nivara mean (Table 2.1). The NIL results provide evidence that this single
locus is sufficient to trigger flowering under a long day length.

Table 2.1. Flowering time of Near-isogenic lines (NILs).
O. nivara and O. rufipogon are parental plants that serve as controls. NIL-NN refers to a NIL
with O. nivara alleles at the flowering time QTL. NIL-RR refers to a NIL with O. rufipogon
alleles at the flowering time QTL.

Plant
O. nivara
O. rufipogon
NIL-NN
NIL-RR

Days to flower
104.0 (2.45)
did not flower
142.3 (6.51)
did not flower

High resolution fine mapping is a classic means for identifying the causal genetic
elements influencing trait variation (Salvi et al., 2002; Li et al., 2006; Mackay et al., 2009). The
flowering time of the recombinant population (n=56), as well as the parental species and F1s is
provided in Figure 2.3. Days to flowering was recorded as the days past germination. The O.
nivara parents flowered between 98-109 days, the F1s between 147-182 days, and the O.
rufipogon parents did not flower under these long day conditions. The recombinant plants
displayed two distinct groups for flowering time dates. An early group flowered between 112
and 126 days and a later group from 138-155days. The early and late groups were scored as
having the O. nivara or F1 phenotype, respectively. Integrating the phenotype and genotype of
the recombinants allowed for fine mapping of the genetic region (Fig. 2.2). Based on this
analysis, the causal gene is located between the SNP markers at 31.0cM and 31.58 cM (physical
positions 6370921 and 6512441, respectively).

47

Within the 141kb sub-region there are 16 genes in O. sativa (Table 2.2). Only one of
these genes stood out as a clear flowering time candidate based on the gene annotation. This
gene, FT-LIKE 1 (FTL1), is a close homolog of the Arabidopsis major flowering regulator, FT.
Analysis of the coding region sequence revealed 5 polymorphisms between O. rufipogon and O.
nivara (Table 2.3). Of these polymorphisms, 3 are nonsynonymous and are candidates for
influencing gene function.

Table 2.2. Genes in the fine mapped region.
Genes in the fine mapped region. Gene ID refers to the accession number available from the
MSU Rice Genome Annotation Project. Gene annotation provides information on the putative
function of the gene.
Gene ID
LOC_Os01g11790
LOC_Os01g11800
LOC_Os01g11810
LOC_Os01g11820
LOC_Os01g11830
LOC_Os01g11840
LOC_Os01g11850
LOC_Os01g11860
LOC_Os01g11870
LOC_Os01g11880
LOC_Os01g11900
LOC_Os01g11910
LOC_Os01g11920
LOC_Os01g11940
LOC_Os01g11946
LOC_Os01g11952

Gene annotation
GDSL-like lipase/acylhydrolase, putative, expressed
retrotransposon protein, putative, unclassified, expressed
powdery mildew resistant protein 5, putative, expressed
expressed protein
oryzain alpha chain precursor, putative, expressed
xylem cysteine proteinase 2 precursor, putative, expressed
expressed protein
DJ-1 family protein, putative, expressed
expressed protein
DJ-1 family protein, putative, expressed
hhH-GPD superfamily base excision DNA repair protein, expressed
basic helix-loop-helix, putative, expressed
retrotransposon protein, putative, unclassified, expressed
osFTL1 FT-Like1 homologous to Flowering Locus T gene; contains
Pfam profile PF01161: Phosphatidylethanolamine-binding protein,
expressed
ABC transporter, ATP-binding protein, putative, expressed
histone-lysine N-methyltransferase ATX5, putative, expressed

48

Discussion
Previously, I have identified three QTL that control flowering time differentiation
between O. rufipogon and O. nivara, located on chromosomes 1, 6 and 9 (Grillo et al., 2009).
Through consecutive rounds of backcrossing I have introgressed O. nivara alleles for the major
flowering time QTL into an O. rufipogon background, resulting in a near-isogenic line. Nearisogenic lines (NILs) provide a novel approach for determining the phenotypic effect of an
individual allele. This procedure is especially relevant given the phylogenetic direction of
evolution. Phylogenetic inference suggests that O. nivara has recently diverged from an O.
rufipogon-like ancestor and that early flowering is a derived trait (Barbier et al., 1991; Zhu &
Ge, 2005). Therefore, the generation of a NIL closely mimics the process of O. nivara alleles
arising in the background of the common ancestor resembling O. rufipogon. By growing these
NILs under a constant long day condition, I have demonstrated that this allele is sufficient at
providing photoperiod insensitive flowering. That is, NILs with O. rufipogon alleles at this locus
do not flower under the long day length, whereas O. nivara alleles do flower. It should be noted
that while the NILs with O. nivara alleles were able to flower under the long day length, they
still flowered considerably later than the O. nivara parents (Table 2.1). This may be explained
by the other QTL on chromosome 6 and 9 that control days to flowering, both of smaller effect
than the QTL under investigation in the present study. O. nivara alleles at these loci likely
promote earlier flowering, after flowering has been achieved by being photoperiod insensitive.
One of these QTL harbors a candidate gene that lends support for this hypothesis. The QTL on
chromosome 6 encompasses Hd1 as a candidate, with O. nivara alleles possessing a loss of
function mutation in this gene (Grillo et al., 2009). Takahashi et al. (2009) found Hd1 to be one
of the few flowering time genes associated with flowering time variation in cultivated rice, with

49

loss of function alleles providing early flowering. The NILs presented here contain O. rufipogon
alleles at the two other QTL which act to delay flowering.

Figure 2.1. Lifecycle of Oryza species.
Yearly precipitation and day-length corresponding with the estimated flowering phenology. Data
from Cuttack, India where CRRI is located (20ºN, 85ºE). Day-length is for the 1st day of the
month (http://www.qpais.co.uk/modb-iec/dayleng.htm). Precipitation is the monthly average
from 1867-1973 (www.worldclimate.com). Plant characteristics estimated from the literature,
discussions with experts, and per observation (Oka & Chang 1960; Vaughn 1994)

50

The results of the NIL experiment suggest that flowering time in this system can be
portioned into two distinct components: 1) the ability to flower given a photoperiod sensitivity
requirement and 2) once flowering is achieved, variation in the days to flower.
A comprehensive understanding of the genetic basis of adaptation requires moving
beyond QTL to identify the causal genes and mutations controlling trait variation (Barton &
Keightley, 2002; Mackay et al., 2009; Rockman, 2012). In addition to revealing the phenotypic
effects of individual loci, NILs are useful starting material for fine mapping QTL regions (Salvi
et al., 2002; Werner et al., 2005; Xia et al., 2012). QTL regions can be quite large and
encompass hundreds of genes. The QTL region under investigation in the present study was
9cM long (2.2Mbp) and includes over 300 genes. By screening a large recombinant fine
mapping population I have substantially narrowed down the QTL region to an approximately
141kbp region that encompasses 16 genes (Table 2.2). The molecular function of these 16 genes
has not been rigorously assessed, although they have been annotated with putative functions
based on their sequence identity and expression patterns (RGAP7). Of these genes there is one
strong candidate, FT-like1 (FTL1) which is homologous to the Arabidopsis gene FT (Doi et al.,
2004; Chardon & Damerval, 2005). In O. sativa this gene is highly expressed in the young
inflorescence and remains expressed in all floral tissues (www.plexdb.org). Furthermore, this
gene is located at approximately 31.5cM which is closely aligned to the position of the QTL
peak LOD, 31cM.

51

Table 2.3. Polymorphisms in the FTL1coding region.
Position, position of the polymorphism in FTL1 based on the O. sativa sequence. AA, amino
acid number of the polymorphism. O. sativa AA, O. rufipgon AA, O. nivara AA, amino acid
abbreviations of these respective accessions.

Position in
gene
72
121
385
449
696

Exon
1
1
1
1
4

AA
#
24
41
129
150
232

R

N

S

A
G
T
A
G

C
A
C
C
C

C
G
C
C
G

O. rufipogon
AA
T
D
L
D
Q

O. nivara
AA
T
N
L
A
H

O. sativa
AA
T
D
L
A
Q

The coding region of FTL1 was sequenced from O. rufipogon and O. nivara, revealing 5
SNPs that varied between them. Of these polymorphisms, three are nonsynonmous resulting in a
shift in amino acid (Table 2.3). All three of these polymorphisms could impact protein function
as the substitutions changed the amino acid polarity and/or charge. Characterizing the role of
protein coding and cis-regulatory variation is of significant interest in studying the genetic basis
of adaptation.

Figure 2.2. Fine Mapping Results.
Fine mapping results. Each line represents an individual recombinant. The numbers at the top
indicate the location in cM of SNP and microsatellite markers used for fine mapping. The color
indicates the genotype of the recombinant at that marker. The blue color represents a
heterozygote and yellow represents an O. nivara homozygote. Green indicates the interval that
recombination occurred in. The phenotype is given as being either that of the O. nivara
homozygote (NN) or the heterozygote (RN).
52

Figure 2.2 (cont’d)
27
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
H
H
H
H
H
H
H
H
H
N
N
N
N
N
N
N
N
N
N
H

29
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
H
H
H
H
N
N
N
N
N
N
N
N
N
H
H
H
H
H
H
H

30.65
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
H
H
H
H
N
N
N
N
N
N
N
N
H
H
H
H
H
H
H
H

31
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
H
H
N
N
N
N
N
N
N
N
H
H
H
H
H
H
H
H
H
H

31.25
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
H
H
H
H
H
H
H
H
H
H
H
53

31.58
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
H
N
N
N
N
N
N
N
N
N
H
H
H
H
H
H
H
H
H
H
N

32.9
N
N
N
N
N
N
N
N
N
N
N
N
H
H
H
H
H
N
N
N
N
N
N
N
N
N
H
H
H
H
H
H
H
H
H
H
N

36
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
N
N
N
N
N
N
N
N
N
H
H
H
H
H
H
H
H
H
H
N

Phenotype
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
NN
RN
RN
RN
RN
RN
RN
RN
RN
RN
RN
RN

Hoekstra and Coyne (2007) highlighted that considerable emphasis has been placed on cisregulatory evolution; however, sufficient data are lacking to make generalizations. In regards to
in cultivated rice, both types of mutations have been documented in impacting flowering time
variation. For example loss of function mutations in the Hd1coding region disrupt protein
function, whereas promoter region variation impacts expression of Hd3a and thereby flowering
time (Takahashi et al., 2009). Functional studies are required to confirm that FTL1 is the causal
gene underlying the QTL, as well as to identify the specific mutation responsible.
FLOWERING LOCUS T, and its homolog Hd3a, have proven to be critical regulators of
flowering time in both rice and Arabidopsis (Kardailsky et al., 1999; Kojima et al., 2002;
Tamaki et al., 2007). This gene is conserved across diverse angiosperms and has been
documented as the mobile florigen signal (Izawa, 2007; Amasino, 2010). Both FT and Hd3a
encode for a very similar phosphatidylethanolamine-binding protein (PEBP) (Chardon &
Damerval, 2005). These proteins are found in both plants and animals, and while their molecular
function is not entirely understood they have been shown to be highly involved with flowering in
angiosperms. The PEBP family in Arabidopsis is quite small and consists of 5 members, all
involved in flowering (Chardon & Damerval, 2005). Conversely, this family is much more
complex in rice and other grass genomes (Doi et al., 2004). The FT-LIKE genes are the largest
PEBP subfamily in rice, consisting of 13 members (FTL1-13) including Hd3a and RICE
FLOWERING LOCUS T1 (RFLT1/FTL3) (Chardon & Damerval, 2005). Despite sharing a
highly conserved protein, PEBP genes can have quite distinct effects on flowering. For example,
in Arabidopsis FT promotes flowering in long day lengths, where its close homolog
(TERMINAL FLOWER1) delays flowering under these conditions (Kobayashi 99).

54

Here, I have provided evidence for FTL1 as a candidate gene within the fine mapped
region. A strong argument can be made for the role of FTL1 in flowering based on its close
sequence similarity to FT/Hd3a including the PEBP domain, and its documented expression
during the onset of flowering in cultivated rice. Moreover, Doi et al. (2004) have observed that
EARLY HEADING DATE 1, a rice gene involved in short-day promotion of flowering, can
induce FTL1 expression in an Hd1-deficient background. In the present work, another FT LIKE
gene, FTL8, was also a potential candidate within the large flowering time QTL on chromosome
1. The recombinant mapping results excluded FTL8 as a causal gene as it does not reside within
the fine mapped region. Interestingly, the remaining flowering time QTL on chromosome 9 also
encompasses an FT LIKE gene (FTL4) (Chardon & Damerval, 2005). Functional experiments
and high resolution mapping are needed to confirm the role of this gene in underlying flowering
time variation at this QTL. Although FT and other members of PEBP family are crucial for
flowering in Arabidopsis, they are not typically associated with variation for flowering time
among natural accessions (Strange et al., 2011). Compared to Arabidopsis, the FT LIKE genes
are more diverse in rice and I have provided evidence for their potential role in governing
flowering time variation in the wild rice species.

55

Figure 2.3. Days to flowering in the recombinant mapping population.

Flowering time is a key trait for survival in the seasonally dry habitats that O. nivara
inhabits. Phylogenetic analysis suggests that the common ancestor of the focal species closely
resembled O. rufipogon by being perennial and photoperiod sensitive (short day) (Barbier et al.,
1991; Zhu & Ge, 2005). It is likely that propagules of the common ancestor frequently dispersed
into seasonally dry habitats along the periphery of wetlands. The dry season begins 2-3 months
prior to the flowering period of O. rufipogon (Kariali et al., 2008). Because of this, propagules
of the common ancestor would be unable to persist in these habitats without a major shift in
flowering time, allowing them to set seed prior to the onset of the dry season. A shift in

56

flowering time would thereby be a necessary initial step in order to adapt to this new habitat.
Under this scenario minor adjustments in flowering time would not increase survival. O.
rufipogon is an obligate short day plant, initiating flowering after an approximately 11 hour day
length. Conversely, O. nivara can flower under any day length provided a sufficient period of
vegetative growth (Oka & Chang, 1960). Thus flowering time divergence between these species
is based on photoperiod sensitivity. Previously I have identified a large effect QTL that controls
days to flowering in this system. The analysis of NILs has revealed that this locus is sufficient at
reducing photoperiod sensitivity to allow flowering under a long day length. Therefore, it is
highly likely that a mutation at this locus was an initial first step in adapting to seasonally dry
habitats. This notion is in agreement with Orr’s model of adaptation that predicts that large
effect mutations occur early on during an adaptive walk(Orr, 1998).
Understanding the genetic basis of adaptation requires identifying the causal genes and
mutations controlling adaptive traits (Mackay et al., 2009; Rockman, 2012). Here, I have
provided strong evidence for FTL1 as a causal gene underlying a major QTL for flowering time,
a key trait in this system. Additionally I have identified nonsynonymous polymorphisms in the
coding region of this gene that could be the causal mutations. Further experiments are required
to validate the role of FTL1 and specific mutations that control phenotypic differentiation
between the focal species. Such experiments could include conducting transgenic
complementation, gene knockdown with VIGS or RNAi, assessing gene expression levels
throughout floral initiation, and thoroughly examining the promoter region for cis-regulatory
variation. Beyond this, variation in this gene could be assessed in numerous populations of each
species. Lastly field experiments are required to confirm the role of flowering time and specific
flowering time loci in influencing fitness.

57

LITERATURE CITED

58

LITERATURE CITED

Amasino R. 2010. Seasonal and developmental timing of flowering. Plant Journal 61(6): 10011013.
Barbier P, Morishima H, Ishihama A. 1991. PHYLOGENETIC-RELATIONSHIPS OF
ANNUAL AND PERENNIAL WILD-RICE - PROBING BY DIRECT DNA
SEQUENCING. Theoretical and Applied Genetics 81(5): 693-702.
Barrett SCH, Harder LD, Worley AC. 1996. The comparative biology of pollination and
mating in flowering plants. Philosophical Transactions of the Royal Society of London
Series B-Biological Sciences 351(1345): 1271-1280.
Barton NH, Keightley PD. 2002. Understanding quantitative genetic variation. Nature Reviews
Genetics 3(1): 11-21.
Böhlenius H, Huang T, Charbonnel-Campaa L, Brunner AM, Jansson S, Strauss SH,
Nilsson O. 2006. CO/FT regulatory module controls timing of flowering and seasonal
growth cessation in trees. Science 312(5776): 1040-1043.
Bradshaw HD, Otto KG, Frewen BE, McKay JK, Schemske DW. 1998. Quantitative trait
loci affecting differences in floral morphology between two species of monkeyflower
(Mimulus). Genetics 149(1): 367-382.
Bradshaw HD, Schemske DW. 2003. Allele substitution at a flower colour locus produces a
pollinator shift in monkeyflowers. Nature 426(6963): 176-178.
Bradshaw WE, Emerson KJ, Catchen JM, Cresko WA, Holzapfel CM. 2012. Footprints in
time: comparative quantitative trait loci mapping of the pitcher-plant mosquito,
Wyeomyia smithii. Proceedings of the Royal Society B-Biological Sciences 279(1747):
4551-4558.
Chardon F, Damerval C. 2005. Phylogenomic analysis of the PEBP gene family in cereals.
Journal of Molecular Evolution 61(5): 579-590.

