Dissertation Christiane Kiefer 2008 Version 23 copy right

Dissertation Christiane Kiefer 2008 Version 23 copy right
Evolution and Phylogeography of the
North American genus Boechera
(Brassicaceae)
and the Evolution of Apomixis
Dissertation
Christiane Kiefer
born in Stuttgart
2008
Dissertation
submitted to the
Combined Faculties for the Natural Sciences and for Mathematics
of the Ruperto-Carola University of Heidelberg, Germany
for the degree of
Doctor of Natural Sciences
presented by
Diplom-Biologin Christiane Kiefer
born in Stuttgart/Germany
Oral-examination:: .......................................
1
Evolution and Phylogeography of the
North American genus Boechera (Brassicaceae)
and the Evolution of Apomixis
Referees:
Prof. Dr. Marcus Koch
Prof. Dr. Thomas Rausch
2
My PhD thesis is dedicated to my parents
Erich Kiefer 09.12.1939 – 21.07.2007
“Nichts ist so schlecht, dass es nicht für etwas anderes wieder gut ist”
(Nihil adeo est malum, non sit aliquid boni)
He taught me to see the bright side – no matter what.
Karin Kiefer geb. Nagel
She saw the interest for biology in me when I was a little child.
She constantly supported my hunger for knowledge on nature.
My life is dedicated to Science
3
Der unermesslich reichen, stets sich erneuernden Natur gegenüber wird der Mensch,
soweit er auch in der wissenschaftlichen Erkenntnis fortgeschritten sein mag, immer
das sich wundernde Kind bleiben und muss sich stets auf neue Überraschungen
gefasst machen.
(No matter how far humans progress in their scientific understanding, they will always
remain marvelling children in the face of the infinite wealth and continual change within
nature, and will always have to be ready for new surprises.)
Max Planck (1858 - 1947)
4
Table of contents
0.1 Summary
0.2 Zusammenfassung
6
7
1. Introduction
1.1 Phylogeography
1.1.1 An introduction to phylogeography
1.1.2 Ice ages and vegetation history in North America
1.1.3 Present Day Ecoregions of North America
1.1.4 Phylogeography in North America
1.2 The Brassicaceae Family
1.3 The North American genus Boechera
1.4 Apomixis in Boechera
1.5 Aims of the Dissertation
1.6 Literature cited
8
8
8
8
19
22
22
23
24
25
26
2. Molecular marker based Studies of the North American genus Boechera
2.1 cpDNA Gene Pool Analysis (submitted to MPE)
2.2 ITS and Single Copy Gene Studies (to be submitted to AmJBot)
2.3 Eastern versus western North American Boechera
(submitted to TAXON)
33
33
55
84
3. Cytogenetic Analyses of apomictic Boechera (to be submitted to Heredity)
106
4. Overview of Scientific Contributions
120
5. Contents supplementary material and raw data on the DVD included
in the thesis
121
6. Acknowledgements
122
Supplementary data from the chapters, a distribution map showing all accessions
used in this thesis and an electronic copy of the thesis itself are given on the DVD in
the back of the book.
5
Dissertation Christiane Kiefer – Summary
0.1 Summary
Boechera is a North American genus of the Brassicaceae named after the Danish
botanist Tyge Böcher who cytogenetically studied members of the genus in great
detail. Until the end of the 20th century most species which belong today to
Boechera were still included in the genus Arabis. Molecular studies in the late
1990s revealed that Arabis was polyphyletic and subsequently a majority of North
American Arabis species was transferred into Boechera.
Boechera inhabits a wide range of habitats and reproduces sexually as well as
apomictic (asexual reproduction via seeds).
The aims of the study presented inhere were (a) to reconstruct the phylogeographic
history of the genus based on chloroplast DNA marker sequences and (b) to
investigate the nrDNA ITS gene pool and reconstruct phylogenies based on nrDNA
ITS and single copy gene introns.
In the course of the experiments it appeared that eastern North American Boechera
constitute different evolutionary lineages than the species centred in western North
America. Hence the split of eastern and western North American Boechera species
became a third subject (c).
In our continental wide phylogeographic studies we detected a large amount of
haplotype sharing indicating recurrent hybridisation on the one hand and nondifferentiation of haplotypes since speciation on the other hand. We concluded that
the chloroplast gene pool in Boechera pre-dates speciation in respect to the
investigated markers. Unrelated from taxon identity we could show that the
evolutionary lineages detected have a different phylogeographic history in terms of
glacial refugia and recently recolonized areas.
The study based on nrDNA ITS and introns of two single copy genes enabled us to
obtain deeper insights into Boechera phylogeny and ITS type distribution across
taxa. We could show that species specific lineages exist although the relationship
among them is poorly resolved. Comparing gene tree topologies this indicates rapid
speciation which probably happened in the second half of the quaternary. Hybrids
could be identified by the comparison of the different marker systems together with
chloroplast DNA types from an earlier study.
The comparison of eastern and western North American Boechera based on DNA
marker sequences showed that eastern North American Boechera represent two
evolutionary lineages within the genus Boechera (cpDNA) or among the tribe
Boechereae (nrDNA). We also included the Siberian taxa Borodinia and Boechera
falcata in the analysis and could show that they were resolved within Boechera.
The final task in the thesis was to unravel the origin of two apomict specific
chromosomes (Het and Del). BAC-FISH analyses revealed that Het is a homologue
of Boechera stricta linkage group 1 and Del appears only in 15 chromosome
apomicts and is a fragment of Het.
The thesis presented inhere offers deeper insights into the evolution of this complex
genus and offers a good basis for ongoing and future research projects dealing with
evolution and expression of apomixis as well as genome evolution in Boechera.
6
Dissertation Christiane Kiefer – Zusammenfassung
0.2 Zusammenfassung
Boechera ist eine nordamerikanische Gattung aus den Brassicaceae. Die Gattung
wurde nach dem dänischen Botaniker Tyge Böcher benannt welcher sie in den
1950er Jahren cytogenetisch untersuchte. Bis zum Ende des 20. Jahrhunderts
wurde Boechera zur Gattung Arabis gezählt, bis molekularbiologisch gezeigt
werden konnte, dass Arabis polyphyletisch ist. Im Folgenden wurden daher Taxa
welche bisher zu Arabis gerechnet wurden in die bereits 1971 beschriebene
Gattung Boechera überführt. Boechera lebt in einer großen Anzahl verschiedener
Habitate und kann sich auf sexuellem wie auch apomiktischen Wege (asexuelle
Fortpflanzung über Samen) fortpflanzen.
Die Ziele der vorliegenden Studie waren (a) die phylogeographische Geschichte der
Gattung basierend auf chloroplastidären Markersequenzen zu rekonstruieren und
(b) den nrDNA ITS Genpool zu untersuchen und phylogenetische Untersuchungen
basierend auf nrDNA ITS und Single Copy Gen Intronsequenzen zu rekonstruieren.
Im Verlauf der Experimente wurde deutlich, dass nordostamerikanische Boechera
Arten in anderen evolutionären Linien zu finden sind als die Arten, welche ihren
Verbreitungsschwerpunkt im westlichen Nordamerika haben. Daher ergab sich als
dritter Projektteil (c) die Untersuchung der phylogenetischen Beziehungen der
beiden Artengruppen.
In der kontinentweiten phylogeographischen Studie konnte vermehrt Haplotyp
Sharing festgestellt werden welches entweder als Indikator fortwährender
Hybridisierung oder nicht-Differenzierung de Haplotypen seit Artbildung (in Bezug
auf den verwendeten Marker) interpretiert werden kann. Es folgt, dass die
Differenzierung des Chloroplasten Genpools vor der Artbildung stattfand.
Unabhängig von Taxonidentität konnte gezeigt werden, dass die detektierten
evolutionären Linien eine unterschiedliche phylogeografische Vergangenheit in
Bezug auf Refugialgebiete und nacheiszeitlich wieder besiedelte Gebiete haben.
Die auf nrDNA ITS und Single Copy Gen Intronen basierende Studie gestattete es
einen tieferen Einblick in die Phylogenie und Verteilung von ITS Typen in einzelnen
Taxa zu erhalten. Es konnte gezeigt werden, dass artspezifische Linien existieren
obwohl ihre Beziehung zueinander nicht aufgelöst werden kann. Der Vergleich der
Topologie unterschiedlicher Genbäume implizierte, dass die Differenzierung
unterschiedlicher Linien über einen kurzen Zeitraum in der zweiten Hälfte des
Quartärs erfolgte. Hybriden konnten durch den Vergleich der unterschiedlichen
Markersysteme mit Chloroplasten-DNA Haplotypen aus dem ersten Projektteil
identifiziert werden.
Der Vergleich NW- und NO –Amerikanischer Boechera Arten basierend auf DNAMarker Sequenzen ergab, dass beide Artengruppen in unterschiedlichen
evolutionären Linien zu finden sind. Es konnte ebenfalls gezeigt werden, dass die
sibirischen Taxa Borodinia und Boechera falcata beide innerhalb der Gattung
Boechera stehen.
Schlussendlich konnte mittels BAC-FISH Analysen die Herkunft zweier für
Apomikten spezifischen Chromosomen geklärt werden (Het und Del). Het ist ein
Homolog der Boechera stricta Kopplungsgruppe 1 und Del ist ein Bruchstück davon
welches in Apomikten mit 15 Chromosomen vorkommt.
Insgesamt gibt die in diesem Rahmen vorgestellte Arbeit tiefere Einblicke in die
Evolution dieser komplexen Gattung und bietet eine Basis für momentane und
zukünftige Projekte zur Erforschung der Evolution und Expression von Apomixis
sowie Genomevolution in der Gattung Boechera.
7
Dissertation Christiane Kiefer – Introduction
1. Introduction
1.1 Phylogeography
1.1.1 An Introduction to Phylogeography
The quaternary period beginning 1.8 million years ago has been marked by
temperature oscillations caused by different inclinations of Earth towards the sun
(Milankovic theory, Seibold and Seibold, 2005). During periods in which global
temperatures decreased ice sheets extended from the poles into lower latitudes and
in montaneous regions from higher to lower altitude. The coverage of landmasses by
ice shields and the general drop in temperature had a major impact on fauna and
flora. Animals and plants were forced to retreat into glacial refugia until rising
temperatures allowed them to recolonize their former distribution ranges . The effect
of those quaternary migrations is visible until today. They influenced along with
habitat preferences the formation of present species distribution (Hewitt, 200, Hewitt,
2004).
Species distribution alone may be studied in biogeography. But how are geographical
distribution and phylogeny related? This is precisely what phylogeographic studies try
to answer. The term phylogeography itself was introduced in 1987 by John Avise
(Avise, 1987). Since then a multitude of studies revealed migration patterns of plants
and animals allover the globe (Avise, 1998). Phylogeographic inference makes use
of the fact that quaternary migrations not only left their footprints in species
distribution but also in the distribution of DNA sequence types. Typically plant studies
employ non-coding pieces of chloroplast DNA as so called marker systems.
Chloroplast DNA is a desirable molecule in angiosperms since in most cases it is
uniparentaly (maternally) inherited (Birky, 1995) and the “migration of the seed” may
be followed. The relatedness of DNA sequences calculated from mutations together
with the abundance of different DNA sequence types in a region enable the inference
of possible phylogeographic scenarios.
1.1.2 Ice Ages and Vegetation History in North America
During the quaternary North America was affected by two main glacials termed
Illinoian and Wisconsin. The Illinoian includes two periods of glaciation and lasted
from 300,000 to 130,000 years ago (Liesiecki et al. 2005). During the Illinoian the
Laurentide ice shield covered about 85% of the area of Illinois and reached into
Kansas during the maximum extent (Stiff and Hansel, 2004). The Sangamon or
Eemian interglacial is temporarily equivalent to the Ipswichian Stage in Great Britain
and the Riss-Würm interglacial in the European Alps. It began about 128,000130,000 years ago. Northern hemisphere winters were slightly warmer and wetter
than today. By 115,000 years ago the glacial era had returned (Lang, 1994). The last
glacial period was the Wisconsin period which started about 100,000 years ago and
8
Dissertation Christiane Kiefer – Introduction
ended between 10,000 and 15,000 years ago. Glaciation reached its maximum
extent by 18,000 years ago and affected mainly the northern but to some extent also
the southern hemisphere. Canada and the northern United States were almost
completely covered by the Laurentide ice-sheet while Alaska remained mostly icefree (Adams and Faure, 1997). Local glaciations existed in the Rocky Mountains and
the Cordilleran ice shield as well as in the Sierra Nevada in northern California (ice
fields and ice caps, James et al. 2002).
The so-called Pinedale or Fraser glaciation was the last glaciation which affected the
Rocky Mountains. It lasted from approximately 30,000 to 10,000 years ago (Brief
geologic history, Rocky Mountain National Park). It mainly consisted of mountain
glaciers merging into the Cordilleran ice shield (Ice Age Floods, From: U.S. National
Park Service Website). The Cordilleran ice shield itself was responsible for
phenomena like the glacial lake Missoula which cause the Missoula floodings after
breakage of the ice dam closing the lake (about 40 times between 15,000 and 13,000
years ago) (Waitt, 1985). The Wisconsin glacial period which refers to the last
extension of continental glaciers in the Laurentide ice shield had three glacial
maxima called Tahoe, Tenaya and Tioga. The Tahoe reached its maximum by
70,000 years ago and the Tioga began about 30,000 years ago and ended 10,000
years ago. At the hight of glaciation the Bering Strait was covered by ice and allowed
migration of humans and mammals. At the hight of the Wisconsin glaciation Canada,
the Upper Midwest and New England as well as parts of Montana and Washington
were covered in ice. In southwestern Saskatchewan and southeastern Alberta the
Laurentide and Cordilleran ice shield met and formed the cypress hills. The Great
Lakes resulted from pooled water from the melting glaciers (Damery, 2004).
With the quaternary climatic oscillations vegetation changed dramatically in North
America as can be seen from pollen records. About 40,000 years ago (C14 time
estimate) spruce and jack pine forests seemed to cover most of the eastern United
States and for Tennessee and North Carolina mixed, temperate forests are reported.
South of this in Texas southern pine forest with oak an hickory is reported. Eastern
Canada was still covered by an ice-shield extending as far as the Great Lakes
(Delcourt and Delcourt, 1981). 25,000 to 28,000 years ago, shortly before the last
glacial maximum most of the eastern United States possibly had an open woodland
vegetation. A mixed cool temperate forest belt seems to have existed across the
southern Appalachian Mountains.
In the western Cordillera lake levels were higher but had not reached their maximum
hight yet. In northern Arizona vegetation belts had already declined. In the north
western Cordillera forest cover was less than today with more steppe and coldtolerant species (Whitlock and Bartlein, 1997).
9
Dissertation Christiane Kiefer – Introduction
Fig. 1.1 Vegetation in North America during the last glacial maximum as given in
Adams and Faure, 1997.
Vegetation types for the period during the last glacial maximum (18,000-15,000 C14
years ago; 21,000-17,000 calender years ago) are given in figure 1.1 according to
Adams and Faure, 1997. Most of Canada (including the islands along the west coast)
and the Northern United States were covered by an ice shield. Alaska remained ice
free and was covered by polar desert and dry tundra in the central part (Ritchie &
Cwynar 1982). In eastern North America open woodlands with pine and spruce
replaced the present deciduous forests (Jackson et al. 1997). Open spruce woodland
10
Dissertation Christiane Kiefer – Introduction
Fig. 1.2 Present day potential vegetation in North America according to Adams and
Faure, 1997.
extended further west into the prairie zone. The north-west of the United States was
dominated by dry alpine tundra and polar desert (Whitlock and Bartlein, 1997) with
scattered areas of cold-tolerant conifer woodland in the central Rocky Mountains.
Further south in the Cordillera a mosaic of semi-desert scrub and sparse conifer
woodland was found (Thompson et al. 1993). The south-west of North America was
moister than today. Open conifer woodlands and scrub covered areas which are
semi-desert today (for comparison of vegetation types compare figure 1.1 and figure
1.2).
Since 3000 years ago vegetation developed towards what is found today in North
America.
11
Dissertation Christiane Kiefer – Introduction
1.1.3 Present Day Ecoregions of North America
Landscape may be subdivided into ecoregions which are characterized by a certain
type of flora, fauna and environmental factors. As a result of several mapping
projects the land masses of Earth have been subdivided into 867 ecoregions of
which 116 are located in North America. Analyses in the present study were based
on ecoregions as described in Ricketts et al. 1999. In the following details on the
larger ecoregions which play a major role in the study and an outline on present
vegetation is given. Representative pictures are given in figures 1.4 and 1.5.
Temperate Broadleaf and Mixed Forests
New England/Acadian Forest
A hilly to montaneous area in Eastern Canada with a mosaic of forest types and nonforest habitats. Northern hardwood and spruce forests predominate. Wide distribution
of Picea rubens and Pinus resinosa distinguish this ecoregion from the adjacent
Great Lakes ecoregion with predominantly deciduous forest.
Glaciers shaped the topography of this landscape with its characteristic mountains
and plateaus. Soils are mostly poor and swamps and lakes are frequent. Summers
are warm and moist, winters are cold and snowy.
Great Lakes Forests (East and West)
The climate in this ecoregion is affected by the Great Lakes resulting in longer
growing season and effects on average temperature as well as timing and amount of
precipitation. The region was entirely covered in glaciers and the ground is covered
by a thick layer of deposits of glacial drift.
The Western Great Lakes region is characterized by mixed forests with Populus
tremuloides, Betula papyrifera, Pinus banksiana, Picea glauca, Picea mariana and
Abies balsamea. Wetlands are widespread in this region.
The Eastern Great Lakes forests lie between the boreal forests and the broadleaf
deciduous forests and are therefore transitional. Parts of the forests are coniferous
while others form a mosaic of pure deciduous stands in habitats with good soils.
Temperate Coniferous Forests
Sierra Nevada
The Sierra Nevada Range is about 100 miles long and 50 miles wide and runs
southwest to northwest approximately along the border of the US American States
California and Nevada. The eastern escarpment is much steeper than the western. It
is home of one of the most diverse coniferous forests on Earth with an extraordinary
range of habitats. Different communities are distributed in elevation belts on both
sides of the range.
12
Dissertation Christiane Kiefer – Introduction
3400
Alpine
Sub-alpine
2700
Red Fir
2000
East Side Pine
Sierra Nevada
Mixed Conifer
High Desert
1300
Ponderosa Pine
650
Oak Grass Woodlands
Grasslands
Agricultural
Riparian
Fig. 1.3 Sierra Nevada plant communities elevations belts; heights are given in metres;
changed after LeBlanc and Harris, http://www.cnr.berkeley.edu/departments/espm/
extension/ECOLOGY.HTM
Klamath Siskiyou Forests
The Klamath Siskiyou forests are one of the biodiversity hotspots on Earth due to its
complex terrain, geology, climate and biogeographic history. It is located in southern
Oregon and northern California. The regions was not affected by quaternary
glaciations and therefore provided a refuge and long periods of favourable conditions
for many taxa. Climatic shifts over time made this ecoregion a transition zone of
several major biota (Great Basin, Oregon Coast Range, Cascade Range, Sierra
Nevada, California Central Valley). 3,500 plant species are known from this region
including many endemics which only occur on single mountains. Serpentine soils are
present in some spots.
Wasatch and Uinta Montane Forests
The Wasatch/Uinta ecoregion is a block of high montane habitats bordering the
Great Basin in the East, reaching from southern Idaho to the isolated ranges of the
Colorado Plateau. The dominant vegetation is coniferous forest with Pinus
ponderosa, Pseudotsuga menziesii, Abies lasiocarpa and Picea engelmann. Pinus
flexilis occurs rarely. The Wasatch/Uinta range is arid due to the rain shadow of the
Sierra Nevada located 500 miles further west.
13
Dissertation Christiane Kiefer – Introduction
14
Dissertation Christiane Kiefer – Introduction
Colorado Rockies Forest
The Colorado Rockies are the region in which the highest elevations of the Rocky
Mountains are found. Vegetation is present in vertical zonation being a consequence
of abrupt elevational gradients. The dominant vegetation type is coniferous forest
along with mountain meadows, foothill grasslands, riparian woodlands and uppertreeline alpine tundra communities.
North Central Rockies Forest
The Northern Central Rockies stretch over 600 miles from north to south from
western Canada into the United States. Climatic conditions vary from east to west,
the west experiencing the influence of the Pacific while the east is more continental.
Mean annual temperature is 2 °C higher in the western part (3.5 °C versus 5.5 °C).
Glaciation left large valleys throughout the region. The dominant vegetation type is
coniferous forest. Typically the tree species found in the Cascades are also found in
the Northern Central Rockies. Mountain meadows, alpine habitats, foothill grasslands
and riparian woodlands are also present. The vertical zonation is pronounced as a
consequence of abrupt changes in elevation.
South Central Rockies Forest
The South Central Rockies cover the area western Wyoming, eastern Idaho and
central Montana. A second portion is located in central and eastern Idaho south of
the Clearwater River. Flatlands and mountains change abruptly resulting in a
dramatic vertical zonation. The Black Hills unit of this ecoregion is the lowest in
elevation but contains a distinct, floristic diversity with floral elements of the Great
Basin, Eastern deciduous, Boreal, Rocky Mountain, and Southern Great Plains. The
dominant vegetation type is coniferous forest. Relative to other regions within the
Rocky Mountains the South Central Rockies are dry with a predominantly continental
climate. Summers are short, winters are long and cold. Parts of the region are
influenced by geothermal activities (Yellowstone National Park).
Cascade Mountains Leeward Forest
This ecoregion is located along the leeward side of the Cascade Mountains. A strong
climatic gradient exists from the moist coastal climate to the semi-arid continental
climate in the southern interior. Alpine tundra, montane forests and parklands with
scattered ponderosa pine are common.
Eastern Cascades Forest
The eastern Cascades span the eastern slopes of the Cascade Mountains in
Oregon and Washington meeting the Cascade Mountains Leeward in the North. The
Eastern Cascades are a series of steep, rugged mountains rising up to 2,700 m and
volcanic peaks reaching a hight of 4,300 m. In some areas serpentine soils are
15
Dissertation Christiane Kiefer – Introduction
found. The climate is in general mild. The natural vegetation is a mosaic of shrub
lands, grasslands and coniferous forests. Ponderosa pine is very common but
species composition of the forests varies along with environmental gradients of
physical factors such as temperature and moisture.
Blue Mountains Forest
The Blue Mountains ecoregion is characterized by several basin and range areas,
alluvial fans and floodplains. The relief is highly variable. Plant communities in the
different ecosystems in the Blue Mountains are sagebrush-, Pinyon-juniper-,
ponderosa pine-, Douglas fir-, western larch-, spruce-fir- and lodgepole pine
communities as well as chaparral-mountain shrub, mountain meadows, mountain
grasslands and alpine communities.
Temperate Grasslands/Savanna/Shrub
Canadian Aspen Forests and Parklands
The ecoregion is classified by a sub humid, low-boreal ecoclimate. Summers are
short and warm, winters are cold and long. The region is mostly underlain by
Cretaceous shale. Lakes are frequent in shallow depressions. Vegetation is
characterized by Populus tremuloides followed in number by Populus balsamifera
together with an understory of mixed herbs and shrubs. Picea glauca and Abies
balsamea are the climax species. However, they are not frequent due to fire. Poorly
drained sites are covered with sedges, willows, spruce and Larix laricina.
Xeric Shrublands/Deserts
Great Basin Shrub Steppe
The Great Basin is the most northerly of the four North American deserts. The Great
Basin has affinities with the cold vegetation unlike the other deserts which are more
associated with warm-temperate or tropical-subtropical vegetation. The topography
of the Great Basin is diverse. A series of parallel ranges and their intervening valleys
characterizes the landscape. The Great Basin is bordered by the central Rockies and
the Colorado Plateau in the east, the Columbia Plateau in the north and the
Cascades and Sierra Nevada in the west. Towards the south it meets the Mojave
desert at the Colorado River drainage. Vegetation is dominated by cold temperate
species like sagebrush, saltbrush and winterfat. Evolutionary tied to warmer places
are genera like Chrysothamnus, Coleogyne, Tetradymia and Grayia which also occur
in the Great Basin.
16
Dissertation Christiane Kiefer – Introduction
Fig. 1.5 Pictures representing dry ecoregions. Pictures all obtained from
http://www.nationalgeographic.com/wildworld/, photographer given in brackets after
the ecoregion name. a Great Basin (John Morrison), b Snake/Columbia Plateau (John
Morrison), c Colorado Plateau (John Morrison), d Mojave Desert (John Morrison),
e Chihuahua Desert (wwf-Canon, Edward Parker), f Sonora Desert (David Olson),
g Wyoming Basin (USFWS), h California Montane Chaparral and Woodlands (John
Morrison), i Yukon Interior Dry Forest (J. Peepre)
Snake/Columbia Shrub Steppe
The Snake/Columbia shrub steppe is a mostly arid ecoregion north of the Great
Basin. Towards the east it’s boundary is the Continental Divide. It is situated in a
major river system. However, due to the rain shadow of the Cascade Mountains the
ecoregion receives only little water. The dominant vegetation is Artemisia species,
often associated with Agropyron species, Festuca idahoensis and other
bunchgrasses. In the montaneous parts of this ecoregion juniper and bunchgrass
communities are found. In some parts of montaneous areas Pseudotsuga menziesii,
Abies lasiocarpa and Populus tremuloides occur.
Colorado Plateau Shrublands
The Colorado Plateau is an elevated, northward-tildes plateau with arid to semiarid
climate. It has developed a great relief through erosion by rivers such as the Grand
Canyon. The major vegetation types are woodlands dominated by Pinus edulis and
several juniper species. Between the trees the ground is sparsely covered. The
17
Dissertation Christiane Kiefer – Introduction
montaneous areas are covered by ponderosa pine forests in the south and lodgepole
pine and aspen forests further north. The lowest montaneous zone supports
grasslands with many bare areas.
Mojave Desert
The Mojave desert is the smallest of the four North American deserts. Dominant
plants of the Mojave are creosote bush (Larrea tridentata), brittlebush (Atriplex
polycarpa), desert holly (Atriplex hymenelytra), white burrobush (Hymenoclea
salsola), and Joshua tree (Yucca brevifolia). The Mojave supports numerous species
of cacti. Elevations range from below sea level to 1,600 m. Most regions are between
600 and 1,200 m, this being the reason why the Mojave is referred to as a high
desert.
Chihuahua Desert
The Chihuahua desert reaches from the southern United States deep into Mexico.
The Chihuahua desert is in general cooler and wetter than the other North American
deserts due to its higher elevation. Larrea divarecata, Florensia cernua, Prosopis
articulata and Acacia species characterize the landscape. The North American
Chihuahua desert has an extensive grass component as opposed to the Mexican
part which supports mainly a vegetation composed of cacti, yucca and shrubs. The
Chihuahua desert is of recent origin.
Sonora Desert
The Sonora Desert reaches from southeastern California into the western two thirds
of southern California and into Mexico in the states of Sonora and parts of Baja
California. The Sonora desert is divided into two sections characterized by the
availability of water. Temperatures are relatively high in winter and summer and
rainfall patterns are irregular. Lower-elevation areas are dominated by dense
communities of creosote and white bursage. On slopes Cercidium species and
ironwood (Olneya tesota) are found.
Wyoming Basin Shrub Steppe
The Wyoming Basin ecoregion is a high, open, arid country. It is nearly completely
surrounded by montaneous ecoregions. The aridity is due to the rain shadow of the
Rocky Mountains. The dominant vegetation type is sagebrush, often associated with
wheatgrasses or fescue.
18
Dissertation Christiane Kiefer – Introduction
Boreal Forest/Taiga
Yukon Interior Dry Forests
The Yukon Interior Dry Forest lies mainly within the Yukon Territory. The climate is
cold and semi-arid. The mean annual temperature is -3 °C. Elevations in this
ecoregion are generally above 1,000 m. White and Black spruce form the dominant
vegetation. South-facing slopes are often characterized by grassland communities. In
the colder alpine regions Dryas hookeriana, dwarf shrubs, forbs, grasses and lichens
cover the ground.
Fig. 1.6 Geographical Location of ecoregions in North America according to the wwf
ecoregion map (Olson et al. 2001); 1 New England/Acadian Forest, 2 Great Lakes
Forest, 3 Sierra Nevada, 4 Klamath/Siskiyou Forests, 5 Wasatch Uinta, 6 Colorado
Rockies, 7 North Central Rockies, 8 South Central Rockies, 9 Cascade Mountains
Leeward, 10 Eastern Cascades, 11 Blue Mountains, 12 Canadian Aspen Forest and
Parkland, 13 Great Basin, 14 Colorado Plateau, 15 Mojave Desert, 16 Chihuahua
Desert, 17 Sonora Desert, 18 Wyoming Basin, 19 Yukon Interior Dry Forest
1.1.4 Phylogeography in North America
Besides Europe North America is the second best studies continent in
phylogeography. Western North America is highly diverse in its landscape and
ecosystems and therefore species show complex distribution patterns. However,
three major patterns exist which are shared by several species.
Mesic Forest Disjunction pattern - Species of mesic forests are found in the Pacific
coastal and interior Rocky Mountains which are 300 km apart separated by the arid
Columbia Basin. Possible explanations for this pattern are ancient vicariance with a
19
Dissertation Christiane Kiefer – Introduction
formerly continuous distribution of the mesic species or inland dispersal, assuming a
recent colonization of the interior Rockies by mesic species (Brunsfel et al. 2001).
Cascade/Sierra pattern - It has been shown for numerous species that they reach
their southern or northern distribution limit at the Sierra/Cascades transition. The
reason for this may be an environmental gradient which existed during the Cenozoic,
or the presence of multiple refugia (Brunsfeld et al. 2001; Calsbeek et al. 2003).
Northern Rocky Mountain pattern - A greater part of the diversity of mesic forest
species occurs south of the the limits of the last Cordilleran ice-shield. The highest
diversity is found in the Clearwater Range. However, some plant species from
coastal ranges only occur north of the last glaciation. In general the biogeographic
discontinuity in the Rocky Mountains spans two elevation belts – middle elevation
(mesic-temperate species) and subalpine habitats (Cascade/Sierra species). The
genetic pattern present might either be explained by multiple refugia or by recent
colonization (Brunsfeld et al. 2001).
Another common pattern is the split into eastern and Western North American clades
which came into secondary contact in western North America as it is found for
example in several bird species (Hull and Girman, 2005; Spellman et al. 2007).
A comparison of several phylogeographic studies centred in western North America
is given in table 1.1.
