i

PHYLOGENY, BIOGEOGRAPHY, AND REPRODUCTIVE BIOLOGY OF THE

COSMOPOLITAN FLOWERING PLANT GENUS STELLARIA L.

by

MATHEW THOMAS SHARPLES

B.A., University of Massachusetts, 2008

A dissertation submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirement for the degree of

Doctor of Philosophy

Department of Ecology and Evolutionary Biology

2019

ii

This dissertation entitled: Phylogeny, biogeography, and reproductive biology of the cosmopolitan flowering plant genus Stellaria L. written by Mathew Thomas Sharples has been approved for the Department of Ecology and Evolutionary Biology

______Dr. Erin A. Tripp

______Dr. Jeffry Mitton

______Dr. Mitchell McGlaughlin

______Dr. Stacey D. Smith

______Dr. William Bowman

Date 4 November 2019

The final copy of this thesis has been examined by the signatories, and we find that both the content and the form meet acceptable presentation standards of scholarly work in the above mentioned discipline.

iii

Sharples, Mathew Thomas (Ph.D., Ecology and Evolutionary Biology)

Phylogeny, biogeography, and reproductive biology of the cosmopolitan flowering plant genus

Stellaria L.

Dissertation directed by Associate Professor and COLO Herbarium Curator Dr. Erin A. Tripp

The flowering plant genus Stellaria L. (Caryophyllaceae; the “starworts”) numbers around 112 species and exhibits a cosmopolitan distribution. To gain familiarity with the

Caryophyllaceae and a broad number of temperate plant taxa overall in the field, I conducted a floristic inventory of the South San Juan Wilderness of southern Colorado, documenting 533 species of mostly native vascular plants present there, including five species of Stellaria.

Research on Stellaria beyond Colorado was then conducted across five continents based on field and museum work, with the proximate goal of forming a comprehensive phylogenetic hypothesis of evolutionary relationships within this and other flowering plant lineages based on RADseq data. This research revealed the existence of five clades of a core genus Stellaria, offered clarification of taxonomic boundaries of starworts, provided evidence for resurrection of two plant genera related to Stellaria (Adenonema and Mesostemma), and also provided evidence for description of two new plant genera, Nubelaria and Rabelera of Eurasia. Circumscription of core

Stellaria allowed for addressing of downstream evolutionary hypotheses. Namely, a well- sampled phylogeny of Stellaria enabled the biogeographical history of starworts to be reconstructed as well as the history of petal evolution. Analyses revealed a Miocene origin of

Stellaria likely in northeastern Eurasia with myriad dispersals and radiations across the world subsequently, and also suggested widespread habitat and climate lability in the genus. Data also iv revealed numerous, recurrent losses of petals in unrelated starwort lineages, and a field experiment on the alpine tundra of Niwot Ridge (Boulder County, Colorado) suggested that these losses may be associated with transitions from outcrossing to self-fertilizing pollination systems.

v

ACKNOWLEDGMENTS

This work could not have been completed without the assistance from a great many people and institutions. Curator of COLO Erin Tripp encouraged this work from Day One and therefore holds my utmost and foremost gratitude. The remainder of my Ph.D. committee comprised of professors William Bowman, Jeffry Mitton, Mitchell McGlaughlin, and Stacey

Smith (as well as her lab members) provided extensive helpful feedback and other assistance throughout this dissertation, as did Nolan Kane. Stephanie Mayer served as both a teaching and life mentor throughout this dissertation. Early-era lab members Yi-Hsin Erica Tsai, Heather

Stone, and Vanessa Díaz, as well as late-era lab members Louise Barton, Matthew Schreiber,

Carly Anderson Stewart, Sophia Warsh, Justin Williams, and Yongbin Zhuang are particularly thanked for edifying discussions and other assistance over the years. Tyson Burch also provided early mentorship. A magnificent array of botanists and herbaria hosted and helped me throughout this work, including but not limited to ALTB (Alexander Shmakov, Sergei Smirnov), CANB

(Cathy Miller), CAS (Tom Daniel, Rebecca Peters, Debra Trock), CHR (Ines Schönberger),

COLO (Ryan Allen, Dina Clark, Tim Hogan), E (Leslie Scott), IBSC (Sunan Huang, Deng

Yunfei), K (Marie Briggs), KUN, LE (Galina Konechnaya, Roman Ufimov, Marina Yarichina),

MEL (Pina Milne), MO (Jim Solomon), NSK (Andrey Erst, Dmitri Shaulo), NY (Jackie

Kallunki, James Lendemer), P (Vanessa Invernon), PE, UC (Kim Kersh), VBGI (Evgeny

Boltenkov, Valentina Kalinkina), and VLA. Funding was generously provided by the American

Society of Plant Taxonomists, the Botanical Society of America, the Department of Ecology and vi

Evolutionary Biology at the University of Colorado (Boulder), a University of Colorado

(Boulder) Cynthia H. Schultz Graduate School Small Grant, the University of Colorado Museum of Natural History, and the University of Colorado Undergraduate Research Opportunities

Program. Finally, I wish to thank my parents and sister for being fully supportive of my endeavors across the years.

vii

CONTENTS

INTRODUCTION...... 1

CHAPTER

I. VASCULAR FLORA OF THE SOUTH SAN JUAN MOUNTAINS (COLORADO, U.S.A.): A FLORISTIC INVENTORY OF TWO SOUTHERN ROCKY MOUNTAINS SLOPES

Introduction...... 3

Methods...... 22

Results and Discussion...... 24

Checklist of the Vascular Flora of the South San Juan Mountains...37

II. Part I: PHYLOGENETIC RELATIONSHIPS WITHIN AND DELIMITATION OF THE COSMOPOLITAN FLOWERING PLANT GENUS STELLARIA L. (CARYOPHYLLACEAE): CORE STARS AND FALLEN STARS

Introduction...... 99

Materials and Methods...... 104

Results...... 116

Discussion...... 126

Revised Taxonomic Concepts...... 141

Part II: TAXONOMIC OBSERVATIONS WITHIN STELLARIA L. (CARYOPHYLLACEAE): INSIGHTS FROM ECOLOGY, GEOGRAPHY, MORPHOLOGY, AND PHYLOGENY SUGGEST WIDESPREAD PARALLELISM IN STARWORTS AND ERODE PREVIOUS INFRAGENERIC CLASSIFICATIONS

viii

Introduction...... 152

Materials and Methods...... 153

Results and Discussion...... 157

III. BIOGEOGRAPHIC AND ECOLOGICAL HISTORY OF THE COSMOPOLITAN ANGIOSPERM GENUS STELLARIA L. (CARYOPHYLLACEAE)

Introduction...... 183

Materials and Methods...... 193

Results...... 201

Discussion...... 216

IV. THE EVOLUTION OF PETAL LOSS IN THE COSMOPOLITAN GENUS STELLARIA L. (CARYOPHYLLACEAE)

Introduction...... 233

Materials and Methods...... 235

Results...... 242

Discussion...... 251

BIBLIOGRAPHY...... 261

APPENDICES AND SUPPLEMENTARY FIGURES...... 285

ix

TABLES

Table

1. Number of species in each family in the South San Juan Mountains....25

2. The eight largest genera in the South San Juan Mountains...... 26

3. Number of families and species in each Division in the South San Juan Mountains...... 26

4. Number of species and percentage of the flora in each distribution type in the South San Juan Mountains...... 30

5. A generalized classification of distribution types in the South San Juan Mountains...... 30

6. List of core Stellaria plus outgroup accessions sampled, including specimen ages and data yields...... 105

7. Estimate of number of core Stellaria species inhabiting major geographical regions as well as total global Stellaria...... 111

8. Known clades of core Stellaria...... 158

9. Stellaria species sampled for biogeographical analysis...... 202

10. BioGeoBEARS model comparisons...... 209

11. Ancestral state reconstruction of climate occupation and habitat results in Stellaria...... 213

12. Field data collected for pollination biology of S. irrigua and S. longipes...... 243

13. AIC weights of four ancestral state reconstruction models of petal evolution in Stellaria...... 247

x

FIGURES

Figure

1. Map of study area in the South San Juan Mountains...... 5

2. Wolf Creek Pass climate summary...... 9

3. Volcanic tundra of the South San Juan Mountains...... 13

4. Forest types of the South San Juan Mountains...... 14

5. Morphological diversity of some Stellaria species...... 100

6. Map of localities of Stellaria specimens sampled...... 117

7. Phylogenetic hypothesis across select members of Caryophyllaceae based on RAD loci...... 119

8. Phylogenetic hypothesis of relationships within core Stellaria...... 120

9. Phylogenetic hypothesis of relationships within the Larbreae clade of core Stellaria...... 121

10. Scatter plot of relationship between specimen age and number of RAD reads in Caryophyllaceae...... 125

11. Phylogeny of core Stellaria...... 155

12. Phylogeny of the Larbreae clade of core Stellaria...... 156

13. Some habitat diversity in Stellaria...... 184

14. Worldwide map of number of Stellaria species in each region...... 187

15. Morphological diversity of Stellaria in Oceania...... 191

16. Divergence time estimates across select Caryophyllaceae...... 204

17. Divergence time estimates across Stellaria...... 206

18. Ancestral area reconstruction of Stellaria...... 210 xi

19. Ancestral reconstructed ecology of Stellaria...... 214

20. Study area hosting two sympatric species of Stellaria on the tundra of Niwot Ridge...... 237

21. Fertility of Stellaria irrigua and S. longipes at Niwot Ridge...... 244

22. Experimental field manipulations of S. longipes and S. irrigua...... 245

23. Bud pollination in Stellaria irrigua...... 248

24. Ancestral reconstruction of petal evolution in Stellaria...... 249

1

INTRODUCTION

The study of plant diversity is amongst the most ancient of human pursuits (Weber and

Wittmann, 1992). Central to understanding plant diversity is the study of quantifying which plants occupy a given landscape, known as “floristics”. A floristic inventory provides a baseline dataset for what species inhabited a given place during a given point in time, and such knowledge is central to fields such as ecology, evolution, conservation, and ethnobotany, whose studies all rely on robust knowledge of what species live where and when. Across the summers of 2013, 2014, and 2016 I compiled such a floristic inventory for an understudied mountain range of southern Colorado, the South San Juan Mountains. This inventory provides baseline knowledge of the species inhabiting the area in contemporary times and provides dry plant specimens for subsequent researchers with myriad interests.

Central to understanding plant diversity in a modern context is the reconstruction of evolutionary relationships amongst organisms and the proper naming of those organisms, known together as systematic biology. Systematic biology is possible only because floristic inventories have provided the baseline requisite knowledge of what plant species live where. One flowering plant genus that has been collected widely through floristic inventories and other efforts, but which lacks a robust hypothesis of evolutionary relationships within the genus, is Stellaria in the

Carnation family (Caryophyllaceae). Worldwide sampling of field sites and museum collections of this genus and relatives enabled me to construct a robust hypothesis of evolutionary relationships within Stellaria based on ca. 92% of species sampling in the genus at present. This 2 comprehensive sampling approach identified for the first time major clades within Stellaria, resulted in a number of taxonomic changes in Stellaria, and provided evidence for description of two new genera, Nubelaria M.T.Sharples & E.Tripp and Rabelera M.T.Sharples & E.Tripp.

Once evolutionary relationships have been reconstructed within a group, downstream evolutionary questions may be pursued. One use of phylogenetic hypotheses is to reconstruct ancestral states across evolutionary history to better understand how traits and other aspects of a group evolve through time. Nearly complete phylogenetic species sampling of Stellaria allowed for particularly robust reconstruction of the evolutionary history of geographical areas as well as aspects of its ecology: its propensity for cold environments and its flexibility in inhabiting wet or dry habitats. Stellaria is one of the most widespread and cosmopolitan of all plant genera and has alternatively been hypothesized as originating in the broad Himalaya region or in the southern hemisphere of the Old World. We conducted divergence time estimation on the phylogeny of

Stellaria to date the timing of major biogeographical events and found diversity in the broad

Himalaya to be primary derived from a surprising context. We also found ecological transitions to be rampant across the history of the genus and inferred multiple, recent origins of Arctic species from colder southern areas.

Across its worldwide distribution, Stellaria also exhibits intriguing variation in petal morphology. While most species exhibit showy white petals, others do not produce petals or alternatively produce reduced or malformed petals. As above, our phylogenetic hypothesis of the genus allowed us to reconstruct the history of petal evolution across the genus, and we recovered a surprising pattern of changes in petals across the world. Field experimental data provided a complement towards helping explain why such changes may be occurring in markedly different localities. 3

CHAPTER 1

INTRODUCTION

Biodiversity inventories are imperative to conduct as soon as possible across all

Kingdoms of life and throughout the planet, due in large part to the dramatic nature of habitat loss induced by various pressures of the Anthropocene (Raven and Wilson, 1992). Botanical inventories are a significant form of such inventories, as the vascular plant communities present in a region often define what other biota exist in that landscape. Despite a rich history of botanical exploration and a lower amount of biodiversity than in more tropical latitudes, North

America north of Mexico is still generating many important baseline floristic inventories (Ertter,

2000). Such inventories are still documenting new species in North America annually, but perhaps more critically, botanical inventories produce numerous other scientific resources. These include providing a baseline reference collection for a given area during a given time, contributing distributional and morphological knowledge for regional and other floras, providing raw specimens for systematic study, and otherwise contributing to natural history collections in innumerable ways (Ertter, 2000; Funk, 2004; Willis et al., 2008).

The South San Juan Mountains comprise the southern subset of the Eastern San Juan

Mountains of southern Colorado (Fig. 1, inset). Within these mountains lies the South San Juan

Wilderness (henceforth: Wilderness), which was designated as Wilderness by Congress in 1980

(Fig. 1, white line boundaries). The Wilderness is a relatively vast and wild region that harbors 4 large swaths of road- and trail-less terrain and has only a handful of proximate human settlements: the villages Jasper, Platoro, Chromo, and the sizable town of the area, Pagosa

Springs. Despite Colorado’s rich history of botanical exploration, the South San Juan Wilderness region has remained underexplored botanically. This is likely attributable to the area’s great distance from major metropolitan areas, the loose, unstable, and rugged volcanic rock terrain, and the area’s relative inaccessibility as assured through its Wilderness designation.

Although a floristic inventory of all vascular plant species in the South San Juan

Wilderness region has never been completed, two previous floristic inventories sampled some portions of the Wilderness (Douglas, 1992; Flaig, 2007). In 1992, a floristic inventory of the

Conejos River basin of Conejos county was completed (Douglas, 1992). The headwaters and middle-upper elevations of the Conejos River drainage overlap with some of the eastern part of the study area extent presented here. In 2003 and 2004, inventory work across all of the eastern

San Juan Mountains was completed by Flaig (2007), but this inventory did not extensively catalogue the South San Juan Wilderness itself. Lastly, Heil and colleagues collected in parts of the San Juan River drainage included in the study area as part of a regionwide Four Corners

Flora treatment throughout the 1990s and 2000s (Heil et al., 2013). The present study therefore builds the most comprehensive inventory of the South San Juan Wilderness flora to date. As with any floristic survey conducted by a lone individual, both areal and species gaps inevitably exist. However, attempts to mitigate botanical gaps were achieved by conducting extensive herbarium database searches for additional taxa in the study area, and these taxa are mostly attributable to the inventory efforts of the three studies mentioned above combined with the efforts of their colleagues.

Geographical Extent 5

Fig. 1. Map of the South San Juan Mountains. Blue lines represent major highways delimiting the study area. The study area’s southern extent is delimited by the New Mexico border, the southernmost horizontal black line. Blue lines follow the tracks of Highways 84 and 160 on the western slope, and Highway 17 on the eastern slope. The red line on the eastern slope approximately follows Rio Grande National Forest Road 250. White lines within the blue and red lines represent the South San Juan Wilderness boundary. The black line vertically bisecting the map depicts the Archuleta (left) and Conejos (right) county border, and this border generally follows the Continental Divide ridge. The upper horizontal black lines delimit Mineral (left) and

Rio Grande (right) counties. Red points indicate Sharples’ collection localities. Green shading is

National Forest Land, white is private land, orange shading is Bureau of Land Management land, and purple shading is Native American land. The inset shows the study area location within

Colorado.

6

The study area comprises approximately 158790 acres (=64260 hectares, 248 square miles) in terms of land protected by the boundaries of the South San Juan Wilderness. The

Wilderness was the specific area most thoroughly explored and botanized (white lines, Fig. 1).

The total study area including land inside and outside of Wilderness comprises approximately

480000 acres (=194400 hectares, 750 square miles), including private lands which nonetheless remain largely-unexplored botanically (e.g., the extensive Tierra Amarilla Land Grant lands in white in the southern portion of the study area map). Elevations range from 7200 ft (2195 m) on the western slope near the junction of the Rio Blanco with U.S. Highway 84 to 13307 ft (4056 m) atop Summit Peak. All taxa in the checklist are found east of U.S. Routes 84 and 160 in

Archuleta county, west of Rio Grande National Forest Service Road 250 in Conejos County, as well as west of Colorado State Highway 17 south of where it meets Forest Road 250 (Fig. 1).

The northern boundary of the study area is Bonito Pass just south of Wolf Creek Pass and just north of the Wilderness boundary, east to Summitville, and along Wightman Fork to its junction with the Alamosa River and with US Forest Service Road 250, and the southern boundary is the

Colorado border with New Mexico. The only land included outside of areas in Archuleta or

Conejos County as demarcated by the features and roads listed above is the small portion of the

South San Juan Wilderness itself that pokes into Mineral and Rio Grande counties. Few taxa (27) collected under Sharples’ numbers just outside of these demarcations, yet found by other investigators’ records to reside within these demarcations, are both included under Sharples’ collection numbers and are accompanied by a record from another investigator in the checklist.

This includes areas just west of U.S. Route 84 and habitats along the lower Alamosa River valley

(Fig. 1). Notably, the study area defined here does not include the southern/eastern San Juan

Mountains of southwestern Rio Grande County. 7

The South San Juan Mountains include some of the southernmost extent of the Southern

Rocky Mountains, a distinct subregion of the Rocky Mountains Floristic Province (Takhtajan,

1986). The Southern Rocky Mountains extend from the Medicine Bow Mountains of southeastern Wyoming in the north to the Sangre de Cristo mountains of northern New Mexico in the south. The western extent of the Southern Rocky Mountains is often delimited as the La

Sal Mountains in far eastern Utah, and/or the San Francisco Peaks of Northern Arizona, a notable disjunct station of the Southern Rocky Mountains flora. Notably, the South San Juan

Mountains include the southernmost high points of the North American portion of the

Continental Divide. The Wilderness protects ca. 42 miles of the Continental Divide’s alpine, subalpine and montane habitat.

The South San Juan Mountains are bound by San Luis Park to the east, Wolf Creek Pass and the continuation of the San Juan chain along the continental divide to the north, the San Juan

River valley to the west, and the New Mexico border with Colorado to the south; the Continental

Divide abruptly drops into a featureless undulation at the New Mexico border. The South San

Juan Wilderness encompasses the high mountain core of this broad region. A large portion of the

South San Juan Wilderness’ protected area is found east of the Continental Divide (Conejos

County Wilderness portions, Fig. 1). To the south, the Wilderness boundaries form

“pseudopodia” around a significant section of private land known as the Tierra Amarilla Land

Grant, which is further subdivided into various hunting ranches and at times has provided novel floristic records (both exotic and native) not found elsewhere in the South San Juans.

Major drainages of the study region are Park Creek, the Alamosa River, the Conejos

River, Rio Chama, Rio Blanco, and the Navajo River. Only the Rio Blanco and Navajo rivers flow into the San Juan River; the other four drainages empty into the Rio Grande. Only one of 8 these drainages, the Conejos, has been comprehensively botanized before (Douglas, 1992).

Thirty-two lakes exist within the Wilderness, all part of the headwaters of these drainages

(USDA, 2017a).

Modern Climate

The weather station at 10640 feet on Wolf Creek Pass is near the northern end of the study area and its data best represent the climate of the subalpine zone of the region (National

Climatic Data Station #059181-5; WRCC; http://www.wrcc.dri.edu/cgi-bin/cliMAIN.pl?co9181; accessed Jan 2017). The climate of the region from 1957–2001 was characterized by frigid winter temperatures (average 18.13°F, Dec–Feb) and mild summer temperatures (average

50.9°F, Jun–Aug), with cold autumns (average 35.47°F, Sep–Nov) and springs (30.4°F, Mar–

May) (Fig. 2). The average yearly temperature of the pass was 33.73°F (1°C), and only the months of June–September maintained an average low temperature above freezing. In terms of precipitation, the area is characterized by occasional heavy winter snowfalls as well as regular monsoonal rains and thunderstorms after June and throughout the summer. The average wettest month for 1957–2001 was March, when the area is prone to heavy spring snowstorms, and the average driest month was June, before the regional thunderstorm season begins (Fig. 2). The average yearly precipitation from 1957–2001 was 45.39 in/115.3 cm (high of 58.3" and a low of 29.62"), with the majority of the moisture falling as snow. PRISM data report a range of 57–

140 cm/year in the study area, seemingly, but not exclusively, across an elevation gradient from low to high (Blair and Bracksieck, 2011).

Orogeny and Glacial History

The Southern Rocky Mountains began attaining their current elevation throughout the 9

Fig. 2. Climate summary of the late twentieth century at Wolf Creek Pass near the northern end of the study area. Data provided by the Western Regional Climate Center (WRCC, 2017).

10

Paleogene (ca. 66 to 23 million years ago). Prior to the Paleogene, Colorado was part of an intercontinental seaway, and uplift subsequently helped lead to the dissolving of the seaway during the Laramide orogeny (early Paleogene) (Gonzales and Karlstrom, 2011). By the end of the Eocene (ca. 56 to 34 million years ago), vulcanism had spread throughout the southwestern quadrant of Colorado and small parts of northern New Mexico, and thus, the modern South

San Juan Mountains are volcanic in origin, sharing a common orogenetic history with the greater

San Juan Mountains. The study area is near the southern limit of this San Juan Volcanic Field, although the western limits of the study area start reaching into sedimentary rock dating prior to vulcanism in the area (Lipman and McIntosh, 2011). Vulcanism and continued uplift further heightened the ancestral San Juan Mountains during the Laramide orogeny (Blair and Gillam,

2011).

Four composite volcanos were active in the South San Juan Mountains around 35 to 30 million years ago (Blair and Bracksieck, 2011). They formed andesite across much of the west and north part of the study area. Much of the the rest of the study area (particularly parts of the eastern portion) is geologically represented by newer (30 to 23 million years ago) ash

(ignimbrite) and lava (basalt) flows that formed atop the base layers of andesite (Lipman and

McIntosh, 2011). Most of the study area thus has Oligocene volcanic rock as its substrate.

Two ancient calderas are found in the northeastern portion of the study area: the Platoro and

Summitville Calderas. The former encompasses about 25 kilometers in diameter, is composed of intrusive rock (ancient cooled magma), and is centered around the Platoro town and reservoir; the latter is embedded within the former, and is 10 kilometers in diameter (Bethke, 2011). These calderas were active from around 33 million years ago to 28.5 million years ago (Lipman and

McIntosh, 2011). Much of the current landscape of the South San Juans represents the eroded 11 and eroding remnants of this ancient, uplifted volcanic activity, further dramatized by the subsiding of the San Luis Valley as the Rio Grande Rift formed (Blair and Gillam, 2011).

Glaciers are no longer present in the South San Juans, and it is unclear when the first

Pleistocene glaciations affected these mountains. There is little evidence of glaciation in the entirety of the San Juan Mountains prior to approximately 650000 years ago (Blair and Gillam,

2011). The last glaciation in the region, ending around 20000 years ago, covered the ridges, high plateaus, and valleys of the eastern slope of the study area, likely explaining why the high plateaus are covered in rich wetlands (Blair and Gillam, 2011). Most of these wetlands are pristine and harbor otherwise rare or absent taxa in the study area. Glaciation was sparse on the west side of the Continental Divide, only present right up against the Divide itself. The Chalk

Mountains subrange of the western slope was too low for glaciation (Blair and Gillam, 2011).

The Chalk Mountains harbor no alpine tundra habitat at present, with a highpoint of only 12,037 feet and with its ridges forested.

Habitat Diversity of the South San Juan Mountains

There are several general common habitat types in the South San Juan Mountains, ranging from lowland woodlands to volcanic tundra (Figs. 3–4). The range of approximate extent of each habitat is given when appropriate. Elevations are given as approximates, as elevation-based habitat boundaries differ depending on whether they are found on the eastern vs. western side of the Continental Divide, whether they have a north vs. south aspect, and numerous other microclimatic factors, and therefore habitat boundaries inevitably intergrade and vary.

While a few taxa are here mentioned explicitly, see the checklist for some more details regarding which plants are common in which habitats. 12

Ponderosa-Oak Forest.—Much of the lower elevations of the western slope are covered in this high diversity habitat, but only a small portion is protected within the boundaries of

Wilderness. This forest type often has a savannah quality to it, with individual trees of Pinus ponderosa interspersed throughout, yet rarely forming dense, dark stands. Quercus gambelii is a dominant small oak tree found in the ponderosa pine understory and forest openings of the western slope, and it occasionally forms large stands of bright forest without any neighboring ponderosas. Many Eurasian aliens are found along this habitat’s roadsides, and indeed within some of the forest understory. This habitat intergrades with the next at elevations in the mid- to upper 8000s (feet), but may stretch above 9000' in pockets. Pinus ponderosa forests are found at the lowest elevations of the study area’s eastern slope, but Quercus gambelii is conspicuously absent on the east side of the Continental Divide, and very little ponderosa pine forest habitat is found within the Wilderness of the eastern slope.

Montane Mixed Forest.—Much of the region’s woody plant diversity is found in this habitat, found above yet intergrading with the ponderosa pine zone, and found below and intergrading with the subalpine spruce zone. Populus tremuloides and Pseudotsuga menziesii are some of the more conspicuous tree species living in this zone, and Acer glabrum, Actaea rubra,

Aquilegia elegantula, Mahonia repens and Rubus parviflorus are a few of the more common forest understory species. Populus tremuloides sometimes forms dense swaths of clonally formed forest and can dominate large areas of elevations ranging from the upper 8000s to the lower

10000s. Truly mixed forests with codominance of many tree species (Abies spp., Picea spp.,

Pinus spp. in addition to Populus tremuloides and Pseudotsuga menziesii) are most common below ca. 10000' on the western slope yet above the ponderosa-oak forest.

Subalpine Spruce Forest.—This is the dominant habitat type above 10000' and below the 13

Fig. 3. Volcanic tundra adjacent to the Continental Divide.

14

Fig. 4. Some forest types of the South San Juan Mountains. Top left: spruce-dominated subalpine forest; Top right: subalpine mixed forest; Bottom right: ponderosa-oak forest; Bottom left: aspen-dominant forest.

15 treeline. Picea engelmannii is the pervasive tree species in this habitat, occasionally mixed in with patches of Populus tremuloides or Abies spp. It is one of the lowest plant diversity habitats found in the region, with a rather barren understory present in the drier spruce forests. Extensive areas of subalpine spruce understory are otherwise often covered by continuous stretches of

Vaccinium myrtillus, Aconitum columbianum, Arnica cordifolia, Delphinium barbeyi,

Pedicularis racemosae, and Polemonium pulcherrimum. Some patches of subalpine forest are exceptionally moist, with pyroloid Ericaceae preferring the understory of the wettest spruce forests.

Barren Montane Slopes.—Such habitats are dry (well-drained and with bare sedimentary soils) and exposed, with sparsely-vegetated slopes occurring intermittently below the middle subalpine elevations (ca. 11000'), and are commonly found on south-facing slopes. Agastache pallidiflora, Aliciella pinnatifida, and Penstemon barbatus are notable entities more or less restricted to such habitats.

Lowland Riparian Zones.—This habitat encompasses the borderland areas of the various rivers and their tributaries (see “Geographical Extent”) of the study area as well as more stagnant wetlands such as ponds, marshes, and bogs, generally up to ca. 10000'. Alnus incana, Carex spp.,

Juncus spp., Potamogeton spp., and Salix spp. are some common members of the low elevation wetlands.

High Elevation Riparian Zones.—The most common riparian areas near and above treeline are the snowfed streamlets, along which Cardamine cordifolia, Rhodiola rhodantha,

Rorippa spp., Salix spp., and Swertia perennis can often be found. Small ponds are found interspersed throughout the alpine, but these support low plant diversity. Below treeline, both ponds and streamsides become increasingly more diverse with lower elevation. 16

Highland Scree & Talus.—This habitat encompasses rockfields from the lower subalpine to the top of Summit Peak. Many taxa are shared between the subalpine and alpine versions of these habitats, but some taxa are only found truly above treeline. Talus constitutes largely unvegetated fields and slopes of medium to large boulders, while screes constitute slopes and ridges of small, loose rock fragments and usually yield higher plant diversity than talus. Alpine screes of San Juan Volcanic Field origin are floristically unique in these mountains. Angelica grayi, Claytonia megarhiza, Ligularia soldanella, Penstemon hallii, Rhodiola integrifolia, and

Trisetum spicatum inhabit barren, rocky highland areas. Most such rockfields are found above

11000', and they become increasingly dominant across the landscape on alpine slopes above

12000' close to the Continental Divide.

Stable Tundra.—Stable, grassy tundra begins in openings of the krummholz of Picea engelmannii. Krummholz can end as low as 12000', or as high as nearly 12300', depending on aspect and other factors contributing to landscape heterogeneity. Where the treeline ends, stable tundra habitat is contiguous along the Continental Divide in the Wilderness and extends as high as the apex of Summit Peak. Scree and talus also share this distribution, and the two habitats are intermixed along the Divide. Geum rossii dominates wide expanses of green, well-drained areas of the tundra, as do many members of the Poaceae.

Persistent Snowfields.—Late-summer snowfields occur on the tundra and are found on steep, often northfacing (or otherwise obliquely-angled) slopes through July, usually August, and sporadically perennially. Ranunculus macauleyi is endemic to the barren, sopping ground adjacent to such snowfields, at least in the South San Juans.

Montane Meadows.—Meadow habitats generally below 10500' in elevation, where

Dasiphora fruticosa comprises the dominant shrubby vegetation and where herbaceous diversity 17 is high, are found sporadically in the South San Juans. These meadows tend to dry out in summer. Expanses of this habitat are especially present in the lower river valleys of the southeastern part of the study area.

Dry Subalpine Meadows.—This habitat is found in small to large sections of forest openings and open river valleys above 10000' and up to the treeline. Such meadows floristically intergrade with tundra at their upper limits. These meadows are renowned across the Southern

Rocky Mountains for their picturesque displays of charismatic and diverse flowers. Castilleja spp., Erigeron spp., Potentilla spp., Senecio spp., and Veratrum californicum are common in such meadows.

Wet Subalpine Meadows.—Marshy subalpine meadows are extensive in Conejos County, particularly on the windswept plateaus extending eastwards from the main Continental Divide ridge. Much ground on these plateaus was formerly covered by glaciers (Blair and Gillam,

2011). Sedges and rushes dominate here, with muddy depressions and boggy ponds common throughout. Plants not occurring elsewhere in the study area are found in the diverse wetlands of this region. Particularly exemplary are the vast meadows surrounding Red Lake in the Conejos part of the Wilderness. Eriophorum spp. are found solely in such habitats.

Multi-Habitat Generalists.—While best described as a life history strategy rather than a habitat type, some plants are very successful habitat generalists. These tend to occur in both enclosed forests and open meadows across an elongate elevational gradient and across different types of disturbance. Achillea millefolium and Campanula rotundifolia are noticeably common throughout elevations and habitats.

Natural Disturbance 18

Historical fire intervals and historical fire intensities of forests in the South San Juans vary according to elevation and the concomitant forest type found at various elevations. Human repression of fire has altered these historical regimes, and reported fire intervals refer to pre-

European settlement patterns. The ponderosa-oak and lower mixed forests are subject to frequent events of low-intensity understory conflagrations, with a mean fire interval of every 3 to 11 years

(Korb and Wu, 2011). Historically savannah-type vegetation in the ponderosa-oak forest prevented high intensity canopy fires, but European settlement and subsequent fire repression have built up high fuel loads in this forest type which will result in more extreme fire intensities someday. This phenomenon also applies to the lower mixed forests, which are dense forests in the region at present; historically they maintained more of a savannah aspect. Aspen forests burn relatively infrequently due to their typically moister character and the low flammability of aspen tissues, but they do experience episodic high intensity burns every 70–100 or more years (Korb and Wu, 2011). Aspens quickly regenerate after moderate-intensity fire due to their clonal- rhizomatous nature. Upper mixed forests and subalpine spruce-fir forests burn most infrequently of all the forest types found in the South San Juan Mountains. When they do burn, though, it is a product of both multiple years of drought and decades of build-up of fuel load. These fires are thus high-intensity crown fires and occur every two or more centuries, on account of their ordinarily wet boreal climate (Korb and Wu, 2011). The subalpine fire regime is the only one that appears to be unaltered at present, meaning that burn frequency and intensity appear to be similar to what they were before European settlement. Historically, the higher mixed forests could experience moderate severity fires every several decades (14–63 years), or high intensity crown fires once every multiple hundreds of years, as these forests are transitional between the warmer, lower mixed forests and the subalpine boreal forests (Korb and Wu, 2011). The highest 19 subalpine boreal forests burn only two or three times per millenium. They are naturally dense and are not subject to frequent surface fires. Understory snow is still present into June during most years in subalpine forests, and the summer is usually characterized by consistent monsoonal rains before the dry autumn cools the weather down significantly, warding off fire throughout the year. Essentially, the South San Juan forests pattern is: the higher the elevation, the longer the mean fire interval, and the higher the intensity of a given burn when it does occur.

There is extensive spruce beetle-kill on the eastern slope of the study area at present, and currently available public USDA Forest Service aerial survey maps suggest widespread spruce beetle kill of tens of thousands of acres across subalpine forest throughout the Wilderness since

1995 (USDA, 2017b); many expanses of the Conejos subalpine zone in particular are becoming subalpine meadows as a result. In the worst-afflicted areas, fewer than 20% of the spruce trees appear to still be living in a given acre, though seedling establishment still seems to be strong

(personal observation). Aspen forests in the region have been affected by bark beetles and borers as a consequence of regional drought, and have experienced some decline (Korb and Wu, 2011).

Historical ranges of variability in regards to extent of booms and busts of such forest pest outbreaks are unknown, and thus it is unclear how abnormal the situation currently is.

Winds can play a role in influencing forest structure in the South San Juans, as anywhere.

During the field seasons in which the vascular flora checklist was compiled, a large subalpine forest blowdown was present on the eastern slope of the range in the vicinity of Victoria Lake.

Korb and Wu (2011) put forth the idea that subalpine forests are disturbed more often by winds and insect infections than by their rare fire events. Treefall is also influenced by the impact of heavy blizzards in the San Juans. Flooding and mudslides are possible sporadically in the study area after heavy precipitation events during the monsoon season. Evidence of mudslides can be 20 seen along the road to the Fish Creek trailhead and in scattered parts of the western and eastern slopes.

Human Presence and Impacts

The Summitville Superfund site has had the most pernicious environmental impact of all human presence in the region. A former mine sitting near treeline at 11500', its open pits have resulted in extensive metal leaching into the upper portions of the Alamosa River drainage

(Bethke, 2011). Summitville was an underground mining operation from the late 1800s until the

1980s, when open pits were excavated to extract gold and silver from low-concentration ore with cyanide solutions. The ground water was infiltrated by the cyanide solutions, and the surface water of the Wightman Fork was also contaminated on multiple occasions. Cyanide is not a long-term contaminant, but more serious contaminants of surface water, groundwater, and soils through acid mine drainage from waste dumps and tunnels include aluminum, cadmium, copper, iron, lead, manganese, nickel, and zinc (Bethke, 2011). The deforestation of the area that was necessary to expose minerals has acidified and mobilized heavy metals, the result of which is additional acidic and metallic drainage downstream of the site. Parts of the Alamosa River are unfit for aquatic life. The $100 million-dollar joint cleanup between the State of Colorado and the Environmental Protection Agency, as well as the decommissioning process, is ongoing at present.

The village of Platoro is the only human settlement within the study area, sitting at 9900' along the upper Conejos River. The Platoro Reservoir above the town dams the river. Platoro is nearly 135 years old but has no census population data. The name Platoro is said to reflect the

Spanish words for silver and gold, the discovery of which in the area was the impetus for the 21 founding of the town. It has slowly transitioned from a mining town to a tourist town, with an intermittent period when both mining and tourism were active.

The Wilderness and National Forests encompassing the study area are surrounded by and interspersed with private land, respectively, which takes the form of various ranches and

“suburbs” of Pagosa Springs and other small settlements in the region. These properties can be hubs for exotic plant activity, and most of the exotic flora component is found along populated and well-traveled roadsides in the national forest lands. Trailheads also harbor many exotic plants.

All-Terrain Vehicle use is popular and common on Forest Road 250 and along road corridors of the study area’s east side generally, and the Platoro and Elwood Pass areas have a noticeably constant ATV presence during the height of summer. The forest and county roads on the western slope are usually terminal and do not provide good looping opportunities for motorists. One manifestation of this contrast is that dispersed, free camping is not allowed along the Conejos River portion of Forest Road 250 at present, and this portion is host to many campgrounds as a result. Only informal campsites exist on the western slope. Fishing is also popular on the eastern slope along the Conejos River, and there is infrastructure that supports this beloved activity. The Alamosa River does not seem to host many fisherpeople, perhaps in part due to the Summitville catastrophe.

In the regions where the Wilderness bounds private lands of the Tierra Amarilla Land

Grant, evidence of recent logging both within and without the putative boundaries of the

Wilderness was observed. Logging also occurs in the southeastern portion of the study area at present in areas badly affected by spruce beetle kill. 22

Other than the areas of the Continental Divide from the Summit Peak region north to the

Long Trek Mountain area, much of the tundra is found on ridges that are less than 12500' high.

With an average starting treeline of 12000', vast areas of tundra are susceptible to disappearance if treeline rises in the region.

Other than backpacking, horsepacking, fishing, hunting, and mountain-climbing, direct anthropogenic impacts within the Wilderness are felt from cattle grazing. This practice continues to occur within the boundaries of the Wilderness and is particularly prevalent throughout the river valleys of the eastern slope, especially in the wide drainages of the various upper Forks and tributaries of the Conejos River. A flora loving disturbance, including native and non-native taxa, tends to congregate at the more heavily-grazed areas.

As mentioned, human presence varies across the Continental Divide slopes of the South

San Juans; roads and trails are too terminal on the western slope of the study area—in addition to the fact that many small, highly private “ranches” are common on that slope- for tourist-level recreation to be as prominent as on the eastern slope. Backpacking and horsepacking, however, are still quite popular out of the Fish Lake Trailhead, with dayhiking also being moderately popular in the area. Off-trail recreational hiking seems to be almost nonexistent at this time, as no signs of recent human activity were encountered during remote cross-country botanizing excursions.

METHODS

A total of 1151 collections of vascular plants in the study region were made by the author and at times Dr. Erin Tripp (indicated in the checklist by Sharples’ collection numbers) throughout the summers of 2013 and 2014, as well as during a small portion of the summer of 23

2016. The South San Juan Wilderness itself was the area most comprehensively botanized, but surrounding National Forest land was also extensively botanized (see Fig. 1). On each foray, previously unexplored terrain was covered, although a few sites were visited more than once during distinctly different periods of the growing season to search for taxa with restricted early or late season phenologies. Collections were pressed and subsequently dried at a temperature of

110°F (43°C) for approximately 30 hours. Taxa were identified using the COLO reference collections, guided by Colorado Flora (Weber and Wittman, 2012a & b), Flora of Colorado

(Ackerfield, 2015), the Flora of North America (Flora of North America Editorial Committee,

1993+), and other resources as necessary. Species not collected by the author were searched for in the online regional herbarium database SEINet (Southwest Environmental Information

Network, 2017), and were added to the flora if found to occur in the study area. This was done systematically for each plant family in the two counties of interest, Archuleta and Conejos.

Species found only by one investigator at one locality in the study area were studied in person to determine the accuracy of the identification. Foreign ornamental plants restricted to private land and not naturalized in the area were not included in the flora. No collections prior to the twentieth century were included.

The study area map was created in ArcGIS using publicly-available map layers of

Colorado’s county boundaries, roads, and public lands (Esri, 2016). Taxon distributions were determined using the Flora of North America when available, and the USDA Plants Database

(USDA, 2017c), Colorado Flora (Eastern and Western Slopes), and SEINet were utilized whenever a Flora of North America treatment was still pending, as well as to corroborate distribution information given in the extant Flora of North America treatments. The Flora of 24

U.S.S.R. (Komarov and Shishkin, 1936) and the Flora of China (eFloras, 2008) were additionally consulted at times when Asia was concerned. The Flora of North America’s opinion is reported as concerns the distribution of species inhabiting Europe (inhabiting Europe distinguishes a circumboreal distribution from a North American-Asian distribution). Commonality was ranked qualitatively and quantitatively according to the system described in the checklist portion below.

RESULTS AND DISCUSSION

A total of 744 species of vascular plants in 84 plant families have been identified as occurring in the study area (Table 1). This figure totals about one-third of the Colorado flora (ca.

2300 spp.: Weber and Wittmann, 2010a; 2677 spp. in Ackerfield, 2015). The flora of the South

San Juans is approximately as diverse as the Sangre de Cristo Mountains of Colorado (Hogan, unpublished data). The largest families in the flora are the Asteraceae (131), Poaceae (82),

Cyperaceae (50), Brassicaceae (37) and Fabaceae (28), as is typical of temperate northern hemisphere landscapes and of the findings of other floristic work in Colorado (e.g., Flora of

U.S.S.R., 1936; Hogan, 1992; Flora of North America, 1993+; Clark, 1996; Weber and

Wittmann, 2010a & b). Carex (41) is by far the largest genus in the South San Juans flora, with the second largest genus, Erigeron, numbering only 17 species (Table 2). Nearly all of the species diversity in the flora is attributable to the flowering plants (712 spp.) (Table 3).

Fourteen species are reported here as new to the South San Juan Mountains of Archuleta and Conejos counties (=the study area). These species are not new to either county. Six species included in this figure are only new to the Archuleta portion of the South San Juan Mountains:

Carex echinata, Cryptogramma acrostichoides, Descurainia californica, Plantago lanceolata and Tripleurospermum inodorum. The following 8 species are new in general to the South San 25

Table 1. Number of species in each family in the study area.

Asteraceae: 135 Hydrophyllaceae: 5 Papaveraceae: 2 Poaceae: 83 Malvaceae: 5 Phrymaceae: 2 Cyperaceae: 50 Montiaceae: 5 Scrophulariaceae: 2 Brassicaceae: 37 Amaryllidaceae: 4 Verbenaceae: 2 Fabaceae: 28 Apocynaceae: 4 Alismataceae: 1 Ranunculaceae: 26 Dryopteridaceae: 4 Anacardiaceae: 1 Rosaceae: 26 Liliaceae: 4 Cactaceae: 1 Caryophyllaceae: 23 Potamogetonaceae: 4 Cannabaceae: 1 Plantaginaceae: 21 Adoxaceae: 3 Ceratophyllaceae: 1 Polygonaceae: 19 Crassulaceae: 3 Convolvulaceae: 1 Apiaceae: 17 Equisetaceae: 3 Dennstaedtiaceae: 1 Salicaceae: 17 Nyctaginaceae: 3 Elaeagnaceae: 1 Juncaceae: 16 Primulaceae: 3 Euphorbiaceae: 1 Orobanchaceae: 15 Pteridaceae: 3 Fagaceae:1 Onagraceae: 14 Rubiaceae: 3 Haloragaceae: 1 Ericaceae: 14 Sparganiaceae: 3 Hydrangeaceae: 1 Polemoniaceae: 13 Araceae: 2 Hydrocharitaceae: 1 Saxifragaceae: 13 Asparagaceae: 2 Hypericaceae: 1 Boraginaceae: 11 Berberidaceae: 2 Lentibulariaceae: 1 Gentianaceae: 11 Betulaceae: 2 Linaceae: 1 Pinaceae: 10 Campanulaceae: 2 Nymphaeaceae: 1 Orchidaceae: 10 Celastraceae: 2 Rhamnaceae: 1 Ophioglossaceae: 6 Cornaceae: 2 Santalaceae: 1 Caprifoliaceae: 6 Cupressaceae: 2 Sapindaceae: 1 Grossulariaceae: 6 Geraniaceae: 2 Selaginellaceae: 1 Lamiaceae: 6 Iridaceae: 2 Typhaceae: 1 Violaceae: 6 Isoëtaceae: 2 Urticaceae: 1 Amaranthaceae: 5 Melanthiaceae: 2 Zannichelliaceae: 1

26

Table 2. The eight largest genera in the study area.

Carex L. Erigeron L. Poa L. Salix L. Juncus L. Ranunculus L. Artemisia L.; Draba L.

Table 3. Total number of families and species in each Division.

Angiospermophyta 75 (727) Gymnospermophyta 2 (12) Monilophyta 5 (17) Lycopodiophyta 2 (3) All 84 (759)

27

Juan Mountains: Asclepias subverticillata, Barbarea vulgaris, Chamaesyce serpyllifolia,

Collinsia parviflora, Hordeum jubatum, Malva neglecta, Polygonum minimum, and Ranunculus cardiophyllus.

Thirty-nine collections were new county occurrence vouchers for various species.

Twenty-eight species are new to Archuleta, and 11 are new to Conejos. Their identities can be found in the checklist below and are identified as either “New to AA” or “New to CN” therein.

In a few instances, a species was collected newly in both CN and AA counties, and this is indicated when so. Numerous county vouchers are new to the South San Juan Mountains as well but are not listed with the 14 above.

Forty-two collections (45, if Carex echinata, Descurainia californica, and Polygonum minimum above are considered) represented species newly collected from the Wilderness, as opposed to surrounding National Forest lands. Taxa classified this way were already known from the county in which they were collected, but were unknown from within the Wilderness boundaries of that county. Of these, 28 species were new to the Wilderness as a whole (either county), 9 species were new to the Wilderness in Archuleta County, and 5 species were new within the Wilderness in Conejos County. The first category of species new to the Wilderness as a whole is designated by “New to Wilderness (AA or CN)” in the checklist, and the second two categories are designated by “New to AA or CN Wilderness”.

Records located on SEINet were not always accurate. Over 100 species were removed after checking anomalous SEINet records in person. Such records were removed for three reasons. Sometimes specimens were misidentified. More often, it was found that herbarium label annotations were not up to date. That is, quite a few specimens were originally identified as a species that would have been anomalous in the study area, but had been re-annotated and 28 identified as a more common species in the study area subsequently. The reannotated concept was not updated on SEINet since the specimen record had originally been uploaded, and thus the original, older species identification was given on the public database. Periodically, records on

SEINet were also found to have locality information erroneously translated onto the public database from the original specimen label. These observations highlight the importance of making in-person herbarium visits during the process of conducting biodiversity inventories using public online databases.

Despite anthropogenic disturbance in the region as discussed earlier, the Wilderness boundaries harbor a nearly pristine pre-Columbian flora. A total of 533 species of vascular plants are found within its boundaries. Almost all of these (521 spp.) are native to Colorado; 2.25% of the species in the Wilderness are therefore Eurasian aliens. These are almost exclusively confined to the immediate boundaries of footpaths, though Verbascum thapsus is found away from trails at times in low, open areas. If a walker leaves the trail, virtually no Eurasian aliens are to be encountered except along the odd section of floodplain or otherwise highly-disturbed drainage. The flora of the broader South San Juans outside of Wilderness boundaries according to Figure 1 includes another 211 species. A total of 41 of these are Eurasian aliens: 19.4% is thus the relative percentage of Eurasian aliens in the flora outside of Wilderness. In the lower elevations outside of Wilderness boundaries, Eurasian aliens are widespread and common along roads and trails, as well as throughout other habitats such as forest understory, meadows, and waterways. This could be due to a synergistic effect of lower elevations (=milder climates) supporting more invasive species by virtue of the relative non-invasibility of more extreme climates of higher elevations in addition to the high level of anthropogenic disturbance to which the lower elevations of the study area are subject. Anthropogenic disturbance is strongest lower 29 on the western slope than in other parts of the study area. Haying, cattle grazing, private residences and ranches, as well as the usual recreational activities are widespread here. Thus, qualitatively, Eurasian aliens are in highest concentration low on the western slope, and it would be worth investigating further which factors exactly contribute to this phenomenon.

Floristic Distributions

Distribution remarks are not to be regarded as static facts, and reflect only the current state of knowledge (Table 4). Sometimes, a given classification may have seemed too arbitrary, and therefore similar classifications are combined in a more general view of floristic affinities

(Table 5). That is to say, at times a taxon exhibited a distribution that seemed to have both feet in either a Western United States or a Western North America distribution type, for example, perhaps being found in only the western United States except for a population front extending into southwestern Canada. Until fine-scale distributional extent is mapped across all countries and until taxonomy is fully resolved for all species, there will be a degree of imprecision in making conclusions about distributional extent. The following distributional categories were identified within the South San Juan Mountains flora. Acronyms in bold are found for each species within the checklist.

WNA—Species classified within a Western North America distribution typically range from western (northwestern) Mexico throughout the western United States, and into high latitudes of Alaska, Yukon, and/or the Northwest Territories. Sometimes these taxa are not found in Mexico, or they do not quite range up to the northernmost State and/or Provinces, yet are otherwise regionally widespread beyond the boundaries of the western United States.

WUSA—Taxa with a Western United States distribution are usually restricted to the mountainous western half of the contiguous United States. At times, though, some of these 30

Table 4: Number of species and percentage of the flora classified in each distribution type.

Western North America 118 16%

Western United States 98 13% Circumboreal 91 12%

Western & Northern North America 70 9% Eurasian Alien 59 8% Widespread North America 55 7%

Four Corners 54 7% United States Rocky Mountains 49 6% Southern Rocky Mountains 44 6%

North American-Asian 29 4% Temperate Northern Hemisphere 29 4% Southwestern North America 27 4% Cosmopolitan 16 2%

Anomalous 11 1% The Rocky Mountains 9 1%

Table 5: A more generalized classification of distributions.

Western North America 216 28% North America 125 16%

Northern Hemisphere 120 16% Southwestern North America 81 11% Eurasian Alien 59 8% Rocky Mountains 58 8%

Southern Rocky Mountains 44 6% North American-Asian 29 4% Cosmopolitan 16 2%

Anomalous 11 1%

31 species will range into the southwestern corner of Canada, and thus tread the line between having a Western United States or a Western North America distribution. Likewise, a species may not range into Canada at all, but will range slightly into northwestern Mexico. Both of these example situations are ordinarily classified as Western United States, as they are greatly restricted outside of the contiguous United States’ latitudes.

WNNA—Species found in Western and Northern North America exhibit what could be called a circumboreal distribution that lacks the Eurasian extent. Thus, these are species that occur in the western United States, sometimes western Mexico, large portions of Canada

(sometimes all of Canada), and at times large portions of the northeastern United States. The exact boundary of this distribution may in rare cases overlap with what might be considered

Temperate North America (see below).

TNA—Widespread or Temperate North American species are found throughout Canada, the United States, and often parts of Mexico. Occasionally these species do not range into the farthest northern parts of the continent, or into the warm, subtropical parts of the southeastern

United States. When absent from the southeastern United States but found throughout the rest of

North America north of Mexico, the range can begin to touch toes with a Western and Northern

North America distribution. However, Temperate North American species are widely distributed outside of the boreal climate zones, unlike Western and Northern North American species.

EUA—Eurasian aliens are taxa not native to North America, and are usually, but not always, found in disturbed areas (overgrazed meadows, roadsides, car camping areas, trailsides, settled areas, etc.) at lower elevations. Their origin may be from Europe, Asia, both, or unknown.

Exotic aliens from the southern hemisphere do not appear to be part of the flora. Aliens are essentially unknown from the alpine. 32

4CR—Four Corners taxa are usually restricted to the states of Arizona, Colorado, New

Mexico, and Utah, but one of these states may be absent in the range of a Four Corners species.

Sometimes these species are found in the Four Corners states but also extend into one or another neighboring state, such as Texas, Nevada, southern Wyoming, or northern Mexico. If multiple of these are all extended into, the species is more wont to be classified as Southwestern North

American (see below). Many of these species are found at lower elevations. Taxa with a distribution closely restricted around the actual Four Corners boundary are classified here.

SWNA—Southwest North American species are usually found through the southwestern

United States and into northern Mexico, encompassing a much broader area than Four Corners taxa, while often overlapping with a Four Corners distribution. Table 5 combines these distributions.

CIRC—Circumboreal species are found in arctic-alpine, boreal, and/or high latitude areas throughout the northern hemisphere and sometimes beyond (e.g., disjunctions in South

America). Circumboreal taxa rarely extend into more temperate, southerly, non-mountainous regions of the northern hemisphere. “Circumpolar” species restricted to arctic and alpine areas across the northern hemisphere are also classified here. In the study area, circumboreal species tend to be found above 10,000 feet.

TNH—Temperate or Widespread Northern Hemisphere taxa range widely throughout the hemisphere, from arctic to often subtropical latitudes. Though found in much warmer areas than circumboreal taxa, these distributions overlap extensively, and are combined in Table 5.

NAA—North American – Asian species are usually found in portions of western North

America and eastern Asia, but may be more widespread throughout North America as well.

These distributions often look circumboreal sans a presence in western Asia and Europe. 33

COSM—Cosmopolitan species are widespread throughout the northern and southern hemispheres, and tend to be seedless plants or aquatic angiosperms. Most of these taxa are found natively on all continents outside of Antarctica, but may be missing from Africa, Oceania, and/or the lowland tropics.

ANOM—Anomalously-distributed species are: Lemna minor, Packera tridenticulata,

Draba crassifolia, Erysimum capitatum, Arenaria lanuginosa, Cerastium arvense, Oxytropis lambertii, Botrychium pallidum, Veronica peregrina, and Viola labradorica.

USRM—A United States Rocky Mountains distribution ranges from New Mexico or

Colorado up north into Montana and Idaho, and usually does not extend eastwards or westwards from these states. At times, a breach into Alberta, or a disjunction in the Cascades, may still be classified here. These taxa are absent from the Sierra Nevada and the Great Basin, however.

ROMO—Taxa ranging throughout the Rocky Mountains extend from throughout the

United States Rocky Mountains and up north to nearly the border of the Yukon. They are not found in the Pacific territories. As only a small number of taxa exhibit this distribution, these plants should probably be considered as having an extended United States Rocky Mountains distribution.

SRM—The Southern Rocky Mountains encompass as much as parts of four states: the high country of western Colorado, northern New Mexico, southern Wyoming, and eastern Utah.

In some cases, northern Arizona (the San Francisco Mountains) is considered part of this range.

Taxa categorized as having a Southern Rocky Mountains distribution may be even further narrowly restricted within the Southern Rockies, e.g., found only in parts of Colorado, or otherwise not found through the full extent of the Southern Rockies. A number of taxa here are indeed endemic solely to southwestern Colorado. These narrow endemics are listed below. 34

Much of the South San Juan flora thus has a species-rank relationship to broad temperate areas of the three northern continents, both natively and through introduction; North American,

Northern Hemisphere, Eurasian Alien, and American – Asian taxa comprise almost half of the flora with 331 species (Table 5). Most of the rest of the flora is restricted to part or all of Western

North America, totaling 387 species, and this component only sometimes reaches into Canada.

The most restricted of taxa are those which exhibit a Four Corners, Southwestern North America or Southern Rocky Mountains distribution, totaling 117 species.

Percentages of distribution indicated in Table 5 are similar to those of the Gore Range of north-central Colorado (Hogan, 1992). The Gore Range flora, however, seems to lack the

Southwestern North American/Four Corners element, and has a lower percentage of Eurasian aliens (in 1992: 4%). The flora of Mesa de Maya in southeastern Colorado contrasts with the

Gore Range and the South San Juans (Clark, 1996). A quarter of that flora has a Great Plains distribution, a distribution so inconspicuous in the South San Juans flora that it is lumped under

“Anomalous”. The Mesa de Maya flora also has a conspicuous Eastern North America element which is lacking in the South San Juans. The Mesa de Maya has a higher percentage of Eurasian aliens (13%), presumably on account of its lower elevation and higher concentration of roads and ranchlands.

Plants endemic to the greater San Juan Mountains (to include, in some cases, the Elk

Mountains, and/or the Sangre de Cristos) and found within the study area include Corydalis caseana (disjunct elsewhere), Draba smithii, Ligularia soldanella, Penstemon hallii, Penstemon harbourii, Phlox caryophylla, Ranunculus macauleyi, Stellaria irrigua sensu W.A. Weber,

Trautvetteria caroliniensis (disjunct elsewhere), and Trifolium brandegeei. 35

A few taxa are conspicuously absent from the study area despite occurring nearby.

Cactaceae is virtually absent (1 collection). Arceuthobium (Viscaceae) is altogether absent. In the alpine, several species appearing in neighboring mountain ranges are conspicuously absent or very uncommon. Some of these taxa include Dryas octopetala, Primula angustifolia, Castilleja haydenii, and Eremogone fendleri.

Sensitive Species

The Wilderness and surroundings harbor a flora that is more or less unthreatened across its distribution at present. There are no Federally-Endangered, Threatened, or Proposed species in the flora, and there are only three species in the flora on the Forest Service Regionally

Sensitive List (Region 2): Astragalus proximus (Fabaceae), Draba smithii (Brassicaceae), and

Eriophorum gracile (Cyperaceae). Sixteen additional species are tracked by the Colorado

Natural Heritage Program (www.cnhp.colostate.edu/download/list/vascular.asp; accessed Jan

2017): Artemisia laciniata, Asclepias hallii, Botrychium pinnatum, Carex limosa, Carex oreocharis, Carex retrorsa, Castilleja lineata, Cryptogramma stelleri, Draba malpighiacea,

Draba streptobrachia, Grindelia arizonica, Luzula subcapitata, Minuartia macrantha, Phlox caryophylla, Townsendia glabella, and Trautvetteria carolinensis.

Agency Recommendations

A higher amount of vascular plant diversity seems to be found in the ponderosa-oak zone habitats of the western slope of the study area than in the habitats above it. Conversely, these habitats are the most disturbed and least protected of those found in the study area. Only a small portion of the South San Juan Wilderness in Archuleta county extends down into this lowland habitat type. Though this habitat is widespread just west of the study area, none of it is protected from anthropogenic disturbances such as grazing, logging, motorized uses, and settlement 36 activities, and many rare or endemic species are found in the ponderosa-oak savannahs both west and east of Highway 84. A high proportion of the striking ponderosa-oak habitat found on the western slope in this region that harbors lowland Four Corners endemic plant species is covered in private land, tribal land, or disturbed National Forest land. It is therefore recommended that western boundaries of the Wilderness be extended lower in some areas to encompass more of this habitat, particularly in the southwestern parts of the study area, so that it can be preserved as well as possible. An alternate or complementary suggestion may be to protect more of this habitat type west of Highway 84 and/or in far northern New Mexico. It might be furthermore recommended that cattle grazing be more restricted on the western slope of the study area, as this type of disturbance in particular seems to be a very successful vector for the conveyance of

Eurasian exotic species into relatively pristine public lands. In effect, the question might be raised as to whether such exotics would be found on public lands in the region (besides by roads or trails) if not for the way that cattle seem to facilitate the spread of exotics into various habitats.

On another note, although this work has vouchered many new populations of most species of the

South San Juans flora, fewer than 100 species were found to constitute new discoveries of some sort to the region. Cumulatively, this checklist represents a species count resulting from scores of decades’ worth of dedicated collecting work in the region, and it should be considered current and nearly complete. Mineral and Rio Grande county portions of the San Juans may yield further new discoveries, however.

37

CHECKLIST OF THE VASCULAR FLORA OF THE SOUTH SAN JUAN MOUNTAINS

The checklist is alphabetically sorted by family, and within family, sorted alphabetically by genus and species. Elevations and collection numbers match one another respectively if multiple are given. More collections for certain taxa are not necessarily indicative of those taxa being more common. All vouchers cited as Sharples or Sharples & Tripp are deposited at the

University of Colorado Museum Herbarium (COLO), with numerous duplicates deposited at RM and NY. Acronyms found at the end of a species’ listing refer back to floristic distributions described above.

Nomenclature.—Genera and specific epithets typically follow the Flora of North

America, when available. Inconsistencies with the Flora of North America or taxa not yet treated by the FNA are given based on Colorado Flora or/and Flora of Colorado. Family circumscriptions follow the Angiosperm Phylogeny Group (APG IV, 2016). Subspecies are generally not recognized unless they represent a highly distinct entity of a much more widespread species complex.

Rarity.—When the taxon was not personally collected by the author, SEINET and COLO database searches serve as a metric for ranking commonality (more common=more databased collections), but full, unquestioning credence should not be given to this system, due to collector bias. Otherwise, commonness is ranked according to qualitative field observations and by the following system:

Abundant—Dominant and found virtually everywhere in a given life zone.

Common—Widespread and nearly always found throughout a given habitat, yet not quite a dominant species. Usually found often during a given foray into a given habitat. Known from more than 10 populations in the study area. 38

Occasional/Scattered/Intermittent—Interspersed across a given life zone(s), yet not always found during an extensive excursion into that life zone. Known from 5–10 populations across investigators in the study area.

Uncommon—A surprise to encounter at all; sparsely-distributed across the landscape as a whole or quite restricted to an uncommon habitat type. Known from 2–5 populations in total in the study area, including Sharples’ collections.

Rare—One population or one restricted region known.

Checklist Symbols.—An asterisk (*) before the genus and epithet denotes a taxon unknown from within the Wilderness. A (W) after the distribution listing denotes an alien taxon within Wilderness.

Other South San Juans Investigators & Affiliations.—The below investigators have collected species in the study area not collected by me. Collection numbers do not refer to cumulative number of collections made by each investigator throughout the study area, but rather to number of species contributed to the checklist.

Douglas: Patricia Douglas, Colorado State University (CS, RM); 51 collections

Flaig: Jeanette Flaig, University of Wyoming (RM); 37 collections

Harrington: Harold David Harrington, Colorado State University (CS); 6 collections

Hartman: Ronald L. Hartman, University of Wyoming (RM); 5 collections

Heil: Kenneth Heil, San Juan College (SJNM); 27 collections

Komarek: Susan Komarek, Fort Lewis College (FLD); 2 collections

Nelson: Burrell E. Nelson, University of Wyoming (RM); 35 collections

O’ Kane: Steve O’ Kane Jr, San Juan College (SJNM); 6 collections

Denver Botanic Gardens: (DBG); 2 collections 39

Common Additional Acronyms—AA = Archuleta County; AZ = Arizona; CA =

California; CN = Conejos County; CO = Colorado; E = Eastern; MX = Mexico; N = Northern;

NE = Northeast; NV = Nevada; PNW = Pacific Northwest; S = South; SSJ = South San Juan

Mountains; TX = Texas; U = Utah; W = Western; WY = Wyoming.

ADOXACEAE – 3 species

Adoxa moschatellina L., rare. ~11500', Douglas 1703 (CS). CIRC.

Sambucus caerulea Raf., uncommon. 9200', Hartman 79726 (RM). WUSA.

Sambucus racemosa L. var. melanocarpa (A. Gray) McMinn, common in forest understory.

9100 & 11100', Sharples 23 & 478. TNH.

ALISMATACEAE – 1 species

*Alisma triviale Pursh, rare in low boggy areas. 8700', Sharples 1306. WNNA.

AMARANTHACEAE – 5 species

Chenopodium atrovirens Rydb., common in forest meadows. 10150 & 9250', Sharples 959 &

1082. Essentially WUSA.

Chenopodium capitatum (L.) Ambrosi [C. overi], scattered in open areas. 12050’, Sharples &

Tripp 1128. WNNA natively.

*Chenopodium foliosum (Moench) Asch., uncommon. 10000', Sharples 150. EUA.

Chenopodium fremontii S. Watson, common in low elevation open areas. 9600 & 8500',

Sharples 690 & 950. WNA.

Monolepis nuttalliana (Schult.) Greene, scattered across lowlands. 8750 & 9600', Sharples 246

& 691. New to Wilderness (CN). TNA.

AMARYLLIDACEAE– 4 species 40

Allium acuminatum Hook., common in the ponderosa-oak zone. 8700', Sharples 803. WUSA.

*Allium cernuum Roth, uncommon in ponderosa-oak zone openings. 7760–7920', Nelson 58992

(RM). TNA.

Allium geyeri S. Watson, common in open areas. 8000, 7800 & 9300', Sharples 101, 187 & 796.

WNA.

Allium rubrum Osterh. [A. geyeri var. tenerum], common in upper elevation openings. 11500 &

11100', Sharples 316 & 851. WUSA.

ANACARDIACEAE – 1 species

*Toxicodendron rydbergii (Small ex Rydb.) Greene, rare. O’ Kane 3952 (SJNM). WNNA.

APIACEAE – 16 species

Angelica grayi (J.M. Coult. & Rose) J.M. Coult. & Rose, uncommon, restricted to alpine talus.

12600 & 12000', Sharples 925 & 968. SRM.

Angelica pinnata S. Watson, rare in low, forested areas. 8800', Sharples 1077. USRM.

*Cicuta maculata L., rare in low western slope wetlands. 7400', Sharples 1303. TNA.

Conioselinum scopulorum (A. Gray) J.M. Coult. & Rose, scattered in lower elevation forests.

9200–9240', Nelson 62576 (RM). USRM.

Cymopterus bakeri (J.M. Coult. & Rose) M.E. Jones [Oreoxis], common in high meadows. O’

Kane 5041 (SJNM). Endemic around 4CR.

Cymopterus lemmonii (J.M. Coult. & Rose) Dorn [Pseudocymopterus montanus], abundant in meadows. 10350, 7750, 11100, 9050 & 11100', Sharples & Tripp 12 and Sharples 89, 356, 524

& 850. 4CR.