59

Colosimo PF, Peichel CL, Nereng K, Blackman BK, Shapiro MD, Schluter D, Kingsley
DM. 2004. The genetic architecture of parallel armor plate reduction in threespine
sticklebacks. Plos Biology 2(5): 635-641.
Doi K, Izawa T, Fuse T, Yamanouchi U, Kubo T, Shimatani Z, Yano M, Yoshimura A.
2004. Ehd1, a B-type response regulator in rice, confers short-day promotion of
flowering and controls FT-Iike gene expression independently of Hd1l. Genes &
Development 18(8): 926-936.
Fisher RA. 1930. The genetical theory of natural selection. Chicago, IL: University of Chicago
Press.
Goodwillie C, Ritland C, Ritland K. 2006. The genetic basis of floral traits associated with
mating system evolution in Leptosiphon (Polemoniaceae): An analysis of quantitative
trait loci. Evolution 60(3): 491-504.
Grillo MA, Li CB, Fowlkes AM, Briggeman TM, Zhou AL, Schemske DW, Sang T. 2009.
Genetic architecture for the adaptive origin of annual wild rice, Oryza nivara. Evolution
63(4): 870-883.
Hall MC, Willis JH. 2006. Divergent selection on flowering time contributes to local adaptation
in Mimulus guttatus populations. Evolution 60(12): 2466-2477.
Hayama R, Izawa T, Shimamoto K. 2002. Isolation of rice genes possibly involved in the
photoperiodic control of flowering by a fluorescent differential display method. Plant and
Cell Physiology 43(5): 494-504.
Hayama R, Yokoi S, Tamaki S, Yano M, Shimamoto K. 2003. Adaptation of photoperiodic
control pathways produces short-day flowering in rice. Nature 422(6933): 719-722.
Hoekstra HE, Coyne JA. 2007. The locus of evolution: Evo devo and the genetics of
adaptation. Evolution 61(5): 995-1016.
Izawa T. 2007. Adaptation of flowering-time by natural and artificial selection in Arabidopsis
and rice. Journal of Experimental Botany 58(12): 3091-3097.
Izawa T, Oikawa T, Sugiyama N, Tanisaka T, Yano M, Shimamoto K. 2002. Phytochrome
mediates the external light signal to repress FT orthologs in photoperiodic flowering of
rice. Genes & Development 16(15): 2006-2020.
60

Kalisz S, Wardle GM. 1994. Life-history variation in Campanula americana (Campanulaceae)
- population differentiation. American Journal of Botany 81(5): 521-527.
Kardailsky I, Shukla VK, Ahn JH, Dagenais N, Christensen SK, Nguyen JT, Chory J,
Harrison MJ, Weigel D. 1999. Activation tagging of the floral inducer FT. Science
286(5446): 1962-1965.
Kariali E, Kuanar SR, Mohapatra PK. 2008. Individual tiller dynamics of two wild Oryza
species in contrasting habitats. Plant Production Science 11(3): 355-360.
Kobayashi Y, Kaya H, Goto K, Iwabuchi M, Araki T. 1999. A pair of related genes with
antagonistic roles in mediating flowering signals. Science 286(5446): 1960-1962.
Kojima S, Takahashi Y, Kobayashi Y, Monna L, Sasaki T, Araki T, Yano M. 2002. Hd3a, a
rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of
Hd1 under short-day conditions. Plant and Cell Physiology 43(10): 1096-1105.
Komiya R, Yokoi S, Shimamoto K. 2009. A gene network for long-day flowering activates
RFT1 encoding a mobile flowering signal in rice. Development 136(20): 3443-3450.
Lacey EP. 1988. Latitudinal variation in reproductive timing of a short-lived monocarp, Daucus
carota (Apiaceae). Ecology 69(1): 220-232.
Leinonen PH, Remington DL, Leppala J, Savolainen O. 2013. Genetic basis of local
adaptation and flowering time variation in Arabidopsis lyrata. Molecular Ecology 22(3):
709-723.
Li CB, Zhou AL, Sang T. 2006. Rice domestication by reducing shattering. Science 311(5769):
1936-1939.
Linnen CR, Poh YP, Peterson BK, Barrett RDH, Larson JG, Jensen JD, Hoekstra HE.
2013. Adaptive Evolution of Multiple Traits Through Multiple Mutations at a Single
Gene. Science 339(6125): 1312-1316.
Mackay TFC, Stone EA, Ayroles JF. 2009. The genetics of quantitative traits: challenges and
prospects. Nature Reviews Genetics 10(8): 565-577.

61

Marquis RJ. 1988. PHENOLOGICAL VARIATION IN THE NEOTROPICAL
UNDERSTORY SHRUB PIPER-ARIEIANUM - CAUSES AND CONSEQUENCES.
Ecology 69(5): 1552-1565.
McIntyre GI, Best KF. 1978. STUDIES ON FLOWERING OF THLASPI-ARVENSE L .4.
GENETIC AND ECOLOGICAL DIFFERENCES BETWEEN EARLY-FLOWERING
AND LATE-FLOWERING STRAINS. Botanical Gazette 139(2): 190-195.
McKay JK, Richards JH, Mitchell-Olds T. 2003. Genetics of drought adaptation in
Arabidopsis thaliana: I. Pleiotropy contributes to genetic correlations among ecological
traits. Molecular Ecology 12(5): 1137-1151.
Mitchell-Olds T, Schmitt J. 2006. Genetic mechanisms and evolutionary significance of natural
variation in Arabidopsis. Nature 441(7096): 947-952.
Mitchell-Olds T, Willis JH, Goldstein DB. 2007. Which evolutionary processes influence
natural genetic variation for phenotypic traits? Nature Reviews Genetics 8(11): 845-856.
Munguia-Rosas MA, Ollerton J, Parra-Tabla V, De-Nova JA. 2011. Meta-analysis of
phenotypic selection on flowering phenology suggests that early flowering plants are
favoured. Ecology Letters 14(5): 511-521.
Oka HI, Chang WT. 1960. Survey of variations in photoperiodic response in wild Oryza
species. Bot. Bull. Acad. Sinica 1: 1-14.
Oka HI, Chang WT. 1961. HYBRID SWARMS BETWEEN WILD AND CULTIVATED
RICE SPECIES, ORYZA PERENNIS AND O SATIVA. Evolution 15(4): 418-&.
Olsson K, Ågren J. 2002. Latitudinal population differentiation in phenology, life history and
flower morphology in the perennial herb Lythrum salicaria. Journal of Evolutionary
Biology 15(6): 983-996.
Orr HA. 1998. The population genetics of adaptation: The distribution of factors fixed during
adaptive evolution. Evolution 52(4): 935-949.
Orr HA. 2001. The genetics of species differences. Trends in Ecology & Evolution 16(7): 343350.

62

Orr HA. 2005. The genetic theory of adaptation: A brief history. Nature Reviews Genetics 6(2):
119-127.
Phillips PC. 2005. Testing hypotheses regarding the genetics of adaptation. Genetica 123(1-2):
15-24.
Rockman MV. 2012. THE QTN PROGRAM AND THE ALLELES THAT MATTER FOR
EVOLUTION: ALL THAT'S GOLD DOES NOT GLITTER. Evolution 66(1): 1-17.
Salvi S, Tuberosa R, Chiapparino E, Maccaferri M, Veillet S, van Beuningen L, Isaac P,
Edwards K, Phillips RL. 2002. Toward positional cloning of Vgt1, a QTL controlling
the transition from the vegetative to the reproductive phase in maize. Plant Molecular
Biology 48(5): 601-613.
Simpson GG, Dean C. 2002. Flowering - Arabidopsis, the rosetta stone of flowering time?
Science 296(5566): 285-289.
Slate J. 2005. Quantitative trait locus mapping in natural populations: progress, caveats and
future directions. Molecular Ecology 14(2): 363-379.
Stebbins GL. 1974. Flowering plants: evolution above the species level. Cambridge, MA: The
Belknap Press of Harvard University Press.
Stinchcombe JR, Weinig C, Ungerer M, Olsen KM, Mays C, Halldorsdottir SS,
Purugganan MD, Schmitt J. 2004. A latitudinal cline in flowering time in Arabidopsis
thaliana modulated by the flowering time gene FRIGIDA. Proceedings of the National
Academy of Sciences of the United States of America 101(13): 4712-4717.
Strange A, Li P, Lister C, Anderson J, Warthmann N, Shindo C, Irwin J, Nordborg M,
Dean C. 2011. Major-effect alleles at relatively few loci underlie distinct vernalization
and flowering variation in Arabidopsis accessions. Plos One 6(5).
Takahashi Y, Teshima KM, Yokoi S, Innan H, Shimamoto K. 2009. Variations in Hd1
proteins, Hd3a promoters, and Ehd1 expression levels contribute to diversity of flowering
time in cultivated rice. Proceedings of the National Academy of Sciences of the United
States of America 106(11): 4555-4560.
Tamaki S, Matsuo S, Wong HL, Yokoi S, Shimamoto K. 2007. Hd3a protein is a mobile
flowering signal in rice. Science 316(5827): 1033-1036.
63

Tanksley SD. 1993. Mapping polygenes. Annual Review of Genetics 27: 205-233.
Van Dijk H, Hautekeete N. 2007. Long day plants and the response to global warming: rapid
evolutionary change in day length sensitivity is possible in wild beet. Journal of
Evolutionary Biology 20(1): 349-357.
Vaughan DA. 1994. The wild relatives of rice: a genetic resources bankbook. Philippines:
International Rice Research Institute.
Werner JD, Borevitz JO, Warthmann N, Trainer GT, Ecker JR, Chory J, Weigel D. 2005.
Quantitative trait locus mapping and DNA array hybridization identify an FLM deletion
as a cause for natural flowering-time variation. Proceedings of the National Academy of
Sciences of the United States of America 102(7): 2460-2465.
Wilczek AM, Roe JL, Knapp MC, Cooper MD, Lopez-Gallego C, Martin LJ, Muir CD,
Sim S, Walker A, Anderson J, Egan JF, Moyers BT, Petipas R, Giakountis A,
Charbit E, Coupland G, Welch SM, Schmitt J. 2009. Effects of genetic perturbation
on seasonal life history plasticity. Science 323(5916): 930-934.
Xia ZJ, Watanabe S, Yamada T, Tsubokura Y, Nakashima H, Zhai H, Anai T, Sato S,
Yamazaki T, Lu SX, Wu HY, Tabata S, Harada K. 2012. Positional cloning and
characterization reveal the molecular basis for soybean maturity locus E1 that regulates
photoperiodic flowering. Proceedings of the National Academy of Sciences of the United
States of America 109(32): E2155-E2164.
Yano M, Katayose Y, Ashikari M, Yamanouchi U, Monna L, Fuse T, Baba T, Yamamoto
K, Umehara Y, Nagamura Y, Sasaki T. 2000. Hd1, a major photoperiod sensitivity
quantitative trait locus in rice, is closely related to the arabidopsis flowering time gene
CONSTANS. Plant Cell 12(12): 2473-2483.
Zheng XM, Ge S. 2010. Ecological divergence in the presence of gene flow in two closely
related Oryza species (Oryza rufipogon and O. nivara). Molecular Ecology 19(12): 24392454.
Zhou HF, Zheng XM, Wei RX, Second G, Vaughan DA, Ge S. 2008. Contrasting population
genetic structure and gene flow between Oryza rufipogon and Oryza nivara. Theoretical
and Applied Genetics 117(7): 1181-1189.
Zhu QH, Ge S. 2005. Phylogenetic relationships among A-genome species of the genus Oryza
revealed by intron sequences of four nuclear genes. New Phytologist 167(1): 249-265.
64

CHAPTER 3
Genetic architecture of flowering time differentiation between locally
adapted populations of Arabidopsis thaliana

The following chapter was published in New Phytologist
Grillo MA, Li CB, Hammond M, Wang LJ, Schemske DW. 2013. Genetic architecture of
flowering time differentiation between locally adapted populations of Arabidopsis thaliana. New
Phytologist 197(4): 1321-1331.

65

Abstract
To gain an understanding of the genetic basis of adaptation, we conducted quantitative
trait locus (QTL) mapping for flowering time variation between two winter annual populations
of Arabidopsis thaliana that are locally adapted and display distinct flowering times. QTL
mapping was performed with large (n=384) F2 populations with and without vernalization, in
order to reveal both the genetic basis of a vernalization requirement and that of variation in
flowering time given vernalization. In the Non-Vernalization treatment, none of the Sweden
parents flowered, while all of the Italy parents and 42% of the F2s flowered. We identified three
QTL for flowering without vernalization, with much of the variation being attributed to a QTL
co-localizing with FLOWERING LOCUS C (FLC). In the Vernalization treatment all parents
and F2s flowered, and six QTL of small to moderate effect were revealed, with underlying
candidate genes that are members of the vernalization pathway. We found no evidence for a role
of FRIGIDA in regulating flowering times. These results contribute to a growing body of
evidence aimed at identifying ecologically relevant genetic variation for flowering time in
Arabidopsis, and set the stage for functional studies to determine the link between flowering time
loci and fitness.

Introduction
Throughout a species’ range, populations adapt to local conditions. Understanding the
genetic basis of such local adaptation has been a major goal in evolutionary biology for over a
century. The last decade has witnessed profound advances in molecular genetic approaches for
finding genomic regions associated with adaptive traits, allowing for the rigorous assessment of
this long-standing question (Barton & Keightley, 2002).

66

Plants have developed sophisticated physiological, developmental and genetic
mechanisms to optimize the time of flowering (Bastow & Dean, 2003). Because the transition
from vegetative growth to flowering is crucial for plant reproductive success, plants integrate
different environmental cues to achieve a flowering response that is adapted to local conditions.
The observation that the flowering time of many plant populations varies with latitude or altitude
suggests that this trait contributes to geographic adaptation (Lacey, 1988; Kalisz & Wardle,
1994; Olsson & Ågren, 2002). Hence, understanding the pathways and genetic mechanisms that
contribute to natural variation in flowering time is a central goal of many studies in plant
evolutionary genomics.
The model plant Arabidopsis thaliana (hereafter referred to as Arabidopsis), serves as a
workhorse for the study of plant molecular biology. This species is distributed widely
throughout northern latitudes and exhibits substantial natural variation in flowering time. Two
different approaches have been used to characterize the life history and flowering behavior of
Arabidopsis (Nordborg & Bergelson, 1999). The ecological criterion is based on the timing of
seed germination and flowering; winter annuals germinate in fall, overwinter as rosettes, and
flower in late spring, while summer annuals germinate in spring and flower in mid to late
summer. Based on this criterion, Arabidopsis has a winter annual life history throughout most of
its range, although summer annuals have also been described (Napp-Zinn, 1985; Pigliucci, 2003;
Koornneef et al., 2004; Shindo et al., 2007). Climatic data from the southern edge of the native
range, indicates that while soil temperatures very rarely fall below freezing they do fall below
4°C (Ågren & Schemske, 2012), which is a sufficient temperature for a vernalization response
(Nordborg & Bergelson, 1999; Shindo et al., 2005).

67

In contrast, the physiological criterion classifies plants as winter annuals if vernalization,
i.e. a period of cold temperature at the seed or rosette stage, is required for flowering, and as
summer annuals if plants can flower without vernalization. These categories are also referred to,
respectively, as “early” or “late” flowering (Clarke et al., 1995; Gazzani et al., 2003) because the
vegetative phase is shorter in plants that do not require vernalization. Some populations flower
without vernalization in the lab, and would therefore be classified as summer annuals by the
physiological criterion, but are winter annuals under the ecological criterion (Schemske and
Ågren, pers. comm.). We suggest that the ecological criterion for classification of life history be
adopted universally, because it best reflects the actual timing of growth and flowering in the
field.
The genetic basis of flowering time in Arabidopsis has received considerable attention
(Simpson & Dean, 2002; Sung & Amasino, 2004; Amasino, 2010). Analysis of Arabidopsis
mutants has characterized multiple genetic pathways that regulate the transition from vegetative
to reproductive growth. The photoperiod and vernalization pathway regulate the response to
environmental signals, while the autonomous and gibberellin pathways respond to endogenous
signals and are functionally independent from environmental cues (Simpson & Dean, 2002).
These pathways form a complex network of more than 100 genes that regulate flowering.
However, only a few genes identified from mutant screens have also been associated with natural
variation in flowering time. Of these, FRIGIDA (FRI) and FLOWERING LOCUS C (FLC) are
thought to be major regulators of flowering in natural populations (Burn et al., 1993; Lee et al.,
1993; Clarke & Dean, 1994; Lempe et al., 2005). FLC, a transcription factor encoding a MADsbox protein, represses the expression of other transcription factors promoting flowering
(Michaels & Amasino, 1999). Functional alleles at FRI delay flowering by activating the strong

68

expression of FLC (Johanson et al., 2000). Vernalization represses FLC expression by
epigenetic modification, reducing its sensitivity to FRI and thus promotes flowering (Amasino,
2004).
Molecular analysis of natural accessions has revealed considerable allelic variation for
FRI and FLC, and many mutations are known that cause inactivation of either FRI or FLC. Such
mutants are often significantly associated with early flowering (Johanson et al., 2000; Le Corre
et al., 2002; Gazzani et al., 2003; Lempe et al., 2005). Over 70% of early flowering accessions
contain loss-of-function alleles at the FRI locus (Shindo et al., 2005). In a geographic survey of
the relationship between flowering time and FRI, active FRI alleles were associated with a
latitudinal cline in flowering, but no such cline was detected in accessions with nonfunctional
FRI (Stinchcombe et al., 2004). In the presence of functional FRI alleles, FLC allelic variation
contributes to a latitudinal cline in flowering, and FLC expression is most highly correlated with
flowering time variation (Caicedo et al., 2004; Lempe et al., 2005; Shindo et al., 2005). These
results suggest that FRI and FLC are targets of natural selection (Stinchcombe et al., 2004;
Izawa, 2007). Although FRI and FLC are important determinants of natural variation in
flowering time for some Arabidopsis populations, extensive molecular and genetic analysis of
Arabidopsis accessions has indicated that other loci also contribute to flowering time (Lempe et
al., 2005; Werner et al., 2005a; Werner et al., 2005b; Li et al., 2006)
Quantitative trait locus (QTL) mapping has proven to be a powerful method for
identifying the genetic basis of quantitative trait variation (Tanksley, 1993; Barton & Keightley,
2002; Mitchell-Olds & Schmitt, 2006), and numerous QTL studies have been conducted
examining flowering time variation in Arabidopsis. The majority of these studies include
mapping populations derived from one or both of the laboratory strains Columbia (Col) and

69

Landsberg erecta (Ler), both of which could be classified as summer annuals based on a
physiological criterion alone (Lister & Dean, 1993). Knowing where in the life cycle variation
in flowering time is manifest is a critical step in assessing the mechanisms responsible for
population differences in flowering time. For example, in plants requiring vernalization, the
developmental and physiological pathways that contribute to variation in flowering time among
populations may differ from those of populations that do not require vernalization. It is thus
important to understand the life histories of populations grown under conditions of their native
environments.
Recently, Ågren & Schemske (2012) have presented the results of a multi-year reciprocal
transplant study between Arabidopsis populations originating from Sweden and Italy – the
geographic limits of the native distribution. This experiment is the first of its kind involving
Arabidopsis (Lowry, 2012), and has demonstrated strong adaptive differentiation of these
populations to their source environment. In this study, freezing tolerance and flowering time
were identified as putative adaptive traits conferring local adaptation. At the Italian field site, the
Italy population flowered 33 and 50 days prior to Sweden, in the two years of study; at the
Swedish field site the Italy population flowered 3 days prior to Sweden in both years (Ågren &
Schemske, 2012). Here we report the results of extensive genetic mapping experiments designed
to examine the genetic basis of variation in flowering time differences between these
populations. Given the winter annual life history of the study populations, we employed a two
tiered approach to dissect the genetic basis of flowering time. We first searched for QTL
required for plants to flower in the absence of vernalization, and then identified QTL that
contribute to variation in flowering following vernalization. In this way, we are able to decouple
flowering time variation from a vernalization requirement.