Table 1.1 Comparison of phylogeographic studies centred in North America
species
area under
investigation
Northern Rocky
Mountains
refuge area
remark
marker system
reference
Clearwater Range
might be one of
several refugia
mtDNA sequence
data
Good and Sullivan,
2001
Cardamine
constancei
(Brassicaceae)
South and North
central Rocky
Mountains
“Greater
Clearwater
Refugium”
range fluctuations
with Pleistocenic
ice-ages;
recolonization
through Bitterroot
Range and
Continental Divide
multicompartmented
refugium
cpDNA sequence
data
Brunsfeld and
Sullivan, 2005
Salix melanopsis
(Salicaceae)
South and North
central Rocky
Mountains;
Pacific North
West
North American
continent wide
“Greater
Clearwater
Refugium”
no endemic
haplotypes in
Bitterroot Range;
cpDNA sequence
data
Brunsfeld et al.,
2007
refugia east and
west of Great
Basin in Pinus
ponderosa
stands; multible
refugia possible
long-term isolated
populations;
divergence of
northern and
southern clades
forced by ice-ages
mtDNA sequence
data
Spellman and
Klicka, 2007
mountain
ranges along
US.-México
Border
-
northward
migration from
México and
subsequent
isolation on skyislands
mtDNA and 9
microsatellites
Holycross and
Douglas, 2007
Tamias ruficaudus
(Rodentia,
Sciuridae)
Sitta carolinensis
(Aves)
Crotalus willardi
obscurus
(Reptilia)
20
Dissertation Christiane Kiefer – Introduction
species
refuge area
remark
marker system
reference
Sidalcea
(Malvaceae)
area under
investigation
western North
America
-
nrDNA
Andreasen and
Bladwin, 2003
Sitta pygmaea
(Aves)
western North
America
southern
California (like
Pinus ponderosa)
mtDNA sequence
data
Spellman and
Klicka, 2006
Martes americana
(Mammalia)
western North
America
North Pacific
Coast
mtDNA sequence
data
Stone et al. 2002
Dendroctonus
rufipennis
(Curculionidae)
North American
continent wide
mtDNA sequence
data
Maroja et al. 2007
Pinus flexilis
northwestern
North America
secondary contact
zone in Colorado
mtDNA sequence
data
Mitton et al. 2000
Aglenopsis aperta
(Araneae)
western North
America
postglacial
expansion out of
eastern refuge into
Rocky Mountains
mtDNA sequence
data
Ayoub and
Riechert, 2004
Poecile gambeli
(Aves)
western North
America
Beringia, Rockies
and central
eastern United
States south of
LGM
east of the
Rockies in
Colorado and
Kansas,
numerous sites in
Great Basin and
Fremont Co.,
Colorado and
Bighorn Co.,
Wyoming
east of the Rocky
Mountains,
between Rocky
Mountains and
Sierra Nevada,
west of the Sierra
Nevada
eastern (Rocky
Mountain) and
western (Sierra
Nevada) refuge
supports a
migration from
México along
Sierra Nevada one
one hand and
Rocky Mountains
on the other hand
for some species
refugia were
expected to be in
south Sierra
Nevada and
Arizona/New
Mexico
two colonization
events for pacific
NW
secondary contact
in Pacific
northwest
mtDNA sequence
data
Spellman et al.
2007
Accipiter striatus
velox (Aves)
North American
continent wide
-
mtDNA sequence
data
Hull and Girman,
2005
Tamias amoenus
(Rodentia,
Sciuridae)
NW North
America
-
mtDNA sequence
data
Demboski and
Sullivan, 2003
Lycaeides
(Lepidoptera)
western North
America
three refugia:
central North
America, east of
Great Lakes,
Sierra Nevada
eastern and
western clade;
postglacial
introgression in
Mono Lake region
distinct eastern
and western
group; recent
expansion of
eastern group
phylogeography
shaped by
vicariance;
east/west
dichotomy
glaciations
fostered
divergence,
postglacial
expansion lead to
gene flow and
reticulation
mtDNA, SSCP
analysis
Nice et al., 2005
Rana luteiventris
(Amphibia,
Ranidae)
Great Basin and
adjacent
mountain
ranges
-
one Bonneville
Basin and one
Lahontan Basin
clade (Nevada);
one Rocky
Mountain clade;
no genetic
structure for more
recent events
mtDNA sequence
data
Bos and Sites,
2001
Batrachoseps
(Amphibia)
California
-
large number of
contact zones
mtDNA sequence
data
Jockusch and
Wake, 2002
21
Dissertation Christiane Kiefer – Introduction
species
Veronica alpina
(Scrophulariaceae)
area under
investigation
western North
America
refuge area
remark
marker system
reference
maybe northern
Rocky Mountains;
three separate
phylogeographic
groups: Cascades,
northern and
southern Rocky
Mountains; British
Columbia as
secondary contact
zone
cpDNA sequence
data; AFLP
Albach et al. 2006
Eastern North America has been subject to a larger number of phylogeographic
studies than western North America. Phylogeographic patterns are mainly shaped by
the Appalachian Mountains. Due to its north-south orientation the Appalachian
Mountains did not represent an obstacle to migration into southern areas like for
example the Alps in Europe. Together with the high complexity of the landscape this
fact results into a multitude of phylogeographic patterns rather than the defined
patterns that have been inferred for Europe (Soltis et al. 2006). Phylogeographic
patterns in eastern North America have been reviewed and some recurrent patterns
have also been found in eastern North America (Soltis et al. 2006). Among these
were the Apalachicola River Basin discontinuity in Florida exhibited by a number of
fish and turtle species and a phylogeographic split along the Tombigbee River in
Alabama. A third pattern is referred to as the Appalachian Mountain discontinuity
which refers to an east/west split over the mountain chain probably caused by two
different glacial refugia on the opposing sites.
1.2 The Brassicaceae Family
The Brassicaceae (mustard family) comprises approximately 3500 species
distributed among approximately 350 genera of which many are monotypic. Among
them are well known as important crop plants such as Brassica oleracea or
ornamental plants like Aubrieta deltoides. Furthermore the Brassicaceae include the
most studied organism studied in plant molecular biology: the Thale Cress
Arabidopsis thaliana. Brassicaceae are found worldwide except in the tropics and
Antarctica. Their centre of diversity is the Irano-Turranian floristic region and some
authors suppose an origin of the whole Brassicaceae family for this region (Hedge et
al. 1976, Franzke et al. accepted).
Past taxonomic systems of the Brassicaceae followed morphological characters such
as fruit or hair shape (Jahnchen, Schulz, Hayek). Traditionally the Brasicaceae have
been subdivided into a tribal system. With the availability of molecular markers it was
possible to reconstruct a family wide ITS (Internal Transcribed Spacer) based
phylogeny. This lead to the restructuring of old tribes and also the description of new
tribes such as the Boechereae (Bailey et al. 2006). However, although the resolution
within the tribes was well, the relationships among the tribes was only poorly
resolved. Therefore the “Brassicaceae working group” (an international group of
collaborating researchers) has defined a TOP 100 list of Brassicaceae representing
22
Dissertation Christiane Kiefer – Introduction
the different tribes for which several nuclear markers shall be amplified and
sequenced in order to obtain a better resolution in phylogenetic tree reconstruction
(Marcus Koch, personal communication).
1.3 The North American Genus Boechera (Brassicaceae)
The genus Boechera comprises 110 species of which 108 are confined to the North
American continent (Windham and Al-Shehbaz 2006, 2007a, 2007b). The other two
species occur in Greenland (Boechera holboellii) and Siberia (Boechera falcata),
respectively. Boechera belongs to the tribe Boechereae which also includes the
genera Nevada, Pennellia, Sandbergia and Cusikiella (Bailey et al. 2006).
The plants are herbaceous and biennial to perennial with simple or branched
trichomes which may be stalked or sessile. Stems can be simple or branched. Basal
leaves are petiolate, simple, entire or dentate (rarely lyrat-pinnatified) while cauline
leaves are sessile or rarely very short petiolate. Inflorescences are sometimes in
panicles, fruiting pedicels can be erect, ascending, divaricate, or reflexed. Petals are
white, pink or purple. Flower morphology follows the typical Brassicaceae flower (four
sepals, four petals, two short +four long stamen and a fused pistil made up of two
carpels). The ovary is superior. Fruits are siliques (Al-Shehbaz, New Flora of North
America, unpublished). Examples of Boechera morphology are given in figure 1.7.
Fig. 1.7 Examples of Boechera morphology illustrated by a Boechera stricta; USDA
NRCS. Wetland flora: Field office illustrated guide to plant species. USDA Natural
Resources Conservation Service. Provided by NRCS National Wetland Team, Fort
Worth, TX; b Boechera holboellii var. collinsii; Britton, N.L., and A. Brown. 1913. An
illustrated flora of the northern United States, Canada and the British Possessions.
Vol. 2: 183. Courtesy of Kentucky Native Plant Society; c Boechera laevigata; Britton,
N.L., and A. Brown. 1913. An illustrated flora of the northern United States, Canada
and the British Possessions. Vol. 2: 181. Courtesy of Kentucky Native Plant Society;
23
Dissertation Christiane Kiefer – Introduction
In the past Boechera used to be included into the genus Arabis and species number
varied between 50 and 80 according to the author although Boechera as a genus
was already established in 1976 (Löve and Löve, 1976). Later molecular data proved
that a greater part of North American Arabis were not related to each other (Koch et
al. 1999, 2000). Since then species were repeatedly transferred into the genus
Boechera, named after the Danish botanist Tyge Böcher who cytogenetically studied
the plants in great detail (Böcher, 1951). The most recent description of the whole
genus was done for the new edition of the Flora of North America (Al-Shehbaz, New
Flora of North America, unpublished), giving a more extensive and complete
description of the genus than the series of papers published by Windham and AlShehbaz (2006, 2007a, 2007b).
In Boechera the base chromosome number is x=7. However, plants do not only occur
on the diploid level. Aneuploidy is a common phenomenon not only represented by
the many triploids but also in chromosome numbers deviating from a multiple of
seven. Some diploid plants and the aneuploids reproduce apomictically (asexual
reproduction via seeds). In the new description of Boechera sexually and asexually
reproducing plants were separated into different species. Hence 72 species are
sexually reproducing while 38 species which are assumed to be of hybrid origin are
asexually reproducing.
Boechera lives in a wide range of different habitats reaching from dry desert and
shrub steppe to forest and alpine habitats.
1.4 Apomixis in Boechera
Boechera may reproduce sexually or asexually via apomixis. Cytogenetic studies
showed that apomicts contain chromosomes from Boechera stricta and Boechera
holboellii relatives (sensu Rollins, 1993) and my therefore be assumed to be of hybrid
origin (Kantama et al. 2007). Morphological evidence also supports this view
(Windham and Al Shehbaz, 2006. 2007a, 2007b).
Often apomixis is associated with high ploidy levels. However, in Boechera apomicts
can be diploid, triploid or aneuploid (Dobeš et al. 2006). Apomixis in Boechera is
facultative and sexual and asexual reproduction may occur in the same individual
(Schranz et al. 2005). Hence genomes can be recombined leading to different
numbers of parental chromosomes in the hybrid offspring (e.g. four B. stricta
chromosomes and ten B. holboellii related chromosomes; Kantama et al. 2007).
Microsporogenesis in Boechera apomicts is disturbed and leads to a number of
malformed pollen grains along with a number of spherical pollen grains opposed to
the elongated ones produced by sexuals (Voigt et al. 2007). Pollen size was also
taken as a criterion to distinguish sexual and asexual Boechera species for the new
species circumscriptions (Windham and Al Shehbaz, 2006).
It was hypothesized that Boechera apomicts are characterized by a B chromosome
(Sharbel et al. 2004, Sharbel et al. 2005). B chromosomes are extra chromosomes
24
Dissertation Christiane Kiefer – Introduction
which are inherited independently, are dispensable and do not pair with other
chromosomes during meiosis (Jones and Houben, 2003). However, cytogenetic
studies could show that 2n and 2n + 1 apomicts are rather characterized by the
presence of a heterochromatic chromosome and in the cases of the 2n + 1 apomicts
a deletion chromosome. According to their most prominent features the
chromosomes were named Het and Del (Kantama et al. 2007). Meiotic studies
suggest that the Het chromosome pairs regularly with another chromosome while the
Del chromosome forms a heteromorphic trivalent. Since Boechera chromosomes are
very small (3-6 µm) and uniform in shape and size (except the Het and Del) it is
unknown to which other chromosome/chromosome part Het and Del are
homologous. Since the Het chromosome is present only in apomicts it is tempting to
hypothesize that it plays a role in expressing apomixis. However, nothing which
would support this fact has been proven so far.
1.5 Aims of the Dissertation
The aim of the research conducted in the framework of this thesis was to unravel the
phylogeographic and evolutionary history for as many members of the genus
Boechera as possible by using cpDNA data following a study which had been based
on Boechera divaricarpa (sensu Rollins, 1993), Boechera stricta and Boechera
holboellii (sensu Rollins, 1993) (Dobeš et al. 2004). Centres of genetic diversity
should be revealed for the whole genus and should be compared with species
diversity data.
A second aim was to nrDNA ITS gene pool and to reconstruct a phylogeny for the
genus by using nrDNA and nuclear single copy gene markers (introns from genes
neighbouring the loci encoding the vrn1 and ELF3 proteins; primers according to Eric
Schranz, Thomas Mitchell-Olds and Bao-Hua Song).
To get an insight into the genomic structure of apomicts BAC-FISH analyses were
carried out in order to identify translocations and the origin of the Het and Del
chromosome (project part funded by the DAAD).
25
Dissertation Christiane Kiefer – Introduction
1.6 Literature cited
Albach, D. C., Schönswetter, P., Tribsch, A. 2006. Comparative phylogeography of
the Veronica alpina complex in Europe and North America. Molecular Ecology 15:
3269–3286.
Al-Shehbaz, I. A. submitted. Boechera A. Löwe and D. Löwe. In: Flora North
America.
Andreasen K., Baldwin, B. G. 2003. Reexamination of relationships, habital evolution,
and phylogeography of checker mallows (Sidalcea; Malvaceae) based on molecular
phylogenetic data. American Journal of Botany 90: 436–444.
Avise, J. C., Arnold, J., Ball, R. M., Bermingham, E., Lamb, T., Neigel, J. E., Reeb, C.
A., Saunders, N. C. 1987. Intraspecific phylogeography: the mitochondrial DNA
bridge between population genetics and systematics. Annual Reviews in Ecology and
Systematics 18: 489–522.
Avise, J. C. 1998. The history and purview of phylogeography: a personal reflection.
Molecular Ecology 7: 371–379.
Ayoub, N. A., Riechert, S. 2004. Molecular evidence for Pleistocene glacial cycles
driving diversification of a North American desert spider, Agelenopsis aperta.
Molecular Ecology 13: 3453–3465.
Bailey, C. D., Koch, M. A., Mayer, M., Mummenhoff, K., O’Kane, S. L., Warwick, S.
I., Windham, M. D., Al-Shehbaz, I. A. 2006. A Global nrDNA ITS Phylogeny of the
Brassicaceae. Molecular Biology and Evolution 23: 2142–2160.
Birky, C. W. Jr. 1995. Uniparental Inheritance of mitochondrial and chloroplast genes:
Mechanisms and Evolution Proceedings of the National Academy of Sciences 92:
11331–11338.
Böcher, T. W. 1951. Cytological and embryological studies in the amphi-apomictic
Arabis holboellii complex. Kong Danske Vidensk Selsk, 6: 1–59.
Bos, D. H., Sites, J. W. Jr. 2001. Phylogeography and conservation genetics of the
Columbia spotted frog (Rana luteiventris; Amphibia, Ranidae). Molecular Ecology 10:
1499-1513.
26
Dissertation Christiane Kiefer – Introduction
Brunsfeld, S. J., Sullivan, J., Soltis, D. E., Soltis, P. S. 2001. Comparative
phylogeography of north-western North America: a synthesis. in: Integrating ecology
and evolution in a spatial context. Silvertown J.; Antonovics J. Blackwell Science
LTD. ISBN 0-632-05824-2.
Brunsfeld, S. J., Miller, T. R., Carstens, B. C., 2007. Insights into the Biogeography of
the Pacific Northwest North America: Evidence from the Phylogeography of Salix
melanopsis. Systematic Botany 32: 129–139.
Brunsfeld, S. J., Sullivan, J. 2005. A multi-compartmented glacial refugium in the
northern Rocky Mountains: Evidence from the phylogeography of Cardamine
constancei (Brassicaceae). Conservation Genetics 6: 895–904.
Calsbeek, R., Thompson, J. N., Richardson, J.E. 2003. Patterns of molecular
evolution and diversification in a biodiversity hotspot: the California Floristic Province.
Molecular Ecology 12: 1021–1029.
Delcourt P. A., Delcourt H. R. 1981. Vegetation maps for eastern North America:
40,000 yr BP to the present. Geobotany II. Ed; R.C. Romans. Plenum.
Lisiecki, L.E., 2005. "Ages of MIS boundaries." LR04 Benthic Stack Boston
University, Boston, MA
Dembiski, J. R., Sullivan, J. 2003. Extensive mtDNA variation within the yellow-pine
chipmunk, Tamias amoenus (Rodentia: Sciuridae), and phylogeographic inferences
for western North America. Molecular Phylogenetics and Evolution 26: 389–408.
Dobeš, C. H., Mitchell-Olds, T., Koch, M. A., 2004. Extensive chloroplast haplotype
variation indicates Pleistocene hybridization and radiation of North American Arabis
drummondii, A. X divaricarpa, and A. holboellii (Brassicaceae). Molecular Ecology
13: 349–370.
Dobeš, C., Koch, M. A., Sharbel, T. F. 2006. Embryology, karyology, and modes of
reproduction in North American genus Boechera (Brassicaceae): a compilation of
seven decades of research. Annals Miss. Bot. Gard. 93: 517–533.
Franzke, A., German, D., Al-Shehbaz, I. A.,Mummenhoff, K. accepted. Arabidopsis
family ties: Molecular phylogeny and age estimates in the Brassicaceae. TAXON
Good, J. M., Sullivan, J., 2001. Phylogeography of the red-tailed chipmunk (Tamias
ruficaudus), a northern Rocky Mountain endemic. Molecular Ecology 10: 2683–2695.
27
Dissertation Christiane Kiefer – Introduction
Hedge, I. C. 1976. A systematic and geographical survey of the Old World
Cruciferae. In The biology and chemistry of the Cruciferae (eds. Vaughan JG, Mac
Leod AJ, Jones BMG), pp. 1–45. Academic Press, London.
Hewitt, G. 2000. The genetic legacy of the Quaternary ice ages. Nature 405: 907–
913.
Hewitt, G. M. 2004. Genetic consequences of climatic oscillations in the Quaternary.
Philosophical Transaction of the Royal Society of London B. Biological Sciences
359:183–95.
Holycross, A. T., and Douglas, M. E. 2007. Geographic isolation, genetic divergence,
and ecological non-exchangeability define ESUs in a threatened sky-island
rattlesnake. Biological Conservation 134: 142–154.
Hull, J. M., Girman, D. J. 2005. Effects of Holocene climate change on the historical
demography of migrating sharp-shinned hawks (Accipiter striatus velox) in North
America. Molecular ecology 14: 159–170.
Jackson S. T., Overpeck J. T., Webb T. III, Keattch S. E., Anderson K. H. 1997.
Mapped plant-macrofossil and pollen records of late Quaternary vegetation change in
eastern North America. Quaternary Science Reviews v.16 p.1–70.
James, L. A., Harbor, J., Fabel, D., Dahms, D., Elmore, D. 2002. Late Pleistocene
Glaciations in the Northwestern Sierra Nevada, California. Quaternary Reserach 57:
409–419.
Jokusch, E., Wake, D. B. 2002. Falling apart and merging: diversification of slender
salamanders (Plethodontidae: Batrachoseps) in the American West. Biological
Journal of the Linnean Society 76: 361–391.
Jones, N., Houben, A. 2003. B chromosomes in plants: escapees from the A
chromosome genome? TRENDS in Plant Science 8: 417–423.
Kantama, L., Sharbel, T. F., Schranz, M. E., Mitchell-Olds, T., de Vries, S., de Jong,
H. 2007. Diploid apomicts of the Boechera holboellii complex show large scale
chromosome substitutions and different aberrant chromosomes. Proceedings of the
National Academy of Sciences 104: 14026–14031.
28
Dissertation Christiane Kiefer – Introduction
Koch, M., Bishop, J., Mitchell-Olds, T. 1999. Molecular systematics and evolution of
Arabidopsis and Arabis. Plant Biology: 1: 529–537.
Koch, M., Haubold, B., Mitchell-Olds, T. 2000. Comparative evolutionary analysis of
chalcone synthase and alcohol dehydrogenase loci in Arabidopsis, Arabis and
related genera. Molecular Biology and Evolution. 17: 1483–1498.
Lang, G. 1994 in Quartäre Vegetationsgeschichte Europas, Gustav Fischer Verlag,
Jena. ISBN 3-334-60405-5.
Lisiecki, L. E., Raymo, M. E. 2005. "A Pliocene-Pleistocene stack of 57 globally
distributed benthic d18O records. Paleoceanography." vol. 20, PA1003,
doi:10.1029/2004PA001071
Löve, A., Löve, D. 1976. Botaniska Notiser 128: 513.
Maroja, L. S., Bogdanowicz, S. M., Wallin, K. F., Raffa, K. F., Harrison, R. G. 2007.
Phylogeography of spruce beetles (Dendroctonus rufipennis Kirby) (Curculionidae:
Scolytinae) in North America. Molecular Ecology 16, 2560–2573.
Mitton, J. B., Kreiser, B. R., Latta, R. G. 2000. Glacial refugia of limber pine (Pinus
flexilis James) inferred from the population structure of mitochondrial DNA. Molecular
Ecology 9, 91–97.
Nice, C. C., Anthony, N., Gelembiuk, G., Ratermans, D., Ffrench-Constant, R. 2005.
The history and geography of diversification within the butterfly genus Lycaeides in
North America. Molecular Ecology 14, 1741–1754.
Olson, D. M., Dinerstein, E., Wikramanayke, E. D., Burgess, N. D., Powell, G. V. N.,
Underwood, E. C., D’Amico, J., Itoua, I., Strand, H. E., Morrison, J. C., Loucks, C. J.,
Allnutt, T. F., Ricketts, T. H., Kura, Y., Lamoreux, J. F., Wettengel, W. W., Hedao, P.,
Kassem, K. R. 2001. Terrestrial ecoregions of the world: New map of life on earth.
Bioscience 51: 933–938.
Ricketts, T. H., Loucks, C. J., Dinerstein, E. 1999. Terrestrial Ecoregions of North
America: A conservation Assessment. Island Press.
Richard B. Waitt, Jr. 1985. Case for periodic, colossal jökulhlaups from Pleistocene
glacial Lake Missoula. Geological Society of America Bulletin 96: 1271–1286.
29
Dissertation Christiane Kiefer – Introduction
Ritchie J. C., Cwynar L. C. 1982. The late Quaternary vegetation of the North Yukon,
in: Palaeoecology of Beringia. ed.s Hopkins, D. M., Matthews, J. V., Schweger, C. E.,
Young, S. B. Academic press, London p.113–126.
Rollins, R. C. 1993. The Cruciferae of Continental North America. Stanford
University Press, Stanford.
Schranz, M. E., Dobeš, C. H., Koch, M. A., Mitchell-Olds, T. 2005. Sexual
reproduction, hybridization, apomixis and polyploidization in the genus Boechera
(Brassicaceae). American Journal of Botany 92: 1797–1810.
Seibold, E., Seibold, I. 2005. Milankovitch’s Strahlenkurve und deren geologische
Deutung - Anfänge in Deutschland. International Journal of Earth Sciences 3: 495–
503.
Sharbel, T.F., Voigt, M.-L., Mitchell-Olds, T., Kantama, L., de Jong, H. (2004) Is the
aneuploid chromosome in apomictic Boechera holboellii a genuine B chromosome?
Cytogenetic Genome Research 106: 173–183.
Sharbel, T. F., Mitchell-Olds, T., Dobeš, C., Kantama, L., de Jong, H. 2005.
Biogeographic distribution of polyploidy and B chromosomes in the apomictic
Boechera holboellii complex. Cytogenetic Genome Research 109: 283–292.
Soltis, D. E., Morris, A. B., McLachlan, J. S., Manos, P. S., Soltis, P. S. 2006.
Comparative phylogeography of unglaciated eastern North America. Molec. Ecol.
15: 4261–4293.
Spellmann, G. M., Klicka, J. 2006. Testing hypotheses of Pleistocene population
history using coalescent simulations: phylogeography of the pygmy nuthatch (Sitta
pygmaea). Proceedings of the Royal Society B. 273: 3057–3063.
Spellmann, G. M., Klicka, J. 2007. Phylogeography of the white-breasted nuthatch
(Sitta carolinensis): Diversification in North American pine and oak woodlands.
Molecular Ecology 16: 1729–1740.
Spellmann, G. M., Riddle, B., Klicka, J. 2007. Phylogeography of the mountain
chickadee (Poecile gambeli): diversification, introgression, and expansion in
response to Quaternary climate change. Molecular Ecology 16: 1055–1068.
30
Dissertation Christiane Kiefer – Introduction
Stiff, B. J., Hansel, A. K. 2004. Quaternary glaciations in Illinois. in Ehlers, J., and
P.L. Gibbard, eds., pp. 71-82, Quaternary Glaciations: Extent and Chronology 2: Part
II North America, Elsevier, Amsterdam. ISBN 0-444-51462-7.
Stone, K. D., Flynn, R. W., Cook, J. A. 2002. Post-glacial colonization of
northwestern North America by the forest-associated American marten (Martes
americana, mammalia: Carnivora: Mustelidae). Molecular Ecology 11: 2049-2063.
Thompson, R. S., Whitlock, C., Bartlein, P. J., Harrison, S. P., Spaulding, W. G.
1993. Climatic changes in the western United States since 18,000 yr. B.P. p.468-513.
In; Global Climates Since the Last Glacial Maximum. ed.s; Wright H.E. Jr., Kutzbach
J.E., Webb T. III, Ruddiman W.F., Street-Perrott F.A. & Bartlein P.J.: University of
Minnesota Press, Minneapolis.
Voigt, M.-L., Melzer, M., Rutten, T., Mitchell-Olds, T., Sharbel, T. F. 2007.
Gametogenesis in the apomictic Boechera holboellii complex: the male perspective.
Pp. 235--258 in: Hörandl, E., Grossniklaus, U., can Dijk, P., Sharbel, T.F. (eds),
Apomixis: Evolution, Mechanisms and Perspectives. International Association for
Plant Taxonomy, Koeltz Scientific Books.
Whitlock C. & Bartlein P.J. 1997. Vegetation and climate change in northwest
America during the past 125 kyr. Nature, v.388 p.57–61.
Windham, M. D., Al-Shehbaz, I. A. 2006. New and noteworthy species of Boechera
(Brassicaceae) I: sexual diploids. Harvard Papers in Botany 11: 61–68.
Windham, M. D., Al-Shehbaz, I. A. 2007a. New and Noteworthy Species of Boechera
(Brassicaceae) II: Apomictic Hybrids. Harvard Papers in Botany 11: 257–274.
Windham, M. D., Al-Shehbaz, I. A., 2007b. New and Noteworthy Species of
Boechera (Brassicaceae) III: Additional Diploids and Apomictic Hybrids. Harvard
Papers in Botany 12: 251–274.
web-information:
Adams J.M. & Faure H. 1997. (ed.s), QEN members. Review and Atlas of
Palaeovegetation: Preliminary land ecosystem maps of the world since the Last
Glacial
Maximum.
Oak
Ridge
National
Laboratory,
TN,
USA.
http://www.esd.ornl.gov/ern/qen/adams1.html Retrieved in September 2008
31
Dissertation Christiane Kiefer – Introduction
Brief geologic history, Rocky Mountain National Park http://www.nature.nps.gov
/geology/parks/romo/index.cfm#geology Retrieved in September 2008
Ice Age Floods, From: U.S. National Park Service Website Retrieved September
http://vulcan.wr.usgs.gov/Glossary/Glaciers/IceSheets/description_ice_sheets.html
Le Blanc and Harris, http://www.cnr.berkeley.edu/departments/espm/extension/
ECOLOGY.HTM
Damery,
2004
Formation
of
the
Great
Lakes.http://www.emporia.edu
/earthsci/student/damery1/gl_form.html#Pre-Wisconsin_Drainage Retrieved on 16th
October 2008
32
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
2. Molecular Marker Based Studies of the North American Genus
Boechera (Brassicaceae)
2.1 cpDNA Gene pool Analyses
Manuscript submitted to MPE; supplementary data are included on the DVD in the
back of the dissertation.
PHYLOGEOGRAPHIC STRUCTURE OF THE CHLOROPLAST DNA GENE POOL
IN NORTH AMERICAN BOECHERA – A GENUS AND CONTINENTAL WIDE
PERSPECTIVE
Christiane Kiefer1, Christoph Dobeš1,3, Timothy F. Sharbel2 and Marcus A. Koch1
1
Heidelberg Institute of Plant Science, Biodiversity and Plant Systematics,
Heidelberg University, Im Neuenheimer Feld 345, 69120 Heidelberg, Germany
2
Apomixis Research Group, Dept. of Cytogenetics and Genome Analysis, Leibnitz
Institute of Plant Genetics and Crop Plant Research (IPK), 06466 Gatersleben,
Germany
3
Present address: Department of Pharmacognosy, University of Vienna,
Althanstrasse 14, A-1090 Vienna, Austria
keywords: Boechera / Brassicaceae / cpDNA / phylogeography / hybridisation
Corresponding author:
Marcus A. Koch
Heidelberg Institute of Plant Sciences, Department Biodiversity and Plant
Systematics, University of Heidelberg, Im Neuenheimer Feld 345, D-69120
Heidelberg, Germany.
Phone:
+49-6221-54-4655
FAX:
+49-6221-54-5508
eMAIL:
[email protected]
Running Title: cpDNA and species diversity analyses in Boechera
33
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Abstract
Continental wide phylogeographic studies of plants in North America are rare. In our
study we examined the phylogeographic history of Boechera (Brassicaceae) on a
continental wide scale testing if it is possible to do an analysis for 57 of the currently
accepted taxa simultaneously. We detected a large amount of haplotype sharing
indicating recurrent hybridisation on the one hand and non-differentiation of
haplotypes since speciation on the other hand. Hence, the chloroplast gene pool in
Boechera predates speciation and therefore justifies the simultaneous analysis of a
large number of taxa. Unrelated from taxon identity we can show that the
evolutionary lineages detected have a different phylogeographic history in terms of
glacial refugia and recently recolonized areas.
