SEXUAL DIMORPHISM in the MOSS <Em>BRYUM ARGENTEUM</Em

University of Kentucky UKnowledge

Theses and Dissertations--Biology Biology

2017

SEXUAL DIMORPHISM IN THE MOSS BRYUM ARGENTEUM AND ITS IMPLICATIONS FOR SEX RATIO BIAS

Jonathan David Moore III University of Kentucky, [email protected] Author ORCID Identifier: https://orcid.org/0000-0002-8690-6318 Digital Object Identifier: https://doi.org/10.13023/ETD.2017.240

Right click to open a feedback form in a new tab to let us know how this document benefits ou.y

Recommended Citation Moore, Jonathan David III, "SEXUAL DIMORPHISM IN THE MOSS BRYUM ARGENTEUM AND ITS IMPLICATIONS FOR SEX RATIO BIAS" (2017). Theses and Dissertations--Biology. 43. https://uknowledge.uky.edu/biology_etds/43

This Doctoral Dissertation is brought to you for free and open access by the Biology at UKnowledge. It has been accepted for inclusion in Theses and Dissertations--Biology by an authorized administrator of UKnowledge. For more information, please contact [email protected]. STUDENT AGREEMENT:

I represent that my thesis or dissertation and abstract are my original work. Proper attribution has been given to all outside sources. I understand that I am solely responsible for obtaining any needed copyright permissions. I have obtained needed written permission statement(s) from the owner(s) of each third-party copyrighted matter to be included in my work, allowing electronic distribution (if such use is not permitted by the fair use doctrine) which will be submitted to UKnowledge as Additional File.

I hereby grant to The University of Kentucky and its agents the irrevocable, non-exclusive, and royalty-free license to archive and make accessible my work in whole or in part in all forms of media, now or hereafter known. I agree that the document mentioned above may be made available immediately for worldwide access unless an embargo applies.

I retain all other ownership rights to the copyright of my work. I also retain the right to use in future works (such as articles or books) all or part of my work. I understand that I am free to register the copyright to my work.

REVIEW, APPROVAL AND ACCEPTANCE

The document mentioned above has been reviewed and accepted by the student’s advisor, on behalf of the advisory committee, and by the Director of Graduate Studies (DGS), on behalf of the program; we verify that this is the final, approved version of the student’s thesis including all changes required by the advisory committee. The undersigned agree to abide by the statements above.

Jonathan David Moore III, Student

Dr. D. Nicholas McLetchie, Major Professor

Dr. David F. Westneat, Director of Graduate Studies

SEXUAL DIMORPHISM IN THE MOSS BRYUM ARGENTEUM AND ITS IMPLICATIONS FOR SEX RATIO BIAS

______

DISSERTATION ______

A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biology in the College of Arts and Sciences at the University of Kentucky

By Jonathan David Moore III

Lexington, KY

Director: Dr. D. Nicholas McLetchie, Associate Professor of Biology

Lexington, KY

2017

ABSTRACT OF DISSERTATION

SEXUAL DIMORPHISM IN THE MOSS BRYUM ARGENTEUM AND ITS IMPLICATIONS FOR SEX RATIO BIAS

In dioecious plants, selection due to sex function differences has produced sex- specific life histories, morphologies, and physiologies. In many dioecious seed plants, dimorphisms and population sex ratios have been plausibly linked, but similar links are not yet apparent in dioecious bryophytes. Population sex ratio bias is often expected to favor the sex with lower investment in sexual reproduction, especially in resource-poor environments. Unlike in seed plants, bryophyte males may have higher average reproductive investment than females, which typically have low offspring production rates due to sperm limitation. However, traits aside from reproductive investment such as shoot and leaf arrangement may be differentially selected and could influence life history and sex ratio, but these are rarely tested. My questions concentrated on the dimorphic traits responsible for sex ratio bias and their links to sex function. My studies, using the moss Bryum argenteum, included field and greenhouse experiments investigating sex ratio bias and morphological plasticity along a light/canopy openness (exposure) gradient, a greenhouse comparison of clump morphology and water-holding capacity, and a field and growth chamber study on sex-specific responses to stress (high temperature and desiccation). The sex ratio of urban Lexington, KY was highly female-biased, did not correlate with exposure, and was not linked with pre-zygotic reproductive investment. Leaf characteristics of B. argenteum plastically responded to exposure but were not sex- specific. However, juvenile females produced shoots at a faster rate and grew taller in high light. Juvenile male shoots held more external water than female shoots, but this did not predict mature clump water-holding capacity. Male clumps were shorter, denser, and held less water than females likely to shed sperm-laden water for sexual reproduction. Clump height did not trade off with reproductive investment, adding evidence that sex- specific size is linked with other aspects of sex function. Although chlorophyll fluorescence data (a measure of the status of photosystem II) from both field and growth chamber experiments indicated subtle sex-specific stress recovery responses among sexually immature and mature plants, differences were weaker than predicted and sexually mature shoots did not fare worse than vegetative shoots. The sex differences in size, clump morphology, and clump water-holding capacity very likely affect survival, growth, competitive ability, and ultimately adult sex ratio bias.

KEYWORDS: dioecious, gamete dispersal, divergent selection, stress tolerance, spatial segregation of the sexes

Jonathan David Moore III

May 15, 2017

SEXUAL DIMORPHISM IN THE MOSS BRYUM ARGENTEUM AND ITS IMPLICATIONS FOR SEX RATIO BIAS

Jonathan David Moore III

D. Nicholas McLetchie, PhD Director of Dissertation

David F. Westneat, PhD Director of Graduate Studies

May 15, 2017

ACKNOWLEDGEMENTS

I would like to recognize many folks for their assistance in completing this

dissertation. For their support, I would like to thank my parents, Elizabeth Tyler Moore

and Jon Moore. My mother especially fostered my love for plants from the earliest stages

of my life and has encouraged me throughout my academic career. My wife, Sarah

Moore, has been a great source of support, a confidant, and more than patient during the

final stages of research and writing. I could not have asked for a better advisor than

Nicholas McLetchie, and I thank him for the years of training he has provided me, his encouragement, his patience, and the many laughs. I thank my committee members,

Scott Gleeson, Tim Phillips, and Phil Crowley for helping me plan my research and for providing feedback on manuscript drafts. I also thank Pradeep Kachroo for serving as my outside examiner for my dissertation defense. I am grateful to Alexandra Hurst for helping develop the stock of plants used for my research, and my undergraduate students

John Krzton-Presson, Taylor Walker, and Jennifer Davis for help in data collection and preliminary experiments. I offer a special thanks to my collaborators and some of the finest researchers I know, REU students Morgan Franke and Leslie Kollar, who received funding from NSF REU DBI-1062890 (PI Westneat). My fellow lab members, Chris

Stieha, Deric Miller, Alexandra Hyatt, and Rose Marks have been wonderful to work with and have provided helpful discussion of ideas and feedback on manuscripts. I also want to thank fellow members of the “plant group,” Carol Baskin, Megan Seifert, and

Jim Shaffer, for their constructive criticism of my work. Chapter Two (unpublished) and

Five (published) were improved with comments of four and two anonymous reviewers,

respectively. Amanda Ensminger, Craig Sargent, Jeremy Van Cleve, Joshua Lambert,

III and Isaac Nakamura provided assistance with statistics. I am grateful for the funding I received from the Biology Department’s Getrude Flora Ribble fund to conduct and present my research. I also thank the Graduate School and the American Bryological and

Lichenological Society for travel funding. I thank the College of Agriculture for greenhouse space, and David Westneat, Craig Sargent, and Scott Gleeson for use of lab equipment. Beverly Taulbee and Jacqueline Burke were invaluable for helping me navigate through graduate school. My teaching advisors, Andy Bouwma-Gearhart,

Jennifer Osterhage, Kay Shenoy, and Carol Baskin, have each uniquely contributed to my abilities as a teacher. To all others that I owe a debt of gratitude but have failed to name, my apologies and thank you. Finally, I must thank God for bringing each of the individuals above and many more into my life and for the myriad of blessings and opportunities that have helped me reach this point.

TABLE OF CONTENTS

Acknowledgements ...... III List of Tables ...... VIII List of Figures ...... X

CHAPTER ONE: INTRODUCTION ...... 1

CHAPTER TWO: SEXUAL DIMORPHISM ACROSS LAND PLANTS Abstract ...... 6 Introduction ...... 7 Sex ratio and sexual dimorphism ...... 11 Life history ...... 15 Angiosperms ...... 17 Gymnosperms ...... 20 Bryophytes ...... 22 Morphology...... 25 Angiosperms ...... 27 Gymnosperms ...... 28 Bryophytes ...... 29 Physiology...... 30 Angiosperms ...... 32 Gymnosperms ...... 37 Bryophytes ...... 39 Conclusion and future directions ...... 41

CHAPTER THREE: THE EFFECTS OF AN URBAN EXPOSURE GRADIENT ON LEAF MORPHOLOGY, PIGMENT PHYSIOLOGY, AND POPULATION SEX RATIO OF BRYUM ARGENTEUM HEDW. (BRYACEAE) Abstract ...... 46 Introduction ...... 47 Methods...... 50 Study species ...... 50 Collection procedures and site characterization ...... 51 Characterization of light environment and canopy openness ...... 52 Assessment of morphological traits on field collected plants ...... 53 Pigment extraction and quantification ...... 55 Assessment of life history, morphological plasticity, and sex ratio in the greenhouse ...... 55 Statistical analyses ...... 57 Results ...... 61 Field exposure ...... 61 Sex ratio ...... 61 Leaf morphology in the field ...... 63 Pigment content in the field ...... 68 Morphology and growth in a greenhouse common garden ...... 71 Relationships between traits in the field and common garden ...... 77 Discussion ...... 78 Field sex ratio patterns ...... 79 Sex-differences in common garden ...... 80 Variation among habitats in the relationship between morphology and exposure

V in the field ...... 81 Overall morphology and growth responses to exposure in the field and light intensity in common garden ...... 82 Pigment responses to light intensity in the field ...... 86 Implications of genetic differentiation ...... 89 Conclusions ...... 91

CHAPTER FOUR: DO SEX DIFFERENCES IN RESPONSE TO HEAT AND DESICCATION STRESS OCCUR IN THE MOSS BRYUM ARGENTEUM HEDW. (BRYACEAE)? Abstract ...... 93 Introduction ...... 94 Methods...... 96 Study species ...... 96 Field collection procedures and habitat characterization ...... 97 Recovery of field-collected shoots from desiccation ...... 98 Recovery of field-collected shoots from wet heat ...... 99 Stress responses of shoots from growth-chamber-grown sexually immature clumps ...... 99 Stress responses of shoots from growth-chamber-grown sexually mature clumps ...... 100 Statistical analyses ...... 101 Results ...... 103 Field studies ...... 103 Growth-chamber-grown sexually immature clumps ...... 107 Growth-chamber-grown sexually mature clumps ...... 112 Discussion ...... 122 Sex-specific responses to heat stress and desiccation ...... 123 Comparing stage-specific and sex-specific responses ...... 126 Responses of Bryum argenteum to desiccation and heat ...... 127 Conclusions ...... 128

CHAPTER FIVE: DOES SELECTION FOR GAMETE DISPERSAL AND CAPTURE LEAD TO A SEX DIFFERENCE IN CLUMP WATER-HOLDING CAPACITY? Abstract ...... 130 Introduction ...... 131 Methods...... 134 Study species ...... 134 Field collections ...... 135 Experimental growing conditions ...... 136 Assessing clump characteristics and water-holding capacity ...... 136 Shoot external water-holding capacity...... 138 Extrapolating water-holding capacity from the shoot level to the clump level ...138 Statistical analyses ...... 139 Results ...... 141 Clump characteristics and water-holding capacity ...... 141 Shoot external water-holding capacity...... 143 Extrapolated clump water-holding capacity ...... 144 Discussion ...... 145 Divergent selection ...... 146 Stage effects, water-holding capacity, and clump morphology ...... 146 Conceptual connection to seed plants ...... 150 Implications for other species of bryophytes ...... 151 Consequences for sex ratio ...... 152 Potential new insights for arthropod sperm dispersers ...... 154 VI

Conclusions ...... 155

CHAPTER SIX: CONCLUSIONS AND FUTURE DIRECTIONS Sex differences ...... 156 Sexually dimorphic traits among sexually immature plants ...... 158 Ecology ...... 160

APPENDIX 1: Chlorophyll a:b ratios for three Populus species ...... 163

APPENDIX 2: Chlorophyll and carotenoid extraction and quantification methods ...... 164

REFERENCES ...... 166

VITA ...... 187

VII

LIST OF TABLES

Table 2.1. Summary of patterns of population sex ratios, life history traits, and growth forms across relevant land plant divisions ...... 17 Table 2.2. Summary of morphological patterns and growth forms across relevant land plant divisions ...... 26 Table 2.3. Summary of physiological patterns and growth forms across relevant land plant divisions ...... 32 Table 3.1. Mean exposure ± SE for each habitat ...... 61 Table 3.2. Habitat and overall sex ratios for Bryum argenteum as proportion male with associated standard errors ...... 63 Table 3.3. Habitat trait means of Bryum argenteum from the field and trait least square means from the greenhouse common garden with associated standard errors ...... 66 Table 3.4. Partial correlation coefficients of field leaf morphological variables of Bryum argenteum ...... 68 Table 3.5. Results of the doubly multivariate analysis for greenhouse morphology and growth of Bryum argenteum ...... 72 Table 3.6. Correlation coefficients for pairwise correlations among leaf morphology and measures of growth of Bryum argenteum in the greenhouse common garden ...... 77 Table 4.1. Results from the repeated measures analysis for recovery of photosystem II using Fv/Fm from desiccation of field-collected females and males of Bryum argenteum ...... 105 Table 4.2. Results of repeated measures analysis for responses to wet hot conditions of field-collected females and males of Bryum argenteum ...... 106 Table 4.3. Repeated measures analysis of the responses to wet cool, dry cool, wet hot, and dry hot conditions of female and male shoots of Bryum argenteum from growth-chamber-grown sexually immature clumps ...... 108 Table 4.4. Repeated measures analysis of the responses to dry and wet hot alone of female and male shoots of Bryum argenteum from growth-chamber-grown sexually immature clumps ...... 111 Table 4.5. Repeated measures analysis of responses to wet cool, dry cool, dry hot, and wet hot treatments of female and male vegetative and sex-expressing shoots of Bryum argenteum from growth-chamber-grown sexually mature clumps ...... 116 Table 4.6. Repeated measures analysis of responses to dry cool, dry hot, and wet hot treatments of female and male vegetative and sex-expressing shoots of Bryum argenteum from growth-chamber-grown sexually mature clumps ...... 119 Table 4.7. Repeated measures analysis of responses to dry cool, dry hot, and wet hot treatments of female and male sex-expressing shoots alone of Bryum argenteum from growth-chamber-grown sexually mature clumps ...... 121 Table 4.8. Repeated measures analysis of responses to dry cool, dry hot, and wet hot treatments of female and male vegetative shoots alone of Bryum argenteum from growth-chamber-grown sexually mature clumps ...... 122 Table 5.1. Least squares means ± SE of clump morphological characteristics and proportion of shoot apices expressing sex of Bryum argenteum ...... 143

VIII

Table 5.2. Partial correlation coefficients based on the residuals from the MANOVA on isolated shoot water capacity, extrapolated clump water-holding capacity, actual clump water-holding capacity, morphological characteristics, and proportion of shoot apices expressing sex ...... 143 Table A1.1. Chlorophyll a:b ratios calculated for three Populus species ...... 163

LIST OF FIGURES

Figure 2.1. Divergent selection’s effects upon life history, morphology, and physiology with links between traits and their effect on sex ratios ...... 13 Figure 3.1. Photo of typical habitat of Bryum argenteum taken at the outdoor amphitheater behind Memorial Hall on the University of Kentucky campus ...... 51 Figure 3.2. Drawing of a Bryum argenteum leaf showing how morphological measurements were made ...... 54 Figure 3.3. The influence of exposure and five habitat types on relative achlorophyllous area, leaf length:width ratio, and relative apex length of Bryum argenteum in the field ...... 65 Figure 3.4. The influence of exposure and five habitat types on total chlorophyll per dry mass and total carotenoids per dry mass of Bryum argenteum in the field ...... 70 Figure 3.5. Means ± SE by habitat origin of high and low light intensity effects on leaf area, leaf length:width ratio, and relative apex length of Bryum argenteum in the greenhouse common garden ...... 73 Figure 3.6. Mean ± SE by sex of high and low light intensity effects on shoot length, three week shoot production, and three month shoot density of Bryum argenteum in the greenhouse common garden ...... 75 Figure 3.7. Mean ± SE of relative apex length of Bryum argenteum in the greenhouse common garden and field...... 85 Figure 4.1. Photosystem II recovery using Fv/Fm after rehydration of desiccated, field-collected females and males of Bryum argenteum ...... 105 Figure 4.2. Photosystem II responses (Fv/Fm) to wet hot conditions of field-collected females and males of Bryum argenteum ...... 107 Figure 4.3. Photosystem II recovery using Fv/Fm of female and male shoots from growth-chamber-grown sexually immature clumps of Bryum argenteum that were hydrated or rapid dried then exposed to wet cool, dry cool, dry hot, or wet hot conditions for 1 h ...... 109 Figure 4.4. Shoot production of sexually immature males and females of Bryum argenteum after initial hydration or rapid dry and exposure to wet cool, dry cool, dry hot, or wet hot conditions for 1 h ...... 112 Figure 4.5. Photosystem II recovery using Fv/Fm of sex-expressing shoots from growth-chamber-grown females and males of Bryum argenteum that were hydrated or rapid dried then exposed to wet cool, dry cool, dry hot or wet hot conditions for 1 h ...... 114 Figure 4.6. Photosystem II recovery using Fv/Fm of mature vegetative shoots from growth-chamber-grown females and males of Bryum argenteum that were hydrated or rapid dried then exposed to wet cool, dry cool, dry hot, or wet hot conditions for 1 h ...... 115 Figure 4.7. Photosystem II recovery using Fv/Fm of sex-expressing and vegetative shoots of growth-chamber-grown Bryum argenteum clumps without respect to sex that were hydrated or rapid dried then exposed to wet cool alone and pooled dry cool, dry hot, and wet hot treatments for 1 h ...... 118 Figure 5.1. Violin plot and overlaying box plot of isolate means of actual clump

water-holding capacity in a 1 cm2 area of Bryum argentuem females and males ...... 142

Figure 5.2. Violin plot and overlaying box plot of isolate means of external water-holding capacity of isolated young shoots of Bryum argenteum females and males standardized by their length ...... 144 Figure 5.3. Violin plot and overlaying box plot of isolate means of extrapolated clump water-holding capacity in a 1 cm2 area of Bryum argenteum females and males ...... 145 Figure 5.4. Photographs of a representative female and male of Bryum argenteum ...... 150

CHAPTER ONE: INTRODUCTION

In dioecious plants, differences between the sexes in sexual reproductive function has led to the evolution of sexually dimorphic traits (Delph 1999; Barrett and Hough 2013).

Though these dimorphic traits are frequently subtle, and thus went unrecognized much longer than in animal systems (Willson 1991), they can lead to skewed population sex ratios (Delph 1999; Barrett and Hough 2013). Understanding the causes of skewed population sex ratios is important for dioecious species of conservation concern and for predicting adaptive potential of populations to environmental change (e.g. Ellstrand and

Elam 1993; Iszkuło et al. 2009; Wall et al. 2013) via the effects of sex ratio on effective population sizes (Sinclair et al. 2012). Much previous research has sought, often successfully, to link sex differences in reproductive investment to sexually dimorphic traits and skewed sex ratios (i.e. higher investment leading to lower representation; Delph

1999; Barrett and Hough 2013). However, many species of plants in all major divisions possess dimorphic traits and have skewed population sex ratios that do not intuitively link to sex differences in total reproductive investment, using carbon as the currency (e.g.

Stark et al. 2000; Gauquelin et al. 2002; Harris and Pannell 2008).

In Chapter Two, I used a comparative approach across the major divisions of land plants (angiosperms, gymnosperms, and bryophytes), focusing on species with genetically determined unisexuality, to understand more fully the patterns and causes of sexual dimorphism in life history, morphology, and physiology and their effects on population sex ratios. Most studies on sexual dimorphism in dioecious plants have focused on one division while ignoring or making cursory reference to other divisions

(e.g. McLetchie 1992; McLetchie and Puterbaugh 2000). Because angiosperms are the

most studied, I used their patterns as a comparative scaffold upon which to map trends in

the other two divisions. From this approach, a clear correlation between life histories,

sex ratios and growth habit emerged, which suggests that the causes of sexual

dimorphism and sex ratio are similar between shrubby angiosperms and shrubby

gymnosperms and between herbaceous (clonal and abiotically pollinated) angiosperms

and bryophytes. These correlations have not, to my knowledge, been previously found

and permitted the prediction of expected patterns and their causes in these groups.

One of these predictions is that female-biased sex ratios in both herbaceous angiosperms and bryophytes, which may be exacerbated by abiotic stress, is related to higher realized male reproductive investment compared to females, which was previous hypothesized for bryophytes (McLetchie 1992; Stark et al. 2000; Stark et al. 2005).

Higher male reproductive investment can occur in the former through nitrogen expenditure on reproduction (as opposed to the typical measured carbon expenditure;

Harris and Pannell 2008) and in the later via low fertilization rates for females and higher male total pre-zygotic reproductive investment (e.g. McLetchie 1992; Stark et al. 2000).

Thus, bryophyte studies can increase understanding of female-biased, herbaceous angiosperm systems. In Chapter Three, I used the dioecious moss Bryum argenteum, where males have higher pre-zygotic reproductive investment than females (Horsley et al.

2011), to test whether males become rarer as exposure (light intensity and canopy openness) increases in the field. Through field and greenhouse experiments on the same isolates, I also quantified leaf morphology, pigment content, and growth, hypothesizing that stressful conditions would amplify sex differences. Sex ratio was female-biased but uncorrelated with exposure, and is more likely explained by larger size and faster shoot

production of females compared to males. Male rarity prevented assessment of field sex

differences, and morphology was not sex-specific in the greenhouse. The morphology of

B. argenteum is light-reflective at high exposure, which is similar to seed plant

morphological strategies, and also probably prevents damage during desiccation. Indeed,

plants are reflective enough for high light plants to compensate by increasing

photosynthetic pigment production. Finally, traits were genetically differentiated among

habitats, suggesting adaptation to novel anthropogenic microhabitats as shown for other

bryophytes and flowering plants (e.g. Hassel et al. 2005; Brzyski et al. 2014; Marks et al.

2016; Jordan 1992; Linhart and Grant 1996; Wittig 2004; Cheptou et al. 2008).

In Chapter Four, using field-collected and cultured plants in four experiments, I explicitly tested the hypothesis that females are more stress tolerant than males by subjecting B. argenteum juvenile, vegetative, and sex-expressing male and female shoots to heat and desiccation stress, tracking recovery through time using chlorophyll fluorescence to measure potential quantum efficiency of photosystem II (Fv/Fm).

Traditionally, sexually dimorphic traits in plants have been thought to arise via reproductive tradeoffs only after sexual maturity (Dawson and Geber 1999; Case and

Ashman 2005). My studies provide evidence to the contrary because cultured juveniles were subtly dimorphic (females recover faster than males at 24 h) in response to the greatest stress (wet heat). This dimorphic response, however, did not translate into higher growth rate for females compared to males after wet heat stress, so a statement of which sex is more stress tolerant is not possible. The strongest sex differences were found among cultured sex-expressing shoots with females showing faster recovery under the greatest stress (wet and dry heat). Cultured vegetative shoots from mature clumps were

much less dimorphic, which may relate to developmental trajectory (i.e. low dimorphism

if sex-expression is not imminent) or currently low resource expenditure (i.e. slow growth

rate and/or no investment in sexual reproduction). Physiological sex-differences could be

manifested during periods of high resource expenditure for juvenile (high growth rate)

and sex-expressing shoots (reproduction). Sex-differences in juveniles may result from genetic correlations. There were no sex-differences evident among field-collected, stress- hardened plants.

Though sex-specific stress tolerance among single shoots are as yet not strongly linked to biased sex ratios, bryophytes, particularly mosses, rarely exist as single shoots.

Because aggregations of shoots (clumps) are arguably the functional unit in nature and are likely single-genotype, they are subject to selection. As with branch or canopy morphology in seed plants (Niklas 1985 a, b), bryophyte clump morphology can affect abiotic dispersal or capture of male gametes (i.e. mating success), and thus may be sexually dimorphic. In Chapter Five, using greenhouse-grown males and females, I examined whether sex differences in clump morphology and water-holding characteristics in B. argenteum were consistent with divergent selection for relatively low male water-holding capacity for gamete dispersal and high female capacity for gamete capture. I hypothesized that differences in water-holding capacity would not be due to individual juvenile shoot-level characteristics, but rather to mature clump morphology. I used juvenile shoot water-holding capacity to predict clump capacity, and compared predicted with actual clump capacity. Juvenile male shoots held more water per unit length, and male clumps had higher shoot density, which extrapolated to higher predicted clump water-holding capacity for males compared to females. However, female clumps

4 held more water and were taller with more robust shoots. Actual clump capacity correlated positively with clump height and shoot cross-sectional area. The sex difference in actual clump capacity and its unpredictability from juvenile shoots are consistent with my hypothesis that males hold less water than females to facilitate sexual reproduction. These results provide conceptual connections to other plant groups and implications for connecting divergent selection to female-biased sex ratios in B. argenteum and other bryophytes, including the potential for advantages in overgrowth, growth rates, or preparation for desiccation for females compared to males.

In Chapter Six, I explore the implications of my research for sexual dimorphism and skewed sex ratios in dioecious bryophytes and other plants. I also suggest future directions for research that would illuminate when sexual dimorphism arises in plants and the selection pressures that have selected for dimorphic traits.

CHAPTER TWO: SEXUAL DIMORPHISM ACROSS LAND PLANTS

Abstract

A significant number of land plants are dioecious and sexually dimorphic, but causes of sexual dimorphism across the major divisions of land plants remain unclear with most studies focusing on one division, ignoring or making cursory reference to other divisions.

In this review, I use a comparative approach to understand more fully the patterns and causes of sex differences in life history, morphology, and physiology in all land plants with genetically determined unisexuality. These plants include members of the angiosperm, gymnosperm, and bryophyte divisions. Because angiosperms are the most studied, I use their patterns as a scaffold upon which to map the other two divisions. This integrative approach reinforces the emerging pattern that sex ratios and life histories correlate with growth habit, suggesting fruitful analogies between 1) woody angiosperms and gymnosperms (including shrubby angiosperms and conifers), and 2) between herbaceous (clonal and abiotically pollinated) angiosperms and bryophytes. These novel correlations lead to predictions on the causes of sexual dimorphism and biased population sex ratios across these major divisions of land plants. Further, these predictions can focus research efforts in areas where knowledge is scarce—in morphology and physiology in gymnosperms and bryophytes and in below-grown morphology and physiology in all divisions.

KEY WORDS: dioecious; growth habit; life history; morphology; physiology

Introduction

Darwin’s publication of The Descent of Man and Selection in Relation to Sex

(1871) sparked great interest in sexual dimorphism, and although the causes of sexual

dimorphisms in animals are well understood, the causes of sexual dimorphism across

major divisions of land plants are not. Much work in animals participating in obvious

male-male competition or female mate choice has focused on sexual selection to explain

conspicuous sexual dimorphism. More recently, sexual selection was proposed to occur

in sedentary animals with external fertilization (Levitan 2010). Because sexual selection

does not clearly explain all dimorphism, animal biologists have proposed other

hypotheses (Shine 1989; Vollrath 1998). For example, sexual selection is often

implicated for greater male size compared to females (Darwin 1871; Searcy 1979; Shine

1979; Berry and Shine 1980; Weckerly 1998), but fecundity selection is favored to

explain larger female size (Shine 1988; Head 1995; Reeve and Fairbairn 1999; Kupfer

2009).

With much internal and external controversy, plant biologists applied sexual

selection to plants beginning in the 1970’s (Charnov 1979; Willson 1979). The delay and

controversy were perhaps because dimorphisms are often less pronounced in plants

(Willson 1991) than in animals and thus more difficult to define. Many studies focused on sexual selection of floral morphology and displays, especially male fitness through larger displays (see Andersson and Iwasa 1996), but some have explored sexual selection’s effects on other sexual dimorphisms such as age to maturity, stress tolerance, metabolic rate, nitrogen use efficiency, meristem production and conversion, and canopy morphology (e.g. Willson 1991; Vasiliauskas and Aarssen 1992; Dawson and Geber

1999). Sexual selection via intrasexual competition for mates causes adjustments to life

histories, morphologies, or physiologies to maximize mating success. Usually,

discussions of sexual selection in plants focus less on trade-offs between mating success and survival than in animal systems (Moore and Pannell 2011).

Commonly, plant biologists emphasize hypotheses such as the cost of sexual reproduction (CoS) to explain sexual dimorphisms. The CoS hypothesis posits that differential reproductive investment causes plant sexual dimorphisms (e.g. Ornduff 1996;

Dawson and Geber 1999; Delph 1999; Stark et al. 2000; Bisang and Ehrlén 2002; Delph et al. 2005), assuming one sex (usually females) invests more in sexual reproduction than the other (usually males). That is, fecundity selection dictates life history, morphological, or physiological adjustments by one or both sexes due to tradeoffs. These adjustments may lead to survival costs or sex-specific niche partitioning (Bierzychudek and Eckhart

1988; Dawson and Geber 1999; Obeso 2002; Barrett and Hough 2013).

Lloyd and Webb (1977) define secondary sexual characters (equivalent to sexual

dimorphism in this review) in flowering plants as those outside either the gynoecium or

the androecium, female and male sex organs respectively, which include both juvenile

and post-reproductive traits. Dimorphic floral traits, bryophyte perianths, or

gymnosperm cone scales, which protect sex organs or directly facilitate sexual

reproduction, would be among these but have been reviewed (mainly in angiosperms)

elsewhere (see Willson 1979, 1991; Andersson and Iwasa 1996; Eckhart 1999). This review focuses on secondary sexual characteristics outside these protective or facilitative structures across all land plants, which include sex differences in life histories, morphologies, and physiologies.

Plants have a variety of sexual systems (hermaphroditism, monoecy, dioecy,

gynodioecy, androdioecy etc.) and sex determining mechanisms (chromosomal, non-

chromosomal genetic, labile sex, incomplete sex separation etc., Meagher 1988). For

simplicity I focus on plants with (confirmed or assumed) genetically determined dioecy

which is still a large number of plant species. Approximately 6% of angiosperms (Sakai

and Weller 1999), 36% of conifers and nearly all species of other gymnosperm lineages

(Givnish 1980; Leslie et al. 2013), and more than 58% of bryophytes (Wyatt and

Anderson 1984) are dioecious. I group bryophytes together because of similarities in life

cycle, growth habit, and ecology but recognize their phylogenetic status is debated (Cox

et al. 2014; Ruhfel et al. 2014). Sex determination is chromosomal in bryophytes (Allen

1917; Cameron and Wyatt 1990; Tanurdzic and Banks 2004) and chromosomal, non- chromosomal genetic, and environmental in other lineages (Lee 1954; Abraham and

Mathew 1962; Tanurdzic and Banks 2004; Gangopadhyay et al. 2007). I draw analogies between gametophyte-dominant bryophytes and sporophyte-dominant seed plants (Jesson and Garnock-Jones 2012) (see Box 2.1 for dioecy vs. dioicy). Discussion of ferns is excluded here because sex in their gametophytes is environmentally or sporophytically determined and sporophytes are sexless (Tanurdzic and Banks, 2004).

Box 2.1. Dioecious vs. Dioicous Both dioecious and dioicous mean “two houses” from the same Greek roots with the former passing into English from Greek via Latin transliteration. However, some researchers have insisted on their being only applied to specific groups (Wyatt 1985). Dioecious has been applied to plant species that have physiologically independent sporophytes (sporophyte-dominant) that produce microspores or megaspores on separate individuals, consequently individuals are either male or female. Dioicous has been used by some to describe plant species that have physiologically independent gametophytes (gametophyte- dominant) that produce antheridia and archegonia on separate individuals (note: by definition all seed plants are dioicous, having male and female gametophytes, i.e. pollen or embryo sacs, respectively). To maintain the analogy between all groups of land plants and for simplicity, I use dioecious to refer to any plant species with unisexual individuals in its dominant life stage.

I have used the work in angiosperms, the most studied division of land plants, as a scaffold upon which to map gymnosperms and bryophytes. Sexual dimorphism has been well reviewed in angiosperms (e.g. Lloyd and Webb 1977; Willson 1991; Dawson and

Geber 1999; Delph 1999; Geber 1999; Barrett and Hough 2013), but reviews in other lineages are lacking. I provide an integrated review that synthesizes the work done throughout the land plants and makes ideas accessible to researchers focusing on only one group. Interestingly, Lloyd and Webb (1977) referenced bryophytes in their seminal work on sexually dimorphic traits in plants, but few seed plant authors have continued in their footsteps by following the bryophyte literature on that subject. Bryologists usually only make cursory use of literature on sexually dimorphic traits and their ecological correlates in angiosperm and gymnosperm and often fail to explore similarities between these divisions and bryophytes (e.g. McLetchie 1992; McLetchie and Puterbaugh 2000;

Stark et al. 2005b; Horsley et al. 2011; but see Holá et al. 2014).

I touch on how sex ratio patterns in the three divisions can indicate sexual dimorphism, and then I review relevant secondary sexual dimorphism in 1) life histories,

2) morphologies, and 3) physiologies. In each section, I begin with angiosperms and then move across the taxonomic tree to gymnosperms and end with bryophytes. I draw parallels between angiosperms and other plant taxa while discussing trends and future directions. I highlight cases where gymnosperms and bryophytes may be well suited to answer questions that are unanswered in angiosperms and vice versa. I show the comparability of 1) woody angiosperms and gymnosperms (including between shrubby species), 2) herbaceous clonal angiosperms (especially those with abiotic pollination and seed dispersal) and bryophytes, and 3) wind pollinated seed plants and water fertilized bryophytes.

Sex ratio and sexual dimorphism

Because of the inextricable linkage to sexual dimorphism in many but not all cases, I discuss sex ratio throughout this review. In this section, I provide background on the link between sex ratio bias and sexual dimorphism. More detail on the types of dimorphism leading to a particular sex ratio bias is provided in the sections following this one.

Population sex ratio bias varies in direction and extent within and among major plant divisions. However, I note, because dimorphic species can have unbiased sex ratios

(Delph 1999), an absence of sex ratio bias does not indicate a lack of sexual dimorphism.

Seed or spore sex ratio or germination bias is an important cause of sex ratio bias in plants (e.g. McLetchie 1992; Taylor 1994) and must be ruled out before using sex ratio

11 bias of adult population as evidence for life history dimorphism. If adult population sex ratio is skewed more than offspring sex ratio, dimorphic life histories such as differences in survival, competitive ability, sex expression, or growth rate are likely causes. These life history differences are potentially caused in turn by sexually dimorphic morphologies or physiologies, and all are ultimately caused by divergent selection pressures (see Figure

2.1). However, if dimorphisms are absent, then seeking bias in seeds, spores, or germination rates is warranted to account for adult sex ratio bias. If male and female plants germinate at a sex ratio of unity, a predominantly female-biased population should have dimorphic traits allowing females to survive over males (Gehring and Linhart 1993;

Harris and Pannell 2008). The opposite is true for male-biased populations where the effect of offspring sex ratio has been accounted for. Species may also exhibit spatial segregation of the sexes (populations or subpopulations dominated by one sex) through niche partitioning (Sinclair et al. 2009) or differences in competitive ability (Mercer and

Eppley 2010). Therefore, similar patterns of population sex ratio among different species lead to the hypothesis of similar sexual dimorphisms among these species with similar causes, which will be outlined in relevant sections below.

Adult Sex Ratios

Sex-Specific Life Histories

Sex-Specific Sex-Specific Morphologies Physiologies

Divergent Selection (Natural and/or Sexual)

Figure 2.1. Divergent selection’s effects upon life history, morphology, and physiology with links between traits and their effect on sex ratios. Black arrows indicate selection upon traits (ultimate causes), and gray arrows indicate influences of traits on other traits and adult population sex ratios (proximate causes).

In angiosperms, male bias in population sex ratio is twice as common as female bias (Field et al. 2013). Male bias is associated with longer-lived growth forms (woody species, i.e. trees) and with biotic pollination and seed dispersal (Geber 1999; Obeso

2002; Barrett et al. 2010; Field et al. 2013). One can hypothesize that these traits lead to a sex-specific CoS (females pay a higher cost due to fruit and seed production), causing lower female survival, especially in aging populations (Allen and Antos 1993). In other angiosperm growth forms, female bias is associated with clonality in herbaceous plants and with abiotic pollen and seed dispersal in both herbaceous plants and shrubs (Sinclair et al. 2012; Field et al. 2013). Field et al. (2013) found that nonclonal herbaceous angiosperm sex ratios did not differ from 1:1, but they indicate the result may be due to low levels of abiotic pollination (9 of 33 species) in their data set. Abiotic pollination in herbaceous species could increase male investment in sexual reproduction (Dawson and

Geber 1999), leading to lower male survival and female-biased populations.

Sex ratios among gymnosperms are frequently male-biased, following the patterns of long-lived woody angiosperms, especially among cycads (Ornduff 1985, 1987;

Grobbelaar et al. 1989; Ornduff 1990; Tang 1990; Ornduff 1991; Vasiliauskas and

Aarssen 1992; Ward 2007) and in Welwitschia (Henschel and Seely 2000), but there are

exceptions. Some cycads have 1:1 sex ratios (Ornduff 1987; Watkinson and Powell

1997), although to my knowledge, a female-biased sex ratio has not been reported in cycads. However, some conifers have 1:1 or female-biased sex ratios (e.g. Ortiz et al.

1998; Gauquelin et al. 2002; Hilfiker et al. 2004; Quinn and Meiners 2004; Sousa et al.

2004; Iszkuło and Boratyński 2011). The cases of female-biased gymnosperms may be

comparable to female bias among angiosperm shrubs with wind pollination and abiotic

seed dispersal (Sinclair et al. 2012). All the cases cited above of 1:1 or female-biased sex

ratios in gymnosperms except one (Sousa et al. 2004) were of species that are large shrubs or small trees, most having wind pollination though with biotic seed dispersal.

Variation in sex ratios within species occurs among conifers (Quinn and Meiners 2004;

Nanami et al. 2005; Ward 2007; Kang and Shin 2012), with males often predominating at stressful sites (Lawton and Cothran 2000; Ortiz et al. 2002; Nuñez et al. 2008; Iszkuło et al. 2009). These patterns suggest dimorphic traits in many gymnosperms will resemble

patterns in male-biased angiosperms (i.e. those leading to lower female survival).

Bryophytes mostly have female-biased population sex ratios (Bisang and Hedenäs

2005) but see (Wyatt 1977; Cameron and Wyatt 1990; Alvarenga et al. 2013; Holá et al.

2014), similar to clonal herbaceous angiosperms with abiotic pollination and seed

dispersal. Herbaceous habit, clonality, and primarily abiotic fertilization (arthropod

augmentation has been found in two species, Rosenstiel et al. 2012) and spore dispersal

in bryophytes (Slack 2011) are likely contributing factors causing female-biased sex

ratios in bryophytes.

Despite observed female-biased population sex ratios, a 1:1 spore and subsequent germination sex ratio is expected for bryophytes. Meiotic sex determination leaves little time for differential investment in male and female spores (Hedenäs and Bisang 2011), although it remains untested in most species (but see Allen 1919; Stark et al. 2010). Sex- specific germination (i.e. spore abortion or differential investment) occurs in bryophytes mostly to a female advantage (Newton 1972; Longton and Greene 1979; Ramsay 1979;

McLetchie 1992; Shaw and Gaughan 1993; McLetchie 2001) and rarely male advantage

(Newton 1972). Germination patterns can vary by population (Shaw and Beer 1999),

owing probably to genetic distortion of spore sex ratio (McDaniel et al. 2007).

Nevertheless, adult sex ratios can become more biased than at germination through life

history differences such as sex-specific mortality (McLetchie 1992). In summary,

population sex ratios are generally male-biased in woody angiosperms and many

gymnosperms and female-biased in herbaceous angiosperms and bryophytes (see Table

2.1). Adult sex ratios can serve as evidence for post-germination life history differences.

Life history

Life history theory is based on the idea that organisms partition finite amounts of

resources to competing traits and events (i.e. trade-offs). Selection acts upon these allocation patterns to maximize fitness (Delph 1999). Trade-offs with reproduction

(CoS) are common explanations for sex-specific life history differences. Even if initial male reproductive investment is higher (e.g. flower and inflorescence size or number plus pollen), females generally incur greater costs because pollen (or sperm in bryophytes)

demands fewer resources than eggs, seeds, and fruits (or sporophytes in bryophytes)

(Lloyd and Webb 1977; Delph 1999; Stark et al. 2000; Obeso 2002). Life histories affect

population dynamics and genetic variation (Eriksson 1989) and ultimately the likelihood

of species or population persistence (Grime 1977, 1979). Sex-specific life histories can

lead to sex-specific population dynamics, skewed sex ratios, and spatial segregation of

the sexes (Delph 1999; Bowker et al. 2000; McLetchie et al. 2002). Life history patterns in long-lived woody angiosperms are fairly well-established, although patterns in female- biased shrubs are not. In herbaceous angiosperms and bryophytes, patterns are emerging, but are not, in my opinion, as firmly established as in long-lived woody species. In gymnosperms, life history patterns are relatively understudied, but are expected to follow patterns in woody angiosperms. These patterns along with growth form and sex ratio trends are summarized in Table 2.1.