70

Our study addressed the following questions: 1) what is the number, location, and
magnitude of effect for QTL contributing to flowering time? 2) What are the genetic regions
responsible for differences in vernalization response and do these contribute to flowering time
variation after vernalization? 3) Do QTL co-localize with FRI and FLC as is seen in other
studies? 4) Do candidate genes underlying QTL differ between parental populations in coding
region sequence?

Materials and Methods
This experiment is part of an ongoing effort to understand the mechanisms of adaptation
in natural populations of Arabidopsis. In this study, we used one population collected from
central Italy, Castelnuevo di Porto (N 42°07’ E 12°29’), and one from north-central Sweden,
Rödåsen (N 62°48’ E 18°12’), as mapping parents. These individuals are derived from the same
maternal lines utilized by Ågren & Schemske (2012) -- additional information on these
populations is presented there. Both the Italy and Sweden populations are winter annuals; plants
germinate from late August to early September in Sweden and from late October to midNovember in Italy, and in both regions overwinter as rosettes. The flowering phenology differs
between the two areas, with the Italy population flowering and fruiting much earlier (MarchApril) than the Sweden populations (May-June) (Ågren & Schemske, 2012). A plant from Italy
was utilized as the maternal parent and was crossed to Sweden to produce an F1 generation. A
large F2 population was then generated by autonomous self-pollination of an F1.
Arabidopsis seeds were cold stratified for 4 days at 6°C and then were sown on nutrient
agar and then allowed to germinate in a growth chamber under long-day conditions (16h).
Twenty-day old seedlings were transplanted into 2 inch pots and were allowed to grow for an

71

additional 4 weeks (total age ~ 7 week), at which time they were randomly assigned to Nonvernalization or Vernalization or treatments. The Non-vernalization treatment proceeded at 22°C
under a long-day condition with 16 hours of light (total PAR= ~ 125 µmol m-2 s-1). Plants in the
Vernalization treatment experienced 5 weeks of vernalization at 6°C with a short day condition
of 10 hours of light (Total PAR= ~ 50 µmol m-2 s-1). After this 5 week period, the conditions
were switched back to long days as described above in the Non-vernalization treatment. For
each treatment there were 50 plants from each parental population and 384 F2 hybrids
The dates of bolting and of first flowering were recorded for each plant. The date of
bolting was defined as the date the inflorescence extended beyond the rosette. Date of first
flowering was the date on which the first flower opened, defined as the stage at which the petals
reflexed to expose the stigma. In the Non-vernalization treatment, plants were coded as 1 if they
flowered and 0 if they did not flower. For the Vernalization treatment, plants were coded as
above, and the date that plants flowered was recorded. For data analysis, plants were scored as
the number of days relative to the first plant that bolted or flowered.
Simple sequence repeats (SSR) were utilized for genotyping the mapping population. 99
SSR primer pairs were obtained from the TAIR website and another 202 SSR primer pairs were
designed after searching for microsatellites in the Arabidopsis genome of the Columbia
accession. The total, 301 SSR loci, were screened for polymorphisms between the Italy and
Sweden mapping parents. 103 polymorphic markers were identified; of these 71 were used for
segregation analysis in 96 F2 plants. Segregation distortion was examined by chi-square tests (P
< 0.05), to identify markers that deviated from an expected 1:2:1 ratio (Lu et al., 2002). In the
end, 62 non-distorted markers were selected to construct the linkage map, providing even
coverage of the Arabidopsis genome. DNA was isolated using FastDNA kits (Q-BIOgene,

72

Carlsbad, CA). SSRs were amplified by polymerase chain reaction (PCR) and were visualized
on 3 % agarose gels stained with ethidium bromide.
The genetic map was constructed based on 62 SSR markers and 768 F2 plants from the
combined Vernalization and Non-vernalization treatment. The linkage analysis was performed
based on 706 F2 individuals using JoinMap 3.0 (Van Ooijen & Voorips, 2001). The linkage
groups were separated using the highest logarithm of odds (LOD), 10.0. Recombination values
were converted to genetic distances using the Kosambi mapping function. Five linkage groups
were obtained corresponding to the five Arabidopsis chromosomes. For QTL analysis the
mapping populations included 384 F2 plants in both the Vernalization and Non-vernalization
treatments. In the Vernalization treatment QTL analysis was conducted with the raw flowering
time data and quantile normalized flowering times. Both the raw and transformed data yielded
essentially the same results, with the same number of QTL being identified and the effect sizes
and peak LOD positions changing only slightly. Here we present QTL results with the raw
flowering time data. QTL analysis was performed with the R/qtl package (Broman et al., 2003).
We used a Haley-Knott regression (Haley & Knott, 1992) in the "stepwiseqtl" function to detect
QTL, with the normal and binary model being utilized in the Vernalization and Nonvernalization treatments, respectively. Genome-wide significance (P =0.01) thresholds for QTL
detection were determined with 10,000 permutations. For all QTL, we determined 2-LOD
confidence intervals, as well as estimates of additive and dominance allelic effects. Epistasis
between QTL was tested using two-way analysis of variance (ANOVA) (Westerbergh &
Doebley, 2002; Morgan & Mackay, 2006). The markers closest to the LOD peaks of QTL were
tested for interaction.

73

A comprehensive list of potential candidate genes was compiled by searching TAIR 10
(Brachi et al., 2010). Candidates were selected that contain “flowering” or “vernalization” in
their name and/or description. This list includes a total of 165 unique genes. To reveal
candidate genes underlying QTL, the sequence of SSR markers flanking QTL were located
within the Columbia genome. The locations of the flowering/vernalization genes were located
in the Columbia genome using the Chromosome Map Tool, available on TAIR. Genes that lie
within markers flanking QTL were considered potential candidates. Candidate genes largely fell
into three categories: 1) Genes that are involved in basic biological processes with
mutants/transgenics showing a suite of abnormal phenotypes, including flowering; additionally
published work is lacking or no direct evidence of a role in regulating flowering time or
development has been documented. 2) Genes that are involved in development during the
transition from vegetative growth to flowering; typically these genes have been well
characterized, yet have no known role in controlling the initiation of flowering. 3) Genes that are
members of the photoperiod or vernalization pathway in controlling flowering time, most of
which have been well studied and identified as candidates in published QTL studies, or
associated with natural variation in flowering time. These categories were classified as weak,
medium or strong candidates respectively.
The coding regions of several strong candidate genes underlying QTL were sequenced in
an effort to identify putative casual polymorphisms (Salome et al., 2011). Sequencing primers
were designed to amplify the gene regions based on the Columbia genome. Gene regions were
PCR amplified using a high fidelity polymerase (PfuUltra II Fusion HS DNA Polymerase,
Stratagene). Successful amplification was determined by running PCR products on a 1.5%
agarose gel. DNA bands were then cut from the gel and purified (Wizard SV Gel and PCR

74

Clean-Up, Promega). PCR products were sequenced at the MSU Genomics Facility. To ensure
the correct gene region was amplified, the sequence was compared to the Columbia genome via a
BLAST search. Sequence data was analyzed with the Staden package Pregap4 and Gap4.
Polymorphisms were compared between the mapping parents in BioEdit.

Table 3.1. Results of quantitative trait locus (QTL) analysis of flowering time.
Ch, chromosome on which the QTL is located. Position, peak logarithm of odds (LOD) position
on the chromosome in cM. Support interval, 2 LOD. PVE, percentage phenotypic variance
explained by the QTL. Nearest marker, simple sequence repeat (SSR) marker closest to the QTL

Treatment QTL
NonVern

Additive
Support LOD
Dominance Nearest
Ch Position
PVE effect
interval score
effect
marker
(se)
1

70

NVF2

5

8

NVF3
Vern

NVF1

5

60

VF1

1

80.04

VF2

3

18.8

VF3

3

63.19

VF4

5

7.9

VF5

5

37.88

VF6

5

68.54

60.0082.00
5.9610.00
48.0168.0
69.7183.59
11.631.44
56.7463.19
1.9912.05
26.0944.49
62.7174.0

9

5.9

41.1

32.8

9.1

6

4.8

3.1

6.7

4.4

9.2

6.1

15.3

10.5

4.5

2.9

4.5

2.9

75

-1.50
(0.25)
-3.80
(0.49)
-1.30
(0.24)
-1.50
(0.33)
1.62
(0.36)
-2.0
(0.33)
-3.42
(0.40)
-1.88
(0.45)
-1.78
(0.40)

-0.16
(0.34)
-0.42
(0.50)

F5114
NGA158

0.78 (0.32)

C1W9

0.36 (0.51)

MSAT1.5

-1.84
(0.50)
-1.08
(0.47)
0.27 (0.50)
-0.95
(0.52)
-0.39
(0.48)

MSAT3.2
NGA6
NGA158
T5E15
MCK7-1

Results
The days to bolting and flowering were shown to be highly correlated (R2=0.94), with
plants flowering an average of 6.2 days after bolting. Because of this, only days to flowering
was used for analysis. In the Non-vernalization treatment, all (42/42) of the parental plants from
Italy (Fig. 3.1) and 43% (165/384) of the F2 plants flowered (Fig. 3.2), but none of the Sweden
parents flowered (0/42). Thus, the Italy population is able to flower without vernalization. For
the Vernalization treatment, all of the parents from Italy (44/44) and Sweden (40/40), as well as
all of the F2s, flowered. Since all plants flowered in this treatment, the five week cold treatment
was sufficient for saturating vernalization. While all plants flower in this treatment, there is
considerable flowering time differentiation between populations, with the mean flowering date
for Italy being 28 days earlier than Sweden (Fig. 3.3). Transgressive segregation was not
detected in this mapping population, which is in contrast to other studies (El-Lithy et al., 2006;
O'Neill et al., 2008; Brachi et al., 2010; Salome et al., 2011; Seymour et al., 2012).
A total of 9 QTL were identified in this study with 3, and 6, QTL being identified in the
Non-vernalization and Vernalization treatment, respectively (Table 3.1). Of these, 8 were
unique, as one of the QTL in the Vernalization (VF4) and Non-vernalization (NVF2) treatments
mapped to the same chromosomal location, with overlapping LOD support intervals and
identical QTL peaks. While this shared QTL co-localizes between treatments, the effect size
differs, with a PVE of 32.8 and 10.5, in the Non-vernalization and Vernalization treatments,
respectively. Two other pairs of QTL between treatments (NFF1/VF1 and NVF3/VF6) have
slightly overlapping LOD support intervals, but distinct LOD peaks and so are considered
separate QTL.

76

QTL effect sizes ranged from 2.9 to 32.8 (PVE) with 7 of the 9 QTL having effect sizes
under 10 PVE. No QTL were detected on chromosome 2 or 4. With the exception of one QTL
(VF2), the Italy alleles result in early flowering and the Sweden alleles in later flowering.
Significant epistasis between QTL was identified in both treatments. In the Non-vernalization
treatment, all three of the QTL were involved in an epistatic interaction. In this treatment the
large effect QTL (NVF2) which co-localizes with FLC, interacts with both of the other QTL. In
the Vernalization treatment there is a marginally significant interaction of the QTL that colocalizes with FLC (VF4) and VF1, as well as one other interacting pair of QTL.

Figure 3.1. Proportion of plant flowering.
Proportion of Arabidopsis thaliana populations that flowered in the Non- vernalization
treatment. The number above each bar indicate the percentage of each group that flowered.

77

In this study, 7 of the 9 QTL intervals harbor at least one strong candidate gene (Table
3.1). Of these, all but two, DAY NEUTRAL FLOWERING (DNF) and LATE FLOWERING
(LATE), have been discussed as candidates in previous QTL studies. The QTL VF2
encompasses DNF, as well as two other common candidates, Vernalization Insensitive 3-like 1
(VIL1) and REDUCED VERNALIZATION RESPONSE 1 (VRN1). LATE is the only strong
candidate underlying NVF3, which has recently been characterized as controlling expression of
certain flowering genes in leaf vasculature and floral meristem identity genes at the shoot apex
(Weingartner et al., 2011). Two QTL in the Vernalization treatment (VF3 and VF5) do not
possess a strong candidate gene. A possible, yet unlikely candidate for VF5 is SERRATED
LEAVES AND EARLY FLOWERING (SEF); SEF mutants show a suite of abnormal flowering
phenotypes (March-Diaz et al., 2007). Two developmental genes underlie VF3, LATE
MERISTEM IDENTITY2 (LMI1) and GLABROUS INFLORESCENCE STEMS (GIS), but they
have not been studied in regards to controlling flowering time.
Recently, Salome et al. (2011) identified flowering time QTL found across a range of
Arabidopsis accessions and found that these genomic regions include the candidate genes FRI,
FLC, MAF1, and MAF2-5. They sequenced and compared the coding regions of these
candidates across accessions (Salome et al., 2011). For comparison to their results, we
sequenced the coding regions of the same genes (FRI, FLC, MAF1, and MAF2), and three other
strong candidates underlying QTL in our mapping population (VIN3, VIL1, and FT).
Polymorphisms between the mapping parents are presented in Table 3.2. All of the genes except
for FLC have at least one non-synonymous substitution between the Italy and Sweden mapping
parents. Many of the polymorphisms found between the mapping parents were also identified by
Salome et al., (2011). Only one indel was identified; a 3bp indel in exon 5 of VIN3 (Table 3.2).

78

Discussion
Here we report a QTL mapping study of flowering time differentiation utilizing a large
mapping population generated from two natural accessions of winter annuals. In the present
study, QTL mapping was conducted with and without a vernalization treatment, allowing us to
decouple a vernalization requirement from flowering time variation per se. In the Nonvernalization treatment all of the parental plants from Italy and none from Sweden flowered.
Plants from Italy are therefore able to flower in the absence of vernalization, yet in nature they
behave as true winter annuals-over wintering as rosettes and flowering in the spring (Ågren &
Schemske, 2012). Shindo et al., (2005) examined vernalization response in accessions collected
over a wide latitudinal range. When plants did not receive vernalization, all of the accessions
collected above 62° N did not flower, all of the accessions collected below 45° N did flower, and
intermediate latitudes displayed a range of responses. The Italy and Sweden populations in this
study were collected at 42°07’N and 62°48’N, respectively.

Table 3.2 Polymorphisms of candidate gene coding regions between the Italy and Sweden
Arabidopsis thaliana mapping parents.
Position, position of the polymorphism in the Col genome. AA, amino acid number of the
polymorphism. Italy, Sweden and Col, sequence of these respective accessions. Italy, Sweden
and Col AA, amino acid abbreviations of these respective accessions.

79

Table 3.2 (cont’d)
Gene
FLC
FRI

Position
3173805
269462
269469
269962269963

Exon AA
8
1
1

It

Sw

156 G A
146 C G
148 G G

Col
G
A
A

Italy
AA
V
A
M

Sw
AA
V
G
M

Col
AA
V
E
I

18 G A

G

V

I

V

4

129 G G

C

G

G

G

28958968
28959821
25982475
8877617
8876928
8876600
8876598

6
9
1
2
3
3
3

141
189
21
159
363
472
473

G
G
T
G
G
A
G

G
G
T
G
G
A
G

K
P
V
V
I
R
Q

E
P
F
G
R
R
R

E
P
F
G
R
R
R

3

518 C C

T

P

P

S

23248316
23248305
23247644

3
3
4

79 C A
82 G G
274 A C

A
A
A

L
E
M

M
E
L

M
E
M

4

298 A A

G

T

T

A

2324698723246893

VIN3

1

23247572

MAF2
VIL1

24331561

8876464

MAF1

16bp
del

24333548

FT

1 312-

5

QG

R

QG

502503

A
A
G
T
T
G
A

3bp
del

While all of the parents from Italy flowered, it should be noted that there was variation in
the time required to initiate flowering, with plants flowering over a period of 36 days (Suppl.
Fig. 3.1). Under the constant temperature and day length conditions of the Non-vernalization
treatment, plants initiate flowering through an autonomous mechanism which likely does not
80

provide as consistent of a flowering response as environmental cues such as photoperiod and
temperature. In the Vernalization treatment all of the Italy and Sweden plants flower, as do the
F2s. Following vernalization, the parental plants from Sweden flower on average 28 days later
than those from Italy (Figure 3.3). The earlier flowering of Italy after a vernalization treatment
suggests the Italy plants are able to initiate the transition from vegetative to flowering stages
after a much shorter period of cold exposure than Sweden. This is consistent with data from the
Italian field site, where the Italy populations flower substantially earlier than Sweden (Ågren &
Schemske, 2012).
A major goal of evolutionary biology has been to characterize the number and effect size
of genes responsible for adaptive variation. Numerous QTL studies of adaptive traits have
provided support for Orr’s (1998) model; predicting an exponential distribution of effect sizes
with few of large effect (Barton & Keightley, 2002). For nearly 20 years, Arabidopsis
researchers have been dissecting the genetic basis of flowering time. We outline the traditional
linkage (QTL) mapping publications; this includes 26 publications, with 98 different QTL
experiments. These studies differ in the mapping parents utilized, size and type of mapping
populations, vernalization treatment, photoperiod, and growing conditions. The average number
of QTL identified is 4.03; we have detected a typical number of QTL, with studies ranging
between one (O'Neill et al., 2008; Balasubramanian et al., 2009) and ten QTL (Weinig et al.,
2002). Our results are in agreement to most studies where at least one large effect QTL is
identified, but in contrast to studies that detect a higher occurrence of large effect QTL (Loudet
et al., 2002; Salome et al., 2011; Strange et al., 2011). The ability to detect QTL of small effect
is limited by the size of the mapping population (Beavis, 1998). Many previous QTL studies for
flowering time involved mapping populations with less than 150 individuals, and therefore

81

limited ability to detect small effect QTL. Here we are able to identify QTL with effect sizes as
low as 2.9% PVE. The distribution of QTL effect sizes in the present study provides additional
support for Orr’s (1998) model. Additionally, we have detected a moderate amount of epistasis
between QTL. The QTL that co-localize with FLC (NVF2 and VF4) are involved in the
majority of the significant QTL x QTL interactions. This may be expected due the central role of
FLC in genetic pathways of floral initiation (Simpson & Dean, 2002).