34
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Introduction
Phylogeography examines the spreading of taxa in space and time by relating
molecular genetic analyses to geography. The term phylogeography was introduced
only about 20 years ago (Avise et al. 1987) and since then a multitude of studies
investigated migration patterns of various plant and animal taxa (Avise 1998). Many
of these studies focused on the evolutionary history of European and North American
taxa during the Pleistocene, and the later introduced field of comparative
phylogeography enabled researchers to reveal common patterns and trends. The
distribution of species as we find it today is not only shaped by habitat preferences
but also by the glaciation cycles with alternating cool and warm periods during the
Quaternary which forced plants and animals to retreat into mostly southern located
glacial refugia during colder periods. Warm periods allowed the taxa to recolonize
their original distribution ranges again before they had to retreat into their refugia
again during the next glacial cycle. However, plants adapted to colder habitats could
also migrate and colonize new areas during periods in which temperatures
decreased again in the changing climate and changing landscape which continuously
provided new geographically defined corridors. The later fact might have facilitated
that not only migration, but also speciation was promoted (Jordon-Thaden and Koch,
accepted). These quaternary migrations left their footprints until today expressed in
the geographical distribution of DNA-based polymorphisms, a phenomenon often
described as “northern purity verses southern richness” (Hewitt, 2001).
Phylogeographic patterns in Europe were largely influenced by the Alps as an eastwest oriented barrier for southwards migration and the Pyrenees as a barrier
between the Iberian Peninsula and central Europe. Those two major obstacles lead
to clear phylogeographic patterns in several plant and animal taxa eg. refuge areas
close to the eastern margin of the Alps or on the Balkans or differentiation of Iberian
from central European populations (Hewitt, 2001; Taberlet et al. 1998).
Phylogeographic patterns in North America are less clear partly due to the northsouth orientation of North American mountain chains which meant that plants and
animals could migrate along the mountain chains into southern regions without
having to cross them. However, a major refuge area for several plant and animal taxa
was found in the southern central Rocky Mountains where the canyons were just
deep enough to maintain warmer temperatures suitable for survival (Ayoub and
Riechert, 2004; Brunsfeld and Sullivan, 2005; Mitton et al. 2000). For other plant and
animal taxa glacial refugia were found in the Colorado Rockies and the eastern Great
Basin (Limber pine, Mitton et al. 2000) and the eastern Great Lakes (Lycaeides, Nice
et al. 2005).
Most phylogeographic studies focus on the migration of a single species. Often only
small study areas are covered due to sampling of parts of the distribution ranges only
or a narrow distribution of the species. However, in order to investigate large scale
35
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
phylogeographic patterns it is necessary to study a system with a wide distribution
range (e.g. Koch and Matschinger, 2007; Jakob and Blattner, 2006). One of the few
continental wide phylogeographic studies in North America used three members of
the genus Boechera (Brassicaceae): the former Boechera holboellii sensu Rollins
(Rollins, 1993), Boechera divaricarpa sensu Rollins (Rollins, 1993) and Boechera
stricta, respectively (Dobeš et al. 2004a). This study suggested the Rocky Mountains
as a primary centre of genetic diversity and the Sierra Nevada as a secondary centre
of genetic diversity (Dobeš et al. 2004a). Those three species were taken as
representative for the whole genus since they have the widest distribution range of all
Boechera species.
In the herein presented study we test if it is possible to perform a genus-wide
phylogeographic study of Boechera by including 57 taxa which represent half of the
currently accepted 110 species (Al-Shehbaz, unpublished). If it is possible to do a
genus-wide phylogeographic analysis Boechera as a monophyletic group with a
continental-wide distribution may be an excellent model system for examining largescale phylogeographic patterns on the North American continent. Hence we would
like to compare patterns revealed by the large-scale study to other smaller studies
and see to what extent the results are congruent.
Furthermore genetic and species diversity should be compared for revealing to which
extent they are congruent. Congruence of the centres of genetic and species
diversity could be interpreted as long term stable populations or as regions subject to
multiple colonization in which speciation took place over a long time. Incongruencies
on the other hand could be interpreted as either more recent speciation events since
which no divergence of cpDNA haplotypes has happened yet or as populations in
which speciation had already happened and which then went through bottlenecks
and lost part of their genetic diversity.
Boechera together with seven other genera belongs to the tribe Boechereae (Bailey
et al. 2006). The Boechereae are almost exclusively found in North America (one
species in Siberia and Greenland only, respectively). According to the most recent
taxonomic concept Boechera comprises 110 species. Crossing experiments (Roy,
1995; Schranz et al. 2005) and studies of nrDNA ITS variation (Koch et al. 2003)
showed the high potential for hybridisation within the various members of the genus.
Hybrids are often triploid and reproduce asexually by apomixis (Dobeš et al. 2006).
Until the complete revision of Boechera for the new Flora of North America (AlShehbaz, unpublished; Windham and Al-Shehbaz 2006, 2007a, 2007b) triploid
hybrids were lumped together with the diploid taxa to which they were most similar.
Taxonomically they were often treated as a variety. Today the triploid apomicts are
separated from the diploid sexuals and are described as distinct species. Hence the
110 recognized species contain 72 sexually reproducing diploid species and 38
triploid apomictic species. These Boechera apomicts have been investigated in
36
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
various studies focusing on genome evolution (Kantama et al. 2007) or addressing
the phenomenon of apomixis (Voigt et al. 2007). Here it is still unclear if apomixis is a
driving evolutionary agent for speciation or simply the result of hybridization and
genomic/genetic stabilization. Apomixis/polyploidy seems to have arisen repeatedly
from a sexual background (Sharbel et al. 2001; Dobeš et al. 2004b) and has been
extensively described for the former “Boechera holboellii sensu Rollins” (Rollins
1993), an almost artificial taxon (Al-Shebaz pers. Comm.). Hence, inhere we also
look for apomictic lineages within Boechera by reconstructing networks and
phylogenetic trees based on chloroplast DNA marker sequences trnL-F and rpoC1
including more than 1300 accessions collected from herbarium specimens
representing 57 taxa.
Materials and Methods
PLANT MATERIAL
Leaf material was obtained from herbarium specimens from GH, MO and DAO and
from collections of Thomas Mitchell-Olds (Duke University, North Carolina, USA).
Corresponding accession details are listed in supplement table 1, the distribution of
the samples is given in figure 7 (online material). Herein we analysed in total 1286
accessions.
DNA EXTRACTION AND SEQUENCING
Total DNA extraction and PCR reactions were done as described in Dobeš et al.
(2004) with some minor changes. Prior to sequencing PCR products were purified
with the NucleoFast Kit (Macherey & Nagel). Cycle sequencing was done in our lab
or at the Genome Centre, IPK Gatersleben, using the DYEnamic ET Terminator
Cycle Sequencing Kit (Amersham Biosciences) and running a MegaBASE 500
sequencer. Sequencing was done in forward and reverse reaction.
DNA MARKER SELECTION
We used various DNA regions from the chloroplast genome to reconstruct maternal
phylogenetic lineages allowing also to reconstruct past migration history as the
chloroplast genome is inherited maternally in the Brassicaceae. The trnLF region was
analysed to characterize cpDNA haploytypes for further assessment of genetic
variation and diversity. The same DNA region was also used to reconstruct haplotype
networks and infer phylogenetic hypothesis. In order to increase significance of
networks and phylogenetic trees with its major evolutionary lineages we also added
sequence data from the rpoC1 intron to a subset of trnLF haplotypes.
37
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
ALIGNMENTS AND HAPLOTYPE DEFINITION
The forward and reverse sequences were aligned and trimmed. The trnL intron and
the trnLF intergenic spacer regions were assembled into one sequence and missing
bases at the 3’ and 5’ prime ends were substituted with N. The alignment was done
manually using the program GenDoc according to the alignment published by (Dobeš
et al. 2004a). Haplotypes were defined by running the program TCS 1.21 (Clement et
al. 2000). Gaps were set as a 5th state and the connection limit to 95%. In Boechera
(as in other Brassicaceae genera e.g. Arabidopsis, Cardaminopsis, refer to (Koch et
al. 2007a, 2007b)). trnF pseudogenes are present in the trnL-trnF intergenic spacer.
In past studies in our group (Dobeš et al. 2004a, Schmickl et al. 2008) it was shown
that pseudogene copies were gained and lost independently several times across the
phylogenetic tree. As the mechanism through which those pseudogenes arise and
multiply or are deleted is unknown, we omitted them from the analysis. Neglecting
the pseudogenes lead to the collapse of some haplotypes into single nodes in the
network analysis. Those haplotypes are hereafter referred to as suprahaplotypes.
Twenty suprahaplotypes (S1 to S20) were defined (see supplement table 3). For the
rpoC1 intron forward and reverse sequences were also assembled and trimmed and
subsequently aligned manually. For a combined analysis rpoC1 and trnLF sequence
data of the same individual were fused into a combined alignment to obtain more
variable sites for the analysis.
DNA-BASED PHYLOGENETIC RECONSTRUCTIONS AND NETWORK ANALYSIS
Phylogenetic reconstruction based on (1) the trnLF dataset and (2) a combined
dataset of a subset of the trnLF dataset together with rpoC1 sequences was done by
running a heuristic search with parsimony as optimality criterion in PAUP4.0beta
(Swofford 2001) under default parameters. The initial max trees was set to 1000.
Trees were rooted with Halimolobus perplexa as outgroup. The trnF pseudogene
region (see Koch et al. 2007a, 2007b) spanning bases 837 to 1001 in the alignment
was excluded in both analyses. A strict consensus tree was calculated from the 1000
shortest trees. For confirmation of the tree structures a bootstrap analysis was done.
Number of bootstrap replicates was set to 1000 and the optimality criterion was
parsimony. The search was also done with default parameters.
Network reconstruction was done by running the program TCS1.21 (Clement et al.
2000). Two analyses were carried out. In the first analysis the trnLF dataset was
analysed alone. First the alignment was split into three sub-alignments according to
the parsimony analysis. Then the pseudogene region was excluded and the
connection limit was set to 95%. After network reconstruction the sub-networks were
rejoined according to the parsimony analysis. In the second analysis a subset of the
trnLF dataset was combined with rpoC1 sequences in order to get a higher resolution
and to confirm the backbone of the network. trnF pseudogenes were again excluded.
In the network analysis the connection limit was set to six steps as more steps lead
38
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
to inaccuracies. Haplotypes that were excluded from the network due to the lower
connection limit were fused to their most likely position according to the phylogenetic
tree. Those added connections are shown as dotted lines.
For the identification of species specific lineages suprahaplotype identity and species
identity were entered in a table showing suprahaplotype identity, species identity and
percentage of suprahaplotype per species.
GENETIC AND SPECIES DIVERSITY
All geographical information available from the annotation on the herbarium vouchers
was entered into a BioOffice 2.0.4 (Biogis Consulting) database. Missing
geographical coordinates were added according to the descriptions on the herbarium
vouchers. Using BioOffice, haplotypes were plotted into the WWF ecoregions map
(Olson et al. 2001), and distribution ranges of haplotype lineages were later drawn by
hand on a satellite picture (Photo Courtesy of NASA).
For diversity statistics the dataset was divided into subsets according to ecoregions
as defined in the WWF ecoregions map version (Olson et al. 2001).
Genetic diversity was calculated as gene diversity and nucleotide diversity as
implemented in Arlequin vers. 2.0 (Excoffier et al. 2005) separately for ecoregions
and lineages. In cases where the number of accessions per region was below eight,
two adjacent regions with similar ecosystems were analysed together.
Phylogeogographic inferences were made by applying the concept of comparing
gene diversity and nucleotide diversity according to Avise (2000). Scenarios were
derived for all ecoregions where a sufficient number of samples was present.
Following the taxonomic concept of Windham and Al-Shehbaz and the distribution
ranges of species as given in their publications (Windham and Al-Shehbaz 2006,
2007a, 2007b) species number per state was recorded. Species numbers where
colour coded and states were shaded according to the colouring scheme. Species
diversity was calculated with and without apomictic species.
Results
CALIFORNIA AND NEVADA ARE THE ALPHA BIODIVERSITY HOTSPOT FOR
BOECHERA
According to the current taxonomic concept (Al-Shehbaz, submitted) for Boechera,
110 species are accepted of which 72 are sexually reproducing and 38 are apomictic.
Fifty-seven species are found in California, thus making it the species biodiversity
hotspot for Boechera. Fewer species have been reported from other states, including
Nevada (46), Oregon (32),Utah (30), Idaho (26) and Wyoming (23). Between 10 and
20 species were found in Colorado, Washington, Montana and Utah. Species
diversity decreased to the north, south and east from California (Figure 1a).
39
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Omitting the apomictic species from the analysis did not change the relative species
diversity among the states (data not shown).
Figure 1 (A) Species diversity calculated as number of species per state (B) Gene
diversity lineage I (C) Gene diversity lineage II (D) Gene diversity lineage III; increasing
colour depth = increasing gene diversity. Evolutionary lineages are shown and
correspond to Fig. 2.
40
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
DEFINITION OF HAPLOTYPES, PSEUDOGENES AND SUPRAHAPLOTYPES
The analysis of the trnL-trnF (total alignment length 1148 bp, see supplement Table 2
for the alignment) plastidic region allowed us to identify 241 cpDNA haplotypes.168
of those were found to be singletons and 92 haplotypes were shared by several
accessions. Haplotype coding was done according to Dobeš et al. (2004a)
(haplotypes A to DU) and Schranz et al. (2005) (haplotypes DV to GS) (see
supplement Table 3 for haplotype names and genebank accession numbers). Up to 3
trnF pseudogenes in the trnL-F intergenic spacer were found adjacent to the
functional trnF gene between alignment positions 785 and 1100 (copy I 785-900,
copy II 901-991, copy III 992-1000). Haplotypes differing only in pseudogene number
or mutations in this region were collapsed into suprahaplotypes (supplement Table
4).
For a higher resolution and statistical support in the phylogenetic reconstruction and
network analysis, 303 accessions from the trnLF dataset representing 127
haplotypes were combined with rpoC1 sequences. The total length of the rpoC1
alignment alone was 821 bp, the combined alignment had a total length of 1972 bp
(alignment see supplement Table 5). The additional variable characters in the rpoC1
region resulted in the definition of 147 combined haplotypes. Combined trnLF/rpoC1
haplotypes were given numbers which are listed in supplement Table 6 together with
the corresponding trnLF haplotype name and the genebank accession numbers of
the rpoC1 types.
LOW RESOLUTION IN THE BACKBONE: RPOC1 AND TRNLF TREES
Resolution of the phylogenetic trees obtained from the analyses of trnLF (Figure 4,
online material) was low. However it could be enhanced by the combination of the
trnLF and rpoC1 dataset (Figure 5, online material). Both new and previously
described evolutionary lineages (I, II and III as in (7)) could be identified, and
grouping between them was used to divide the dataset for subsequent network
analyses.
NETWORK STRUCTURE REVEALS SIX ANCESTRAL cpDNA LINEAGES
Network analysis of the trnLF dataset revealed a complex pattern of lineages as well
as several starlike nodes in the network (Figure 2).
Suprahaplotype 8 was found to be in the centre of the network to which all identified
lineages (I to VI) were connected. Lineage I consisted of two frequent
suprahaplotypes (S1 and S7) and a number of haplotypes and suprahaplotypes
derived from them. The same was true for lineage II, although the network structure
of this lineage was considerably less complex than that of lineage I. Most accessions
41
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Figure 2 Network Analysis based on trnLF chloroplast DNA; the network was
reconstructed using TCS1.21. Node size corresponds to number of accessions
carrying the (supra)haplotype. Portions of the network are coloured according to their
lineage identity. The same colour code was used for displaying the distribution ranges
in the top figures.
in lineage II were found either in S9 or S10. From these two suprahaplotypes a range
of haplotypes differentiated by a single mutation step were derived, a pattern
consistent with rapid range expansion (Koch et al. 2006). In Lineage III the frequency
of suprahaplotypes increased with the distance from the centre of the network, an
unusual finding as central/old haplotypes are typically most frequent. The most
frequent haplotypes were all derived from each other. From S13, S14 and S16
several haplotypes were derived, giving the same pattern as found in lineage II for S9
and S10. The network analysis of the combined dataset for trnLF and rpoC1 yielded
the same result, although the addition of the rpoC1 sequence produced additional
mutation steps due to more variable characters (Figure 6, online material).
HAPLOTYPE LINEAGE DISTRIBUTION RANGES BROADLY OVERLAP
The distribution ranges of the cpDNA lineages I, II, III and VI overlapped in western
North America (the Great Basin and adjacent mountain ranges). Lineage I occurred
in the main distribution range of Boechera and only 15 individuals carrying lineage I
haplotypes were found in the Yukon and the Northern Rocky Mountains. Lineage II
(suprahaplotypes 9 and 10) and III (suprahaplotypes 13, 14, 15) were also found in
the range of lineage I, but also occurred north of the last glaciation. The most derived
haplotypes within these lineages were nonetheless found in more southern regions.
Lineage VI was distributed in the north-eastern range of lineage I (Figure 2).
42
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
PATTERNS OF GENETIC DIVERSITY ARE DIFFERENT IN LINEAGES I, II AND III
Centres of gene diversity were found to be different for lineages I, II and III, and
followed a slight west to east trend. Lineage I had its centre of gene diversity in the
Klamath-Siskiyou region north of the Sierra Nevada (p=0.94), followed by the the
Snake/Columbia Plateau and the Wasatch/Uinta Range (p=0.89 and p=0.85) (figure
1b). The centre of gene diversity for lineage II was was found to be the Colorado
Plateau and the adjacent Colorado Rocky Mountains (p=0.78), and decreased
towards the west, north and south (Figure 1c).
For lineage III the highest gene diversity was found in the southeastern central Rocky
Mountains (Figure 1d) (p=3.0), and followed the Sierra Nevada and the other
mountains ranges surrounding the Great Basin.
In general the gene diversity was lowest in regions north of the last glaciation. The
highest number of singletons in lineage I and II was found in the centre of gene
diversity, while in Lineage III this was found in the Sierra Nevada which had the
second highest gene diversity.
RANGE
EXPANSION,
STABLE
POPULATIONS
AND
MULTIPLE
COLONIZATIONS: PHYLOGEOGRAPHIC ANALYSIS BY THE COMPARISON OF
GENE AND NUCLEOTIDE DIVERSITY
Following Avise (2000) the comparison of gene (h) and nucleotide (p) diversity allows
the inference of possible phylogeographic scenarios, including: (A) high h and p –
stable population over long term with large Ne (effective population size) or multiple
colonizations and admixed populations, (B) high h and low p – rapid population
growth from small ancestral population with small Ne, (C) low h and high p –
transient bottleneck in large ancestral population or admixture from small,
geographical subdivided population and (D) low h and low p – rapid long distance
dispersal and migration or prolonged or severe bottleneck. For regions south of the
last glaciation the derived phylogeographic scenario often differed for lineages I, II
and III within one region (Figure 3, and supplement Tables 7 and 8). For regions
north of the last glaciation we inferred rapid range expansion except for the
ecoregion south of the Great Lakes for lineage II.
SPECIES
AND
HAPLOTYPE-DIFFERENTIATION
–
TWO
UNLINKED
PROCESSES?
Our network analysis demonstrated a mixture of few species specific lineages and
suprahaplotypes that are shared by up to 27 taxa. True species specific lineages
were rare. In lineage I suprahaplotype 20 and its derivatives represented a lineage
specific to B. microphylla. Lineage II was dominated by B. stricta individuals. Lineage
IV was specific for B. canadensis, and lineage V was characterized by several
Eastern North American Boechera, most of them with private haplotypes (B.
laevigata, B. missouriensis, B. perstellata, B. serotina). Lineage III contained
43
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
only
Figure 3 Phylogeographic scenarios derived according to Avise (2000). A-C Scenarios
inferred for lineages I, II, III, D Summary of the most important phylogeographic
implications.
only one species-specific lineage ((Boechera fendleri var. spatifolia, now Boechera
spatifolia (14)). Suprahaplotypes 2, 9, 10, 11, 19 and 20 were shown to be speciesspecific or predominantly found in one species.
Discussion
Phylogeographic studies often consider only one single species because different
species may have different habitat preferences and life cycles and hence may show
a different response to climatic events such as the Quaternary ice ages. If several
species are considered phylogeographic histories of species are first inferred for
single species and later compared to each other (comparative phylogeography). For
Boechera we tested if it is possible to conduct a phylogeographic study
simultaneously for 57 taxa. Our network analysis based on cpDNA markers trnL-F
and rpoC1 revealed six evolutionary lineages all arising from a central
suprahaplotype. Comparing haplotype identity and species identity showed that
haplotype sharing is very common within Boechera. This can either be explained by
frequent hybridisation which was previously reported for Boechera or by the fact that
haplotype differentiation predates speciation. Both possibilities seem to hold true for
44
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Boechera. Since the cpDNA haplotypes are older than the species in many cases
they represent an old gene pool. Therefore we conclude that a phylogeographic
study based on this cpDNA marker system can be conducted for the whole group of
57 taxa simultaneously if the dataset is subdivided according to the evolutionary
lineages inferred.
Phylogeographic analysis was done by the comparison of gene and nucleotide
diversity (Avise, 2000) for the major evolutionary lineages I, II and III separately.
Lineages IV and V were treated separately since they comprise only species
restricted to Eastern North America (Kiefer et al. submitted). Gene diversity is based
on the number of different DNA sequences and their frequency in a region while
nucleotide diversity is a measure of differences among haplotypes within one region.
The higher the gene diversity, the more different sequences were detected, the
higher the nucleotide diversity, the more distantly related are the haplotypes. If gene
and nucleotide diversity are both high this is taken as an indication for a long term
stable population because new haplotypes had time to evolve (high gene diversity)
and differentiate (high nucleotide diversity). Another possible interpretation may be
multiple colonization events into this region also because several (high gene
diversity) haplotypes which had previously differentiated (high nucleotide diversity)
arrived from different source regions. On the other hand a region where gene and
nucleotide diversity are low has either experienced a severe bottleneck during which
most of the genetic diversity was lost or recent rapid range expansion where only a
limited number of haplotypes reached the colonized area and time was too short for
differentiation.
Our analysis revealed that the phylogeographic history of the evolutionary chloroplast
DNA lineages I, II and III is different although their distribution ranges overlap in most
parts. The lineages differed in respect to the centres of gene diversity and also in
their distribution of singleton haplotypes. However, all three lineages had their lowest
gene and nucleotide diversity north of the last glaciation indication recent range
expansion. This is a common pattern found in phylogeographic studies often referred
to as southern richness versus northern purity caused by the retreat of taxa into
refugia south of the ice shield during the quaternary ice ages (Hewitt, 2001). One
exception was the Great Lakes region where we detected a high gene diversity for
lineage II. This was already noted in the previous phylogeographic study and
interpreted as a potential glacial refuge in this region (Dobeš et al. 2004a). The
eastern Great Lakes region was also shown to be a glacial refuge for butterflies
(Lycaeides; Nice et al. 2005).A small isolated range of lineage I haplotypes in the
Yukon interior dry forest in western Canada should also be noted. Only individuals
carrying haplotype M included in suprahaplotype 1 (S1) were found in this northern
enclave separated from the distribution range of lineage I by several hundred
kilometres. Haplotype M is an old haplotype since S1 is central to lineage I. Its
45
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
presence in the Yukon interior dry forest might indicate a larger preglacial distribution
range of lineage I reaching into the north of Canada. However, long-distance
dispersal of this haplotype from the Cascades or the central Rocky Mountains where
it is most abundant cannot be ruled out.
South of the last glacation no common pattern was found for the three evolutionary
lineages. Klamath-Siskiyou was found to be the centre of genetic diversity for lineage
I. This ecoregion is known for its extraordinary biodiversity and it was not or at least
not heavily affected by the quaternary ice-ages (Ricketts et al. 1999). Therefore it
may have offered a suitable habitat to Boechera over a long period of time.
Comparison of h and p indicates a long term persistent population for lineage I in
Klamath/Siskiyou, which also hosted the largest number of singleton haplotypes.
However, for lineages II and III rapid long distance dispersal was inferred, thus
illustrating the different phylogeographic histories of each evolutionary lineage.
Considering the haplotype distribution ranges it is likely that lineage III invaded
Klamath/Siskiyou from the Sierra Nevada (shared haplotype FE) or the central and
southern Cascades postglacially. The same may hold true for lineage II. This leads
us to the conclusion that Klamath/Siskiyou possibly offered possibilities for
hybridisation between lineage I and lineage II or III haplotype carriers.
The Colorado Plateau was the centre of genetic diversity for lineage II and both gene
and nucleotide diversity were high. Evidence for a long term stable population of
lineage II also comes from the presence of singleton haplotypes not only on the
Colorado Plateau but especially in the adjacent Colorado Rocky Mountains. The
Colorado Rocky Mountains were described as a refuge for limber pine (Mitton et al.
2000) which fits well with a long term persistent population of Boechera lineage II. On
the other hand the Colorado Rockies were also described as a genetically poor area
(desert spider, Ayoub and Riechert 2004), a scenario which also applies to lineage I
which was genetically poor in the Colorado Rocky Mountains. A high gene diversity
was found for lineages I and III in the Colorado Plateau, although a high gene and
haplotype diversity can be interpreted in two ways; a long term stable population or
multiple colonization events. The latter interpretation seems to be true for lineages I
and III because the suprahaplotypes present in the region do not form a continuous
lineage but have missing “connecting” suprahaplotypes between them. The long term
persistence of lineage II and the possibility of several subsequent colonization events
by lineage I or II haplotype carriers offers multiple possibilities of hybridisation
between individuals carrying haplotypes from different lineages.
The southern central Rocky Mountains comprise a region which has been addressed
numerous times in phylogeographic studies in the past. In our analysis the southern
central Rocky Mountains were divided into an eastern and a western section. The
southeastern Rocky Mountains were shown to be the centre of genetic diversity for
lineage III. Gene diversity and nucleotide diversity indicated a long term stable
population or multiple colonization events. The eastern part of the southern central
46
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Rocky Mountains has only been described as a glacial refuge for limber pine (Mitton
et al. 2000) which might be explained by the fact that in the Eastern part of the
southern central Rocky Mountains glaciation is assumed to have started in the
Windriver Range (Leonard 2007). The high diversity in this ecoregion seems to be
influenced by haplotypes that are also found in more southerly regions of the Rocky
Mountains like the Colorado Rockies or the Wasatch/Uinta Mountains.
The southwestern central Rocky Mountains are separated from the southeastern
central Rocky Mountains by the Montana Valley and Foothill Grasslands which are in
parts characterized by a semi-dry landscape similar to that found on the
Snake/Columbia plateau (Olson et al. 2001). The gene diversity in Boechera was
lower in the southwestern central Rockies than in the southeastern central Rockies
and the percentage of singleton haplotypes was about the same. However,
nucleotide diversity was lower, so we rapid population growth from ancestral
population with small Ne (Avise 2000) for this ecoregion. This may indicate a glacial
refuge for Boechera with a small remaining population from which it expanded rapidly
after deglaciation within this part of the Rocky Mountains chain. This hypothesis is
likely as the western part of the southern central Rocky Mountains (Salmon River
Mountains, Clearwater Range) is known as a refuge area (e.g. Salix melanopsis
(Brunsfeld et al. 2007); Cardamine constancei (Brunsfeld and Sullivan 2005); redtailed chipmunk (Good and Sullivan 2001); pine beetle (Maroja et al. 2007)) and as
genetic contact zone (Brunsfeld and Sullivan 2005).
In terms of haplotype composition the southwestern central Rocky Mountains are
similar to the northern central Rocky Mountains. The northern central Rocky
Mountains stick out among other ecoregions as it is the only one in which all major
suprahaplotypes from lineage III are present, thus making it a potential centre of
origin or refuge area. However, only one singleton haplotype occurs here compared
to the Sierra Nevada, where nine singleton haplotypes occur in addition to all but one
suprahaplotype. Therefore we hypothesize that lineage III evolved in the Sierra
Nevada, and from there colonized today’s distribution range to reach the northern
central Rocky Mountains via the southwestern central Rocky Mountains. It may have
survived in the Bitterroot Range or adjacent ranges, where S15 (the suprahaplotype
absent from the Sierra Nevada) evolved to become the major postglacial colonization
haplotype.
In summary our continent-wide phylogeographic study of Boechera shows that within
this genus there are three lineages with a unique phylogeographic history which
shows migration and colonization for a time predating speciation in this genus.
Migration seems not to have happened only one way. As differentiation in a lineage
proceeded haplotypes moved further into new habitats and at the same time
backwards into their old range leading to the overlap of distribution ranges of
haplotypes today.
47
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
In comparison to other phylogeographic studies the results for Boechera are in
agreement with some of them while they are in conflict with others. Boechera’s
phylogeographic history is in concordance with patterns which were often revealed in
the past such as the glacial refugium in the southern central Rocky Mountains. More
phylogeographic studies covering a larger area would be needed for a better
comparison or for analysing common patterns as it was done for eastern North
America (Soltis et al. 2007).
In order to connect the obtained gene diversity data to their biological background we
calculated species diversity per state and compared haplotype and species identity.
California was identified as the state with the highest species diversity for Boechera,
even when apomictic species were excluded from the analysis. California is known
for its botanical biodiversity and is among the 25 biodiversity hotspots on Earth
(Calsbeek et al. 2003). The high biodiversity in California can be explained by the
presence of multiple habitat types (elevation belts of the Sierra Nevada) which may
also be the reason for the high species diversity in Boechera. The largest portion of
the adjacent state of Nevada, which is second in Boechera biodiversity, is covered by
the Great Basin, a highly structured landscape which displays a mixture of coldtemperate and desert vegetation which shows the influence of the adjacent Mojave
Desert (Ricketts et al. 1999). This particular mixing of species with different climate
preferences was also true for Boechera (see supplement table 9) and is correlated
with the high species diversity in Nevada.
Comparison of species and haplotype identity revealed few species specific
haplotypes or lineages, in conjunction with elevated levels of haplotype sharing.
Suprahaplotype 8, which is central to the network and connects all lineages is the
oldest suprahaplotype according to coalescent theory (Posada and Crandall 2000). It
occurs only in the southern Great Basin and is shared by several species, and has
been estimated to be 0.7 to 2 million years old (Dobeš et al. 2004a). This places the
origin of all Boechera lineages and also all derived haplotypes into the pleistocene.