Heracleum maximum W. Bartram, common in mixed forest understory. 9200, 8750 & 9250',

Sharples 448, 543 & 552. TNA. 41

Ligusticum porteri J.M. Coult. & Rose, common in upper elevation forests and openings. 9800,

10350 & 12050', Sharples 153, 415 & 1059. USRM; N MX.

*Lomatium grayi (J.M. Coult. & Rose) J.M. Coult. & Rose, scattered in ponderosa-oak lands.

7924', Flaig 3859 (RM). WUSA.

*Orogenia linearifolia S. Watson, rare in ponderosa-oak lands. Heil 23434 (SJNM). WUSA.

*Osmorhiza chilensis Hook. & Arn. [O. berteroi], uncommon in mixed forest. 8780–9200',

Hartman 79699 (RM). WNNA.

Osmorhiza depauperata Phil., common in forested areas. 8600', Sharples 813. WNNA.

Oxypolis fendleri (A. Gray) A. Heller, common in forest understory. 8950, 11550 & 9250',

Sharples 304, 398 & 556. 4CR; S WY.

*Pastinaca sativa L., rare in wet, disturbed areas. 7600', Sharples 1360. EUA.

Podistera eastwoodiae (J.M. Coult. & Rose) Mathias & Constance, scattered on tundra. 12350',

Sharples 828. SRM.

*Sium suave Walter, scattered in wet ground at low elevations on the western slope. 8700',

Sharples 1308. TNH.

APOCYNACEAE – 4 species

Apocynum androsaemifolium L., scattered across the mixed forest zone. 9700 & 9500', Sharples

599 & 692. TNA.

Asclepias hallii A. Gray, uncommon. ~10800', Douglas 1702 (RM). 4CR; E NV.

Asclepias speciosa Torr., rare along roads and trails. 9050', Sharples 286. WNA.

*Asclepias subverticillata (A. Gray) Vail, rare on bare low elevation slopes. 7850', Sharples

1352. New to SSJ, found in AA. SWNA.

ARACEAE – 2 species 42

*Lemna minuta Kunth, scattered in lowland bogs and ponds. 8750 & 8950', Sharples 562 &

1314. ANOM: temperate Americas.

*Lemna trisulca L., rare in low elevation riparian areas. 8720–9200', Nelson 62500 (RM).

COSM.

ASPARAGACEAE – 2 species

Maianthemum racemosum (L.) Link subsp. amplexicaule (Nutt.) LaFrankie, common in mixed forest understory. 9100, 7750 & 9300', Sharples & Tripp 24 and Sharples 119 & 799. WNA.

Maianthemum stellatum (L.) Link, common at lower elevations in forest understory. 10100, 8900

& 10350', Sharples & Tripp 54 and Sharples 64 & 418. TNA.

ASTERACEAE – 131 species

Achillea millefolium L., abundant across open areas. 8650, 9200 & 10250', Sharples 282, 421 &

482. TNH.

Agoseris aurantiaca (Hook.) Greene, common in subalpine meadows. 11600, 11500 & 12800',

Sharples 332, 462 & 636. WNA.

Agoseris glauca (Pursh) Raf., scattered in higher meadows. 11400', Sharples 648. WNA.

Amauriopsis dissecta (A. Gray) Rydb. [Bahia], scattered at low elevations. 8800 & 9200',

Sharples 659 & 1347. SWNA.

*Ambrosia psilostachya DC, rare along low elevation roads. Heil 12631 (SJNM). TNA.

Anaphalis margaritacea (L.) Benth. & Hook. f., common in open areas at lower elevations.

10450 & 9500', Sharples 1040 & 1085. Widespread NAA.

*Antennaria anaphaloides Rydb., uncommon in lowlands. 8650 & 9450', Sharples 298 & 485.

New to AA & CN (new to SSJ). WUSA.

Antennaria corymbosa E.E. Nelson, scattered in meadows. 12000', Sharples 340. WUSA. 43

Antennaria media Greene, scattered in high rocky areas. 12800', Sharples 923. WNA.

Antennaria microphylla Rydb., scattered across lowlands. 9515', Douglas 1248 (CS). WNNA.

Antennaria parvifolia Nutt., common on dry, open ground. 9100', Sharples & Tripp 35. WNA.

Antennaria rosea Greene, uncommon in open areas. 9550', Sharples 163. WNNA.

Antennaria rosulata Rydb., uncommon along roads. Bayer et al. CO-407 (RM). 4CR.

Antennaria umbrinella Rydb., scattered in upper meadows. 11475', Sharples 1362. WNA.

*Arctium minus (Hill) Bernh., rare along low western slope roads. 8535–8730', Flaig 2909

(RM). EUA.

Arnica cordifolia Hook., common in subalpine forest understory. 9100 & 10350', Sharples 28 &

412. Greater WNA

Arnica latifolia Bong., uncommon. 9700', Sharples 955. New to AA Wilderness. WNA.

Arnica mollis Hook. [A. ovata], common in subalpine openings. 11800, 11700 & 12550',

Sharples 389, 476 & 1132. WNA.

Arnica parryi A. Gray, rare in high meadows. 12000', Douglas 1369 (CS). WNA.

*Artemisia biennis Willd., rare. Radely 57 (CS). Provenance unclear; TNH and beyond.

Artemisia borealis Pall. [A. campestris, Oligosporus groenlandicus], common in tundra meadows. 12300, 12400, 12550 & 12150', Sharples 351, 875, and Sharples & Tripp 1138 &

1140. CIRC.

Artemisia carruthii Wood ex Carruth, common along lower elevation roads and openings. 10650

& 7800', Sharples & Tripp 1118 and Sharples 1358. SWNA.

Artemisia dracunculus L., common on dry mixed forest slopes. 9800 & 8450', Sharples 689 &

1076. CIRC; cultivated elsewhere. 44

Artemisia franserioides Greene, common in open areas of lowlands. 10600', Sharples 1104. New to AA Wilderness. 4CR.

Artemisia frigida Willd., uncommon in open lower elevation sites. 10000', Sharples 687. New to

Wilderness (CN). TNH.

Artemisia laciniata Willd. subsp. parryi (A. Gray) W.A. Weber [A. parryi], uncommon on bare ground. 9950', Sharples 688. Subspecies: SRM; species: CIRC.

Artemisia ludoviciana Nutt., scattered in open areas in the ponderosa-oak zone. 8250', Sharples

1079. TNA.

Artemisia michauxiana Besser, uncommon on rocky sites. 9100, 12050 & 12150', Sharples 443,

1060 and Sharples & Tripp 1135. New to AA and CN. WNA.

Artemisia scopulorum A. Gray, common in subalpine meadows. 12650 & 11900', Sharples 863

& 960. USRM.

*Bidens cernua L., rare along ponderosa-oak zone roads. 9200', Sharples 1346. TNH.

Brickellia grandiflora (Hook.) Nutt., common on loose, rocky slopes at mid-elevation. 10200,

9800 & 10600', Sharples 614, 670 & 1105. WUSA.

Carduus nutans L., infrequent in wilderness, but common along roadsides. 7450 & 8650',

Sharples 539 & 721. EUA (W).

Chaenactis alpina (A. Gray) M.E. Jones, uncommon on loose, volcanic, high ridges. New to AA.

12350, 11950 & 12200', Sharples 353, 914 & 1061. WUSA.

*Chrysothamnus depressus Nutt., rare along lowland roads. Heil 12715 (SJNM). SWNA.

Cirsium arvense (L.) Scop., common along roads and other low disturbed areas. 8600', Sharples

503. EUA (W). 45

*Cirsium canescens Nutt., common in low disturbed areas on the western slope. 8600', Sharples

505. Patchy WUSA.

Cirsium parryi (A. Gray) Petr. [C. pallidium], common in the mixed forest zone. 9800, 10700 &

9200', Sharples 529, 673 & 971. 4CR; not UT.

*Cirsium scariosum Nutt., rare. 8130–8535', Flaig 2857 (RM). WNA.

Cirsium scopulorum (Greene) Cockerell ex Daniels [C. eatonii var. eriocephalum], common in high elevation openings. 12200', Sharples 683. SRM.

*Cirsium vulgare (Savi) Ten., uncommon in the ponderosa-oak zone. 8600', Sharples 1078.

EUA.

*Conyza canadensis (L.) Cronquist, rare along lowland roadsides. 8100 & 7900', Sharples 189 &

709. TNA natively.

Crepis nana Richardson, rare on alpine screes. 12274', Douglas 2068 (CS). NAA.

*Crepis occidentalis Nutt., common in low, dry sites on the western slope. 8000 & 7800',

Sharples 105 & 195. WNA.

*Cyclachaena xanthifolia (Nutt.) Fresen., rare along low roads of the western slope. 8200',

Sharples 703. TNA.

Dieteria bigelovii (A. Gray) D.R. Morgan & R.L. Hartm. [Machaeranthera], common on open slopes. 9500, 8450 & 10700', Sharples 662, 1074 and Sharples & Tripp 1147. New to CN

Wilderness (1147). 4CR.

*Ericameria nauseosa (Pall. ex Pursh) G.L. Nesom & G.I. Baird [Chrysothamnus], common in low elevation open, dry areas. 8450 & 9600', Sharples 1073 & 1112. WNA.

*Ericameria parryi (A. Gray) G.L. Nesom & G.I. Baird, uncommon in low elevation meadows.

8800', Sharples 658. WUSA. 46

Erigeron compositus Pursh, common on cliffs and rocks at upper elevation. 10050, 10500, 12250

& 11950', Sharples & Tripp 53, and Sharples 146, 825 & 915. NAA.

Erigeron coulteri Porter, common in meadows and open areas. 11600, 9150, 9300 & 11600',

Sharples 330, 576, 798 & 1113. WUSA.

*Erigeron divergens Torr. & A. Gray, scattered across open sites in the ponderosa-oak zone.

7800', Sharples 1357. WNA.

Erigeron elatior (A. Gray) Greene, uncommon. 8950–9050', Nelson 62462 (RM). SRM.

Erigeron eximius Greene, scattered in montane forests. 9500–9800', Flaig 2980 (RM). 4CR.

Erigeron flagellaris A. Gray, common on exposed sites. 8700, 8600, 7800 & 9200', Sharples &

Tripp 17, and Sharples 68, 108 & 423. WNA.

Erigeron formosissimus Greene, common in lower elevation openings. 9500', Sharples 1084.

USRM.

Erigeron glacialis (Nutt.) A. Nelson [E. peregrinus], common in subalpine open areas. 11100 &

11650', Sharples 479 & 587. WNA.

Erigeron grandiflorus Hook. [E. simplex], common in meadows. 9200 & 12150', Sharples 422 &

906. WUSA.

Erigeron leiomerus A. Gray, common on rocky sites. 10600, 8600, 12800 & 12700', Sharples

776, 809, 921 & 966. WUSA.

*Erigeron lonchophyllus Hook., rare in the ponderosa-oak zone. Heil & Mietty 12700 (SJNM).

NAA.

Erigeron melanocephalus (A. Nelson) A. Nelson, common in high open areas. 12200 & 12150',

Sharples 682 & 1364. SRM. 47

Erigeron pinnatisectus (A. Gray) A. Nelson, common in high, rocky sites. 11850, 11700, 11950

& 12700', Sharples 214, 471, 916 & 965. SRM.

Erigeron speciosus (Lindl.) DC, uncommon in the montane. 8950–9050', Nelson 62468 (RM).

WNA.

Erigeron subtrinervis Rydb. ex Porter & Britton, common in forest openings. 9500 & 10700',

Sharples 517 and Sharples & Tripp 1378. WUSA.

Erigeron ursinus D.C. Eaton, rare in the subalpine. 10250', Sharples 1372. New to CN. USRM.

Erigeron vagus Payson, rare on tundra. 12900', Sharples 861. Restricted WUSA.

Gnaphalium exilifolium A. Nelson, scattered in low elevations wetlands. 9200–9240', Nelson

62550 (RM). Broader 4CR.

*Grindelia arizonica A. Gray, rare in ponderosa-oak forest openings. 7760–7920', Nelson 59011

(RM). SWNA.

*Grindelia decumbens Greene, rare in lowland gravels. 9023', Douglas 1260 (CS). SRM.

Helianthella parryi A. Gray, uncommon in the montane. 10000', Douglas 1082 (CS). 4CR.

Helianthella quinquenervis (Hook.) A. Gray, common in middle elevation openings. 10350 &

9500', Sharples 420 & 521. WUSA.

*Helianthus annuus L., common along pondersa-oak zone roads. 7800', Sharples 1355. TNA natively.

*Helianthus pauciflorus Nutt. [H. rigidus], rare along ponderosa-oak zone roads. Heil & Mietty

28141 (SJNM). TNA.

Heliomeris multiflora Nutt., scattered in lowlands. 10000 & 9350', Sharples 672 & 999. WUSA.

*Herrickia glauca (Nutt.) Brouillet [Eucephalus], uncommon in the ponderosa-oak zone. 8450',

Sharples 1072. USRM. 48

Heterotheca pumila (Greene) Semple, uncommon on high, open slopes. 10950', Sharples 582.

SRM.

Heterotheca villosa (Pursh) Shinners, common in low elevation openings. 8300, 8000 & 7450',

Sharples 112, 197 & 535. WNNA.

Hieracium fendleri Sch. Bip., uncommon in ponderosa-oak forests. 8000', Sharples 103. SWNA.

Hieracium triste Willd. ex Spreng. [H. gracile, Chlorocrepis tristis], uncommon in high openings. 11811', Douglas 1758 (RM). WNA.

*Hymenopappus newberryi (A. Gray) I.M. Johnst., rare in the ponderosa-oak zone. 7300 &

7350', Harrington 8072 (CS) and Sharples 804. SRM.

Hymenoxys hoopesii (A. Gray) Bierner, common in forest openings. 11700, 9450, 9500 & 9400',

Sharples 376, 486, 514 & 794. WUSA.

*Hymenoxys richardsonii (Hook.) Cockerell var. floribunda (A. Gray) K.F. Parker, uncommon in lowland openings. 7500, 8700 & 8750', Kelly 198 (DGB) and Sharples 250 & 608. Species:

USRM; variety: 4CR.

*Lactuca serriola L., uncommon along low elevation roads. 7800', Sharples 1356. EUA.

Leucanthemum vulgare Lam., scattered along roadsides. 10450 & 9150', Sharples 407 & 565.

EUA (W).

Ligularia amplectens (A. Gray) W.A. Weber, common in high elevation meadows. 11450 &

12250', Sharples 585 & 724. USRM.

Ligularia bigelovii (A. Gray) W.A. Weber, common in upper subalpine meadows. 11700',

Sharples 628. 4CR; S WY.

Ligularia holmii (Greene) W.A. Weber, uncommon on loose alpine substrata. 12100', Sharples

909. New to AA. Greater USRM. 49

Ligularia soldanella (A. Gray) W.A. Weber, scattered on screes. 11850 & 12350', Sharples 388

& 866. SRM: the volcanic ranges.

Ligularia taraxacoides (A. Gray) W.A. Weber, scattered on loose alpine terrain. 12000 &

12300', Sharples 625 & 631. SRM.

Madia glomerata Hook., uncommon in disturbed areas of the ponderosa-oak zone. 8400 &

8650', Sharples 714 & 718. WNNA.

*Mulgedium pulchellum (Pursh) G. Don [Lactuca tatarica], rare in aspen forest. 9150–9400',

Nelson 62392 (RM). TNA.

Oreochrysum parryi (A. Gray) Rydb. [Solidago], common at upper elevations. 11700, 11000,

12050, 9600 & 11600', Sharples 627, 1042, 1058, 1092, and Sharples & Tripp 1124. Greater

4CR.

Packera crocata (Rydb.) W.A. Weber & Á Löve, uncommon in high meadows. 11600', Sharples

1114. Greater USRM.

Packera dimorphophylla (Greene) W.A. Weber & Á Löve, common in higher elevation meadows. 11800 & 11200', Sharples 344 & 852. Greater USRM.

Packera neomexicana (A. Gray) W.A. Weber & Á Löve, common in open lowland areas. 9100,

8000, 9200, 8450 & 9050', Sharples & Tripp 21, and Sharples 104, 450, 746 & 768. SWNA.

Packera streptanthifolia (Greene) W.A. Weber & Á Löve, scattered across elevations and habitats. 9250, 12000 & 9600', Sharples 548, 626 & 1091. WNA.

Packera tridenticulata (Rydb.) W.A. Weber & Á Löve, scattered on the eastern slope; all records are likely a hybrid with Packera neomexicana. 9800', Sharples 239. ANOM: USA Great Plains. 50

Packera werneriifolia (A. Gray) W.A. Weber & Á Löve, common in high meadows. 10825,

12300, 12000, 12500 & 12050', Sharples 219, 347, 822, 922 and Sharples & Tripp 1365.

WUSA.

*Pericome caudata A. Gray, rare at exposed low elevation sites. 7900', Sharples 705. SWNA.

Pyrrocoma crocea (A. Gray) Greene, common in open areas at lower elevations. 8600, 9700 &

7600', Sharples 509, 1000 and Sharples & Tripp 1359. 4CR; S WY.

*Ratibida columnifera (Nutt.) Wooton & Standl., rare along ponderosa-oak zone roadsides.

7460', Heil 12610 (SJNM). TNA.

Rudbeckia ampla A. Nelson, common along corridors at low elevations on the western slope.

8800 & 8650', Sharples 512 & 716. Greater USRM.

*Rudbeckia hirta L., uncommon in the Conejos valley. 9000', Douglas 1070 (CS). TNA.

Rydbergia grandiflora (Torr. & A. Gray) Greene, common on stable soils on tundra. 12400 &

11500', Sharples 229 & 458. USRM.

Senecio atratus Greene, common in open areas. 8700 & 9100', Sharples 301 & 442. Greater

SRM.

Senecio crassulus A. Gray, common in upper elevation openings. 11600, 12250 & 12150',

Sharples 331, 723 and Sharples & Tripp 1136. WUSA.

Senecio eremophilus Richardson, scattered across lower open areas. 9250, 10300 & 10650',

Sharples 729, 1002 and Sharples & Tripp 1119. Greater ROMO.

Senecio fremontii Torr. & A. Gray, scattered on gravelly riparian ground below treeline. 11200

& 10200', Sharples 367 & 615. WNA.

*Senecio integerrimus Nutt., rare on lowland loams. 9000', Douglas 1076 (CS). WNA.

Senecio serra Hook., uncommon in mixed forest. 9200–9240', Nelson 62540 (RM). WUSA. 51

Senecio triangularis Hook., common along wooded subalpine streams. 11550, 10800 & 10150',

Sharples 401, 686 & 1100. WNA.

*Senecio wootonii Greene, scattered in openings below treeline. 10150 & 11000', Sharples &

Tripp 42 and Sharples 372. SWNA.

*Solidago canadensis L., rare on floodplains. 7990', M Smith 94-MS-292 (CS). TNA.

Solidago multiradiata Aiton, scattered in high meadows. 11900', Sharples & Tripp 1145.

WNNA.

Solidago simplex Kunth [S. spathulata], common at lower elevations. 8650, 9250, 11350 &

9550', Sharples 278, 547, 1003 & 1089. WNNA

Solidago velutina DC, rare along low roadsides. 7900', Sharples 704. WUSA; MX.

*Symphyotrichum ascendens (Lindl.) G.L. Nesom [Aster, Virgulus], uncommon in montane forests. 8535–8730', Flaig 2884 (RM). WNA.

*Symphyotrichum falcatum (Lindl.) G.L. Nesom, uncommon in the ponderosa-oak zone. 8130–

8535', Flaig 2865 (RM). WNA.

Symphyotrichum foliaceum (DC) G.L. Nesom [Aster], common at high elevations. 10200 &

11000', Sharples 1012 & 1368. WNA.

*Symphyotrichum laeve (L.) Á. Löve & D. Löve, uncommon in the ponderosa-oak zone. 8200',

Sharples 698. TNA.

*Symphyotrichum lanceolatum (Willd.) G.L. Nesom [Aster hesperius], uncommon in ponderosa- oak lands. 8700', Heil 12646 (CS). TNA.

Symypyhotrichum spathulatum (Lindl.) G.L. Nesom, uncommon.10700', Sharples 1108. WNA.

Taraxacum officinale F.H. Wigg., abundant across elevations. 10350 & 11500', Sharples &

Tripp 8 and Sharples 457. EUA (W). 52

Taraxacum scopulorum (A. Gray) Rydb., uncommon on tundra. Douglas 2050 (RM). ROMO;

Canadian arctic and Greenland.

*Tetradymia canescens DC, rare in low, dry areas. 7500 & 8750', Gilbert 66 (CS) and Sharples

609. WUSA.

*Tetraneuris ivesiana Greene, scattered across the lowest western slope elevations. 7300 &

8550', Harrington 8080 (CS) and Sharples 132. 4CR.

Tonestus pygmaeus (Torr. & A. Gray) A Nelson, common on tundra. 12300', Sharples 629 &

680. USRM.

*Townsendia glabella A. Gray, uncommon in ponderosa-oak zone openings. 8600' (Sharples),

O’Kane 3742 (SJNM) and Sharples 131. SRM.

Townsendia hookeri Beaman, rare in low openings. 9100', Sharples 760. New to SSJ and

Wilderness. WNA.

*Tragopogon dubius Scop., infrequent in low, dry sites. 9500–9800 & 9000', Flaig 2957 (RM) and Sharples 60. EUA.

Tragopogon pratensis L., uncommon along low pathways. 9050', Sharples 523. EUA (W).

*Tripleurospermum inodorum (L.) Sch. Bip., rare along roadsides. 9350', Sharples 1345. New to

SSJ of AA. EUA.

*Wyethia arizonica A. Gray, common in ponderosa-oak openings. 7900', Sharples 98. 4CR.

BERBERIDACEAE – 2 species

*Berberis fendleri A. Gray, common on the lowest slopes. 9100 & 8550', Sharples & Tripp 36 and Sharples 134. 4CR; not AZ.

Mahonia repens (Lindl.) G. Don, common in low elevation forest understory and openings. 8900

& 10000', Sharples 65 & 152. WNA. 53

BETULACEAE – 2 species

Alnus incana (L.) Moench subsp. tenuifolia (Nutt.) Breitung, common tree of low riparian zones.

9500, 9100 & 8600', Sharples 164, 772 & 806. Subspecies: WNA; species: CIRC.

*Betula occidentalis Hook., rare on the western slope. Heil 13099 (SJNM). WNNA.

BORAGINACEAE – 11 species

*Cynoglossum officinale L., rare around cattle tanks. 7650', Sharples 123. Found just west of study area, but new to eastern AA and likely found in the study area on private land. EUA.

Eritrichium nanum (Vill.) Schrad. ex Gaudin, scattered on wind-swept ridges. 12700, 12300 &

12200', Sharples 651, 678 & 868. CIRC.

Hackelia floribunda (Lehm.) I.M. Johnst., uncommon. 8750', Sharples 544. WNA.

Lappula occidentalis (S. Watson) Greene [L. redowskii], common in open, low-elevation sites.

7800, 8750, 9100 & 8350', Sharples 194, 253, 445 & 764. TNA.

*Lithospermum incisum Lehm., common in lowlands. 8600 & 7200', Sharples 72 & 751. TNA.

Lithospermum multiflorum Torr. ex A. Gray, common at lower elevations. 8750 & 9100',

Sharples 245 & 453. 4CR.

Mertensia ciliata (James ex Torr.) G. Don, scattered across elevations. 8650', Sharples 279.

WUSA.

Mertensia franciscana A. Heller, common in subalpine forests. 9800 & 9450', Sharples 156 &

561. 4CR.

Mertensia lanceolata DC [M. bakeri, M. fusiformis, M. viridis], common across elevations.

10300 & 11800', Sharples 202 & 346. ROMO.

*Myosotis scorpioides L., rare in low elevation creeklets. 8150', Sharples 951. New to SSJ, AA, and S CO. EUA. 54

*Plagiobothrys scouleri (Hook. & Arn.) I.M. Johnst., uncommon in disturbed areas. 7650 &

9000', Sharples 139 & 997. WNNA.

BRASSICACEAE – 37 species

*Arabis pycnocarpa M. Hopkins [A. eschscholtiziana, A. hirsuta], uncommon in montane openings. 9150–9400', Nelson 62399 (RM). NAA.

*Barbarea vulgaris W.T. Aiton, rare near creeks. 9500', Sharples 176. New to SSJ (AA). EUA.

Boechera grahamii (Lehm.) Windham & Al-Shehbaz [B. divaricarpa], rare on floodplains. 8600

& 9200', Sharples 844 & 1380. New to Wilderness (AA & CN). WNNA.

Boechera spatifolia (Rydb.) Windham & Al-Shehbaz, scattered in lowland openings. 8600',

Sharples 811. New to Wilderness (AA). SRM.

Boechera stricta (Graham) Al-Shehbaz [B. drummondii], common in open areas. 8700, 10825,

11550, 11200, 9200, 9100 & 12150', Sharples & Tripp 18, Sharples 218, 323, 369, 433, 761 and

Sharples & Tripp 1133. WNNA.

*Camelina microcarpa DC., uncommon in disturbed sites in the ponderosa-oak zone. 7760–

7920', Nelson 59017 (RM). EUA.

*Capsella bursa-pastoris (L.) Medik., common at lower elevations. 8000', Sharples 82. EUA.

Cardamine cordifolia A. Gray, abundant along streams, particularly throughout the subalpine.

8500, 7600, 9800, 11550 & 12000', Sharples 74, 93, 157, 322 & 1051. WNA.

Descurainia californica (A. Gray) O.E. Schulz, rare in lowlands. 9150', Sharples 573. New to

SSJ of AA, and new to Wilderness. WNA.

Descurainia incana (Bernh. ex Fisch. & C.A. Mey.) Dorn [D. richardsonii]., common in open areas. 11600, 11550, 10250 & 10800', Sharples 312, 321, 481 & 685. WNNA. 55

Descurainia incisa (Engelm. ex A. Gray) Britton, uncommon on low elevation barren sites.

8500', Sharples 783. WNA.

Descurainia sophia (L.) Webb ex Prantl, scattered in disturbed ground. 7450', Sharples 540.

EUA (W).

*Draba albertina Greene, rare. 11352', Douglas & Wilken 1217 (CS). WNA.

Draba aurea Vahl ex Hornem., common in open areas. 10050, 10250, 9900, 11500 & 12900',

Sharples & Tripp 49, 57, and Sharples 578, 855 & 860. WNNA.

Draba crassifolia Graham, scattered on bare ground along high creeksides. 12250 & 11500',

Sharples 640 & 854. New to AA. ANOM: WNNA, and N Europe.

Draba fladnizensis Wulfen, rare on tundra. Heil 12351 (SJNM). CIRC.

Draba helleriana Greene, common in open areas. 12000, 11500 & 8400', Sharples 333, 465 &

755. 4CR; not UT.

*Draba malpighiacea Windham & Al-Shehbaz, rare in the subalpine. 10663', Douglas & Wilken

1136 (CS). SRM.

Draba smithii Gilg & O.E. Schulz, rare in high openings. 11600', Sharples 1005 & 1109. New to

Wilderness (AA). SRM.

Draba spectabilis Greene, scattered on high slopes. 11484', Douglas 1713 (CS). SRM.

Draba streptobrachia R.A. Price, common on high, bare ground. 10750, 11900 & 12400',

Sharples 206, 385 & 858. SRM.

Draba streptocarpa A. Gray, uncommon above treeline. 12300', Sharples 677. New to CN.

SRM. 56

Erysimum capitatum (Douglas ex Hook.) Greene, common in open sites of upper elevations.

10250, 10750, 11500, 11400 & 12900', Sharples & Tripp 47 and Sharples 204, 460, 817 & 1068.

ANOM.

*Erysimum repandum L., uncommon in the ponderosa-oak zone. 7924', Flaig 3866 (RM). EUA.

*Hesperidanthus linearifolius (A. Gray) Rydb., restricted to low, dry elevations. 8750', Sharples

605. 4CR; not UT; MX.

*Lepidium alyssoides A. Gray, rare at the lowest elevations. 8850' (Sharples), Hartwell 2118

(DBG) and Sharples 602. Broader 4CR.

*Lepidium campestre (L.) W.T. Aiton, uncommon in the ponderosa-oak zone. 7760–7920',

Nelson 59010 (RM). EUA.

*Lepidium densiflorum Schrad., uncommon along roads. 9250', Sharples 870. TNA.

*Lepidium ramosissimum A. Nelson, scattered along corridors. 8200', Sharples 702. WNNA.

*Lepidium virginicum L., scattered across disturbed ground. 9500–9800', Flaig 2959 (RM).

TNA.

Noccaea fendleri (A. Gray) Holub [Thlaspi montanum], ubiquitous in the early season. 10350,

11250 & 10150', Sharples 6, 365 & 814. WUSA.

Rorippa alpina (S. Watson) Rydb., scattered. 10350 & 11900', Sharples 837 & 970. New to

Wilderness (AA). USRM.

Rorippa curvipes Greene, common in higher, muddy ground. 11400 & 10350', Sharples 645 &

838. WNA.

Rorippa palustris (L.) Besser, common in low, moist areas. 9450, 8600, 9250 & 10150', Sharples

171, 268, 869 & 1101. TNH. 57

Smelowskia americana Rydb., occasional in rocky tundra areas. 11850, 12300, 11900 & 13100',

Sharples 213, 349, 384 & 652. WNA.

*Thlaspi arvense L., scattered in disturbed sites of the ponderosa-oak zone. 7650 & 7450',

Sharples 126 & 537. EUA.

Turritis glabra L., rare in the mixed forest zone. 9200–9240', Nelson 62556 (RM). TNH.

CACTACEAE – 1 species

*Opuntia fragilis (Nutt.) Haw., rare in the ponderosa-oak zone. Heil 11795 (SJNM). WNNA.

CAMPANULACEAE – 2 species

Campanula parryi A. Gray, scattered in low meadows. 9200', Sharples 426. USRM.

Campanula rotundifolia L., common in open sites across elevations. 9200 & 11900', Sharples

425 & 644. TNH.

CANNABACEAE – 1 species

*Humulus lupulus L., uncommon in low-elevation screes. 8050 & 1354', Sharples 710 & 1354.

TNH.

CAPRIFOLIACEAE – 6 species

*Linnaea borealis L., rare in subalpine forest. 10000', Gierisch 1020 (RM). CIRC.

Lonicera involucrata (Richardson) Banks ex Spreng., occasional in the subalpine. 8500 &

11100', Sharples 75 & 354. WNNA.

Symphoricarpos rotundifolius A. Gray, abundant in the mixed forest zone. 8550, 9700, 9500 &

9900', Sharples 116, 292, 518 & 580. SWNA.

Valeriana acutiloba Rydb., common across subalpine forests. 12000, 10350 & 11300', Sharples

337, 417 & 592. WUSA. 58

Valeriana edulis Nutt., scattered in open areas across elevations. 12050, 8800, 9550 & 9000',

Sharples 236, 511, 526 & 601. WNA.

Valeriana occidentalis A. Heller, uncommon in open sites. 10300', Sharples 200. WUSA.

CARYOPHYLLACEAE – 23 species

Arenaria lanuginosa (Michx.) Rohrb. [Spergulastrum], common in open areas. 8750, 10650,

9700, 8250, 8950, 11000, 10000 & 9150', Sharples 931, Sharples & Tripp 1120, 1297, 1305,

1310, 1369, 1371 & 1379. ANOM: warm parts of the Americas.

Cerastium arvense L. [C. strictum], common in upper elevation openings. 10050, 10825, 12400,

10200 & 10200', Sharples 48, 217, 230, 616 & 1020. ANOM: Americas & Europe.

Cerastium beeringianum Cham. & Schltdl., common on tundra. 11950 & 12150', Sharples 912 and Sharples & Tripp 1143. WNNA.

Cerastium fontanum Baumg., uncommon. 9150', Sharples 571. EUA (W).

*Eremogone fendleri (A. Gray) Ikonn., rare. 7200 & 9050', Harrington 8592 (CS) and Sharples

288. 4CR; into WY.

Minuartia macrantha (Rydb.) House [Alsinanthe], uncommon on rocky streambanks. 9150 &

10200', Sharples 574 & 619. 4CR.

Minuartia obtusiloba (Rydb.) House, common on open subalpine ridges and slopes and throughout the alpine zone. 11850, 12150, 12050, 12000 & 12150', Sharples 386, 834, 1066,

1116 & 1141. NAA.

Minuartia rubella (Wahlenb.) Hiern, scattered in dry upper elevation areas. 11350, 11950 &

12150', Sharples 221, 913 & 1142. New to AA. CIRC.

*Moehringia macrophylla (Hook.) Fenzl [Arenaria], rare. Heil et al. 19915 (SJNM). NAA. 59

Pseudostellaria jamesiana (Torr.) W.A. Weber & R.L. Hartm., uncommon in lowland forest.

9050', Sharples 525. WUSA.

Sagina saginoides (L.) H. Karst., scattered along streams. 8650', Sharples 284. CIRC.

Silene acaulis (L.) Jacq., common in the alpine. 11400 & 12900', Sharples 223 & 635. CIRC.

Silene drummondii Hook., scattered in understory of montane forests. 8500 & 10150', Sharples

779 & 1001. WNA.

Silene hitchguirei Bocquet, rare in gravelly soil at upper elevations. 11418', Douglas 1778 (CS).

USRM; into Alberta.

*Silene latifolia Poir., rare along roadsides. Heil 12419 (SJNM). EUA.

Silene menziesii Hook., rare. 10150', Sharples 815. New to AA Wilderness. WNA.

Silene scouleri Hook., common below treeline. 10900, 8400 & 10000', Sharples 1008, 1071 &

1099. WNA.

Spergularia rubra (L.) J. Presl & C. Presl, uncommon on disturbed subalpine ground. 11050',

Sharples 1361. EUA (W).

Stellaria crassifolia Ehrh., uncommon. 9700' from Sharples 1295. CIRC.

Stellaria nom. nov. [Alsine polygonoides Greene ex Rydb.; Stellaria irrigua Bunge sensu W.A.

Weber], scattered on eroded alpine tuffs and screes. 12900, 11950, 12450, 12700, 11900, 12500,

11850, 12200, 12500 & 12200' from Sharples 654, 918, 926, 967, 969, 1010, 1036, 1050, 1062

& 1367. SRM; volcanic ranges.

Stellaria longifolia Muhl. ex Willd., scattered in wet meadows. 9550', Sharples 1318. CIRC.

Stellaria longipes Goldie, scattered in open areas across elevations. 10850 & 9100', Sharples 835

& 872. New to AA Wilderness. CIRC.

Stellaria umbellata Turcz., scattered. 9700 & 10800', Sharples 1294 & 1301. NAA. 60

CELASTRACEAE – 2 species

Parnassia fimbriata K.D. Koenig, rare. 9898', Flaig 2510 (RM). WNA.

Paxistima myrsinites (Pursh) Raf., uncommon in mixed forest understory. 10200 & 9600',

Sharples 216 & 748. WNA.

CERATOPHYLLACEAE – 1 species

*Ceratophyllum demersum L., rare in low elevation ponds. 8950', Sharples 1313. New to AA.

COSM.

CONVOLVULACEAE – 1 species

*Convolvulus arvensis L., scattered along ponderosa-oak zone roads. 7300–7500 & 7200',

Hartman 77410 (RM) and Sharples 143. EUA.

CORNACEAE – 2 species

*Cornus canadensis L., rare. 9600', Komarek 503 (FLD). NAA.

Cornus sericea L., scattered across low elevations. 9200', Sharples 306. TNA.

CRASSULACEAE – 3 species

Rhodiola integrifolia Raf. [Tolmachevia], common in rocky habitats in middle-high elevations.

13300', Sharples 634. NAA.

Rhodiola rhodantha (A. Gray) H. Jacobsen [Clementsia], common along high creeks and moist meadows. 10200, 12200 & 10200', Sharples 618, 1009 & 1014. USRM.

Sedum lanceolatum Torr., common in rocky, well-drained sites across elevations. 9150',

Sharples 568. WNA.

CUPRESSACEAE – 2 species

Juniperus communis L., scattered throughout forest understory and open areas. 9100 & 9200',

Sharples 33 & 434. TNH. 61

*Juniperus scopulorum Sarg., scattered in the ponderosa-oak zone. 7600', Sharples 137. WNA.

CYPERACEAE – 50 species

Carex albonigra Mack., uncommon in high openings. 11850', Sharples 211. WNA.

Carex aquatilis Wahlenb., common in high boggy areas. 11350, 11500, 12250 & 11750',

Sharples 396, 461/464, 907 & 1027. CIRC.

*Carex athrostachya Olney, uncommon in wetlands. 11350', Sharples 397. WNA.

Carex atrosquama Mack., rare in high elevation wetlands. 11100', Sharples 357. WNA.

Carex aurea Nutt., uncommon in low elevation wetlands. 9000', Douglas 1074 (CS). WNNA.

Carex bella L.H. Bailey, scattered in wetlands. 9550', Sharples 953. SWNA.

Carex canescens L., scattered in subalpine wetlands. 11557–11814', Flaig 7974 (RM). COSM.

Carex chalciolepis Holm [C. heteroneura], scattered in high elevation wetlands. 12189', Douglas

1722 (CS). Greater USRM.

*Carex disperma Dewey, rare in seeps. 9442', Flaig 4740 (RM). CIRC.

*Carex douglasii Boott, rare. 8789', Flaig 4652 (RM). WNA.

Carex ebenea Rydb., common in high, wet areas. 10200, 11600 & 12000', Sharples 728, 1018 &

1053. Greater 4CR.

Carex echinata Murray [C. angustior], rare in moist meadows. 10150', Sharples 958. New to SSJ of AA (and Wilderness). CIRC.

*Carex egglestonii Mack., rare in montane openings. 9800–10200', Nelson 62329 (RM). SRM.

Carex elynoides Holm, scattered in high meadows. 11218', Flaig 4863 (RM). USRM.

*Carex geyeri Boott, uncommon along watercourses at lower elevations of the western slope.

8780–9200', Hartman 79711 (RM). Mostly WUSA. 62

Carex haydeniana Olney, common in wet sites and high meadow areas. 11850', Sharples 1038.

Mostly WUSA.

Carex illota L.H. Bailey, scattered across high wetlands. 11750', Sharples 1032. Mostly WUSA.

*Carex inops L.H. Bailey [C. pensylvanica], uncommon at low elevations. 7924', Flaig 3904

(RM). Mostly WUSA.

Carex jonesii L.H. Bailey, scattered in riparian areas. 10900', Sharples 624. WUSA.

Carex limosa L., uncommon in high wetlands. 10998', Harrington 1830 (CS). CIRC.

Carex micropoda C.A. Mey. [C. crandallii, C. pyrenaica], common on rocky tundra. 12400 &

12200', Sharples 1047 & 1064. NAA.

Carex microptera Mack., uncommon in montane wetlands. 8780–9200', Hartman 79617 (RM).

WNA.

Carex nelsonii Mack., uncommon in high wetlands. 12274', Douglas 2048 (CS). USRM.

Carex nigricans C.A. Mey., scattered in high openings. 12189', Douglas 12189 (CS). WNA.

Carex nova L.H. Bailey, common in higher elevation wetlands. 11400', Sharples 646. WUSA.

*Carex occidentalis L.H. Bailey, uncommon in ponderosa-oak zone wetlands. 10111', Flaig

3752 (RM). WUSA.

Carex oreocharis Holm, rare on high rock slopes. 11812', Douglas 1755 (CS). SRM.

Carex pachystachya Cham. ex Steud., scattered in lower forests. 7600, 8600, 8750 & 9800',

Sharples 94, 264, 936 & 1095/1096. NAA.

*Carex pellita Muhl. ex Willd. [C. lanuginosa], rare in low wetlands. 9000', Douglas 1065 (CS).

TNA.

Carex perglobosa Mack., rare on tundra screes. 11850', Sharples 1037. New to CN. SRM. 63

Carex phaeocephala Piper, uncommon in open areas. 11600', Sharples & Tripp 1121. New to

CN. WNA.

Carex praeceptorum Mack., uncommon in high openings. 11814–12237', Flaig 7949 (RM).

Essentially WUSA.

*Carex praegracilis W. Boott, rare in montane forests. 9442', Flaig 4704 (RM). TNA.

Carex retrorsa Schwein., scattered in ponderosa-oak wetlands. 9400', Sharples 789. WNNA.

Carex rossii Boott, uncommon in wetlands below treeline. 11484', Douglas 1743 (CS). WNNA.

Carex saxatilis L., scattered in high marshlands. 11750', Sharples 1376. CIRC.

Carex scopulorum Holm, scattered in high wetlands. 11400 & 11750', Sharples 647 & 1031.

WNA.

Carex siccata Dewey [C. foenea], scattered in wetlands. 10350', Sharples & Tripp 9. WNNA.

Carex stevenii (Holm) Kalela [C. norvegica], rare in high marshes. Harrington 1834 (CS).

USRM.

Carex utriculata Boott, common in riparian areas. 9500, 9500, 8600 & 9800', Sharples 158, 178,

265 & 1093. CIRC.

Carex vernacula L.H. Bailey [C. foetida], scattered on high saturated ground. 12250', Sharples

908. Parts of WUSA.

Eleocharis acicularis (L.) Roem. & Schult., uncommon along pond shores. 10200', Sharples

663. New to CN Wilderness. TNH.

Eleocharis palustris (L.) Roem. & Schult. [E. xyridiformis], occasional in low elevation wetlands. 9450 & 10000', Sharples 172 & 1370. New to CN Wilderness. TNH.

Eleocharis quinqueflora (Hartmann) O. Schwarz [E. pauciflora], occasional in high marshes.

11800', Sharples 1115. CIRC. 64

Eriophorum angustifolium Honck., uncommon in high bogs. 11500', Sharples 1363. CIRC.

Eriophorum gracile W.D.J. Koch ex Roth, uncommon in high bogs. 12200', Sharples 1366.

CIRC.

Kobresia myosuroides (Vill.) Fiori, uncommon in high meadows. 11850', Sharples 215. CIRC.

*Schoenoplectus acutus (Muhl. ex Bigelow) Á. Löve & D. Löve [Scirpus], rare around low ponds. Heil 12677 (SJNM). TNA.

Schoenoplectus tabernaemontani (C.C. Gmel.) Palla, rare in low wetlands. 7650', Sharples 124.

COSM.

*Scirpus microcarpus J. Presl & C. Presl, uncommon along floodplains. 8720–9200 & 8600',

Nelson 62496 (RM) and Sharples 270. TNA.

DENNSTAEDTIACEAE – 1 species

Pteridium aquilinum (L.) Kuhn, common in understory of open mixed forest on the western slope. 8850', Sharples 722. COSM.

DRYOPTERIDACEAE – 4 species

*Athyrium filix-femina (L.) Roth, uncommon. 9843', Douglas 2144 (CS). TNH.

Cystopteris fragilis (L.) Bernh. [C. reevesiana], common in rock crevices across elevations.

11600, 10050 & 8600', Sharples 328, 667 & 808. COSM.

*Gymnocarpium dryopteris (L.) Newman, rare in gorges. 9500', S. Komarek 480 (FLD). CIRC.

*Woodsia oregana D.C. Eaton, rare. 9500–9800', Flaig 2955 (RM). WNNA.

ELAEAGNACEAE – 1 species

Shepherdia canadensis (L.) Nutt., scattered in mixed forest understory. 9100, 9600 & 9300',

Sharples 30, 749 & 800. WNNA.

EQUISETACEAE – 3 species 65

Equisetum arvense L., infrequent along creeks in montane forests. 9100', Sharples 786. TNH.

*Equisetum hyemale L., scattered along riparian corridors. 9500', Sharples 491. TNH.

*Equisetum laevigatum A. Braun, uncommon along low riparian areas. 8780–9200', Hartman

79671 (RM). WNA.

ERICACEAE - 14 species

Arctostaphylos uva-ursi (L.) Spreng., scattered in low forest openings. 9150', Sharples 770.

TNH.

Chimaphila umbellata (L.) W.P.C. Barton, uncommon in ponderosa understory. New to AA

Wilderness. 8550', Sharples 805. TNH.

Gaultheria humifusa (Graham) Rydb., rare. 11484', Douglas 2067 (CS). WNA.

Kalmia microphylla (Hook.) A. Heller, rare. 11484', Douglas 2066 (CS). WNA.

Moneses uniflora (L.) A. Gray, scattered in subalpine forest understory. 10850', Sharples 594.

CIRC.

Orthilia secunda (L.) House, uncommon in moist forest understory. 9700', Sharples 734 and

Sharples & Tripp 1296. CIRC.

Pterospora andromedea Nutt., rare. 11000', Douglas 1353. WNNA.

Pyrola asarifolia Michx. [P. rotundifolia], scattered in wet forest understory. 9700', Sharples

597. NAA.

Pyrola chlorantha Sw., scattered in wet forest understory. 9450', Sharples 558. CIRC.

Pyrola minor L., scattered in subalpine forest understory. 10850', Sharples 1373. CIRC.

*Pyrola picta Sm., rare. 9843', Douglas 2143a (CS). Mostly WUSA.

Vaccinium cespitosum Michx., uncommon on stable tundra slopes. 12350 & 12050', Sharples

829 and Sharples & Tripp 1131. New to AA. WNNA. 66

Vaccinium myrtillus L., abundant in subalpine forest understory and clearings. 10000, 11350,

10800 & 9550', Sharples 151, 777, 928 & 954. CIRC.

Vaccinium scoparium Leiberg ex Coville, common in subalpine openings. 11450 & 11600',

Sharples 1029 and Sharples & Tripp 1125. WNA.

EUPHORBIACEAE – 1 species

*Chamaesyce serpyllifolia (Pers.) Small, rare along low roadsides. 8200', Sharples 701. New to

SSJ. TNA.

FABACEAE – 28 species

*Astragalus agrestis Douglas ex G. Don, uncommon. 8789', Flaig 4591 (RM). NAA.

Astragalus alpinus L., common in forest understory. 9100, 10350, 9150 & 10200', Sharples &

Tripp 27 and Sharples 416, 771 & 1023. CIRC.

*Astragalus bisulcatus (Hook.) A. Gray, common in the ponderosa-oak zone. 7750', Sharples

753. Mostly WUSA.

*Astragalus lonchocarpus Torr., common in the ponderosa-oak zone. 8600, 7200 & 8500',

Sharples 129, 750 & 781. 4CR; NV.

Hedysarum occidentale Greene, occasional on eroded slopes and other openings. 10800, 8600, &

10200', Sharples 406, 807 & 1041. New to AA Wilderness. Greater USRM.

Lathyrus leucanthus Rydb., common in forest understory. 8000 & 10850', Sharples 76 & 531.

New to Wilderness (AA). 4CR; S WY.

*Lotus wrightii (A. Gray) Greene, scattered in the ponderosa-oak zone. 7721 & 7950', Flaig

4924 (RM) and Sharples 778. 4CR.

Lupinus argenteus Pursh, common in open areas below treeline. 9200', Sharples 427. WUSA. 67

*Lupinus kingii S. Watson, scattered in ponderosa-oak savannah. 7750', Sharples 86. 4CR; E

NV.

*Lupinus polyphyllus Lindl. [including L. ammophilus], uncommon at low elevation. 8730 &

8150', Flaig 3779 (RM) and Sharples 758. WNNA.

Lupinus sericeus Pursh, common in the ponderosa-oak zone. 7800', Sharples 111. Mostly

WUSA.

*Medicago lupulina L., uncommon on disturbed ground of the western slope. 7750 & 8250',

Sharples 87 & 700. EUA.

*Medicago sativa L., rare. 8130–8535', Flaig 2858 (RM). EUA.

*Melilotus albus Medik., rare along ponderosa-oak zone roads. Heil 12603 (SJNM). EUA.

*Melilotus officinalis L., scattered along roadsides. 8730, 8400 & 9500', Flaig 3781 (RM) and

Sharples 136 & 257. EUA.

Oxytropis lambertii Pursh, uncommon at low elevation dry sites. 8750 & 9100', Sharples 610 &

769. ANOM: central North America.

Oxytropis podocarpa A. Gray, rare. 12274', Douglas 2070 (CS). WNNA.

Oxytropis splendens Douglas ex Hook., rare. 9200', Sharples 428. New to Wilderness (CN).

WNNA.

Thermopsis montana Nutt., common in mixed forest zone. 8700, 8000 & 9200', Sharples &

Tripp 16 and Sharples 101 & 451. WUSA.

Trifolium attenuatum Greene, common in the upper subalpine and on tundra. 10750, 12050,

11900, 12900 & 12050', Sharples 203, 235, 534, 1067 & 1282. New to AA Wilderness. SRM.

Trifolium brandegeei S. Watson, scattered in the upper subalpine and on tundra. 10750, 12050,

11250 & 12800', Sharples 205, 234, 363 & 637. SRM. 68

*Trifolium hybridum L., uncommon. 8780–9200 & 8650', Hartman 79625 (RM) and Sharples

1283. EUA.

*Trifolium longipes Nutt., common in lowlands. 8000 & 9550', Sharples 77 & 159. WUSA.

Trifolium nanum Torr., uncommon on barren tundra. 12050', Sharples 238. USRM.

Trifolium parryi A. Gray, uncommon on tundra. 12400', Sharples 1046. USRM.

Trifolium pratense L., common along low disturbed corridors, e.g. floodplains and trails. 8650,

9250 & 9300', Sharples 280, 557 & 802. EUA (W).

Trifolium repens L., common aong disturbed ground in the mixed forest zone. 8650, 9200, 9250

& 9500', Sharples 281, 447, 549 & 1087. EUA (W).

Vicia americana Muhl. ex Willd., common in mixed forest understory. 7750, 9100 & 9100',

Sharples 85, 452 & 788. New to CN Wilderness. TNA.

FAGACEAE – 1 species

Quercus gambelii Nutt., a dominant tree of the western slope lowlands. 8000 & 8600', Sharples

83 & 715. Greater 4CR.

GENTIANACEAE – 10 species

Frasera speciosa Douglas, scattered on the western slope at low-middle elevations. 8600 &

9500', Sharples 502 & 513. WUSA.

Gentiana algida Pall., common in high, open areas. 11650 & 12500', Sharples 1006 & 1011.

NAA.

*Gentiana fremontii Torr. [G. aquatica], rare. 8789', Flaig 4623 (RM). Mostly WUSA.

Gentiana parryi Engelm., common in high meadows. 11900', Sharples 643. SRM.

Gentiana prostrata Haenke, uncommon in “understory” of tundra vegetation. 12450, 12450 &

12700', Sharples 650, 1044 & 1144. New to AA (MS 1044). CIRC. 69

Gentianella acuta (Michx.) Hiitonen [G. amarella], common in high openings. 11600, 11800,

11900 & 12400', Sharples 674, 1007, 1043 & 1063. CIRC.

Gentianella heterosepala (Engelm.) Holub [G. amarella], uncommon in forest openings. 9000',

Sharples 944. 4CR; WY.

Gentianella tenella (Rottb.) Börner, rare on tundra. 12400', Sharples 1049. New to AA. CIRC.

Gentianopsis thermalis (Kuntze) Iltis, common in open areas. 11250 & 8750', Sharples 666 &

996. USRM.

Swertia perennis L., common in moist subalpine meadows. 11900 & ~11600', Sharples 642 &

725. CIRC.

GERANIACEAE – 2 species

Geranium caespitosum E. James, scattered in low areas. 9000, 8900, 8750 & 9300', Sharples 61,

67, 244 & 289. Greater SRM.

Geranium richardsonii Fisch. & Trautv., common in open areas. 9800, 8600, 11600 & 9500',

Sharples 155, 273, 311 & 516. WNA.

GROSSULARIACEAE – 6 species

*Ribes inerme Rydb., uncommon in lowlands. 10050', Sharples & Tripp 52. WNA.

Ribes lacustre (Pers.) Poir., rare. 11850', Sharples 831. WNNA.

Ribes laxiflorum Pursh [R. coloradense], rare. 10992–12569', Flaig 2478 (RM). NAA.

*Ribes leptanthum A. Gray, uncommon at low elevation. 9100', Sharples & Tripp 38. 4CR.

Ribes montigenum McClatchie, scattered in forest understory, openings, and subalpine ridges.

9500' & 12000', Sharples 175 & 821. Essentially WUSA.

Ribes wolfii Rothr., common at low elevations. 9100, 10500, 10300 & 9300', Sharples & Tripp

26 and Sharples 149, 773 & 801. 4CR; disjunct in PNW. 70

HALORAGACEAE – 1 species

*Myriophyllum sibiricum Kom., scattered in low lakes. 9600', Heil 12742 (SJNM). NAA.

HYDRANGEACEAE – 1 species

Jamesia americana Torr. & A. Gray, rare. 10400', Sharples 774. New to Wilderness (CN).

SWNA.

HYDROCHARITACEAE – 1 species

*Elodea canadensis Michx., uncommon in lakes. O’Kane et al. 6911 (SJNM). TNA.

HYDROPHYLLACEAE – 5 species

*Hydrophyllum capitatum Douglas ex Benth., rare. 9515', Douglas 1899 (CS). WUSA.

Hydrophyllum fendleri (A. Gray) A. Heller, common in montane forests and openings. 9800 &

9200', Sharples 154 & 840. Essentially WUSA.

*Phacelia bakeri (Brand) J.F. Macbr., rare. 9250 & 10450', Heil 24980 (SJNM) and Sharples

613. SRM

Phacelia heterophylla Pursh, common in low forests. 9600 & 9150', Sharples 488 & 566.

WUSA.

Phacelia sericea (Graham) A. Gray, common on and around unstable rock slopes. 11600, 8600

& 12000', Sharples 314, 820 & 843. WNA.

HYPERICACEAE – 1 species

*Hypericum scouleri Hook., rare on rocky floodplains. 7850', Sharples 496. WNA.

IRIDACEAE – 1 species

Iris missouriensis Nutt., common in open sites from low-middle elevations. 9200', Sharples 15 &

441. Essentially WUSA.

ISOËTACEAE – 2 species 71

Isoëtes bolanderi Engelm., scattered in subalpine ponds. 11650, 11500 & 11450', Sharples 402,

1026 & 1030. USRM.

Isoëtes echinospora Durieu, uncommon in high eastern slope lakes. 11484', Douglas 2072 (CS).

CIRC.

JUNCACEAE – 16 species

Juncus arcticus Willd. [J. balticus], uncommon at low elevations. 9200', Sharples 432. New to

Wilderness (CN). TNH; beyond.

Juncus bufonius L., rare in lower riparian areas. 9020', M. Smith 94MS-96 (CS). COSM.

Juncus castaneus Sm., rare in montane wetlands. 9020', M. Smith 94MS-87 (CS). CIRC.

Juncus confusus Coville, occasional in mixed forest wetlands. 9160–9560', Nelson 62436 (RM).

WNA.

Juncus drummondii E. Mey., common on tundra. 12000, 11600 & 12400', Sharples 336, 727 &

1048. WNA.

*Juncus dudleyi Wiegand, rare. Heil 11485 (SJNM). TNA.

Juncus ensifolius Wikstr. [J. saximontanus, J. tracyi], uncommon along streamcourses. 9250 &

7900', Sharples 550 & 708. CIRC.

*Juncus hallii Engelm., uncommon in riparian areas. Heil 12452 (SJNM). Greater USRM.

Juncus interior Wiegand, rare in marshy areas. 9200–9240', Nelson 62606 (RM). TNA.

*Juncus longistylis Torr., rare. 10000–10080 & 8600', Nelson 62295 (RM) and Sharples 271.

WNNA.

Juncus mertensianus Bong., common in higher wet areas. 11600', Sharples 726. WNA.

Juncus nevadensis S. Watson, uncommon in low wetlands. 9200–9240', Nelson 62557 (RM).

WNA. 72

Juncus parryi Engelm., scattered in subalpine meadows. 11250', Sharples 361. WUSA.

Luzula parviflora (Ehrh.) Desv., common in subalpine forests. 11550 & 11000', Sharples 399 &

848. CIRC.

Luzula spicata (L.) DC, common on tundra. 12000 & 11850', Sharples 335 & 1034. CIRC.

Luzula subcapitata (Rydb.) H.D. Harr., rare in bogs. Lundquist & Rink L72 (SJNM). SRM.

LAMIACEAE – 6 species

Agastache pallidiflora (A. Heller) Rydb., common in openings of the western slope. 10950,

9000, 9650 & 8500', Sharples 581, 600, 946 & 1350. 4CR; TX.

Dracocephalum parviflorum Nutt., common in meadows. 8800', Sharples 1081. TNA.

*Glechoma hederacea L., rare in low, wet soils. 7800', Sharples 113. New to AA (and SSJ).

EUA.

Mentha arvensis L., scattered on disturbed ground on the western slope. 8600 & 9200', Sharples

504 and Sharples & Tripp 1348. CIRC.

Prunella vulgaris L., occasional in wet areas of lower elevations. 9200, 8150 & 8900', Sharples

429, 952 & 974. Native range unclear; TNH and beyond.

*Stachys pilosa Nutt., uncommon along stream courses. 7850', Sharples 497. New to SSJ of AA.

TNA.

LENTIBULARIACEAE – 1 species

Utricularia macrorhiza Leconte [U. vulgaris], uncommon in lower elevation ponds. 10000',

Douglas 2077 (CS). NAA.

LILIACEAE – 4 species

Calochortus gunnisonii S. Watson, common in the ponderosa-oak zone. 8600', Sharples 500.

USRM. 73

Lloydia serotina (L.) Salisb. ex Rchb., rare on subalpine ridges. 11484', Douglas 2092 (CS).

CIRC.

Prosartes trachycarpa S. Watson, common at low elevations. 9100', Sharples & Tripp 22.

WUSA.

Streptopus amplexifolius (L.) DC [S. fassettii], uncommon in dense forest understory. 9550',

Sharples 1317. CIRC.

LINACEAE – 1 species

Linum lewisii Pursh, scattered in low openings. 8000, 9500 & 9250', Sharples 495, 519 & 1083.

WNNA.

MALVACEAE – 4 species

*Malva neglecta Wallr., rare along roads. 9350', Sharples 1344. New to SSJ. EUA.

Sidalcea candida A. Gray, common in the ponderosa-oak zone. 9100, 8600 & 9500', Sharples

494, 508 & 522. Aberrant 4CR.

Sidalcea neomexicana A. Gray, common in the ponderosa-oak zone. 7450 & 8200', Sharples 538

& 697. SWNA.

*Sphaeralcea fendleri A. Gray, rare. Heil 12635 (CS). Aberrant 4CR.

MELANTHIACEAE – 2 species

Veratrum californicum Durand, abundant in subalpine meadows. 10350 & 8800', Sharples 414 &

510. WUSA.

Zigadenus elegans Pursh, common in openings across elevations. 7600, 11600, 10850 & 12500',

Sharples 138, 329, 530 & 859. TNA.

MONTIACEAE – 5 species 74

*Claytonia lanceolata Pursh, uncommon on the western slope. Heil & Heil 21879 (SJNM).

WUSA.

Claytonia megarhiza (A. Gray) Parry ex S. Watson, abundant across alpine scree and talus.

11850, 12100 & 12300', Sharples 387, 823 and Sharples & Tripp 1139. WNA.

*Lewisia nevadensis (A. Gray) B.L. Rob., uncommon in ponderosa-oak understory. 8000',

Sharples 81. Patchy WUSA.

Lewisia pygmaea (A. Gray) B.L. Rob., uncommon in subalpine openings. 11850', Sharples 962.

New to AA Wilderness. WNA.

Montia chamissoi (Ledeb. ex Spreng.) Greene, rare. 9200', Sharples 439. New to Wilderness

(CN). WNA.

NYCTAGINACEAE – 2 species

Mirabilis linearis (Pursh) Heimerl, rare. 9023', Douglas 1252 (CS). TNA.

*Mirabilis melanotricha (Standl.) Spellenb., uncommon on low screes and outcroppings. 8050 &

8500', Sharples 711 & 1351. New to AA and SSJ. SWNA.

NYMPHAEACEAE – 1 species

Nuphar polysepala Engelm., uncommon in ponds. 11000', Sharples 403. WNA.

ONAGRACEAE – 13 species

Chamerion angustifolium (L.) Holub, common on open ground. 9000, 9000 & 9250', Sharples

305, 483 & 554. TNH.

Chamerion latifolium (L.) Holub, uncommon on rocky ground. New to AA. 8650, 9100 &

12150', Sharples 275, 871 & 910. CIRC.

Epilobium anagallidifolium Lam. [E. alpinum], uncommon in alpine riparian zones. 11850',

Sharples 964. CIRC. 75

Epilobium brachycarpum C. Presl, scattered in low meadows. 9200–9240', Nelson 62544 (RM).

WNNA.

Epilobium ciliatum Raf., common in low open areas. 9550 & 8750', Sharples 527 & 545. TNA; beyond.

Epilobium clavatum Trel., rare in high, rocky areas. 10350', Sharples 836. New to the South San

Juans counties (found in AA). WNA.

Epilobium halleanum Hausskn., uncommon. Heil & Mietty 13274 (SJNM). Mostly WUSA.

Epilobium hornemannii Rchb. [including E. lactiflorum], common at middle elevations. 10350',

Sharples 1286. CIRC.

Epilobium saximontanum Hausskn., common in riparian areas. 11350, 9250 & 8600', Sharples

395, 551 & 810. Mostly WUSA.

*Gayophytum diffusum Torr. & A. Gray, uncommon low on the western slope. 8130–8535',

Flaig 2861 (RM). Essentially WUSA.

*Oenothera caespitosa Nutt., common in low openings. 8950' (Sharples), Heil 13301 and

Sharples & Tripp 58. Essentially WUSA.

*Oenothera flava (A. Nelson) Garrett, uncommon in low areas. 9100', Sharples & Tripp 39.

Essentially WUSA.

*Oenothera villosa Thunb., uncommon in lowlands. 7850 & 7900', Sharples 498 & 706. TNA natively.

OPHIOGLOSSACEAE – 6 species

*Botrychium hesperium (Maxon & R.T. Clausen) W.H. Wagner & Lellinger, rare. 10500', Root

91-27 (DBG). Mostly WNA. 76

Botrychium lanceolatum (S.G. Gmel.) Angstr. subsp. lanceolatum, uncommon. ~12100',

Douglas 2010 (CS). CIRC.

Botrychium lunaria (L.) Sw., uncommon. ~12100', Douglas 2011 (CS). COSM.

*Botrychium minganense Vict., rare. Reeves 10662 (SJNM). WNNA.

*Botrychium pallidum W.H. Wagner, rare. 10200', Root 1206 (DBG). ANOM: disjunct from eastern/central Canada.

*Botrychium pinnatum H. St. John, uncommon. Heil et al. 12456 (SJNM). WNA; Greenland.

ORCHIDACEAE – 10 species

*Calypso bulbosa (L.) Oakes, rare in montane forests. 8600', Heil & Mietty 13277 (SJNM).

CIRC.

Corallorhiza maculata Raf., common forest understory. 8000, 9600, 9900 & 10850', Sharples

100, 487, 579 & 595. TNA.

Corallorhiza trifida Châtel., rare. 9187', Douglas 2016a (CS). CIRC.

Corallorhiza wisteriana Conrad, uncommon. Heil & Mietty 13278 (SJNM). TNA.

Goodyera oblongifolia Raf., scattered in lowland forest understory. 9500', Sharples 1086.

WNNA.

Listera cordata (L.) R. Br., uncommon in subalpine forest. 10200', Sharples 874. CIRC.

Platanthera huronensis (Nutt.) Lindl. [P. aquilonis, P. hyperborea], scattered in moist forests.

10550, 8950 & 9550', Sharples 596 and Sharples & Tripp 1311 & 1329. WNNA.

Platanthera obtusata (Banks ex Pursh) Lindl., rare. 11198', Douglas 1343 (CS). CIRC.

Platanthera purpurascens (Rydb.) Sheviak & W.F. Jennings [P. saccata, P. stricta], common in low meadows. 9350 & 10150', Sharples 795 & 956. SWNA. 77

Spiranthes romanzoffiana Cham., uncommon in riparian meadows. 8750 & 11000', Sharples 994

& 1024. New to AA. WNNA.

OROBANCHACEAE – 14 species

Castilleja linariifolia Benth., scattered across lowlands. 9200', Sharples 449. WUSA.

Castilleja lineata Greene, uncommon in montane meadows. 9200', Sharples 430. 4CR; not UT.

Castilleja miniata Douglas ex Hook., common below treeline. 9550, 10300 & 10200', Sharples

167, 201 and Sharples & Tripp 1117. WNA.

Castilleja occidentalis Torr., common in high meadows. 11500', Sharples 318. USRM.

Castilleja rhexifolia Rydb., common in openings. 8650, 11500, 11800 & 11450', Sharples 296,

320, 341, 583 & 584. USRM.

Castilleja sulphurea Rydb., common. 8650 & 9650', Sharples 277 & 949. USRM.

Orthocarpus luteus Nutt., scattered across lowland savannah. 8400 & 8650', Sharples 712 &

719. New to Wilderness (AA). WNA.

Pedicularis bracteosa Benth. subsp. paysoniana (Pennell) W.A. Weber, scattered in the subalpine zone. 11650', Sharples 589. New to AA. Mostly WUSA.

Pedicularis groenlandica Retz., common in wet, open areas below treeline. 8950, 9650, &

11650', Sharples 303, 307 & 586. WNNA.

Pedicularis parryi A. Gray, common on loose ground in the subalpine zone. 11200 & 11700',

Sharples 366 & 472. Greater USRM.

Pedicularis procera A. Gray, scattered throughout mixed forests. 9600', Sharples 598. Greater

4CR.

Pedicularis racemosa Douglas ex Benth. subsp. alba Pennell, common in dry areas of subalpine forests. 11500, 10350 & 10850', Sharples 375, 411 & 532. Mostly WUSA. 78

Pedicularis scopulorum A. Gray, rare on tundra. New to AA. 12650', Sharples 862. SRM.