Figure 3.2 Days to flowering in the F2 plants.

82

Understanding the role of genotype-environment interaction has been a long-standing
goal in evolutionary biology as this can influence local adaptation (Kawecki & Ebert, 2004).
Several studies in Arabidopsis have characterized significant QTL x environment interactions
(Ungerer et al., 2002; Weinig et al., 2002; Li et al., 2006). The ability to detect QTL colocalizing with FRI can depend on the environmental conditions. For example, in the
extensively utilized mapping population, Col x Ler, detection of FRI varies with vernalization
and photoperiod (Stratton, 1998; Weinig et al., 2002; Weinig et al., 2003; Juenger et al., 2005).
Here we report distinct QTL profiles between the Vernalization and Non-vernalization
treatments, with only one QTL shared between them. The effect size of this shared QTL differs
substantially between the treatments (Table 3.1). In the Non-vernalization treatment the QTL is
of large effect, with a PVE of 32.8%, as opposed to 10.5% in the Vernalization treatment. FLC
stands out as a candidate gene for the shared QTL (NVF2 and VF4). FLC is a strong inhibitor of
flowering and is routinely identified in studies examining flowering time variation. Another
candidate in the Non-vernalization treatment is FT; this QTL was not detected after
vernalization.
In an analogous experiment, Strange et al., (2011) examined the effect of different
vernalization treatments on mapping populations of four accessions from Sweden crossed with
Col. Overall, they also found that the effect size of QTL decreased after vernalization. More
specifically, the effect size of QTL co-localizing with FLC and FT went from having a strong
effect with no vernalization, to being not significant after 14 weeks of vernalization (Strange et
al., 2011). Our results suggest that differences between Italy and Sweden at the FLC locus
allows for populations from Italy to flower in the absence of vernalization. Vernalization acts to
repress FLC by epigenetic modification, allowing for the transition from vegetative to flowering

83

states (Amasino, 2004). FLC allelic variation has been shown to be associated with a latitudinal
cline in flowering time variation. In southern populations alleles that weaken FLC allow for
flowering without vernalization. It is to be expected that FLC would have a smaller effect size in
the Vernalization treatment because vernalization would provide the strong repression of FLC in
both the parental populations. The Non-vernalization treatment yields QTL with candidates
involved in the photoperiod regulation of flowering (FLC & FT) where the Vernalization
treatment involves candidates of the vernalization pathway (VRN1, VIL1, VIN3). In the present
work only vernalization was varied between treatments; future studies should include
photoperiods similar to those in the habitats of both populations.
Many of the QTL identified here have major candidate genes that map to the same
chromosomal location (Table 3.1). It is possible that these candidate genes are the causal genetic
elements underlying QTL, however, functional experiments are required to confirm their role in
flowering time variation. As indicated above, FLC is a prime candidate for the shared QTL in
both the Vernalization and Non-vernalization treatment. We did not identify any nonsynonymous polymorphisms in the FLC coding region which is in agreement with other studies
examining FLC variation (Gazzani et al., 2003; Caicedo et al., 2004; Strange et al., 2011).
In addition to FLC, the Non-vernalization treatment yields FT as a possible candidate
gene. FT has been identified as a key component of the photoperiod pathway of floral induction
across plant species (Kardailsky et al., 1999; Kobayashi et al., 1999; Böhlenius et al., 2006;
Tamaki et al., 2007) and is highly conserved (Kobayashi et al., 1999). We have identified one
non-synonmous polymorphism in the FT coding region; however, it is unlikely to have an effect
as both amino acids are neutral and non-polar (Table 3.2). Recent studies have failed to detect
non-synonomous polymorphisms in the coding region of FT, but rather find considerable cis-

84

regulatory variation (Schwartz et al., 2009; Adrian et al., 2010; Strange et al., 2011). In our
study, the QTL (NVF1) that encompasses FT is of relatively small effect size and only was
identified in the Non-vernalization treatment. FLC acts to repress FT expression until it is
inhibited by vernalization or an autonomous mechanism. It is logical that these candidates were
detected in the Non-vernalization treatment as allelic variation between the mapping parents
would provide differential sensitivity of FT to FLC in the absence of vernalization. With
vernalization the effect of FT allelic variation could be negligible once FLC is sufficiently
inhibited.
In the Vernalization treatment there are several candidate genes identified, including all
members of the FLC/MAF clade, which Salome et al., (2011) have highlighted as being
common across QTL mapping studies. MAF1, also known as FLM, is a candidate underlying
VF1. MAF1 is a homologue of FLC and also acts to inhibit flowering in a similar fashion
(Scortecci et al., 2001). We have identified two synonymous substitutions and one nonsynonymous substitution in the coding region (Table 3.2). Given that there is functional
redundancy between FLC and MAF1, it would be expected that weakened alleles at both genes
would be necessary to allow for earlier flowering.
The QTL, on the end of chromosome V, (VF6), encompasses seven potential candidate
genes. MAF2-4 act to repress flowering after short cold spells (Ratcliffe et al., 2001), and
polymorphisms at these loci have been shown to be associated with diverse flowering times in
Arabidopsis accessions (Caicedo et al., 2009). Other strong candidates at the locus include
VIN3, VIP4, and ELF5. We have sequenced the coding regions of MAF2 and VIN3. In MAF2
there is one non-synonymous substitution between Italy in Sweden, however, there is no change

85

in polarity or charge. On the other hand, in the VIN3 coding region there are 2 non-synonymous
substitutions, as well as a 3bp deletion (Table 3.2).
Three genes, VRN1, VIL1, and DNF underlie the LOD support interval of VF2. VRN2
and VIL1 act to promote flowering and are upstream members of the vernalization pathway
(Gendall et al., 2001). DNF acts in the repression of CO during the early part of the day (Morris
et al., 2010). For successful vernalization it is important to note that two distinct processes must
occur: 1) the initial repression of FLC and 2) the maintenance of FLC inhibition following
vernalization. VIL1 (along with VIN3) is necessary for the chromatin modification of
FLC/MAF1 repression.
The genetics of flowering time in Arabidopsis has been examined extensively, and the
molecular genetic pathways have been well characterized by a variety of means, including QTL
mapping. However, almost all of these QTL studies have involved the lab strains Col, Ler or
both as mapping parents and been performed under controlled laboratory conditions. In the
present study we identify two QTL that do not have a clear candidate gene (VF3 and VF5). Two
floral developmental genes underlie this QTL, but neither would be considered a strong
candidate, or have been described in previous QTL studies. A possible candidate for VF5 is
SEF, which has not been thoroughly studied to date. While it is rare to reveal novel loci involved
in flowering time, it is possible; as is demonstrated in the present work through mapping natural
populations, and a recent study that investigated natural accessions crossed with Col in the field
(Brachi et al., 2010).
Elucidating the role of cis-regulatory and protein coding variation is a central goal in
studying the genetic basis of adaptation. Hoekstra and Coyne (2007) highlight that considerable
emphasis has been placed on cis-regulatory evolution, however, sufficient data is lacking to

86

make generalizations. Here, we examine only the protein coding regions of several known
candidate genes underlying QTL. Rigorous functional experiments are required to identify the
causal genes and mutations controlling flowering time adaptation. Nonetheless, in several cases
a lack of protein coding variation suggests the role of cis-regulatory changes. In FT and FLC
there are no structural polymorphisms between the mapping parents, which is in agreement with
studies characterizing regulatory variation of these genes (Gazzani et al., 2003; Michaels et al.,
2003; Schwartz et al., 2009; Adrian et al., 2010; Strange et al., 2011). In contrast, VIN3 could
potentially represent protein coding changes responsible for flowering time variation, as has been
well described in FRI (Johanson et al., 2000; Gazzani et al., 2003).

Figure 3.3. Days to flowering in the Vernalization treatment.
Days to flowering of the Arabdiopsis thaliana F2’s in the Vernalization treatment. The gray
bars represent the flowering duration of the respective parents, with the black triangle
representing the mean flowering date.

87

While many genes control flowering time, few have been shown to be associated with
natural flowering time variation. The regulatory gene FRI has received the most attention as a
major determinant of flowering time in natural populations of Arabidopsis (Clarke & Dean,
1994; Johanson et al., 2000; Le Corre et al., 2002; Simpson & Dean, 2002). Functional FRI
alleles result in the accumulation of FLC mRNA which inhibits flowering. Vernalization acts to
reduce the sensitivity of FLC to FRI and thereby promotes flowering. Several lines of evidence
support the role of FRI in controlling natural flowering time variation including 1) the
identification of several naturally occurring, independently derived null FRI alleles (Le Corre et
al., 2002), 2) a large number of non-synonomous polymorphisms in the first exon of FRI (Le
Corre et al., 2002), 3) the identification of QTL co-localizing with FRI in several QTL studies
with different parental populations varying in flowering time (Salome et al., 2011), and 4)
evidence for FRI alleles influencing fitness depending on the FLC allelic background and
environmental conditions (Korves et al., 2007). Despite this substantial body of research
demonstrating the contribution of FRI in flowering time variation, we found no role of FRI
allelic variation controlling flowering time differentiation in the Italy and Sweden populations.
It has been well established that coding region variation in exon 1 of FRI results in nonfunctional alleles and early flowering (Johanson et al., 2000; Le Corre et al., 2002; Gazzani et
al., 2003). We have identified one polymorphism in FRI between the Italy and Sweden mapping
parents which results in a Glycine to Alanine substitution at amino acid 146 (Table 3.2). Both
these amino acids are neutral and nonpolar, and are unlikely to affect protein function.
Furthermore, no QTL were found to co-localize with FRI.
This may be surprising given that a latitudinal cline in flowering time has been predicted
to be associated with FRI allele functionality (Karlsson et al., 1993; Nordborg & Bergelson,

88

1999; Pigliucci & Marlow, 2001). That is, populations in southern latitudes are expected to
harbor null FRI alleles providing flowering in the absence of vernalization, and northern
populations will contain functional FRI alleles. However, several studies have been unable to
detect such a cline in flowering time (Karlsson et al., 1993; Nordborg & Bergelson, 1999;
Pigliucci & Marlow, 2001; Stinchcombe et al., 2004). In the most extensive study, Stinchcombe
et al. (2004) identify functional FRI alleles to be associated with a cline in flowering, but no
association was detected for non-functional FRI alleles, as is predicted. Furthermore, the
populations from southern latitudes with functional FRI alleles actually flowered earlier than
those from northern latitudes. The variation in flowering time is explained by an epistatic
interaction between the functional FRI alleles and FLC (Caicedo et al., 2004). Our results are in
agreement with those of Stinchcombe et al. (2004) and Caicedo et al. (2004). In the present
study, FRI is not identified as a candidate gene, but FLC is in both the Vernalization and NonVernalization treatment. This makes sense when taking life history into consideration. Given
that both of the parental populations are winter annuals, FRI functional alleles would be favored
to delay flowering until the appropriate time in the spring (Simpson & Dean, 2002). Weak
alleles of FLC in the Italy mapping parent, reduce sensitivity to FRI and promote flowering after
mild periods of cold compared to Sweden (Amasino, 2004).
In summary, we have presented a QTL mapping study of flowering time variation
between two winter annual accessions of Arabidopsis collected from the northern and southern
geographic limits of their native range. Our study benefits from a large mapping population size
and is thus capable of accurately estimating effect sizes. Although we identified both large and
small effect QTL, it is almost certain that we have missed many QTL of very small effect (≈1%
PVE). We find the genetic architecture of adaptation to depend heavily on the environmental

89

conditions, namely the vernalization treatment. Previous QTL studies have emphasized the role
of FRI and FLC in determining flowering time differentiation. While substantial variation is
explained by a QTL co-localizing with FLC, we find no role of FRI. Furthermore, it should be
noted that genetic basis of flowering time in this system is complex and includes several QTL
with numerous other candidate genes. Future research should be conducted to thoroughly
examine the causal genetic elements underlying QTL as well as field studies to determine the
relationship of flowering time and fitness. An integration of these pursuits will provide a
comprehensive understanding of the genetic basis of flowering time in Arabidopsis.

Acknowledgments
We would like to thank M. Cameron for assistance with figure development, K. Califf for
assistance with sequence alignment, C. Oakley for help with R/qtl, and R. Amasino and D.
Weigel for helpful comments on the manuscript as well as two anonymous reviewers.

90

LITERATURE CITED

91

LITERATURE CITED

Adrian J, Farrona S, Reimer JJ, Albani MC, Coupland G, Turck F. 2010. cis-Regulatory
elements and chromatin state coordinately control temporal and spatial expression of
FLOWERING LOCUS T in Arabidopsis. Plant Cell 22(5): 1425-1440.
Ågren J, Schemske DW. 2012. Reciprocal transplants demonstrate strong adaptive
differentiation of the model organism Arabidopsis thaliana in its native range. New
Phytologist 194(4): 1112-1122.
Amasino R. 2004. Vernalization, competence, and the epigenetic memory of winter. Plant Cell
16(10): 2553-2559.
Amasino R. 2010. Seasonal and developmental timing of flowering. Plant Journal 61(6): 10011013.
Balasubramanian S, Schwartz C, Singh A, Warthmann N, Kim MC, Maloof JN, Loudet O,
Trainer GT, Dabi T, Borevitz JO, Chory J, Weigel D. 2009. QTL mapping in new
Arabidopsis thaliana advanced intercross-recombinant inbred lines. Plos One 4(2).
Barton NH, Keightley PD. 2002. Understanding quantitative genetic variation. Nature Reviews
Genetics 3(1): 11-21.
Bastow R, Dean C. 2003. Plant sciences - Deciding when to flower. Science 302(5651): 16951697.
Beavis WD 1998. QTL analyses: power, precision, and accuracy. In: Paterson AH ed. Molecular
Dissection of Complex Traits. Boca Raton, FL: CRC Press, 145-162.
Böhlenius H, Huang T, Charbonnel-Campaa L, Brunner AM, Jansson S, Strauss SH,
Nilsson O. 2006. CO/FT regulatory module controls timing of flowering and seasonal
growth cessation in trees. Science 312(5776): 1040-1043.

92

Brachi B, Faure N, Horton M, Flahauw E, Vazquez A, Nordborg M, Bergelson J, Cuguen
J, Roux F. 2010. Linkage and association mapping of Arabidopsis thaliana flowering
time in nature. Plos Genetics 6(5).
Broman KW, Wu H, Sen S, Churchill GA. 2003. R/qtl: QTL mapping in experimental crosses.
Bioinformatics 19(7): 889-890.
Burn JE, Bagnall DJ, Metzger JD, Dennis ES, Peacock WJ. 1993. DNA methylation,
vernalization, and the initiation of flowering. Proceedings of the National Academy of
Sciences of the United States of America 90(1): 287-291.
Caicedo AL, Richards C, Ehrenreich IM, Purugganan MD. 2009. Complex rearrangements
lead to novel chimeric gene fusion polymorphisms at the Arabidopsis thaliana MAF2-5
flowering time gene cluster. Molecular Biology and Evolution 26(3): 699-711.
Caicedo AL, Stinchcombe JR, Olsen KM, Schmitt J, Purugganan MD. 2004. Epistatic
interaction between Arabidopsis FRI and FLC flowering time genes generates a
latitudinal cline in a life history trait. Proceedings of the National Academy of Sciences of
the United States of America 101(44): 15670-15675.
Clarke JH, Dean C. 1994. Mapping FRI, a locus controlling flowering time and vernalization
response in Arabidopsis thaliana. Molecular & General Genetics 242(1): 81-89.
Clarke JH, Mithen R, Brown JKM, Dean C. 1995. QTL analysis of flowering time in
Arabidopsis thaliana Molecular & General Genetics 248(3): 278-286.
El-Lithy ME, Bentsink L, Hanhart CJ, Ruys GJ, Rovito DI, Broekhof JLM, van der Poel
HJA, van Eijk MJT, Vreugdenhil D, Koornneef M. 2006. New Arabidopsis
recombinant inbred line populations genotyped using SNPWave and their use for
mapping flowering-time quantitative trait loci. Genetics 172(3): 1867-1876.
Gazzani S, Gendall AR, Lister C, Dean C. 2003. Analysis of the molecular basis of flowering
time variation in Arabidopsis accessions. Plant Physiology 132(2): 1107-1114.
Gendall AR, Levy YY, Wilson A, Dean C. 2001. The VERNALIZATION 2 gene mediates the
epigenetic regulation of vernalization in Arabidopsis. Cell 107(4): 525-535.
93

Haley CS, Knott SA. 1992. A simple regression method for mapping quantitative trait loci in
line crosses using flanking markers. Heredity 69: 315-324.
Hoekstra HE, Coyne JA. 2007. The locus of evolution: Evo devo and the genetics of
adaptation. Evolution 61(5): 995-1016.
Izawa T. 2007. Adaptation of flowering-time by natural and artificial selection in Arabidopsis
and rice. Journal of Experimental Botany 58(12): 3091-3097.
Johanson U, West J, Lister C, Michaels S, Amasino R, Dean C. 2000. Molecular analysis of
FRIGIDA, a major determinant of natural variation in Arabidopsis flowering time.
Science 290(5490): 344-347.
Juenger TE, Sen S, Stowe KA, Simms EL. 2005. Epistasis and genotype-environment
interaction for quantitative trait loci affecting flowering time in Arabidopsis thaliana.
Genetica 123(1-2): 87-105.
Kalisz S, Wardle GM. 1994. Life-history variation in Campanula americana (Campanulaceae)
- population differentiation. American Journal of Botany 81(5): 521-527.
Kardailsky I, Shukla VK, Ahn JH, Dagenais N, Christensen SK, Nguyen JT, Chory J,
Harrison MJ, Weigel D. 1999. Activation tagging of the floral inducer FT. Science
286(5446): 1962-1965.
Karlsson BH, Sills GR, Nienhuis J. 1993. Effects of photoperiod and vernalization on the
number of leaves at flowering in 32 Arabidopsis thaliana (Brassicaceae) ecotypes.
American Journal of Botany 80(6): 646-648.
Kawecki TJ, Ebert D. 2004. Conceptual issues in local adaptation. Ecology Letters 7(12): 12251241.
Kobayashi Y, Kaya H, Goto K, Iwabuchi M, Araki T. 1999. A pair of related genes with
antagonistic roles in mediating flowering signals. Science 286(5446): 1960-1962.