While age estimates for other haplotypes are imprecise and characterized by large
confidence intervals, the relative ages between haplotypes can nonetheless be
compared. Species-specific lineages connected directly to S8 are likely relatively old
species as they represent the earliest split possible within Boechera. These early
diverged species-specific lineages include lineage II (specific to B. stricta; haplotypes
are rarely shared), lineage IV (specific to B. canadensis), and lineage V which
represents another eastern North American branch within Boechera (B. laevigata, B.
missouriensis and B. shortii). (Supra)haplotypes specific to Boechera microphylla
and Boechera spatifolia were derived from suprahaplotypes further away from the
centre of the network, thus implying that these species have a more recent origin.
48
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Although some species- and lineage-specific haplotypes were detected, they are the
exception since haplotype sharing is the rule. One explanation for haplotype sharing
could be frequent hybridisation and subsequent backcrossing, which may lead to
chloroplast capture. Alternatively the age of some species may be so young that
significant differentiation between haplotypes has not yet happened. This seems to
be well illustrated by S16 (lineage III) which is shared by 50% of the taxa in our
analysis.
Tracing species abundance through the suprahaplotypes in the network, it is
remarkable that one group of species appears in all major nodes except for the
species-specific lineages. Those species are - according to Rollins’ taxonomic
system (Rollins 1993) – the five varieties of B. holboellii which have all obtained
species rank (Windham and Al-Shehbaz 2006). Their presence in all major nodes of
the network means that these species or former varieties encompass most of the
chloroplast DNA genetic variability found in the genus Boechera. We therefore think it
is likely that Boechera holboellii sensu Rollins represents the ancestral types from
which most other Boechera species evolved.
Plotting apomictic species on the network revealed that they are not found
within one lineage but rather scattered across the tree. There were indications from
microsatellite studies that apomixis arose repeatedly from a sexual background after
hybridisation (Sharbel et al. 2001; Dobeš et al. 2004b). At least the repeated
independent expression of apomixis is also supported by chloroplast DNA data.
However, since chloroplasts are maternally inherited it may be more precise to say
that there is no single maternal apomictic lineage. Since chloroplast differentiation
predates speciation in most cases in Boechera nuclear data would be needed to
really prove the absence of apomictic lineages. This is work currently being in
progress.
Summary and Outlook
Based on cpDNA haplotype variation and differentiation we were able to reconstruct
major processes in Boechera evolutionary history and biogeography. For some taxa
these reconstructions were even possible on species-level. However, more research
is needed to link species definition (alpha taxonomy) with phylogeny and
biogeography. Our ongoing research and preliminary results based on ITS (internal
transcribed spacer of nuclear ribosomal DNA) and various single copy genes indicate
that simple gene trees and even coalescent theory might not help to resolve any
species. It is more likely that a continental wide population genetics approaches will
lead to a deeper understanding of the relationship of single Boechera species.
ACKNOWLEDGEMENTS
Funding over the last years from DFG (German Research Foundation) and ASF
(Austrian Science Foundation) to M. A. K. was greatly acknowledged. We are very
49
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
grateful for suggestions and comments made by Dr. Ihsan Al-Shehbaz (Missouri
Botanical Gardens/MO/USA).
LITERATURE CITED
Al-Shehbaz, I.A., submitted. Boechera A. Löwe and D. Löwe. In: Flora North
America.
Avise, J., 2000. In: Phylogeography: The History and Formation of Species. Harvard
University Press, Cambridge, MA.
Avise, J.C., Arnold, J., Ball, R.M., Bermingham, E., Lamb, T., Neigel, J.E., Reeb,
C.A., Saunders, N.C., 1987. Intraspecific phylogeography: the mitochondrial
DNA bridge between population genetics and systematics. Annual Reviews in
Ecology and Systematics 18, 489-522.
Avise, J.C., 1998. The history and purview of phylogeography: a personal reflection.
Molecular Ecology 7, 371-379.
Ayoub, N.A., Riechert, S., 2004. Molecular evidence for Pleistocene glacial cycles
driving diversification of a North American desert spider, Agelenopsis aperta.
Molecular Ecology 13, 3453-3465.
Bailey, C.D., Koch, M.A., Mayer, M., Mummenhoff, K., O’Kane Jr., S.L., Warwick,
S.I., Windham, M.D., Al-Shehbaz, I.A.., 2006. Toward a global phylogeny of
the Brassicaceae. Molecular Biology and Evolution 23, 2142-2160.
Brubaker, L.B., Anderson, P.M., Edwards, M.E., Lozhkin, A.V., 2005. Beringia as a
glacial refugium for boreal trees and shrubs: new perspectives from mapping
pollen data. Journal of Biogeography 32, 833-848.
Brunelle, A., Whitlock, C., 2003. Postglacial fire, vegetation, and climate history in the
Clearwater Range, Northern Idaho, USA. Quaternary Research 60, 307-318.
Brunsfeld, S.J., Miller, T.R., Carstens, B.C., 2007. Insights into the Biogeography of
the Pacific Northwest North America: Evidence from the Phylogeography of
Salix melanopsis. Systematic Botany 32, 129-139.
Brunsfeld, S.J., Sullivan, J., 2005. A multi-compartmented glacial refugium in the
northern Rocky Mountains: Evidence from the phylogeography of Cardamine
constancei (Brassicaceae). Conservation Genetics 6, 895-904.
Calsbeek, R., Thompson, J.N., Richardson, J.E., 2003. Patterns of molecular
evolution and diversification in a biodiversity hotspot: the California Floristic
Province. Molecular Ecology 12, 1021-1029.
Clement, M., Posada, D., Crandall, K.A., 2000. TCS: a computer program to estimate
gene genealogies. Molecular Ecology 9, 1657-1660.
Dobeš, C.H., Mitchell-Olds, T., Koch, M.A., 2004a. Extensive chloroplast haplotype
variation indicates Pleistocene hybridization and radiation of North American
Arabis drummondii, A. X divaricarpa, and A. holboellii (Brassicaceae).
Molecular Ecology 13, 349-370.
50
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Dobeš, C.H., Koch, M.A., Mitchell-Olds, T., 2004b. Intraspecific diversification in
North American Boechera stricta (= Arabis drummondii), Boechera x
divaricarpa, and Boechera holboellii (Brassicaceae) inferred from nuclear and
chloroplast molecular markers—an integrative approach. American Journal of
Botany 91, 2087-2101.
Dobeš, C., Koch M., Sharbel, T., 2006. Embryology, karyology, and modes of
reproduction in North American genus Boechera (Brassicaceae): a compilation
of seven decades of research. Annals of the Missouri Botanical Gardens 93,
517-533.
Excoffier, L., Laval, G., Schneider, S., 2005. Arlequin (version 3.0): An integrated
software for population genetics data analysis. Evolutionary Bioinformatics
Online 1, 47-50.
Good, J.M., Sullivan, J., 2001. Phylogeography of the red-tailed chipmunk (Tamias
ruficaudus), a northern Rocky Mountain endemic. Molecular Ecology 10, 26832695.
Hewitt, G., 2000. The genetic legacy of the Quaternary Ice Ages. Nature 405: 907913.
Hewitt, G.M., 2001. Speciation, hybrid zones and phylogeography - or seeing genes
in space and time. Molecular Ecology 10, 537-549.
Hull, J.M., Girman, D.J., 2005. Effects of Holocene climate change on the historical
demography of migrating sharp-shinned hawks (Accipiter striatus velox) in
North America. Molecular Ecology 14, 159-170.
Jakob, S.S., Blattner, F.R., 2006. A Chloroplast Genealogy of Hordeum (Poaceae):
Long-Term Persisting Haplotypes, Incomplete Lineage Sorting, Regional
Extinction, and the Consequences for Phylogenetic Inference. Molecular
Biology and Evolution 23, 1602-1612.
James, L.A., Harbor, J., Fabel, D., Dahms, D., Elmore, D., 2002. Late Pleistocene
Glaciations in the Northwestern Sierra Nevada, California. Quaternary
Research 57, 409-419.
Jordon-Thaden, I., Koch, M. accepted. Diversity patterns in the genus Draba: A first
global perspective. Plant Ecology and Diversity.
Kantama, L., Sharbel, T.F., Schranz, M.E., Mitchell-Olds, T., de Vries, S., de Jong,
H., 2007. Diploid apomicts of the Boechera holboellii complex show large
scale chromosome substitutions and different aberrant chromosomes.
Proceedings of the National Academy of Sciences 104, 14026-14031.
Koch, M., Dobeš, C., Mitchell-Olds, T., 2003.: Multiple hybrid formation in natural
populations: Concerted evolution of the internal transcribed spacer of nuclear
ribosomal DNA (ITS) in North American Arabis divaricarpa (Brassicaceae).
Molecular Biology and Evolution 20, 338-350.
51
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Koch, M., Kiefer, C., Vogel, J., Ehrich, D., Brochmann, C., Mummenhoff, K., 2006.
Three times out of Asia Minor - the phylogeography of Arabis alpina L.
(Brassicaceae). Molecular Ecology 15, 825-839.
Koch, M.A, Dobeš, C., Kiefer, C., Schmickl, R., Klimes, L., Lysak, M.A., 2007a.
SuperNetwork identifies multiple events of plastid trnF (GAA) pseudogene
evolution in the Brassicaceae. Molecular Biology and Evolution 24, 63-73.
Koch, M.A., Matschinger, M., 2007b. Evolution and genetic differentiation among
relatives of Arabidopsis thaliana. Proceedings of the National Academy of
Sciences 104, 6272-6277.
Leonard, E., 2007. Modelled patterns of Late Pleistocene glacier inception and
growth in the Southern and Central Rocky Mountains, USA: sensitivity to
climate change and paleoclimatic implications. Quaternary Science Reviews
26, 2152-2166.
Maroja, L.S., Bogdanowicz, S.M., Wallin, K.F., Raffa, K.F., Harrison, R.G., 2007.
Phylogeography of spruce beetles (Dendroctonus rufipennis Kirby)
(Curculionidae: Scolytinae) in North America. Molecular Ecology 16, 25602573.
Mitton, J.B., Kreiser, B.R., Latta, R.G., 2000. Glacial refugia of limber pine (Pinus
flexilis James) inferred from the population structure of mitochondrial DNA.
Molecular Ecology 9, 91-97.
Nice, C.C., Anthony, N., Gelembiuk, G., Ratermans, D., Ffrench-Constant, R., 2005.
The history and geography of diversification within the butterfly genus
Lycaeides in North America. Molecular Ecology 14, 1741-1754.
Olson, D.M., Dinerstein, E., Wikramanayke, E.D., Burgess, N.D., Powell, G.V.N.,
Underwood, E.C., D’Amico, J., Itoua, I., Strand, H.E., Morrison, J.C., Loucks,
C.J., Allnutt, T.F., Ricketts, T.H., Kura, Y., Lamoreux, J.F., Wettengel, W.W.,
Hedao, P., Kassem, K.R., 2001. Terrestrial ecoregions of the world: New map
of life on earth. Bioscience 51, 933-938.
Osborn, G., Bevis, K., 2001. Glaciation in the Great Basin of the Western United
States. Quaternary Science Reviews 20, 1377-1410.
Phillips, F.M., Zreda, M.G., Smith, S.S., Elmore, D., Kubik, P.W., Sharma, P., 1990.
Cosmogenic chlorine-36 chronology for glacial deposits at Bloody Canyon,
eastern Sierra Nevada California. Science 248, 1529-1532.
Posada, D., Crandall, K.A., 2001. Intraspecific phylogenetics: Trees grafting into
networks. Trends in Ecology and Evolution 16, 37-45.
Ricketts, T.H., Loucks, C.J., Dinerstein, E., 1999. Terrestrial Ecoregions of North
America: A conservation Assessment. Island Press.
Rollins, R.C., 1993. The Cruciferae of Continental North America. Stanford University
Press, Stanford, CA.
Roy, B.A., 1995. The breeding system of six species of Arabis (Brassicaceae).
American Journal of Botany 82: 869–877.
52
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Schmickl, R., Kiefer, C., Dobeš, C., Koch, M.A., 2008. Evolution of trnF(GAA)
pseudogenes in cruciferous plants. Plant Systematics and Evolution: DOI
10.1007/s00606-008-0030-2.
Schranz, M.E., Dobeš, C.H., Koch, M.A., Mitchell-Olds, T., 2005. Sexual
reproduction, hybridization, apomixis and polyploidization in the genus
Boechera (Brassicaceae). American Journal of Botany 92, 1797-1810.
Sharbel, T.F., Mitchell-Olds, T., 2001. Recurrent polyploid origins and chloroplast
phylogeography in the Arabis holboellii complex (Brassicaceae). Heredity 87,
59-68.
Soltis, D.E., Morris, A.B., McLachlan, J.S., Manos, P.S., 2006. Comparative
phylogeography of unglaciated eastern North America. Molecular Ecology 15,
4261-4293.
Swofford, D.L., 2002. PAUP*: Phylogenetic Analysis Using Parsimony (*and Other
Methods), Version 4. Sinauer Associates, Sunderland, Massachusetts.
Thompson, R.S., Anderson, K., 2000. Biomes of western North America at 18,000,
6,000 and 0 14C yr B.P. reconstructed from pollen and packrat midden data. J
ournal of Biogeography 27, 555-584.
Taberlet, P., Fumagalli, L., Wust-Saucy, A.G., Cosson, J.F., 1998. Comparative
phylogeography and postglacial colonization routes in Europe. Molecular
Ecology 7, 453-464.
Voigt, M.L., Melzer, M., Rutten, T., Mitchell-Olds, T., Sharbel, T.F. In: Hörandl, E.,
Grossniklaus, U., Van Dijk, P., Sharbel, T.F., 2007. in Apomixis: Evolution,
Mechanisms and Perspectives, eds Hörandl E, Grossniklaus U, Van Dijk P,
Sharbel TF (Koeltz, Koenigstein, Germany), pp 235–258.
Windham, M.D., Al-Shehbaz, I.A., 2006. New and noteworthy species of Boechera
(Brassicaceae) I: sexual diploids. Harvard Papers in Botany 11, 61-68.
Windham, M.D., Al-Shehbaz, I.A., 2007a. New and Noteworthy Species of Boechera
(Brassicaceae) II: Apomictic Hybrids. Harvard Papers in Botany 11, 257-274.
Windham, M.D., Al-Shehbaz, I.A., 2007b. New and Noteworthy Species of Boechera
(Brassicaceae) III: Additional Diploids and Apomictic Hybrids. Harvard Papers
in Botany 12, 251-274.
Legends supplementary material
Figure 4 (online material) Parsimony analysis of the trnLF dataset calculated from
1000 shortest trees.
Figure 5 (online material) Parsimony analysis of the combined trnLF-rpoC1 dataset
calculated from 1000 shortest trees.
53
Dissertation Christiane Kiefer – cpDNA Gene Pool Analysis
Figure 6 (online material) Network analysis based on the combined trnLF-rpoC1
dataset. Lineages identified were the same as in the analysis based on trnLF only.
The number of the combined trnLF-rpoC1 haplotype is given in the corresponding
node. Accessions included and their trnLF haplotype names are given in supplement
table 5.
Figure 7 (online material) Distribution of all samples included in this analysis.
Table 1 (supplement material) Accession List stating accession number, origin of
herbarium voucher, geographic coordinates, haplotype evolutionary lineage and
ecoregion in which the sample was placed.
Table 2 (supplement material) Annotated alignment of all trnLF haplotypes including
pseudogene region.
Table 3 (supplement material) Genebank accession details of haplotypes and
voucher information of the individual in which the haplotype was found first.
Table 4 (supplement material) Suprahaplotype codes and haplotypes included within
them.
Table 5 (supplement material) Combined alignment of rpoC1 and trnLF
Table 6 (supplement material) Combined trnLF-rpoC1 haplotypes and accession
numbers of individuals in which they were found
Table 7 (supplement material) Haplotype abundance and haplotype names for each
ecoregion investigated and gene and nucleotide diversity based on those numbers; a
separate table for every lineage identified is included. A column also states the
phylogeographic scenario inferred according to (16).
Table 8 (supplement material) Phylogeographic scenarios inferred for ecoregions
and lineages along with information on ecoregions and a comparison to other
phylogeographic studies.
Table 9 (supplement material) Species Distribution in the United States and Canada.
1 indicates the presence of a species in a state or province. The colour code is given
below the table.
54
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
2.5 ITS and Single Copy Gene Analyses
This manuscript will be submitted to American Journal of Botany.
Supplementary material is given on the DVD in the back of the thesis.
Analyses which are in progress for the manuscript submission include Bayesian
analyses for all datasets used in this study and a lineage through time plot for the
ITS dataset.
Running Title:
ITS and single copy gene intron studies in Boechera
Molecular systematics and evolutionary history of North American
Boechera (Brassicaceae) – nuclear encoded ITS and single copy
non coding intron sequences characterize old lineages and single
species.
CHRISTIANE KIEFER, CHRISTOPH DOBEŠ*, AND MARCUS A. KOCH#
Heidelberg Institute of Plant Science, Biodiversity and Plant Systematics, Heidelberg
University, Im Neuenheimer Feld 345, 69120 Heidelberg, Germany
*
Present address: Department of Pharmacognosy, University of Vienna,
Althanstrasse 14, A-1090 Vienna, Austria
# Corresponding author:
Marcus A. Koch
Heidelberg Institute of Plant Sciences, Department of Biodiversity and Plant
Systematics, University of Heidelberg, Im Neuenheimer Feld 345, D-69120
Heidelberg, Germany.
Phone:
+49-6221-54-4655
FAX:
+49-6221-54-5508
eMAIL:
[email protected]
55
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Manuscript received _______; revision accepted _______.
Acknowledgements
We are grateful to the curators of the herbaria GH, MO and DAO for providing leaf
material, and Ihsan Al-Shehbaz for continuous advice and support during our studies
on Boechera .
Abstract:
110 species are currently described for the North American genus Boechera
(Brassicaceae). Several hybrid combinations are assumed based on morphological
evidence. However, little is known about the phylogenetic relatedness of the species
to each other. We used nrDNA ITS and intron sequences of two single copy genes to
obtain deeper insights into Boechera phylogeny and ITS type distribution across
taxa. We could show that species specific lineages exist although the relationship
among them is poorly resolved. Comparing tree topologies this indicates rapid
speciation which probably happened in the second half of the quaternary. Hybrids
could be identified by the comparison of the different marker systems together with
chloroplast DNA types from an earlier study.
We hypothesize that a continental wide population biology approach would be
needed to unravel the complex relationships of evolutionary lineages in this diverse
genus.
keywords: Boechera / Brassicaceae / ITS / single copy genes / hybridisation
56
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
INTRODUCTION
As complex as the geography of the North American continent is the genus Boechera
(Brassicaceae) living in this heterogeneous environment. Boechera is part of the tribe
Boechereae that encompasses seven closely related genera, Anelsonia, Cusickiella,
Nevada, Phoenicaulis, Polyctenium, Sandbergia (= Halimolobus perplexa, AlShehbaz, 2007) (Bailey et al. 2006). They are all almost exclusively North American
and have a base chromosome number of x=7 (Dobeš et al., 2006, Warwick et al.,
2006) and live in a wide range of habitats reaching from desert and shrub steppe to
montane meadow and habitats above timberline.
.With few exceptions members of the tribe Boechereae are characterized by
branched trichomes and entire leaves. Most of them are perennials with a welldeveloped basal rosette (Al-Shehbaz & al., 2006). The taxonomic history of
Boechera has been eventful. For a long time the taxa were treated as members of
the genus Arabis. Arabis holboellii was described by Hornemann in 1827
(Hornemann, 1827) and later transferred into the genus Boechera which was
established in 1976 with Boechera holboellii as type species (Löve and Löve, 1976).
However, when North American Arabis species were revised, such as Arabis
retrofracta, Graham, they were described as varieties of Boechera holboellii which
was then called Arabis again, in this case A. holboellii (Hornemann) A. Löve & D.
Löve var. retrofracta (Graham) Rydberg. Most detailed descriptions of the species
were available from Rollins (1993) who treated today’s Boechera as a separate
series of Arabis (Rollins, 1941). Analyses based on molecular data finally showed
that Arabis was polyphyletic (Koch et al. 1999, 2000) and subsequently taxa related
to Arabis drummondii (sensu Rollins, 1993) and Arabis holboellii (sensu Rollins,
1993) were transferred into Boechera (Al-Shehbaz, 2003). Boechera holboellii (sensu
Rollins 1993) was a highly polymorphic taxon with a number of varieties with
chromosome numbers of 2n=14 and 21 and a range of aneuploids (Rollins, 1941;
Sharbel and Mitchell-Olds, 2001), and as in other taxa diploids and triploids were
lumped into single species (Rollins, 1993). This view changed and in the forthcoming
treatment of FNA (Flora North America) B. holboellii is restricted to Greenland (AlShehbaz, pers. comm.).
The formation of hybrids is a frequent phenomenon in Boechera, with homoploid
hybridisation leading to the formation of diploid hybrids which often show lack of
meiotic reduction of gametes and thus provide the first step in triploid formation
(Dobeš et al. 2006, 2007a, 2007b). Imbalanced chromosome numbers in triploids
preclude balanced meiosis and hence sexual reproduction, and thus triploid
Boechera are apomictic (asexual reproduction via seed). Interestingly, diploid
apomixis, which is virtually absent in the plant kingdom (e.g. in Paspalum rufum,
Siena et al. 2008), is also found in Boechera (Böcher 1951).
57
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Boechera was completely revised for the new flora of North America (Al-Shehbaz,
unpublished; Windham and Al-Shehbaz, 2006, 2007a, 2007b), and in this most
recent treatments of the genus apomictically and sexually reproducing taxa have
been separated resulting into 72 species which reproduce sexually and 38 species
characterized by apomixis.
In the past few years various attempts have been started to unravel also the
evolutionary history and systematics of more widely distributed members of
Boechera such as B. stricta (= B. drummondii) and B. holboelllii (Koch et al. 2003,
Dobeš et al. 2004a, 2004b). These studies were extended including other taxa
(Schranz et al. 2005) also employing a broader spectrum of molecular and
cytogenetic approaches (Song et al. 2006).
Studies based on chloroplast DNA gave insights into past events which shaped
haplotype distribution in Boechera such as post-glacial range expansion or past
range fragmentation (Dobeš et al. 2004a, 2004b). Further analyses including more
Boechera taxa were consistent with those findings and revealed a large amount of
haplotype sharing among species which was attributed either to the haplotype
differentiation predating speciation and incomplete lineage sorting and/or frequent
hybridisation (Dobeš et al. 2004a, Kiefer et al. submitted).
Based on the previously obtained insights into the evolution of the maternal lineage
(Dobeš et al. 2004a, Kiefer et al. submitted) we wanted to apply biparentally
inherited markers in order to get an insight into speciation processes. In our
analysis we employed nrDNA ITS1 and ITS2 as well as intron sequences from
genes neighbouring the ELF3 locus (At2g25920; according to the ancestral crucifer
karyotype from Schranz et al. 2007 this locus should be in genome block h which in
Boechera stricta is on chromosome 5) and neighbouring the VRN1 locus
(At3g18900, Boechera stricta chromosome 3; Schranz et al. 2007). VRN1 and ELF3
are both genes potentially involved in adaptation, so they may have an evolutionary
significance. VRN genes mediate vernalization, the process by which a long period
of cold induces a mitotically stable state that leads to accelerated flowering during
later development. VRN1 encodes a protein that binds DNA in vitro in a nonsequence-specific manner and functions in stable repression of the major target of
the vernalization pathway (Levy et al. 2002). ELF genes are involved in plant
circadian clocks (elf = early flowering) and are found with different variants (Hazen
et al. 2005).ELF3 encodes a circadian clock-regulated nuclear protein that functions
in an Arabidopsis PHYB signal transduction pathway (Liu et al. 2001).
We chose sequences from neighbouring genes because VRN1 and ELF3 are
members of gene families while the adjacent genes were present as single copies
only. That way we prevented the amplification of paralogues. In the following we refer
to those genes by their locus (ELF3 neighbouring gene = At2g25920 and At3g18900
= VRN1 neighbouring gene) since their proteins have unknown function.
58
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
By choosing on the one hand nrDNA ITS1 and ITS2 and on the other hand single
copy gene introns we obtained two entirely different marker systems. The ITS on the
one hand is known to have the possibility to undergo concerted evolution (reviewed
by Elder and Turner 1995). Concerted evolution means that over time the thousands
of copies of ITS are adapted. Since the ITS is biparentally inherited this means that
the ITS copies either adapt to the maternal type, to the paternal type or to a mixture
of both types. However, also both copies may be conserved (Feliner et al., 2004).
This characteristic of the ITS is of special interest in Boechera since a lot of hybrids
can be expected that could carry either one of the parental ITS types, a recombinant
ITS type or several ITS types. Indeed it was shown in the past that single Boechera
individuals carried up to eight different ITS types (Koch et al. 2003).
Since the ITS is inherited biparentally it may be interpreted as a marker of present or
past gene flow (Koch et al. 2005). Hence, biogeographic analyses of the ITS data
may show which areas were in contact in the past or where gene flow persists to the
present day. The pattern of ITS type distribution along with species identity may
suggest if range fragmentation and expansion had an effect in speciation.
Integrative analysis of the different marker systems along with biogeographic
analyses may help us understand how Boechera shaped its biodiversity.
MATERIALS AND METHODS
Plant material.—Leaf material was obtained from herbarium specimens from GH,
MO and DAO. Corresponding accession details are listed in supplementary table 1,
the geographic distribution is given in supplementary figure 1. In total we analysed
964 vouchers including outgroup specimens, but we did not obtain DNA sequence
information for all accessions and all loci (refer to supplementary table 1).
DNA extraction.—Total DNA was obtained from a 0.5x0.5 cm2 piece of dried leaf
tissue from single individuals. Extraction followed the CTAB method of Doyle &
Doyle (1987), but some modifications were applied, involving grinding of only a
0.5x0.5 cm2 piece of dry leaf tissue in 2ml tubes using a Retsch swing mill (MM
200), addition of two units of ribonuclease (RNAse A) to the resolved DNA, and
washing of the DNA pellet twice with 70% ethanol. DNA was finally dissolved in 5070µl Bidest or low TE-buffer (Tris-EDTA) for long-term storage.
PCR conditions.—PCR reactions were performed in a volume of 25 µl containing
1x GoTaq buffer (Promega, Madison, USA), 2 mM MgCl2, 5 pmol of each primer, 5
nmol dNTPs (1.25 nmol of each dNTP) and one unit Taq DNA polymerase (GoTaq,
Promega), and variable concentrations of template (50 to 400 ng) using a PTC-200
thermal cycler (MJ-Research). Thermal cycling started with a denaturation step at
59
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
95°C lasting three min; followed by 30 cycles each comprising 30 s denaturation at
95°C, 30 s annealing at 48°C and elongation at 72°C. Amplification ended with an
elongation phase at 72°C lasting 10 min, and a final hold at 4°C.
ITS1 and ITS2 were amplified as described in Dobeš & al. (2004b) with PCR
products spanning the complete ITS1 and ITS2 as well as the intervening 5.8s
rRNA gene. PCR products were checked on agarose gels (1% agarose in TAE).
As additional nuclear markers we chose intron sequences from genes flanking the
VRN1 and ELF3 loci. Primer sequences for the intron in the single copy locus
adjacent to VRN1 were forward GCACTTGACCATCTCTTCAGATAA and reverse
AGTCCTTCGACGCAAACTG. They are placed in the flanking coding regions and
were obtained from Th. Mitchell-Olds and Eric Schranz, Duke University, USA. The
intron in the gene neighbouring the ELF3 gene was amplified using the primer
sequences
forward
TTTGTTGTTGCATATGGTTGT
and
reverse
TGCTTTACATGACTTGCTCTTA also optimized and obtained from Th. MitchellOlds and Eric Schranz, Duke University, USA. PCR products were checked on
agarose gels (1% agarose in TAE). PCR products were all purified with the
NucleoFastKit (Macherey-Nagel, Germany).
Cycle Sequencing.—Cycle sequencing was done with the DYEnamic ET
Terminator Cycle Sequencing Kit (Amersham Biosciences) using the PCR primers
for the cycle sequencing reaction. Samples were resolved in 10 µl Loading Solution
and then run on a MegaBace 500 Sequencer.
Alignments and ITS type definition.—Forward and reverse sequences were
aligned, edited by hand and trimmed to a common length. New ITS-types were
named following the nomenclature as introduced previously Koch et al. (2003) or
were assigned to ITS-types already described and published by these authors
(Koch et al. 2003; Dobeš et al. 2004b). An ITS type was defined as any sequencing
differing from the other sequences by at least one mutation. Sequences with
ambiguous sites were also named as an own ITS type since parents could not be
determined in most cases by comparing to sequences without ambiguous sites and
no cloning was done.
The alignments used for the phylogenetic analysis were made manually using the
program GenDoc (Nicholas & Nicholas, 1997).
The alignment of the ITS sequences followed a previously published alignment
(Koch et al. 2003). However, new gaps were introduced when needed. The ITS
alignment was subdivided for the different types of analyses.
The alignments for the intron sequence were done by eye since sequence similarity
was very high and no doubtful positions were present.
60
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
ITS, AT2G25920 and AT3G18900 phylogenetic tree reconstructions.—
Phylogenetic reconstruction based on ITS was done twice with two different
alignments. In the first analysis the alignment was once kept as a whole including all
identified sequence types and also sequences with ambiguous sites. In the second
analysis sequences with ambiguous sites and sequences originating from potential
apomicts were excluded. We decided to follow this approach because ambiguous
sites may lead to a false position of the accession in the tree. Also apomicts are
assumed to be of hybrid origin and may therefore contain recombinant ITS types
which would also alter the structure of the tree since they cause conflict in the data.
The analysis of both ITS alignments including ITS1 and ITS2 as well as the 5.8s
rRNA gene was done in PAUP4.0beta (Swofford, 2002) running a parsimony
analysis. The maximum number of trees retained was limited to 10,000. For the
heuristic search sequences were added randomly in 1000 replicates during which
10 trees were saved in each replicate. Chuckscore was set to one. Gaps were
treated as missing. TBR was used as branch-swapping algorithm. Starting trees
were obtained via stepwise addition. All characters had equal weight.
For phylogenetic reconstructions based on the ITS dataset the alignment was
supplemented with sequences of closely related genera obtained from Genebank
as well as from our own dataset to determine the relationship of Boechera to other
genera of the tribe Boechereae (AF146515, AF146514 = Cusickiella douglasii,
DQ452059 = Anelsonia eurycarpa, DQ452061 = Nevada holmgrenii, DQ452066 =
Cusickiella quadricostata and AJ232927, AJ232926 = Halimolobus perplexa var.
lemhiensis (hereafter Sandbergia), AY230615 = Polyctenium williamsiae, and
AF183109 = Polyctenium fremontii). Capsella rubella (AJ232913) served as
appropriate outgroup.