*Rhinanthus minor L., uncommon, in the Chama valley. 8750', Sharples 995. EUA.

PAPAVERACEAE – 2 species

Corydalis aurea Willd., scattered in low, dry areas. 9100', Sharples & Tripp 37. TNA.

Corydalis caseana A. Gray subsp. brandegeei (S. Watson) G. Ownbey, scattered, mostly in moist meadows. 9100 & 9150', Sharples 493 & 570. Patchy WUSA; subspecies: SRM.

PHRYMACEAE – 2 species

Mimulus guttatus DC, common near flowing water. 8600, 8700, 11500 & 10550', Sharples 266,

300, 400 & 1102. WNA.

Mimulus tilingii Regel, scattered in high areas by flowing water. 11500', Sharples 319. WNA.

PINACEAE – 10 species

Abies bifolia A. Murray, uncommon mixed forest component. 10700', Sharples & Tripp 1148.

New to CN. ROMO.

Abies concolor (Gordon & Glend.) Lindl. ex Hildrebr., uncommon mixed forest component.

8750', Sharples 930. New to Wilderness (AA). WUSA.

Picea engelmannii Parry ex Engelm., abundant/dominant tree of higher elevations. 10500 &

9550', Sharples 145 & 166. WNA.

Picea pungens Engelm., scattered along montane drainages. 8600', Sharples 272. 4CR.

*Pinus aristata Engelm., rare on barren subalpine slopes. 11000', Sharples 373. SRM.

Pinus contorta Douglas ex Loudon, rare in subalpine forest. 10450', Sharples 408. WNA.

Pinus flexilis E. James, scattered on open ridges. 11100 & 10600', Sharples 391 & 775. New to

Wilderness (AA). Essentially WUSA. 79

Pinus ponderosa Douglas ex P. Lawson & C. Lawson, abundant component of the lowest forests. 7900 & 9200', Sharples 99 & 438. Essentially WUSA.

Pinus strobiformis Engelm., uncommon in the subalpine zone. 10700', Sharples 1106. New to

Wilderness (AA). Greater 4CR.

Pseudotsuga menziesii (Mirb.) Franco, common component of lower forests. 8950 & 9800',

Sharples 59 & 241. WNA.

PLANTAGINACEAE – 21 species

Besseya alpina (A. Gray) Rydb., uncommon in alpine cliffs and talus. 12250, 13000 & 12350',

Sharples 826, 833 & 864. New to AA. SRM.

Callitriche palustris L. [C. verna], scattered in ponds. 9450 & 11475', Sharples 168 and Sharples

& Tripp 1375. COSM.

Chionophila jamesii Benth., scattered on tundra. New to AA. 12900', Sharples 639. SRM.

*Collinsia parviflora Lindl., uncommon in the ponderosa-oak zone. 7900 & 8400', Sharples 97

& 756. New to SSJ. WNNA.

Hippuris vulgaris L., occasional in semi-stagnant waters. 8750 & 9400', Sharples 563 & 790.

New to AA Wilderness. TNH.

*Linaria vulgaris Mill., uncommon roadside weed. 8200', Sharples 696. EUA.

Penstemon barbatus (Cav.) Roth, scattered on arid ground at low elevation. 7800 & 9500',

Sharples 191 & 661. 4CR; TX.

Penstemon crandallii A. Nelson [P. caespitosus, P. teucrioides], common on arid, low sites.

8100, 9700, 8500 & 9050', Sharples 117, 294, 612 & 766. New to Wilderness (CN). 4CR.

Penstemon hallii A. Gray, scattered across semi-barren portions of the tundra. 12300, 11700 &

12300', Sharples 350, 474 & 630. SRM. 80

Penstemon harbourii A. Gray, infrequent on unstable tuff slopes and alpine cliffs. New to SSJ; new to CN and AA (and Wilderness). 11700, 11650 & 12100', Sharples 657, 819 & 911. SRM: volcanic ranges.

Penstemon rydbergii A. Nelson, scattered at low elevation. 9100', Sharples 455. New to

Wilderness (CN). WUSA.

Penstemon strictus Benth. [P. strictiformis], common in low, open areas. 7650, 8100 & 9200',

Sharples 142, 181 & 431. Greater 4CR.

Penstemon whippleanus A. Gray, scattered across open areas of upper elevations. 11200',

Sharples 368. USRM.

*Plantago lanceolata L., scattered along roads. 8200', Sharples 699. New to SSJ of AA. EUA.

Plantago major L., common along disturbed corridors. 8250 & 9150', Sharples 542 & 567. EUA

(W).

Plantago tweedyi A. Gray, uncommon. 11352', Douglas & Wilken 1221 (CS). USRM.

Veronica americana Schwein. ex Benth., common in low riparian areas. 8600, 9200 & 9400',

Sharples 267, 440 & 791. NAA.

*Veronica anagallis-aquatica L. [V. catenata], scattered in low wetlands. 8950', Sharples 1309.

COSM; native status unclear.

*Veronica peregrina L., rare. 10171', Douglas 2121 (CS). ANOM: Americas natively.

Veronica serpyllifolia L., common in low riparian areas. 7600 & 9550', Sharples 92 & 161.

COSM.

Veronica wormskjoldii Roem. & Schult. [V. nutans], ubiquitous on tundra and high meadows.

12000, 11650 & 11600', Sharples 334, 588 & 675. WNNA.

POACEAE – 82 species 81

Achnatherum hymenoides (Roem. & Schult.) Barkworth, occasional at low elevation. 8550 &

8400', Sharples 130 & 784. WNA.

Achnatherum lettermanii (Vasey) Barkworth, uncommon in the montane zone. 9800–10200',

Nelson 62330 (RM). WUSA.

Achnatherum nelsonii (Scribn.) Barkworth, uncommon. 10400', Sharples 620. WNA.

Achnatherum pinetorum (M.E. Jones) Barkworth, rare. 10171', Douglas 2093 (CS). WUSA.

*Agropyron cristatum (L.) Gaertn., uncommon along roadsides. 7924 & 8700', Flaig 3885 (RM) and Sharples 252. EUA.

Agrostis exarata Trin., uncommon near water. 8950–9050', Nelson 62466 (RM). WNA.

Agrostis gigantea Roth, uncommon along western slope corridors. 8950–9050', Nelson 62467

(RM). EUA (W).

Agrostis idahoensis Nash, uncommon. 11557–11814', Flaig 7979 (RM). WUSA.

Agrostis scabra Willd., scattered. 9200–9240', Nelson 62622 (RM). NAA.

*Agrostis stolonifera L., scattered in low disturbed sites. 8535–8730', Flaig 2897 (RM). EUA.

Alopecurus aequalis Sobol., common in wet, muddy areas. 8700', Sharples 299. TNH.

*Alopecurus pratensis L., scattered in low disturbed areas. 8000', Sharples 84. EUA.

Anthoxanthum hirtum (Shrank) Y. Schouten & Veldkamp [Hierochloë hirta; H. odorata], uncommon in meadows. 11200', Sharples 853. New to Wilderness (AA). CIRC.

Beckmannia syzigachne (Steud.) Fernald, uncommon along roads and trails. 8700', Sharples

1307. NAA.

Blepharoneuron tricholepis (Torr.) Nash, scattered. 10400', Sharples 622. SWNA.

Bromus carinatus Hook. & Arn., uncommon in montane forests. 9000', Sharples 942. New to

Wilderness (AA). WNA. 82

Bromus ciliatus L., scattered in the montane. 8950–9050', Nelson 62489 (RM). TNA.

*Bromus inermis Leyss., common along low roads. 7650, 8600 & 8200', Sharples 125, 507 &

695. EUA.

*Bromus japonicus Thunb., rare along ponderosa-oak zone roads. 7760–7920', Nelson 59013

(RM). EUA.

*Bromus lanatipes (Shear) Rydb., rare. 9800–10200', Nelson 62334 (RM). SWNA.

Bromus porteri (J.M. Coult.) Nash [B. anomalus], uncommon in open areas. 10400 & 9800',

Sharples 621 & 1098. New to Wilderness (CN & AA). WNA.

Bromus pumpellianus Scribn., rare in high meadows. 11700', Sharples 378. NAA.

*Bromus tectorum L., scattered in disturbed lowlands. 8745 & 7650', Flaig 3910 (RM) and

Sharples 122. EUA.

Calamagrostis canadensis (Michx.) P. Beauv., common in clearings and riparian areas. 9000',

Sharples 940. NAA.

Calamagrostis purpurascens R. Br., rare. 12107', Douglas 1717 (CS). NAA.

*Calamagrostis scopulorum M.E. Jones, uncommon. 10171', Douglas 2119 (CS). USRM.

Dactylis glomerata L., uncommon along disturbance corridors. 8750', Sharples 937. EUA (W).

Danthonia intermedia Vasey, occasional in high meadows. 11600', Sharples 676. WNNA.

Danthonia parryi Scribn., uncommon in open areas. 9800', Douglas 1098 (CS). USRM.

Deschampsia cespitosa (L.) P. Beauv., common in moist areas. 12150, 10200 & 11700',

Sharples 903, 1021 & 1381. CIRC.

Elymus elymoides (Raf.) Swezey, common in dry, open areas at low elevation. 8100 & 9800',

Sharples 182 & 1097. WNA. 83

Elymus glaucus Buckley, scattered in wet areas of low western slope forests. 8600, 8750 &

9000', Sharples 812, 934 & 943. WNNA.

*Elymus repens (L.) Gould [Elytrigia], rare. 8720–9200', Nelson 62520 (RM). EUA.

Elymus scribneri (Vasey) M.E. Jones, common on tundra. 13300, 12350, 12100 & 11850',

Sharples 633, 867, 917 & 1055. WUSA.

Elymus trachycaulus (Link) Gould ex Shinners, common in open areas. 11700, 11900, 10200 &

11600', Sharples 379, 961, 1015 and Sharples & Tripp 1126. WNNA.

*Festuca arizonica Vasey, scattered in lowlands. 7760–7920', Nelson 58994 (RM). SWNA.

Festuca brachyphylla Schult. & Schult. f., common on alpine tundra. 12150, 12800, 11450 &

11600', Sharples 900, 920, 1028 and Sharples & Tripp 1122. CIRC.

Festuca idahoensis Elmer, rare. 9800', Douglas 1094 (RM). Mostly WUSA.

Festuca minutiflora Rydb., occasional on loose alpine slopes. 11900', Sharples 380. WNA.

*Festuca rubra L., rare. 8789', Flaig 4669 (RM). TNH.

Festuca saximontana Rydb., uncommon in high, wet openings. 11850', Sharples 1035. WNNA.

Festuca sororia Piper, uncommon on the lower western slope. 8750', Sharples 935. 4CR.

Festuca thurberi Vasey, common in low openings. 8400 & 8650', Sharples 713 & 719. SRM.

*Glyceria borealis (Nash) Batch., rare in low riparian areas. O Kane et al. 6915 (SJNM).

WNNA.

Glyceria elata (Nash) M.E. Jones, scattered in the montane zone of the western slope. 9000',

Sharples 941. Mostly WUSA.

Glyceria grandis S. Watson, uncommon in low ponds. 8700', Sharples 1349. WNNA.

Glyceria striata (Lam.) Hitchc., scattered in low wetlands. 9020', M. Smith 94MS-91 (CS). TNA. 84

*Hesperostipa comata (Trin. & Rupr.) Barkworth, uncommon in ponderosa-oak savannah. 7924

& 8100', Flaig 3884 (RM) and Sharples 180. WNNA.

*Hordeum brachyantherum Nevski, uncommon in the ponderosa-oak zone. 7721', Flaig 4943

(RM). Mostly WNA.

*Hordeum jubatum L., scattered along roads. 8600', Sharples 501. New to SSJ. NAA.

Koeleria macrantha (Ledeb.) Schult., common in low openings. 8500', Sharples 780. TNH.

Melica porteri Scribn., uncommon on floodplains. 9250', Sharples 730. SWNA.

*Muhlenbergia andina (Nutt.) Hitchc., rare on floodplains. 7990', 94MS-299 (CS). Mostly

WUSA.

*Muhlenbergia asperifolia (Nees & Meyen ex Trin.) Parodi, rare in low talus. Heil & Heil 14045

(SJNM). WNNA.

*Muhlenbergia montana (Nutt.) Hitchc., rare in rocky openings of mixed forest. 8730–9500',

Flaig 2946 (RM). Greater SWNA.

*Oryzopsis asperifolia Michx., rare in montane forest. 9100', Sharples & Tripp 19. WNNA.

*Pascopyrum smithii (Rydb.) Barkworth & D.R. Dewey [Elymus], uncommon at low elevation.

7924', Flaig 3883 (RM). TNA.

Phleum commutatum Gaudin [P. alpinum], common in higher open sites. 11250, 12800 &

10200', Sharples 362, 638 & 1017. CIRC.

Phleum pratense L., common in disturbed lowlands. 8200 & 8750', Sharples 694 & 938. EUA

(W).

Poa abbreviata R. Br. [P. pattersonii], uncommon at and above treeline. 11850', Sharples 1404.

New to AA. CIRC. 85

Poa alpina L., common at and above treeline. 11100, 12600, 10200 & 11850', Sharples 849,

924, 1019 & 1056. CIRC.

Poa arctica R. Br., scattered in moist, high openings. 11850', Sharples 1033. CIRC.

*Poa bulbosa L., rare along ponderosa-oak zone roads. Heil 13116 (SJNM). EUA.

*Poa compressa L., uncommon at low elevation. 7721', Flaig 4942 (RM). EUA.

Poa cusickii Vasey [P. epilis], rare in the alpine. 12150', Sharples 902. WNA.

Poa fendleriana (Steud.) Vasey, scattered in open areas. 9100 & 10150', Sharples 762 & 816.

WNA.

Poa glauca Vahl, uncommon above treeline. 10350 & 13300', Sharples & Tripp 14 and Sharples

632. CIRC.

Poa interior Rydb. [P. nemoralis], scattered. 11814–12237', Flaig 7964 (RM). WNNA.

Poa leptocoma Trin., uncommon at high elevation. 12150', Sharples & Tripp 1134. Mostly

WNA.

Poa lettermanii Vasey, rare in tundra meadows. 11814–12237', Flaig 7959 (RM). WUSA.

Poa palustris L., scattered in low, moist areas. 9160–9560', Nelson 62428 (RM). CIRC.

Poa pratensis L. [P. agazizensis], common in open low elevation areas. 8550, 9100, 8750 &

8750', Sharples 115, 787, 932 & 933. TNH.

Poa reflexa Vasey & Scribn., scattered in forests. 11458–11557', Flaig 7997 (RM). Greater

USRM.

Poa secunda J. Presl, uncommon. 10350', Sharples 839. WNNA; disjunct elsewhere.

Poa wheeleri Vasey [P. nervosa], rare. 10000–10080', Nelson 62293 (RM). WNA.

*Psathyrostachys juncea (Fisch.) Nevski, rare in disturbed sites of western slope lowlands.

7650', Sharples 140. Found west of study area, but constitutes a new AA record. EUA. 86

*Schedonorus arundinaceus (Schreb.) Dumort. [Festuca], uncommon along ponderosa-oak zone roads. 7924', Flaig 3888 (RM). EUA.

*Thinopyrum intermedium (Host) Barkworth & D.R. Dewey [Elymus hispidus], uncommon.

10000–10080', Nelson 62293 (RM). EUA.

Torreyochloa pallida (Torr.) G.L. Church, uncommon in wet montane soils. 9160–9560', Nelson

62431 (RM). Sporadic TNA.

Trisetum spicatum (L.) K. Richt. [T. montanum], common on high, barren areas and screes.

12150, 12100, 10200, 11850 & 12550', Sharples 899, 919, 1022, 1057 and Sharples & Tripp

1137. CIRC.

Trisetum wolfii Vasey, rare in high marshes. 11000', Harrington 1836 (CS). WUSA.

*Triticum aestivum L., rare in moist river valley areas. 9000', Douglas 1066 (CS). EUA.

POLEMONIACEAE – 13 species

Aliciella pinnatifida (Nutt. ex A. Gray) J.M. Porter, uncommon on barren substrates. 9300, 9800

& 11300', Sharples 290, 669 & 1110. New to AA and CN Wilderness. USRM.

Collomia linearis Nutt., common in the lowlands. 8000, 8000, 9450 & 9500', Sharples 79, 80,

170 & 520. WNNA.

Ipomopsis aggregata (Pursh) V.E. Grant, common in low, open, dry areas. 9800, 9200 & 9450',

Sharples 240, 424 & 484. WNA.

Leptosiphon nuttallii (A. Gray) J.M. Porter & L.A. Johnson, common on dry slopes and ridges in the subalpine. 12150 & 9500', Sharples 345 & 489. WUSA.

*Phlox caryophylla Wherry, uncommon on low, barren slopes. 7750', Sharples 752. SRM: SW

CO/NW NM endemic. 87

Phlox condensata (A. Gray) E. E. Nelson, common on tundra. 12050 & 11700', Sharples 237 &

473. Patchy SWNA.

Phlox pulvinata (Wherry) Cronquist, uncommon on cliffs. 9400', Sharples & Tripp 44. WUSA.

*Polemonium brandeegei (A. Gray) Greene, rare; discovered on one inaccessible rock outcropping. 9400', Sharples & Tripp 41. New to CN and SSJ. SRM.

Polemonium confertum A. Gray, uncommon on tundra. 12250', Sharples 824. SRM.

Questionably distinct from P. viscosum.

Polemonium foliosissimum A. Gray, scattered across lowlands. 9500', Sharples 490 & 515.

Greater USRM.

*Polemonium occidentale Greene [P. caeruleum], uncommon in low boggy areas. 9550',

Sharples 1319. New to AA. WUSA.

Polemonium pulcherrimum Hook., abundant in subalpine spruce shade. 10050, 11200 & 11500',

Sharples 50, 220 & 459. WNA.

Polemonium viscosum Nutt., common on tundra. 12050 & 11600', Sharples 233 & 469. WUSA.

POLYGONACEAE – 19 species

Bistorta bistortoides (Pursh) Small, common in upper elevation openings. 11500 & 11100',

Sharples 317 & 358. Mostly WUSA.

Bistorta vivipara (L.) Delarbre, common in upper elevation openings. 9650 & 12200', Sharples

309 & 649. CIRC.

*Eriogonum alatum Torr., scattered across low elevations. 8000', Sharples 106. SWNA.

*Eriogonum lonchophyllum Torr. & A. Gray, common low on the western slope. 7200 & 7800',

Rink 2353 (FLD) and Sharples 192. 4CR; not AZ. 88

*Eriogonum racemosum Nutt., common in the ponderosa-oak zone. 8600, 8850, & 8450',

Sharples 506, 604 & 1075. 4CR; NV.

*Eriogonum umbellatum Torr., uncommon in the ponderosa-oak zone. 7721 & 8000', Flaig 4953

(RM) and Sharples 196. WUSA.

Oxyria digyna (L.) Hill, common in rocky higher elevation areas. 11200 & 10200', Sharples 370

& 617. CIRC.

*Persicaria amphibia (L.) Gray, uncommon in lower lakes. 9550', Sharples 1316. TNH.

*Persicaria lapathifolia (L.) Gray, rare along ponds. 10171', Douglas 2114 (CS). COSM.

*Polygonum argyrocoleon Steud. ex Kunze, uncommon at low elevation on the western slope.

8535–8730', Flaig 2894 (RM). EUA.

Polygonum aviculare L., common weed of low grazed roadsides and trailsides. 8650 & 9250',

Sharples 720 & 998. EUA (W).

Polygonum douglasii Greene [P. sawatchense], common on open ground in the lowlands. 9650

& 9250', Sharples 945 & 1088. WNNA.

Polygonum minimum S. Watson, rare or overlooked. 9150', Sharples 575. New to SSJ (AA

Wilderness). WNA.

Polygonum polygaloides Meisn., uncommon in low, barren, rocky areas. 9650', Sharples 1288.

New to Wilderness (AA). WNA.

Rumex acetosella L., uncommon in open areas. 11100, 8750 & 9650', Sharples 480, 546 & 948.

New to CN. EUA (W).

Rumex crispus L., common in lowlands. 7200 & 9200', Sharples 144 & 436. EUA (W).

Rumex densiflorus Osterh., common in high open areas. 11600, 11500, 11050 & 11600',

Sharples 327, 463, 1025 and Sharples & Tripp 1123. SRM. 89

Rumex occidentalis S. Watson, rare in low moist meadows. 9200–9240', Nelson 62575 (RM).

WNNA.

Rumex triangulivalvis (Danser) Rech. f., common in low openings. 8550, 7450 & 9000',

Sharples 255, 541 & 939. TNA.

POTAMOGETONACEAE – 4 species

Potamogeton alpinus Balb., rare in higher lakes. Lundquist & Rink 141 (SJNM). CIRC.

Potamogeton gramineus L., uncommon in stagnant waters. 9450', Sharples 169. New to AA.

CIRC.

*Potamogeton natans L., rare along subalpine ponds. 10170', Douglas 2124 (RM). CIRC.

*Potamogeton pusillus L., uncommon in ponds. 8950', Sharples 1312. New to AA. COSM.

PRIMULACEAE – 3 species

Androsace septentrionalis L., common throughout. 10325, 8000, 9150 & 12250', Sharples 1, 78,

569 & 641. CIRC.

*Dodecatheon pulchellum (Raf.) Merr., rare in wet meadows. 9500', Sharples 179. WNA.

Primula parryi A. Gray, common along high streamsides. 11600', Sharples 326. USRM.

PTERIDACEAE – 3 species

*Cheilanthes feei T. Moore, rare in low cliff walls. 9050', Sharples 287. Found east of study area, but new to CN. WNA.

Cryptogramma acrostichoides R. Br., common on cliffs. 9400 & 9450', Sharples & Tripp 45 &

Sharples 559. New to SSJ of AA (559). NAA.

*Cryptogramma stelleri (S.G. Gmel.) Prantl, rare. 9843', Douglas 2137 (CS). CIRC.

RANUNCULACEAE – 26 species 90

Aconitum columbianum Nutt., common in understory and openings of high forest. 11500 &

10350', Sharples 315 & 409. WUSA.

Actaea rubra (Aiton) Willd., scattered in mixed forest understory. 9100 & 9200', Sharples 29 &

841. WNNA.

*Anemone cylindrica A. Gray, rare. Heil 12636 (SJNM). WNNA.

Anemone multifida Poir., rare on well-drained ridgetops. New to Wilderness (AA). 11900',

Sharples 533. WNNA; disjunct elsewhere.

*Anemone patens L. [Pulsatilla], uncommon in lowlands. 9100', Sharples & Tripp 32. CIRC.

Aquilegia coerulea E. James, occasional across habitats. 9450 & 11700', Sharples 173 & 475.

Greater USRM.

Aquilegia elegantula Greene, common component of mixed forest understory. 9100 & 10350',

Sharples 20 & 410. 4CR.

Caltha leptosepala DC, abundant in higher riparian areas. 10350 & 11500', Sharples & Tripp 2 and Sharples 456. WNA.

*Clematis columbiana (Nutt.) Torr. & A. Gray, uncommon in low forest. 9100', Sharples &

Tripp 25. Greater USRM.

*Clematis hirsutissima Pursh, uncommon in the ponderosa-oak zone. 8730 & 8400', Flaig 3773

(RM) and Sharples 785. Greater USRM.

Delphinium barbeyi (Huth) Huth, common in forest understory and clearings. 11200 & 10350',

Sharples 310 & 413. 4CR; S WY.

*Delphinium nuttallianum Pritz., common in ponderosa-oak woodlands. 8000 & 8500', Sharples

71 & 759. WNA. 91

Ranunculus alismifolius Geyer ex Benth., common in wet subalpine areas. 10350, 11100, 11800,

11500 & 11300', Sharples & Tripp 3 and Sharples 207, 343, 466 & 590. Mostly WUSA.

Ranunculus aquatilis L., uncommon aquatic. 7850 & 10200', Sharples 499 & 664. CIRC.

*Ranunculus cardiophyllus Hook., uncommon in mid-elevation wetlands. 10350', Sharples &

Tripp 10. New to SSJ. Greater ROMO.

Ranunculus flammula L. [R. reptans], rare in subalpine wetlands. 11250', Sharples 665. New to

Wilderness (CN). CIRC.

Ranunculus gmelinii DC, uncommon around beaver ponds. 9160–9560', Nelson 62419 (RM).

CIRC.

Ranunculus hyperboreus Rottb., scattered in lowland riparian habitats. 8600 & 9400', Sharples

263 & 792. CIRC.

Ranunculus inamoenus Greene, common in moist middle elevation areas. 10350, 10050 & 9300',

Sharples & Tripp 11, 51 and Sharples 797. WNA.

Ranunculus macauleyi A. Gray, common downslope of persistent snowfields in barren areas.

11100, 11500 & 12200', Sharples 208, 467 & 681. SRM; S CO/N NM endemic.

Ranunculus macounii Britton, uncommon. 9160–9560', Nelson 62422 (RM). WNNA.

Ranunculus pensylvanicus L. f., uncommon. 8720–9200', Nelson 62497 (RM). WNNA.

Ranunculus uncinatus D. Don ex G. Don, uncommon. 11218', Flaig 4807 (RM). WNA.

Thalictrum fendleri Engelm. ex A. Gray (T. occidentale, T. venulosum), common in mixed forest understory. 10250 & 10350', Sharples & Tripp 55 and Sharples 419. Mostly WUSA.

Trautvetteria caroliniensis (Walter) Vail, occasional in subalpine understory communities.

11100, 9250 & 10500', Sharples 477, 555 & 623. NAA. 92

Trollius albiflorus (A. Gray) Rydb., uncommon in marshy snowmelt. 10450 & 12100', Sharples

& Tripp 43 and Sharples 832. ROMO.

RHAMNACEAE – 1 species

Ceanothus fendleri A. Gray, common in dry, low areas. 8000', Sharples 198. SWNA.

ROSACEAE – 24 species

*Amelanchier alnifolia (Nutt.) Nutt. ex M. Roem., common in lowlands. 7750', Sharples 88.

WNA.

*Amelanchier utahensis Koehne, scattered across lowlands. 7924 & 8900', Flaig 3865 (RM) and

Sharples 63. WUSA.

*Crataegus erythropoda Ashe, rare along roads. 7300 & 7650', Sharples 1321 & 1322. SRM.

Dasiphora fruticosa (L.) Rydb., common in dry, open areas. 8900, 11000 & 9200', Sharples 62,

405, 435. CIRC.

Fragaria virginiana Mill., common across the forested elevations. 10350 & 11250', Sharples 13

& 364. TNA.

Geum macrophyllum Willd., common in mixed forest riparian areas. 9500, 8600 & 9400',

Sharples 177, 269 & 793. CIRC.

Geum rossii (R. Br.) Ser., ubiquitous on high slopes and tundra. 11850, 11800 & 12300',

Sharples 210, 342 & 679. NAA.

Geum triflorum Pursh [Erythrocoma], uncommon. 7650', Sharples 141. WNNA.

Holodiscus dumosus (S. Watson) A. Heller, scattered across low, open areas. 9200 & 9800',

Sharples 660 & 671. SWNA.

Ivesia gordonii (Hook.) Torr. & A. Gray, rare on tundra. 12350', Sharples 352. New to CN.

WUSA. 93

Potentilla anserina L., uncommon at low elevations. 8650', Sharples 276. COSM.

Potentilla concinna Richardson, common in bare, rocky areas. 9400, 11850, 11400 & 12400',

Sharples & Tripp 40 and Sharples 212, 818 & 827. New to Wilderness (AA and CN). Essentially

WUSA.

Potentilla glaucophylla Lehm. [P. diversifolia], scattered near wetlands. 9650', Sharples 1281.

WNA; Greenland.

Potentilla hippiana Lehm., common at low elevations. 7800 & 9050', Sharples 107 & 767.

WNNA.

Potentilla nivea L., rare on alpine tundra. 12999', Harrington 1789 (RM). CIRC.

Potentilla pulcherrima Lehm. [P. gracilis], common in mixed forest zone openings. 7800 &

9650', Sharples 185 & 308. WNA.

Potentilla subjuga Rydb., scattered across higher elevations. 11814–12237', Flaig 7939 (RM).

SRM; disjunct in Alberta.

*Prunus virginiana L., common at the lowest elevations. 8730 & 8900', Flaig 3778 (RM) and

Sharples 66. TNA.

*Rosa acicularis Lindl. [R. sayi], scattered across low elevations. Heil & Mietty 28142 (SJNM).

CIRC.

Rosa woodsii Lindl., common across lower elevations. 8600, 7450 & 9200', Sharples 128, 188 &

437. WNNA.

Rubus parviflorus Nutt., common in mixed forest understory. 9100 & 9800', Sharples 492 &

528. WNNA.

Rubus idaeus L., scattered at lower elevations. 8650 & 9600', Sharples 283 & 668. New to CN

Wilderness. TNH. 94

Sibbaldia procumbens L., common on stable tundra and high meadows. 12000, 11100 & 11600',

Sharples 339, 359 and Sharples & Tripp 1127. CIRC.

Sorbus scopulina Greene, scattered across lower elevations. 7750', Sharples 121. WNA.

RUBIACEAE – 3 species

Galium boreale L. [G. septentrionale], common in lowlands. 9100 & 9250', Sharples 454 & 553.

CIRC.

*Galium trifidum L., scattered across low wet areas. 9550', Sharples 162. CIRC.

Galium triflorum Michx., uncommon at lower elevations. 9150', Sharples 572. TNH.

SALICACEAE – 17 species

*Populus angustifolia E. James, common in low river valleys. 8800 & 8750', Sharples 1299 &

1302. WNA.

Populus tremuloides Michx., abundant hardwood tree of middle elevations and higher. 9200',

Sharples 842. TNA.

*Salix bebbiana Sarg., rare along pond borders. 8800', Sharples 1298. NAA.

*Salix boothii Dorn, uncommon in riparian areas. 10350 & 9100', Sharples & Tripp 4 & 31.

Essentially WUSA.

Salix brachycarpa Nutt., common in high openings and wetlands. 11650, 11550 & 12150',

Sharples 226, 325 & 904. WNNA.

Salix drummondiana Barratt ex Hook., common along low waterways. 8450, 9150 & 8750',

Sharples 747, 973 & 1284. WNA.

Salix eriocephala Michx. [including S. ligulifolia, S. lutea], rare along streamcourses. 9150',

Sharples 972. WUSA.

*Salix exigua Nutt., scattered by low streams. 7800 & 7450', Sharples 110 & 536. WNA. 95

Salix geyeriana Andersson, uncommon along creeks. 9100', Sharples 763. WUSA.

Salix glauca L., uncommon in high boggy areas. 11100 & 11300', Sharples 355 & 593. CIRC.

Salix lasiandra Benth., scattered across lower elevation creeksides. 7800 & 8750', Sharples 109

& 929. New to Wilderness (AA). WNA.

Salix monticola Bebb, common in riparian areas. 9550, 8600, 9650, 10550 & 10200', Sharples

165, 274, 947, 1103 & 1287. 4CR; WY.

Salix petrophila Rydb. [S. arctica], scattered on tundra. 11600', Sharples 470. Mostly WUSA.

Salix planifolia Pursh, common in high openings. 11550, 12450 & 12150', Sharples 324, 865 &

901. WNNA.

Salix reticulata L. [S. nivalis], common in high grassy areas. 11650 & 12350', Sharples 228 &

830. CIRC.

Salix scouleriana Barratt ex Hook., scattered along creeksides. 10200', Sharples 1013. WNA.

*Salix wolfii Bebb, uncommon in riparian areas. 11550', Sharples 1285. WUSA.

SANTALACEAE – 1 species

*Comandra umbellata (L.) Nutt., common in the ponderosa-oak zone. 7650 & 8100', Sharples

96 & 183. TNA.

SAPINDACEAE – 1 species

Acer glabrum Torr., scattered in lower forests. 10500 & 9450', Sharples 148 & 560. WNA.

SAXIFRAGACEAE – 13 species

Heuchera parvifolia Nutt., common in high rocky areas. 11650' & 12050', Sharples 371 & 1130.

Essentially WUSA.

*Lithophragma tenellum Nutt., uncommon in dry, open lowlands of the western slope. 8700'

(Sharples), Heil 23432 (SJNM) and Sharples 757. Essentially WUSA. 96

Micranthes odontoloma (Piper) A. Heller, scattered in the subalpine near water. 11600 & 12000',

Sharples 468 & 1052. WNA.

Micranthes rhomboidea (Greene) Small, common in open subalpine areas. 10350, 11850 &

11100', Sharples 5, 209 & 360. Greater USRM.

Mitella pentandra Hook., scattered. 10998', Harrington 1823 (CS). WNA.

Mitella stauropetala Piper, common in forest understory. 7750 & 10200', Sharples 118 & 873.

USRM; PNW.

Saxifraga adscendens L. [Muscaria], rare. COJ 01-960 (SJNM). CIRC.

Saxifraga bronchialis L. subsp. austromontana (Wiegand) Piper [Ciliaria austromontana], common across altitudes on rock walls. 9100, 12300, 9900 & 12050', Sharples 34, 348, 577 &

1129. ROMO.

Saxifraga cernua L., uncommon on tundra. 13100', Sharples 653. CIRC.

Saxifraga cespitosa L. [Muscaria], uncommon on tundra. 12274', Douglas 2064 (CS). CIRC.

Saxifraga chrysantha A. Gray, rare on tundra. 13000', Harrington 1792 (CS). USRM.

Saxifraga flagellaris Willd. subsp. crandallii (Gand.) Hultén, uncommon on high, unstable, loose substrata. 12900 & 11850', Sharples 655 & 1054. MS 1054 is new to AA. CIRC.

Subspecies: USRM.

Saxifraga hyperborea R. Br. [S. rivularis], uncommon in high riparian rock crevices. 12050 &

12200', Sharples 927 & 1065. CIRC.

SCROPHULARIACEAE – 2 species

Scrophularia lanceolata Pursh, uncommon. 10826', Douglas 1700 (RM). TNA.

Verbascum thapsus L., common along low roadsides. 7850 & 8250', Sharples 693 & 1080. EUA

(W). 97

SELAGINELLACEAE – 1 species

Selaginella densa Rydb., common on tundra and rock crevices. 10250 & 12150', Sharples &

Tripp 56 and Sharples 905. Greater ROMO.

SPARGANIACEAE – 3 species

Sparganium angustifolium Michx., scattered in ponds and bogs. 8750', Sharples 564. COSM.

*Sparganium emersum Rehmann, rare in low elevation waters. 8950', Sharples 1315. New to

AA. COSM.

Sparganium natans L. [S. minimum], uncommon in lakes. Lundquist & Rink 136 (SJNM).

COSM.

TYPHACEAE – 1 species

*Typha latifolia L., uncommon in low elevation riparian areas. 7900', Sharples 707. COSM.

URTICACEAE – 1 species

Urtica dioica L. subsp. gracilis (Aiton) Selander, scattered in rocky areas below treeline. 9100',

Sharples 444. TNA.

VERBENACEAE – 1 species

*Verbena macdougalii A. Heller, rare at low elevations. 8750' (Sharples), Penland 4651 (COLO) and Sharples 254. 4CR.

VIOLACEAE – 6 species

Viola adunca Sm., common in lowlands. 10350 & 7650', Sharples 7 & 90. WNNA.

Viola canadensis L., common in low forests. 7650 & 8450', Sharples 91 & 745. TNA.

Viola labradorica Schrank, common in high elevation openings. 11650, 12000 & 11850',

Sharples 227, 338 & 963. ANOM: disjunct from E North America.

*Viola nuttallii Pursh, rare. O Kane 3740 (SJNM). Greater USRM. 98

Viola palustris L. [V. macloskeyi], uncommon. O Kane 7348 (SJNM). CIRC.

Viola nephrophylla Greene [V. soraria], uncommon in riparian areas. 9550', Sharples 160.

WNNA.

ZANNICHELLIACEAE – 1 species

Zannichellia palustris L., rare in subalpine ponds. 10007’, Douglas 2076 (CS). COSM.

99

CHAPTER 2 (PART I)

INTRODUCTION

The flowering plant genus Stellaria L. (Caryophyllaceae, tribe Alsineae sensu Harbaugh et al., 2010) is a cosmopolitan group of herbaceous annuals or perennials distributed across a wide variety of ecosystems and spanning tropical, temperate, and high latitudes and elevations in all hemispheres. Stellaria is morphologically diverse (Fig. 5) and is marked by a complex taxonomic history replete with convergent evolution of many characters. Although the name

Alsine L. was published with the name Stellaria by Linnaeus in the same publication, as the names are currently typified the former is a heterotypic synonym of the latter, which has been adopted for the caryophyllaceous lineage known colloquially as the “starworts” or sometimes the

“chickweeds” (Sprague, 1920; McNeill, 1962). The choice between two competing names with equal priority is determined by the first such choice to be effectively published (Art. 11.5 of the

ICN, Turland et al., 2018). Villars (1789: 615) included Alsine media L., the first designated type of Alsine, and Stellaria graminea L., the conserved type of Stellaria (Tikhomirov, 2016), in a single genus, Stellaria, giving that name precedence over Alsine.

Stellaria remains poorly delimited, with disparate estimates of the total number of starwort species occurring thoughout the literature. Many of these estimates range from ~120 spp. (Ankei, 1982; Morton, 2005; Holmgren et al., 2012; Mabberley, 2017), to 150–200 spp.

(Mahdavi et al., 2010), to upwards of 190 spp. (Shilong and Rabeler, 2001). The last attempt at a 100

Fig. 5. Morphological diversity of some Stellaria species as captured in their multifarious habitats: A) Stellaria brachypetala B) Stellaria crassifolia C) Stellaria crispa D) Stellaria decipiens E) Stellaria decumbens F) Stellaria filiformis G) Stellaria flaccida H) Stellaria gracilenta I) Stellaria graminea J) Stellaria humifusa K) Stellaria irrigua L) Stellaria littoralis

M) Stellaria longifolia N) Stellaria media O) Stellaria nemorum P) Stellaria nepalensis Q)

Stellaria nitens R) Stellaria persica S) Stellaria pubera T) Stellaria pungens U) Stellaria radians

V) Stellaria roughii W) Stellaria sanjuanensis X) Stellaria vestita. All photos taken by Mat

Sharples except for “S”, taken by Erin Tripp.

101

102 worldwide classification of Stellaria, that of Pax and Hoffmann (1934), hypothesized “over 100 species” to be found in the genus. The lack of recent comprehensive analysis of diversity within

Stellaria is striking given its moderate species diversity and the extremely broad geographical distribution of the genus altogether. Many floras, regional taxonomic treatments, and other studies have considered local diversity of Stellaria such that there exist robust hypotheses of morphological species delimitation within regional taxonomic accounts (e.g., Schischkin, 1936;

Chater and Heywood, 1964; Garnock-Jones et al., 1988; Shilong and Rabeler, 2001; Chinnappa et al., 2005; Morton, 2005; Miller and West, 2012), but a modern investigation of global scope is lacking.

Greenberg and Donoghue (2011) advanced a family-wide molecular phylogenetic study of Caryophyllaceae. Their results demonstrated broad paraphyly of Stellaria as currently circumscribed, with species resolved in clades both closely and more distantly related to a core, monophyletic radiation of Stellaria (hereafter, “core Stellaria” or “core stars”). In their study, which represents the densest molecular sampling of species of Stellaria yet published, approximately one-third of species were included in phylogenetic analysis. The broad paraphyly of species of Stellaria mixed among other white-flowered alsinaceous genera (e.g., Arenaria L.,

Cerastium L., Minuartia Loefl., and Pseudostellaria Pax) in their study and in others (Fior et al.,

2006; Harbaugh et al., 2010; Dillenberger and Kadereit, 2014; Sadeghian et al., 2015; Zhang et al., 2017) made clear that expanded taxon sampling was needed to more fully understand relationships and evolutionary trends within this group. Greenberg and Donoghue’s (2011) study was based on nrITS in addition to five chloroplast markers, which resulted in a number of nodes with low (<70) bootstrap support and some unresolved relationships. Thus, a thoroughly- sampled, well-resolved, and well-supported evolutionary hypothesis of relationships among 103 species of Stellaria remains to be generated. The absence of such work has contributed to poor understanding of numerous species and a lack of robust estimates of extant species worldwide.

Restriction site-associated DNA sequencing (henceforth RADseq) has provided powerful data for resolving phylogenetic relationships in the genomic era. RADseq has been shown to be effective at recovering evolutionary relationships of both understudied and problematic taxa at various phylogenetic scales in flowering plants and other lineages (e.g., Cariou et al., 2013;

Eaton and Ree, 2013; Cruaud et al., 2014; Hipp et al., 2014; Díaz-Arce et al., 2016; Hou et al.,

2016a; Massatti et al., 2016; Wessinger et al., 2016; Eaton et al., 2017; Nieto-Montes de Oca et al., 2017; Tripp et al., 2017; Daniel and Tripp, 2018; Beeler et al., in press). These and other

RADseq studies have generally recovered increased phylogenetic resolution in the groups under study than had been possible with alignment of comparatively few loci, those derived primarily from Sanger sequencing. We previously used RADseq data to resolve phylogenetic relationships within a species complex of Stellaria in combination with outgroups spanning core Stellaria and two related genera (Sharples and Tripp, 2019b). We therefore sought a similar approach in the present investigation to advance knowledge of phylogenetic relationships across the genus as a whole, including several phylogenetic neighborhoods of a priori interest to us (Sharples, companion paper [this issue]). This companion paper provides detailed discussion of phylogenomic patterns to emerge from extended taxon sampling within several historically difficult and geographically widespread species complexes such as the Stellaria borealis

Bigelow complex, the Stellaria cuspidata Willd. group and New World relatives, the circumboreal Stellaria longipes Goldie complex, the cosmopolitan and ruderal Stellaria media

(L.) Villars group, and the scandent members of the Asian Stellaria vestita Kurz group. 104

The overarching objectives of the present investigation were thus as follows: (1) to better elucidate phylogenetic delimitation of core Stellaria L. and related lineages using RADseq data, a near complete species-level sampling of core stars taxa, and a sizeable sampling of outgroup genera combined with non-core Stellaria samples (hereafter, “false stars”); (2) to produce a robust estimate of the number of species comprising core Stellaria based on phylogenetic results in combination with extensive fieldwork, study of literature, and study of herbarium material; and (3) to assess the monophyly or lack thereof of infrageneric taxa (i.e., subgenera, sections, series) proposed by prior authors (especially: Endlicher, 1840; Fenzl, 1842; Pax and Hoffmann,

1934; Schischkin, 1936) in light of new molecular data, with dense taxon sampling focused on several historically difficult species complexes as well as lineages of widespread species, particularly those identified above. Altogether, our results demonstrate that RADseq holds great promise for resolving relationships across Caryophyllaceae as a whole.

MATERIALS AND METHODS

Field and Herbarium Samples

Over the course of 18 distinct expeditions spanning four continents and approximately

203 days between 2013–2018, fieldwork was undertaken to facilitate acquaintance with starwort species and outgroups in their native habitats as well as to fresh collect extant diversity. Newly collected specimens from mesic climates were dried under low heat to prevent tissue rot and to preserve DNA integrity; those collected from arid localities were air-dried in collection newspapers without supplemental heat. Following processing, identification, and curation, these herbarium voucher specimens were deposited at the following institutions: COLO, IBSC, NY and/or RM. For species that we were unable to locate in the field, numerous herbaria were visited 105

Table 6. List of core Stellaria plus outgroup accessions sampled in this study (n=164), including the number of RAD reads containing the EcoR1 cutsite and its protector base (=“CAATT”), the year in which the herbarium voucher was collected, and the age of the material when it was sequenced. An (x) in front of a species’ name indicates that the taxon is incorrectly placed in the genus Stellaria and is in need of revision. A “(1)”, “(2)”, and other subsequent parenthetical numbers or geographical labels after the taxonomic authority are referable to tip labels with multiple accessions per species, as per Figs. 7–9 and per the appendix.

Species Read# Year Age Adenonema cherleriae (Fisch. ex Ser.) M.T.Sharples & E.Tripp 1,552,202 2005 13 Adenonema petraeum Bunge 1,259,375 2007 10 Arenaria lanuginosa (Michx.) Rohrb. 3,176,702 2014 3 Cerastium beeringianum Cham. & Schltdl. 1,471,144 2014 1 Cerastium fontanum Baumg. 5,541,581 2001 16 Cerastium parvum (Pedersen) M.T.Sharples & E.Tripp 1,080,067 1971 47 Cerastium pauciflorum Steven ex Ser. 5,825,301 2015 2 Cerastium sp. Altai 5,670,434 2015 3 Holosteum umbellatum L. 4,887,157 2018 1 Mesostemma gypsophiloides (Fenzl) M.T.Sharples & E.Tripp 1,552,202 1990 28 Mesostemma martjanovii (Krylov) Ikonn. 4,301,866 2015 2 Nubelaria arisanensis (Hayata) M.T.Sharples & E.Tripp 907,059 1968 49 Nubelaria diversiflora (Maxim.) M.T.Sharples & E.Tripp 5,934,257 2000 18 Nubelaria wushanensis (F.N.Williams) M.T.Sharples & E.Tripp (1) 3,259,047 1998 20 Nubelaria wushanensis (F.N.Williams) M.T.Sharples & E.Tripp (2) 624,316 2000 18 Pseudostellaria heterophylla (Miq.) Pax 1,744,884 1993 25 Pseudostellaria jamesiana (Torr.) W.A.Weber & R.L.Hartm. 1,452,010 2013 2 Rablera holostea (L.) M.T.Sharples & E.Tripp 2,375,356 1986 31 Sabulina macrantha (Rydb.) Dillenb. & Kadereit 6,580,398 2013 4 Silene baccifera (L.) Roth 3,306,138 2017 1 Silene menziesii Hook. 4,130,102 2014 3 Stellaria alaschanica Y.Z.Zhao 2,675,591 2006 11 Stellaria alaskana Hultén 6,511,732 1988 30 Stellaria alsine Grimm (NA) 977,874 2011 6 Stellaria alsine Grimm (CH) 708,557 1996 21 Stellaria altimontana N.S.Pavlova 2,077,376 1979 39 (x) Stellaria americana (Porter ex B.L.Rob.) Standl. 122,784 2016 1 Stellaria anagalloides C.A.Mey. ex Rupr. 2,012,568 1977 41 Stellaria angarae Popov 3,171,560 2003 15 Stellaria angustifolia Hook. 6,701,440 2016 1 (x) Stellaria antillana Urb. 2,125,463 2004 13 Stellaria aquatica (L.) Scop. (1) 3,585,882 2012 5 Stellaria aquatica (L.) Scop. (2) 5,542,426 2017 1 106

Stellaria arenarioides S.L.Chen, Rabeler & Turland 3,825,691 1993 24 Stellaria arvalis F. Phil. 574,702 1986 31 Stellaria borealis Bigelow 3,814,110 1966 51 Stellaria brachypetala Bunge 5,308,473 2015 2 Stellaria bungeana Fenzl 170,661 2015 1 Stellaria calycantha (Ledebour) Bongard 1,279,504 1992 25 Stellaria chilensis Pedersen (1) 5,962,159 1976 42 Stellaria chilensis Pedersen (2) 7,194,697 2000 17 Stellaria chinensis Regel 2,710,420 1992 25 Stellaria ciliatosepala Trautv. 430,886 1957 60 Stellaria congestiflora H.Hara 4,205,611 1996 22 Stellaria corei Shinners 527,600 1980 37 Stellaria crassifolia Ehrhart (1) 10,488,377 1984 33 Stellaria crassifolia Ehrhart (2) 7,022,879 2015 3 Stellaria crassifolia Ehrhart (3) 1,991,190 2015 3 Stellaria crassipes Hultén 2,142,706 1925 93 Stellaria crispa Cham. & Schltdl. (US) 7,550,962 2016 1 Stellaria crispa Cham. & Schltdl. (RU) 1,265,340 2000 18 Stellaria cuspidata Willd. ex Schltdl. 2,367,244 1939 79 Stellaria dahurica Willd. ex Schltdl. 449,002 2009 8 Stellaria debilis d’Urv. 517,916 1999 18 Stellaria decipiens Hook.f. 507,872 1985 32 Stellaria decumbens Edgew. 875,986 1998 19 Stellaria dianthifolia F.N.Williams (1) 181,502 1976 42 Stellaria dianthifolia F.N.Williams (2) 986,604 2017 1 Stellaria discolor Turcz. 8,444,675 2006 12 Stellaria edwardsii R.Br. 1,529,073 1964 53 Stellaria eschscholtziana Fenzl (1) 1,878,393 2003 15 Stellaria eschscholtziana Fenzl (2) 2,746,601 2000 18 Stellaria fenzlii Regel 2,515,632 1996 21 Stellaria filicaulis Makino 3,034,243 1991 27 Stellaria filiformis (Benth.) Mattf. 9,827,521 2016 1 Stellaria fischeriana Ser. 1,389,398 1976 41 Stellaria flaccida Hook. (1) 710,771 1988 29 Stellaria flaccida Hook. (2) 2,937,125 2016 1 (x) Stellaria fontinalis (Short & R.Peter) B.L.Rob. 4,461,142 1992 25 Stellaria gracilenta Hook.f. 3,905,836 2017 1 Stellaria graminea L. 1,796,904 2003 14 Stellaria henryi F.N.Williams 3,825,504 2011 7 Stellaria hintoniorum B.L.Turner 675,170 1994 23 (x) Stellaria howardii Maguire 6,349,554 2004 14 Stellaria humifusa Rottbøll (1) 829,225 1968 49 Stellaria humifusa Rottbøll (2) 6,124,275 2015 2 Stellaria infracta Maxim. (1) 1,188,432 2007 11 Stellaria infracta Maxim. (2) 750,903 1984 34 Stellaria irazuensis Donn.Sm. 1,195,152 1995 22 Stellaria irrigua Bunge (1) 9,386,090 2011 6 Stellaria irrigua Bunge (2) 1,117,502 1974 43 Stellaria irrigua Bunge (3) 3,229,238 2015 1 107

Stellaria irrigua Bunge (4) 5,158,013 2015 1 Stellaria lanata Hook.f. 10,570,930 1990 27 Stellaria lanipes C.Y.Wu & H.Chuang 11,474,005 1979 38 Stellaria leptoclada (Benth.) C.H.Mill. & J.G.West (1) 2,376,139 2010 8 Stellaria leptoclada (Benth.) C.H.Mill. & J.G.West (2) 3,802,685 2014 4 Stellaria littoralis Torr. (1) 757,152 1936 81 Stellaria littoralis Torr. (2) 2,515,622 2015 1 Stellaria longifolia Muhl. ex Willd. (1) 1,447,963 1979 38 Stellaria longifolia Muhl. ex Willd. (2) 6,584,361 1970 47 Stellaria longifolia Muhl. ex Willd. (3) 3,009,094 2016 1 Stellaria longifolia Muhl. ex Willd. (4) 10,434,121 2017 1 Stellaria longifolia Muhl. ex Willd. (5) 3,037,194 1996 22 Stellaria longipes Goldie 2,047,289 2015 1 Stellaria mainlingensis L.H.Zhou 1,259,581 1972 45 (x) Stellaria mannii Hook.f. 4,712,826 2008 10 Stellaria maximowiczii Kozhevn. (1) 362,650 1954 63 Stellaria maximowiczii Kozhevn. (2) 10,518,666 2017 1 Stellaria media (L.) Villars (1) 2,476,597 2015 1 Stellaria media (L.) Villars (2) 6,103,400 2017 1 Stellaria media (L.) Villars (3) 5,206,300 2017 1 (x) Stellaria monosperma Buch.-Ham. ex D.Don 3,461,400 1993 24 Stellaria montioides (Edgew. & Hook.f.) S.A.Ghazanfar 1,394,884 1964 54 Stellaria multiflora Hook. 639,239 1956 61 Stellaria neglecta Weihe (1) 512,465 2012 5 Stellaria neglecta Weihe (2) 2,193,119 1984 33 Stellaria nemorum L. 3,222,017 2015 1 Stellaria nepalensis Majumdar & Vartak 3,869,274 2017 1 Stellaria nipponica Ohwi 3,640,038 1970 47 Stellaria nitens Nutt. 2,325,389 1997 20 Stellaria nubigena Standl. 5,409,834 1983 34 Stellaria omeiensis C.Y.Wu & Y.W.Tsui ex P.Ke (1) 5,775,136 2017 1 Stellaria omeiensis C.Y.Wu & Y.W.Tsui ex P.Ke (2) 3,923,014 2017 1 Stellaria omeiensis C.Y.Wu & Y.W.Tsui ex P.Ke (3) 3,841,358 2017 1 Stellaria palustris Ehrh. ex Retz. 5,262,501 2006 11 Stellaria papillata C.H.Mill. & J.G.West (1) 6,260,355 2003 15 Stellaria papillata C.H.Mill. & J.G.West (2) 3,680,133 2009 9 Stellaria parviflora Banks & Sol. ex Hook.f. 1,166,927 2017 1 Stellaria patens D.Don (1) 452,408 2004 13 Stellaria patens D.Don (2) 3,384,626 2017 1 Stellaria peduncularis Bunge 3,475,208 2015 1 Stellaria persica Boiss. (1) 8,105,152 2004 13 Stellaria persica Boiss. (2) 694,196 2015 3 Stellaria petiolaris Hand.-Mazz. 900,820 2006 11 Stellaria pilosoides S.L.Chen, Rabeler & Turland (1) 5,911,295 1937 81 Stellaria pilosoides S.L.Chen, Rabeler & Turland (2) 3,344,671 2017 1 Stellaria pilosoides S.L.Chen, Rabeler & Turland (3) 6,095,303 2017 1 Stellaria polyantha (Edgew.) M.T.Sharples & E.Tripp 1,412,691 1954 63 Stellaria porsildii C.C.Chinnappa 1,143,392 1973 44 Stellaria prostrata Baldwin 2,543,437 1994 23 108

Stellaria pubera Michx. 1,038,600 1998 19 Stellaria pungens Brongn. 1,231,359 1966 51 Stellaria pusilla Em. Schmid 2,244,905 1987 30 Stellaria radians L. 4,544,628 2003 14 Stellaria recurvata Willd. ex Schltdl. 1,175,747 1999 19 Stellaria roughii Hook.f. 5,827,944 2017 1 Stellaria ruscifolia Willd. ex Schltdl. 1,645,000 2004 13 Stellaria salicifolia Y.W.Tsui & P.Ke 6,265,003 2007 10 Stellaria sanjuanensis M.T.Sharples & E.Tripp 8,757,745 2014 3 Stellaria sennii Chiov. 1,105,900 1986 31 Stellaria serpyllifolia Willd. ex Schltdl. 1,416,343 1985 33 Stellaria sessiliflora Y.Yabe 918,540 1919 98 Stellaria sikkimensis Hook.f. 5,518,224 1992 25 Stellaria sitchana Steudel (1) 2,453,279 1965 52 Stellaria sitchana Steudel (2) 3,523,480 1991 26 Stellaria cf. sitchana Bigelow 4,043,638 2017 1 Stellaria soongorica Roshev. (1) 625,296 1987 30 Stellaria soongorica Roshev. (2) 1,956,484 1958 59 Stellaria souliei F.N.Williams 3,408,659 1940 77 Stellaria sp. (1) 3,770,833 2017 1 Stellaria sp. (2) 7,170,649 2017 1 Stellaria sp. cf. alsine Russia 6,904,381 2015 3 Stellaria tetrasticha (Mattf.) M.T.Sharples & E.Tripp 3,950,070 1985 32 Stellaria tomentosa Maxim. 1,698,288 1958 59 Stellaria uchiyamana Makino 504,862 1986 32 Stellaria uda F.N.Williams 830,707 1998 19 Stellaria cf. uda F.N.Williams 9,610,707 2017 1 Stellaria venezuelana Steyerm. 1,099,115 1981 37 Stellaria vestita Kurz (1) 545,171 1989 28 Stellaria vestita Kurz (2) 1,200,499 2005 12 Stellaria weddellii Pedersen 490,224 1990 28 Stellaria winkleri (Briq.) Schischk. (1) 2,480,822 1958 60 Stellaria winkleri (Briq.) Schischk. (2) 1,121,944 1988 30

109 across North America, Eurasia, and Oceania (ALTB, CAS, CHR, COLO, E, IBSC, K, KUN, LE,

MO, NSK, NY, P, PE, RM, UC, VBGI, and VLA; Appendix 1; herbarium codes follow Thiers, continuously updated) to study material and acquire leaf tissue samples. These specimens spanned a range of decades (Table 6; Appendix 1) and preservation techniques, the latter of which were unknown to us.

Taxon Sampling

Within core Stellaria, we attempted to sample as much of the extant core Stellaria species diversity as possible in this investigation. Our list of taxa to sample was in part based on extensive study of global treatments of Stellaria (Table 7) in combination with study of protologues of species of Stellaria and related genera. Extending the sampling strategy further, we included putative synonyms of widespread and variable species complexes, where possible, to better understand their taxonomic and evolutionary structure beyond simply evaluating potential monophyly of synonyms. In 28 instances, we sampled multiple representatives per species (typically two to three, but up to five distinct accessions in the case of Stellaria longifolia

Muhl. ex Willd.) to test for reciprocal monophyly of species using RADseq data (Table 6); these taxa (all of which belonged to core Stellaria except one) often had broad geographical ranges, and in those cases we attempted to sample material from geographically different portions of those ranges. Contrasting with the above, we emphasized species-level relationships in this study and thus did not prioritize subspecific or varietal taxa during taxon sampling. However, extreme infraspecific variants of a select number of taxa (e.g., some members of the Stellaria decumbens

Edgew. group) were included. 110

Outside of core Stellaria as informed by prior works (Fior et al., 2006; Harbaugh et al.,

2010; Greenberg and Donoghue, 2011; Dillenberger and Kadereit, 2014; Sadeghian et al., 2015;

Zhang et al., 2017; M. Sharples, unpub. data in preparation for this study), we sampled putative outgroup material not in a comprehensive but rather in a strategic manner that emphasized a broad diversity of clades both closely and more distantly related to core stars. Sampled outgroups of alsinoid Caryophyllaceae related to core Stellaria included representatives of Arenaria,

Cerastium, Holosteum, Pseudostellaria, Sabulina, and others detected in the current study.

According to all available evidence (prior phylogenetic works cited above), including results reported here, the former type species of Stellaria (S. holostea L.: see Tikhomirov, 2016) and several other false stars must be considered as part of those outgroups. Two representatives from the genus Silene L. were included for phylogenetic rooting purposes, being least closely related to Stellaria than other outgroups included here (Table 6; Greenberg and Donoghue, 2011;

Dillenberger and Kadereit, 2014).

Our concept of the number of species in Stellaria presented herein was based on the following: study of original type material and protologues; ecological, geographical, morphological, and reproductive features of studied plants; results of molecular phylogenetic analyses from this and prior studies; study of regional starwort and outgroup treatments; and consideration of the opinions of various taxonomic authorities who have worked on Stellaria and relatives in prior works (e.g., Schischkin, 1936; Chater and Heywood, 1964; Garnock-Jones et al., 1988; Shilong and Rabeler, 2001; Chinnappa et al., 2005; Morton, 2005; Miller and West,

2012; others as listed in Table 7 and elsewhere).

Wet Lab Procedures

Between November 2015 and April 2018, DNA extractions were conducted following the 111

Table 7. Estimate of number of core Stellaria species inhabiting major geographical regions as well as total global Stellaria. The Net Species column is additive across each row of geographical Region and therefore reflects the increase in species count added by occurrences of species that were not present in any of the preceding rows. The following regional treatments and other resources were considered when producing estimates. Africa: Wild, 1961; Australia: Miller and West, 2012; Central America: Baker and Burger, 1983, Beaman, 2009, Duke, 2013,

Rodríguez-Jiménez, 2013; Europe: Chater and Heywood, 1964; Japan: Zoku, 1965; New

Zealand: Garnock-Jones et al., 1988; North America: Morton, 2005; South America: Brako and

Zarucchi, 1993, Volponi, 1993, Jørgensen and León-Yánez, 1999, Idárraga-Piedrahita, 2001,

Nee and Jørgensen, 2014, Timaná, 2017; Southern Asia: Hara and Febbs, 1979, Grierson, 1984,

Rechinger, 1988, Shilong and Rabeler, 2001, Larsen, 2002, Ghazanfar and Nasir, 2011,

Majumdar, 2007–onwards; Former U.S.S.R. (minus Europe): Schischkin, 1936, Kozhevnikov,

1994, Pavlova, 1996, Malyschev and Peschkova, 2003, Vlasova, 2012. The caryophyllaceous floras of South America and southern Asia, in particular, are incompletely known at present and may yet harbor undescribed Stellaria diversity.

Region Native spp.# Sampled# Unsampled# Net species Africa 1 1 X 1 Australia 7 7 X 8 Central America 4 4 X 12 Europe 11 11 X 23 Japan 15 15 X 33 New Zealand 5 4 1 38 North America 19 19 X 51 South America 18 9 9 68 Southern Asia 49 44 5 105 U.S.S.R.(-EU) 31 31 X 112 Total net sampled: 97 15 112

112

CTAB procedure (Doyle and Doyle, 1987). A double-digest RADseq protocol adapted from

Parchman et al. (2012) and further customized (Tripp et al., 2017) was used for library preparations, as follows. Genomic DNA extractions containing an average of 150 ng/μl of DNA were subjected to restriction site double digestion with the enzymes EcoRI and MseI. Custom- designed barcodes of variable length (7–10 bases) were ligated onto restriction fragments to facilitate multiplexing of 96 total samples in a pooled library (Tripp et al., 2017). The barcoded restriction-ligation reaction product was PCR amplified, samples were pooled, and the resultant

PCR product was run on a 1% mixed half-and-half agarose/high-resolution agarose gel. DNA fragment size selection followed two procedures for a total of five RAD libraries prepared and sequenced during this study. For the first two libraries sequenced, DNA smears were size selected from the gel by hand, targeting the 200–500 base pair region. Gel fragments were then purified using a QIAquick gel extraction kit according to the manufacturer’s protocol (Qiagen,

Valencia, California), and resulting libraries were subjected to an Agilent 2200 TapeStation

D1000 ScreenTape to quantify fragment size distribution as well as to undergo other quality control. For the latter three libraries sequenced, purified gel fragments were further size selected using a BluePippin, again targeting the 200–500 bp fragment region and subject to quality control with an Agilent 2200 TapeStation using high sensitivity D5000 ScreenTape. All pooled libraries were sequenced at the University of Colorado Biofrontiers Next-Generation Sequencing

Facility on an Illumina HiSeq 2000 with a V3 100-Cycle single read sequencing kit (first library) or a NextSeq 500 with a V2 High-Output 75-Cycle single read sequencing kit (latter four libraries).

Bioinformatic Analyses and Molecular Alignment Preparation 113

Raw .fastq reads were first assessed for quality, quantity, and other sequencing statistics using FastQC (Andrews, 2010). As FastQC results indicated high sequencing quality of libraries used in this study, reads were then filtered and Illumina adapters trimmed using cutadapt v1.13

(Martin, 2011). Given different read lengths derived from the two sequencing platforms, reads from the HiSeq reaction were further trimmed using cutadapt v1.13 to mirror the read lengths of the NextSeq data. Reads were then demultiplexed with barcode input files using fastq-multx, which is a module distributed within the ea-utils package (Aronesty, 2011). Some of these libraries included samples from other researchers and taxa, and these samples were discarded after demultiplexing, as were all unmatched.fastq files to which reads lacking barcodes were sequestered by fastq-multx. Due to variable barcode lengths, final filtered and processed reads had maximum lengths between 82 and 85 bp; these were not further trimmed to a uniform length. Many processed reads were shorter in length, ranging between our specified minimum read length (-m 35 in cutadapt) up to 81 bp, depending on whether or not the second cutsite was present.

Curated RAD reads were processed using iPyRAD v.0.7.23 (Eaton, 2018) to construct

DNA alignments suitable for phylogenetic inference. Parameters that were necessary to alter from the default were “datatype” (parameter [7], changed to “ddrad”) and “restriction_overhang”

(parameter [8], changed to “CAATT,TAC”). Parameters voluntarily changed from defaults yet held static across analyses were parameter [12] (“mindepth_majrule”), parameter [22]

(“max_SNPs_locus”), and parameter [24] (“max_shared_Hs_locus”). We lowered [12] to “2” to make use of RAD loci with lower read depth, raised [22] to “40” to allow for more variable and/or uncertain portions of loci across core stars and outgroups to be retained, and lowered [24] to “0.2” to minimize potential paralogs in our alignments. We then conducted numerous 114 additional exploratory analyses to optimize parameters, particularly varying two other settings relating to clustering thresholds within and across RADseq samples and relating to missing data thresholds: parameter [14], “clust_threshold”, and parameter [21], “min_samples_locus” (final parameters differing from the default are described below). Results of these exploratory analyses are not shown, but we note that our prior work using RAD loci in Stellaria and other systems has shown that altering these parameters does not substantially impact overall phylogenetic relationships, but rather primarily impacts bootstrap support values and resolvedness (Tripp et al., 2017; Sharples and Tripp, 2019b). Our exploratory analyses yielded similar outcomes to these two earlier published studies.

Because inclusion of distant relatives of core Stellaria sometimes dramatically reduced support values (likely owing to locus dropout combined with variable read depths: see Eaton et al., 2017), we assembled three different aligned matrices to infer three different phylogenies, each with different objectives. The first analysis (“outgroups”), which emphasized the inclusion of core stars outgroups, was conducted to place various species currently ascribed to Stellaria in a broad phylogenetic context of the entirety of Caryophyllaceae. The outgroups analysis included

27 outgroup tips plus 19 core Stellaria tips (n=46) and was based on a clustering threshold

[parameter 14] of 0.86 and minimum taxa [parameter 21] set to 12. The second analysis (“core stars”), which spanned the breadth of core Stellaria with only one outgroup included, was conducted to specifically resolve relationships within the hypothesized core Stellaria clade. The core stars analysis spanned 104 core Stellaria tips in addition to the outgroup (Cerastium pauciflorum Steven ex Ser.) and was based on a clustering threshold [parameter 14] of 0.87 and minimum taxa [parameter 21] set to 22. The third analysis (“Larbreae”) focused in on the clade representing what is often treated as Stellaria subsection Larbreae (Fenzl in Endl.) Pax & 115

Hoffmann (or Stellaria section Larbreae Fenzl in Endl.), the most species-rich clade of core

Stellaria, to resolve relationships within this radiation. More samples representing multiple individuals of taxa of Larbreae than were included in the core stars analysis were added to this third analysis, resulting in 106 tips in total. This analysis was based on a clustering threshold

[parameter 14] of 0.9 and minimum taxa [parameter 21] set to 15.