94

Koornneef M, Alonso-Blanco C, Vreugdenhil D. 2004. Naturally occurring genetic variation
in Arabidopsis thaliana. Annual Review of Plant Biology 55: 141-172.
Korves TM, Schmid KJ, Caicedo AL, Mays C, Stinchcombe JR, Purugganan MD, Schmitt
J. 2007. Fitness effects associated with the major flowering time gene FRIGIDA in
Arabidopsis thaliana in the field. American Naturalist 169(5): E141-E157.
Lacey EP. 1988. Latitudinal variation in reproductive timing of a short-lived monocarp, Daucus
carota (Apiaceae). Ecology 69(1): 220-232.
Le Corre V, Roux F, Reboud X. 2002. DNA polymorphism at the FRIGIDA gene in
Arabidopsis thaliana: Extensive nonsynonymous variation is consistent with local
selection for flowering time. Molecular Biology and Evolution 19(8): 1261-1271.
Lee I, Bleecker A, Amasino R. 1993. Analysis of naturally-occurring late flowering in
Arabidopsis thaliana. Molecular & General Genetics 237(1-2): 171-176.
Lempe J, Balasubramanian S, Sureshkumar S, Singh A, Schmid M, Weigel D. 2005.
Diversity of flowering responses in wild Arabidopsis thaliana strains. Plos Genetics 1(1):
109-118.
Li Y, Roycewicz P, Smith E, Borevitz JO. 2006. Genetics of local adaptation in the laboratory:
flowering time Quantitative Trait Loci under geographic and seasonal conditions in
Arabidopsis. Plos One 1(2).
Lister C, Dean C. 1993. Recombinant inbred lines for mapping RFLP and phenotypic markers
in Arabidopsis thaliana. Plant Journal 4(4): 745-750.
Loudet O, Chaillou S, Camilleri C, Bouchez D, Daniel-Vedele F. 2002. Bay-0 x Shahdara
recombinant inbred line population: a powerful tool for the genetic dissection of complex
traits in Arabidopsis. Theoretical and Applied Genetics 104(6-7): 1173-1184.
Lowry DB. 2012. Local adaptation in the model plant. New Phytologist 194(4): 888-890.

95

Lu H, Romero-Severson J, Bernardo R. 2002. Chromosomal regions associated with
segregation distortion in maize. Theoretical and Applied Genetics 105(4): 622-628.
March-Diaz R, Garcia-Dominguez M, Florencio FJ, Reyes JC. 2007. SEF, a new protein
required for flowering repression in Arabidopsis, interacts with PIE1 and ARP6. Plant
Physiology 143(2): 893-901.
Michaels SD, Amasino RM. 1999. FLOWERING LOCUS C encodes a novel MADS domain
protein that acts as a repressor of flowering. Plant Cell 11(5): 949-956.
Michaels SD, He YH, Scortecci KC, Amasino RM. 2003. Attenuation of FLOWERING
LOCUS C activity as a mechanism for the evolution of summer-annual flowering
behavior in Arabidopsis. Proceedings of the National Academy of Sciences of the United
States of America 100(17): 10102-10107.
Mitchell-Olds T, Schmitt J. 2006. Genetic mechanisms and evolutionary significance of natural
variation in Arabidopsis. Nature 441(7096): 947-952.
Morgan TJ, Mackay TFC. 2006. Quantitative trait loci for thermotolerance phenotypes in
Drosophila melanogaster. Heredity 96(3): 232-242.
Morris K, Thornber S, Codrai L, Richardson C, Craig A, Sadanandom A, Thomas B,
Jackson S. 2010. DAY NEUTRAL FLOWERING represses CONSTANS to prevent
Arabidopsis flowering early in short days. Plant Cell 22(4): 1118-1128.
Napp-Zinn K. 1985. Handbook of Flowering. Boca Raton, FL: CRC.
Nordborg M, Bergelson J. 1999. The effect of seed and rosette cold treatment on germination
and flowering time in some Arabidopsis thaliana (Brassicaceae) ecotypes. American
Journal of Botany 86(4): 470-475.
O'Neill CM, Morgan C, Kirby J, Tschoep H, Deng PX, Brennan M, Rosas U, Fraser F, Hall
C, Gill S, Bancroft I. 2008. Six new recombinant inbred populations for the study of
quantitative traits in Arabidopsis thaliana. Theoretical and Applied Genetics 116(5): 623634.

96

Olsson K, Ågren J. 2002. Latitudinal population differentiation in phenology, life history and
flower morphology in the perennial herb Lythrum salicaria. Journal of Evolutionary
Biology 15(6): 983-996.
Orr HA. 1998. The population genetics of adaptation: The distribution of factors fixed during
adaptive evolution. Evolution 52(4): 935-949.
Pigliucci M. 2003. Selection in a model system: Ecological genetics of flowering time in
Arabidopsis thaliana. Ecology 84(7): 1700-1712.
Pigliucci M, Marlow ET. 2001. Differentiation for flowering time and phenotypic integration in
Arabidopsis thaliana in response to season length and vernalization. Oecologia 127(4):
501-508.
Ratcliffe OJ, Nadzan GC, Reuber TL, Riechmann JL. 2001. Regulation of flowering in
Arabidopsis by an FLC homologue. Plant Physiology 126(1): 122-132.
Salome PA, Bomblies K, Laitinen RAE, Yant L, Mott R, Weigel D. 2011. Genetic
architecture of flowering-time variation in Arabidopsis thaliana. Genetics 188(2): 421U313.
Schwartz C, Balasubramanian S, Warthmann N, Michael TP, Lempe J, Sureshkumar S,
Kobayashi Y, Maloof JN, Borevitz JO, Chory J, Weigel D. 2009. Cis-regulatory
changes at FLOWERING LOCUS T mediate natural variation in flowering responses of
Arabidopsis thaliana. Genetics 183(2): 723-732.
Scortecci KC, Michaels SD, Amasino RM. 2001. Identification of a MADS-box gene,
FLOWERING LOCUS M, that represses flowering. Plant Journal 26(2): 229-236.
Seymour DK, Filiault DL, Henry IM, Monson-Miller J, Ravi M, Pang AD, Comai L, Chan
SWL, Maloof JN. 2012. Rapid creation of Arabidopsis doubled haploid lines for
quantitative trait locus mapping. Proceedings of the National Academy of Sciences of the
United States of America 109(11): 4227-4232.

97

Shindo C, Aranzana MJ, Lister C, Baxter C, Nicholls C, Nordborg M, Dean C. 2005. Role
of FRIGIDA and FLOWERING LOCUS C in determining variation in flowering time of
Arabidopsis. Plant Physiology 138(2): 1163-1173.
Shindo C, Bernasconi G, Hardtke CS. 2007. Natural genetic variation in arabidopsis: Tools,
traits and prospects for evolutionary ecology. Annals of Botany 99(6): 1043-1054.
Simpson GG, Dean C. 2002. Flowering - Arabidopsis, the rosetta stone of flowering time?
Science 296(5566): 285-289.
Stinchcombe JR, Weinig C, Ungerer M, Olsen KM, Mays C, Halldorsdottir SS,
Purugganan MD, Schmitt J. 2004. A latitudinal cline in flowering time in Arabidopsis
thaliana modulated by the flowering time gene FRIGIDA. Proceedings of the National
Academy of Sciences of the United States of America 101(13): 4712-4717.
Strange A, Li P, Lister C, Anderson J, Warthmann N, Shindo C, Irwin J, Nordborg M,
Dean C. 2011. Major-effect alleles at relatively few loci underlie distinct vernalization
and flowering variation in Arabidopsis accessions. Plos One 6(5).
Stratton DA. 1998. Reaction norm functions and QTL-environment interactions for flowering
time in Arabidopsis thaliana. Heredity 81: 144-155.
Sung SB, Amasino RM. 2004. Vernalization and epigenetics: how plants remember winter.
Current Opinion in Plant Biology 7(1): 4-10.
Tamaki S, Matsuo S, Wong HL, Yokoi S, Shimamoto K. 2007. Hd3a protein is a mobile
flowering signal in rice. Science 316(5827): 1033-1036.
Tanksley SD. 1993. Mapping polygenes. Annual Review of Genetics 27: 205-233.
Ungerer MC, Halldorsdottir SS, Modliszewski JL, Mackay TFC, Purugganan MD. 2002.
Quantitative trait loci for inflorescence development in Arabidopsis thaliana. Genetics
160(3): 1133-1151.
Van Ooijen JW, Voorips RE. 2001. Joinmap 3.0: software for the calculation of genetic
98

linkage maps. Wageningen, the Netherlands: Plant Research International.
Weingartner M, Subert C, Sauer N. 2011. LATE, a C2H2 zinc-finger protein that acts as floral
repressor. Plant Journal 68(4): 681-692.
Weinig C, Stinchcombe JR, Schmitt J. 2003. QTL architecture of resistance and tolerance
traits in Arabidopsis thaliana in natural environments. Molecular Ecology 12(5): 11531163.
Weinig C, Ungerer MC, Dorn LA, Kane NC, Toyonaga Y, Halldorsdottir SS, Mackay TFC,
Purugganan MD, Schmitt J. 2002. Novel loci control variation in reproductive timing
in Arabidopsis thaliana in natural environments. Genetics 162(4): 1875-1884.
Werner JD, Borevitz JO, Uhlenhaut NH, Ecker JR, Chory J, Weigel D. 2005a. FRIGIDAindependent variation in flowering time of natural Arabidopsis thaliana accessions.
Genetics 170(3): 1197-1207.
Werner JD, Borevitz JO, Warthmann N, Trainer GT, Ecker JR, Chory J, Weigel D. 2005b.
Quantitative trait locus mapping and DNA array hybridization identify an FLM deletion
as a cause for natural flowering-time variation. Proceedings of the National Academy of
Sciences of the United States of America 102(7): 2460-2465.
Westerbergh A, Doebley J. 2002. Morphological traits defining species differences in wild
relatives of maize are controlled by multiple quantitative trait loci. Evolution 56(2): 273283.

99

CHAPTER 4
Functional traits underlying mating system differentiation
between the wild relatives of rice

100

Introduction
Angiosperms display a remarkable diversity of reproductive traits and strategies.
Understanding the evolutionary mechanisms that have shaped this diversity has been a central
goal in studies of plant evolution (Barrett, 2010a). Given the sessile nature of plants,
reproduction is of particular significance, as it and subsequent seed dispersal are the only
opportunity for alleles to move within and among populations, as well as the physical landscape
(Levin & Kerster, 1974).
The mating system shift to predominantly selfing from outcrossing taxa is regarded as the
most common transition in plant evolution (Stebbins, 1974). Outcrossing rates can vary
considerably among closely related species and even populations, suggesting that it can respond
quickly to natural selection (Jain, 1976; Goodwillie et al., 2005). Additionally, the transition to
selfing is associated with characteristic shifts in floral biology, life history, and ecology (Barrett
et al., 1996). There are two classic explanations for the advantage of self-fertilization: 1) selfing
may be beneficial as it can provide a means for reproductive assurance when pollinators or mates
are scarce; a hypothesis first proposed by Darwin (1876) and 2) an inherent genetic transmission
advantage, demonstrated by Fisher (1941), as selfing individuals function as both the maternal
and paternal parents. In contrast, the evolutionary consequences of self-fertilization can be
profound, as selfing can drastically reduce the effective population size (Hamrick & Loveless,
1987; Schoen & Brown, 1991) and nucleotide diversity (Wright et al., 2008), influence
population genetic structure (Hamrick & Godt, 1996), and restrict broad scale rates of
diversification (Charlesworth, 1992; Takebayashi & Morrell, 2001; Igic et al., 2008). Most
directly, selfed offspring can suffer from inbreeding depression, the primary genetic cost of selffertilization (Lloyd, 1979; Charlesworth & Charlesworth, 1987; Husband & Schemske, 1996).

101

Lande and Schemske (1985) demonstrated the joint evolution of inbreeding depression
and selfing, with reduced levels of inbreeding depression as selfing increases. This theoretical
model predicts two stable endpoints to mating system evolution: predominant outcrossing and
selfing. Subsequent analysis of empirical studies provided support for this bimodal distribution
of outcrossing rates (Schemske & Lande, 1985). Taken together these findings prompted a flurry
of theoretical and empirical investigation into the distribution of mating systems, with much
emphasis on the evolutionary significance of mixed mating (Barrett, 2003). Recent surveys of
outcrossing rates have demonstrated a higher occurrence of mixed mating and do not support a
bimodal distribution, particularly for animal pollinated (entomophilous) taxa (Vogler & Kalisz,
2001; Goodwillie et al., 2005). However, this finding should be interpreted with caution; for
example there is a disproportionate lack of mating system estimates of obligate outcrossing
species (self-incompatible or dioecious), as well as other potential sources of bias in available
data (Igic & Kohn, 2006). Nonetheless, a bimodal distribution of outcrossing rates is well
supported for wind-pollinated (anemophilous) plants (Aide, 1986; Vogler & Kalisz, 2001;
Goodwillie et al., 2005). This observation is attributed to the predictable nature of wind as a
pollen vector, as opposed to animal vectors that can fluctuate over time (Aide, 1986; Friedman &
Barrett, 2009b).

“Our knowledge of the reproductive ecology of most wind-pollinated
plants is quite rudimentary, and the functional relationships between pollination
and mating are poorly understood.”-Friedman and Barrett 2009

102

For nearly 150 years, the vast body of literature devoted to plant mating systems and
pollination biology has focused almost exclusively on animal-pollinated systems (Friedman &
Barrett, 2009b; Barrett, 2010b). This is unfortunate given that approximately 10% of all
angiosperms are wind-pollinated (Ackerman, 2000), including our most important crops (rice,
wheat, maize, etc) and wind-pollination is a derived state, having evolved at least 65 times from
animal-pollinated ancestors (Linder, 1998). There are also intrinsic differences in floral biology
between wind- and animal-pollinated species that are particularly interesting to study within a
mating system context, including: 1) the presence/absence of attractive floral structures that
evolve with the outcrossing rate, 2) fundamental differences in floral anatomy that may
determine the functional traits that control selfing, and 3) the role of geitonogamy (mating
between flowers of the same plant) and mechanisms that provide this mode of selfing.
In animal-pollinated systems, comparative studies of floral biology between selfing and
outcrossing groups have been conducted in a wide range of species (Darwin, 1876; Ritland &
Ritland, 1989; Lyons & Antonovics, 1991; Goodwillie, 1999; Armbruster et al., 2002; VallejoMarin & Barrett, 2009). Through these efforts it has been well established that selfing plants
produce small flowers with highly reduced attractive floral traits (e.g. showy petals, nectar
production, attractive scents, etc), when compared to outcrossers (Ornduff, 1969; Goodwillie et
al., 2010). To explain this syndrome, sex allocation theory has been developed; which predicts
reduced allocation to both, male function and attractive structures, as selfing increases
(Charlesworth & Charlesworth, 1987; Charlesworth & Morgan, 1991). Correlations between
attractive floral characters and functional traits that provide selfing, are common and can
influence, or even limit, an evolutionary response to selection (Conner, 2006). In windpollinated plants attractive floral traits are absent all together; therefore the transition to a selfing

103

syndrome may be relatively less complex than in animal-pollinated species. Comparative studies
of floral biology between selfing and outcrossing groups are lacking in wind-pollinated plants.
In addition to comparative work, more mechanistic experiments have been aimed at
elucidating the mechanisms of self-fertilization in animal-pollinated taxa. In hermaphroditic selfcompatible plants, the most common means for achieving autogamous (fertilization within an
individual flower) selfing, is through a reduction in herkogamy (the physical separation of stigma
and anthers) (Ritland & Ritland, 1989; Dole, 1990; Holtsford & Ellstrand, 1992; Kohn & Barrett,
1994; Belaoussoff & Shore, 1995; Karron et al., 1997; Chang & Rausher, 1999; Kalisz et al.,
1999; Barrett, 2003). Conversely, herkogamy in wind-pollinated groups has been unexplored,
and its potential to contribute to selfing is unclear due to fundamental differences in floral
biology; namely dry, unclumped pollen (Crane, 1986; Ackerman, 2000) and in the case of
grasses, poricidal anthers that arrange pollen in a unique fashion and release it through terminal
pores (Kirpes et al., 1996). Therefore, physical contact between stigma and anthers may not
provide a selfing mechanism as reliable as that seen in animal-pollinated plants. In windpollinated groups, inflorescence architecture is the functional trait that has received the most
attention in regards to floral biology (Niklas, 1985; Niklas, 1987; Friedman & Harder, 2004;
Friedman & Harder, 2005; Cresswell et al., 2010). Grasses, in particular, display a diversity in
‘openness’ of inflorescences ranging from diffuse to compact (Doust & Kellogg, 2002).
Research involving inflorescence architecture in grasses has focused solely on wind
aerodynamics and strategies for pollen removal and receipt, and has not been investigated in
terms of selfing rates. It is likely that close proximity of flowers in compact inflorescences could
enhance the opportunity for geitonogamy.