Table 2.1. Summary of patterns of population sex ratios, life history traits, and growth forms across relevant land plant divisions.

Pre- Average Growth Earliest zygotic realized Taxa Sex Ratio Size rate maturity RI Total RI RI Survival Angio. Woody Ma Ma Ma Ma Ma Fa Fa Ma Herb. Fa Fa Fb c Ma Mb c Fa Fa Fb c Gymno. Conifers Vb c Mb c Mb c Mb c U Fb c Fb c U Cycads Mb U U Mb c Vb Fb Fb U Bryo. Fa Fa Fb c Mb c Mb Fa Md U Note: The category “woody angiosperm” does not take into account female-biased shrubs with abiotic pollination and seed dispersal due to lack of data. RI = sexual reproductive investment. Survival does not account for mortality due to abiotic stress (see Table 2.3). Male > female is represented by an M and female > male is represented by an F. V represents “variable,” meaning the pattern varies between species, environmental conditions, or reproductive status and requires further investigation. U represents “undetermined,” meaning I am not aware of any studies demonstrating a difference between the sexes. a Pattern cited in previous reviews or other articles b To my knowledge, a pattern reviewed here c Few studies but with clear pattern d I know of only one study

Angiosperms—Angiosperms exhibit sex differences in timing of flowering, individual

size, and longevity (Delph 1999). The last two correlate with plant form: woody vs.

herbaceous (Harris and Pannell 2008; Hesse and Pannell 2011; Sinclair et al. 2012;

Barrett and Hough 2013) or clonal herbaceous (Field et al. 2013). The connection

between CoS, life history, and sex ratio bias is apparent for predominantly male-biased,

long-lived, woody angiosperms. Typically, among woody angiosperms, males flower

sooner, at smaller size, or more frequently (Faliński 1980; Allen and Antos 1993; Delph

1999), and invest more pre-zygotically in sexual reproduction than females (Cruden and

Lyon 1985), which is tangible evidence of lower cumulative stored resource costs

because of flower-only investment. Alternatively, sexual selection can cause male

17 precociousness (i.e. maturation at a smaller size; Delph 1999; Geber 1999), which, if not costly, can lead to (observationally) male-biased sex ratios (Nicotra 1998). If precociousness reduces survival, older populations will be female-biased (for an animal example see Myers 1984). In contrast, females must limit survival cost by growing large enough to support flowers and starch or lipid rich fruit and seeds, which require greater resources such as stored carbon compared to relatively inexpensive pollen (Faegri and van der Piji 1979; Wallace and Rundel 1979; Case and Ashman 2005). One caveat is that inflorescence structures such as calyces and fruits can photosynthesize to offset expenditure partially (Obeso 2002; Case and Ashman 2005; Gehring and Delph 2006).

Female resource expenditure often manifests as lower fecundity the next season (i.e. fewer or no flowers), growth rate, survival, and smaller overall sizes than males with differences in life history sometimes exacerbated by stress (see Physiology), leading to male-biased population sex ratios. Frequently, sex-specific life histories appear only after maturity, which supports and explains the popularity of the CoS hypothesis (Lloyd and Webb 1977; García and Antor 1995; Delph 1999; Obeso 2002; Barrett and Hough

2013; Munné-Bosch 2015).

However, in shrubs with abiotic pollination and fruit dispersal, female-biased population sex ratios are unexplained by the life-history patterns detailed above. Sinclair et al. (2012) suggested that inbreeding in these shrubby species leads to the evolution of female-biased sex ratios. At least among some Salix species with mixed insect and wind pollination, there is evidence that seed sex ratio is female-biased (Ueno et al. 2007;

Myers-Smith and Hik 2012; Che-Castaldo et al. 2015), and some show similar sexual dimorphisms to male-biased trees such as higher female reproductive biomass (Sakai et

al. 2006) and higher male growth under water limitation (Dudley 2006), indicating that

sexual dimorphism is not the causes of adult sex ratio bias. However, females of other

Salix species grow at the same rate as males (Sakai et al. 2006) or are more likely to

suffer lower mortality due to herbivory (Boecklen et al. 1990) and high UV stress

(Randriamanana et al. 2015). Data on seed sex ratios from other shrubby species would

permit the testing of this hypothesis further. If seed sex ratios do not predict adult

population sex ratios, then dimorphic life histories are the likely causes of the biases.

Perhaps pollen dispersal, limited due to height of these wind-pollinated shrubs (Niklas

1985), prevents females from fully realizing their full reproductive costs thereby

increasing growth or survival.

In predominantly female-biased herbaceous species, the connection between CoS, life history, and sex ratio is also less discernible. As expected, in herbaceous species,

males mature earlier (Putwain and Harper 1972; Delph 1999; Harris and Pannell 2008),

have higher pre-zygotic investment (Fox and Harrison 1981; Gross and Soule 1981;

Korpelainen 1992; Delph et al. 1993; Gehring and Linhart 1993), and invest less biomass

in reproduction than females (Obeso 2002; but see Delph et al. 1993; Sánchez Vilas and

Pannell 2010). In contrast to woody angiosperms, herbaceous females commonly grow

larger (Bram and Quinn 2000; Obeso 2002; Field et al. 2013), can grow faster (Doust and

Doust 1987), are better competitors (Doust et al. 1987; Eppley 2006), and have higher

survival (Putwain and Harper 1972; Stehlik and Barrett 2005; but see Meagher and

Antonovics 1982). Water stress may also cause higher female survival (Decker and

Pilson 2000) or growth rate (Liu and Duan 2013), but there are exceptions (e.g. Houssard et al. 1992). In Mercurialis annua (dioecious when diploid, monoecious or

androdioecious when polyploid, Pannell 1997), males are smaller than females but have similar reproductive effort in terms of biomass. However, males allocate more nitrogen

(N) (absolutely and relatively) to inflorescences and less to total vegetative tissue than females (Harris and Pannell 2008), which plays a key role in the size dimorphism (see

Physiology for more detail). Yet there are currently too few studies on too few herbaceous angiosperm species to consider these patterns general.

Gymnosperms—Many gymnosperm life history differences are comparable to those in woody angiosperms. Males are typically precocious in conifers (Nanami et al. 2005;

Iszkuło and Boratyński 2011), cycads (Clark and Clark 1987; Tang 1990; Watkinson and

Powell 1997), and Welwitschia (Henschel and Seely 2000). Among cycads, total pre- fertilization female strobili mass may be greater than (Tang 1990) or smaller than

(Ornduff 1996) male strobili mass. However, females have higher total reproductive investment after seed (with fleshy sarcotesta) maturity (Ornduff 1985, 1987; Clark and

Clark 1988; Ornduff 1996), the cost of which reduces female leaf production compared to males in Zamia integrifolia (Ornduff 1996). Though infrequently tested in conifers (e.g.

Ortiz et al. 2002; Rovere et al. 2003; Ward 2007), total female sexual reproductive

investment, due to larger female cones, is expected to exceed male investment (Biswas and Johri 1997). Males grow larger (Vasiliauskas and Aarssen 1992; Iszkuło et al. 2009) or faster than females in conifers (Rovere et al. 2003; Montesinos et al. 2006; Nuñez et al. 2008; Cedro and Iszkuło 2011; Iszkuło and Boratyński 2011) and Welwitschia

(Henschel and Seely 2000). Conifer population sex ratios can reflect stress-induced

higher male growth or survival (i.e. stress-induced life history differences, Nuñez et al.

2008; Iszkuło et al. 2009; but see Marion and Houle 1996), which is likely the result of similar reproductive investment patterns as woody angiosperms.

If these sex differences arise after sexual maturity (i.e. multiple reproductive bouts), then CoS hypothesis is supported (Cipollini and Whigham 1994; Delph 1999;

Stark et al. 2000), which is generally the case. However in the conifer, Juniperus virginiana, juvenile growth rate is higher for males than females (Quinn and Meiners

2004), which is evidence of genetic correlation between juvenile and adult characteristics. That is, larger male size could be selected for rather than simply resulting from lower reproductive costs. Vasiliauskas and Aarssen (1992) suggest sexual selection for larger size (canopy emergence) to increase pollen dispersal. Alternatively, slower juvenile female growth rate could be favored to store resources for greater reproductive investment at adulthood, but these ideas need to be tested. Juvenile characteristics could contribute more to adult dimorphism than is currently understood and data on these would differentiate between dimorphism originating through direct tradeoffs with reproduction or differential development. The development of genetic tools for sex identification or long-term tracking of juvenile individuals of long-lived species through maturity is necessary to answer this question.

Male-biased population sex ratios are often explained by earlier maturity or greater male survivorship in harsh or resource-poor environments due to lower stored resource demand. The cases of 1:1 population sex ratios could reflect a lack of sex- specific advantage under favorable conditions, but female-biased sex ratios require more explanation. As mentioned previously of shrubby female-biased angiosperms, seed or seedling sex ratios and life-history dimorphism bias adult sex ratios. In J. thurifera,

female bias is hypothesized to result from greater female nutrient recycling via cone

decomposition (versus male nutrient export via pollen dispersal) leading to relatively

higher growth or survival in females compared to males (Gauquelin et al. 2002). This explanation resembles the one for female-biased sex ratio in the herbaceous angiosperm,

Mercurialis annua. In cases of female-biased sex ratios in angiosperms and gymnosperms, sexually selected elevated pollen production (male-male competition and high nutrient export) may also increase male growth and ultimately survival costs especially if combined with reduced average female reproductive investment via low seed production or seed abortion (Ortiz et al. 1998; Ward 2007; Harris and Pannell 2008).

Data is needed for these species on reproductive investment of non-carbon resources, whether these resources are recycled, and prevalence of pollen limitation, which will inform research seeking dimorphic life histories in female-biased angiosperm shrubs.

Bryophytes—In bryophytes, many sex-specific life histories have been documented and can lead to female dominance. I note that while 40% of hornworts are dioecious

(Villarreal and Renner 2013) with evidence of spatial segregation of the sexes (Renzaglia and McFarland 1999), all bryophyte sexual dimorphism studies are from mosses and liverworts. As in angiosperms and gymnosperms, some moss males express sex earlier, at smaller size, and more prolifically than females (Shaw and Beer 1999; Horsley et al.

2011 and references therein). Male pre-zygotic sexual reproductive investment in mosses exceeds that of females by as much as six (Stark et al. 2000) or 24 times (Horsley et al.

2011), but total investment (including gametophyte-dependent sporophytes) is higher for females (Bowker et al. 2000; Stark 2002; Horsley et al. 2011), which is consistent with

22 angiosperms (Antos and Allen 1994; Nicotra 1999; Sanchez Vilas and Pannell 2011) and gymnosperms. Sporophytes, analogous to seed plant female reproductive structures, can photosynthesize to offset their costs in mosses (Krupa 1969; Proctor 1977), hornworts

(Thomas et al. 1978), and some liverworts (Thomas et al. 1979). Alternatively, total female investment and costs may be lower because of frequent sporophyte abortion. For example, Syntrichia caninervis females aborting all sexual offspring express sex more frequently than males, indicating a tradeoff between current reproductive investment and future sex expression (Stark et al. 2000; Stark et al. 2001). In Ceratodon purpureus, males but not females exhibit a tradeoff between sexual and vegetative tissue investment

(McDaniel 2005). If females have higher overall reproductive costs and most females produce sporophytes, then, according to the CoS hypothesis, adult sex ratios should be male-biased (for a study meeting the expectation see Alvarenga et al. 2013). However, as in herbaceous (especially clonal or wind pollinated) angiosperms, female-bias is common in bryophytes. Therefore, differences in expected total sexual reproduction investment

(based on biomass) are insufficient to explain sex ratio patterns in bryophytes and herbaceous angiosperms (see also Physiology).

Consistent with herbaceous angiosperm patterns, where size dimorphisms occur, female bryophytes are often larger (McLetchie 1992; Shaw and Gaughan 1993; Bowker et al. 2000; Glime and Bisang 2014). At the extreme, some bryophyte males are epiphytic dwarfs upon females (seemingly obligatory and genetically fixed in some species), which could be caused by conflicting forces between cytoplasmic (maternal) and nuclear (maternal and paternal) genomes (Hedenäs and Bisang 2011). Female-biased adult sex ratios are common among these taxa (Ramsay and Berrie 1982). Another

possible explanation, even in less size-dimorphic species such as Sphaerocarpos texanus,

is that males have evolved for coexistence closer to females (McLetchie 1992; Hedenäs

and Bisang 2011) as suggested in some animals (Vollrath 1998). While proximity is

useful for sperm dispersal in bryophytes, it increases male-female competition. To avoid fitness reductions through competition with mates, considering sperm dispersal distance limits access to mates further away where direct competition can be avoided, males may senesce early or be poorer competitors than females. A trait that simultaneously increases fitness but decreases survival bears marks of sexual selection and may be especially advantageous for males of annual plants where there is no potential for future reproductive opportunity. Alternatively, females could actively suppress males (Une

1985; McLetchie 1992). These are fascinating possibilities for those female-biased angiosperms with growth forms similar to bryophytes.

Female growth can be higher in bryophytes. Females have been found to regenerate better from detached leaves than males in one moss, potentially causing higher female growth during colonization or after damage (Stark et al. 2004; Stark et al. 2005b), and to have higher protonemal growth in another, though the plants were hybrids from two very disparate populations (McDaniel et al. 2008). Females of the liverwort,

Marchantia inflexa, have greater growth rates and meristematic tip production than males

(McLetchie and Puterbaugh 2000; but for variation in dimorphisms across habitats see

Brzyski et al. 2014), which, along with greater predicted competitive ability (McLetchie

et al. 2002; Crowley et al. 2005), could explain female-biased adult sex ratios. In the same species, female gemmae (asexual propaguales) have higher substrate attachment and survival after desiccation than males, but male adults produce more gemmae (Stieha

et al. 2014). Ceratodon purpureus females are overall larger and produce fewer but

larger ramets than males (Shaw and Gaughan 1993). The sex-specific life history

differences cited in this paragraph were found using sexually-immature plants from

regenerated tissue, which is again unexpected according to the CoS hypothesis.

To explain female-biased sex ratios in general and the increasingly higher female

bias in drier habitats (Stark et al. 2001; Stark et al. 2005a; Stark et al. 2005b; Stark et al.

2010), several researchers (e.g. McLetchie 1992; Stark et al. 2000; Stark et al. 2001),

recognizing females fail to realize all reproductive costs due to male rarity (Longton

1981) or sporophyte abortion, proposed a modification to the CoS hypothesis termed the

cost of realized sexual reproduction hypothesis (CoRS). Only sperm limitation or high

levels of sporophyte abortion are required to reduce average female realized cost below

average male cost. With only centimeters of sperm dispersal distance, spore dispersal

followed by clonal growth can cause sperm limitation. However, one study showed no

sex difference in pre-zygotic reproductive investment in a female-biased species (Bisang

et al. 2006), but associated photosynthetic leaves were included in pre-zygotic investment

estimates. Interestingly, CoRS could explain female-bias in long-lived angiosperms with

pollen limitation, but pollen limitation when it occurs in angiosperms may be less

extreme than sperm limitation in bryophytes (see Carlsson-Graner et al. 1998).

Morphology

Sex-specific morphologies under consideration here (i.e. those not directly linked

to sexual reproduction) have lacked attention in all taxa, but what is known suggests this

will be a productive area of study. Sex differences in morphology can arise from sex- specific reproductive requirements or physiological needs. Like sex ratio, morphological

dimorphism can differ between populations because of environmental variation (Brus et

al. 2011), but sex-specific morphological responses to changes in environment remain

understudied. Females, due to higher carbon requirements, may produce larger leaves or

adjust their canopy morphology to maximize light interception and photosynthesis

(Dawson and Geber 1999). Sex-specific morphologies are often inseparably linked to physiological differences and ultimately to life history differences. Therefore, differences in morphology are also important for population dynamics and persistence via their life history consequences. Morphological patterns are not nearly as explored as life histories and only in long-lived woody angiosperms are they relatively firm. In herbaceous angiosperms, gymnosperms, and bryophytes morphological sex-differences are not well understood. Morphological patterns correlated with growth forms are summarized in Table 2.2.

Table 2.2. Summary of morphological patterns and growth forms across relevant land plant divisions.

Stomata/Pore Taxa density Leaf size Branching Branch size Angio. Woody Fa Vb Ma Fa Herb. U Vb U U Gymno. Conifers Fb c Fb c U U Cycads U Fb c Mb c U Bryo. U Fb c Mb c Fb c Note: The category “woody angiosperm” does not take into account female-biased shrubs with abiotic pollination and seed dispersal due to lack of data. Male > female is represented by an M and female > male is represented by an F. V represents “variable,” meaning the pattern varies between species, environmental conditions, or reproductive status and requires further investigation. U represents “undetermined,” meaning I am not aware of any studies demonstrating a difference between the sexes. a Pattern cited in previous reviews or other articles b To my knowledge, a pattern reviewed here c I know of only one study

Angiosperms—Canopy morphology can be sex-specific (Wallace and Rundel 1979;

Kohorn 1994; Bond and Maze 1999; Harris and Pannell 2010), although few taxa have

been studied. A case study is provided by Leucadendron, a genus with biotic and abiotic

pollination, increased branching creates smaller branches but larger floral displays,

increasing mating success via pollinator attraction or higher windborne pollen dispersal

(Dawson and Geber 1999) while increasing male mortality as demonstrated in an insect- pollinated species (Bond and Maze 1999). Sexual selection for male investment in flowers and pollen (Moore and Pannell 2011) could explain this canopy structure

dimorphism (i.e. non-floral traits) via tradeoffs. In contrast, stouter stems are likely

selected for in females to conduct more water to serotinous “cones” protecting mature

seeds for later dispersal (dimorphism increases with serotiny, Harris and Pannell 2010).

Degree of dimorphism is also increased in wind-pollinated species compared to insect-

pollinated (Bond and Midgley 1988; Tonnabel et al. 2014), indicating stronger selection

on males in wind-pollinated species.

Often angiosperm female leaf traits reflect selection for greater carbon

assimilation including larger individual leaf areas, higher specific leaf areas, or higher

stomata densities (Wallace and Rundel 1979; Dawson and Geber 1999; Dormann and

Skarpe 2002; Li et al. 2007; for larger male leaf areas see Dawson and Bliss 1989;

Cipollini and Whigham 1994). Larger leaves could correlate with larger stems, needed to

support fruit, but this may be synergistic with the need for more carbon (Bond and

Midgley 1988).

Patterns in herbaceous angiosperms are unclear but there are examples of females

having larger leaves (biomass or area), which can be genetically correlated with flower

size (Dawson and Geber 1999; Bram and Quinn 2000; but see Randriamanana et al.

2014), larger branches (Pickering 2000), and more branches (Bram and Quinn 2000).

Larger leaf area in males may be connected with overall larger growth rates compared to

females, which prevents generalizing the direction of the dimorphism.

There are hints of sex differences in root morphology related to sex-specific physiologies. However, the lack of studies prevents discerning broad patterns in root morphology correlated with sex and growth habit. Some differences likely correlate with below-ground biomass and root:shoot ratios (Mercer and Eppley, 2010). For instance, females of Corema album were inferred through isotope analysis to have deeper roots and potentially a higher root:shoot ratio than males, indicating greater water requirements

(Álvarez-Cansino et al. 2010). Sex differences in resource acquisition in the soil can

explain higher mycorrhizal colonization and associated morphology in females relative to

males (Eppley et al. 2009; Vega-Frutis and Guevara 2009; see also Physiology). Though

the proximate causes are unknown, Eppley et al. (2009) suggest either females of

Distichlis spicata have longer roots or simply attract more fungal associations than males.

This pattern may be general (Vega-Frutis et al. 2013).

Gymnosperms—Few conifer studies assess morphological sex differences outside

reproductive structures. Gauquelin et al. (2002) found that males of female-biased

Juniperus thurifera had greater crown projection areas than females despite females

being taller, but the reason is unknown. In contrast, J. virginiana males are taller than

females, which could result from sexual selection for wind pollination (Vasiliauskas and

Aarssen 1992). In wind-pollinated species, a lower density or emergent crown could

increase pollen dispersal via greater airflow (Niklas 1985). However, canopy emergence

may trade off against root growth, lateral crown growth, or risk wind-throw and lightning strike. If this is the case, then perhaps sexual selection has co-opted an existent trait

(greater male size) originally produced by sex-specific CoS. Co-opting of ecologically derived traits by sexual selection has recently been suggested as important for animal evolution and diversity (Bonduriansky 2011). At the leaf level, Taxus baccata females had longer leaves with larger area and higher stomatal density than males (Iszkuło et al.

2009), which provides evidence of a compensatory increase in photosynthesis due to high

CoS.

For cycads, knowledge of morphological sex differences is scant. However,

Newell (1985) suggested that males branch more for greater cone production. Although not tested in Zamia pumila, males of Z. integrifolia have higher branching than females

(Ornduff 1996). Greater ramification by males seems to reflect sexual selection pressures, as in angiosperms, to produce more cones (more pollen) to increase mating success. For sex-specific leaf morphology, Z. pumila females had more leaflets and larger leaves than males (Newell 1985). Sex-specific gymnosperm root morphologies are largely unexplored.

Bryophytes—Data on morphological differences for bryophytes is scant, but those reported connect to physiology and life history differences. At a large (for bryophytes) scale, Ceratodon purpureus males produce greater numbers of smaller shoots (Shaw and

Gaughan 1993), which likely leads to clump morphological differences. Clump morphological differences are largely unexplored in bryophytes. The link for bryophytes

between greater male ramification and sex expression (analogous to greater floral

display) and mating success (or survival) should also be studied. Marchantia inflexa

females produce more meristematic tips and males produce more asexual propagules than females (McLetchie and Puterbaugh 2000). Greater asexual propagule production and dispersal could be a sexually selected strategy to increase male gamete production through sexual maturation of asexual offspring (a form of male-male spatial competition for access to fertile females), trading off with persistence and competitive ability.

Our knowledge of other dimorphisms above or below ground are very limited.

One study on leaf morphology showed that C. purpureus males have shorter leaves than females (Shaw and Beer 1999), following angiosperm and gymnosperm patterns, but I know of no other similar studies. Bryophytes lack roots but rhizoids can be considered analogous to roots, but sex differences in these are unstudied. Even though mycorrhizae associate with bryophytes (Schüßler 2000; Russell and Bulman 2005; Davey and Currah

2006), I know of no studies on the interaction between morphological sex differences and mycorrhizal colonization.

Physiology

Although convenient, the separate consideration of sex-specific life histories, morphologies, and physiologies is artificial because physiological differences may cause or be caused by morphological differences, and both ultimately affect life history differences (Figure 2.1). For instance, stress can induce life history differences, which are rooted in sex-specific physiology. I note some physiological differences are likely controlled by sex-determining hormones and could be adaptive or simply side-effects

(Munné-Bosch 2015). Pressures to be the “best” male or female possibly lead to sex-

30 specific physiological requirements. Natural or sexual selection then dictates sex-specific resource allocation. One of the ways natural selection increases female fitness is greater carbon investment in sexual reproduction (see Life History). This is frequently true for angiosperms, bryophytes, and gymnosperms, although not often tested in conifers.

However, sexual selection may dictate higher N investment in pollen (or bryophytes sperm) for males to increase success in mating or male-male competition (see

Morphology and below). Physiological patterns in angiosperms are better explored than morphologies, but, as I discuss below, patterns of sex-differences in physiology can be highly context dependent and in general are more likely to shift with more studies than those in life history. At this time, there is very little I can discern about patterns in gymnosperms and bryophytes, and those I present are based on generally few studies.

These patterns correlated with growth forms are summarized in Table 2.3.

Table 2.3. Summary of physiological patterns and growth forms across relevant land plant divisions.

Overall reproductive C Pre-Zygotic N Photosynthetic Water Use Abiotic Stress Taxa allocation allocation rate Efficiency Tolerance Angio. Woody Fa Ma Va Ma Ma Herb. Fa Mb c Va Fe Fb d Gymno. Conifers Fa c U U U Ma c Cycads Fb c U U U U Bryo. Fb c U U U Fa c Note: The category “woody angiosperm” does not take into account female-biased shrubs with abiotic pollination and seed dispersal due to lack of data. Note abiotic stress tolerance may also be inferred from information presented in the Life History section. Male > female is represented by an M and female > male is represented by an F. V represents “variable,” meaning the pattern varies between species, environmental conditions, or reproductive status and requires further investigation. U represents “undetermined,” meaning I am not aware of any studies demonstrating a difference between the sexes. a Pattern cited in previous reviews or other articles b To my knowledge, a pattern reviewed here c Few studies but with clear pattern d Few studies but with a tentative pattern e I know of only one study

Angiosperms—Females benefit from greater photosynthetic capacity and more photosynthetic machinery during reproduction because of their relatively higher investment (Dawson and Geber 1999). For several tree and herbaceous species, females had more chlorophyll (Sysov and Sycev 1970; Cha 1987; Kumar et al. 2006;

Randriamanana et al. 2014), which could increase photosynthetic rate. However, counter examples are known including in one of the same genera (Barradas and Correia 1999; Xu et al. 2008; Xu et al. 2010) and in one herbaceous plant (Dawson and Geber 1999), which could indicate differential nitrogen (N) investment in reproduction (more in fruits) trading off with chlorophyll production see (Gehring and Monson 1994; Barradas and

Correia 1999). In one tree species, chlorophyll a:b ratio was lower in males than females

under drought (Xu et al. 2008) and in high UV-B (Xu et al. 2010), but a:b ratio was not

clearly different between the sexes in the field for three related species (analysis

performed by me, see Appendix 1; Cha 2000). For two herbaceous species, males had

higher a:b ratios either compared with unpollinated females (Dawson and Geber 1999) or

females at any time (Kumar et al. 2006). Considering that some studies above (Xu et al.

2008; Xu et al. 2010; Randriamanana et al. 2014) used non-reproductive clones of mature

trees, the discovery of sex-differences in chlorophyll is puzzling and could be due to

epigenetic effects (Juvany and Munne-Bosch 2015). Studies of true seedling juveniles

compared to non-reproductive clones of mature trees (for which members of the

Salicaceae seem particularly well suited) are necessary to determine at what stage and by

what mechanism these sex differences occur.

Sex differences in photosynthetic traits such as water use efficiency (WUE) need

to be contextualized by environmental variables and other physiological or even morphological measurements for interpretation (Zimmerman and Lechowicz 1982;

Juvany and Munne-Bosch 2015). For example, males had higher photosynthesis at leaf

level but females were estimated to gain more carbon over the whole plant (Nicotra et al.

2003). Generally woody angiosperm females have lower WUE than males to increase

their photosynthesis especially when supporting fruit, potentially lowering survival in

drier environments (Dawson and Geber 1999; Juvany and Munne-Bosch 2015).

However, there are notable exceptions (e.g. Letts et al. 2008; Xu et al. 2010) including higher female WUE yet lower photosynthesis under drought stress while males are seemingly better competitors (Correia and Diaz Barradas 2000). In one herbaceous

33 species, females had higher WUE than males, which gain more carbon but still have lower growth rates than females (Gehring and Monson 1994).

The joint use of δ13C with intrinsic WUE may be particularly informative as one can track long term WUE and the other can detect differences during changes in growing season and reproductive status. For example, Xu et al. (2008) showed in P. cathayana that males and females (from cuttings and not currently flowering) did not differ in intrinsic WUE while females had higher δ13C but lower net photosynthesis under drought. Juvany and Munne-Bosch (2015) made a convincing case that patterns of sex- specific physiological advantage under abiotic and biotic stress correlated with growth form are not fully fleshed out. They suggest consideration of modularity, ecological context, and sectoriality (the restriction of substance transport to subunits within plant architecture) are necessary to untangle sex-specific physiological stress responses. The lack of pattern is a bit surprising considering sex-specific life history patterns are well known and correlate with growth form.

While females require more carbon, males may require more N during flowering

(and potentially total) because pollen is N rich (Faegri and van der Piji 1979; Wallace and

Rundel 1979; Dawson and Geber 1999; Case and Ashman 2005). Females may require more total N because their large reproductive investment (Antos and Allen 1990;

Cipollini and Whigham 1994; Randriamanana et al. 2014). However, even if female N reproductive investment exceeds that of males, early season male N investment may reduce photosynthesis and carbon fixation (Case and Ashman 2005). Indeed, because wind-pollination requires more pollen than insect-pollination, male wind-pollinated

34 plants’ reproductive N costs can exceed female cost, influencing sexual dimorphism

(Barrett and Hough 2013).

Because herbaceous plants require more N and phosphorous (P) per unit carbon than woody plants to build tissues (Kerkhoff et al. 2006), I predict they will exhibit N- and P-limitation more than woody plants. There is some evidence that growth form correlates with N limitation, although the pattern is not clear cut (trees and grasses are more N limited than forbs and shrubs) (Xia and Wan 2008). Nevertheless, N can play an important role in dimorphism in herbaceous angiosperms. As mentioned above (see Life

History), the smaller Mercurialis annua males allocate 70% more N to inflorescences than females despite similar reproductive effort by biomass and reduce root growth less than females in response to higher soil N (Harris and Pannell 2008; but see Hogan et al.

1998). Seemingly males trade off investment in inflorescence N with growth, limiting male size relative to females (i.e. higher N costs). In Carex picta, N and carbon allocation combined with defoliation during flowering reduces tiller growth for males but not females, although N was not directly tested (Delph et al. 1993). Males in Silene dioica have higher reproductive effort with regard to P and deplete their underground N stores more than females (Hemborg and Karlsson 1999). Although the role of P and its interplay with N promises to be fruitful, only a few have begun to explore this relationship (e.g. Randriamanana et al. 2014). These studies show that trade-offs in resources other than carbon and N are important in causing sexual dimorphism and are worthy of consideration for employment in the CoS hypothesis (Case and Ashman 2005;

Barrett and Hough 2013).

Below ground physiological sex differences occur in Carica papaya, where

females had more mycorrhizal colonization and higher hyphae growth than males.

Hyphae were related positively with soil N:P, suggesting sex-differences in physiology

(mycorrhizal provision included) (Vega-Frutis and Guevara 2009). Interestingly, the sex differences in mycorrhizae in C. papaya begin to occur in juvenile plants and are thus inconsistent with the CoS hypothesis. Emerging patterns indicate females may fine tune mycorrhizal colonization better than males (Vega-Frutis et al. 2013), which could indicate a tighter control on the balance between resource acquisition and mutualist provisioning. Below ground intrasexual or interspecies interactions could also be important. For instance, in M. annua, close proximity to other conspecific females increases female reproductive investment (Hesse and Pannell 2011), which could relate to female-specific soil microbial community responses (Sánchez Vilas and Pannell 2010).

In other physiologically related biotic interactions, males are favored by herbivores and invest less than females in physiological defenses such as tannins and phenolics (Cornelissen and Stiling 2005; Vega-Frutis et al. 2013). Males generally attract more pollinators (and potentially herbivores) than females through higher upfront reproductive investment, including fragrance (a resource cost trading off with traits outside reproductive structures) (Ashman 2009). Patterns remain unclear whether one sex has physiological advantages for pathogen defense. Vega-Frutis et al. (2013) make a compelling case for the fruitfulness of studying the interaction between multiple biotic antagonists and mutualists.

Gymnosperms—Even though few gymnosperm studies focus on physiological sex

differences, some morphological studies have implied them. The leaf morphological sex

differences (see Morphology) indicate similar potential as angiosperms for higher female

photosynthetic capacity compared to males, but this has not been tested. Rovere et al.

(2003) also alluded to preliminary data possibly showing greater photosynthetic capacity

in Austrocedrus chilensis females especially during cone development, but apparently

this was not confirmed. One study indicated that leaves of Ginkgo biloba females had

higher moisture, chlorophyll, and carotenoid content during the entire growing season

(Mori et al. 2000), but whether these differences translate to photosynthetic differences is

unknown. Similarly, some cycad females have larger leaves than males, but I know of no

studies on physiological consequences such as carbon gain. Stress-induced physiological

sex-differences are likely important in gymnosperms, but there are few studies. Lower

WUE has been suggested, but not tested, to explain greater drought sensitivity of young

Juniperus thurifera females (Rozas et al. 2009). In J. communis ssp. communis, females

at a dry site appeared to have lower WUE (inferred from lower δ13C) than males.

However, males were more stressed at a wet site than females such that male δ13C was

lower (though not significantly) than female δ13C and sex ratio was highly female-biased

(Hill et al. 1996). These results correspond with the spatial segregation found for the

riparian angiosperm Acer negundo var. interior (Dawson and Ehleringer 1993). These

examples warrant similar studies among other gymnosperm taxa. These studies are

congruent with patterns, though they remain less definitive, for female woody

angiosperms which often have larger leaves, higher photosynthetic rates, and lower WUE than males.

Male angiosperms spend a higher proportion of their reproductive investment on

pollinator attraction than females, and similarly cycads exhibit physiological differences

related to pollinator attraction. For instance, the sexes of Macrozamia lucida differ in

cone thermogenesis (Roemer et al. 2005). While metabolically active longer, female cones peaked earlier and at lower temperature than males. Male cones peaked more optimally in time for pollinator attraction, but which sex incurs greater metabolic costs is unknown. Furthermore, because cycads pollinators are often destructive (e.g. beetles), male cycads, like male angiosperms, invite more damage. While damage is costly, male cone destruction by pollinators can lower cone maintenance costs, decreasing time between reproductive bouts (Marler 2010). The prevalence of physiological sex differences of this type among cycads is another area needing further exploration.

Male gymnosperms like male angiosperms may be sexually selected to increase pollen production. Following the prediction from angiosperms (Case and Ashman 2005),

I predict that even though females may have greater total N investment in reproduction, males will invest more N earlier, potentially decreasing growing season carbon gain.

This prediction is supported by evidence from monoecious Pseudotsuga menziesii var. glauca, which invests more N per unit carbon for male reproduction than female

(McDowell et al. 2000). Therefore, decreases in growth or survival due to male pollen investment could cause female-biased sex ratios, potentially explaining the female bias of

J. thurifera mentioned above. However, all gymnosperms are woody and thus potentially less N-limited than herbaceous plant taxa. Nevertheless, T. baccata shows sex differences in N use. For females but not males, N content positively correlates with specific leaf area (Iszkuło et al. 2009).

Below ground physiology remains relatively unexplored in gymnosperms. A

study on J. monosperma reported that heavy infestation by the hemiparsitic mistletoe,

Phoradendron juniperum, appears more harmful to females because (not directly tested)

females lower photosynthate allocation to mycorrhizae more than males (Gehring and

Whitham 1992). In Ephedra trifurca, a gnetophyte, gall-forming midges damage males more than females (consistent with angiosperm patterns). Degree of dimorphism decreases at higher water and N sites, suggesting sex-specific nutritional needs and responses to stress (Boecklen and Hoffman 1993). In other organism interactions, females of J. communis are more susceptible to rabbit and pathogenic fungi damage than males (Ward 2007), which suggests sex differences in secondary compounds. I expect similarities in dimorphic physiology as in life history between woody angiosperms and gymnosperms to become more apparent as more studies are done.

Bryophytes—As in gymnosperms, few physiological sex differences have been reported for bryophytes (see Groen et al. 2010a), but several morphological differences imply sex differences in photosynthesis. Ceratodon purpureus males have shorter leaves than females, potentially lowering male photosynthesis compared to females. Marchantia inflexa has a dimorphic relationship between thallus support tissue thickness and photosynthetic rate such that females had a significant negative relationship while males had a non-significant positive trend (Groen et al. 2010a). Males of M. inflex produce more gemmae cups (carbon export) while females produce more meristematic tips, which presumably increases the vegetative body and photosynthetic area of a female. Males of the same species have lower chlorophyll a:b ratios than females, which is characteristic of

low-light plants despites males seemingly occupying more open habitat areas (Groen et

al. 2010b). Even though differences in WUE and photosynthetic level have been

reported in angiosperms, these have not been investigated in bryophytes.

Because many bryophytes enter physiological dormancy during drying events, recovery ability is very important (Proctor et al. 2007). Two species have females with greater desiccation tolerance than males, one from regenerating leaves (Newton 1972)

and one in gemmae (Stieha et al. 2014). A similar pattern was found in vegetative plants

originally collected from their native habitat but not in plants collected from a recent

range expansion (Marks et al. 2016). Contrastingly, one study showed that males tolerate

dry heat better than females (Stark et al. 2009). Differences in regeneration ability and stress tolerance favoring females over males (or in some cases vice versa) are important proximate causes of life history differences especially in drier habitats (Stark et al. 2001;

Stark et al. 2005a; Stark et al. 2005b; Stark et al. 2010) and must also be determined physiologically, but the mechanisms behind these differences are unknown.

While evidence is accumulating for N’s involvement in sexual dimorphism in herbaceous angiosperms, N’s involvement in bryophytes is uncertain. Bryophytes can also be considered herbaceous (lower C:N), and their sperm could be N-rich like seed plant pollen. Sexual selection for increased bryophyte male reproductive N investment could lead to lower male growth and survival, which may explain female-biased bryophyte sex ratios. I would then predict nitrophily among male bryophytes.

Although lacking roots, bryophytes do have below ground portions. To my knowledge, sex differences in below ground physiology have not been explored.

However, given other plant lineages exhibit sex differences in mycorrhizal interaction

likely involving differing physiologies, future research may reveal physiological below

ground sex differences for bryophytes as well. Other biotic interactions also deserve

attention. There is evidence of a sex difference in pathogen infection rate in S. caninervis

(more infection for males) (Stark et al. 2005b), and animal mediated sperm dispersal

occurs in bryophytes (Cronberg et al. 2006). In C. purpureus, fertile female clumps had

more volatile organic compounds (VOCs) and springtails preferred female scent over

male. These same VOCs have also been identified in angiosperms (Rosenstiel et al.

2012). Rosenstiel et al. (2012) noted that mutualistic microbes might induce sex-specific

VOCs in bryophytes as in other plant divisions, and, following Ashman (2009), they suggest female bryophytes have more VOCs due to mate limitation. Researchers have speculated that bryophyte products including mucilage, sugars, sperm (from males), starch, and fatty acids function as rewards to microarthropods (Cronberg et al. 2006;

Cronberg 2012; Rosenstiel et al. 2012). Through tradeoffs, these sex-specific physiologies involving fertilization could affect other physiological traits such as nutrient use efficiencies, WUE, and photosynthetic rate. Bryophytes are generally attacked by few herbivores with the exception of sporophyte capsules (Davidson et al. 1990). The relationship between sex-specific physiology and biotic interactions is a rich area for future studies with many emerging analogies among vascular plants.

Conclusion and future directions

Parallels can be drawn between gymnosperms and woody angiosperms because of similar growth habits, sex ratios, and life history patterns. In most cases the CoS hypothesis explains secondary sexual characteristics in these divisions, although the appearance of juvenile sex differences or among individuals not expressing sex in several

41 species including in physiology warrants further research. The lack of ability to identify sex of juvenile has likely limited the collection of data on sexual dimorphism, and more of these traits could be arising before sexual maturity than is currently known. Some research conducted on non-flowering saplings propagated from mature trees suggest that some sex differences are not simply due to direct tradeoffs or responses to reproductive investment. Whether these differences are dictated at sexual maturity or as juveniles is unknown. Considering their ability dedifferentiate into the juvenile phase as well as be clonally propagated as adults (Giles 1971), bryophytes could shed light on how and when some sex-differences arise.

The seemingly anomalous female-biased sex ratios of some angiosperm and gymnosperm shrubs suggest they could be good model systems for each other.

Considering the ease of propagation, members of the Salicaceae can be and already have been good systems to work with. Currently, information is lacking for generalization about the causes of these sex ratio biases, and data on seed/seedling sex ratios, rates of inbreeding, and life history differences including realized total reproductive investment of both carbon and non-carbon resources are needed. Considering their scarcity in general, more studies are needed on reproductive investment in dioecious conifers.

Parallels can also be drawn between the herbaceous angiosperms (especially those with abiotic pollination and clonal growth) and the bryophytes evidenced by female- biased sex ratios—so much so that bryophytes may serve as tractable model species for herbaceous angiosperms. Sex differences within these growth habit groups contrast with the typical predictions of the CoS hypothesis, but may be explained by realized costs or sex-specific morphology and physiology due to contrasting sex functions. There is

accumulating evidence that this is the case in bryophytes, but data is needed on pollen

limitation and realized cost of reproduction, considering carbon and non-carbon

resources, in herbaceous angiosperms. If inbreeding has not caused the evolution of

biased seed or spore sex ratios, sexual selection could also explain female-bias through

favoring early senescence of males or smaller size to prevent competition with offspring-

bearing females, especially in annual plants. For sexual selection to operate in this

fashion, variation in mating strategy and success must be present, and so identifying these

strategies as they relate to life histories, morphology, and physiology is another possible

future direction for studies.

While above ground morphological differences in angiosperms except for the

effect of stress on sex-specific morphology have been explored to an extent, information is lacking in gymnosperms and bryophytes. Fruitful work may be done to define the link in the latter two divisions between differing resource needs (physiology) and morphology. The role of sexual selection in determining differences in morphology deserves more attention in all land plants but especially gymnosperms and bryophytes.