Phylogenetic Reconstruction

Assembled, aligned, and concatenated DNA sequences of RAD loci were used for phylogenetic reconstruction implemented in RAxML v.8.2.8, specifying a GTRCAT model of sequence evolution (Stamatakis, 2014). The GTRCAT model performs at least equally well to

GTR + Γ in RAxML but is more efficient than the latter, especially with large datasets such as ours (Stamatakis, 2006). A rapid bootstrapping analysis with 96 bootstrap replicates (this replicate number optimized for best computational output on the University of Colorado’s HPC, given 24 cores per node) was conducted and the most highly supported ML tree was exported to compare the effects of parameter settings (primarily missing data thresholds through alteration of parameter [21] in iPyRAD) on support values and relationships recovered. Bootstrap values were appraised along the following lines: poor=<70; fair=70–79; good=80–94; high=95–100. The two

Silene species were used to root the output from the outgroups analysis, and the Cerastium pauciflorum sample was used to root the core stars analysis output. The tree resultant from the

Larbreae analysis was rooted using Stellaria alaskana Hultén, which the previous analyses had revealed as sister to the rest of the Larbreae clade. Phylogenetic trees were visualized in FigTree v1.4.3 (Rambaut, 2016).

Specimen Age 116

To assess whether age of herbarium specimen had an impact on number of RAD reads generated, we analyzed the number of useable RAD reads in terms of the age of the specimen.

The number of usable reads per sample was approximated by using the command “grep” upon trimmed specimen .fastq files; this yielded the number of RAD reads containing the EcoR1 cutsite as well as its protector base (=“CAATT”) per file. Age of specimen DNA was quantified by subtracting the year of specimen collection from the year of DNA sequencing. We then fit a linear model to the data. Basic linear model and plotting functions were implemented using R version 3.5.1 (R Core Team, 2018).

All bioinformatic work was conducted using the Summit Supercomputer at the

University of Colorado (Anderson et al., 2017). Individual .fastq files for each sample used in this study have been deposited as a Sequence Read Archive (SRA #PRJNA547948; see also

SRA #PRJNA473254 for select samples also included in Sharples and Tripp, 2019b) within

GenBank.

RESULTS

Utility of RAD Loci—A total of 164 samples representing 97 accepted species-rank names of core Stellaria and 27 outgroups was ultimately included in our ddRADseq sampling across Stellaria’s natural distribution (Fig. 6; Table 6). This study serves as the first presentation of RAD loci from 151 of our 164 total specimens sampled (other RAD loci published for

Caryophyllaceae are limited to those from Trucchi et al., 2017, and Sharples and Tripp, 2019b, to our knowledge). Phylogenetically analyzed specimens spanned a diversity of collection dates and presumably a variety of preservation techniques (Table 6). We found, however, that low read 117

Fig. 6. Localities of specimens sampled in this study, somewhat approximating the worldwide distribution of Stellaria. Outgroups (i.e., not member to core Stellaria) are not plotted.

118 samples (<300,000 total reads containing the EcoR1 cutsite, these likely attributable to stochastic laboratory processes rather than specimen age) generally performed poorly in phylogenetic analyses, contributing to low support values at nodes otherwise involving high quality samples

(>1,000,000 reads containing the EcoR1 cutsite). For the former, we either resequenced these to yield greater numbers and read depths of RAD loci or discarded the samples from final analyses.

One notable exception was our sample of Stellaria bungeana Fenzl, a sample that we were consistently unable to successfully sequence to greater read depth but which nonetheless was consistently recovered in the clade it was predicted to belong (based on morphological features) with high bootstrap support (Table 6). Otherwise, phylogenetic inference across our alignments generally resulted in high phylogenetic resolution accompanied by good to high support values throughout the phylogenies (Figs. 7–9). Reciprocal monophyly of taxa was recovered for all included species with multiple representatives (Figs. 7–9); this was interpreted as corroborative evidence that RAD loci were recovering our best estimate of phylogenetic relationships. In two cases (Stellaria alsine and Stellaria crispa; Figs. 8–9), non-monophyly of tips revealed unpredicted phylogenetic patterns rather than failure of RADseq methods (Sharples, companion paper [this issue]).

Phylogenetic Analyses: Outgroups

The alignment used to infer the outgroups phylogeny (Fig. 7) comprised 5,916 RAD loci,

438,029 base pairs, 110,097 variable sites (=SNPs), and 64.58% missing data. Owing to the combination of parameters used to output an alignment that was phylogenetically informative across outgroups, some support values within core Stellaria from the outgroups analysis were poor, although core Stellaria itself was nonetheless recovered as monophyletic with the highest 119

Fig. 7. Phylogenetic hypothesis across select members of “Alsinoideae” and two outgroups in

“Caryophylloideae” based on RAD loci. The star indicates the ancestral node of core Stellaria.

Scale bar: substitutions per site.

120

Fig. 8. Phylogenetic hypothesis of relationships within core Stellaria, with all sampled diversity falling into five major clades. Scale bar: substitutions per site.

121

Fig. 9. Phylogenetic hypothesis of relationships within the largest infrageneric group within core

Stellaria, the Larbreae clade. Sister to Stellaria alaskana are 14 major clades within this lineage, several of which are given informal names here for the first time. Scale bar: substitutions per site.

122 support (Fig. 7). Most other relationships outside of core Stellaria were similarly recovered with high or good support (Fig. 7).

Our analyses strongly support continued recognition of a core Stellaria clade sister to the genus Cerastium and others (Fig. 7). In contrast to members of Stellaria clearly ascribable to the core clade (see below), many other “Stellaria” were here recovered outside of Cerastium + core

Stellaria (Fig. 7; Table 6). The most appropriate generic names for some of these taxa are not completely clear based on our present taxon sampling and thus await further study with expanded outgroup inclusion. Our results, however, demonstrate that Stellaria holostea, long considered to be the type species of the generic name, constitutes a monotypic entity that was resolved external to Stellaria + Cerastium, confirming a similar placement demonstrated in

Greenberg and Donoghue (2011). Elsewhere, our phylogenetic data in combination with morphological features support recognition of a new genus, described below as Nubelaria M. T.

Sharples & E. Tripp. Additionally, our results support continued recognition of the genus

Mesostemma Vved., many species of which have been described under Stellaria in the past (e.g.,

Schischkin, 1936; Rechinger, 1988), as well as resurrection of the genus Adenonema (Bunge,

1835). Finally, Stellaria parva Pedersen was recovered as a member of Cerastium and the new combination is made below (see Revised Taxonomic Concepts).

Phylogenetic Analyses: Core Stars

The alignment used to infer the phylogeny of the core stars (Fig. 8) comprised 25,775

RAD loci, 1,931,158 base pairs, 567,879 variable (=SNPs) sites, and 64.82% missing data.

Based on our phylogenetic analyses, knowledge derived from extensive fieldwork and study of herbarium material (including scores of type specimens), and information gleaned from all major worldwide floristic treatments of Stellaria, we estimate core Stellaria to comprise ca. 112 species 123

(Table 7). Our estimate takes into account our phylogenetic results, traditionally difficult and uncertain species boundaries, future synonymies that seem likely based on our results, species descriptions, and unsampled diversity as identified through the literature and herbarium study. It is likely that this is a slight underestimate, as infraspecific taxa were not an explicit focus of our study and some intraspecific taxa may ultimately better be treated at the species level (see

Discussion regarding S. decumbens lineages). Our estimate of course cannot properly account for species yet to be described. We estimate that up to 13% of core stars species diversity remains to be sampled in future work (Table 7).

Five major clades within core stars were recovered (Fig. 8), and these are here designated with informal clade names prior to establishing a formal infrageneric classification. The bulk of core Stellaria diversity falls into the Larbreae clade (n=75 species sampled here; Figs. 8–9;

Table 6). The second most species-rich monophyletic group was the Petiolares clade (n=14 species sampled here). These two clade names are based on previous morphological infrageneric hypotheses (Endlicher, 1840) and represent the only two lineages across Stellaria that were both recovered as monophyletic and that were more or less in concordance with prior classifications

(Endlicher, 1840; Fenzl, 1842; Pax and Hoffmann, 1934; Schischkin, 1936). The three remaining clades were small in comparison to the Larbreae and Petiolares clades. The Insignes clade was recovered as sister to the remainder of core Stellaria and comprised four species (Fig. 8). Sister to the Larbreae clade was a small lineage (the Nitentes clade) of three species. Finally, sister to the Larbreae + Nitentes clades was a lineage previously treated as the genera Plettkea Mattf. and

Pycnophyllopsis Skottsb. (see Revised Taxonomic Concepts, below), currently represented by phylogenetic data from only two species but predicted to contain several more based on a recent 124 taxonomic treatment (Timaná, 2017). The unknown status of how many species belong to this lineage contributes the largest potential inaccuracy in our core Stellaria species estimate.

Phylogenetic Analyses: Larbreae

The alignment used to infer the Larbreae clade phylogeny (Fig. 9) comprised 40,419

RAD loci, 3,012,940 base pairs, 753,575 variable sites (=SNPs), and 66.55% missing data.

Multiple samples of a given species in the Larbreae clade were consistently recovered as reciprocally monophyletic in our analyses, with the notable exception of Stellaria irrigua Bunge in relation to Stellaria sanjuanensis M. T. Sharples & E. Tripp (Sharples and Tripp, 2019b).

Intraspecific relationships were almost always recovered with 100% bootstrap support (rarely falling into the upper 90s; Fig. 9). Earliest diverging within this clade was a monotypic lineage containing S. alaskana, a taxon endemic to Beringia. Nested within the clade sister to S. alaskana, we recovered 14 subclades of Larbreae; informal names are here given to these 14 clades following previous designations when possible (i.e., largely those recognized by

Schischkin 1936 as series within subsection Larbreae). We recovered five clades lacking such previous appellations, however; these consisted of taxa from eastern portions of the Eastern

Hemisphere (e.g., China, New Zealand, Pan-Himalaya) that do not range into Schischkin’s

(1936) U.S.S.R. study area and which have never been subsequently treated. The status of

Schischkin’s (1936) and others’ classifications is largely subject of the companion paper, and these are only further referenced briefly below.

Impact of Specimen Age on RAD Loci

Although specimen age had a statistically significant negative impact on final data yields

(p=0.0000574, r²=0.08978; Fig. 10; Table 6), in practice, older specimens were not typically a hindrance to phylogenomic inference in Stellaria (Figs. 7–9; Table 1). Read numbers containing 125

Fig. 10. Scatter plot of relationship between specimen age and number of RAD reads with the

EcoR1 cutsite generated per sample. A weak yet significant relationship between specimen age and data yields was recovered: p=0.0000574, r²=0.08978.

126 the EcoR1 cutsite varied the most amongst our most recent samples, from less than one million reads to over 10,000,000 reads per sample (Fig. 10; Table 6). Older specimens (60+ years old), though never producing more than 6,000,000 reads containing the EcoR1 cutsite, nonetheless performed satisfactorily during phylogenomic inference (Figs. 7–9; visual comparison with

Table 6). In this study, we retrieved adequate, usable data from specimens as old as one hundred years (e.g., Stellaria sessiliflora Y. Yabe, collected in 1919) and from many other specimens several decades old.

DISCUSSION

Phylogenetic Circumscription of Core Stellaria

Our results confirm (Greenberg and Donoghue, 2011) that the name Stellaria is massively paraphyletic while also clarifying the breadth of a “core” Stellaria genus. Our sampling strategy inclusive of many different outgroup lineages and approximately 87% of core

Stellaria species resulted in a more narrow definition of Stellaria compared to the broad concepts often seen in recent literature. Inclusion of “false stars” (see below) within Stellaria has resulted in inflated estimates of species numbers in prior works (e.g., Shilong and Rabeler, 2001; Morton,

2005; Mahdavi et al., 2010). Our current estimate of 112 total species of Stellaria includes “core stars” only. Based on our and previous studies, we here delimit core Stellaria to include only and all species in the clade sister to Cerastium + Holosteum + Dichodon + Moenchia (Fig. 7;

Harbaugh et al., 2010; Greenberg and Donoghue, 2011; Dillenberger and Kadereit, 2014; Arabi et al., 2018). Although we do not describe any new species in the current study, we deem it very likely that uncatalogued or undescribed species diversity in Stellaria remains to be discovered in 127 herbaria, in remote areas, or masquerading under other generic names (particularly as members of Arenaria, Cerastium, Minuartia, Pseudostellaria, and so on). The clade of core Stellaria spans five major subclades according to our results, and these include 97 species here sampled.

Additional, other small clades may be discovered pending full phylogenetic sampling of the genus such that we consider five to be a minimum number at present.

Morphological Circumscription of Core Stellaria

Members of core Stellaria are distinct from other white-flowered alsinoids in having deeply-bifid petals (these sometimes absent) usually in combination with six-valved capsules

(some diversity depicted in Fig. 5). Species within core Stellaria are distinguished primarily from species in the main sister genus Cerastium L. in having bilobed petals that are distinct/dissected nearly the full petal length (giving the appearance of ten rather than five petals), three styles, and six straight capsule valves typically splitting down much of the length of the mature fruit (see Key to the genera of Tribe Alsineae at the end of this article). However, occasional individuals bearing four styles can be observed in many core Stellaria species such that the oft-cited number of three styles is not infallible in uniting all members of this clade; furthermore, other aberrations exist, such as the five styles typical of Stellaria aquatica (L.)

Scop. (≡ Myosoton aquaticum (L.) Moench.), confirmed in this study as embedded within the

Petiolares clade (Fig. 8). In contrast, Cerastium has petals bilobed to approximately only 1/3 of the distance from the petal apex to the base, usually five styles, and usually curved capsules with

10 teeth. Teeth is the preferred term in Cerastium to “valves” as in Stellaria, as these teeth are generally restricted to only the most apical regions of the fruit and do not split further down the length of the fruit. Stellaria is distinguished from another close relative, Minuartia L. sensu lato, primarily by the latter having entire petals and bearing capsule valves equaling the number of 128 styles (i.e., usually three). Pseudostellaria is so-named due to its mixture of Stellaria and

Cerastium features, namely its three styles and six capsule valves in combination with shallowly- bilobed petals, but unlike these other genera, Pseudostellaria typically has few ovules (< three) per ovary rather than the numerous (>10) ovules ordinarily found in Stellaria and Cerastium.

Type Species of Stellaria

Stellaria holostea, the former type species of the generic name, was selected as such

(Hitchcock and Green, 1929: 155) on the basis of being “one of the best known of the eight original species” of Stellaria. As with prior phylogenetic work, our results recovered this taxon as sister to the clade containing core Stellaria and Cerastium; this lineage is consistently recovered as monotypic across this and earlier studies, despite having some morphological affinity to Pseudostellaria (Fig. 7; Dillenberger and Kadereit, 2014; Greenberg and Donoghue

2011). Although S. holostea bears three styles, its petal lobes are not bifid beyond the petal mid- point. The designation of this taxon as type of Stellaria was thus unfortunate, but a recent proposal, with which we are in full agreement, to conserve the name Stellaria with a new type

(Stellaria graminea L.) has been presented (Tikhomirov, 2016) and accepted (Applequist, 2017:

505; Wilson, 2017: 742; Shenzhen Code Appendix III, https://naturalhistory2.si.edu/botany/codes-proposals/). This new type species is here resolved as a member of core Stellaria and embedded in the Larbreae clade (Figs. 8–9). Thus, the S. holostea lineage requires a new generic designation, proposed below as Rabelera.

Five Clades of Core Stars

Stellaria radians L. was recovered as part of the Insignes clade, which is sister to the remainder of core Stellaria, but this taxon is sometimes classified in a monotypic genus,

Fimbripetalum (Turcz.) Ikonn. (Pavlova, 1996). Despite its singular floral morphology of having 129 five or more lobes per petal (Fig. 5), it is closely related to other early diverging members of core

Stellaria that bear deeply bilobed petals, namely Stellaria corei Shinners, Stellaria pubera

Michx., and S. sessiliflora. This finding renders the apomorphic floral morphology of S. radians inappropriate as a means of recognizing a separate, monotypic genus for this taxon;

Fimbripetalum is thus appropriately considered a synonym of Stellaria. The phylogenetic standing of the Insignes clade has interesting implications in terms of the group’s biogeography, however, which is highly suggestive of a Tertiary eastern Asian-eastern North American disjunction (M. Sharples and E. Tripp, in prep.).

We recovered a Petiolares clade, united in part with the morphological synapomorphy of bearing distinctly petiolate leaves. The Petiolares clade formed a monophyletic group that was largely congruent with morphology-based hypotheses of its constituents (e.g., Schischkin, 1936;

Fig. 8). Within the Petiolares clade is the geographically widespread Stellaria media complex, which was here resolved as sister to Stellaria nemorum L. of northern Eurasia. The clade containing S. nemorum + the S. media complex was in turn resolved as sister to a clade containing three other taxa: Stellaria sennii Chiov. of Africa that is sister to a clade of Stellaria bungeana + Stellaria aquatica of Siberia and broad Eurasia, respectively. This result was unexpected given that distributions and morphologies of S. nemorum and S. bungeana are quite similar, such that a sister relationship was predicted. Stellaria aquatica, widely placed in the monotypic genus Myosoton even in contemporary treatments (e.g., Shilong and Rabeler, 2001;

Morton, 2005), was here confirmed to be embedded in the Petiolares clade (Greenberg and

Donoghue, 2011; Wan and Liu, 2017; Wang et al., 2017). Sister to these two clades are additional members of the Petiolares clade that are members of the Stellaria cuspidata group

(Sharples, companion paper [this issue]), some species of which remain dubiously resolved (Fig. 130

8). Nonetheless, members of this complex across North and South America form a well- supported clade within Petiolares. Fenzl in Endlicher (1840) placed the petiolate Stellaria flaccida Hook. of Australia in Stellaria section Petiolares Fenzl in Endl. as well, but our inferences instead place S. flaccida in an unrelated clade containing other Australian taxa in the

Larbreae clade (Fig. 9). This result suggests that petioles are a homoplasious character in

Stellaria, an observation further supported by petiole presence elsewhere in the Larbreae clade.

Stellaria nitens Nutt. formed a clade with Stellaria nubigena Standl. plus an accession previously identified as Stellaria alsine Grimm (=Stellaria sp. cf. alsine here) that was sister to the Larbreae clade (Fig. 8). As neither of the former two taxa have been placed within an infrageneric classification (see companion paper for further discussion of S. alsine), we refer to this lineage as the Nitentes clade, after the ostensibly grass-like and reduced-petaled lowland desert species S. nitens of western North America. We further recovered a Plettkeae clade from

South America that was sister to the Nitentes + Larbreae clades. This clade potentially contains several more than just the two species for which we currently have phylogenetic evidence

(Timaná, 2017; Table 7; Fig. 8; see below).

Stellaria alaskana was recovered in a consistent phylogenetic position across analyses, as sister to the remainder of the Larbreae clade (Fig. 8). We refer to this monotypic lineage as the

Alaskanae lineage within the Larbreae clade. Schischkin (1936) proposed multiple series within what he called “Stellaria subsection Larbreae Fenzl in Endl.” (best interpreted as Stellaria section Larbreae Fenzl in Endl. per Greuter, 1995). We here apply many of these series names to the informal clades within the Larbreae clade that our analyses recovered, but no previous, morphology-based infrageneric hypothesis within Larbreae was perfectly congruent with our informal concept of these names. These clades are: Anagalloideae (1 species), Foliaceo- 131 bracteatae (9 spp.), Gramineae (3 spp.), Parviflorae (3 spp.), Pedunculares (12 spp.), Uliginosae

(1 sp.), and Umbellatae (7 spp.). Several other clades of taxa generally restricted to Asia are recognized for the first time here: Brachypetalae (3 spp.), Chinenses (3 spp.), Decumbentes (8 spp.), Patentes (7 spp.), and Vestitae (6 spp.). A clade Maori (4 spp.) of New Zealand is also newly recognized here, and a clade Spinescentes (7 spp.) from Australia is here expanded from an infrageneric name first used to only include Stellaria pungens Brongn. (Endlicher, 1840).

Relationships within these clades are discussed in further detail in the companion paper.

The Varieties of Stellaria decumbens Edgew.

Stellaria decumbens of the Himalayas has previously been divided into several distinct varieties. Our analyses support the view that some of these represent species distinct from the original S. decumbens var. decumbens Edgew., particularly S. decumbens var. polyantha Edgew.

& Hook. f. and S. decumbens var. pulvinata Edgew. & Hook. f. Stellaria decumbens var. pulvinata has always been distinct in the Himalayas by virtue of its pulvinate morphology; it is the only species of Stellaria in the Himalayas with a dense cushion habit. As Stellaria decumbens var. pulvinata was not the closest relative of S. decumbens var. decumbens, and in combination with its morphological distinctness, our data corroborate a species rank for this taxon, proposed earlier by Kozhevnikov (1994) as Stellaria maximowiczii Kozhevn. (Stellaria pulvinata Grub. already existed) (Fig. 9). Stellaria decumbens var. polyantha, on the other hand, was recovered in our analyses in a Larbreae clade (the Patentes clade) unrelated to the clade containing S. decumbens var. decumbens and Stellaria maximowiczii (the Decumbentes clade)

(Fig. 9). Based on this evidence, alongside morphological differences, we propose a new status

(Stellaria polyantha) for this taxon below. It is possible that other varieties in the Stellaria 132 decumbens complex unsampled here (e.g., S. decumbens var. aciculuaris Edgew. & Hook. f.) may also be distinct species from S. decumbens var. decumbens pending further scrutiny.

Pycnophyllopsis and Pycnophyllum

Greenberg and Donoghue (2011) recovered a sample of Pycnophyllopsis cryptantha

(Mattf.) Timaná (as Plettkea cryptantha Mattf.) from Peru as embedded within core Stellaria. In agreement with their result, our work also suggests inclusion of one or more Andean genera within core Stellaria: Pycnophyllopsis tetrasticha (Mattf.) Timaná sampled here was strongly supported as a member of the core stars (Fig. 8). Pycnophyllopsis species inhabit exceptionally high elevations and are cushion-forming plants endemic to the Andes of Peru, Bolivia, Chile, and

Argentina (Molinari-Novoa, 2016; Timaná, 2017). This growth form is rare within core Stellaria, where it has evolved in a limited number of additional instances in response to climate and ecology (e.g., S. maximowiczii above). We thus transfer Pycnophyllopsis cryptantha (Mattf.)

Timaná and Pycnophyllopsis tetrasticha to Stellaria (see Revised Taxonomic Concepts) but refrain from treatment of other Pycnophyllopsis taxa until full phylogenetic sampling is conducted, as the remaining species currently ascribed to this genus either differ significantly in floral characters from the former two, or they differ too subtely to justify separation as distinct species as-yet (Timaná, 2017). All Pycnophyllopsis species are currently included in our probable core stars species count, however, as it is likely that future phylogenetic data will place them as close relatives of P. cryptantha and P. tetrasticha (Table 7). The potentially related genus Pycnophyllum J. Rémy comprises approximately 10 species (Timaná, 2017) that are similarly restricted to high elevation Andean habitats and have cushion growth forms but, unlike most species of Pycnophyllopsis, members of Pycnophyllum are dioecious (which applies also to two weakly separable Pycnophyllopsis species) and differ in overall cushion habit, in addition to 133 other differences (Timaná, 2017). We were unable to sample Pycnophyllum for the present investigation, but prior work suggests that at least two species of Pycnophyllum arose from basal nodes in the phylogeny of Caryophyllaceae and are unrelated to Alsineae (Harbaugh et al.,

2010).

Discrepancies with Earlier Work

Greenberg and Donoghue (2011) recovered Stellaria nitens as sister to the Larbreae clade, with Plettkea cryptantha sister to Larbreae + S. nitens, and two samples of Stellaria alsine sister to all of the former but with low support of some nodes. Our results recovered different relationships in this phylogenetic region between the Larbreae and Petiolares clades, with our

“Plettkea” sample always resolved as sister to all of the aforementioned lineages (Figs. 7–8).

Support values for our topology, which was consistently recovered across analyses, were always high (Figs. 7–8). In another major topology difference, the clade containing S. aquatica (sensu

Greenberg and Donoghue, 2011) was found to be sister to the remainder of the Petiolares clade; our results instead recovered the Stellaria cuspidata group as sister to the remainder of the

Petiolares clade (Fig. 8). Some other discrepancies between our and their studies likely involve misidentifications. For example, we reidentified Stellaria longifolia Muhl. ex Willd. (sensu

Greenberg and Donoghue, 2011) to instead represent Stellaria borealis. Paraphyletic S. aquatica with respect to the placement of S. bungeana (sensu Greenberg and Donoghue, 2011) likely instead represents monophyletic S. aquatica without S. bungeana included (Fig. 8); however, as the voucher information for the samples of relevance here is incomplete or unpublished, we cannot definitively confirm this. And lastly, the close relationship of Stellaria chinensis Regel to

S. media (sensu Greenberg and Donoghue, 2011) instead is here strongly resolved with S. chinensis as member to the Larbreae clade, which is more likely based on geographical evidence 134

(Fig. 9). As we have not seen the S. chinensis specimen personally, we may only speculate that this is the case. However, in and among starworts, and especially in relation to S. media specimens, misidentifications are rampant in herbarium accessions, especially given pronounced morphological homoplaisy throughout the alsinoids and within difficult species complexes (M.

Sharples, pers. obs.; Sharples, companion paper [this issue]). Additionally, lower support values across the Sanger-based phylogenies of Greenberg and Donoghue (2011) also help explain the discrepancies with the present study.

A New Genus of Caryophyllaceae

Greenberg and Donoghue (2011) recovered a clade of two species therein referred to as

Stellaria to be sister to a clade containing Stellaria, Cerastium, Dichodon, Holosteum, Moenchia, and Stellaria holostea L. (≡ Rabelera holostea here). These two species, Stellaria arisanensis

Hayata and Stellaria diversiflora Maxim., warranted further study. When we included these species in either broader outgroup sampling or core Stellaria analyses, they consistently formed a distinct lineage external to the Stellaria + Cerastium clade (Fig. 7; only outgroup-focused analysis shown). Stellaria wushanensis F. N. Williams of China was also recovered as belonging to this clade of false stars. A new genus, Nubelaria M. T. Sharples & E. Tripp, is proposed below to accommodate this small lineage of eastern Asian plants.

Resurrection of Adenonema and Retention of Mesostemma

Our analyses support continued recognition of the Asian genus Mesostemma Vved., which is more closely related to Pseudostellaria than to either Stellaria or Cerastium, but which is often lumped into Stellaria in the literature (e.g., Schischkin, 1936; Fig. 7). Species of

Mesostemma are distinguishable from core Stellaria by having four-merous flowers (amongst other differences): they bear four sepals, four shallowly-bilobed petals, and two styles. Although 135 some southwestern Asian taxa would have been optimally included in our analyses, collections from those regions are generally scarce and/or historical in the herbaria visited across this study.

Our analyses recovered Mesostemma martjanovii (Krylov) Ikonn. (≡ Stellaria martjanovii

Krylov) from the Altai Mountains as monophyletic with Stellaria gypsophiloides Fenzl from eastern Asia; a new combination for the latter is proposed below. At least ten additional species of Mesostemma from central and western Asia were not included in our sequence database and await further work (e.g., Rechinger, 1988).

Our analyses support resurrection of the genus Adenonema Bunge, which, unlike

Mesostemma, is a name whose use has gradually declined. The name has, however, been used as a paraphyletic subgenus (Fenzl in Endl., 1840) or section of Stellaria (e.g., in Pax and Hoffmann,

1934), including such members as true Adenonema (Fig. 7), Stellaria decumbens (firmly embedded within the Larbreae clade: Fig. 9), and Stellaria dicranoides (Cham. & Schltdl.) Fenzl

(Pax and Hoffmann, 1934). Unfortunately, samples of Stellaria dicranoides were obtained but not successfully sequenced in this study. However, based on morphology (three capsules valves, densely-cushioned growth form, single-seeded ovary, etc.), S. dicranoides is likely to be more closely allied to Minuartia and relatives than it is to core Stellaria or closely related genera.

Bunge (1835) erected the genus Adenonema to accommodate a distinct lineage marked by glandular trichomes on the stamens and five-valved, single-seeded capsules (his Adenonema petraeum Bunge, from Central Asia, was first described as Stellaria petraea Bunge in 1830). We agree that this combination of characters is unknown in core Stellaria; glandular trichomes on the stamens alone are absent from core stars, and glands are very rare elsewhere in the core stars.

Fenzl in Endlicher (1840) synonymized the one species in Bunge’s genus Adenonema with

Stellaria, but the reasoning for this change was not conveyed. In the present study, our data 136 support Bunge’s (1835) decision to erect a new genus; we recovered two species formerly ascribed to Stellaria but with morphology fitting the concept of Adenonema, Stellaria cherleriae

(Fisch. ex Ser.) F. N. Williams and Stellaria petraea, as monophyletic and sister to the clade containing Mesostemma + Pseudostellaria. Below, we propose the new combination for S. cherleriae in Adenonema. We furthermore retain A. cherleriae as a species distinctive from

Adenonema petraeum based on its overall growth form and stature as well as in bearing a multi- branching cyme (as opposed to solitary flowers; see also Schischkin, 1936). It is unclear how many additional species may belong to this genus beyond the two recognized here. Other names associated with subgenus Adenonema (Bunge) Fenzl in Endl. are discussed below or are currently considered within the genus Pycnophyllopsis (Timaná, 2017).

False Stars

Although several lineages of starworts and related genera are marked by generally reliable suites of morphological features, exceptions exist. For example, Stellaria aquatica is a five-styled starwort, whereas Cerastium parvum (Pedersen) M. T. Sharples & E. Tripp

(combination proposed below) is a three-styled member of that genus. Moreover, potentially monotypic or otherwise taxonomically depauperate genera (e.g., Nubelaria, Rabelera holostea, etc.) share some characteristics with these much more diverse genera. These factors have contributed amply to taxonomic confusion in the past, and they are furthermore indicative of the evolutionary lability of habit and morphology in alsinoid Caryophyllaceae previously recognized by others (e.g., Dillenberger and Kadereit, 2014). Several species currently treated under

Stellaria (but inappropriately so, hence “false stars”) have qualities somewhat intermediate between core Stellaria and other genera, as in the cases of Mesostemma and Adenonema. For example, some false stars have petal lobes bifid approximately half the length of the petals while 137 still bearing three styles. Some false stars also have six capsule valves that split down the entire fruit length (as in Stellaria and Pseudostellaria) and may have one to numerous ovules per ovary, encompassing the ovule variation found across Stellaria, Cerastium, and Pseudostellaria.

Our phylogenetic work and morphological investigations show that these false stars that bear white, bilobed petals can be distinguished from core Stellaria (despite their similiarities in some floral, fruit and habit traits to core stars) by their lack of petals that are bilobed to the base.

Conversely, some alsinoids that completely lack petals have been difficult to place into proper genera. These taxa are difficult to place especially if lacking fruiting material (e.g., to help distinguish them from Minuartia and other genera). For example, we lack molecular sequence data for the apetalous Stellaria sibirica (Rgl. & Til.) Schischk., but study of herbarium material suggests it is unrelated to core stars; it bears three capsule valves and is filed under Minuartia at the VLA herbarium. In another example, both Greenberg and Donoghue (2011) and Dillenberger and Kadereit (2014) found that the apetalous Stellaria fontinalis (Short & R. Peter) B. L. Rob.

(together with other petalous or apetalous species, i.e., Stellaria howardii Maguire, Stellaria obtusa Engelm., and Stellaria ovata Willd.) were embedded within a paraphyletic grade of

“Minuartia”, closely related to Schiedea and species of Colobanthus, Sagina, and Scleranthus.

Our results confirm that S. fontinalis and S. howardii are distantly related to core stars; we furthermore demonstrate that Stellaria antillana Urb. is closely related to S. howardii (Fig. 7).

Other revisions are necessary: Stellaria fontinalis and S. howardii do not form a clade in our study. The former instead was resolved as closely related to Sabulina macrantha (Rydb.)

Dillenb. & Kadereit (≡ Minuartia macrantha (Rydberg) House) (Fig. 7). Clearly, false stars comprise a series of non-monophyletic groups that are in need of much expanded taxon sampling for better phylogenetic understanding. A key to the genera of Tribe Alinseae as currently 138 circumscribed is found at the end of this article and may aid in adding more clarity to proper generic boundaries in at least the starwort tribe, if not other alsinoid tribes.

Other false stars also warrant discussion here. First, the three entities included in Stellaria section Oligosperma Boiss. by Pax and Hoffmann (1934) should not be considered representatives of core Stellaria. The appropriate genus or genera for two of these taxa is currently unclear, but all three species have shallow petal notches; the third species, Stellaria kotschyana Fenzl, is likely better placed in Mesostemma (Rechinger, 1988). Species affiliated with Stellaria subgenus Schizotechium Fenzl in Endl. are also inappropriately placed among core

Stellaria due to bearing glandular hairs and petals cleft halfway down the petal length, amongst other characters. Pax and Hoffmann (1934) included Stellaria crispata Wall. and Stellaria paniculata Edgew. in what they called Stellaria section Schizotechium Fenzl (best interpreted as

Stellaria subgenus Schizotechium Fenzl in Endl. per Greuter, 1995), both of which are now treated as varieties of Stellaria monosperma Buch.-Ham. ex D. Don (Shilong and Rabeler, 2001) and which thus do not belong to core stars (Fig. 7). Pax and Hoffmann (1934) also included

Stellaria drymarioides Thwait. of Sri Lanka in S. section Schizotechium. Stellaria drymarioides is now considered a synonym of Stellaria mannii Hook.f., the latter a species not allied to core

Stellaria but instead allied with S. monosperma (Fig. 7; together, these taxa are false stars).

Finally, the two entities included in Stellaria subgenus Leucostemma (Benth.) Fenzl in Endl. by

Pax and Hoffmann (1934; as Stellaria section Leucostemma (Benth.) Fenzl) do not belong to core Stellaria; they instead share many features with Mesostemma and may even warrant treatment as the genus Leucostemma pending further study (Rechinger, 1988).

RADseq Phylogenomics 139

As to be expected based on ample success in prior studies at the species level (e.g.,

Cariou et al., 2013; Eaton and Ree, 2013; Cruaud et al., 2014; Hipp et al., 2014; Hou et al., 2015;

Díaz-Arce et al., 2016; Hou et al., 2016a; Massatti et al., 2016; Wessinger et al., 2016; Eaton et al., 2017; Nieto-Montes de Oca et al., 2017; Tripp et al., 2017; Daniel and Tripp, 2018; Beeler et al., in press), RADseq yielded high quality, high quantity data needed to resolve starwort relationships with consistently high support. Despite the fact that 68% of our samples were derived from herbarium materials, we were able to include these in the present study with as much success as with fresh samples, and specimen age had only a minimal impact on read number (Fig. 10). Specimen age is important to explore in terms of utilizing old and possibly degraded specimens in phylogenetic analysis when no other specimens are available, perhaps as a function of general rarity or extinction; thus our results support the findings of others that specimens many decades old are amenable to next-generation sequencing approaches (Staats et al., 2013; Besnard et al., 2014; Beck and Semple, 2015; Bakker et al., 2016; multiple RADseq studies cited above). We contribute to this developing body of work by demonstrating that specimens of variable ages can be processed through the RADseq pipeline across members of alsinoid Caryophyllaceae.

However, a few barriers prevented complete phylogenetic resolution in this study. First, missing data levels inherent in RADseq datasets combined with comparatively low number of restriction site reads for some samples meant that a few of these samples yielded erratic/inconsistent behaviors through various exploratory phylogenetic parameters. Most such low-read samples were ultimately discarded from our analyses, while other lower-read samples used here generally performed well (e.g., S. bungeana; Table 6; Fig. 8). Despite shortcomings with some such samples, a significant body of work has demonstrated that phylogenetic results 140 remain robust to high levels of missing data, given that the data present bear adequate phylogenetic signal (e.g., Wiens, 2003; Wiens and Morrill, 2011; Rubin et al., 2012). This finding is supported in Stellaria and other Caryophyllaceae by the fact that phylogenetic relationships were more or less congruent across widely variable missing data levels explored over the course of this work (Sharples and Tripp, 2019b; other data not shown).

Second, taxa and clades in the Asian radiations of the Larbreae clade are not yet fully resolved here, perhaps due in part to rapid divergence times in those groups (M. Sharples and E.

Tripp, in prep.; Fernández-Mazuecos et al., 2018). Generally lower bootstrap values and phylogenetic inconsistency were recovered in these groups compared with other portions of the starwort phylogeny. Other approaches beyond the strictly ML approach utilized here combined with denser geographical and taxonomic sampling may in the future paint a clearer picture of diversification dynamics in this part of the starwort phylogeny. For example, coalescent based approaches (e.g., Lambert et al., 2015; Fernández-Mazuecos et al., 2018) with focused sampling on groups of interest (Fig. 9) should be considered in the future. We expect that alternate approaches likely will not yield significantly divergent patterns from those herein reconstructed, though (M. Sharples and E. Tripp, in prep.; unpub. data). Third, hybridization is likely to have impacted the evolutionary history of starworts (Chinnappa et al., 2005), potentially complicating phylogenetic analysis. Stellaria fenzlii Regel, for example, was phylogenetically inconsistent, sometimes being found in the Foliaceo-bracteatae clade, and other times being found as a monotypic lineage not immediately related to Foliaceo-bracteatae (data not shown). Two accessions of Stellaria soongorica Roshev. that were always recovered as reciprocally monophyletic behaved similarly, with their phylogenetic inconsitency confined to phylogenetic neighborhoods related to the Pedunculares clade. Though such behavior is often attributed to 141 high levels of missing data (Hipp et al., 2014), we speculate that their irregular placement in our results may instead perhaps in part be attributable to a hybrid origin of these taxa (see results in

McDade, 1992), as neither S. fenzlii nor S. soongorica were characterized by low data yields

(Table 7).

Our work nonetheless provides a robust phylogenetic framework within which to ask numerous downstream questions and to test myriad hypotheses about starworts and related groups. Our results suggest that the evolution of Stellaria exhibits numerous homoplasious transitions amongst some of the starworts’ most conspicuous morphological characters, and it is likely that this phenomenon of recurrent homoplasy may be symptomatic across entire lineages of the alsinoid Caryophyllaceae (Dillenberger and Kadereit, 2014; Sharples, companion paper

[this issue]). Potential explanatory correlates of such widespread convergence and parallelism within the alsinoids in general should be sought in future investigations, particularly as relates to their reproductive biology.

REVISED TAXONOMIC CONCEPTS

I. New Genera

Nubelaria M. T. Sharples & E. Tripp, gen. nov. TYPE SPECIES: Nubelaria arisanensis

(Hayata) M. T. Sharples & E. Tripp (≡ Cerastium arisanense Hayata, Mat. Fl. Formos.: 35–36.

1911). TAIWAN. Arisan, 1907, T. Kawakami s.n (holotype: TI-00394 [photo!]).

Perennial or annual herbs. Stems up to 60 cm, glabrous to sparsely glandular and/or sparsely coated in uniseriate multicellular trichomes, these also occurring on petioles and foliage.

Leaves distinctly petiolate, petioles of mature leaves at least 1 cm (to 2+ cm), often of similar 142 length to or longer than leaf blades; blades broadly ovate with cordate to cuneate leaf bases, at least 1 cm long and up to 2–3.5 cm; blades covered in warty papillae, these sometimes also found on stems; leaf blade apices with a small (1–2 mm) yet conspicuous mucro. Inflorescences of solitary flowers arising from leaf axils. Flowers bisexual. Sepals 5, 5.5–8 mm, green. Petals 5,

5–10.5 mm, bilobed near apex, lobes 1–2.5 mm (ca. 1/5th of petal length), white. Stamens usually

10, ca. 5 mm. Styles usually 3, 2.75–3 mm; ovaries 1.5–2 mm. Fruits exceeding length of sepals; capsule valves 6, splitting apically. Seeds brown to dark purple, usually many per fruit, ca. 1 mm in diameter, testa cells conspicuously ridged.

Distribution and Habitat

The three species are restricted to maritime eastern Asia, to include mountains of Taiwan

(N. arisanensis), lower elevations of Japan (N. diversiflora), and mountains of eastern China (N. wushanensis).

Etymology

The name of this genus originates from the Latin word for “cloud”, reflective of both the maritime climate this genus inhabits and the cloudy muddling of generic boundaries due to parallel morphological evolution in this and other small Stellaria- and Cerastium-like clades.

Notes

Nubelaria primarily differs from Stellaria in its consistently petiolate leaves, leaves with warty papillae and mucros, inflorescences that are always solitary, and petals merely notched rather than deeply bilobed.

Nubelaria arisanensis (Hayata) M. T. Sharples & E. Tripp, comb. nov. Cerastium arisanense

Hayata, Mat. Fl. Formos.: 35–36. 1911. Stellaria arisanensis (Hayata) Hayata, Icon. Pl. Formos.

3: 40. 1913. TYPE: TAIWAN. Arisan, 1907, T. Kawakami s.n (holotype: TI-00394 [photo!]). 143

Nubelaria diversiflora (Maxim.) M. T. Sharples & E. Tripp, comb. nov. Stellaria diversiflora

Maxim., Bull. Acad. Imp. Sci. Saint-Pétersbourg, sér. 3, 18: 379–380, in nota. 1873. TYPE:

JAPAN. Nippon media, 1866, Tschonowski s.n. (lectotype, designated here: K000723434!, isolectotype: LE!). Additional syntype: JAPAN. Ko-isi-wara, 1863, collector unknown (LE!).

= Stellaria diandra Maxim., Bull. Acad. Imp. Sci. Saint-Pétersbourg, sér. 3, 18: 379. 1873.

TYPE: JAPAN. Yokohama. In montibus Hakone, 1862, Maximowicz s.n. (lectotype, designated here: LE01011767!, isolectotypes: K000723436!, LE01011766!, P05006592!, P05006594!).

Notes

Although doubt was expressed in the protologue (Maximowicz, 1873) as to whether S. diversiflora was distinct from S. diandra (the former further being described as a footnote on the same page as the latter), S. diandra was later synonymized into S. diversiflora, and S. diversiflora has been subsequently used in regional floras and herbaria (e.g., Zoku, 1965; collections housed at CAS, MO, PE, etc.). We continue this convention by making the new combination Nubelaria diversiflora, though we have been unable to locate the publication in which the two names were synonymized.

The S. diandra lectotype is here designated due to having more material and more plentiful flowers and fruits than the other isolectotypes. The S. diversiflora lectotype is arbitrarily designated, as it is no better and no worse than the LE specimens.

Nubelaria wushanensis (F. N. Williams) M. T. Sharples & E. Tripp, comb. nov. Stellaria wushanensis F. N. Williams, J. Linn. Soc., Bot. 34: 434. 1899. TYPE: CHINA. District of North

Wushan, 1889, A. Henry 7047 (holotype: K000723674!). 144

Rabelera (L.) M. T. Sharples & E. Tripp, gen. nov. TYPE SPECIES: Rabelera holostea (L.) M.

T. Sharples & E. Tripp. (≡ Stellaria holostea L., Sp. Pl. 1: 422. 1753). Ab. loco, Herb. Linn. No.

584.4 (lectotype: LINN[photo!]; designated by Jonsell and Jarvis in Jarvis et al. 1993: 91).

Perennial herbs. Stems square, up to 60 cm in height, often branching and creeping, glabrous lower and sometimes with cilia on angles of upper stems. Leaves sessile, narrowly lanceolate, 3–11.5 cm x 4–12 mm, with lower leaves longer than upper leaves, and small stiff cilia found along midribs and margins. Inflorescence terminal, cymose, flowers only loosely clustered, pedicilate, pedicels often covered with simple and slightly hooked trichomes. Bracts leafy and smaller than the stem leaves. Flowers bisexual. Sepals 5, 6–10 mm, green. Petals 5, 9–

18 mm, up to twice as long as the sepals, bilobed up to half of petal length, white. Stamens usually 10, ca. 4–7 mm. Styles 3, ca. 3–4.5 mm. Ovary globose, 1.5–2.5 mm. Fruit globose, capsule valves 6, green (before splitting) to tawny (after splitting), approximately equaling the sepals. Seeds brown to dark purple, 1–3 mm in diameter.

Distribution and Habitat

The sole species of Rabelera is distributed widely in both temperate Europe and Asia and is associated with forested areas and their margins.

Etymology

The genus is named in honor of Richard K. Rabeler and his many contributions to our knowledge of both Stellaria and close relatives as well as to the Caryophyllaceae broadly. Rich has been mentor to and contributed greatly to the first author’s research.

Notes 145

Rabelera primarily differs from Stellaria in having square and ciliate stems, elongate leaves that are always sessile and lanceolate, bracts that are always leafy, and petals that are only bilobed one half of the petal length.

Rabelera holostea (L.) M. T. Sharples & E. Tripp, comb. nov. Stellaria holostea L., Sp. Pl. 1:

422. 1753. Alsine holostea (L.) Britton, Mem. Torrey Bot. Club 5(10): 150. 1894. TYPE: Ab. loco, Herb. Linn. No. 584.4 (lectotype: LINN[photo!]; designated by Jonsell and Jarvis in Jarvis et al. 1993: 91).

II. New Combinations

Adenonema cherleriae (Fisch. ex Ser.) M. T. Sharples & E. Tripp, comb. nov. Arenaria cherleriae Fisch. ex Ser. in Candolle, Prodr. 1: 409. 1824. Stellaria cherleriae (Fisch. ex Ser.) F.

N. Williams, Bull. Herb. Boissier, sér. 2 7: 830. 1907. TYPE: RUSSIA. In Dahuriae apricis frequentissima primo vere, 1819, Fischer s.n. (holotype: G00213044 [photo!]).

Notes

Three varieties of Adenonema petraeum were proposed in the original description of

Adenonema. The first, var. alpinum Bunge, was considered by Bunge to represent his original

Stellaria petraea Bunge, with its solitary flowers and compact habit; we concur with this assessment. The second, var. cherleriae (Fisch.) Bunge (as “Cherlerieae”), purportedly also corresponded with Arenaria cherleriae (Bunge, 1835), which has subsequently been treated at the species level in Stellaria (Pavlova, 1996). However, Bunge then listed Adenonema cherleriae var. uniflora Fisch. under var. cherleriae, and type material of this latter variety (K000723695!) 146 does not resemble type material of Arenaria cherleriae, the latter considered to be the basionym of Stellaria cherleriae (Pavlova, 1996). Our sampled material of “Stellaria cherleriae” resembles type material of Arenaria cherleriae, with its multi-branched inflorescences and caulescent stature; this contrasts with the var. cherleriae of Adenonema petraeum that Bunge collected, which instead is close in resemblance to his var. alpinum, i.e., Adenonema petraeum. A third taxon, var. fasciculata (Fisch.) Bunge, was also referenced by Bunge, based on a description first presented by Fischer.

Fischer’s original treatment of two of these entities provides some clarity. Type material collected by Fischer housed in De Candolle’s herbarium includes both Adenonema cherleriae and Adenonema petraeum on the same sheet, both originally presented as varieties of Arenaria cherleriae. Fischer’s Arenaria cherleriae var. uniflora (G00213045 [photo!]) corresponds to

Bunge’s Adenonema petraeum, while Fischer’s Arenaria cherleriae var. fasciculata, according to his protologue, matches no specimens to our knowledge. His locality description for this latter variety, however, matches verbatim a different collection on the previously referenced sheet determined by Fischer as “var. major?”, suggesting that var. fasciculata and unpublished “var. major?” are the same entities. The conclusion that they are likely the same considers that Fischer listed only two varieties of Arenaria cherleriae, and that no additional type material outside of that already referenced is currently known to us. In any case, the G00213045 Fischer specimen is clearly referable to Bunge’s Adenonema petraeaum, while G00213044 is referable to the new combination proposed here, Adenonema cherleriae.

Cerastium parvum (Pedersen) M. T. Sharples & E. Tripp, comb. nov. Stellaria parva Pedersen,

Bot. Tidsskr. 57(1): 44–46. 1961. TYPE: ARGENTINA. Prov. Corrientes, dep. Mercedes, 147 estancia La Concepción, 11 September 1957, Pedersen 4608 (holotype: C10009268 [photo!], isotypes: A00037998 [photo!], CTES0028849 [photo!], US00103231 [photo!]).

Notes

This species was described in Stellaria¸ yet its shallowly notched petals combined with phylogenetic evidence places it within Cerastium. Previous descriptions are otherwise inconclusive as to which genus this species is member, with inadequate description of capsule morphology, and a variable style count (Pedersen, 1961; Morton, 2005). Our finding that this species belongs in Cerastium places a species with three to four styles in a genus widely defined as having five styles, further suggesting that style number is homoplasious in alsinoid

Caryophyllaceae and thus rendering it of questionable taxonomic utility. The obovate to oblanceolate leaves of C. parvum also represent a character state not found in core Stellaria.

Mesostemma gypsophiloides (Fenzl) M. T. Sharples & E. Tripp, comb. nov. Stellaria gypsophiloides Fenzl, in Ledeb. Fl. Ross. 1: 380. 1842. TYPE: MONGOLIA. Bunge s.n.

(lectotype: LE-!; designated by Kozhevnikov in Grubov 1994: 31).

Stellaria cryptantha (Mattf.) M. T. Sharples & E. Tripp, comb. nov. Plettkea cryptantha Mattf.,

Schriften Vereins Naturk. Unterweser 7: 15. 1934. Pycnophyllopsis cryptantha (Mattf.) Molinari,

Polish Bot. J. 61(2): 276. 2016. TYPE: PERU. Morococha, Raimondi 8392 (holotype: B†).

NEOTYPE: PERU. Lima, Huarochirí, Paso de Anticona, Ticlio, 11°35′ S, 76°15′ W, 4850 m, 16

Dec 1998, Timaná & Tate 3767 (USM; isoneotypes: BM, E, F, K, LPB-523 [photo!], MICH,

MO-1100075!, TEX, US; designated by Timaná 2017: 6). 148

Stellaria tetrasticha (Mattf.) M. T. Sharples & E. Tripp, comb. nov. Plettkea tetrasticha Mattf.,

Schriften Vereins Naturk. Unterweser 7: 21. 1934. Pycnophyllopsis tetrasticha (Mattf.) Timaná,

Lundellia 20(1): 4–24. 2017. TYPE: PERU. Ancash, Cajatambo, Schwarze Cordillere über

Ocros, auf dürfig bewachsenem steinigem Boden, 4500–4800 m, Weberbauer 2804a (holotype:

B†). NEOTYPE: PERU. Ancash, Yungay Province, Huascaran National Park, Llanganuco sector, Quebrada Demanda, west of Chacraraju base camp, 4750–4900 m, 9°01′ S, 77°36′ W, 13

Apr 1985, Smith & V. Cautivo 10287 (MO- 052068!, isoneotype: F; designated by Timaná 2017:

8).

III. New Rank & Lectotypification

Stellaria polyantha (Edgew. & Hook. f.) M. T. Sharples & E. Tripp, stat. nov. Stellaria decumbens var. polyantha Edgew. & Hook. f., Fl. Brit. India 1(2): 235. 1874. TYPE: INDIA.

Kashmir, 1822, Royle s.n. (lectotype, designated here: K000723637!). Additional syntype:

INDIA. Sikkim, J. D. Hooker s.n. (not located).

IV. Other Synonymy and Lectotypfication

Stellaria sikkimensis Hook. f., Fl. Brit. India 1(2): 230. 1874. TYPE: INDIA. Sikkim Himalaya, alt. 5–9000 ft., Herb. Griffith, J. D. Hooker s.n. (lectotype, designated here: P05436791!, isolectotypes: E00317573!, P05436790!, P05436793!).

= Stellaria zangnanensis L. H. Zhou, Bull. Bot. Lab. N. E. Forest. Inst., Harbin 9: 56. 1980.

TYPE: CHINA. Xizang, Zayu, alt. 2300 m., 26 July 1973, Qinghai-Xizang Expedition 73-909

(holotype: PE01187867!, isotype: PE01187868!). 149

Notes

Type materials of these two taxa are indistinguishable. The name Stellaria zangnanensis has been previously misidentified and applied to taxa as different from it as Stellaria vestita.

However, comparison of type material indicates that S. zangnanensis is the equivalent of S. sikkimensis but occurring in southeastern Tibet, where S. sikkimensis has previously been unknown to occur (Hara and Febbs, 1979; Grierson, 1984; Shilong and Rabeler, 2001; Ghazanfar and Nasir, 2011; Majumdar, 2007–onwards). No reference to S. sikkimensis is made in the protologue of S. zangnanensis, indicating that the author was potentially unaware of specimens of that taxon. As S. sikkimensis is previously known from the Himalaya of northeastern India,

Nepal, and Bhutan, it is not unreasonable to expect the species to occur just across the Chinese border. Both of these names represent a very distinct entity in Stellaria, most starkly characterized by the dense indumentum of golden uniseriate multicellular trichomes and often ochre stems.

The lectotype of S. sikkimensis is chosen because it has the most material out of the four available and bears both flowers and fruits.

V. The Genera of Tribe Alsineae (sensu Harbaugh et al., 2010; Pusalkar and Singh, 2015;

Sadeghian et al., 2015; Zhang et al., 2017; Arabi et al., 2018)

A key to currently recognized, new, and resurrected genera of tribe Alsineae is provided.

Widespread use of “usually” bespeaks the fact that numerous of the characters presented have evolved in polyphyly or bear marked exceptions within and across genera. Our and previous results support the existence of at least 14 genera in Alsineae, with more to be described ahead.

Stellaria is the largest genus in the tribe. 150

KEY TO THE GENERA OF TRIBE ALSINEAE

1. Petals usually deeply bilobed more than half the length of the petals; petals 5 (lacking or minute in numerous species); styles usually 3 (5 in S. aquatica); capsule valves 5 or 6 2 1. Petals bilobed up to 1/2 the length of the petals or entire; petals 4 or 5 and usually present; styles 2–5; capsule valves or teeth 2, 3, 4, 5, 6 or 10 3

2. Herbs or rarely dense cushions in alpine habitats, annual or perennial; leaves sessile or petiolate; petals variable; capsule valves 6 (5 in S. aquatica, and then herbaceous); seeds many; cosmopolitan Stellaria 2. Woody at the base and underground, perennial; leaves sessile; petals +/- minute; capsule valves 5; seeds few (~1–2); restricted to greater Siberia Adenonema

3. Petals 5 (4 occasionally in Moenchia) and usually present, entire to merely fringed, jagged, or emarginate; herbs 4 3. Petals 4 or 5 and usually present, shallowly notched to bilobed 1/5–1/2 of the petal length; herbs or woody at the base 8

4. Styles 4 or 5, equaling petal number; annuals with linear to narrowly lanceolate leaves; seeds numerous (>30); native to Mediterranean, alien elsewhere Moenchia 4. Styles 2 or 3, less than the petal number; life cycle and leaves variable; seed number variable 5

5. Styles usually 2; seeds few (<4) 6 5. Styles usually 3; seeds few to numerous 7

6. Capsule valves 2; annual; native to southwestern and central Asia and alien elsewhere Lepyrodiclis Fenzl 6. Capsule valves double the style number; annual or perennial; restricted to the Pan-Himalaya Odontostemma Benth.

7. Caespitose perennials; inflorescence solitary; petals white to pink and violet, entire; seeds fewer, <10; restricted to higher elevations in the Pan-Himalaya Shivparvatia Pusalkar & D. K. Singh (≡Solitaria (McNeill) Sadeghian & Zarre) 7. Annual; inflorescence terminal and umbellate; petals white, entire to jagged; seeds numerous, >30; native to Mediterranean climates of the Old World and alien elsewhere Holosteum

8. Styles 2; capsule valves 4; petals 4 or 5, bilobed up to 1/2 petal length; xerophytic perennials; seeds few (~1–2); restricted to mountains of central, western, and southern Asia Mesostemma 8. Styles variable; capsule valves or teeth 3, 4, 5, 6, or 10; petals 5, variously bilobed (but never beyond ½ petal length); annuals or perennials; seeds few to numerous 9 151

9. Petals only shallowly notched or bilobed up to 1/5 petal length; capsule valves 6 (variable in Pseudostellaria); seeds many or few 10 9. Petals bilobed beyond 1/5 length and up to 1/2 length; capsule valves usually 6 or of 10 apical teeth; seeds many 12

10. Leaves petiolate, with numerous warty projections on surfaces; leaf apices with a distinct mucro; flowers solitary; styles 3; seeds many; plants restricted to Maritime Eastern Asia Nubelaria 10. Leaves +/- sessile (sometimes tapering to the base); leaves lacking a mucro and warts; flowers solitary or multiple; plants otherwise not as above; styles 2 or 3; seeds few; Eurasia and western North America 11

11. Stamens usually 10; styles 2 or 3; cleistogamous and chasmogamous flowers usually present; roots tuberous, cormlike; capsules valves variable; Eurasia and western North America Pseudostellaria 11. Stamens 5; styles 3; chasmogamous flowers only; roots slender; capsule valves six; restricted to California and Idaho Hartmaniella M. L. Zhang & Rabeler

12. Styles usually 5 (3 in C. parvum); capsules usually apically curved and dehiscing by 10 teeth; cosmopolitan Cerastium 12. Styles usually 3; capsules straight and dehiscing via valves twice as many as style number; plants of western Eurasia (alien elsewhere) 13

13. Leaves generally less than 3 cm; petals shallowly to deeply notched Dichodon 13. Leaves over 3 cm and up to 11.5 cm; petals bilobed halfway to their base Rabelera

152

CHAPTER 2 (PART II)

INTRODUCTION

Recent phylogenetic work across Stellaria L. (the “starworts”) and relatives (Sharples and Tripp, companion paper [this issue]) recovered five major clades of core Stellaria: the

Insignes (four species), Nitentes (three species), Petiolares (14 species sampled), Plettkeae (one species sampled), and Larbreae clades (75 species sampled). RADseq data confirmed this core clade of Stellaria sister to Cerastium and others (Greenberg and Donoghue, 2011), and based on the taxonomic diversity of our sampling, it is possible to further interpret some patterns and findings in numerous species complexes that comprise much of extant Stellaria diversity. Of the five major clades of starworts, only two had previous generic subdivisional names that were largely congruent with the monophyletic groups recovered in our phylogenetic reconstructions, namely the sizeable Petiolares and Larbreae clades. Within the largest clade, Larbreae, most previously proposed generic subdivisions were found to be paraphyletic or polyphyletic, and it was evident that many were untenable and require extensive reworking or outright abandonment.

These subdivisions of Stellaria were problematic largely due to (I) including members from multiple different clades in a single taxon, or (II) the subdivisional names in question originally were described based on taxa that are not members of the core genus Stellaria (and were often not validly published in the first place). These previous hypotheses of delimitation of groups within Stellaria are partly the subject of the present paper, and it will be seen that the lack of 153 monophyly of these previous groupings is largely due to recurrent, parallel evolution of taxonomically informative characters in unrelated clades of core Stellaria.

Further complicating the infrageneric taxonomic situation of Stellaria, many widespread and species-rich lineages within Stellaria include some questionably distinct taxa. For example, previous work had suggested the need of both splitting and lumping of many of the taxa and synonyms recognized in the large and circumboreal Stellaria longipes Goldie species complex

(Hultén, 1943; Chinnappa and Morton, 1991; Chinnappa et al., 2005; Morton, 2005). Because of the complicated taxonomic history of this group, many of its putative species and synonyms were included in our phylogenetic sampling. Other widespread and variable species complexes with plentiful synonyms were of similar interest. These included the New World Stellaria cuspidata Willd. ex Schltdl. group, with many close relatives within the group and uncertain species boundaries across an expansive geographical range (Volponi, 1993), and the cosmopolitan and weedy Stellaria media complex, with close relatives variously considered as distinct species from or subspecies of S. media (Chater and Heywood, 1964; Sobey, 1981). Other morphologically variable species complexes in Stellaria investigated here include the circumboreal Stellaria longifolia Muhl. ex Willd. group, members of the circumboreal Stellaria borealis Bigelow group and relatives, and the ground-scandent members of Asian Stellaria sharing similarities with the common and widespread Stellaria vestita Kurz. The present paper thus further aims to shed additional light upon Stellaria evolution by offering some insight into how many starwort lineages were recovered within historically difficult groups.

MATERIALS AND METHODS

154

Our best estimate of phylogenetic relationships within Stellaria (Figs. 11–12) combined with previous inferences (see Sharples and Tripp, companion paper [this issue]) warranted recognition of a core Stellaria clade comprising ca. 112 species. We sampled 97 of these species

(~87% of the genus), providing an extensive baseline for comparison with previous taxonomic work. Phylogenies were generated using a ddRADseq protocol (Tripp et al., 2017; Sharples and

Tripp, 2019b) and were inferred with tens of thousands of loci using a maximum likelihood approach (Sharples and Tripp, companion paper [this issue]). A full list of specimens and names sampled for molecular work discussed here may also be found in Table 6 and Appendix 1 of the companion paper.

Observations presented here are additionally informed by ecology, geography, and morphology. This information was derived from numerous literature resources including regional floras, regional taxonomic treatments, and protologues (e.g., Pax and Hoffmann, 1934;

Schischkin, 1936; Wild, 1961; Chater and Heywood, 1964; Zoku, 1965; Hara and Febbs, 1979;

Grierson, 1984; Garnock-Jones et al., 1988; Rechinger, 1988; Volponi, 1993; Kozhevnikov,

1994; Pavlova, 1996; Jørgensen and León-Yánez, 1999; Idárraga-Piedrahita, 2001; Shilong and

Rabeler, 2001; Larsen, 2002; Malyschev and Peschkova, 2003; Chinnappa et al., 2005; Morton,

2005; Majumdar, 2007–onwards; Ghazanfar and Nasir, 2011; Miller and West, 2012; Vlasova,

2012; Duke, 2013; Rodríguez-Jiménez, 2013; Nee and Jørgensen, 2014; Timaná, 2017).

Literature perusal was accompanied by study of type and other herbarium material from the following institutions: ALTB, CAS, CHR, COLO, E, IBSC, K, KUN, LE, MO, NSK, NY, P, PE,

RM, UC, VBGI, and VLA (abbreviations follow Thiers, continuously updated). Concepts put forth herein were augmented further by extensive acquaintance with field populations of 155

Fig. 11. Phylogeny of core Stellaria, with the Petiolares clade emphasized and Larbreae collapsed. Thinner branches reflect areas of lower bootstrap support; see companion paper for exact values.

156

Fig. 12. Phylogeny of the Larbreae clade. Thinner branches reflect areas of lower bootstrap support; see companion paper for exact values.

157 numerous Stellaria species, made across 18 expeditions spanning four continents and approximately 203 days between 2013 and 2018. Species are here considered distinct if they were recovered as separately evolving lineages and additionally differed in at least two other ways (e.g., ecologically, geographically, morphologically) (de Queiroz, 2007). Invalidly published infrageneric names are specified in double quotations throughout, including many of the series published by Schischkin (1936).

RESULTS AND DISCUSSION

The Stellaria cuspidata Group

Most starwort diversity within the Petiolares clade occurs within the New World

Stellaria cuspidata group, nine of whose members were sampled (out of a probable 11 species in total) (Fig. 11; Table 8: Cuspidatae). Stellaria cuspidata was among the first species described from this New World clade (Schlechtendal, 1816). Its description was based on material from the

Andes in Ecuador, but it has been confused at times as a New World occurrence or introduction of Stellaria nemorum L. of western Eurasia (Brako and Zarucchi, 1993; Duke, 2013; Nee and

Jørgensen, 2014), as both species are petiolate and relatively large statured with petals exceeding the sepals. I consider S. nemorum entirely absent from the New World, however, as do most large floras (e.g., Schischkin, 1936; Shilong and Rabeler, 2001; Morton, 2005); moreover, our phylogenetic evidence rejects a close relationship between the broader S. cuspidata group and S. nemorum (Fig. 11).

The Flora of North America (Morton, 2005) states that S. cuspidata also occurs in southern North America in addition to South America, and that the co-distributed Stellaria prostrata Baldwin should be treated as Stellaria cuspidata Willd. ex Schltdl. subsp. prostrata 158

Table 8. Known clades of core Stellaria.

Clade Nested Sampled Constituent Uniting Geographical

Lineage Species Feature(s) Extent

Insignes S. corei, S. pubera, S. Large flowers and Eastern Asia-

radians, S. sessiliflora stature eastern North

America

Larbreae Alaskanae S. alaskana Cryophily, large Beringia

flowers

Larbreae Anagalloideae S. anagalloides Leaves crisped Caucasus

Larbreae Brachypetalae S. brachypetala, S. Cryophily, Central Asia

pusilla, S. winkleri compact growth

Larbreae Chinenses S. chinensis, S. henryi, S. Conspicuous China

salicifolia petals

Larbreae Decumbentes S. arenarioides, S. Petals mostly Eastern Asia

congestiflora, S. reduced

decumbens, S. infracta, S.

lanata, S. maximowiczii,

Stellaria sp., S. uda

Larbreae Foliaceo- S. borealis, S. calycantha, Foliaceous bracts, Circumboreal

bracteatae S. crispa RU, S. crispa +/- cryophilic

US, S. fenzlii, S. humifusa,

S. littoralis, S. nipponica,

S. sitchana 159

Larbreae Gramineae S. graminea, S. palustris, Leaves much Northern

S. persica longer than wider; Eurasia

petals conspicuous

Larbreae Maori S. decipiens, S. gracilenta, Variable New Zealand

S. parviflora, S. roughii

Larbreae Parviflorae S. discolor, S. filicaulis, S. Leaves much Circumboreal

longifolia longer than wider

Larbreae Patentes S. mainlingensis, S. Variable Eastern Asia

montioides, S. nepalensis,

S. patens, S. polyantha, S.

sikkimensis, S. souliei

Larbreae Pedunculares S. altimontana, S. Conspicuous Circumboreal

angarae, S. petals, cryophily

ciliatosepala/S. edwardsii,

S. crassipes, S. dahurica,

S. eschscholtziana, S.

fischeriana, S. longipes, S.

peduncularis, S. porsildii,

S. ruscifolia, S.

soongorica?