104

Self-fertilization can arise from either geitonogamy or autogamy. Large floral displays
function in attracting pollinators across long distances, but can result in geitonogamy when
pollinators visit subsequent flowers on the same plant. Theoretical work has demonstrated that
geitonogamy is unlikely to be favored by selection (Lloyd, 1992) and this mode of selfing can
have profound costs through pollen and seed discounting (Lloyd & Schoen, 1992; Dejong et al.,
1993; Harder & Barrett, 1995; Barrett, 2003; Goodwillie et al., 2005). Thus, geitonogamous
selfing does not provide reproductive assurance in the same manner as autogamy; although there
examples where this is the case, such as monecious Begonia species (Ågren & Schemske, 1993).
Geitonogamy can be quite common in animal-pollinated systems (Harder & Barrett, 1995;
Eckert, 2000; Dorken et al., 2002; Karron et al., 2004) and is suspected to promote the evolution
of dichogomy or dioecy, particularly in large plants (Barrett, 2003). In a similar fashion, wind
could provide geitonogamous selfing between flowers in close proximity, however, this has not
been investigated in wind-pollinated taxa until recently. Friedman and Barrett (2009a) have
identified substantial geitonogamous selfing in Carex spp., and propose that, unlike animalpollinated species, geitonogamy in wind-pollinated systems may often provide reproductive
assurance, particularly in plants with unisexual flowers. More experiments in wind-pollinated
plants are needed that dissect the contributions of autogamy and geitonogamy, in order to
understand their role in selfing.
Here, I present a comparative study of floral biology between two wind-pollinated sister
species, Oryza nivara and O. rufipogon, that have recently diverged in their mating system
(Morishima & Barbier, 1990; Vaughan, 1994). The goals of this study are to characterize
differences in floral biology that may contribute to the mating system, and to partition the role of
geitonogamy and autogamy in providing selfing. Additionally, I provide a mechanistic

105

framework for mating system divergence in these species. To my knowledge, this is the first
study of comparative floral biology between a wind-pollinated selfing and outcrossing group.
This line of research enhances my previous work with these species and provides a foundation
for a comprehensive understanding of mating system evolution in this system.
The sister species, Oryza rufipogon and O. nivara , are wetland grasses distributed
throughout the tropics and subtropics of Asia. Oryza nivara, is an annual that inhabits ephemeral
pools containing standing water throughout the monsoon season, but then desiccates during the
dry season (Vaughan, 1994). Conversely, O. rufipogon, is a perennial that inhabits lowland
marshes and wetlands that retain sufficient water levels for the plants to persist throughout the
year. Both species are self-compatible and hermaphroditic, but differ in their mating system.
Barbier (1989) measured outcrossing rates for one O. nivara and two O. rufipogon populations in
Thailand. Multilocus outcrossing rates were estimated with 8 isozyme loci based on the mixedmating model (Ritland & Jain, 1981). O. rufipogon had substantially higher outcrossing rates
(tm=0.50 and 0.56) than O. nivara (tm=0.04), which is consistent with earlier estimates based on
other approaches (Oka & Morishima, 1967). Phylogenetic analysis has indicated that O. nivara
has recently diverged from a common ancestor resembling O. rufipogon (Barbier et al., 1991;
Zhu & Ge, 2005), which is consistent with broader observations of annual selfers evolving from
perennial outcrossers (Barrett et al., 1996). Previously, I have mapped the genetic basis of
several adaptive traits that differ between these species, including three mating system traits:
anther length, inflorescence architecture, and inflorescence exsertion (Grillo et al., 2009). In the
present study, I expand upon this work by examining a broad range of floral morphological and
physiological traits that could potentially influence the mating system. A list of traits and their
putative function are outlined in Table 1. These traits are compiled from preliminary

106

observations of floral biology in the focal species and a review of relevant literature on the
subject.

Materials and Methods
An exhaustive survey of potential traits that may influence mating system divergence
between O. rufipogon and O. nivara was conducted. Unless otherwise mentioned, these data
were obtained from one accession of each species collected in India and described in Grillo et al
(2009). Based on previous greenhouse observations of both species, these two individuals were
selected as being representative of typical O. rufipogon and O. nivara populations. Statistical
differences between trait means were determined through one-way ANOVA, implemented in R.
Multivariate analysis of variance (MANOVA), implemented in JMP 10, was performed to
identify overall differences in floral phenotypes. Plants were grown and studied over a three
year period at Michigan State University in greenhouses and growth chambers. All the mating
system traits require plants to be in flower, and this was achieved by exposing plants to short day
length in the growth chamber.
Anther length, anther basal pore size, pollen size, stigma length, and stigma papillae
number were measured from freshly collected samples with a compound light microscope.
Length measurements were obtained with a lens ocular micrometer and were converted to the
appropriate units (µm or mm). Stigma color was measured qualitatively as purple or white.
Anther position was scored as ‘1’ if anthers fell below the flower in a pendent-like position, or as
‘0’ if they remained upright above the stigma. Inflorescence openness was measured by the
angles between primary branches and rachis. Inflorescence exsertion was measured by the
distance from the flag leaf node to the first primary branch node. For all traits the number of

107

measurements is given in Table 2. A subset of the traits (anther length, inflorescence exsertion,
and inflorescence openness) was measured from multiple populations of each species, including
five populations of O. nivara and two populations of O. rufipogon from Thailand, and one
population of each species from India (Table 3).
The timing of anther dehiscence was examined before anthesis. Under the constant
conditions of the growth chamber, both species began flowering 4 hours after first light. All
flowers of an inflorescence typically complete flowering within a 4 day period, thus each day a
subset of flowers open (Grillo per. obs.). On the final day of this period, approximately 10
flowers remain and there is a very high probability that they will open on that day. Thus, it is
reasonably possible to predict when a flower will open on the last day of flowering. When this
was the case I waited until the first flower opened for that day and then promptly opened the
remaining flowers. Anthers and stigmas of these flowers were collected and examined under a
microscope. Early dehiscence was scored as 0, 1, 2, 3,or 4 based on the relative number of
pollen grains on the stigma prior to flower opening (0 = 0, 1 = <20, 2 = 20-40, 3 = 40-80, 4 =
>80 pollen grains, respectively).
Pollen viability was measured by pollen staining with a fluorescent dye (RodriguezRiano & Dafni, 2000).

Pollen was collected at the following time intervals (minutes following

anthesis), 0, 10, 20, as well as prior to anthesis as described for ‘early dehiscence’. To store
pollen, anthers were harvested with forceps and then shaken over a microcentrifuge tube
containing 1mL of 10% sucrose. Storage of pollen in sucrose solution prevents against further
degradation and maintains pollen grains at the collection state (Rodriguez-Riano & Dafni, 2000).
A Fluorescein Diacetate (FDA) solution was created with 60mg of FDA in 30mL of deionized
water. Droplets of FDA solution were added to the pollen collection tubes (suspended in 1mL

108

10% sucrose) until the solution turned milky gray (Rodriguez-Riano & Dafni, 2000). The
stained pollen were then allowed to incubate for 5 minutes, before being viewed on a fluorescent
microscope (purple-indigo beam). Pollen grains were scored as being viable if they absorbed the
stain. For each measurement approximately 100 pollen grains were scored. Measurements at
each time point were repeated 5 times by collecting pollen from fresh flowers. Pollen staining
was attempted with Potassium Iodide, as has been conducted in O. sativa (Song et al., 2004);
however, this effort was unsuccessful due to consistent over-staining. Additionally, pollen
germination was attempted with a standard pollen germination media (Khatun & Flowers, 1995),
yet this was unsuccessful as well. Pollen staining indicates the presence of a cytoplasm and is
merely a proxy for pollen viability (Rodriguez-Riano & Dafni, 2000). To validate the pollen
staining results a pollination study was conducted. For this, approximately 30 minute-old pollen
was deposited on fresh stigmas (at anthesis) of O. rufipogon. The donor O. rufipogon flowers
were emasculated at anthesis prior to anther dehiscence, to prevent self-pollination.
Accomplishing this in O. nivara is highly unlikely and so this experiment was attempted only
with O. rufipogon as the maternal plant. This experiment was conducted 10 times for each
species.
Stigma receptivity was measured by the peroxide test (Kearns & Inouye, 1993). Stigmas
were collected and placed in a droplet of 15% hydrogen peroxide on a microscope slide. The
slides where then viewed under a dissecting microscope and the number of bubbles emitted over
a 30 second period was recorded. Hydrogen peroxide reacts with enzymes on the surface the
stigma, generating bubbles of oxygen gas. Stigma receptivity is indicated by the number bubbles
emitted (i.e. high receptivity yields lots of bubbles). Stigmas were collected at the following
time intervals (minutes following anthesis), 0, 15, 30, 45, 60, as well as prior to anthesis as

109

described above. To verify actual stigma receptivity, stigmas of O. rufipogon were hand
pollinated with fresh pollen 45 minutes post anthesis. The flowers were emasculated prior to
anther dehiscence and the experiment was carried out 10 times. O. nivara was not included in
this experiment for reasons indicated above (see pollen viability). As a control fresh O.
rufipogon stigmas were pollinated with fresh pollen at the time of anthesis.
In order to distinguish between geitonogamy and autogamy, manipulations of floral
number were conducted (Schoen & Lloyd, 1992). To determine the rate of autogamy
inflorescences were trimmed so that only one flower remained. These plants were then isolated
in separate greenhouse rooms and allowed to flower. Plants were marked on the day of anthesis,
and approximately one week later, flowers were collected to determine if there was successful
self-fertilization. To determine the rate of geitonogamy, plants were cut back so that only a
single inflorescence was allowed to flower, and again plants were isolated from each other in
separate greenhouse rooms. After all flowers had opened, the inflorescence was bagged, and
seed set was scored once they had matured. Under these conditions, there is the opportunity for
both autogamy and geitonogamy. Additionally, plants with several inflorescences were allowed
to flower in a greenhouse room with dozens of simultaneously flowering plants and abundant
wind generated by fans.

110

Table 4.1. Potential mating system traits and their putative function.
These traits were compiled from examination of literature and detailed observations of floral
biology in O. rufipogon and O. nivara.

Trait
Inflorescence
openness
Inflorescence
exsertion
Anther length
Filament length
Anther position
Early dehiscence
Basal pore length
Apical pore length
Pollen size
Pollen viability
Stigma length
Stigma papillae
density
Stigma color
Stigma receptivity
Dichogamy

Putative Function
Open inflorescences disperse and capture pollen in the wind.
Exserted inflorescences from the foliage promote pollen
dispersal/capture
Long anthers contain more pollen to promote outcrossing
Long filaments may allow anthers to fall below the flower
Position in relation to the stigma controls opportunity for selfing
Anthers release pollen prior to anthesis promoting autogamous selfing
Pollen deposited out the basal pore fall onto the stigma causing selfing
Pollen dispensed by apical pore may promote outcrossing
Can influence pollen viability and aerodynamic properties in the wind
Prolonged pollen viability can promote outcrossing over the large
distance
Large stigmas for capturing pollen in the wind
Increases surface area for pollen capture to promote outcrossing
Anthocyanins may protect against desiccation and promote
outcrossing
Duration of receptivity restricts the ability for outcrossing
Temporal separation of male and female functions

Results
Trait means are provided in Table 2 for all of the morphological and physiological traits
under investigation. Additionally, Table 3 gives the values for anther length, inflorescence
exsertion, and inflorescence openness for additional populations of each species. Overall, there
was a significant difference between the multivariate floral phenotypes of O. nivara and O.
rufipogon (MANOVA: F8,45 = 7.5; P<0.0001).
111

Unlike O. rufipogon, early dehiscence occurred prior to floret opening in O. nivara,
resulting in pollen deposited on the stigma (Fig. 1). Early dehiscence was scored based on the
relative number of pollen grains deposited on the stigma prior to anthesis. O. nivara had an early
dehiscence value of 2.33, compared to 0.1 in O. rufipogon. While this difference is significant, it
is likely to be an underestimate for O. nivara. A measurement of ‘0’ could either indicate that
the anthers did not dehisce and deposit pollen on the stigma, or that the particular flower was
sampled too early to witness early dehiscence. Alternatively, the proportion of flowers with any
pollen deposited on the stigma was examined. In this case, 55% of O. nivara flowers (n=20)
exhibited early dehiscence, compared to 6.4% (n=31) in O. rufipogon. In O. rufipogon, early
dehiscence was observed in only two flowers, and in both cases few pollen grains were
deposited. These instances could have been caused by contamination from pollen in the air, or
physical damage of the stigma during manual flower opening. Nevertheless, even if this is not
the case, early dehiscence occurs rarely in O. rufipogon.
In a time course study, pollen viability did not differ between species until after 20
minutes, at which time O. rufipogon had an average of 62% pollen viability, compared to only
21% in O. nivara (Fig. 2). This observation is consistent with results presented by Song et al
(2004), in which O. rufipogon was approximately 38% viable after 20 minutes, compared to 0%
in O. sativa. The ability for pollen grains to absorb a stain is merely a proxy for the actual pollen
viability, and is likely to be an overestimate (Song et al 2004). A functional experiment was
conducted by placing 30-minute old pollen on fresh stigmas. In this experiment there was no
significant difference between species (60% and 70% in O. nivara and O. rufipogon,
respectively). The large number of pollen grains deposited on the stigma may have allowed for
seed production, despite the presumably low pollen viability of O. nivara .

112

Oryza rufipogon and O. nivara had similar rates of decrease in stigma receptivity
following anthesis (Fig. 3). As indicated by the assay, stigmas have low levels of receptivity
after 45 minutes. To test the functional significance of the stigma receptivity values, a handpollination study was conducted with stigmas 45 minutes post anthesis. Only one O. rufipogon
stigma that was hand pollinated resulted in the production of a seed (n=10), validating the stigma
receptivity assay. In contrast, all of the controls resulted in seed production when stigmas were
hand-pollinated at the time of anthesis (n=10). Similar experiments were not conducted in O.
nivara due to the high occurrence of prior selfing (see discussion) and the fact that both species
generate identical rates of decline in stigma receptivity. This rapid decrease in receptivity is not
unexpected as flowers are open for just 1 hour. In contrast, stigma receptivity did differ between
species prior to anthesis, with O. nivara being considerably more receptive (Fig. 3).
By manipulating flower number, I can partition the selfing rate into contributions from
geitonogamy and autogamy (Fig. 4). If autogamous selfing is predominant, one would predict a
consistent level of seed set regardless of flower number. Alternatively, if geitonogamy is
predominant, seed set should increase with increases in flower number/pollen availability
(Schoen & Lloyd, 1992). In the treatments with only a single flower, seed set averaged 72.5%
and 10.3% in O. nivara and O. rufipogon, respectively. Since there is only one flower, all
resulting seed set is generated through autonomous selfing. In the treatments with a single
inflorescence, there is the opportunity for geitonogamy and autogamy. Under these conditions
there is a significant increase in the average seed set for both species, however, seed set of O.
rufipogon (37.7%) did not approach levels comparable to O. nivara (85.4%). These results
clearly demonstrate that autogamy is the predominant form of selfing in O. nivara. While
geitonogamy did provide additional seed set, this may be an artifact of damage induced by the

113

floral manipulations. In the single-flower experiments the inflorescence was cut back, which
could have had detrimental effects on the remaining flower. Conversely, there were no such
manipulations in the single-inflorescence experiment. This does not impact interpretation of the
role for autogamy between species, as both experienced the same manipulations in the singleflower experiments, but yielded drastically different results. In the case of O. rufipogon,
geitonogamy increased seed set in the single-inflorescence experiment, although the effects were
limited - likely due to pollen limitation. This is demonstrated in the whole-plant experiments,
where there was vastly more available pollen. It should be noted that there is opportunity for
autogamy, geitonogamy, and outcrossing in this treatment, however; this is inconsequential as
pollen limitation is the key factor being addressed. Under these conditions, seed set in O.
rufipogon did not differ significantly from O. nivara.

Figure 4.1. Functional traits of Oryza species.
Functional traits controlling mating system differentiation in O. nivara and O. rufipogon. O.
nivara traits are in the left column and O. rufipogon on the right. A) O. nivara floret
demonstrating anther size. B) O. rufipogon floret demonstrating anther size. C) O. nivara
panicle shape and exertion. D) O. rufipogon panicle shape and exertion. E) O. nivara stigma
length, papillae density and stigma color. F) O. rufipogon stigma length, papillae density and
stigma color. G. Basal pore of O. nivara. H. Early dehiscence of O. nivara, stigma coated in
pollen prior to floret opening.

114

Figure 4.1 (cont’d).

115

Discussion
The development of polymorphic markers and statistical methodologies has allowed for
the reliable estimation of outcrossing rates (t) in plants (Ritland & Jain, 1981; Ritland, 2002). It
is widely accepted that outcrossing rates below 20% (t ≤ 0.2) and above 80% (t ≥ 0.8), represent
predominantly selfing and outcrossing mating systems, respectively (Goodwillie et al., 2005).
The estimate of multilocus outcrossing rate in O. nivara (tm=0.04) confirms its status as being
predominantly selfing. As an annual with small population sizes (Morishima & Barbier, 1990;
Vaughan, 1994; Zhou et al., 2008) selection for reproductive assurance likely drove the
evolution of selfing in Oryza nivara. Despite being substantially more outcrossing than O.
nivara; mating system estimates from O. rufipogon (tm=0.50 and 0.56) would be classified as
mixed-mating (Barbier, 1989). It is possible that these estimates may be less than the actual
outcrossing rate, due to the occurrence of geitonogamy and/or biparental inbreeding.
Geitonogamy is thought to be responsible for a significant amount of the variation in outcrossing
rates in plants (Barrett, 2003). This may be the case in O. rufipogon, as I have demonstrated that
autogamy is rare in this species and large clonal ramets are common in nature (Xie et al., 2001),
providing ample opportunity for geitonogamy. Similarly, outcrossing events between close
relatives (biparental inbreeding) can result in ‘apparent selfing’, as is observed in selfincompatible species with outcrossing estimates less than 1 (Ritland, 1984; Goodwillie et al.,
2005). Populations of O. rufipogon can be quite large (Oka & Morishima, 1967; Vaughan, 1994;
Zheng & Ge, 2010), however increasingly their habitats are being fragmented and disturbed. For
example, Akimoto et al. (1999) have monitored one such population in Thailand that became
severely fragmented. During a ten year period the observed outcrossing rate decreased from
0.538 to 0.236, due in part to an increase in biparental inbreeding. Outcrossing rates should be

116

measured in more populations of each species to better estimate the mating system. Regardless,
I am confident that O. rufipogon possess attributes consistent with an outcrossing strategy,
despite the observed estimates that would be considered mixed-mating.

Table 4.2. Trait means between O. rufipogon and O. nivara.
Trait means between O. nivara and O. rufipogon, standard deviation given in parenthesis, n=the
number of measurements. Pollen viability is measured at 20 minutes post anthesis. Stigma
receptivity is at an ‘early dehiscence’ stage. Significance tested with one-way anova, ∗, P<0.05;
∗∗, P<0.01; ∗∗∗, P<0.001.
Trait
Anther length (mm)
Anther position (%)
Basal pore length (mm)
Early dehiscence
Pollen size (mm)
Pollen viability (%)
Stigma length (mm)
Stigma papillae density
Stigma pigmentation
Stigma receptivity
Inflorescence openness (°)
Inflorescence exsertion (cm)

O. nivara
2.19 (0.22) n=30
40 (n=20)
0.35 (0.11) n=32
2.33 (2.0) n=20
0.047 (0.002) n=54
19.2 (2.5) n=5
1.2 (0.19) n=30
29.8 (3.43) n=20
white
3.25 (0.5) n=5
5.1 (4.1) n=20
1.2 (0.8) n=20

O. rufipogon
4.39 (0.46) n=30
71.4 (n=20)
0.14 (0.07) n=30
0.1 (0.3) n=31
0.031 (0.003) n=50
42.2 (2.4) n=5
3.02 (0.3) n=30
38.3 (3.15) n=20
purple
0.428 (0.46) n=5
45.6 (10.2) n=20
13.9 (6.3) n=20

Significance
***
**
***
**
***
*
***
**
***
***
***
***

In wind-pollinated plants, male fitness is thought to be directly related to investment in
pollen production (Charlesworth & Charlesworth, 1981), and because of this male-biased sex
allocation is commonly observed (Burd & Allen, 1988). As an illustration, anthers of O.
117

rufipogon are approximately 60% of total flower length. Anthers of O. nivara are half the length
of O. rufipogon (Table 2) and likely house half the amount of pollen, although there are slight
differences in pollen size (see below). This decrease in anther length fits predictions of sex
allocation theory, with decreased investment in male function as selfing increases (Charlesworth
& Morgan, 1991).

Figure 4.2. Pollen viability time course.
Pollen viability time course for O. nivara and O. rufipogon. Percent pollen viability measured
using FDA florescent staining. Pollen was collected at anther dehiscence and was fixed in 10%
sucrose at 10 minute intervals. For the ‘Pre-anthesis’ timepoint, anthers from unopened florets
were collected at peak flowering.

Pollen

O. rufipogon
O. nivara

0

10

15
Time

118

20

Oryza rufipogon and O. nivara possess poricidal anthers that open at pores on each end.
There is no difference in apical pore size between these species (Table 2). However, there is a
considerable difference in basal pore size, with O. nivara having a basal pore twice that of O.
rufipogon. In a study examining different breeds of cultivated rice, large basal pore size resulted
in increased self-pollen deposited on the stigma (Matsui & Kagata, 2003). In O. rufipogon,
anthers hang in a pendent-like fashion, below the flower (Fig. 5 and Table 2). In this case, pollen
is predominantly dispensed through the apical pore due to gravity. Alternatively, anthers of O.
nivara remain upright and pollen falls through the basal pore. The basal pore of O. rufipogon is
quite small and often does not open, as it likely serves no function. In these species, the stamen
filaments are the same length and the position of anthers upon anthesis (pendent or upright) is a
likely result of pollen mass.

Figure 4.3. Stigma receptivity time course.
Stigma viability time course for O. nivara and O. rufipogon. Stigmas were collected at 15
minute time intervals after floret opening. Stigma receptivity was measured through a
peroxidase detection assay by counting the number of bubbles emitted from the stigma surface
during a 30 second time interval. Stigma viability scored as: 0=0 1= <50, 2=50-100, 3=100-200,
4=>200 bubbles emitted.

119

Stigma receptivity

Figure 4.3 (cont’d).

O. nivara
O. rufipogon

Pre

0

15

30

45

60

Time
Furthermore, in O. nivara, the basal pore opens and pollen is released before the flower
opens, contributing to prior selfing (Fig. 5). Prior selfing is autonomous selfing that occurs
before the opportunity for the receipt of outcross pollen (Lloyd & Schoen, 1992). In animalpollinated systems, this happens due to early spatial and developmental overlap between male
and female functions and can occur prior to anthesis (Fishman & Wyatt, 1999; Fishman et al.,
2002). Matsui and Kagata (2003) observed early dehiscence in cultivated rice. Similarly, O.
nivara flowers that are artificially opened reveal considerable pollen deposited on the stigma
(Fig. 1). This phenomenon is not observed in O. rufipogon, and is likely a necessary step for
autogamous selfing in O. nivara.
Unlike animal-pollinated plants, pollen in wind-pollinated taxa lack ornamentation and
exhibit a narrower range of pollen sizes (Harder, 1998; Ackerman, 2000). Pollen size evolution
is dictated by the aerodynamic properties of wind, in order to promote successful pollen dispersal
and receipt (Friedman & Barrett, 2009b). For example small pollen grains have low settling
velocity, allowing them to travel farther, and low inertia to facilitate removal from anthers

120

(Niklas, 1985). Oryza rufipogon has smaller pollen grains which may be beneficial for
outcrossing over long distances. The larger pollen grains of O. nivara may be the result of
relaxed selection on pollen size. Although, it is possible that larger pollen grains would settle out
of the airstream more quickly; better ensuring deposition on the stigma within the same flower.
Typically, anthesis in wind pollinated taxa is highly synchronous and abbreviated when
compared to animal pollinated systems (Dowding, 1987). Accordingly, pollen of windpollinated species is only viable for a short duration (Dafni & Firmage, 2000). Pollen viability in
O. nivara decreased rapidly, and after 20 minutes (post anther dehiscence) was only 1/3 as viable
as O. rufipogon (Fig. 2). The loss of viability in O. nivara closely resembled that observed in O.
sativa by Song et al. (2004). Both, O. nivara and O. sativa are highly selfing, and are unlikely to
experience selection for prolonged pollen viability. Conversely, O. rufipogon pollen retains high
viability after 20 minutes, allowing for successful fertilization after pollen dispersal (Fig. 2).
Song (2001) identified similar levels of pollen viability for O. rufipogon.
Grasses possess unique feathery stigmas that are exserted from the flower (Friedman &
Harder, 2004). Wind tunnel experiments have demonstrated that feathery stigmas have higher
pollen collection efficiencies than solid stigmas; as they maintain a high surface area, but
produce a thinner boundary layer (Niklas, 1985). Stigmas in the focal species differ in their
length, papillae number, pigmentation, and receptivity (Table 2). Oryza rufipogon has larger
stigmas with a higher density of papillae, than O. nivara, which may aid in capturing outcross
pollen. Additionally stigmas of Oryza rufipogon are dark purple in coloration, due the presence
of anthocyanins (Fig. 1). Anthocyanins can function to protect plant tissue from environmental
stress, including drought and oxidative damage from UV exposure (Stapleton & Walbot, 1994;
Hughes et al., 2010). In order to ensure successful pollen germination, stigmas must maintain a

121

specialized enzymatic surface. Wind-pollinated stigmas are likely prone to desiccation and UV
radiation by being exserted outside the flower. Stigma pigmentation in O. rufipogon may be an
adaptation to promote successful pollen germination and outcrossing. Stigmas of O. nivara lack
pigmentation and are white. As a selfer, O. nivara would not experience selection to maintain
pigmentation and allocating resources to anthocyanins could incur a cost. Variation in stigma
color is quite common across grass species (Grillo per. obs.) and I hypothesize is associated with
the mating system. Stigma receptivity also differed between the focal species, prior to anthesis.
O. nivara has considerably higher stigma receptivity at this stage which is expected if
fertilization occurs due to prior selfing. Overall, there is a diversity of stigma traits in grasses
that have not been examined in regards to pollination and mating system (Friedman & Harder,
2004). O. rufipogon and O. nivara encompass much of this diversity and serve as a potential
model for understanding the functional relevance of stigma characteristics.

Table 4.3. Trait means for additional populations of each species
Species

Location

O. nivara

Thailand
(Changmai)
Thailand
(Changmai)
Thailand
(Kahnken)
Thailand
(Rayong)
Thailand
(Rayong)
India (Cuttack)
Thailand
(Changmai)
Thailand
(Changmai)
India (Cuttack)

O. rufipogon

Anther length
(mm)
2.60 (0.20) n=8

Inflorescence
openness (°)
21.3 (3.1) n=10

Inflorescence
exsertion (cm)
6.1 (1.3) n=10

2.31 (0.28) n=15

12.4 (3.8) n=10

5.8 (1.8) n=10

2.12 (0.19) n=20

15.4 (6.1) n=9

7.4 (2.3) n=9

2.43 (0.32) n=6

27.2 (4.3) n=10

5.5 (1.1) n=10

2.52 (0.31) n=11

20.7 (6.8) n=11

6.3 (2.1) n=11

2.24 (0.23) n=6
4.93 (0.51) n=10

8.2 (2.1) n=4
71.4 (12.3) n=4

3.7 (0.9) n=4
16.3 (1.1) n=4

4.51 (0.43) n=10

62.3 (10.7) n=4

17.8 (1.7) n=4

4.42 (0.44) n=6

50.6 (4.8) n=3

16.5 (1.3) n=3

122

Grasses display a range of inflorescence architectures, ranging from diffuse to compact,
representing different means for pollen removal and receipt (Friedman & Harder, 2005;
Friedman & Barrett, 2009b). Wind tunnel experiments have demonstrated that compact panicles
act to obstruct airflow, thereby creating eddies on the leeward side of the panicle. This allows
pollen to settle out of the airstream and become deposited on stigmas. Alternatively, diffuse
panicles disturb airflow less, and pollen simply comes in direct contact by windward stigmas
(Niklas, 1987). By disturbing airflow less, diffuse panicles may be for more effective at pollen
dispersal (Friedman & Harder, 2005). Surveys of O. rufipogon populations revealed consistent
openness of panicles that may promote pollen dispersal in such a way (Table 3). On the whole
O. nivara panicles are quite compact, although there is variation for this trait (Table 3).
Certainly, the close proximity of simultaneously open flowers within a compact panicle presents
an opportunity for geitonogamy.
In addition to inflorescence architecture, plant height can influence pollination efficiency.
In general, horizontal wind speed increases with height above ground (Burd & Allen, 1988).
Therefore, taller plants experience higher wind speeds than shorter ones in the same habitat. For
this reason, grasses generally protrude their inflorescence above the boundary layer of their
foliage (i.e. panicle exsertion) to promote pollen dispersal (Friedman & Harder, 2004; Friedman
& Harder, 2005). Tiller length is greater in O. rufipogon (Grillo et al., 2009), although this may
not affect plant height as O. rufipogon inhabits marshes and tillers essentially float. Panicle
exsertion is likely more important in this regard for the focal species. Panicles of O. nivara have
limited exsertion (Table 2, Fig. 1) and are essentially nested within foliage, which is unlikely to
provide effective pollen dispersal. The reverse is true for O. rufipogon, consistent with an
outcrossing mating system.

123

Self-compatible plants with many flowers in anthesis simultaneously present an
opportunity for geitonogamous selfing (Dejong et al., 1993). This is surely the case in grasses
that display highly synchronous bouts of flowering and often have large plant sizes (Friedman &
Barrett, 2009b). For this reason, it is suspected that avoidance of geitonogamy has driven the
common occurrence of dioecy and self-incompatibility in grasses (Friedman & Harder, 2005;
Friedman & Barrett, 2009b). Nonetheless, there are certainly numerous self-compatible
hermaphroditic grasses that are prone to this mode of selfing. In order to distinguish between
geitonogamy and autogamy, manipulative experiments of floral number are needed (Schoen and
Lloyd 1992). Such experiments are rarely carried out, but the few examples reveal high levels of
geitonogamous selfing (Harder & Barrett, 1995; Eckert, 2000; Dorken et al., 2002; Karron et al.,
2004). I have demonstrated that in O. nivara, high levels of selfing are achieved through
autogamy. In contrast self-fertilization in O. rufipogon is due to geitonogamy and increases with
flower number (Fig. 4). As described previously, geitonogamy in O. rufipogon could influence
the observed outcrossing rate.
The transition from outcrossing to selfing is regarded as the most common trend in plant
evolution (Stebbins, 1974; Barrett, 2003). This is exemplified by the recent divergence of O.
nivara (selfing) from an O. rufipogon-like (outcrossing) ancestor (Barbier et al., 1991; Zhu &
Ge, 2005). Therefore, it is important to identify the functional traits that underlie this shift, and
to consider the means by which they have evolved. In O. nivara a suite of traits act in concert to
provide autogamous self-fertilization: the timing of anther dehiscence, anther basal pore size, and
stigma receptivity. Before the flower even opens the anthers of O. nivara become dehiscent with
the opening of the basal pore (Fig. 5). Pollen is released through the basal pore, which is quite
large in this species. In turn, the stigma is receptive at this time as well. Upon anthesis the

124

anthers in O. nivara remain upright and continue to deposit pollen onto the stigma of the same
flower. Thus, autogamy is achieved in O. nivara through prior selfing.

Figure 4.4. Autogamy and geitonogamy.
Partionioning of selfing into autogamy and geitonogamy through floral number manipulations.
In the ‘single flower’ treatment inflorescences were trimmed to a single flower, allowing only for
autogamous selfing. In the ‘single inflorescence’ treatment there is opportunity for autogamy
and geitonogamy. In the ‘many plants’ treatment numerous plants flowered simultaneously and

Seed set

so there was no pollen limitation and high seed sets were achieved.

Single
Flower

Single
Panicle

125

Whole
Plant

Based on the properties of the mating system traits, it is possible to predict the order in
which they may have evolved. Here I describe two possible alternatives for the evolution of
autogamy: 1) an indirect path where an increase in selection for geitonogamy provides a bridge
to autogamy, or 2) the direct evolution of autogamy.
Selfing likely became beneficial in this system concurrently with a shift to seasonally dry
habitats that do not allow plants to persist as perennials. Populations of O. nivara are often quite
small and potential mates can be scarce today, as they likely were for the original founders of
these habitats. Under such conditions there would be selection for reproductive assurance
through selfing. A similar situation is observed in populations that exhibit increased selfing rates
along range margins, where there are few mates available (Barrett, 2003). For the O. rufipogonlike ancestor, selfing would be achieved through geitonogamy in these circumstances. Changes
in inflorescence architecture could influence the rates of geitonogamy and variation for this trait
is common in natural populations (Table 3). Selection for compact panicles would thereby
provide close proximity of flowers and aerodynamic properties favorable for high
geitonogamous selfing. Once appreciable levels of self-fertilization were achieved through
geitonogamy, there would be selection for reduced investment in allocation to male function.
Large anthers are likely to be energetically costly and excess pollen production (for outcrossing)
would be unnecessary under these conditions. As anthers began to shorten they would weigh
less and maintain a more upright position. At this point gravity would no longer allow pollen to
exit through the apical pore; rather the basal pore would be more important in this regard.
Selection for large basal pores would allow pollen to fall directly onto the stigma of the same
flower, thus providing autogamous selfing. After this point, alterations in the timing of anther

126

dehiscence and stigma receptivity would result in prior selfing. In this way geitonogamy would
serve as a bridge to autogamous selfing.

Figure 4.5. Flower caricatures of O. nivara and O. rufipogon
Floral caricatures of O. rufipogon and O. nivara, pre- and post-anthesis. O. nivara has smaller
anthers than O. rufipogon and the anther pores open, releasing pollen through the basal pore
prior to anthesis. Upon anthesis the anthers remain upright and dispense pollen through the basal
pore onto the stigma. O. rufipogon has much larger anthers that do not dehisce prior to anthesis.
Upon anthesis the large basal pores hang pendent-like below the flower and pollen is dispensed
through the apical pore.

127

Alternatively, autogamy may have evolved directly in an O. rufipogon-like ancestor. In
this scenario selfing would have to occur prior to anthesis, as the large basal pores of O.
rufipogon would fall below the flower upon opening. Therefore, anther dehiscence would be the
first trait to evolve. Early anther dehiscence would be unlikely to provide reliable autogamous
selfing alone and would have to be accompanied closely by an increase in basal pore size to
deposit pollen on the stigma. Additionally, earlier stigma receptivity would enhance the rate of
autogamy. Once autogamous selfing became predominant there would be selection for reduced
anther size.
In either case traits that promote outcrossing are expected to evolve after autogamy has
been established. These traits include pollen viability and stigma characteristics (length, papillae
density, pigmentation).
In the work presented here, I have identified traits associated with a shift in mating
system between O. nivara and O. rufipogon. Additionally, I have demonstrated that autogamy is
the mode of selfing displayed by O. nivara. This research sets the stage for future experiments
to gain a more comprehensive understanding of mating system evolution in this system. In the
summer of 2013 I will conduct experiments aimed at 1) identifying the specific functional traits
controlling the degree of autogamous selfing and 2) mapping the genetic basis of mating system
trait variation including the autogamous selfing rate. To date I have generated a mapping
population with 200 F2 lines and at least 4 clones of each line. This summer I will finish
measuring the mating system traits in the F2 population. Additionally, I will determine the
autogamous selfing rate by manipulating flower number as described here. I will also complete
genotyping these F2s with microsatellite markers that I have previously developed (Grillo et al.,
2009). Through these pursuits I will be able to determine which traits are associated with the

128

autogamous selfing rate. I will also characterize the genetic architecture of mating system trait
differentiation and identify which specific trait loci co-localize with mating system (autogamy)
loci.