Bootstrap analysis was also run in PAUP4.0beta with the same settings as the
parsimony analysis in 1000 replicates.
Phylogenetic reconstruction based on the AT2G25920 intron data was done in the
same way as with the ITS dataset with the exception that only Polyctenium fremontii
and Sandbergia perplexa were used as representatives of other Boechereae.
Polyctenium was used as outgroup.
For the AT3G18900 intron a Bayesian analysis was run using MrBayes through the
program TOPALi (Milne et al. 2004). The SYM model was selected according to the
model selection option in TOPALi, burn-in period was set to 25%, number of
generations was set to 100,000 and number of runs was two. In this analysis
Cusickiella douglasii and Sandbergia perplexa were included as additional species
from the Boechereae. Cusickiella were used as outgroup.
In addition to the separate analyses a combined analysis including only accessions
for which at least one single copy gene intron was sequenced was done using the
same parsimony settings in PAUP 4.0beta as described above.
61
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
ITS network reconstructions.— Network reconstruction was done by running the
program TCS1.21 (Clement & al. 2000) with the reduced alignment including only
sequences without ambiguous sites or from individuals which were most likely
apomicts according to the revision of the genus for the new Flora of North America
(Al-Shehbaz, unpublished). Gaps were included as 5th state , and the connection
limit was set to 95%. An initial analysis on this alignment resulted into a network
with numerous circular connections. Therefore, the analysis was run on subsets of
the alignment representing significantly supported subgroups recovered in the
parsimony analysis.
Sequence analysis of subgroups.— In order to get an insight into the divergence
of sequences within lineages, the original alignment was divided into subgroups
according to the phylogenetic tree based on the ITS dataset. The analysis of the
alignment was also done by using TOPALi with the display summary option (Milne
et al. 2004).
Geographical analysis.—All information from the herbarium vouchers was entered
into the database BioOffice (Biogis Consulting, Version 2.0.4). Missing geographical
coordinates were added according to the descriptions on the herbarium vouchers.
Using BioOffice haplotypes were plotted on North America maps included in the
ArcView Package Version 8.
RESULTS
ITS-type definition.— In total 289 ITS types were detected among the 964
accessions and at least 63 species (only counting the ones which could without any
doubt be assigned to a taxon defined by Al-Shehbaz, submitted) including ITS types
characterized by ambiguous sites. Of those 39 were shared by several species, 30
were specific to one species and 220 were singletons. Of the 39 ITS-types shared by
several species 21 were dominated by one taxon.
57 sequences contained ambiguous sites in forward and reverse sequence and were
therefore assumed to be of hybrid origin; only three of them occurred twice, all others
were singletons.
All genebank accession numbers, corresponding ITS-types, and accessions which
share the ITS type and the ecoregion in which they were collected as well as cpDNA
haplotype identity are shown in supplementary table 2. Colour codes indicate if the
ITS type occurred once (yellow), if it was species specific (orange) or if it was shared
by several taxa (red). Bold letters indicate the presence of ambiguous codes in the
sequence.
62
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Alignment analyses.— The ITS alignment including all sequence types detected in
this study had a total length of 725 bp. 362 characters were constant, 154 variable
characters parsimony-uninformative and 209 characters were parsimony informative
(supplementary table 3). In total the alignment included 350 ITS sequence types
representing Boechera accessions and 29 accessions representing the outgroup and
other Boechereae.
The alignment which only included sequences without ambiguous sites and
accessions which were most likely not apomictic (according to the
taxonomic/cytological descriptions published by Windham and Al-Shehbaz, 2006,
2007a, 2007b; Dobeš et al. 2006) included only 168 ITS sequence types and 29
sequencing representing other Boechereae and the outgroup. The alignment had a
total length of 721 base pairs including gaps (supplementary table 4). 439 of the
characters were constant, and of the variable characters 143 were parsimonyuninformative and 139 were parsimony informative.
AT2G25920 was only sequenced for 210 accessions representing Boechera species
and five accessions representing Sandbergia and Polyctenium. The total length of
the AT2G25920 alignment was 564 bp. Six indels were coded separately and added
as an AT-matrix to the alignment giving it a new total length of 570 bp
(supplementary table 5). 331 characters were constant, 79 variable characters were
parsimony-uninformative while 154 characters were parsimony-informative (with gap
coding 160).
AT3G18900 was only sequenced for 111 accessions representing Boechera and four
accessions representing Sandbergia and Cusickiella. The AT3G18900 alignment had
a total length of 631 characters. Four gaps were coded in an AT matrix increasing the
alignment length to 635 bp (supplementary table 6). 494 characters were found to be
constant (including gap coding 495), 55 of the variable characters were parsimonyuninformative while 82 (including gap coding 86) were parsimony informative.
Phylogenetic reconstruction based on ITS.— The phylogenetic reconstruction
based on ITS separated genera of the Boechereae as well as Boechera species into
several major lineages (figure 1). However the relationship between the genera and
within Boechera remained unresolved, apart from the bulk of Cusickiella accessions
being sister to all other genera included in this study. Eastern North American
Boechera, Anelsonia, Nevada, Polyctenium, Sandbergia, one Cusickiella accession
and also Boechera repanda were represented by lineages arising from a polytomy
together with a large of group of Boechera species centred in western North America.
Within the large Boechera group 18 lineages containing more than one ITS-type
arose from the polytomy. 44 ITS-types were not assigned to any lineage but were
found separate on the polytomy. Some of the recovered lineages were speciesspecific while other lineages were dominated by one species and again others
contained a multitude of species.
63
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
The most frequent ITS-types were generally shared by several species, though
sometimes dominated by one or two taxa.
Taxa dominating in one group of ITS types or a single ITS-type were Boechera
puberula, Boechera subpinnatifida, Boechera sparsiflora, Boechera breweri,
Boechera koehleri, Boechera stricta, Boechera pendulocarpa, Boechera lemmonii,
Boechera retrofracta, Boechera microphylla, Boechera perennans, Boechera
pendulina, Boechera demissa or oxylobula, Boechera cobrensis, Boechera
platysperma, Boechera pulchra, Boechera glaucovalvula, Boechera rectissima,
Boechera constancei and Boechera davidsonii.
Fig.1 Parsimony analysis of the ITS dataset reduced to sequences without ambiguous
codes and originating from accessions which were most likely not apomictic. Capsella
was used as outgroup, other genera of the Boechereae, Boechera
missouriensis/laevigata, Boechera canadensis and Boechera repanda are in the left
lower corner, ITS types from Boechera species centred in western North America are
sorted alphabetically by the name of the ITS type. Taxa for which it was not clear if
they were apomictic or sexual are indicated in red
64
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
The phylogenetic reconstruction including all ITS-sequence detected in this study is
included in the supplementary material on the DVD accompanying the thesis.
Phylogenetic Reconstruction based on AT2G25920.— In the phylogenetic
reconstruction based on the AT2G25920 intron Boechera laevigata, Boechera
missouriensis, Boechera canadensis and Boechera repanda were found on a
polytomy together with a large group containing Sandbergia and Boechera species
centred in western North America. Sandbergia was sister to the large Boechera
group. A small group containing accessions assigned to Boechera stricta, Boechera
lemmonii, Boechera puberula and Boechera lasiocarpa was placed in sister position
to the remaining Boechera accessions (figure 2).
The resolution among taxa within the western North American Boechera group was
poor. In total 19 lineages containing more than one accession were found. 76
additional AT2G25920 sequence types were unresolved on the polytomy. As in the
phylogenetic reconstruction based on the ITS some species-specific lineages were
recovered while other lineages represented several taxa. Taxa representing own
lineages were Boechera microphylla, Boechera lemmonii, Boechera cobrensis,
Boechera glaucovalvula, Boechera schistacea, Boechera stricta, Boechera
suffrutescens (shared with one individual of Boechera constacei and Boechera
koehleri), Boechera rectissima (together with one Boechera arcuata accession) and
Boechera crandallii.
Phylogenetic reconstruction based on AT3G18900.— The phylogenetic
reconstruction based on the AT3G18900 intron placed Boechera repanda as sister to
the remaining Boechera accessions and Sandbergia. The Boechera accessions were
split into two groups, Sandbergia being sister to one of the groups (figure 3).
Relationships in the group being sister to Sandbergia remained unresolved. On the
polytomy a specific lineage for Boechera lemmonii, Boechera lyallii and Boechera
stricta was found. Another lineage contained accessions representing Boechera
koehleri, Boechera breweri, Boechera sparsiflora but also Boechera microphylla and
Boechera microphylla var. macounii. AT3G18900 intron types specific to Boechera
crandallii and Boechera cobrensis were exclusively found within the Boechera
lineage being sister to Sandbergia. However they were found to be unresolved on the
polytomy among accessions representing B. perennans, B. pallidifolia, B. williamsii,
B. rectissima, B. selbyi, B. lignifera, B. pallidifolia and B. schistacea.
Boechera missouriensis and Boechera laevigata were sister to the second clade
recovered by the AT3G18900 based phylogenetic reconstruction. Boechera
canadensis was sister to the remaining Boechera accessions in the second clade.
Boechera rigidissima, Boechera constancei and Boechera suffrutescens formed a
sister clade to a polytomy from which five lineages arose. The first lineage was a
single Boechera macounii accession, the second lineage contained Boechera pendu65
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Figure 2 Parsimony analysis of the AT2G25920 intron. Polyctenium was used as
outgroup. The colours of the branches indicate the lineage of the cpDNA haplotype
that was found in this accession (Kiefer et al. submitted); lineage I blue, lineage II
yellow, lineage III red, lineage IV pink, lineage V purple, central haplotype and directly
derived grey
66
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Figure 3 Bayesian analysis of the AT3G18900 intron run through the program TOPALi
(Milne et al. 2004). Cusickiella was used as outgroup. The colours of the branches
indicate the lineage of the cpDNA haplotype that was found in this accession (Kiefer et
al. submitted); lineage I blue, lineage II yellow, lineage III red, lineage IV pink, lineage V
purple, central haplotype and directly derived grey. Numbers indicate posterior
probabilities
lina (six accessions), Boechera fendleri (five) and Boechera perennans (two
accessions).
The third lineage contained six accessions representing Boechera formosa, two
Boechera pulchra and one Boechera lincolnensis accession. The fourth lineage
67
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
contained one accession of Boechera inyoensis, Boechera fendleri, Boechera
shockleyi, and Boechera sparsiflora. The fifth lineage contained two Boechera
pygmaea accessions.
Combined analysis of ITS, AT3G18900 intron and AT2G25920 intron.— The
combined analysis resulted in a completely unresolved polytomy (data not shown).
Network analysis of the ITS dataset.— The network analysis was run using the
same alignment as for the phylogenetic reconstruction based on ITS data with the
non-Boechera taxa and eastern North American Boechera taxa being omitted. This
analysis resulted into a network with a mixture of clearly separated lineages (ab/bv,
er/ly/lz, ks/hl/gy/kr, pq/la/lb/lc/po/pu, rb/pp arising from ITS type ad; pg/ph/pi/lk and
na/nb/nc arising from ie which is connected to ad via ee; ar/gx/pf/sa/kn/mm/ha/mv
connected to h; nh/ky/su/rl/ld connected to h) and reticulate structures (e.g.
op/ml/mn/mo/ie/hm/ee connected to ad) (figure 4). Especially the ITS types ac, v, g
and f all themselves being origin for separate lineages and connected to h could not
be resolved. Most of the reticulate pattern was caused by repeated mutations in
alignment position 216 and 444. Hence, the alignment was subdivided into sets of
ITS-types according to the phylogenetic reconstruction based on ITS sequences. ITS
type ac and v gave both rise to four separate lineages and were connected to h by
two mutation steps (figure 5a and 5b). ITS type g was connected to ITS type h by
one mutation step and was origin for four lineages. ITS type bt which was connected
to g by one mutation step was also connected to a series of missing ITS types which
connected a group of four ITS types to the network (z/mp/le and og/sx). Those two
groups of ITS types were also connected to ITS type h by several missing ITS types
(figure 5c). Finally ITS-type f was connected to ITS type h by two mutation steps and
gave rise to 11 lineages (figure 5d).
Geographic distribution of subgroups.— Distribution ranges of most ITS-types
overlapped. However, for the most frequent ITS-types a centre of distribution could
be determined (Table 1). ac, au, eu, f, g, gx and op had had their major occurrence in
the Great Basin and on the Snake/Columbia Plateau. ab, bt, l and t had their
distribution centre in the Rocky Mountains. c and r occurred to the same extent in the
Great Basin or Snake/Columbia Plateau and the Rocky Mountains. az was found in
the Rocky Mountains, the Cascades and in general north of the last glaciation. e was
exclusively found north of the last glaciation. ev was typically found in the deserts as
well as on the Colorado Plateau. z had its centre of distribution in the Sierra Nevada.
h and ad were both central to the network analysis and both had several lineages
arising from them. h had a continuous distribution range but it occurred mostly in the
Great Basin, the Rocky Mountains and also north of the last glaciation. ad had a
disjunct distribution range and was mainly present in the western Great Basin (mainly
68
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Figure 4 Parsimony network analysis of the analysis of the alignment reduced to
sequences without ambiguous codes except N and sequences originating only from
accessions which were most likely not apomictic according to their taxonomic
determination; only the resolved section of the network is shown. Species names in
the squares indicate the taxa which predominantly shared the ITS types included in
the squares
Boechera cobrensis) and on the Colorado Plateau (mainly Boechera pallidifolia and
Boechera selbyi = Boechera gracilenta).
Three of the subgroups from the network analysis were analysed in greater
geographic detail. ITS type ac was central to the first group (figure 6a). It had a
disjunct distribution range and was found in Klamath/Siskiyou (Boechera
subpinnatifida and Boechera koehleri), the western Great Basin (Boechera
69
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
retrofracta and Boechera pinetorum), the northern Wasatch/Uinta Range (Boechera
lignifera and Boechera retrofracta) and the southern central Rockies (Boechera
retrofracta). Derived from it were ITS types eu and ru in Klamath Siskiyou (Boechera
subpinnatifida) and in the western Great Basin and on the western Snake/Columbia
Plateau (Boechera puberula, Boechera pulchra, Boechera retrofracta). ITS type au
was also derived from ITS type ac and occurred in the western half of the Great
Basin (Boechera puberula and others). ITS type oh also derived from ITS type ac
occurred in the northern Wasatch/Uinta range (Boechera lasiocarpa).
ITS-type g was derived from ITS type h by one mutation step. It was mainly found in
Boechera pendulina individuals in the central Great Basin and the Wasatch Uinta
Range (figure 6b). However, moving further southeast into the southern
Wasatch/Uinta Range, on the Colorado Plateau and the Arizona Mountains, ITS type
g was detected in Boechera fendleri individuals. The lineage containing ITS types is
and ke was derived from ITS type g. Those ITS types were restricted to Boechera
fendleri in the Mojave Desert. ITS types kd, rd and gz were also derived from ITS
type g. They had a circular connection to ITS type g, so it is unclear if they all belong
to the same lineage or if they are individually derived from ITS type g. They were
found in Boechera fendleri (one individual), Boechera pendulina (one individual) and
also in three other species (three individuals) on the northern Colorado Plateau. ITS
type ho was derived from ITS type g and found in the Chihuahua Desert in one
individual of Boechera fendleri. Another larger lineage arising from ITS type g was
dominated by Boechera pendulocarpa individuals (bt, bw, bx, bu, eq). Those ITS
types had a scattered distribution in the western and north eastern Great Basin,
Sierra Nevada, Klamath/Siskiyou, the Colorado Rockies, Okanogan Forest, Eastern
Cascades, Montana Valley Foothill Grasslands and Wasatch/Uinta Range with a
focus in the southern central Rockies.
ITS type gx was derived from ITS type h by two mutation steps and was origin to a
lineage of mainly Boechera sparsiflora, Boechera breweri and Boechera koehleri
individuals (figure 6c). ITS type gx was dominated by Boechera sparsiflora
individuals. It occurred along the western border of the Great Basin, reaching into the
Northern Sierra Nevada and the northwestern Snake/Columbia Plateau. In one
accession of Boechera breweri it occurred also in Klamath/Siskiyou. ITS types mm,
ha and mv as well as ITS types kn and na, and pf made up three lineages derived
from ITS type gx. They all represented a mixture of Boechera breweri and Boechera
koehleri individuals (along with one Boechera retrofracta accession) in
Klamath/Siskiyou and the Northern Sierra Nevada.
70
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Figure 5 Parsimony network analysis based on ITS sequences; subgroups were
defined by the groups recovered in the parsimony analysis. ITS type h was added to
all analysis for rooting; A ITS type ac and derived lineages, B Boechera stricta specific
lineage, C ITS type g and derived lineages with additional ITS types that have a
circular connection to a lineage arising from ITS type g, D ITS type f and derived
lineages; species names in the squares indicate the taxa which predominantly shared
the ITS types included in the squares
71
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Figure 6 Geographical distribution ranges of selected ITS lineages; in each figure a
simplified version of the network section is given and colour coded according to the
distribution ranges indicated in the map. Next to each distribution range a list of the
species sharing the ITS type in the encircled region is given. The number in brackets
indicates the number of individuals; A distribution range of ITS type ac and derived
lineages, B distribution range of ITS type g and derived lineages, C distribution range
of ITS type gx and derived lineages
Divergence of sequences within subgroups.— Divergence of sequences within
subgroups defined by the phylogenetic reconstruction based on the ITS dataset was
calculated as average p-distances (Table 2). The core Boechera group (all Boechera
species included in this study centred in western North America had an average
divergence of 0.011. All subgroups had an average divergence between 0,002 and
0.005.
Assuming a mutation rate for the ITS of approximately 0.5% to 2.5% nucleotide
divergence per 1 million years (for comparison: highest substitution rates with 5.3 3
10–9 substitutions/site/year, Wendel, Schnabel, and Seelanan 1995; lowest rates
with 0.35 3 1029 substitutions/site/year, Suh et al. 1993) a p-distance of 0.01
correspond to 1 million years (as a mean). It is obvious that this calculation is highly
biased, and it is used herein simply as a very rough indicator.
Comparison of cpDNA-types, ITS types and single copy allele identity.—
Phylogenetic reconstruction based on nrDNA ITS, AT2G25920 and AT3G18900
revealed lineages specific to some species or a group of species as well as putative
hybrids which showed up among specific lineages. ITS-types and intron sequence
types as well as their position in network and phylogenetic reconstruction were
compared for additional information on the nature of a specific lineage or hybrid. For
information of the maternal side cpDNA haplotypes as defined earlier (Kiefer et al.
submitted) were also included in the comparison (supplementary table 7).
The comparison revealed that some lineages were uniform and recovered at least in
the phylogenetic reconstruction based on two of the marker sequences while other
species had a complex mixture of sequence types from different lineages.
72
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Table 1 Distribution ranges of major ITS types; numbers indicate the number of
individuals carrying this ITS type in the named region. Bold numbers indicate the
centres of distribution defined by number of individuals.
ITS
type
Klamath
Siskiyou
ab
ac
ad
ar
au
az
bt
bx
c
e
eu
ev
f
g
gx
h
i
l
op
r
t
v
z
Great
Basin
Snake
3
1
1
1
6
2
1
6
12
2
13
1
3
6
8
4
8
6
9
30
1
8
7
3
-
Sierra
Nevada
1
2
1
1
1
2
8
6
1
2
6
Deserts
Rocky
Mtns.
9
3
2
-
4
3
1
6
12
2
5
2
36
25
5
8
4
1
Cascades
N of
glaciation
Wasatch
Uinta
1
5
1
38
25
2
-
3
4
1
3
1
4
5
1
2
2
-
3
1
7
1
1
2
7
3
-
Colorado
Plateau
12
2
8
4
3
2
1
-
other
1
2
1
3
2
1
10
1
1
1
Table 2 Divergence time estimates of subgroups calculated from average sequence
divergence calculated through the program TOPALi (Milne et al. 2004). The first
column gives the most frequent ITS type of a group, the second column gives the
taxa dominating this lineage, the third column gives the average pairwise distances
and the fourth column gives the age estimate assuming an average divergence of 1%
per 1 million years (Koch et al. 2003).
phylogroup
predominant taxa
all core Boechera
ac group
az group
bt group
ev group (network)
f group (network)
gx group
B. puberula
B. stricta
B. pendulocarpa
B. perennans
B. sparsiflora
B. breweri
B. koehleri
B. platysperma
B. rigidissima
B. pulchra
B. arcuata
B. sparsiflora
B. rectissima
B. shockleyi
B. inyoensis
kv group
ky group
le group
ml group
73
average pairwise
distance
0.0111
0.0045
0,0042
0,0035
0,0029
0,0051
0,0035
age estimate
1 my
450,000 y
420,000 y
350,000 y
290,000 y
510,000 y
350,000 y
0,0083
830,000 y
0,0044
0,0050
440,000 y
500,000 y
0,0031
310,000 y
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
Boechera cobrensis shared ITS type ad and ITS-types derived from it. In the
phylogenetic reconstruction based on the AT2G25920 intron the accessions also
clustered in one lineage. AT3G18900 was sequenced for three of the accessions and
put them together in one of the recovered lineages but in an unresolved relationship.
The cpDNA haplotype was exclusively haplotype ci, the most frequent cpDNA type in
the cpDNA analysis (Kiefer et al. submitted).
If Boechera fendleri shared ITS type g it was unresolved in the analysis based on
AT2G25920 intron but occurred in one lineage in the phylogenetic reconstruction
based on the AT3G18900 intron. The accession which shared ITS type f was also
unresolved in the analysis based on the AT2G25920 intron. The accession which
carried ITS type sn showed up in a group of mixed species which was recovered in
the phylogenetic reconstruction based on the AT2G25920 intron as well as on the
AT3G18900 intron. This individual also carried a haplotype from the evolutionary
chloroplast lineage II (Kiefer et al. submitted). Boechera formosa was also an
example for a simple lineage. The ITS types were either ad or derived from ad, in
AT2G25920 the relationship of the sequences representing these accessions was
unresolved and in the phylogenetic reconstruction based on AT3G18900 they
constituted a separate lineage within clade 2.
Boechera lemmonii was interesting in so far that several individuals were found to be
in a separate lineage in the phylogenetic reconstructions based on either one of the
employed markers. However, two accessions carried cpDNA haplotype from lineage
I unlike the other accessions which all carried haplotypes from chloroplast lineage III
(Kiefer et al. submitted). One individual which carried a different ITS type showed up
in a lineage of mixed species in the phylogenetic reconstruction based on
AT2G25920. Three other individuals were found in one ITS-lineage but showed up in
different lineages in the phylogenetic reconstruction based on the AT2G25920 intron.
However, if the chloroplast DNA haplotype was known it was found to be in
chloroplast lineage III.
Boechera lignifera had a wild mixture of ITS types from different lineages, among
them sequences with ambiguous codes. Also in the phylogenetic reconstruction
based on the AT2G25920 intron and the AT3G18900 intron the accessions showed
up in different positions.
Boechera microphylla (in the sense of Rollins 1993) shared ITS type h mainly with
Boechera retrofracta. Commonly Boechera microphylla accessions shared
chloroplast haplotypes from lineage I. In the phylogenetic reconstruction based on
the AT2G25920 intron Boechera microphylla accessions were found in a separate
lineage except for two individuals which showed up in a lineage of mixed species or
unresolved. The individual carrying the AT2G25920 allele whose relationship was
unresolved carried a chloroplast haplotype from lineage III. The individual whose
AT2G25920 allele was found in the mixed lineage showed up in an unresolved
position in clade 2 of the phylogenetic reconstruction based on the AT3G18900 intron
74
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
while other accessions which carried ITS type h and were in a separate lineage
based on the AT2G25920 intron were found in clade on in a lineage with Boechera
sparsiflora, Boechera breweri and Boechera koehleri in clade 1.
Boechera pallidifolia consistently showed up together with Boechera selbyi
(according to Al-Shehbaz unpublished = Boechera gracilenta).
Boechera suffrutescens showed up consistently with Boechera constacei also when
the chloroplast DNA haplotype was from different lineages.
Boechera perennans, Boechera pendulina and Boechera fendleri had a complex
relationship with each other. Typical for Boechera pendulina and Boechera fendleri
was ITS type g. However it also carried ITS type ev in a small region on the western
Colorado Plateau. Boechera perennans mainly carried ITS type ev. Boechera
pendulina mainly carried chloroplast lineage I haplotypes while Boechera perennans
typically carried lineage III haplotypes. However, Boechera perennans, Boechera
pendulina and Boechera fendleri showed up together in one lineage in clade 2 in the
analysis of AT3G18900 data.
DISCUSSION
We based our phylogenetic reconstructions on three different marker systems all with
different properties or locations: (a) nrDNA ITS which is present in the genome in
thousands of copies and is subject to concerted evolution, a fact which may cause
trouble in defining true lineages, (b) an intron of AT3G18900, a single copy gene on
Boechera stricta chromosome 3 not undergoing concerted evolution, and (c) an
intron of AT2G25920, a single copy gene on Boechera stricta chromosome 5 not
undergoing concerted evolution. Neither the analysis of ITS1 and ITS2 nor the
analysis of AT2G25920 resulted into a well resolved tree. The analysis of
AT3G18900 resulted into two clades which were not or only little resolved.
An earlier study of nrDNA in Boechera including Boechera stricta, Boechera holboellii
sensu Rollins (Rollins 1993) and Boechera divaricarpa sensu Rollins (Rollins 1993)
had already shown that there was no resolution among ITS types and only few ITS
types were joined in separate groups (Dobeš et al. 2004b). Unresolved trees are a
common phenomenon, usually regarded as experimental failure (Rokas and Carroll
2006). However, the lack of resolution indicates that the time span between
branching events was extremely small which is a useful information. If a polytomy is
caused by truly simultaneous cladogenesis it is referred to as a hard polytomy while
a polytomy caused by too few characters and superimposed substitutions it is called
a soft polytomy (Maddison, 1989). Since the polytomy was recovered for ITS and
AT2G25920 in parallel and also in the phylogenetic reconstruction based on
AT3G18900 internal branches are shorter than terminal branches the polytomy
detected in our dataset can be regarded as a hard polytomy showing the rapid
75
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
cladogenesis in Boechera. Many of the lineages we recovered were species specific.
Therefore, the shape of the trees cannot only be interpreted as rapid cladogenesis
but indeed as rapid speciation.
Every phylogenetic reconstruction based on sequence data does not necessarily
represent the phylogenetic history of the species. It only shows relationships among
the different alleles of the piece of DNA under investigation. However, if several DNA
sequences support the same tree topology it is likely sequence evolution and species
evolution are congruent in this case. With so many unresolved lineages it is hard to
find congruent topologies. Nevertheless we found lineages which were recovered two
or all of the phylogenetic reconstructions. However, often in one dataset a lineage
was resolved while it was unresolved in the other. If a “good lineage” could be
identified and only few individuals showed up among other lineages or had
ambiguous sites in their ITS sequence these were interpreted as hybrids. That way
we could identify hybrids for example in Boechera fendleri, Boechera lemmonii,
Boechera microphylla, Boechera perennans, and Boechera pendulina.
The nrDNA ITS dataset was also subjected to a network analysis. The network was
found to have two central ITS types, ad and h. From both centres several species
specific or mixed lineages originated. The lineages originating from ITS type ad were
mostly without conflicting data and circular connections were rare. On the other hand
connections to lineages originating from ITS type h were often in conflict with each
other. These conflicts were mainly due to two alignment positions, namely 216 and
444. Boechera stricta typically had the combination A in 216 and T in 444. The
majority of accessions had the combination G in 216 and T in 444. This means that
position 216 can be used to tell apart the Boechera stricta lineage from the other
lineages. Hence it follows that a group with A in position 216 is derived from the
Boechera stricta lineage, such as the lineage including ITS type f. On the other hand
a group with G in position 216 is derived from the majority, such as the lineage
including ITS type g. The lineages including ITS types f and g both have G in position
444. There are three possibilities which may explain this. First of all a mutation might
have occurred in parallel in position 444 changing T into G in parallel. The second
possibility is that in the lineage including most accessions the mutation in position
444 happened first giving rise to ITS type g. Later hybridisation with a Boechera
stricta individual and concerted evolution may have produced ITS type f. Both
possibilities are equally parsimonious because they involve two steps. However,
Boechera stricta is also characterized by a mutation in position 161 which is neither
present in ITS type f or g or any of the ITS types derived from them. Hence, it is most
likely that the mutation in position 444 occurred in parallel in two lineages.
76
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
ITS lineages in which the central ITS type was shared by several species were also
analysed geographically. ITS type ac was such a case. It showed a fragmented
distribution in Klamath/Siskiyou, the western Great Basin, the northern
Wasatch/Uinta Range and with one accession in the southern central Rockies.
Interestingly ITS type ac was mainly found in Boechera retrofracta individuals. Two
lineages dominated by Boechera puberula split from ITS type ac in the western Great
Basin and on the Snake/Columbia Plateau co-occurring with ITS type ac. In
Klamath/Siskiyou ITS type ac was shared by Boechera subpinnatifida (two
accessions) and Boechera koehleri (one accession). Boechera subpinnatifida also
carried ITS type ru derived from one of the Boechera puberula types in
Klamath/Siskiyou. It seems possible that either the distribution range of ac and eu
used to reach Klamath/Siskiyou in the past and after range diffraction Boechera
subpinnatifida evolved in this ecoregion. However, it could also be that the
distribution range of ac was never larger and that ITS types ac and eu were carried
into Klamath/Siskiyou by the widespread Boechera retrofracta. Since we do not have
information about the speciation process itself we cannot rule out one of the
possibilities.
Another interesting example is ITS type g and its derived lineages. ITS type g is
typical for Boechera pendulina. It occurs in the central Great Basin, the
Wasatch/Uinta Range, on the Colorado Plateau and in the Arizona Mountains.
Remarkably there seems to be a boundary to the south beyond which ITS type g is
found in Boechera fendleri. Additionally from three other separate lineages
originating from ITS type g two are specific to Boechera fendleri and only in the
southern range. The third lineage includes a mix of Boechera fendleri, Boechera
pendulina and other species. North of the distribution range of ITS type g lies the
distribution range of a lineage dominated by Boechera pendulocarpa. So unlike in the
lineage including ITS type ac it is not range fragmentation which may have played a
role in speciation but rather range expansion to the north and south.