Larbreae Spinescentes S. angustifolia, S. Variable Australia

filiformis, S. flaccida, S. 160

leptoclada, S. multiflora,

S. papillata, S. pungens

Larbreae Uliginosae S. alsine Riparian habitats Asia-North

America

Larbreae Umbellatae S. alaschanica?, S. Variable Circumboreal

crassifolia, S. debilis, S. and South

irrigua, S. sanjuanensis, America

S. tomentosa?, S.

uchiyamana?

Larbreae Vestitae S. dianthifolia, S. lanipes, +/- Scandent Eastern Asia

S. omeiensis, S. petiolaris,

S. pilosoides, S. vestita

Nitentes S. sp. cf. alsine, S. nitens, Scarious bracts, North

S. nubigena petals small or America and

lacking Europe

Petiolares Cuspidatae S. arvalis, S. chilensis, S. Petioles, large Subtropical to

cuspidata, S. hintoniorum, petals, large tropical New

S. irazuensis, S. prostrata, stature World

S. recurvata, S.

venezuelana, S. weddellii

Petiolares Mediae S. media, S. neglecta Petioles, reduced Cosmopolitan

petals 161

Petiolares S. aquatica, S. bungeana, Petioles, large Old World

S. nemorum, S. sennii petals, large

stature

Plettkeae S. tetrasticha Cushion growth, Andes

cryophily

162

(Baldwin) J. K. Morton (Morton, 2004). The various views taken in North and South American treatments of Stellaria have contributed to persistent ambiguity in this group, however, with either S. prostrata or S. cuspidata recognized in a given flora (S. prostrata: Beaman, 2009; S. cuspidata: Volponi, 1993; Luteyn, 1999; Nee and Jørgensen, 2014) and with them sometimes appearing in sympatry (Baker and Burger, 1983; Brako and Zarucchi, 1993; Idárraga-Piedrahita,

2011; Jones et al., 2014), or with S. prostrata considered a subspecies or a synonym of S. cuspidata (Morton, 2005; Rodríguez-Jiménez, 2013). In North America, some narrowly- restricted lineages within this complex have been recognized as distinct species, namely Stellaria hintoniorum B. L. Turner of northern Mexico, Stellaria irazuensis Donn. Sm. of Costa Rica, and

Stellaria miahuatlana B. L. Turner of Oaxaca, the former two having been sampled in our work.

Greater taxonomic diversity, however, exists in this group in South America, which hosts a broader distribution of the Cuspidatae clade starworts as well as more morphological variation across species (e.g., Brako and Zarucchi, 1993; Volponi, 1993; Luteyn, 1999; Idárraga-

Piedrahita, 2011; Nee and Jørgensen, 2014). Stellaria arvalis F. Phil. is the southernmost member of these starworts, absent from the northern Andes but ranging to Patagonia from the central Andes (Volponi, 1993). In addition to S. cuspidata from Ecuador, species sampled in our phylogenetic work included S. arvalis, Stellaria chilensis Pedersen of the central Andes,

Stellaria recurvata Willd. ex Schltdl. and Stellaria serpyllifolia Willd. ex Schltdl. of the northern

Andes, as well as Stellaria weddellii Pedersen of Bolivia. Other distinct taxa with affinities to the

Cuspidatae clade from South America currently remain unsampled.

Stellaria arvalis was recovered with high support as sister to all members of the New

World Petiolares clade (=Cuspidatae), with S. weddellii in turn recovered as sister to the remainder of the New World species (Fig. 11). Embedded within the remainder of New World 163 species, a S. prostrata-S. hintoniorum clade was recovered as sister (albeit weakly supported) to another pair of species farther to the south: Stellaria irazuensis was recovered as sister to the

Venezuelan Stellaria venezuelana Steyerm. (Fig. 11). Our sample of S. hintoniorum from near the type locality in Nuevo León, Mexico, strongly resembled type material of both S. hintoniorum and of S. prostrata from Florida. According to the protologue of S. hintoniorum, this species was essentially considered to be a restricted, high elevation lineage of S. prostrata

(identified as S. cuspidata there), with which it is otherwise fully sympatric (Turner, 1991). It is thus possible that these two species are synonymous, and if supported through further work, S. prostrata has priority. Sister to these North American plus Venezuelan samples were three South

American species. In this clade, two samples of S. chilensis formed a highly supported monophyletic pair (Fig. 11), and S. recurvata and S. serpyllifolia formed another highly supported pair, confirming Volponi’s (1993) conclusion that they are synonymous. These two individuals were in turn closely related to S. cuspidata from near its type locality in Ecuador

(Fig. 11).

North American entities in this group are thus phylogenetically distinct from South

American S. cuspidata group members. This was unsurprising, given the disjunction across the lowlands associated with the greater Darien Gap and adjacent areas in the New World Petiolares clade members; starworts are entirely absent from lowland tropical habitats. Our results suggest that the name S. prostrata may represent the widespread “S. cuspidata” group member as it is known north of the Darien Gap, as North American material sampled from Mexico in this study was not immediately related to S. cuspidata and relatives from the northern Andes (Fig. 11).

These results suggest that North American (plus Venezuelan) members of the New World 164

Petiolares group may form a clade embedded within a widespread South American grade, but denser and broader sampling in this group across the New World is clearly needed.

The Stellaria media Complex

Much confusion has surrounded species boundaries in the cosmopolitan Stellaria media complex, consisting of that species, Stellaria neglecta Weihe, and Stellaria pallida (Dumort.)

Crép., with these latter considered either distinct species or subspecies of the former (e.g.,

Schischkin, 1936; Chater and Heywood, 1964; Sobey, 1981; Morton, 2005); numerous other subspecies and likely synonyms are currently underexplored. Our results and other evidence suggest that S. neglecta (and potentially also S. pallida, which remains unsampled with RADseq data) may represent morphological variation within a broad concept of S. media. Our samples included specimens with morphologies attributable to S. media (three samples) and S. neglecta

(two samples), yet neither of these taxa was recovered as monophyletic in our analyses (Fig. 11).

The hypothesis that S. media, S. neglecta, and S. pallida represent morphological variation within one very widespread lineage also is supported somewhat by our result showing no geographic cohesiveness within our five sampled specimens; not only was S. neglecta non- monophyletic, but both of our samples were from North America and yet were not sister taxa

(Fig. 11). Likewise, samples of S. media from the same continent were not sister taxa, indicating widespread and seemingly random worldwide dispersal in this group (Fig. 11). In terms of S. pallida, petal presence or absence is a salient feature distinguishing S. media and S. pallida, with keys describing S. media as “usually” having petals, and S. pallida as “usually” not having petals, creating no clear dichotomy between them (Schischkin, 1936; Chater and Heywood,

1964; Shilong and Rabeler, 2001; Morton, 2005), although seed differences between the two are reported to be consistent across their ranges (R. Rabeler, pers. comm.). However, additional 165 evidence beyond differences in seed characters should be accumulated before S. pallida can be considered a distinct species from the paraphyletic entities of S. media and S. neglecta.

As further display of S. media’s morphological plasticity, in cultivation it may produce zero to five petals over the life of a given plant, and other morphologically plastic features such as stature and anthocyanin content seem to be attributable to growing environment/ecology only

(M. Sharples, pers. obv.; unpub. data). Likewise, stamen number has been used to distinguish species in the S. media complex, but under cultivation it varies as much as petals in this group

(M. Sharples, unpub. data). As mentioned, other putatively delimiting characters between

“species” in the S. media group mostly revolve around seed morphology (i.e., size, color, ornamentation), but relatively slight seed differences in themselves are evidently poor characters for separating species of Stellaria. For example, much variation exists in seed size within experimental populations of other Stellaria species within a single alpine field site in Colorado at

Niwot Ridge (M. Sharples et al., in prep.), and differences in seed color may be attributable to different developmental stages/gradations under observation.

A specimen with the stature of S. neglecta from Ohio (yet originally identified as S. pallida) was sister to a clade of two sample pairs each. In each of these sample pairs, a sample of

“S. media” or “S. neglecta” from North America formed a clade with a sample from Asia (Fig.

11). That is, samples from neither North America nor Asia formed geographical clades. Our phylogenetic results also reject the idea that Stellaria media is member to the genus Alsine (Löve and Löve, 1974a), as the S. media group is deeply embedded within the Petiolares clade of core

Stellaria (Fig. 11).

I thus hypothesize that the Stellaria media complex is extremely variable across its cosmopolitan range, which includes six continents, Madagascar, Greenland, New Zealand, and 166 numerous other islands/landmasses (e.g., Sobey, 1981), and that its variability is largely attributable to different environmental factors. Great morphological variation is common in other widespread, ruderal (human-loving) angiosperms (e.g., Weber and Wittmann, 2012).

Nonetheless, the S. media group would yet benefit from a comprehensive phylogenetic survey including representatives across a greater range of geographical, morphological, and taxonomic variation.

Foliaceo-bracteatae and the Stellaria borealis Group

Fenzl (1842), Schischkin (1936), and subsequent authors placed numerous taxa of the

Larbreae clade into further generic subdivisions. The Foliaceo-bracteatae clade is the only one that somewhat reflects any of Fenzl’s (1842) subdivisions of his Stellaria [unranked] Larbreae

Fenzl (“Divisio”). However, traditional concepts of Stellaria [unranked] Foliaceo-bracteatae

Fenzl (“Subdivisio”) are still paraphyletic, with many members forming a clade but with other members dispersed elsewhere in the Stellaria phylogeny. Fenzl (1842) included S. borealis,

Stellaria crassifolia Ehrh., Stellaria crispa Cham. & Schltdl., Stellaria eschscholtziana Fenzl,

Stellaria humifusa Rottb., and Stellaria ruscifolia Willd. ex Schltdl. in this group. Schischkin

(1936) added Stellaria calycantha (Ledeb.) Bong., Stellaria ebracteata Kom. (recently placed in

Pseudostellaria and not further considered: Zhang et al., 2017), and Stellaria winkleri (Briq.)

Schischk. in his concept of this group, which he recognized as a series. These taxa all have foliaceous bracts, but this character arises in parallel among numerous unrelated starworts (M.

Sharples, unpub. data; inferences from results herein). The Foliaceo-bracteatae clade recognized here, with a name suggestive of leafy bracts, thus only encompasses some of the foliaceous bract diversity in Stellaria. 167

Our sampling recovered a Foliaceo-bracteatae clade that included S. borealis, S. calycantha, S. crispa, S. humifusa, Stellaria littoralis Torr., and Stellaria sitchana Steud. (Fig.

12). Schischkin (1936), in contrast to Fenzl (1842), considered S. eschscholtziana and S. ruscifolia as members of a different series (“Stellaria ser. Ruscifoliae Roshev.”), and these two species were recovered allied with the S. longipes group (discussed below). Stellaria crassifolia is part of the Umbellatae clade rather than a close ally of the Foliaceo-bracteatae clade (Fig. 12).

Additionally, S. winkleri is unrelated to the Foliaceo-bracteatae clade despite its foliaceous bracts, and RADseq data supported placement of S. littoralis in the Foliaceo-bracteatae group for the first time (Fig. 12). In contrast to Torrey (1857) and Morton (2005), this California coastal endemic is neither closely related to Stellaria pubera Michx. nor to Stellaria dichotoma

L., also supported by previous analysis (Sharples and Tripp, 2019b).

The name Stellaria borealis generally does not appear in Eurasian literature, and it was first described from the northeastern United States (Bigelow, 1824). The name Stellaria calycantha, in contrast, appears in Eurasian literature to represent a very closely related entity to the former, and S. calycantha is also widely reported as being sympatric with S. borealis across western North America (e.g., Morton and Rabeler, 1989; Morton, 2005). Phylogenetic results did not place all entities under these names as a monophyletic group, however. Individuals of

Stellaria borealis subsp. sitchana (Steud.) Piper & Beattie formed a distinct clade not immediately related to other samples attributable to S. borealis and S. calycantha (Fig. 12).

Based on this result, resurrection of the name Stellaria sitchana Steud. is justified. This taxon is more closely tied to the coasts of the Pacific Northwest of North America than other members of this complex (Morton and Rabeler, 1989). Stellaria crispa of western North America was recovered as sister to S. sitchana. In addition, an individual referable to S. crispa of the Far East 168 of Russia was recovered as a lineage sister to S. crispa + S. sitchana, and as this sample did not form a clade with the individual of S. crispa from western North America (Idaho) (Fig. 12), this name is rendered paraphyletic. Further comparative work between Asian and North American populations is required to clarify the number of “crispa-like” species.

It was unexpected to recover S. borealis-S. calycantha as sister to S. humifusa rather than as closely related to S. sitchana. One individual of S. borealis from Greenland and one individual of S. calycantha from Oregon formed a clade, and Stellaria littoralis was sister in turn to S. borealis-S. calycantha + S. humifusa (Fig. 12). Stellaria humifusa and S. littoralis are both distinct morphologically and ecologically, and as previously stated, expectations based on morphology and geography were to recover S. sitchana as closely related to S. borealis-S. calycantha. The finding that S. borealis and S. sitchana are not one another’s closest relatives suggests that S. sitchana has evolved in close morphological parallelism with the circumboreal S. borealis-S. calycantha group. Stellaria fenzlii Regel was usually recovered as sister to S. borealis-S. calycantha + S. humifusa + S. littoralis, and Stellaria nipponica Ohwi (endemic to

Japan) was sister to the remainder of the Foliaceo-bracteatae clade altogether (Fig. 12).

A specimen from New Zealand was resolved as closely related to the two S. sitchana samples from North America. No other members of the Foliaceo-bracteatae clade are yet known from the southern hemisphere. The specimen is not immediately separable from S. sitchana and does not warrant a new description; with S. sitchana from western North America it shares lanceolate leaves on average four centimeters long, solitary axillary inflorescences, petals shorter than the sepals, and a riparian habitat. We lacked sufficient populational sampling of S. sitchana to potentially determine whether the New Zealand sample was of natural or introduced provenance in Oceania. However, given that the population was found in boggy areas amongst a 169 weedy parking-lot flora (and in sympatry with the naturalized Prunella vulgaris L.), given its morphological affinities, and that its phylogenetic branch length was not especially protracted

(M. Sharples, unpub. data), this sample likely represents a previously unrecognized anthropogenic introduction of S. sitchana in the southern hemisphere. This is nonetheless startling, as neither S. sitchana nor S. borealis nor any other member of the Foliaceo-bracteatae clade is considered a ruderal weed, and the presence of the New Zealand population raises the question as to how the introduction was first established.

Additional cryptic diversity potentially awaits species-level recognition in the Foliaceo- bracteatae clade, considering the diversity uncovered in our phylogenetic results and the yet unsampled worldwide variation in this clade. Namely, future sampling of individuals identified within the broad S. borealis complex across Eurasia and elsewhere in North America would contribute towards a settled taxonomy for this group. It will be informative to infer their phylogenetic placement in relation to the populations from North America and Greenland, in relation to S. sitchana, and in relation to the other members of the Foliaceo-bracteatae clade.

Pedunculares and the Circumboreal Stellaria longipes Complex

The “Stellaria ser. Ruscifoliae Roshev.” group recognized by Schischkin (1936) included

Stellaria arctica Schischk. (=S. longipes), S. eschscholtziana, Stellaria fischeriana Ser., and S. ruscifolia. A clade of the latter three was, however, recovered embedded within an enlarged concept of “Stellaria ser. Pedunculares Schischk.” (Fig. 12). The original “S. ser. Pedunculares” group included much of the broad S. longipes group bearing scarious bracts: Stellaria ciliatosepala Trautv., Stellaria dahurica Willd. ex Schltdl., Stellaria edwardsii R. Br., Stellaria laxmanni Fisch. ex Ser. (=S. ciliatosepala), Stellaria peduncularis Bunge, and Stellaria soongorica Roshev. (Schischkin, 1936). All members of the “S. ser. Pedunculares” and “S. ser. 170

Ruscifoliae” groups that were sequenced (with the possible exception of S. soongorica: see below), in addition to others, formed a broad Pedunculares clade (Fig. 12; Table 8).

Stellaria longipes was first described from Ontario, Canada, and has subsequently been subject to conflicting opinions of species delimitation across its circumboreal range and across its many regional variants and close relatives. Hultén (1943) produced the first extensive taxonomic treatment of this and related species across their ranges, having studied a diversity of

North American and European Stellaria material. Hultén split species in this complex based on salient morphological features, e.g., bracts scarious vs. leafy, plants glabrous vs. pubescent, and differences in petal length in relation to sepal length. Although Chinnappa, Morton, and others have demonstrated remarkable morphological plasticity possible within North American S. longipes (summarized in Chinnappa and Morton, 1991; Chinnappa et al., 2005), and although S. longipes is currently considered to be inclusive of numerous synonyms (including some proposed by Hultén as distinct species; see Morton, 2005), our results support various species being split from S. longipes, based mainly on geographical and phylogenetic data combined with previous morphological justifications.

In North America, S. longipes from Colorado, Stellaria porsildii C. C. Chinnappa from

Arizona, S. edwardsii of Yukon, and S. ciliatosepala from Alaska were sequenced (the latter two currently considered synonymous with S. longipes: Morton, 2005). Our specimen from Colorado differed from type material from Ontario (K barcode 000723598!) in its inflorescence structure, and the specimen of S. ciliatosepala from Alaska was morphologically akin to the type material of S. longipes from Ontario. The following Asian taxa in this broad group also were included in our analyses: Stellaria altimontana N. S. Pavlova, Stellaria angarae Popov, S. eschscholtziana,

S. fischeriana, and S. peduncularis. From the Beringia region, Hultén (1943) hypothesized that S. 171 alaskana and S. ruscifolia, though perhaps related to the S. longipes group, were not immediately related to the S. longipes species complex. Indeed, Chinnappa and Morton (1991) and Morton

(2005) refrained from synonymizing these two taxa under S. longipes. Both of these also were included in our RADseq dataset, as well as Stellaria crassipes Hultén of Europe.

Our phylogenetic results support Hultén (1943) in his recognition of some species as distinct from S. longipes. His S. crassipes, as well as S. ruscifolia, were recovered as lineages distinct from S. longipes of North America (Fig. 12). Our data placed S. ruscifolia in an Asian clade sister to a North American clade, while S. crassipes always formed a strongly supported clade with S. altimontana of eastern Siberia, indicating that these latter two might represent the same lineage and be synonymous. In the North American clade, samples of S. ciliatosepala and

S. edwardsii from arctic North America formed a clade sister to one containing S. longipes from

Colorado and S. porsildii. Again, given the remarkable morphological plasticity demonstrated in this group (Chinnappa et al., 2005), it remains unclear which of these four North American taxa represent distinct species and which represent synonyms. Hultén’s S. alaskana, on the other hand, occupies a unique phylogenetic position in all of Stellaria, forming a monotypic lineage sister to the remainder of the Larbreae clade (Figs. 11–12).

The clade sister to the North American clade comprised three eastern Asian species.

Stellaria eschscholtziana was recovered as sister to S. ruscifolia, and though overlapping in general geography, the former is quite distinct from the latter in its dense pubescence and its montane-volcanic ecological requirement. Stellaria eschscholtziana may thus represent another edaphic paraspecies (a monophyletic evolutionary offshoot from a more widespread species) in

Stellaria (Sharples and Tripp, 2019b). Stellaria fischeriana was sister to S. eschscholtziana + S. ruscifolia (Fig. 12). The S. crassipes + S. altimontana clade was sister to the North American 172 clade plus the eastern Asian clade, and S. peduncularis of the Altai Mountains was sister to these three clades (Fig. 12). Stellaria angarae was in turn sister to S. peduncularis + the aforementioned, and S. dahurica of Siberia was always recovered as sister to the remainder of these Pedunculares clade members. These findings suggest a possible Siberian origin of this group, and denser sampling of these entities and other synonyms across their distributions may better disentangle taxonomic boundaries in this group. For now, phylogenetic results support that

S. ciliatosepala and S. edwardsii might accurately be considered synonyms of one another and potentially of S. longipes more broadly (Morton, 2005); that S. longipes of Colorado and S. porsildii may represent the same lineage; and that S. altimontana and S. crassipes are potentially the same species. Our closest sampled specimen to the type locality of S. longipes in Ontario was the Colorado specimen, but without sampling from eastern Canada, the phylogenetic standing of

“true” Stellaria longipes remains unknown.

Parviflorae

The Parviflorae clade, described first as “Stellaria ser. Parviflorae” by Schischkin

(1936), included Stellaria diffusa Willd. ex Schltdl. (=S. longifolia) and S. fenzlii. These two taxa are not closely related, with S. longifolia recovered as closely related to the Pedunculares clade, and S. fenzlii recovered in the Foliaceo-bracteatae clade. Stellaria longifolia was found to be closely related to Stellaria discolor Turcz. and Stellaria filicaulis Makino of eastern Asia, even when multiple samples of S. longifolia across its circumboreal range were included (Fig. 12).

This result was unexpected, as S. longifolia had been hypothesized earlier along with S. porsildii as being a hybrid progenitor of S. longipes (Chinnappa, 1992). Our finding that S. longifolia is not immediately related to the North American S. longipes group suggests that viable hybridization across members of different clades in Stellaria may be possible (that is, if 173

Chinnappa’s hypothesis is borne out by further molecular work); Morton and Rabeler (1989) also reported that S. borealis and S. longifolia can hybridize across a greater phylogenetic distance, although these are sterile (though robust). Two strongly supported clades within S. longifolia were also recovered, one North American and the other Eurasian, indicating that at least two geographical lineages are present within this taxon.

The Gramineae and Brachypetalae Clades

"Stellaria ser. Gramineae Roshev.” included Stellaria brachypetala Bunge, Stellaria fontana Popov, a synonym of S. filicaulis (Stellaria jaluana Nakai), Stellaria graminea L.,

Stellaria imbricata Bunge, Stellaria palustris Ehrh. ex Retz., and Stellaria persica Boiss.

(Schischkin, 1936). The validly published Stellaria subsect. Gramineae Tzvelev essentially retained the “S. ser. Gramineae” members of Schischkin (1936) but added S. angarae, which is not supported by our results (see above) (Tzvelev, 2000). The phylogenetic placement of S. fontana and S. imbricata remains unresolved for the time being. However, based on morphology,

I predict that they are not most closely related to S. graminea and its relatives. Of the remaining taxa, only the namesake species S. graminea, S. palustris, and S. persica comprised a monophyletic group, forming a Gramineae clade in our work (Fig. 12). However, since Stellaria has been conserved with a new type (S. graminea: see Sharples and Tripp, this issue), the name

Gramineae will need to be abandoned as an illegitimate infrageneric name when a taxonomic revision of the starworts is undertaken (Article 22.1: Turland et al., 2018).

Stellaria brachypetala of central Asia instead was resolved within a lineage containing other members of the Larbreae clade closely related to the Parviflorae and Pedunculares clades.

This small clade usually was recovered as sister to these latter two clades rather than as a relative of the Gramineae clade (Fig. 12; Schischkin, 1936). We recovered other central Asian taxa as 174 forming a clade with S. brachypetala that initially were expected to be closely related to the

Decumbentes clade (see below) based on similar distribution and habit with the latter. Thus

Stellaria winkleri (Briq.) Schischk. (previously considered a member of “S. ser. Foliaceo- bracteatae”: Schischkin, 1936) was recovered with S. brachypetala, and Stellaria pusilla Em.

Schmid was sister to this pair of species (Fig. 12). As this small clade is distinct on phylogenetic and morphological grounds, we referred to it as the Brachypetalae clade.

Stellaria soongorica is related to the Brachypetalae, Parviflorae, and Pedunculares clades, but our two samples of this taxon were subject to phylogenetic inconsistency, sometimes recovered as more closely related to the Pedunculares clade, or alternatively more closely related to the Brachypetalae clade. A possible explanation for this inconsistency is proposed in the companion paper, as well as potential resolutions for determining this taxon’s true evolutionary history. This species was usually placed as closely related to members of the Pedunculares clade, however, and this placement was originally hypothesized by Schischkin (1936).

Uliginosae

Stellaria alsine Grimm (=Stellaria uliginosa Murray) and Stellaria anagalloides C. A.

Mey. ex Rupr. were the only members placed in “Stellaria ser. Uliginosae Schischk.”

(Schischkin, 1936). The two are not closely related, and S. anagalloides therefore should not be considered a member of the Uliginosae clade (Fig. 12), thus likely making “S. ser. Uliginosae

Schischk.” monotypic. Specimens identified as “Stellaria alsine” revealed an unexpected phenomenon, however, with samples appearing in two unrelated phylogenetic positions, distinguishable by geography moreso than morphology. One specimen from European Russia was recovered in the Nitentes clade (Fig. 11), and two specimens from eastern Asia and eastern

North America formed a monophyletic group sister to the Maori and Spinescentes clades (Fig. 175

12; see below). This suggests that at least two different “Stellaria alsine” taxa have arisen convergently, this perhaps being the most dramatic example of parallelism yet discerned within starworts. Given that Stellaria nubigena Standl. was recovered as sister to the European sample of “S. alsine”, and given its morphological similarity to S. alsine, it is possible that the name S. nubigena is representative of one or more introduced European populations of S. alsine in

Central America. The finding that samples of S. alsine from eastern Asia and eastern North

America formed a monophyletic group may further be representative of a second Tertiary disjunction between these two regions in starworts (Sharples and Tripp, companion paper, this issue; Sharples and Tripp, in prep.) Further systematic investigation into the various “Stellaria alsine” taxa with dense molecular sampling across their distributions is needed.

Umbellatae

"Stellaria ser. Umbellatae Schischk.” (=Stellaria section Irrigua Lazkov) was proposed

(Schischkin, 1936) as a small group including S. irrigua and S. umbellata (now considered synonyms of one another: see Sharples and Tripp, 2019b). An Umbellatae clade containing S. irrigua also includes the newly recognized Stellaria sanjuanensis M. T. Sharples & E. Tripp

(Fig. 12). Stellaria debilis d’Urv. from southern South America was predicted to be related to or part of the Cuspidatae clade due to geography more so than morphology. We found instead that it is allied to S. crassifolia, a circumboreal entity not known from the southern hemisphere.

Stellaria crassifolia and S. debilis formed a clade sister to S. irrigua + S. sanjuanensis in all analyses, forming a core Umbellatae clade. Stellaria crassifolia and S. humifusa are rather similar morphologically and can be found in sympatry. They usually share broadly elliptic leaves, flowers with petals equaling or exceeding the sepals, and similar stature. It was thus unexpected to find that S. humifusa belongs to the Foliaceo-bracteatae clade while S. crassifolia 176 belongs to the Umbellatae clade; this further bespeaks the prevalence of parallelism within

Stellaria. The Umbellatae clade may be larger than these four species (see below).

The Spinescentes and Maori Clades

The Spinescentes clade previously has been recognized as a small subdivision of Stellaria

(Stellaria section Spinescentes Fenzl ex Endl.) for Stellaria pungens Brongn. (and Stellaria squarrosa Hook., now considered a synonym of the former) of Australia (Endlicher, 1840). This species was recovered within a clade of other Australian plants. We referred to these Australian species as the Spinescentes clade, this being the only previous name for a generic subdivision applicable to these stars. However, Stellaria pungens is the only “pungent” or “spinescent” starwort, and thus this name is not descriptive of this clade as a whole. No taxa from New

Zealand previously have been given a taxonomic assignment or appellation, and we refer to a

Maori clade within the Larbreae clade that encompasses most of the starwort diversity in New

Zealand. The great span of forms to be found in Oceania is discussed further in a future work (M.

Sharples and E. Tripp, in prep.).

Other Unplaced Larbreae

Stellaria anagalloides from the Caucasus was recovered as a monotypic lineage sister to the Brachypetalae-Gramineae-Parviflorae-Pedunculares clades (Fig. 12). As mentioned, it was placed by Schischkin (1936) into “S. ser. Uliginosae”, which is untenable based on our results.

Otherwise in the Larbreae clade, numerous Asian core Stellaria have been placed in Stellaria subsection Larbreae (Fenzl in Endl.) Pax & Hoffmann (or S. section Larbreae Fenzl in Endl.) according to original descriptions or subsequent treatments, with no further placement into series or other infrageneric subdivisions. Therefore, many of the lineages recovered in the Larbreae clade lacked previous names. Stellaria sikkimensis Hook. f., for example, was placed by Pax and 177

Hoffmann (1934) in their S. subsection Larbreae, and as this taxon does not occur within the then Soviet Union, it also was not considered by Schischkin (1936). Stellaria lanata Hook. f. and

Stellaria tomentosa Maxim., as other examples, were also placed in S. section Larbreae in protologues without being further treated subsequently.

We recovered several largely Asian clades related to the Umbellatae clade that were species-rich and morphologically diverse (Fig. 12). However, we could not recover consistently the relationships amongst these subclades with confident backbone support (including precise placement of the Umbellatae clade), despite distinct morphologies of species in Asia. Some of our unsampled Stellaria diversity is expected to belong within some of these clades pending further work, perhaps contributing to imperfect phylogenetic resolution at present. These newly recognized clades are now discussed.

The Decumbentes, Vestitae, and Patentes Clades

We recognized two previous varieties of Stellaria decumbens Edgew. of the Himalayas at species rank based on phylogeny and morphology: Stellaria maximowiczii Kozhevn. (=Stellaria decumbens var. pulvinata Edgew. & Hook. f.) and Stellaria polyantha (Edgew.) M. T. Sharples

& E. Tripp (Fig. 12). Other starworts from the Himalayas were expected to be closely related to

S. decumbens due to morphological affinity, namely Stellaria arenarioides S. L. Chen, Rabeler

& Turland and Stellaria congestiflora H. Hara, which share compact growth forms, reduced petals, and alpine habitat with the former. These latter two species formed a clade with S. decumbens, comprising part of a Decumbentes clade (Fig. 12). Other members of the

Decumbentes clade were a potentially undescribed species, Stellaria maximowiczii, and others mentioned below. 178

Several Asian taxa sharing a broad geographical range with members of the Decumbentes clade are characterized by a ground-scandent growth form. This habit is exemplified by Stellaria vestita, one of the most widespread starworts occurring in southeastern Asia (Shilong and

Rabeler, 2001). This species exhibits pronounced morphological variation. Many specimens have ovate leaves, with large portions of the plants bearing a dense covering of stellate trichomes. This indumentum is often so dense that the leaf surface cannot be seen through the vestiture. In contrast, many specimens have lanceolate or elliptic leaves and/or bear only a sparse covering of stellate trichomes, with older leaves scarcely speckled with them at times; however, ovate and lanceolate leaves often may be observed on the same specimen. Yet S. vestita is easy to identify due to the presence of stellate indumentum combined with a scandent habit and ovate to lanceolate leaves, amongst other features. This combination of characters is found otherwise only in the Japanese S. tomentosa group, and therefore geography can be used to quickly distinguish these taxa.

Other scandent Asian taxa also are distinguished by relatively complex pubescence (e.g., lanate, golden, or pilose trichomes). Some species considered as potential relatives of S. vestita based on similar growth form, complex pubescence, and geographical distribution were Stellaria delavayi Franch., Stellaria infracta Maxim., S. lanata, Stellaria monosperma Buch.-Ham. ex D.

Don, Stellaria nepalensis Majumdar & Vartak, Stellaria omeiensis C. Y. Wu & Y. W. Cui ex P.

Ke, Stellaria patens D. Don, Stellaria petiolaris Hand.-Mazz., Stellaria pilosoides S. L. Chen,

Rabeler & Turland, S. tomentosa, Stellaria uchiyamana Makino, and Stellaria yunnanensis

Franch.

Stellaria yunnanensis was consistently recovered on long branches outside of core

Stellaria (data not shown) and therefore was not considered further as a member of Stellaria 179 according to phylogenetic and morphological evidence. Similarly, Stellaria monosperma was supported in our analyses as more closely related to Pseudostellaria than to core Stellaria (see also Greenberg and Donoghue, 2011) and it might ultimately belong to a lineage formerly considered a section of Arenaria (Zhang et al., 2017). Although we were not able to successfully sequence samples of S. delavayi, it is here predicted to be related to S. monosperma due to sharing with it very large leaves (smaller leaves generally present in core Stellaria), glandular pubescence (rare in core Stellaria), and its merely two-cleft petals (Shilong and Rabeler, 2001;

Sharples and Tripp, companion paper [this issue]; study of type material at P).

Our results otherwise indicated that habit and morphological features shared with S. vestita have evolved in parallel within Asia (and indeed elsewhere). Out of the remaining vestita- like species, the most unexpected result was placement of the S. tomentosa group as unrelated to

S. vestita; again, this group is more similar to S. vestita than any other starworts, sharing similar habits, leaf shapes, and indumentum of stellate trichomes, albeit bearing floral differences and occurring in Japan rather than southeastern Asia. The overall similarity was such that the

Japanese representatives were expected to fall within the same clade as S. vestita, if not be sister to it. Instead, the two Japanese taxa were recovered in a clade sister to the core Umbellatae clade

(Fig. 12), though this relationship was not always highly supported, and sometimes this species pair was recovered elsewhere in the Asian Larbreae group (data not shown). Wherever Stellaria tomentosa-S. uchiyamana were recovered, though, they were never close relatives of S. vestita, pointing to another exceptional example of close parallelism within starworts. Stellaria infracta too was expected to be closely related to S. vestita, as not only does it share dense stellate indumentum, but it also exhibits a similar distribution in China and has been described as being decumbent (Shilong and Rabeler, 2001). It was recovered as a close relative of S. vestita (Fig. 180

12), but it was never borne out as sister to S. vestita, further indicative of parallel evolution of stellate trichomes (these are otherwise unknown in the genus).

Stellaria nepalensis was not closely related to other scandent “vestita-like” taxa despite sharing a very similar habit with S. vestita; it was instead recovered as most closely related to S. patens, with S. polyantha in turn sister to both of these taxa, comprising part of a Patentes clade

(Fig. 12). Although S. patens is also somewhat similar in habit and morphology to S. vestita, the former is covered in simple (not stellate) trichomes. Stellaria lanata, similar to S. patens in its lanate indumentum (but this a much denser, more woolly pubescence in the former), was recovered in an entirely different clade from the Patentes clade. Instead, S. lanata was recovered as part of Decumbentes and as sister to all of the high-altitude members of this clade, suggesting that a scandent-habit ancestor gave rise to a small alpine radiation of plants with compact growth in the Himalayas.

Stellaria vestita itself was the namesake for a Vestitae clade (Fig. 12). Stellaria petiolaris and S. pilosoides were recovered in sister clades within the Vestitae clade, and though similar in morphology and habit, they were not recovered as sister taxa, being instead more closely related to S. omeiensis and S. vestita, respectively (Fig. 12). Stellaria omeiensis is morphologically similar to S. vestita but lacks stellate trichomes, instead being glabrous or sparsely pubescent with simple hairs. Other than scandent growth, these two taxa share relatively large, sessile, lanceolate-ovate leaves, axillary and scarious-bracteate cymes, as well as flowers with petals shorter than the sepals. The scandent growth form unites all members of this Vestitae clade to which Stellaria dianthifolia F. N. Williams is recovered as sister (Fig. 12).

Other Implausible Taxonomic Groupings of Starworts 181

Stellaria subgenus Adenonema (Bunge) Fenzl in Endl. is resurrected as a genus in the companion paper. The status of Stellaria section Pseudalsine Boiss. remains unclear. After morphological study of Stellaria alsinoides Boiss. & Buhse (type of this monotypic section), this taxon is similar to Stellaria nitens Nutt. of North America and to Stellaria filiformis (Benth.)

Mattf. of Australia in overall habit, neither of which are closely related, once more indicative of parallel evolution within starworts (Fig. 11). Although perhaps allied to (or even a member of)

Leucostemma and/or Mesostemma in having two styles and four-merous flowers (Nevski, 1937;

Rechinger, 1988), the annual habit and small flowers of S. alsinoides would seem to exclude it from these two genera. Stellaria alsinoides and the validly of potentially recognizing it as the monotypic Tytthostemma alsinoides (Boiss. & Buhse) Nevski can be further addressed once additional sequence data are generated. Stellaria section Holosteae Fenzl in Endl. only included

S. holostea; since S. holostea is not a member of core Stellaria (Sharples and Tripp, companion paper [this issue]), this section can now be ignored.

Fenzl’s (1842) Stellaria “Subdivisio” Scarioso-bracteatae Fenzl in his Stellaria

“Divisio” Larbreae Fenzl likely was meant to encompass that group’s diversity outside of his

Stellaria “Subdivisio” Foliaceo-bracteatae Fenzl as he knew it, i.e., those “Divisio Larbreae” species with scarious bracts. As with foliaceous bracts, scarious bracts evolved in parallel across

Stellaria, and any classification using them is therefore non-monophyletic. Schischkin (1936) did not recognize Fenzl’s “Scarioso-bracteatae” group and instead described multiple smaller

“series” in Larbreae, which largely were applied to the clade names discussed here. Stellaria subsection Discolores N. Vlassova (=“Stellaria ser. Discolores Schischk.”) has been considered monotypic (Stellaria discolor), but as discussed, S. discolor is a member of a clade containing S. filicaulis and S. longifolia. “Stellaria ser. Dichotomae Roshev.”, with S. dichotoma as the type 182 species and with two-cleft petals, is not a member of core Stellaria (Sharples and Tripp, companion paper [this issue]).

The traditional Stellaria subgenus “Eustellaria Fenzl in Endl.” (invalid per article 21.3 in the Shenzhen code: Turland et al., 2018) essentially contains the breadth of the core Stellaria lineage sister to Cerastium and others. “Subgenus Eustellaria” is thus synonymous with Stellaria itself and therefore has little meaning as a taxonomic concept. Once the phylogenetic positions of

S. alsinoides and other lineages are known, a complete reworking of the subdivisional classification of core Stellaria can be undertaken. At present it is clear, though, that traditional morphology-based classifications are problematic in Stellaria on account of recurrent parallel evolution of characters used to delimit these lineages, and additional phylogenetic surprises are yet to be expected in future work.

183

CHAPTER 3

INTRODUCTION

The genus Stellaria L. (Caryophyllaceae, the “starworts”) is a cosmopolitan and herbaceous flowering plant clade numbering around 110 species (Sharples and Tripp, 2019a).

Starworts favor open habitats such as meadows, forest clearings and edges, arctic and alpine tundra, desert-steppe, bare rock slopes, and ruderal ecosystems, but some species also occur in closed habitats such as temperate and cloud forest understories (Fig. 13). Although favoring open habitats, starworts occur in most major ecosystems of the world with the notable exception of lowland tropical forests, and these are probably the only terrestrial barrier to successful dispersal within Stellaria. Stellaria is native to all six habitable continents (Fig. 14) and its taxonomic diversity is greatest in mountainous areas. This diversity is concentrated in high mountains and at higher latitudes, and thus starworts typify the cryophilous (cold-loving) plant biota of northern ecosystems (Fig. 13). In fact, species in Stellaria are amongst both the farthest northern and highest elevation vascular plants in the world (Chinnappa and Morton, 1984; Dolezal et al.,

2016). Conversely, some species are extremely successful cosmopolitan ruderal components, and

Stellaria media is found ubiquitously across urban and other disturbed areas (Ridley, 1930;

Sobey, 1981; regional floras cited herein). Stellaria elatinoides of New Zealand is the only extinct species, this likely through habitat change and subsequent movement of Eurasian aliens

(including of S. media) into its habitat (Lange et al., 2010; M. Sharples, pers. obs.). 184

Fig. 13. Some habitat diversity of Stellaria, with example occupants of each habitat in parentheses. a) Altai Mountains, Russia, alpine tundra wetlands (S. brachypetala, S. irrigua, S. peduncularis) b) Altai Mountains, Russia, subalpine forests (S. brachypetala, S. crassifolia) c)

Altai Mountains, Kazakhstan, mesic subalpine meadows (S. persica) d) countryside north of

Saint Petersburg, Russia, ruderal meadows with patchy boreal forest (S. alsine, S. graminea, S. media, S. nemorum) e) lowlands of the Sonoran Desert, Arizona (S. nitens) f) Murray River

Valley, Australia, arid mallee shrublands and savannah (S. filiformis, S. multiflora) g) northern

Norway, arctic seashores (S. crassifolia, S. humifusa) h) South Island, New Zealand, alpine talus slopes (S. gracilenta, S. roughii) i) Banks Peninsula, New Zealand, temperate rain forest dominated by Fuchsia (S. decipiens) j) high Himalaya, Nepal, the highest altitude screes on

Earth supporting angiosperms (5500 meters+; S. arenaria, S. decumbens, S. pulvinata) k) northern California, Mediterranean-climate coastline (S. littoralis) l) Australian Alps, eucalypt savannah (S. multiflora, S. pungens) m) Kybeyan Mountains, Australia, temperate rain forest dominated by Dicksonia (S. flaccida) n) Kunming Botanical Garden, China, concrete crevices (S. aquatica, S. graminea, S. media) o) subtropical monsoonsal cloud forest, Nepal (S. nepalensis, S. vestita).

185

186

187

Fig. 14. Worldwide approximate numbers of Stellaria species. The first number (if two are present) denotes the number of widespread, native species occupying a given area; the number in brackets denotes the number of endemic species occupying a given area. Percent endemism per land area: Africa: 100%; Australia: 100%; Greater Himalaya: 69%; Maritime Asia: 52%; New

Zealand: 80%; North America: 57%; Northern Eurasia: 20%; South America: 91%.

188

Across its expansive biome and taxonomic diversity, Stellaria exhibits several biogeographic patterns that have long been of interest to phytogeographers. For one, it is very species-rich in the mountains of central and southern Asia; almost half of all starwort species are present there, and a large percentage of those (ca. 70%) are endemic to the region (Fig. 14).

Second, starworts exhibit a puzzling distribution across the southern hemisphere; they occur in southern South America, Oceania, and sub-Saharan Africa, yet previous work suggests that they must be too young for this distribution to be explained through Gondwanan vicariance (Jordan &

Macphail, 2003; Frajman et al., 2009; Valente et al., 2010; Gizaw et al., 2016; Jia et al., 2016).

Third, the genus is diverse across the mountainous amphi-Beringian arc of northeastern Asia and western North America, and it also occurs in eastern North America, central America, and throughout the Andes chain (Fig. 14). Stellaria is moreover well-represented across the Arctic generally and is found across most mountain chains worldwide.

Pax and Hoffman (1934) first proposed the area of origin of Stellaria to be in the mountains of the eastern part of central Asia, where they stated the most species are to be found.

As the greatest diversity of Stellaria is found across the Himalaya proper, throughout the mountain ranges adjoining and comprising the Tibetan Plateau and other parts of western China, and northwards into the Altai and Sayan Mountains of southern Siberia, this center of diversity likely represents Pax and Hoffman’s hypothesized origin area of Stellaria. We refer to this region as the Pan-Himalaya, since all of the major mountain chains of central, southern, and eastern Asia as well as the entirety of the Tibetan Plateau owe their loftiness to the Indian-

Eurasian plate collision, the formation of the Himalaya, and subsequent uplift of the entire region throughout the past 50 million years (Zhisheng et al., 2001; Mulch and Chamberlain, 2006;

Favre et al., 2015; Renner, 2016). Stellaria, though very species-rich in this part of Asia, 189 decreases in taxonomic diversity westwards into more arid regions of western Asia, and also decreases in diversity to the south and north of the Pan-Himalaya (Fig. 14). For example, diversity of Stellaria in tropical southeastern Asia is restricted to widespread weeds and the presence of Stellaria vestita in high mountains (Larsen, 2002). In contrast to the above, Croizat

(1952) hypothesized the origin area of Stellaria as lying in the southern hemisphere, which he described as feeding the center of diversity in Eurasia. Croizat considered a triangle centered around the Near East-Eastern Siberia-Southwestern China as an important location of secondary diversification of Stellaria, which may be interpreted as the Pan-Himalaya. Pax and Hoffman’s

(1934) origin hypothesis is thus in contradiction with that of Croizat. The competing hypotheses can be framed in the following manner: 1) The significant species-rank starwort diversity in the

Pan-Himalaya is indicative of an origin there, or rather 2) is indicative of a secondary center of diversification.

The presence of Stellaria pollen in Tertiary, pre-tundra Beringia (here, Alaska and far northeastern Russia) and other far northern deposits suggests long-term success of the genus in the far north, given its present wide distribution across the Arctic (e.g., Bennike and Böcher,

1990; Matthews and Ovenden, 1990; Fig. 14). This distribution may have been facilitated by repeated Quaternary and Tertiary Bering Land Bridge iterations, which have facilitated movement of plethora plant lineages (e.g., Murray, 1995; DeChaine, 2008; Eidesen et al., 2013;

Wen et al., 2016). Numerous taxa with distributions well represented in the Arctic have been hypothesized as or demonstrated to have arrived in the Arctic from origins in the southern mountains of Asia and North America (Tolmatchev, 1960; Murray, 1995; Weber, 2003; Tkach et al., 2014). Such mountains have been more generally hypothesized to have given rise to cryophilic taxa that were “pre-adapted” for dispersal to cold regions of the far north once 190 northern forest gave rise to northern tundra and extensive glacial ice sheets during the Tertiary-

Quaternary transition into the Pleistocene (Löve and Löve, 1974; Murray, 1995). Founding stocks of the Arctic flora have alternatively been shown as autochthonous in origin, with subsequent movement further south inferred (e.g., Hou et al., 2016a; Hou et al., 2016b). It is therefore of interest to elucidate which of these competing hypotheses better explains starwort diversity in the far north, and to infer the role that Beringia played in facilitating Old World-New

World movements of starworts.

An increasing body of evidence infers that long distance dispersal (LDD) has been instrumental in forming present-day plant distributions across a breadth of taxonomic scales

(Abbott and Brochmann 2003; Richardson et al., 2003; De Queiroz, 2005; Popp et al., 2011; Le

Roux et al., 2014; Tripp and McDade, 2014; Dupin et al., 2017). Although Caryophyllaceae may have originated in Oceania (Jordan and Macphail, 2003), the split of the two major subgroups in

Caryophyllaceae—the caryophylloids and the alsinoids—has been dated at the Middle to Late

Eocene, suggesting that Stellaria had not yet evolved by the time Gondwana had fully separated.

This observation invokes LDD to explain the presence of Stellaria in Oceania (here, Australia and New Zealand), but there is nonetheless an incredible diversity of morphological forms to be found there. This diversity parallels much of the morphological diversity to be found in this genus across the rest of its worldwide distribution (Fig. 15). We therefore hypothesized that numerous LDD dispersal events from diverse geographical areas may help explain the great diversity of forms to be found in Australia and New Zealand.

As Stellaria exhibits extraordinary biome and habitat diversity across its cosmopolitan distribution (Figs. 13–14), we were additionally interested in whether two aspects of climate and habitat were conserved phylogenetically in Stellaria. Although the genus is often characteristic 191

Fig. 15. Some morphological diversity of Stellaria in Oceania, this diversity spanning much of the global breadth of morphological diversity of the genus. A: Stellaria angustifolia of AU, sharing morphological affinity with S. graminea, S. palustris, and others. B: Stellaria pungens of

AU, sharing pungent leaf apices with S. decumbens and others. C: Stellaria filiformis of AU, a basal-leaved, much-branching oddity in the genus. D: Stellaria elatinoides, extinct taxon of NZ, with the habit of S. media and allies. E: Stellaria flaccida of AU, an indefinite, indeterminate ground-sprawler like many other Stellaria worldwide. F: Stellaria multiflora, a unique apetalous annual of AU perhaps approaching S. nitens of western North America. G: Stellaria roughii of alpine NZ, the only starwort to share succulence with S. sanjuanensis and S. humifusa. H:

Stellaria decipiens of NZ, which independently evolved the ground-sprawling habit from S. flaccida. I: Stellaria parviflora of NZ (and, improbably, South America), close to and questionably distinct from S. decipiens above. J) Stellaria gracilenta of NZ, somewhat similar to

S. longipes and allies. Stellaria leptoclada and Stellaria papillata of AU are not shown. 192

193 of cryophilic floras, there are also numerous temperate, aridland, as well as sub-tropical and cloud forest starworts. Additionally, and across biomes, many Stellaria species preferentially occupy either wet or dry habitats (e.g., Ankei, 1982). We thus explored whether these two aspects of ecology characterized multiple branching events on the starwort phylogeny, or alternatively if these two aspects were more conserved and tended to unite clades.

We therefore had multiple goals in reconstructing the biogeographic history of Stellaria across its widespread and dynamic distribution. 1) We sought to test whether the center of diversity of the genus in southern, central, and eastern Asia may have represented the origin area of the genus, or whether this diversity was derived secondarily from origins elsewhere. 2) We sought to infer from where and how many colonizations of the Arctic have occurred in Stellaria, or alternatively to infer if lineages evolved in situ in the Arctic. Peripheral to this, we sought to infer how many times and in which direction(s) the Bering Land Bridges may have facilitated starwort movement. 3) We sought to elucidate how many dispersal events occurred to Oceania to explain the morphological diversity there. 4) We sought to explore the degree to which shifts in habitat and climate tolerances were conserved or not in starworts. Our nearly full taxon sampling allows for broad biogeographical inferences across Stellaria and represents the most comprehensive biogeographical study in the Caryophyllaceae to date.

MATERIALS AND METHODS

Divergence Time Estimation

To estimate divergence times within Stellaria to facilitate hypothesis evaluation, we conducted fossil-constrained calibration analyses using the primary macrofossil described in

Jordan and Macphail (2003). There is a paucity of macrofossils or identified microfossils in 194

Caryophyllaceae generally (Huang et al., 2013), with all divergence time studies heretofore using this same primary calibration point (Frajman et al., 2009; Valente et al., 2010; Gizaw et al.,

2016; Jia et al., 2016). We thus used similar methods as in these previous Caryophyllaceae studies, which nonetheless allowed evaluation of whether RAD data (see below) converged on similar divergence time estimates to those inferred with fewer loci. Our outgroup sampling outside of core Stellaria was skewed towards alsinoid Caryophyllaceae (Stellaria and allies), though two species of caryophylloid Caryophyllaceae (in the genus Silene) and five species of

Caryophyllaceae sister to the alsinoids and caryophylloids were included in our sampling. The fossil reported by Jordan and Macphail (2003) is representative of the common ancestor of alsinoid and caryophylloid Caryophyllaceae and could thus be applied to our phylogeny. It was dated from the middle-late Eocene and has been used to calibrate the crown split of alsinoids and caryophylloids at 34 Mya (Frajman et al., 2009; Valente et al., 2010; Gizaw et al., 2016; Jia et al., 2016). We were unable to use fossil data from our ingroup because of limited material from the Stellaria tribe (Alsineae), which is represented only by scant and recent palynological specimens at present (e.g., Bennike and Böcher, 1990; Matthews and Ovenden, 1990).

To infer the phylogeny of both Caryophyllaceae and Stellaria, we utilized previously published ddRADseq data (Sharples and Tripp, 2019a) in addition to samples newly sequenced for the present study (Appendix 2). Library generation of new samples followed Sharples and

Tripp (2019a). In contrast with our prior work, we conducted phylogenomic analyses here targeting only one accession per species. We utilized iPyRAD v.0.7.23 to construct RAD locus alignments (Eaton, 2018). Our iPyRAD parameters files for each analysis differed from the default by specifying datatype as “ddrad”, restriction overhang as “CAATT,TAC”, and minimum read depth as 3. Minimum number of taxa sharing a locus was changed for each analysis (n=20 195 in the Caryophyllaceae analysis out of 50 species, n=50 in the Stellaria analysis out of 103 species), and the clustering threshold of RAD loci among and across taxa also differed between the two analyses (0.85 in the Caryophyllaceae analysis, and 0.89 in the Stellaria analysis).

BEAST v.2.5.1 was then used to estimate divergence times across Caryophyllaceae and Stellaria

(Bouckaert et al., 2014). Calibrated Yule and birth-death speciation models were evaluated and relaxed molecular clock models with a lognormal distribution were chosen for each. A

GTR+G+I model of molecular sequence evolution was specified since RAD loci exhibit a mixture of conserved and variant sites (Rubin et al., 2012; Zhang et al., 2018; Frajman et al.,

2019; Paetzold et al., 2019). The offset for the calibration point at the most recent common ancestor (MRCA) of alsinoids and caryophylloids (represented by the split of Silene and the large clade sister to it, here) was specified at 34 Mya with a lognormal prior distribution and a standard deviation of 1, and this split was constrained as monophyletic to run the analysis. A lognormal distribution was chosen to both represent the minimum age estimate for the primary calibration while recognizing uncertainty towards the tail end of the age estimate of this fossil.

We specified a mean age of 12 Mya with a normal prior distribution (standard deviation of 2) for the MRCA of Stellaria based on prior results (Frajman et al., 2009; Gizaw et al., 2016; Jia et al.,

2016; Valente et al., 2016), including those recovered herein. A normal distribution was chosen to better evenly incorporate uncertainty around our secondary calibration point rather than considering it as a strict minimum bound. Stellaria was likewise constrained as monophyletic to run the analysis. Four independent MCMC runs spanning 50 million generations each (with sampling every 5000th generation) were used to explore tree space under the specified parameters for both the Caryophyllaceae and Stellaria analyses. Convergence across MCMC runs was assessed by ESS values over 200 as visualized in Tracer v.1.7 (Rambaut et al., 2018), 196 and results from multiple runs were combined. Maximum clade credibility trees were exported in

TreeAnnotator v.1.8.4 (Rambaut and Drummond, 2016) (burn-in: 20%) and visualized in

FigTree v.1.4.3 (Rambaut, 2016).

Phylogenetic alignments and divergence time analyses were conducted on two separate computing systems. Initial processing of RADseq data was conducted on the Summit supercomputer at the University of Colorado, Boulder (Anderson et al., 2017). BEAST analyses were conducted on the CIPRES gateway (Miller et al., 2010).

Ancestral Area and Present Distribution

To understand the biogeographical history of Stellaria, we conducted ancestral area reconstruction on our nearly fully-sampled phylogenetic hypothesis of the genus (Sharples and

Tripp, 2019a). To this we added new RAD samples of Stellaria species that have not been previously sampled phylogenetically: S. fontana, S. himalayensis, S. parviflora of South

America, S. reticulivena, and S. weberbaueri (=Pycnophyllopsis). Several Stellaria species remain unsampled. This resulted in a dataset of 104 tips/species including one species of

Cerastium as an outgroup based on prior work (Greenberg and Donoghue, 2011; Sharples and

Tripp, 2019a). Cerastium is the largest genus in the clade sister to Stellaria and has been inferred as originating somewhere in Eurasia (Scheen et al., 2004). The Cerastium clade is more closely related to Dichodon, Holosteum, and Moenchia than to Stellaria (Harbaugh et al., 2010;

Greenberg and Donoghue, 2011; Zhang et al. 2017; Arabi et al. 2018); these smaller genera are all most diverse in or restricted to what we code as Northern Eurasia here (see below), and

Cerastium is also diverse in this area. Therefore, our Cerastium outgroup tip was coded as present in Northern Eurasia to reflect a probable origin there of the clade sister to Stellaria. 197

Areas of extant taxa of Stellaria were derived from the following floras and taxonomic treatments: An Enumeration of the Flowering Plants of Nepal (Hara, 1979), Flora of Bhutan

(Grierson, 1984), Flora of China (Shilong and Rabeler, 2001), Flora Europaea (Chater and

Heywood, 1964), Flora Iranica (Rechinger, 1988), Flora of Japan (Zoku, 1965), Flora

Malesiana (Larsen, 2002), Flora of New Zealand (Garnock-Jones, 1988), Flora of North

America (Morton, 2005), Flora U.S.S.R. (Schischkin, 1936), Konspekt flory Aziatskoi Rossii

(Vlasova, 2012), Plantae Vasculares Orientis Extremi Sovietici (Pavlova, 1996), and a recent

Australian treatment (Miller and West, 2012). Use of such materials was informed by extensive study of herbarium specimens at panoply institutions (ALTB, CAS, CHR, COLO, E, IBSC, K,

KUN, LE, MO, NSK, NY, P, PE, RM, UC, VBGI and VLA; herbarium codes follow Thiers, continuously updated), as well as by the myriad expeditions conducted by the first author

(Sharples and Tripp, 2019a). We coded each starwort species as occupying one or more of eight geographical areas. These were: Africa (A), Southeastern North America and Central America

(E), Pan-Himalaya (H), Maritime Asia (M), Oceania (O), South America (S), Northern Eurasia

(U), and Western and Boreal North America (W). We split North America into two because while most of the species there are concentrated in the western and northern (i.e., boreal) portions of the continent, there are distinct and endemic lineages in the temperate eastern and southern portions of the continent (including Central America), and these two divisions in North

America reflect markedly different climates and floristic compositions. Species in South

America generally occur in restricted pockets of distribution throughout the Andes; the coded area “S” thus represents the Andes generally, as no South American starworts are native to lowland tropical areas, tepuis, pampas, or hyperarid habitats. Oceania represents Australia or

New Zealand. We split Eurasia into multiple subareas because these represent distinct floristic 198 regions harboring marked degrees of endemism in Stellaria (Fig. 2) (Schischkin, 1936; Zoku,

1965; Hara, 1979; Grierson, 1984; Takhtajan, 1986; Pavlova, 1996; Shilong and Rabeler, 2001;

Vlasova, 2012). Northern Eurasia included Europe, the Caucasus, and Siberia west of maritime regions, as continental Europe alone almost exclusively harbors more widespread taxa (see discussion). Maritime Asia comprised coastal areas of China, northeastern Russia, as well as the

Korean Peninsula and Japan. The Pan-Himalaya represents the region described above. Although it is more conventional to refer to this region as the greater Qinghai-Tibetan Plateau, we note that

Stellaria diversity across the Plateau proper is low compared to the high diversity in adjoining and extended mountain ranges, including the Himalaya proper. If a species was found to largely occur in one or more of the delimited areas but merely peripherally seeped into another area, we coded it as the former area(s) and not as present in its peripheral area. We also coded widespread alien weeds (e.g., S. aquatica, S. graminea, S. media) as originating in northern Eurasia and base this upon first descriptions from Europe as well as reports of occurrences outside of this area described as “introduced”. Stellaria media, for one, is reported from Europe prior to the last glaciation (Ridley, 1930).

The input phylogenetic tree plus area database were analyzed with the R package

BioGeoBEARS to reconstruct ancestral areas of nodes across Stellaria (Matzke, 2013; R Core

Team, 2019). The limit on the maximum number of areas a taxon could occupy was specified at

3, as no starwort is thought to natively occupy all areas utilized in analyses, and several taxa presently occupy up to 3 areas (e.g., Hou et al., 2016b). These widespread, more or less circumboreal species likely never occupied more than the 3 coded areas, this based on major climatic and geographical limits associated with these species. For example, S. humifusa, occupying polar and boreal coastal portions of Maritime Asia, northern Eurasia, and western 199

North America, is highly unlikely to have ever occupied Africa, Oceania, or elsewhere.

Dispersal-Extinction-Cladogenesis (DEC), DIVA-like, and BayArea-like models with and without the “J” founder parameter were run. Model likelihoods were compared using the Akaike information criterion (AIC) and models with lower AIC values were considered more likely.

Ancestral Reconstruction of Ecology

To test if the ecologies of starworts were conserved and to better understand how

Stellaria has occupied so many landmasses worldwide, we reconstructed the ancestral states of two aspects of their niche: their propensity for cryophily and for occupying wet or dry habitats.

Climate and habitat occupancy were scored based on the above resources for delimiting starwort areas. We defined cryophily as occupying habitats subject to average low temperatures below freezing for five or more months of the year. In the northern hemisphere this characterizes boreal, subalpine, alpine, and/or tundra biomes. In the southern hemisphere, cryophily is rare in

Stellaria and is manifest in occupying subalpine or alpine areas (Stellaria is notably absent from

Antarctic tundra). The “temperate” character state was coded for species occupying more moderate ecosystems, including deserts, grasslands, savannahs, temperate forests, cloud forests, and middle to lower montane habitats. Species were coded as occupying dry habitats if moisture is limited for them throughout much of the growing season (e.g., rocky slopes, screes and taluses, open grassy slopes, sandy or barren non-riparian soils, outcrops, ridgetops, forest margins) or if a species was generalist (found across moisture gradients, e.g., on forest floors, forest margins and grasslands, roadsides, and riparian corridors). Taxa were coded as occupying wet habitats if they lived in substrates with constant moisture throughout much or all of the growing season (e.g., creek and stream banks, marine or freshwater shores, swampy woods, boggy or marshy areas, wet mossy areas, wet gravels, rivulets, wet meadows). In a few instances there were ambiguities 200 in deciding what state a given starwort species should be coded for habitat, namely with those species occupying both wet and dry habitats. A majority-rule approach was taken in these instances: if a species was almost always reported from riparian areas but an occasional specimen was collected from a “dry slope” or some such habitat, it was coded as wet (e.g., S. longifolia). In a few instances, a paucity of both specimens and descriptions of species poorly known in the wild meant that detailed habitat information was unavailable. As it is usually specially noted when a starwort prefers wet habitats (whether in protologues, on specimen labels, or in regional treatments), species lacking available information (e.g., S. montioides) were by default coded as preferring dry habitats.

Analyses and figure export were conducted with R version 3.5.2 (R Core Team, 2019) using the packages phytools (Revell, 2012) and ape (Paradis and Schliep, 2019). Equal rates

(ER), all rates different (ARD), and irreversible models were tested using the function “fitMk”, then AIC scores were compared to determine which model most likely explained ecological evolution in starworts. Stochastic mapping under the best performing models in each case was then conducted with the “make.simmap” function and posterior probabilities of 1000 replicates were plotted onto the maximum clade credibility tree generated in BEAST. A transition between character states was inferred when an ancestral state reconstructed with greater than 50% probability for one of our two states gave rise to the alternative character state, and these transitions were considered well supported if both branch posterior probabilities were greater than 80% and reconstructed ancestral node probabilities were greater than 70%. Phylogenetic signal of climate and habitat were quantified with the “phylosig” function in phytools (Pagel,

1999; Revell, 2012). We then tested for correlated evolution between climate and ecology using the “fitPagel” function in phytools (Pagel, 1994; Revell, 2012). The full list of taxa, their 201 biogeographic areas, coded ecologies, and GenBank accession numbers can be found in Table 9 and in the Appendices.

RESULTS

Phylogeny

The Caryophyllaceae phylogeny was inferred from an alignment comprising 41,627 nucleotides, and the Stellaria phylogeny was inferred from an alignment comprising 80,676 nucleotides. The Stellaria phylogeny represents sampling from 92% of species in the genus.

Calibrated Yule model results are reported since they were not significantly different from birth- death results. Analyses in BEAST yielded posterior probabilities that were generally high across nodes of both phylogenies (Figs. 16–17), and phylogenetic relationships and support values were generally congruent with previously published phylogenomic hypotheses within

Caryophyllaceae and within Stellaria using a maximum likelihood approach (Sharples and Tripp,

2019a). The same major clades of Stellaria as in Sharples and Tripp (2019a) were recovered, though relationships within some of these clades differed. Although the Caryophyllaceae tree bore some low backbone support, phylogenetic relationships were congruent with extant knowledge, and support was maximum across shallower nodes (Fig. 16). The Stellaria phylogeny exhibited maximum backbone support, but some areas of shallower divergence were poorly supported, particularly in the main Pan-Himalaya, S. cuspidata group, and S. longipes group radiations (Fig. 17).

Divergence Times and Ancestral Areas

The split between Silene and the alsinoid Caryophyllaceae was here estimated at 35.6

Mya (95% HPD: 34.03–38.85) (Fig. 16). The age of the split between Stellaria and the clade 202

Table 9. Stellaria species sampled in Chapter 3. General distributions, areas specifically sampled for each species, and GenBank SRA sample accession numbers are given (SRAs PRJNA547948,

PRJNA473254, and PRJNA578194). Taxonomic boundaries and number of species under the broad worldwide name “S. alsine” are still unclear, hence uncertainty is expressed under

“distribution” for those names.

Species Distribution Sampled Area GenBank SRA# Stellaria alaschanica Y.Z.Zhao China Gansu, China SAMN11982351 Stellaria alaskana Hultén Alaska & Yukon Alaska, U.S.A. SAMN11982352 Stellaria alsine Grimm ~E Asia-North America North Carolina, U.S.A. SAMN11982353 Stellaria sp. cf. alsine Grimm ~N Eurasia Pskov, Russia SAMN11982476 Stellaria altimontana N.S.Pavlova Russian Far East Khabarovsk Krai, Russia SAMN11982355 Stellaria anagalloides C.A.Mey. ex Rupr. Caucasus Abkhazia, Georgia SAMN11982357 Stellaria angustifolia Hook. Australia Australian Capital Territory SAMN11982359 Stellaria aquatica (L.) Scop. Eurasia & beyond Gandaki Pradesh, Nepal SAMN11982362 Stellaria arenarioides S.L.Chen, Rabeler & Turland Himalaya Qinghai, China SAMN11982363 Stellaria arvalis F.Phil. Southern South America Los Lagos, Chile SAMN11982364 Stellaria borealis Bigelow ~W & N North America Greenland SAMN12277172 Stellaria brachypetala Bunge Central Asia Altai Republic, Russia SAMN11982365 Stellaria bungeana Fenzl N Asia Altai Republic, Russia SAMN09373290 Stellaria calycantha (Ledebour) Bongard ~Circumboreal Oregon, U.S.A. SAMN11982366 Stellaria chilensis Pedersen Central Andes Cajamarca, Peru SAMN11982368 Stellaria chinensis Regel China Jiangxi, China SAMN11982369 Stellaria congestiflora H.Hara Himalaya Sikkim, India SAMN11982371 Stellaria corei Shinners E United States Alabama, U.S.A. SAMN11982372 Stellaria crassifolia Ehrh. Circumboreal Finnmark, Norway SAMN11982374 Stellaria crassipes Hultén Europe Lappland, Sweden SAMN11982376 Stellaria crispa Cham. & Schltdl. W North America Idaho, U.S.A. SAMN11982377 Stellaria sp. cf. crispa Cham. & Schltdl. Northeastern Russia Kamchatka Krai, Russia SAMN11982378 Stellaria dahurica Willd. ex Schlecht. Central Asia Republic of Buryatia, Russia SAMN11982380 Stellaria debilis d’Urv. S South America Tierra del Fuego, Argentina SAMN11982381 Stellaria decipiens Hook.f. Oceania Antipodes Island, N.Z. SAMN11982382 Stellaria decumbens Edgew. Himalaya Yunnan, China SAMN11982383 Stellaria dianthifolia F.N.Williams China Sichuan, China SAMN11982385 Stellaria discolor Turcz. Far East of Asia Primorsky Krai, Russia SAMN11982386 Stellaria eschscholtziana Fenzl Far East of Asia Kamchatka Krai, Russia SAMN11982389 Stellaria fenzlii Regel Far East of Asia Kuril, Russia SAMN11982390 Stellaria filicaulis Makino Far East of Asia Primorsky Krai, Russia SAMN11982391 Stellaria filiformis (Benth.) Mattf. Australia Victoria SAMN11982392 Stellaria fischeriana Ser. Far East of Asia Kamchatka Krai, Russia SAMN11982393 Stellaria flaccida Hook. Australia New South Wales SAMN11982395 Stellaria fontana Popov Central Asia Bamyan, Afghanistan SAMN13053022 Stellaria gracilenta Hook.f. New Zealand Canterbury, N.Z. SAMN11982397 Stellaria graminea L. Cosmopolitan Weed Wisconsin, U.S.A. SAMN09373291 Stellaria henryii F.N.Williams China Shaanxi, China SAMN11982398 Stellaria himalayensis Majumdar Himalaya Province No. 1, Nepal SAMN13053023 Stellaria hintoniorum B.L.Turner Mexico Nuevo Leon, Mexico SAMN11982399 Stellaria humifusa Rottbøll Circumboreal Finnmark, Norway SAMN11982402 Stellaria infracta Maxim. China Sichuan, China SAMN11982403 Stellaria irazuensis Donn.Sm. Central America San José, Costa Rica SAMN11982405 Stellaria irrigua Bunge E Asia-W North America Altai Republic, Russia SAMN11982409 Stellaria lanata Hook.f. Himalaya Yunnan, China SAMN11982410 Stellaria lanipes C.Y.Wu & H.Chuang Himalaya Yunnan, China SAMN11982411 Stellaria leptoclada C.H.Mill. & J.G.West Australia New South Wales, Australia SAMN11982413 Stellaria littoralis Torr. W North America California, U.S.A. SAMN11982415 Stellaria longifolia Muhl. ex Willd. Circumboreal Colorado, U.S.A. SAMN11982418 Stellaria longipes Goldie W & N North America Colorado, U.S.A. SAMN09373313 Stellaria edwardsii R.Br. W & N North America Yukon, Canada SAMN11982387 203

Stellaria mainlingensis L.H.Zhou Himalaya Tibet SAMN11982421 Stellaria maximowiczii Kozhevn. Himalaya Sichuan, China SAMN11982424 Stellaria media (L.) Villars Cosmopolitan Weed Uttarakhand, India SAMN11982426 Stellaria multiflora Hook. Australia South Australia SAMN11982430 Stellaria neglecta Weihe Widespread Weed Louisiana, U.S.A. SAMN11982432 Stellaria nemorum L. Eurasia Leningrad, Russia SAMN09373315 Stellaria nepalensis Majumdar & Vartak Himalaya Province No. 3, Nepal SAMN11982433 Stellaria nipponica Ohwi Japan Chubu, Japan SAMN11982434 Stellaria nitens Nutt. W North America Nevada, U.S.A. SAMN11982435 Stellaria nubigena Standl. Central America Cartago, Costa Rica SAMN11982436 Stellaria palustris Ehrh. ex Retz. Eurasia Lower Silesia, Poland SAMN11982440 Stellaria papillata C.H.Mill. & J.G.West Australia Victoria, Australia SAMN11982441 Stellaria parviflora Banks & Sol. ex Hook.f. Oceania Canterbury, New Zealand SAMN11982443 Stellaria sp. cf. parviflora Banks & Sol. ex Hook.f. S South America Tierra del Fuego, Argentina SAMN13053025 Stellaria patens D.Don Himalaya Gandaki Pradesh, Nepal SAMN11982445 Stellaria peduncularis Bunge W North America Altai Republic, Russia SAMN09373316 Stellaria persica Boiss. W Asia Samtskhe-Javakheti, Georgia SAMN11982446 Stellaria petiolaris Hand.-Mazz. China Yunnan, China SAMN11982448 Stellaria pilosoides S.L.Chen, Rabeler & Turland China Yunnan, China SAMN11982451 Stellaria polyantha (Edgew.) M.T.Sharples & E.Tripp Himalaya Province No. 1, Nepal SAMN11982452 Stellaria porsildii C.C.Chinnappa W North America Arizona, U.S.A. SAMN11982453 Stellaria prostrata Baldwin S North America Neuvo Leon, Mexico SAMN11982454 Stellaria pubera Michx. E North America North Carolina, U.S.A. SAMN11982455 Stellaria pungens Brongn. Australia South Australia SAMN11982456 Stellaria pusilla Em. Schmid Central Asia Xinjiang, China SAMN11982457 Stellaria radians L. Eastern Asia Sakhalin Island, Russia SAMN11982458 Stellaria recurvata Willd. ex Schltdl. Northern Andes Pastaza, Ecuador SAMN11982459 Stellaria reticulivena Hayata Himalaya Gasa, Bhutan SAMN13053024 Stellaria roughii Hook.f. New Zealand Canterbury, N.Z. SAMN11982460 Stellaria ruscifolia Willd. ex Schltdl. Far East (Asia) & W Alaska Kuril, Russia SAMN11982461 Stellaria salicifolia Y.W.Tsui & P. Ke E Asia Gansu, China SAMN11982462 Stellaria sanjuanensis M.T.Sharples & E.Tripp W North America Colorado, U.S.A. SAMN11982463 Stellaria sennii Chiov. Africa Bale, Ethiopia SAMN11982464 Stellaria sessiliflora Y.Yabe Japan & Jeju (Korea) Kanto, Japan SAMN11982466 Stellaria sikkimensis Hook.f. Himalaya West Bengal, India SAMN11982467 Stellaria sitchana Steudel W North America California, U.S.A. SAMN11982469 Stellaria soongorica Roshev. Central Asia Naryn, Kyrgyzstan SAMN11982472 Stellaria souliei F.N.Williams China Yunnan, China SAMN11982473 Stellaria tetrasticha (Mattf.) M.T.Sharples & E.Tripp N Andes Ancash, Peru SAMN11982477 Stellaria tomentosa Maxim. Japan Kansai, Japan SAMN11982478 Stellaria uchiyamana Makino Japan Chūgoku, Japan SAMN11982479 Stellaria uda F.N.Williams China Sichuan, China SAMN11982480 Stellaria venezuelana Steyerm. Venezuela Sucre, Venezuela SAMN11982482 Stellaria vestita Kurz S Asia Yunnan, China SAMN11982484 Stellaria weberbaueri (Muschl.) M.T.Sharples & N Andes La Paz, Bolivia SAMN13053026 E.Tripp Stellaria weddellii Pedersen S Andes La Paz, Bolivia SAMN11982485 Stellaria winkleri (Briq.) Schischk. Central Asia Xinjiang, China SAMN11982486

204

Fig. 16. Divergence time estimates across select Caryophyllaceae based on our primary macrofossil calibration, with 50 species included. A maximum clade credibility tree generated in

BEAST based on 4 independent 50,000,000 MCMC generations is shown. Posterior probability values of branch support are depicted; most branches are maximally supported except along the backbone. Stellaria likely originated ca. 12 Mya in the Miocene. The bottom axis denotes millions of years ago; the Holocene is not depicted. Blue bars depict the 95% HPD around node heights. 205

206

Fig. 17. Divergence time estimates across Stellaria based on the calibration point derived from the analysis depicted in Figure 16 (and others) and inclusive of 103 Stellaria species and one outgroup. A maximum clade credibility tree generated in BEAST based on 4 independent

50,000,000 MCMC generations is shown. Posterior probability values of branch support are depicted; most branches are maximally supported except along the backbone of the very rapid

Pan-Himalaya radiation. The bottom axis denotes millions of years ago; the Holocene is not depicted. Blue bars depict the 95% HPD around node heights. Larger branch numbers indicate clade names referenced in the text: 1, Insignes; 2, Petiolares; 3, Plettkeae; 4, Nitentes; 5,

Larbreae. 207

208 containing Cerastium was estimated at 16.68 Mya (12.96–19.08) (Fig. 16). The origin node of

Stellaria was placed at 12.34 Mya (9.39–15.60) (Fig. 16; 11.40 Mya in Fig. 17) and represents the deepest split in the genus, giving rise to the eastern Asian-eastern North American Insignes clade (Figs. 16–17). All models in BioGeoBEARS with the +J parameter added performed significantly better than models lacking this parameter (Table 10). However, ambiguous placement of the origin area of Stellaria was also recovered across models but appears to have been somewhere between northern Eurasia and the New World under the BayArea-like+J model

(Table 10; Fig. 18). Although DIVA-like+J and DEC+J were more likely than the BayArea- like+J model (Table 10), we present results from the BayArea-like+J model, as it displays fewer and more realistic area likelihoods along the Stellaria backbone and all models were in agreement in terms of general patterns recovered at shallower nodes. The DIVA-like+J and

DEC+J results can be found as Supplementary Figures 1–2.

The wide Eurasian and New World Petiolares clade split from the rest of Stellaria 9.95

Mya (95% HPD: 6.07–13.77) (Fig. 17) from a likely northern Eurasian origin (Fig. 18), and subsequent deep splits in the Stellaria phylogeny were dated thereafter (divergence of South

American Plettkeae, 7.64 Mya (4.26–10.48); divergence of North American – European

Nitentes, 6.63 Mya (3.99–9.49)) (Fig. 17). The core clade containing the majority of Stellaria diversity (Larbreae) then originated 5.40 Mya (3.27–7.86) likely in northern Eurasia and subsequently colonized the Pan-Himalaya, Maritime Asia, and North America multiple times each, but Stellaria never reached Africa again and only reached South America once more (Figs.

17–18). A large secondary diversification in the Pan-Himalaya was extremely rapid, giving rise to at least 33 extant species within the past 1.89 Mya (1.04–2.83) (Figs. 17–18), a few of which eventually extended into extra-Asian distributions (e.g., S. crassifolia, S. irrigua). Across these 209

Table 10. BioGeoBEARS model comparisons. Chi-squared comparisons resulted as follows:

DEC vs. DEC+J (p=2.2e-08), DIVALIKE vs. DIVALIKE+J (p=2.7e-10), BAYAREALIKE vs.

BAYAREALIKE+J (p=1.4e-22).

Model LnL #params d e j AICc AICc_wt DEC -183.9114 2 0.017043428 4.667348e-03 0 371.9416 3.021939e-07 DEC+J -168.2697 3 0.008305894 1.000000e-12 0.01967126 342.7794 6.498105e-01 DIVALIKE -188.8309 2 0.022337605 1.000000e-12 0 381.7807 2.206841e-09 DIVALIKE+J -168.8883 3 0.009811970 1.000000e-12 0.02001794 344.0166 3.500490e-01 BAYAREALIKE -224.4975 2 0.019376805 2.057007e-01 0 359.6625 7.144549e-25 BAYAREALIKE+J -176.7113 3 0.006288775 1.000000e-12 0.02696913 344.6045 1.401717e-04

210

Fig. 18. Ancestral area reconstruction of Stellaria inferred with BioGeoBEARS under the

BayArea-like+J model. Node pies represent the probability that an ancestral area was occupied.

The origin of the genus was reconstructed ambiguously in the New World and in northern

Eurasia, with deep nodes retaining that ambiguity. However, ancestral nodes of most subclades in Stellaria were reconstructed with high probabilities as belonging to one area. Black asterisks indicate inferred dispersal events from the Old World into the New World; red asterisks indicate such events in the opposite direction. The asterisk representing the dispersal from Oceania to

South America is the only movement between the Old and New Worlds geographically far removed from the Bering Land Bridges, and these bridges are invoked to explain 14 other movements between the Old and New Worlds. 211

212 latter as well as other dispersal events between the New World and Asia, usage of Bering Land

Bridges was inferred 14 times across divergence dates (Fig. 18). All sampled Stellaria of

Oceania formed a clade related to the Pan-Himalayan radiation and was dated at 2.95 Mya

(1.62–4.43), indicating only one colonization of the region. No area besides Africa was reconstructed as a fully irreversible sink area that prohibited dispersal elsewhere, although all areas tended to “trap” lineages into radiating there.

Ecology

Trait irreversible models for habitat and climate were rejected (Table 11). The common ancestor of Stellaria was reconstructed as likely inhabiting temperate areas, but whether it preferred wet or dry habitats could not be reconstructed unambiguously (Fig. 19). From a temperate ancestral state, numerous transitions to cryophily occurred, including twice in the

Petiolares clade as well as in the most recent common ancestor of the Plettkeae, Nitentes, and

Larbreae clades (Fig. 19). However, transitions back to temperate climates from cryophily were three times more common than transitions to cryophily from temperate climates, as shown by support for an ARD over an ER model (p=0.016). Stochastic mapping agreed with this finding and estimated 37.868 changes between climates, recovering 11.468 transitions from temperate to cryophilic and 26.4 transitions from cryophilic back to temperate on average. This is best explained through the patterns recovered within the ancestrally cryophilic Larbreae clade: nine reoccupancies of temperate conditions were likely from a cryophilic ancestor in the primary Pan-

Himalaya radiation alone (Fig. 19). Although most subclades of Stellaria were more or less united by cryophily or temperateness (Pagel’s λ= 0.842; p= 2.101504e-07), shifting of climate within these clades was still possible (Fig. 19). In contrast, preference for wet or dry habitats exhibited much less phylogenetic structure and is more evolutionarily labile in starworts than 213

Table 11. Ancestral state reconstruction of climate occupation and habitat results based on maximum likelihood. The favored model for each analysis is bolded. T=temperate, C=cryophilic,

D=dry, W=wet.

Climate Model AIC Habitat Model AIC ER 122.1245 ER 138.3012 ARD 118.3394 ARD 139.7170 T->C 147.6368 D->W 172.1938 C->T 130.7445 W->D 176.2227

214

Fig. 19. Reconstructed ecology of Stellaria. Temperateness is depicted on the reconstruction on the left and habitat is depicted on the right. Habitat has undergone more evolutionary fluidity in

Stellaria than temperateness, but both were evolutionarily labile and underwent numerous transitions across the history of Stellaria. The ARD model was better supported and is plotted for cryophily, whereas the ER model was better supported and is plotted for habitat. Tests for correlated evolution supported independent evolution of climate occupancy and habitat as follows: independent vs interdependent (p=0.41), independent vs climate dependent (p=0.21), independent vs habitat dependent (p=0.56). Topology is identical to that in Figure 18. Cyan: cryophily and wet habitats; Red: temperate and dry habitats. Branch numbers on the cryophily phylogeny indicate clade names referenced in the text: 1, Insignes; 2, Petiolares; 3, Plettkeae; 4,

Nitentes; 5, Larbreae. 215

216 climatic occupancy (Pagel’s λ= 6.717495e-05; p=1), with numerous transitions between habitats reconstructed (Fig. 19). ARD and ER models were not significantly different for habitat

(p=0.444). As ancestral node state probabilities for habitat were ambiguous across the backbone of Stellaria, transitions based on these probabilities were difficult to quantify. However, stochastic mapping recovered 88.432 changes between wet and dry states, with 46.828 transitions from dry to wet and 41.604 transitions from wet to dry recovered on average. Lastly, our two reconstructed climate and habitat states in Stellaria did not undergo correlated evolution and the trait independent model was most highly supported (Fig. 19).

DISCUSSION

Origins and the New World

The origin of Stellaria was reconstructed ambiguously lying in northern Eurasia or the

New World (Fig. 18). Starworts likely originated in a temperate context during the middle to late

Miocene (ca. 12 Mya; Figs. 16 & 19), a time several degrees warmer than present and before the formation of perennial northern hemisphere ice (Zachos et al., 2001; Favre et al., 2015).

Temperate forests covered this broad possible origin area of Stellaria prior to the Pliocene

(Murray, 1995; McIver and Basinger, 1999; Wen et al., 2010), which agrees with our reconstruction of a temperate ancestral state during the Miocene in this genus (Fig. 19). Colder lineages then began to diverge in Stellaria as Northern Hemisphere ice sheets started to form, especially from the Pliocene onwards (Figs. 17 & 19). The earliest-diverging lineage in Stellaria

(the Insignes clade) is disjunct in the northern hemisphere between eastern Asia and eastern

North America (see also below). The majority of such disjunctions are thought to have originated in eastern Asia and utilized the Miocene Bering Land Bridge to migrate into the New World, and 217 the minority of scenarios inferring directionality from the New World into the Old World via

Beringia is biased towards conifers (Wen et al., 2010). As such, we propose a potential origin of the genus Stellaria in northeastern Eurasia. Our ancestral area reconstructions, though ambiguous, nonetheless demonstrate that neither Pax and Hoffmann (1934) nor Croizat (1952) rightly hypothesized the origin area of Stellaria. Under this scenario of northern origin, one lineage of starworts (part of the Petiolares clade and including S. aquatica and S. nemorum) remained and diversified in northern Eurasia, while the other lineage of Petiolares (the broad S. cuspidata group) diversified in the Americas (Fig. 18). The former group then gave rise to the S. sennii lineage that migrated to Africa to become the sole, albeit widespread, native starwort lineage there.

After divergence of the Petiolares clade, the common ancestor of the Plettkeae and

Nitentes clades in addition to S. alaskana also diversified between northern Eurasia and the New

World, with one lineage migrating into and radiating in the Andes as cushion plants (Plettkeae), and the others spreading mostly across western North America (Fig. 18). Most other Stellaria clades then remained and diversified in the Old World. However, from an Old World origin,

Stellaria longifolia then reached western and boreal North America, and the Stellaria longipes group also sent derivates into western and boreal North America (Fig. 18). Both S. crassifolia and S. irrigua (or the MRCA of the two) also moved into the New World (see below). Stellaria irrigua reached its limit in southern North America and spawned S. sanjuanensis (Sharples and

Tripp, 2019b), and S. crassifolia also reached a similar limit and likely gave rise to S. debilis of southern South America (Fig. 18). Altogether, at least 13 movements into the New World are inferred from a possible origin in northern temperate Asia, and the Bering Land Bridges were available for all of them (except for one dispersal in the southern hemisphere: see below). 218

Several of these movements are inferred from the very recent Pleistocene and in groups that are still present in Beringia (see below). Results in Stellaria thus accord with the hypothesis that circumboreal taxa originated in recent, Pleistocene times, a result also suggested in Cerastium

(Scheen et al., 2004; see below). More, and recent, movements within species will likely be detected once a phylogeographic approach is taken towards inferring the history of these groups.

The connection between South America and North America via establishment of the

Panamanian landbridge was secured by three million years ago (O’Dea et al., 2016), which greatly facilitated dispersal of terrestrial organisms between the two continents (Cody et al.,

2010). The common ancestor of the circumboreal S. crassifolia and South American S. debilis was dated after the completion of the Isthmus of Panama (0.82 Mya; Fig. 17) and thus this particular dispersal to South America may have utilized this landbridge. However, both plants and animals had been dispersing between both continents for much longer than the formation of the Isthmus and such dispersals have been dated to 20 Mya and older (Cody et al., 2010; O’Dea et al., 2016). Some members of Cerastium in South America have also derived from a North

American origin, but it is unclear when this dispersal occurred (Scheen et al., 2004). This north to south migration pattern likely explains the presence of the Plettkeae clade of Stellaria in the

Andes, but the migration of this lineage into South America was dated far before the formation of the Isthmus of Panama (ca. 8 Mya; Fig. 17), indicating trans-oceanic dispersal. Due to poor bootstrap support along the backbone of the S. cuspidata group of species, it is difficult to infer whether this group originated in North or South America from a Eurasian origin, although the former is more parsimonious. Nonetheless, the Darien Gap disjunction in the S. cuspidata complex was dated in our study at approximately 2 Mya (Figs. 17–18), and in total these results 219 suggest that migration of Stellaria between North and South America occurred both prior to and after the formation of the Panamanian land bridge.

Pan-Himalaya

Some widespread angiosperm lineages with significant standing diversity in the Pan-

Himalayan mountains originated there. Lagotis (Plantaginaceae) likely originated there, with subsequent dispersal throughout parts of Eurasia and into North America (Li et al., 2014). The deepest splits in the Gentiana (Gentianceae) phylogeny arose in the Pan-Himalaya center of diversity, with numerous subsequent dispersals across the world occurring in parallel (Favre et al., 2016). A large radiation of Rhodiola (Crassulaceae) also originated in the Pan-Himalaya and dispersed repeatedly into Europe and North America, achieving a widespread northern temperate distribution with a distinct center of diversity in the mountains of southern and eastern Asia

(Zhang et al., 2014; György et al., 2018). Tkach et al. (2014) found multiple origins of circumboreal or Arctic members of Pedicularis (Orobanchaceae) from the Pan-Himalaya, an area that also represents the center of diversity of Pedicularis worldwide. The diverse circumboreal Saxifraga hirculus group (Saxifragaceae) also spawned from a Pan-Himalayan origin center (Hedberg, 1992; Ebersbach et al., 2017), and numerous taxa of the European Alps are hypothesized to derive from Asian or Pan-Himalayan founding stocks (Kadereit et al., 2008).

Several other investigated lineages with circumboreal and Pan-Himalayan diversity have also moved northwards out of the southern Asian mountains (e.g., Koenigia, Polygonaceae: Fan et al.,

2013; Pyrola, Ericaceae: Liu et al., 2014).

In contrast, other taxa with significant standing diversity in the Pan-Himalaya derived that diversity from ancestral origins elsewhere. In Cassiope (Ericaceae), a northern origin of the genus gave rise to diversity in the southern Asian mountains (Hou et al., 2016a), and Diapensia 220

(Diapensiaceae) also diversified in the Pan-Himalaya from an inferred northern origin (Hou et al., 2016b). Scrophularia (Scrophulariaceae) originated and radiated in western Asia, subsequently moving into the Himalaya, Europe, and into the New World via Beringia

(Scheunert and Heubl, 2017), and a similar pattern has been suggested for Allium (Hauenschild et al., 2017). Saxifraga as a whole originated in North America before colonizing and diversifying in the Pan-Himalaya on several different occasions (Ebersbach et al., 2017). This pattern of secondary rather than primary diversification in the Pan-Himalaya has also been recovered in other groups such as Caragana (Fabaceae: Zhang et al., 2015) and Maianthemum

(Asparagaceae: Meng et al., 2008).

Aligning with the immigration rather than the origin scenario, the Pan-Himalaya was recovered as a major secondary center of diversity in Stellaria (Fig. 18). This was counter to our original favored hypothesis due to the great morphological variation to be found in this region, which was predicted to reflect multiple diversification events in the region (or one ancestral diversification with subsequent movements elsewhere, as in Gentiana). This morphological variation led us to hypothesize that a secondary burst of evolutionary diversification was unlikely there, as such a secondary burst may be predicted to be united by more morphological cohesiveness than what is in fact found there; the span of morphological diversity in the Pan-

Himalaya encompasses much of the breadth of morphological diversity to be found across the remainder of global starworts outside of that region, as with the species of Oceania (Fig. 15).

Specifically, taxa with, e.g., ground-scandence, petiolate or sessile leaves, petal presence or absence, and alpine cushion growth can be found both within the Pan-Himalaya and in starwort taxa occurring outside of that region. 221

Our findings suggest that the most outstanding altitudinal gradient on Earth has been host to the greatest amount of species diversity in Stellaria, here shown to have likely originated rapidly within the Pleistocene. These results are consistent with the hypothesis that some angiosperm lineages of the Pan-Himalaya colonized the area (especially from more northerly areas) prior to or during the onset of the Pleistocene (Sun, 2002; reviewed in Wen et al., 2014;

Hou et al., 2016b; Hauenschild et al., 2017). Of interest, in Stellaria it appears that a cryophilic ancestor may have been prerequisite for initial colonization of and diversification within the

Himalaya (Fig. 19). However, several shifts back to temperateness then occurred, and even united some small radiations. Late Miocene/Pliocene uplift phases of the Himalaya gave rise to alpine elevations above 5,000 meters and provided an even greater diversity of ecologies to inhabit (Zhisheng et al., 2001; Favre et al., 2015), and this elevational gradient is often invoked to help explain rapid diversifications. However, high elevations in the region were present long before Stellaria arrived there (Favre et al., 2015; Renner, 2016). Nonetheless, the most species- rich radiation of Stellaria occurred in the Pan-Himalaya and has given rise to nearly one third of extant species in the genus during the Pleistocene (Fig. 17). Other rapid diversifications in high mountain systems likely occurred in very recent times, i.e., within the past few million years

(reviewed in Hughes and Atchison 2015; Xing and Ree, 2017), and this pattern may prove common across flowering plant evolution. Rapid radiation in the Caryophyllaceae has previously been demonstrated in Dianthus of the Mediterranean Basin (Valente et al., 2010), and diversification rate analysis with a full starwort phylogeny may one day reveal similarly high diversification rates in Stellaria, as well as in other large and diverse caryophyllaceous lineages such as Silene (Wen et al., 2014). The massive Pan-Himalayan mountain chains remain underexplored botanically and may also harbor yet undescribed Stellaria diversity. 222

Maritime Asia

In Maritime Asia we uncovered a previously unrecognized disjunction in the clade sister to the rest of core Stellaria: the Insignes group of two starworts of far eastern Asia and two of eastern North America (Fig. 18). The disjunct distribution of Insignes in these two regions accords with a broad pattern of temperate Tertiary disjunctions between them, as shown in dozens of other flowering plant lineages (e.g., Li, 1952; Wen et al., 2010; Manos and Meireles,

2015). Specifically, the late Miocene to early Pliocene date recovered here for the disjunction in

Stellaria aligns with dates attributed to numerous other angiosperm disjunctions between the two regions (Wen et al., 2010; Li et al., 2017). This finding also aligns with the observed pattern that

Miocene and younger disjunctions between these two areas are associated with temperate taxa, and that older disjunctions inferred between these two areas are associated with tropical or subtropical taxa (Wen et al., 2010). The younger disjunctions are often explained by vicariance of aforementioned once more continuous temperate northern forests spanning the area between present day temperate forests of eastern Asia and eastern North America and across Beringia

(Murray, 1995). This Insignes clade diverged from the rest of Stellaria over 11 Mya and currently only contains four species, suggesting a long period of evolutionary stasis in this clade.

Conversely, much extinction may have occurred in the history of Insignes.

Elsewhere in Maritime Asia, the Stellaria tomentosa-S. uchiyamana group colonized

Japan from the Pan-Himalaya (Fig. 18), and other widespread or circumboreal species have breached the region from inferred origins elsewhere (see below). The clade containing largely western North American species such as S. crispa, S. littoralis, and S. sitchana originated in

Maritime Asia before diversifying more broadly in North America and elsewhere (Fig. 18). The

S. longifolia complex also likely generated in Maritime Asia (Fig. 18). Thus the northeast of Asia 223 hosted a dynamic role in the evolutionary history of Stellaria, serving as both a source and a sink area in independent lineages.

Oceania

In contrast with the broad Pan-Himalayan clade, which was characterized by short branch lengths and recalcitrance to highly-supported backbone resolution, relationships within the

Oceania clade of Stellaria were well-supported and appear to have undergone comparatively long periods of non-reticulate evolution as observed by the longer branches recovered between tips and most nodes (Fig. 17). The great morphological diversity in Oceania thus represents divergence between close relatives rather than multiple independent colonization events. As stated, the morphological diversity here spans a breadth of forms to be found across Stellaria worldwide, with features such as succulence, indefinite ground-scandence, apetaly, and petioles found both in Oceania and elsewhere in the world (Fig. 15; see also Pan-Himalaya). Non- succulence, definite growth, petaly, and sessile leaves are all the more common conditions across starworts worldwide. Sister to the Oceanic clade was S. alsine, a somewhat weedy, “drab”, and widespread species that lacks succulence, lacks ground-scandence, usually bears petals, and has sessile leaves, and due to the above factors we are currently unable to infer from where dispersal to Oceania occurred. It was possibly from Malesia, however, given its proximity to Australia and the contemporary presence of S. alsine there, although this occurrence is at high elevation and thought to represent recent introduction (Larsen, 2002). Detailed phylogeographic work is necessary to better potentially recognize if and from where S. alsine may have dispersed to

Oceania, and indeed to better delimit the taxonomy of S. alsine-like species (Sharples, 2019).

We found evidence of only one dispersal from Oceania back to an extra-Oceanic landmass after the initial founder event, inferred in the amphi-Pacific species S. parviflora, 224 which was not recovered as monophyletic here (Fig. 18). The relatively young age of Stellaria– and especially the young age of all of the Oceania taxa—inferred in our analyses combined with the lack of relatedness between such groups both reject the possibility that Stellaria taxa in

Oceania and southern South America represent elements of an ancient Gondwanan flora. As it is unclear if Australia and New Zealand were colonized separately, either one or two initial founder events to Oceania are inferred, and if one, then at least two additional dispersal events between

Australia and New Zealand must have occurred. However, since relationships recovered here within Oceania differ from previous work (Sharples and Tripp, 2019a), it remains unclear exactly which taxa there comprise which clades, thereby obscuring the number of dispersal events between Australia and New Zealand that may have occurred. Nonetheless, migration from

Oceania to South America occurred after Oceania was colonized and this represents one of the longest inferred dispersal events in the history of flowering plants.

Circumboreal and Arctic Taxa

Cerastium has been implicated in at least two movements into the New World from the

Old World via Beringia (Scheen et al., 2004). As mentioned, Stellaria has also colonized the

New World multiple times from the Old World (Fig. 18), and Beringia is implicated in these starwort movements. The Bering Land Bridge last closed around 10,000 years ago (Elias et al.,

1996) and it has been available as a corridor for migration throughout much of the evolutionary history of Stellaria (Fig. 17; Wen et al., 2016). Given the ability of Stellaria to disperse across short (and long) stretches of ocean (see also below), the current Bering Strait may still serve as a gateway of sorts between Asia and North America. Directionality of initial dispersal from eastern

Asia into North America (rather than vice versa) is supported in various temperate angiosperm lineages exhibiting Miocene eastern Asian-eastern North American disjunctions (Wen et al., 225

2016) as well as in numerous other lineages present on both sides of Beringia (Murray, 1995). A general bias in directionality from the Old World to the New World was also supported in

Stellaria (Fig. 18). However, dispersal back to the Old World, though infrequent, is not prohibited and has occurred in the circumpolar S. humifusa as well as in the S. calycantha group

(Fig. 18). These occurred in the extremely recent past and Beringia is thereby also implicated in their movements.

The circumboreal and arctic starworts have arisen multiple times in unrelated clades rather than reflecting in situ radiation within the Arctic, a pattern that explains species diversity in the Arctic in multiple unrelated lineages (Hoffmann and Röser, 2009). Taxa with circumboreal distributions are thought to be of Pliocene or younger age, as establishment of the biomes associated with their distributions is a recent phenomenon (Abbott and Brochmann, 2003; Liu et al., 2014). The multiple circumboreal groups with Arctic presence in Stellaria were all dated as diverging with their closest relatives during the Pleistocene, in accord with the hypothesis of recent ages of circumboreal taxa. These lineages were inferred from several different origin points. Circumboreal Stellaria crassifolia is deeply embedded within the large Pan-Himalayan clade and appears to have originated there, as is S. irrigua, which is broadly distributed across cold habitats of eastern Asia and western North America, including in Beringia (Fig. 18;

Sharples and Tripp, 2019b). Stellaria longifolia likely originated in Maritime Asia before spreading widely, and the Stellaria longipes group likely originated in central Asia before colonizing cold areas across the northern hemisphere (Fig. 18). In contrast, Stellaria humifusa of

Arctic and other cool coastlines appears to have spawned in western North America before spreading back into the Old World, and the S. borealis group appears to have also originated in

North America before migrating into the Old World (Fig. 18). Multiple, parallel, unrelated 226 colonizations of the Arctic in Stellaria accord with findings in other genera such as Artemisia

(Tkach et al., 2008), Pedicularis (Tkach et al., 2014), Pyrola (Liu et al., 2014), and Ranunculus

(Hoffmann et al., 2010). These colonizations are generally inferred as expansions of southern species into northern areas, although in the absence of complete taxon sampling such inferences may have limited power at present. In Stellaria, though, recent origins of circumboreal taxa from southern mountains are implicated in S. crassifolia, S. irrigua, S. longipes, and possibly also in the S. borealis group and S. longifolia, all taxa that currently inhabit both mountains and lowlands and which were all reconstructed with cryophilic ancestors (see below).

Observations on Endemism

Some areas of Stellaria occurrence lack species endemism while others are replete with endemism. Most notable of the former is Europe, which harbors only one possible endemic (S. crassipes) despite being moderately diverse (n=11 stars). The lack of endemism across Europe is striking, particularly in context of its many cold and isolated mountain ranges such as the Alps,

Balkans, and Pyrenees, which harbor endemic taxa in other diverse caryophyllaceous lineages

(e.g., Arenaria, Cerastium, Dianthus, Minuartia, Paronychia, and Silene: Walters, 1964). In contrast, endemism within Australia and New Zealand is extremely high; all taxa of Australia are endemic there, and only one species native to New Zealand occurs elsewhere (Patagonia) (Fig.

14). The high level of endemism in Oceania is likely attributable to its isolation, as it is the only area harboring high endemism in Stellaria that is unconnected with other landmasses. Likewise, endemism in South America and Africa is also high, though less striking than in Oceania due to a cohesive similarity of morphological forms in the former and only a single widespread species occurring in the later. In the Pan-Himalaya, two-thirds of the starworts occurring there are endemic to the region (Fig. 14), and other more widespread species occurring there have 227 distributions largely restricted to neighboring areas. North America and Maritime Asia exhibit a mixed pattern, in which a little over half of the starworts are endemic to those areas while the remaining percentage of taxa occur elsewhere (Fig. 14).

Narrow restriction to one region is especially marked in some Stellaria species. Stellaria sanjuanensis is narrowly endemic to loose alpine substrata in volcanic portions of the Southern

Rocky Mountains (Sharples and Tripp, 2019b), and Stellaria roughii is restricted to rocky mountain slopes of the South Island of New Zealand (Garnock-Jones, 1988). Stellaria eschscholtziana is similarly restricted to rocky volcanic areas of greater Kamchatka (Schischkin,

1936; Pavlova, 1996), and just across Beringia S. alaskana is found only in a limited portion of

Alaska and the Yukon (Morton, 2005). Stellaria leptoclada of Australia is essentially restricted to the Great Dividing Range of New South Wales (Miller and West, 2012), and S. nipponica is restricted to a few alpine areas in Japan (Zoku, 1965). In total, well over half of all Stellaria species only occur in one area as defined here, and numerous others are narrowly limited to small areas within those confines. Factors contributing to narrow and widespread distributions across Stellaria should be investigated to provide further insight into its evolutionary history.

Habitat and Climate

The ancestral state of Stellaria was ambiguously reconstructed for the two aspects of ecology explored in this study, though there was slightly higher than 50% support that the ancestral starwort inhabited temperate environments (Fig. 19). On a broad phylogenetic and geographic scale, ecological requirements of flowering plants are largely evolutionarily conserved across myriad lineages (Crisp et al., 2009; de Casas et al., 2016). On a finer scale, e.g. in the Burseraceae and Proteaceae, the predominance of ecological conservatism has also been demonstrated (Weeks et al., 2014; Cardillo et al., 2017), although exceptions exist 228

(Cucurbitaceae, Holstein and Renner, 2011; Amaryllidaceae, Jara-Arancio et al., 2014) and patterns of both conservatism and high amounts of shifts have been demonstrated in the

Caesalpinoideae (Fabaceae, Souza-Neto et al., 2016). Stellaria globally inhabits almost every major biome type besides lowland tropical forests and many types of habitats in between (Fig.

13). Although we reconstructed binary states of just two macroecological conditions in Stellaria, ecological conservatism in the genus appears to be the exception rather than the rule. First,

Stellaria exhibits rampant transitions between wet and dry habitats within and across clades and such habitat occupancy exhibited no phylogenetic signal (Fig. 19). This is in contrast with

Pedicularis, within which soil moisture requirements are generally conserved within clades

(Tkach et al., 2014), but is in alignment with Ranunculus within which such transitions are likely common (Hoffmann et al., 2010) as well as with Anacardiaceae (Weeks et al., 2014). Second, transitions between temperate and cryophilic states were also evolutionarily labile in Stellaria, though they were more conserved than transitions between wet and dry habitats and did exhibit phylogenetic signal (Fig. 19). Cryophily was supported as the ancestral state for aforementioned lineages that occur in the Arctic or that are representative of cold mountains (Fig. 19). As these lineages did not evolve in situ in the Arctic, this finding supports the idea that cryophily may have been prerequisite for colonization of the far north or high elevations, hypothesized also in

Pyrola (Liu et al., 2014), Ranunculus (Hoffmann et al., 2010), and others (Hoffmann and Röser,

2009). Our data reject the possibility that occupancy of either wet or dry habitats was prerequisite for reaching cold places. Climate and habitat occupancy were not correlated (Fig.

19), and the ecological lability of lineages across Stellaria as well as the lack of dependent evolution between these two ecological factors has surely contributed to its widespread worldwide success. 229

Certain Stellaria clades nonetheless exhibit some degree of ecological conservatism. The aforementioned Insignes clade is united by temperateness. The circumboreal Stellaria longipes group and relatives is united by cryophily but exhibits less conservatism in terms of wet or dry habitat preference (Fig. 19). The Stellaria cuspidata complex, ranging across the length of South

America and southern North America, was instead largely temperate throughout, giving rise to cryophily only in the S. weddellii lineage of high elevations of Bolivia, but likewise exhibited less conservatism in terms of wet or dry habitat preference (Fig. 19). Both of these clades involve morphologically similar multiploid taxa and species boundaries in each are still unclear, suggesting that clades of similar species in Stellaria may exhibit more conserved ecologies than better-defined species in clades of morphologically dissimilar species. Indeed, these species complexes may be driving our finding of phylogenetic signal in the climate dataset.

Star Dispersal

Few flowering plant genera have achieved such a widespread native distribution as

Stellaria, a distribution broader even than Ranunculus (buttercups), Senecio (groundsels), and

Vaccinium (blueberries). While the latter two are clearly dispersal by wind (via a pappus) or herbivores (through a fleshy fruit), seed dispersal ability in Stellaria is unassuming yet has been underappreciated. Despite an apparent lack of adaptation for such purposes, our results support rampant LDD within Stellaria, sometimes across sea barriers. As with related genera and nearly all members of Caryophyllaceae, Stellaria has dry fruits at maturity, capsules that simply release seeds through the agency of ovary splitting. The seeds are not specially equipped for further movement (e.g., to further float on the wind), lacking an aerodynamic shape or comose adaptations (Mahdavi et al., 2012). Other members of Caryophyllaceae, these too lacking apparent adaptations for frugivory or wind-mediated seed dispersal, have been supported under 230 scenarios of LDD dispersal (Hagen et al., 2001; Westergaard et al., 2011; Jia et al., 2016). The ability of a plant lineage to disperse widely must be considered under a wider range of factors than traditional ones such as presence of fleshy fruit, presence of comose seeds, presence of pappus, presence of dust-like seeds, and so on.

In some instances, water-based dispersal may be an important agent helping to shape

Stellaria distributions. For example, S. humifusa preferentially occupies far northern coastlines, and individuals of this species are regularly submerged in saltwater (Fig. 13). Thus its seeds are regularly subjected to long periods of immersion in sea water before germination, and it may be expected that these seeds at times travel great distances in sea water, helping to explain its extensive northern-coastal distribution. Many starworts inhabit arctic and alpine environments, and in the arctic, wind is thought to transport seeds across vast oceanic distances over the winter sea ice; in both arctic and alpine contexts, wind transport across glaciers and snow is also thought to be common (Brochmann and Steen, 1999). Thus starworts sympatric with S. humifusa

(e.g., S. alsine, S. calycantha, S. crassifolia) can also potentially be blown across vast portions of sea ice during the winter, facilitating establishment of widespread circumboreal distributions.

Most species of Stellaria are not coastal and therefore lack the latent dispersal ability these northern coastal taxa have. However, it is clear that numerous starworts are tightly associated with freshwater (Fig. 19) and may disperse down elevational gradients by traveling along water corridors.

Although dispersal investigations in Stellaria and Caryophyllaceae in general are lacking, some observations relate to the cosmopolitan S. media, and extrapolations from that species and a few others can be hypothesized upon the whole genus. Wind, and in particular strong events such as whirlwinds and cyclones, can transport entire individuals of S. media, and ants have been 231 shown to carry around its seeds (Ridley, 1930). Indeed, it has been noted that strong winds can move metal objects and even horses (Brochmann and Steen, 1999). Stellaria media and S. aquatica have been reported as pseudo-epiphytes, i.e., growing in small patches of soil in canopies of temperate willows and other trees, and they have been hypothesized as arriving there via wind, via nest-building activities of birds, and/or via ant transportation (Ridley, 1930).

Unlike S. humifusa, S. media does not preferentially inhabit seashores (but rather, can live almost anywhere, including seashores), yet it can germinate after immersion in sea water up to 90 days and its seedlings can float (Ridley, 1930).

Stellaria is surely widely dispersed through the agency of herbivory despite its dry fruits, i.e., its capsules rather than berries. Pigs, horses, cattle, deer, magpies, wide-ranging desert grouse, both common and australasian quail, and sparrows have been found to disperse seeds of

S. media and others through herbivory (Ridley, 1930), and we might expect that the whole genus is well-dispersed through these and numerous other vertebrate agents (and also in the context of secondary dispersal). Circumboreal Stellaria longifolia has been reported as preferred forage of reindeer (Ridley, 1930). Other agents of starwort dispersal have been proposed, such as in mud with seeds adhering to seabirds to explain occurrence of the S. cuspidata complex and other starworts on various islands (Ridley, 1930). Lastly, numerous species of Stellaria are characterized by a habit in which they sprawl or climb along the ground and other plants indefinitely. In this manner they likely commonly travel along with other vegetation such as that attached to large animals (e.g., with Galium spp.) or that moving along with flash flooding and other catastrophic events. Other than dispersal agents in Stellaria, its dynamic reproductive biology with numerous parallel inferred instances of predominant self-fertilization systems worldwide must have also been critical in establishing in new areas repeatedly (Sharples et al., in 232 prep.). Equivocal ancestral states of outcrossing for both the Oceania and the Pan-Himalaya clades may also suggest that a lability between outcrossing and selfing in Stellaria has facilitated its colonization and diversification in these areas in particular. These agents in total likely help explain much of the natural cosmopolitan distribution of Stellaria.

Future Directions

Our study uses phylogenomic data applied to genus-wide historical reconstruction in the

Carnation family and will be a useful comparative baseline for future work in other clades of

Caryophyllaceae. RADseq data yielded divergence time estimates within Caryophyllaceae that were congruent with earlier work based on fewer nucleotides. Other lineages of Caryophyllaceae are diverse across the distribution of Stellaria, and future work will be of interest for comparison with the patterns recovered in our study, such as in Silene, which also exemplifies a center of diversity in the Pan-Himalaya.

233

CHAPTER 4

INTRODUCTION

Conspicuous petals, showy tepals, petaloid floral whorls, or colorful bracts function in attracting diverse groups of animal pollinators (Faegri and van der Pijl, 1979; Fenster et al.,

2004; Friedman and Barrett, 2009). Marked petal variation within genera, including diverse forms, colors, sizes, and degrees of fusion, have evolved in response to different pollinator functional groups (e.g., Ruellia, Tripp and Manos, 2008; Lobelioideae, Lagomarsino et al., 2017;

Brassica rapa [experimental], Gervasi and Schiestl, 2017; Gesneria and Rhytidophyllum, Joly et al., 2018; Penstemon, Wessinger et al., 2019). Showy, attractive tepals or petals are thought to have been present in the ancestral angiosperm flower (Sauquet et al., 2017), and early angiosperms have further been reconstructed as insect-pollinated (Hu et al., 2007). However, evolutionary loss of petals is rampant and characterizes many unrelated lineages in the

Alismatales, Cucurbitales, Fagales, Myrtales, Poales, Ranunculales, Rosales, and Saxifragales, in addition to several others (Linder and Rudall, 2005; Endress and Matthews, 2006; Zhang et al.,

2013; Du and Wang, 2014).

Across angiosperms, abiotic pollination such as via wind or water is often invoked to explain the evolution of apetaly (Culley et al., 2002; Friedman and Barrett, 2009; Du and Wang,

2014). Wind-pollinated lineages are often characterized by additional features such as unisexual flowers, densely-flowered inflorescences, few ovules per ovary, and exserted stamens, whereas 234 water-pollinated lineages are further characterized by unisexual flowers (Culley et al., 2002;

Friedman and Barrett, 2009; Soza et al., 2012; Du and Wang, 2014). Instances of the evolution of apetaly that do not involve shifts to abiotic pollination (from animal-pollinated ancestors) often involve co-option of other floral whorls, such as stamens or sepals, for pollinator attraction

(e.g., Ranunculaceae, Zhang et al., 2013). A fourth scenario to invoke the evolution of apetaly involves shifts to predominantly self-pollinating systems, whose potential correlates involve reductions of all floral whorls in addition to other changes that help facilitate transfer of pollen to stigma(s) of the same flower (Ornduff, 1969; Wyatt, 1988).

The flowering plant family Caryophyllaceae is marked by numerous evolutionary transitions between petaly and apetaly, both within and across myriad genera (Greenberg and

Donoghue, 2011). One lineage with a high representation of apetalous species is the cosmopolitan genus Stellaria L. (colloquially, the “starworts”). Stellaria includes more than 110 species and is found across a marked diversity of biomes and habitats (Sharples and Tripp,

2019a; Sharples and Tripp, in prep.). Although most species of Stellaria bear showy petals, approximately one-fourth of the genus is characterized by petal reductions, ranging from fully absent to partial petal development to petal plasticity in some instances. These petal reductions are not unique to any one geographical region or biome. Within Stellaria, petaly has been reconstructed as ancestral (Greenberg and Donoghue, 2011), but infrageneric sampling in that study was limited to only one-third of taxa.

Whereas other angiosperms manifest syndromes (e.g., wind or water pollination) that correlate with apetaly, such suites of characters do not readily apply to apetalous Stellaria (M.

Sharples, pers. obs.). Stellaria lacks, e.g., unisexual flowers, densely-flowered inflorescences, few ovules per ovary, high stature, reflexed sepals, and well-exserted stamens that are 235 characteristic of wind-pollinated systems, and the lack of unisexual flowers also suggests against water-pollination. Likewise, expected floral differences between outcrossing progenitors and self-pollinating derivatives (Ornduff, 1969; Wyatt, 1988) do not seem to hold in Stellaria (M.

Sharples, pers. obs.); for example, apetalous species do not necessarily bear fewer flowers per inflorescence, different flower colors, shorter pedicels, higher ploidies, or floral scent differences than in petalous species. These failures to conform to better known syndromes yield uncertainties regarding ecological and evolutionary drivers of petal reductions in Stellaria.

To our knowledge, empirical information is lacking in regards to reproductive biology of apetalous Caryophyllaceae outside of Schiedea, the only known wind-pollinated member of the family (Weller et al., 1998). The lack of work is striking, as apetaly is both rampant and reconstructed as ancestral within Caryophyllaceae (Greenberg and Donoghue, 2011). The range of petal variation present in Stellaria makes this system optimal in which to explore reiterative petal gains and losses, along with the relationship of such changes to reproductive strategies. In this study, we (1) describe the pattern of petal evolution in Stellaria using a comprehensive species-level sampling of the lineage, and (2) compare empirically seed sets resulting from pollinator exclusion and outcrossing treatments in an arctic-alpine population of two sympatric starwort species, one apetalous and the other petalous. Given self-pollination rates are high in arctic-alpine angiosperms compared to other ecosystems, we predicted that both the petalous and apetalous species in our experimental study would exhibit at least some degree of self- fertilization. Our results provide a descriptive baseline of petal evolution across Stellaria as well as yield insight into how apetalous starworts are primarily reproducing.

MATERIALS AND METHODS 236

Pollination Biology

Within Stellaria, the showy petalous starworts have features suggestive of biotic pollination systems in producing large white petals and nectar discs likely attractive to pollinators. It has been demonstrated that Stellaria pubera Michx. of the eastern United States— which bears white, unfused, rotate corollas exceeding the sepals—is generalist-pollinated by bee- flies (Bombylius), flies (Hylemya, Tachinidae), hoverflies (Syrphidae), cuckoo bees (Nomada), mining bees (Andrena), halictid and Osmia bees, skipper, marble, and lycaenid butterflies, among other animal species (Campbell, 1985; Motten, 1986). Stellaria longipes Goldie, also showy-flowered, is likewise a generalist-pollinated species visited by mosquitoes (Aedes spp.), flies (Brachicoma sp., Bufolucilia silvarum, Dolichopus dasyops, Delia liturata, Lucilia illustris,

Phaenicia sericata, and Pollenia rudis), mining bees (Andrena sp.), wasps (Cryptus arcticus), bumblebees (Megabombus polaris and Pyrobombus ternarius), and hoverflies (Syritta pipiens)

(Chinnappa and Morton, 1984). The cosmopolitan Stellaria media (L.) Villars, which often bears petals subequaling the sepals in length, is also visited by several insect lineages (Sobey, 1981). In contrast to the above, no published empirical study exists regarding pollination biology of any apetalous starwort, to our knowledge. These observations served as bases for preparation of our trait matrix and subsequent analyses as well as motivation for experimental work (see below).

Field Methods

To infer reproductive biology of petalous and apetalous species of Stellaria, experimental field work was conducted at Niwot Ridge (Niwot Ridge Long Term Ecological Research site,

Colorado, U.S.A.) with two sympatric starworts: Stellaria irrigua Bunge and Stellaria longipes

(Fig. 20). Both species produce bisexual flowers, but the latter is predominantly outcrossing with 237

Fig. 20. Study area hosting two sympatric species of Stellaria on the tundra of Niwot Ridge Long

Term Ecological Research Site, University of Colorado field station. A. The petalous Stellaria longipes, which occurs on drier soils atop and along the visible embankment. B. The apetalous

Stellaria irrigua, which occurs in wet soils at the base of and along the embankment (below which is a perennial marsh). Pollinator exclusion bags are visible in the foreground of both panels.

238 showy petals (Philipp, 1980; Chinnappa and Morton, 1984) whereas the former has not been studied from a reproductive perspective and lacks petals. The Niwot Ridge site was optimal for field experimentation because species sympatry at the site reduces environmental variation and associated impacts on reproductive dynamics including output. The field site was located on treeless tundra just above the krummholz zone and is located at 40.06020, -105.58955 (WGS 84) at ca. 11,400 ft. (~3,473 m) above sea level. None of our study individuals were located within the limits of other, ongoing field experimental studies conducted at Niwot Ridge.

Control Dataset. A control dataset consisted of number of mature seeds (scaled by total ovule number: unfertilized ovules in our species remain attached to fallow parts of a given capsule) quantified from unmanipulated individuals of both species at Niwot Ridge. Care was taken to quantify only capsules that bore fully mature ovules but whose valves had not yet begun to dehisce, ensuring that all fertilized ovules were considered. This control dataset generated the expected seed set per ovary and represented baseline fertility of both species at Niwot Ridge in absence of experimental manipulation.

Experimental Treatments. Two experimental treatments consisting of selfing and outcrossing manipulations were advanced in order to test the relative contributions of selfed versus outcrossed seeds to total reproductive output of our two focal populations. In selfing treatments, inflorescences or individual flowers were excluded from pollinators using bags while in bud stage, i.e., prior to anthesis. Bags made of synthetic organza fabric were utilized over other options due to preliminary trials that revealed their superiority at Niwot Ridge (M.

Sharples, unpub. data). These bags were (1) resistant to degradation from monsoonal tundra conditions, (2) of a small gauge size that conferred breathability and sunlight access but yet (3) retained all seeds of dehisced capsules of both species. In selfing treatments, no hand-pollination 239 techniques were implemented due to the miniscule and fragile nature of flowering material, especially in the context of S. irrigua (flowers <0.5 cm in diameter; M. Sharples, unpub. data); as such, these treatments consisted only of pollinator exclusion. Because inflorescences of tundra populations of S. irrigua (these typically 3–6 flowered at Niwot Ridge) are similarly miniscule, we sometimes bagged entire inflorescence units in our selfing treatments rather than individual flowers. Individual flowers were typically bagged for S. longipes given they are larger in size and usually solitary. When quantifying seed set from selfing bags of S. irrigua, we scaled numbers of captured seeds by number of mature capsules.

Outcrossing treatments were conducted between open flowers of intraspecific individuals that were located at least 5 meters distant from one another. A mature flower was severed and carried to another flowering individual, where direct contact between anthers shedding pollen of the severed flower and fully unfurled, receptive stigmas of the receiving flower was made for a period spanning 30 seconds, then flowers were bagged. Crosses were conducted on only one flower per individual, and outcrossed flowers were excluded from further manipulation after treatment. To ensure that seed quantification from these bags resulted only from outcrossing treatment, all remaining flowers of the inflorescence were removed prior to experimental pollination and bagging.

Fertility Indices

Seed set was quantified using dissecting microscopes at the University of Colorado

Herbarium (COLO). Rather than reporting absolute seed count, we compared results using a fertility index, which accounts for variability in ovule number across individuals of both species and yields a proportion of fertilized ovules per fruit, as follows:

FI = # of developed seeds per ovary total # of ovules per ovary 240

We used T-tests implemented in R (R Core Team, 2019) to ascertain differences between the means of selfing versus outcrossing treatments for both species as well as to control data. We then converted fertility indices into a standard index for gauging self compatibility versus self incompatibility in flowering plants (i.e., the Index of Self Incompatibility, ISI: Lloyd, 1965;

Raduski et al., 2011; Grossenbacher et al., 2017). The ISI was calculated following

Grossenbacher et al. (2017):

ISI = 1 - % seed set from selfed fruits % seed set from outcrossed fruits

We interpreted values >0.8 as reflecting self incompatibility and values <0.2 as reflecting strong self compatibility (Raduski et al., 2011; Grossenbacher et al., 2017). Though ISI tends to be strongly bimodal in this manner, it is a continuous scale of plant breeding system classification.

Trait Matrix

Petal presence/absence was scored based on protologues, study of type material, extensive field observations of 45 Stellaria species in native habitats, regional descriptions, and local taxonomic treatments referenced in previous work (Sharples and Tripp, 2019a), as well as study of additional collections housed at the following herbaria: ALTB, CAS, CHR, COLO, E,

IBSC, K, KUN, LE, MO, NSK, NY, P, PE, RM, UC, VBGI, and VLA (herbarium codes follow

Thiers, continuously updated). Petal variation within Stellaria involves, if present, length of petals (i.e., if exceeding, equaling, or subequaling the sepal length, then “showy”; e.g., S. cuspidata, S. pubera), or if reduced, petals that are ca. one half or less than the length of the sepals (here, “reduced”; e.g., S. congestiflora, S. lanata, S. sikkimensis), or complete absence

(e.g., S. filiformis, S. multiflora). Finally, some species (e.g., S. irrigua, S. nitens) exhibit plasticity in petal development, with some individuals in a given population producing some 241 flowers with one or more minute petals while other individuals completely lack petals. Species bearing petals greater than half the length of the sepals were scored as petalous (i.e., showy to potential visitors). In addition to true apetaloids, species bearing petals

(2019), and trait codings are provided in Appendix 4.

Phylogenetic and Ancestral State Reconstruction

To describe the pattern of petal evolution in Stellaria, we utilized a comprehensive sampling of species diversity with previously published ddRADseq data (Sharples and Tripp,

2019a). Briefly, we utilized iPyRAD v.0.7.23 to construct a RAD locus alignment (Eaton, 2018) and inferred the Stellaria phylogeny using BEAST v.2.5.1 (Bouckaert et al., 2014) following the methods detailed in our biogeographical study of Stellaria (Sharples and Tripp, in prep.). The maximum clade credibility tree was rooted with Cerastium following Greenberg and Donoghue

(2011) and Sharples and Tripp (2019a; also references therein). Cerastium was coded as petalous given that nearly all members of the sister group of Stellaria to which it belongs bear showy petals. Our final tree was represented by 104 tips. Species sampled and GenBank accession numbers of specimens are found in Appendix 5.

To estimate ancestral state and lability of petal evolution in Stellaria, we reconstructed ancestral states of petal presence/absence using the packages phytools and ape in R v.3.5.2

(Revell, 2012; Paradis and Schliep, 2019; R Core Team, 2019). Ancestral states were 242 reconstructed using the “fitMk” function to infer discrete character evolution. Equal rates, all rates different, and irreversible models were tested, then AIC scores were compared to determine which model most likely explained petal evolution in starworts. Stochastic mapping under the best performing model was then conducted with the “make.simmap” function and posterior probabilities of 1000 replicates were plotted onto the maximum clade credibility tree. A transition between character states was inferred when an ancestral state was reconstructed with greater than 50% likelihood for one of our two characters and gave rise to the alternate tip state.

A transition was considered well supported if both posterior probabilities of branch support were greater than 80% and reconstructed ancestral state posterior probabilities were greater than 70%.

Phylogenetic signal of petaly versus apetaly was quantified with the “phylosig” function in phytools (Pagel, 1999; Revell, 2012).

RESULTS

Field Experiment

Control trials indicated that Stellaria longipes produces more ovules per ovary than

Stellaria irrigua (Table 12; Fig. 21) at Niwot Ridge but that S. irrigua is more fertile than S. longipes (p= 6.611e-16) (Table 12; Fig. 21). Field experimental results revealed stark differences between reproductive strategies of the two species. The high fertility of Stellaria irrigua is attributable to a great propensity for self-pollination (Table 12; Fig. 21). No significant difference in means was found between the control dataset and either treatment dataset for this species (Table 12; Fig. 21), supporting the hypothesis that self-pollination is predominant in the life history of S. irrigua. In contrast, Stellaria longipes was almost entirely infertile in pollinator exclusion trials (Table 12; Appendix 6; Figs. 21–22): only a single seed was observed across 91 243

Table 12. Field data collected for pollination biology of S. irrigua and S. longipes. Note that ISI is strongly bimodal as predicted from floral morphology. Abbreviations: FI, fertility index; ISI, index of self incompatibility; C, control; S, selfing; O, outcrossing; n, number of capsules. See methods for details.

Stellaria irrigua Stellaria longipes Total Ovules (5–) 12.7 (–22); n=218 (11–) 25.3 (–38); n=220 Control FI (0.2–) 0.678 (–1) [0.182]; n=85 (0.095–) 0.415 (–0.857) [0.186]; n=104 Selfing FI (0.176–) 0.707 (–1) [0.212] (0–) 0.0009 (–0.083) [0.0087]; n=91 Outcrossed FI (0.285–) 0.716 (–1) [0.213] (0.103–) 0.486 (–0.87) [0.235] C & S p=0.308; n=187 p=2.2e-16; n=195 C & O p=0.157; n=116 p=0.8215; n=129 S & O p=0.618; n=133 p=2.514e-10; n=116 ISI 0.0126 0.9815

244

Fig. 21. Fertility indices and ovule variation of S. irrigua and S. longipes from Niwot Ridge.

Stellaria longipes ovaries are denser with ovules but less fertile than S. irrigua ovaries: a greater percentage of ovules per ovary developed into mature seeds in the S. irrigua control dataset than in the S. longipes control dataset. Fertility of S. irrigua was statistically equal across control, selfing, and outcrossing datasets. In contrast, pollinator-excluded ovaries of S. longipes were essentially infertile, and outcrossing and control datasets both produced similar seed sets.

245

Fig. 22. Experimental manipulations including selfing and outcrossing treatments of Stellaria longipes and S. irrigua from the field. A. Capsule of S. longipes resulting from outcrossing treatment. Dark coloration of capsule and extension beyond the sepals signifies fruit maturation.

B. Pollinator-excluded flower (without maturing fruit) of S. longipes after normal maturation time interval, with fallowness indicated by tawny coloration of the sepals and lack of a visible, darkened ovary. C. View inside of a fertile, cross-pollinated capsule of S. longipes. Note larger ovules attached to the central column, which matured into seeds; smaller ovules were not fertilized. D. View underneath the tawny sepals of a pollinator-excluded flower of S. longipes.

Note absence of a darkened, elongated ovary. Unfertilized ovules are visible through fracture in ovary wall. E. Stellaria irrigua inflorescence within pollinator exclusion treatment bag. Bag gauge size was small enough to retain all seeds if capsules dehisced prior to data collection.

Circle depicts small cluster of mature seeds, which are much smaller than mature S. longipes seeds. F. Capsule of S. irrigua with one mature seed visible and several unfertilized ovules visible inside fractured capsule husk.

246 pollinator-excluded ovaries of this species (Table 12; Appendix 6; Fig. 21). Outcrossing treatments for S. longipes resulted in somewhat higher fertility of capsules compared to control data, but this difference was not significant (Table 12; Fig. 21). The differences between selfing and both the control and outcrossing S. longipes datasets were highly significant (Table 12; Fig.

21). After converting data to ISI values, the two species yielded the strongly bimodal distribution that was expected based on floral morphology wherein S. longipes valued close to 1 and S. irrigua valued close to 0 (Table 12), further demonstrating that S. longipes is strongly self- incompatible and that S. irrigua is strongly self-compatible. Self-fertilization of S. irrigua is likely assured in the bud (Fig. 23). Neither of our treatments—selfing and outcrossing—yielded ovule numbers outside of the range of ovule variation found in the control individuals (Appendix

6), thus our pollination bags likely had little to no deleterious effects on photosynthetic or reproductive output of these miniscule plants. No insects or evidence of other animals were observed within pollination bags after weeks of exposure, confirming that the bags were effective excluders of potential pollinators of either species.

Ancestral State Reconstruction

The all rates different model outperformed other models (p=0.0065 compared with the

ER model) (Table 13). Ancestral state reconstruction under this model indicated that a showy, petalous species was most likely ancestral in Stellaria, and the earliest diverging clade 1 was reconstructed with showy petals with high probability (Fig. 24; Supp. Fig. 3). Petals were then reduced, lost, or evolved plasticity in myriad clades across the rest of Stellaria (Fig. 24). Two losses occurred in the ancestrally petalous clade 2 (Fig. 24). Following divergence of clade 2, three small lineages in clades 3 and 4 also underwent petal loss or reduction independently (Fig.

24). Apetaly then evolved in multiple subclades of clade 5 across taxa worldwide with 247

Table 13. AIC weights of four ancestral state reconstruction models tested. The ARD model is bolded as the best performing model. P=petalous, A=apetalous.

Model AIC ER 130.6855 ARD 125.2854 P->A 143.1919 A->P 139.5136

248

Fig. 23. A & B. Bud pollination in Stellaria irrigua as observed upon dissection of fresh flowers prior to floral anthesis. Note pollen grains of dehiscent anthers adhering to nearby stigmas. C.

Evidence for pollen deposition onto stigmas within a single flower of S. longipes; no fruits matured following such pollen transfer events. Filled arrows indicate a stigma with pollen deposition, and unfilled arrows indicate an anther shedding pollen.

249

Fig. 24. Ancestral state reconstruction of petal evolution across Stellaria based on ca. 92% sampling of species diversity in the genus. Showy, petalous lineages (black) are reconstructed as the most likely ancestral state, with numerous subsequent losses or reductions of petals (red) across clades evident. Pie charts at each node depict the posterior probabilities that petaly or apetaly was reconstructed as the ancestral state for that node under the ARD model. Arrow indicates an evolutionary regain of petals not evident in our ASRs, but is instead based on results in Sharples and Tripp (2019b). Black stars indicate our two focal experimental taxa. Branch numbers indicate clades recognized in Sharples and Tripp (2019a): 1, Insignes (petalous ancestor, no changes); 2, Petiolares (petalous ancestor, two losses); 3, Plettkeae (apetalous ancestor, one loss in this ancestor); 4, Nitentes (petalous ancestor, two losses); 5, Larbreae

(petalous ancestor); 6, S. cuspidata group (Cuspidatae) (petalous ancestor, no changes); 7, S. longipes group (Pedunculares) and sister clades (petalous ancestor, no changes); 8, Maori,

Spinescentes, and Uliginosae (petalous ancestor, possibly six losses); 9, Foliaceo-bracteatae

(possibly petalous ancestor, possibly one loss and three regains); 10, rapid radiation in Asia

(petalous ancestor, possibly seven losses and three regains). 250

251 occasional reversals back to petaly (Fig. 24), though few reversals were supported overall and were limited to clades 9 and 10 (Fig. 24). The number of losses and regains in a largely Asian radiation (clade 10) is unclear due to uncertainty near the backbone and low support for relevant clades (Fig. 24; Supp. Fig. 3; Sharples and Tripp, 2019a). Altogether, based on node and branch support probabilities, petal losses were supported 19 times and reversals back to petaly were supported five times (but see discussion). Stochastic mapping, however, simulated 98.05 changes between states, with changes from petaly to apetaly occurring 47.73 times and changes from apetaly to petaly occurring 50.32 times on average. A shift from petaly to apetaly did not unite any species-rich clades, but in contrast, all members of several species-rich clades were united by petaly, including the Stellaria cuspidata Willd. ex Schltdl. group of the New World as well as the large clade containing S. longipes (Fig. 24). Though petaly is labile across Stellaria worldwide, petal presence or absence exhibited moderate phylogenetic signal (Pagel’s λ= 0.436; p=0.007), likely due to petaly uniting certain species-rich clades.