Analogies can be drawn between males of wind-pollinated angiosperms (both woody and herbaceous) and water-fertilized bryophytes. These males should have morphologies reflecting selection to increase mating success. In wind pollinated species morphologies that allow flowing air to remove pollen increase mating success (Friedman and Barrett

2009) and in water-fertilized species morphologies that allow flowing water to carry away pollen or sperm should increase mating success. In bryophytes, males with smaller leaves and more streamlined profiles or lower shoot densities could have better sperm dispersal. Below-ground morphology is more difficult to assess and is potentially a rich

area for future study in all taxa, particularly, because it will allow the contextualization of

physiological measurements such as WUE (i.e. one sex could have a less conservative water strategy because it is better tapped into ground-water than the other).

For dioecious angiosperms, we have only begun to understand the effects of sex-

specific resource requirements and biotic interactions on physiology and their link to life

history and sex ratios. Often it seems, one sex fares better under abiotic or biotic stress,

but linking physiology to life history patterns has proven challenging. In angiosperm and

gymnosperm trees, males predominantly do better under drought. In bryophytes, males

appear to do more poorly under water stress and some sex-differences favoring females in desiccation tolerance have been identified, but the mechanisms of these differences are not understood. More studies are needed on patterns of instantaneous as well as long- term WUE, photosynthesis, sex-specific use of biochemical pathways, secondary compounds, and mutualist recruitment and provisioning. While carbon and water are important resources that affect many aspects of sexual dimorphism, the importance of other resources is becoming more apparent. Because pollen is N expensive and sperm potentially is also, a parallel can be drawn. The role of N costs in sexual dimorphism in male bryophytes is as yet uninvestigated. The first step in testing the role of N for males is to test the prediction that male bryophytes should be nitrophiles (N abundance alleviating the potential tradeoff between sexual investment and growth/survival). Sexual selection could increase N allocation in both angiosperm and bryophyte males at the expense of growth and survival. As is often the case, very little work has yet been done in gymnosperms and bryophytes. Bryophytes are among the oldest groups of land plants whose colonization of land is thought to have been facilitated by mycorrhizal mutualists

(Wang et al. 2010), and yet very little is known about the interaction between bryophytes

and their mutualists let alone sex-differences. I hypothesize that bryophyte females have more beneficial relationships with microbes, leading to female-biased sex ratios.

In summary, there is much to be learned from the application of knowledge from these different divisions of land plants and the in-depth comparisons drawn by the authors may be the first of their kind. My hope is the synthesis of information from the angiosperms, gymnosperms, and bryophytes will inform and inspire researchers working on sexual dimorphism in these three divisions.

CHAPTER THREE: THE EFFECTS OF AN URBAN EXPOSURE GRADIENT ON LEAF MORPHOLOGY, PIGMENT PHYSIOLOGY, AND POPULATION SEX RATIO OF BRYUM ARGENTEUM HEDW. (BRYACEAE)

Abstract

In dioecious seed plants, the sex with higher reproductive investment can be rarer in stressful environments, but similar associations are not well established for bryophytes.

Using the dioecious moss Bryum argenteum, where males likely investment more in

reproduction than females, I tested whether males became rarer as exposure increased. I

also quantified leaf morphology and pigment content, hypothesizing that stressful

conditions would amplify sex differences. I assessed sex, leaf morphology, pigment

content, and exposure level (light intensity and canopy openness) of field-collected plants

and leaf morphology and growth rate on the same isolates in a greenhouse under two

light intensities. Sex ratio was female-biased but uncorrelated with exposure. Male rarity

prevented assessment of field sex differences. Leaf length:width ratio related negatively

to both exposure and light. Leaf achlorophyllous area and apex length related positively

to exposure. The latter, though unaffected by light, was longer in the greenhouse than the

field. High light increased total chlorophyll and carotenoids, but did not affect pigment

ratios. Females tended to grow larger and faster than males, which likely causes female-

bias rather than sex-specific responses to exposure. Leaf morphological responses to

high exposure likely protect against photodamage during desiccation and create deep

shade conditions, increasing photosynthetic pigments. Apex function in B. argenteum

apparently differs from awns in other xerophytic bryophyte species. Lastly, traits

differed genetically among habitats, which connects to others studies showing

anthropogenic microhabitat differentiation in both seed plants and bryophytes.

KEY WORDS: sex differences; spatial segregation of the sexes; light responses; stress

gradient; life history

Introduction

In sessile organisms such as dioecious plants, finding a mate is highly dependent

on the proximity of opposite-sex neighbors, which is linked to the local population sex

ratio. When population sex ratios vary along environmental gradients, the likelihood that

nearest neighbors are the opposite sex also varies. If gamete dispersal distances are small

and distance between neighbors are relatively large on average, highly skewed local

population sex ratios can greatly reduce the opportunity for sexual reproduction. In

plants, sex ratios vary along environmental gradients including, in some species, the

spatial segregation of the sexes (SSS) (Sinclair et al. 2012). Unsurprisingly, the reduction of sexual reproduction and/or the reduction of effective population size caused by highly skewed population sex ratios can decrease genetic diversity and species persistence

(Kimura and Crow 1963; Ellstrand and Elam 1993; Sinclair et al. 2012). Thus, understanding the causes of sex ratio variation assists in endeavors such as rare species conservation (e.g. Iszkuło et al. 2009; Wall et al. 2013) and prediction of adaptation responses to environmental change (Ellstrand and Elam 1993).

The causes of environmentally linked sex ratio variation are likely related to sex- specific environmental limitations caused by differences in sexual reproductive function.

Female reproductive function differs from that of males by including post-zygotic maturation and dispersal of offspring in addition to mating success. These differences are associated with sex-specific physiology, morphology, and life histories (Dawson and

Geber 1999; Delph 1999). Most studies on sex-specific environmental limitation and sex

47 ratio variation are on seed plants, particularly angiosperms. In seed plants, the most obvious sexually dimorphic trait, reproductive investment, is expected via tradeoffs to influence sex differences in environmental limitation (Dawson and Geber 1999).

Research has often focused on the connection between higher female reproductive costs, sex-specific traits that support or trade off with reproductive investment, and sex ratio variation along environmental gradients. Generally, sex ratios in seed plants are more male-biased in higher-stress habitats, supporting the idea that tradeoffs with reproduction interact with traits that influence stress tolerance (Bierzychudek and Eckhart 1988;

Dawson and Geber 1999; Barrett and Hough 2013). However, the exact traits responsible for apparent sex differences in stress tolerance remain elusive (Juvany and

Munne-Bosche 2015), and few studies have assessed whether these traits are present in non-reproductive individuals or juveniles (but see, Quinn and Meiners 2004; Vega-Frutis and Guevara 2009).

The connection between sex ratio variation, sexually dimorphic traits, and environmental limitation remains a relatively unexplored frontier in bryophytes (but see

Groen et al. 2010 a,b). Sex ratio variation including SSS is well documented in bryophytes, and in some cases, the sexes are completely isolated (Bisang and Hedenäs

2005; Renzaglia and McFarland 1999; Bowker et al. 2000). However, examples demonstrating the influence of environmental gradients on sex ratio variation are surprisingly limited (but see Pettet 1967; Watson 1975). In contrast to seed plants and despite higher female potential reproductive cost, bryophyte males likely have higher average actual sexual reproductive investment due to sperm-limited female reproductive success (McLetchie 1992; Stark et al. 2000). Using similar reasoning as used for seed

plants, higher actual male investment in sexual reproduction could lead via tradeoffs to lower male stress-tolerance compared to females with females predominating in higher-

stress habitats (Stark et al. 2005), which is the opposite of seed plant patterns. For

example, although male individuals occurred too infrequently for a conclusive

assessment, males of the desert moss Syntrichia caninervis grew only in shrub understory

while females were found in both understory and full sun (Bowker et al. 2000; Stark et al.

2005). Additionally, a few studies have shown that males tolerate drying stress less than

females (Newton 1972; Stieha et al. 2014; Marks et al. 2016). In each case though, the

sex difference was present in juveniles, indicating physiological sex differences that

could contribute to environmentally linked demography may arise before sexual maturity

(i.e. before direct tradeoffs with reproduction).

In the western United States, the moss Bryum argenteum has female-biased sex ratios overall, and though its population sex ratios are never male-biased, populations become less female-biased as elevation increases and climate becomes more mesic (Stark et al. 2010). Males of B. argenteum also invest more in sexual reproduction pre-

zygotically than females (up to 24 times, Horsley et al. 2011), suggesting that if females

are sperm limited, average reproductive investment is higher for males than females.

Based on the pattern between climate variation and sex ratio, I hypothesized males would

also be less frequent at high exposure sites compared to more mesic sites at a local scale.

Using B. argenteum collected from five physically different habitats at the same

elevation within an urban environment, I asked: (1) Does female bias increase with

increases in local light intensity/canopy openness levels? (2) Does leaf (phyllid)

morphology correlate with light/canopy openness levels, and do key pigment ratios and

concentrations change predictably with light intensity? (3) Are patterns in morphology

and pigment physiology sex-specific? Because variation in morphology and pigment

physiology in B. argenteum along environmental gradients is not well understood, it was

necessary to assess broad patterns first in order to contextualize any sex differences in

morphology or pigments. Additionally, using non-sex-expressing shoots, I asked if leaf morphology and life history traits responded plastically to light, were sex-specific, or varied genetically in a greenhouse common garden. For traits linked to environmental gradients in either the field or greenhouse, I expected sex differences would be most apparent at higher light intensity/canopy openness (field) or higher light (greenhouse), which are potentially more stressful.

Methods

Study species—Bryum argenteum (Hedw.) is a moss with unisexual individuals

(dioecious) and a widespread distribution in natural and human-disturbed habitats (Figure

3.1) including disturbed soil, tree bases, pavement, masonry, and rooftops. The sex ratio of B. argenteum is female-biased in the Western USA with males absent at low elevation and proportion of males increasing with elevation and in urban habitats (Stark et al.

2010). The plant is silvery when dry due to the absence of chlorophyll from leaf tips

(achlorophyllous), although chlorophyll can be present almost to the leaf apex under some conditions. Morphology varies within the species potentially due to phenotypic plasticity rather than genetic differentiation (Crum and Anderson 1981). Costae in B. argenteum may end well below the leaf apex or be percurrent to excurrent, the latter being a true awn (Crum and Anderson 1981). For the sake of simplicity, elongated leaf

apices are hereafter referred to as “apices” regardless of costa length. Sexual and asexual

reproductive investment differs genetically among populations (Horsley et al. 2011).

1 cm

Figure 3.1. Photo of typical habitat of Bryum argenteum taken at the outdoor amphitheater behind Memorial Hall on the University of Kentucky campus. The inset shows B. argenteum (silvery, light green; clump indicated by yellow arrow) growing in a stone crevice among other mosses (dark green) and lichens.

Collection procedures and site characterization—I collected B. argenteum clumps

(aggregations of shoots) from the University of Kentucky campus and the immediately adjacent area (approximately 385 ha) between June 6 and June 12, 2012. I divided the campus into four quadrants of similar size. Clumps were haphazardly collected no closer together than approximately 3m by beginning at the center of the four quadrants and sampling each quadrant for four hours, walking along footpaths. A coin flip determined the direction taken at forks in the footpaths thus minimizing sampling bias. To the best

of my ability, all potential habitats were explored along the paths including horizontal

hard surfaces, topsoil, tree bases, vertical hard surfaces, and greenhouse roofs. In the study area, most tree base clumps occurred on trees of the genus Quercus, section

Lobatae. Differences between habitats were determined visually. During sampling, I

observed only one potentially mixed-sex clump (i.e. old sporophytes present) and only one male clump expressing sex. In total, I collected 95 clumps of B. argenteum. Two

clumps were from a unique habitat (discarded cloth) and were excluded from further

analysis. Of the remaining 93, I randomly selected 15 clumps from each of four habitats

(horizontal hard surfaces, tree bases, vertical hard surfaces and greenhouse roofs) and all

seven clumps from topsoil (a total of 67 clumps) for further measurements and planting

in a controlled greenhouse common garden experiment.

Characterization of light environment and canopy openness—I characterized the light

regime (% canopy openness and estimated photosynthetic photon flux density (PPFD):

mol m-2 day-1) of each selected clump using hemispherical canopy photographs taken

with a digital camera (Coolpix 950, Nikon, Tokyo, Japan) equipped with a 180° fish-eye

lens and the program WinSCANOPY (Regent Instruments, Ville de Québec, Québec,

Canada). Because canopy photographs were taken in summer, the assessment period for

PPFD was May 1 to June 9, 2012, corresponding to when trees are likely fully leafed

(data from the USA National Phenology Network, www.usanpn.org) up to collection. In

ideal (greenhouse) conditions B. argenteum grows approximately 0.8 mm per month

(Chapter Five, published as Moore et al., 2016), so the maximum new growth of field

plants during the light level assessment period was about 1 mm. I was unable to obtain a

canopy photograph for one clump, so in total 66 clumps had corresponding light regimes.

Greenhouse roof clumps were collected from two sets of greenhouses, one set of which

had recently been painted (Classic Kool Ray Greenhouse Paint, Continental Products,

Euclid, Ohio, USA), potentially altering the light environment experienced by the plants.

The majority of greenhouse roof clumps (12 of 15) had paint on them, but trait ranges were similar between clumps with and without paint, and greenhouse roofs on the whole had similar trait ranges as other habitats (see Results). Because vascular plant leaves can adjust both pigments (Naidu and DeLucia 1997; Oguchi et al. 2003) and morphology

(Nicotra et al. 2011) and bryophytes can adjust pigments (Kershaw and Weber 1986;

Martínez-Abaigar et al. 1994) in response to short-term light changes, it seems likely that

B. argenteum would respond to light changes during the assessment period.

Assessment of morphological traits on field collected plants—From the middle of each

selected clump, one shoot was removed. Mature leaves were chosen from immediately

below the growing tip to avoid taking currently expanding leaves or those shaded by new

growth. I measured leaf width, leaf length, leaf apex length, total leaf area, and

proportion of leaf area that was achlorophyllous (relative achlorophyllous area; Figure

3.2) using a dissecting microscope, a digital camera, and a Macintosh computer with

Image J (developed by U.S. National Institutes of Health and available at

http://imagej.nih.gov/ij/). To assess leaf shape changes, I used the ratio of leaf length to

leaf width, which is independent of leaf size. Leaf length:width ratio varies within (e.g.

Harvey et al. 1988; Kern et al. 2004) and among species and is taxonomically important

(Hickey 1973; Shaw 1984; Ferguson et al. 2000), but I did not have a prediction for its

53 potential relationship with exposure in B. argenteum. Awns in the moss Syntrichia caninervis capture dew (Tao and Zhang 2012), and assuming apices function similarly in

B. argenteum, I predicted increasing apex length with increasing exposure. Preliminary partial correlations indicated a strong positive correlation between apex length and leaf length, so I standardized apex length by dividing by leaf length. In many seed plants, leaf area decreases in higher light (Sultan 2001), lessening exposure to unneeded and damaging solar radiation (Potters et al. 2007), and I predicted that leaf area of B. argenteum would decrease as exposure increases. In B. argenteum, leaf tips lacking chlorophyll is an apparent adaptation to protect against high light damage (Scott 1982), and I predicted this area (relative to total leaf area) would increase with exposure. I did not make specific predictions regarding the direction of sex-differences in leaf morphology.

Figure 3.2. Drawing of a Bryum argenteum leaf showing how morphological measurements were made.

Pigment extraction and quantification—I extracted pigments from an additional five shoots from the center of each clump using 0.5 ml of DMSO overnight following Groen et al. (2010b). Total chlorophyll (a+b) and total carotenoid concentration were quantified using standard equations for low resolution spectrophotometers (Wellburn 1994; Tait and

Hik 2003). After extraction, shoots were dried at 60°C for 24 hours and weighed to the nearest 0.0001 mg to standardize pigment amounts, see Appendix 2 for details. I calculated chlorophyll a:b ratio, total carotenoid:total chlorophyll ratio, and total chlorophyll and total carotenoids per dry mass. In general, vascular plants have higher chlorophyll a:b ratio and carotenoid:chlorophyll ratio in high light compared to low light due to the relative reduction of chlorophyll b and the relative increase of carotenoids for photoprotection in high light (Boardman 1977; Dale and Causton 1992; Sarijeva et al.

2007). Bryophytes from high light habitats can also have higher carotenoid:chlorophyll ratios (Marschall and Proctor 2004). Total chlorophyll on a dry mass basis generally decreases as light intensity increases because photosynthesis is more limited by carbon fixation reactions than light energy (Boardman 1977; Dale and Causton 1992; Sarijeva et al. 2007). I expected B. argenteum to respond similarly to increasing light intensity.

Assessment of life history, morphological plasticity, and sex ratio in the greenhouse—

In the greenhouse common garden, plants were grown from a single shoot removed from the center of each of the same clumps, likely representing the same genetic individuals, used to assess field traits. Using identical plants increases the comparability between the two experiments. Each shoot was cut into two pieces bearing four leaves each in most cases to control for size. These pieces were planted on steam sterilized local topsoil in

55 individual pots (59 mL). One clone of each isolate was placed in each of two treatments achieved with light-filtering lids: high light (100% transmission) and shade (24% transmission, 0.6 neutral density, Lee Filters, Burbank, California, USA). The pots were randomized on a capillary watering mat in the greenhouse under a PVC frame overlain with a plastic drop cloth that provided additional shade and further prevented contamination from propagule rain. Total shoot number was assessed three weeks after planting. After approximately three months, plants were assessed for shoot density, leaf area, leaf length, leaf width, apex length, and shoot length. Because some replicates had begun to express sex, shoots were carefully chosen to avoid those with sex structures so that comparisons were among non-expressing shoots only.

Relative achlorophyllous leaf area was not measured because initial observations indicated both high and low light leaves were green throughout. Plants were then cultured until sex expression. Additionally, to augment sample size for sex ratio assessment, samples from all remaining field-collected clumps (n = 26) were planted three weeks after the light experiment plants and cultured to sexual maturity. These plants were placed in the greenhouse on a capillary mat with clear neutral density filters.

A survey between 9 and 13 months after planting of the light experiment plants allowed the assignment of sex to ~92% of isolates, although the sex of three isolates (~3%) were assigned after an additional 12 months. Plants cultured from the center of each field- collected clump were considered to represent the majority sex of that clump, according to

Stark et al. (2010). This procedure is further supported by my observations of only one field-collected clump showing evidence of both sexes.

Statistical analyses—Statistical analyses were conducted in SAS 9.4 (SAS Institute,

Cary, North Carolina, USA) unless otherwise noted.

Exposure—Because canopy openness and total light are highly correlated and because

preliminary analyses indicated very similar (though not identical) results for PPFD or

openness alone, I used principal components analysis to create a combination of light and

openness (loading = 0.97 for each and explaining 93.4 % of the variation), referred to

hereafter as “exposure.” I used an ANOVA to test for differences between habitats in

exposure and grouped the habitats based on pairwise t-tests.

Sex ratio—I used logistic regression to test the likelihood of plants being male or female

against habitat and exposure. Three isolates of unidentifiable sex and one isolate without

a canopy photograph were excluded from the analysis. Additionally, to achieve model

convergence, topsoil plants (all 5 were female) were excluded from the analysis (Total N

= 58). Sex was the dependent variable and exposure and habitat were the independent

variables. The interaction between exposure and habitat was not significant and was

dropped from the model.

Additionally, I conducted three goodness-of-fit tests (G-tests) for sex ratio variation and deviation of the overall sex ratio from unity using a Microsoft Excel spreadsheet according to Sokal and Rolf (1995). Based on germination sex ratio, an adult sex ratio near unity is expected for Bryum argenteum (Stark et al. 2010). The heterogeneity G-test assesses whether sex ratio varies among habitats. The total G-test test indicates whether the level of sex ratio skewness in each habitat could happen by

57 chance alone. The pooled G-test considers the overall sex ratio. This analysis included plants of known sex but without a canopy photograph, and thus only habitat was used as the independent variable. Plants failing to express sex (four clumps) were excluded from the analysis. A more conservative analysis assuming all plants of unidentifiable sex were male did not differ in interpretation (analysis not shown).

Field grown plant morphology—I used a MANOVA to test for overall effects of exposure and habitat on morphology. I did not test for a sex effect because only 8 of 67 samples were male, severely reducing power. Therefore, all 67 plants regardless of sex were included in the analysis. I tested the effects of habitat, exposure, and their interaction on relative apex length (apex length/leaf length), leaf length:width ratio, relative achlorophyllous area (achlorophyllous area/leaf area), and total leaf area.

Relative apex length was log transformed and relative achlorophyllous area was logit transformed to improve normality. I tested for correlations among dependent variable pairs while accounting for the effects of independent and other dependent variables using partial correlations. A univariate ANCOVA was performed for each dependent variable with the same independent variables as the MANOVA. Differences in the slope of the response to exposure between habitats were tested using contrasts.

Field pigment content—I conducted two MANOVA’s for pigment content, one for the two ratios (chlorophyll a:b and carotenoids:chlorophyll) and one for the two total contents by biomass (total chlorophyll and total carotenoids). For pigment ratios, three datum points (one each from greenhouse roofs, tree bases, and vertical hard surfaces) were dropped due to negative pigment values due to poor extractions (total N= 64). For

total pigment content, the same three points were excluded and additionally five other

points (three from greenhouse roofs and two from tree bases) were excluded due to loss

of shoots during weighing (final N = 59). In each case, the independent variables were

habitat, total light, and their interaction. I used total light instead of the principle

component of light and canopy openness (exposure) because light most directly

influences pigment variation. Univariate analyses were also performed for each

dependent variable with the same independent variables. All dependent variables were

log-transformed to improve normality.

Light environment effects on morphology in the greenhouse—I performed a doubly

multivariate repeated measures analysis in PROC GLM to test for overall effects of light,

habitat, and sex. Dependent variables included three measures of leaf morphology: leaf

area, length:width ratio, and relative apex length, and two measures of growth: shoot

length and shoot number at three weeks. The doubly multivariate analysis is very

conservative because PROC GLM drops all rows with missing data points. Missing data

points within an isolate were due to replicate mortality. This reduced the total number of

isolates from 64 to 42 of which only three were male. Shoot density at three months was

not included in the doubly multivariate analysis because it was significantly positively

correlated with three week shoot production (see Table 3.6), would be redundant, and

likely does not reflect final clump shoot density (Moore et al. 2016). Furthermore, shoot

production at three weeks is likely to be more tightly correlated with early growth rate.

Using PROC MIXED, which does not drop rows with missing data points, I tested for the

effects of sex, habitat, light level, the interaction between habitat and light, and the

59 interaction between sex and light on each dependent variable including shoot density.

Because my initial hypotheses included sex-specific responses to light, I included sex and the interaction of sex with light despite low male sample size and non-significance in the multivariate analysis (see Table 3.5). The interaction between habitat and sex was not included because only two habitats had more than one male. Sex was unidentifiable for three isolates due to poor health or death, and these were excluded from the analyses.

Shoot length, leaf length:width ratio, leaf area, and shoot density were log-transformed and shoot number at three weeks was square-root-transformed to improve normality.

Because of non-significance in the multivariate analysis, non-significant interactions between sex and light were dropped from univariate analyses.

Additionally, I conducted analyses of covariance to test for relationships between field morphological traits and those measured on greenhouse common garden plants (i.e. signatures of genetic influence on field-measured traits). Leaf length:width ratio, leaf area, and relative apex length were measured in both field and common garden. Because trait relationships with exposure varied by habitat for both leaf length:width ratio and relative apex length (see Results), each habitat of origin was tested separately, using trait values from high light common garden plants to maximize sample size. Field traits were used as dependent variables with exposure as the independent variable and the matching trait value from the same isolate in common garden as the covariate. Transformations were as described above.

Results

Field exposure—The five designated habitat types differed significantly in exposure

(F4,61 = 31.13, P < 0.0001). These five habitats also differed significantly in openness

(F4,61 = 33.16, P < 0.0001) and light intensity (F4,61 = 25.62, P < 0.0001) when tested

alone. For exposure, pairwise t-tests demonstrated that greenhouse roofs and tree bases

are the most and least exposed, respectively. All other habitats, while having a range of

exposure more broad than either greenhouse roofs or tree bases, did not differ from one

another (Table 3.1).

Table 3.1. Mean exposure ± SE (principal component scores of light and canopy openness) for each habitat. Superscript letters indicate significant differences (P < 0.05) among habitats by t-test.

Habitat Mean ± se Greenhouse roofs 1.80 ± 0.04a Horizontal hard surfaces -0.22 ± 0.27b Topsoil -0.28 ± 0.43b Vertical hard surfaces -0.11 ± 0.28b Tree bases -1.45 ± 0.09c

Sex ratio—As mentioned above, only one clump out of 95 had any apparent evidence of

sexual reproduction (senesced sporophytes) at collection, suggesting that sexual

reproduction is present but infrequent. From a non-destructive survey of collected

clumps, only one was identifiable as likely exclusively male, and no female sex structures

beyond those bearing sporophytes were observed. However, sex structures may be

hidden by new growth (personal observation), requiring an intensive destructive survey

for identification. The absence of sexual reproduction provides confidence that in most

cases culturing one shoot will reveal the majority sex of a clump. In the greenhouse

common garden, some plants expressed sex by 7 weeks after planting (10 clones

61 representing 9 unique isolates). In common garden, three isolates (one from horizontal hard surfaces and two from topsoil) produced no observable sex structure due to poor health or death of clones and were excluded from the sex ratio analyses.

By logistic regression, the overall likelihood of being a particular sex was not associated with exposure level. One habitat type (topsoil) had only females among those plants that expressed sex and was excluded from the logistic regression due to a lack of convergence (total N = 58). The overall model including habitat, exposure, and their interaction was not significant using likelihood ratio (Χ2 = 4.77, df = 7, P = 0.69). No significant effect was found for habitat (Χ2 = 3.08, df = 3, P = 0.38), exposure (Χ2 = 0.29, df = 1, P = 0.59), or their interaction (Χ2 = 1.16, df = 3, P = 0.76). Dropping the interaction also did not lead to significance for the model (Χ2 = 3.61, df = 4, P = 0.46).

Furthermore, when sex-expressing plants lacking a canopy photo were added

(total N = 89), the overall sex ratio was highly female-biased, but sex ratio was not heterogeneous among habitats. Extreme deviance from unity in all five habitats is very unlikely to be due to chance alone. Overall and individual habitat sex ratios are summarized in Table 3.2.

Table 3.2. Habitat and overall sex ratios for Bryum argenteum as proportion male with associated standard errors. Tests for individual habitats and pooled sex ratios indicate significant deviation from unity. The total G-test assesses the likelihood that as extreme deviance from unity in all five habitats would be observed due to chance. The heterogeneity G-test assesses whether sex ratio is heterogeneous among habitats.

Source n Proportion male G df P Greenhouse roof 16 0.125 ± 0.085 9.370 1 0.0022 Horizontal hard 34 0.118 ± 0.056 21.664 1 <0.0001 Topsoil 5 0 ± 0 4.876 1 0.027 Tree bases 16 0.125 ± 0.085 9.370 1 0.0022 Vertical hard 18 0.222 ± 0.101 5.538 1 0.019 Pooled 89 0.135 ± 0.036 49.468 1 <0.0001 Total -- -- 50.817 5 <0.0001 Heterogeneity -- -- 1.349 5 0.93

Leaf morphology in the field—The overall MANOVA was significant for the effect of

habitat (Wilks’ λ = 0.50, F16,162.56 = 2.59, P = 0.0013), exposure (Wilks’ λ = 0.78, F4,53 =

3.65, P = 0.011), and the interaction (Wilks’ λ = 0.57, F16,162.56 = 2.06, P = 0.012) on the

morphological variables. However, the results of the univariate analyses were mixed.

Total leaf area—Total leaf area was not significantly affected by habitat (F4,56 = 0.79, P =

0.53), exposure (F1,56 = 0.31, P = 0.58), or the interaction (F4,56 = 1.56, P = 0.20).

Relative achlorophyllous area—Relative achlorophyllous area was significantly affected

by habitat (F4,56 = 2.66, P = 0.042), exposure (slope = 0.35, F1,56 = 8.17, P = 0.0060,

2 overall R = 0.30), and the interaction (F4,56 = 3.61, P = 0.011). In both tree bases (slope

2 = 0.44, F1,13 = 4.91, P = 0.045, R = 0.27) and greenhouse roofs (slope = 1.55, F1,13 =8.35,

P = 0.013, R2 = 0.39) relative achlorophyllous area had significant positive relationships with exposure (Figure 3.3A). For horizontal hard surfaces, the relationship tended to be

2 negative (slope = -0.12, F1,13 = 3.57, P = 0.082, R = 0.22). The contrast between the

63 slopes of tree bases and greenhouses tended toward significance (F1,56 = 3.53, P = 0.065).

The slopes of both greenhouse roofs and tree bases were different from the slope of horizontal hard surfaces (F1,56 = 8.97, P = 0.0041 and F1,56 = 5.67, P = 0.021, respectively). Because the slopes of the relationship between exposure and proportion achlorophyllous leaf area differ between habitats, a comparison between habitat means is less meaningful. A better comparison would seek habitat differences at common exposure level, but not all habitats have overlapping ranges of exposure. Therefore, any projection of trait values onto a common exposure level would not be biologically relevant. When I ignored the relationship of achlorophyllous leaf area with exposure, pairwise t-tests showed a significant difference only between horizontal hard surfaces and tree bases (P < 0.05; Table 3.3), indicating tree bases were among the smallest and horizontal hard surfaces among the greatest in relative achlorophyllous area.

A 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Relative achlorophyllousarea leaf -2.5 -1.5 -0.5 0.5 1.5 2.5 B 4 3.5 3 2.5 2 1.5 1

Leaf length:width ratio 0.5 0 -2.5 -1.5 -0.5 0.5 1.5 2.5

Greenhouse roofs C Horizontal hard 0.3 Topsoil Vertical hard 0.25 Tree bases

length 0.2

0.15 apex 0.1

Relative 0.05

0 -2.5 -1.5 -0.5 0.5 1.5 2.5 Exposure

Figure 3.3. The influence of exposure and five habitat types on (A) relative achlorophyllous area, (B) leaf length:width ratio, and (C) relative apex length of Bryum argenteum in the field. Habitats include (isolate n in parentheses): greenhouse roofs (15), horizontal hard surfaces (15), topsoil (7), vertical hard surfaces (15), and tree bases (14). Solid lines represent significant (P < 0.05) linear relationships and dashed lines represent a tendency (0.05 < P < 0.1) for linear relationships between a leaf morphological traits and exposure within a habitat.

Table 3.3. Habitat trait means of Bryum argenteum from the field and trait least square means from the greenhouse common garden (CG) with associated standard errors. Habitats were greenhouse roofs (GH), horizontal hard surfaces (HH), topsoil (TS), vertical hard surfaces (VH), and tree bases (TB). Isolate sample sizes for field and common garden are denoted by nF and nC, respectively. To avoid pseudo-replication, habitat least square means were estimated for common garden plants using untransformed data in the same statistical model as reported for transformed data (see Methods). Superscript letters indicate significant differences among habitats using pairwise t-tests.

2 Hab nF, nC Leaf area (mm ) Rel. achl. area Length:width Relative apex length

Field CG Field Field CG Field CG GH 15, 14 0.11 ± 0.01 0.05 ± 0.01 0.35 ± 0.05ab 1.71 ± 0.22b 2.24 ± 0.26b 0.08 ± 0.01c 0.32 ± 0.03b HH 15, 13 0.11 ± 0.01 0.06 ± 0.01 0.40 ± 0.04a 2.34 ± 0.18a 2.44 ± 0.26b 0.09 ± 0.01bc 0.39 ± 0.03a TS 7, 5 0.11 ± 0.02 0.07 ± 0.01 0.36 ± 0.08ab 2.19 ± 0.17ab 3.60 ± 0.41a 0.12 ± 0.01ab 0.36 ± 0.04ab VH 15, 14 0.10 ± 0.01 0.05 ± 0.01 0.33 ± 0.05ab 1.77 ± 0.09b 3.15 ± 0.23a 0.15 ± 0.02a 0.31 ± 0.02b

66 TB 14, 15 0.14 ± 0.02 0.05 ± 0.01 0.24 ± 0.04b 1.82 ± 0.13b 2.51 ± 0.24ab 0.08 ± 0.01c 0.39 ± 0.02a

Leaf length:width ratio—Leaf length:width ratio was significantly affected by habitat

(F4,56 = 3.25, P = 0.018) and had a trend with exposure (slope = - 0.47, F1,56 = 3.71, P =

2 0.059, R = 0.25), but the interaction was not significant (F4,56 = 1.78, P = 0.15). Plants

on greenhouse roofs tended to have a negative relationship between leaf length:width

2 ratio and exposure (slope = - 2.69, F1,13 = 4.55, P = 0.053, R = 0.26; Figure 3.3B). The

relationship between exposure and leaf length:width ratio was not significant for any

other habitat. Pairwise t-tests among all five habitats, ignoring exposure, showed that

horizontal hard surfaces had significantly greater length:width ratios than greenhouse

roofs, tree bases, and vertical hard surfaces (P < 0.05 for each; Table 3.3).

Relative apex length—Relative apex length was significantly affected by habitat (F4,56 =

3.86, P = 0.0077), tended toward a positive relationship with exposure (slope = 0.15, F1,56

= 3.96, P = 0.051, overall R2 = 0.40), and there was a slight tendency for an interaction

(F4,56 = 2.07, P = 0.098). Greenhouse roofs showed a positive relationship between exposure and relative apex length (slope = 1.92, R2 = 0.37, P = 0.017), while plants from

vertical hard surfaces tended to have a negative relationship (slope = - 0.19, R2 =0.26, P =

0.064; Figure 3.3C). The slope of greenhouse roofs was different from the slope of

vertical hard surfaces (F1,56 = 7.21, P = 0.0095). These contrasts should be interpreted

cautiously due to the non-significant interaction between habitat and exposure. Using

pairwise t-tests among all five habitats, ignoring exposure, vertical hard surfaces had

significantly greater relative apex lengths than horizontal hard surfaces, greenhouse roofs,

and tree bases. Topsoil plants also had significantly greater relative apex lengths than

greenhouse roofs and tree bases (P < 0.05 in both cases; Table 3.3). Surprisingly, tree

67 bases and greenhouse roofs did not differ in relative apex length. Relative apex length was significantly positively correlated with proportion achlorophyllous area (Table 3.4).

All other partial correlations were not significant.

Table 3.4. Partial correlation coefficients of field leaf morphological variables of Bryum argenteum. Residuals were used from the MANOVA testing the effects of habitat, exposure, and the interaction on relative apex length, leaf length:width ratio, relative achlorophyllous area, and total leaf area.

Relative Variables Length:width achlorophyllous Leaf area area Relative 0.07ns 0.40* — 0.22ns apex length

Length:width 0.19ns — 0.16ns

Relative achlorophyllous — 0.0006ns area Notes: *P < 0.01, nsP ≥ 0.1

Pigment content in the field—Chlorophyll a:b ratio averaged 2.77 ± 0.16 and ranged from 1.29 to 9.88. Carotenoids:chlorophyll ratio averaged 0.28 ± 0.01 and ranged from

0.14 to 0.72. The MANOVA from the pigment ratios indicated that there were no significant effects of total light (Wilks’ λ = 0.98, F2, 52 = 0.42, P = 0.66), habitat (Wilks’ λ

= 0.88, F8, 104 = 0.83, P = 0.58), or the interaction (Wilks’ λ = 0.84, F8, 104 = 1.22, P =

0.29). Chlorophyll a:b ratio and carotenoid:chlorophyll ratio are significantly positively correlated using the MANOVA partial correlation (r = 0.75, P < 0.0001).

Across all exposure levels and habitats, the average total chlorophyll content was

8.09 ± 1.00 µg · mg-1 and ranged from 0.40 to 37.4 µg · mg-1. Total carotenoid content averaged 2.04 ± 0.24 µg · mg-1 and ranged from 0.26 to 9.30 µg · mg-1. The MANOVA from total chlorophyll and total carotenoids together was significant for the effect of total

light (Wilks’ λ = 0.85, F2, 47 = 4.07, P = 0.023), habitat (Wilks’ λ = 0.70, F8, 94 = 1.22, P =

0.0280), and the interaction (Wilks’ λ = 0.67, F8, 94 = 2.60, P = 0.013). There was a

significant positive partial correlation between total chlorophyll and total carotenoids (r =

0.95, P < 0.0001).

Total chlorophyll—Total chlorophyll tended to relate positively to light intensity (slope =

2 0.19, F1, 48 = 3.82, P = 0.057, overall R = 0.27), but there were no significant effects of habitat (F4, 48 = 1.72, P = 0.16) or the interaction (F1, 48 = 1.75, P = 0.16). Only

greenhouse roofs had a significant positive relationship between light and total

2 chlorophyll (slope = 0.96, F1, 9 = 8.91, P = 0.015, R = 0.50; Figure 3.4A). In all other

habitats, the relationship was not significant.

A 40

) 35 1 - 30

µg · mg µg 25 20 15 10

Total chlorophyll ( 5 0 0 20 40 60 B 10 9 ) Greenhouse roofs 1 - 8 Horizontal hard 7 Topsoil Vertical hard µg · mg µg 6 Tree bases 5 4 3 2 Total carotenoids ( carotenoids Total 1 0 0 20 40 60 -2 -1 Light intensity as PPFD (mol · m · day )

Figure 3.4. The influence of exposure and five habitat types on (A) total chlorophyll per dry mass and (B) total carotenoids per dry mass of Bryum argenteum in the field. Habitats include (isolate n in parentheses): greenhouse roofs (11), horizontal hard surfaces (15), topsoil (7), vertical hard surfaces (13), and tree bases (12). A solid line represents a significant (P < 0.05) linear relationship and a dashed line represents a tendency (0.05 < P < 0.1) for a linear relationship between a pigment trait and exposure within a habitat.

Total carotenoids—For total carotenoids, there was a significant positive relationship

2 with light intensity (slope = 0.20, F1, 48 =6.23, P = 0.016, overall R = 0.30), a significant effect of habitat (F4, 48 = 2.69, P = 0.042), and a tendency for the interaction (F4, 48 = 2.47,

P = 0.057). Greenhouse roofs had a positive relationship between total carotenoids and

2 light (slope = 1.07, F1, 9 = 14.52, P = 0.0042, R = 0.62), but tree bases tended to have a negative relationship between total carotenoids and light (slope = - 0.059, F1, 10 = 3.80, P

= 0.080, R2 = 0.28; Figure 3.4B). All other within-habitat relationships were non- significant. The slope of greenhouse roofs was different from the slope of tree bases (F1,

48 = 7.65, P = 0.0080). Using pairwise t-tests of habitat means and ignoring light,

horizontal hard surfaces (isolate n = 15, 2.94 ± 0.69 µg · mg-1) had significantly more

carotenoids than both vertical hard surfaces (isolate n = 13, 1.28 ± 0.90 µg · mg-1) and

topsoil (isolate n = 7, 1.06 ± 0.62 µg · mg-1), and tree bases (isolate n = 12, 2.21 ± 1.48

µg · mg-1) had significantly more carotenoids than vertical hard surfaces (P < 0.05 in

each case).

Morphology and growth in a greenhouse common garden—The doubly multivariate

repeated measures analysis including leaf area, length:width ratio, relative apex length,

shoot length and shoot number at three weeks showed significant effects of response (a

dummy variable representing all dependent variables), the interaction between response

and light, and the three-way interaction between response, light, and habitat. The

response effect simply shows that the dependent variables behave differently. The

interactions of response with light and with light and habitat indicate that light and the

interaction between light and habitat have overall significant effects across the dependent

variables. There was also a tendency for habitat to have an overall effect on the

dependent variables, which is shown by the nearly significant interaction between

response and habitat. The effect of sex and interactions with sex were not significant,

which is not surprising considering only three males were included. See Table 3.5.

Table 3.5. Results of the doubly multivariate analysis for greenhouse morphology and growth of Bryum argenteum. “Response” is a dummy variable representing all five morphological and growth variables (leaf area, length:width ratio, relative apex length, shoot length and shoot number at three weeks). An interaction between “response” and another independent variable shows that there is a significant overall effect of the independent variable when considering morphology and growth responses together.

Source Wilks' λ df F P Response 0.03 5, 32 195.97 <0.0001 Response*habitat 0.41 20, 107.08 1.65 0.055 Response*sex 0.87 5, 32 0.94 0.47 Response*light 0.65 5, 32 3.39 0.014 Response*light*habitat 0.40 20, 107.08 1.71 0.042 Response*light*sex 0.87 5, 32 0.96 0.46

Total leaf area—Leaves had significantly greater area in high light compared to low light

2 (0.067 ± 0.005 vs. 0.041 ± 0.003 mm , F1, 55 = 24.14, P < 0.0001), but leaf area was not

affected by sex (F1, 55 = 0.42, P = 0.52), habitat (F4, 55 =0.86, P = 0.49), or the interaction

between habitat and light (F4, 55 =0.82, P = 0.52; Figure 3.5A). The interaction between

sex and light was not significant and was dropped from the model.

A 0.14

0.12 )

2 0.1

0.08

0.06 Leaf area (mm Leaf area 0.04

0.02 12 12 10 11 5 3 14 11 13 12 0 B 6

4 ratio

2 length:width

Leaf 1 12 12 10 11 5 3 14 11 13 12 0

C 0.5 0.45 High light 0.4 Low light 0.35 0.3 0.25 0.2 0.15 Relative awn length 0.1 0.05 12 12 10 11 5 3 14 11 13 12 0 Greenhouse Horizontal Topsoil Vertical hard Tree bases roofs hard surfaces surfaces

Figure 3.5. Means ± SE by habitat origin of high and low light intensity effects on (A) leaf area, (B) leaf length:width ratio, and (C) relative apex length of Bryum argenteum in the greenhouse common garden. Only isolates that expressed sex were used. Sample sizes for means appear on their associated bar.

Leaf length:width ratio—Length:width ratio was significantly affected by habitat (F4, 55

=3.43, P = 0.014) and associated negatively with light level (2.57 ± 0.13 in high light vs.

3.19 ± 0.16 in low light, F1, 55 = 18.00, P < 0.0001), but was not affected by sex (F1, 55 =

2.69, P = 0.11) or the interaction between habitat and light (F4, 55 =1.55, P = 0.20; Figure

3.5B). The interaction between sex and light was not significant and was dropped. T-

tests showed that greenhouse roof and horizontal hard surface plants had smaller leaf

length:width ratios than those from vertical hard surfaces and topsoil (P < 0.05 in each

case; Table 3.3).

Relative apex length—Relative apex length tended to be affected by habitat (F4, 55 = 2.49,

P = 0.054), but not by light level (0.35 ± 0.01 in high light vs. 0.34 ± 0.02 in low light,

F1, 55 = 0.05, P = 0.82), sex (F1, 55 = 0.14, P = 0.71), or the interaction between habitat and light level (F4, 55 = 1.19, P = 0.33; Figure 3.5C). The non-significant interaction between

sex and light was dropped from the model. T-tests showed that greenhouse roofs plants

had significantly smaller relative apex lengths than horizontal hard surfaces, vertical hard

surfaces, and tree bases (P < 0.05 for each; Table 3.3).

Shoot length—There were no significant effects of light level (F1, 55 = 0.56, P = 0.46), sex

(F1, 55 = 1.14, P = 0.29), habitat (F4, 55 = 1.18, P = 0.33), or the interaction of habitat and

light level (F4, 55 = 0.83, P = 0.51) on shoot length. The interaction between light level

and sex was significant (F1, 55 =4.37, P = 0.041; Figure 3.6A). T-tests showed that high

light females had longer shoots than low light females (P < 0.05) and tended to have

longer shoots than high light males (P = 0.054).

A 5 4.5 4 3.5 3 2.5 47 2

Shoot length (mm) 1.5 1 46 3 7 0.5 0

B 25

15 56 10

Shoot number at 3 weeks 5 53 8 8

C 140

Female ) 120 2 -

cm Male 100

80 53 60

Shoot density (shoots · 20 49 5 8

0 Low light High light

Figure 3.6. Mean ± SE by sex of high and low light intensity effects on (A) shoot length, (B) three week shoot production, and (C) three month shoot density of Bryum argenteum in the greenhouse common garden. Only isolates that expressed sex were used. Sample sizes for means appear on their associated bar.

Shoot production rate and shoot density—After three weeks, shoot number was higher in

high light than low light (F1,58 = 28.79, P < 0.0001) and females tended to have more

shoots than males (F1,58 = 3.26, P = 0.076; Figure 3.6B). Shoot production rate tended to

be affected by habitat (F4,58 = 2.34, P = 0.066), but not by its interaction with light

intensity (F4,58 = 1.54, P = 0.20). The non-significant interaction between sex and light was dropped. Using pairwise t-tests, plants from vertical walls (isolate n = 15, 18.89 ±

3.12) produced more shoots (P < 0.05 for each) than plants from horizontal hard surfaces

(isolate n = 14, 11.80 ± 3.46), greenhouse roofs (isolate n = 15, 9.34 ± 3.37), and tree

bases (isolate n = 15, 6.88 ± 3.27).

Shoot density followed shoot production rate. After three months, shoot density

tended to be affected by sex (F1,58 = 2.91, P = 0.093) and the interaction between sex and

light (F1,58 = 3.32, P = 0.074), but was not significantly affected by habitat (F4,58 = 1.38, P

= 0.25), light level (F1,58 = 0.14, P = 0.71), or their interaction (F4,58 = 0.69, P = 0.60).

Using pairwise t-tests, high light females had higher shoot density than low light females and high light males (P < 0.05 for each; Figure 3.6C).

Correlations between leaf morphology and growth—Leaf area was significantly

positively correlated with shoot length, shoot production at three weeks, and shoot density. Furthermore, shoot length, shoot production, and shoot density were significantly positively correlated with each other. Lastly, leaf length:width ratio tended

to be positively correlated with both shoot length and shoot density. Greenhouse

correlations are summarized in Table 3.6.

76 Table 3.6. Correlation coefficients for pairwise correlations among leaf morphology (leaf length:width ratio and relative apex length) and measures of growth (shoot length, shoot production rate, and shoot density) of Bryum argenteum in the greenhouse common garden. Relative apex Shoot Shoot Variables Length:width Shoot length length production density Leaf area -0.10ns -0.14ns 0.55*** 0.22* 0.36**

Length:width -0.12ns 0.19t 0.05ns 0.18t

Relative apex -0.09ns 0.02ns 0.06ns length Shoot Length 0.41*** 0.53*** Shoot 0.47*** production Notes: nsP ≥ 0.1, tP < 0.07, *P < 0.05, **P < 0.001, ***P < 0.0001.

Relationships between traits in the field and common garden—Significant relationships between common garden and field traits were only found for greenhouse roofs (analyses for other habitats not shown). For leaf area, the effect of exposure slightly tended to be

2 positive (slope = 0.12, F1, 9 = 3.37, P = 0.099, overall R = 0.59) and common garden leaf area was significantly positively related to field leaf area (slope = 0.17, F1, 9 = 6.77, P =

0.03). For leaf length:width ratio, the effect of exposure was significantly negative (slope

2 = - 3.35 , F1, 9 = 9.82, P = 0.012, overall R = 0.52), and although common garden leaf length:width ratio was not significant related to field length:width ratio (F1, 9 = 0.01, P =

0.93), I note the improvement in the overall R2 compared to the previous analysis of exposure alone. For relative apex length, the effect of exposure was significantly positive

2 (slope = 0.58, F1, 9 = 5.34, P = 0.046, overall R = 0.56), and common garden relative apex length tended to relate positively to field relative apex length (slope = 0.79, F1, 9 =

4.00, P = 0.077).

Discussion

The sex ratio of Bryum argenteum in Lexington, KY, while highly female-biased, does not vary along an exposure gradient but may be explained by sex-specific patterns in shoot size and production rates identified in a common garden experiment. Though morphological variation along environmental gradients in bryophytes are often noted, these are rarely tested in both field and common garden studies (but see e.g.

Vanderpoorten and Jacquemart 2004; Buryová and Shaw 2005; Groen et al. 2010 a,b;

Reynolds and McLetchie 2011;). Because field assessment of morphology in B. argenteum is likely complicated by responses to a variety of environmental factors, my study highlights the advantage of coupling field studies with common garden experiments in bryophytes. Additionally, including the same genetic individuals in both experiments increased comparability and allowed me to connect common garden traits to field traits (i.e. genetic signature) in some cases. In the field, leaf morphology and pigment physiology varied with exposure and habitat. I found that leaf length:width ratio but not relative apex length responded plastically to light intensity. Apices are plastic to other environmental variables and appear to serve a different function in B. argenteum compared to other moss species. Intriguingly, when strong patterns were present, total chlorophyll and total carotenoids increased with light intensity, indicating that B. argenteum increasingly self-shades as light increases. In the common garden, genetic differentiation among habitats of origin may be linked to novel selection pressures created by anthropogenic environmental alteration, paralleling studies in both seed plants and bryophytes.

Field sex ratio patterns—Because B. argenteum sex ratios are less female-biased among more mesic sites in the Western US (Stark et al. 2010), I hypothesized that sex ratio would vary similarly along a local canopy openness/light (exposure) gradient within a single-climate. However, I found no segregation of the sexes (SSS) along the gradient and no variation in sex ratio among habitats, which does not support my hypothesis.

Males were present in every habitat except topsoil and all along the exposure gradient.

The overall sex ratio of 0.135 ± 0.036 (proportion male) in urban Lexington, KY was highly female-biased and is comparable to and slightly more female-biased than the montane and pooled city sex ratios found by Stark et al. (2010). Furthermore, by culturing field-collected samples to sex expression, my study minimizes the potential inaccuracies of surveying only sex-expressing plants in the field (Stark et al. 2010; Field et al. 2012). My results suggest that SSS in B. argenteum does not occur on a local scale at least within a temperate climate. Perhaps environmental conditions in Kentucky are too mild to promote SSS along environmental gradients. A climate intermediate between that of Kentucky and the lowland desert of the Western USA where populations are all female (Stark et al. 2010) might promote local-scale SSS more readily. Alternatively, sporophyte production may be relatively high in Kentucky’s temperate climate, permitting more frequent colonization of habitats by males and females via spores, which are produced at a 0.5 sex ratio (Stark et al. 2010). However, the genetic differentiation among habitats (see further discussion below) argues either against high colonization rates or for high selection against genotypes with certain phenotypes.

Sex differences in common garden—Despite my study having relatively few males, some sex-specific patterns or tendencies were present. Contrary to my expectations, these sex-specific patterns were only related to measures of growth and not morphological responses to light intensity. Shoot length was positively associated with light intensity for females while males in both light intensities have comparable shoot lengths to females in low light. Further, females tend to produce shoots faster than males.

The shoot length difference is consistent with the sex differences found for clump height

(Chapter Five) and sexually mature shoot length (Horsley et al., 2011). Furthermore, the shoot length difference was not related to a tradeoff with sexual reproduction because no males were expressing sex, though some females were, when shoot length was measured.

Moore et al. (2016) also found no correlation between clump height and proportion of shoots expressing sex. The tendency for females to produce more shoots than males in the first three weeks was followed by females having higher shoot density than males at three in high light. The latter result is inconsistent with the higher male shoot density found for male clumps in another study (Chapter Five). However, the density in the present study was measured at an earlier time and may reflect higher female growth rate rather than shoot density at latter stages of clump formation. Together these results allow me to hypothesize that at early stages females are able to capitalize on high light better than males to produce longer and more shoots. The shoot length difference is suggestive of the formation of morphological sex differences potentially related to water retention for fertilization before sexual maturity (Chapter Five). Juvenile sex differences in growth rate are also present in Juniperus virginiana (gymnosperm) where males grow faster than females potentially to increase future mating success through canopy emergence (Quinn

and Meiners 2004). If females are able to produce more and taller shoots than males at early growth stages, females may have a competitive advantage over males. The overgrowth of males by females would help explain the female-biased sex ratios in B. argenteum and is worthy of future study.

Variation among habitats in the relationship between morphology and exposure in the field—Among greenhouse roofs in the field, the control for extraneous variables afforded by uniform substrate (glass and steel), lack of plant competitors, and a canopy of almost exclusively permanent buildings likely explains why significant trait patterns occurred there most often (except relative achlorophyllous area among tree bases). Furthermore, genetic variation among individuals in leaf area and relative apex length (tendency) were detectable only among greenhouse roofs. Tree bases likewise have fairly uniform substrate and canopy type (mostly Quercus section Lobatae) but also house more cryptogam competitors. Higher variation in substrate, soil moisture, canopy type, and competition from vascular plants and cryptogams within other habitats potentially mask exposure’s influence on traits. If winter canopy affects early summer traits, then canopy variation (deciduous trees vs. buildings) may explain the high variability within some habitats and similar trait ranges across habitats.

In addition, the measured traits are likely influenced by unmeasured traits (e.g. shoot density), their relationship to exposure, and genetic variation in plasticity among habitats, which was evident when all traits were considered together in common garden.

Shoot density affects water relations and self-shading and is positively related to exposure in other mosses (Tobias and Niinemets 2010; Elumeeva et al. 2011). However,

81 exposure’s effect on field shoot density is unpredictable from common garden data because clumps were assessed at early stages of formation, shown by the direction of the sex difference (i.e. at later stage, male clumps are denser than females; Chapter Five).

Overall morphology and growth responses to exposure in the field and light intensity in common garden—In the field, leaf morphology generally varied with exposure as predicted (when patterns were discernable), and responses to light intensity were more consistent across habitats of origin in the common garden, aiding in interpretation of field results. In common garden, one aspect of leaf morphology, leaf length:width ratio, responded to light intensity in a manner consistent with field patterns. The combination of field patterns with the lack of achlorophyllous tips in common garden plants suggests that achlorophyllous area responds to the moisture changes along the exposure gradient.

For relative apex length, results were not consistent between field and common garden, suggesting a different function contrary to my expectation. Morphological plasticity was also evident in common garden, which was expected in B. argenteum because of the climatic breadth it occupies. Male rarity in the field prevented testing for sex differences in morphology and physiology.

Field patterns and plasticity in common garden show that in B. argenteum leaf length:width ratio increases (i.e. more lance-shaped) at lower light intensities. The similarity to predicted self-shading avoidance responses of monocot-like angiosperms, which are expect to increase leaf length:width ratio in low light, is striking (Takenaka

1994). Therefore, B. argenteum is likely similarly avoiding self-shading in low light, which is corroborated by light-limited shoot production rates and leaf areas in the

common garden. In high light, a smaller leaf length:width ratio (i.e. more deltoid) could

be interpreted as increasing self-shading and preventing light damage in lower leaves.

Furthermore, my observations indicate that deltoid leaves are positioned more vertically and have a more shingle-like arrangement, potentially reducing water loss of individual shoots and of the clump by allowing higher shoot density. In bryophytes, a shingle-like

leaf arrangement (Schofield 1982; Romero et al. 2006) and high shoot density are

considered adaptations to decrease water loss (Elumeeva et al. 2011). Similarly, in

Juniperus sp. (gymnosperm) leaves shift morphology from needle-like as juveniles to

scale-like as adults presumably to achieve higher photosynthetic rates as juveniles while

conserving water as adults (Miller et al. 1995).

In the field, relative achlorophyllous area related positively to exposure within

both greenhouse roofs and tree bases. In common garden, plants were constantly wet and

had little to no achlorophyllous area. Together these patterns suggest that

achlorophyllous area may respond to moisture or desiccation frequency rather than light.

However, there was a weak tendency for a negative relationship between relative

achlorophyllous area and exposure within horizontal hard surfaces. Reflective

achlorophyllous leaf tips are considered an adaptation to protect against light exposure in

xeric environments (Robinson 1971; Glime 2007a) and may perform an analogous

function to the reflective trichomes of some vascular plants (Fahmy 1997; Rossatto and

Kolb 2010). Because achlorophyllous tips are seemingly suited for light reflection, their

lack of relationship with light in common garden and potential relationship with moisture

are initially counterintuitive. However, in desiccation tolerant plants, the protection of

photosystems from light damage during desiccation is critical (Glime 2007a). In B.

argenteum, though light reflection during hydration may be important, the greater silvery

appearance of dry individuals indicates a greater need for light protection. Future

experiments are needed to test the effects of moisture availability or desiccation

frequency on achlorophyllous area.

In bryophytes of arid regions or exposed habitats, awns are thought to reflect

light, prevent water loss (Glime 2007a, 2015), and, for the moss Syntrichia caninervis,

demonstrably gather dew (Tao and Zhang 2012), but the patterns in the present study

suggest that apex function in B. argenteum does not follow these expectations. In the

field, relative apex length related positively to exposure within greenhouse roofs and

positively correlated with relative achlorophyllous area. However, the least and most

exposed habitats (tree bases and greenhouse roofs, respectively) had the shortest relative

apex lengths, and vertical hard surfaces tended to have a negative relationship with

exposure. Though seemingly indicative of a reflective role for apices, the positive

correlation with relative achlorophyllous area may simply be the result of apices being

the distal portion of achlorophyllous tips. In common garden, relative apex length had no

relationship with light and was shortest again among greenhouse roofs (along with

vertical hard surfaces). Furthermore, a paired t-test showed that high light common

garden plants had higher relative apex lengths than their field counterparts (t53 = 17.94, P

< 0.0001; Figure 3.7), which is inconsistent with field greenhouse roof patterns and suggests that relative apex length responds positively to moisture or nutrient availability

(common garden substrate was constantly moist soil).

0.4 Greenhouse roofs 0.2 0.35 0.15 0.3 0.1 0.05 0.25 0 0.2 1.3 1.5 1.7 1.9 Exposure 0.15 Relative Relative awn length 0.1

0.05

0 Common garden Field

Figure 3.7. Mean ± SE of relative apex length of Bryum argenteum in the greenhouse common garden and field. Inset: line of best fit for the relationship between relative apex length and exposure for greenhouse roofs in the field.

The relative increase of apex length of cultured B. argenteum contrasts the loss of awns for cultured S. caninervis (Reynolds and McLetchie 2011). Thus, while awns are important for water gain in S. caninervis, apices apparently relate to water loss in B. argenteum because they are shorter in more water-stressed conditions (field) than in the relatively mesic greenhouse common garden. I speculate that differences in leaf structure, including costa length and morphology, and leaf arrangement influence whether awns/apices contribute to water gain or loss. The clasping hydrated leaves of B. argenteum (Crum and Anderson 1981), compared to the spreading hydrated leaves of S. caninervis (Mishler 2007), may allow the apices to draw water through a capillary continuum formed by leaves, shoots, and rhizoids. Further investigation is needed to clarify the primary function of elongated apices in B. argenteum.

Based on seed plant patterns (Balaguer et al. 2001; Bragg and Westoby 2002), I expected a negative relationship between leaf area and exposure in the field or light

intensity in the greenhouse, but there was no pattern in the field and leaf area was

positively related to light intensity in the greenhouse. The lack of relationship in the field

suggests that B. argenteum does not adjust leaf area in response to exposure. The

positive relationship of leaf area with light intensity in the greenhouse is consistent with

light-limited growth evident also in the positive relationships between light intensity and

both shoot production and shoot density as mentioned above.

Pigment responses to light intensity in the field—Contrary to my predictions, patterns in

pigment ratios and absolute amounts were not consistent either with seed plant or many

bryophyte studies. The increase of pigment amounts with light intensity in one habitat is

especially puzzling. However, I suggest as did Schroeter et al. (2012) that B. argenteum

is essentially creating via changes in morphology its own shade—one that is increasingly

deep as light increases. Thus, the photosynthetic regions of plants at high light actually

experience a low-light environment.

In contrast to seed plant (Boardman 1977; Dale and Causton 1992; Sarijeva et al.

2007) and bryophyte patterns (Post 1990; Martínez-Abaigar 1994; Marschall and Proctor

2004; Núñez-Olivera et al. 2005; Groen et al. 2010b), increasing light intensity did not result in increased chlorophyll a:b and carotenoid:chlorophyll ratios in the present study.

The two pigment ratios were also highly positively correlated, which has been observed across several species of bryophytes (Marschall and Proctor 2004). Generally, in seed

plants, high light intensity is expected to increase chlorophyll a:b and

carotenoid:chlorophyll ratios through reductions in the accessory pigment chlorophyll b

and relative increases in light-protective carotenoids (Dale and Causton 1992; Demmig-

Adams 1998). However, no change in chlorophyll a:b ratio with light intensity (Tobias

and Niinemets 2010) and even a decrease in higher light (Rincón 1993; Hamerlynck et al.

2002) have been found in some bryophytes. Furthermore, no difference in chlorophyll a:b ratios and carotenoid:chlorophyll ratios between sun and shade were reported for

Antarctic B. argenteum var. muticum (Schroeter et al. 2012). The same study also reported similar averages to the present study for chlorophyll a:b ratio but slightly higher averages for carotenoid:chlorophyll ratios. My results for both ratios are most consistent with shade leaves of vascular plants (Demmig-Adams 1998). Therefore, leaf morphology variation, rather than shifts in chlorophyll a:b ratio, may permit B. argenteum to occupy high-exposure habitats, but the contribution of changes in xanthophyll cycle carotenoid ratios to high light acclimation, which were important for

Antarctic B. argenteum var. muticum (Schroeter et al. 2012), cannot be ruled out.

In addition to pigment ratios, total chlorophyll did not decrease as predicted with increasing light as found in seed plants (Boardman 1977; Dale and Causton 1992;

Sarijeva et al. 2007) and bryophytes (Glime 1984; Kershaw and Weber 1986; Martínez-

Abaigar et al. 1994; Tobias and Niinemets 2010) but rather increased in plants from greenhouse roofs. Total carotenoids increased similarly with light in plants from greenhouse roofs. Overall, total carotenoids positively correlated with total chlorophyll, leading to the relatively constant carotenoid:chlorophyll ratio. Because relative achlorophyllous leaf area increased with exposure in the same habitat, I expected that total pigments would decrease. Schroeter et al. (2012) observed a similar increase of chlorophyll and carotenoids, although on a per area basis, with light intensity in B. argenteum var. muticum and hypothesized that, as a plant with shade-adapted

photosystems (from a vascular plant perspective), the reflective achlorophyllous leaf tips

attenuated light even at high exposure enough to require increased photosynthetic

pigments (i.e. a shade response). Their hypothesis links well with the patterns I found for

relative achlorophyllous area. The question arises, however, as to why B. argenteum

would attenuate light to the extent of creating the equivalent of deep shade in full

sunlight. As discussed above, increases in achlorophyllous tips could protect against

light during desiccation. In high light environments, increased reflectivity during

desiccation may increase reflectivity during hydration as well, effectively attenuating

light more than required even for the essentially shade-type photosystem. In other words,

photoprotection during desiccation trades off with photosynthetic light availability during

hydration. I speculate that B. argenteum’s apparently unique system of light attenuation

is critical for its ability to colonize extreme habitats such as greenhouse roofs where B.

argenteum is the only plant species occurring in my study sites. An analogous pattern

occurs in some seed plants where pubescent vs. glabrous leaves in the same species have

higher chlorophyll, which is interpreted to result from light attenuation or quality change

due to pubescence (Liakopoulos et al. 2006; Martin et al. 2007), but this pattern is not

consistent (Hoof et al. 2008; Guerfel et al. 2009). Additionally, other morphological

changes related to reducing water loss (Elumeeva et al. 2011) such as increased shoot

density, which I did not measure in the field, could also increase self-shading in high light, increasing the need for more light-gathering pigments.

Implications of genetic differentiation—While demonstrating that B. argenteum is phenotypically plastic for several traits in response to light intensity, the common garden study confirmed, the existence of genetic differences in morphology and life history that could be ecotypic. Anthropogenically altered environments such as urban areas challenge plants with novel habitats and disturbance regimes, leading to plastic shifts in phenotype (Geng et al. 2007) or ecotypic variation via novel selection pressures

(Hufbauer et al. 2011). Previous work has noted that B. argenteum varies in morphology among habitats (Crum and Anderson 1981), but the contribution of plasticity and genetics to this variation is not well understood. In the field, relative achlorophyllous area, leaf length:width ratio, and relative apex length varied by habitat, and habitat significantly interacted with exposure for relative achlorophyllous area and nearly so for relative apex length. These differences hint at ecotypic variation, but they may be due to phenotypic plasticity linked to other environmental factors such as substrate type. Culture in common garden allowed the minimization of field effects because plants regenerated through the juvenile phase (protonemata). Trait variation in common garden linked to source habitat supports the idea that differences among habitats have promoted the evolution of ecotypes or provide strong filters on naturally occurring genetic variation.

Specifically, leaf length:width ratio was and relative apex length tended to be habitat- specific. In the common garden, greenhouse roof plants were among those with the smallest leaf length:width ratios, which aligns with their high exposure level in the field.

Surprisingly, vertical hard surface plants tended to have greater leaf length:width ratios than greenhouse roofs. Perhaps vertical walls strongly limit solar radiation through all seasons, leading to selection for relatively more lance-shaped leaves to limit self-shading.

Contrary to my predictions, greenhouse roofs and vertical walls had the smallest relative

apex lengths in common garden, perhaps indicating selection to reduce water loss (see

discussion above).

Additionally, habitat tended to affect shoot production rate, which suggests that

life history may vary among habitats as well. Life history traits, including growth rate,

have been shown to differ genetically among populations of B. argenteum from different

geographical regions (Longton 1981; Horsley et al. 2009) and from populations within a

region (Shaw et al 1989; Shaw and Albright 1990), but my study provides evidence that

life history may vary on a smaller scale. Tree bases and vertical walls represented the

lower and upper ranges of shoot production rate, respectively. Perhaps the unique

challenges of vertical walls including frequent disturbance, small substrate capillary water storage capacity/conduction (short hydration periods), and selection against large clump size (heavy clumps may fall) favor individuals with fast growth rates and sparse growth habitat. Indeed, most often B. argenteum lives as isolated shoots on vertical walls while other habitats more regularly allow clump formation (personal observation), which should permit longer active growth periods. A reciprocal transplant study after regeneration in common garden according to Såstad et al. (1999) would help clarify the adaptive significance, if any, of the differences between source habitats. My work here links to a growing body of work showing anthropogenically-derived microhabitat differentiation in seed plants (e.g. Jordan 1992; Linhart and Grant 1996; Wittig 2004;

Cheptou et al. 2008) and bryophytes (Hassel et al. 2005; Brzyski et al. 2014; Marks et al.

2016).

Conclusions

In B. argenteum from an urban area in a mesic temperate climate, the sexes do not

segregate along an exposure gradient, and males are not excluded by abiotic

environmental conditions from any exposure level and are likely present in all habitats

(though no males were found on topsoil). The sexes exhibited no sex-specific leaf morphological responses to light intensity. Rather, the greater shoot length and production of females relative to males suggests that females can outperform or outcompete males leading to female bias in all habitats. The presence of these sex differences among juveniles is likely connected to selection for future mating success linked to relationships with water, the gamete dispersal medium. The maintenance of males in the metapopulation then would depend on successful colonization through spore or asexual propagules, and the uniformity of sex ratio across habitats in the present study suggests colonization rate is also fairly uniform. Field patterns, when apparent, suggest that B. argenteum leaves respond to an increase in exposure by increasing relative achlorophyllous area, and decreasing length:width ratio. The pattern for leaf length:width ratio was confirmed in common garden as a response to light itself.

Relative achlorophyllous area likely responds to other light-correlated environmental factors potentially including dryness. Relative apex length likely also responds to moisture, but appears to increase rather than decrease as expected with increasing moisture, suggesting apices function differently in B. argenteum from awns in some other mosses. I found no evidence that B. argenteum decreases leaf area in order to avoid damage from high exposure, but leaf area can be light-limited. Contrasting expectations for vascular plants and bryophytes, the patterns in pigmentation suggest that B.

argenteum adjusts to high light by creating its own shade which becomes increasingly

deep as light increases. The apparent overcompensation in light protection likely protects

the photosystems during periods of desiccation. Despite short distances and the potential

for a uniform colonization rate, some traits vary genetically among habitats, suggesting strong selection against certain phenotypes. The potential link between genetic differentiation and anthropogenically-influenced microhabitats in B. argenteum parallels similar trait differentiation in vascular plants.

CHAPTER FOUR: DO SEX DIFFERENCES IN RESPONSE TO HEAT AND DESICCATION STRESS OCCUR IN THE MOSS BRYUM ARGENTEUM HEDW. (BRYACEAE)?

Abstract

In dioecious seed plants, apparent sex differences in stress tolerance correlate with

differential reproductive investment. In bryophytes, females are predicted to have lower

average reproductive investment due to low fertilization rates and higher stress-tolerance

than males. I tested the responses of shoots from field-collected (variable maturity) and

growth-chamber-grown (sexually immature and mature) clumps of the dioecious moss

Bryum argenteum to desiccation, dry heat, and wet heat, predicting that (1) females

would be less impacted and recover more quickly from stress than males, and (2) sex-

differences would be strongest among sex-expressing plants. Stress recovery was tracked using potential quantum efficiency of photosystem II (Fv/Fm). Additionally, regenerative

shoot production of sexually immature shoots was assessed after treatment. Female

sexually immature and sex-expressing shoots recovered more quickly from severe heat

stress than males with a stronger pattern among sex-expressing shoots. However,

regenerative shoot production was not sex-specific among sexually immature shoots.

Dimorphic patterns were absent from field-collected shoots and negligible for mature

vegetative shoots. Sex-expressing shoots had female-like while mature vegetative shoots had male-like stress recovery patterns, which may be related to lower resource pools for vegetative vs. sex-expressing shoots. At this time, a link between sex-specific stress responses and stress tolerance remains unclear. The sex-specific responses of sexually immature shoots may be genetically correlated with sexually mature resource investment patterns and linked with high resource expenditure (accelerated young shoot growth), which could explain the weakness of sex differences among mature vegetative shoots.

KEY WORDS: chlorophyll fluorescence; sexual dimorphism; physiology; bryophyte;

dioecious, stress tolerance; life-stage-specific traits

Introduction

In dioecious plants, sex differences in stress tolerance can lead to biased sex ratios and spatial segregation of the sexes in certain environments (Sinclair et al. 2012; Barrett and Hough 2013). Among seed plants, greater female reproductive investment compared to males is expected to cause higher female mortality and greater stress sensitivity than males through resource tradeoffs. This expectation is supported by frequent male-biased sex ratios in stressful habitats (Bierzychudek and Eckhart 1988; Delph 1999; Dawson and

Geber 1999; Barrett and Hough 2013) and less negative impacts of abiotic stressors for males compared to females in several species (Dawson and Geber 1999; Juvany and

Munne-Bosche 2015).

However, pinpointing when in the life cycle sex differences in stress tolerance arise is logistically difficult in long-lived dioecious species because of the lack of genetic

sex-markers, thus requiring sexually immature individuals to be measured and then

tracked until sex-expression. Consequently, the majority of studies have focused on

sexually mature plants (Dawson and Geber 1999; Barrett and Hough 2013; Juvany and

Munne-Bosche 2015). Very few studies have sought and found sex differences of any

kind among sexually immature individuals of long-lived species (e.g. growth rate in

Juniperus virginiana; Quinn and Meiners 2004), but a few have shown sex-specific

physiological stress responses in non-flowering plants derived from cuttings of mature

trees (e.g. Xu et al. 2008; Xu et al. 2010). These studies indicate the possibility that

differences in physiology related to stress tolerance are not directly linked to resource

tradeoffs with reproduction. Furthermore, the responses to physiological stress that link

to apparent sex differences in stress tolerance are not well understood (Juvany and

Munne-Bosche 2015).

Because of their high incidence of dioecy, biased population sex ratios, relatively

short life cycle, tractability for controlled experiments, and ability to recapitulate

development through a clonal juvenile phase (protonemata), bryophytes are ideal systems

to test for physiological sex differences related to stress tolerance and the timing of these

differences (i.e. before vs. after sexual maturity). Similar to seed plants, sex-specific

stress tolerance in bryophytes is predicted to correlate with reproductive investment

patterns, which are likely to be similar to seed plants (Stark et al. 2000). However, bryophyte females unlike males may rarely invest fully in sexual reproduction due to low fertilization rates caused by short sperm dispersal distances (Wyatt 1977; Reynolds 1980;

McLetchie 1992; Stark et al. 2000). Thus, females are expected to invest less than males on average in sexual reproduction and be more stress tolerant (McLetchie 1992; Stark et al. 2000, 2010). Bryophyte sex ratios, which are often female biased (Bisang and

Hedenäs 2005) and increasingly so in stressful environments (Stark et. al 2010), appear to reflect this investment difference. Nevertheless, several studies on bryophytes have demonstrated higher stress tolerance in females compared to males among sexually immature individuals (Newton 1972; Stieha et al. 2014; Marks et al. 2016), which is incongruent with predictions based on resource tradeoffs with reproduction.

The cosmopolitan moss Bryum argenteum has female-biased sex ratios and the degree of female-bias varies with an elevational gradient. In the western USA, low-land desert populations are entirely female while populations at mid-elevation have males

present. The climate is more mesic at mid-elevation compared to low-elevation sites, suggesting that males are less stress tolerant than females (Stark et al. 2010).

Furthermore, the combination of greater male pre-fertilization reproductive investment

(up to 24x more than females; Horsley et al. 2012) and low sporophyte production by females (see Results) suggests that the apparent sex difference in stress tolerance is consistent with a direct tradeoff with sexual reproduction.

I used B. argenteum to address two hypotheses: (1) the sexes differ in physiological responses to desiccation, dry heat, and wet heat stress, and (2) the presence of a sex difference depends on reproductive status. High temperatures during hydration can be especially stressful in desiccation tolerant bryophytes because high temperatures typically occur during desiccation, which, though also potentially stressful, prevents heat damage (Glime 2007b; Reynolds and McLetchie 2011). I predicted that females would be less impacted and recover more quickly from stressors than males, and that the sex difference would be strongest when males and females were investing in sex structures.

To test my hypotheses, I conducted three experiments. The first was an experiment where field-collected plants were tested for sex differences in recovery from desiccation and wet heat. The second and third tested the recovery responses of shoots from growth- chamber-grown sexually immature and sexually mature clumps, respectively, to desiccation, dry heat, and wet heat.

Methods

Study species—Bryum argenteum Hedw. (Bryaceae) is a dioecious, acrocarpous,

desiccation tolerant moss with a distribution spanning all seven continents where it is

found in human disturbed and natural habitats, including hot and cold deserts (Crum and

Anderson 1981; Longton 1981; Stark et al. 2010, Horsley et al. 2011). Sex ratio from

spore germination is near unity (Stark et al. 2010). Clonal reproduction occurs via specialized short, deciduous branches (bulbils) or fragmentation (Horsley et al. 2011).

Expression of sex is day-neutral and occurs readily in lab culture (Chopra and Bhatla

1981). Males invest more in sexual reproduction pre-zygotically (up to 24x) than females

(Horsley et al. 2011). Adult populations are generally highly female-biased (Stark et al.

2010).

Field collection procedures and habitat characterization—I collected plants from the

University of Kentucky campus and adjacent areas (approximately 385 ha). Using roads,

I divided campus into four approximately equal quadrants and walked along footpaths searching all possible habitats for plants during the course of four hours in each quadrant.

A coin flip at forks in paths was used to reduce bias. In total, I collected 96 clumps

(aggregations of shoots) of B. argenteum that were spaced at least 3m apart to reduce the chance of collecting clones. Only one clump showed evidence of being mixed-sex

(senesced sporophytes were present) and one was identifiably male. The sex of all others were not immediately identifiable by non-destructive surveys. The habitat of each clump was characterized as one of five types based on substrate: greenhouse roofs, horizontal hard surfaces, topsoil, vertical hard surfaces, and tree bases. I randomly selected 15 clumps from all habitats except topsoil, which had only seven clumps. All seven clumps from topsoil were used. Light environment (PPFD) and canopy openness was quantified for each clump (except one for which a photo could not be obtained) using hemispherical canopy photographs (180° fish-eye lens and digital camera, Coolpix 950, Nikon, Tokyo,

Japan) and the program WinSCANOPY (Regent Instruments, Ville de Québec, Québec,

Canada). Because canopy photographs were taken in summer and many clumps were under deciduous tree canopies, the PPFD was estimated from May 1 to June 9, 2012, which corresponded to the period from leaf-out of deciduous trees (data from the USA

National Phenology Network, www.usanpn.org) to clump collection. One shoot from each clump was planted on steam-sterilized local topsoil in 59 mL pots on a capillary mat in a greenhouse for culture to sex expression. Most clumps in nature are single sex

(Chapter Three), and I assumed that the sex of plants taken from the center of each clump would represent the majority sex, following Stark et al. (2010).

Recovery of field-collected shoots from desiccation—Clumps were generally dry when collected. Field-hydrated clumps were allowed to dry in the lab at room temperature inside paper packets. Following dehydration, three shoots were removed from the center of each clump and rehydrated. Dark-adapted samples (at least 20 min) were assessed for potential quantum efficiency of photosystem II (Fv/Fm) at 0, 2, 4, 6, 12, 18, 24, 30, 36, 48,

60, and 72 h following rehydration using an OS5-FL modulated chlorophyll fluorometer

(Opti-Sciences, Tyngsboro, MA, USA). Between readings, samples were allowed to recover in a 96-well plate in distilled water in a growth chamber (14 h/10 h, 20°C/17°C day/night with fluorescent lighting at approximately 50 µmol·m–2·s–1). Reductions in

Fv/Fm (i.e. photoinhibition) have been connected to a variety of plant stressors in both vascular plants (Maxwell and Johnson 2000) and bryophytes (Murray et al. 1993;

Reynolds and McLetchie 2011, Marks et al. 2016).

Recovery of field-collected shoots from wet heat—To test the effects of wet heat, three

shoots were again taken from the center of each field-collected clump and rehydrated for

72 h at room temperature. An initial Fv/Fm reading was taken, and then samples were

subjected to wet heat stress. Wet heat stress was induced by placing the hydrated shoot

sample in an oven at 40°C for 1 h in the dark. Recovery occurred in a growth chamber

as above and was tracked by chlorophyll fluorescence at 0, 2, 4, 6, 12, 18, 24, 30, 36, 48,

60, and 72 h after wet heat treatment.

Stress responses of shoots from growth-chamber-grown sexually immature clumps—

Nineteen male and 19 female plants were chosen from a stock population of B.

argenteum (for cultural details see Chapter Five, published as Moore et al. 2016). Stock

plants were maintained in a growth chamber for more than 1 y to remove field effects. A

single shoot tip was taken from the stock plants and planted on fresh steam-sterilized topsoil in a 9.6 mL Petri dish and placed in a growth chamber with 14 h/10 h, 20°C/17°C day/night with fluorescent lighting at approximately 50 µmol·m–2·s–1. Clumps were

allowed to dry before being watered. As soon as all cultures produced sufficient shoots,

five shoots from each clump were placed into each of four treatments in 96-well plates:

1) hydrated, room temperature; 2) hydrated, 45°C; 3) dehydrated, room temperature; 4) dehydrated, 45°C. Room temperature on the day of treatment was approximately 22°C, and relative humidity was approximately 53%. Few bryophytes tolerate greater than 45°C

when hydrated (Glime 2007b). For the hydrated treatment, shoots were placed in

distilled water for 1 h. For the dry treatments, shoots were dehydrated at room

temperature and ambient relative humidity for 1 h. Next, hydrated and dehydrated shoots

were placed in the dark either on a benchtop at room temperature or in an oven at 45°C

for 1 h. On the same day, plants were treated and assayed in three sets of 12-13 isolates

representing approximately equal numbers of males and females with samples of each

isolate in all four treatments (total samples for each set = 48-52) due to limitations in the number of fluorometer clips. During recovery, samples were placed in 96-well plates in distilled water back in the growth chamber where they were grown. I assayed Fv/Fm

before and directly after treatment and then at 24, 48, 72, and 96 h after treatment. After

96 h, each set of five shoots was planted on fresh steam-sterilized soil in a new 9.6 mL

Petri dish and placed back in the growth chamber to measure growth after stress. After

10 days, these plants were assayed for shoot production.

Stress responses of shoots from growth-chamber-grown sexually mature clumps—The four replicates planted for assessment of growth rate from the first growth chamber experiment yielded four replicates each of 17 males (2 male isolates died) and 19 females. These plants all completed another juvenile phase as protonemata and were cultured under the same conditions in a growth chamber until they expressed sex. To encourage fast growth and shorten the time to the production of sex structures, clumps were not allowed to dry during culture. Shoots were pooled across all available replicates of the same isolate to permit the collection of enough shoots for reliable chlorophyll fluorescence readings. Groups of up to four shoots with and without sex structures were placed in each of the same 4 treatment groups and allowed to recover as above. Thus, each isolate had groups of sex-expressing and groups of non-sex-expressing shoots in each of the treatments. Because of the large number of samples, shoots were treated and

100

assayed in 6 sets of 5-7 isolates (approximately equal numbers of males and females) in

all four treatments (40-56 samples in each set) staggered over the course of

approximately one month. During this month room temperature varied between

approximately 22-24°C and 55-62% relative humidity. Because of the potential for

variation between sets treated on different days, these were treated as blocks.

Measurements of Fv/Fm were taken before and directly after treatment and then at 24, 48,

72, 96, and 240 h (10 d) after treatment.

Statistical analyses—All statistical analyses were conducted in the program SAS 9.3 or

9.4 (SAS Institute, Cary, North Carolina, USA). In all cases, Fv/Fm values were logit-

transformed to improve normality.

Field-collected shoots—Light (PPFD) and % canopy openness were highly positively

correlated (analysis not shown; see Chapter Three). I used principal components analysis

to combine them into a single value (“exposure,” loading = 0.97 for each, explaining 93.4

% of variation). Three isolates (1 from horizontal hard surfaces and 2 from topsoil) failed

to express sex due to poor health or death in greenhouse culture, and these were excluded

from the analyses (final isolate N = 64). Both recovery from desiccation and wet heat

stress were analyzed using multivariate repeated measures ANOVA. Time was the

within-subject (i.e. repeated within-isolate) factor. The model of between-subject (i.e.

between isolate) factors was determined by using model reduction; an initial model was

fitted including sex, habitat, exposure, and the interactions of exposure with sex and

habitat. Interactions of sex and habitat were excluded because both sexes were not

101

present in all habitats. Non-significant interactions were dropped. For desiccation

recovery, the final model included sex, habitat, exposure, and the interaction between

habitat and exposure. Plants were considered to be recovered from desiccation at the first

point that clearly did not differ (P > 0.1) from the average of subsequent times (i.e. the

time at which Fv/Fm ceases to change). The dependent variables were the Fv/Fm readings

at each time point. For recovery from wet heat shock, the final between-subject factor

model included only sex, habitat, and exposure. Effect of the heat stress was tested using

a contrast between the initial and after heat-shock time points. Changes through time were tracked by comparing each time with the average of subsequent times.