Last but not least ITS type gx, also derived from ITS type h, is a second example for
a lineage where species differentiated in the Klamath/Siskiyou ecoregion. ITS type gx
is dominated by Boechera sparsiflora and is distributed along the most western part
of the Great Basin reaching into the Sierra Nevada and into Klamath/Siskiyou with
one Boechera breweri accession. The three lineages derived from ITS type gx occur
only in the Klamath/Siskiyou and the Sierra Nevada and with one exception are
shared between Boechera koehleri and Boechera breweri.
Since ITS type gx is derived from ITS type h and ITS type h is absent from
Klamath/Siskiyou it seems likely that first Boechera sparsiflora differentiated in the
western Great Basin from the ancestor carrying ITS type h followed by migration or
more likely dispersal into Klamath/Siskiyou where Boechera koehleri and Boechera
breweri evolved. Alternatively the distribution range of Boechera breweri and
Boechera koehleri was larger in the past reaching into the western Great Basin. Then
77
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
glaciation of the Sierra Nevada and Southern Cascades separated the unglaciated
Klamath Siskiyou from the unglaciated Great Basin and lead to the extinction of the
two species in this range.
Average sequence divergence of ITS types allowed for a rough age estimate. Taking
the average value of 1% sequence divergence per one million years (Koch et al.
2003) the age of ITS lineages representing species centred in western North America
is one 1my. The age of the identified lineages is between 500,000 and 200,000
years. This time falls entirely into the second half of the quaternary. During this time
at least three glaciation cycles took place (Bintanja and van der Wal, 2008). The
quaternary ice ages forced plant and animal taxa to migrate into and out of glacial
refugia (Hewitt, 2000). Changing climate promoted the formation of new habitats
which supported speciation (Hewitt, 2004, Willis and Niklas, 2004). Regarding the
time frame in which the ITS lineages developed makes it likely that Boechera’s
diversity as we find it today is also a result of those climatic oscillations.
SUMMARY
The phylogeny of Boechera could neither be resolved by nrDNA ITS nor by
sequencing two introns from single copy genes. Species specific lineages could be
detected but it cannot be determined which lineages are sister to each other.
Interpreting the structure of the trees with either unresolved lineages in the case of
ITS and the AT2G25920 intron or short internal branches as in the case of
AT3G18900 indicates rapid cladogenesis and in this case rapid speciation.
Even if Boechera underwent radiation the split of lineages was successive. To
unravel those quick processes is not possible with sequencing marker sequences. A
continental wide population genetics approach might be the most promising
approach.
LITERATURE CITED
AL-SHEHBAZ, I. A. (submitted) Boechera A. Löve and D. Löve. In: Flora North
America.
AL-SHEHBAZ, I. A. 2003. Transfer of most North American species of Arabis to
Boechera (Brassicaceae). Novon 13: 381-391.
78
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
AL-SHEHBAZ I. A., BEILSTEIN, M. A., AND E. A. KELLOGG. 2006. Systematics
and phylogeny of the Brassicaceae (Cruciferae): an overview. Plant Systematics
and Evolution 259: 89–120.
AL-SHEHBAZ, I. A. 2007. The North American genus Sandbergia (Boechereae,
Brassicaceae). Harvard Papers of Botany 12: 425–427.
BAILEY, C. D., KOCH, M. A., MAYER, M., MUMMENHOFF, K., O’KANE, S. L.,
WARWICK, S. I., WINDHAM, M. D., AND I. A. AL-SHEHBAZ. 2006. A Global
nrDNA ITS Phylogeny of the Brassicaceae. Molecular Biology and Evolution 23:
2142–2160.
BINTANJA, R., AND R. S. W. VAN DER WAL 2008. North American Ice Sheet
Dynamics and the onset of 100,000 year glacial cycles. Nature 454: 869–872.
BÖCHER, T. W. 1951. Cytological and embryological studies in the amphiapomictic Arabis holboellii complex. Kong Danske Vidensk Selsk, 6: 1–59.
CLEMENT, M., POSADA, D., AND K. A. CRANDALL. 2000. TCS: a computer
program to estimate gene genealogies. Molecular Ecology 9: 1657–1659.
DOBEŠ, C., MITCHELL-OLDS, T. AND M. A. KOCH. 2004a. Extensive chloroplast
haplotype variation indicates Pleistocene hybridization and radiation of North
American Arabis drummondii, A. ×divaricarpa, and A. holboellii (Brassicaceae).
Molecular Ecology 13: 349–370.
DOBEŠ, C., MITCHELL-OLDS, T. AND M. A. KOCH. 2004b. Intraspecific
diversification in North American Boechera stricta (= Arabis drummondii), Boechera
x divaricarpa, and Boechera holboellii (Brassicaceae) inferred from nuclear and
chloroplast molecular markers—an integrative approach. American Journal of
Botany 91: 2087–2101.
DOBEŠ, C., KOCH, M.A., AND T. F. SHARBEL. 2006. Embryology, karyology, and
modes of reproduction in North American genus Boechera (Brassicaceae): a
compilation of seven decades of research. Annals of the Missouri Botanical
Gardens 93: 517–533.
DOBEŠ C., KOCH M., AND C. KIEFER. 2007. Apomixis and radiation at low ploidy
levels exemplified in the evolutionary model genus Boechera (Brassicaceae).
Regnum Vegetabile 147: 391–412.
79
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
DOBEŠ, C., SHARBEL, T. F, AND M. A. KOCH. 2007. Towards understanding the
dynamics of hybridization and apomixis in the evolution of genus Boechera
(Brassicaceae). Systematics and Biodiversity 5: 321–332.
DOYLE, J. J. AND J. L. DOYLE. 1987. A rapid DNA isolation procedure for small
amounts of fresh leaf tissue. Phytochemical Bulletin 19: 11–15.
ELDER J. F., AND B. J. TURNER. 1995. Concerted evolution of repetitive DNA
sequences in eukaryotes. Quarterly Reviews of Biology 70: 297–320.
FELINER G. N., LARENA B. G., AND J. F. AGUILAR. 2004. Fine-scale
geographical structure, intra-individual polymorphism and recombination in nuclear
ribosomal internal transcribed spacers in Armeria (Plumbaginaceae). Annals of
Botany 93: 189–200.
HAZEN, S. P., BOREVITZ, J. O., HARMON, F. G., PRUNEDA-PAZ, J. L., SCHULTZ,
T. F., YANOVSKY, M. J., LILJEGREN, S. J., ECKER, J. R., AND STEVE A. KAY
2005. Rapid Array Mapping of Circadian Clock and Developmental Mutations in
Arabidopsis. Plant Physiology 138: 990–997.
HEWITT, G. 2000. The genetic legacy of the Quaternary ice ages. Nature 405: 907–
913.
HEWITT, G.M. 2004. Genetic consequences of climatic oscillations in the
Quaternary. Philosophical Transaction of the Royal Society of London B. Biological
Sciences 359:183–95.
HORNEMANN, J. W. 1827. Flora Danica. 11 (32): 1879.
KIEFER, C., DOBEŠ, C., SHARBEL, T. F., M. KOCH. (submitted) Phylogeographic
structure of the chloroplast DNA gene pool in North American Boechera – a genus
and continental wide perspective.
KOCH, M., BISHOP, J. AND T. MITCHELL-OLDS. 1999. Molecular systematics and
evolution of Arabidopsis and Arabis. Plant Biology: 1: 529–537.
KOCH, M., HAUBOLD, B., AND T. MITHCHELL-OLDS. 2000. Comparative
evolutionary analysis of chalcone synthase and alcohol dehydrogenase loci in
Arabidopsis, Arabis and related genera. Molecular Biology and Evolution. 17: 1483–
1498.
80
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
KOCH, M., DOBEŠ, C., AND T. MITCHELL-OLDS. 2003. Multiple hybrid formation in
natural populations: Concerted evolution of the internal transcribed spacer of nuclear
ribosomal DNA (ITS) in North American Arabis divaricarpa (Brassicaceae). Molecular
Biology and Evolution 20, 338–350.
KOCH, M., KIEFER, C., VOGEL, J., EHRICH, D., BROCHMANN, C., AND K.
MUMMENHOFF. 2006. Three times out of Asia Minor - the phylogeography of Arabis
alpina L. (Brassicaceae). Molecular Ecology 15: 825–839.
LEVY, Y. Y., MESNAGE, S., MYLNE, J. S., GENDALL, A. R., AND C. DEAN. 2002.
Multiple Roles of Arabidopsis VRN1 in Vernalization and Flowering Time Control
Science: 297(5579): 243–246
LIU, X. L., COVINGTON, M. F., FANKHAUSER, C., CHORY, J., AND D. R.
WAGNER. 2001. ELF3 Encodes a Circadian Clock–Regulated Nuclear Protein That
Functions in an Arabidopsis PHYB Signal Transduction Pathway; The Plant Cell13:
1293–1304.
LÖVE, A. AND D. LÖVE. 1976. Botaniska Notiser 128: 513.
MADDISON, W. P. 1989. Reconstructing Character Evolution in polytomous
cladograms. Cladistics 5: 365–377.
MILNE, I., WRIGHT, F., ROWE, G., MARSHAL, D. F., HUSMEIER, D. AND G.
MCGUIRE. 2004. TOPALi: Software for Automatic Identification of Recombinant
Sequences within DNA Multiple Alignments, Bioinformatics 20: 1806–1807
ROKAS, A., AND B. C. CARROLL. 2006. Bushes in the tree of life. Public Library of
Sciences Biology 4: 1899–1904.
NICHOLAS, K.B. AND H. B. JR. NICHOLAS. 1997. GeneDoc: A tool for editing and
annotating multiple sequence alignments. Distributed by the author.
ROLLINS, R. C. 1941. A monographic study of Arabis in western North America.
Rhodora, 43: 289–325 348–411 425–481.
ROLLINS, R.C. 1993. The Cruciferae of Continental North America. Stanford
University Press, Stanford.
81
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
SCHRANZ, M. E., WINDSOR, A. J., SONG, B.-H., LAWTON-RAUH, A., AND T.
MITCHELL-OLDS. 2007. Comparative Genetic Mapping in Boechera stricta, a
Close Relative of Arabidopsis. Plant Physiology 144: 286–298.
SIENA, L., SARTOR, M., ESPINOZA, F., QUARIN, C., AND J. ORTIZ. 2008.
Genetic and embryological evidences of apomixis at the diploid level in paspalum
rufum support recurrent auto-polyploidization in the species. Sexual Plant
Reproduction 21: 205–215.
SHARBEL, T. F., AND T. MITCHELL-OLDS. 2001. Recurrent polyploid origins and
chloroplast phylogeography in the Arabis holboellii complex (Brassicaceae). Heredity
87: 59–68.
SONG, B. H., CLAUSS, M. J., PEPPER, A. AND T. MITCHELL-OLDS. 2006.
Geographic patterns of microsatellite variation in Boechera stricta, a close relative
of Arabidopsis. Molecular Ecology 15: 357–369
SUH, Y., THEIN, L. B., REEVE, H. E., AND E. A. ZIMMER. 1993. Molecular
evolution and phylogenetic implications of internal transcribed spacer sequences of
ribosomal DNA in Winteraceae. American Journal of Botany 80:1042–1055.
SWOFFORD, D.L. 2002. PAUP*: Phylogenetic Anaylsis Using Parsimony (*and
Other Methods), Version 4.0B10. Sinauer Associates, Sunderland, Massachusetts
WARWICK, S. I., FRANCIS, A., AND I. AL-SHEHBAZ. 2006. Brassicaceae species
checklist and database on CD-ROM. Plant Systematics and Evolution 259: 249–258.
WENDEL, J. F., A. SCHNABEL, AND T. SEELANAN. 1995. An unusual ribosomal
DNA sequence from Gossypium gossypioides reveals ancient, cryptic, intergenomic
introgression. Molecular Phylogenetics and Evolution 4:298–313.
WILLIS, K.J., AND K. J. NIKLAS. 2004. The role of quaternary environmental
change in plant macroevolution: the exception or the rule? Philosophical
Transactions of the Royal Society of London B. 359, 159–172
WINDHAM, M.D. AND I. A. AL-SHEHBAZ. 2006. New and noteworthy species of
Boechera I: sexual diploids. Harvard Papers of Botany 11: 61–88.
WINDHAM, M.D. AND I. A. AL-SHEHBAZ. 2007a. New and noteworthy species of
Boechera (Brassicaceae) II: apomictic hybrids. Harvard Papers of Botany 11: 257–
274.
82
Dissertation Christiane Kiefer – ITS and single copy gene intron analysis
WINDHAM, M.D. AND I. A. AL-SHEHBAZ. 2007b. New and noteworthy species of
Boechera (Brassicaceae) III: additional sexual diploids and apomictic hybrids.
Harvard Papers of Botany 12: 235–257.
LEGENDS SUPPLEMENTARY MATERIAL
Supplementary figure 1 Geographic distribution of accessions used in this study
Supplementary table 1 Accession list showing internal accession number, ITS type,
trnLF haplotype as well as origin of the herbarium specimen, taxon information and
geographical location.
Supplementary Table 2 ITS type gene bank accession numbers and accessions
sharing the ITS-type as well as cpDNA haplotype identified and the ecoregion to
which the accession was assigned. Colour codes indicate if the ITS type occurred
once (yellow), if it was species specific (orange) or if it was shared by several taxa
(red). Bold letters indicate the presence of ambiguous codes in the sequence.
Supplementary table 3 Annotated Alignment including all ITS types recovered in this
study and in Koch et al. 2003.
Supplementary table 4 ITS Alignment reduced to ITS types without ambiguous
codes other than N and originating from accessions which were most likely not
apomictic according to the species determination.
Supplementary table 5 Alignment of the intron sequences from the intron of
At2g25920 neighbouring the ELF3 gene.
Supplementary table 6 Alignment of the intron sequences from the intron of
At3g18900 neighbouring the VRN1 gene.
Supplementary table 7 Table listing material used for the single copy gene intron
analysis together with ITS types and chloroplast haplotypes.
83
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
2.3 Eastern versus western North American Boechera
Manuscript submitted to TAXON;
supplementary data are included on the DVD in the back of the dissertation.
Running title:
Systematics of eastern North American Boechera
Boechera or not? Phylogeny and phylogeography of eastern North
American Boechera species (Brassicaceae)
Christiane Kiefer, Christoph Dobeš, and Marcus A. Koch#
Heidelberg Institute of Plant Science, Biodiversity and Plant Systematics,
Heidelberg University, Heidelberg, Germany
Correspondence:
#
Marcus A. Koch. Heidelberg Institute of Plant Science, Department of Biodiversity
and Plant Systematics, Im Neuenheimer Feld 345, D-69120 Heidelberg, Germany.
Fax: +49-(0)-6221-54-5508; E-mail: [email protected]
1
Manuscript received _______; revision accepted _______.
Acknowledgements
We are grateful to the curators of the herbaria GH, MO and DAO for providing leaf
material, and Ihsan Al-Shehbaz for continuous advice and support during our studies
on Boechera . Funding from DFG to M.A.K. (German Research Foundation) was
greatly acknowledged.
84
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
Abstract
The North American genus Boechera comprises according to current taxonomic
classification 110 species. Most of them are centered in western North America,
with some species extending their distribution range from the west into eastern
North America and seven species exclusively occur in eastern North America. Past
phylogenetic studies included at most two of the eastern species and placed them
as sister to two representatives of western North American Boechera. Our studies
based on chloroplast DNA markers (trnL intron, trnLF IGS) and nuclear ribosomal
DNA ITS1 and ITS2 show that eastern North American Boechera represent two
evolutionary lineages within the genus Boechera (cpDNA) or among the tribe
Boechereae (nrDNA). We included also ancestral Polyctenium and other
Boechereae like Sandbergia and Borodinia, for which close relationships to
Boechera are assumed. Based on our data we suggest to retain eastern North
American Boechera within Boechera rather than transferring them into a new
genus. Some phylogeographic conclusions are also drawn like a possible migration
scenario of Boechera into eastern North America. The Siberian species Boechera
falcata as well as the Siberian genus Borodinia were shown to be placed within
Boechera.
KEYWORDS: Boechera, Borodinia, Brassicaceae, cpDNA, ITS, phylogeny,
phylogeography, Sandbergia, trnLF
85
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
INTRODUCTION
Boechera A. Löve & D. Löve is a North American genus from the Brassicaceae
(mustard family) and is currently subject to a broad range of studies such as
apomixis research (Sharbel & al. 2004, Voigt & al. 2007), phylogeography (Dobeš &
al., 2004a), drought tolerance (Knight & al., 2006), host-fungi interactions (Roy,
2001), and response to herbivory (Siemens & al., 2003). Furthermore, genome and
chromosome evolution were studied in great detail (Kantama & al., 2007).
Population studies were carried out using the sexual species B. stricta (Song & al.,
2006) and crossing experiments revealed the high potential for hybridisation within
Boechera (Schranz & al., 2005). Based on recent molecular studies and analysis of
morphological data (Al-Shehbaz & al. 2006) Boechera belongs to the tribe
Boechereae together with six other genera [Anelsonia, Cusickiella, Nevada,
Phoenicaulis, Polyctenium, Sandbergia (= Halimolobus perplexa, Al-Shehbaz,
2007)]. The Boechereae are almost exclusively North American, have a base
chromosome number of x=7 (Dobeš & al., 2006, Warwick & al., 2006) and are
characterized with few exceptions by branched trichomes and entire leaves. Most of
the members of the tribe are perennials with a well-developed basal rosette (AlShehbaz & al., 2006). Boechera has its highest species diversity in western North
America. However, seven of the 110 currently accepted species (Windham & AlShehbaz, 2006, 2007a, 2007b) are restricted to eastern North America (Boechera
laevigata, Boechera missouriensis, Boechera perstellata, Boechera burkii,
Boechera shortii, Boechera serotina, Boechera canadensis; a summary of the
distribution ranges of eastern North American Boechera species and their habitats
and chromosome numbers is given in Table 1). Eastern and western North
American Boechera differ from each other in life form, habitat preferences, as well
as morphology. The eastern species are biennials of forest habitats and have
dentate obovate to broadly oblanceolate leaves up to 16 cm long. In contrast,
western Boecheras are mainly perennials of open habitats and have entire, much
smaller leaves (Rollins, 1993).
A phylogenetic study based on the plastid ndhF placed eastern North American B.
laevigata and B. shortii as sister to Boechera platysperma and Anelsonia (Beilstein
& al., 2006). This sister group relationship of B. laevigata and B. platypserma was
also shown in a study based on several plastidic and nuclear markers which in
addition included Boechera stricta which was phylogenetically closest to B.
platysperma (Bailey & al., 2006). In both studies Polyctenium was sister to all other
members of tribe Boechereae, followed by Cusickiella and Nevada as next closest
sisters. However, the phylogenetic position of Cusickiella seems to depend strongly
on the applied marker system which is obvious from an earlier phylogenetic study
where three different marker systems were applied (trnLF region, nrDNA ITS,
pistillata intron 1; Bailey & al., 2002). No other eastern North American Boechera
species except B. laevigata and B. shortii were included in any of these
86
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
phylogenetic reconstructions so no information about their relatedness is available
so far. Furthermore Boechera falcata, the only Siberian Boechera described so far,
as well as the Siberian genus Borodinia for which a close relationship to Boechera
was assumed from morphological data, have not been included into a molecular
phylogenetic study yet. At the generic level past studies of the chloroplast DNA
marker trnLF revealed three major evolutionary lineages within Boechera (Dobeš &
al., 2004) and were complemented by the analysis of nuclear encoded loci (Koch &
al., 2003, Dobeš & al., 2004a). These studies included only B. stricta (sexual,
diploid species), B. holboellii sensu Rollins and B. divaricarpa sensu Rollins
(Rollins, 1993) but none of the eastern North American relatives.
Table 1 Distribution ranges, habitats and chromosome numbers for eastern North
American Boechera.
species
distribution
habitat
Boechera
laevigata
southern Quebec
southwards into Georgia
and Alabama, west into
Oklahoma and Kansas,
and north into Minnesota
from Maine and Vermont
south into Georgia, west
through Oklahoma, and
north into Wisconsin
from Canada’s southern
Quebec and Ontario
south into Florida,
westwards into Texas,
and north into Nebraska
and Minnesota
Maryland, Pennsylvania.,
Tennessee, Virginia and
West Virginia
Virginia and West
Virginia
in wooded areas and
on cliffs, sandhills,
bluffs, and limestone
ledges
Boechera
missouriensis
Boechera
canadensis
Boechera
burkii
Boechera
serotina
Boechera
shortii
from New York south into
Alabama, west into
Kansas, and north into
South Dakota and
Minnesota
Boechera
perstellata
Kentucky and
Tennessee
87
in wooded areas and
on cliffs, sandhills,
bluffs, and limestone
ledges
Riparian habitats
in rocky areas,
wooded slopes, and
on river banks
on shale barrens and
wooded slopes of
crumbling shale
in rich woods, on
stream banks, lake
shores, steep
wooded slopes,
limestone bluffs and
cliffs, river flood
plains, and shaded
bottomlands
hillsides and
calcareous bluffs
Base
chromoso
me number
n=x=7
n=x=7
n=x=7
Not
available
2n=2x=14
2n=2x=12
(counted in
Arabis
dentata
(Smith
1938)=
Boechera
shortii)
2n=2x=14
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
Palynological studies have shown that during quaternary glaciation cycles
forest habitats underwent relocations in eastern North America as tree species
retreated into refuge areas further south avoiding the colder temperatures (Jackson
& al. 2000), and one might assume that the Boecheras followed migration of these
habitat types. However, some tree species as for example Quercus rubra remained
further North closer to the ice shield during the LGM (Magni & al. 2005). Actually,
nothing is known about the phylogeographic history of eastern North American
Boechera species. For B. stricta, whose distribution range extends into north
eastern America chloroplast DNA data indicated a potential refuge area near the
Great Lakes region (Dobeš & al. 2004a). However, this was not true for B. holboellii
varieties investigated in the same study. A well known phylogeographic pattern in
eastern North America is the Atlantic and Gulf Coastal Plain disjunction to the Great
Lakes region. Prunus (Shaw & Small 2005) as well as Trillium grandiflorum, a
herbaceous woodland species (Griffin & Barrett 2004), seem to follow this pattern
among many others.
Inhere, first we aim to reconstruct the phylogenetic relationships between the
eastern North American Boechera species and those centred in western North
America as well as the most of the remaining Boechereae based on cpDNA (trnL
intron and trnL-F intergenic spacer) and nrDNA (ITS1, ITS2). Here we included also
Borodinia and Boechera falcata, for their phylogenetic positions within Boechereae
are still unclear.. Second, we would like to verify the morphology-based generic
assignment of eastern North American Boechera using the same data. Third, based
on a phylogeographic analysis of the chloroplast DNA we will try to identify potential
refuge areas of eastern North American Boechera species.
MATERIALS AND METHODS
Plant material.—Leaf material was obtained from herbarium specimens from GH,
MO and DAO. Corresponding accession details are listed in Table S1 (Online
supplementary material). From B. burkii we obtained no and from B. serotina, we
obtained only limited material, and only some molecular DNA sequence data from
herbarium material of B. serotina were obtained successfully.
DNA extraction.—Total DNA was obtained from a 5x5 mm2 piece of dried leaf
tissue from single individuals. Extraction followed the CTAB method of Doyle &
Doyle (1987), but some modifications were applied, involving grinding of only a 5x5
cm2 piece of of dry leaf tissue in 2ml tubes using a Retsch swing mill (MM 200),
addition of two units of ribonuclease (RNAse A) to the resolved DNA, and washing
of the DNA pellet twice with 70% ethanol. DNA was finally dissolved in 50-70µl low
TE-buffer or low Tris-buffer for long-term storage.
88
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
PCR conditions.—PCR reactions were performed in a volume of 25 µl containing
1x GoTaq buffer (Promega), 2 mM MgCl2, 5 pmol of each primer, 5 nmol dNTPs
(1.25 nmol of each dNTP) and one unit Taq DNA polymerase (GoTaq, Promega),
and variable concentrations of template (50 to 400 ng) using a PTC-200 thermal
cycler (MJ-Research). Thermal cycling started with a denaturation step at 95°C
lasting three min; followed by 30 cycles each comprising 30 s denaturation at 95°C,
30 s annealing at 48°C and elongation at 72°C. Amplification ended with an
elongation phase at 72°C lasting 10 min, and a final hold at 4°C.
The trnL intron was amplified using the forward primer 5‘-CGA AAT CGG TAG ACG
CTA CG-3‘ and the reverse primer 5‘-GGG GAT AGA GGG ACT TGA AC-3‘ (primer
c and d according to Taberlet & al., 1991), which anneals in the first and second
exon of the trnL gene, respectively. Sequences comprised the complete intron and
the second exon of the trnL gene. For amplification of the trnL-F intergenic spacer
(IGS) primers 5‘-GGT TCA AGT CCC TCT ATC CC-3‘ (primer e according to
Taberlet & al., 1991) and 5‘-GAT TTT CAG TCC TCT GCT CTA C-3’ (designed in
Dobeš & al., 2004a) annealing in the second exon of the trnL gene and the trnF
gene, respectively, were used. Amplified sequences included the complete IGS and
the first 18 bases of the trnF gene. PCR products were checked for length and
concentrations on agarose gels (1% agarose in TAE). PCR products were all
purified with the NucleoFast Kit (Macherey & Nagel, Germany).
ITS1 and ITS2 were amplified together as described in Dobeš & al. (2004b).
Sequences comprised the complete ITS1 and ITS2 as well as the 5.8s rRNA gene.
PCR products were checked on agarose gels (1% agarose in TAE). PCR products
were all purified with the NucleoFastKit (Macherey & Nagel, Germany).
Cycle Sequencing.—Cycle sequencing was done with the DYEnamic ET
Terminator Cycle Sequencing Kit (Amersham Biosciences) using the PCR primers
for the cycle sequencing reaction. Samples were resolved in 10 µl Loading Solution
and then run on a MegaBace 500 Sequencer.
Alignments and chloroplast haplotype/ITS type definition.—After sequencing,
the forward and reverse sequences were aligned, edited by hand and trimmed to a
common length. The trnL intron and the trnL-F IGS region were assembled into one
sequence (trnLF region) and missing bases at the joining 3’- and 5’-primer ends
were substituted with Ns. Duplicate sequences were removed and cpDNA and ITS
haplotypes were defined based on single nucleotide and indel polymorphisms.
Haplotypes designation introduced by Dobeš & al. (2004) and Schranz & al. (2005)
were assigned to those haplotypes already published by these authors and found
again in this study. The alignments used for the phylogenetic analysis were made
manually using the program GenDoc (Nicholas & Nicholas, 1997) according to the
89
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
alignment published by Dobeš & al. (2004) including one to three trnF pseudogenes
at the 3´-end of the trnLF intergenic spacer. ITS types were defined by the same
method as described for trnL-F haplotypes.
TrnLF Phylogenetic reconstruction.—For reconstruction of phylogenetic
relationships of taxa belonging to the Boechereae sequences of Sandbergia
(EU850744, EU850745), Cusickiella (EU850746, EU850747; EU850748,
EU850749; EU850750, EU850751; EU850754, EU850755; EU850752, EU850753)
and Polyctenium (EU850756, EU850757; EU850758, EU850759; EU850760,
EU850761; EU850762, EU850763; EU850764, EU850765; EU850766, EU850767;
EU850768, EU850769) as well as several sequences representing members of the
Halimolobeae (AF307555, AF307549, AF307551, AF307543, AF307545,
AF307538, AF307530, AF307531) were added to the alignment used for the
network analysis (see below). Capsella bursa-pastoris (AY122454, DQ310514) was
used as outgroup. The region of the trnL second exon was omitted from the
analysis (positions 530-610) because many missing or ambiguous data were
present in this region. Tree reconstructions were done by using TREECON for
Windows 1.3b (Van de Peer & De Waachter, 1994). A neighbour-joining (NJ)
analysis (Kimura-2-parameter distance; transition/transversion ratio calculated and
set automatically from the data; gaps excluded) was run excluding the pseudogene
region and its further 3´-region (bp. 802-1214) because of uncertain homology of
the pseudogenes found in different genera (see haplotype definition for more
information on pseudogenes). A bootstrap analysis running 1000 replicates was
performed with TREECON using the same distance measurements.
In order to estimate the potential influence of the various gaps on tree topology we
performed the same analysis running TREECON with insertion/deletions taken into
account.
TrnLF network reconstructions.—For network analysis the most common
haplotypes (AY257725, AY257736, AY257718, AY257692, AY257710, AY257694,
AY257712, AY257697, AY257703, AY257761, AY257764, AY257775, DQ012052,
AY257778) identified by Dobeš & al. (2004) were added to the alignment to infer the
relationship of haplotypes of eastern North American Boechera and previously
identified haplotypes. Network reconstruction was done by running the program
TCS1.21 (Clement & al. 2000), in which the gaps were coded and the connection
limit was set to 95%. The analysis was run excluding the trnF pseudogenes (see
haplotype definition for more information on pseudogenes).
ITS phylogenetic tree reconstructions.—Phylogenetic reconstructions based on
ITS1 and ITS2 were done by running parsimony analyses as implemented in
PAUP4.0beta (Swofford, 2002). The maximum of trees retained was limited to
90
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
10,000. For the heuristic search sequences were added randomly in 1000 replicates
during which 10 trees were saved in each replicate. Chuckscore was set to one.
Gaps were treated as missing. TBR was used as branch-swapping algorithm.
Starting trees were obtained via stepwise addition. Six characters were excluded
from the analysis (505-510). All characters had equal weight.
For phylogenetic reconstructions based on the ITS dataset alone the alignment was
supplemented with sequences of closely related genera obtained from Genebank to
determine the relationship of eastern North American Boechera species to
representatives from western North America and to other genera of the tribe
Boechereae (DQ452606, AF146515, AF146514 = Cusickiella douglasii, DQ452059
= Anelsonia eurycarpa, DQ452061 = Nevada holmgrenii, DQ452066 = Cusickiella
quadricostata and AJ232927, AJ232926 = Halimolobus perplexa var. lemhiensis
(hereafter Sandbergia)). Three Capsella species served as outgroup (Capsella
rubella AJ232913, Capsella bursa-pastoris AF137570, Capsella grandiflora
AM905718). In order to resolve relationships among the Boechereae and
Halimolobeae Halimolobos (AF307634, AF307635), Mancoa (AF307631),
Sphaerocardamum (AF307611, AF307612) and Pennellia (AF307627, AF307629),
Arabis tricornuta (AF307628) were added to the analysis.
Bootstrap analysis was also run in PAUP4.0beta with the same settings as the
parsimony analysis in 1000 replicates.