129

LITERATURE CITED

130

LITERATURE CITED

Ackerman JD. 2000. Abiotic pollen and pollination: ecological, functional, and evolutionary
perspectives. Plant Systematics and Evolution 222(1-4): 167-185.
Ågren J, Schemske DW. 1993. OUTCROSSING RATE AND INBREEDING DEPRESSION
IN 2 ANNUAL MONOECIOUS HERBS, BEGONIA-HIRSUTA AND BSEMIOVATA. Evolution 47(1): 125-135.
Aide TM. 1986. THE INFLUENCE OF WIND AND ANIMAL POLLINATION ON
VARIATION IN OUTCROSSING RATES. Evolution 40(2): 434-435.
Akimoto M, Shimamoto Y, Morishima H. 1999. The extinction of genetic resources of Asian
wild rice, Oryza rufipogon Griff.: A case study in Thailand. Genetic Resources and Crop
Evolution 46(4): 419-425.
Armbruster WS, Mulder CPH, Baldwin BG, Kalisz S, Wessa B, Nute H. 2002. Comparative
analysis of late floral development and mating-system evolution in Tribe Collinsieae
(Scrophulariaceae SL). American Journal of Botany 89(1): 37-49.
Barbier P. 1989. GENETIC-VARIATION AND ECOTYPIC DIFFERENTIATION IN THE
WILD RICE SPECIES ORYZA-RUFIPOGON .1. POPULATION DIFFERENTIATION
IN LIFE-HISTORY TRAITS AND ISOZYMIC LOCI. Japanese Journal of Genetics
64(4): 259-271.
Barbier P, Morishima H, Ishihama A. 1991. PHYLOGENETIC-RELATIONSHIPS OF
ANNUAL AND PERENNIAL WILD-RICE - PROBING BY DIRECT DNA
SEQUENCING. Theoretical and Applied Genetics 81(5): 693-702.
Barrett SCH. 2003. Mating strategies in flowering plants: the outcrossing-selfing paradigm and
beyond. Philosophical Transactions of the Royal Society of London Series B-Biological
Sciences 358(1434): 991-1004.
Barrett SCH. 2010a. Darwin's legacy: the forms, function and sexual diversity of flowers.
Philosophical Transactions of the Royal Society B-Biological Sciences 365(1539): 351368.

131

Barrett SCH. 2010b. Understanding plant reproductive diversity. Philosophical Transactions of
the Royal Society B-Biological Sciences 365(1537): 99-109.
Barrett SCH, Harder LD, Worley AC. 1996. The comparative biology of pollination and
mating in flowering plants. Philosophical Transactions of the Royal Society of London
Series B-Biological Sciences 351(1345): 1271-1280.
Belaoussoff S, Shore JS. 1995. FLORAL CORRELATES AND FITNESS CONSEQUENCES
OF MATING-SYSTEM VARIATION IN TURNERA-ULMIFOLIA. Evolution 49(3):
545-556.
Burd M, Allen TFH. 1988. SEXUAL ALLOCATION STRATEGY IN WIND-POLLINATED
PLANTS. Evolution 42(2): 403-407.
Chang SM, Rausher MD. 1999. The role of inbreeding depression in maintaining the mixed
mating system of the common morning glory, Ipomoea purpurea. Evolution 53(5): 13661376.
Charlesworth B. 1992. EVOLUTIONARY RATES IN PARTIALLY SELF-FERTILIZING
SPECIES. American Naturalist 140(1): 126-148.
Charlesworth D, Charlesworth B. 1981. ALLOCATION OF RESOURCES TO MALE AND
FEMALE FUNCTIONS IN HERMAPHRODITES. Biological Journal of the Linnean
Society 15(1): 57-74.
Charlesworth D, Charlesworth B. 1987. INBREEDING DEPRESSION AND ITS
EVOLUTIONARY CONSEQUENCES. Annual Review of Ecology and Systematics 18:
237-268.
Charlesworth D, Morgan MT. 1991. ALLOCATION OF RESOURCES TO SEX
FUNCTIONS IN FLOWERING PLANTS. Philosophical Transactions of the Royal
Society of London Series B-Biological Sciences 332(1262): 91-102.
Conner JK 2006. Ecological genetics of floral evolution. In: Harder LD, Barrett SCH eds. The
Ecology and Evolution of Flowers: Oxford University Press.
Crane PR 1986. Form and function in wind dispersed pollen. In: Blackmore S, Ferguson IK eds.
Pollen and Spores: Form and Function. London: Academic, 179-202.

132

Cresswell JE, Krick J, Patrick MA, Lahoubi M. 2010. The aerodynamics and efficiency of
wind pollination in grasses. Functional Ecology 24(4): 706-713.
Dafni A, Firmage D. 2000. Pollen viability and longevity: practical, ecological and evolutionary
implications. Plant Systematics and Evolution 222(1-4): 113-132.
Darwin C. 1876. The effects of cross and self fertilisation in the vegetable kingdom. London,
UK: John Murray Publishers.
Dejong TJ, Waser NM, Klinkhamer PGL. 1993. GEITONOGAMY - THE NEGLECTED
SIDE OF SELFING. Trends in Ecology & Evolution 8(9): 321-325.
Dole JA. 1990. ROLE OF COROLLA ABSCISSION IN DELAYED SELF-POLLINATION OF
MIMULUS-GUTTATUS (SCROPHULARIACEAE). American Journal of Botany
77(11): 1505-1507.
Dorken ME, Friedman J, Barrett SCH. 2002. The evolution and maintenance of monoecy and
dioecy in Sagittaria latifolia (Alismataceae). Evolution 56(1): 31-41.
Doust AN, Kellogg EA. 2002. Inflorescence diversification in the panicoid "bristle grass" clade
(Paniceae, Poaceae): Evidence from molecular phylogenies and developmental
morphology. American Journal of Botany 89(8): 1203-1222.
Dowding P. 1987. WIND POLLINATION MECHANISMS AND AEROBIOLOGY.
International Review of Cytology-a Survey of Cell Biology 107: 421-437.
Eckert CG. 2000. Contributions of autogamy and geitonogamy to self-fertilization in a massflowering, clonal plant. Ecology 81(2): 532-542.
Fisher RA. 1941. Average excess and average effect of a gene substitution. Annals of Eugenics
11: 53-63.
Fishman L, Kelly AJ, Willis JH. 2002. Minor quantitative trait loci underlie floral traits
associated with mating system divergence in Mimulus. Evolution 56(11): 2138-2155.
Fishman L, Wyatt R. 1999. Pollinator-mediated competition, reproductive character
displacement, and the evolution of selfing in Arenaria uniflora (Caryophyllaceae).
Evolution 53(6): 1723-1733.
133

Friedman J, Barrett SCH. 2009a. The consequences of monoecy and protogyny for mating in
wind-pollinated Carex. New Phytologist 181(2): 489-497.
Friedman J, Barrett SCH. 2009b. Wind of change: new insights on the ecology and evolution
of pollination and mating in wind-pollinated plants. Annals of Botany 103(9): 1515-1527.
Friedman J, Harder LD. 2004. Inflorescence architecture and wind pollination in six grass
species. Functional Ecology 18(6): 851-860.
Friedman J, Harder LD. 2005. Functional associations of floret and inflorescence traits among
grass species. American Journal of Botany 92(11): 1862-1870.
Goodwillie C. 1999. Multiple origins of self-compatibility in Linanthus section Leptosiphon
(Polemoniaceae): Phylogenetic evidence from internal-transcribed-spacer sequence data.
Evolution 53(5): 1387-1395.
Goodwillie C, Kalisz S, Eckert CG 2005. The evolutionary enigma of mixed mating systems in
plants: Occurrence, theoretical explanations, and empirical evidence. Annual Review of
Ecology Evolution and Systematics. Palo Alto: Annual Reviews, 47-79.
Goodwillie C, Sargent RD, Eckert CG, Elle E, Geber MA, Johnston MO, Kalisz S, Moeller
DA, Ree RH, Vallejo-Marin M, Winn AA. 2010. Correlated evolution of mating
system and floral display traits in flowering plants and its implications for the distribution
of mating system variation. New Phytologist 185(1): 311-321.
Grillo MA, Li CB, Fowlkes AM, Briggeman TM, Zhou AL, Schemske DW, Sang T. 2009.
Genetic architecture for the adaptive origin of annual wild rice, Oryza nivara. Evolution
63(4): 870-883.
Hamrick JL, Godt MJW. 1996. Effects of life history traits on genetic diversity in plant
species. Philosophical Transactions of the Royal Society of London Series B-Biological
Sciences 351(1345): 1291-1298.
Hamrick JL, Loveless MD. 1987. ASSOCIATIONS BETWEEN THE BREEDING SYSTEM
AND THE GENETIC-STRUCTURE OF TROPICAL TREE POPULATIONS. American
Journal of Botany 74(5): 642-642.
Harder LD. 1998. Pollen-size comparisons among animal-pollinated angiosperms with different
pollination characteristics. Biological Journal of the Linnean Society 64(4): 513-525.
134

Harder LD, Barrett SCH. 1995. MATING COST OF LARGE FLORAL DISPLAYS IN
HERMAPHRODITE PLANTS. Nature 373(6514): 512-515.
Holtsford TP, Ellstrand NC. 1992. GENETIC AND ENVIRONMENTAL VARIATION IN
FLORAL TRAITS AFFECTING OUTCROSSING RATE IN CLARKIATEMBLORIENSIS (ONAGRACEAE). Evolution 46(1): 216-225.
Hughes NM, Reinhardt K, Feild TS, Gerardi AR, Smith WK. 2010. Association between
winter anthocyanin production and drought stress in angiosperm evergreen species.
Journal of Experimental Botany 61(6): 1699-1709.
Husband BC, Schemske DW. 1996. Evolution of the magnitude and timing of inbreeding
depression in plants. Evolution 50(1): 54-70.
Igic B, Kohn JR. 2006. The distribution of plant mating systems: Study bias against obligately
outcrossing species. Evolution 60(5): 1098-1103.
Igic B, Lande R, Kohn JR. 2008. Loss of self-incompatibility and its evolutionary
consequences. International Journal of Plant Sciences 169(1): 93-104.
Jain SK. 1976. EVOLUTION OF INBREEDING IN PLANTS. Annual Review of Ecology and
Systematics 7: 469-495.
Kalisz S, Vogler D, Fails B, Finer M, Shepard E, Herman T, Gonzales R. 1999. The
mechanism of delayed selfing in Collinsia verna (Scrophulariaceae). American Journal of
Botany 86(9): 1239-1247.
Karron JD, Jackson RT, Thumser NN, Schlicht SL. 1997. Outcrossing rates of individual
Mimulus ringens genets are correlated with anther-stigma separation. Heredity 79: 365370.
Karron JD, Mitchell RJ, Holmquist KG, Bell JM, Funk B. 2004. The influence of floral
display size on selfing rates in Mimulus ringens. Heredity 92(3): 242-248.
Kearns CA, Inouye DW. 1993. Techniques for pollination biologists. Niwot, CO: University
Press of Colorado.

135

Khatun S, Flowers TJ. 1995. THE ESTIMATION OF POLLEN VIABILITY IN RICE. Journal
of Experimental Botany 46(282): 151-154.
Kirpes CC, Clark LG, Lersten NR. 1996. Systematic significance of pollen arrangement in
microsporangia of Poaceae and Cyperaceae: Review and observations on representative
taxa. American Journal of Botany 83(12): 1609-1622.
Kohn JR, Barrett SCH. 1994. POLLEN DISCOUNTING AND THE SPREAD OF A
SELFING VARIANT IN TRISTYLOUS EICHHORNIA-PANICULATA - EVIDENCE
FROM EXPERIMENTAL POPULATIONS. Evolution 48(5): 1576-1594.
Lande R, Schemske DW. 1985. THE EVOLUTION OF SELF-FERTILIZATION AND
INBREEDING DEPRESSION IN PLANTS .1. GENETIC MODELS. Evolution 39(1):
24-40.
Levin DA, Kerster HW. 1974. Gene flow in seed plants. Evol. Biol. 7: 139-220.
Linder HP 1998. Morphology and the evolution of wind pollination. In: Owens SJ, Rudall PJ
eds. Reproductive biology in systematics, conservation and economic botany: Kew:
Royal Botanic Gardens, 123-135.
Lloyd DG. 1979. SOME REPRODUCTIVE FACTORS AFFECTING THE SELECTION OF
SELF-FERTILIZATION IN PLANTS. American Naturalist 113(1): 67-79.
Lloyd DG. 1992. SELF-FERTILIZATION AND CROSS-FERTILIZATION IN PLANTS .2.
THE SELECTION OF SELF-FERTILIZATION. International Journal of Plant Sciences
153(3): 370-380.
Lloyd DG, Schoen DJ. 1992. SELF-FERTILIZATION AND CROSS-FERTILIZATION IN
PLANTS .1. FUNCTIONAL DIMENSIONS. International Journal of Plant Sciences
153(3): 358-369.
Lyons EE, Antonovics J. 1991. BREEDING SYSTEM EVOLUTION IN LEAVENWORTHIA
- BREEDING SYSTEM VARIATION AND REPRODUCTIVE SUCCESS IN
NATURAL-POPULATIONS OF LEAVENWORTHIA-CRASSA (CRUCIFERAE).
American Journal of Botany 78(2): 270-287.
Matsui T, Kagata H. 2003. Characteristics of floral organs related to reliable self-pollination in
rice (Oryza sativa L.). Annals of Botany 91(4): 473-477.
136

Morishima H, Barbier P. 1990. MATING SYSTEM AND GENETIC STRUCTURE OF
NATURAL POPULATIONS IN WILD RICE ORYZA-RUFIPOGON. Plant Species
Biology 5(1): 31-40.
Niklas KJ. 1985. THE AERODYNAMICS OF WIND POLLINATION. Botanical Review
51(3): 328-386.
Niklas KJ. 1987. POLLEN CAPTURE AND WIND-INDUCED MOVEMENT OF COMPACT
AND DIFFUSE GRASS PANICLES - IMPLICATIONS FOR POLLINATION
EFFICIENCY. American Journal of Botany 74(1): 74-89.
Oka HI, Morishima H. 1967. VARIATIONS IN BREEDING SYSTEMS OF A WILD RICE
ORYZA PERENNIS. Evolution 21(2): 249-&.
Ornduff R. 1969. Reproductive biology in relation to systematics. Taxon 18: 121-133.
Ritland C, Ritland K. 1989. VARIATION OF SEX ALLOCATION AMONG 8 TAXA OF
THE MIMULUS-GUTTATUS SPECIES COMPLEX (SCROPHULARIACEAE).
American Journal of Botany 76(12): 1731-1739.
Ritland K. 1984. THE EFFECTIVE PROPORTION OF SELF-FERTILIZATION WITH
CONSANGUINEOUS MATINGS IN INBRED POPULATIONS. Genetics 106(1): 139152.
Ritland K. 2002. Extensions of models for the estimation of mating systems using n independent
loci. Heredity 88: 221-228.
Ritland K, Jain S. 1981. A MODEL FOR THE ESTIMATION OF OUTCROSSING RATE
AND GENE-FREQUENCIES USING N INDEPENDENT LOCI. Heredity 47(AUG):
35-52.
Rodriguez-Riano T, Dafni A. 2000. A new procedure to asses pollen viability. Sexual Plant
Reproduction 12(4): 241-244.
Schemske DW, Lande R. 1985. THE EVOLUTION OF SELF-FERTILIZATION AND
INBREEDING DEPRESSION IN PLANTS .2. EMPIRICAL OBSERVATIONS.
Evolution 39(1): 41-52.

137

Schoen DJ, Brown AHD. 1991. WHOLE-FLOWER AND PART-FLOWER SELFPOLLINATION IN GLYCINE-CLANDESTINA AND G-ARGYREA AND THE
EVOLUTION OF AUTOGAMY. Evolution 45(7): 1651-1664.
Schoen DJ, Lloyd DG. 1992. SELF-FERTILIZATION AND CROSS-FERTILIZATION IN
PLANTS .3. METHODS FOR STUDYING MODES AND FUNCTIONAL-ASPECTS
OF SELF-FERTILIZATION. International Journal of Plant Sciences 153(3): 381-393.
Song ZP, Lu BR, Wang B, Chen JK. 2004. Fitness estimation through performance
comparison of F-1 hybrids with their parental species Oryza rufipogon and O-sativa.
Annals of Botany 93(3): 311-316.
Stapleton AE, Walbot V. 1994. FLAVONOIDS CAN PROTECT MAIZE DNA FROM THE
INDUCTION OF ULTRAVIOLET-RADIATION DAMAGE. Plant Physiology 105(3):
881-889.
Stebbins GL. 1974. Flowering plants: evolution above the species level. Cambridge, MA: The
Belknap Press of Harvard University Press.
Takebayashi N, Morrell PL. 2001. Is self-fertilization an evolutionary dead end? Revisiting an
old hypothesis with genetic theories and a macroevolutionary approach. American
Journal of Botany 88(7): 1143-1150.
Vallejo-Marin M, Barrett SCH. 2009. Modification of flower architecture during early stages
in the evolution of self-fertilization. Annals of Botany 103(6): 951-962.
Vaughan DA. 1994. The wild relatives of rice: a genetic resources bankbook. Philippines:
International Rice Research Institute.
Vogler DW, Kalisz S. 2001. Sex among the flowers: The distribution of plant mating systems.
Evolution 55(1): 202-204.
Wright SI, Ness RW, Foxe JP, Barrett SCH. 2008. Genomic consequences of outcrossing and
selfing in plants. International Journal of Plant Sciences 169(1): 105-118.
Xie ZW, Lu YQ, Ge S, Hong DY, Li FZ. 2001. Clonality in wild rice (Oryza rufipogon,
Poaceae) and its implications for conservation management. American Journal of Botany
88(6): 1058-1064.

138

Zheng XM, Ge S. 2010. Ecological divergence in the presence of gene flow in two closely
related Oryza species (Oryza rufipogon and O. nivara). Molecular Ecology 19(12): 24392454.
Zhou HF, Zheng XM, Wei RX, Second G, Vaughan DA, Ge S. 2008. Contrasting population
genetic structure and gene flow between Oryza rufipogon and Oryza nivara. Theoretical
and Applied Genetics 117(7): 1181-1189.
Zhu QH, Ge S. 2005. Phylogenetic relationships among A-genome species of the genus Oryza
revealed by intron sequences of four nuclear genes. New Phytologist 167(1): 249-265.

139