DISCUSSION

Our study details the pattern of petal evolution across a nearly complete phylogenetic sampling of Stellaria. We here estimate that the presence of petals was likely ancestral within

Stellaria, with numerous subsequent losses occurring in unrelated lineages worldwide. In the few instances in which petal loss was followed by regain, petaly never again underwent subsequent transition. Furthermore, apetalous lineages are typically species-poor in Stellaria. Experimental data from our field study of two species—one apetalous and one petalous—indicate that showy, petalous Stellaria species are predominantly outcrossing, in agreement with prior work

(Chinnappa and Morton, 1984; Campbell, 1985; Motten, 1986). In contrast, the apetalous species 252 is largely selfing in the tundra biome of Niwot Ridge, Colorado. These field data lend support to the hypothesis that apetaly may frequently be associated with self-pollination, which has to date been little explored in apetalous starworts.

Pollination Biology

Arctic-alpine flowering plant lineages are often characterized by high rates of self- pollination (Löve and Löve, 1974b; Petersen, 1977; Molau, 1993; Brochmann and Steen, 1999;

Grundt et al., 2005; Jónsdóttir, 2011; Peng et al., 2014). This phenomenon is generally attributed to ensuring sexual reproduction in an extreme environment where pollinator availability varies as a function of interannual conditions and day-to-day volatility. Although not an explicit focus of our study, during our field experiment, we failed to observe any pollinators on either of our focal species at the Niwot Ridge tundra site. On the Arctic island of Svalbard, self-compatibility is so pronounced that of all the bisexual plant species occurring there, none are thought to be self- incompatible (Brochmann and Steen, 1999). Our data collected at the arctic-alpine Niwot Ridge site provide strong evidence for a preponderance of self-pollination in the apetalous S. irrigua, along with a lack of self-pollination in the petalous, outcrossing S. longipes. These data concur with a pattern of higher reproductive output in predominantly self-pollinating plants than in outcrossing species in the Arctic or arctic analogue regions (Table 12; Fig. 21; Molau, 1993;

Jónsdóttir, 2011). Our results also provide additional support for Molau’s (1993) findings that later flowering species of the arctic-alpine are more likely to be self-compatible than earlier flowering, showier-flowered species in the same ecosystems, the latter of which are more dependent on outcrossing for reproductive assurance. At Niwot Ridge, where the growing season spans only approximately two months, S. longipes is typically flowering by late June whereas S. irrigua does not flower until middle to late July, when many early-flowering species are already 253 in fruit (M. Sharples: pers. obs.). Molau (1993) also found that later flowering, more self- compatible species were often tied to late-thawing habitats such as snow beds. This is also in agreement with our findings: that S. longipes of early-thawing sites is predominantly outcrossing, whereas moisture-loving S. irrigua of late-thawing snowbeds, marshes, and the like (Sharples and Tripp, 2019b) is predominantly selfing, a result also recovered in an alpine system in the sister genus Cerastium (Totland and Schulte-Herbrüggen, 2003).

Self-pollination occurs throughout varied ecosystems but may be especially overrepresented in some environments such as aridlands and Mediterranean climates (Stebbins,

1957; Rick, 1988; Arroyo et al., 2006), islands (Barrett, 1996; Grossenbacher et al., 2017), naturalized habitats (Stebbins, 1957; Razanajatovo et al., 2016), in addition to aforementioned arctic-alpine regions (Löve and Löve, 1974b; Petersen, 1977; Molau, 1993; Brochmann and

Steen, 1999; Grundt et al., 2005; Jónsdóttir, 2011; Peng et al., 2014). Self-pollination is also overrepresented in environments typically characterized as high-stress or limited by one or more factors, e.g., nutrient stress, water limitation, temperature extremes, habitats with pronounced seasonality, peripheral or marginal distributions, areas of pollinator unpredictability or high competition, areas with a sole colonist, and regions where population genetic bottlenecks are frequent (Stebbins, 1957; Jain, 1976; Wyatt, 1988; Lloyd and Schoen, 1992; Barrett, 1996;

Aarssen, 2000; Silvertown and Charlesworth, 2001; Moeller and Geber, 2005; Arroyo et al.,

2006; Brys and Jacquemyn, 2011; Levin, 2012; Spigler and Kalisz, 2017). However, our two focal species geographically range into less “limiting” environments than the arctic-alpine context of our Niwot Ridge study site, with no differences in floral morphology from the alpine populations studied here. That is, S. irrigua ranges into boreal forests and S. longipes into myriad habitats below the treeline, but flowers below the arctic-alpine zone are not readily distinct from 254 flowers in arctic-alpine zones. Thus, in the context of high rates of self-pollination recovered in

S. irrigua, occupation of pollinator-limited, arctic-alpine habitats does not in itself correlate with loss of petals in this species, and future studies should take into consideration relevant information regarding the full range of niche of this species.

Dissection of unopened flowers of S. irrigua strongly suggests that bud pollination likely facilitates self-pollination in this lineage rather than alternatives, such as cleistogamy or parthenogenesis. Rather than producing flowers that never open but instead self-pollinate in a permanent closed state as in true cleistogamy (e.g., Acleisanthes, Danthonia, Lespedeza,

Lithospermum, Medicago, Ruellia, Schiedea, Viola; reviewed in Culley and Klooster, 2007), flowers of S. irrigua develop mature anthers that deposit pollen onto receptive stigmas of the same flower prior to floral anthesis (Fig. 23), and flowers then open normally as in chasmogamous species. Bud pollination has earlier been observed in Stellaria media, Cerastium caespitosum Gilib., and numerous other unrelated species that occur in the cool, subarctic regions of the Faroe Islands (Hagerup, 1951), and has also been reported in a segregate of S. media (S. pallida: Schultz, 1906). It has also been reported from the extreme arctic climate of

Svalbard (e.g., of Draba norvegica Gunnerus; Brochmann and Steen, 1999) as well as in a peripheral Arctic population of Campanula uniflora L., a genus which is otherwise self- compatible (Ægisdóttir and Thórhallsdóttir, 2006). Bud pollination may be a widespread mechanism of selfing in cold climates generally and its relative preponderance in such ecosystems should be further assessed.

The phenomenon of bud pollination followed by ephemeral flower opening in S. irrigua is in stark contrast with the wide diversity of self-pollination systems found in other closely related Caryophyllaceae. Pseudostellaria, a close relative of Stellaria, produces both showy 255 chasmogamous as well as true cleistogamous flowers on a single individual (Luo et al., 2012).

Some species of Schiedea are also reported to be cleistogamous (Wagner et al., 2005). Self- pollination in certain populations of Arenaria uniflora (Walt). Muhl. neither occurs in bud nor in flowers that never open but rather occurs while flowers are in the process of opening and is at times further assured through early closing of the sepals, which forces into contact mature male and female floral whorls (Wyatt, 1984). Elsewhere in starworts, Stellaria alsine Grimm has been described as spontaneously self-pollinating in the opened state due to “great movements” of the anthers “to and fro” (Hagerup, 1951: p. 36).

In S. longipes, we observed numerous instances wherein flowers subjected to self- pollination treatment, upon later dissection, yielded evidence of pollen deposition on stigmas of the same flower (Fig. 23). Such instances, however, yielded no reproductive output (Fig. 21).

Whether pollen deposition occurred in bud or during floral anthesis is unknown, but a lack of physical contact between pollen and stigma cannot explain a lack of selfing in studied individuals of this species. In other populations of this species, both protandry and gynodioecy along with male-sterile individuals have been documented (Philipp, 1980; Chinnappa and

Morton, 1984), suggesting multiple strategies at work in preventing self-pollination across the wide geographical range of S. longipes. Although some self-pollination has been reported in greenhouse experiments with this species (Chinnappa and Morton, 1984), it is still unclear whether utilized material is truly conspecific with southern Rocky Mountains plants herein studied (Sharples, 2019). Neither male sterility, gynodioecy, nor protandry were observed in our experimental field population, but we cannot rule out these phenomena elsewhere on Niwot

Ridge or throughout the southern Rocky Mountains, where S. longipes is common. Nonetheless, from our experiment it is clear that individuals of S. longipes on the tundra of Niwot Ridge are 256 essentially obligate outcrossers, whereas individuals of S. irrigua are precociously self- pollinating.

Pollination between flowers of the same individual (geitonogamy) is generally prohibited in S. irrigua by the very limited time period in which flowers of the terminal cymes are open

(Sharples et al., unpub. data). This contrasts with flowers of Stellaria longipes, which are open for longer durations. These two strategies are consistent with observations comparing self- pollinating versus outcrossing flowers of Arenaria uniflora mentioned above (Wyatt, 1984). In S. irrigua, typically only one flower is open at a time (some exceptions exist for plants with many- flowered inflorescences, these particularly below treeline; M. Sharples, unpub. data). For S. irrigiua, none of our selfing treatments produced more than 7 flowers per bag in total (range: 1–

7). This implies that self-pollination between flowers on the same individual (versus within- flower selfing) likely contributed very little to resulting seed set. For S. longipes geitonogamy is similarly generally prohibited because only a single flower is typically borne per individual plant

(exceptions exist, and these individuals primarily with a second flower). Nonetheless, our results indicate a near total lack of selfing across all individuals of S. longipes, regardless of number of flowers per individual (Appendix 6). Showy petals on often solitary flowers in S. longipes facilitates outcrossing, this at times with individuals of different ploidies (Chinnappa et al.,

2005), which in combination have likely facilitated the proliferation of this species across a widespread circumboreal distribution (Löve and Löve, 1974b), while in S. irrigua, a lack of petals and a high, essentially obligate rate of self fertility across its entire inflorescence allow it to succeed across a similarly expansive distribution.

Evolution of Apetaly in Stellaria 257

Transitions to a self-compatible pollination system from a self-incompatible ancestor are evolutionarily far more frequent than the converse, as inferred here in Stellaria as well as in prior work (e.g., Rick, 1988; Igić et al., 2003; Ferrer and Good-Avila, 2006; Hörandl, 2010). Indeed, the transition from self-incompatibility to self-compatibility may be among the most common evolutionary transitions in angiosperms (Stebbins, 1957; Stebbins, 1970) and is highly convergent within and across lineages (Wyatt, 1988). Paradoxically, the shift to selfing has been vigorously debated as representing an evolutionary dead end in flowering plants (e.g., Stebbins,

1957; Takebayashi and Morrell, 2001; Igić and Busch, 2013) as well as in other systems (Tripp,

2016). Our data rejected trait-irreversible ASR models and recovered rampant transitions between petalous and apetalous states. However, apetalous lineages underwent little to no diversification, and transitions from apetaly back to petaly were only one-fourth as frequent as petal losses from petalous ancestors based on evaluation of probabilities rather than stochastic mapping averages (Fig. 24). The mechanism for recurrent petal losses in such cases may be simple: if a species has become predominantly self-pollinating from an outcrossing progenitor state, petal production may confer no further adaptive benefits but instead only cost (Ornduff,

1969).

Nonetheless, aberrant petals that tend to be minute (and fewer than the five typical of petalous species) are occasionally produced in otherwise apetalous species. Such malformed petals may serve to facilitate rare outcrossing thereby introducing new genetic diversity into lineages that are apetalous and otherwise largely selfing. As capsules with 100% fertility were relatively rare in our selfed S. irrigua datasets (and absent from the S. longipes datasets altogether), bud pollination in this case does not exclude the possibility of outcrossing. Most apetalous species of Stellaria have wide to wider geographical distributions than closely-related, 258 petalous species (Sharples and Tripp, 2019b; Sharples, unpub. data; see also Ornduff, 1969 and

Randle et al., 2009), suggesting that despite presumed predominance of self-pollination and higher rates of inbreeding, ranges and/or ecological amplitude are minimally affected. On the contrary, Stellaria irrigua, for example, occurs across an amphi-Beringian arc from the mountains and arctic latitudes of eastern Asia to western North America, oftentimes occurring near the upper elevational limits of vascular plant growth (Sharples and Tripp, 2019b).

The artic-alpine habitats in which S. irrigua proliferates lend insight into occasional opportunities for outcrossing in this species. Although the flowers of S. irrigua are not suggestive of water pollination (syndrome sensu Du and Wang, 2014), and despite our experiment supporting high levels of self-fertilization in this species, S. irrigua often grows in dense, localized clusters very close to the ground surface and outcrossing between individuals within a dense cluster may be facilitated by water in some instances. The specific epithet indeed refers to its “irrigated” quality, and S. irrigua preferentially inhabits areas that are wet throughout the growing season, including riparian margins and areas of late snowmelt. During heavy precipitation events or during particularly dramatic snowmelt days, it is plausible that water may sporadically facilitate pollen transfer between individuals of these low-growing plants.

Furthermore, field observations (M. Sharples, pers. obs.) in regions such as the Himalaya, which experiences a heavy monsoon during the alpine growing season, suggest that rain-facilitated pollen dispersal may occur between proximate flowers of different individuals as well as within single flowers during heavy precipitation events. In fact, rain pollination has been reported in multiple unrelated species occupying another area (the Faroe Islands) with cold, rainy growing seasons that may be lacking in available pollinators (Hagerup, 1950; Hagerup, 1951). The cold climate of Niwot Ridge experiences what is locally referred to as a summer monsoon, which 259 often results in heavy thunderstorms capable of producing localized torrents in wet areas (M.

Sharples, pers. obs.). Of interest, autogamy in species closely tied to water is thought to be prerequisite for a transition to a fully water-based pollination system (Du and Wang, 2014). Our study site at Niwot Ridge also experiences consistently pronounced winds (Petersen, 1977) such that limited, stochastic wind-mediated pollen dispersal may occasionally facilitate outcrossing.

Regardless of outcrossing opportunities, however, Stellaria irrigua predominantly assures sexual reproduction through a preponderance of bud pollination.

Based on taxon sampling in an earlier study (Sharples and Tripp, 2019b), it is likely that reappearance of petals from an apetalous progenitor occurred in at least one instance that is not evident from our ancestral reconstruction here (Fig. 24). One of our focal species, S. irrigua, is a geographically widespread lineage from which a narrowly distributed, southern Rocky

Mountains endemic species Stellaria sanjuanensis M.T.Sharples & E.Tripp likely derived. Upon phylogenetic sampling across much of the distribution of S. irrigua, data indicate that S. sanjuanensis regained petals from its presumed apetalous progenitor lineage (Sharples and Tripp,

2019b, Fig. 5 therein), a result clearly not recovered in the present study with only a single tip representing S. irrigua in the ASR dataset (Fig. 24). Not only does this suggest that loss of petals is not necessarily accompanied by loss of genetic architecture needed to manufacture petals in

Stellaria, but this observation also highlights potential underestimation of evolutionary transitions when a single sample is used as a placeholder to represent a widespread and evolutionarily dynamic species complex, emphasizing the need for properly considered taxon sampling (Hardy, 2006). We did not include multiple samples of S. irrigua for the present study so as not to bias taxon sampling towards multiple tips for only one species, but perhaps one solution would be to include multiple samples to represent widespread species and one or few 260 samples to represent narrowly distributed species to better account for evolutionary diversity within a genus.

Once complete sampling and further understanding of the life history of Stellaria species are achieved, we may more adequately infer whether petal reductions and/or plasticity are intermediate states towards full apetaly, as well as investigate other floral modifications potentially associated with apetaly. Additionally, the genetic basis of apetaly remains ripe for future study, as does additional field investigation targeting correlates of petal plasticity.

Stellaria thus represents an incipient model system for further understanding the evolution of apetaly.

261

BIBLIOGRAPHY

Aarssen, L. W. 2000. Why are most selfers annuals? A new hypothesis for the fitness benefit of selfing. Oikos 89: 606–612. DOI: 10.1034/j.1600-0706.2000.890321.x.

Abbott, R. J. and C. Brochmann. 2003. History and evolution of the arctic flora: in the footsteps of Eric Hultén. Molecular Ecology 12: 299–313. DOI: 10.1046/j.1365-294X.2003.01731.x.

Ackerfield, J. 2015. Flora of Colorado. Bot. Misc. 41. Fort Worth, Texas, U.S.A.: Botanical Research Institute of Texas Press.

Ægisdóttir, H. H. and T. E. Thórhallsdóttir. 2006. Breeding system evolution in the Arctic: a comparative study of Campanula uniflora in Greenland and Iceland. Arctic, Antarctic, and Alpine Research 38: 305–312. DOI: 10.1657/1523-0430(2006)38[305:BSEITA]2.0.CO;2.

Anderson, J., P. J. Burns, D. Milroy, R. Ruprecht, T. Hauser, and H. J. Siegel. 2017. Deploying RMACC Summit: an HPC resource for the Rocky Mountain region. In Proceedings of PEARC17. New Orleans, LA, USA, July 9–13, 2017, 7 pages. DOI: 10.1145/3093338.3093379.

Andrews, S. 2010. FastQC: A quality control tool for high throughput sequence data. Available from: http://www.bioinformatics.babraham.ac.uk/projects/fastqc.

Angiosperm Phylogeny Group IV. 2016. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG IV. Botanical Journal of the Linnean Society 181:1–20.

Ankei, T. 1982. Habitat gradient and reproductive habits of the seven Stellaria species in Japan. The Botanical Magazine (Shokubutsugaku zasshi) 94: 35–48.

Applequist, W. L. 2017. Report of the Nomenclature Committee for Vascular Plants: 69. Taxon 66: 500–513. DOI: 10.12705/662.17.

Arabi, Z., F. Ghahremaninejad, R. K. Rabeler, I. Sokolova, G. Heubl, and S. Zarre. 2018. On the taxonomic status of the genus Dichodon (Caryophyllaceae: tribe Alsineae): morphological and molecular evidence reassessed. Phytotaxa 360: 220–236. DOI: 10.11646/phytotaxa.360.3.3.

Aronesty, E. 2011. Command-Line tools for processing biological sequencing data, ea-utils. Durham, North Carolina: Expression Analysis. Available from: https://expressionanalysis.github.io/ea-utils/.

262

Arroyo, M. T. K., M. S. Muñoz, C. Henríquz, I. Till-Bottraud, and F. Pérez. 2006. Erratic pollination, high selfing levels and their correlates and consequences in an altitudinally widespread above-tree-line species in the high Andes of Chile. Acta Oecologica 30: 248–257. DOI: 10.1016/j.actao.2006.05.006.

Baker, R. A. and W. C. Burger. 1983. Caryophyllaceae. Pp. 227–247 in Flora Costaricensis, ed. W. C. Burger. Fieldiana Botany, New Series No. 13. Chicago: Field Museum of Natural History.

Bakker, F. T., D. Lei, J. Yu, S. Mohammadin, Z. Wei, S. van de Kerke, B. Gravendeel, M. Nieuwenhuis, M. Staats, D. E. Alquezar-Planas, and R. Holmer. 2016. Herbarium genomics: Plastome sequence assembly from a range of herbarium specimens using an Iterative Organelle Genome Assembly pipeline. Biological Journal of the Linnean Society 117: 33–43. DOI: 10.1111/bij.12642.

Barrett, S. C. 1996. The reproductive biology and genetics of island plants. Philosophical Transactions of the Royal Society B 351: 725–733. DOI: 10.1098/rstb.1996.0067.

Beaman, J. H. 2009. Caryophyllaceae. In Flora de Nicaragua, vol. 1: Introducción, Gimnospermas y Angiospermas (Acanthaceae-Euphorbiaceae). St. Louis: Missouri Botanical Garden Press. Available online from: http://www.tropicos.org/Project/FN (last accessed Sept. 2018).

Beck, J. B. and J. C. Semple. 2015. Next-generation sampling: Pairing genomics with herbarium specimens provides species-level signal in Solidago (Asteraceae). Applications in Plant Sciences 3(6): 1500014. DOI: 10.3732/apps.1500014.

Beeler, R. B., M. T. Sharples, and E. A. Tripp. In press. Introgression among three western North American bilberries (Vaccinium section Myrtillus). Systematic Botany.

Bennike, O. and J. Böcher. 1990. Forest-tundra neighbouring the North Pole: Plant and insect remains from the Plio-Pleistocene Kap København Formation, North Greenland. Arctic 43: 331– 338. DOI: 10.14430/arctic1629.

Besnard, G., P.-A. Christin, P.-J. G. Malé, E. Lhuillier, C. Lauzeral, E. Coissac, and M. S. Vorontsova. 2014. From museums to genomics: Old herbarium specimens shed light on a C3 to C4 transition. Journal of Experimental Botany 65: 6711–6721. DOI: 10.1093/jxb/eru395.

Bethke, P. 2011. Mineralization in the Eastern San Juan Mountains. In: R. Blair and G. Bracksieck. The Eastern San Juan Mountains: Their geology, ecology, and human history. Boulder, Colorado, U.S.A.: University Press of Colorado.

Bigelow, J. 1824. Florula Bostoniensis. A collection of plants of Boston and its vicinity, with their generic and specific characters, principal synonyms, descriptions, places of growth, and time of flowering, and occasional remarks, 2nd ed. Boston: Cummings, Hilliard, & Co.

263

Blair, R. and G. Bracksieck. 2011. The Eastern San Juan Mountains: Their geology, ecology, and human history. Boulder, Colorado, U.S.A.: University Press of Colorado.

Bouckaert, R. R., J. Heled, D. Kuehnert, T. G. Vaughan, C-H. Wu, D. Xie, M. A. Suchard, A. Rambaut, and A. J. Drummond. 2014. BEAST 2: A software platform for Bayesian evolutionary analysis. PloS Computational Biology 10: e1003537. DOI: 10.1371/journal.pcbi.1003537.

Brako, L. and J. L. Zarucchi. 1993. Catalogue of the Flowering Plants and Gymnosperms of Peru. St. Louis: Missouri Botanical Garden Press.

Brochmann, C. and S. W. Steen. 1999. Sex and genes in the flora of Svalbard - implications for conservation biology and climate change. Det Norske Videnskaps-Akademi. I. Matematisk Naturvitenskapelig Klasse, Skrifter, Ny serie 38: 33–72.

Brys, R. and H. Jacquemyn. 2011. Variation in the functioning of autonomous self-pollination, pollinator services and floral traits in three Centaurium species. Annals of Botany 107: 917–925. DOI: 10.1093/aob/mcr032.

Bunge, A. V. 1835. Verzeichniss der im Jahre 1832, im östlichen Theile des Altai-Gebirges gesammelten Pflanzen: ein Supplement zur Flora Altaica. Mémoires présentés à l'Académie impériale des Sciences de St. Petersbourg par divers Savans et dans ses assemblées 2: 523–610. Campbell, D. 1985. Pollinator sharing and seed set of Stellaria pubera: Competition for pollination. Ecology 66: 544–553.

Cardillo, M., P. H. Weston, Z. K. M. Reynolds, P. M. Olde, A. R. Mast, E. M. Lemmon, A. R. Lemmon, and L. Bromham. 2017. The phylogeny and biogeography of Hakea (Proteaceae) reveals the role of biome shifts in a continental plant radiation. Evolution 71: 1928–1943. DOI: 10.1111/evo.13276.

Cariou, M., L. Duret, and S. Charlat. 2013. Is RAD-seq suitable for phylogenetic inference? An in silico assessment and optimization. Ecology and Evolution 3: 846–852. DOI: 10.1002/ece3.512.

Chater, A. O. and V. H. Heywood. 1964. 6. Stellaria L. Pp. 133–136 in Flora Europaea, vol. 1 (Lycopodiaceae to Platanaceae), eds. T. G. Tutin, V. H. Heywood, N. A. Burges, D. H. Valentine, S. M. Walters, and D. A. Webb. Cambridge, UK: Cambridge University Press.

Chinnappa, C. C. 1992. Stellaria porsildii, sp. nov., a new member of the S. longipes complex (Caryophyllaceae). Systematic Botany 17: 29–32. DOI: 10.2307/2419062.

Chinnappa, C. C. and J. K. Morton. 1984. Studies on the Stellaria longipes complex (Caryophyllaceae)—Biosystematics. Systematic Botany 9: 60–73.

Chinnappa, C. C. and J. K. Morton. 1991. Studies on the Stellaria longipes complex (Caryophyllaceae) – taxonomy. Rhodora 93: 129–135.

264

Chinnappa, C. C., G. M. Donald, R. Sasidharan, and R. J. N. Emery. 2005. The biology of Stellaria longipes (Caryophyllaceae). Canadian Journal of Botany 83: 1367–1383. DOI: 10.1139/b05-117.

Clark, D. 1996. A floristic survey of the Mesa de Maya region, Las Animas County, Colorado. Natural History Inventory of Colorado 17:1–44.

Colorado Natural Heritage Program. 2015. CNHP tracked vascular plant species available at www.cnhp.colostate.edu/download/list/vascular.asp. Last accessed 15 Jan 2017. Colorado State University, Fort Collins, Colorado, U.S.A.

Crisp, M. D., M. T. K. Arroyo, L. G. Cook, M. A. Gandolfo, G. J. Jordan, M. S. McGlone, P. H. Weston, M. Westoby, P. Wilf, and H. P. Linder. 2009. Phylogenetic biome conservatism on a global scale. Nature Letters 458: 754–760. DOI: 10.1038/nature07764.

Croizat, L. 1952. Manual of Phytogeography; or, an Account of Plant-Dispersal Throughout the World. The Hague: Dr. W. Junk.

Cruaud, A., M. Gautier, M. Galan, J. Foucaud, L. Sauné, G. Genson, E. Dubois, S. Nidelet, T. Deuve, and J.-Y. Rasplus. 2014. Empirical assessment of RAD sequencing for interspecific phylogeny. Molecular Biology and Evolution 31: 1272–1274. DOI: 10.1093/molbev/msu063.

Culley, T. M. and M. R. Klooster. 2007. The cleistogamous breeding system: A review of its frequency, evolution, and ecology in angiosperms. The Botanical Review 73: 1–30. DOI: 10.1663/0006-8101(2007)73[1:TCBSAR]2.0.CO;2.

Culley, T. M., S. G. Weller, and A. K. Sakai. 2002. The evolution of wind pollination in angiosperms. TRENDS in Ecology & Evolution 17: 361–369. DOI: 10.1016/S0169- 5347(02)02540-5.

Daniel, T. F. and E. A. Tripp. 2018. Louteridium (Acanthaceae: Acanthoideae: Ruellieae: Trichantherinae): Taxonomy, phylogeny, reproductive biology, and conservation. Proceedings of the California Academy of Sciences 65: 41–106.

De Casas, R. R., M. E. Mort, and D. E. Soltis. 2016. The influence of habitat on the evolution of plants: a case study across Saxifragales. Annals of Botany 118: 1317–1328. DOI: 10.1093/aob/mcw160.

De Queiroz, K. 2005. The resurrection of oceanic dispersal in historical biogeography. TRENDS in Ecology and Evolution 20: 68–73. DOI: 10.1016/j.tree.2004.11.006.

De Queiroz, K. 2007. Species concepts and species delimitation. Systematic Biology 56: 879– 886. DOI: 10.1080/10635150701701083.

265

DeChaine, E. G. 2008. A bridge or a barrier? Beringia’s influence on the distribution and diversity of tundra plants. Plant Ecology and Diversity 1: 197–207. DOI: 10.1080/17550870802328660.

Díaz-Arce, N., H. Arrizabalaga, H. Murua, X. Irigoien, and N. Rodríguez-Ezpeleta. 2016. RAD- seq derived genome-wide nuclear markers resolve the phylogeny of tunas. Molecular Phylogenetics and Evolution 102: 202–207. DOI: 10.1016/j.ympev.2016.06.002.

Dillenberger, M. S. and J. W. Kadereit. 2014. Maximum polyphyly: Multiple origins and delimitation with plesiomorphic characters require a new circumscription of Minuartia (Caryophyllaceae). Taxon 63: 64–88. DOI: 10.12705/631.5.

Dolezal, J., M. Dvorsky, M. Kopecky, P. Liancourt, I. Hiiesalu, M. Macek, J. Altman, Z. Chlumska, K. Rehakova, K. Capkova, J. Borovec, O. Mudrak, J. Wild, and F. Schweingruber. 2016. Vegetation dynamics at the upper elevational limit of vascular plants in Himalaya. Scientific Reports 6: 24881. DOI: 10.1038/srep24881.

Douglas, P. 1992. Vascular flora of the Conejos River Basin, San Juan Mountains, Conejos County, Colorado. Unpublished Master’s Thesis, Colorado State University, Fort Collins, Colorado, U.S.A.

Doyle, J. J. and J. L. Doyle. 1987. A rapid isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11–15.

Du, Z.-Y. and Q.-F. Wang. 2014. Correlations of life form, pollination mode and sexual system in aquatic angiosperms. PLoS ONE 9(12): e115653.

Duke, J. A. 2013. Caryophyllaceae. In Flora of Panama (World Flora Online). Available from: http://www.tropicos.org/Project/FOPWFO (last accessed July 2018).

Dupin, J., N. J. Matzke, T. Särkinen, S. Knapp, R. G. Olmstead, L. Bohs, and S. D Smith. 2017. Bayesian estimation of the global biogeographical history of the Solanaceae. Journal of Biogeography 44: 887–899. DOI: 10.1111/jbi.12898.

Eaton, D. A. R. 2018. iPyRAD: Interactive assembly and analysis of RADseq data sets. Available from: https://ipyrad.readthedocs.io/installation.html.

Eaton, D. A. R. and R. H. Ree. 2013. Inferring phylogeny and introgression using RADseq data: An example from flowering plants (Pedicularis: Orobanchaceae). Systematic Biology 62: 689– 706. DOI: 10.1093/sysbio/syt032.

Eaton, D. A. R., E. L. Spriggs, B. Park, and M. J. Donoghue. 2017. Misconceptions on missing data in RAD-seq phylogenetics with a deep-scale example from flowering plants. Systematic Biology 66: 399–412. DOI: 10.1093/sysbio/syw092.

266

Ebersbach, J., A. N. Muellner-Riehl, I. Michalak, N. Tkach, M. H. Hoffmann, M. Röser, H. Sun, and A. Favre. 2017. In and out of the Qinghai-Tibet Plateau: divergence time estimation and historical biogeography of the large arctic-alpine genus Saxifraga L. Journal of Biogeography 44: 900–910. DOI: 10.1111/jbi.12899.

Eidesen, P. B., D. Ehrich, V. Bakkestuen, I. G. Alsos, O. Gilg, P. Taberlet, and C. Brochmann. 2013. Genetic roadmap of the Arctic: plant dispersal highways, traffic barriers and capitals of diversity. New Phytologist 200: 898–910. DOI: 10.1111/nph.12412.

Elias, S. A., S. K. Short, C. H. Nelson, and H. H. Birks. 1996. Life and times of the Bering land bridge. Nature 382: 60–63. DOI: 10.1038/382060a0.

Endlicher, S. 1840. Stellaria. P. 969 in Genera plantarum secundum ordines naturales disposita, ed. S. Endlicher. Vindobonae: Apud Fr. Beck Universitatis Bibliopolam.

Endress, P. K. and M. L. Matthews. 2006. First steps towards a floral structural characterization of the major rosid subclades. Plant Systematics and Evolution 260: 223–251. DOI: 0.1007/s00606-006-0444-7.

Ertter, B. 2000. Floristic surprises in North America north of Mexico. Annals of the Missouri Botanical Garden 87:81–109.

ESRI. 2016. ArcGIS. www.esri.com.

Faegri, K. and L. van der Pijl. 1979. Principles of pollination ecology (3rd ed.). Oxford: Pergamon Press.

Fan, D-M., J-H. Chen, Y. Meng, J. Wen, J-L. Huang, and Y-P. Yang. 2013. Molecular phylogeny of Koenigia L. (Polygonaceae: Persicarieae): Implications for classification, character evolution and biogeography. Molecular Phylogenetics and Evolution 69: 1093–1100. DOI: 10.1016/j.ympev.2013.08.018.

Favre, A., M. Päckert, S. U. Pauls, S. C. Jähnig, D. Uhl, I. Michalak, and A. N. Muellner-Riehl. 2015. The role of the uplift of the Qinghai-Tibetan Plateau for the evolution of Tibetan biotas. Biological Reviews 90: 236–253. DOI: 10.1111/brv.12107.

Favre, A., I. Michalak, C-H. Chen, J-C. Wang, J. S. Pringle, S. Matuszak, H. Sun, Y-M. Yuan, L. Struwe, and A. N. Muellner-Riehl. 2016. Out-of-Tibet: the spatio-temporal evolution of Gentiana (Gentianaceae). Journal of Biogeography 43: 1967–1978. DOI: 10.1111/jbi.12840.

Fenster, C. B., W. S. Armbruster, P. Wilson, M. R. Dudash, and J. D. Thomson. 2004. Pollination syndromes and floral specialization. Annual Review of Ecology, Evolution, and Systematics 35: 375–403. DOI: 10.1146/annurev.ecolsys.34.011802.132347.

267

Fenzl, E. 1842. Stellaria L. Pp. 375–395 in Flora Rossica; sive, Enumeratio Plantarum in totius Imperii Rossici Provinciis Europaeis, Asiaticis et Americanis hucusque observatorum, vol. 1, ed. C. F. Ledebour. Stuttgart: Sumptibus Librariae E. Schweizerbart.

Fernández-Mazuecos, M., G. Mellers, B. Vigalondo, L. Sáez, P. Vargas, and B. J. Glover. 2018. Resolving recent plant radiations: Power and robustness of genotyping-by-sequencing. Systematic Biology 67: 250–268. DOI: 10.1093/sysbio/syx062.

Ferrer, M. M. and S. V. Good-Avila. 2006. Macrophylogenetic analyses of the gain and loss of self-incompatibility in the Asteraceae. New Phytologist 173: 401–414. DOI: 10.1111/j.1469- 8137.2006.01905.x.

Fior, S., P. O. Karis, G. Casazza, L. Minuto, and F. Sala. 2006. Molecular phylogeny of the Caryophyllaceae (Caryophyllales) inferred from chloroplast matK and nuclear rDNA ITS sequences. American Journal of Botany 93: 399–411. DOI: 10.3732/ajb.93.3.399.

Flaig, J. H. 2007. A vascular plant inventory of the Eastern San Juan Mountains and vicinity in Southern Colorado. Unpublished Master’s Thesis, University of Wyoming, Laramie, Wyoming, U.S.A.

Flora of China Editorial Committee. 2008. Flora of China, published on the Internet and available at http://www.efloras.org. Last accessed 15 Jan 2017. Missouri Botanical Garden, St. Louis, Missouri, & Harvard University Herbaria, Cambridge, Massachusetts, U.S.A.

Flora of North America Editorial Committee. 1993+. Flora of North America north of Mexico, 19+ volumes. Online edition. Accessed from Dec 2013 to Oct 2016. New York, U.S.A. and Oxford, U.K.

Frajman, B., F. Eggens, and B. Oxelman. 2009. Hybrid origins and homoploid reticulate evolution within Heliosperma (Sileneae, Caryophyllaceae)—A multigene phylogenetic approach with relative dating. Systematic Biology 58: 328–345.

Frajman, B., E. Záveská, A. Gamisch, T. Moser, The STEPPE Consortium, and P. Schönswetter. 2019. Integrating phylogenomics, phylogenetics, morphometrics, relative genome size and ecological niche modeling disentangles the diversification of Eurasian Euphorbia seguieriana s.l. (Euphorbiaceae). Molecular Phylogenetics and Evolution 134: 238–252. DOI: 10.1016/j.ympev.2018.10.046.

Friedman, J. and S. C. H. Barrett. 2009. Wind of change: news insights on the ecology and evolution of pollination and mating in wind-pollinated plants. Annals of Botany 103: 1515–1527.

Funk, V. 2004. 100 uses for an herbarium (well at least 72). A. S. P. T. Newsletter 17(2):17–19.

Garnock-Jones, P. J. 1988. Caryophyllaceae. Pp. 473–510 in Flora of New Zealand, vol. 4: Naturalised Pteridophytes, Gymnosperms, Dicotyledons, eds. C. J. Webb, W. R. Sykes, and P. J. Garnock-Jones. Christchurch: Botany Division, D.S.I.R. 268

Gervasi, D. D. L. and F. P. Schiestl. 2017. Real-time divergent evolution in plants driven by pollinators. Nature Communications 8: 14691. DOI: 10.1038/ncomms14691.

Ghazanfar, S. A. and Y. J. Nasir. 2011. Stellaria. In Flora of Pakistan. Available from: http://www.tropicos.org/Name/40004952?projectid=32 (last accessed July 2018).

Gizaw, A., C. Brochmann, S. Nemomissa, T. Wondimu, C. A. Masao, F. M. Tusiime, A. A. Abdi, B. Oxelman, M. Popp, and D. Dimitrov. 2016. Colonization and diversification in the African ‘sky islands’: insights from fossil-calibrated molecular dating of Lychnis (Caryophyllaceae). New Phytologist 211: 719–734. DOI: 10.1111/nph.13937.

Gonzales, D. A. and K. E. Karlstrom. 2011. A legacy of mountains past and present in the San Juan Region. In: R. Blair and G. Bracksieck. The Eastern San Juan Mountains: Their geology, ecology, and human history. Boulder, Colorado, U.S.A.: University Press of Colorado.

Greenberg, A. K. and M. J. Donoghue. 2011. Molecular systematics and character evolution in the Caryophyllaceae. Taxon 60: 1637–1652.

Greuter, W. 1995. Silene (Caryophyllaceae) in Greece: a subgeneric and sectional classification. Taxon 44: 543–581. DOI: 10.2307/1223499.

Grierson, A. J. C. 1984. 6. Stellaria L. Pp. 205–209 in Flora of Bhutan: Including a Record of Plants from Sikkim, vol. 1 pt. 2, eds. A. J. C. Grierson and D. G. Long. Edinburgh: Royal Botanic Garden.

Grossenbacher, D. L., Y. Brandvain, J. R. Auld, M. Burd, P.-O. Cheptou, J. K. Conner, A. G. Grant, S. M. Hovick, J. R. Pannell, A. Pauw, T. Petanidou, A. M. Randle, R. R. de Casas, J. Vamosi, A. Winn, B. Igic, J. W. Busch, S. Kalisz, and E. E. Goldberg. 2017. Self-compatibility is over-represented on islands. New Phytologist 215: 469–478. DOI: 10.1111/nph.14534.

Grundt, H. H., R. Elven, and C. Brochmann. 2005. A rare case of self-incompatibility in arctic plants: Draba palanderiana (Brassicaceae). Flora 200: 321–325. DOI: 10.1016/j.flora.2004.10.001.

György, Z., E. G. Tóth, N. Incze, B. Molnár, and M. Höhn. 2018. Intercontinental migration pattern and genetic differentiation of arctic-alpine Rhodiola rosea L.: A chloroplast DNA survey. Ecology and Evolution 8: 11508–11521. DOI: 10.1002/ece3.4589.

Hagen, A. R., H. Giese, and C. Brochmann. 2001. Trans-Atlantic dispersal and phylogeography of Cerastium arcticum (Caryophyllaceae) inferred from RAPD and SCAR markers. American Journal of Botany 88: 103–112. DOI: 10.2307/2657131.

Hagerup, O. 1950. Rain-pollination. Det Kongelige Danske Videnskabernes Selskab: Biologiske Meddelelser 18(5): 1–19.

269

Hagerup, O. 1951. Pollination in the Faroes in spite of rain and poverty in insects. Det Kongelige Danske Videnskabernes Selskab: Biologiske Meddelelser 18(15): 1–48.

Hara, H. and M. C. Febbs. 1979. Stellaria. Pp. 57–59 in An Enumeration of the Flowering Plants of Nepal, vol. 2, eds. H. Hara, L. H. J. Williams, and collaborators. London: Trustees of British Museum (Natural History).

Harbaugh, D. T., M. Nepokroeff, R. K. Rabeler, J. McNeill, E. A. Zimmer, and W. L. Wagner. 2010. A new lineage-based tribal classification of the family Caryophyllaceae. International Journal of Plant Sciences 171: 185–198. DOI: 10.1086/648993.

Hardy, C. R. 2006. Reconstructing ancestral ecologies: challenges and possible solutions. Diversity and Distributions 12: 7–19. DOI: 10.1111/j. 1366-9516.2006.00207.

Hauenschild, F., A. Favre, J. Schnitzler, I. Michalak, M. Freiburg, and A. N. Muellner-Riehl. 2017. Spatio-temporal evolution of Allium L. in the Qinghai-Tibet-Plateau region: Immigration and in situ radiation. Plant Diversity 39: 167–179. DOI; 10.1016/j.pld.2017.05.010.

Hedberg, K. O. 1992. Taxonomic differentiation in Saxifraga hirculus L. (Saxifragaceae)—a circumpolar Arctic-Boreal species of central Asiatic origin. Botanical Journal of the Linnean Society 109: 377–393.

Heil, K. D., S. L. O’Kane Jr., and L. M. Reeves. 2013. Flora of the Four Corners region: Vascular plants of the San Juan River drainage. St. Louis, Missouri, U.S.A.: Missouri Botanical Garden Press.

Hipp, A. L., D. A. R. Eaton, J. Cavender-Bares, E. Fitzek, R. Nipper, and P. S. Manos. 2014. A framework phylogeny of the American oak clade based on sequenced RAD data. PLoS ONE 9: e102272. DOI: 10.1371/journal.pone.0102272.

Hitchcock, A. S. and M. L. Green. 1929. Standard-species of Linnean genera of Phanerogamae. Pp. 110–199 in International Botanical Congress, Cambridge (England), 1930. Nomenclature. Proposals by British botanists. London: Printed under the authority of H.M.S.O. by Wyman and Sons.

Hoffmann, M. H. and M. Röser. 2009. Taxon recruitment of the arctic flora: an analysis of phylogenies. New Phytologist 182: 774–780. DOI: 10.1111/j.1469-8137.2009.02782.x.

Hoffmann, M. H., K. B. von Hagen, E. Hörandl, M. Röser, and N. V. Tkach. 2010. Sources of the Arctic flora: Origins of Arctic species in Ranunculus and related genera. International Journal of Plant Sciences 171: 90–106. DOI: 10.1086/647918.

Hogan, T. 1992. A floristic survey of the Eagles Nest Wilderness Area in the southern Gore Range of central Colorado. Natural History Inventory of Colorado 12:1–35.

270

Holmgren, N. H., P. K. Holmgren, J. L. Reveal, and Collaborators. 2012. Intermountain Flora; Vascular Plants of the Intermountain West, U.S.A.: Subclasses Magnoliidae-Caryophyllidae, vol. 2 (A). Bronx: The New York Botanical Garden Press.

Holstein, N. and S. S. Renner. 2011. A dated phylogeny and collection records reveal repeated biome shifts in the African genus Coccinia (Cucurbitaceae). BioMed Central Evolutionary Biology 11: 28. DOI: 10.1186/1471-2148-11-28.

Hörandl, E. 2010. The evolution of self-fertility in apomictic plants. Sexual Plant Reproduction 23: 73–86. DOI: 10.1007/s00497-009-0122-3.

Hou, Y., M. D. Nowak, V. Mirré, C. S. Bjorå, C. Brochmann, and M. Popp. 2015. Thousands of RAD-seq loci fully resolve the phylogeny of the highly disjunct arctic-alpine genus Diapensia (Diapensiaceae). PLoS ONE 10(10): e0140175. DOI: 10.1371/journal.pone.0140175.

Hou, Y., M. D. Nowak, V. Mirré, C. S. Bjorå, C. Brochmann, and M. Popp. 2016a. RAD-seq data point to a northern origin of the arctic-alpine genus Cassiope (Ericaceae). Molecular Phylogenetics and Evolution 95: 152–160. DOI: 10.1016/j.ympev.2015.11.009.

Hou, Y., C. S. Bjorå, H. Ikeda, C. Brochmann, and M. Popp. 2016b. From the north into the Himalayan-Hengduan Mountains: fossil-calibrated phylogenetic and biogeographical inference in the arctic-alpine genus Diapensia (Diapensiaceae). Journal of Biogeography 43: 1502–1513. DOI: 10.1111/jbi.12715.

Hu, S., D. L. Dilcher, D. M. Jarzen, and D. W. Taylor. 2007. Early steps of angiosperm- pollinator coevolution. Proceedings of the National Academy of Sciences, USA 105: 240–245. DOI: 10.1073/pnas.0707989105.

Huang, Y-J., Y-S. Liu, F. M. B. Jacques, T. Su, Y. Xing, and Z. Zhou. 2013. First discovery of Cucubalus (Caryophyllaceae) fossil, and its biogeographical and ecological implications. Review of Palaeobotany and Palynology 190: 41–47. DOI: 10.1016/j.revpalbo.2012.11.011.

Hultén, E. 1943. Stellaria longipes Goldie and its allies. Botaniska Notiser 2: 251–270.

Hughes, C. E. and G. W. Atchison. 2015. The ubiquity of alpine plant radiations: from the Andes to the Hengduan Mountains. New Phytologist 207: 275–282. DOI: 10.1111/nph.13230.

Idárraga-Piedrahita, Á. 2011. Caryophyllaceae. Pp. 396–397 in Flora de Antioquia: catálogo de las plantas vasculares, vol. 2: listado de las plantas vasculares del departamento de Antioquia, eds. Á. Idárraga-Piedrahita, R. del C. Ortiz, R. Callejas Posada, and M. Merello. St. Louis: Missouri Botanical Garden Press.

Igić, B., L. Bohs, and J. R. Kohn. 2003. Historical inferences from the self-incompatibility locus. New Phytologist 161: 97–105. DOI: 10.1046/j.1469-8137.2003.00952.x.

271

Igić, B and J. W. Busch. 2013. Is self-fertilization an evolutionary dead end? New Phytologist 198: 386–397. DOI: 10.1111/nph.12182.

Jain, S. K. 1976. Evolution of inbreeding in plants. Annual Review of Ecology and Systematics 7: 469–495. DOI: 10.1146/annurev.es.07.110176.002345.

Jara-Arancio, P., M. T. K. Arroyo, P. C. Guerrero, L. F. Hinojosa, G. Arancio, and M. A. Méndez. 2014. Phylogenetic perspectives on biome shifts in Leucocoryne (Alliaceae) in relation to climatic niche evolution in western South America. Journal of Biogeography 41: 328–338. DOI: 10.1111/jbi.12186.

Jarvis, C. E., F. R. Barrie, D. M. Allan, and J. L. Reveal. 1993. A List of Generic Names and Their Types. Regnum Vegetabile 127. Königstein: Koeltz Scientific Books.

Jia, S-W., M-L. Zhang, E. V. Raab-Straube, and M. Thulin. 2016. Evolutionary history of Gymnocarpos (Caryophyllaceae) in the arid regions from north Africa to central Asia. Biological Journal of the Linnean Society 119: 511–522. DOI: 10.1111/bij.12834.

Joly, S., F. Lambert, H. Alexandre, J. Clavel, E. Léveillé‐Bourret, and J. L. Clark. 2018. Greater pollination generalization is not associated with reduced constraints on corolla shape in Antillean plants. Evolution 72: 244–260. DOI: 10.1111/evo.13410.

Jones, S. D., J. K. Wipff, and P. M. Montgomery. 2014. Vascular Plants of Texas: A comprehensive checklist including synonymy, bibliography, and index. Austin: University of Texas Press.

Jónsdóttir, I. S. 2011. Diversity of plant life histories in the Arctic. Preslia 83: 281–300.

Jordan, G. J. and M. K. Macphail. 2003. A Middle-Late Eocene inflorescence of Caryophyllaceae from Tasmania, Australia. American Journal of Botany 90: 761–768.

Jørgensen, P. M. and S. León-Yánez. 1999. Catalogue of the Vascular Plants of Ecuador. St. Louis: Missouri Botanical Garden Press.

Kadereit, J. W., W. Licht, and C. H. Uhink. 2008. Asian relationships of the flora of the European Alps. Plant Ecology and Diversity 1: 171–179. DOI: 10.1080/17550870802328751.

Komarov, V. L. and B. K. Shishkin. 1936. Flora of U.S.S.R. Volume VI: Centrospermae. Landau, N. 1970. English Translation from the Russian. Keter Press: Jerusalem, Israel.

Korb, J.E. and R.Y. Wu. 2011. Fire, climate, and forest health. In: R. Blair and G. Bracksieck. The Eastern San Juan Mountains: Their geology, ecology, and human history. University Press of Colorado: Boulder, Colorado, U.S.A.

Kozhevnikov, Y. P. 1994. Caryophyllaceae. Pp. 10–35 in Plantae Asiae Centralis, vol. 11, ed. V. I. Grubov. Enfield: Science Publishers. 272

Lagomarsino, L. P., E. J. Forrestel, N. Muchhala, and C. C. Davis. 2017. Repeated evolution of vertebrate pollination syndromes in a recently diverged Andean plant clade. Evolution 71: 1970– 1985. DOI: 10.1111/evo.13297.

Lambert, S. M., T. W. Reeder, and J. J. Wiens. 2015. When do species-tree and concatenated estimates disagree? An empirical analysis with higher-level scincid lizard phylogeny. Molecular Phylogenetics and Evolution 82: 146–155. DOI: 10.1016/j.ympev.2014.10.004.

Lange, P. D., P. Heenan, D. Norton, J. Rolfe, and J. Sawyer. 2010. Threatened Plants of New Zealand. Christchurch: Canterbury University Press.

Larsen, K. 2002. Caryophyllaceae. In Flora Malesiana, ser. 1, vol. 16, ed. H. P. Nooteboom. Leiden: National Herbarium of the Netherlands.

Le Roux, J. J., D. Strasberg, M. Rouget, C. W. Morden, M. Koordom, and D. M. Richardson. 2014. Relatedness defies biogeography: the tale of two island endemics (Acacia heterophylla and A. koa). New Phytologist 204: 230–242.

Levin, D. A. 2012. Mating system shifts on the trailing edge. Annals of Botany 109: 613–620. DOI: 10.1093/aob/mcr159.

Li, H. L. 1952. Floristic relationships between eastern Asia and eastern North America. Transactions of the American Philosophical Society 42: 371–429. DOI: 10.2307/1005654.

Li, G-D., C. Kim, H-G. Zha, Z. Zhuo, Z-L. Nie, and H. Sun. 2014. Molecular phylogeny and biogeography of the arctic-alpine genus Lagotis (Plantaginaceae). Taxon 63: 103–115. DOI: 10.12705/631.47.

Li, P., Z-C. Qi, L-X. Liu, T. Ohi-Toma, J. Lee, T-H. Hsieh, C-X. Fu, K. M. Cameron, and Y-X. Qiu. 2017. Molecular phylogenetics and biogeography of the mint tribe Elsholtzieae (Nepetoideae, Lamiaceae), with an emphasis on its diversification in East Asia. Scientific Reports 7: 2057. DOI: 10.1038/s41598-017-02157-6.

Linder, H. P. and P. J. Rudall. 2005. Evolutionary history of Poales. Annual Review of Ecology, Evolution, and Systematics 36: 107–124. DOI: 10.1146/annurev.ecolsys.36.102403.135635.

Lipman, P.W. and W.C. McIntosh. 2011. Tertiary volcanism in the Eastern San Juan Mountains. In: R. Blair and G. Bracksieck. The Eastern San Juan Mountains: Their geology, ecology, and human history. University Press of Colorado: Boulder, Colorado, U.S.A.

Liu, Z-W., D. D. Jolles, J. Zhou, H. Peng, and R. I. Milne. 2014. Multiple origins of circumboreal taxa in Pyrola (Ericaceae), a group with a Tertiary relict distribution. Annals of Botany 114: 1701–1709. DOI: 10.1093/aob/mcu198.

273

Lloyd, D. G. 1965. Evolution of self-compatibility and racial differentiation in Leavenworthia (Cruciferae). Contributions from the Gray Herbarium of Harvard University 195: 3–134. Lloyd, D. G. and D. J. Schoen. 1992. Self- and cross-fertilization in plants. I. Functional dimensions. International Journal of Plant Sciences 153: 358–369.

Löve, Á. and D. Löve. 1974a. Nomenclatural adjustments in the Yugoslavian flora. II. Pteridophytes and Dicotyledons. Preslia 46: 123–138.

Löve, Á. and D. Löve. 1974b. Origin and evolution of the arctic and alpine floras. In J. D. Ives and R. G. Barry (Eds.), Arctic and alpine environments (Pp. 571–603). London: Methuen.

Luo, Y., F.-H. Bian, and Y.-B. Luo. 2012. Different patterns of floral ontogeny in dimorphic flowers of Pseudostellaria heterophylla (Caryophyllaceae). International Journal of Plant Sciences 173: 150–160. DOI: 10.1086/663166.

Luteyn, J. L. 1999. Páramos: A checklist of plant diversity, geographical distribution, and botanical literature. Bronx: The New York Botanical Garden Press.

Mabberley, D. L. 2017. The Plant Book, ed. 4. Cambridge, UK: Cambridge University Press.

Mahdavi, M., M. Assadi, F. Fallahian, and T. Nejadsattari. 2012. The systematic significance of seed micro-morphology in Stellaria L. (Caryophyllaceae) and its closest relatives in Iran. The Iranian Journal of Botany 18: 302–310.

Majumdar, N. C. 2007–onwards. Caryophyllaceae. In eFlora of India. Available from: http://efloraindia.nic.in/efloraindia/taxonList.action?id=456&type=2 (last accessed July 2018).

Malyschev, L. I. and G. A. Peschkova. 2003. Flora of Siberia, vol. 6: Portulacaceae– Ranunculaceae. Enfield: Science Publishers.

Manos, P. S. and J. E. Meireles. 2015. Biogeographic analysis of the woody plants of the Southern Appalachians: Implications for the origins of a regional flora. American Journal of Botany 102: 780–804. DOI: 10.3732/ajb.1400530.

Martin, M. 2011. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet.journal 17: 10–12. DOI: 10.14806/ej.17.1.200.

Massatti, R., A. A. Reznicek, and L. L. Knowles. 2016. Utilizing RADseq data for phylogenetic analysis of challenging taxonomic groups: A case study in Carex sect. Racemosae. American Journal of Botany 103: 337–347. DOI: 10.3732/ajb.1500315.

Matthews, J. V. and L. E. Ovenden. 1990. Late Tertiary plant macrofossils from localities in arctic/subarctic North America: A review of the data. Arctic 43: 364–392.

274

Matzke, N. J. 2013. Probabilistic historical biogeography: New models for founder-event speciation, imperfect detection, and fossils allow improved accuracy and model-testing. Frontiers of Biogeography 5: 242–248. DOI: 10.21425/F5FBG19694.

Maximowicz, C. J. 1873. Diagnoses des nouvelles plantes du Japon et de la Mandjourie. Décade XIV. Bulletin de l'Académie impériale des sciences de St.-Pétersbourg, sér. 3, 18: 371–402.

McDade, L. A. 1992. Hybrids and phylogenetic systematics II. The impact of hybrids on cladistic analysis. Evolution 46: 1329–1346. DOI: 10.1111/j.1558-5646.1992.tb01127.x.

McIver, E. E. and J. F. Basinger. 1999. Early Tertiary floral evolution in the Canadian high Arctic. Annals of the Missouri Botanical Garden 86: 523–545. DOI: 10.2307/2666184.

McNeill, J. 1962. Taxonomic studies in the Alsinoideae: I. Generic and infra-generic groups. Notes from the Royal Botanic Garden Edinburgh 24: 79–155.

Meng, Y., J. Wen, Z-L. Nie, H. Sun, and Y-P. Yang. 2008. Phylogeny and biogeographic diversification of Maianthemum (Ruscaceae: Polygonatae). Molecular Phylogenetics and Evolution 49: 424–434. DOI: 10.1016/j.ympev.2008.07.017.

Miller, M. A., W. Pfeiffer, and T. Schwartz. 2010. Creating the CIPRES Science Gateway for inference of large phylogenetic trees. In: Proceedings of the Gateway Computing Environments Workshop (GCE), 14 Nov. 2010, New Orleans, LA, pp. 1–8.

Miller, C. H. and J. G. West. 2012. A revision of the genus Stellaria (Caryophyllaceae) in Australia. Journal of the Adelaide Botanic Gardens 25: 27–54.

Moeller, D. A. and M. A. Geber. 2005. Ecological context of the evolution of self-pollination in Clarkia xantiana: Population size, plant communities, and reproductive assurance. Evolution 59: 786–799. DOI: 10.1554/04-656.

Molau, U. 1993. Relationships between flowering phenology and life history strategies in tundra plants. Arctic and Alpine Research 25: 391–402. DOI: 10.1080/00040851.1993.12003025.

Molinari-Novoa, E. A. 2016. Further nomenclatural changes regarding Peruvian endemics. Polish Botanical Journal 61: 275–277. DOI: 10.1515/pbj-2016-002.

Morton, J. K. and R. K. Rabeler. 1989. Biosystematic studies on the Stellaria calycantha (Caryophyllaceae) complex. I. Cytology and cytogeography. Canadian Journal of Botany 67: 121–127. DOI: 10.1139/b89-018.

Morton, J. K. 2004. New combinations in North American Caryophyllaceae. Sida 21: 887–888.

Morton, J. K. 2005. Stellaria. Pp. 96–114 in Flora of North America north of Mexico, vol. 5, eds. Flora of North America Editorial Committee. New York: Oxford University Press.

275

Motten, A. F. 1986. Pollination ecology of the spring wildflower community of a temperate deciduous forest. Ecological Monographs 56: 21–42.

Mulch, A. and C. P. Chamberlain. 2006. The rise and growth of Tibet. Nature 439: 670–671. DOI: 10.1038/439670a.

Murray, D. F. 1995. Causes of Arctic plant diversity: Origin and evolution. In S. S. Chapin and C. Körner (Eds.), Arctic and Alpine Biodiversity: Patterns, Causes, and Ecosystem Consequences (Ecological Studies Vol. 113: pp. 21–32). Berlin: Springer. DOI: 10.1007/978-3- 642-78966-3.

Nee, M. and P. M. Jørgensen. 2014. Caryophyllaceae. Pp. 490–497 in Catálogo de las plantas vasculares de Bolivia, vol. 1, eds. P. M. Jørgensen, M. Nee, and S. Beck. St. Louis: Missouri Botanical Garden Press.

Nevski, S. A. 1937. Materials for the flora of Kugitang and its foothills (Materialy k flore Kugitanga i yego predgoriy). Trudy Botanicheskogo Instituta Akademii Nauk SSSR. Ser. 1. Flora i sistematika vysshikh rasteniy 4: 199–346.

Nieto-Montes de Oca, A., A. J. Barley, R. N. Meza-Lázaro, U. O. García-Vázquez, J. G. Zamora-Abrego, R. C. Thomson, and A. D. Leaché. 2017. Phylogenomics and species delimitation in the knob-scaled lizards of the genus Xenosaurus (Squamata: Xenosauridae) using ddRADseq data reveal a substantial underestimation of diversity. Molecular Phylogenetics and Evolution 106: 241–253. DOI: 10.1016/j.ympev.2016.09.001.

O'Dea, A., H. A. Lessios, A. G. Coates, R. I. Eytan, S. A. Restrepo-Moreno, A. L. Cione, L. S. Collins, A. de Queiroz, D. W. Farris, R. D. Norris, R. F. Stallard, M. O. Woodburne, O. Aguilera, M. P. Aubry, W. A. Berggren, A. F. Budd, M. A. Cozzuol, S. E. Coppard, H. Duque- Caro, S. Finnegan, G. M. Gasparini, E. L. Grossman, K. G. Johnson, L. D. Keigwin, N. Knowlton, E. G. Leigh, J. S. Leonard-Pingel, P. B. Marko, N. D. Pyenson, P. G. Rachello- Dolmen, E. Soibelzon, L. Soibelzon, J. A. Todd, G. J. Vermeij, J. B. Jackson. 2016. Formation of the Isthmus of Panama. Science Advances 2(8): e1600883. DOI: 10.1126/sciadv.1600883.

Ornduff, R. 1969. Reproductive biology in relation to systematics. Taxon 18: 121–133.

Paetzold, C., K. R. Wood, D. A. R. Eaton, W. L. Wagner, and M. S. Appelhans. 2019. Phylogeny of Hawaiian Melicope (Rutaceae): RAD-seq resolves species relationships and reveals ancient introgression. Frontiers in Plant Science: 10.1074. DOI: 10.3389/fpls.2019.01074.

Pagel, M. D. 1994. Detecting correlated evolution on phylogenies: a general method for the comparative analysis of discrete characters. Proceedings of the Royal Society B 255: 37–45. DOI: 10.1098/rspb.1994.0006.

Pagel, M. D. 1999. Inferring the historical patterns of biological evolution. Nature 401: 877–884. DOI: 10.1038/44766. 276

Paradis, E. and K. Schliep. 2019. ape 5.0: an environment for modern phylogenetics and evolutionary analyses in R. Bioinformatics 35: 526–528. DOI: 10.1093/bioinformatics/bty633.

Parchman, T. L., Z. Gompert, J. Mudge, F. D. Schilkey, C. W. Benkman, and C. A. Buerkle. 2012. Genome-wide association genetics of an adaptive trait in lodgepole pine. Molecular Ecology 21: 2991–3005. DOI: 10.1111/j.1365-294X.2012.05513.x.

Pavlova, N. S. 1996. 12. Stellaria L. Pp. 65–85 in Plantae Vasculares Orientis Extremi Sovietici, tomus 8, ed. S. S. Charkevicz. Saint Petersburg: Russian Academy of Sciences.

Pax, F. and K. Hoffmann. 1934. 43. Stellaria L. Pp. 320–322 in Die natürlichen Pflanzenfamilien, ed. 2, eds. A. Engler and H. Harms. Leipzig: Engelmann.

Pedersen, T. M. 1961. New species of Hydrocleis, Scirpus and Stellaria. Botanisk Tidsskrift 57: 38–48.

Peng, D-L., X-K. Ou, B. Xu, Z-Q. Zhang, Y. Niu, Z-M. Li, and H. Sun. 2014. Plant sexual systems correlated with morphological traits: Reflecting reproductive strategies of alpine plants. Journal of Systematics and Evolution 52: 368–377. DOI: 10.1111/jse.12046.

Petersen, B. 1977. Pollination of Thlaspi alpestre by selfing and insects in the alpine zone of Colorado. Arctic and Alpine Research 9: 211–215. DOI: 10.1080/00040851.1977.12003915.

Philipp, M. 1980. Reproductive biology of Stellaria longipes Goldie as revealed by a cultivation experiment. New Phytologist 85: 557–569.

Popp, M., V. Mirré, and C. C. Brochmann. 2011. A single Mid-Pleistocene long-distance dispersal by a bird can explain the extreme bipolar disjunction in crowberries (Empetrum). Proceedings of the National Academy of Sciences, USA 108: 6520–6525. DOI: 10.1073/pnas.1012249108.

Pusalkar, P. K. and D. K. Singh. 2015. Taxonomic rearrangement of Arenaria (Caryophyllaceae) in Indian Western Himalaya. The Journal of Japanese Botany 90: 77–91.

R Core Team. 2018–onwards. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Available from: https://www.R-project.org/.

Raduski, A. R., E. B. Haney, and B. Igić. 2011. The expression of self-incompatibility in angiosperms is bimodal. Evolution 66: 1275–1283. DOI: 10.1111/j.1558-5646.2011.01505.x.

Rambaut, A. 2016. FigTree version 1.4.1. Available from: http://tree.bio.ed.ac.uk/software/figtree/.

Rambaut, A. and A. J. Drummond. 2016. TreeAnnotator version 1.8.4. Available from: https://github.com/beast-dev/beast-mcmc/releases/download/v1.10.4/BEAST.v1.10.4.zip. 277

Rambaut, A., A. J. Drummond, D. Xie, G. Baele, and M. A. Suchard. 2018. Posterior summarization in Bayesian phylogenetics using Tracer 1.7. Systematic Biology 67: 901–904. DOI:10.1093/sysbio/syy032.

Randle, A. M., J. B. Slyder, and S. Kalisz. 2009. Can differences in autonomous selfing ability explain differences in range size among sister-taxa pairs of Collinsia (Plantaginaceae)? An extension of Baker’s Law. New Phytologist 183: 618–629. DOI: 10.1111/j.1469- 8137.2009.02946.x.

Raven, P.H. & E.O. Wilson. 1992. A 50-year plan for biodiversity surveys. Science 258:1099- 1110.

Razanajatovo, M., N. Maurel, W. Dawson, F. Essl, H. Kreft, J. Pergl, P. Pyšek, P. Weigelt, M. Winter, and M. van Kleunen. 2016. Plants capable of selfing are more likely to become naturalized. Nature Communications 7: 13313. DOI: 10.1038/ncomms13313.

Rechinger, K. H. 1988. Flora Iranica: Caryophyllaceae, vol. 163. Graz: Akademische Druck- u. Verlagsanstalt.

Renner, S. S. 2016. Available data point to a 4-km-high Tibetan Plateau by 40 Ma, but 100 molecular-clock papers have linked supposed recent uplift to young node ages. Journal of Biogeography 43: 1479–1487. DOI: 10.1111/jbi.12755.

Revell, L. J. 2012. phytools: an R package for phylogenetic comparative biology (and other things). Methods in Ecology and Evolution 3: 217–223. DOI: 10.1111/j.2041- 210X.2011.00169.x.

Richardson, J. E., M. F. Fay, Q. C. B. Cronk, and M. W. Chase. 2003. Species delimitation and the origin of populations in island representatives of Phylica (Rhamnaceae). Evolution 57: 816– 827. DOI: 10.1111/j.0014-3820.2003.tb00293.x.

Rick, C. M. 1988. Evolution of mating systems in cultivated plants. Pp. 133–147 in Plant Evolutionary Biology, Gottlieb, L. D. and S. K. Jain (eds.). Dordrecht: Springer. DOI: 10.1007/978-94-009-1207-6_6.

Ridley, H. R. 1930. The Dispersal of Plants Throughout the World. Ashford, Kent: L. Reeve & Co., Ltd.

Rodríguez-Jiménez, C. 2013. 79. Caryophyllaceae. Online pre-release text of Flora Mesoamericana, vol. 2 (1). Available from: http://www.tropicos.org/docs/meso/caryophyllaceae.pdf (last accessed July 2018).

Rubin, B. E. R., R. H. Ree, and C. S. Moreau. 2012. Inferring phylogenies from RAD sequence data. PLoS ONE 7: e33394. DOI: 10.1371/journal.pone.0033394.

278

Sadeghian, S., S. Zarre, R. K. Rabeler, and G. Heubl. 2015. Molecular phylogenetic analysis of Arenaria (Caryophyllaceae: tribe Arenarieae) and its allies inferred from nuclear DNA internal transcribed spacer and plastid DNA rps16 sequences. Botanical Journal of the Linnean Society 178: 648–669. DOI: 10.1111/boj.12293.

Sauquet, H., M. von Balthazar, S. Magallón, J. A. Doyle, P. K. Endress, E. J. Bailes, E. B. de Morais, K. Bull-Hereñu, L. Carrive, M. Chartier, G. Chomicki, M. Coiro, R. Cornette, J. H. L. El Ottra, C. Epicoco, C. S. P. Foster, F. Jabbour, A. Haevermans, T. Haevermans, R. Hernández, S. A. Little, S. Löfstrand, J. A. Luna, J. Massoni, S. Nadot, S. Pamperl, C. Prieu, E. Reyes, P. dos Santos, K. M. Schoonderwoerd, S. Sontag, A. Soulebeau, Y. Staedler, G. F. Tschan, A. W.-S. Leung, and J. Schönenberger. 2017. The ancestral flower of angiosperms and its early diversification. Nature Communications 8: 16047. DOI: 10.1038/ncomms16047.

Scheen, A-C., C. Brochmann, A. K. Brysting, R. Elven, A. Morris, D. E. Soltis, P. S. Soltis, and A. A. Victor. 2004. Northern Hemisphere Biogeography of Cerastium (Caryophyllaceae): Insights from Phylogenetic Analysis of Noncoding Plastid Nucleotide Sequences. American Journal of Botany 91: 943–952. DOI: 10.3732/ajb.91.6.943.

Scheunert, A. and G. Heubl. 2017. Against all odds: reconstructing the evolutionary history of Scrophularia (Scrophulariaceae) despite high levels of incongruence and reticulate evolution. Organisms Diversity and Evolution 17: 323–349. DOI: 10.1007/s13127-016-0316-0.

Schischkin, B. K. 1936. Stellaria. Pp. 299–326 in Flora of the U.S.S.R., vol. 6, ed. V. L. Komarov. Translated into English by Dr. N. Landau in 1970. Jerusalem: Israel Program for Scientific Translations. Originally published as Flora S.S.S.R. Moscow-Leningrad: Izdatel’stvo Akademii Nauk SSSR. https://www.biodiversitylibrary.org/bibliography/43751.

Schlechtendal, D. F. K. 1816. Ueber die Stellarien und Arenarien Arten, welche in der Willdenowschen Pflanzensammlung aufbewahrt werden. Der Gesellschaft Naturforschender Freunde zu Berlin Magazin für die neuesten Entdeckungen in der Gesammten Naturkunde 7: 190–213.

Schultz, A. 1906. Das Blühen von Stellaria pallida (Dum.). Berichte der Deutschen Botanischen Gesellschaft 24: 245–255. DOI: 10.1111/j.1438-8677.1906.tb06513.x.

Sharples, M. T. 2019. Taxonomic observations within Stellaria L. (Caryophyllaceae): Insights from ecology, geography, morphology, and phylogeny suggest widespread parallelism in starworts and erode previous infrageneric classifications. Systematic Botany, in press.

Sharples, M. T. and E. A. Tripp. 2019a. Phylogenetic relationships within and delimitation of the cosmopolitan flowering plant genus Stellaria L. (Caryophyllaceae): Core stars and fallen stars. Systematic Botany, in press.

Sharples, M. T. and E. A. Tripp. 2019b. RAD sequencing rejects long distance disjunction in Stellaria (Caryophyllaceae) and yields support for a new Southern Rocky Mountains endemic. Taxon. 279

Shilong, C. and R. K. Rabeler. 2001. Stellaria. Pp. 11–29 in Flora of China, vol. 6, eds. W. Zhengyi, P. H. Raven, and H. Deyuan. Beijing: Science Press; St. Louis: Missouri Botanical Garden Press.

Silvertown, J. and D. Charlesworth. 2001. Introduction to plant population biology (4th ed.). Oxford: Blackwell Science.

Sobey, D. G. 1981. Biological flora of the British Isles: Stellaria media (L.) Vill. Journal of Ecology 69: 311–335. DOI: 10.2307/2259833.

Southwest Environmental Information Network, SEINet – Arizona Chapter. 2013–2016. Available at http://swbiodiversity.org/seinet/index.php. Accessed from Dec 2013 to June 2017.

Souza-Neto, A. C., M. V. Cianciaruso, and R. G. Collevatti. 2016. Habitat shifts shaping the diversity of a biodiversity hotspot through time: insights from the phylogenetic structure of Caesalpinoideae in the Brazilian Cerrado. Journal of Biogeography 43: 340–350. DOI: 10.1111/jbi.12634.

Soza, V. L., J. Brunet, A. Liston, P. S. Smith, and V. S. Di Stilio. 2012. Phylogenetic insights into the correlates of dioecy in meadow-rues (Thalictrum, Ranunculaceae). Molecular Phylogenetics and Evolution 63: 180–192. DOI: 10.1016/j.ympev.2012.01.009.

Spigler, R. B. and S. Kalisz. 2017. Persistent pollinators and the evolution of complete selfing. American Journal of Botany 104: 1783–1786. DOI: 10.3732/ajb.1700332.

Sprague, T. A. 1920. L.—Stellaria or Alsine. Bulletin of Miscellaneous Information (Royal Botanic Gardens, Kew) 1920: 308–319. DOI: 10.2307/4107531.

Staats, M., R. H. J. Erkens, B. V. D. Vossenberg, J. J. Wieringa, K. Kraaijeveld, B. Stielow, J. Geml, J. E. Richardson, and F. T. Bakker. 2013. Genomic treasure troves: Complete genome sequencing of herbarium and insect museum specimens. PLoS ONE 8: e69189. DOI: 10.1371/journal.pone.0069189.

Stamatakis, A. 2006. Phylogenetic models of rate heterogeneity: A high performance computing perspective. Pp. 278–286 in Proceedings of the 20th IEEE International Parallel & Distributed Processing Symposium (IPDPS2006). Washington, DC: IEEE Computer Society Press.

Stamatakis, A. 2014. RAxML version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30: 1312–1313. DOI: 10.1093/bioinformatics/btu033. Stebbins, G. L. 1957. Self fertilization and population variability in the higher plants. The American Naturalist 91: 337–354. DOI: 10.1086/281999.

Stebbins, G. L. 1970. Adaptive radiation of reproductive characteristics in angiosperms, I: Pollination mechanisms. Annual Review of Ecology and Systematics 1: 307–326. DOI: 10.1146/annurev.es.01.110170.001515. 280

Takebayashi, N. and P. L. Morrell. 2001. Is self-fertilization an evolutionary dead end? Revisiting an old hypothesis with genetic theories and a macroevolutionary approach. American Journal of Botany 88: 1143–1150. DOI: 10.2307/3558325.

Takhtajan, A. 1986. Floristic regions of the world (English translation from original Russian). Berkeley, California, U.S.A.: University of California Press.

Thiers, B. [continuously updated]. Index Herbariorum: A global directory of public herbaria and associated staff. New York Botanical Garden’s Virtual Herbarium. http://sweetgum.nybg.org/science/ih/.

Tkach, N. V., M. H. Hoffmann, M. Röser, A. A. Korobkov, and K. B. von Hagen. 2008. Parallel evolutionary patterns in multiple lineages of Arctic Artemisia L. (Asteraceae). Evolution 62: 184–198. DOI: 10.1111/j.1558-5646.2007.00270.x.

Tkach, N. V., R. H. Ree, P. Kuss, M. Röser, and M. H. Hoffmann. 2014. High mountain origin, phylogenetics, evolution, and niche conservatism of arctic lineages in the hemiparasitic genus Pedicularis (Orobanchaceae). Molecular Phylogenetics and Evolution 76: 75–92. DOI: 10.1016/j.ympev.2014.03.004.

Tikhomirov, V. N. 2016. (2427) Proposal to conserve the name Stellaria (Caryophyllaceae) with a conserved type. Taxon 65: 389–390. DOI: 10.12705/652.21.

Timaná, M. E. 2017. Nomenclatural notes on the Andean genera Pycnophyllopsis and Pycnophyllum (Caryophyllaceae). Lundellia 20: 4–24. DOI: 10.25224/1097-993X-20.1.4.

Tolmatchev, A. I. 1960. Der Autochthone Grundstock der arktischen Flora und ihre Beziehungen zu den Hochgebirgsfloren Nord- und Zentralasiens. Botanisk Tidsskrift 55: 269–276.

Torrey, J. 1856. No.4: Description of the General Botanical Collections. Pp. 61–182 in Reports of explorations and surveys: to ascertain the most practicable and economical route for a railroad from the Mississippi River to the Pacific Ocean, made under the direction of the Secretary of War, vol. 4, eds. War Department [U.S.], United States Army Corps of Engineers, J. Henry, and S. F. Baird. Washington, D.C.: A. O. P. Nicholson, Printer.

Totland, Ø and B. Schulte-Herbrüggen. 2003. Breeding system, insect flower visitation, and floral traits of two alpine Cerastium species in Norway. Arctic, Antarctic, and Alpine Research 35: 242–247. DOI: 10.1657/1523-0430(2003)035[0242:BSIFVA]2.0.CO;2.

Tripp, E. A. 2016. Is asexual reproduction an evolutionary dead end in lichens? The Lichenologist 48: 559–580. DOI: 10.1017/S0024282916000335.

Tripp, E. A. and P. S. Manos. 2008. Is floral specialization an evolutionary dead end? Pollination system transitions in Ruellia (Acanthaceae). Evolution 62: 1712–1737. DOI: 10.1111/j.1558- 5646.2008.00398.x. 281

Tripp, E. A. and L. A. McDade. 2014. A rich fossil record yields calibrated phylogeny for Acanthaceae (Lamiales) and evidence for marked biases in timing and directionality of intercontinental disjunctions. Systematic Biology 63: 660–684. DOI: 10.1093/sysbio/syu029.

Tripp, E. A., Y. E. Tsai, Y. Zhuang, and K. G. Dexter. 2017. RADseq dataset with 90% missing data fully resolves recent radiation of Petalidium (Acanthaceae) in the ultra-arid deserts of Namibia. Ecology and Evolution 7: 7920–7936. DOI: 10.1002/ece3.3274.

Trucchi, E., B. Frajman, T. H. A. Haverkamp, P. Schönswetter, and O. Paun. 2017. Genomic analyses suggest parallel ecological divergence in Heliosperma pusillum (Caryophyllaceae). New Phytologist 216: 267–278. DOI: 10.1111/nph.14722.

Turland, N. J., J. H. Wiersema, F. R. Barrie, W. Greuter, D. L. Hawksworth, P. S. Herendeen, S. Knapp, W.-H. Kusber, D.-Z. Li, K. Marhold, T. W. May, J. McNeill, A. M. Monro, J. Prado, M. J. Price, and G. F. Smith (eds.). 2018. International Code of Nomenclature for algae, fungi, and plants (Shenzhen Code) adopted by the Nineteenth International Botanical Congress Shenzhen, China, July 2017. Regnum Vegetabile 159. Oberreifenberg: Koeltz Botanical Books.

Turner, B. L. 1991. A new species of Stellaria (Caryophyllaceae) from northeastern Mexico. Phytologia 71: 483–485.

Tzvelev, N. N. 2000. О роде Звездчатка (Stellaria L., Caryophyllaceae) в Восточной Европе. Byulleten' Moskovskogo Obshchestva Ispytatelei Prirody Otdel Biologicheskii 105: 69–72.

USDA, South San Juan Wilderness description. 2017a. www.fs.usda.gov/recarea/sanjuan/recarea/?recid=81038. Accessed May 2017.

USDA, Forest Health Protection And Its Partners. 2017b. https://www.fs.usda.gov/detail/r2/forest-grasslandhealth/?cid=fsbdev3_041629. Accessed May 2017.

USDA, NRCS. 2017c. The PLANTS Database. http://plants.usda.gov. National Plant Data Team, Greensboro, North Carolina, U.S.A. Accessed from May 2013 to Jan 2017.

Valente, L. M., V. Savolainen, and P. Vargas. 2010. Unparalleled rates of species diversification in Europe. Proceedings of the Royal Society B 277: 1489–1496. DOI: 10.1098/rspb.2009.2163.

Villars, D. 1789. Histoire des plantes de Dauphiné, vol. 3 pt. 2. Grenoble: Chez l'auteur.

Vlasova, N. V. 2012. Caryophyllaceae. Pp. 69–92 in Konspekt flory Aziatskoi Rossii: Sosudistye Rasteniya [Synopsis of Flora of Asian Russia: Vascular Plants], ed. K. S. Baikov. Novosibirsk: Sibirskoye Otdeleniye Rossiyskoy Akademii Nauk.

Volponi, C. R. 1993. Stellaria cuspidata (Caryophyllaceae) and some related species in the Andes. Willdenowia 23: 193–209.

282

Wagner, W. L., S. G. Weller, and A. Sakai. 2005. Monograph of Schiedea (Caryophyllaceae- Alsinoideae). Systematic Botany Monographs 72: 1–169.

Walters, S. M. 1964. Caryophyllaceae. In T. G. Tutin, V. H. Heywood, N. A. Burges, D. H. Valentine, S. M. Walters, and D. A. Webb (Eds.), Flora Europaea (Vol. 1, Lycopodiaceae to Platanaceae, pp. 115–204). Cambridge: Cambridge University Press.

Wan, C. and J. Liu. 2017. The morphology and systematics of the pollen of Stellaria. Palynology 41: 533–546. DOI: 10.1080/01916122.2017.1298540.

Wang, L., Y.-Y. Zhao, and J.-X. Liu. 2017. Embryology of Myosoton and Stellaria and its taxonomic significance (Caryophyllaceae). Phytotaxa 306: 124–134. DOI: 10.11646/phytotaxa.306.2.2.

Weber, W. A. 2003. The Middle Asian element in the southern Rocky Mountain flora of the western United States: A critical biogeographical review. Journal of Biogeography 30: 649–685. DOI: 10.1046/j.1365-2699.2003.00864.x

Weber, W.A. and R. C. Wittmann. 1992. Catalog of the Colorado flora: a biodiversity baseline. Boulder, Colorado, U.S.A.: University Press of Colorado.

Weber, W. A. and R. C. Wittmann. 2012a. Colorado flora: eastern slope. 4th Edition. Boulder, Colorado, U.S.A.: University Press of Colorado.

Weber, W. A. and R. C. Wittmann. 2012b. Colorado flora: western slope. 4th Edition. Boulder, Colorado, U.S.A.: University Press of Colorado.

Weeks, A., F. Zapata, S. K. Pell, D. C. Daly, J. D. Mitchell, and P. V. A. Fine. 2014. To move or to evolve: contrasting patterns of intercontinental connectivity and climatic niche evolution in “Terebinthaceae” (Anacardiaceae and Burseraceae). Frontiers in Genetics 5: Article 409. DOI: 10.3389/fgene.2014.00409.

Weller, S. G., A. K. Sakai, A. E. Rankin, A. Golonka, B. Kutcher, and K. E. Ashby. 1998. Dioecy and the evolution of pollination systems in Schiedea and Alsinidendron (Caryophyllaceae: Alsinoideae) in the Hawaiian Islands. American Journal of Botany 85: 1377– 1388.

Wen, J., S. Ickert-Bond, Z-L. Nie, and R. Li. 2010. Timing and modes of evolution of Eastern Asian-North American biogeographic disjunctions in seed plants. In M. Long, H. Gu, and Z. Zhou (Eds.), Darwin’s heritage today: Proceedings of the Darwin 200 Beijing International Conference (Pp. 252–269). Beijing: Higher Education Press.

Wen, J., J-Q. Zhang, Z-L. Nie, Y. Zhong, and H. Sun. 2014. Evolutionary diversifications of plants on the Qinghai-Tibetan Plateau. Frontiers in Genetics 5: article 4. DOI: 10.3389/fgene.2014.00004.

283

Wen, J., Z-L. Nie, and S. M. Ickert-Bond. 2016. Intercontinental disjunctions between eastern Asia and western North America in vascular plants highlight the biogeographic importance of the Bering land bridge from late Cretaceous to Neogene. Journal of Systematics and Evolution 54: 469–490. DOI: 10.1111/jse.12222.

Wessinger, C. A., C. C. Freeman, M. E. Mort, M. D. Rausher, and L. C. Hileman. 2016. Multiplexed shotgun genotyping resolves species relationships within the North American genus Penstemon. American Journal of Botany 103: 912–922. DOI: 10.3732/ajb.1500519.

Wessinger, C., M. D. Rausher, and L. C. Hileman. 2019. Adaptation to hummingbird pollination is associated with reduced diversification in Penstemon. Evolution Letters 3: 521–533. DOI: 10.1002/evl3.130.

Westergaard, K. B., I. G. Alsos, M. Popp, T. Engelskjøn, K. I. Flatberg, and C. Brochmann. 2011. Glacial survival may matter after all: nunatak signatures in the rare European populations of two west-arctic species. Molecular Ecology 20: 376–393. DOI: 10.1111/j.1365- 294X.2010.04928.x.

Western Regional Climate Center. 2017. Wolf Creek Pass Climate Summary. www.wrcc.dri.edu/cgi-bin/cliMAIN.pl?co9181. Last Accessed Jan 2017.

Wiens, J. J. 2003. Missing data, incomplete taxa, and phylogenetic accuracy. Systematic Biology 52: 528–538. DOI: 10.1080/10635150390218330.

Wiens, J. J. and M. C. Morrill. 2011. Missing data in phylogenetic analysis: Reconciling results from simulations and empirical data. Systematic Biology 60: 719–731. DOI:10.1093/sysbio/syr025.

Wild, H. 1961. Caryophyllaceae. In Flora Zambesiaca, vol. 1 pt. 2, eds. A. W. Exell and H. Wild. London: Crown Agents for Overseas Governments and Administrations.

Willis, C. G., Ruhfel, B., Primack, R.B., A.J. Miller-Rushing, and C. C. Davis. 2008. Phylogenetic patterns of species loss in Thoreau’s woods are driven by climate change. Proceedings of the National Academy of Sciences, USA 105:17029–17033.

Wilson, K. L. 2017. Report of the General Committee: 18. Taxon 66: 742–744. DOI: 10.12705/663.15.

Wyatt, R. 1984. The evolution of self-pollination in granite outcrop species of Arenaria (Caryophyllaceae). I. Morphological correlates. Evolution 38: 804–816. Wyatt, R. 1988. Phylogenetic aspects of the evolution of self-pollination. Pp. 109–131 in Plant Evolutionary Biology, Gottlieb, L. D., and S. K. Jain (eds.). Dordrecht: Springer. DOI: 10.1007/978-94-009-1207-6_5.

284

Xing, Y. and R. H. Ree. 2017. Uplift-driven diversification in the Hengduan Mountains, a temperate biodiversity hotspot. Proceedings of the National Academy of Sciences, USA 114: E3444–E3451. DOI: 10.1073/pnas.1616063114.

Zachos, J., M. Pagani, L. Sloan, E. Thomas, and K. Billups. 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science 292 (5517): 686–693. DOI: 10.1126/science.1059412.

Zhang, J-Q., S-Y. Meng, G. A. Allen, J. Wen, and G-Y. Rao. 2014. Rapid radiation and dispersal out of the Qinghai-Tibetan Plateau of an alpine plant lineage Rhodiola (Crassulaceae). Molecular Phylogenetics and Evolution 77: 147–158. DOI: 10.1016/j.ympev.2014.04.013.

Zhang, M., J. Xue, Q. Zhang, and S. C. Sanderson. 2015. Inferring ancestral distribution area and survival vegetation of Caragana (Fabaceae) in Tertiary. Plant Systematics and Evolution 301(7): 1831–1842. DOI: 10.1007/s00606-015-1196-z.

Zhang, M-L., X.-G. Zeng, C. Li, S. C. Sanderson, V. V. Byalt, and Y. Lei. 2017. Molecular phylogenetic analysis and character evolution in Pseudostellaria (Caryophyllaceae) and description of a new genus, Hartmaniella, in North America. Botanical Journal of the Linnean Society 184: 444–456. DOI: 10.1093/botlinnean/box036.

Zhang, R., W. Guo, P. Wang, L. Li, X. Duan, Q. Du, L. Zhao, H. Shan, S. A. Hodges, E. M. Kramer, and H. Kong. 2013. Disruption of the petal identity gene APETALA3-3 is highly correlated with loss of petals within the buttercup family (Ranunculaceae). Proceedings of the National Academy of Sciences, USA 110: 5074–5079.

Zhang, Y. M., A. I. H. Bass, D. Catalina Fernández, and B. J. Sharanowski. 2018. Habitat or temporal isolation: Unraveling herbivore-parasitoid speciation patterns using double digest RADseq. Ecology and Evolution 8: 9803–9816. DOI: 10.1002/ece3.4457.

Zhisheng, A., J. E. Kutzbach, W. L. Prell, and S. C. Porter. 2001. Evolution of Asian monsoons and phased uplift of the Himalaya-Tibetan plateau since Late Miocene times. Nature 411: 62–66. DOI: 10.1038/35075035.

Zoku, H. 1965. 11. Stellaria L. Pp. 428–431 in Flora of Japan (in English), ed. J. Ohwi. Washington, DC: Smithsonian Institution.

285

APPENDICES

Appendix 1. Specimens sampled in Chapter 2 (tip labels bolded, alphabetically organized by genus and species), with collectors and collection numbers, herbarium, and geographical derivations. Herbarium accession numbers are included (if available) when collector number is absent. Adenonema_cherleriae; A. B. Galanin s.n. (VBGI accession 100984), SAMN11982332, Zabaykalsky Krai, Russia. Adenonema_petraeum; Sino-Russian Altai Expedition 2007256 (ALTB), SAMN11982333, Xinjiang, China. Arenaria_lanuginosa; M. Sharples & E. Tripp 1120 (COLO), SAMN11982334, Colorado, USA. Cerastium_beeringianum; M. Sharples 912 (COLO), SAMN09373285, Colorado, USA. Cerastium_fontanum; Y.-Y. Huang 449 (MO), SAMN11982335, Taiwan. Cerastium_parvum; T. Pedersen 9828 (UC), SAMN11982336, Corrientes, Argentina. Cerastium_pauciflorum; M. Sharples & S. Smirnov 1233 (COLO), SAMN11982337, Altai Republic, Russia. Cerastium_sp_Altai; M. Sharples 1212 (COLO), SAMN11982338, Altai Republic, Russia. Holosteum_umbellatum; M. Sharples 1552 (COLO), SAMN11982339, Colorado, USA. Mesostemma_gypsophiloides; T. Nyegayeva s.n. (VLA accession 118560), SAMN11982340, Primorsky Krai, Russia. Mesostemma_martjanovii; M. Sharples & S. Smirnov 1268 (COLO), SAMN11982341, Altai Republic, Russia. Nubelaria_arisanensis; M. T. Kao 7236 (UC), SAMN11982342, Taiwan. Nubelaria_diversiflora; S. Tsugaru & T. Takahashi 29153 (MO), SAMN11982343, Kansai, Japan. Nubelaria_wushanensis_1; Y. N. Xiong 2114 (MO), SAMN11982344, Jiangxi, China. Nubelaria_wushanensis_2; Xu Honggui 230119 (MO), SAMN11982345, Sichuan, China. Pseudostellaria_heterophylla; V. Yakubov s.n. (VLA accession 117596), SAMN11982346, Primorsky Krai, Russia. Pseudostellaria_jamesiana; M. Sharples 525 (COLO), SAMN09373286, Colorado, USA. Rabelera_holostea; Laine & Lempiäinen s.n. (NY), SAMN11982347, Southwest Finland, Finland. Sabulina_macrantha; M. Sharples 619 (COLO), SAMN11982348, Colorado, USA. Silene_baccifera; M. Sharples & S. Huang 1524 (COLO), SAMN11982349, Yunnan, China. Silene_menziesii; M. Sharples 815 (COLO), SAMN11982350, Colorado, USA. Stellaria_alaschanica; Baishuijiang Expedition Team 265 (PE), SAMN11982351, Gansu, China. Stellaria_alaskana; C. Parker 88-94 (COLO), SAMN11982352, Alaska, USA. Stellaria_alsine_NA; D. Poindexter 11-86 (NY), SAMN11982353, North Carolina, USA. Stellaria_alsine_CH; L. Zheng-yu 15566 (CAS), SAMN11982354, Sichuan, China. Stellaria_altimontana; Kharkevich s.n. (VLA accession 117713), SAMN11982355, Khabarovsk Krai, Russia. Stellaria_americana; M. Sharples 1330 (COLO), SAMN11982356, Montana, USA. Stellaria_anagalloides; V. Vašák s.n. (COLO accession 431878), SAMN11982357, Abkhazia, Georgia. Stellaria_angarae; A. Belikova & Roenko s.n. (VBGI accession 101530), SAMN11982358, Zabaykalsky Krai, Russia. Stellaria_angustifolia; M. Sharples & N. Neve 1385 (COLO), SAMN11982359, Australian Capital Territory, Australia. Stellaria_antillana; P. Acevedo-Rdgz. 14206 (NY), SAMN11982360, Santiago, Dominican Republic. Stellaria_aquatica_1; M. Nee 59467 (NY), SAMN11982361, Wisconsin, USA. Stellaria_aquatica_2; M. Sharples & R. Basnet 1509 (COLO), SAMN11982362, Gandaki Pradesh, Nepal. Stellaria_arenarioides; T. Ho & al. 902 (CAS), SAMN11982363, Qinghai, China. Stellaria_arvalis; P. Puchagrán 14302 (MO), 286

SAMN11982364, Los Lagos, Chile. Stellaria_borealis; L. Jørgensen & S. Larsson 66-1545 (NY), SAMN12277172, Greenland. Stellaria_brachypetala; M. Sharples & S. Smirnov 1239 (COLO), SAMN09373288, Altai Republic, Russia. Stellaria_bungeana; M. Sharples & S. Smirnov 1235 (COLO), SAMN09373290, Altai Republic, Russia. Stellaria_calycantha; D. Mansfield 92-461 (NY), SAMN11982366, Oregon, USA. Stellaria_chilensis_1; D. Steward 5-5 (COLO), SAMN11982367, Valparaiso, Chile. Stellaria_chilensis_2; O. Orozco & I. Sanchez- Vega 166 (MO), SAMN11982368, Cajamarca, Peru. Stellaria_chinensis; N. Min-xiang 92081 (NY), SAMN11982369, Jiangxi, China. Stellaria_ciliatosepala; J. Koranda & H. Sherman 23964 (NY), SAMN11982370, Alaska, USA. Stellaria_congestiflora; Edinburgh Expedition to Northern Sikkim 343 (CAS), SAMN11982371, Sikkim, India. Stellaria_corei; W. Dennis 2402 (NY), SAMN11982372, Alabama, USA. Stellaria_crassifolia_1; I. Krasnoborov s.n. (NSK), SAMN11982373, Sakha Republic, Russia. Stellaria_crassifolia_2; M. Sharples 1184 (COLO), SAMN11982374, Finnmark, Norway. Stellaria_crassifolia_3; M. Sharples & S. Smirnov 1254 (COLO), SAMN11982375, Altai Republic, Russia. Stellaria_crassipes; G. Samuelsson s.n. (COLO accession 131698), SAMN11982376, Lappland, Sweden. Stellaria_crispa_US; M. Sharples 1326 (COLO), SAMN11982377, Idaho, USA. Stellaria_crispa_RU; O. Mochalova 37 (VLA), SAMN11982378, Kamchatka Krai, Russia. Stellaria_cuspidata; C. Penland 552 (COLO), SAMN11982379, Bolívar, Ecuador. Stellaria_dahurica; S. Kazanovskii 552 (NSK), SAMN11982380, Republic of Buryatia, Russia. Stellaria_debilis; F. Biganzoli 766 (MO), SAMN11982381, Tierra del Fuego, Argentina. Stellaria_decipiens; C. D. Meurk 85/17 (CHR), SAMN11982382, Antipodes Island, New Zealand. Stellaria_decumbens; D. Boufford & al. 29169 (CAS), SAMN11982383, Yunnan, China. Stellaria_dianthifolia_1; Qingzang Scientific Research Team 12828 (PE), SAMN11982384, Xizang, China. Stellaria_dianthifolia_2; M. Sharples & S. Huang 1532 (COLO), SAMN11982385, Sichuan, China. Stellaria_discolor; V. Verkholat s.n. (VBGI accession 123939), SAMN11982386, Primorsky Krai, Russia. Stellaria_edwardsii; E. Hultén s.n. (NY barcode: 3317780), SAMN11982387, Yukon, Canada. Stellaria_eschscholtziana_1; N. Brummitt & A. Farjon 207 (MO), SAMN11982388, Kamchatka Krai, Russia. Stellaria_eschscholtziana_2; A. Psheninova & Firsov s.n. (VBGI accession 12866), SAMN11982389, Kamchatka Krai, Russia. Stellaria_fenzlii; S. Gage 1909 (NY), SAMN11982390, Kuril, Russia. Stellaria_filicaulis; S. Kharkevich & T. Buch 221 (MO), SAMN11982391, Primorsky Krai, Russia. Stellaria_filiformis; M. Sharples & N. Neve 1391 (COLO), SAMN11982392, Victoria, Australia. Stellaria_fischeriana; Kozhevnikov & Gorikov 19785 (VBGI), SAMN11982393, Kamchatka Krai, Russia. Stellaria_flaccida_1; W. Greuter 21321 (NY), SAMN11982394, New South Wales, Australia. Stellaria_flaccida_2; M. Sharples & N. Neve 1387 (COLO), SAMN11982395, New South Wales, Australia. Stellaria_fontinalis; M. Pyne 92-042 (NY), SAMN11982396, Tennessee, USA. Stellaria_gracilenta; M. Sharples & J. Reid 1398 (COLO), SAMN11982397, Canterbury, New Zealand. Stellaria_graminea; M. Nee 52333 (NY), SAMN09373291, Wisconsin, USA. Stellaria_henryi; Bashan Collection Team 5046 (PE), SAMN11982398, Shaanxi, China. Stellaria_hintoniorum; Hinton & al. 24017 (MO), SAMN11982399, Nuevo Leon, Mexico. Stellaria_howardii; P. Acevedo-Rdgz. 13993 (NY), SAMN11982400, Independencia, Dominican Republic. Stellaria_humifusa_1; N. S. Probaiova 1594 (VLA), SAMN11982401, Sakhalin, Russia. Stellaria_humifusa_2; M. Sharples 1190 (COLO), SAMN11982402, Finnmark, Norway. Stellaria_infracta_1; D. Boufford & al. 38447 (CAS), SAMN11982403, Sichuan, China. Stellaria_infracta_2; Yellow Plateau Team 3024 (MO), SAMN11982404, Shanxi, China. Stellaria_irazuensis; R. Aguilar 3814 (NY), SAMN11982405, San José, Costa Rica. Stellaria_irrigua_1; P. Lesica 10735 (NY), 287

SAMN11982406, Montana, USA. Stellaria_irrigua_2; S. Kharkevich 230 (NY), SAMN11982407, Kamchatka Krai, Russia. Stellaria_irrigua_3; M. Sharples & E. Tripp 1277 (COLO), SAMN11982408, Colorado, USA. Stellaria_irrigua_4; M. Sharples & S. Smirnov 1250 (COLO), SAMN11982409, Altai Republic, Russia. Stellaria_lanata; Dulong Jiang I. T. 2227 (CAS), SAMN11982410, Yunnan, China. Stellaria_lanipes; L. Qin & D. Xiangfu 790520 (KUN), SAMN11982411, Yunnan, China. Stellaria_leptoclada_1; A. Bean 30433 (CANB), SAMN11982412, New South Wales, Australia. Stellaria_leptoclada_2; R. Purdie 9567 (CANB), SAMN11982413, New South Wales, Australia. Stellaria_littoralis_1; A. Eastwood s.n. (NY), SAMN11982414, California, USA. Stellaria_littoralis_2; M. Sharples 1158 (COLO), SAMN11982415, California, USA. Stellaria_longifolia_1; H. H. Iltis & al. 662 (NY), SAMN11982416, Irkutsk, Russia. Stellaria_longifolia_2; S. Welsh & J. Rigby 10662 (NY), SAMN11982417, Yukon, Canada. Stellaria_longifolia_3; M. Sharples 1318 (COLO), SAMN11982418, Colorado, USA. Stellaria_longifolia_4; M. Sharples & E. Boltenkov 1519 (COLO), SAMN11982419, Primorsky Krai, Russia. Stellaria_longifolia_5; X. Lin 6107 (MO), SAMN11982420, Heilongjiang, China. Stellaria_longipes; M. Sharples 1276 (COLO), SAMN09373313, Colorado, USA. Stellaria_mainlingensis; Qingzang Team 4043 (HNWP), SAMN11982421, Xizang, China. Stellaria_mannii; S. Trigui & al. SMT 418 (MO), SAMN11982422, Diana, Madagascar. Stellaria_maximowiczii_1; L. Swan 4 (CAS), SAMN11982423, Province No. 1, Nepal. Stellaria_maximowiczii_2; M. Sharples & S. Huang 1540 (COLO), SAMN11982424, Sichuan, China. Stellaria_media_1; M. Sharples 1280 (COLO), SAMN11982425, Colorado, USA. Stellaria_media_2; M. Sharples & R. Basnet 1515 (COLO), SAMN11982426, Uttarakhand, India. Stellaria_media_3; M. Sharples & S. Huang 1521 (COLO), SAMN11982427, Yunnan, China. Stellaria_monosperma; P. Hua 1598 (KUN), SAMN11982428, Yunnan, China. Stellaria_montioides; W. Sugong 8431 (KUN), SAMN11982429, Yunnan, China. Stellaria_multiflora; H. Eichler 12833 (UC), SAMN11982430, South Australia, Australia. Stellaria_neglecta_1; M. Nee 59115 (NY), SAMN11982431, Ohio, USA. Stellaria_neglecta_2; R. Dale Thomas & al. 88147 (NY), SAMN11982432, Louisiana, USA. Stellaria_nemorum; M. Sharples & R. Ufimov 1169 (COLO), SAMN09373315, Leningrad Oblast, Russia. Stellaria_nepalensis; M. Sharples & R. Basnet 1506 (COLO), SAMN11982433, Province No. 3, Nepal. Stellaria_nipponica; H. Ohba & al. 70756 (MO), SAMN11982434, Chubu, Japan. Stellaria_nitens; P. Leary & al. 4656 (NY), SAMN11982435, Nevada, USA. Stellaria_nubigena; L. Gómez 19872 (MO), SAMN11982436, Cartago, Costa Rica. Stellaria_omeiensis_1; M. Sharples & S. Huang 1523 (COLO), SAMN11982437, Yunnan, China. Stellaria_omeiensis_2; M. Sharples & S. Huang 1526 (COLO), SAMN11982438, Yunnan, China. Stellaria_omeiensis_3; M. Sharples & S. Huang 1529 (COLO), SAMN11982439, Yunnan, China. Stellaria_palustris; E. Koziol 1602 (NY), SAMN11982440, Lower Silesia, Poland. Stellaria_papillata_1; N. Walsh 5816 (MEL), SAMN11982441, Victoria, Australia. Stellaria_papillata_2; D. Osler 123 (MEL), SAMN11982442, Victoria, Australia. Stellaria_parviflora; M. Sharples & J. Reid 1395 (COLO), SAMN11982443, Canterbury, New Zealand. Stellaria_patens_1; First Darwin Nepal Fieldwork Training Expedition 282 (CAS), SAMN11982444, Province No. 1, Nepal. Stellaria_patens_2; M. Sharples & R. Basnet 1513 (COLO), SAMN11982445, Gandaki Pradesh, Nepal. Stellaria_peduncularis; M. Sharples & S. Smirnov 1248 (COLO), SAMN09373316, Altai Republic, Russia. Stellaria_persica_1; T. Gviniashvili 40 (NY), SAMN11982446, Samtskhe- Javakheti, Georgia. Stellaria_persica_2; M. Sharples 1228 (COLO), SAMN11982447, East Kazakhstan, Kazakhstan. Stellaria_petiolaris; D. Boufford & al. 35171 (CAS), 288

SAMN11982448, Yunnan, China. Stellaria_pilosoides_1; T. T. Yü 14104 (KUN), SAMN11982449, Sichuan, China. Stellaria_pilosoides_2; M. Sharples & S. Huang 1528 (COLO), SAMN11982450, Yunnan, China. Stellaria_pilosoides_3; M. Sharples & S. Huang 1530 (COLO), SAMN11982451, Yunnan, China. Stellaria_polyantha; L. Swan 290 (CAS), SAMN11982452, Province No. 1, Nepal. Stellaria_porsildii; Lehto & al. 11484 (NY), SAMN11982453, Arizona, USA. Stellaria_prostrata; Hinton & al. 24112 (NY), SAMN11982454, Nuevo Leon, Mexico. Stellaria_pubera; R. Thomas 156210 (NY), SAMN11982455, North Carolina, USA. Stellaria_pungens; I. Wilson 485 (NY), SAMN11982456, South Australia, Australia. Stellaria_pusilla; Qingzang Team 870451 (KUN), SAMN11982457, Xinjiang, China. Stellaria_radians; B. Legler 672 (NY), SAMN11982458, Sakhalin, Russia. Stellaria_recurvata; D. Neill & QCNE interns 12012 (MO), SAMN11982459, Pastaza, Ecuador. Stellaria_roughii; M. Sharples & J. Reid 1399 (COLO), SAMN11982460, Canterbury, New Zealand. Stellaria_ruscifolia; V. Verkholat 2989 (VBGI), SAMN11982461, Kuril, Russia. Stellaria_salicifolia; D. Boufford 37796 (CAS), SAMN11982462, Gansu, China. Stellaria_sanjuanensis; M. Sharples 1070 (COLO), SAMN11982463, Colorado, USA. Stellaria_sennii; T. Mesfin 5378 (MO), SAMN11982464, Bale, Ethiopia. Stellaria_serpyllifolia; A. Sagástegui & al. 12834 (MO), SAMN11982465, Lambayeque, Peru. Stellaria_sessiliflora; T. Makino 32282 (UC), SAMN11982466, Kanto, Japan. Stellaria_sikkimensis; D. Long & al. 1140 (CAS), SAMN11982467, West Bengal, India. Stellaria_sitchana_1; C. L. Hitchcock 23962 (NY), SAMN11982468, Montana, USA. Stellaria_sitchana_2; B. Ertter 10409 (MO), SAMN11982469, California, USA. Stellaria_cf_sitchana; M. Sharples & J. Reid 1400 (COLO), SAMN11982470, West Coast, New Zealand. Stellaria_soongorica_1; J. Morefield & al. 5267 (CAS), SAMN11982471, Xinjiang, China. Stellaria_soongorica_2; Ubukeeva & Filatova 1 (MO), SAMN11982472, Naryn, Kyrgyzstan. Stellaria_souliei; K. Feng 6851 (PE), SAMN11982473, Yunnan, China. Stellaria_sp_1; M. Sharples & S. Huang 1533 (COLO), SAMN11982474, Sichuan, China. Stellaria_sp_2; M. Sharples & S. Huang 1538 (COLO), SAMN11982475, Sichuan, China. Stellaria_sp_cf_alsine; G. Konyechnaya s.n. (COLO), SAMN11982476, Pskov, Russia. Stellaria_tetrasticha; Smith & Cautivo 10287 (MO), SAMN11982477, Ancash, Peru. Stellaria_tomentosa; M. Hiroe 12917 (UC), SAMN11982478, Kansai, Japan. Stellaria_uchiyamana; S. Tsugaru & Y. Shinno 7139 (MO), SAMN11982479, Chūgoku, Japan. Stellaria_uda; D. Boufford & al. 28625 (CAS), SAMN11982480, Sichuan, China. Stellaria_cf_uda; M. Sharples & S. Huang 1536 (COLO), SAMN11982481, Sichuan, China. Stellaria_venezuelana; G. Davidse & A. González 19543 (MO), SAMN11982482, Sucre, Venezuela. Stellaria_vestita_1; B. Bai & al. 278 (CAS), SAMN11982483, Yunnan, China. Stellaria_vestita_2; Gaoligong Shan Biodiversity Survey 22709 (CAS), SAMN11982484, Yunnan, China. Stellaria_weddellii; S. Beck 17895 (MO), SAMN11982485, La Paz, Bolivia. Stellaria_winkleri_1; L. Anren & Z. Annan 6336 (KUN), SAMN11982486, Xinjiang, China. Stellaria_winkleri_2; S. Wu & al. 2666 (KUN), SAMN11982487, Xinjiang, China.

289

Appendix 2. Specimens of non-Stellaria species sampled as well as of Stellaria specimens newly sampled here for Chapter 3. Collectors and collection numbers, herbarium, GenBank BioSample accession numbers for SRAs PRJNA547948, PRJNA473254, and PRJNA578194, and geographical derivations are given. Herbarium accession numbers are included (if available) when collector number is absent. Adenonema_cherleriae; A. B. Galanin s.n. (VBGI accession 100984), SAMN11982332, Zabaykalsky Krai, Russia. Adenonema_petraeum; Sino-Russian Altai Expedition 2007256 (ALTB), SAMN11982333, Xinjiang, China. Arenaria_lanuginosa; M. Sharples & E. Tripp 1120 (COLO), SAMN11982334, Colorado, USA. Cerastium_beeringianum; M. Sharples 912 (COLO), SAMN09373285, Colorado, USA. Cerastium_fontanum; Y.-Y. Huang 449 (MO), SAMN11982335, Taiwan. Cerastium_parvum; T. Pedersen 9828 (UC), SAMN11982336, Corrientes, Argentina. Cerastium_pauciflorum; M. Sharples & S. Smirnov 1233 (COLO), SAMN11982337, Altai Republic, Russia. Cerastium_sp_Altai; M. Sharples 1212 (COLO), SAMN11982338, Altai Republic, Russia. Gypsophila_sp; M. Sharples & S. Smirnov 1780 (COLO), SAMN13053015, Altai Republic, Russia. Holosteum_umbellatum; M. Sharples 1552 (COLO), SAMN11982339, Colorado, USA. Honckenya_peploides; M. Sharples 1182 (COLO), SAMN13053016, Finnmark, Norway. Mesostemma_gypsophiloides; T. Nyegayeva s.n. (VLA accession 118560), SAMN11982340, Primorsky Krai, Russia. Mesostemma_martjanovii; M. Sharples & S. Smirnov 1268 (COLO), SAMN11982341, Altai Republic, Russia. Nubelaria_arisanensis; M. T. Kao 7236 (UC), SAMN11982342, Taiwan. Nubelaria_diversiflora; S. Tsugaru & T. Takahashi 29153 (MO), SAMN11982343, Kansai, Japan. Nubelaria_wushanensis; Y. N. Xiong 2114 (MO), SAMN11982344, Jiangxi, China. Paronychia_congesta; J. M. Poole 2940 (TEX), SAMN13053017, Texas, USA. Paronychia_herniarioides; J. R. Bozeman 9501 (TEX), SAMN13053018, Georgia, USA. Pseudostellaria_heterophylla; V. Yakubov s.n. (VLA accession 117596), SAMN11982346, Primorsky Krai, Russia. Pseudostellaria_jamesiana; M. Sharples 525 (COLO), SAMN09373286, Colorado, USA. Pycnophyllum_bryoides; S. G. Beck 32419 (TEX), SAMN13053019, Potosí, Bolivia. Rabelera_holostea; Laine & Lempiäinen s.n. (NY), SAMN11982347, Southwest Finland, Finland. Sabulina_macrantha; M. Sharples 619 (COLO), SAMN11982348, Colorado, USA. Silene_baccifera; M. Sharples & S. Huang 1524 (COLO), SAMN11982349, Yunnan, China. Silene_menziesii; M. Sharples 815 (COLO), SAMN11982350, Colorado, USA. Spergularia_rubra; M. Sharples 1361 (COLO), SAMN13053020, Colorado, USA. Stellaria_antillana; P. Acevedo-Rdgz. 14206 (NY), SAMN11982360, Santiago, Dominican Republic. Stellaria_dichotoma; M. Sharples & S. Smirnov 1253 (COLO), SAMN13052021, Altai Republic, Russia. Stellaria_fontinalis; M. Pyne 92-042 (NY), SAMN11982396, Tennessee, USA. Stellaria_howardii; P. Acevedo-Rdgz. 13993 (NY), SAMN11982400, Independencia, Dominican Republic. Stellaria_mannii; S. Trigui & al. SMT 418 (MO), SAMN11982422, Diana, Madagascar. Stellaria_monosperma; P. Hua 1598 (KUN), SAMN11982428, Yunnan, China. Stellaria_fontana; K. H. Rechinger 18536 (K), SAMN13053022, Bamyan, Afghanistan. Stellaria_himalayensis; LKSRD 348 (E), SAMN13053023, Province No. 1, Nepal. Stellaria_reticulivena; H. Hara & al. 2613 (E), SAMN13053024, Gasa, Bhutan. Stellaria_sp_cf_parviflora; D. M. Moore 1943 (K), SAMN13053025, Tierra del Fuego, Argentina. Stellaria_weberbaueri; S. G. Beck 14884 (TEX), SAMN13053026, La Paz, Bolivia.

290

Appendix 3. Coded character states for all Stellaria species included in ancestral reconstructions for Chapter 3. Climate: 0=temperate, 1=cryophilic. Habitat: 0=dry, 1=wet. Area: A, Africa; E, Eastern North America and Central America; H, Pan-Himalaya; M, Maritime Asia; O, Oceania; S, South America; U, Northern Eurasia; W, Western and Northern North America. name climate habitat area Cerastium_sp 0 0 U S_sessiliflora 0 1 M S_radians 0 1 M S_pubera 0 0 E S_corei 0 0 E S_prostrata 0 0 E S_hintoniorum 0 0 E S_venezuelana 0 1 S S_irazuensis 0 0 E S_weddellii 1 0 S S_chilensis 0 1 S S_recurvata 0 1 S S_cuspidata 0 0 S S_arvalis 0 1 S S_sennii 0 0 A S_aquatica 0 1 U S_bungeana 1 0 U S_nemorum 0 1 U S_neglecta 0 0 U S_media 0 0 U S_tetrasticha 1 0 S S_weberbaueri 1 0 S S_nubigena 0 1 E S_cf_alsine 0 1 U S_nitens 0 0 W S_alaskana 1 0 W S_souliei 0 0 H S_montioides 1 0 H S_omeiensis 0 0 H S_dianthifolia 0 0 H S_lanipes 0 0 H S_vestita 0 0 H S_petiolaris 0 0 H S_pilosoides 0 0 H S_patens 0 0 H S_nepalensis 0 0 H 291

S_polyantha 1 0 H S_decumbens 1 0 H S_uda 1 1 H S_maximowiczii 1 0 H S_lanata 0 1 H S_infracta 0 0 H S_sp_HM 1 0 H S_mainlingensis 1 1 H S_arenarioides 1 0 H S_congestiflora 1 0 H S_uchiyamana 0 0 M S_tomentosa 0 0 M S_henryi 0 0 H S_salicifolia 0 0 H S_chinensis 0 0 H S_debilis 1 1 S S_crassifolia 1 1 U, W S_irrigua 1 1 H, M, W S_sanjuanensis 1 0 W S_sikkimensis 0 1 H S_himalayensis 0 0 H S_alaschanica 0 0 H S_reticulivena 0 0 H S_nipponica 1 0 M S_crispa_RU 1 1 M S_sitchana 0 1 W S_crispa 1 1 W S_littoralis 0 0 W S_humifusa 1 1 M, U, W S_calycantha 1 1 U S_fenzlii 1 0 M S_borealis 1 1 W S_alsine 0 1 E, H, M S_flaccida 0 1 O S_parviflora_SA 0 1 S S_roughii 1 0 O S_decipiens 0 1 O S_parviflora 0 0 O S_filiformis 0 0 O S_pungens 0 0 O S_papillata 0 1 O 292

S_angustifolia 0 1 O S_multiflora 0 0 O S_leptoclada 0 1 O S_gracilenta 1 0 O S_brachypetala 1 1 H S_winkleri 1 1 H S_fontana 1 1 H S_pusilla 1 1 H S_discolor 1 1 M S_filicaulis 0 0 M S_longifolia 1 1 M, U, W S_angarae 1 1 M S_peduncularis 1 0 H S_soongorica 1 0 H S_crassipes 1 0 U S_ruscifolia 1 1 M S_eschscholtziana 1 0 M S_longipes 1 0 W S_porsildii 1 0 W S_edwardsii 1 0 M, W S_altimontana 1 0 M S_fischeriana 1 1 M S_dahurica 1 1 H S_graminea 0 0 U S_persica 1 1 U S_palustris 0 1 U S_anagalloides 1 1 U

293

Appendix 4. Coded character states for petaly vs. apetaly in Stellaria for Chapter 4. 0=petals showy, 1=petals absent/plastic/reduced. name petals Cerastium_sp 0 S_sessiliflora 0 S_radians 0 S_pubera 0 S_corei 0 S_prostrata 0 S_hintoniorum 0 S_venezuelana 0 S_irazuensis 0 S_weddellii 0 S_chilensis 0 S_recurvata 0 S_cuspidata 0 S_arvalis 0 S_sennii 1 S_aquatica 0 S_bungeana 0 S_nemorum 0 S_neglecta 0 S_media 1 S_tetrasticha 1 S_weberbaueri 1 S_nubigena 0 S_cf_alsine 1 S_nitens 1 S_alaskana 0 S_souliei 0 S_montioides 0 S_omeiensis 0 S_dianthifolia 0 S_lanipes 1 S_vestita 0 S_petiolaris 0 S_pilosoides 0 S_patens 0 S_nepalensis 0 S_polyantha 1 S_decumbens 1 294

S_uda 0 S_maximowiczii 1 S_lanata 1 S_infracta 0 S_sp_HM 0 S_mainlingensis 0 S_arenarioides 1 S_congestiflora 1 S_uchiyamana 0 S_tomentosa 1 S_henryi 0 S_salicifolia 0 S_chinensis 0 S_debilis 0 S_crassifolia 0 S_irrigua 1 S_sanjuanensis 0 S_sikkimensis 1 S_himalayensis 0 S_alaschanica 0 S_reticulivena 0 S_nipponica 0 S_crispa_RU 1 S_sitchana 1 S_crispa 1 S_littoralis 0 S_humifusa 0 S_calycantha 1 S_fenzlii 0 S_borealis 1 S_alsine 1 S_flaccida 0 S_parviflora_SA 1 S_roughii 0 S_decipiens 1 S_parviflora 1 S_filiformis 1 S_pungens 0 S_papillata 1 S_angustifolia 0 S_multiflora 1 295

S_leptoclada 0 S_gracilenta 0 S_brachypetala 0 S_winkleri 0 S_fontana 0 S_pusilla 0 S_discolor 0 S_filicaulis 0 S_longifolia 0 S_angarae 0 S_peduncularis 0 S_soongorica 0 S_crassipes 0 S_ruscifolia 0 S_eschscholtziana 0 S_longipes 0 S_porsildii 0 S_edwardsii 0 S_altimontana 0 S_fischeriana 0 S_dahurica 0 S_graminea 0 S_persica 0 S_palustris 0 S_anagalloides 0

296

Appendix 5. GenBank BioSample accession numbers for SRAs PRJNA547948, PRJNA473254, and PRJNA578194 for specimens sampled in Chapter 4.

Species Sampled Area GenBank SRA# Stellaria alaschanica Y.Z.Zhao Gansu, China SAMN11982351 Stellaria alaskana Hultén Alaska, U.S.A. SAMN11982352 Stellaria alsine Grimm North Carolina, U.S.A. SAMN11982353 Stellaria sp. cf. alsine Grimm Pskov, Russia SAMN11982476 Stellaria altimontana N.S.Pavlova Khabarovsk Krai, Russia SAMN11982355 Stellaria anagalloides C.A.Mey. ex Rupr. Abkhazia, Georgia SAMN11982357 Stellaria angustifolia Hook. Australian Capital Territory SAMN11982359 Stellaria aquatica (L.) Scop. Gandaki Pradesh, Nepal SAMN11982362 Stellaria arenarioides S.L.Chen, Rabeler & Turland Qinghai, China SAMN11982363 Stellaria arvalis F.Phil. Los Lagos, Chile SAMN11982364 Stellaria borealis Bigelow Greenland SAMN12277172 Stellaria brachypetala Bunge Altai Republic, Russia SAMN11982365 Stellaria bungeana Fenzl Altai Republic, Russia SAMN09373290 Stellaria calycantha (Ledebour) Bongard Oregon, U.S.A. SAMN11982366 Stellaria chilensis Pedersen Cajamarca, Peru SAMN11982368 Stellaria chinensis Regel Jiangxi, China SAMN11982369 Stellaria congestiflora H.Hara Sikkim, India SAMN11982371 Stellaria corei Shinners Alabama, U.S.A. SAMN11982372 Stellaria crassifolia Ehrh. Finnmark, Norway SAMN11982374 Stellaria crassipes Hultén Lappland, Sweden SAMN11982376 Stellaria crispa Cham. & Schltdl. Idaho, U.S.A. SAMN11982377 Stellaria sp. cf. crispa Cham. & Schltdl. Kamchatka Krai, Russia SAMN11982378 Stellaria dahurica Willd. ex Schlecht. Republic of Buryatia, Russia SAMN11982380 Stellaria debilis d’Urv. Tierra del Fuego, Argentina SAMN11982381 Stellaria decipiens Hook.f. Antipodes Island, N.Z. SAMN11982382 Stellaria decumbens Edgew. Yunnan, China SAMN11982383 Stellaria dianthifolia F.N.Williams Sichuan, China SAMN11982385 Stellaria discolor Turcz. Primorsky Krai, Russia SAMN11982386 Stellaria eschscholtziana Fenzl Kamchatka Krai, Russia SAMN11982389 Stellaria fenzlii Regel Kuril, Russia SAMN11982390 Stellaria filicaulis Makino Primorsky Krai, Russia SAMN11982391 Stellaria filiformis (Benth.) Mattf. Victoria SAMN11982392 Stellaria fischeriana Ser. Kamchatka Krai, Russia SAMN11982393 Stellaria flaccida Hook. New South Wales SAMN11982395 Stellaria fontana Popov Bamyan, Afghanistan SAMN13053022 Stellaria gracilenta Hook.f. Canterbury, N.Z. SAMN11982397 Stellaria graminea L. Wisconsin, U.S.A. SAMN09373291 Stellaria henryii F.N.Williams Shaanxi, China SAMN11982398 Stellaria himalayensis Majumdar Province No. 1, Nepal SAMN13053023 Stellaria hintoniorum B.L.Turner Nuevo Leon, Mexico SAMN11982399 Stellaria humifusa Rottbøll Finnmark, Norway SAMN11982402 Stellaria infracta Maxim. Sichuan, China SAMN11982403 Stellaria irazuensis Donn.Sm. San José, Costa Rica SAMN11982405 Stellaria irrigua Bunge Altai Republic, Russia SAMN11982409 Stellaria lanata Hook.f. Yunnan, China SAMN11982410 Stellaria lanipes C.Y.Wu & H.Chuang Yunnan, China SAMN11982411 Stellaria leptoclada C.H.Mill. & J.G.West New South Wales, Australia SAMN11982413 Stellaria littoralis Torr. California, U.S.A. SAMN11982415 Stellaria longifolia Muhl. ex Willd. Colorado, U.S.A. SAMN11982418 Stellaria longipes Goldie Colorado, U.S.A. SAMN09373313 Stellaria edwardsii R.Br. Yukon, Canada SAMN11982387 Stellaria mainlingensis L.H.Zhou Tibet SAMN11982421 Stellaria maximowiczii Kozhevn. Sichuan, China SAMN11982424 Stellaria media (L.) Villars Uttarakhand, India SAMN11982426 297

Stellaria multiflora Hook. South Australia SAMN11982430 Stellaria neglecta Weihe Louisiana, U.S.A. SAMN11982432 Stellaria nemorum L. Leningrad, Russia SAMN09373315 Stellaria nepalensis Majumdar & Vartak Province No. 3, Nepal SAMN11982433 Stellaria nipponica Ohwi Chubu, Japan SAMN11982434 Stellaria nitens Nutt. Nevada, U.S.A. SAMN11982435 Stellaria nubigena Standl. Cartago, Costa Rica SAMN11982436 Stellaria palustris Ehrh. ex Retz. Lower Silesia, Poland SAMN11982440 Stellaria papillata C.H.Mill. & J.G.West Victoria, Australia SAMN11982441 Stellaria parviflora Banks & Sol. ex Hook.f. Canterbury, New Zealand SAMN11982443 Stellaria sp. cf. parviflora Banks & Sol. ex Hook.f. Tierra del Fuego, Argentina SAMN13053025 Stellaria patens D.Don Gandaki Pradesh, Nepal SAMN11982445 Stellaria peduncularis Bunge Altai Republic, Russia SAMN09373316 Stellaria persica Boiss. Samtskhe-Javakheti, Georgia SAMN11982446 Stellaria petiolaris Hand.-Mazz. Yunnan, China SAMN11982448 Stellaria pilosoides S.L.Chen, Rabeler & Turland Yunnan, China SAMN11982451 Stellaria polyantha (Edgew.) M.T.Sharples & E.Tripp Province No. 1, Nepal SAMN11982452 Stellaria porsildii C.C.Chinnappa Arizona, U.S.A. SAMN11982453 Stellaria prostrata Baldwin Neuvo Leon, Mexico SAMN11982454 Stellaria pubera Michx. North Carolina, U.S.A. SAMN11982455 Stellaria pungens Brongn. South Australia SAMN11982456 Stellaria pusilla Em. Schmid Xinjiang, China SAMN11982457 Stellaria radians L. Sakhalin Island, Russia SAMN11982458 Stellaria recurvata Willd. ex Schltdl. Pastaza, Ecuador SAMN11982459 Stellaria reticulivena Hayata Gasa, Bhutan SAMN13053024 Stellaria roughii Hook.f. Canterbury, N.Z. SAMN11982460 Stellaria ruscifolia Willd. ex Schltdl. Kuril, Russia SAMN11982461 Stellaria salicifolia Y.W.Tsui & P. Ke Gansu, China SAMN11982462 Stellaria sanjuanensis M.T.Sharples & E.Tripp Colorado, U.S.A. SAMN11982463 Stellaria sennii Chiov. Bale, Ethiopia SAMN11982464 Stellaria sessiliflora Y.Yabe Kanto, Japan SAMN11982466 Stellaria sikkimensis Hook.f. West Bengal, India SAMN11982467 Stellaria sitchana Steudel California, U.S.A. SAMN11982469 Stellaria soongorica Roshev. Naryn, Kyrgyzstan SAMN11982472 Stellaria souliei F.N.Williams Yunnan, China SAMN11982473 Stellaria tetrasticha (Mattf.) M.T.Sharples & E.Tripp Ancash, Peru SAMN11982477 Stellaria tomentosa Maxim. Kansai, Japan SAMN11982478 Stellaria uchiyamana Makino Chūgoku, Japan SAMN11982479 Stellaria uda F.N.Williams Sichuan, China SAMN11982480 Stellaria venezuelana Steyerm. Sucre, Venezuela SAMN11982482 Stellaria vestita Kurz Yunnan, China SAMN11982484 Stellaria weberbaueri (Muschl.) M.T.Sharples & E.Tripp La Paz, Bolivia SAMN13053026 Stellaria weddellii Pedersen La Paz, Bolivia SAMN11982485 Stellaria winkleri (Briq.) Schischk. Xinjiang, China SAMN11982486

298

Appendix 6. Data derived from our field experiment. Each cell displays total number of fertilized ovules over total number of ovules (fertilized plus unfertilized), organized from lowest to highest seed and ovule number. These fractions were used to derive fertility indices. Bolded cells in the selfing datasets represent fruits bagged with other flowers of an inflorescence.

S. irrigua S. longipes CONTROL SELFING CROSSING CONTROL SELFING CROSSING 3/5 3/6 5/8 8/11 0/11 3/12 3/5 7/7 4/9 5/14 1/12 4/13 3/5 2/8 5/9 6/16 0/15 5/14 5/5 5/8 6/9 6/16 0/15 13/22 2/6 6/8 11/11 4/18 0/15 7/23 3/6 6/8 10/12 9/18 0/16 11/24 3/7 6/8 5/13 4/19 0/17 3/26 3/8 7/8 5/13 5/19 0/17 21/26 3/8 3/9 11/13 7/19 0/17 17/27 4/8 6/9 12/13 5/20 0/17 19/27 5/8 6/9 4/14 6/20 0/18 4/28 6/8 7/9 8/14 2/21 0/18 12/28 8/8 7/9 10/14 9/21 0/18 3/29 7/9 8/9 12/14 15/21 0/18 9/29 8/9 2/10 13/14 18/21 0/19 15/30 2/10 5/10 13/14 6/22 0/19 21/30 6/10 5/10 5/15 7/22 0/20 4/31 7/10 6/10 6/15 7/22 0/20 16/31 5/11 7/10 10/15 8/22 0/21 19/31 6/11 7/10 11/15 10/22 0/21 24/31 6/11 8/10 12/15 15/22 0/21 27/31 7/11 8/10 12/15 4/23 0/21 19/33 8/11 8/10 12/15 6/23 0/21 22/34 9/11 8/10 14/15 11/23 0/21 28/34 11/11 9/10 12/16 11/23 0/21 19/38 4/12 9/10 13/16 11/23 0/21 4/12 10/10 14/16 15/23 0/21 7/12 3/11 9/17 16/23 0/22 7/12 3/11 14/17 5/24 0/22 8/12 4/11 18/18 7/24 0/22 8/12 6/11 20/20 9/24 0/22 8/12 7/11 15/24 0/23 8/12 7/11 17/24 0/23 8/12 7/11 18/24 0/23 9/12 7/11 19/24 0/23 9/12 8/11 4/25 0/24 9/12 8/11 4/25 0/24 10/12 8/11 5/25 0/24 299

10/12 8/11 6/25 0/24 10/12 9/11 6/25 0/24 11/12 9/11 6/25 0/24 11/12 9/11 12/25 0/24 12/12 10/11 15/25 0/25 6/13 10/11 4/26 0/25 6/13 10/11 6/26 0/25 7/13 10/11 6/26 0/25 7/13 5/12 8/26 0/25 7/13 5/12 9/26 0/25 10/13 7/12 11/26 0/25 10/13 9/12 14/26 0/25 11/13 11/12 18/26 0/25 11/13 4/13 20/26 0/25 12/13 4/13 21/26 0/25 12/13 9/13 22/26 0/26 3/14 11/13 8/27 0/26 8/14 11/13 9/27 0/26 8/14 11/13 10/27 0/26 11/14 12/13 10/27 0/26 11/14 12/13 11/27 0/27 11/14 12/13 11/27 0/27 11/14 3/14 13/27 0/27 11/14 6/14 14/27 0/27 11/14 7/14 14/27 0/27 12/14 10/14 15/27 0/27 12/14 11/14 16/27 0/27 12/14 11/14 18/27 0/27 13/14 11/14 19/27 0/27 10/15 11/14 5/28 0/28 10/15 11/14 5/28 0/28 11/15 11/14 5/28 0/28 12/15 12/14 9/28 0/28 13/15 12/14 10/28 0/28 9/16 12/14 10/28 0/28 11/16 12/14 12/28 0/28 11/17 13/14 13/28 0/28 11/17 13/14 16/28 0/28 12/17 6/15 16/28 0/29 13/18 7/15 17/28 0/29 14/18 12/15 18/28 0/29 16/18 12/15 5/29 0/29 7/19 12/15 7/29 0/29 11/19 13/15 7/29 0/29 11/19 13/15 8/29 0/29 300

13/21 14/15 9/29 0/30 14/22 14/15 9/29 0/30 15/15 12/29 0/30 4/16 14/29 0/31 15/16 10/30 0/31 3/17 16/30 0/31 6/17 16/30 0/32 12/17 6/31 0/33 13/17 10/31 0/34 14/17 12/31 14/17 12/31 15/17 16/31 15/17 24/31 15/17 5/32 13/18 10/32 15/18 15/32 7/19 11/33 14/19 9/34 16/20 6/35 16/35 15/36 301

Supplementary Fig. 1. Ancestral area reconstruction of Stellaria inferred with BioGeoBEARS under the DEC+J model. Topology and pie colors are identical to Fig. 18.

302

Supplementary Fig. 2. Ancestral area reconstruction of Stellaria inferred with BioGeoBEARS under the DIVA-like+J model. Topology and pie colors are identical to Fig. 18.

303

Supplementary Fig. 3. Posterior probabilities of branch support within Stellaria and representative of the input tree for ancestral reconstruction analyses of Chapter 4.