Geographical analysis.—All information from the herbarium vouchers was entered
into the database BioOffice (Biogis Consulting, Version 2.0.4). Missing geographical
coordinates were added according to the descriptions on the herbarium vouchers.
Using BioOffice haplotypes were plotted on North America maps included in the
ArcView Package Version 8.
RESULTS
Alignments.—The length of the alignment of the combined trnLF region used for
intrageneric analysis in Boechera was 975 bp (Table S2, online suppl. material; trnL
intron: position 1 to 506; trnL second exon: position 507 to 556; the trnL-F IGS:
position 557 to 960, the partial sequence of the trnF gene from 961 to 975) in the
alignment for the network analysis. In the trnL-F IGS one to three pseudogenic
copies of trnF were present as in all other Boechera species and closely related
genera like Halimolobus. The pseudogene region (alignment position 626-933) was
excluded from all further analyses except for haplotype definition because addition
and deletion of copies does not coincide with the phylogenetic signal of regions
other than the pseudogenic one (cf. Dobeš & al., 2007).
91
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
The trnLF alignment for the phylogenetic analysis including the other genera from
the Boechereae and Capsella species as outgroup had a total length of 1214 bp.
The alignment is given in Table S4.
The ITS alignment had a total length of 693 bp (Table S3, online suppl. material;
partial sequence of the 18s rRNA gene: position 1 to 23; ITS1: position 24 to 296;
5.8s rRNA gene: positions 297 to 469; ITS2: position 470 to 693).
TrnLF phylogenetic reconstruction.—Phylogenetic reconstruction based on the
trnLF dataset is shown in Fig. 1. The representatives of the Halimolobeae included in
this study were sister to the Boechereae with moderate bootstrap support. The
Figure 1 Neighbour-joining tree based on cpDNA marker system trnLF; for the
outgroup (Capsella) and the Halimolobeae (Bailey & al. 2002) genebank accession
numbers are indicated, for the Boechereae either the haplotype name (Boechera
species except Boechera falcata) or the accession number of the individual in our
study is given; for information on the individuals carrying those haplotypes refer to
supplementary Table 1. Arabis tricornuta (AF307555= has been proposed to be
transferred to the genus Pennellia (Al-Shehbaz 2003).
92
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
genus Cusickiella constituted two lineages representing C. douglasii and C.
quadricostata, respectively, at the base of the Boechereae and was paraphyletic to
a clade comprising Polyctenium fremontii, Sandbergia and Boechera. Cusickiella
was not grouped into a monophyletic lineage and Polyctenium was placed ancestral
to all remaining taxa of the Boechereae. Sandbergia grouped together with
Boechera lineage I haplotypes (sensu Dobeš & al. 2004a). Boechera lineage I and
Sandbergia were sister to a clade comprising Boechera lineages II and III, all
eastern North American Boechera species as well as two haplotypes (AA, AB)
found exclusively in the southern Great Basin. Siberian Boechera falcata was most
closely related to haplotype CI (AY257778), a wide spread haplotype from cpDNA
evolutionary lineage III.
Analysis running under the OPTION “insertion/deletions” resulted in an almost
identical topology (data not shown) with two notable changes: 1) Sphaerocardamum
and remaining Halimolobeae are monophyletic with weak bootstrap support (BS
53%) and sister to Boechereae. 2) Sandbergia is ancestral to all other Boechera
species with weak bootstrap support as well (BS 41%), a finding consistent with ITS
data as described below.
ITS Phylogenetic reconstructions.—In the ITS dataset 687 characters were
included in the analysis with 461 characters found to be constant, 69 parsimonyuninformative characters and 157 parsimony-informative characters. We excluded
characters 505-510, which could be hardly aligned (Supplementary Table 3.) The
phylogenetic reconstruction recovered a strict consensus tree based on 1,000
shortest trees with a good resolution among terminal accessions corresponding to
species or groups of species, but in contrast with a very poorly supported
relationships among them (Fig. 2) (CI=0.76, RI=0.92, RC=0.70, tree length=370).
Cusickiella was sister to the Halimolobeae and Boechereae (bootstrap support (BS)
= 40%). The Halimolobeae and Boechereae were also found to have a sister group
relationship. However this was only weakly supported (BS 38%).
Within the Boechereae several lineages were identified: three lineages were
found to arise from a basal polytomy. The first lineage included Anelsonia, the
second Sandbergia and Polyctenium, and the third Borodinia, Nevada and eastern
and western North American Boechera. Borodinia was on a polytomy with a lineage
comprising B. perstellata, B. shortii, B. missouriensis, B. laevigata and B. serotina
and a lineage comprising B. falcata, western North American Boechera and B.
canadensis. The splits differentiating major groups had low to no bootstrap support.
In general it is noteworthy that among the eastern North American Boechera
species the ITS is not species-specific. This is also true for most western North
American Boechera species (Kiefer & Koch, unpublished).
93
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
Figure 2 Parsimony analysis of nrDNA data; strict consensus tree of 10,000 shortest
trees. Capsella was used as outgroup. For the Halimolobeae, Cusickiella,
Polyctenium, Anelsonia and two Sandbergia accessions genebank accession
numbers are given; for Boechera species either a genebank accession number is
given if it was analysed in a previous study or the internal accession number from
supplementary Table 1 is given.
94
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
TrnLF network.—The network analysis revealed six different lineages. It recovered
the previously described lineages I, II and III (Dobeš & al. 2004a) and two additional
lineages described herein as lineages IV and V (Fig. 3) which contained the eastern
North American Boechera species and a lineage containing only one haplotype.
Haplotypes AA and AB (combined here to one single haplotype because they differ
only in one mutation in the excluded pseudogene region) were found to be in the
centre of the network. Lineage IV contained five haplotypes that were exclusive to
B. canadensis with the most internal haplotype FX being connected to AA/AB by
seven mutational steps. Lineage V was connected to the central haplotypes AA/AB
by one mutational step (haplotype EF). Haplotype FL was derived from haplotype
EF by one mutational steps and haplotype HM was derived from haplotype FL by
one mutational step. Haplotype Z (three mutational steps) and GH (four mutational
steps) were also derived from haplotype EF. Haplotype EF was private to B.
laevigata with one exception where it was also found in a B. missouriensis
individual. Haplotype FL was found in B. missouriensis, B. perstellata, and one
individual of B. laevigata. Haplotype GH and Z were found in B. shortii. Haplotype Z
was also found in one B. divaricarpa accession [taxonomy according to Rollins
(1993)], a hybrid species of multiple origin (see Windham & Al-Shehbaz, 2007b).
Figure 3 Network analysis of trnLF
data
to
reconstruct
phylogeographic history of eastern
North American Boechera species.
These species are found in three
distinct lineages which like the
haplotype lineages typical for
western North American Boechera
species originate from Haplotype
AA/AB. Haplotype names are given
as
letter
codes;
individuals
carrying these haplotypes can be
found in supplementary Table 1.
Boxes with dashed lines encircle
lineages, boxes with solid lines
indicate haplotypes specific to a
single species or a group of closely
related species.
Phylogeographic analysis.—For phylogeographic analysis, the geographic
distribution of haplotypes was plotted onto a map (Fig. 4a, b). For B. canadensis,
haplotype EI was found north of the last maximum glaciation in New York and north
of Lake Erie in Ontario and southern Quebec. Only one accession in Iowa was
found south, but still close, to the maximum extent of the glaciers during the last
95
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
glaciation. Haplotype EK, derived from EI, was found at the western edge of the
distribution range of EI also in Iowa. Haplotypes FU and FX were found in Arkansas
and Georgia, and haplotype FV was found in Mississippi. Haplotype FW was found
east of the Appalachian Mountains in North Carolina.
Boechera laevigata haplotype EF was found throughout the sampled area. The
northern range of EF extended along lakes Erie and Ontario, as well as into New
York State and southern Quebec. The southern range of EF included Virginia, North
Carolina, and Tennessee. In North Carolina bordering Tennessee, the haplotype EF
was also shared by one B. missouriensis accession. Haplotype FL was distributed
south of the last glaciation co-occurring with EF. It was found in Tennessee, South
Carolina, North Carolina, and Oklahoma. In Tennessee FL was also shared by B.
perstellata and one B. laevigata accession. Boechera shortii haplotype GH was
found in Missouri, and its haplotype Z was found north of Erie Lake. Haplotype Z,
shared by B. divaricarpa, was found in Minnesota.
Figure 4 Distribution ranges of
haplotypes found in eastern North
American Boechera. The extension
of the last glaciation is shown as a
red
line,
the
Appalachian
Mountains are shown as shaded
area, species names in figure (B)
are coded with two letters;
different
symbols
represent
different haplotypes whose name
is given in the figures; asterisks
indicate singleton haplotypes. (A)
Lineage IV; haploytypes in lineage
IV were all found in B. canadensis
accessions. Haplotypes found in
the south are connected to the
centre of the network shown in
figure 3; north of the extension of
the last glaciation only haplotype
EI is found. (B) Lineage V; north of
the extension of the last glaciation
only haplotypes EF and Z were
found. South of the extension of
the last glaciation haplotype EF
occurred together with haplotype
FL and two derived haplotypes.
96
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
DISCUSSION
Phylogenetic Reconstruction.— Eastern and western North American Boechera
species differ significantly in their life history, habitat preferences and morphology.
The question is if this differentiation can also be recognized in the divergence of
marker sequences and how Eastern North American Boechera are related to other
genera within the Boechereae. Our study based on cpDNA placed B. laevigata, B.
missouriensis, B. shortii and B. perstellata inside Boechera as sister to Boechera
chloroplast lineages II and III and B. canadensis. In the nrDNA dataset on the other
hand B. laevigata, B. missouriensis, B. perstellata, B. shortii and B. serotina were
sister to ITS-lineages representing western North American Boechera. In a familywide phylogeny B. laevigata was found to be sister to B. platysperma and B. stricta
(Bailey & al. 2006). A phylogenetic study based on ndhF placed B. laevigata and B.
shortii as sister group to B. platysperma and Anelsonia (Beilstein & al. 2006). Both
results are in agreement with our result obtained for the ITS analysis.
Boechera canadensis was not included into previous phylogenetic studies. Inhere
B. canadensis was carrying haplotypes nested within Boechera, sister to Boechera
chloroplast lineage II. Incongruencies between phylogenies based on different
marker systems can occur due to different reasons. Within Boechera it seems likely
that we have to explain distribution of chloroplast DNA variation among species as a
result of the distribution of ancestral genetic variation predating separation of the
various lineages carrying these haplotypes today. This phenomenon has been
described in detail not only for Boechera (Dobeš & al, 2004), but also for the genus
Arabidopsis (Koch & Matschinger, 2007).
The Siberian B. falcata has also not been included in phylogenetic analyses so far.
In our analyses it was placed with western North American Boechera species for
cpDNA as well as nrDNA ITS. The cpDNA haplotype found in B. falcata was found
to be related to haplotype CI (AY257778) found in evolutionary lineage III. This
haplotype is among the most widespread haplotypes within Boechera as was
shown in previous phylogeographic studies (Dobeš & al., 2004a). It also extends
into Alaska which might hint on a migration into Siberia via the Bering Strait. The
ITS sequence from B. falcata was sister to ITS sequences found in western North
American Boechera which suggests an early split of B. falcata and western North
American Boechera species.
Resolution between Boechera and other members of the Boechereae was low and
our analyses were partially in conflict with the analyses performed in previous
studies (Bailey & al., 2002; Bailey & al., 2006; Beilstein & al., 2006). However, we
can show that the Siberian genus Borodinia groups together with Boechera.
Hereby, the placement of Borodinia within the Boechereae or even Boechera based
on morphological data can be confirmed by molecular data. Furthermore, we found
weak evidence that Sandbergia is basal to Boechera and Borodinia. The placement
97
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
of Cusickiella was dependent on which marker system was applied. The cpDNA
dataset suggested to place Cusickiella within tribe Boechereae while the analysis of
nrDNA data places it even outside the Halimolobeae. This discrepancy in the
placement of Cusickiella was also observed in a study of the Halimolobeae where
Cusickiella and B. stricta (former Arabis drummondii) were used as representatives
of the Boechereae (Bailey & al. 2002). In a family wide phylogeny based on several
nuclear and plastidic marker systems Cusickiella was resolved within the
Boechereae and Polyctenium was sister to Cusickiella and Boechera (Bailey & al.,
2006). This is in conflict with both of our phylogenetic analyses. However, bootstrap
support for the relationships among the Boechereae was almost low in our datasets
and it was also low in the family wide phylogeny. At this point we can conclude only
that Cusickiella is paraphyletic and Polyctenium has been considered for any future
taxonomical changes within this group. The family wide phylogeny was already
based on ten different marker systems (five from the plastid including trnLF, five
from the nuclear genome including ITS), and might also reflect the general conflict
between nuclear and plastid derived phylogenetic hypothesis. So it is unlikely that
the addition of more sequence data from the various genomes would lead to a
higher bootstrap support or better resolution for the relationships among the
Boechereae.
Phylogeographic Analysis based on the trnLF Network.—Our sampling of
eastern North American Boechera species covered their entire distribution ranges,
and so we could draw some conclusions on their phylogeographic history.
Phylogeographic studies of eastern North America have revealed a multitude of
different patterns in the past (Soltis & al., 2005). Many of the patterns are shaped by
the Appalachian mountains that in some cases acted as barriers to gene flow
(Griffin & al., 2004; Godbout & al., 2005). In the case of B. laevigata (haplotype EF
only) and B. missouriensis (haplotypes FL and EF only) this mountain chain seems
to have had little effect on haplotype distribution and genetic differentiation as the
same haplotypes are found east end west of the ridge of the Appalachian mountain
chain. In B. canadensis sampling density is not sufficient to draw any conclusions
on that particular pattern.
A common pattern of phylogeographic analyses is northern purity versus southern
richness as an effect of the quaternary ice ages (southern refugia and recolonization
of northern areas) (Hewitt, 2000). For B. canadensis the basal haplotypes were all
found south of the last glaciation in southeastern North America and Iowa while north
of the last glaciation one single derived haplotype was found. The second chloroplast
DNA lineage for eastern North American Boechera was remarkably poor in
haplotypes considering the fact that it contains four species and it was not possible to
draw any conclusions on the phylogeographic history of this lineage. The extreme
haplotype poverty in this lineage might hint at recent differentiation of those taxa. In
98
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
the past B. laevigata and B. missouriensis were sometimes even treated as varieties
of the same species (Arabis laevigata (Muhl. ex. Willd.) DC. var. missouriensis
(Greene) H.E. Ahles; Ahles, 1964). This view is also supported by some haplotype
sharing between B. missouriensis and B. laevigata. Haplotype sharing of B. laevigata
and B. missouriensis occurred in Tennessee in the Interior Plateau ecoregion
adjacent to the southwestern Appalachian Mountains. Rollins (1993) already noted
that “… in North Carolina in particular, the two species appear to come close together
morphologically …”. In fact this might indicate hybridisation of the two species in this
region or regions close by (like in Tennessee) as is also indicated by the chloroplast
data.
B. laevigata and B. missouriensis ecologically favour forest habitats. From the
descriptions on herbarium specimens and from the placement of accessions on the
used ecoregion map (Olson & al., 2001) the analysed individuals were all collected in
temperate broad leaf and mixed forest in particular in connection with various
Quercus species. During the last glacial maximum Quercus was abundant in the
Mississippi valley and Florida but it also occurred in patches in between (Jackson &
al. 2000). If B. laevigata and B. missouriensis had to retreat into refugia during the
last glacial maximum they may have followed Quercus in one of the scattered
populations and may have hybridized there as a consequence of close spatial
contact.
B. shortii is only represented by two individuals in the study. Its haplotypes are
derived from EF and form a lineage with many missing haplotypes. From two
individuals no phylogeographic history can be inferred but it is still noteworthy that
haplotype Z found in B. shortii is also shared by one triploid apomictic hybrid that
previously had been identified as B. divaricarpa. According to the currently accepted
description this species is restricted to Washington, Idaho, and Montana, south to
Wyoming, Utah, Nevada, and California (Windham & Al-Shehbaz, 2007b).
However, in the past B. divaricarpa was used as a “waste basket species”
(Windham & Al-Shehbaz, 2007) and many hybrids of different parentage were
determined as B. divaricarpa. The individual described as B. divaricarpa should be
revisited for species identity as typically B. divaricarpa shares haplotypes from
lineages I, II or III (Dobeš & al., 2004).
Haplotype AA/AB is central to the cpDNA network and connects the five identified
lineages to each other. Today it is only found in the southern Great Basin. This
raises the question where eastern North American Boechera species evolved and
how they arrived in their current distribution range. Most Boechera species
originated in western North America, an area from which lineages IV and V reached
the eastern part of the continent, as evidenced by the distribution range of central
haplotype AA/AB from which lineages IV and V are derived. From our data the
99
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
following hypothesis about the evolutionary relationships between western and
eastern North American haplotypes may be developed: Haplotype EF probably
originated in western North America prior to glaciation or during interglacial periods
and later migrated towards the East. Haplotypes Z and GH also differentiated from
EF probably when lineages IV and V were not fully differentiated from lineages I, II,
and III and the distribution rages of species from all lineages overlapped. During
that time, haplotype Z was captured by species that normally carry haplotypes from
lineages I, II or III like for example “B. divaricarpa” or its ancestral species.
Haplotypes EF and Z became extinct in the West because of repeated glacial
cycles and survived only in eastern North America. Haplotype Z was also
diminished and many of the intermediate haplotypes leading from EF to Z were lost.
SUMMARY
Our phylogenetic reconstructions based on chloroplast DNA and nrDNA were not
congruent. This can most likely be explained by the different evolutionary histories
of the employed marker systems. Support for ancestral splits between lineages in
the phylogenetic reconstruction based on nrDNA data was low while more shallow
splits had a good support. Therefore we can conclude that B. laevigata, B.
missouriensis, B. shortii, B. serotina and B. perstellata constitute one lineage of
eastern North American Boechera while B. canadensis constitutes a second one.
This is also supported by chloroplast DNA data. However the relationships among
those lineages and their relationship to western North American Boechera remain
unclear. Therefore we suggest to retain B. canadensis and the group of Boechera
species related to B. laevigata within Boechera instead of transferring them into
different genera as has been suggested elsewhere. The Siberian Boechera species
B. falcata and the Siberian genus Borodinia were shown to be placed within
Boechera.
Phylogeographic analysis revealed a higher chloroplast diversity in southern regions
for B. canadensis and a uniform haplotype distribution for B. missouriensis and B.
laevigata. The route through which eastern North American Boechera arrived in
their current habitats is unclear. Since the closest related haplotype AA/AB is only
found in the southern Great Basin in the western United States either this haplotype
had a wider distribution range into the east or eastern North American Boechera
evolved closer to the distribution range of AA/AB, migrated into eastern North
America and became extinct in the western parts of North America during the
quaternary glaciation cycles.
100
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
LITERATURE CITED
Ahles, H.E. 1964. Journal of the Elisha Mitchell Scientific Society 80: 172.
Al-Shehbaz, I.A. (submitted) Boechera A. Löve and D. Löve. In: Flora North
America.
Al-Shehbaz I.A., Beilstein, M.A. & Kellogg, E.A. 2006. Systematics and
phylogeny of the Brassicaceae (Cruciferae): an overview. Pl. Syst. Evol. 259: 89-120
Al-Shehbaz, I.A. 2003. Transfer of most North American species of Arabis to
Boechera (Brassicaceae). Novon 13: 381--391.
Al-Shehbaz, I.A. 2007. The North American genus Sandbergia (Boechereae,
Brassicaceae). Harvard Papers Bot. 12: 425--427.
Bailey, C.D., Koch, M.A., Mayer, M., Mummenhoff, K., O’Kane, S.L., Warwick,
S.I., Windham, M.D. & Al-Shehbaz, I.A. 2006. A Global nrDNA ITS Phylogeny of
the Brassicaceae. Mol. Biol. Evol. 23: 2142--2160.
Bailey, C.D.M., Price, R. A. & Doyle, J. J. 2002. Systematics of the Halimolobine
Brassicaceae: Evidence from Three Nuclear Loci and Morphology. Syst. Bot. 27:
318--332.
Beilstein, M.A., Al-Shebbaz, I.A. & Kellogg, E.A. 2006. Brassicaceae phylogeny
and trichome evolution. Am. J. Bot. 93, 607--619.
Clement, M., D. Posada, & Crandall, K.A. 2000. TCS: a computer program to
estimate gene genealogies. Molec. Ecol. 9: 1657--1659.
Dobeš, C., Mitchell-Olds, T. & Koch, M.A. 2004a. Extensive chloroplast haplotype
variation indicates Pleistocene hybridization and radiation of North American Arabis
drummondii, A. ×divaricarpa, and A. holboellii (Brassicaceae). Molec. Ecol 13: 349-370.
Dobeš, C., Mitchell-Olds, T. & Koch, M.A. 2004b. Intraspecific diversification in
North American Boechera stricta (= Arabis drummondii), Boechera x divaricarpa,
and Boechera holboellii (Brassicaceae) inferred from nuclear and chloroplast
molecular markers—an integrative approach. Am. J. Bot. 91: 2087—2101.
101
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
Dobeš, C., Koch, M.A., Sharbel, T.F. 2006. Embryology, karyology, and modes of
reproduction in North American genus Boechera (Brassicaceae): a compilation of
seven decades of research. Annals Miss. Bot. Gardens 93: 517--533.
Dobeš, C., Kiefer, C., Kiefer, M. & Koch, M.A. 2007. TrnF pseudogenes in North
American genus Boechera (Brassicaceae): mechanistic aspects of evolution. Pl.
Biol. 9: 1--14.
Doyle, J. J. & Doyle, J.L. 1987. A rapid DNA isolation procedure for small amounts
of fresh leaf tissue. Phytochem. Bull. 19: 11--15.
Griffin S. R. & Barrett, S.C.H. 2004 Post-glacial history of Trillium grandiflorum
(Melanthiaceae) in eastern North America: Inferences from Phylogeography. Am. J.
Bot. 91: 465--473.
Godbout, J., Jamarillo-Correa, J.P., Beaulieu J. & Bousquet, J. 2005 A
mitochondrial DNA minisatellite reveals the postglacial history of jack pine (Pinus
banksiana), a broad-range North American conifer. Molec. Ecol. 14: 3497--3512.
Hewitt, G. 2000 The genetic legacy of the Quaternary ice ages. Nature 405: 907-913.
Jackson, S.T., Webb, R.S., Anderson, K.H., Overpeck, J.T., Webb III, T.,
Williams, J.W., & Hansen, B.C.S. 2000. Vegetation and environment in Eastern
North America during the Last Glacial Maximum. Quat. Science Rev. 19: 489--508.
Kantama L., Sharbel T.F., Schranz M.E., Mitchell-Olds T., de Vries S. & de Jong,
H. 2007 Diploid apomicts of the Boechera holboellii complex show large scale
chromosome substitutions and different aberrant chromosomes. Proc. Natl. Acad.
Sci. USA 104: 14026--14031.
Knight, C., Vogel, H., Kroymann, K., Shumate, A., Witsenboer, H., & MitchellOlds, T. 2006. Expression profiling and local adaptation of Boechera holboellii
populations for water use efficiency across a naturally occurring water stress
gradient. Molec. Ecol. 15: 1229--1237.
Koch, M.A., Dobeš, C., Kiefer, C., Schmickl, R. & Klimes, L. 2006. SuperNetwork
identifies multiple events of plastid trnF (GAA) pseudogene evolution in the
Brassicaceae. Mol. Biol. Evol. 24: 63--73.
102
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
Nicholas, K.B. & Nicholas, H. B. Jr. 1997. GeneDoc: A tool for editing and
annotating multiple sequence alignments. Distributed by the author.
Olson, D.M., Dinerstein, E., Wikramanayke, E.D., Burgess, N.D., Powell, G.V.N.,
Underwood, E.C., D’Amico, J., Itoua, I., Strand, H.E., Morrison, J.C., Loucks,
C.J., Allnutt, T.F., Ricketts, T.H., Kura, Y., Lamoreux, J.F., Wettengel, W.W.,
Hedao, P. & Kassem, D.K.R. 2001. Terrestrial ecoregions of the world: New map
of life on earth. BioScience 51: 933--938.
Magni, C.R., Ducousso, R., Petit, J.R. & Kremer, A. 2005. Chloroplast DNA
variation of Quercus rubra L. in North America and comparison with Fagaceae.
Molec. Ecol. 14: 513--524.
Rollins, R.C. 1993. The Cruciferae of Continental North America. Stanford
University Press, Stanford.
Roy, B.A. 2001. Patterns of association between crucifers and their flower mimic
pathogens: Host jumps are more common than coevolution or cospeciation.
Evolution 55: 41--53
Schmickl, R., Kiefer, C., Dobeš, C. & Koch, M.A. (in press) Evolution of trnF(GAA)
pseudogenes in cruciferous plants. Pl. Syst. Evol.: doi:10.1007/s00606-008-0030-2.
Schranz, M.E., Dobeš, C., Koch, M.A. & Mitchell-Olds, T. 2005. Sexual
reproduction, hybridization, apomixis, and polyploidization in the genus Boechera
(Brassicaceae). Am. J. Bot. 92: 1797--1810.
Sharbel, T.F., Voigt, M.-L., Mitchell-Olds, T., Kantama, L. & de Jong, H. 2004. Are
the aneuploid chromosomes in the apomictic Boechera holboellii complex B
chromosomes? Camacho J.P.M. (ed): B Chromosomes in the Eukaryote Genome
(special issue of Cytogenetic and Genome Research). Cytogen. Genome Res.
106(2-4): 173--183.
Shaw, J. & Small, R.L. 2005. Chloroplast DNA phylogeny and phylogeography of
the North American Plums (Prunus subgenus Prunus section Prunucerasus,
Rosaceae). Am. J. Bot. 92 (12): 2011--2030.
Siemens, D.H., Lischke, H., Maggiulli, N., Schürch, S. & Roy, B.A. 2003. Costs of
resistance and tolerance under competition: the defense-stress benefit hypothesis.
Evol. Ecol. 17: 247--263.
103
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
Smith, F.H. 1938. Some chromosome numbers in the Cruciferae. Am. J. Bot. 25:
220--221.
Soltis, D.E., Morris, A.B., McLachlan, J.S., Manos, P.S. & Soltis, P.S. 2006.
Comparative phylogeography of unglaciated eastern North America. Molec. Ecol.
15: 4261--4293
Song, B.H., Clauss, M.J., Pepper, A. & Mitchell-Olds, T. 2006. Geographic
patterns of microsatellite variation in Boechera stricta, a close relative of
Arabidopsis. Molec. Ecol. 15: 357--369
Swofford, D.L. 2002. PAUP*: Phylogenetic Anaylsis Using Parsimony (*and Other
Methods), Version 4.0B10. Sinauer Associates, Sunderland, Massachusetts
Van de Peer, Y & De Wachter, R. 1994. TREECON for Windows: a software
package for the construction and drawing of evolutionary trees for the Microsoft
Windows environment. Comput. Applic. Biosci. 10: 569-570.
Voigt, M.-L., Melzer, M., Rutten, T., Mitchell-Olds, T. & Sharbel, T.F. 2007.
Gametogenesis in the apomictic Boechera holboellii complex: the male perspective.
Pp. 235--258 in: Hörandl, E., Grossniklaus, U., can Dijk, P., Sharbel, T.F. (eds),
Apomixis: Evolution, Mechanisms and Perspectives. International Association for
Plant Taxonomy, Koeltz Scientific Books.
Warwick, S.I., Francis, A., Al-Shehbaz, I. 2006. Brassicaceae species checklist
and database on CD-ROM. Pl. Syst. Evol. 259: 249--258.
Windham, M.D. & Al-Shehbaz, I.A. 2006. New and noteworthy species of
Boechera I: sexual diploids. Harv. Papers Bot. 11: 61--88.
Windham, M.D. & Al-Shehbaz, I.A. 2007a. New and noteworthy species of
Boechera (Brassicaceae) II: apomictic hybrids. Harv. Papers Bot. 11: 257--274.
Windham, M.D. & Al-Shehbaz, I.A. 2007b. New and noteworthy species of
Boechera (Brassicaceae) III: additional sexual diploids and apomictic hybrids. Harv.
Papers Bot. 12: 235--257.
104
Dissertation Christiane Kiefer – Eastern versus western North American Boechera
Supplementary Table 1
Accession details of the plants included in the study and their haplotype information.
Supplementary Table 2
TrnLF alignment used for network analysis; sequence features are indicated above
the alignment.
Supplementary Table 3
ITS-Alignment used for phylogenetic reconstruction; sequence features are
indicated above the alignment.
Supplementary Table 4
TrnLF Alignment used for phylogenetic reconstruction; sequence features are given
above the alignment.
105
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
3. Cytogenetic Analyses of apomictic Boechera (to be submitted to
Heredity)
Manuscript to be submitted to Heredity.
Het-Hunting in Boechera apomicts – Searching for the Origin of a
Heterochromatized and a Supernumerary Chromosome
Christiane Kiefer1, Eric Schranz2, Terezie Mandakova3, Martin Lysak3, Hans de
Jong4#
1
University of Heidelberg, Heidelberg Institute f Plant Sciences, Biodiversity and
Systematics, Im Neuenheimer Feld 345, 69120 Heidelberg, Germany
2
University of Amsterdam, Instituut voor Biodiversiteit en Ecosysteem Dynamica;
Kruislaan 318, 1098 SM Amsterdam, Netherlands
3
Masaryk University, Department of Experimental Biology, Kamenice 5, 62500 Brno,
Czek Republic
4
Wageningen University, Laboratory of Genetics, Arboretumlaan 4, 6703 BD
Wageningen, the Netherlands
Corresponding author:
Hans de Jong; Wageningen University, Laboratory of Genetics, Arboretumlaan 4,
6703 BD Wageningen, the Netherlands. Tel. +31 629023329; e-mail.
[email protected]
Running Title: Heterochromatic Chromosome in Boechera Apomicts
Keywords: Boechera, apomixis, heterochromatic, cytogenetics
Wordcount:
106
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
Abstract
Usually apomixis, a mode of asexual reproduction, is associated with higher ploidy
levels among plants. Apomixis is a frequent mode of reproduction among North
American Boechera, however, in Boechera apomixis on the diploid and triploid level
is common. Past cytogenetic studies revealed that some apomictic Boechera
accessions are aneuploid and have a small extra chromosome which was assumed
to be a B chromosome. Furthermore it was shown that all diploid and aneuploid
apomictic accessions have a heterochromatized (Het) chromosome which pairs with
another chromosome in meiosis. In a BAC-FISH appproach we reveal that the Het
chromosome is a homologue of Boechera stricta linkage group 1 which is missing a
fragment (genome block D) in 15 chromosome apomicts. The missing fragment was
apparently transfered to a triplicate centromere or has fomed a neo-centromere by
itself. The fragment corresponds to the previously described Del chromosome.