Shoots from growth-chamber-grown sexually immature clumps—I used a doubly- repeated multivariate ANOVA to test for differences among treatments and between sexes through time. Sex was the between-subject (i.e between-isolate) factor, and treatment and time were the within-subject (i.e. within-isolate repeated) factors. The dependent variables were the Fv/Fm readings at each time point. After conducting the overall analysis, significantly stressed treatments were separated from non-stressed treatments to seek sex differences. Stressed treatments were further separated if they differed from one another. Contrasts were conducted between the times before and after treatment to test for sex differences in the magnitude of stress experienced by the treatments. Sex differences in recovery rate were assessed by contrasts between the first time after treatment and the final measurement (96 h). Other contrasts were selected by visual inspection of graphs.

102

Shoot production 10 d after planting was analyzed using a repeated measures

ANOVA in PROC MIXED. Out of 152 replicates, only 5 died. These were marked as missing data points. Sex was the between subject factor, and treatment and the interaction of treatment and sex were the within subject factors. Treatments were compared using Tukey-Kramer corrected pairwise comparisons. Shoot production was square-root-transformed to improve normality.

Shoots from growth-chamber-grown sexually mature clumps—I tested for overall treatment, stage (sex-expressing vs. vegetative), and sex through time using a doubly multivariate repeated measures ANOVA with set as a blocking effect. Treatment and stage were within-subject effects, and sex and set were between-subject effects. The dependent variables were Fv/Fm readings through time. Because the initial repeated measures analysis indicated differences between stages (see Results), these were analyzed separately. As above, significantly stressed treatments were analyzed together to assess sex-specific stress responses. Stressed treatments that differed from one another were further analyzed separately. Contrasts were conducted as described above, and significant results are presented below.

Results

Field studies—In my field study, in most cases, the sex of collected plants was not known until several months after they were collected and cultured to sexual maturity, and only one clump was likely mixed sex, having senesced sporophytes. Of the 64 isolates only eight were males, thus few males were used in the recovery from desiccation and wet heat stress.

103

Recovery from desiccation—A univariate ANOVA at the initial Fv/Fm reading indicated

no difference in recovery between the sexes in the field-collected plants (F1, 58 = 0.69, P =

0.4108). The plants clearly showed a general increase in Fv/Fm through time, indicated by a significant time effect, but there were no significant interactions with time. The overall

Fv/Fm level was not affected by sex but tended to have a negative relationship with

exposure (slope = - 0.098, F1, 61 = 3.77, P = 0.058). Habitat tended to affect the

relationship with exposure differently, and only tree bases tended to have a negative

relationship between exposure and average Fv/Fm (slope = - 0.18, F1, 12 =3.76, P = 0.077).

For a summary of the overall repeated measures analysis, see Table 4.1. Although the overall time effect was significant, Fv/Fm did not begin to increase significantly until 12

h, indicating a lag in time before physiological repairs are manifested. The plants were

nearly fully recovered by 18 h with nearly 60% of the total average increase in Fv/Fm

(initial through 72 h) occurring between 6 and 18 h vs. approximately 29% between 18

and 72 h (Figure 4.1).

104

Table 4.1. Results from the repeated measures analysis for recovery of photosystem II using Fv/Fm from desiccation of field-collected females (n = 54) and males (n = 8) of Bryum argenteum.

Source Wilks' λ df F P A -- Within subject effects Time 0.58 11, 41 2.70 0.010 Time*sex 0.94 11, 41 0.25 0.99 Time*habitat 0.46 44, 158.81 0.82 0.77 Time*exposure 0.78 11, 41 1.07 0.41 Time*habitat*exposure 0.45 44, 158.81 0.83 0.76

B -- Between subject effects Sex -- 1, 51 2.18 0.15 Habitat -- 4, 51 1.38 0.25 Exposure -- 1, 51 3.77 0.058 Habitat*exposure -- 4, 51 2.20 0.082

0.8 0.7 0.6 * 0.5 m * *

/ * v

F 0.4 Female 0.3 Male 0.2 0.1 0 -10 0 10 20 30 40 50 60 70 80 Time after rehydration (h)

Figure 4.1. Photosystem II recovery using Fv/Fm after rehydration of desiccated, field- collected females (n = 54) and males (n = 8) of Bryum argenteum. Points are means ± SE of untransformed data. Asterisks indicate significant (P < 0.05) differences between a time point and the average of all subsequent times without regard to sex.

Recovery from wet heat stress—Before exposure to wet hot conditions, there were no

differences between the plants in Fv/Fm due to sex (F1, 56 = 1.13, P = 0.29), habitat (F4, 56

= 1.76, P = 0.15), or exposure (F1, 56 = 1.00, P = 0.32). The response varied through time,

105

but generally remained high, and the reduction (difference between before and after

treatment) in Fv/Fm due to the wet hot treatment of approximately 3% was not significant

(F1, 55 = 1.18, P = 0.28), indicating no signs of stress, but Fv/Fm decreased slightly

through time (Figure 2). The interactions of sex, habitat, and exposure with time were

not significant. The overall average response tended to be affected by habitat, but not by

sex or exposure (Table 4.2). Using pairwise t-tests on overall averages across all time points, tree bases (isolate n = 15, 0.738 ± 0.008) had significantly higher (P < 0.05 in

each case) Fv/Fm than vertical hard surfaces (isolate n = 14, 0.701 ± 0.013), horizontal

hard surfaces (isolate n = 13, 0.700 ± 0.016), and greenhouse roofs (isolate n = 15, 0.696

± 0.010). Plants from topsoil (isolate n = 5, 0.732 ± 0.017) did not differ from the other

four habitats, and greenhouse roofs, horizontal hard surfaces, and vertical hard surfaces

were not different from one another (P > 0.05 in each case).

Table 4.2. Results of repeated measures analysis for responses to wet hot conditions (40°C for 1 h) of field-collected females (isolate n = 54) and males (isolate n = 8) of Bryum argenteum.

Source Wilks' λ df F P A -- Within subject effects Time 0.57 11, 45 3.03 0.0042 Time*sex 0.78 11, 45 1.16 0.34 Time*habitat 0.45 44, 174.11 0.92 0.61 Time*exposure 0.80 11, 45 1.00 0.46 B -- Between subject effects Sex -- 1, 55 0.02 0.89 Habitat -- 4, 55 2.53 0.051 Exposure -- 1, 55 1.76 0.19

106

0.9 0.8 * * * * 0.7 0.6

0.5 Female

Fv/Fm 0.4 Male 0.3 0.2 0.1 0 -5 15 35 55 75 Time after wet heat shock (h)

Figure 4.2. Photosystem II responses (Fv/Fm) to wet hot conditions (40°C for 1 h) of field-collected females (n = 54) and males (n = 8) of Bryum argenteum. Points are means ± SE of untransformed data. Asterisks indicate significant (P < 0.05) differences between a time point and the average of all subsequent times without regard to sex. The dashed line represents the mean Fv/Fm across both sexes before treatment.

Growth-chamber-grown sexually immature clumps—Before treatment, there were no

significant effects of treatment groups (F3, 32 = 1.53, P = 0.22), sex (F1, 34 = 0.42, P =

0.52), or the interaction between treatment group and sex (F3, 32 = 0.61, P = 0.61). There

was an overall time effect, indicating at least one treatment resulted in changes in Fv/Fm

over time. Considering the average across all times, wet cool and dry cool did not differ

from one another and showed no signs of stress, dry hot was intermediately stressed, and

wet hot was the most stressed. Furthermore, considering before and after treatment, wet

cool and dry cool had non-significant reductions in Fv/Fm of 2.36% (F1, 33 = 1.64, P =

0.21) and 4.96% (F1, 32 = 0.86, P = 0.36), respectively. Dry hot and wet hot had

significant reductions of 34.41% (F1, 31 = 53.96, P < 0.0001) and 77.09% (F1, 35 = 198.06,

P < 0.0001), respectively. Sex had no effect on the average Fv/Fm over all time points

and had no significant interactions (Table 4.3; Figure 4.3).

107

Table 4.3. Repeated measures analysis of the responses to wet cool, dry cool, wet hot, and dry hot conditions of female (isolate n = 15) and male (isolate n = 14) shoots of Bryum argenteum from growth-chamber-grown sexually immature clumps.

Source Wilks' λ df F P A -- Within subject effects Treatment 0.09 3, 25 83.76 <0.0001 Treatment*sex 0.91 3, 25 0.80 0.51 Time 0.07 5, 23 66.08 <0.0001 Time*treatment 0.03 15, 13 27.28 <0.0001 Time*sex 0.95 5, 23 0.26 0.93 Time*treatment*sex 0.40 15, 13 1.32 0.31 B -- Between subject effects Sex -- 1, 27 0.60 0.44

108

A 0.9 Wet 0.8 a 0.7 0.6

m 0.5 /F v F 0.4 0.3 0.2 0.1 0 -24 0 24 48 72 96 B 0.9 Dry 0.8 a 0.7 0.6

m 0.5 /F v F 0.4 0.3 0.2 0.1 0 -24 0 24 48 72 96 0.9 Dry hot C 0.8 0.7 0.6

m 0.5 b /F v F 0.4 0.3 0.2 0.1 0 -24 0 24 48 72 96 0.9 Wet hot

D 0.8 0.7 Female 0.6 Male m 0.5 /F v F 0.4 * 0.3 0.2 c 0.1 0 -24 0 24 48 72 96 Time after treatment (h)

109

Figure 4.3. Photosystem II recovery using Fv/Fm of female and male shoots from growth-chamber-grown sexually immature clumps of Bryum argenteum that were hydrated or rapid dried (22°C and 53% RH) then exposed to A) wet cool (hydrated, 22°C; female n = 19, male n = 16), B) dry cool (desiccated, 22°C; female n = 16, male n = 18), C) dry hot (desiccated, 45°C; female n = 18, male n = 15), or D) wet hot (hydrated, 45°C; female n = 18, male n = 19) conditions for 1 h. Points are means ± SE of untransformed data. Lower case letters indicate treatment groupings by pairwise contrasts (P < 0.0001). An asterisk indicates a sex by time interaction (P < 0.05) when contrasting a particular time point with the average of all subsequent points. Dashed lines are means of Fv/Fm across both sexes before treatment.

Significant differences between treatments at 96 h show that stressed plants did not recover fully during the assessment period (Wilks’ λ = 0.45, F3, 32 = 12.86, P <

0.0001). Wet and dry cool still had the highest Fv/Fm and did not differ from one another

(F1, 34 = 0.19, P = 0.67). Dry hot was significantly lower than wet cool (F1, 34 = 7.35, P =

0.01) and tended to be lower than dry cool (F1, 34 = 3.71, P = 0.063). Wet hot was significantly lower than wet cool (F1, 34 = 28.55, P < 0.0001), dry cool (F1, 34 = 19.00, P =

0.0001), and dry hot (F1, 34 = 4.77, P = 0.036).

Sex-specific patterns—For wet cool alone, which was not stressed, there were no significant effects of sex (F1, 33 =1.02, P = 0.32), sex by time (Wilks’ λ = 0.88, F5, 29 =

0.81, P = 0.55), or sex by time contrasts between times (analyses not shown), suggesting any sex differences in other treatments were triggered by stress (Figure 4.3A).

Considering only the treatments causing significant stress (dry and wet hot), the results were similar and no overall sex effects or interactions with sex were significant (Table

4.4). Because the sexes appeared to recover differently in the two stressful treatments between 24 and 48 h, I conducted a contrast at this point. The contrast comparing the results at 24 h with the average of subsequent times indicated an interaction between time, treatment (dry hot vs. wet hot), and sex (F1, 30 = 4.45, P = 0.043), which should be

110

interpreted cautiously due to the overall non-significant interaction between sex,

treatment, and time. Dry hot alone showed no overall sex (F5, 27 = 1.92, P = 0.18) or time

by sex (Wilks’ λ = 0.85, F5, 27 = 0.99, P = 0.44) effects, and though the sexes appear to

differ in slope between after treatment and 24 h, the difference was not significant (F1, 31

= 1.39, P = 0.23; Figure 4.3C). In the wet hot treatment, though there were no overall

sex (F1, 35 = 0.00, P = 0.99) or time by sex effects (Wilks’ λ = 0.88, F5, 31 = 1.06, P =

0.40), females recover significantly faster leading to males having higher Fv/Fm before

and up to 24 h with females surpassing males after 48 h, though the differences are not

significant (analyses not shown) at each time point (Figure 4.3D). When considering

both stressed treatments across the last three times (48, 72, and 96 h), females tended to

have higher Fv/Fm than males (F1, 30 = 3.69, P = 0.064; Figure 4.3C and 4.3D).

Table 4.4. Repeated measures analysis of the responses to dry and wet hot alone of female (isolate n = 19) and male (isolate n = 17) shoots of Bryum argenteum from growth-chamber-grown sexually immature clumps.

Source Wilks' λ df F P A -- Within subject effects Treatment 0.53 1, 30 26.67 <0.0001 Treatment*sex 0.99 1, 30 0.27 0.60 Time 0.08 5, 26 59.00 <0.0001 Time*treatment 0.31 5, 26 11.40 <0.0001 Time*sex 0.84 5, 26 0.97 0.45 Time*treatment*sex 0.86 5, 26 0.88 0.51 B -- Between subject effects Sex -- 1, 30 1.69 0.20

Effects of stress on shoot production—Shoot production was significantly affected by

treatment (F3, 36 = 19.42, P < 0.0001), but not by sex (F1, 36 = 0.28, P = 0.60) or the

interaction (F3, 36 = 0.72, P = 0.55). Overall, shoot production reflected the stress level measured by overall Fv/Fm caused by each of the treatments. Wet cool and dry cool did

111

not differ from each other and had the highest shoot production. Shoot production was

intermediate for dry hot and lowest for wet hot (Figure 4.4).

25 a 20 a Female 15 Male

10 b

Number of shoots of Number 19 19 18 19 c 5 19 18 19 0 16 Wet cool Dry cool Dry hot Wet hot

Figure 4.4. Shoot production of sexually immature males and females of Bryum argenteum after initial hydration or rapid dry (22°C and 53% RH) and exposure to wet cool (hydrated, 22°C), dry cool (desiccated, 22°C), dry hot (desiccated, 45°C), or wet hot (hydrated, 45°C) conditions for 1 h. Bars are means ± SE of untransformed shoot production. Isolate sample sizes are indicated on their associated bars. Lower case letters indicate significant differences (P < 0.05) among treatments by pairwise comparisons using Tukey-Kramer adjustments.

Growth-chamber-grown sexually mature clumps—Before treatment, sex-expressing shoots tended to have higher Fv/Fm than vegetative shoots (0.715 ± 0.008 vs. 0.707 ±

0.009, respectively; Wilks’ λ = 0.90, F1, 29 = 3.34, P = 0.078), but there were no

significant effects of treatment groups (Wilks’ λ = 0.87, F3, 27 = 1.32, P = 0.29), sex (F1, 29

= 0.02, P = 0.88), set (F5, 29 = 2.03, P = 0.10), sex by treatment group (Wilks’ λ = 0.97,

F3, 27 = 0.27, P = 0.84), set by treatment group (Wilks’ λ = 0.55, F15, 74.94 = 1.20, P =

0.29), stage by treatment group (Wilks’ λ = 0.91, F3, 27 = 0.89, P = 0.46), stage by

treatment group by sex (Wilks’ λ = 0.97, F3, 27 = 0.26, P = 0.86), or stage by treatment

group by set (Wilks’ λ = 0.70, F15, 74.94 = 0.70, P = 0.78).

Considering all treatments across time and both stages, there was an overall time

effect, indicating that in at least one treatment Fv/Fm changed through time. Overall

112

Fv/Fm significantly differed between treatments and was highest for wet cool, intermediate for dry cool, and lowest for dry hot and wet hot, which did not differ from each other (Figures 4.5 and 4.6), unlike in the immediately previous experiment.

Furthermore, considering time points before and after treatment, wet cool had a non- significant increase in Fv/Fm of 2.34% (F1, 29 = 0.90, P = 0.35), while Fv/Fm was reduced

48.09% for dry cool (F1, 27 = 173.53, P = < 0.0001), 67.28% for dry hot (F1, 29 = 591.55, P

< 0.0001), and 55.47% for wet hot (F1, 29 = 311.82, P < 0.0001). The reduction was significantly greater for dry hot compared to wet hot (F1, 27 = 7.14, P = 0.013). Estimates of percent reductions are means of isolate mean reductions (both sex-expressing and vegetative shoots). Overall, sex-expressing shoots had higher Fv/Fm than vegetative shoots (0.546 ± 0.006 vs. 0.528 ± 0.006, respectively), and the responses of sex- expressing and vegetative shoots differed through time, but there was no overall effect of sex or its interactions. The overall multivariate repeated measures analysis for both stages and all treatments is summarized in Table 4.5.

113

0.8 Wet cool A a 0.7 0.6 0.5 m /F

v 0.4 F 0.3 0.2 0.1 0 -24 0 24 48 72 96 120 144 168 192 216 240 0.8 Dry cool B 0.7 0.6 0.5 m /F

v 0.4 F b 0.3 0.2 0.1 0 -24 0 24 48 72 96 120 144 168 192 216 240 0.8 Dry hot, wet hot combined C * 0.7 0.6 0.5

m *

/F 0.4 v Female F 0.3 c Male 0.2 0.1 0 -24 0 24 48 72 96 120 144 168 192 216 240 Time after treatment (h)

Figure 4.5. Photosystem II recovery using Fv/Fm of sex-expressing shoots from growth- chamber-grown females (n = 19) and males (n = 17) of Bryum argenteum that were hydrated or rapid dried (22°C and 53% RH) then exposed to A) wet cool (hydrated, 22°C), B) dry cool (desiccated, 22°C), C) dry hot (desiccated, 45°C) or wet hot (hydrated, 45°C) conditions for 1 h. A and B show means ± SE and C shows least square means ± SE (dry and wet hot together) of untransformed data. Lower case letters indicate treatment groupings by pairwise contrasts (P < 0.0001). An asterisk indicates a sex by time (P < 0.05) interaction when contrasting a time point with the average of all subsequent points. An asterisk over a bracket indicates a sex by time interaction (P < 0.05) when comparing the times at the bracket ends. Dashed lines are means of Fv/Fm across both sexes before treatment.

114

0.8 Wet cool A a * 0.7 0.6 0.5 m /F

v 0.4 F 0.3 0.2 0.1 0 -24 0 24 48 72 96 120 144 168 192 216 240 B 0.8 Dry cool 0.7 0.6 0.5 m F

/ 0.4 v b

F 0.3 0.2 0.1 0 -24 0 24 48 72 96 120 144 168 192 216 240 C 0.8 Dry hot, wet hot combined 0.7 0.6 * 0.5 m

/F 0.4 v Female F 0.3 c Male 0.2 0.1 0 -24 0 24 48 72 96 120 144 168 192 216 240 Time after treatment (h)

Figure 4.6. Photosystem II recovery using Fv/Fm of mature vegetative shoots from growth-chamber-grown females and males of Bryum argenteum that were hydrated or rapid dried (22°C and 53% RH) then exposed to A) wet cool (hydrated, 22°C; female n = 19, male n = 17), B) dry cool (desiccated, 22°C; female n = 18, male n = 16), C) dry hot (desiccated, 45°C; female n = 19, male n = 17), or wet hot (hydrated, 45°C; female n = 19, male n = 17) conditions for 1 h. A and B show means ± SE and C shows least square means ± SE (dry and wet hot together) of untransformed data. Lower case letters indicate treatment groupings among treatments by pairwise contrasts (P < 0.05). An asterisk indicates a sex by time interaction (P < 0.05) when contrasting a particular time point with the average of all subsequent points. An asterisk over a bracket indicates a sex by time interaction (P < 0.05) when comparing the times at the bracket ends. Dashed lines are means of Fv/Fm across both sexes before treatment.

115

Table 4.5. Repeated measures analysis of responses to wet cool, dry cool, dry hot, and wet hot treatments of female (isolate n = 18) and male (isolate n = 16) vegetative and sex- expressing shoots of Bryum argenteum from growth-chamber-grown sexually mature clumps.

Source Wilks' λ df F P A -- Within subject effects Treatment 0.044 3, 25 179.05 < 0.0001 Treatment*stage 0.94 3, 25 0.57 0.64 Treatment*sex 0.85 3, 25 1.44 0.26 Treatment*set 0.12 15, 69.42 5.27 < 0.0001 Treatment*stage*sex 0.93 3, 25 0.59 0.63 Treatment*stage*set 0.58 15, 69.42 1.00 0.48 Time 0.016 6, 22 218.61 < 0.0001 Time*treatment 0.0075 18, 10 73.84 < 0.0001 Time*stage 0.57 6, 22 2.78 0.036 Time*sex 0.80 6, 22 0.93 0.49 Time*set 0.13 30, 90 2.02 0.0060 Time*treatment*stage 0.15 18, 10 3.08 0.037 Time*treatment*sex 0.52 18, 10 0.51 0.90 Time*treatment*set 0.001 90, 53.03 1.85 0.0082 Time*treatment*stage*sex 0.49 18, 10 0.57 0.86 Time*treatment*stage*set 0.0071 90, 53.03 1.05 0.44 Time*stage*sex 0.87 6, 22 0.54 0.78 Time*stage*set 0.26 30, 90 1.20 0.25 Stage 0.78 1, 27 7.71 0.0098 Stage*sex 1.00 1, 27 0.06 0.80 Stage*set 0.65 5, 27 2.94 0.030 B -- Between subject effects Sex -- 1, 27 0.02 0.90 Set -- 5, 27 6.05 0.0007

Treatments effects remained evident at 10 days (F3, 27 = 5.44, P = 0.0046), and

pairwise contrasts showed that wet cool tended to have higher Fv/Fm than dry cool (F1, 29

= 3.46, P = 0.073), and had significantly higher Fv/Fm than dry hot (F1, 29 = 7.68, P =

0.0096) and wet hot (F1, 29 = 0.55, P = 0.46). Dry cool did not significantly differ from

dry hot (F1, 29 = 2.00, P = 0.17) or wet hot (F1, 29 = 0.55, P = 0.46), and neither did dry hot

and wet hot (F1, 29 = 0.87, P = 0.36).

116

Differences between sex-expressing and vegetative shoots—Wet cool alone had no

significant effects of sex (F1, 29 = 0.14, P = 0.71), sex by stage (Wilks λ = 1.00, F1, 29 =

0.02, P = 0.90), sex by time (Wilks λ = 0.85, F6, 24 = 0.72, P = 0.64), stage by time (Wilks

λ = 0.70, F6, 24 = 1.74, P = 0.15), or sex by stage by time (Wilks λ = 0.80, F6, 24 = 0.98, P

= 0.46), but sex-expressing shoots had significantly higher Fv/Fm than vegetative shoots

(Wilks λ = 0.77, F1, 29 = 8.89, P = 0.0058; Figure 4.7A). A comparison of the time after

treatment with the average of subsequent times indicated that vegetative shoots drop

more than sex-expressing shoots at later times (Figure 4.7A). A contrast between the time directly after treatment and 10 days showed that sex-expressing shoots differed from vegetative shoots in their recovery after stress (F1, 27 = 8.42, P = 0.0073). Considering

only stressed treatments (dry cool, dry hot, and wet hot), sex-expressing shoots tended to have higher overall Fv/Fm, but did not differ from vegetative shoots in recovery overall

through time (Table 4.6). However, sex-expressing shoots recovered significantly faster than vegetative shoots overall and the means of sex-expressing shoots and vegetative shoots switch positions after 24 h (Figure 4.7B). This difference in recovery was manifested as higher Fv/Fm among sex-expressing shoots relative to vegetative shoots 10 days after treatment (F1, 29 = 9.52, P = 0.044), which tended to persist after accounting for

initial differences between stages (Wilks λ = 0.89, F1, 29 = 3.68, P = 0.065).

117

A 0.8 Wet cool * 0.7 0.6 0.5 m /F v 0.4 F 0.3 0.2 0.1 0 -24 0 24 48 72 96 120 144 168 192 216 240 B 0.8 Dry cool, dry hot, wet hot combined 0.7 * 0.6 0.5 m /F v 0.4 * F Sex-expressing 0.3 Vegetative 0.2 0.1 0 -24 0 24 48 72 96 120 144 168 192 216 240 Time after treatment (h)

Figure 4.7. Photosystem II recovery using Fv/Fm of sex-expressing (n = 34) and vegetative shoots (n =34) of growth-chamber-grown Bryum argenteum clumps without respect to sex that were hydrated or rapid dried (22°C and 53% RH) then exposed to A) wet cool alone (hydrated, 22°C) and B) pooled dry cool (desiccated, 22°C), dry hot (desiccated, 45°C), and wet hot (hydrated, 45°C) treatments for 1 h. Points are least square means ± SE. An asterisk indicates a stage by time interaction (P < 0.05) when contrasting a particular time point with the average of all subsequent times. An asterisk over a bracket indicates a stage by time interaction (P < 0.05) when comparing the times at the bracket ends. Least square means before treatment of sex-expressing and vegetative shoots are represented by light gray or dark gray dashed lines, respectively.

118

Table 4.6. Repeated measures analysis of responses to dry cool, dry hot, and wet hot treatments of female (isolate n = 18) and male (isolate n = 16) vegetative and sex- expressing shoots of Bryum argenteum from growth-chamber-grown sexually mature clumps.

Source Wilks' λ df F P A -- Within subject effects Treatment 0.36 2, 26 22.68 < 0.0001 Treatment*stage 0.94 2, 26 0.83 0.45 Treatment*sex 0.88 2, 26 1.75 0.19 Treatment*set 0.31 10, 52 4.08 0.0004 Treatment*stage*sex 0.94 2, 26 0.86 0.43 Treatment*stage*set 0.63 10, 52 1.36 0.23 Time 0.020 6, 22 182.85 < 0.0001 Time*treatment 0.21 12, 16 4.90 0.0020 Time*stage 0.65 6, 22 1.94 0.12 Time*sex 0.79 6, 22 0.95 0.48 Time*set 0.17 30, 90 1.71 0.028 Time*treatment*stage 0.29 12, 16 3.32 0.014 Time*treatment*sex 0.67 12, 16 0.67 0.76 Time*treatment*set 0.035 60, 78.7 1.37 0.096 Time*treatment*stage*sex 0.73 12, 16 0.49 0.89 Time*treatment*stage*set 0.039 60, 78.7 1.32 0.12 Time*stage*sex 0.88 6, 22 0.52 0.78 Time*stage*set 0.32 30, 90 0.99 0.49 Stage 0.89 1, 27 3.30 0.081 Stage*sex 1.00 1, 27 0.06 0.81 Stage*set 0.76 5, 27 1.67 0.18 B -- Between subject effects Sex -- 1, 27 0.00 0.98 Set -- 5, 27 6.97 0.0003

Sex-specific patterns among sex-expressing shoots—When considering sex-expressing

shoots alone, there were no significant effects of sex (F1, 29 = 0.01, P = 0.92), sex by time

(Wilks’ λ = 0.69, F6, 24 = 1.77, P = 0.15), or sex by treatment (Wilks’ λ = 0.84, F3, 27 =

1.71, P = 0.19), but there was a tendency for the treatment by sex by time interaction

(Wilks’ λ = 0.21, F18, 12 = 2.45, P = 0.059). For wet cool alone, there were no significant effects of sex (F1, 29 = 0.10, P = 0.75) or sex by time (Wilks’ λ = 0.95, F6, 24 = 0.23, P =

119

0.96; Figure 4.5A). Considering all significantly stressed treatments together (dry cool,

dry hot, and wet hot) for sex-expressing shoots, there were slight tendencies for treatment

by sex and sex by time (Table 4.7). There were signficant interactions of sex with time

between 48 and 72 h (F1, 29 = 7.52, P = 0.010) and when comparing 48 h to the average of

subsequent times (F1, 29 = 8.72, P = 0.0062), suggesting that females recovery more

quickly after stress and surpass males starting at 48 h, but these interactions should be

interpreted cautiously considering the overall non-significant time by sex interaction. For dry cool alone, there were no significant effects of sex (F1, 29 = 2.17, P = 0.15) or sex by

time (Wilks’ λ = 0.89, F6, 24 = 0.48, P = 0.82; Figure 4.5B). Because dry and wet hot did

not differ overall in their effects on Fv/Fm, although initially dry hot caused a greater

decrease, and showed visually similar patterns in recovery, I considered these treatments

together. For dry and wet hot together, there were no overall significant effects of sex

(F1, 29 = 0.89, P = 0.34), sex by time (Wilks’ λ = 0.74, F6, 24 = 1.40, P = 0.26), sex by

treatment (Wilks’ λ = 0.95, F1, 29 = 1.50, P = 0.23), or sex by time by treatment (Wilks’ λ

= 0.92, F6, 24 = 0.33, P = 0.92), but females recovered faster from stress when comparing

the time after treatment with 10 d with females surpassing males after 48 h (Figure 4.5C).

When considering the last three times (72, 96, 240 h) and both dry and wet hot, females had significantly higher Fv/Fm than males (F1, 29 = 4.62, P = 0.040).

120

Table 4.7. Repeated measures analysis of responses to dry cool, dry hot, and wet hot treatments of female (isolate n = 19) and male (isolate n = 17) sex-expressing shoots alone of Bryum argenteum from growth-chamber-grown sexually mature clumps.

Source Wilks' λ df F P A -- Within subject effects Treatment 0.38 2, 28 22.57 < 0.0001 Treatment*sex 0.84 2, 28 2.57 0.094 Treatment*set 0.59 10, 56 1.68 0.11 Time 0.03 6, 24 144.17 < 0.0001 Time*treatment 0.20 12, 18 5.83 0.0005 Time*sex 0.66 6, 24 2.09 0.092 Time*set 0.20 30, 98 1.61 0.043 Time*treatment*sex 0.51 12, 18 1.43 0.24 Time*treatment*set 0.07 60, 88.07 1.15 0.27 B -- Between subject effects Sex -- 1, 29 0.00 0.98 Set -- 5, 29 8.00 < 0.0001

Sex-specific patterns among vegetative shoots—Among vegetative shoots alone, there was no overall significant effects of sex (F1, 27 = 0.00, P = 0.99), sex by time (Wilks’ λ =

0.79, F6, 22 = 0.95, P = 0.48), sex by treatment (Wilks’ λ = 0.99, F3, 25 = 0.07, P = 0.98), or sex by treatment by time (Wilks’ λ = 0.60, F18, 10 = 0.37, P = 0.97). For wet cool alone, while there were no overall sex (F1, 29 = 0.10, P = 0.75) or sex by time effects (Wilks’ λ =

0.76, F6, 24 = 1.26, P = 0.31), but males drop slightly more in Fv/Fm than females at the first time after treatment (Figure 4.6A). Similarly, when considering stressed (dry cool, dry hot, and wet hot) vegetative shoots, there were no overall effects of sex, sex by time, sex by treatment, or sex by treatment by time (Table 4.8). For dry cool alone, there were no significant effects of sex (F1, 27 = 0.03, P = 0.86) or sex by time (Wilks’ λ = 0.95, F6, 22

= 0.20, P = 0.97; Figure 4.6B). For dry and wet hot together (analyzed as above), there were no overall significant effects of sex (F1, 29 = 0.01, P = 0.94), sex by time (Wilks’ λ =

0.76, F6, 24 = 1.30, P = 0.30), or sex by time by treatment (Wilks’ λ = 0.93, F6, 24 = 0.30, P

121

= 0.93), but the slope of females was greater than males from 72-96 h (F1, 29 = 4.63, P =

0.040; Figure 4.6C). When considering the last two times (96 and 240 h) of both dry and wet hot together, females and males did not differ in Fv/Fm (F1, 29 = 1.39, P = 0.25).

Table 4.8. Repeated measures analysis of responses to dry cool, dry hot, and wet hot treatments of female (isolate n = 18) and male (isolate n = 16) vegetative shoots alone of Bryum argenteum from growth-chamber-grown sexually mature clumps.

Source Wilks' λ df F P A -- Within subject effects Treatment 0.71 2, 26 5.39 0.011 Treatment*sex 1.00 2, 26 0.04 0.96 Treatment*set 0.39 10, 52 3.08 0.0038 Time 0.032 6, 22 110.01 < 0.0001 Time*treatment 0.37 12, 16 2.27 0.064 Time*sex 0.79 6, 22 0.97 0.47 Time*set 0.24 30, 90 1.30 0.17 Time*treatment*sex 0.89 12, 16 0.15 0.99 Time*treatment*set 0.032 60, 78.7 1.42 0.072 B -- Between subject effects Sex -- 1, 27 0.02 0.88 Set -- 5, 27 3.17 0.022

Discussion

In dioecious plants, sex differences in stress tolerance have often been posited as explanations for differential mortality and biased sex ratios in certain habitats

(Bierzychudek and Eckhart 1988; Delph 1999; Barrett and Hough 2013). However, the physiological causes of the apparent sex differences in stress tolerance (i.e. sex ratio bias in stressful habitats) are not well investigated (Juvany and Munne-Bosche 2015). Though it is generally thought that sexually mature plants due to reproductive tradeoffs will be more sexually dimorphic, including in traits related to stress responses (Delph 1999;

Obeso 2002; Case and Ashman 2005), the potential for such traits to arise before sexual

122

maturity has not been adequately tested. The scarcity of studies is likely because

measuring and tracking sexually immature individuals until sex-expression in long-lived species is logistically difficult. My study, using chlorophyll fluorescence, demonstrated that the sexes of Bryum argenteum are subtly dimorphic in their responses to heat stress, both wet and dry, and the strength of these responses depends on the development stage.

Furthermore, the stress response pattern of females resembles that of sex-expressing

shoots (combined across both sexes), suggesting females deal with stress in a similar way

to sex-expressing shoots. Curiously, young sexually immature and sexually mature

shoots exhibit similar patterns of recovery and are more dimorphic in their response to

stress than vegetative shoots from sexually mature clumps. Responses to stress also

varied among experiments, which suggests plastic changes in hardening.

Sex-specific responses to heat stress and desiccation—After stress, the responses of

cultured sexually immature and sex-expressing shoots were sexually dimorphic, though

weaker than expected, but vegetative shoots from sex-expressing clumps were much less

dimorphic. As expected because of concurrent investment in sexual reproduction, the

dimorphic pattern was strongest among sex-expressing shoots. Field-collected plants,

which were not currently expressing sex, were evidently hardened to wet heat stress and

showed no sex differences in recovery from desiccation or in response to wet heat. In the

most stressful treatments (wet heat for sexually immature and wet and dry heat for sex-

expressing shoots), the patterns suggest not an overall maintenance of higher potential

quantum efficiency (Fv/Fm) of one of the sexes, but rather that female rate of recovery

surpasses that of males a few days after high stress events, leading to higher Fv/Fm for

123

females compared to males among sexually immature (tendency) and sex-expressing

shoots after 48-72h of recovery. A higher Fv/Fm for females compared to males under

water stress conditions was also found for the dioecious angiosperm shrub, Corema

album (Álvarez-Cansino et al. 2012), although the initial response to stress and the speed

of recovery were not tested. Among sexually immature shoots, females surpass males

earlier (24 h) than among sex-expressing shoots (48 h), which may be attributable to

overall greater dehardening in the second growth chamber experiment. These patterns

suggest that females repair damage to photosystem II more quickly after stress than

males. In sex-expressing shoots, though not sexually immature shoots, faster female

repair is consistent with the idea that males invest more nitrogen in reproduction than

females (numerous sperm vs. few eggs), which has also been proposed for seed plants

(Case and Ashman 2005), limiting nitrogen availability for photosynthetic machinery

production/repair. Size may also play a role because female shoots are typically bigger

than male shoots especially when sex-expressing (Horsley et al. 2011; Chapter Five),

though I did not measure size here.

The difference in degree of dimorphism between sexually immature, vegetative,

and sex-expressing shoots is puzzling. The prevailing paradigm in dioecious plants is that

sexual dimorphism will arise after sexual maturity (Delph 1999; Obeso 2002; Case and

Ashman 2005). My study connects with literature in both seed plants (Quinn and

Meiners 2004; Vega-Frutis and Guevara 2009) and bryophytes (Newton 1972; Shaw and

Gaughan 1993; Stieha et al. 2014; Marks et al. 2016; Chapter Five) that shows the

occurrence of sex differences at immature stages. Nevertheless, the question remains:

why are sexually immature and sex-expressing shoots more dimorphic than vegetative

124

shoots from mature clumps? I propose two possible explanations: (1) sexual dimorphism

manifests at times of high resource expenditure, (2) all shoots begin sexually dimorphic

and move on in development either to express sex or become vegetative shoots (i.e.

sexually dimorphic physiology shuts off). The first explanation is connected to high

growth rates among sexually immature shoots and high reproductive investment of sex-

expressing shoots. Perhaps physiological sex differences that affect stress responses,

though always present, are only obvious when plants are allocating relatively high

amounts of resources to other traits (i.e. growth or reproduction). Though sex differences

were not tested, Stark et al. (2016) demonstrated that stress tolerance strategy shifts

between developmental stages in the moss Syntrichia pagorum (inducible vs. constitutive desiccation tolerance in young immature and mature vegetative shoots, respectively), which they hypothetically linked to the high growth rates of immature shoots. The second explanation proposes that sexually dimorphic physiology is present in sexually immature shoots because it is preparatory for sex-expression or is genetically correlated with the physiology of sexually mature individuals, but that in vegetative shoots their developmental trajectory has silenced (though perhaps not irreversibly) this sex difference because they will not express sex.

Because the sex differences in stress recovery among sexually immature shoots did not correspond to differences in shoot production rates, which generally reflected the relative severity of the treatments, a statement of which sex is more stress tolerant is not possible from the current study. If females tolerate stress better, the difference in shoot production rates without stress (females non-significantly higher than males) should be exacerbated by stress, but this was not the case. In the moss Syntrichia caninervis, though

125

females have higher shoot production from regenerating leaves than males under mesic

conditions, this difference is diminished by wet heat stress (Stark and McLetchie 2006), and males gain biomass more quickly after desiccation and heat to 80-120°C (Stark et al.

2009), suggesting males recover better after heat stress than females. In the present study, the lack of control for initial shoot size may have limited detection of small sex

differences in shoot production.

Comparing stage-specific and sex-specific responses—When comparing recovery

patterns of sex-expressing and vegetative shoots, there is a striking similarity to the

comparison between males and females, suggesting that females and sex-expressing

shoots have similar physiological strategies for dealing with stress. Sex-expressing

shoots recover faster after stress (considering dry cool, dry hot, and wet hot together) and

have higher Fv/Fm at 10 d than vegetative shoots. Unlike females relative to males, sex- expressing shoots have higher Fv/Fm than vegetative shoots even in the absence of stress,

which may be related to the photosynthetic requirements of their reproductive

investment. In the dioecious angiosperm tree Ilex aquifolium, higher Fv/Fm of non-

fruiting vs. fruiting branches among females was interpreted as a photosynthetic source-

sink relationship (Obeso et al. 1998). A chronically low Fv/Fm (i.e. as in vegetative

shoots) has been considered a consequence of higher photoprotection and potentially

higher stress tolerance (Gilmore and Ball 2000; Hamerlynck et al. 2002; Reynold and

McLetchie 2011). My field data lends some credence to the interpretation of higher

photoprotection by showing that higher exposure reduces Fv/Fm chronically. However,

higher recovery among sex-expressing shoots does not support higher stress tolerance for

126

vegetative shoots. As with females and males, size may play a role in recovery because

sex-expressing shoots are often larger than vegetative shoots, though I did not control for it. Furthermore, because vegetative shoots allocate no resources to sexual reproduction, a tradeoff between reproductive investment and photosynthetic repair as proposed above to explain slower male vs. female recovery will not hold for vegetative shoots. Another explanation could be that vegetative shoots have lower resource pools than sex- expressing shoots (i.e. the reason vegetative shoots do not express sex), allowing sex- expressing shoots to invest more in repair despite reproductive investment.

Responses of Bryum argenteum to desiccation and heat—When considering responses to and recovery from desiccation across my experiments, B. argenteum shows evidence of dehardening to desiccation stress when grown in mesic conditions. Among field-grown plants, the majority of desiccation recovery occurred within 18 hours after hydration which is consistent with the field desiccation recovery time (24h) found for S. caninervis

(Reynold and Mcletchie 2011). In the first growth chamber experiment using sexually immature shoots, a rapid dry to approximately 53% relative humidity, which can be very stressful to some desiccation tolerant bryophytes (Pressel and Duckett 2010; Stark et al.

2013), did not measurably stress the plants. On the other hand, in the second growth chamber experiment, both vegetative and sex-expressing shoots were stressed by rapid desiccation to approximately 59% relative humidity and Fv/Fm did not plateau until 96h,

never fully recovering. The difference between experiments may be due to allowing

plants to dry before being watered in the first but not in the second (for rapid sexual

127

maturation). However, some effect of sex-expression cannot be ruled out, although

vegetative shoots (lacking reproductive organs) showed similar stress patterns.

Responses to heat depended on whether plants were field or growth chamber

grown and hydration status. For field-grown B. argenteum collected during early

summer, 40°C for one hour caused no detectable stress. Similar temperatures up to 41°C

on summer days have been observed for partially hydrated plants in full sun and are

likely to be common (J. Moore unpublished data), which may explain the apparent wet

heat tolerance of field-grown plants. In the first growth chamber experiment, heat during

both desiccation and hydration reduced Fv/Fm with wet heat causing worse stress, which

is consistent with the idea that the desiccated dormant state provides protection from heat

damage (Glime 2007b; Stark et al. 2009). Contrastingly, in the second growth chamber experiment, dry heat was just as stressful as wet heat and initially reduced Fv/Fm

approximately 11% more, further indicating that plants were dehardened to desiccation

and thus lost its heat damage protection. The greater stress due to wet heat among

growth chamber plants compared to field-collected plants may be attributable to

dehardening, but the 5°C increase in temperature for growth chamber plants was

potentially near the upper limit of their tolerance as temperatures between 40-45°C can

be lethal for several species of mosses (Glime 2007b).

Conclusions

Males and females of B. argenteum are subtly sexually dimorphic in their response to stress caused by heat and desiccation. However, whether one sex is more stress tolerant than the other is not clear from the current study. Further investigation using more refined measures of growth after stress would be required to interpret the

128 patterns of change in potential quantum efficiency. Stress responses are dimorphic in sexually immature shoots and more strongly in sexually mature shoots. The latter supports the hypothesis that reproductive investment manifests sex differences in dioecious species. However, the former suggests some correlation in sexually immature shoots with later mature physiological characteristics, which may be present in sexually immature shoots due to high growth rate. Females responses to severe stress follow a similar pattern to sex-expressing shoots (across all stress treatments and both sexes) while vegetative shoots resemble males. The reason for this resemblance is currently uncertain, but may relate to nutrient acquisition or expenditure. Additionally, while I expected shoots bearing sex structures to be more vulnerable to stress due to tradeoffs with reproduction, I found no evidence to support this idea as sex-expressing shoots appear to recover faster from stress than vegetative shoots. My studies also suggest that B. argenteum can tolerate quite severe desiccation and heat stress, but that the tolerance to these stresses can be lost due to dehardening.

129

CHAPTER FIVE: DOES SELECTION FOR GAMETE DISPERSAL AND CAPTURE LEAD TO A SEX DIFFERENCE IN CLUMP WATER-HOLDING CAPACITY?

Abstract

Differences in male and female reproductive function can lead to selection for sex-

specific gamete dispersal and capture traits. These traits have been explored from shoot to

whole plant levels in wind-pollinated species. While shoot traits have been explored in

water-fertilized species, little is known about how whole plant morphology affects

gamete dispersal and capture. I used the dioecious, water-fertilized plant Bryum argenteum to test for differences in clump morphology and water-holding characteristics consistent with divergent selection. I hypothesized that sex-specific clump morphology, arising at maturity, produces relatively low male water-holding capacity for gamete dispersal and high female capacity for gamete capture. I measured isolated young shoot and clump water-holding capacity and clump morphological characteristics on greenhouse-grown plants. Young shoot capacity was used to predict clump capacity, which was compared with actual clump capacity. Young male shoots held more water per unit length, and male clumps had higher shoot density, which extrapolated to higher clump water-holding capacity. However, female clumps held more water and were taller with more robust shoots. Actual clump capacity correlated positively with clump height and shoot cross-sectional area. The sex difference in actual clump capacity and its unpredictability from younger shoots are consistent with my hypothesis that males should hold less water than females to facilitate sexual reproduction. These results provide conceptual connections to other plant groups and implications for connecting divergent selection to female-biased sex ratios in B. argenteum and other bryophytes.

130

KEY WORDS: biased population sex ratio; Bryaceae; bryophyte; Bryum argenteum; dioecious; divergent selection; gamete dispersal; sexual dimorphism

Introduction

The inherent differences in male and female reproductive function can lead to divergent selection pressure to optimize traits for dispersal of gametes in males and capture of male gametes as well as offspring maturation in females (Lloyd and Webb

1977; Friedman and Barrett 2009). Using divergent selection to explain morphological differences in dioecious plants, particularly flowering plants, is fairly common

(Vasiliauskas and Aarssen 1992; Eckhart 1999; Li et al. 2007; Harris and Pannell 2010).

Specifically, selection for gamete dispersal and capture manifests as sexually dimorphic flowers or analogous sex structures, inflorescences or cones, and plant architecture.

However, how divergent selection has acted upon whole plant architecture in dioecious species with water-dispersed gametes is particularly underexplored.

In insect-pollinated flowering plants, selection related to gamete dispersal often drives males to produce larger, showier flowers or larger floral displays than females to increase pollinator visitation (Andersson and Iwasa 1996; Delph 1999; Eckhart 1999). In wind-pollinated plants, divergent selection has led to exserted stamens for pollen dispersal and feathery stigmas for pollen capture (Niklas 1985a). Inflorescences of wind- pollinated flowering plants are also sometimes markedly dimorphic (Friedman and

Barrett 2009) as seen in the hop plant, Humulus lupulus (Shephard et al. 2000). In conifers, morphological differences have been linked to promotion of pollen shedding in male cones and direction of pollen-containing air currents to ovules in female cones

(Niklas 1985a, b). Similar morphological specialization occurs in inflorescences of water-

131

pollinated sea grasses (Ackerman 1997). Selection for gamete dispersal has also been

postulated to affect plant architecture (Vasiliauskas and Aarssen 1992; Bond and Maze

1999; Dawson and Geber 1999). Selection for larger floral displays in insect-pollinated plants can lead to architectural differences by increasing ramification in males (Bond and

Maze 1999). In wind-pollinated plants, branch or whole plant architecture may affect pollen dispersal and capture (Niklas 1985a, b). Importantly, canopy architecture can cause intra- and interspecific differences in water relations (Tausend et al. 2000), light interception (Valladares et al. 2002; Pearcy et al. 2005), and susceptibility to wind

(Sellier and Fourcaud 2009) or lightning damage (Yanoviak et al. 2015). Therefore,

sexually dimorphic canopy architecture is potentially linkable to sex differences in other

traits that could cause sex-specific demography, but this link is understudied.

Dioecious plants with water-facilitated gamete dispersal, such as bryophytes, must solve very similar problems as wind-pollinated seed plants. Their dependence on the aquatic medium is evidenced by free-water-adapted sperm dehiscence and dispersal mechanisms (Muggoch and Walton 1942; Paolillo 1977; Cronberg et al. 2008). However, in two moss species, the presence of both free water and springtails (microarthropods) increased fertilization over free water alone (Cronberg et al. 2006; Rosenstiel et al. 2012).

Therefore, bryophyte gamete dispersal and capture strategies are, in general, analogous to those in wind-pollinated seed plants. Just as air currents carry pollen, water currents carry bryophyte sperm to be captured by females. In bryophytes, most work has focused on sperm dispersal rather than capture mechanisms. Adaptations produced by divergent selection at the single shoot level include the splash cups in males of some mosses, which help sperm disperse through splashing of rain drops (Reynolds 1980; Andersson 2002).

132

Presumably, the strikingly dimorphic sex structures of the liverwort genus Marchantia

have been selected to assist sperm dispersal through splashing and sperm capture by

directing water to female sex organs (Glime 2013). However, little is known about how

whole plant architecture is adapted to facilitate gamete dispersal or capture and whether

these are linkable to any other sex-specific trait.

In this study, I focus on clump (an aggregation of shoots) stage adaptations for sperm dispersal and capture. Clump formation is common in bryophytes and potentially comprises a single genotype unit whose characteristics are relevant for sperm dispersal and capture. Clump characteristics, including water-holding capacity in capillary spaces, are well known to affect water relations in bryophytes (Elumeeva et al. 2011). As in seed plants, bryophyte shoots lacking sex organs influence the behavior of the gamete dispersal medium, making a bryophyte clump analogous to a seed plant inflorescence or cone, branch, or even a whole canopy. Therefore, clump morphology could be sex- specific due to different roles of the sexes in sexual reproduction. A male should allow water to flow away from its clump to assist in sperm dispersal, which could result in short, low-density clumps with small shoots for males. On the other hand, a mature female should allow sperm-laden water to infiltrate its clump and retain it, which facilitates sperm capture and fertilization. For females, greater water-holding capacity could be achieved with a tall, high-density clump with robust shoots. However, because of the intimate connection between water and life processes such as growth and survival in plants, this water-use pattern connected to sex function has potentially negative survival consequences for males and positive consequences for females. Higher water- holding capacity in females over males could lead to more favorable growing and

133 survival conditions which might explain the commonness of female-biased sex ratios among dioecious bryophytes (Bisang and Hedenäs 2005).

I used the moss Bryum argenteum to test for differences between male and female clumps in water-holding capacity, asking three questions: (1) Do females have higher clump water-holding capacity than males? (2) Do differences in water- holding capacity arise from water-holding differences present in young shoots or do differences emerge as clumps form? (3) What clump traits could contribute to an actual difference in clump water-holding capacity? I expected that females would have higher clump water-holding capacity than males. However, I did not have a prediction for differences in young shoot water-holding capacity as these were not sex expressing shoots. I did not expect to be able to predict clump sex differences by extrapolating from the isolated, young shoot stage (i.e. we expected clump water-holding capacity to emerge from clump development) because sexual reproduction occurs after clump formation. For clump morphology, we predicted female clumps to be denser, taller, and have more robust shoots than males. Because water-holding capacity is potentially connected to growth and survival, testing for this difference is the first step in connecting divergent selection on clump morphology to demography. Additionally, demonstrating that the sex difference arises after shoots have aggregated into clumps would lend further support for the hypothesis that selection has shaped that stage to facilitate sexual reproduction.

Methods

Study species—Bryum argenteum Hedw. (Bryaceae) is a cosmopolitan, dioecious, clump-forming moss with a wide distribution in natural and human-disturbed habitats, preferring exposed areas such as brick, sidewalks, and dry logs (Crum and Anderson

134

1981; Crum 1983; Pisa et al. 2014). Bryum argenteum has asexual propagules (bulbils) for local dispersal likely by water (Selkirk et al. 1998) and perhaps by human disturbance

(Burnell et al. 2004). Long-distance dispersal, potentially continental in range, is by sexually produced spores (Skotnicki et al. 1998). Sexual reproduction in the Kentucky population is uncommon (personal observation), and the sex ratio is highly female-biased

(unpublished data), which agrees with Stark et al. (2010). Animal assistance in sperm transfer has been recently demonstrated in B. argenteum (Cronberg et al. 2006;

Rosenstiel et al. 2012); however, animals are not required for successful fertilization

(Chopra and Bhatla 1981; Rosenstiel et al. 2012).

Field collections—Plants were collected from urban areas in Kentucky, United States

(Lexington: n = 19 males and 19 females, 298 m a.s.l., 38°02’04”N 84°30’27.4”W;

Bowling Green: n = 5 males and 5 females, 166.7 m a.s.l., 36°54’52.0”N 86°21’01.6”W; and Hazard: n = 1 male and 1 female, 283 m a.s.l., 37°15’07.1”N 83°11’42.7”W).

Generally clumps of plants with sporophytes (i.e. evidence of both sexes) were collected no less than 3 m apart, reducing the chance of collecting clones, if evidence of both sexes existed (i.e. sporophytes were present). Collecting plants by looking for sporophytes inherently biases our stock plants by increasing our chances of collecting males and females with compatible phenotypes for coexistence, which may not represent the complete phenotypic range in nature. However, six male and six female plants were collected from clumps without sporophytes (i.e. no evidence of sexual reproduction or coexistence of the sexes), and sex was determined by identification of sex structures under a dissecting microscope or culturing to sex expression. The vast majority of clumps

135

in nature appear to be single sex, and several hundred were examined before finding

sexes coexisting. After collection and identification of sex, a single sex expressing shoot

or shoot fragment was placed in a Petri dish on local, steam-sterilized topsoil and

cultured in a common environment (growth chamber with 14 h/10 h, 20°C/17°C

day/night with fluorescent lighting at about 50 µmol·m–2·s–1) to ensure the removal of

field habitat effects.

Experimental growing conditions—A total of 25 male and 25 female isolates with four

to six replicates each were grown in a greenhouse. Each replicate was grown from one

bulbil or similar-sized shoot fragment on local, steam-sterilized topsoil in an individual

pot (59 mL) fitted with a lid containing a neutral density filter (67.7% transmission, 0.15

neutral density; Lee Filters, Burbank, California, USA) to provide shade and prevent

contamination. The pots were randomized on a capillary watering mat over which was

placed a PVC frame and plastic drop cloth to provide additional shade and further prevent

contamination by propagule rain. Eleven replicates died, but no isolate was left with less

than three replicates alive.

Assessing clump characteristics and water-holding capacity—Plants were allowed to

grow in the greenhouse until clumps were fully formed (about nine months). To estimate

shoot density, shoot cross-sectional area, and proportion of shoot apices expressing sex I

used a digital camera (Coolpix 950, Nikon, Tokyo, Japan), a dissecting microscope, and a

PC computer with Image J (U. S. National Institutes of Health, Bethesda, Maryland,

USA; available at http://imagej.nih.gov/ij/). I measured average clump height by taking a

136

1 × 1 cm square out of each clump and measuring height at every 2 mm for a total of 5

measurements. To estimate water held within a clump but external to the shoots (clump

water-holding capacity), a second 1 × 1 cm square was excised from the center of each

clump. Shoots were left attached to soil to preserve clump structure. Through a pilot

experiment, blotting was found to be a flawed method likely due to interference of

attached soil and because evaporation from filter paper occurred before all external water

was removed from a clump sample. Instead I used centrifugation to quantify clump

water-holding capacity in a manner similar to Dilks and Proctor (1979) but with a few

modifications. The second excised section of clump was carefully handled to preserve

original shoot organization and capillary space. Samples were saturated with water,

enclosed in 12-well plates wrapped in parafilm, and frozen at –20°C for at least 12 h.

Freezing allowed plant material and especially water above soil level to be removed with

a razor blade and placed into a centrifuge tube (0.5 mL) in which a pinhole had been

made at the bottom. These tubes were then placed into larger centrifuge tubes (1.5 mL)

and centrifuged (Fisher Scientific 05-090-128; Hampton, New Hampshire, USA) for 5 min at 6000 rpm. With a radius of rotation of ≈3.6 cm, 6000 rpm produces ≈1450 × g.

Following the equations of Edmunds and Bath (1976), I calculated that capillaries 10 µm

in radius and larger were completely emptied during centrifugation and that all others

were reduced to <0.001 of their original capacity (see Nobel 1983; Mankiewicz 1987).

Sheathing leaf bases and tomentum of some mosses have capillaries with roughly a 10

µm radius (Proctor 1982). In a pilot study, well-blotted hydrated plants yielded no water

at 1450 × g or even 5220 × g, so internal water was not extracted during centrifugation.

Water collected from each clump sample was weighed to the nearest 0.1 mg.

137

Shoot external water-holding capacity—From the same set of plants, I also determined young shoot external water-holding capacity. After 1 month’s growth, one shoot was taken from the center of each replicate. At the time of sampling, the plants had yet to form clumps or express sex and were growing as isolated shoots having originated from protonemata (juvenile stage). Shoot external water retention was too small to be estimated by centrifugation, so I used blotting. Shoots were hydrated in distilled water for

1 h. Hydrated shoots were then removed from the water and blotted immediately on filter

paper (Whatman Grade 541). The outline of each blot was traced in pencil and area was

estimated using a dissecting microscope and a Macintosh computer with Image J. By

adding a known volume of water (0.25–5 µL) to filter paper, I calculated a constant by

which blot area could be multiplied to estimate volume. For the volumes used in the pilot

study, the linear relationship between actual and estimated volume was nearly 1:1 (0.97)

when the intercept was set to zero and had an R2 = 0.98. The length of each blotted shoot

was measured to standardize external water-holding capacity. I note that these values are potentially overestimated due to small volumes of water picked up on forceps wicking onto filter paper.

Extrapolating water-holding capacity from the shoot level to the clump level—To test whether differences between young shoots growing in isolation scale up to the clump, I used shoot external water-holding capacity per unit length to extrapolate from isolated young shoot to clump water-holding capacity. For each replicate, I matched shoot water capacity with clump height and shoot density. In cases where shoot data for a replicate was unavailable due to either a lack of shoots at time of isolated shoot sampling or a

138

small number of scribal errors, average for shoot water capacity across all available

isolate replicates was substituted. This substituted average shoot water capacity was then

linked with actual clump height and shoot density recorded for the same replicate at 9

months. As a result, 21 of 50 isolates had one replicate using the isolate average shoot

water capacity, only one isoolate had two. To get an extrapolated clump capacity, I

multiplied shoot water capacity by clump height and then by shoots per square

centimeter. Because shoot water-holding capacity is potentially overestimated, extrapolated clump water-holding capacity values may similarly be overestimated.

However, the apparent overestimation could also be due to differences in isolated shoots and those growing in clumps. Nevertheless, I was still able to determine an expected

direction of sex difference in clump water-holding capacity.

Statistical analyses—All statistical analyses were performed in the program SAS 9.3 or

9.4 (SAS Institute, Cary, North Carolina, USA). To test for overall sex differences in

water-holding capacity and clump morphology, I used a MANOVA. Dependent variables

were isolated shoot water-holding capacity, clump water-holding capacity, extrapolated

water-holding capacity, average clump height, shoot cross-sectional area, shoot density,

and proportion of shoot apices expressing sex. The overall MANOVA for sex was tested

with isolate nested within sex as the error term. I also used partial correlation coefficients

from the MANOVA to test for correlations among traits while controlling for the independent and other dependent variables. Data were transformed as described below.

139

Clump characteristics and water-holding capacity—I performed individual univariate

tests for each of the dependent variables again testing sex with isolate nested within sex

as the error term. In each case, the isolate effect was considered random. Clump water- holding capacity, clump average height, and shoot cross-sectional area were log-

transformed. Density was square-root-transformed and proportion of shoots expressing sex was arcsine square-root-transformed to improve normality.

Young shoot external water-holding capacity—The sex effect was tested using a nested

ANOVA. Isolate nested within sex was considered random and was used as the error term to test the sex effect. Water-holding capacity per unit length data were log- transformed.

Extrapolated clump water-holding capacity—The analysis for extrapolated clump water-

holding capacity was run identically as the univariate test for actual clump water-holding capacity. The dependent variable, extrapolated clump water-holding capacity, was log-

transformed. The analysis was run with and without the samples for which young shoot

water-holding capacity was represented by the isolate mean. The analysis presented in the

results includes values using a isolate average for shoot external water-holding capacity

(as discussed above). This analysis did not differ in interpretation from and was slightly

more conservative than the analysis excluding estimates using isolate averages for shoot

external water-holding capacity (not shown).

Additionally, I performed a Spearman rank correlation between extrapolated and

actual clump water-holding capacity within each isolate with sufficient replicates (25

140

female and 23 male isolates) to test how well extrapolated clump water-holding capacity

predicted the actual. Ranks were only assigned to replicates that were present both in

extrapolated and actual clump water-holding capacity measures. The results from these correlations were summarized using a one-sided exact binomial test. Assuming the null

hypothesis that there is no correlation between extrapolated and actual clump capacity is

true, I used an expected frequency of 0.05 for significant correlations within isolate. This

test allowed me to test whether there were more significant correlations than due to

chance alone.

Results

Clump characteristics and water-holding capacity—The overall MANOVA for the

effect of sex was significant (Wilks’ λ = 0.243, F7, 42 = 18.70, P < 0.0001). There were

also differences overall among isolates (Wilks’ λ = 0.04, F336, 1046.4 = 1.79, P < 0.0001).

For clump water-holding capacity, females held more water than males (F1, 48 = 14.07, P

= 0.0005; Figure 5.1).

141

Figure 5.1. Violin plot and overlaying box plot of isolate means of actual clump water- holding capacity in a 1 cm2 area of Bryum argentuem females (isolate n = 25) and males (isolate n = 25). Boxes mark the bounds of the 1st and 3rd quartiles. Whiskers mark the farthest data point within the 1st and 3rd quartiles minus or plus (respectively) 1.5 times the interquartile range. The median is represented by a solid line and the least squares mean by a dashed line.

Morphology that contributed to clump canopy attributes differed between males and females for each of our clump measurements. Females compared with males were taller (F1, 48 = 26.27, P < 0.0001), had a greater shoot cross-sectional area (F1, 48 = 25.51,

P < 0.0001), had lower shoot density (F1, 48 = 38.31, P < 0.0001), and had a lower proportion of visible shoot apices expressing sex (F1, 48 = 31.51, P < 0.0001).

Morphology and sex expression data are summarized in Table 5.1. Clump water-holding capacity was significantly positively correlated with both clump height and shoot cross- sectional area. Clump height was significantly positively correlated with shoot cross- sectional area. Lastly, shoot density was significantly negatively correlated with both clump height and shoot cross-sectional area and positively with proportion of shoot apices expressing sex. Partial correlations are summarized in Table 5.2.

142

Table 5.1. Least squares means ± SE of clump morphological characteristics and proportion of shoot apices expressing sex of Bryum argenteum. Trait Female Male Average clump height (mm) 8.21 ± 0.26 5.96 ± 0.27 Shoot cross-sectional area (mm2) 0.28 ± 0.01 0.18 ± 0.01 Shoot density (shoots·cm–2) 136.42 ± 6.26 196.50 ± 6.29 Proportion shoot apices expressing sex 0.056 ± 0.012 0.256 ± 0.012

Table 5.2. Partial correlation coefficients based on the residuals from the MANOVA on isolated shoot water capacity (ISC), extrapolated clump water-holding capacity (ECC), actual clump water-holding capacity (ACC), morphological characteristics, and proportion of shoot apices expressing sex (sex expression). Sex Variables ECC ACC Clump height Shoot area Shoot density expression

ISC 0.92 *** −0.10 ns −0.02 ns 0.18 * −0.03 ns −0.06 ns

ECC 0.11 ns 0.25 ** 0.17 * 0.13 ns −0.01 ns

ACC 0.60 *** 0.27 ** −0.08 ns 0.14 ns

Clump height 0.22 ** −0.36 *** 0.01 ns

Shoot area −0.30 ** −0.01 ns Shoot density 0.16 * Notes: *P < 0.05, **P < 0.01, ***P < 0.0001, ns P > 0.05

Shoot external water-holding capacity—Young isolated male shoots held more water per unit length externally than females (F1, 48 = 5.92, P = 0.019; Figure 5.2). This result

must be interpreted cautiously because the overall ANOVA was not significant (F49, 173 =

0.87, P = 0.716).

143

Figure 5.2. Violin plot and overlaying box plot of isolate means of external water- holding capacity of isolated young shoots of Bryum argenteum females (isolate n = 25) and males (isolate n = 25) standardized by their length. Boxes mark the bounds of the 1st and 3rd quartiles. Whiskers mark the farthest data point within the 1st and 3rd quartiles minus or plus (respectively) 1.5 times the interquartile range. The median is represented by a solid line and the least squares mean by a dashed line.

Extrapolated clump water-holding capacity—Extrapolated clump water-holding

capacity was greater for males than females (F1, 48 = 4.43, P = 0.041; Figure 5.3).

However, the overall ANOVA was not significant (F49, 163 = 1.04, P = 0.408). As noted

above (see Methods), these extrapolated clump water-holding capacities show an

apparent overestimation either due to overestimation of shoot water-holding capacity or

developmental changes from isolated shoots to those within clumps. Extrapolated clump

water-holding capacity was not correlated with actual clump water-holding capacity in

the partial correlations (Table 2). Furthermore, only five of 48 Spearman rank

correlations within isolates between extrapolated and actual clump capacity were

significant (P < 0.0001 in each case). Of these, three were negatively and two were positively correlated. The binomial test summarizing the results of the rank correlations was not significant (P = 0.091), suggesting very little evidence for more significant

within-isolate correlations than expected due to chance.

144

Figure 5.3. Violin plot and overlaying box plot of isolate means of extrapolated clump water-holding capacity in a 1 cm2 area of Bryum argenteum females (isolate n = 25) and males (isolate n = 25). Boxes mark the bounds of the 1st and 3rd quartiles. Whiskers mark the farthest data point within the 1st and 3rd quartiles minus or plus (respectively) 1.5 times the interquartile range. The median is represented by a solid line and the least squares mean by a dashed line.

Discussion

Female clumps hold more water than males, which supports my hypothesis that

divergent selection has shaped male and female clumps for relatively lower and higher

water-holding capacity, respectively. This result contrasts my findings for young shoots

and extrapolated clump water-holding capacity. Thus, the sex difference in actual clump

water-holding capacity arises at the clump stage. My findings analogously parallel

morphological differences found in other species of land plants, and I explore the

conceptual connection below. The sex difference in clump water-holding capacity also

has implications for explaining sex ratio bias in this species and potentially many others.

Lastly, considering that Bryum argenteum utilizes arthropods for sperm dispersal, the

results offer additional insights into this newly discovered interaction.

145

Divergent selection—The sex difference in clump water-holding capacity is consistent with the hypothesis that divergent selection has acted upon B. argentum clumps due to differences in sex function. Because an individual moss isolate produces a clump of shoots, selection can act upon the size, length, and arrangement of these shoots, which, even if not directly physiologically integrated, behave in many ways like a unit. Bryum argenteum, like other dioecious plants, has to solve the problem of gamete dispersal for males and gamete capture for females. Because water is the dispersal agent for bryophyte sperm, which can swim only a few centimeters (Wyatt 1977; Reynolds 1980; McLetchie

1996), movement of sperm laden water from male to female is especially important. For successful fertilization, females must also remain hydrated. Lower water stress is also beneficial for maturation of subsequent offspring (Stark et al. 2000, 2007). A male can gain higher sexual fitness by allowing water to move away, but a female that facilitates infiltration and retention of sperm-laden water increases sexual fitness via higher fertilization and offspring maturation rates compared with a female that does not. For males, fitness gain is traded off against hydration and thus growth and survival.

Conversely, for females, there is synergy between growth, survival, and sexual reproduction.

Stage effects, water-holding capacity, and clump morphology—Interestingly, actual water-holding capacity contrasted starkly with our extrapolated water-holding capacity, which further supports the idea that selection is acting at the clump stage to facilitate sexual reproduction. The clump stage is the point at which plants disperse or capture sperm, making this stage the most important one for interactions with water related to

146

sexual reproduction. Extrapolated clump water-holding capacity combines young shoot water-holding capacity with mature clump morphology to test whether clump water-

holding differences arise from differences present in young shoots. According to the

extrapolation, the greater isolated shoot water-holding capacity of males ought to scale up

to a greater clump water-holding capacity despite females’ greater average clump height.

This contrast with actual clump water-holding capacity is evidence that clump capacity

arises at this later stage of maturity. One explanation for the difference between actual

and extrapolated clump water-holding capacity is that overall clump morphology

overrides the shoot sex difference. That is, greater female water-holding capacity could be an emergent clump property with selection acting on interactions between shoots rather than simply the additive effects of individual shoot properties. Accordingly, females could have higher capillary capacity between shoots due to their height and other interactions between shoots. That is, capillary spaces between shoots could be relatively more important to the sex difference in clump water-holding capacity than the additive effect of water trapped between clasping leaves of individual shoots alone. However, spaces between shoots would need to be small enough to retain a significant amount of water through capillary action. The propensity of B. argenteum to produce rhizoids along older shoot bases (personal observation) could assist significantly in making use of spaces between shoots for water retention, and the morphology of these rhizoids could differ between the sexes. Additionally, even though as measured it is a shoot morphological difference, the interaction between more robust female shoots may create a clump morphological difference allowing them to hold more water than males. I do not believe these size differences are simply due to differential reproductive investment

147

(Horsley et al. 2011) because the sex differences between sex-expressing plants is

consistent with data from the few plants that did not express sex (data not shown) and

previous experiments (unpublished data). There is also no correlation between proportion of sex-expressing apices and clump height.

The other possible interpretation is that individual shoot morphology changes during the process of clump formation and the differences in clump water-holding capacity are mostly due to additive shoot properties, though water between shoots would still likely contribute. If this is the case, then mature females should hold more water per length of individual shoots than mature males to produce, via greater length, a difference in clump water-holding capacity. However, I did not measure water-holding capacity of individual shoots from mature clumps. The strong positive correlation between clump height and water-holding capacity could lend support to either this explanation or the previous one. These two explanations are not mutually exclusive but rather represent two ends of a spectrum. In either case, the reversal of direction in sex difference between extrapolated and actual clump water-holding capacity combined with the lack of correlation in both the partial correlations and within-isolate rank correlations suggests that clump water-holding capacity is not predictable from young shoot water-holding characteristics generally and in terms of sex differences. Future experiments tracking shoot and clump water-holding capacity and morphology, including internal clump morphology, as clump formation proceeds would shed light on the contribution of developmental processes and interactions between shoots in clumps to the sex difference in water-holding capacity.

148

For males to use water to carry sperm away, I predicted loosely packed, short,

slender shoots. For females to facilitate infiltration and retention of sperm-laden water, I

predicted densely packed, tall, robust shoots with fertility sex expression rates lower than

males. Males, however had higher shoot density, which likely results from small shoot

cross-sectional area paired with tight packing (Figure 5.4). This pattern could be explained simply by within-isolate resource competition between shoots that self-thin as they grow taller, which is supported by a positive correlation of shoot cross-sectional area with clump height and negative correlations of shoot density with both clump height and shoot cross-sectional area. Males may also produce more but smaller shoots to increase sex structure number, which also occurs in angiosperms (Bond and Maze 1999). My data

on sex structure production is consistent with a previous study on B. argenteum by

Horsley et al. (2011). I had imagined that male clumps would be more diffuse, but quick

drying resulting from a diffuse clump could be detrimental. If male shoots remain tightly

packed to prevent water loss, clumps may remain short to decrease water flow resistance

for sperm dispersal, which is consistent with our primary hypothesis. Some species of

aquatic mosses have shorter, denser clumps to decrease resistance to water flow

(Mägdefrau 1982). Sperm dispersal could proceed in this manner: a coherent sperm mass is released from an antheridium, floats to the top of the water column, and disperses across the water surface to be carried away by current, which is more easily available due to a short clump profile. None of these potential reasons need be mutually exclusive and may all contribute to sex-specific clump attributes.

149

Figure 5.4. Photographs of a representative female (A) and male (B) of Bryum argenteum. Arrows indicate a shoot apex with sex organs. While there was variation within males and females, these photos illustrate the sex differences I found. Note the smaller cross-sectional area and denser packing of male vs. female shoots. Note also the leaves subtending sex organs; these are longitudinally curled with longer awns in females but are more clasping around antheridia with shorter awns in males.

Conceptual connection to seed plants—If my interpretation is correct that selection due to gamete dispersal and capture requirements has shaped the clump stage of B. argenteum, my results connect broadly to studies in other land plant lineages. As

mentioned above, in seed plants, divergent selection has shaped flowers, inflorescences,

cones, and canopies in monoecious and dioecious wind- and water-pollinated plants to

facilitate pollen movement from male to female sex organs (Niklas 1985a, b; Niklas and

Buchmann 1985; Bond and Maze 1999). Being composed of both shoots with and without sex organs, bryophyte clumps function like the inflorescences, cones, or even the branches or canopies of seed plants in that all these direct movement of an abiotic

dispersal medium. The morphology of short, dense male B. argenteum clumps that helps

sperm reach water current is analogous to adaptations of some grass inflorescences

(exserted stamens), male pine cones, and flowering plant catkins to move pollen beyond

boundary layers (Niklas 1985a, b). On the other hand, tall, less-dense female clumps with

150 more robust shoots could slow the flow of and retain more sperm-laden water. The interaction with water for sperm capture case is conceptually analogous to morphological adaptions that direct air flow for pollen deposition in seed plants such as Simmondsia chinensis (leaves and peduncles) and pine (cones and leaves) (Niklas 1985b; Niklas and

Buchmann 1985).

Implications for other species of bryophytes—My interpretation of sex-specific clump attributes in B. argenteum applies most immediately to clump traits in many other species of bryophytes. Studies on the function of different clump morphologies have largely been confined to comparison among species or within species without regard to sex (e.g. Dilks and Proctor 1979; Rice et al. 2001; Rice and Schneider 2004; Elumeeva et al. 2011;

Michel et al. 2013), and we know of no other studies on how clump morphology might affect sperm dispersal or capture. As an example, our interpretation could help explain why females of the moss Ceratodon purpureus produce fewer but larger shoots than males (Shaw and Gaughan 1993).

In addition, my interpretation of clump traits may also apply to sex structure traits. In bryophytes, studies on specialized morphological adaptations have usually focused on sperm dispersal ability including that of splash cups in some mosses

(Reynolds 1980; Andersson 2002), but to my knowledge no studies have documented the connection between female sex structure morphology and function. In B. argenteum, leaf arrangement around sex organs is dimorphic (Figure 5.4) such that male leaves clasp around antheridia much more closely, while female leaves around archegonia flare out and curl near an elongated apex, creating a partial tube. Future experiments could show

151

that these differences are morphological adaptations to shed water in males and retain or

direct water in females. Similar morphological adaptions could be found in many species

of dioecious mosses or even liverworts. For instance, the dense tomentum-like scales

surrounding the archegonia of Marchantia inflexa seem well suited for retention of

sperm-laden water (J. D. Moore and D. N. McLetchie, personal observation). On the

other hand, male sex structures of M. inflexa completely lack these scales, which could

permit relatively unhindered water flow, and the male structure of M. polymorpha has

been suggested to function like a splash cup (Glime 2013).

Consequences for sex ratio—The sex differences in clump water-holding capacity and other traits also suggest potential explanations for the frequency of female-biased sex ratios among dioecious clump-forming bryophytes. To my knowledge, no other study has attempted to connect sex-specific clump morphology and water-holding capacity to skewed sex ratios. In my study populations, B. argenteum is most often found as

apparently single sex clumps (Chapter Three), which could be caused by asexual and

sexual propagule dispersal or intersexual competitive exclusion as suggested for other

species (McLetchie et al. 2001; Crowley et al. 2005). For B. argenteum, the sex

difference in clump height suggests that females could overgrow males, which could by

itself contribute to female-biased sex ratios. These single-sex clumps, no matter their

ultimate cause, most often allow formation of the sex-specific patterns we found. Water-

holding capacity alone, due to its potential physiological consequences leading to female

advantage, may explain sex ratio bias in B. argenteum including female-only desert

populations (Stark et al. 2010). In principle but without regard to sex, intraspecific clump

152

physiological differences due to water-holding capacity have been shown in other

species. For instance, higher water-holding capacity in larger clumps leads to longer

physiologically active periods in both Leucobryum glaucum (Rice and Schneider, 2004)

and Grimmia pulvinata (Zotz et al. 2000), but not always higher carbon fixation in G. pulvinata. A counter example is provided by a greenhouse study of 22 Artic bryophyte species among which clump desiccation rate correlated negatively with shoot size and positively with shoot density (Elumeeva et al. 2011). However, because B. argenteum is only comparable in size to the smallest species in that study (Philonotis caespitosa and

Paludella squarrosa) (Crum and Anderson 1981), intersexual variation in B. argenteum

is likely comparatively small. Assuming higher female water-holding capacity for B.

argenteum leads to longer photosynthetically active periods and greater carbon fixation,

females would grow faster under water-limiting conditions, which are quite frequent.

Additionally, slower drying allows for greater preparation for desiccation (Farrant et al.

1999; Oliver et al. 2000; Cruz de Carvalho et al. 2011), which could lead to faster female recovery after rewetting. Field trials for B. argenteum in natural habitats after the manner

of Rice and Schneider (2004) are needed to determine whether higher female water-

holding capacity leads to female advantage. Studies seeking similar clump characteristics

to those presented here and their consequences in other species will bear out the general applicability of our results to sex-specific life histories and skewed sex ratios among bryophytes.

153

Potential new insights for arthropod sperm dispersers—Because some bryophytes can use both abiotic and animal gamete dispersers (Cronberg et al. 2006; Rosenstiel et al.

2012), their gamete dispersal system may be analogous to those of flowering plants that use both insects and wind for pollen dispersal (Culley et al. 2002). Therefore, my findings here may not only be important for strict water dispersal of male gametes but also for dispersal via potential animal mutualists already identified for B. argenteum and

C. purpureus. These small arthropods prefer female volatile compounds to those of males, but the reason is unknown (Rosenstiel et al. 2012). Dioecious cycads present an apparently opposite pattern where male strobili have higher thermogenic metabolism

(presumably to volatilize organic compounds) and peak earlier, while females remain active longer. The dimorphic pattern may guide pollinators to collect pollen first and then to move to female strobili (Roemer et al. 2005). Males of dioecious flowering plants are also typically expected to have larger floral displays than females for pollinator attraction

(Barrett and Hough 2013). The attraction of springtails to females over males in B. argenteum is surprising considering the typically larger and more numerous male sex structures. Because springtails require moist habitats (Chikoski et al. 2006), I make the novel suggestion that their attraction to females can also be driven by higher water retention, making females a safer habitat relative to males. If male and female clumps are close, arthropods may move from male to female clumps as clumps dry, allowing them to disperse sperm during migration. Both the similarity and contrast with seed plant systems suggests that much interesting work remains to determine how B. argenteum uses animals to transfer sperm and how well developed the attraction system is.

154

Conclusions

I have demonstrated under common garden conditions that female clumps hold more water than males on average. I found clump morphological differences that could help explain the difference in water-holding capacity. Females have taller clumps, more robust shoots, and lower shoot density. I also found evidence that the sex difference in water- holding capacity emerges as clumps form and is not predictable based on isolated young shoots. These results have great implications for connecting divergent selection to life history differences and female-biased sex ratios in B. argenteum and other bryophytes. If females are able to hold more water, they may also stay hydrated for longer periods increasing growth rates and chances of survival. These sex differences in morphology and water-holding capacity are consistent with my hypothesis that male sexual fitness is increased by retaining less water, while female sexual fitness is increased by retaining more water.

This chapter was previously published under the same title by JD Moore,

LM Kollar, and DN McLetchie in

American Journal of Botany Vol. 108 no. 8: 1-9.

155

CHAPTER SIX: CONCLUSIONS AND FUTURE DIRECTIONS

Differential selection due to the different reproductive functions of the sexes has

led to the evolution of a myriad of sexual dimorphisms in plants, which can lead to

skewed population sex ratios. In seed plants, male-bias has been associated with higher

female reproductive investment than males, but a number of species in seed plants and

especially bryophytes have female-biased population sex ratios. In this dissertation, I

found several dimorphic traits that can offer some explanation for female-biased sex ratios in bryophytes and potentially for other plant groups along with trait relationships that aid in understanding the ecology of Bryum argenteum and other mosses.

Sex differences—Bryum argenteum females tend to produce more shoots as young sexually immature plants especially in higher light (Chapter Three), have more robust shoots, grow taller, and have higher clump-level external water-holding capacity than males (Chapter Five, published as Moore et al. 2016). Male clumps have smaller shoots but higher density than females (Chapter Five). These dimorphic traits suggest that females could colonize space more quickly than males during early growth and then either 1) continue growing faster than males via longer hydrated active periods, 2) overgrow males, or 3) survive better than males in competition with other plants. The first suggestion could be confirmed by future experiments testing for sex-specific drying rates and photosynthetic activity using mature clumps either in the lab or in field trials following Rice and Schneider (2004). The second suggestion could be addressed in future studies by planting males and females together to assess whether one sex overgrows the other, and the third could be addressed by growing males and females with

156 a common competitor. If either of the first two suggestions is true, then the persistence of males in all habitats would require that males grow separated from females, potentially reducing sexual reproduction, and all three suggestions would require that males recolonize via spore or asexual propagules. The contribution of colonization via spore and asexual propagules, including propagule production, dispersal, and germination rates, to male persistence needs further investigation.

Despite faster female recovery from severe stress than males (using Fv/Fm), it is still unclear whether one sex is more stress tolerant (Chapter Four). Though stress indeed reduces shoot production in B. argenteum, I found no difference between the sexes in shoot production after stress. The difference between the sexes in size, which I did not control for, could have prevented the detection of sex differences in shoot production after stress. Future experiments should attempt to connect the patterns in stress tolerance to measures of growth while controlling for size, which could include using shoots of a standard size or measuring differences in biomass gained after stress treatment (Stark et al. 2009). Other measures of damage caused by heat and desiccation stress, including estimating the amount of tissue surviving stress, photosynthetic rates, photosynthetic nitrogen use efficiency, or concentration of chlorophyll, which is known to be destroyed in bryophytes by stress (Glime 2007b; Stark et al. 2016), would flesh out the sex- differences in stress response. The same measurements could be made to compare sex- expressing with vegetative shoots from mature clumps to explain why sex-expressing shoots have higher Fv/Fm in general and recovery more quickly from extreme stress.

Nevertheless, contrary to my predictions, I found no evidence of increasingly female- biased sex-ratios in harsher (based on exposure) microhabitats (Chapter Three). The lack

157

of pattern could be attributable to the mildness of central Kentucky’s climate or no

meaningful sex-difference in stress tolerance. The latter could occur if inducible

hardening to stresses masks differences present in a dehardened state.

Sexually dimorphic traits among sexually immature plants—Though previous studies

have shown that males invest more in reproduction pre-zygotically than females, which should translate to higher average reproductive investment for males given low fertilization rates, the appearance of sex differences among sexually immature individuals

in my studies (Chapters Three, Four, and Five) and in other species (e.g. Newton 1972;

Stieha et al. 2014; Marks et al. 2016) argue against the idea that sexual dimorphism in B.

argenteum and other bryophytes is a simple consequence of resource allocation tradeoffs

at the time of reproduction. Dimorphic traits among sexually immature individuals that I

found include sex differences in growth rate, stress responses, and single-shoot external water-holding capacities. The prevalence of sex differences among sexually immature individuals in seed plants is largely unexplored probably due to the difficulty identifying sex, but my studies (among others on bryophytes) suggest sex differences among sexually immature individuals should be explored more fully in other plant groups.

Herbaceous species of seed plants with short life-cycles seem ideally suited to allow faster sex identification through flowering. For longer-lived species where sex-

expression can take several years, genetic sex markers would be necessary in many cases

to identify the sex of sexually immature plants.

158

Sexually immature traits may be non-adaptively genetically correlated with sexually mature adult traits (i.e. by-products of selection on adult traits) or adaptively correlated, enhancing future fitness (Case and Ashman 2005). Extensive knowledge of the function of such traits and their connections to future fitness would be necessary to determine whether they are non-adaptively or adaptively correlated with adult traits.

Resource partitioning to sexual reproduction, generally assumed to begin at sexual maturity, is only part of the equation for fitness. Resource storage for future expression of sex could begin at immature stages and differ among the sexes. For instance, future studies should attempt to connect the growth rate differences at immature stages (Chapter

Three) with future reproductive investment. Additionally, these studies should seek connections between immature growth rates and photosynthetic rate or photosynthetic nitrogen use efficiency (nitrogen could be an important currency for males; Chapter

Two), and compare these with the same traits in adults.

Other traits that begin in immature plants could include traits like branch and whole-plant morphology, which affects the transfer of male gametes for plants utilizing abiotic vectors (for wind-pollinated plants, see Niklas 1985a, b). The morphology necessary for male gamete transfer must be in place before expression of sex in order to be effective. In water-fertilized systems, mating success may be affected by shoot size and arrangement in a clump. For males, clump morphology needs to permit the carrying away of sperm-laden water, and for females, clump morphology should favor holding

water. The arrangement of the clump is dependent on shoot characteristics produced as

sexually immature plants and also depends on the traits of non-sex-expressing mature

vegetative shoots. To demonstrate the adaptive significance of sexually dimorphic traits

159

among immature plants in B. argenteum, a necessary next step would be to connect those

traits to the formation of dimorphic clumps, which would require tracking clump

development through time. To verify the hypothesis that dimorphic clumps facilitate

mating success, one could use males of two different water holding capacities and test

their mating success against a common female. In Chapter Five, there is evidence of two

different water holding strategies for males (Figure 5.1), though not explored there,

which could provide the opportunity to identify males of different water-holding

strategies.

Similarly, dimorphic traits among sexually immature individuals in abiotically-

pollinated seed plant species could contribute to sex-specific morphologies that facilitate the transfer of pollen from male to female. Regardless of when they arise, the consequences of sex-specific morphology in seed plants is not well understood, but could contribute to the development of female-biased sex ratios in these taxa, which occur among gymnosperm and angiosperm shrubs and in clonal herbaceous angiosperms

(Chapter Two; Sinclair et al. 2012; Field et al. 2013). Further studies are needed to assess the connection between sex-specific morphology and life history traits leading to skewed population sex ratios.

Ecology—Additionally, Chapter Three offers insight on the acclimation of moss leaves to high exposure/light levels, and corroborates earlier work (Schroeter et al. 2012) suggesting that B. argenteum uses changes in leaf morphology to create its own shade in high light, which fits with the paradigm that all bryophytes are essentially shade plants

(from a vascular plant perspective). I found that chlorophyll and carotenoids increase as

160

light increases, likely due to the shade created by shifts in leaf morphology, but pigment

ratios did not change. However, I did not measure the relative contribution of different carotenoid pigments, and future experiments on changes in xanthophyll cycle pigments, which were important for B. argenteum var. muticum (Schroeter et al. 2012) in response

to high light, are needed to understand the physiological acclimation mechanisms of B.

argenteum.

Changes in morphology included greater leaf achlorophyllous areas and smaller

leaf length:width ratios as field exposure increased, which create a reflective surface and

avoid excess light, respectively. Using the same isolates from the field in a common

garden indicated leaf length:width is plastic to light, but leaf achlorphyllous area may

respond to moisture. The role of the leaf apex in the acclimation of B. argenteum to

increased exposure is not yet clear but may also respond to moisture, though opposite the

expectation from another moss (Glime 2007a, 2015; Reynolds and McLetchie 2011; Tao

and Zhang 2012). The response of leaf achlorphyllous area and apex length to moisture

could be addressed in a future study by manipulating moisture regimes of plants in a

common garden such that some plants stay constantly moist while other receive wet-dry

cycles.

There was considerable variation in the field unaccounted for by the

environmental gradient I measured. One potential reason is that leaf morphology

interacts with whole clump morphology, which includes shoot density. Shoot density is

an important factor for water relations in other species of mosses (Elumeeva et al. 2011) and could also influence self-shading. Future experiments should test how shoot density affects leaf morphology by accounting for density in the field. Common garden

161

experiments could use isolates (both males and females) that have been previously

identified (using data from Chapter Five) as differing genetically in shoot density placed

in different light and moisture regimes.

In Chapter Three, I found evidence of genetic variation in morphology and

growth among habitats despite the small distances between them. To test whether the

variation is ecotypic, reciprocal transplants could be performed using individuals from

habitats identified as differing genetically (Chapter Three). These individuals could then be tracked for survival and other measures of fitness, including biomass gain, size, and reproductive output. If individuals derived from one habitat and transplanted to another do more poorly than native individuals, then the genetic variation is likely adaptive.

In summary, this dissertation shows the value and tractability of using B. argenteum to address questions related to the evolution of sex-specific traits and their connection to skewed population sex ratios not only in dioecious bryophytes but seed plant systems as well.

162

APPENDIX 1: CHLOROPHYLL A:B RATIOS FOR THREE POPULUS SPECIES

Table A1.1. Chlorophyll a:b ratios calculated for three Populus species using data reported in Table 1 of Cha (1987). I analyzed the a:b ratios using t-tests in JMP 12 (SAS Institute, Cary, North Carolina, USA) for each of the species (n = 9 females and n = 9 males for each). Means ± standard errors.

P. angustifolia P. sargentii P. tremuloides Female 0.90 ± 0.02 1.03 ± 0.04 0.82 ± 0.03 Male 0.95 ± 0.04 0.98 ± 0.03 0.90 ± 0.01 t-ratio 1.27 -1.12 2.22 Adjusted DF 10.71 15.78 10.21 P 0.230 0.280 0.0501

163

APPENDIX 2: CHLOROPHYLL AND CAROTENOID EXTRACTION AND QUANTIFICATION

I extracted five shoots from each collected clump of Bryum argenteum by placing them in

1.5 ml microcentrifuge tubes with 0.5 ml of DMSO overnight in the dark. A well plate

reader was used due to its ability to handle up to 96 small (300 µL) samples at once.

However, Wellburn (1994) equations for a 1-4 nm resolution spectrophotometer require

absorbances taken at 480, 649, 665 nm, and the well plate reader only has 465, 650, and

660 nm available. In order to convert absorbances from the well plate reader, I created linear equations after the manner of Groen et al. (2010). I extracted approximately 0.75 g of fresh green B. argenteum tissue overnight in DMSO. I then diluted the resulting solution to give a series of relative concentrations: 5, 10, 20, 30, 40, 50, 60, 70, 80, 90,

and 100 %. The absorbances of these solutions were taken both on a spectrophotometer

(Spectronic 601, Milton Roy Co., Rochester, NY, U.S.A.) and on the well plate reader

(300 µL per well). I then used equations from Wellburn (1994) to estimate pigment

concentrations. Linear equations were created using concentrations estimated from the

well plate reader as the independent variable and concentrations estimated from the

spectrophotometer as the dependent variable. In each case the relationship was

significant (P < 0.0001). Equations were as follows:

2 Ya = 1.52317 Xa – 0.08376 (R = 0.995)

2 Yb = 1.20971 Xb – 0.06966 (R = 0.992)

2 Yc = 1.04403 Xc + 0.01502 (R = 0.997)

164

In each equation, Y stands for the converted concentration and X stands for the concentration estimated using absorbances available on the well plate reader. The letters a, b, and c stand for chlorophyll a, b, and total carotenoids, respectively.

165

REFERENCES Abraham, A., and Mathew, P.M. 1962. Cytological studies in the cycads: sex chromosomes in Cycas. Ann. Bot. 26(102): 261-266. Ackerman, J.D. 1997. Submarine pollination in the marine angiosperm Zostera marina (Zosteraceae). I. The influence of floral morphology on fluid flow. Am. J. Bot. 84(8): 1099–1109. doi:10.2307/2446153. Allen, C.E. 1917. A chromosome difference correlated with sex differences in Sphaerocarpos. Science 46(1193): 466-467. Allen, C.E. 1919. The basis of sex inheritance in Sphaerocarpos. Proc. Am. Philos. Soc. 58(5): 289-316. Allen, G.A., and Antos, J.A. 1993. Sex ratio variation in the dioecious shrub Oemleria cerasiformis. Am. Nat. 141(4): 537-553. doi: 10.1086/285490. Alvarenga, L.D.P., Pôrto, K.C., and Zartman, C.E. 2013. Sex ratio, spatial segregation, and fertilization rates of the epiphyllous moss Crossomitrium patrisiae (Brid.) Müll.Hal. in the Brazilian Atlantic rainforest. J. Bryol. 35(2): 88-95. doi: 10.1179/174328213x13662092820316. Álvarez-Cansino, L., Díaz Barradas, M.C., Zunzunegui, M., Esquivias, M.P., and Dawson, T.E. 2012. Gender-specific variation in physiology in the dioecious shrub Corema album throughout its distributional range. Funct. Plant Biol. 39(12): 968-978. Álvarez-Cansino, L., Zunzunegui, M., Díaz Barradas, M.C., and Esquivias, M.P. 2010. Physiological performance and xylem water isotopic composition underlie gender-specific responses in the dioecious shrub Corema album. Physiol. Plant. 140(1): 32-45. doi: 10.1111/j.1399-3054.2010.01382.x. Andersson, K. 2002. Dispersal of spermatozoids from splash-cups of the moss Plagiomnium affine. Lindbergia 27: 90–96. Andersson, M., and Iwasa, Y. 1996. Sexual selection. Trends Ecol. Evol. 11(2): 53-58. doi: 10.1016/0169-5347(96)81042-1. Antos, J.A., and Allen, G.A. 1990. A comparison of reproductive effort in the dioecious shrub Oemleria cerasiformis using nitrogen, energy and biomass as currencies. Am. Midl. Nat. 124(2): 254-262. doi: 10.2307/2426174. Antos, J.A., and Allen, G.A. 1994. Biomass allocation among reproductive structures in the dioecious shrub Oemleria cerasiformis -- A functional interpretation. J. Ecol. 82: 21-29. Ashman, T.-L. 2009. Sniffing out patterns of sexual dimorphism in floral scent. Funct. Ecol. 23(5): 852-862. doi: 10.1111/j.1365-2435.2009.01590.x. Balaguer, L., Martínez‐Ferri, E., Valladares, F., Pérez-Corona, M.E., Baquedano, F.J., Castillo, F.J., and Manrique, E. 2001. Population divergence in the plasticity of the response of Quercus coccifera to the light environment. Funct. Ecol. 15(1): 124-135. Barradas, M.C.D., and Correia, O. 1999. Sexual dimorphism, sex ratio and spatial distribution of male and female shrubs in the dioecious species Pistacia lentiscus L. Folia Geobot. 34(1): 163-174. doi: 10.1007/bf02803082. Barrett, S.C.H., and Hough, J. 2013. Sexual dimorphism in flowering plants. J. Exp. Bot. 64(1): 67-82. doi: 10.1093/jxb/ers308.

166

Barrett, S.C., Yakimowski, S.B., Field, D.L., and Pickup, M. 2010. Ecological genetics of sex ratios in plant populations. Philos. Trans. R. Soc., B 365(1552): 2549-2557. doi: 10.1098/rstb.2010.0002. Berry, J.F., and Shine, R. 1980. Sexual size dimorphism and sexual selection in turtles (Order Testudines). Oecologia 44: 185-191. Bierzychudek, P., and Eckhart, V. 1988. Spatial segregation of the sexes of dioecious plants. Am. Nat. 132(1): 34-43. Bisang, I., and Ehrlén, J. 2002. Reproductive effort and cost of reproduction in female Dicrunum polysetum. Bryologist 105(3): 384-397. Bisang, I., Ehrlén, J., and Hedenäs, L. 2006. Reproductive effort and costs of reproduction do not explain female-biased sex ratios in the moss Pseudocalliergon trifarium (Amblystegiaceae). Am. J. Bot. 93(9): 1313-1319. doi: 10.3732/ajb.93.9.1313. Bisang, I., and Hedenäs, L. 2005. Sex ratio patterns in dioicous bryophytes re-visited. J. Bryol. 27(3): 207-219. doi: 10.1179/174328205x69959. Biswas, C., and Johri, B.M. 1997. The Gymnosperms. Springer-Verlag, Berlin. Boecklen, W.J., and Hoffman, M.T. 1993. Sex-biased herbivory in Ephedra trifurca: the importance of sex-by-environment interactions. Oecologia 96: 49-55. Boecklen, W.J., Price, P.W., and Mopper, S. 1990. Sex and drugs and herbivores: sex- biased herbivory in arroyo willow (Salix lasiolepis). Ecology 71(2): 581-588. doi: 10.2307/1940311. Bond, W.J., and Maze, K.E. 1999. Survival costs and reproductive benefits of floral display in a sexually dimorphic dioecious shrub, Leucadendron xanthoconus. Evol. Ecol. 13(1): 1-18. doi: 10.1023/a:1006581412580. Bond, W.J., and Midgley, J. 1988. Allometry and sexual differences in leaf size. Am. Nat. 131(6): 901-910. Bonduriansky, R. 2011. Sexual selection and conflict as engines of ecological diversification. Am. Nat. 178(6): 729-745. doi: 10.1086/662665. Boardman, N.K. 1977. Comparitive photosynethesis of sun and shade plants. Annu. Rev. Plant Physiol. 28: 355-377. Bowker, M.A., Stark, L.R., McLetchie, D.N., and Mishler, B.D. 2000. Sex expression, skewed sex ratios, and microhabitat distribution in the dioecious desert moss Syntrichia caninervis (Pottiaceae). Am. J. Bot. 87(4): 517-526. Bragg, J.G and Westoby, M. 2002. Leaf size and foraging for light in a schlerophyll woodland. Funct. Ecol. 16(5): 633-639. Bram, M.R., and Quinn, J.A. 2000. Sex expression, sex-specific traits, and the effects of salinity on growth and reproduction of Amaranthus cannabinus (Amaranthaceae), a dioecious annual. Am. J. Bot. 87(11): 1609-1618. Brus, R., Ballian, D., Zhelev, P., Pandža, M., Bobinac, M., Acevski, J., Raftoyannis, Y., and Jarni, K. 2011. Absence of geographical structure of morphological variation in Juniperus oxycedrus L. subsp. oxycedrus in the Balkan Peninsula. Eur. J. For. Res. 130(4): 657-670. doi: 10.1007/s10342-010-0457-1. Brzyski, J.R., Taylor, W., and McLetchie, D.N. 2014. Reproductive allocation between the sexes, across natural and novel habitats, and its impact on genetic diversity. Evol. Ecol. 28(2): 247-261. doi: 10.1007/s10682-013-9672-9.

167

Burnell, K.D., Yelverton, F.H., Neal, J.C., Gannon, T.W., Scott, J., and McElroy, J.S. 2004. Control of silvery-thread moss (Bryum argenteum Hedw.) in creeping bentgrass (Agrostis palustris Huds.) putting greens. Weed Technol. 18(3): 560– 565. doi:10.1614/WT-03-082R1. Buryová, B. and Shaw, A.J. 2005. Phenotypic plasticity in Philonotis fotana (Bryopsida: Bartramiaceae). J. Bryol. 27(1): 13-22. Cameron, R.G., and Wyatt, R. 1990. Spatial patterns and sex ratios in dioecious and monoecious mosses of the genus Splachnum. Bryologist 93(2): 161-166. Carlsson-Graner, U., Elmqvist, T., and Agren, J. 1998. Floral sex ratios, disease and seed set in dioecious Silene dioica. J. Ecol. 86(1): 79-91. Case, A.L., and Ashman, T.-L. 2005. Sex-specific physiology and its implications for the cost of reproduction. Edited by E.G. Reekie and F.A. Bazzaz. Elsevier Academic Press, Boston. pp. 129-157. Cedro, A., and Iszkuło, G. 2011. Do Females Differ from Males of European Yew (Taxus baccata L.) in Dendrochronological Analysis? Tree-Ring Res. 67(1): 3-11. doi: 10.3959/2009-9.1. Cha, Y. 1987. Leaf chlorophyll contents of three Populus (Salicaceae) species occuring on elevational gradient. Korean J. Ecol. 10(119-127). [in Korean with English abstract] Charnov, E.L. 1979. Simultaneous hermaphroditism and sexual selection. Proc. Natl. Acad. Sci. USA 76(5): 2480-2484. Che-Castaldo, C., Crisafulli, C.M., Bishop, J.G., and Fagan, W.F. 2015. What causes female bias in the secondary sex ratios of the dioecious woody shrub Salix sitchensis colonizing a primary successional landscape? Am. J. Bot. 102(8): 1309- 1322. doi: 10.3732/ajb.1500143. Cheptou, P., Carrue, O., Rouifed, S., and Cantarel, A. 2008. Rapid evolution of seed dispersal in an urban environment in the weed Crepis sancta. Proc. Natl. Acad. Sci. USA 105(10): 3796-3799. Chikoski, J. M., S. H. Ferguson, and L. Meyer. 2006. Effects of water addition on soil arthropods and soil characteristics in a precipitation-limited environment. Acta Oecol. 30(2): 203–211. doi:10.1016/j.actao.2006.04.005. Chopra, R. N., and S. C. Bhatla. 1981. Effect of physical factors on gametangial induction, fertilization and sporophyte development in the moss Bryum argenteum grown in vitro. New Phytol. 89(3): 439-447. Cipollini, M.L., and Whigham, D.F. 1994. Sexual dimorphism and cost of reproduction in the dioecious shrub Lindera benzoin (Lauraceae). Am. J. Bot. 81(1): 66-75. Clark, D.A., and Clark, D.B. 1987. Temporal and environmental patterns of reproduction in Zamia skinneri, a tropical rain forest cycad. J. Ecol. 75: 135-149. Clark, D.B., and Clark, D.A. 1988. Leaf production and the cost of reproduction in the neotropical rain forest cycad, Zamia skinneri. J. Ecol. 76(4): 1153-1163. Cornelissen, T., and Stiling, P. 2005. Sex-biased herbivory: a meta-analysis of the effects of gender on plant-herbivore interactions. Oikos 111(3): 488-500. doi: 10.1111/j.1600-0706.2005.14075.x. Correia, O., and Diaz Barradas, M.C. 2000. Ecophysiological differences between male and female plants of Pistacia lentiscus L. Plant Ecol. 149(2): 131-142. doi: 10.1023/a:1026588326204.

168

Cox, C.J., Li, B., Foster, P.G., Embley, T.M., and Civán, P. 2014. Conflicting phylogenies for early land plants are caused by composition biases among synonymous substitutions. Syst. Biol. 63(2): 272-279. doi: 10.1093/sysbio/syt109. Cronberg, N. 2012. Animal-mediated fertilization in bryophytes–parallel or precursor to insect pollination in angiosperms? Lindbergia 35: 76-85. Cronberg, N., Natcheva, R., and Berggren, H. 2008. Observations regarding the life cycle of silvermoss Bryum argenteum. In Bryology in the new millennium. Edited by H. Mohamed, B. B. Baki, A. Nasrulhaq-Boyce, and P. K. Y. Lee. University of Malaya, Kuala Lumpur. pp. 347–352. Cronberg, N., Natcheva, R., and Hedlund, K. 2006. Microarthropods mediate sperm transfer in mosses. Science 313: 1255. Crowley, P.H., Stieha, C.R., and McLetchie, D.N. 2005. Overgrowth competition, fragmentation and sex-ratio dynamics: a spatially explicit model. J. Theor. Biol. 233(1): 25-42. doi: 10.1016/j.jtbi.2004.09.017. Cruden, R.W., and Lyon, D.L. 1985. Patterns of biomass allocation to male and female functions in plants with different mating systems. Oecologia 66(2): 299-306. Crum, H. 1983. Mosses of the Great Lakes forest, 3rd ed. University of Michigan, Ann Arbor, Michigan. Crum, H.A., and Anderson, L.E. 1981. Mosses of Eastern North America. Columbia University Press, New York. pp. 570-572. Cruz de Carvalho, R., Branquinho, C., and Marques da Silva, J. 2011. Physiological consequences of desiccation in the aquatic bryophyte Fontinalis antipyretica. Planta 234(1): 195–205. doi:10.1007/s00425-011-1388-x. Culley, T.M., Weller, S.G., and Sakai, A.K. 2002. The evolution of wind pollination in angiosperms. Trends Ecol. Evolut. 17(8): 361–369. doi:10.1016/S0169- 5347(02)02540-5. Dale, M.P., and Causton, D.R. 1992. Use of the chlorophyll a/b ratio as a bioassay for the light environment of a plant. Funct. Ecol. 6(2): 190-196. Darwin, C. 1871. The descent of man and selection in relation to sex. Murray, London. Davey, M.L., and Currah, R.S. 2006. Interactions between mosses (Bryophyta) and fungi. Can. J. Bot. 84(10): 1509-1519. doi: 10.1139/b06-120. Davidson, A.J., Harborne, J.B., and Longton, R.E. 1990. The acceptability of mosses as food for generalist herbivores, slugs in the Arionidae. Bot. J. Linn. Soc. 104(1-3): 99-113. Dawson, T.E., and Bliss, L.C. 1989. Patterns of water use and the tissue water relations in the dioecious shrub, Salix arctica: the physiological basis for habitat partitioning between the sexes. Oecologia 79(3): 332-343. doi: 10.1007/BF00384312. Dawson, T.E., and Ehleringer, J.R. 1993. Gender-specific physiology, carbon isotope discrimination, and habitat distribution in boxelder, Acer Negundo. Ecology 74(3): 798-815. doi: 10.2307/1940807. Dawson, T.E., and Geber, M.A. 1999. Sexual dimorphism in physiology and morphology. In Gender and sexual dimorphism in flowering plants. Edited by M.A. Geber and T.E. Dawson and L.F. Delph. Springer, Berlin. pp. 174-215. Decker, K.L., and Pilson, D. 2000. Biased sex ratios in the dioecious annual Croton texensis (Euphorbiaceae) are not due to environmental sex determination. Am. J. Bot. 87(2): 221-229.

169

Delph, L.F. 1999. Sexual dimorphism in life history. In Gender and sexual dimorphism in flowering plants. Edited by M.A. Geber and T.E. Dawson and L.F. Delph. Springer, Berlin. pp. 149-173. Delph, L.F., Gehring, J.L., Arntz, A.M., Levri, M., and Frey, F.M. 2005. Genetic correlations with floral display lead to sexual dimorphism in the cost of reproduction. Am. Nat. 166 (S4): S31-S41. doi: 10.1086/444597. Delph, L.F., Lu, Y., and Jayne, L.D. 1993. Patterns of resource allocation in a dioecious Carex (Cyperaceae). Am. J. Bot. 80(6): 607-615. Demmig-Adams, B. 1998. Survey of thermal energy dissipation and pigment composition in sun and shade leaves. Plant Cell Physiol. 39(5): 474-482. Dilks, T.J.K., and Proctor, M.C.F. 1979. Photosynthesis, respiration and water content in bryophytes. New Phytol. 82(1): 97–114. doi:10.1111/j.1469-8137.1979.tb07564.x Dormann, C.F., and Skarpe, C. 2002. Flowering, growth and defence in the two sexes: consequences of herbivore exclusion for Salix polaris. Funct. Ecol. 16(5): 649- 666. Doust, L.L., and Doust, J.L. 1987. Leaf demography and clonal growth in female and male Rumex acetosella. Ecology 68(6): 2056-2058. Doust, J.L., O'Brien, G., and Doust, L.L. 1987. Effect of density on secondary sex characteristics and sex ratio in Silene alba (Caryophyllaceae). Am. J. Bot. 74(1): 40-46. Dudley, L.S. 2006. Ecological correlates of scecondary sexual dimorphism in Salix glauca (Salicaceae). Am. J. Bot. 93(12): 1775-1783. doi: 10.3732/ajb.93.12.1775. Eckhart, V.M. 1999. Sexual dimorphism in flowers and inflorescences. In Gender and sexual dimorphism in flowering plants. Edited by M.A. Geber and T.E. Dawson and L.F. Delph. Springer, Berlin. pp. 123-148. Edmunds, W. M., and A. H. Bath. 1976. Centrifuge extraction and chemical analysis of interstitial waters. Environ. Sci. Technol. 10: 467–472. doi:10.1021/es60116a002. Ellstrand, N.C. and Elam, D.R. 1993. Population genetic consequences of small population size: Implications for plant conservation. Annu. Rev. Ecol. Syst. 24: 217-242. Elumeeva , T. G. , Soudzilovskaia, N.A., During, H.J., and Cornelissen, J.H.C. 2011. The importance of colony structure versus shoot morphology for the water balance of 22 subarctic bryophyte species. J. Veg. Sci. 22(1): 152 – 164. Eppley, S.M. 2006. Females make tough neighbors: sex-specific competitive effects in seedlings of a dioecious grass. Oecologia 146(4): 549-554. doi: 10.1007/s00442- 005-0026-3. Eppley, S.M., Mercer, C.A., Haaning, C., and Graves, C.B. 2009. Sex-specific variation in the interaction between Distichlis spicata (Poaceae) and mycorrhizal fungi. Am. J. Bot. 96(11): 1967-1973. doi: 10.3732/ajb.0900076. Eriksson, O. 1989. Seedling dynamics and life histories in clonal plants. Oikos 55: 231- 238. Faegri, K., and van der Piji, L. 1979. The principles of pollination of biology. Pergamon Press, New York. Fahmy, G.M. 1997. Leaf anatomy and its relation to the ecophysiology of some non- succulent desert plants from Egypt. J. Arid Environ. 36(3): 499-525.

170

Faliński, J.B. 1980. Vegetation dynamics and sex structure of the population of pioneer dioecious woody plants. Vegetatio 43: 23-38. Farrant, J., Cooper, K., Kruger, L.A., and Sherwin, H.W. 1999. The effect of drying rate on the survival of three desiccation-tolerant angiosperm species. Ann. Bot. 84(3): 371–379. doi:10.1006/anbo.1999.0927. Ferguson, M.E., Maxted, N., van Slageren, M., and Robertson, L.D. 2000. A re- assessment of the taxonomy of Lens Mill. (Leguminosae, Papilionoideae, Vicieae). Bot. J. Linn. Soc. 133(1): 41-59. Field, D.L., Pickup, M., and Barrett, S.C.H. 2013. Comparative analyses of sex-ratio variation in dioecious flowering plants. Evolution 67(3): 661-672. doi: 10.1111/evo.12001. Fox, J.F., and Harrison, A.T. 1981. Habitat assortment of sexes and water balance in a dioecious grass. Oecologia 49(2): 233-235. doi: 10.1007/bf00349194. Friedman, J., and Barrett, S.C.H. 2009. Wind of change: new insights on the ecology and evolution of pollination and mating in wind-pollinated plants. Ann. Bot. 103(9): 1515-1527. doi: 10.1093/aob/mcp035. Gangopadhyay, G., Roy, S.K., Ghose, K., Poddar, R., Bandyopadhyay, T., Basu, D., and Mukherjee, K.K. 2007. Sex determination in Carica papaya and Cycas circinalis in pre-flowering stage by ISSR and RAPD. Curr. Sci. 92(4): 524-526. García, M.B., and Antor, R.J. 1995. Sex ratio and sexual dimorphism in the dioecious Borderea pyrenaica (Dioscoreaceae). Oecologia 101: 59-67. doi: 10.1007/BF00328901. Gauquelin, T., Bertaudière-Montès, V., Badri, W., and Montès, N. 2002. Sex ratio and sexual dimorphism in mountain dioecious thuriferous juniper (Juniperus thurifera L., Cupressaceae). Bot. J. Linn. Soc. 138: 237-244. doi: 10.1046/j.1095- 8339.2002.138002237.x. Geber, M.A. 1999. Theories of the evolution of sexual dimorphism. In Gender and sexual dimorphism in flowering plants. Edited by M.A. Geber and T.E. Dawson and L.F. Delph. Springer, Berlin. pp. 97-122. Gehring, J.L., and Delph, L.F. 2006. Effects of reduced source-sink ratio on the cost of reproduction in females of Silene latifolia. Int. J. Plant Sci. (Chicago, IL, U. S.) 167(4): 843-851. doi: 10.1086/503784. Gehring, J.L., and Linhart, Y.B. 1993. Sexual dimorphisms and response to low resources in the dioecious plant Silene latifolia (Caryophyllaceae). Int. J. Plant Sci. (Chicago, IL, U. S.) 154(1): 152-162. Gehring, J.L., and Monson, R.K. 1994. Sexual differences in gas exchange and response to environmental stress in dioecious Silene latifolia (Caryophyllaceae). Am. J. Bot. 81(2): 166-174. Gehring, C.A., and Whitham, T.G. 1992. Reduced mycorrhizae on Juniperus monosperma with mistletoe: the influence of environmental stress and tree gender on a plant parasite and a plant-fungal mutualism. Oecologia 89(2): 298-303. doi: 10.1007/bf00317231. Geng, Y., Pan, X., Xu, C., Zhang, W., Li, B., Chen, J., Lu, B., and Song, Z. 2007. Phenotypic plasticity rather than locally adapted ecotypes allows the invasive alligator weed to colonize a wide range of habitats. Biol. Invasions 9(3): 245-256.

171

Giles, K.L. 1971. Dedifferentiation and regeneration in bryophytes: A selective review. N. Z. J. Bot. 9(4): 689-694. doi: 10.1080/0028825X.1971.10430231. Gilmore, A.M. and Ball, M.C. 2000. Protection and storage of chlorophyll in overwintering evergreens. Proc. Natl. Acad. Sci. USA 97(20): 11098-11101. Givnish, T.J. 1980. Ecological constraints on the evolution of breeding systems in seed plants: dioecy and dispersal in gymnosperms. Evolution 34(5): 959-972. doi: 10.2307/2408001. Glime, J.M. 1984. Theories on adaptations to high light intensity in the aquatic moss Fontinalis. J. Bryol. 13(2): 257-262. Glime, J.M. 2007a. Effects of high intensity. In Bryophyte Ecology. Volume 1. Physiological ecology. E-book sponsored by Michigan Technological University and the International Association of Bryologists [online]. Edited by J.M. Glime. Available from http://www.bryoecol.mtu.edu/ [accessed 19 January 2017]. Glime, J.M. 2007b. Temperature: heat. In Bryophyte Ecology. Volume 1. Physiological ecology. E-book sponsored by Michigan Technological University and the International Association of Bryologists [online]. Edited by J.M. Glime. Available from http://www.bryoecol.mtu.edu/ [accessed 8 April 2017]. Glime, J. M. 2013. Marchantiophyta. In Bryophyte Ecology. Volume 1. Physiological ecology. E-book sponsored by Michigan Technological University and the International Association of Bryologists. Edited by J.M. Glime. Available from http://www.bryoecol.mtu.edu/[accessed 21 December 2015]. Glime, J.M. 2015. Water relations: leaf strategies—structural. In Bryophyte Ecology. Volume 1. Physiological ecology. E-book sponsored by Michigan Technological University and the International Association of Bryologists. Edited by J.M. Glime. Available from http://www.bryoecol.mtu.edu/ [accessed 22 February 2017]. Glime, J.M., and Bisang, I. 2014. Sexuality: Size and sex differences. In Bryophyte Ecology. Volume 1. Physiological ecology. Ebook sponsored by Michigan Technological University and the International Association of Bryologists [online]. Edited by J.M. Glime. Available from http://www.bryoecol.mtu.edu/ [accessed 05 February 2015]. Grime, J.P. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. Am. Nat. 111(982): 1169-1194. Grime, J.P. 1979. Plant strategies and vegetation processes. John Wiley & Sons, New York. Grobbelaar, N., Meyer, J.J.M., and J, B. 1989. Coning and sex ratio of Encephalartos transvenosus at the Modjadji Nature Reserve. S. Afr. J. Bot. 55(1): 79-82. Groen, K.E., Stieha, C.R., Crowley, P.H., and McLetchie, D.N. 2010a. Sex-specific plant responses to light intensity and canopy openness: implications for spatial segregation of the sexes. Oecologia 162(3): 561-570. doi: 10.1007/s00442-009- 1473-z. Groen, K.E., Stieha, C.R., Crowley, P.H., and McLetchie, D.N. 2010b. Sex-specific plant responses to two light levels in the liverwort Marchantia inflexa (Marchantiaceae). Bryologist 113(1): 81-89. doi: 10.1639/0007-2745-113.1.81. Gross, K.L., and Soule, J.D. 1981. Differences in biomass allocation to reproductive and vegetative structures of male and female plants of a dioecious, perennial herb, Silene alba (Miller) Krause. Am. J. Bot. 68(6): 801-807.

172

Guerfel, M., Baccouri, O., Boujnah, D., Chaïbi, W., and Zarrouk, M. 2009. Impacts of water stress on gas exchange, water relations, chlorophyll content, and leaf structure in the two main Tunisian olive (Olea europaea L.) cultivars. Sci. Hort. 119(3): 257-263. Hamerlynck, E.P., Csintalan, Z., Nagy, Z., Tuba, Z., Goodin, D., and Henebry, G.M. 2002. Ecophysiological consequences of contrasting microenvironments on the desiccation tolerant moss Tortula ruralis. Oecologia 131(4): 498-505. Harris, M.S., and Pannell, J.R. 2008. Roots, shoots and reproduction: sexual dimorphism in size and costs of reproductive allocation in an annual herb. Proc. R. Soc. B 275(1651): 2595-2602. doi: 10.1098/rspb.2008.0585. Harris, M.S., and Pannell, J.R. 2010. Canopy seed storage is associated with sexual dimorphism in the woody dioecious genus Leucadendron. J. Ecol. 98(2): 509-515. doi: 10.1111/j.1365-2745.2009.01623.x. Harvey, S.J., Nowierski, R.M., Mahlberg, P.G., and Story, J.M. 1988. Taxonomic evaluation of leaf and latex variability of leafy spurge (Euphorbia spp.) for Montana and European accessions. Weed Sci. 36(6): 726-733. Hassel, K., Pederson, B., and Söderström, L. 2005. Changes in life-history traits in an expanding moss species: phenotypic plasticity or genetic differentiation? A reciprocal transplantation experiment with Pogonatum dentatum. Ecography 28(1): 71-80. Head, G. 1995. Selection on fecundity and variation in the degree of sexual size dimorphism among spider species (Class Araneae). Evolution 49(4): 776-781. Hedenäs, L., and Bisang, I. 2011. The overlooked dwarf males in mosses—unique among green land plants. Perspect. Plant Ecol. Evol. Syst. 113: 121-135. doi: 10.1016/j.ppees.2011.03.001. Hemborg, Å., and Karlsson, P.S. 1999. Sexual differences in biomass and nutrient allocation of first-year Silene dioica plants. Oecologia 118(4): 453-460. doi: 10.1007/s004420050748. Henschel, J.R., and Seely, M.K. 2000. Long-term growth patterns of Welwitschia mirabilis, a long-lived plant of the Namib Desert (including a bibliography). Plant Ecol. 150: 7-26. doi: 10.1023/a:1026512608982. Hesse, E., and Pannell, J.R. 2011. Sexual dimorphism in a dioecious population of the wind-pollinated herb Mercurialis annua: the interactive effects of resource availability and competition. Ann. Bot. 107(6): 1039-1045. doi: 10.1093/aob/mcr046. Hickey, L.J. 1973. Classification of the architecture of dicotyledonous leaves. Am. J. Bot. 60(1): 17-33. Hilfiker, K., Gugerli, F., Schütz, J.-P., Rotach, P., and Holderegger, R. 2004. Low RAPD variation and female-biased sex ratio indicate genetic drift in small populations of the dioecious conifer Taxus Baccata in Switzerland. Conserv. Genet. 5(3): 357- 365. doi: 10.1023/B:COGE.0000031144.95293.1b. Hill, P.W., Handley, L.L., and Raven, J.A. 1996. Juniperus communis L. ssp. communis at Balnaguard, Scotland: Foliar carbon discrimination (δ13C) and 15-N natural abundance (δ15N) suggest gender-linked differences in water and N use. Bot. J. of Scotl. 48(2): 209-224. doi: 10.1080/03746609608684842.

173

Hogan, K.P., García, M.B., Cheeseman, J.M., and Loveless, M.D. 1998. Inflorescence photosynthesis and investment in reproduction in the dioecious species Aciphylla glaucescens (Apiaceae). N. Z. J. Bot. 36(4): 653-660. doi: 10.1080/0028825x.1998.9512602. Holá, E., Vesalainen, T., Těšitel, J., and Laaka-Lindberg, S. 2014. Sex ratio, sex-specific pattern in vegetative growth and gemma production in an aquatic liverwort, Scapania undulata (Marchantiophyta: Scapaniaceae). Bot. J. Linn. Soc. 175: 229- 241. doi: 10.1111/boj.12159. Hoof, J., Sack, L., Webb, D.T., and Nilsen, E.T. Contrasting structure and function of pubescent and glabrous variaties of Hawaiian Metrosideros polymorpha (Myrtaceae) at high elevation. Biotropica 40(1): 113-118. Horsley, K., Stark, L.R., and McLetchie, D.N. 2011. Does the silver moss Bryum argenteum exhibit sex-specific patterns in vegetative growth rate, asexual fitness or prezygotic reproductive investment? Ann. Bot. 107(6): 897-907. doi: 10.1093/aob/mcr027. Houssard, C., Escarré, J., and Vartanian, N. 1992. Water stress effects on successional populations of the dioecious herb, Rumex acetosella L. New Phytol. 120(4): 551- 559. doi: 10.1111/j.1469-8137.1992.tb01805.x. Hufbauer, R.A., Facon, B., Ravigné, V., Turgeon, J., Foucaud, J., Lee, C.E., Rey, O., and Estoup, A. 2012. Anthropogenically induced adaptation to invade (AIAI): contemporary adaptation to human-altered habitats within the native range can promote invasions. Evol. Appl. 5(1): 89-101. Iszkuło, G., and Boratyński, A. 2011. Initial period of sexual maturity determines the greater growth rate of male over female in the dioecious tree Juniperus communis subsp. communis. Acta Oecol. 37(2): 99-102. doi: 10.1016/j.actao.2011.01.001. Iszkuło, G., Jasińska, A.K., Giertych, M.J., and Boratyński, A. 2009. Do secondary sexual dimorphism and female intolerance to drought influence the sex ratio and extinction risk of Taxus baccata? Plant Ecol. 200(2): 229-240. doi: 10.1007/s11258-008-9447-5. Jesson, L.K., and Garnock-Jones, P.J. 2012. Can classifications of functional gender be extended to all land plants? Perspect. Plant Ecol. Evol. Syst.14: 153-160. doi: 10.1016/j.ppees.2011.10.003. Jordan, N. 1992. Path analysis of local adaptation in two ecotypes of the annual plant Diodia teres Walt. (Rubiaceae). Am. Nat. 140(1): 149-165. Juvany, M., and Munne-Bosch, S. 2015. Sex-related differences in stress tolerance in dioecious plants: a critical appraisal in a physiological context. J. Exp. Bot. 66(20): 6083-6092. doi: 10.1093/jxb/erv343. Kang, H., and Shin, S. 2012. Sex ratios and spatial structure of the dioecious tree Torreya nucifera in Jeju Island, Korea. J. Ecol. Field Biol. 35(2): 111-122. doi: 10.5141/JEFB.2012.015. Kerkhoff, A.J., Fagan, W.F., Elser, J.J., and Enquist, B.J. 2006. Phylogenetic and growth form variation in the scaling of nitrogen and phosphorus in the seed plants. Am. Nat. 168(4): E103-E122. doi: 10.1086/507879. Kern, S.O., Hovenden, M.J., and Jordan, G.J. 2004. The impacts of leaf shape and arrangement on light interception and potential photosynthesis in southern beech (Nothofagus cunninghamii). Funct. Plant Biol. 31(5): 471-480.

174

Kershaw, K.A. and Weber. 1986. Seasonal changes in the chlorophyll content and quantum efficiency of the moss Brachythecium rutabulum. J. Bryol. 14(1): 151- 158. Kimura M. and Crow J.F. 1963. The measurement of effective population number. Evolution 17(3): 279-288. Kohorn, L.U. 1994. Shoot morphology and reproduction in jojoba: advantages of sexual dimorphism. Ecology 75(8): 2384-2394. Korpelainen, H. 1992. Patterns of resource allocation in male and female plants of Rumex acetosa and R. acetosella. Oecologia 89(1): 133-139. doi: 10.1007/BF00319025. Krupa, J. 1969. Photosynthetic activity and productivity of the sporophyte of Funaria hygrometrica during ontogenesis. Acta Soc. Bot. Poloniae 38: 207-215. Kumar, N., Gupta, S., and Tripathi, A.N. 2006. Gender-specific responses of Piper betle L. to low temperature stress: changes in chlorophyllase activity. Biol. Plant. 50(4): 705-708. doi: 10.1007/s10535-006-0111-4. Kupfer, A. 2009. Sexual size dimorphism in caecilian amphibians: analysis, review, and directions for future research. Zoology 112(5): 362-369. doi: 10.1016/j.zool.2008.12.001. Lawton, R.O., and Cothran, P. 2000. Factors influencing reproductive activity of Juniperus virginiana in the Tennessee Valley. J. Torrey Bot. Soc. 127(4): 271- 279. doi: 10.2307/3088645. Lee, C.L. 1954. Sex chromosomes in Ginkgo biloba. Am. J. Bot. 41(7): 545-549. Leslie, A.B., Beaulieu, J.M., Crane, P.R., and Donoghue, M.J. 2013. Explaining the distribution of breeding and dispersal syndromes in conifers. Proc. R. Soc. B 280(1770): 20131812. doi: 10.1098/rspb.2013.1812. Letts, M.G., Phelan, C.A., Johnson, D.R.E., and Rood, S.B. 2008. Seasonal photosynthetic gas exchange and leaf reflectance characteristics of male and female cottonwoods in a riparian woodland. Tree Physiol. 28(7): 1037-1048. doi: 10.1093/treephys/28.7.1037. Levitan, D.R. 2010. Sexual selection in external fertilizers. Edited by D. Westneat and C. Fox. Oxford University Press, New York. pp. 343-364. Li, C., Xu, G., Zang, R., Korpelainen, H., and Berninger, F. 2007. Sex-related differences in leaf morphological and physiological responses in Hippophae rhamnoides along an altitudinal gradient. Tree Physiol. 27(3): 399-406. doi: 10.1093/treephys/27.3.399. Liakopoulos, G., Nikolopoulos, D., Klouvatou, A., Vekkos, K.-A., Manetas, Y., and Karbourniotis G. 2006. The photoprotective role of epidermal anthocyanins and surface pubescences in young leaves of grape vine (Vitis vinifera). Ann. Bot. 98(1): 257-265. Liu, J., and Duan, J. 2013. Humulus scandens gender differences in response to water stress in the vegetative growth stage. Acta Prataculturae Sinica 22(2): 243-249. Linhart, Y.B. and Grant, M.C. 1996. Evolutionary significance of local genetic differentiation in plants. Annu. Rev. Ecol. Syst. 27: 237-277. Lloyd, D.G., and Webb, C.J. 1977. Secondary sex characters in plants. Bot. Rev. 43(2): 177-216. doi: 10.1007/bf02860717. Longton, R.E. 1981. Interpopulation variation and in morphology and physiology in the cosmopolitan moss Bryum argenteum Hedw. J. Bryol. 11(3): 501-520.

175

Longton, R.E., and Greene, S.W. 1979. Experimental studies of growth and reproduction in the moss Pleurozium schreberi (Brid.) Mitt. J. Bryol. 10: 321-338. Mägdefrau, K. 1982. Life-forms of bryophytes. In Bryophyte ecology. Edited by A.J.E. Smith. Chapman and Hall, London. pp. 45–58. Mankiewicz, P.S. 1987. The low pressure field porometer: A new, low cost technique for characterizing external capillary water conduction in whole colonies of bryophytes and other small plants. Bryologist 90(3): 253–262. doi:10.2307/3242936. Marion, C., and Houle, G. 1996. No differential consequences of reproduction according to sex in Juniperus communis var. depressa (Cupressaceae). Am. J. Bot. 83(4): 480-488. Marks, R., Burton, J., McLetchie, D. N. 2016. Sex differences and plasticity in dehydration tolerance: insight from a tropical liverwort. Ann. Bot. 118(2): 347- 356. Marler, T.E. 2010. Cycad mutualist offers more than pollen transport. Am. J. Bot. 97(5): 841-845. doi: 10.3732/ajb.0900251. Martin, R.E., Asner, G.P., and Sack, L. 2007. Genetic variation in leaf pigment, optical and photosynthetic function among diverse genotypes of Metrosidros polymorpha grown in a common garden. Oecologia 151(3): 387-400. Martínez-Abaigar, J., Núñez-Olivera, E, and Sánchez-Díaz, M. 1994. Seasonal changes in photosynthetic pigment composition of aquatic bryophytes. J. Bryol. 18(1): 97- 113. Maxwell, K. and Johnson, G.N. 2000. Chlorophyll fluorescence—a practical guide. J. Exp. Bot. 51(345): 659-668. McDaniel, S.F. 2005. Genetic correlations do not constrain the evolution of sexual dimorphism in the moss Ceratodon Purpureus. Evolution 59(11): 2353-2361. doi: 10.1111/j.0014-3820.2005.tb00945.x. McDaniel, S.F., Willis, J.H., and Shaw, A.J. 2007. A linkage map reveals a complex basis for segregation distortion in an interpopulation cross in the moss Ceratodon purpureus. Genetics 176(4): 2489-2500. doi: 10.1534/genetics.107.075424. McDaniel, S.F., Willis, J.H., and Shaw, A.J. 2008. The genetic basis of developmental abnormalities in interpopulation hybrids of the moss Ceratodon purpureus. Genetics 179(3): 1425-1435. doi: 10.1534/genetics.107.086314. McDowell, S.C., McDowell, N.G., Marshall, J.D., and Hultine, K. 2000. Carbon and nitrogen allocation to male and female reproduction in Rocky Mountain Douglas- fir (Pseudotsuga menziesii var. glauca, Pinaceae). Am. J. Bot. 87(4): 539-546. McLetchie, D.N. 1992. Sex ratio from germination through maturity and its reproductive consequences in the liverwort Sphaerocarpos texanus. Oecologia 92(2): 273-278. doi: 10.1007/BF00317375. McLetchie, D. 1996. Sperm limitation and genetic effects on fecundity in the dioecious liverwort Sphaerocarpos texanus. Sex. Plant Reprod. 9(2): 87–92. doi:10.1007/BF02153055. McLetchie, D.N. 2001. Sex-specific germination response in the liverwort Sphaerocarpos texanus (Sphaerocarpaceae). Bryologist 104(1): 69-71. doi: 10.1639/0007-2745(2001)104[0069:SSGRIT]2.0.CO;2.

176

McLetchie, D.N., García-Ramos, G., and Crowley, P.H. 2002. Local sex-ratio dynamics: a model for the dioecious liverwort Marchantia inflexa. Evol. Ecol. 15(4): 231- 254. doi: 10.1023/a:1016000613291. McLetchie, D.N., and Puterbaugh, M.N. 2000. Population sex ratios, sex-specific clonal traits and tradeoffs among these traits in the liverwort Marchantia inflexa. Oikos 90(2): 227-237. doi: 10.1034/j.1600-0706.2000.900203.x. Meagher, T.R. 1988. Sex determination in plants. Edited by J. Lovett Doust and L. Lovett Doust. Oxford University Press, New York. pp. 125-138. Meagher, T.R., and Antonovics, J. 1982. The population biology of Chamaelirium luteum, a dioecious member of the lily family: life history studies. Ecology 63(6): 1690-1700. Mercer, C.A., and Eppley, S.M. 2010. Inter-sexual competition in a dioecious grass. Oecologia 164(3): 657-664. doi: 10.1007/s00442-010-1675-4. Michel, P., Payton, I.J., Lee, W.G., and During, H. J. 2013. Impact of disturbance on above-ground water storage capacity of bryophytes in New Zealand indigenous tussock grassland ecosystems. New Zeal. J. Ecol. 37(1): 114–126. Miller, P.M., Eddleman, L.E., and Miller, J.M. 1995. Juniperus occidentalis juvenile foliage: advantages and disadvantages for a stress-tolerant, invasive conifer. Can. J. For. Res. 25(3): 470-479. Mishler, B.D. 2007. Syntrichia. In Flora of North America North of Mexico, vol. 27. Edited by Flora of North America Editorial Committee [eds.], 1993+. Oxford University Press, New York. pp. 618-626. Montesinos, D., de Luís, M., Verdú, M., Raventós, J., and García-Fayos, P. 2006. When, how and how much: Gender-specific resource-use strategies in the dioecious tree Juniperus thurifera. Ann. Bot. 98(4): 885-889. doi: 10.1093/aob/mcl172. Moore, J.D., Kollar, L.M., and McLetchie, D.N. 2016. Does selection for gamete dispersal and capture lead to a sex difference in clump water-holding capacity? Am. J. Bot. 103(8): 1-9. Moore, J.C., and Pannell, J.R. 2011. Sexual selection in plants. Curr. Biol. 21(5): R176- 182. doi: 10.1016/j.cub.2010.12.035. Mori, M., Suzuki, K., and Kohzaki, R. 2000. Variations in chlorophyll and carotenoid content in the growth process of the ginkgo leaf. J. Jpn. Soc. Food Sci. Technol. 47(6): 448-451. [in Japanese with English abstract] Muggoch, H., and Walton, J. 1942. On the dehiscence of the antheridium and the part played by surface tension in the dispersal of spermatocytes in Bryophyta. Proc. R. Soc. B 130(861): 448–461. doi:10.1098/rspb.1942.0012. Munné-Bosch, S. 2015. Sex ratios in dioecious plants in the framework of global change. Environ. Exp. Bot. 109: 99-102. doi: 10.1016/j.envexpbot.2014.08.007. Murray, K.J., Tenhunen, J.D., and Nowak, R.S. 1993. Photoinhibition as a control on photosynthesis and production if Sphagnum mosses. Oecologia 96(2): 200-207. Myers, R.A. 1984. Demographic consequences of precocious maturation of Atlantic Salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 41(9): 1350-1353. doi: 10.1139/f84-165. Myers-Smith, I.H., and Hik, D.S. 2012. Uniform female-biased sex ratios in alpine willows. Am. J. Bot. 99(7): 1243-1248. doi: 10.3732/ajb.1200107.

177

Naidu, S.L and DeLucia, E.H. 1997. Acclimation of shade-developed leaves on saplings exposed to late-season canopy gaps. Tree Physiol. 17: 367-376. Nanami, S., Kawaguchi, H., and Yamakura, T. 2005. Sex ratio and gender-dependent neighboring effects in Podocarpus nagi, a dioecious tree. Plant Ecol. 177(2): 209- 222. doi: 10.1007/s11258-005-2210-2. Newell, S.J. 1985. Intrapopulational variation in leaflet morphology of Zamia pumila L. in relation to microenvironment and sex. Am. J. Bot. 72(2): 217-221. Newton, M.E. 1972. Sex-ratio differences in Mnium hornum Hedw. and M. undulatum Sw. in relation to spore germination and vegetative regeneration. Ann. Bot. 36: 163-178. Nicotra, A.B. 1998. Sex ratio variation and spatial distribution of Siparuna grandiflora, a tropical dioecious shrub. Oecologia 115: 102-113. doi: 10.1007/s004420050496. Nicotra, A.B. 1999. Reproductive allocation and the long-term costs of reproduction in Siparuna grandiflora, a dioecious neo-tropical shrub. J. Ecol. 87(1): 138-149. doi: 10.1046/j.1365-2745.1999.00337.x. Nicotra, A.B., Chazdon, R.L., and Montgomery, R.A. 2003. Sexes show contrasting patterns of leaf and crown carbon gain in a dioecious rainforest shrub. Am. J. Bot. 90(3): 347-355. doi: 10.3732/ajb.90.3.347. Niklas, K.J. 1985a. The aerodynamics of wind pollination. Bot. Rev. 51(3): 328-386. doi: 10.1007/bf02861079. Niklas, K.J. 1985b. Wind pollination—A study in controlled chaos. Am. Sci. 73(5): 462– 470. Niklas, K.J., and Buchmann, S.L. 1985. Aerodynamics of wind pollination in Simmondsia chinensis (Link) Schneider. Am. J. Bot. 72(4): 530–539. doi:10.2307/2443585. Nobel, P.S. 1983. Biophysical plant physiology and ecology. W. H. Freeman and Company, San Francisco. Núñez-Olivera, E., Arróniz-Crespo, M., Martínez-Abaigar, J., Tomás, R., and Beaucort, N. 2005. Assessing the UV-B tolerance of sun and shade samples of two aquatic bryophytes using short-term tests. Bryologist 108(3): 435-448. Nuñez, C.I., Nuñez, M.A., and Kitzberger, T. 2008. Sex-related spatial segregation and growth in a dioecious conifer along environmental gradients in northwestern Patagonia. Écoscience15(1): 73-80. doi: 10.2980/1195- 6860(2008)15[73:sssagi]2.0.co;2. Obeso, J.R. 2002. The costs of reproduction in plants. New Phytol. 155(3): 321-348. doi: 10.1046/j.1469-8137.2002.00477.x. Obeso, J.R., Alvarez-Santullano, M., and Retuerto, R. 1998. Sex ratios, size distributions, and sexual dimorphism in the dioecious tree Ilex aquifolium (Aquifoliaceae). Am. J. Bot. 85(11): 1602-1608. Oguchi, R., Hikosaka, K., and Hirose, T. 2003. Does the photosynthetic light-acclimation need change in leaf anatomy? Plant Cell Environ. 26(4): 505-512. Oliver, M.J., Velten, J.. and Wood, A.J. 2000. Bryophytes as experimental models for the study of environmental stress tolerance: Tortula ruralis and desiccation-tolerance in mosses. Plant Ecol. 151(1): 73–84. doi:10.1023/A:1026598724487. Ornduff, R. 1985. Male-biased sex ratios in the cycad Macrozamia riedlei (Zamiaceae). Bull. Torrey Bot. Club 112(4): 393-397. doi: 10.2307/2996040.

178

Ornduff, R. 1987. Sex ratios and coning frequency of the cycad Zamia pumila L.(Zamiaceae) in the Dominican Republic. Biotropica 19(4): 361-364. Ornduff, R. 1990. Geographic variation in reproductive behavior and size structure of the Australian cycad Macrozamia communis (Zamiaceae). Am. J. Bot. 77(1): 92-99. Ornduff, R. 1991. Coning phenology of the cycad Macrozamia riedlei (Zamiaceae) over a five-year interval. Bull. Torrey Bot. Club 118(1): 6-11. Ornduff, R. 1996. Gender performance in a cultivated cohort of the cycad Zamia integrifolia (Zamiaceae). Am. J. Bot. 83(8): 1006-1015. Ortiz, P.L., Arista, M., and Talavera, S. 1998. Low reproductive success in two subspecies of Juniperus oxycedrus L. Int. J. Plant Sci. (Chicago, IL, U. S.) 159(5): 843-847. Ortiz, P.L., Arista, M., and Talavera, S. 2002. Sex ratio and reproductive effort in the dioecious Juniperus communis subsp. alpina (Suter) Celak. (Cupressaceae) along an altitudinal gradient. Ann. Bot. 89(2): 205-211. doi: 10.1093/aob/mcf028. Pannell, J. 1997. Mixed genetic and environmental sex determination in an androdioecious population of Mercurialis annua. Heredity 78: 50-56. doi: 10.1038/hdy.1997.6. Paolillo, D.J. 1977. On the release of sperms in Atrichum. Am. J. Bot. 64(1): 81–85. doi:10.2307/2441879. Pearcy, R.W., Muraoka, H., and Valladares, F. 2005. Crown architecture in sun and shade environments: Assessing function and trade-offs with a three-dimensional simulation model. New Phyotol. 166(3): 791–800. doi:10.1111/j.1469- 8137.2005.01328.x Pettet, H.P. and Berrie, G.K. 1967. Sex ratios of Riccia frostii Aust. in Khartoum Province, Sudan. Trans. Brit. Bryol. Soc. 5(2): 332-337. Pickering, C.M. 2000. Sex-specific differences in floral display and resource allocation in Australian alpine dioecious Aciphylla glacialis (Apiaceae). Aust. J. Bot. 48(1): 81-91. doi: 10.1071/BT97121. Pisa, S., Biersma, E.M., Convey, P., Patiño, J., Vanderpoorten, A., Werner, O., and Ros, R.M. 2014. The cosmopolitan moss Bryum argenteum in Antarctica: Recent colonisation or in situ survival? Polar Biology 37(10): 1469–1477. doi:10.1007/s00300-014-1537-3. Post, A. 1990. Photoproctective pigment as an adaptive strategy in Antarctic moss Ceratodon purpureus. Polar Biol. 10: 241-245. Potters, G., Pasternak, T.P., Guisez, Y., Palme, K., and Jansen, M. A. K. 2007. Stress- induced morphogenic responses: Growing out of trouble? Trends Plant Sci. 12(3): 98-105. Pressel, S. and Duckett, J.G. 2010. Cytological insights into the desiccation biology of a model system: moss protonemata. New Phytol. 185(4): 944-963. 14 Proctor, M.C.F. 1977. Evidence on the carbon nutrition of moss sporophytes from CO2 uptake and the subsequent movement of labelled assimilate. J. Bryol. 9: 375-386. doi: 10.1179/jbr.1977.9.3.375. Proctor, M. C. F. 1982. Physiological ecology: Water relations, light and temperature responses, carbon balance. In Bryophyte Ecology. Edited by A.J.E. Smith. Chapman and Hall, London. pp. 333-381.

179

Proctor, M.C.F., Oliver, M.J., Wood, A.J., Alpert, P., Stark, L.R., Cleavitt, N.L., and Mishler, B.D. 2007. Desiccation-tolerance in bryophytes: a review. Bryologist 110: 595-621. Putwain, P.D., and Harper, J.L. 1972. Studies in the dynamics of plant populations: V. Mechanisms governing the sex ratio in Rumex acetosa and R. acetosella. J. Ecol. 60(1): 113-129. Quinn, J.A., and Meiners, S.J. 2004. Growth rates, survivorship, and sex ratios of Juniperus virginiana on the New Jersey Piedmont from 1963 to 2000. J. Torrey Bot. Soc. 131(3): 187-194. doi: 10.2307/4126949. Ramsay, H.P. 1979. Anisospory and sexual dimorphism in the Musci. In Bryophyte Systematics Academic Press, London, UK. pp. 281-316. Ramsay, H.P., and Berrie, G.K. 1982. Sex determination in bryophytes. J. Hattori Bot. Lab. 52: 255-274. Randriamanana, T.R., Nissinen, K., Moilanen, J., Nybakken, L., and Julkunen-Tiitto, R. 2015. Long-term UV-B and temperature enhancements suggest that females of Salix myrsinifolia plants are more tolerant to UV-V than males. Environ. Exp. Bot. 109: 296-305. doi: 10.1016/j.envexpbot.2014.06.007. Randriamanana, T.R., Nybakken, L., Lavola, A., Aphalo, P.J., Nissinen, K., and Julkunen-Tiitto, R. 2014. Sex-related differences in growth and carbon allocation to defence in Populus tremula as explained by current plant defence theories. Tree Physiol. 34(5): 471-487. doi: 10.1093/treephys/tpu034. Reeve, J.P., and Fairbairn, D.J. 1999. Change in seuxal size dimorphism as a correlated response to selection on fecundity. Heredity 83: 697-706. doi: doi:10.1046/j.1365- 2540.1999.00616.x. Renzaglia, K.S., and McFarland, K.D. 1999. Antheridial plants of Megaceros aenigmaticus in the southern Appalachians: anatomy, ultrastructure and population distribution. Haussknechtia Beiheft 9: 307-316. Reynolds, D.N. 1980. Gamete dispersal in Mnium ciliare. Bryologist 83(1): 73-77. Reynolds, L.A. and Mcletchie, D.N. 2011. Short distances between extreme microhabitats do not result in ecotypes in Syntrichia caninervis. J. Bryol. 33(2): 148-153. Rice, S.K., Collins, D., and Anderson, A.M. 2001. Functional significance of variation in canopy structure. Am. J. Bot. 88(9): 1568–1576. doi:10.2307/3558400. Rice, S. K., and Schneider, N. 2004. Cushion size, surface roughness, and the control of water balance and carbon flux in the cushion moss Leucobryum glaucum (Leucobryaceae). Am. J. Bot. 91(8): 1164–1172. doi:10.3732/ajb.91.8.1164. Rincón, E. 1993. Growth responses of six bryophyte species to different light intensities. Can. J. Bot. 71(5): 661-665. Romero, C., Putz, F.E., and Kitajima, K. 2006. Ecophysiology in relation to exposure of pendant epiphytic bryophytes in the canopy of a tropical montane oak forest. Biotropica 38(1): 35-41. Roemer, R., Terry, I., Chockley, C., and Jacobsen, J. 2005. Experimental evaluation and thermo-physical analysis of thermogenesis in male and female cycad cones. Oecologia 144(1): 88-97. doi: 10.1007/s00442-005-0045-0.

180

Rosenstiel, T.N., Shortlidge, E.E., Melnychenko, A.N., Pankow, J.F., and Eppley, S.M. 2012. Sex-specific volatile compounds influence microarthropod-mediated fertilization of moss. Nature 489(7416): 431-433. doi: 10.1038/nature11330. Rovere, A.E., Aizen, M.A., and Kitzberger, T. 2003. Growth and climatic response of male and female trees of Austrocedrus chilensis, a dioecious conifer from the temperate forests of southern South America. Écoscience 10(2): 195-203. Rozas, V., DeSoto, L., and Olano, J.M. 2009. Sex-specific, age-dependent sensitivity of tree-ring growth to climate in the dioecious tree Juniperus thurifera. New Phytol. 182(3): 687-697. doi: 10.1111/j.1469-8137.2009.02770.x. Rozendaal, D.M.A., Hurtado, V.H., and Poorter, L. 2006. Plasticity in leaf traits of 38 tropical tree species in response to light; relationships with light demand and adult stature. Funct. Ecol. 20(2): 207-216. Ruhfel, B.R., Gitzendanner, M.A., Soltis, P.S., Soltis, D.E., and Burleigh, J.G. 2014. From algae to angiosperms-inferring the phylogeny of green plants (Viridiplantae) from 360 plastid genomes. BMC Evol. Biol. 14: 23. doi: 10.1186/1471-2148-14-23. Russell, J., and Bulman, S. 2005. The liverwort Marchantia foliacea forms a specialized symbiosis with arbuscular mycorrhizal fungi in the genus Glomus. New Phytol. 165(2): 567-579. doi: 10.1111/j.1469-8137.2004.01251.x. Sakai, A., Sasa, A., and Sakai, S. 2006. Do sexual dimorphisms in reproductive biomass increase with an increase in altitude? A case of the shrub willow Salix reinii (Salicaceae). Am. J. Bot. 93(7): 988-992. doi: 10.3732/ajb.93.7.988. Sakai, A.K., and Weller, S.G. 1999. A review of terminology, biogeographic patterns, ecological correlates, and phylogenetic approaches. In Gender and sexual dimorphism in flowering plants. Edited by M.A. Geber and T.E. Dawson and L.F. Delph. Springer, Berlin. pp. 1-31. Sánchez Vilas, J., and Pannell, J.R. 2010. Differential niche modification by males and females of a dioecious herb: extending the Jack Sprat effect. J. Evol. Biol. 23(10): 2262-2266. doi: 10.1111/j.1420-9101.2010.02089.x. Sánchez Vilas, J., and Pannell, J.R. 2011. Sexual dimorphism in resource acquisition and deployment: both size and timing matter. Ann. Bot. 107(1): 119-126. doi: 10.1093/aob/mcq209. Sarijeva, G., Knapp, M., and Lichtenthaler, H.K. 2007. Differences in photosynthetic activity, chlorophyll and carotenoid levels, and chlorophyll fluorescence parameters in green sun and shade leaves of Ginkgo and Fagus. J. Plant Physiol. 164: 950-955. Såstad, S.M., Pedersen, B., and Digre, K. 1999. Habitat-specific genetic effects on growth rate and morphology across pH and water-level gradients within a population of the moss Sphagnum angustifolium (Sphagnaceae). Am. J. Bot. 86(12): 1687-1698. Schofield, W.B. 1982. Ecological significance of morphological characters in the moss gametophyte. Bryologist 84(2): 149-165. Schroeter, B., Green, T.G.A., Kulle, D., Pannewitz, S., Schlensog, M., and Sancho, L.G. 2012. The moss Bryum argenteum var. muticum Brid. is well adapted to cope with high light in continental Antarctica. Antarctic Sci. 24(3): 281-291.

181

Schüßler, A. 2000. Glomus claroideum forms an arbuscular mycorrhiza-like symbiosis with the hornwort Anthoceros punctatus. Mycorrhiza 10(1): 15-21. doi: 10.1007/s005720050282. Scott, G.A.M. 1982. Desert bryophytes. In Bryophyte ecology. Edited by A.J.E. Smith. Chapman and Hall, London. pp 105-122. Searcy, W.A. 1979. Sexual selection and body size in male Red-Winged Blackbirds. Evolution 33(2): 649-661. Selkirk, P.M., Skotnicki, M.L., Ninham, J., Connett, M. B., and Armstrong, J.. 1998. Genetic variation and dispersal of Bryum argenteum and Hennediella heimii populations in the Garwood Valley, southern Victoria Land, Antarctica. Antarct. Sci. 10(4): 423–430. doi:10.1017/S0954102098000510. Sellier, D., and Fourcaud, T. 2009. Crown structure and wood properties: Influence on tree sway and response to high winds. Am. J. Bot. 96(5): 885–896. doi:10.3732/ajb.0800226. Shaw, J. 1984. Quantitative taxonomic study of morphology in Epipterygium. Bryologist 87(2): 132-142. Shaw, A.J. and Albright, D.L. 1990. Potential for the evolution of heavy metal tolerance in Bryum argenteum, a moss II. Generalized tolerances among diverse populations. Bryologist 93(2): 187-192. Shaw, J., and Beer, S.C. 1999. Life history variation in gametophyte populations of the moss Ceratodon purpureus (Ditrichaceae) Am. J. Bot. 86(4): 512-521. Shaw, J., Beer, S.C., and Lutz, J. 1989. Potential for the evolution of heavy metal tolerance in Bryum argenteum, a moss I. Variation within and among populations. Bryologist 92(1): 73-80. Shaw, A.J., and Gaughan, J.F. 1993. Control of sex ratios in haploid populations of the moss, Ceratodon purpureus. Am. J. Bot. 80(5): 584-591. Shephard, H.L., Parker, J.S., Darby, P., and Ainsworth, C.C. 2000. Sexual development and sex chromosomes in hop. New Phytol. 148(3): 397–411. doi:10.1046/j.1469- 8137.2000.00771.x. Shine, R. 1979. Sexual selection and sexual dimorphism in the Amphibia. Copeia 1979(2): 297-306. Shine, R. 1988. The evolution of large body size in females: A critique of Darwin's "Fecundity Advantage" model. Am. Nat. 131(1): 124-131. Shine, R. 1989. Ecological causes for the evolution of sexual dimorphism: a review of the evidence. The Q. Rev. Biol. 64(4): 419-461. Sinclair, J.P., Emlen, J., and Freeman, D.C. 2012. Biased Sex Ratios in Plants: Theory and Trends. Bot. Rev. 78(1): 63-86. doi: 10.1007/s12229-011-9065-0. Sinclair, J., Emlen, J.M., and Rinella, M. 2009. Differential phytosociological interactions involving male and female Atriplex bonnevillensis. West. North Am. Nat. 69(4): 475-480. doi: 10.3398/064.069.0407. Skotnicki, M.L., Ninham, J.A., and Selkirk, P.M. 1998. High levels of RAPD diversity in the moss Bryum argenteum in Australia, New Zealand, and Antarctica. Bryologist 101(3): 412–421. doi:10.1639/0007-2745(1998)101[412:HLORDI]2.0.CO;2. Slack, N.G. 2011. The ecological value of bryophytes as indicators of climate change. Edited by Z. Tuba and N.G. Slack and L.R. Stark. Cambridge University Press, New York. pp. 3-12.

182

Sokal, R.R. and Rohlf, F.J. 1994. Biometry, 3rd ed. W.H. Freeman and Company, New York. pp. 715-724. Sousa, V.A., Robinson, I.P., and Hattemer, H.H. 2004. Variation and population structure at enzyme gene loci in Araucaria angustifolia (Bert.) O. Ktze. Silvae Genet. 3(1): 12-19. Stark, L.R. 2002. Phenology and its repercussions on the reproductive ecology of mosses. Bryologist 105(2): 204-218. doi: 10.1639/0007- 2745(2002)105[0204:PAIROT]2.0.CO;2. Stark, L.R., Greenwood, J.L., Brinda, J.C., and Oliver, M.J. 2013. The desert moss Pterygoneurum lamellatum (Pottiaceae) exhibits an inducible ecological strategy of desiccation tolerance: effects of rate of drying on shoot damage and regeneration. American Journal of Botany 100(8): 1522-1531. Stark, L.R., and McLetchie, D.N. 2006. Gender-specific heat-shock tolerance of hydrated leaves in the desert moss Syntrichia caninervis. Physiol. Plant. 126: 187-195. Stark, L.R., McLetchie, D.N., and Eppley, S.M. 2010. Sex ratios and the shy male hypothesis in the moss Bryum argenteum (Bryaceae). Bryologist 113(4): 788-797. doi: 10.1639/0007-2745-113.4.788. Stark, L., McLetchie, N., and Mishler, B. 2001. Sex expression and sex dimorphism in sporophytic populations of the desert moss Syntrichia caninervis. Plant Ecol. 157(2): 183-196. doi: 10.1023/a:1013968421254. Stark, L.R., McLetchie, D.N., and Mishler, B.D. 2005a. Sex expression, plant size, and spatial segregation of the sexes across a stress gradient in the desert moss Syntrichia caninervis. Bryologist 108(2): 183-193. doi: 10.1639/0007- 2745(2005)108[0183:SEPSAS]2.0.CO;2. Stark, L.R., McLetchie, D.N., and Roberts, S.P. 2009. Gender differences and a new adult eukaryotic record for upper thermal tolerance in the desert moss Syntrichia caninervis. J. Therm. Biol. 34(3): 131-137. doi: 10.1016/j.jtherbio.2008.12.001. Stark, L.R., Mishler, B.D., and McLetchie, D.N. 2000. The cost of realized sexual reproduction: assessing patterns of reproductive allocation and sporophyte abortion in a desert moss. Am. J. Bot. 87(11): 1599-1608. doi:10.2307/2656736. Stark, L.R., Nichols, L., McLetchie, D.N., and Bonine, M.L. 2005b. Do the sexes of the desert moss Syntrichia caninervis differ in desiccation tolerance? A leaf regeneration assay. Int. J. Plant Sci. (Chicago, IL, U. S.) 166(1): 21-29. doi: 10.1086/425671. Stark, L.R., Nichols II, L., McLetchie, D.N., Smith, S.D., and Zundel, C. 2004. Age and sex-specific rates of leaf regeneration in the Mojave Desert moss Syntrichia caninervis. Am. J. Bot. 91(1): 1-9. doi: 10.3732/ajb.91.1.1. Stark, L.R., Oliver, M.J., Mishler, B.D. and McLetchie, D.N. 2007. Generational differences in response to desiccation stress in the Desert Moss Tortula inermis. Ann. Bot. 99(1): 53–60. doi:10.1093/aob/mcl238. Stehlik, I., and Barrett, S.C.H. 2005. Mechanisms governing sex-ratio variation in dioecious Rumex nivalis. Evolution 59(4): 814-825. doi: http://dx.doi.org/10.1554/04-417. Stieha, C.R., Middleton, A.R., Stieha, J.K., Trott, S.H., and McLetchie, D.N. 2014. The dispersal process of asexual propagules and the contribution to population

183

persistence in Marchantia (Marchantiaceae). Am. J. Bot. 101(2): 348-356. doi: 10.3732/ajb.1300339. Sultan, S.E. 2001. Phenotypic plasticity for plant development, function and life history. Trends Plant Sci. 5(9): 537-542. Sysov, S.S., and Sycev, M.V. 1970. The pigment complex in leaves of male and female individuals of dioecious plants. Ukranian Bot. J. 27(3): 379-382. [in Ukranian with English abstract] Takenaka, A. 1994. Effects of leaf blade narrowness and petiole length on the light capture efficiency of a shoot. Ecol. Res. 9: 109-114. Tang, W. 1990. Reproduction in the cycad Zamia pumila in a fire-climax habitat: an eight-year study. Bull. Torrey Bot. Club117(4): 368-374. Tanurdzic, M., and Banks, J.A. 2004. Sex-determining mechanisms in land plants. Plant Cell 16: S61-S71. doi: 10.1105/tpc.016667. Tao, Y and Zhang, Y.M. 2012. Effects of leaf hair points of a desert moss on water retention and dew formation: implications for desiccation tolerance. J. Plant Res. 125: 351-360. Tausend, P.C., Goldstein, G., and Meinzer, F.C. 2000. Water utilization, plant hydraulic properties and xylem vulnerability in three contrasting coffee (Coffea arabica) cultivars. Tree Physiol. 20(3): 159–168. doi:10.1093/treephys/20.3.159. Taylor, D.R. 1994. The genetic basis of sex ratio in Silene alba (=S. latifolia). Genetics 136(2): 641-651. Thomas, R.J., Stanton, D.S., and Grusak, M.A. 1979. Radioactive tracer study of sporophyte nutrition in hepatics. Am. J. Bot. 66: 398-403. Thomas, R.J., Stanton, D.S., Longendorfer, D.H., and Farr, M.E. 1978. Physiological evaluation of the nutritional autonomy of a hornwort sporophyte. Bot. Gaz. (Chicago) 139(3): 306-311. Tobias, M. and Niinemets, Ü. 2010. Acclimation of photosynthetic characteristics of the moss Pleurozium schreberi to among-habitat and within-canopy light gradients. Plant Biol. 12: 743-754. Tonnabel, J., Mignot, A., Douzery, E.J.P., Rebelo, A.G., Schurr, F.M., Midgley, J., Illing, N., Justy, F., Orcel, D., and Olivieri, I. 2014. Convergent and correlated evolution of major life-history traits in the angiosperm genus Leucadendron (Proteaceae). Evolution 68(10): 2775-2792. Ueno, N., Suyama, Y., and Seiwa, K. 2007. What makes the sex ratio female-biased in the dioecious tree Salix sachalinensis? J. Ecol. 95(5): 951-959. doi: 10.1111/j.1365-2745.2007.01269.x. Une, K. 1985. Sexual dimorphism in the Japanese species of Macromitrium Brid.(Musci: Orthotrichaceae). J. Hattori Bot. Lab. 59: 487-513. Valladares, F., Skillman, J.B., and Pearcy, R.W. 2002. Convergence in light capture efficiencies among tropical forest understory plants with contrasting crown architectures: a case of morphological compensation. Am. J. Bot. 89(8): 1275– 1284. doi:10.3732/ajb.89.8.1275 Vanderpoorten, A. and Jacquemart, A.-L. 2004. Evolutionary mode, tempo, and phylogenetic association of continuous morphological traits in the moss genus Amblystegium. J. Evol. Biol. 17: 279-187.

184

Vasiliauskas, S.A., and Aarssen, L.W. 1992. Sex ratio and neighbor effects in monospecific stands of Juniperus virginiana. Ecology 73(2): 622-632. Vega-Frutis, R., and Guevara, R. 2009. Different arbuscular mycorrhizal interactions in male and female plants of wild Carica papaya L. Plant Soil 322(1-2): 165-176. doi: 10.1007/s11104-009-9903-6. Vega-Frutis, R., Munguía-Rosas, M.A., Varga, S., and Kytöviita, M.-M. 2013. Sex- specific patterns of antagonistic and mutualistic biotic interactions in dioecious and gynodioecious plants. Perspect. Plant Ecol. Evol. Syst. 15(1): 45-55. doi: 10.1016/j.ppees.2012.10.004. Villarreal, J.C., and Renner, S.S. 2013. Correlates of monoicy and dioicy in hornworts, the apparent sister group to vascular plants. BMC Evol. Biol. 13(1): 239-239. doi: 10.1186/1471-2148-13-239. Vollrath, F. 1998. Dwarf males. Trends Ecol. Evol. 13(4): 159-163. doi: 10.1016/s0169- 5347(97)01283-4. Wall, W. W., Hohmann, M. G., Walker, A.S., and Gray, J. B. 2013. Sex ratios and population persistence in the rare shrub Lindera subcoriacea Wofford. Plant Ecol. 214: 1105-1114. Wallace, C.S., and Rundel, P.W. 1979. Sexual dimorphism and resource allocation in male and female shrubs of Simmondsia chinensis. Oecologia 44(1): 34-39. doi: 10.1007/bf00346394. Wang, B., Yeun, L.H., Xue, J.Y., Liu, Y., Ane, J.M., and Qiu, Y.L. 2010. Presence of three mycorrhizal genes in the common ancestor of land plants suggests a key role of mycorrhizas in the colonization of land by plants. New Phytol. 186(2): 514- 525. doi: 10.1111/j.1469-8137.2009.03137.x. Ward, L.K. 2007. Lifetime sexual dimorphism in Juniperus communis var. communis. Plant Species Biol. 22(1): 11-21. doi: 10.1111/j.1442-1984.2007.00171.x. Watkinson, A.R., and Powell, J.C. 1997. The life history and population structure of Cycas armstrongii in monsoonal northern Australia. Oecologia 111(3): 341-349. doi: 10.1007/s004420050244. Watson, M.A. 1975. Annual periodicity of incremental growth in the moss Polytrichum commune. Bryologist 78(4): 414-422. Weckerly, F.W. 1998. Sexual-size dimorpism: influence of mass and mating systems in the most dimorphic mammals. J. Mammal. 79(1): 33-52. doi: 10.2307/1382840. Wellburn, A.R. 1994. The spectral determination of chlorophylls a and b, as well as total carotenoids, using various solvents with spectrophotometers of various resolutions. J. Plant Physiol. 144: 307-313. Willson, M.F. 1979. Sexual selection in plants. Am. Nat. 113(6): 777-790. Willson, M.F. 1991. Sexual selection, sexual dimorphism and plant phylogeny. Evol. Ecol. 5(1): 69-87. doi: 10.1007/BF02285247. Wittig, R. 2004. The origin and development of the urban flora of Central Europe. Urban Ecosyst 7(4): 323-339. Wyatt, R. 1977. Spatial pattern and gamete dispersal distances in Atrichum angustatum, a dioicous moss. Bryologist 80(2): 284-291. Wyatt, R. 1985. Terminology for bryophyte sexuality: toward a unified system. Taxon 34(3): 420-425.

185

Wyatt, R., and Anderson, L.E. 1984. Breeding systems in bryophytes. Edited by A.F. Dyer and J.G. Duckett. Academic Press, London. pp. 39-64. Xia, J., and Wan, S. 2008. Global response patterns of terrestrial plant species to nitrogen addition. New Phytol. 179(2): 428-439. doi: 10.1111/j.1469-8137.2008.02488.x. Xu, F., Guo, W., Xu, W., Wei, Y., and Wang, R. 2009. Leaf morphology correlates with water and light availability: What consequences for simple and compound leaves? Prog. Nat. Sci. 19(12): 1789-1798. Xu, X., Peng, G., Wu, C., Korpelainen, H., and Li, C. 2008. Drought inhibits photosynthetic capacity more in females than in males of Populus cathayana. Tree Physiol. 28(11): 1751-1759. doi: 10.1093/treephys/28.11.1751. Xu, X., Zhao, H., Zhang, X., Hanninen, H., Korpelainen, H., and Li, C. 2010. Different growth sensitivity to enhanced UV-B radiation between male and female Populus cathayana. Tree Physiol. 30(12): 1489-1498. doi: 10.1093/treephys/tpq094. Yanoviak, S.P., Gora, E.M., Fredley, J., Bitzer, P.M., Muzika, R.-M., and Carson, W.P. 2015. Direct effects of lightning in temperate forests: A review and preliminary survey in a hemlock-hardwood forest of the northern United States. Can. J. For. Res. 45(10): 1258–1268. doi:10.1139/cjfr-2015-0081. Zimmerman, J.K., and Lechowicz, M.J. 1982. Responses to moisture stress in male and female plants of Rumex acetosella L.(Polygonaceae). Oecologia 53(3): 305-309. Zotz, G., Schweikert, A., Jetz, W., and Westerman, H. 2000. Water relations and carbon gain are closely related to cushion size in the moss Grimmia pulvinata. New Phytol. 148(1): 59–67. doi:10.1046/j.1469-8137.2000.00745.x.

186

VITA

Jonathan D. Moore III

Place of Birth: Kettering, OH, USA

Education B.S. Biology, B.A. Chemistry, University of Kentucky, Lexington, KY 2009 (Summa cum laude)

Professional Positions Teaching Assistant, University of Kentucky, Introductory Biology I Lab (Fall 2009, 2010) Teaching Assistant, University of Kentucky, Introductory Biology II Lab (Spring 2010, 2011) Teaching Assistant, University of Kentucky, Ecology Recitation (Fall 2011) Teaching Assistant, University of Kentucky, Ecology Lab (Spring 2012-2017; Fall 2012, 2014) Teaching Assistant, University of Kentucky, Plant Kingdom Lab (Fall 2013, 2015, 2016) Undergraduate Research mentor, University of Kentucky (Fall 2010, 2013; Spring 2012, 2014) REU mentor, University of Kentucky (Summer 2012, 2013)

Scholastic and Professional Honors Ribble research grant (2010, 2011, 2012) Ribble travel grant (2011, 2014, 2015, 2016) University of Kentucky Graduate travel grant (2014, 2015, 2016) American Bryological and Lichenological Society student travel grant (2014, 2015)

Professional publications Moore, J.D., Kollar, L.M., and McLetchie, D.N. 2016. Does selection for gamete dispersal and capture lead to a sex difference in clump water-holding capacity? Am. J. Bot. 103(8): 1-9.

187

SEXUAL DIMORPHISM in the MOSS &lt;Em&gt;BRYUM ARGENTEUM&lt;/Em

SEXUAL DIMORPHISM in the MOSS <Em>BRYUM ARGENTEUM</Em