Results from the present BAC-FISH analysis are embedded into results from past
cytogenetic analyses. We can show that Boechera does not have a B chromosome.
107
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
Introduction
Meiotic cell division in sexual reproducing organisms requires an even chromosome
number and pairs of homologous chromosomes for a functional first division.
Therefore large scale chromosome mutations, for example translocations or
inversions, will be “punished” by rendering the individual who carrying such a
mutation infertile or of reduced fertility. However, reproduction may also occur
asexually which means that meiosis is not longer required and embryos may develop
spontaneously from a somatic cell or an embryo-sac mother cell (Bicknell and
Koltunow, 2004). This way of reproduction is refered to as apomixis and is rarely
found in the animal kingdom but evolved several times independently throughout the
plant kingdom (Bicknell and Koltunow, 2004). The lack of meiotic cell division means
that uneven chromosome numbers and large scale chromosome mutations can be
tolerated.
In the plant kingdom large apomictic complexes are found within the Asteraceae
(Taraxacum (e.g. van Dijk et al. 1999); Hieracium e.g. Eckardt, 2003), the
Ranunculaceae (Ranunculus cassubicus complex, Paun et al. 2006) and the
Rosaceae (Potentilla e.g. Nylehn et al. 2003; Rubus e.g. Nybom, 1988). However,
although absent in almost all members of the family apomictic reproduction is very
common in one genus of the Brassicaceae, the genus Boechera from North America
(Sharbel et al. 2001, Dobes et al. 2006, Dobes et al. 2007). According to the current
taxonomic circumscription it comprises 110 species of which 38 are apomictic.
Apomicts in Boechera are assumed to be of hybrid origin with various combinations
of parents (Sharbel and Mitchell-Olds, 2001, Windham and Al-Shehbaz 2006,
2007a, 2007b). They are pseudogamous and diplosporous which means that
fertilization of the endosperm is still required and the embryo is formed from an
unreduced embryo-sac morther cell (Naumova et al. 2001). Past studies revealed
that Boechera apomicts are facultative apomicts which means that to a variable
extent they can still take part in sexual reproduction and cross back into sexual
populations (Böcher, 1951, Dobes et al. 2006, Schranz et al. 2005). Up to date it is
unclear what effect this behaviour has on the sexual reproducing populations and if
the potential to become apomicitc can be transferred that way.
Apomicts in Boechera are mostly triploid but diploid or aneuploid apomicts also exist
(Sharbel and Mitchell-Olds, 2001, Dobes et al. 2006). The latter were subject to
cytogenetic investigations in the past which revealed the presence of translocations
(one individual, BH115, with a translocation complex visible during metaphase I in
meiosis) as well as one supernumerary small chromosome referred to as Del
(deletion chromosome) in 15 chromosome apomicts and a heteropycnotic
chromosome referred to as Het (heterochromatized chromosome) found in all
investigated apomicts (Kantama et al. 2007).
108
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
Another characteristic feature was that apomicts had different numbers of parental
chromosomes indicating their ability to occasionally reproduce sexually and therebye
recombining their chromosome sets (Kantama et al. 2007).
The Del chromosome was shown to align with another chromosome pair during
meiosis forming a heteromorphic trivalent which suggested a triplication of at least a
portion of the Del chromosome. Its behaviour fitted that of a translocation trisomic.
The Het chromosome formed a heteromorphic bivalent during meiosis suggesting
that at least parts of it are homologous to another chromosome. The Het
chromosome was heteropycnotic in its behaviour meaning that it showed the
opposite behaviour in contraction and extension compared to the remaining
chromosomes in the chromosome set (Kantama et al. 2007). With its heteromorphic
appearance and heteropycnotic behaviour the Het chromosome seemed to be
comparable to the human Y chromosome (personal communication Hans de Jong).
In the past it was described that the supernumerary chromosome in Boechera is a B
chromosome (Sharbel et al. 2004, 2005).
Since chromosomes in Boechera are comparably small (3-6 µm on average) and do
not display obvious features which can be identified by eye differentiation of
chromosomes in phase contrast or after DAPI staining in the fluorescence
microscope is impossible. Therefore it is not known to what other chromosome or to
what part of another chromosome Het and Del are homologous. Since the Het
chromosome has been only identified in apomictic accessions one feels tempted to
hypothesize that the Het chromosome is involved in the expression of apomixis.
However, proof for this is still lacking, but the identification of the origin of the Het
chromosome would be a first step into the direction of revealing its role in apomixis if
any. Since the Del chromosome is also exclusively found in apomicts the
identification of its origin would be helpful for the characterization of aneuploid 15
chromosome apomicts in Boechera.
For the identification of the origin of Het and Del as well as the detection of further
transloctions we chose a BAC-FISH approach. The genome of the Brassicaceae is
arranged in genome blocks. This genome blocks are syntenic arrays of genes which
are shuffeled as blocks that occur throughout the Brassicaceae family (Schranz et al.
2006). Genetic studies of Boechera stricta indicated the most likely position of
genome blocks in the Boechera genome (Schranz et al. 2007). With repeat free
Arabidopsis thaliana BACs repesenting the genomic blocks we aim to identify (1) the
origin of the Het chromosome, (2) the origin of the Del chromosome and (3) the
presence of translocations within genomes of different diploid and aneuploid
Boechera apomicts.
.
109
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
Materials and Methods
Plant Material
Apomictic and sexual plants were grown in the greenhouse under the natural light
rhythms from April until August. Plants were put into new soil every two months and
fertilized every second week to enhance root growth. The day before root tip
collection the plants were lifted out of their pots to enhance growth of fresh root tips.
Root tips were harvested between 8 and 9 a.m. and stored directly in 2 mM 8hydroxychinolin for 3h at 15 °C. Afterwards they were fixed in Carnoy fixative
(ethanol-acetic aced 3:1) for at least one hour or overnight. Flower buds were
collected between 10 and 11 a.m. and directly fixed in Carnoy fixative.
Details on the origin of the plants and results from previous studies can be found in
Table 1.
Table 1. Accession Details of plant material used in this study
Accession
BH74.40
BH115
ES9
geographical
origin
Ranch Creek,
Granite Co.,
MT, USA
Birch Creek,
Ravalli Co.,
MT, USA
Vipond Park,
Beaverhead
Co., MT, USA
chromosome
number
15
Het
present
yes
Del
present
yes
B. holboellii
chromosomes
4
B. stricta
chromosomes
11
15
yes
yes
5
10
14
yes
7
7
Metaphase and Meiotic Chromosome Preparations
Root tips and flower buds were washed in milli-Q water for 5 min at room
temperature three times and then left in 10 mM citrate buffer pH=4.5 for 5 min to
equilibrate. Root tips were digested in 0.02% enzyme mix (0.3% Cellulase RS, 0.3%
Pectolyase, 0.3% Cytohelicase) in the same citrate buffer for 1-2h at 37°C. Flower
Buds were digested in 0.1% of the enzyme mixture also in citrate buffer as described
above for 1-2h at 37°C. Afterwards the material was stored on ice. One to five root
tips or one flower bud were placed on a slide and excess liquid was removed. Then
the material was tapped with a dissection needle resulting in a homogeneous cell
suspension. A drop of 50% acetic acid was added and the material was stirred
carefully until no cytoplasm was visible any more. Then the slide was incubated on a
hot plate at 45 °C for 1-2 min with occasional careful stirring of the acetic acid
droplet. For fixing Carnoy fixative was dropped around and then on top of the acetic
acid droplet and left for a few seconds before the slide was left to dry on the hot
plate. Quality of the slides was assessed under the phase contrast microscope at 40x
magnification.
110
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
Table 2 Arabidopsis thaliana BACs used in this study and their position in the
Boechera stricta genome derived from to the genetic map by Schranz et al. 2007.
BAC name
T25K16
F21M11
F14J9
F2J6
F11A17
T18C15
F8L10
F13N6
T18I24
T30E166
T25B24
T12P18
F5M15
T23E23
F28L5
T19E23
F12K21
T23K8
T6L1
F14O23
F1B16
F23A5
T4P13
F24K9
MVC8
MOE17
MWL2
F16J10
T6P5
T25N22
K21P3
MIO24
K19P17
MHM17
K9B18
MSL3
K9H21
MVP7
K1F13
K9I9
F7O24
T20D16
F27A10
T19L18
T16B12
F11F19
T2P4
F23E6
T8I3
MBD2
MQO24
K15C23
F10E10
K23F3
genome
block
A
A
A
C
C
C
C
D
D
D
D
D
B
B
B
B
B
E
E
E
E
E
F
F
F
F
F
G
G
G
W
W
W
W
W
X
X
X
X
X
I
I
I
I
J
J
J
J
J
V
V
V
V
V
chromosome
in Boechera
stricta
LG 1
LG 1
LG 1
LG 1
LG 1
LG 1
LG 1
LG 1
LG 1
LG 1
LG 1
LG 1
LG 2
LG 2
LG 2
LG 2
LG 2
LG 2
LG 2
LG 2
LG 2
LG 2
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 3
LG 4
LG 4
LG 4
LG 4
LG 4
LG 4
LG 4
LG 4
LG 4
LG 5
LG 5
LG 5
LG 5
LG 5
BAC name
T10F5
T6B13
F6P23
F19F24
F23N11
F2I9
F5O4
F3C11
MJL14
MMJ24
MFJ20
MTO24
T4A2
T10D17
F14D17
T6H20
T21J18
F3A4
F24M12
F28P10
F24I3
T2O9
F16M2
F5I10
F3D13
TsH3
T27D20
T1J1
T3H13
T5L19
F8L21
F7J8
MHF15
T30N20
T19L5
T21H19
F17K4
F22D1
T32G24
T4C12
T19G15
F5H8
MXA21
MUL8
MNF13
MPK23
F25E4
T20K18
F18A5
T6K21
F1N20
F20O9
F8D20
T5J17
111
genome
block
H
H
H
H
H
K
K
K
L
L
L
L
L
M
M
M
M
M
N
N
N
N
N
O
O
O
O
O
P
P
P
R+Q
R+Q
R+Q
R+Q
R+Q
R+Q
R+Q
R+Q
R+Q
R+Q
S
S
S
S
S
T
T
T
U
U
U
U
U
chromosome
in Boechera
stricta
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 5
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 6
LG 7
LG 7
LG 7
LG 7
LG 7
LG 7
LG 7
LG 7
LG 7
LG 7
LG 7
LG 7
LG 7
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
BAC isolation and Labelling
BACs were grown over night at 37 °C from glycerol stocks in 4 mL LB medium with
either chloramphenicol (12.5 µg/ml LB) or kanamycin (10 µg/mL LB). Isolation of
BACs followed the protocol supplied with the High Pure plasmid isolation kit (Roche
diagnostics).
BACs were labelled block wise either indirectly with Biotin (later to be detected with
avidin-TexasRed or Cy5) or DIG (later to be detected with anti-DIG-FITC) or directly
with Cy3 using nick translation mix supplied by Roche diagnostics (catalogue
numbers 11 745 824 910, 11 745 816 910, 11745808910).
Fluorescence In Situ Hybridisation
The FISH procedure followed essentially the protocol outlined in Lysak et al. 2006.
Changes included the following: the washing step after the hybridisation was done
with 50% formamide in SSC for 5 min in each step at 42 °C and the slide was
incubated with 100 µl TNB for 30 min at 37 °C prior to the incubation with the
antibodies to achieve a better blocking.. For identity of BACs and the block to which
they were assigned refer to table 2.
Results
The Het and Del Chromosome in the 15 chromosome apomict BH74.40
In BH74.40 two copies of chromosomes 2, 3, 4, 5, 6, and 7 were detected and the
pattern of blocks was congruent with that described by Schranz et al. 2007 (figure 1
b-f showing LG2, LG3, LG4, LG5 in two parts). One copy of chromosome 1 was
identified which showed the three blocks A, C and D (figure 1a). The Het
chromosome was easily distinguishable from the other chromosomes by its much
brighter DAPI staining after the FISH procedure and showed signals of block A and C
from chromosome 1. However, the second signal for block D was found apart from
blocks A and C on a very small chromosome (figure 1a).
Three signals for block U from chromosome 7 were detected. The signal for block U
did not occur together with the signal for block D on the small chromosome.
A cartoon of the karyotype of BH74.40 representing the major findings from the FISH
analysis is given in figure 2a.
The Het and Del chromosome in the 15 chromosome apomict BH115
In BH115 the Het chromosome was homologous to blocks A and C from
chromosome 1. Block D was found apart from blocks A and C on the Het
chromosome on a small chromosome (figure 1 g). Figure 2b displays a cartoon of
chromosome pairs 1 to 4.
112
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
The Het chromosome in the14 chromosome apomict ES9
The Het chromosome displayed the signals for block A, C and D from chromosome
1. There was no translocation of block D to a different chromosome (figure 1 h).
Figure 2c displays a cartoon of chromosome pairs 1, 2, 3, 4, 6, and 7.
Figure 1 Examples fluorescent microscopy images of BAC-FISH analyses A BH74.40
LG1, The Het chromosome has only signals for blocks A (green) and C (orange), block
D is located apart from it on a small extra chromosome, the second LG1 shows
signals for blocks A, C and D; B BH74.40 LG2; C BH74.40 LG3; D BH74.40 LG4, E and
F BH74.40 LG5 upper and lower section with overlap; G LG1 BH115; the same pattern
as in LG1 of BH74.40 is seen; H LG1 of ES9; blocks A, C and D are located on the Het
chromosome.
Discussion
Supernumerary chromosomes in Boechera were already described as early as 1951
in the elaborate studies of Tyge Böcher (Böcher, 1951), the cytologist whose name
was later given to the genus into which the Greenlandic Arabis species Arabis
holboellii was transferred (Löve and Löve, 1976). Later studies investigated the
nature of supernumerary chromosomes in 2n + 1 and 3n apomicts closer (Sharbel et
al. 2004, Sharbel et al. 2005). But the findings from those studies were contradictory
when the nature of the additional chromosome, an assumed B chromosome, was
described. In the first publication the additional chromosome was thought to be
heterochromatic while later it was identified as a chromosome of small size. It was
thought that the supernumerary chromosome does not pair with a homologue during
meiosis. A cytogenetic study of Boechera apomicts finally described the
heterochromatic chromosome as the Het chromosome and the small chromosome as
the Del (from deletion) chromosome (Kantama et al. 2007). The Het chromosome
was present in all 2n and 2n + 1 apomicts while the Del chromosome was only
present in the 2n + 1 individuals. Unlike the findings from previous studies both Het
113
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
Figure 2 Drawings of the identified orders of blocks on the linkage groups of Boechera
apomicts in comparison to the genetic map (Schranz et al. 2007). A BH74.40, B BH115,
C ES9, D order and distribution of blocks according to the genetic map
and Del were shown to pair with other chromosomes during meiosis (however, not
shown by painting experiments). The Het formed a heteromorphic bivalent while Del
114
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
was associated with two other chromosomes. Genome Painting experiments
suggested that the Het and the Del chromosome originated from Boechera stricta.
A Boechera stricta origin of the supernumerary chromosome was also assumed from
the B chromosome studies (Sharbel et al. 2004, Sharbel et al. 2005). In those studies
the evidence was taken from the presence of a third group of alleles for the markers
MVI11 and T1B9 which clustered with alleles typical for Boechera stricta. Both
markers are localized on Arabidopsis chromosome 3. Comparing their position with
the genetic map of Boechera stricta and the Brassicaceae genome blocks (Schranz
et al. 2007) both markers should be located in block F on Boechera stricta linkage
group 3 (Brassicaceae genome blocks as described in Schranz et al. 2006).
Our central aim in this study was to elucidate the origin of the Het and the Del
chromosome. Based on the genetic map for Boechera stricta (Schranz et al. 2007)
we applied a BAC-FISH approach.
The Het chromosome was easily identified by its bright DAPI stain. In the 2n +1
apomicts BH115 and BH74.40 the signal for genome block A and C were found on
the Het chromosome. Block D was found to be apart on a suspiciously small
chromosome which we identified as the Del chromosome. In the investigated 2n
apomict ES9 the signals for block A, C and D were present on the Het chromosome.
According to the genetic map of Boechera stricta (Schranz et al. 2007) the blocks A,
C and D are located on Boechera stricta linkage group 1. In the genetic map parts of
block A and C were also located on linkage group 2. However, we used only BACs
which would bind to the portions of blocks A and C on linkage group 1.
From our findings we can conclude that the Het chromosome is homologous to
Boechera stricta linkage group 1 in both 2n and 2n + 1 apomicts. However, in 2n + 1
apomicts the Het chromosome consists only of two out of three blocks (A and C).
The third block, block D, was either translocated to another chromosome which then
was degraded to its small size or it broke off the heterochromatic chromosome,
formed a neo-centromere and became a chromosome by itself. This small
chromosome is the Del chromosome.
Since the Del chromosome was shown to pair with two other chromosomes it is more
likely that it is a fusion to a triplicated centromere than a small chromosome by itself
with a neo-centromere although this possibility cannot be ruled out. Neo-centromeres
formation exists in the plant kingdom (Dawe, 2005) and spontaneous breakage of
heterochromatic chromosomes or heterochromatic regions is common(Sacristan,
1971, McCoy et al. 1982, Fluminhan et al. 1996).
Our knowledge on the nature of the Het and Del chromosome together with the
findings of Kantama et al. 2007 rules out that those two heteromorphic chromosomes
in Boechera are B chromosomes. First of all the Het chromosome is simply a
homologue of LG1/a part of LG1, the only difference being that it is
heterochromatized. The Del chromosome pairs with another chromosome pair at
115
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
meiosis indicating a potential triplication. Pairing with other chromosome is against
the definition of a B chromosome. Furthermore, the statement that the “B
chromosome” seems to be dispensable, which is also a criterion in the definition of B
chromosomes, is not necessarily true. The Del chromosome which was interpreted to
be the B chromosome in Sharbel et al. 2005 simply does not exist in 14 chromosome
apomicts. The block which makes up the Del chromosomes is in place in Boechera
stricta linkage group 1 in 2n apomicts. . In other words, no large scale deletions exist
in these accessions. From these finding we henceconclude that the Het and Del
chromosomes should no longer be referred to as B chromosomes.
In our study we could not find any indication on what caused the
heterochromatinization of the Het chromosome. In Drosophila miranda the formation
of neo-Y chromosomes was investigated (Steinemann and Steinemann, 2008). The
authors came up with a model in which the degeneration of the neo-Y chromosome
resulted from an accumulation of point mutations and massive spread of
transposons, especially retrotransposons. In the Drosophila model transposons could
accumulate because recombination of the X and Y chromosome was suppressed.
The enrichment of transposable elements finally resulted into heterochromatinization.
This Drosophila model can give us an idea of what may have happened in Boechera.
Recombination in the pair of linkage group 1 of Boechera stricta is suppressed for an
unknown reason (Schranz et al. 2007). This is the same situation as in Drosophila.
Past genome painting experiments in Boechera suggested the presence of unique
repeats in the Het chromosome (Hans de Jong, personal communication). Following
the Drosophila model they could be the reason for the heterochromatinization of one
of the LG1 chromosomes in Boechera.
Genome subtraction would give us the possibility to find apomict specific sequences.
Further FISH experiments would enable us to see if they are really located on the Het
chromosome and sequence identity may reveal something about their nature, for
example if they originate from transposable elements.
Apomixis has often been described as a dosage effect (e.g. Grimanelli et al. 1997)
rather than resulting from a particular mutation in a gene. If heterochromatinization
silences genes in one of the copies of LG1 it will also produce a dosage effect. In this
scenario the Het chromosome would have an indirect effect on the expression of
apomixis by halving the amount of a gene product required for sexual reproduction.
References
Bicknell, R. A., Koltunow, A. M. 2004. Understanding Apomixis: Recent Advances
and Remaining Conundrums. The Plant Cell 16: S228-S245.
116
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
Böcher, T. W. 1951. Cytological and embryological studies in the amphi-apomictic
Arabis holboellii complex. Kong Danske Vidensk Selsk, 6: 1–59.
Dawe, R. K. 2005. Centromere renewal and replacement in the plant kingdom. PNAS
102: 11573-11574.
Dobeš, C., Koch, M.A., Sharbel, T.F. (2006) Embryology, karyology, and modes of
reproduction in North American genus Boechera (Brassicaceae): a compilation of
seven decades of research. Annals Miss. Bot. Gard. 93: 517-533.
Dobeš, C., Sharbel, T.F, Koch, M.A. (2007) Towards understanding the dynamics of
hybridization and apomixis in the evolution of genus Boechera (Brassicaceae).
Systematics and Biodiversity 5 (3): 321-332.
Eckardt, N. A. (2003) Patterns of Gene Expression in Apomixis. The Plant Cell, Vol.
15, 1499-1501.
Fluminhan, A. Jr., de Aguiar-Perecin, M. L. R., Dos Santos, J. A. 1996. Evidence for
Heterochromatin Involvement in Chromosome Breakage in Maize Callus Culture.
Ann Bot 78: 73-81.
Grimanelli, D., Hernández, M., Perotti, E., Savidan, Y. 1997. Dosage effects in the
endosperm of diplosporous apomictic
Tripsacum (Poaceae). Sexual Plant
Reproduction 10: 279-282.
Kantama, L., Sharbel, T.F., Schranz, M.E., Mitchell-Olds, T., de Vries, S., de Jong,
H., 2007. Diploid apomicts of the Boechera holboellii complex show large scale
chromosome substitutions and different aberrant chromosomes. Proceedings of the
National Academy of Sciences 104, 14026-14031.
Löve, A. and D. Löve. 1976. Botaniska Notiser 128: 513.
Lysak, M., Frasz, P., Schubert, I. 2006 in Methods in Molecular Biology, vol. 323, p.
173-186: Arabidopsis Protocols, Second Edition. Edited by: J. Salinas and J. J.
Sanchez-Serrano, Humana Press Inc., Totowa, NJ
McCoy TJ, Phillips RL, Rines HW. 1982. Cytogenetic analysis of plants regenerated
from oat (Avena sativa) tissue cultures ; high frequency of partial chromosome loss.
Canadian Journal of Genetics and Cytology 24: 37-50.
Naumova, T.N., van der Laak, J., Osadtchiy, J., Matzk, F., Kravtchenko, A.,
Bergervoet, J., Ramulu, K. S., Boutilier, K. 2001 Reproductive development in
117
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
apomictic populations of Arabis holboellii (Brassicaceae). Sexual Plant Reproduction
4: 195-200
Nybom, H. Apomixis versus sexuality in blackberries (Rubus subgen.Rubus,
Rosaceae) 1988. Plant Systematics and Evolution 3: 207-218.
Nylehn, J., Hamre, E., Nordal, I. 2003. Facultative apomixis and hybridization in
arctic Potentilla section Niveae (Rosaceae) from Svalbard. Botanical Journal of the
Linnean Society 142: 373-381.
Paun, O., Stuessy, T. F., Hörandl, E. (2006) The role of hybridization, polyploidization
and glaciation in the origin and evolution of the apomictic Ranunculus cassubicus
complex. New Phytologist 171: 223-236.
Sacristan, M.D. 1971. Karyotypic changes in callus cultures from haploid and diploid
plants of Crepis capillaris (L.) Wallr. Chromosoma 33: 273-283.
Schranz, M. E., Lysak, M. A., Mitchell-Olds, T. 2006. The ABC’s of comparative
genomics in the Brassicaceae: building blocks of crucifer genomes. TRENDS in
Plant Science 11: 535-542.
Schranz, M. E., Windsor, A. J., Song, B.-H., Lawton-Rauh, A., and T. Mitchell-Olds.
2007. Comparative Genetic Mapping in Boechera stricta, a Close Relative of
Arabidopsis. Plant Physiology 144: 286–298.
Sharbel, T. F., Mitchell-Olds, T. (2001) Recurrent polyploid origins and chloroplast
phylogeography in the Arabis holboellii complex (Brassicaceae). Heredity 87: 59-68.
Sharbel, T.F., Voigt, M.-L., Mitchell-Olds, T., Kantama, L., de Jong, H. (2004) Is the
aneuploid chromosome in apomictic Boechera holboellii a genuine B chromosome?
Cytogenetic Genome Research 106: 173-183.
Sharbel, T.F., Mitchell-Olds, T., Dobeš, C., Kantama, L., de Jong, H. (2005)
Biogeographic distribution of polyploidy and B chromosomes in the apomictic
Boechera holboellii complex. Cytogenetic Genome Research 109: 283-292.
van Dijk, Peter J, Tas, Inge C Q, Falque, M., Bakx-Schotman, T. 1999. Crosses
between sexual and apomictic dandelions (Taraxacum). II. The breakdown of
apomixis. Heredity 83(6):715-721.
118
Dissertation Christiane Kiefer – BAC-FISH Analysis in Boechera Apomicts
Windham, M.D., Al-Shehbaz, I.A., 2006. New and noteworthy species of Boechera
(Brassicaceae) I: sexual diploids. Harvard Papers in Botany 11, 61-68.
Windham, M.D., Al-Shehbaz, I.A., 2007a. New and Noteworthy Species of Boechera
(Brassicaceae) II: Apomictic Hybrids. Harvard Papers in Botany 11, 257-274.
Windham, M.D., Al-Shehbaz, I.A., 2007b. New and Noteworthy Species of Boechera
(Brassicaceae) III: Additional Diploids and Apomictic Hybrids. Harvard Papers in
Botany 12, 251-274.
119
Dissertation Christiane Kiefer – Overview Scientific Contributions
4. Overview of Scientific Contributions
Koch, M., Kiefer, C. 2006. Molecules and migration: biogeographical studies in
cruciferous plants. Plant Systematics and Evolution 259: 121-142.
Koch, M., Kiefer, C., Vogel, J., Ehrich, D., Brochmann, C., Mummenhoff, K. 2006.
Three times out of Asia Minor - the phylogeography of Arabis alpina L.
(Brassicaceae). Molecular Ecology; 15: 825-839.
Koch M., Dobeš, C., Kiefer, C., Schmickl, R., Klimes, L., Lysak, M. A. 2007. Plastidic
trnF (GAA) pseudogene evolution indicates an ancient split in the phylogeny of the
Brassicaceae. Molecular Biology and Evolution 24: 63-73.
Dobeš Ch., Koch M., Kiefer, C. 2007. Apomixis and radiation at low ploidy levels
exemplified in the evolutionary model genus Boechera (Brassicaceae). Regnum
Vegetabile 147: 391-412.
Dobeš C., Kiefer, C., Kiefer M., Koch, M. 2007. SuperNetwork identifies multiple
events of plastid trnF (GAA) pseudogene evolution in the Brassicaceae. Plant
Biology 9: 502-515.
Calonje M., Martin-Bavo S., Dobeš C., Gong W., Jordon-Thaden I., Kiefer C., Kiefer
M., Paule J., Schmickl R., Koch M. A. (2008) Non-coding nuclear DNA marker in
phylogenetic reconstructions. Pl. Syst. Evol. Pl. Syst. Evol. 10.1007/s00606-0080031-1
Schmickl R., Kiefer C., Dobeš C., Koch M. A. (2008) Evolution of trnF(GAA)
pseudogenes in cruciferous plants. Pl. Syst. Evol. DOI 10.1007/s00606-008-0030-2.
120
Dissertation Christiane Kiefer – Contents of DVD
6. Contents supplementary material and raw data on the DVD
included in the thesis
Accessionlist
Accessionlist.mdb
Accessionlist.xls
Pictures Herbarium Specimens
DAO and GH
MO
Raw Data (ordered in sub folders according to accession number)
chloroplast markers
nrDNA
nuclear intron sequences
Sampling_Map_all_Accession.jpg
Supplementary Data Chapter 1
includes all supplementary data mentioned in the text in chapter 1
Supplementary Data Chapter 2
includes all supplementary data mentioned in the text in chapter 2
Supplementary Data Chapter 3
includes all supplementary data mentioned in the text in chapter 3
Thesis_Christiane_Kiefer_2008.pdf
121
Dissertation Christiane Kiefer – Acknowledgements
5. Acknowledgements
A PhD thesis is a big challenge and as in almost every situation in life success
depends on teamwork rather than on being a lone wolf. A number of people have
contributed to the success of my research and also to my personal development
over the past three years. These great personalities shall be mentioned in the
following.
First of all I would like to thank Prof. Marcus Koch (HIP, University of Heidelberg),
who gave me the opportunity to work in his lab prior to my time as a PhD student
and then in this challenging project for my PhD thesis. Prof. Marcus Koch has not
merely been an excellent PhD supervisor but also a friend in tough situations.
I would like to thank Prof. Thomas Rausch (HIP, University of Heidelberg) for being
the second referee. Prof. Thomas Rausch roused my interest in botany and played
a big role in my molecular biological education together with his group.
I would like to express my gratitude towards Prof. Hans de Jong (WUR/NL) and Dr.
Eric Schranz (University of Amsterdam/NL) who made it possible for me to spend
five month in the Netherlands in the framework of a DAAD scholarship.
I would also like to express my gratitude and admiration towards Dr. Ihsan AlShebaz who did this amazing work on the taxonomy of Boechera and innumerable
other Brassicaceae.
Furthermore I would like to say thanx to everyone in the labs of Marcus Koch and
Hans de Jong and Claudia Erbar– you have been great colleagues! Without you it
would have been half the fun...
A special thanx goes to Wilma Lasthuis from Wageningen (Dank je wel... Jij bent
een echt goede vriendin!), Dora Szinay from the Lab of Hans de Jong (take care of
that material, my friend!), Roswitha Schmickl from the Koch Lab (we will always be
the unforgettable Duo Infernale) and Andrea Schott from the Strahl Lab and Astrid
Ledebuhr now a gardener in England (wonderful girls’ nights!!!!)
Another very special thanx goes to Dr. Markus Kiefer (who is completely unrelated
to me) and Dr. Steffen Greiner for many great lunch times, coffee breaks and
computer problem solutions (this goes to Markus) and all the jokes. I can’t imagine
to invent the concept of a portable black hole with anybody else....
And Dr. Peter Sack for being a very good friend... thanx for taking care of me.
Last but not least I would like to mention those persons, who are dearest to me in
my life: my family. These are my parents Erich and Karin Kiefer who made me the
person I am and my brothers Lars-Erich, Mirko and Boris and my aunt and her
family. Tough times tied us together – we are an amazing team!
Together with my family I want to thank Rafael Martinez, who came into my life with
a joke. Thank you for being with me now and ever....
122
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement