historical contingencies in the ecology and evolution of

Historical Contingencies in the Ecology and Evolution of

Species Diversity

by

Rachel M. Germain

A thesis submitted in conformity with the requirements

for the degree of Doctor of Philosophy

Department of Ecology and Evolutionary Biology

University of Toronto

© Copyright by Rachel M. Germain 2016

ii

Historical contingencies in the ecology and evolution of species

diversity

Rachel M. Germain

Doctor of Philosophy

Department of Ecology and Evolutionary Biology

University of Toronto

2016

Abstract

Ecologists have long-sought to explain the high diversity of species in biological communities,

given that classic theory predicts that diversity is limited by available niche space. In recent

years, ecologists have looked towards ‘historical contingencies’, the persistent effects of past

ecological and evolutionary processes, as possible mechanisms that maintain diverse

communities, either by relaxing the constraints of niche availability or by adding temporal

dimensions to species’ niches. In this thesis, I use field and greenhouse experiments to explore

three ways in which historical contingencies manifest in annual plant communities. First, my

work on maternal effects shows that abiotic (ch. 2) and biotic (ch.3) conditions in the maternal

generation have diverse effects on offspring phenotypes across an assemblage of species.

Because species differences in environmental responses can facilitate coexistence, these studies

suggest that maternal effects could act as a form of niche differentiation, and motivate future

research to clarify their influences on coexistence outcomes. Second, I performed, to my

knowledge, the first experimental decoupling of dispersal limitation and environmental sorting in

iii

a natural landscape by manipulating entire seed pools of annual plants (ch. 4). In doing so, I was

able to identify the pervasive and scale-specific influences of dispersal limitation that constrain

species distributions in plant communities. Lastly, I used competitive trials to identify

macroevolutionary divergence in competitive interactions among species (ch. 5), and found

evidence that divergence is contingent on historical competitive interactions in ways that are

consistent with character displacement. In sum, my dissertation work has expanded our

understanding of (i) the number of potential niche dimensions that might allow species to

differentiate, (ii) how this differentiation can arise over evolutionary time, and (iii) the interplay

of current and historical conditions in the maintenance of species diversity, and the timescales

over which they play out.

iv

Acknowledgments

I have a long list of friends and colleagues to thank who have helped me complete my PhD and

keep my sanity in the process.

First and foremost, I am indebted to my advisor Benjamin Gilbert. It is rare to find an advisor

who is the combination of a brilliant mentor, an enthusiast of Ben Stiller movies c. 2000, and

willing to shovel (and re-shovel)10 tonnes of soil off of a university driveway when his student

makes poor life choices. Choosing a PhD advisor is one of the most important and most difficult

decisions that a young scientist can make, and I truly believe that working with Ben was one of

the best decisions I have made for myself. He has forever altered the way that I think as a

scientist and my perspective of academia, and it has been exciting to watch his lab develop and

successes accumulate over the past five years.

I will deeply miss my talented lab family: Natalie Jones, Tess Grainger, Denon Start, and Kelly

Carscadden. Natalie is the sister I never had; she has sat ~4ft from me at all times during the last

five years, and I am glad to have shared many of life’s big moments with her. Her successes have

led her to more xeric pastures living the dream at UC San Diego. Tess has taught me a lot about

life and how to live it (buying eggs on a Tuesday); I am going to miss our amped up science

discussions and mutual desire to derail book clubs. Denon is someone that I am both inspired and

annoyed by, but I will admit that that his projects are pretty rad (he’s too young to know what

that means). Kelly is pun-master extraordinaire; I admire her attention to detail, command of the

literature, and quirky sense of humor. Each of these folks will go on to do amazing things in life,

and I look forward to being their colleague in the future.

In addition to the members of my lab, I have made a number of great friends who have offered

comradery, advice, feedback, support, and commiseration in different ways throughout my time

here. This includes, in no particular order, Alex DeSerrano, Adam Cembrowski, Franus Hauser,

Jenn Coughlan, Rebecca Batstone, David "Jello" Timerman, Michael Foisy ♥, Mitch Daniel,

Luke Rogers, Thomson-lab Paul, Kyle Turner, Stephen DeLisle, not Luke Frishkoff (I haven’t

known him long enough), and many others. A special shout out to Alex DeSerrano, who I first

met in 2011 while admiring her use of fish cartoons as a font on a research poster. Grad school is

full of ups and downs but it is great friends and fun that keep things moving.

v

The Department of Ecology & Evolutionary Biology at the University of Toronto is truly

exceptional, and I am glad to have had the opportunity to carry out my graduate work here. I am

grateful for the time and insight of past and present members of my thesis committee, Jason

Weir, John Stinchcombe, and Marc Cadotte. Helen Rodd, Don Jackson, and Locke Rowe are

champions in the improvement of graduate student life in their positions within EEB and SGS. I

also thank my office neighbor and close friend of Ke$ha, Donald L. Mahler, for his tolerance and

for including me in his lab meetings. There are many more faculty and graduate students to thank

who have contributed to my intellectual and interpersonal development.

There are a number of folks at UC Davis and the University of Toronto who have helped me in

the field, laboratory, greenhouse, or administratively. From my time in California, I thank Sharon

Strauss and Susan Harrison for introducing me to a new study system, as well as their lab

members for project feedback and plant identification. At the University of Toronto, I have been

helped by Kitty Lam, Bruce Hall, Andrew Petrie, Bill Cole, Jim Dix, Jill Wheeler, Jenn English.

There is also a small army of undergraduate assistants to thank, most notably Chris Blackford,

Alanna Leale, Abdul Yossof, Yvonne Chan, and Alicia Chou.

I am grateful for my early mentors at the University of Guelph’s Department of Integrative

Biology, Andrew MacDougall, Chris Caruso, Hafiz Maherali, and Mark Sherrard, for their

dedication to undergraduate research training. I am particularly thankful to Chris and Hafiz, who

welcomed me into their labs despite my lackluster grades and helped me develop the skills to be

successful in graduate school, and to Andrew MacDougall, whose laidback enthusiasm for

community ecology and field research compelled me to pursue it in graduate school. I would not

be where I am today without having worked under their guidance at a critical juncture in life.

Last but not least, I could not have completed this work without the support of my mum and

brother, Marie Germain and Adam Germain, as well as my partner, Alex Kushnir. Alex has put

up with me through late nights, bad moods, weekends in the greenhouse, “insufficient funds”,

long fieldwork absences, etc. As someone outside of academia, his perspective of how bizarre

the day-to-day of ecology and graduate school is (“daycare for adults”) reminds me not to take

myself too seriously and to enjoy the ride. It has been fun exploring Toronto together, and I look

forward to new adventures on the west coast with our three cats, who are to blame for any

typographical errors in this thesis.

vi

This research was funded by NSERC, Sigma Xi, the Botanical Society of America, and

fellowships provided by the University of Toronto and the Department of Ecology and

Evolutionary Biology.

vii

Table of Contents

Acknowledgments........................................................................................................................... ii

Table of Contents .......................................................................................................................... vii

List of Figures ..................................................................................................................................x

List of Appendices ....................................................................................................................... xiv

Chapter 1 General Introduction .......................................................................................................1

Conceptual background ...............................................................................................................1

Thesis overview ..........................................................................................................................3

Maternal effects of past environments .................................................................................4

Dispersal histories in a metacommunity ..............................................................................5

Historical interactions and macroevolution .........................................................................6

Synthesis of research objectives ..........................................................................................7

Literature cited ............................................................................................................................8

Chapter 2 Hidden responses to environmental variation: maternal effects reveal species niche

dimensions ................................................................................................................................13

Abstract .....................................................................................................................................13

Introduction ...............................................................................................................................13

Materials and methods ..............................................................................................................16

Study system ......................................................................................................................16

Maternal generation ...........................................................................................................17

Offspring generation ..........................................................................................................18

Phylogenetic tree construction ...........................................................................................19

Data analysis ......................................................................................................................19

Results .......................................................................................................................................21

Seed size, seed number, and their correlation ....................................................................21

Offspring performance .......................................................................................................24

viii

Discussion .................................................................................................................................25

Acknowledgements ...................................................................................................................29

Literature cited ..........................................................................................................................29

Chapter 3 Frequency-dependent maternal effects across species and environments ....................33

Abstract .....................................................................................................................................33

Introduction ...............................................................................................................................34

Methods .....................................................................................................................................37

Study species ......................................................................................................................37

Greenhouse experiment .....................................................................................................38

Statistical analyses .............................................................................................................38

Results .......................................................................................................................................40

Discussion .................................................................................................................................44

Conclusion ................................................................................................................................47



Chapter 4 Pervasive dispersal limitation reduces diversity in an annual plant community ...........52

Abstract .....................................................................................................................................52

Introduction ...............................................................................................................................52


Study system ......................................................................................................................56

Experimental setup.............................................................................................................57

Data analysis ......................................................................................................................58

Results .......................................................................................................................................61

Discussion .................................................................................................................................65



ix

Chapter 5 Species coexistence: macroevolutionary relationships and the contingency of

historical interactions ................................................................................................................72

Abstract .....................................................................................................................................72

Introduction ...............................................................................................................................72


Species selection ................................................................................................................74

Greenhouse growing conditions ........................................................................................75

Experimental design...........................................................................................................76

Solving for stabilizing differences, fitness differences, and coexistence outcomes ..........78

Statistical analyses .............................................................................................................79

Results and discussion ..............................................................................................................79

Data accessibility ......................................................................................................................86

Authors’ contributions ..............................................................................................................86

Funding statement .....................................................................................................................86



Chapter 6 General Conclusions .....................................................................................................91

Maternal effects of past environments ...............................................................................91

Dispersal histories in a metacommunity ............................................................................92

Historical interactions and macroevolution .......................................................................93

Concluding remarks ..................................................................................................................94


Copyright Acknowledgements.....................................................................................................125

x

List of Figures

Figure 1.1. Schematic positioning my thesis chapters by the timescales that historical

contingencies play out, and their impacts on local or regional diversity. The shaded regions

outline the breadth of each chapter. ................................................................................................ 3

Figure 2.1. Effect sizes of (A) seed size, (B) seed number, and (C) seed dormancy responses to

our watering treatments across a phylogeny of 29 species; see Table S1 for species codes. Effect

sizes were calculated using Cohen’s d (i.e., (µdry - µwet)/pooled σ; Hartung et al. 2011). Positive

effect sizes indicate that trait values were higher in the dry compared to wet environment, and

vice versa. The numbers on the phylogeny correspond to taxonomic orders: (1) Poales, (2)

Ranunculales, (3) Fabales, (4) Caryophyllales, (5) Boraginales, (6) Lamiales, and (7) Asterales.

Dashed branches represent unknown relationships due to missing sequence data [SCOLU and

SVIRI], and ‘×’ indicates unavailable trait data rather than a true zero. * P < 0.05, º P < 0.10. .. 22

Figure 2.2. Seed size-seed number correlations. (A) Correlation among species, with values

representing the species-level means with a best-fit line indicating a significant correlation (R2 =

0.45, P < 0.001). (B) Correlations within species; seed size and seed number were standardized

by species and a constant added to visualize the correlation coefficients for each species. Black

and dark grey lines represent significant (P < 0.05) and marginally significant (P < 0.10)

correlations, respectively; light grey lines are non-significant. The dashed line represents no

correlation (r = 0). All data were log transformed (A) prior to standardization (B)..................... 24

Figure 3.1. Cross-species comparison of frequency-dependent (a) maternal provisioning of seed

size and (b) seed number responses. Points are slope coefficients (± 2 × standard error) of species

responses to the frequency of conspecific competitors, and are shaded black, gray, or white to

indicate significant (P < 0.05), marginally-significant (P < 0.10), or non-significant (P > 0.10)

slopes, respectively. In most cases, frequency × soil moisture interactions are non-significant,

and slopes are averaged across soil moisture environments; when significant, wet (upwards

triangle) and dry (downwards triangle) environments are plotted separately. Competitive pairs

are delineated by lines connecting species codes; see Table S3.1 for species codes. ................... 41

xi

Figure 3.2. Correlations of (a) seed size and seed number responses to conspecific frequency (n

= 60 [15 pairs × two species × two soil moisture environments]), and (b) seed size responses to

soil moisture conditions in the presence and absence of competition (n = 25). Each point is a

species, and points that fall in the grey zones are species with opposing directions of responses.

....................................................................................................................................................... 42

Figure 3.3. Variation in the strength of frequency-dependent maternal provisioning of seed size

explained by the phylogenetic distance between competitors, depending on whether they are

native to the same biogeographic region (a,c) or different regions (b,d). The red dashed line

indicates the cut-off between negative and positive frequency dependence, grey dashed lines

delineate pairs of species that were competed, and solid lines denote significant relationships.

Black symbols and lines are species native to California, whereas gray symbols and lines are

species native to Spain. ................................................................................................................. 43

Figure 4.2. Effect of spatial scale of seed pooling on (a) species richness (mean ± se) and (b) the

strength of species-environment relationships (R2 values); observed patterns correspond to

prediction in Fig. 4.1a. Points with the same letter were not significantly different in a multiple

comparisons test. ........................................................................................................................... 61

Figure 4.3. Effect of spatial scale of seed pooling on (a) species composition and (b) the

compositional dissimilarity of plots (β-diversity). Species composition is from the first axis

scores from a PCoA with Jaccard’s dissimilarity index (see biplot in Fig. S4.4a); results with

second axis scores are qualitatively similar. Points with the same letter were not significantly

different in a multiple comparisons test. ....................................................................................... 63

Figure 4.4. Frequency distribution of site occupancies of all species (gray bars) and species that

were significantly associated with the environment (black bars). The red line is the fitted

relationship between the proportion of species with significant species-environment relationships

and site occupancy; this relationship is a significantly hump-shaped (P = 0.008). Note that the x-

axis is on a log-scale. .................................................................................................................... 63

Figure 4.5. Natural differences among halves of the reserve in (a) environmental conditions

(from a PCA) and (b) species composition (from a PCoA). The contours outline the bivariate

normal distributions of each half of the reserve; dashed contours/upwards triangles are sites that

xii

occur in the north half, solid contours/downwards triangles are sites that occur in the south half.

Environmental conditions (P = 0.001) but not species (P = 0.175) differed significantly between

halves (see Results). ...................................................................................................................... 64

Figure 5.1. Biogeographic history alters the evolutionary trajectory of stabilizing and fitness

differences. (a) Stabilizing differences rapidly increase among sympatric species pairs (light

shade), whereas allopatric species pairs (dark shade) show no relationship. (b) Fitness

differences, by contrast, increase over evolutionary time in both sympatric and allopatric pairs,

but are larger on average among allopatric pairs. Stabilizing differences have a maximum of one

(Eq. S5.1; lines are fitted from the logit-transformed data), whereas fitness differences have no

upper limit (Eq. S5.2). Because soil moisture had no effect on stabilizing or fitness differences,

each point is a fitted average across soil moisture environments for each species pair. .............. 80

Figure 5.2. Fitness differences arise through alternate mechanisms in sympatric and allopatric

species pairs. In log-space, fitness differences (solid line) are the sum of competitive (dashed

line) and demographic (dotted line) components (see Materials and methods), shown here as

fitted relationships. (a) In sympatric species pairs, fitness differences are minimal relative to

allopatric species pairs, and appear to accumulate over macroevolutionary time primarily through

differences in demographic rates. (b) In allopatric species pairs, the more rapid increase in fitness

differences over evolutionary time is driven solely by divergence in competitive ability, as the

apparent divergence in demographic rates is non-significant. Values in the shaded area indicate

that the species with the highest fitness had the lowest fitness component. ................................. 82

Figure 5.3. The effect of evolutionary history on coexistence outcomes depends on

biogeographic history. Species coexistence is not influenced by the phylogenetic distances of

species pairs that occur in sympatry (light shade, slope not significantly different from zero), but

the probability of coexistence decreases with the phylogenetic distance of allopatric pairs (dark

shade). Species pairs were grown in wet (squares) and dry (triangles) environments, but soil

moisture does not alter the effect of evolutionary history on coexistence (fitted lines are averaged

across environments). The dashed line indicates the threshold between coexistence (positive

values) and competitive exclusion (negative values); the coexistence metric is given by Eq. 5.2 in

Materials and methods. Points connected by a solid line represent cases in which a species pair

xiii

was predicted to coexist in one but not both environments (3 of 10 sympatric pairs, 0 of 10

allopatric pairs). ............................................................................................................................ 83

xiv

List of Appendices

Appendix A: Supplementary Information to Chapter 2 .................................................................97

Supplementary tables and figures .............................................................................................97

Appendix B: Supplementary Information to Chapter 3 ...............................................................102

Species Selection .....................................................................................................................102

Greenhouse Experiment ..........................................................................................................102

Statistical Analyses .................................................................................................................103

Supplementary tables and figures ...........................................................................................104

Appendix C: Supplementary Information to Chapter 4 ...............................................................109


Appendix D: Supplementary Information to Chapter 5 ...............................................................115

Phylogenetic tree construction ................................................................................................115

Details of lambda estimation ...................................................................................................116

Supplementary equations ........................................................................................................116


CHAPTER 1: GENERAL INTRODUCTION

1

Chapter 1

General Introduction

Conceptual background

The ecological niche entered our lexicon over a century ago, first by Johnson (1910) and by

Grinnell (1917) and Elton (1926) not soon after, to describe the relationship between species and

their environments. Initially, a species’ niche quite literally referred to its place or role in an

ecological community, defined by a particular habitat-type or essential resource; for example,

vultures would occupy the desert-dwelling carrion-eating niche. This initial definition has

undergone several major conceptual revisions, first by Hutchinson (1957) who regarded the

niche as a property of a population or species, rather than a property of a particular environment

(Colwell & Rangel 2009), and later by MacArthur, Levins (1967), and others (Chase & Leibold

2003) to include the positioning of species along multiple resource, predation, and environmental

axes. These classic ideas set the conceptual foundation upon which modern ecology is built, and

have since been expanded to inform predictions of species coexistence (Adler et al. 2007) and

the maintenance of diverse species assemblages (Tilman & Pacala 1993).

An enduring preoccupation of modern ecologists has been to develop an explicit understanding

of how differences among species manifest in ways that maintain biological diversity. Central to

this preoccupation is idea that no two species can occupy the same niche space and stably coexist

through time; some minimum amount of ecological difference must exist to prevent competitive

exclusion (Gause 1932; MacArthur & Levins 1967). For example, Hutchinson (1959) noted that

the gape or body sizes of co-occurring species tended to differ by a ratio of at least 1.3:1 in a

number of taxa, including birds and mammals. Although the practicality of testing this theory has

been met with intense criticism (e.g., Simberloff & Boecklen 1981), more fundamentally, its

conceptual basis has since been revised and expanded in ways that have redefined the field of

ecology.

The most recent revision stems from theoretical work by Chesson (2000) as synthesized in the

highly cited article by Adler et al. (2007), which posits that species’ differences in resource use

can either promote or preclude local coexistence. Stabilizing differences are competitive


2

differences that promote coexistence, and represent the degree to which population growth rates

are reduced by intraspecific vs. interspecific competition; these differences are analogous to

those discussed by Gause (1932) and Hutchinson (1959). By contrast, fitness differences are

competitive difference that preclude coexistence, and arise when differences in overall

competitive ability lead to the exclusion of inferior competitors by dominant species. This new

framework has been impactful because it has allowed ecologists to better predict coexistence

outcomes based on the combined effects of stabilizing and fitness differences (Levine &

HilleRisLambers 2009), and has since been invoked to explain a diversity of ecological

phenomena, including ecosystem functioning (Carroll et al. 2011), species invasions

(MacDougall et al. 2009; Godoy & Levine 2013), and community phylogenetics (Mayfield &

Levine 2010).

A challenge to understanding the determinants of species coexistence and diversity has been

expanding predictions beyond outcomes at “local” scales – the scales at which individuals

interact – to diverse assemblages of species that exist across regional landscapes. Although many

approaches have been proposed to broaden theory to larger spatial scales (e.g., Caley & Schluter

1997; Kneital & Chase 2004; Rajaniemi et al. 2004), Leibold et al. (2004) offered a necessary

synthesis of ideas with the formalization of the metacommunity concept. A metacommunity is

defined as a regional network of local communities linked by the dispersal of multiple interacting

species. Local diversity in a metacommunity is not solely a product of local conditions, but

rather, feedbacks among processes that occur at local and regional scales, and can generate a

suite of potential dynamics and patterns (Pulliam 2000). For example, in heterogenous systems

with high dispersal rates, populations of species frequently persist in localities of low suitability,

causing increases in local diversity that homogenize differences among localities (Mouquet &

Loreau 2003). Leibold et al.’s (2004) framework has strongly influenced the field of ecology for

the last decade with the recognition that most communities likely exist as part of a larger regional

metacommunity.

In addition to spatial context, ecologists have moved towards a more explicit understanding of

temporal processes over multiple timescales (e.g., Hudson & Catadorri 1994), and the ways in

which spatial and temporal processes are or are not analogous (Wolkovich et al. 2014). One

phenomenon that is uniquely temporal in nature is the presence of historical contingencies, or


3

‘ghosts of the past’, that have effects on diversity patterns that persist long after their ecological

drivers have ceased to operate. For example, plant diversity in an experimental grassland was

27% higher in plots that were exposed to seed predators three years earlier, compared to plots

that had never been exposed to seed predators (Howe & Brown 2001). Many other examples of

historical contingencies exist – “ghosts of x past” appears in the titles of 53 ecological

publications (Web of Science 2016), where x includes competition, predation, environments,

land use, dispersal, evolution, and others (e.g., Knicks et al. 2000; Howe et al. 2001). Persistence

of such ecological ghosts can occur because reverting back to the initial state is difficult or

impossible (e.g., evolutionary change [Klingenberg & Gidaszewski 2009]), because recovery is

slow relative to the ecological dynamics of a system, or because ghosts reappear frequently

enough that ecosystems are continually influenced by the interplay of current and historical

conditions. When present but not accounted for, historical contingencies contribute to the

unexplained variation of a biological system, and have potential to obscure important

relationships among ecological processes of the past and present.

Thesis overview

My thesis explores three manifestations of historical contingencies that alter our understanding

of the niche, and play out over distinct spatial and temporal scales (Fig. 1.1), including past

environments through maternal effects (chs. 2 and 3), dispersal histories as affected by spatial

constraints (ch. 4), and competitive interactions on macroevolutionary timescales (ch. 5). As I

will discuss, each project draws either conceptual or empirical links among disparate subfields in

the biodiversity sciences, and together serve to deepen our understanding of the forces that affect

species coexistence and the maintenance of diversity.

Figure 1.1. Schematic positioning my thesis chapters by the timescales at which historical

contingencies play out, and their impacts on local or regional diversity. The shaded regions

outline the breadth of each chapter.


4

Maternal effects of past environments

In chapters 2 and 3, I explore how the biotic and abiotic environments experienced by the

previous generation alter the phenotypes of offspring (aka maternal effects) in an ecological

assemblage of species. Because many phenotypes are tied to fitness, maternal effects have been

shown to interact with the offspring environment to determine offspring fitness in a diversity of

organisms (Herman & Sultan 2011). For example, Daphnia cuculatta produces offspring with

anti-predator structures, such as helmets and neck spines, when exposed to predators in the

maternal generation (Agrawal et al. 1999); this type of maternal effect can greatly increase

offspring survival, but only if predators are also present in the offspring environment. Without

knowing the history of the maternal environment, the factors that affect distributions of

phenotypes and fitnesses of individuals in populations of species that exhibit maternal effects

would be poorly understood.

Maternal effects have been intensively studied for five decades, almost always from an

evolutionary standpoint using genetically-controlled studies of individual species (e.g., Galloway

& Etterson 2007; see reviews by Roach & Wulff 1987 and Mousseau & Fox 1998)—yet their

prevalence across species and possible ecological consequences are largely unexplored. As one

anonymous reviewer of chapter 3 stated, “the study of maternal effects is an inherently

evolutionary question”, implying it is inappropriate or uninteresting to test their effects on

populations or communities. Yet, in the few ecological studies that exist, maternal effects have

been shown to have large effects on population persistence and the outcome of competition (e.g.,

Metz et al. 2015; Van Allen & Rudolf 2015). The importance of maternal effects for

evolutionary processes, and their apparent influence on ecological processes, suggest that

maternal effects should be considered more broadly in population and community studies.

There are many ways in which maternal effects might impact the ecological dynamics of

populations and communities. As discussed earlier, differences among species in resource use or

environmental responses are required for coexistence. Studies of different species have revealed

different types of maternal effects in response to similar sets of maternal and offspring

environments (e.g., Sultan et al. 2006), and raise the possibility that maternal effects are an

additional axis along which species can differentiate. Indeed, preliminary modeling work points


5

towards species-specific differences in maternal effects as a potential mechanism of species

coexistence, even in the absence of any other form of ecological differentiation (Germain et al.

in progress). Other ecological consequences include modifications to species’ realized and

fundamental niches (Sultan et al. 2006), the ability of individuals to disperse (Larios & Venable

2014) and establish (Van Allen & Rudolf 2013) across regional landscapes, and population

fluctuations over time (Ginzburg & Taneyhill 1994). However, more multi-species, ecologically-

motivated tests of maternal effects are needed to explore the full extent of their importance.

Dispersal histories in a metacommunity

In chapter 4, I experimentally disentangle the roles of species differences in environmental niche

use, environmental heterogeneity, and dispersal limitation in a natural metacommunity using a

novel method. These three processes, depending on their relative strengths, underlie a suite of

metacommunity paradigms that regulate the spatial distributions of species across landscapes

(Pulliam 2000; Leibold et al. 2004). More importantly, these processes form the basis of much

larger debates in ecology, on whether species distributions are more strongly constrained by

environments or dispersal (Cottenie 2005), and if the latter, the degree to which dispersal

constraints allow the persistence of functionally-equivalent “neutral” species (Chave 2004). A

strong role for dispersal limitation implies that diversity patterns are more contingent on chance

and dispersal histories than associations with the underlying environment (Hurtt & Pacala 1994),

or that species-environment relationships can only be viewed within the context of habitat

isolation (Tuomisto et al. 2003; Gilbert & Lechwicz 2004); how the findings of small-scale

ecological experiments can be extended to regional communities is less clear in this scenario.

Although ecologists have come closer to providing closure to these debates, definitive answers

have been hindered by the lack of appropriate methods to compare the relative strengths of

regional processes in natural communities. Existing methods use either statistical partitioning of

observational data (Cottenie 2005) or experiments that use simplified species pools (e.g., Tilman

1997; Germain et al. 2013) or environments (e.g., Low-Décarie et al. 2015). However,

observational data is collected from communities that are already the product of simultaneously

operating processes, and mesocosm experiments lack the biological realism of natural species

pools and environments (Grainger & Gilbert 2016). Finding methods that can experimentally


6

differentiate among environmental and dispersal constraints, while making use of natural

variation in species pools and environments across landscapes, would provide novel insight into

biodiversity`s spatial underpinnings.

Historical interactions and macroevolution

In chapter 5, I explore contingencies of evolutionary and biogeographic histories on the intensity

of competitive interactions among species. As with many ideas in evolutionary biology, Charles

Darwin was the first to recognize that a relationship might exist between the ecology and

evolution of species (Darwin 1859). Because species of the same genus tend to possess similar

traits, Darwin predicted that competition would be most intense between closely related species.

Some of the earliest evidence supporting this prediction came almost a century after The Origin

of Species was first published, with the discovery that fewer species per genus are represented

locally than are regionally available (Elton 1946). Although the generality of such patterns and

the processes that generate them have been revisited many times over the last 60 years (Harper et

al. 1961), they remain two of the major unanswered questions in ecology.

The prevailing framework for the relationship between evolutionary history and ecological

interactions in the last decade was developed by Webb et al. (2002), who posited that community

phylogenetic patterns could be used to make inferences about the ecological processes

underlying their formation, such as competition or environmental filtering. Webb et al. (2002)’s

seminal paper resulted in a surge of publications featuring community phylogenetic patterns, as

ecologists jumped onto what is now termed the ‘phylogenetic bandwagon’ (Fox 2012). In fact,

~60% of manuscript submissions to the American Naturalist from 2010-2012 featured analyses

of phylogenetic dispersion patterns (J. Losos, pers. comm.). However, evidence quickly began to

amass that these patterns were often weak to non-existent (Silvertown et al. 2007; Cahill et al.

2008; Vamosi et al. 2009; Anderson et al. 2011), or were opposite to those predicted by Webb et

al. (2002) in cases where the ecological processes were known (Cavender-Bares et al. 2006).

Together, these findings have led many ecologists to question the utility of phylogenetic patterns

in understanding broad-scale ecological processes (e.g., Weiher et al. 2011).

A major criticism of Webb et al.’s (2002) framework is based on recent developments in species

coexistence theory (Adler et al. 2007; sensu Chesson 2000), which posits that competition alone


7

can favour coexistence between species that are closely or distantly related; the community

phylogenetics framework proposed by Webb et al. (2002) is only valid if stabilizing differences

that promote coexistence evolve faster than fitness differences that preclude coexistence

(Mayfield & Levine 2010). Not only was it not known how stabilizing and fitness differences

evolve until very recently (Narwani et al. 2013; Godoy et al. 2014), but there are also a number

of reasons to expect their evolution to differ among competitors of contrasting biogeographic

histories (sympatry vs. allopatry). The idea that evolution would differ among sympatric and

allopatric competitors (Mooney & Cleland 2001) is well-supported in invasion biology (e.g.,

Diez et al. 2008; Strauss et al. 2006), but has yet to be tested in terms of phylogenetic

community assembly and the mechanisms of species coexistence (MacDougall et al. 2009). Such

tests would provide much needed resolution to the field of community phylogenetics, and more

broadly, how historical interactions have modified the macroevolutionary trajectories of

competitive differences among species.

Synthesis of research objectives

My thesis describes four experiments on annual plant communities that take place in the

greenhouse or field, each linked by the common goal of understanding the mechanisms that

maintain species diversity. Though each experiment addresses this goal from a different

perspective, their common themes are rooted in the history of ecological research, including the

roles of species differences, environmental heterogeneity, spatial context, and historical

contingencies that manifest on ecological and evolutionary timescales. In chapters 2 and 3, I test

the prevalence of maternal effects on offspring phenotypes (i.e., seed size, seed dormancy,

seedling performance) in an assemblage of species, and similarities and dissimilarities among

maternal effects in response abiotic and biotic conditions. Chapter 4 was conducted in a natural

plant metacommunity in Northern California, where I used field manipulations to test how

strongly species distributions are constrained by dispersal limitation, the spatial scales at which

dispersal limitation manifests, and whether the experimental removal of dispersal limitation

allows species to sort along environmental gradients. In chapter 5, I used plant competition

experiments to parameterize annual plant models and estimate how competitive differences

structure species coexistence. I then tested the macroevolutionary trajectories of stabilizing

differences, fitness difference, and coexistence outcomes, and determined how these trajectories


8

are modified by the biogeographic histories of competing species. Chapter 6 provides a

conceptual synthesis of the themes of my four data chapters, the implications of their findings,

and directions for future study.

The chapters of my dissertation are presented as stand-alone research papers. As a result, there is

some repetition among chapters in the Introduction, Methods, and Literature cited sections.

Chapters 2 and 5 are published and included with permission from their publishers; Chapter 3 is

currently in review at the American Naturalist, and Chapter 4 is in preparation for submission to

Proceedings of the National Academy of Sciences. The manuscripts that stem from my thesis

chapters were completed in collaboration with Benjamin Gilbert (chs. 2-5), Natalie Jones (ch. 3),

Tess Grainger (ch. 3), Sharon Strauss (ch. 4), and Jason Weir (ch. 5). In all cases, I have fulfilled

the responsibilities required of a primary author, and my co-authors have each contributed

invaluable feedback and assistance at different stages of each project’s completion.

Literature cited

Adler, P.B., HilleRisLambers, J. & Levine, J.M. (2007) A niche for neutrality. Ecology Letters,

10, 95–104.

Agrawal, A.A., Laforsch, C. & Tollrian, R. (1999) Transgenerational induction of defenses in

animals and plants. Nature, 401, 60-63.

Anderson, T.M., Shaw, J. & Olff, H. (2011) Ecology’s cruel dilemma, phylogenetic trait

evolution and the assembly of Serengeti plant communities. Journal of Ecology, 99, 797–

806.

Caley, M.J. & Schluter, D. (1997) The relationship between local and regional diversity.

Ecology, 78, 70–80.

Cahill, J.F., Kembel, S.W., Lamb, E.G. & Keddy, P.A. (2008) Does phylogenetic relatedness

influence the strength of competition among vascular plants? Perspectives in Plant

Ecolology, Evoution, and Systematics, 10, 41–50.

Carroll, I.T., Cardinale, B.J. & Nisbet, R.M. (2011) Niche and fitness differences relate the

maintenance of diversity to ecosystem function. Ecology, 92, 1157-1165.

Cavender-Bares, J., Keen, A. & Miles, B. (2006) Phylogenetic structure of Floridian plant

communities depends on taxonomic and spatial scale. Ecology, 87, 109-122.

Chase, J.M. & Leibold, M.A. (2003) Ecological niches: linking classical and contemporary

approaches. University of Chicago Press, Chicago.


9

Chave, J. (2004) Neutral theory and community ecology. Ecology Letters, 7, 241-253.

Chesson, P. (2000) Mechanisms of maintenance of species diversity. Annual Review of Ecology

and Systematics, 31, 343–366.

Colwell, R.K. & Rangel T.F. (2009) Hutchinson’s duality: the once and future niche.

Proceedings of the National Academy of Sciences, 106,19651-19658.

Cottenie, K. (2005) Integrating environmental and spatial processes in ecological community

dynamics. Ecology Letters, 8, 1175-1182.

Darwin, C. (1859) On the origin of species by means of natural selection, or the preservation of

favoured races in the struggle for life. John Murray, London.

Diez, J.M., Sullivan, J.J., Hulme, P.E., Edwards, G. & Duncan, R.P. (2008) Darwin’s

naturalization conundrum: dissecting taxonomic patterns of species invasions. Ecology

Letters, 11, 674–681.

Elton, C. (1927) Animal Ecology. The MacMillan Company, New York.

Elton, C. (1946) Competition and the structure of ecological communities. Journal of Animal

Ecology, 15, 54–68.

Fox, J. (2012-10-09) Can the Phylogenetic Community Ecology Bandwagon be Stopped? in

Dynamic Ecology: http://dynamicecology.wordpress.com/2012/10/09/

Galloway, L.F. & Etterson, J.R. (2007) Transgenerational plasticity is adaptive in the wild.

Science, 318, 1134-1136.

Gause, G.F. (1932) Experimental studies on the struggle for coexistence. I. Mixed population of

two species of yeast. Journal of Experimental Biology, 9, 389-402.

Germain, R.M., Caruso, C.M. & Maherali, H. (2013) Mechanisms and consequences of water

stress-induced parental effects in an invasive annual grass. International Journal of Plant.

Sciences, 174, 886-895.

Gilbert, B. & Lechowicz, M.J. (2004) Neutrality, niches, and dispersal in a temperate forest

understory. Proceedings of the National Academy of Sciences, 101, 7651-7656.

Ginzburg, L.R. & Taneyhill, D.E. (1994) Population cycles of forest Lepidoptera: a maternal

effect hypothesis. Journal of Animal Ecology, 63, 79-92.

Godoy, O. & Levine, J.M. (2013) Phenology effects on invasion success: insights from coupling

field experiments to coexistence theory. Ecology, 95, 726-736.

Godoy, O., Kraft, N.J.B. & Levine, J.M. (2014) Phylogenetic relatedness and the determinants of

competitive outcomes. Ecology Letters, 17, 836–844.


10

Grainger, T.S. & Gilbert, B. (2016) Dispersal and diversity in metacommunities: linking theory

and practice. Oikos early view: 10.5061/dryad.nq853

Grinnell, J. (1917) The niche-relationships of the California thrasher. Auk, 34, 427–433

Harper, J.L., Clatworthy, J.N., McNaughton, I.H. & Sagar, G.R. (1961) The evolution and

ecology of closely related species living in the same area. Evolution, 15, 209–227.

Herman, J.J. & Sultan, S.E. (2011). Adaptive transgenerational plasticity in plants: case studies,

mechanisms, and implications for natural populations. Frontiers in Plant Science, 2, 102.

Howe, H.F. & Brown, J.S. (2001) Ghosts of granivory past. Ecology Letters, 4, 371-378.

Hudson, P.J. & Catadorri, I.M. (1994) The Moran effect: a cause of population synchrony.

Trends in Ecology and Evolution, 14, 1-2.

Hurtt, G.C. & Pacala, S.W. (1995) The consequences of recruitment limitation: reconciling

chance, history and competitive differences between plants. Journal of Theoretical

Biology, 176, 1–12.

Hutchinson, G.E. (1957) Concluding remarks. Cold Spring Harbor Symposium of Quantitative

Biology, 22, 415-427.

Hutchinson, G.E. (1959) Homage to Santa Rosalina, or, why are there so many animals?

American Naturalist, 93, 145–159

Johnson, R.H. (1910) Determinant evolution in the color pattern of the lady-beetles. Carnegie

Institution of Washington, Washington.

Klingenberg, C.P. & Gidaszewski, N.A. (2009) Testing and quantifying phylogenetic signals and

homoplasy in morphometric data. Systematic Biology, 59, 245-261.

Kneital, J.M. & Chase, J.M. (2004) Trade-offs in community ecology: linking spatial scales and

species coexistence. Ecology Letters, 7, 69-80.

Knicks, S.T. & Rotenberry, J.T. (2000) Ghosts of habitats past: contribution of landscape change

to current habitats used by shrubland birds. Ecology, 81, 220-227.

Leibold, M.A., Holyoak, M., Mouquet, N., Amarasekare, P., Chase, J.M., Hoopes, M.F., Holt,

R.D., Shurin, J.B., Tilman, D., Loreau, M. & Gonzalez, A. (2004) The metacommunity

concept: a framework for multi-scale community ecology. Ecology Letters, 7, 601-613.

Levine, J.M. & HilleRisLambers, J. (2009) The importance of niches for the maintenance of

species diversity. Nature, 461, 254–257.

Lowe-Decarie, E., Kolber, M., Homme, P., Lofanom A., Dumbrell, A., Gonzalez, A. & Bell, G.

(2015) Community rescue in an experimental metacommunity. Proceedings of the

National Academy of Sciences, 112, 14307-14312.


11

MacDougall, A.S., Gilbert, B. & Levine, J.M. (2009) Plant invasions and the niche. Journal of

Ecology, 97, 609–615.

Mayfield, M.M. & Levine, J.M. (2010) Opposing effects of competitive exclusion on the

phylogenetic structure of communities. Ecology Letters, 13, 1085-1093.

MacArthur, R. & Levins, R. (1967) The limiting similarity, convergence, and divergence of

coexisting species. American Naturalist, 101, 377–385.

Metz, J., von Oppen, J. & Tielbörger, K. (2015). Parental environmental effects due to

contrasting watering adapt competitive ability, but not drought tolerance, in offspring of a

semi-arid annual Brassicaceae. Journal of Ecology, 103, 990–997.

Miao, S.L., Bazzaz, F. & Primack, R. (1991) Persistence of maternal nutrient effects in Plantago

major: the third generation. Ecology, 72, 1634-1642.

Mooney, H.A. & Cleland, E.E. (2001) The evolutionary impact of invasive species. Proceedings

of the National Academy of Sciences, 98, 5446-5451.

Mouquet, N. & Loreau, M. (2003) Community patterns in source-sink metacommunities.

American Naturalist, 162, 554-557.

Mousseau, T.A. & Fox, C.W. (1998) Maternal Effects as Adaptations. Oxford University Press,

New York, NY.

Narwani, A., Alexandrou, M.A., Oakley, T. H., Carroll, I.T. & Cardinale, B.J. (2013)

Experimental evidence that evolutionary relatedness does not affect the ecological

mechanisms of coexistence in freshwater green algae. Ecology Letters, 16, 1373–1381.

Pulliam, H.R. (2000) On the relationship between niche and distribution. Ecology Letters, 3,349-

361.

Rajaniemi, T.K., Goldberg, D.E., Turkington, R. & Dyer, A.R. (2005) Quantitative partitioning

of regional and local processes shaping regional diversity patterns. Ecology Letters, 2006,

121-128.

Roach, D.A. & Wulff, R.D. (1987) Maternal effects in plants. Annual Review of Ecology and

Systematics, 18, 209–235.

Silvertown, J., McConway, K, Gowing, D., Dodd, M., Fay, M.F., Joseph, J.A. & Dolphin, K.

(2006) Absence of phylogenetic signal in the niche structure of meadow plant

communities. Proceedings of the Royal Society B: Biological Sciences, 273, 39-44.

Simberloff, D. & Boecklen, W. (1981) Santa Rosalia reconsidered: size ratios and competition.

Evolution, 35, 1206-1228.

Strauss, S.Y., Webb, C.O. & Salamin, N. (2006) Exotic taxa less related to native species are

more invasive. Proceedings of the National Academy of Sciences, 103, 5841–5845.


12

Sultan, S.E., Barton, K. & Wilczek, A.M. (2009) Contrasting patterns of transgenerational

plasticity in ecologically distinct congeners. Ecology, 90, 1831-1839.

Tilman, D. & Pacala, S. (1993) The maintenance of species richness in plant communities.

Species Diversity in Ecological Communities (eds. Ricklefs, R.E. & Schluter, D.), pp. 13-

25. University of Chicago Press, Chicago.

Tilman, D. (1997) Community invisibility, recruitment limitation, and grassland biodiversity.

Ecology, 78, 81-92.

Tuomisto, H., Ruokolainen, K. & Yli-Halla, M. (2003) Dispersal, environment, and floristic

variation of Western Amazonian forests. Science, 299, 241-244.

Vamosi, S.M., Heard, S.B., Vamosi, J.C. & Webb, C.O. (2009) Emerging patterns in the

comparative analysis of phylogenetic community structure. Molecular Ecology, 18, 572–

592.

Van Allen, B.G. & Rudolf, V.H.W. (2013) Ghosts of habitats past: environmental carry-over

effects drive population dynamics in novel habitat. American Naturalist, 181, 596–608.

Van Allen, B.G. & Rudolf, V.H.W. (2015) Habitat-mediated carry-over effects lead to context-

dependent outcomes of species interactions. Journal of Animal Ecology, 84, 1646–1656.

Webb, C.O., Ackerly, D.D., McPeek, M.A. & Donoghue, M.J. (2002) Phylogenies and

community ecology. Annual Review of Ecology and Systematics, 33, 475–505.

Weiher, E., Freund, D., Bunton, T., Stefanski, A., Lee, T. & Bentivenga, S. (2011) Advances,

challenges and a developing synthesis of ecological community assembly theory.

Philosophical Transactions of the Royal Society B: Biological Sciences, 366, 2403–2413.

Wolkovich, E.M., Cook, B.I., McLauchlan, K.K. & Davis, T.J. (2014) Temporal ecology in the

Anthropocene. Ecology Letters, 17, 1365-1379.

CHAPTER 2: MATERNAL EFFECTS TO ABIOTIC CONDITIONS

13

Chapter 2

Hidden responses to environmental variation: maternal effects

reveal species niche dimensions

Published as Germain, R. M. & B. Gilbert. 2014. Hidden responses to environmental variation:

maternal effects reveal species niche dimensions. Ecology Letters 17:662-669.

Abstract

Species responses to fluctuating environments structure population and community dynamics in

variable ecosystems. Although offspring number is commonly used to measure these responses,

maternal effects on offspring quality may be an important but largely unrecognized determinant

of long-term population growth. We selected 29 species across a Mediterranean annual plant

phylogeny, and grew populations of each species in wet and dry conditions to determine

responses in seed number and maternal effects (seed size, seed dormancy, and seedling growth).

Maternal effects were evident in over 40% of species, but only 24% responded through seed

number. Despite a strong trade-off between seed size and seed number among species, there was

no consistent trade-off within species; we observed correlations that ranged from positive to

negative. Overall, species in this plant guild show a complex range of responses to

environmental variation that may be underestimated when only seed number responses are

considered.

Introduction

Environmental variability is ubiquitous in nature (Vasseur & Yodzis 2004), and plays a central

role in determining the structure and function of many biological communities (Hobbs et al.

2007; Elmendorf & Harrison 2009). When species differ in their responses to the environment,

this variability can maintain biodiversity by favouring different species in different years

(Chesson & Warner 1981; Levine & Rees 2004; Adler et al. 2006; Angert et al. 2009). As a

result, ecologists have devoted significant resources to understanding species-specific responses

to varying environments (Angert et al. 2010; Gremer et al. 2012).


14

Most research on the impacts of environmental variability has focused on numeric responses,

such as the per capita number of offspring produced (Caceres 1997; Levine & Rees 2004; Angert

et al. 2009). However, there is strong evidence that the environment not only affects offspring

quantity, but also offspring quality (Mousseau & Fox 1998), in a phenomenon known as

‘maternal effects’ (Roach & Wulff 1987). A maternal effect is “any phenotypic variation in

offspring that is a consequence of the mother’s phenotype rather than the genetic constitution of

the offspring” (Roff 1998). In plants, for example, increased seed provisioning can alter early

survivorship and competitive ability (Jakobsson & Eriksson 2000; Moles & Westoby 2004).

These changes in offspring quality have the potential to generate responses to environmental

variability that are distinct from changes in offspring quantity, and may therefore play an

important but relatively unexplored role in community dynamics (Ginzburg & Taneyhill 1994).

Previous research has revealed several cases where maternal effects can alter the fitness or

ecology of individuals (e.g., Galloway 2005; Sultan et al. 2009; Dyer et al. 2010). However, the

general importance of maternal effects for ecological dynamics is unknown for two reasons.

First, many studies of maternal effects evaluate their impact on genetically similar individuals,

such as closely related genotypes or clones (e.g., Galloway & Etterson 2007). While these tests

are relevant to the ecology of asexual organisms, it is not clear whether individual-level trends

would be swamped by population-level variation when considered in outcrossing species. In

other words, maternal effects need to be studied at the population level to determine if they are

comparable to demographic effects. Second, most studies of maternal effects include one or two

species that show strong responses (e.g., Galloway 2005). Multi-species tests of maternal effects

are currently lacking, and as a result it is unclear how common they are in natural communities.

In addition to quantifying the prevalence of maternal effects, multi-species tests also allow us to

identify key factors that account for variation in maternal effects expression among species.

Phylogenetic relationships are a likely candidate, given that they are known to structure plasticity

(Kembel & Cahill 2005; Gomez-Mestre et al. 2008) and seed size (e.g., Lord et al. 1995).

Moreover, contemporary hypotheses in the field of phylogenetic community ecology posit that

similar environmental responses among closely related species structures species coexistence

(Mayfield & Levine 2010), suggesting that understanding the phylogenetic basis of numeric

responses and maternal effects should provide insights into community assembly and diversity.


15

Annual plants are ideal for contrasting numeric population responses with maternal effects

because lifetime reproduction can be entirely captured in a single growing season. As such, the

number of offspring produced is estimated by seed number, whereas maternal effects manifest

through changes to seed quality, such as seed size. These differences in seed size often confer

advantages in early growth, survival and competitive ability among species (Leishman 2001;

Moles & Westoby 2002; Moles & Westoby 2004), and have been shown to change the fitness of

individuals (Germain et al. 2013). In addition, studies of seed size and seed number across

species show a consistent trade-off (e.g., Jakobsson & Eriksson 2000; Levine and Rees 2002).

This trade-off can also occur within species (e.g., Germain et al. 2013), but the opposite pattern

has also emerged (Venable 1992), raising the possibility that maternal effects can counteract or

reinforce a species’ numeric response to environmental fluctuations depending on whether they

are positively or negatively correlated to seed number.

In this study, we use annual plants from Mediterranean climate regions to test the consequences

of variation in water availability on seed production. We chose this system because the high

diversity of annual plants in Mediterranean regions is at least partially maintained by species-

specific responses to variable rainfall (e.g., Hobbs et al. 2007; Elmendorf & Harrison 2009). In

addition, several studies have demonstrated the importance of maternal effects in Mediterranean

annuals (e.g., Byers et al. 1997; Germain et al. 2013), but it is unknown how commonly maternal

effects occur across a range of species.

To test the general importance of maternal effects relative to numeric responses, we selected 29

species of both closely and distantly related taxa. We grew these species in a greenhouse under

wet and dry soil moisture conditions, and estimated the size and number of seeds produced, as

well as dormancy and offspring performance. We used these data to address three questions: (1)

How commonly do plants respond to environmental variation with maternal effects, and how

does this compare to numeric responses? (2) Is there a consistent relationship between seed size

and seed number responses within species, and does this relationship tend to be counteracting

(negative correlation) or reinforcing (positive correlation)? and (3) Does phylogenetic

relatedness explain the type and strength of responses to environmental variation?


16

Materials and methods

Study system

We studied the effect of water availability in the maternal generation on offspring of 29 annual

plant species that are native to Mediterranean-climate regions. Mediterranean regions are

diversity hotspots, harbouring ~20% of the world’s total vascular plant species, many of which

are rare or endemic (Cowling et al. 1996). These regions are characterized by cool winters and

hot summers, where most of the annual precipitation falls in the autumn and winter months

(Cowling et al. 1996). The timing and total amount of rainfall can vary up to six-fold between

years (Hobbs et al. 2007), potentially explaining the high turnover in both the density and

composition of species over time in these systems (Cowling et al. 1996; Levine & Rees 2004;

Hobbs et al. 2007; Elmendorf & Harrison 2009).

The species were selected to meet three criteria. First, all species have an annual lifecycle to

ensure that seed production in a single growing season is representative of that individual’s

lifetime reproductive success. Second, each species occurred in annual grasslands with similar

climatic conditions (Leiva et al. 1997) and overlapped geographically with other species in the

study as determined by CalFlora (http://www.calflora.org) and Euro+Med

(http://www.emplantbase.org) plant databases. These criteria ensured that our species had similar

exposure to variation in rainfall over recent evolutionary timescales. Third, species were selected

to span seven Orders [Asterales, Boraginales, Caryophyllales, Fabales, Lamiales, Poales, and

Ranunculales; Table S2.1], to capture the taxonomic diversity of annual species from

Mediterranean-climate regions. Seeds were acquired from commercial suppliers and an

individual donor, and originated from natural populations relatively few generations (usually <3,

all <20) before experimentation (Table S2.1). Although we suspect that our seed sources include

levels of genetic variation typical of natural populations, we are unable to test this assumption

because of a lack of knowledge of the genetic diversity both in the wild and in our populations.

Methods for ensuring genetically representative samples can be found elsewhere (e.g., Project

Baseline (http://www.baselineseedbank.org/)), and remain an important goal for future studies.

We did not cold stratify seeds prior to planting, because doing so is known to induce dormancy

in Mediterranean species (Baskin & Baskin 2001).

http://www.calflora.org/

http://www.emplantbase.org/

http://www.baselineseedbank.org/


17

Maternal generation

We grew the maternal generation in a greenhouse at the University of Toronto from January to

July 2012. The greenhouse was initially set to day/night temperatures of 14/7˚C, with

supplemental high intensity discharge (HID) lighting provided to maintain a 10 h day length.

These conditions gradually increased to a 23/17 ˚C day/night cycle and 14-h day length to

simulate a typical Mediterranean growing season (Fig. S2.1). For each species, 30 seeds were

sown into each of fourteen replicate 2.54-L treepots filled with a 3:2 mixture of sand and

screened topsoil to mimic the sandy loam soils that dominate annual grasslands. Pots were top-

watered daily until saturation during a three week establishment period.

After establishment, we recorded the number of seedlings and thinned each pot to ~eight

individuals. Seedlings were selected systematically to maximize their distance from each other in

a pot, and were thinned by clipping the base of the stem. This low-density environment allowed

us to evaluate the direct effect of soil moisture on a population of individuals, rather than indirect

effects of competition. We randomly assigned pots of each species to receive either a wet or dry

soil moisture regime through drip irrigation lines. Pots in the wet treatment received 175 mL

water twice as often as those in the dry treatment, starting at one and two day intervals which

were extended to seven and 14 days as the growing season progressed (Fig. S2.1). These two soil

moisture regimes represent realistic among-year differences in precipitation, or equally, the 30

year average differences between mesic [662 mm] and xeric [312 mm] sites across the species

ranges in California (http://www.climate-charts.com). We confirmed that the wet (11.1 ± 0.56%;

mean ± SE percent soil moisture content) and dry (5.8 ± 0.42%) treatments were effective using

a volumetric water probe on empty pots. Prior to flowering (~60 days after planting), each pot

was provided with 350 mL of 1500 ppm 20-20-20 NPK fertilizer. Pollination was provided by

commercial colonies of Bombus impatiens, which were active throughout the flowering period.

Bombus impatiens are generalist pollinators, and were observed visiting all insect-pollinated

species in this experiment.

We monitored pots daily, and collected seeds as they matured on the parent plants. Seed samples

were processed in the lab by subsampling collected materials to estimate seed size and seed

number. Subsampling was conducted using a standardized procedure in which each sample was


18

sequentially divided into equal smaller units until manageable subsample sizes were obtained. A

single subsample was then chosen randomly. Seed size was estimated as the mass of all seeds in

the subsample divided by number of seeds in the subsample, whereas seed number was

calculated as the average number of seeds produced per plant in each pot.

Offspring generation

We performed germination and greenhouse experiments on the offspring generation from

January to April 2013. The greenhouse maintained a 15/7˚C day/night temperature regime, with

supplemental high intensity discharge (HID) lighting provided to maintain a 10-h day length

throughout the experiments.

The germination experiment tested for maternal effects on seed viability and dormancy, with the

latter indicating allocation to the seed bank as a bet-hedging strategy for persistence in

unfavourable years (Tielbörger & Valleriani 2005). Thirty seeds from each maternal pot were

haphazardly selected and placed on moist filter paper in a petri plate, with 7 petri plates used per

species by maternal environment combination. The plates were placed on a greenhouse bench

under a shade cloth and scored for germination every two days, and remoistened as needed. After

24 days, once germination plateaued for the majority of species, ungerminated seeds were air-

dried, transferred to fresh filter paper, and moistened with a 500 ppm gibberellic acid solution

(GA3). This naturally occurring plant growth regulator is commonly used to germinate dormant

seeds. Because GA3 application is not always 100% effective for deeply dormant seeds, it

provides a conservative estimate of dormancy (Baskin & Baskin 2001). Seed viability was

scored as the proportion of seeds that germinated under regular and GA3 methods; other viability

methods, such as tissue staining, could not be used due to the very small seed sizes of some

species. Mold growth was spot treated with 0.15% v/v of Previcur® fungicide throughout the

germination trial.

We also tested for maternal effects on offspring performance at the seedling stage. For each petri

plate in the germination trial, two germinants were transplanted into each of two 0.16-L

conetainers filled with a peat-based commercial soil (Sunshine Mix 1, Sun Gro Horticulture

Canada CM Ltd.). This produced fourteen conetainers per species × water level treatment, but

low germination in four species led to unbalanced replication among species. After three weeks,


19

we measured the height and harvested the above-ground biomass of one randomly-selected

individual per conetainer. Many individuals had very low dry mass (≤ 1 mg), which increased

the relative measurement error; we therefore used the height data to test for offspring

performance.

Phylogenetic tree construction

We constructed a phylogenetic tree for 27 of our 29 species using Bayesian methods on

nucleotide sequence data. Two species (Salvia columbariae and S. viridis) were excluded

because sequences were unavailable. Sequences from the ITS1/5.8S/ITS2 coding region were

retrieved from GenBank, aligned using the MUSCLE algorithm in MEGA (version 4.0; Tamura

et al. 2007) and combined into a single sequence matrix for analysis. We then generated a

Bayesian tree in BEAST (version 1.7; Drummond et al. 2002) using a GTR substitution model

selected by ModelTest (version 3.7; Posada & Crandall 1998) and an uncorrelated lognormal

relaxed molecular clock for branch length estimation. Because this coding region is rapidly

evolving, we constrained the tree topology at nodes above the family-level based on well-

resolved relationships in the angiosperms (Stevens 2001). We ran our MCMC chain for

100,000,000 generations, with a 10,000,000 burn-in and data logging every 10,000 generations.

Our analysis generated 10,000 posterior trees, which were summarized into a single maximum

clade credibility tree with median node heights. The resulting tree (Figs. 2.1 and S2.2) is fully

resolved and consistent with Angiosperm Phylogeny Group III classifications (Stevens 2001).

Data analysis

We used generalized least squares (GLS) analyses to test the effects of our maternal watering

treatments on all response variables that were not binomially distributed; those that were

binomially distributed (seed dormancy and viability) were tested with generalized linear models

(GLMs) using a quasibinomial distribution to account for overdispersion. The GLS allows for

heterogeneous variances among treatments, with the varIdent weighting function to account for

the different variances among species. In all analyses, we tested for the effect of maternal

environment (ME) and a species × ME interaction, with significant results indicating common

responses or species-specific responses to the maternal environment respectively.


20

In all analyses where a significant species × ME interaction occurred, we used post-hoc tests to

determine the species that differed and the direction of effect. The large number of species used

in the study made corrections to post-hoc comparisons impractical. As a result, we restrict our

post-hoc tests to comparisons between environments for each species, and present our data in

terms of effect sizes to quantify the magnitude of change relative to the variation. Our effect

sizes were measured using Cohen’s d with a pooled variance (i.e., (µdry - µwet)/pooled σ; Hartung

et al. 2011). Because the dormancy tests had higher variances than expected for proportions, we

used Cohen’s d for these effect sizes as well.

We tested for maternal effects on seed size, seed number, and also seed size-seed number

correlations within and among species. All tests with seed size and seed number were performed

on log-transformed data. The overall test of the correlation between seed size and seed number

within species was done by first standardizing seed size and seed number within species and

performing a regression that included a species × ME interaction. This standardization causes the

slope of the regression line to be equal to the correlation coefficient, and the test therefore

determines if any correlation between seed size and seed number is the same (or different) for

our species (Quinn & Keough 2002). The post-hoc test on seed size-seed number correlations

was performed using Pearson correlations for each species, and the effect size presented is the

correlation coefficient. The correlation among species was calculated using mean values by

species for both seed size and seed number.

Maternal effects were measured in the following generation by measuring seed dormancy,

viability and seedling height. For seedling height, we first tested for differences in height using

GLS, and subsequently tested for whether populations that produced larger seeds tended to

produce taller plants. For this second analysis, we first standardized data within each species,

using mean seed size per pot and mean height of offspring from a given pot. We then performed

a regression that included a species × ME interaction. As with the correlation test above, this test

determines if species differ in the correlation between mean seed size and the height of seedlings.

We tested for a phylogenetic signal in species responses to the maternal environment using the

‘phylosignal’ function in the picante package of R (Kembel et al. 2010), to determine if closely

related species are responding more or less similarly to environmental variation. First, when the


21

species × ME terms were significant, we calculated effect sizes for each species, except the seed

size-seed number correlation, for which we used the correlation coefficients. Second, we

contrasted observed phylogenetic independent contrasts (PICs) to those generated by

randomizing the tips of the tree (10000 randomizations). The rank of the observed contrasts

relative to the randomized contrasts was used to generate p-values, with p-values approaching

one or zero indicating that responses were convergent or conserved, respectively, with a null

expectation of a Brownian motion model of evolution. Because we used the randomizations for

two-tailed tests, the p-values reported are 2*rank observed/10000 for conserved responses, and

2*(1 - rank observed/10000) for convergent responses.

Results

Seed size, seed number, and their correlation

Close to half of species (45%) responded to environmental variation by altering seed size or seed

number, with these two responses being equally common (Fig. 2.1a,b). We detected a significant

maternal environment × species interaction for both responses (seed size: F28,331 = 2.20, P <

0.001, seed number: F27,320 = 2.39, P < 0.001), with seed size responses significant in six of 29

species and seed number responses significant to marginally significant in seven of 28 species (P

= 0.05-0.08 in two of those species). Interestingly, there was no overlap in the species that

responded through seed size and those that responded through seed number. Despite a significant

interaction, we found that seed number decreased by 85% on average in the dry environment

(F1,320 = 20.44, P < 0.001), and only one species significantly increased seed number in this

treatment (Fig. 2.1b). In contrast, seed size showed no mean effect of watering treatment (F1,331 =

2.20, P = 0.136; Fig. 2.1a), meaning that species were equally likely to increase or decrease seed

size in response to moisture.


22

Figure 2.1. Effect sizes of (A) seed size, (B) seed number, and (C) seed dormancy responses to

our watering treatments across a phylogeny of 29 species; see Table S1 for species codes. Effect

sizes were calculated using Cohen’s d (i.e., (µdry - µwet)/pooled σ; Hartung et al. 2011). Positive

effect sizes indicate that trait values were higher in the dry compared to wet environment, and

vice versa. The numbers on the phylogeny correspond to taxonomic orders: (1) Poales, (2)

Ranunculales, (3) Fabales, (4) Caryophyllales, (5) Boraginales, (6) Lamiales, and (7) Asterales.

Dashed branches represent unknown relationships due to missing sequence data [SCOLU and

SVIRI], and ‘×’ indicates unavailable trait data rather than a true zero. * P < 0.05, º P < 0.10.


23

There were significant seed size-seed number correlations, but their direction depended on

whether our focus was among or within species (Fig. 2.2). At the among-species level, seed size

and seed number were negatively correlated (r = -0.67, P < 0.001; Fig. 2.2a), and the slope of

this relationship was not statistically different from -1 (P > 0.05), meaning that the relationship

can be considered linear and is therefore consistent with hypothesized energetic trade-offs

(Levine & Rees 2002). Within species, however, we observed both positive and negative

correlations ranging from r = 0.76 to r = -0.65 (Fig. 2.2b). Overall, there were a similar number

of significant positive (three of 28 species) and negative (two of 28 species) correlations, with

most species having smaller correlation coefficients (-0.5 < r < 0.5) that were not statistically

significant.

We found some support of convergent evolution in the response of seed size, with closely related

species responding less similarly than expected by chance (Blomberg’s K = 0.09, P = 0.057).

This was particularly evident with Vulpia, where one species was less similar to its congeners

than to the other grasses (Fig. 2.1a), but similar patterns were seen in other closely related taxa

(e.g., Fabaceae and Boraginaceae). There was no phylogenetic signal in species responses

through seed number (Blomberg’s K = 0.22, P = 0.86) or seed size-seed number correlations

(Blomberg’s K = 0.33, P = 0.14).


24

Figure 2.2. Seed size-seed number correlations. (A) Correlation among species, with values

representing the species-level means with a best-fit line indicating a significant correlation (R2 =

0.45, P < 0.001). (B) Correlations within species; seed size and seed number were standardized

by species and a constant added to visualize the correlation coefficients for each species. Black

and dark grey lines represent significant (P < 0.05) and marginally significant (P < 0.10)

correlations, respectively; light grey lines are non-significant. The dashed line represents no

correlation (r = 0). All data were log transformed (A) prior to standardization (B).

Offspring performance

The maternal environment had significant effects on seedling height after three weeks of growth

(F1,575 = 15.92, P < 0.0001), with a significant maternal environment × species interaction

(F25,575 = 2.03, P = 0.003; Fig. S2.2). The height responses were only partially consistent with

the seed size effects—one species that had significantly larger seeds in the dry environment


25

(Hordeum vulgare) also had taller offspring when grown from those seeds in the dry

environment. However, this was not the case for many species that showed seed size effects;

only 3 of 26 species showed significant or marginally significant (P = 0.06 for one species)

responses. Nonetheless, a subsequent analysis revealed a positive correlation between seed size

and mean plant height (r = 0.11, P = 0.047, Fig. S2.3), which was consistent across species (non-

significant seed size × species interaction; P = 0.52).

We detected maternal effects on seed dormancy that varied by species (significant maternal

environment × species interaction; F26,282 = 1.96, P = 0.004), with changes to dormancy being

equally frequent as changes to seed size or seed number (Fig. 2.1c). Although seeds tended to

have higher dormancy when the maternal plant was raised in wet environments (six of seven

significant responses; F1,334 = 7.1, P = 0.008), there was one species that showed a significant,

opposite trend. The maternal environment caused large changes to dormancy for some of these

species (from 20-48% for Vulpia octoflora and from 3-9% for Lotus purshianus), whereas other

species shifted from no detectable dormancy to some dormancy (1-9%). Maternal environment

had no detectable effect on seed viability (P = 0.73). There was no phylogenetic signal for

maternal effects on dormancy (Blomberg’s K = 0.194, P = 0.998) or seedling height (Blomberg’s

K = 0.154, P = 0.650).

Discussion

We have shown that the effects of environmental variation on the ecological responses of species

are greatly underestimated when maternal effects are not considered. Ecologically important

maternal effects, such as changes in seed size and dormancy rates, are not only common, but

occur at similar frequencies as changes in seed number (Fig. 2.1). Because these maternal effects

produce responses at the population level, they likely influence long-term growth rates and

stability of populations in ways that cannot be predicted by measuring population growth rates

based on seed number alone. This study represents the first comprehensive investigation of

maternal effects that spans a guild of ecologically similar species, and demonstrates that

ecologists should consider these effects on population dynamics, species coexistence, and

phylogenetic patterns on the landscape.


26

We detected maternal effects on seed size directly, as well as associated differences in

performance. Seed size is known to influence the realized niches of species, with larger seeds

conferring competitive advantages and also contributing to survival in stressful conditions such

as light or nutrient limitation (reviewed in Westoby et al. 1996). We found supporting evidence

that within-species variation in seed size confers a weak but significant advantage to early

growth even when seedlings are grown in benign, non-competitive environments (Fig. S2.3), as

has been demonstrated in cross-species comparisons (Jurado & Westoby 1992). Previous work

on interspecific differences in seed size suggest that the advantages associated with larger seeds

may be less than an equal investment in seed number, with every x-fold increase in seed size

producing roughly x0.6

-fold increase in per capita success in the following generation (Freckleton

& Watson 2001; Levine & Rees 2002). However, the authors of these studies highlight that this

estimate provides a guideline at best; further research on the nature of seed size advantages and

their importance in different offspring environments is needed both within and among species

(Leishman 2001; Moles & Westoby 2004; Germain et al. 2013).

Increased investment in seed size may come at the cost of decreased seed number, a trade-off

that is commonly observed in plants. We only found strong supporting evidence for this trade-off

among species (Fig. 2.2a), which is consistent with previous studies (e.g., Jakobsson & Eriksson

2000). Within species, however, seed size-seed number correlations ranged from positive to

negative (Fig. 2.2b), suggesting that different constraints may be operating at different levels of

ecological organization. For example, Venable (1992) predicted that positive correlations can

arise if variation exists in the resource environment or the size of individual genotypes in a

population. Because changes in soil moisture conditions could induce changes to both seed size

and number, our experimental set-up was well-suited to producing a range of correlations. The

small sizes of our experimental populations (~eight individuals) likely also increased the

likelihood of high-yielding genotypes having large impacts on population-level responses. As

such, the variety of correlations between seed size and seed number likely reflects the combined

effects of species’ allocation strategies across environments as well as individual-level variation

within species.

The clearest signal of the maternal environment on offspring performance was through changes

in dormancy rates. Dormancy is known to have a large impact on population dynamics, but is


27

often thought to be highest with appropriate environmental signals (e.g., Angert et al. 2009) or to

be relatively constant when growing conditions vary unpredictably (Cohen 1966). The finding

that dormancy tended to decreased in response to a dry maternal environment was surprising, as

dormancy is generally thought to increase in response to resource limitation, such as water

availability (Tielbörger & Valleriani 2005). One potential explanation is that other reproductive

responses may influence dormancy. For example, a negative correlation between seed size and

dormancy is predicted by models of bet hedging when large seeds help ensure seedling success

in poor years (Venable 2007). We found limited support for this hypothesis, with approximately

one third of species that responded by changing dormancy rates also responded by altering seed

size (Fig. 2.1a,c). For these species, parents in dry soils produced larger seeds with lower

dormancy rates.

The maternal effects observed in our study have important implications for species coexistence

that are distinct from population-level implications. In particular, coexistence of annual plants in

temporally variable environments requires that (i) species respond differently to environmental

conditions, (ii) they have storage dynamics via seed dormancy, and (iii) species experience a

positive covariance between reproductive potential and intraspecific competition, which allows a

rare species to have high reproductive rates in good years (Chesson 1994; Abrams et al. 2013).

The maternal effects that we observed influence each of these conditions. For example, changes

to seed size doubled the number of species showing significant responses to soil moisture and

generated almost all of the positive responses to drought, meeting condition (i). Similarly, the

effect of the maternal environment on dormancy rates differentially alters both storage (condition

(ii)) and germination of species, with the latter contributing to coexistence by altering condition

(iii; e.g., Angert et al. 2009). Finally, condition (iii) can be met even in species with constant

germination rates so long as the per capita competitive effect and reproductive potential of each

species respond similarly to environmental conditions (Chesson 1994; Chesson et al. 2004;

Abrams et al. 2013). Seed size may generate this covariance, as large seeds have been shown to

produce plants that have a greater competitive effect and higher fecundity (Freckleton &

Watkinson 2001; Levine & Rees 2002).

The phylogenetic signature of maternal effects also offers new insights into the processes that

alter patterns of local species coexistence. Theoretical work predicts that phylogenetic


28

convergence in competitive traits, such as seed size, should deter local coexistence between

closely related species (Mayfield & Levine 2010). In temporally fluctuating environments,

however, the phylogenetic convergence that we observed in seed size responses to soil moisture

may actually promote coexistence between close relatives by favouring different species in

different years, as has been predicted by theory (Abrams et al. 2013). This intricate matching of

competitive ability and environmental conditions leads to contrasting predictions for

phylogenetic effects on coexistence depending on the timescale considered (i.e., single vs.

multiple years). Moreover, the phylogenetic signal observed was only apparent in seed size

responses, suggesting that this maternal effect likely contributes to phylogenetic patterns that

emerge from species interactions in this guild.

Although we have shown that maternal effects are common in this annual guild, there are four

important caveats to our conclusions. First, our study only measured a subset of maternal effects

by quantifying three responses to environmental variation: seed size, seed dormancy, and

seedling growth. Other studies have found that the maternal environment can also alter seed

composition, seed coat characteristics, and epigenetics (Roach & Wulff 1987; Donohue 2009).

As a result, our results likely underestimate the prevalence of maternal effects. Second, we

examine only two environmental conditions. Previous research has shown diverse responses of

annual plants to the timing and patterns of variation in precipitation, for example, and

quantifying the overall importance of maternal effects requires tests across a broader range of

environmental variation (e.g., Hobbs et al. 2007). Third, because we obtained seed from

commercial suppliers (Table S2.1), it is unclear how representative our seeds are of the genetic

diversity found in natural populations. Future work on the ecological importance of maternal

effects should use a standardized seed collection procedure that provides a representative sample

of genetic diversity in populations. Finally, the long-term impacts of maternal effects on

population and community dynamics are speculative in our study, which highlights the need for

long-term studies that quantify these impacts.

In sum, the diversity of responses that we observed suggests that maternal effects can alter

species niches through several pathways, including seed size, dormancy, and offspring

performance. In evolutionary ecology, it is well recognized that maternal effects can affect the

fitness of individuals, often strongly enough to generate artificial signals of natural selection


29

(Roach & Wulff 1987; Donohue 2009). In ecology, however, maternal effects have yet to be

incorporated into most models of population growth and species interactions. Our study

demonstrates that this failure to incorporate maternal effects effectively eliminates over half of

the ways in which species respond to the environment. The varied responses that we observed

suggest that testing the range of maternal effects that function in any guild of species may be

necessary for understanding the complexity of interactions that promote diversity.

Acknowledgements

Many undergraduate students contributed to the completion of this project, most notably Alanna

Leale, Alexandra Mushka, and Christopher Blackford. We also thank Jason Weir for his

assistance with phylogenetic reconstruction, as well as Bruce Hall and Andrew Petrie for

greenhouse assistance, and three anonymous reviewers for providing comments on an earlier

version of this manuscript. Research funding was provided by NSERC (B.G.) and the Botanical

Society of America (R.M.G.) and personal funding was awarded by NSERC-CGS (R.M.G.).

Literature cited

Abrams, P.A., Tucker, C.M. & Gilbert, B. (2013) Evolution of the storage effect. Evolution, 67,

315-327.

Adler, P.B., HilleRisLambers, J., Kyriakidis, P., Guan, Q. & Levine, J.M. (2006) Climate

variability has a stabilizing effect on the coexistence of prairie grasses. Proceedings of

the National Academy of Sciences, 103, 12793-12798.

Angert, A.L., Huxman, T.E., Chesson, P. & Venable, D.L. (2009) Functional tradeoffs determine

species coexistence via the storage effect. Proceedings of the National Academy of

Sciences, 106, 11641-11645.

Angert, A.L., Horst, J.L., Huxman, T.E. & Venable, D.L. (2010) Phenotypic plasticity and

precipitation response in the Sonoran Desert winter annuals. American Journal of Botany,

97, 405-411.

Baskin, C.C. & Baskin, J.M. (2001) Seeds: Ecology, Biogeography, and Evolution of Dormancy

and Germination. Academic Press, San Diego, CA.

Byers, D.L., Platenkamp, G.A.J. & Shaw, R.G. (1997) Variation in seed characters in Nemophila

menziesii: evidence of a genetic basis for maternal effect. Evolution, 51, 1445-1456.

Caceres, C.E. (1997) Temporal variation, dormancy, and coexistence: a field test of the storage

effect. Proceedings of the National Academy of Sciences, 94, 9171-9175.


30

Chesson, P. (1994) Multispecies competition in variable environments. Theoretical Population

Biology, 45, 227-276

Chesson P.L. & Warner, R.R. (1981) Environmental variability promotes coexistence in lottery

competitive systems. American Naturalist, 117, 923-943.

Chesson, P., Gebauer, R.L.E., Schwinning, S., Huntly, N., Wiegand, K., Ernest, M.S.K. et al.

(2004) Resource pulses, species interactions, and diversity maintenance in arid and semi-

arid environments. Oecologia, 141, 236-253.

Cohen, D. (1966) Optimizing reproduction in a randomly varying environment. Journal of

Theoretical Biology, 12, 119-129.

Cowling, R.M., Rundel, P.W., Lamont, B.B., Arroyo, M.K. & Arianoutsou, M. (1996) Plant

diversity in mediterranean-climate regions. Trends in Ecology and Evolution, 11, 362-

366.

Donohue, K. (2009) Completing the cycle: maternal effects as the missing link in plant life

histories. Philosophical Transactions of the Royal Society of London: Biological

Sciences, 364, 1059-1074.

Drummond, A.J., Nicholls, G.K., Rodrigo, A.G. & Solomon, W. (2002) Estimating mutation

parameters, population history and genealogy simultaneously from temporally spaced

sequence data. Genetics, 161, 1307-1320.

Dyer, A.R., Brown, C.S., Espeland, E.K., McKay, J.M., Meimberg, H. & Rice, K.J. (2010) The

role of adaptive trans-generational plasticity in biological invasions of plants.

Evolutionary Applications, 3, 179-192.

Elmendorf, S.C. & Harrison, S.P. (2009) Temporal variability and nestedness in California

grassland species composition. Ecology, 90, 1492-1497.

Freckleton, R. & Watkinson, A. (2001) Predicting competition coefficients for plant mixtures:

reciprocity, transitivity and correlations with life‐history traits. Ecology Letters, 4, 348–

357.

Galloway, L.F. (2005) Maternal effects provide phenotypic adaptation to local environmental

conditions. New Phytologist, 166, 93-100.


Science, 318, 1134-1136.


stress-induced parental effects in an invasive annual grass. International Journal of Plant.

Sciences, 174, 886-895.

Ginzburg, L.R. & Taneyhill, D.E. (1994) Population cycles of forest Lepidoptera: a maternal

effect hypothesis. Journal of Animal Ecology, 63, 79-92.


31

Gomez-Mestre, I., Wiens, J.J. & Warkentin, K.M. (2008) Evolution of adaptive plasticity: risk-

sensitive hatching in Neotropical leaf-breeding treefrogs. Ecological Monographs, 78,

205-224.

Gremer, J.R., Crone, E.E. & Lesica, P. (2012) Are dormant plants hedging their bets?

Demographic consequences of prolonged dormancy in variable environments. American

Naturalist, 179, 315-327.

Hartung, J., Knapp, G. & Sinha, B.K. (2011) Statistical Meta-analysis with Applications. John

Wiley & Sons, Inc., Hoboken, NJ.

Hobbs, R.J., Yates, S. & Mooney, H.A. (2007) Long-term data reveal complex dynamics in

grassland in relation to climate and disturbance. Ecological Monographs, 77, 545–568.

Jakobsson, A. & Eriksson, O. (2000) A comparative study of seed number, seed size, and

recruitment in grassland plants. Oikos, 88, 494-502.

Jurado, E. & Westoby, M. (1992) Seedling growth in relation to seed size among species of arid

Australia. Journal of Ecology, 80, 407–416.

Kembel, S.W. & Cahill, J.F. (2005) Plant phenotypic plasticity belowground: a phylogenetic

perspective on root foraging trade-offs. American Naturalist, 166, 216-230.

Kembel, S.W., Cowan, P.D., Helmus, M.R., Cornwell, W.K., Morlon, H., Ackerly, D.D. et al.

(2010) Picante: R tools for integrating phylogenies and ecology. Bioinformatics, 26,

1463-1464.

Leishman, M.R. (2001) Does the seed size/number trade-off model determine plant community

structure? An assessment of the model mechanisms and their generality. Oikos, 93, 294-

302.

Leiva, M.J., Chapin III, F.S. & Fernandez Ales, R. (1997) Differences in species composition

and diversity among Mediterranean grasslands with different history – the case of

California and Spain. Ecography, 20, 97-106.

Levine, J.M. & Rees, M. (2002) Coexistence and relative abundance in annual plant

assemblages: the roles of competition and colonization. American Naturalist, 160, 452-

467.

Levine, J.M. & Rees, M. (2004) Effects of temporal variability on rare plant persistence in

annual systems. American Naturalist, 164, 350-363.

Lord, J., Westoby, M. & Leishman, M. (1995) Seed size and phylogeny in six temperate floras:

constraints, niche conservatism, and adaptation. American Naturalist, 146, 349-364.


phylogenetic structure of communities. Ecology Letters, 13, 1085-1093.


32

Moles, A.T. & Westoby, M. (2002) Seed addition experiments are more likely to increase

recruitment in larger-seeded species. Oikos, 99, 241-248.

Moles, A.T. & Westoby, M. (2004) Seedling survival and seed size: a synthesis of the literature.

Journal of Ecology, 92, 372-383.

Mousseau, T.A. & Fox, C.W. (1998) The adaptive significance of maternal effects. Trends in

Ecology and Evolution, 13, 403-407.

Quinn, G. & Keough, M.J. (2002) Experimental Design and Data Analysis for Biologists.

Cambridge University Press, Cambridge, UK.

Posada, D. & Crandall, K.A. (1998) MODELTEST: testing the model of DNA substitution.

Bioinformatics, 14, 817-818.


Systematics, 18, 209-235.

Roff, D.A. (1998) The detection and measurement of maternal effects. In: Maternal Effects as

Adaptations. (eds Mousseau, T.A., Fox, C.W.). Oxford University Press, New York, NY,

pp. 83-96.

Stevens, P.F. (2001) Angiosperm Phylogeny Website. Version 12, July 2012 [and more or less

continuously updated since].


plasticity in ecologically distinct congeners. Ecology, 90, 1831-1839.

Tamura, K., Dudley, J., Nei, M. & Kumar, S. (2007) MEGA4: molecular evolutionary genetics

analysis (MEGA) software version 4.0. Molecular Biology and Evolution, 24, 1596-1599.

Tielbörger, K. & Valleriani, A. (2005) Can seeds predict their future? Germination strategies of

density-regulated desert annuals. Oikos, 111, 235-244.

Vasseur, D.A. & Yodzis, P. (2004) The color of environmental noise. Ecology, 4, 1146-1152.

Venable, D.L. (1992) Size-number trade-offs and the variation of seed size with plant resource

status. American Naturalist, 140, 287-304.

Venable, D.L. (2007) Bet hedging in a guild of desert annuals. Ecology, 88, 1086-1090.

Westoby, M., Leishman, M., Lord, J., Poorter, H. & Schoen, D.J. (1996) Comparative ecology of

seed size and dispersal [and discussion]. Philosophilcal Transactions of the Royal Society

B, 351, 1309-1318.

CHAPTER 3: FREQUENCY-DEPENDENT MATERNAL EFFECTS

33

Chapter 3

Frequency-dependent maternal effects across species and

environments

In review as Germain, R. M., T. N. Grainger, N. T. Jones, & B. Gilbert. Frequency-dependent

maternal effects across species and environments. American Naturalist.

Abstract

Maternal provisioning of offspring in response to environmental conditions (“maternal effects”)

has been argued as the ‘the missing link’ in plant life histories. Although recent empirical

evidence suggests that maternal responses to abiotic conditions are common, there is little

understanding of the prevalence of maternal provisioning in response to conditions plants

commonly experience in nature, such as competition. We tested how competition in two soil

moisture environments affects maternal provisioning of offspring seed size by varying the

frequency of conspecifics from 90% (mainly intraspecific competition) to 10% (mainly

interspecific competition) for 15 pairs of annual plant species that differ in phylogenetic

relatedness and biogeographic histories. We found that conspecific frequency affected maternal

provisioning (seed size) in 48% of species, and that these responses included both increased

(20%) and decreased (24%) seed size. In contrast, 68% of species responded to the competitive

environment through changes in per capita fecundity (seed number), which generally decreased

as conspecific frequency increased. We found opposite patterns of frequency-dependent maternal

provisioning among species originating from different biogeographic regions, suggesting that

maternal provisioning might be one trait that differentiates native and non-native species. Given

the prevalence of maternal provisioning of seed size shown here, future study is needed to

understand how these effects contribute to population- and community-level dynamics.


34

Introduction

Ecologists are often interested in relating species performance to competitive interactions and

environmental conditions, with the common goal of understanding the factors that contribute to

biological diversity. Although performance can be evaluated in many ways, the predominant

measure is to quantify the number of offspring produced, with all offspring considered identical

in quality (e.g., Law & Watkinson 1987; Levine & HilleRisLambers 2009). However, variation

in offspring quality in response to maternal environmental conditions is well-documented in

plants and animals, through a phenomenon known as ‘maternal effects’ (Roach & Wulff 1987;

Mousseau & Fox 1998). A maternal effect is the influence of the maternal environment on

offspring phenotypes, rather than or in addition to the influence of the maternal genotype

(Mousseau & Fox 1998). Although the impact of such maternal effects on demographic rates and

community dynamics is an open question (Galloway & Etterson 2007; Kendall et al. 2011),

overlooking maternal effects by focusing solely on offspring quantity has the potential to obscure

important ecological dynamics (Ginzburg & Taneyhill 1994; Benton et al. 2001).

Maternal effects on offspring quality commonly manifest as changes in the size of propagules,

such as seeds or eggs, and result from the maternal provisioning of resources to individual

offspring (Mousseau & Fox 1998; Herman & Sultan 2011). The effects of propagule size on

individual success are often short-lived, primarily manifesting at the juvenile stage; less

frequently they can persist to offspring maturity (Germain et al. 2013) or for multiple

generations (Miao et al. 1991). Regardless of the duration of persistence, maternal provisioning

of offspring size at the propagule stage can provide early life advantages or disadvantages that

affect life histories (Segers & Taborsky 2011; Allen 2012) and components of population

demography such as germination, dormancy rates, and juvenile survival (Westoby et al. 1996;

Smart & Moser 1999). These early life effects are especially important given that the filtering

effects of biotic and abiotic environments that select for some offspring over others can be most

apparent at the juvenile stage (Goldberg et al. 2001; Baldeck et al. 2013).

Although maternal effects are well-documented in response to abiotic conditions (see review by

Herman & Sultan 2011), the effect of competition on maternal provisioning is less understood,

particularly in plants (Weiner et al. 1997; Larios & Venable 2015). Maternal effects in response


35

to competition are likely more complex than responses to abiotic conditions for several reasons.

First, competition might interact with the abiotic environment to affect seed size if competitive

interactions alter resource availability or a species’ response to abiotic conditions. For example,

competitors may exacerbate maternal effects that are driven by a limiting resource (Stratton

1989), such as soil moisture (Fotelli et al. 2001). Most studies of maternal effects are conducted

in low-competition environments (e.g., Germain & Gilbert 2014) even though organisms rarely

occur in the absence of biotic interactors in nature, which suggests that current estimates of

maternal effects are conservative.

Second, the effect of competition on demographic rates often depends on the relative frequency

of conspecific to heterospecific individuals, even when total density is maintained (Law &

Watkinson 1987). The direction of frequency dependence has been used to identify whether

competition is most intense among conspecific (negative frequency dependence; e.g., Ayala

1971) or heterospecific (positive frequency dependence; e.g., Molofsky et al. 2001) individuals;

local coexistence is promoted when competing species show negative frequency dependence

(Levine & HilleRisLambers 2009). In the context of maternal provisioning, it is not known if or

when seed provisioning strategies reinforce or counteract frequency-dependent demographic

rates.

Lastly, the magnitude and direction of maternal effects might also depend on the biogeographic

histories and phylogenetic distances of competing species. Species with different biogeographic

histories (i.e., originating from different regions) often exhibit distinct strategies of resource use

and competitive ability (Funk & Vitousek 2007; Fridley 2012) due to differences in the

evolutionary pressures that exist in different regions (e.g., herbivory, climate; Mitchell et al.

2006). Similarly, phylogenetic distance is a proxy for evolutionary relatedness, with classic and

emerging theories suggesting that the intensity of competition among species is tied to their

evolutionary relatedness (Darwin 1859; Elton 1946; Mayfield & Levine 2010). It should be

noted, however, that unlike classic hypotheses (Darwin 1859; Elton 1946), recent theory does not

predict that competition should be most intense among closely related species. Rather,

contemporary theory predicts that the evolutionary rates of competitive differences determines

both the strength and direction of frequency-dependent demographic rates (Mayfield & Levine

2010; Godoy et al. 2014; Germain et al. 2016). It has recently been shown that frequency-


36

dependent demographic rates depend on biogeographic histories and phylogenetic distances of

competing species (Germain et al. 2016), but the influence of these differing competitive

environments on maternal provisioning of seed size has not been explored.

Although there is a long history of investigating maternal effects from an evolutionary

standpoint, their ecological implications have only recently begun to be explored (e.g., Larios &

Venable 2015; Van Allen & Rudolf 2015). In the evolutionary literature, a common goal has

been to understand how often maternal effects evolve as adaptations (Marshall & Uller 2007;

Mousseau et al. 2009), and to identify their genetic underpinnings (Mousseau & Fox 1998).

These types of evolutionary questions are typically tested in single-species experiments, using

replicated clones or genotypes to identify correlations among maternal effects expression and

fitness. In contrast, a common goal of ecological studies is to understand how prevalent maternal

effects are in natural systems (Germain & Gilbert 2014), and how they might impact population

and community dynamics (Ginzburg & Taneyhill 1994; Van Allen & Rudolf 2013; Van Allen &

Rudolf 2015). This ecological perspective requires that maternal effects be characterized at the

population or species level, rather than among genetically-similar individuals in a population.

Characterizing the strength and direction of maternal effects in an assemblage of species adds to

our understanding of how species interact with each other and their environments.

In this study, we use 25 annual plant species native to the mediterranean-climate regions of

California or Spain to test whether competition in the maternal generation results in maternal

effects on seed size. The seed stage is especially important to the life cycle of an annual plant

because annual plant populations regenerate entirely each year from the seed bank. We grew

fifteen pairs of species at six relative frequencies in two soil moisture environments that are

commonly encountered in mediterranean-climate regions. We quantified the size and number of

seeds produced, and used these data to address three questions: (i) How common are maternal

effects in response to competition, compared to seed number responses? (ii) Are maternal effects

in response to abiotic conditions sensitive to the competitive environment? And (iii) can

variation in the strength of maternal effects among species pairs be explained by differences in

their phylogenetic distances and biogeographic histories?


37

The 25 species considered here have been used previously in experiments designed to understand

macroevolutionary patterns of competitive interactions (Germain et al. 2016) and to quantify

maternal effects in response to abiotic conditions in the absence of competition (Germain &

Gilbert 2014). This previous work shows that maternal effects in the absence of competition are

common (Germain & Gilbert 2014), and that the effect of phylogenetic distance on the intensity

of competition varies with biogeographic history, with the most intense interactions occurring

amongst distant relatives that originate from different regions (Germain et al. 2016). By

extension, we predict that phylogenetic distance and biogeographic history will interactively

explain variation in maternal effects among competitive pairs, if maternal effects are fine-tuned

to competitive differences among species.

Methods

Study species

To test species responses to competition, we competed 15 species pairs out of 25 Mediterranean

annual plant species (Table S3.1) that have been used previously in experiments that investigate

maternal effects in response to abiotic conditions (Germain & Gilbert 2014) and

macroevolutionary patterns of competitive interactions (Germain et al. 2016). Species pairs were

chosen to represent two biogeographic contrasts, with each species in each pair either originating

from a common region (California-California [n = 9]) or distinct regions (California-Spain [n =

6]), with equal sampling across a range of phylogenetic distances (nine to 170 million years since

divergence; x-axes in Fig. 3.3, phylogeny in Fig. S3.1). Additionally, each pair of species within

a biogeographic contrast was selected to be phylogenetically independent from all others (see

Appendix B for details) in order to circumvent the common problem of phylogenetic

pseudoreplication. Seeds were obtained from commercial suppliers and an individual donor, and

were sourced from natural populations consisting of relatively few generations (most < 3, all <

20; Table S2.1) prior to our experiments. It is possible that the genetic diversity of our seed

populations is low compared to natural populations, though we suspect that this is not the case

given the large numbers of individuals used to establish the commercial populations and the

small number of generations that have elapsed.


38

Greenhouse experiment

From January to July 2012, we grew the 15 species pairs in competition in a greenhouse under

two soil moisture levels (wet vs. dry); plants in the wet treatment were watered twice as often as

those in the dry treatment, with the total water received designed to mimic rainfall in mesic sites

during wet and dry years (Germain & Gilbert 2014). The competitive environment was

manipulated by sowing seeds of each species pair at six relative frequency ratios (10:60, 20:50,

30:40, 40:30, 50:20 and 60:10 expected germinants) at a common density of 70 individuals. For

each species pair, we had two replicate pots per combination of soil moisture condition and

frequency ratio, for a total of 420 pots of plants that were randomly assigned to a position in the

greenhouse. All greenhouse growing conditions were chosen to emulate those typical of annual

grassland in Mediterranean-climate regions (Germain & Gilbert 2014; Appendix B). We

monitored pots daily, collecting all seeds produced in each pot as they matured on the parent

plants. At the end of the experiment, all seed material produced in each pot was weighed, and a

random representative subsample was taken to estimate the average size and number of seeds

produced per plant in a given pot (Appendix B).

Concurrent to this experiment, an additional experiment using the same species pairs was

conducted to estimate the impact of soil moisture conditions on maternal provisioning (Germain

& Gilbert 2014). Key differences between the previous experiment and the current one are that

each species was grown individually and at low densities (~seven individuals per pot, compared

to 70 in the current experiment), meaning that competition was greatly relaxed. We include a

summary result from Germain & Gilbert (2014) in this paper to compare maternal effects on

seed size to soil moisture conditions in the presence and absence of competition.

Statistical analyses

Prior to analysis, we transformed the data in two ways to meet model assumptions and facilitate

comparisons among species. First, we log transformed seed size, seed number, and conspecific

frequency to minimize heteroscedasticity and linearize seed size and seed number relationships

with conspecific frequency. Second, to allow us to compare species that differ markedly in seed

production, we standardized the log-transformed seed size and seed number data for each species


39

to a mean of zero and unit variance. For simplicity, we henceforth refer to the standardized log-

transformed data as ‘seed size’, ‘seed number’, and ‘frequency’, unless stated otherwise.

We conducted a cross-species analysis to test if species differed significantly in their seed size

and seed number responses to the competitive and soil moisture environments. For these

analyses, we used the ‘lmerTest’ R package to run linear mixed effects (LME) models, with

either seed size or seed number as response variables, and species (30 levels), soil moisture (two

levels), conspecific frequency (six levels), and their interactions as fixed factors. Pot ID was

included as a random factor, to control for the non-independence of species interacting in a

single pot, as was the identity of the competitive pair (e.g., Lasthenia glabrata vs. L. californica).

Following significant species x frequency and species x soil moisture interactions (see Results),

we ran species-specific analyses. Separately for each species, we then used linear models with

type II sums of squares to test the effects of conspecific frequency, soil moisture conditions, and

their interaction on seed size and seed number, to identify species with significant responses to

the biotic and abiotic environment.

We used major axis regression (MAR; R package ‘lmodel2’) to examine the relationships

between response variables across species, and tested the significance of these relationships

using a Pearson correlation. First, to identify if seed size and seed number responses are

correlated, we performed a MAR with the slopes of species’ seed size responses and the slopes

of species’ seed number responses as variables. Second, we tested whether the presence and

absence of competition alters species’ responses to soil moisture conditions. To do this, we first

calculated species’ average effect sizes of seed size responses to contrasting soil conditions (μdry

– μwet) in the presence of competition using Cohen’s d with pooled variance (Hartung et al.

2011); in cases where a species was used in more than one species pair, a single average effect

size was used. We then used MAR to examine the relationship between these effect sizes and

previously published, identically calculated effect sizes in the absence of competition.

To identify why the strength of frequency-dependent maternal effects varied among species, we

used a LME model, where the response variable was the slope of the relationship between seed

size and conspecific frequency for each species pair. For this analysis, phylogenetic distance,

biogeographic history (native to same vs. different regions), biogeographic origin (native to


40

California vs. Spain) nested within biogeographic history, and soil moisture were included as

fixed factors. The identity of the competitive pair was included as a random factor because each

was grown in two soil moisture environments.

Results

Of the 25 species included in this study, 16 showed significant (P < 0.05) or marginally-

significant (P < 0.10) maternal effects on seed size when in competition (Fig. 3.1a, Table S3.1).

Rather than use post-hoc corrections for multiple tests, we note that we expect 2 or 3 species to

show these results by chance alone. Most species responded to conspecific frequency alone (5

species) or in conjunction with the soil moisture environment in an additive (4 species) or

multiplicative (3 species) manner; four species responded to the soil moisture environment but

not conspecific frequency (Table S3.1). The strength and direction of responses varied among

species, as indicated by significant species x frequency (F29,464 = 2.84, P < 0.001) and species x

soil moisture (F29,410 = 3.19, P < 0.001) interactions in our cross-species statistical model (Table

S3.2). The effect of conspecific frequency on seed size was not consistently negative; species

were similarly likely to increase (five species) or decrease (six species) seed size as frequency

increased (Table S3.1, Fig. 3.1a). Overall, our results demonstrate that for a given plant

community, 64% of species are likely to exhibit maternal effects on seed size in response to

biotic or abiotic conditions.

In contrast to seed size, 21 of the 25 species showed significant or marginally significant

responses through seed number to conspecific frequency (7 species; Fig. 3.1b), soil moisture

conditions (4 species), both additively (6 species), and both interactively (4 species; Table S3.1).

Like seed size responses, the effect of conspecific frequency on seed number depended on the

focal species (significant species x frequency interaction; Table S3.3, F29,716 = 6.63, P < 0.001).

However, in contrast to seed size responses, frequency dependence was negative for most

species (14 out of 17 species; Table S3.1), and was sensitive to the soil moisture environment

(significant frequency x soil moisture interaction; Table S3.3, F1,716 = 4.49, P = 0.034). These

seed number responses were positively correlated with the strength and direction of species’ seed

size responses (r = 0.28, slope = 0.48, P = 0.029; Fig. 3.2a), even though some species showed

opposite seed size and seed number responses (grey regions of Fig. 3.2a).


41

Figure 3.1. Cross-species comparison of frequency-dependent (a) maternal provisioning of seed

size and (b) seed number responses. Points are slope coefficients (± 2 × standard error) of species

responses to the frequency of conspecific competitors, and are shaded black, gray, or white to

indicate significant (P < 0.05), marginally-significant (P < 0.10), or non-significant (P > 0.10)

slopes, respectively. In most cases, frequency × soil moisture interactions are non-significant,

and slopes are averaged across soil moisture environments; when significant, wet (upwards

triangle) and dry (downwards triangle) environments are plotted separately. Competitive pairs

are delineated by lines connecting species codes; see Table S3.1 for species codes.

The strength and direction of seed size responses to contrasting soil moisture environments (wet

vs. dry) in the presence of competition were positively correlated with those in the absence of

competition (r = 0.50, P = 0.010; Fig. 3.2b). The slope of this correlation was less than one

(major axis regression, slope = 0.67), as many species showed stronger responses to soil


42

moisture when grown in the absence of competition (Fig. 3.2b). This result was surprising given

that, in the current experiment, 44% of species altered seed provisioning in response to soil

moisture when in competition, whereas only 21% did so in the absence of competition (Germain

& Gilbert 2014). This suggests that the prevalence but not the magnitude of soil moisture-

induced maternal effects on seed size increase in the presence of competition, possibly because

the effects of soil moisture are dwarfed by those of conspecific frequency.

Figure 3.2. Correlations of (a) seed size and seed number responses to conspecific frequency (n

= 60 [15 pairs × two species × two soil moisture environments]), and (b) seed size responses to

soil moisture conditions in the presence and absence of competition (n = 25). Each point is a

species, and points that fall in the grey zones are species with opposing directions of responses.

The biogeographic history (native to the same or different regions) and origin (California vs.

Spain) of competitors influenced seed size responses to competition, and this effect depended on

phylogenetic relatedness and soil moisture conditions (significant four-way interaction

[phylogenetic distance x biogeographic history x biogeographic origin x soil moisture

environment; F1,60 = 4.55, P = 0.019]; Fig. 3.3, Table S3.4). We did not detect any effect of

phylogenetic relatedness when focal species were competed with species from the same

biogeographic region, regardless of soil moisture treatment (Fig. 3.3a,c). However, when species

from different regions competed in wet soil conditions, closely related Californian and Spanish

species had opposite responses to frequency; California species had negative frequency-


43

dependent responses (smaller seeds at high conspecific frequencies) while their Spanish

competitors had positive frequency-dependent maternal effects (larger seeds at high conspecific

frequencies; Fig. 3.3b). As the phylogenetic distance between species pairs increased, the

direction of frequency-dependent maternal effects switched: species from California and Spain

shifted to positive and negative frequency-dependent seed size responses, respectively (Fig.

3.3b). In contrast, in a dry soil moisture environment, frequency-dependent maternal effects were

consistently positive for California species and negative for Spanish species, and there was no

effect of phylogenetic distance on seed size response (Fig. 3.3d).

Figure 3.3. Variation in the strength of frequency-dependent maternal provisioning of seed size

explained by the phylogenetic distance between competitors, depending on whether they are

native to the same biogeographic region (a,c) or different regions (b,d). The red dashed line

indicates the cut-off between negative and positive frequency dependence, grey dashed lines

delineate pairs of species that were competed, and solid lines denote significant relationships.

Black symbols and lines are species native to California, whereas gray symbols and lines are

species native to Spain.


44

Discussion

There is a substantial body of empirical work investigating maternal effects in response to abiotic

conditions (Herman & Sultan 2011), yet responses to biotic interactions remain understudied,

particularly in plants (Weiner et al. 1997; Larios & Venable 2015). Our results show that

changes to the competitive environment can alter maternal provisioning of seed size and, much

like more commonly measured seed number responses (Law & Watkinson 1987), that the

strength and direction of these provisioning effects depend on identities of competing species.

Below, we discuss how these responses inform classic questions in ecology and generate new

hypotheses for future study.

We detected maternal effects on seed size in response to the frequency of conspecific

competitors in nearly half of the species examined (Fig. 3.1a). Frequency dependence manifests

when the strengths of intraspecific and interspecific competition differ. For example, when

species compete more strongly with conspecific than heterospecific individuals, fecundity should

increase as the frequency of conspecifics decreases and a species becomes relatively rare

(negative frequency dependence; Levine & HilleRisLambers 2009). In the context of

demographic rates, the direction of frequency dependence can indicate whether competition is

more likely to result in coexistence (negative frequency dependence) or exclusion (positive

frequency dependence). In our experiment, negative frequency-dependent seed number

responses that promote coexistence were common (Fig. 3.1b), but negative and positive seed size

responses were equally common (Fig. 3.1a). If seed size is tied to competitive ability, as many

authors suggest (e.g., Westoby et al. 1996), then it is possible that maternal provisioning of seed

size could reinforce or counteract the effects of frequency dependence on demographic rates.

Although our results suggest that seed size responses to conspecific frequency generally act to

reinforce seed number responses (i.e., they are positively correlated, Fig. 3.2a), some species

showed opposite seed size and seed number responses (points in the grey regions of Fig. 3.2a).

This indicates that the long-term coexistence outcome of a given pair of competing species may

be affected by complex and species-specific interactions between seed number and seed size

responses to the maternal competitive environment. For example, while negative frequency

dependence in both seed size and seed number could work in tandem to stabilize plant


45

communities and promote coexistence, opposing seed size and number responses could negate

one another. Although the exact relationship between seed size and demographic rates remains

unknown, previous work suggests that interspecific and intraspecific variation in seed size alters

several important biological parameters, from dormancy to growth and fecundity (Westoby et al.

1996; Eriksson 1999; Germain & Gilbert 2014). Thus, the common but variable seed size

responses observed among species suggest that focusing on seed number responses alone may be

insufficient to capture the full impact of competitive interactions in the offspring generation

(Ginzburg & Taneyhill 1994; Van Allen & Rudolf 2015).

We also found important differences in the prevalence of maternal effects in response to soil

moisture conditions in competitive and non-competitive environments. Specifically, over twice

as many species altered seed provisioning in response to soil moisture in the presence of

competition (44% of species; this study, Table S3.1) compared to in the absence of competition

(21% of species; Germain & Gilbert 2014). This result is important because many tests of

maternal effects have been conducted in non-competitive environments, with individuals grown

alone in pots of soil (e.g., Aarssen & Burton 1990). Because most plants experience competition

in their natural environments, current estimates of the prevalence of maternal effects may be

conservative and most relevant to disturbed environments where plant densities are low.

Additionally, competition appears to dampen maternal effects in response to soil moisture in

some species while causing them to emerge in others (points falling in grey regions of Fig. 3.2b).

This indicates that, contrary to our initial expectations, competition does not simply exacerbate

maternal provisioning of seed size in response to soil moisture, but instead appears to alter the

nature of soil moisture’s effects on seed size in some species.

We found evidence that phylogenetic distance and biogeographic history, two characteristics of

species that have been shown to structure competitive interactions (Strauss et al. 2006; Peay et

al. 2012; Germain et al. 2016), can alter both the strength and direction of seed size provisioning

(Fig. 3.3). In a wet environment, for example, competing with a distant relative that differed in

biogeographic origin led to positive frequency dependence among species native to California,

but negative frequency dependence among species native to Spain (Fig. 3.3b). If increased

maternal investment into individual seeds results in larger, more competitive offspring, then our

results indicate that California species produce more competitive offspring when primarily


46

competing with (i) conspecifics in the presence of a closely related Spanish competitor, or (ii)

distantly related heterospecifics of Spanish origin; the opposite is true for Spanish species.

Although studies to date have shown a general, positive effect of seed provisioning on

competition, whether this effect it is contingent on the environment or depends on the

evolutionary history of a species remains largely unexplored (reviewed in Leishman et al. 2000).

The interactive effects of biogeographic history and phylogenetic distance on maternal effects

shown here reflect the results of previous work in this system on the macroevolution of species’

competitive differences (Germain et al. 2016). From an evolutionary standpoint, our results

suggest that such maternal provisioning strategies to competition differ among plant species, but

may be shared by species from a given bioregion.

From an ecological standpoint, an intriguing hypothesis posed by Dyer et al. (2010) is that

maternal effects might contribute to the invasion success of non-native species. Our experiment

was not designed to test the role maternal provisioning plays during species invasions, but

consistent with this idea, we observe differences among regions in how competition affects

maternal provisioning of seed size. However, much of the strength and direction of frequency-

dependent maternal effects remains unexplained by macroevolutionary relationships, particularly

amongst species of the same biogeographic origin (Fig. 3.3a,c). As such, future work that

explores other differences in the biologies of species and the nature of their competitive

interactions is needed to fully understand when and why maternal effects arise (e.g., Sultan et al.

2009; Metz et al. 2015) and if species from different bioregions differ in optimal provisioning

strategies (Venable 1992).

Although this study advances our understanding of the importance of competition in structuring

maternal provisioning, there are two caveats that should be considered in interpreting our

findings. First, we were unable to identify how changes in seed size translate into differences in

offspring performance, due to the logistical infeasibility of the full factorial experiment that

would be required to test for longer-term impacts of seed size. It is possible that maternal effects

on seed size may not persist beyond the seed stage, but we have shown previously that they can

affect the size of offspring at the establishment stage in the species used here, even in the

absence of competition when such effects are considered weaker (Germain & Gilbert 2014).

Seedling recruitment is an important life stage for plants, especially under competition


47

(Leishman et al. 2000; Goldberg et al. 2001), and thus it is likely that the advantage of larger

seeds can affect demographic rates (e.g., germination, survival, and fecundity). Second, by

focusing on seed size responses, we likely underestimate the overall prevalence of maternal

effects that can manifest in other ways, such as through germination or dormancy rates (Germain

& Gilbert 2014), or through epigenetic effects that can alter the offspring phenotype in more

complex ways (Herman & Sultan 2011). As such, this study should be viewed as an important

first step in characterizing maternal effects in competitive environments that can be used to

inform future work.

Conclusion

Our results suggest that current estimates of maternal effects in non-competitive environments

are conservative, and that frequency-dependent maternal effects have the potential to alter

competitive dynamics. We show that competition can alter maternal provisioning of seed size,

and that these maternal effects are fine-tuned to competitive differences among species, which in

turn are shaped by the abiotic environment, evolutionary history, and biogeographic origin. The

study of maternal effects has received renewed interest due to its exciting potential to explain

population- and community-level responses to heterogeneous environments (Ginzburg &

Taneyhill 1994; Van Allen & Rudolf 2013; Van Allen & Rudolf 2015). Our research suggests

new avenues for future work that would help characterize the full range and impact of maternal

effects, and sheds light on the complex nature of species interactions.

Acknowledgements

We thank C. Blackford, A. Leale, A. Mushka, B. Hall, and A. Petrie for assistance in the

greenhouse and laboratory, and J. Stinchcombe for his feedback on an earlier version of the

manuscript. Funding was provided by NSERC-CGS (R.M.G. & T.N.G.), OGS (N.T.J.), and

NSERC Discovery (B.G.).

Literature cited

Aarssen, L.W. & Burton S.M. (1990) Maternal effects at four levels in Senecio vulgaris

(Asteraceae) grown on a soil nutrient gradient. American Journal of Botany, 77, 1231–

1240.


48

Allen, J.D. (2012) Effects of egg size reductions on development time and juvenile size in three

species of echinoid echinoderms: implications for life history theory. Journal of

Experimental Marine Biology and Ecology, 422, 72-80.

Ayala, F.J. (1971) Competition between species: frequency dependence. Science 171:820–824.

Baldeck, C.A., Harms, K.E., Yavitt, J.B., John, R., Turner, B.L., Valencia, R., Navarrete, H., et

al. 2013. Habitat filtering across tree life stages in tropical forest communities.

Proceedings of the Royal Society B: Biological Sciences, 280, 20130548.

Benton, T.G., Ranta, E., Kaitala, V. & Beckerman, A.P. (2001) Maternal effects and the stability

of population dynamics in noisy environments. Journal of Animal Ecology, 70, 590–599.



Dyer, A.R., Brown, C.S., Espeland, E.K., McKay, J.K., Meimberg, H. & Rice, J.K. (2010) The

role of adaptive trans-generational plasticity in biological invasions of plants.

Evolutionary Applications, 3, 179–192.


Ecology, 15, 54–68.

Eriksson, O. (1999) Seed size variation and its effect on germination and seedling performance

in the clonal herb Convallaria majalis. Acta Oecologica, 20, 61–66.

Fotelli, M.N., Gessler, A., Peuke, A.D. & Rennenberg, H. (2001) Drought affects the

competitive interactions between Fagus sylvatica seedlings and an early successional

species, Rubus fruticosus: responses of growth, water status and delta 13C composition.

New Phytologist, 151, 427–435.

Fridley, J.D. (2012) Extended leaf phenology and the autumn niche in deciduous forest

invasions. Nature, 485, 359–362.

Funk, J.L. & Vitousek, P.M. (2007) Resource-use efficiency and plant invasion in low-resource

systems. Nature, 446, 1079–1081.


Science, 318, 1134–1136.


stress-induced parental effects in an invasive annual grass. International Journal of Plant

Sciences, 174, 886–895.

Germain, R.M. & Gilbert, B. (2014) Hidden responses to environmental variation: maternal

effects reveal species niche dimensions. Ecology Letters, 17, 662–669.

Germain, R.M., Weir, J.T. & Gilbert, B. (2016) Species coexistence: macroevolutionary


49

relationships and the contingency of historical interactions. Proceedings of the Royal

Society B: Biological Sciences, 283, 20160047.

Ginzburg, L.R. & Taneyhill, D.E. (1994) Population-cycles of forest Lepidoptera - a maternal

effect hypothesis. Journal of Animal Ecology, 63, 79–92.



Goldberg, D.E., Turkington, R., Olsvig-Whittaker, L. & Dyer, A.R. (2001) Density dependence

in an annual plant community: variation among life history stages. Ecological

Monographs, 71, 423–446.

Hartung, J., Knapp, G. & Sinha, B.K. (2011) Statistical meta-analysis with applications. John

Wiley & Sons Inc, Hoboken, NJ.

Herman, J.J. & Sultan, S.E. (2011) Adaptive transgenerational plasticity in plants: case studies,

mechanisms, and implications for natural populations. Frontiers in Plant Science, 2, 102.

Kendall, B.E., Fox, G.A., Fujiwara, M. & Nogeire, T.M. (2011) Demographic heterogeneity,

cohort selection, and population growth. Ecology, 92, 1985–1993.

Larios, E. & Venable, D.L. (2015) Maternal adjustment of offspring provisioning and the

consequences for dispersal. Ecology, 96, 2771–2780.

Law, R. & Watkinson, A.R. (1987) Response-surface analysis of two species competition: an

experiment on Phleum arenarium and Vulpia fasciculata. Journal of Ecology, 75, 871–

886.

Leishman, M.R., Wright, I.J., Moles, A.T. & Westoby, M. (2000) The Evolutionary Ecology of

Seed Size. Seeds: The Ecology of Regeneration in Plant Communities (ed. Fenner, M.),

pp. 31-57. CABI, Wallingford.



Marshall, D. & Uller, T. (2007) When is a maternal effect adaptive? Oikos, 116, 1957–1963.


phylogenetic structure of communities. Ecology Letters, 13, 1085–93.

Metz, J., von Oppen, J. & Tielbörger, K. (2015) Parental environmental effects due to

contrasting watering adapt competitive ability, but not drought tolerance, in offspring of a

semi-arid annual Brassicaceae. Journal of Ecology, 103, 990–997.

Miao, S.L., Bazzaz, F. & Primack, R. (1991) Persistence of maternall nutrient effects in Plantago

major: the third generation. Ecology, 72, 1634-1642.


50

Mitchell, C.E., Agrawal, A.A., Bever, J.D., Gilbert, G.S., Hufbauer, R.A., Klironomos, J.N.,

Maron, J.L., et al. (2006) Biotic interactions and plant invasions. Ecology Letters, 9, 726–

740.

Molofsky, J., Bever, J.D. & Antonovics, J. (2001) Coexistence under positive frequency

dependence. Proceedings of the Royal Society B: Biological Sciences, 268, 273-277.

Mousseau, T.A. & Fox, C.W. (1998) Maternal effects as adaptations. Oxford University Press,

USA.

Mousseau, T.A., Uller, T., Wapstra, E. & Badyaev, A.V. (2009) Evolution of maternal effects:

past and present. Philosophical Transactions of the Royal Society B: Biological Sciences,

364, 1035–8.

Peay, K.G., Belisle, M. & Fukami, T. (2012) Phylogenetic relatedness predicts priority effects in

nectar yeast communities. Proceedings of the Royal Society B: Biological Sciences, 279,

749–758.



Segers, F.H.I.D. & Taborsky, B. (2011) Egg size and food abundance interactively affect

juvenile growth and behaviour. Functional Ecology, 25, 166–176.

Smart, A.J. & Moser, L.E. (1999) Switchgrass seedling development as affected by seed size.

Agronomy Journal, 91, 335-338.

Stratton, D.A. (1989) Competition prolongs expression of maternal effects in seedlings of

Erigeron annuus (Asteraceae). American Journal of Botany, 76, 1646–1653.


more invasive. Proceedings of the National Academy of Sciences, 103, 5841–5845.


plasticity in ecologically distinct congeners. Ecology, 90, 1831–1839.

Van Allen, B.G. & Rudolf, V.H.W. (2013) Ghosts of habitats past: environmental carry-over

effects drive population dynamics in novel habitat. American Naturalist, 181, 596–608.

Van Allen, B.G. & Rudolf, V.H.W. (2015) Habitat-mediated carry-over effects lead to context-

dependent outcomes of species interactions. Journal of Animal Ecology, 84, 1646–1656.

Venable, D.L. (1992) Size-number trade-offs and the variation of seed size with plant resource

status. American Naturalist, 140, 287–304.

Weiner, J., Martinez, S., Muller-Scharer, H., Stoll, P., Schmid, B. & Society, B.E. (1997) How

important are environmental maternal effects in plants? A study with Centaurea

maculosa. Journal of Ecology, 85, 133–142.


51

Westoby, M., Leishman, M., Lord, J., Poorter, H. & Schoen, D.J. (1996) Comparative ecology of

seed size and dispersal. Philosophical Transactions of the Royal Society B: Biological

Sciences, 351, 1309–1318

CHAPTER 4: DISPERSAL LIMITATION IN METACOMMUNITIES

52

Chapter 4

Pervasive dispersal limitation reduces diversity in an annual plant

community

Abstract

Metacommunity theory posits that dispersal among habitat patches that links local communities

is a key ‘regional’ process that maintains biological diversity. Yet, regional processes are rarely

tested in natural field conditions because the manipulations required to do so are infeasible for

most systems. In a natural serpentine grassland, we used a seed vacuum to manipulate dispersal

limitation while maintaining realistic spatial structure of species pools and environmental

conditions. We created seed banks that were mixed among thirty communities at spatial scales

varying from <1m to >10km. We found that species richness increased with the spatial scale of

seed addition, with nearly twice as many species present in plots that received seed from large

(>5km) compared to small (<5m) spatial scales. Similarly, the strength of species-environment

relationships increased with the spatial scale of seed addition, from weak associations at small

scales (R2 = 0.25, 18% of species) to strong environmental sorting at large scales (R

2 = 0.43, 38%

of species). These diversity responses to the spatial scale of seed addition were nonlinear, with

no differences both below 5m and above 5km. Further examination of naturally-occurring spatial

structure in this system suggests that these nonlinearities are generally explained by scale-

specific redundancies in environments and species pools. Our novel approach of manipulating

natural communities at different spatial scales reveals (i) nonlinear transitions in the importance

of different ecological processes, (ii) how heterogeneity in species and environments accumulate

relative to one another, and (iii) how this accumulation affects the mechanisms that structure

diversity.

Introduction

A metacommunity is colloquially defined as a “community of communities” (Gilpin & Hanski

1991), or more formally as a regional network of local communities linked by the dispersal of

their constituent species. The recognition that most communities operate as part of a larger


53

metacommunity means that species diversity cannot simply be predicted by biotic and abiotic

conditions at local scales (Caley & Schluter 1997; Leibold et al. 2004). Rather, diversity is the

product of local conditions, regional dispersal that links localities, and their interaction. A large

body of theory has since been developed to understand how local and regional processes together

might structure diversity (e.g., MacArthur & Wilson 1967; Hubbell 2001; Kneital & Chase 2004;

Leibold et al. 2004; Gilbert 2012), and mesocosm experiments and observational data have

provided evidence that is consistent with specific theoretical predictions (e.g., Gilbert &

Lechowicz 2004; Cottenie 2005; Cadotte 2007).

Despite these developments, we have yet to fully understand how local and regional

determinants of diversity play out in actual natural communities, for one key reason: regional

processes, unlike local processes (e.g., Harpole & Tilman 2007), are not often amenable to

experimental manipulation in the field simply due to the inability to move most communities of

organisms. Manipulative tests of local and regional diversity are typically limited to mesocosm

experiments on communities constructed using simplified environments, species pools, or

dispersal patterns (e.g., Low-Décarie et al. 2015; see review by Grainger & Gilbert 2016).

Because it is often unclear how such simplifications affect experimental outcomes (Logue et al.

2011), mesocosm experiments allow essential tests of the range of potential outcomes under

different sets of experimental conditions but might not reflect what occurs in nature.

A feature of natural communities that is particularly difficult to replicate in experiments is the

spatial structure of processes that affect community and metacommunity diversity. Specifically,

as the spatial scale of observation increases, so too do the numbers of species and range of

environments that are sampled (Weins 1989; Lechowicz & Bell 1991) and the role that distances

among localities play in shaping the observed distributions of species (Pinto & MacDougall

2010). As a consequence, the spatial structure of the environment has been hypothesized to

constrain how species are distributed (Pinto & MacDougall 2010) and how they move across

landscapes (Bender & Fahrig 2005). However, finding new ways to combine the power of

manipulative field experiments with the biological realism of natural landscape structure can

address fundamental questions, such as: how widespread is dispersal limitation in a community

and at what spatial scales does it manifest? How strongly do species sort along environmental


54

gradients in the absence of dispersal limitation? And even more basically, at what spatial extents

are communities and metacommunities most appropriately defined?

The above questions can be empirically tested by homogenizing species pools among local

communities to remove any potential dispersal limitation, and tracking the joint responses of

species richness and species-environment relationships to this manipulation (Fig. 4.1). If

dispersal limitation is widespread and the landscape is heterogeneous, then homogenization of

species pools would cause species-environment relationships to increase in strength (Fig. 4.1a) as

species are given access to suitable localities that were previously inaccessible. Alternatively, if

species are not dispersal limited and already well-matched to the environmental conditions (Fig.

4.1b), the strength of species-environment relationships should decrease even as local richness

increases. For species that are neutral in terms of environmental requirements, species-

environment relationships should not change whether species have limited dispersal (Fig. 4.1c)

or are already dispersing evenly among communities (Fig. 4.1d). The approach of homogenizing

species pools and tracking diversity responses to local conditions is similar to that taken by many

experiments that use mesocosm (e.g., Chase 2010) or field (e.g., Germain et al. 2013)

manipulations with artificial species pools (see reviews by Logue et al. 2011 and Grainger &

Gilbert 2016), but has yet to be implemented in the field with natural spatial structure in species

pools and environments.

Figure 4.1. Predicted responses of species richness (top panel) and species-environment

relationships (bottom panel) to the homogenization of the species pool, depending on whether

distance (D) or environment (E) impose limits to naturally-occurring species distributions.


55

Although identifying the mechanisms that structure diversity could be achieved by comparing

communities with and without the homogenization of the regional species pool, species pools

compounded across a range of spatial scales would reveal the spatial scales at which different

mechanisms – dispersal and environmental sorting – are strongest. For example, differences in

species richness and species-environment relationships among treatments in which the spatial

scale of species pooling varies could be due to differences in the ranges of species,

environments, or both at each scale, whereas a lack of differences would indicate that a similar

range of species or environments are present across all spatial scales. There is no reason to

expect responses to change linearly with spatial scale, given that the environmental conditions

(Bell et al. 1993) and dispersal probabilities (Nathan 2006) that constrain species distributions

also scale nonlinearly with distance; Wiens (1986) refers to these nonlinear transitions in the

importance of different ecological processes as ‘domains of scale’. However, it is not known

how species and environments accumulate relative to one another with increasing spatial scale,

and how this spatial scaling affects the mechanisms that structure diversity.

We adapted the ‘hay transfer’ technique from the restoration literature (Coiffait-Gombault et al.

2010) to experimentally homogenize species pools over a range of spatial scales for entire

natural communities of serpentine-associated plants in Northern California. This technique

involved vacuuming all seed and other loose material from field plots, pooling the material

among plots occurring at five spatial scales (~1 m, 5 m, 100 m, 5 km, and 10 km), and then

redistributing it. We performed these manipulations in serpentine plant communities because the

high occurrence of annual species and patchy distribution of serpentine habitat is well-suited to

testing spatial questions (Harrison 1999) with the hay transfer method (Coiffait-Gombault et al.

2010); we refer to these manipulations as ‘seed pooling’ henceforward. In the following growing

season, we collected data on species occupancy and environmental conditions to answer four

questions: (i) how common is dispersal limitation in a metacommunity, and (ii) at what spatial

scales does it manifest? (iii) How strongly do species sort along environmental gradients once

dispersal limitation is experimentally removed? And (iv) how do scale-dependent signals of

dispersal limitation and environmental sorting correspond to the underlying spatial structure of

species pools and environmental conditions? We predicted that environmental conditions would

increasingly explain variation in species distributions as dispersal limitation is removed at


56

increasing spatial scales (Fig. 4.1a). This result would support our hypothesis that diversity

patterns reflect dispersal limitation, and that given the opportunity, species sort deterministically

according to their environmental niche requirements.


Study system

The field experiment was conducted at the 2800-ha McLaughlin Natural Reserve

(http://nrs.ucdavis.edu/McL/) in northern California [38.8739° N, 122.4317° W]. The region is of

a mediterranean climate, featuring cool wet winters (November-March) and hot dry summers

(April-October) with ~750 mm annual rainfall. The landscape is largely comprised of chaparral,

oak woodland, and grassland meadow habitat; common herbivores include mule deer, jack

rabbits, and pocket gophers.

The reserve lies on the San Andreas Fault, and has unique soil chemistry owing to the emergence

and erosion of the Earth’s mantle into the serpentine ultramafic soils that characterize the region.

Serpentine soils have Ca/Mg ratios < 1, as well as low levels of essential nutrients, high heavy

metal content, and poor soil moisture retention. Although these growing conditions are typically

harmful to plant growth, serpentine soils support a unique and diverse flora of native species that

have specialized to tolerate those harmful conditions (Whittaker 1954; Anacker 2014). As a

result, serpentine soils are hypothesized to act as spatial refugia for native species to escape the

competitive effects of exotic European species that now dominate Californian landscapes (e.g.,

Avena spp., Bromus spp. Centaurea solstitialis). Serpentine and non-serpentine soils frequently

occur in close proximity at the McLaughlin reserve (Fig. S4.5), with dense swards of

competitively dominant exotic species in non-serpentine soils likely limiting dispersal among

serpentine sites (Gilbert & Levine 2013).

We focused specifically on serpentine meadow habitat, and observed 113 species in our plots out

of a potential pool of the 310 species that occur in all habitat types at the reserve; 73 of the 113

species were annual plants (Table S4.1). The four most common species observed in

unmanipulated plots were Vulpia microstachys (89% of sites), Hemizonia congesta (79%),

Plantago erecta (69%), and Lasthenia californica (52%). Sixteen of the 113 total species could


57

not be identified or classified as having an annual or perennial life history and were thus

excluded from all analyses; however, these individuals only occurred at single sites and are thus

unlikely to have large effects on diversity patterns.

Experimental setup

In May 2013, we surveyed the reserve for 30 serpentine meadow sites. Site locations were

chosen such that they could be hierarchically grouped at five spatial scales, which resulted in

three groups of five sites each at the northwest and southeast ends of the reserve (Fig. S4.1).

Within each site, we flagged and GPS located eight 0.75 × 0.75 m plots arranged in 2 x 4 block

of plots, with plots separated by 1 m of bare ground. Each plot in a block was randomly assigned

to receive a different experimental treatment (discussed below). In total, there were 240 plots

(eight plots x 30 sites) and five spatial scales for comparison (at the level of the plot [1m], block

of plots at a site [5 m], group of sites [100m], reserve half [5 km], and whole reserve [10 km];

Fig. S4.1).

In late July 2013, once all winter annual species senesced and the majority of summer annual

species had set seed, we harvested all seed and standing vegetation from seven of the eight plots

at each site using garden shears and a powerful gas-powered leaf vacuum (Stihl BG86); the

eighth plot (C2; Fig. S4.1b) was left unaltered to evaluate any unintended effects of the

vacuuming procedure on plant diversity. All collected materials were stored outdoors in paper

bags to allow natural heat stratification until they could be processed (< six weeks).

All collected materials were reintroduced to the seven vacuumed plots per site in two control and

five seed pooling treatments (see Fig. S4.1b). The control treatments were as follows: a ‘vacuum

without replacement’ treatment (C1), to identify individuals that were left behind following

vacuuming, and a ‘vacuum without movement’ treatment (C3), where the collected material was

homogenized at the plot-level and redistributed back onto the source plot. The five spatial

scaling treatments involved the pooling, homogenization, and redistribution of material collected

from either a single plot (1 m; blue plot in Fig. S4.1b), multiple plots at a site (5 m; green plot),

five sites of a single group (100 m; yellow plot), 15 sites from the same half of the reserve (5 km;

orange plot), all 30 sites across the entire reserve (10 km; red plot). The redistributed material

was secured to each recipient plot with jute twine and landscape staples rather than mesh


58

(Coiffait-Gombault et al. 2010) to minimize damage to flowering plants and the risk of trapping

vertebrates.

We surveyed the plot-level community structure and corresponding environmental parameters

during peak biomass in the following growing season. Plots were surveyed April 20 to May 2,

2014, using % cover estimates of each species, since small-statured annuals can occur at

densities up to 5500 individuals m-2

(Bartolome 1979); additional surveys were conducted later

in the season to confirm the identities of late-flowering species. We surveyed the innermost 0.5 ×

0.5 m2 of each plot to account for any edge effects in our analyses. We measured plot-level %

soil moisture content, understory light availability in full sun (PAR), and slope inclination, as

well as site-level elevation, slope aspect, hillside slope steepness, and soil depth; soil moisture

measurements were taken three days after heavy rainfall as a measure of soil moisture retention.

We also performed site-level soil fertility analyses (NO3-N, Olsen-P, X-K, X-Na, X-Ca, X-Mg,

pH, cation exchange capacity, organic matter; UC Davis Analytical Lab) on soil samples

collected and pooled between four plots per site. X-Ca and X-Mg were converted to a ratio of

Ca/Mg. Although site-level estimates of environmental conditions preclude finer-scale estimates,

those variables for which we do have plot-level estimates showed no difference among plots that

occur at the same site (all P > 0.35, Table S4.2).

Data analysis

A presence-absence matrix was created for all 73 annual species in 29 of the 30 sites that were

sampled; data from one site was lost due to a corrupt data file. We used presence-absence data,

rather than an abundance-based estimate, because it is most appropriate for tracking gains and

losses of species (Gilpin & Hanski 1991) in response to manipulations of species pools. The R

package ‘vegan’ was used to create a distance matrix of Jaccard coefficients for all pairwise plot

combinations. A principal coordinates analysis (PCoA) was then performed on this distance

matrix to track changes in species composition, as summarized by the first two axis scores,

among sites and treatments.

We tested the responses of three components of diversity to our species pool manipulations using

linear mixed effects models in the ‘lmerTest’ R package. Species richness, evenness, and

composition (PCoA axis 1 and 2 scores) were response variables in separate analyses, each with


59

spatial seed pooling treatment as a fixed factor and plot nested within site as a random factor.

Because the spatial seed pooling treatment emerged as a significant predictor in all analyses, we

used post-hoc Tukey tests using the ‘multcomp’ package to identify treatment levels that differed

significantly from each other. For species composition, the axis scores summarized a low but still

meaningful proportion of the species composition data (axis 1 = 9.7%, axis 2 = 7.2%; Fig. S4.2).

Prior to testing the responses of species-environment relationships to our seed pooling

treatments, we used complementary multivariate methods in R package ‘vegan’ to reduce the

dimensionality of the environmental variables that were measured. First, we used variance

inflation factors to confirm that multicollinearity was low (VIFs < 10, Neter et al. 1989), and

performed a constrained correspondence analysis (CCA), forward selecting environmental

variables using R function ‘ordistep’, to identify the subset of environmental variables that

significantly influenced species occurrences (all but soil depth). Second, we ran a principal

component analysis (PCA) on the subset of environmental variables identified as meaningful in

the CCA. A screeplot of the PCA output identified the first two axes as interpretable compared to

the commonly-used broken-stick null model (Jackson 1993), with the axis 1 and 2 scores of the

PCA summarizing 51.4% of the variation in the among-site environmental dataset.

We used the first (PCA1) and second (PCA2) PCA axis scores as composite environmental

variables to examine how the seed pooling treatments affected the strength of community- and

species-level responses to environment. To examine community responses as a whole, we first

performed a generalized linear mixed effects (glme) model, with species occurrences as the

binomially-distributed response variable, PCA1, PCA2, species, seed pooling treatment, and

their interactions as fixed effects, and plot nested within site as random effects. The R function

‘drop1’ was sequentially applied until the model that best explained variation in our dataset was

obtained (Table S3). Because the PCA1*species, PCA2*species, and PCA1*PCA2*treatment

interactions were significant, we analyzed each seed pooling treatment separately in a glm with

PCA1*species and PCA2*species predictor variables. We then compared the amount of variation

explained by composite environmental conditions among seed pooling treatments, with the

prediction that R2 values should increase with the spatial scale of seed pooling. ‘Species’ was a

fixed factor in these analysis to facilitate model convergence and because we were directly


60

interested in the amount of variation explained (R2) by species-specific differences in responses

to environmental conditions and not tests of significance per se.

To further examine species-level responses (i.e., how strongly the occurrences of each species

are associated with environmental conditions), we performed separate generalized linear models

for each species and seed pooling treatment, testing the additive effects of PCA1 and PCA2. We

then summed the percentage of species that significantly responded to the environmental

gradient in each treatment, and compared this percentage to the 5% false-discovery rate expected

by chance given the number of tests performed. Because species varied in their post-

manipulation associations with environmental conditions, we tested whether this significance

was tied to differences among species in regional occupancy (# of the 30 sites a species occurred

in) prior to manipulation. To do this, we used a glm with the presences or absences of significant

associations with the environment for all species as the binomially-distributed response variable,

as a function of log-transformed regional occupancy with a second-order polynomial to allow for

a hump-shaped relationship. Allowing for a hump was important because a given species might

poorly associate with the environment because it is either extremely rare or extremely common.

To quantify similarity among sites in environments and species that occurred naturally at distinct

spatial scales, we performed separate permutational multivariate analyses of variance (‘adonis’

function in R package ‘vegan’); this analysis is analogous to a univariate analysis of variance,

except expanded to handle multivariate response variables. The response variables were matrices

of Euclidian (for environment) and Jaccard (for species) distances among sites, and each matrix

was tested for differences among groups of five sites (100 m scale) nested within each half of the

reserve (5 km scale). The species distance matrix was created from plots that received the 1 m

treatment only, as this treatment most closely reflects natural species distributions.

The control plots were used to assess the presence of any unintended effects of the hay transfer

manipulation that were unrelated to the seed pooling treatment. We found no difference in

species richness or composition between the unmanipulated (C2 in Fig. S4.1) plots and our 1 m

treatment plots (all P > 0.998), indicating that the hay transfer manipulation did not affect local

diversity. Similarly, there was no difference among plots that received its own hay back (C3) and

the 1 m treatment plots that received hay from a different adjacent plot (all P > 0.807). As a


61

result, we only report the results from the 1 m treatment plots. The removal without replacement

control plot (C1) had significantly lower species richness than the unmanipulated control plot

(C2) and all treatment plots (all P < 0.002) indicating that are seed vacuum was effective.

Results

Plot-level species richness increased with the spatial scale of seed pooling (F4,112 = 72.4, P <

0.001; Fig. 4.2a), from 11.4 ± 0.73 species (mean ± standard error) in the 1 m treatment to 22.5 ±

1.01 species in the 10 km treatment (Fig. 4.2a). This increase was sigmoidal rather than linear,

with no significant differences in species richness between the two smallest (1 and 5 m; P =

0.98) or the two largest (5 and 10 km; P = 0.68) spatial scales, but highly significant (all P <

0.001) differences among small (≤5 m), intermediate (100 m), and large (≥5 km) spatial scales.

Figure 4.2. Effect of spatial scale of seed pooling on (a) species richness (mean ± se) and (b) the

strength of species-environment relationships (R2 values); observed patterns correspond to

prediction in Fig. 4.1a. Points with the same letter were not significantly different in a multiple

comparisons test.

In an analysis that included all 73 annual species observed in our study, we found the proportion

of variance explained (marginal R2) by our models increased from 0.25 to 0.43 with increasing

spatial scale of seed pooling (Fig. 4.2b). In considering each species separately, we found that

the occurrences of 18% of species were significantly predicted by environmental conditions

when seeds were distributed locally (at 1 m scales). As the spatial scale of seed pooling


62

increased, so too did the percentages of species whose occurrences were significantly associated

with environmental conditions, which peaked at 38% in the second largest spatial scale (Fig.

S4.3)—these significantly-associated species tended be those that occurred at intermediate but

not high or low abundances prior to manipulation (a significant hump-shaped relationship; Χ2 =

9.62, P = 0.008, Fig. 4.4). Taken together, our results show concordance in the effects of seed

pooling on the effects of environmental conditions when species are considered as a community

and separately.

In addition to plot-level species richness (α-diversity), we tested the effects of our seed pooling

treatments on other components of diversity. Species composition was distinct among treatments

at the same spatial scales of seed pooling as species richness, with sigmoidal responses in the

axis 1 and 2 PCoA scores (Fig. 4.3a). These axes were significantly associated with forty-two

plant species (P < 0.05), including many endemic serpentine-associated species (e.g., Clarkia

gracilis, Collinsia sparsiflora, and Navarretia jepsonii). The compositional shifts translated into

a sigmoidal decrease in among-plot species dissimilarity (β-diversity), from 0.80 to 0.55

dissimilarity in comparing the 1 m to 10 km spatial scaling treatments (Fig. 4.3b). Site

occupancy (the number of sites occupied by each species) also increased with the spatial scale of

seed pooling (Fig. S4.4), as did community evenness due to the increasing occurrences of

regionally rare species (shallower slope, Fig. S4.4). The size of the regional species pool (γ-

diversity) was not affected by the seed pooling treatments, with a mean pool size of 60.0 ± 0.8

standard error across treatments (x-intercept in Fig. S4.4). Not all 73 species were observed in all

treatments; on average, each treatment had 1.2 species (± 0.58 standard error) that were not

observed in any other treatment.


63

Figure 4.3. Effect of spatial scale of seed pooling on (a) species composition and (b) the

compositional dissimilarity of plots (β-diversity). Species composition is from the first axis

scores from a PCoA with Jaccard’s dissimilarity index (see biplot in Fig. S4.4a); results with

second axis scores are qualitatively similar. Points with the same letter were not significantly

different in a multiple comparisons test.

Figure 4.4. Frequency distribution of site occupancies of all species (gray bars) and species that

were significantly associated with the environment (black bars). The red line is the fitted

relationship between the proportion of species with significant species-environment relationships

and site occupancy; this relationship is a significantly hump-shaped (P = 0.008). Note that the x-

axis is on a log-scale.


64

We found differences in the naturally-occurring spatial structure of environmental conditions and

species pools that inform our interpretation of the observed diversity responses to our

manipulations. Environmental conditions significantly differed between halves of the reserve (5

km scale; F = 247.68, P = 0.001) and among groups of sites within halves (100 km scale; F =

14.05, P = 0.001), with little overlap in environmental conditions among halves of the reserve

(Fig. 4.5a). Variation in environmental conditions was greater in the south half of the reserve,

possibly due to the greater spatial spread of these sites. Although these are site-level estimates,

for those environmental variables that we do have multiple within-site measurements (5 m scale),

there were no statistically significant differences among plots that occur at the same site (all P >

0.3, Table S4.2) as expected given the biology of serpentine soils.

Figure 4.5. Natural differences among halves of the reserve in (a) environmental conditions

(from a PCA) and (b) species composition (from a PCoA). The contours outline the bivariate

normal distributions of each half of the reserve; dashed contours/upwards triangles are sites that

occur in the north half, solid contours/downwards triangles are sites that occur in the south half.

Environmental conditions (P = 0.001) but not species (P = 0.175) differed significantly between

halves (see Results).

In contrast to environmental conditions, the composition of species significantly differed among

the groups of five sites (100 m scale; F = 1.96, P = 0.001) but not between halves of the reserve

(5 km scale; F = 1.26, P = 0.168; Fig. 4.5b). Despite their lack of differences in species


65

composition, the two halves of the reserve had 37 species in common, but also 12 species that

were unique to the north half and 11 that were unique to the south half; these numbers sum to

less than 73 because some species were not observed in the 1 m treatment. These unique species

were significantly less associated with the axis 1 and 2 PCoA scores than species that were

common to both halves (P = 0.023), which potentially reconciles the high compositional

similarity among halves of the reserve despite the presence of unique species. We discuss the

biological interpretation of this point in the discussion.

Discussion

Many studies seek to identify and compare the factors that constrain species distributions across

landscapes, commonly through controlled experiments or statistical partitioning of observational

data. In a natural metacommunity, our seed pool manipulations resulted in two key findings: (i)

compared to natural patterns, the strength of environmental matching doubles when seeds were

distributed at large spatial scales, and (ii) non-linear diversity responses to the spatial scale of

seed pooling suggests that there are distinct transitions among spatial scales in the forces that

structure diversity. Our findings provide novel insight into how metacommunities and their

constituent species are structured across landscapes, and how that structure is constrained by

distance and the environment.

In the absence of seed pooling, most species in our study (~82%) were not significantly

associated with environmental conditions, consistent with the findings of numerous studies of

individual species (e.g., Borcard et al. 2004; Ramette & Tiedje 2007). This result might be

surprising to many ecologists, given the ubiquity of experimental work demonstrating the

strength of environmental determinants of species’ performances and the composition of

communities (e.g., Harpole & Tilman 1997; Silvertown et al. 2006). However, this type of

observational evidence alone cannot be used to identify the mechanisms underlying the pervasive

lack of environmental associations across species that we observed. The experimental

enhancement of dispersal via seed pool manipulations has allowed us to discriminate among

several potential mechanisms (Fig. 4.1), and supports a model in which dispersal limitation

precludes the colonization of suitable sites. Similar results have been obtained in the canary

violet (Viola praemora) using seed addition trials (Pinto & MacDougall 2010); we extend this


66

perspective beyond individual species to an entire plant community, and show a doubling in the

strength of species-environment relationships (Figs. 4.2b and S4.3) that explains nearly half of

the variation in how species are distributed (Fig. 4.2b). Further work is needed to understand the

factors that determine the distributions of species that remain poorly associated even after

dispersal limitation is removed.

Although not a characteristic of species themselves, we found a clear hump-shaped relationship

between the abundance of species pre-manipulation and how well-matched those species are to

the environment post-manipulation (Fig. 4.4). This makes intuitive sense, given that species

abundance is tied to the relative number of seeds it contributes to the seed pool. For rare species,

this means that seeds have a low probability of ending up in suitable sites when species pools are

homogenized, whereas common species have a high probability of ending up in all sites

regardless of suitability; the latter case could lead to transient or persistent establishment in low

suitability sites (Mouquet & Loreau 2003). Furthermore, the regional abundance of a given

species is in part determined by its degree of environmental specialization simply because

generalists can persist in more sites than specialists. For example, specialist Castilleja

rubicundula occurred in a single site of unusually high percent soil moisture (27.7%, compared

to 13.5% among-site mean ± 0.87 standard error), and was not significantly matched to

environmental conditions. These abundance-based effects are in part an artifact of our seed

pooling manipulation, but are analogous to what would occur in a metacommunity of species

that differ in propagule supply (Hurtt & Pacala 1995) and degree of specialization (Pandit et al.

2009).

Nearly all components of diversity that were examined showed consistent, nonlinear responses to

the spatial scale of seed pooling, including species richness (α diversity), species-environment

relationships, and compositional turnover (β diversity). In all cases, there were no differences in

responses between 1 and 5 m or 5 and 10 km spatial scales of seed pooling, but large differences

among <5 m, 100m, and >5 km, a result that is best explained by first considering the geology of

how serpentine soils are formed. Serpentine soils occur along fault lines and are formed by the

emergence and erosion of the Earth’s mantle; repeated events of emergence and erosion form a

mosaic of soils that are distinct in biochemistry and occur within a matrix of non-serpentine soil.

The resulting landscape is characterized by abrupt transitions in environmental conditions (see


67

photo in Fig. S4.5b), unlike systems where differences accumulate linearly across space

(Lechowicz & Bell 1991).

For plants, the unique geological history of serpentine systems has scale-specific consequences

for natural diversity patterns and responses to our seed pool manipulations. At small spatial

scales (<5 m), conditions are essentially homogenous and within the dispersal abilities of most

plant species in one or few generations (Thomson et al. 2011); our seed pool manipulations had

negligible effects at this scale. At intermediate scales (100 m), our manipulations forced seeds to

cross the boundaries of serpentine outcrops into environments that were highly dissimilar from

their source sites; this allowed the establishment of some species with broad environmental

tolerances. At large scales (>5 km), seeds were sampled from and allowed to access a broad suite

of conditions represented by many serpentine outcrops (Fig. 4.5a); our manipulations caused

diversity to respond at these scales because species were given access to suitable sites in which

they had previously been absent. Together, our evidence indicates that the non-linearity we

observed is tied to the probability that species will encounter suitable sites at each scale as

determined by the underlying structure of the environment.

A common challenge faced by ecologists is identifying the spatial scales at which

metacommunities are most appropriately delineated (Logue et al. 2011), given that the

appropriateness of these delineations are not easily testable (Thuiller et al. 2015) or comparable

among different study systems. In our experiment, we made no assumptions about the spatial

scales at which different ecological processes were operating, and instead tested a range of

spatial scales for signatures of community and metacommunity limits. Our results support the

delineation of regional metacommunities at scales of ~5 km due to the lack of diversity and

species-environment responses to our manipulations above this scale, results that are striking

given that the numbers of seeds that were pooled essentially doubled between the 5 and 10 km

scale. In other words, the imprint of dispersal limitation would not be observed at spatial extents

above 5 km, a conclusion that is concordant with a recent paper on plant metacommunities in the

French Alps (Thuiller et al. 2015). It is important to keep in mind that although these results are

specific to this particular study system, this study provides rare insight into the spatial

delineations of communities and metacommunities in an area that is considered a model system

for biodiversity research (Harrison & Rajakaruna 2011).


68

A limitation of this study is that the data was collected a year after our seed pooling

manipulations were carried out. This means that the diversity patterns we observed are not in an

equilibrium state, and may or may not persist through time. For example, source-sink dynamics

or competitive interactions could be of heightened or diminished importance in this first year

compared to subsequent years, respectively; our results suggest that these concerns are unlikely

to apply to our system. For example, the increase in species-environment relationships at larger

scales, when one might expect homogenizing source-sink effects to overwhelm any

environmental signal, is indicative of the interplay between dispersal limitation and

environmental heterogeneity that characterizes serpentine landscapes. Competitive interactions

tend to be strongest at the seedling stage (Goldberg et al. 2001), suggesting that our

measurements at peak biomass would not obscure the impacts of competition.

The forces that dictate how species are distributed across landscapes have long fascinated

ecologists. In the serpentine system, we used a novel method to identify scale-specific processes

that structure natural plant communities, and how this spatial scaling is explained by the

accumulation of species and environments across space. Because environmental conditions are

tightly linked to the geology of this system, our results highlight the role that geological

complexity plays in the maintenance of diverse species assemblages in a global biodiversity

hotspot. The role of geological complexity (or environmental complexity in general) has been

long-recognized by evolutionary biologists as a mechanism for the generation of biological

diversity (Golberg et al. 2011; Anacker & Strauss 2013; Cacho & Strauss 2014). Our work offers

the ecological counterpart on the forces that maintain diverse assemblages of species en route to

evolutionary change.

Acknowledgements

We thank Susan Harrison, Brian Anacker, Cathy Kohler, Caprice Lee, Erica Case, Stella

Copeland, Paola Arenas, and Laurren Sergison for advice and/or assistance in the field, and

Strauss and Gilbert lab members for project feedback. Funding to R. M. G. provided by NSERC-

CGS D (personal), NSERC-MSFSS (research), and Sigma Xi (research).


69

Literature cited

Anacker, B.L. (2014) The nature of serpentine endemism. American Journal of Botany, 101,

219-224.

Anacker, B.L. & Strauss, S.Y. (2013) The geography and ecology of plant speciation: range

overlap and niche divergence in sister species. Proceedings of the Royal Society B:

Biological Sciences, 281, 1-9.

Borcard, D., Legendre, P., Avois-Jacquet, C. & Tuomisto, H. (2004) Dissecting the spatial

structure of ecological data at multiple scales. Ecology, 85, 1826-1832.

Bartolome, J.W. (1979) Germination and seedling establishment in California annual grassland.

Journal of Ecology, 67, 273–281.

Bell, G., Lechowicz, M.J., Appenzeller, A., Chandler, M. & DeBlois, E. (1993) The spatial

structure of the physical environment. Oecologia, 96, 114-121.

Bender, D.J. & Fahrig, L. (2005) Matrix structure obscures the relationship between interpatch

movement and patch size and isolation. Ecology, 86, 1023-1033.

Cacho, N.I. & Strauss, S.Y. (2014) Occupation of bare habitats, an evolutionary precursor to soil

specialization in plants. Proceedings of the National Academy of Sciences, 111, 15131-

15137.

Caley, M.J. & Schluter, D. (1997) The relationship between local and regional diversity.

Ecology, 78, 70–80.

Cadotte, M.W. (2007) Concurrent niche and neutral processes in the competition-colonization

model of species coexistence. Proceedings of the Royal Society B: Biological Sciences,

274, 2739-2744.

Chase, J.M. (2010) Stochastic community assembly causes higher biodiversity in more

productive environments. Science, 328, 1388-1391.

Coiffait-Gombault, C., Buisson, E. & Dutoit, T. (2010) Hay transfer promotes establishment of

Mediterranean steppe vegetation on soil disturbed by pipeline construction. Restoration

Ecology, 19, 214–222.

Cottenie, K. (2005) Integrating environmental and spatial processes in ecological community

dynamics. Ecology Letters, 8,1175-1182.

Germain, R.M., Johnson, L., Schneider, S., Cottenie, K., Gillis, E.A. & MacDougall, A.S. (2013)

Stochastic variability in plant predation determines the strength of stochastic community

assembly. American Naturalist, 182, 169-179.

Gilbert, B. & Lechowicz, M.J. (2004) Neutrality, niches, and dispersal in a temperate forest

understory. Proceedings of the National Academy of Sciences, 101, 7651-7656.


70

Gilbert, B. & Levine, J.M. (2013) Plant invasions and extinction debts. Proceedings of the

National Academy of Sciences, 110, 1744–1749.

Gilpin, M.E. & Hanski, L.A. (1991) Metapopulation Dynamics: Empirical and Theoretical

Investigations. Academic Press, San Diego.

Goldberg, D.E., Turkington, R., Olsvig-Whittaker, L. & Dyer, A.R. (2001) Density dependence

in an annual plant community: variation among life history stages. Ecological

Monographs, 71, 423-446.

Grainger, T.N. & Gilbert, B. (2016) Dispersal and diversity in metacommunities: linking theory

and practice. Oikos early view: 10.5061/dryad.nq853

Harpole, W.S. & Tilman, D. (2007) Grassland species loss resulting from reduced niche

dimension. Nature, 446, 791-793.

Harrison, S. (1999) Local and regional diversity in a patchy landscape: native, alien, and

endemic herbs on serpentine. Ecology, 80, 70–80.

Hubbell, S. (2001) The unified neutral theory of biodiversity and biogeography. Princeton

University Press, Princeton.

Hurtt, G.C. & Pacala, S.W. (1995) The consequences of recruitment limitation: reconciling

chance, history and competitive differences between plants. Journal of Theoretical

Biology, 176, 1–12.

Harrison, S. & Rajakaruna, N. (2011) Serpentine: The Ecology and Evolution of a Model System.

University of California Press, Berkeley, CA.

Jackson, D.A. (1993) Stopping rules in principal components analysis: a comparison of

heuristical and statistical approaches. Ecology, 74, 2204-2214.

Kneital, J.M. & Chase, J.M. (2004) Trade-offs in community ecology: linking spatial scales and

species coexistence. Ecology Letters, 7, 69-80.

Lechowicz, M.J. & Bell, G. (1991) The ecology and genetics of fitness in forest plants. II.

Microspatial heterogeneity of the edaphic environment. Journal of Ecology, 79, 687-696.

Leibold, M.A., Holyoak, M., Mouquet, N., Amarasekare, P., Chase, J.M., Hoopes, M.F., Holt,

R.D., Shurin, J.B., Tilman, D., Loreau, M. & Gonzalez, A. (2004) The metacommunity

concept: a framework for multi-scale community ecology. Ecology Letters, 7, 601-613.

Logue, J.B., Mouquet, N., Peter, H. & Hillebrand, H. (2011) Empirical approaches to

metacommunities: a review and comparison with theory. Trends in Ecology and

Evolution, 26, 482-491.


71

Lowe-Decarie, E., Kolber, M., Homme, P., Lofanom A., Dumbrell, A., Gonzalez, A. & Bell, G.

(2015). Community rescue in an experimental metacommunity. Proceedings of the

National Academy of Sciences, 112, 14307-14312.

MacArthur, R.H. & Wilson, E.O. (1967) The Theory of Island Biogeography. Princeton

University Press, Princeton.

Mouquet, N. & Loreau, M. (2003) Community patterns in source-sink metacommunities.


Nathan, R. (2006) Long-distance dispersal of plants. Science, 313, 786-788.

Pandit, S., Kolasa, J. & Cottenie, K. (2009) Contrasts between habitat generalists and specialists:

an empirical extension to the basic metacommunity framework. Ecology, 90, 2253-2262.

Pinto, S.M. & MacDougall, A.S. (2010) Dispersal limitation and environmental structure interact

to restrict the occupation of optimal habitat. American Naturalist, 175, 675–686.

Ramette, A. & Tiedje, J.M. (2007) Multi-scale responses of microbial life to spatial distance and

species sorting in a patchy ecosystem. Proceedings of the National Academy of Sciences,

104, 2761-2766.

Silvertown, J., Poulton, P., Johnston, E., Edwards, G., Heard, M. & Biss, P.M. (2006) The Park

Grass experiment 1856-2006: its contribution to ecology. Journal of Ecology, 94, 801-

814.

Thomson, F.J., Moles, A.T., Auld, T.D. & Kingsford, R.T. (2011) Seed dispersal distance is

more strongly correlated with plant height than with seed mass. Journal of Ecology, 99,

1299-1307.

Thuiller, W., Pollock, L.J., Gueguen, M. & Münkemüller, T. (2015) From species distributions

to metacommunities. Ecology Letters, 18, 1321-1328.

Weins, J.A. (1989) Spatial scaling in ecology. Functional Ecology, 3, 385-397.

Whittaker, R.H. (1954) The ecology of serpentine soils. Ecology, 35, 258–288.

Neter, J., Wasserman, W. & Kutner, M.H. (1989) Applied Linear Regression Models, 2nd

edition.

Irwin Homewood, IL.

CHAPTER 5: MACROEVOLUTION AND COMPETITIVE DIVERGENCE

72

Chapter 5

Species coexistence: macroevolutionary relationships and the

contingency of historical interactions

Published as Germain, R.M., Weir, J.T. & B. Gilbert. 2016. Species coexistence:

macroevolutionary relationships and the contingency of historical interactions. Proceedings of

the Royal Socierty B: Biological Sciences, 283, 20160047.

Abstract

Evolutionary biologists since Darwin have hypothesized that closely related species compete

more intensely and are therefore less likely to coexist. However, recent theory posits that species

diverge in two ways: either through the evolution of ‘stabilizing differences’ that promote

coexistence by causing individuals to compete more strongly with conspecifics than individuals

of other species, or through the evolution of ‘fitness differences’ that cause species to differ in

competitive ability and lead to exclusion of the weaker competitor. We tested macroevolutionary

patterns of divergence by competing pairs of annual plant species that differ in their phylogenetic

relationships, and in whether they have historically occurred in the same region or different

regions (sympatric vs. allopatric occurrence). For sympatrically-occurring species pairs,

stabilizing differences rapidly increased with phylogenetic distance. However, fitness differences

also increased with phylogenetic distance, resulting in coexistence outcomes that were

unpredictable based on phylogenetic relationships. For allopatric species, stabilizing differences

showed no trend with phylogenetic distance, whereas fitness differences increased, causing

coexistence to become less likely among distant relatives. Our results illustrate the role of

species’ historical interactions in shaping how phylogenetic relationships structure competitive

dynamics, and offer an explanation for the evolution of invasion potential of non-native species.

Introduction

The diversity of ecological interactions on Earth is the product of ~3.5 billion years of evolution,

with ongoing extinctions matched by the continual divergence of populations and species.

Signatures of this past evolution frequently emerge in the strength of the interactions among


73

current-day species (Peay et al. 2012) in ways that have potential to further perpetuate

divergence and the evolution of interaction strengths (Meyer & Kassen 2007). This dynamic

feedback between the ecology and evolution of organisms is a central theme in

microevolutionary (Gravel et al. 2011; Zuppinger-Dingley et al. 2014), macroevolutionary

(Mahler et al. 2013), and recent ecological perspectives (Johnson & Stinchcombe 2007; Webb et

al. 2002 ; Schoener 2011), as it promises a more complete picture of the processes that generate

and maintain biological diversity.

A longstanding hypothesis in evolutionary biology states that closely related species are more

ecologically similar, and that this similarity leads to intense competition and ultimately exclusion

(Webb et al. 2002; Darwin 1859; Elton 1946; Harper et al. 1961). Despite abundant evidence

that closely related species tend to be similar in terms of functional traits and resource

requirements (Violle et al. 2011), the effect of evolutionary relatedness on the outcome of

competition tends to be weak or absent in experimental (Cahill et al. 2008) and observational

(Anderson et al. 2011; Vamosi et al. 2009) tests. This might occur for two reasons: First,

contemporary ecological theory (Chesson 2000; Mayfield & Levine 2010) suggests that species

may evolve two types of differences, ‘stabilizing differences’ and ‘fitness differences’, that have

opposing effects on competitive outcomes. Stabilizing differences, also known as ‘niche

differences’ (Mayfield & Levine 2010), promote coexistence by causing negative-frequency

dependence in interacting species, and act to stabilize diversity by preventing any one species

from dominating the community. Fitness differences, by contrast, are inequalities in mean

fitness, including intrinsic demographic rates and competitive abilities, which preclude

coexistence by favoring a single dominant species (Chesson 2000; Mayfield & Levine 2010).

Because character displacement may cause species to diverge in both stabilizing differences and

fitness differences, the relationship between coexistence and evolutionary time depends on their

relative evolutionary trajectories (Mayfield & Levine 2010; Narwani et al. 2013; Godoy et al.

2014).

The second reason for inconsistent effects of evolutionary relationships on coexistence and

diversity might stem from the high occurrence of non-native species in many contemporary

communities. Research on coevolution and adaptive radiations predicts that the divergence of

related species depends on a history of competitive interactions, and therefore on whether species


74

have historically occurred in sympatry or in allopatry (Abrams 1986; Thompson 1999; Dayan &

Simberloff 2005). By occurring in sympatry, we refer to the idea that species have had ample

opportunity to interact and potentially influence evolutionary trajectories, rather than a specific

mode of speciation. For the vast majority of taxa, evolutionary histories are not sufficiently

documented to know when species historically occurred in sympatry or allopatry. However,

information on species’ native status in a region and current-day distributional overlap can be

used to identify species with a history of potential interaction; interactions between pairs of

species native to either the same or different regions can therefore be contrasted as a proxy for

the influence of coevolutionary history.

We grew 30 Mediterranean annual plant species both alone and in two-species competition to

estimate stabilizing and fitness differences (Godoy & Levine 2014) in two commonly

encountered environments (wet and dry), and used phylogenetic relationships among species

pairs to test whether the evolutionary trajectories of stabilizing differences, fitness differences,

and coexistence depend on coevolutionary history. We predicted that stabilizing differences

would increase rapidly and predictably with phylogenetic distance among species pairs that have

occurred in sympatry, whereas fitness differences would be constrained (e.g., red queen

hypothesis (Stenseth 1979) and competitive disarmament (Kisdi & Geritz 2001)). In contrast, we

predict a weak or absent relationship between stabilizing differences and phylogenetic distance

for allopatric species, due to the lack of coevolutionary history, and no constraints on fitness

differences. These predictions would correspond to positive and negative relationships between

coexistence and phylogenetic distance, depending on whether species pairs originated from the

same or different biogeographic region, respectively.


Species selection

Species were selected to meet two criteria: First, species of the same biogeographic origin must

have had common affinities for annual grasslands and overlapping geographic distributions

across their native ranges, as determined by CalFlora (http://www.calflora.org; California) and

Euro+Med (http://www.emplantbase.org; Spain) plant databases, to represent a realistic subset of

species that could have potentially interacted over evolutionary time. There were no criteria




75

necessitating the overlap or non-overlap of species of different biogeographic origins in that the

species pool was not designed to discriminate species based on invasion history or lack thereof.

In total, 30 species were included, 20 of which were native to the Central Valley in California

and 10 of which were native to the Mediterranean Basin region of southern Spain (Table S5.1).

These regions were selected from amongst all Mediterranean-climate regions because of their

high similarity in climate, especially rainfall (Germain & Gilbert 2014). Second, species must

have an annual lifecycle to estimate lifetime seed production in a single growing season. We

constructed a phylogenetic tree using Bayesian methods and the ITS1/5.8S/ITS2 and rbcL

sequence regions, as detailed in the Supplementary methods and Fig. S5.1. Seeds were acquired

from independent donors and commercial suppliers (Germain & Gilbert 2014), and were tested

for % germinability under similar soil conditions prior to experimentation. We chose not to cold

stratify the seeds because this method is known to induce rather than break dormancy in

Mediterranean annuals (Baskin & Baskin 2001).

Greenhouse growing conditions

In January 2012, seeds were sown into 12.7 cm diameter, 23 cm deep treepots filled with a 3:2

mixture of sand and screened topsoil, to mimic the sandy loam soils found in annual grasslands.

The topsoil was collected locally (Villacci’s Garden Depot; Scarborough, ON, CA) to ensure that

species were equally naïve to the soil microbiome regardless of origin, a requirement for

unbiased biogeographic comparisons in a common garden experiment. Each pot was randomly

assigned to a position on a bench in the rooftop greenhouse at the University of Toronto

(Toronto, Canada). The greenhouse was maintained at day/night temperatures of 14/7°C, which

was set to gradually increase to 29/17°C on average by the end of the experiment. High intensity

discharge (HID) lighting was provided to maintain a 12 h photoperiod. Each pot was watered

daily to saturation using a drip irrigation system during a three-week establishment period.

After the establishment period, pots of plants were randomly assigned to either a wet or dry soil

moisture regime. Using a drip irrigation system, pots in the wet treatment received 175 mL water

twice as often as those in the dry treatment, starting at one and two day intervals which were

extended to seven and 14 days as the growing season progressed. These two soil moisture

regimes were selected to simulate realistic among-year differences in precipitation, or equally,


76

the 30-year average differences between mesic [662 mm] and dry [312 mm] sites across the

species ranges (http://www.climate-charts.com). We confirmed that the wet (11.1 ± 0.56%; mean

± SE percent soil moisture content) and dry (5.8 ± 0.42%) treatments were effective using a

volumetric water probe (HydroSense™, Campbell Scientific Australia) on empty pots. Prior to

flowering (~60 days after planting), each pot was provided with 350 mL of 1500 ppm 20-20-20

NPK fertilizer (Plant Products, Inc., Brampton, Ontario). Pollination was provided by

commercial colonies of the generalist pollinator Bombus impatiens, which were active

throughout the flowering period (Biobest Canada, Leamington, Ontario). The experiment lasted

220 days owing to our inclusion of some summer annual species with relatively long lifecycles

(Atriplex patula, Chenopodium berlandieri, Crepis capillaris, and Madia elegans).

Experimental design

The experiment was designed to parameterize a series of annual plant models (Levine &

HilleRisLambers 2009). Here we present the model that best fit our data (based on AICc, Table

S5.2), the Beverton-Holt model,

log (𝑁𝑖 𝑡+1

𝑁𝑖 𝑡) = log(𝜆𝑖) − log (1 + 𝛼𝑖𝑖𝑁𝑖 𝑡 + 𝛼𝑖𝑗𝑁𝑗 𝑡) (5.1)

where Ni t and Nj t are the numbers of viable seeds of focal species i and j initially planted, λi is

the finite rate of increase for species i in the absence of competition, αij and αii are the inter- and

intra-specific competition coefficients, respectively, and Ni t+1 is the number of viable seeds of

focal species i in year t+1. Because this equation is symmetric, Nj t+1 can also be calculated by

switching subscripts i and j. To independently estimate λ, we grew each species alone at low

densities and enumerated the number of seeds produced per individual; other models (Table

S5.2) require a similar parameterization for finite rate of increase in low competition.

Specifically, 30 seeds of each species were sown into the pots, and emerged seedlings were

thinned to ~eight maximally-spaced individuals that were allowed to mature and produce seed.

There were seven replicate pots of this low density treatment, and these replicates were used to

calculate the distribution of log-transformed λ for each combination of species × soil moisture

environment (details in Appendix D) to be used as informative priors in our Bayesian analysis

(described below).


77

To estimate competition coefficients (αii, αjj, αij, and αji), we grew species in pairwise

competition in pots at an expected density of 70 plants (based on germination rates from pilot

experiments). This density is comparable to the seeding density found in annual grasslands (2500

to 5500 plants m-2

(Harrison 1999; Bartolome 1979)). Within these pots, we varied the numbers

of individuals of the two species (10:60, 20:50, 30:40, 40:30, 50:20, and 60:10 individuals) to

create six relative frequency ratios, and had two replicates per frequency ratio. The presence of

strong negative frequency-dependence would indicate that stabilizing differences are large

between competing species (Adler et al. 2007).

We used Bayesian modeling (JAGS software v. 3-15, implemented in the ‘rjags’ R package) to

fit annual plant models to our data, using uninformative priors for all fitted parameter estimates,

except log(λ) for which we used the distributions calculated earlier. Similar to previous work

(Godoy & Levine 2014), all competition coefficients (α parameters in Eqn. 5.1) were log-

normally distributed, and other parameters (such as b in models 4 and 6, Table S5.2) were

uniformly distributed but constrained to be positive. We ran four independent Markov chains for

100,000 iterations, with a 10,000 iteration burn-in period based on time to convergence; for each

parameter, the mean of the posterior distribution was used as the best estimate. We then

compared AICc values for the various models by calculating likelihoods for each model using

parameter means. Once we selected the best model (the Beverton-Holt model, as has been

commonly fit in other annual plant studies (Godoy & Levine 2014; Levine & HilleRisLambers

2009)), the estimated parameters were used to calculate stabilizing and fitness differences using

Eqs. S5.1 and S5.2, which are further described below. The entire experiment was replicated in

two soil moisture environments (wet and dry, see Greenhouse growing conditions), and included

900 total pots of plants arranged in a completely randomized design.

We did not compete all possible pairs of the 30 species included in the experiment. Rather, ten

sympatric (both species native to California) and ten allopatric (one species native to California,

the other to Spain) pairs were selected to represent competitive pairs that were phylogenetically

independent (i.e., non-overlapping branch lengths) relative to all other pairs of the same

biogeographic history treatment. Additionally, the same ten Californian focal species were used

in both the sympatric and allopatric pairs, competed against ten other unique species from

California or Spain (e.g., Vulpia microstachys vs. V. octoflora [sympatric] or V. myuros


78

[allopatric]); this was accounted for in the model fitting, with the parameter estimates of all three

species being fit simultaneously (Table S5.3).

Solving for stabilizing differences, fitness differences, and coexistence outcomes

Stabilizing (1-ρ; Eq. S5.1) and fitness (κj/κi; Eq. S5.2) differences were estimated according to

Godoy & Levine (2014) by rearranging the parameters from the Beverton-Holt annual plant

model described in Eq. 5.1. When the strength of intraspecific and interspecific competition is

the same, then √(𝛼𝑖𝑗/𝛼𝑗𝑗 ∙ 𝛼𝑗𝑖/𝛼𝑖𝑖) = 1 and 1-ρ = 0, indicating that there are no stabilizing

differences between competitors. As the relative strength of intraspecific to interspecific

competition increases [√(𝛼𝑖𝑗/𝛼𝑗𝑗 ∙ 𝛼𝑗𝑖/𝛼𝑖𝑖) << 1], 1-ρ approaches 1 indicating that stabilizing

differences are large. In five of 40 cases (20 species pairs × two soil moisture environments), 1-ρ

was < 0, meaning that these species pairs showed evidence of destabilizing effects that are

indicative of priority-based competitive outcomes. For simplicity, we followed the convention of

setting these values to 1 (complete niche overlap (Godoy & Levine 2014)), as this allows

comparison of coexistence outcomes. κj/κi is the average fitness difference between species i and

j, calculated as the product of the demographic ratio ((𝜆𝑗 − 1)/(𝜆𝑖 − 1)), and the competitive

response ratio (√(𝛼𝑖𝑗 ∙ 𝛼𝑖𝑖)/(𝛼𝑗𝑗 ∙ 𝛼𝑗𝑖) ). Our estimate of fitness differences differs from previous

work [23] in that we use Κ, the larger of κj/κi and κi/κj (Eq. S5.3). This was done to simplify the

presentation of our results, but does not alter their outcome or interpretation. Mathematical

derivations of these equations can be found in the appendices of Godoy and Levine (2014).

Species are predicted to coexist locally if both can invade when rare and the other species is at its

equilibrium density, a criterion that is met when (Chesson 2000; Godoy & Levine 2014):

1

𝜌𝛫> 1 (5.2)

In the main text, we use the logarithm of the left-hand side of Eq. 5.2 as our coexistence metric,

so that values greater than zero indicate coexistence whereas those less than zero indicate

competitive exclusion. It is important to note that we consider a model of coexistence in which

all ungerminated seeds are considered inviable, thereby ignoring the contribution that

ungerminated seeds may make to fitness (Godoy & Levine 2014) or to inter-annual stabilization


79

through the storage effect (Chesson & Huntly 1989). We tested the sensitivity of our results

against two alternative models, where the seed bank was set to either realistic species-specific

rates (Table S5.4) or 100% viability to determine Κ, as used previously (Godoy & Levine 2014);

our results are qualitatively similar across all three models (Fig. S5.2). The species-specific rates

for the first alternative model were estimated via a separate germination trial using Petri dishes

and gibberellic acid application (Germain & Gilbert 2014).

Statistical analyses

We analyzed stabilizing differences, fitness differences, and the coexistence metric using linear

mixed effects (LME) models with the R packages ‘lmerTest’ and ‘nlme’. All response variables

were tested against a model with phylogenetic distance, biogeographic history, soil moisture

treatment, and their interactions as fixed effects, and species pair as the random effect since the

same pairs of species were grown in each soil moisture treatment. To meet model assumptions

(linearity and homoscedasticity of errors), stabilizing differences were logit-transformed and

tested with the ‘lmer’ function, whereas fitness differences and the coexistence metric were both

log-transformed and coded to include heteroscedastic variance structure (‘weights’ argument in

the ‘lme’ function). The results of the LMEs were summarized using function ‘Anova’ (type II

analysis of variance) in the ‘car’ package; we report the significant highest-order interactions

only in the main text, and the full model outputs in Tables S5.5 and S5.6. Reported P-values are

calculated from χ2 tests of maximum likelihood ratios.

Results and discussion

Consistent with evolutionary theory, our results support the hypothesis that the

macroevolutionary trajectories of stabilizing differences are mediated by biogeographic history

(phylogenetic distance by biogeographic history interaction χ2

1 = 8.75, P = 0.003). Species that

have had potential to evolve in sympatry rapidly accumulated stabilizing differences with

phylogenetic distance, resulting in nearly complete ecological separation by 90 million years

(Fig. 5.1a, light shade). This was not observed in species pairs that have been evolving in

allopatry (i.e. in California vs. Spain); stabilizing differences showed no relationship with

phylogenetic distance (Fig. 5.1a, dark shade). These lines of phylogenetic evidence are consistent

with theory that competition imposes selective pressure for divergence in resource use among


80

sympatrically-evolving species (Dayan & Simberloff 2005). A relationship would not be

expected between allopatric species pairs because they are evolutionarily naïve to each other;

species native to separate regions have experienced distinct coevolutionary trajectories, such that

stabilizing differences in the native range would not be predictive of those in the introduced

range.

Figure 5.1. Biogeographic history alters the evolutionary trajectory of stabilizing and fitness

differences. (a) Stabilizing differences rapidly increase among sympatric species pairs (light

shade), whereas allopatric species pairs (dark shade) show no relationship. (b) Fitness

differences, by contrast, increase over evolutionary time in both sympatric and allopatric pairs,

but are larger on average among allopatric pairs. Stabilizing differences have a maximum of one

(Eq. S5.1; lines are fitted from the logit-transformed data), whereas fitness differences have no

upper limit (Eq. S5.2). Because soil moisture had no effect on stabilizing or fitness differences,

each point is a fitted average across soil moisture environments for each species pair.

Our results are striking and appear to contradict the few existing experimental studies that test

whether stabilizing differences are explained by phylogenetic relatedness, and find no

relationship (Narwani et al. 2013; Godoy et al. 2014). This apparent contradiction might be

explained by two methodological differences between our study and those that precede it. First,

our study is the first to our knowledge to incorporate information on historical species

distributions. In rerunning our analysis without accounting for differences in biogeographic

history, we find that phylogenetic distance fails to explain variation in stabilizing differences (χ2

1


81

= 0.10, P = 0.756), a result more in line with previous studies. Second, our competitive pairs

were selected to be phylogenetically independent (see Materials and methods). This approach is

necessary because any time there is overlap in evolutionary history among contrasted pairs, the

number of comparisons made will be greater than the number of independent observations from

a phylogenetic perspective. This limits the inferential power of many observational and

experimental tests (Harmon & Glor 2010).

Similarly to stabilizing differences, we found that the pattern of past evolution of fitness

differences depended on biogeographic history. Specifically, fitness differences increased as an

accelerating function of phylogenetic distance [χ2

1 = 17.72, P < 0.001] and were greater overall

amongst allopatric species pairs [χ2

1 = 5.82, P = 0.016; Fig. 5.1b]. In other words, just as co-

evolutionary dynamics lead to a greater probability of stabilizing coexistence, they appear to

constrain the degree of fitness differences that lead to competitive exclusion across the entire

phylogeny.

The effect of biogeographic history on fitness differences could arise from factors other than

coevolutionary dynamics if, for example, the growing conditions in our experiment were more

similar to the ambient environment in either California or Spain. If this were the case, we would

expect absolute fitness differences (κj/κi) to be biased towards species from a particular region.

Further investigation suggests that this is not the case; a post-hoc test showed no consistent

fitness advantage for a particular region (χ2

1 = 0.42, P = 0.515). It appears that the effect of

biogeographic history we observe reflects differences in historical interactions, rather than

experimental conditions favoring species from one region.

Fitness differences arise from the combined effects of differences in demographic rates and

differences in competitive abilities, both of which have been hypothesized to lead to competitive

exclusion by invasive species (Daehler 2003). Despite reports that many invasive species differ

in demographic characteristics, such as seed production (van Kleunen et al. 2010), we found that

allopatric species pairs significantly diverged in competitive ability but the trend in demographic

rates was not significant (Fig. 5.2b, dashed vs. dotted line; Table S5.6). By contrast, sympatric

species pairs significantly diverged in demographic rates but not competitive ability (Fig. 5.2a,

dotted vs. dashed line; Table S5.6), with divergence in overall fitness differences matching


82

divergence in demographic rates. In other words, species that have evolved in sympatry or

allopatry both diverge in fitness differences over evolutionary time, but the fitness component

responsible for this divergence is distinct.

Figure 5.2. Fitness differences arise through alternate mechanisms in sympatric and allopatric

species pairs. In log-space, fitness differences (solid line) are the sum of competitive (dashed

line) and demographic (dotted line) components (see Materials and methods), shown here as

fitted relationships. (a) In sympatric species pairs, fitness differences are minimal relative to

allopatric species pairs, and appear to accumulate over macroevolutionary time primarily through

differences in demographic rates. (b) In allopatric species pairs, the more rapid increase in fitness

differences over evolutionary time is driven solely by divergence in competitive ability, as the

apparent divergence in demographic rates is non-significant. Values in the shaded area indicate

that the species with the highest fitness had the lowest fitness component.

The relationship between species coexistence and phylogenetic distance ultimately depends on

biogeographic history (phylogenetic distance and biogeographic history interaction χ2

1 = 3.89, P

= 0.049), due to differing evolutionary trajectories of stabilizing and fitness differences. For

species pairs that occur in allopatry, coexistence is most likely to occur between close relatives,

and becomes increasingly unlikely over macroevolutionary time (Fig. 5.3, dark shade). In

sympatric species pairs, by contrast, the past evolution of stabilizing and fitness differences has

counteractive effects on coexistence that are nearly perfectly matched, resulting in local


83

coexistence outcomes that are random with respect to phylogeny (Fig. 5.3, light shade; slope = -

0.02 ± 0.01 standard error). Interestingly, our coexistence metric was less variable among

sympatric than among allopatric species pairs, and generally straddled the threshold between

coexistence and exclusion (Fig. 5.3, light shade); in three cases, the soil moisture treatment alone

was enough to cause sympatric species pairs to cross this threshold so that species were predicted

to coexist in at least one environment. Our results highlight the importance of environmental

variation in maintaining species diversity among native species, but suggest that this variation

may be less successful in maintaining diversity among species from different regions.

Figure 5.3. The effect of evolutionary history on coexistence outcomes depends on

biogeographic history. Species coexistence is not influenced by the phylogenetic distances of

species pairs that occur in sympatry (light shade, slope not significantly different from zero), but

the probability of coexistence decreases with the phylogenetic distance of allopatric pairs (dark

shade). Species pairs were grown in wet (squares) and dry (triangles) environments, but soil

moisture does not alter the effect of evolutionary history on coexistence (fitted lines are averaged

across environments). The dashed line indicates the threshold between coexistence (positive

values) and competitive exclusion (negative values); the coexistence metric is given by Eq. 5.2 in

Materials and methods. Points connected by a solid line represent cases in which a species pair

was predicted to coexist in one but not both environments (3 of 10 sympatric pairs, 0 of 10

allopatric pairs).

The use of two sets of environmental conditions in deconstructing the components of species’

competitive dynamics is a strength of our experiment, and is the first to do so to our knowledge;


84

the sensitivity of these components to the underlying environment is heretofore unknown (Kraft

et al. 2014). We found that the soil moisture environment did not affect stabilizing differences,

fitness differences, or coexistence (all P > 0.15), and did not influence how these variables

responded to biogeographic history or phylogenetic distance. The presence of an effect would

have indicated that certain environments cause species to overlap more or less in resource use

(affecting stabilizing differences) or to have stronger or weaker competitive asymmetries

(affecting fitness differences (HilleRisLambers et al. 2012)). Instead, the varied responses

observed across species pairs likely reflect species-specific differences in responses to soil

moisture limitation (Germain & Gilbert 2014); in a dry environment, for example, some species

pairs might experience more overlap in resource use, whereas others might experience less

overlap. Although we do not have the data to identify the exact mechanism of species-specific

differences, it likely has to do with whether or not traits relevant to competition converge or

diverge between species in different environments. Whether the same results would be obtained

in response to other environmental conditions that, unlike soil moisture, are not also an essential

resource, warrants further investigation.

A caveat to our interpretation of our findings is that we can only infer past histories of interaction

from present-day distributional data (Connell 1980), given that macroevolutionary change takes

place on timescales that are not directly observable. For this reason, it is important that we weigh

our results against explanations other than competition for the patterns we observed, such as

neutral evolution or specialization on different environment types. The predictions expected

given these alternative mechanisms differ from those under competition in two key ways. First, if

environmental specialization alone was responsible for divergence, then species would show

clear differences in habitat association, or in their abilities to persist in a common environment in

the absence of competition; neither was true for the species in our experiment. Second, under

either neutral evolution or environmental specialization, stabilizing differences would not be

expected to evolve differently between sympatric and allopatric species pairs, nor would they be

constrained to the narrow range of high stabilization that we observed among sympatric pairs.

For these reasons, we argue that although we cannot definitively rule out alternative

mechanisms, historical interactions likely constrain the evolution of competitive similarities and

dissimilarities among species.


85

Regardless of the mechanism of divergence, the simultaneous evolution of differences that

promote and prevent coexistence, as inferred through phylogenetic relationships, provides new

insight into the diverse patterns of evolutionary relatedness found in natural communities.

Specifically, we find no evidence that closely related species are less likely to coexist; for species

that occur in sympatry, we show that the effect of evolutionary relatedness on coexistence is

unpredictable even though the effects of evolution on the underlying determinants of coexistence

are well understood. This result contradicts common interpretations of over a decade of

observational work in ecology, where patterns of phylogenetic dissimilarity (i.e.,

‘overdispersion’) in communities are typically considered evidence of competitive filtering

(Webb et al. 2002). Although this interpretation has been called into question repeatedly in

recent years (Mayfield & Levine 2010; Kraft et al. 2007), our evidence is consistent with other

recent work (Narwani et al. 2013; Godoy et al. 2014) that competition does not likely result in

phylogenetic overdispersion, at least in our annual plant system. Instead, competition is most

likely to generate patterns of phylogenetic similarity (i.e., ‘underdispersion’) in communities

containing mixed-provenance species, and indeed a survey of the literature (Vamosi et al. 2009)

finds evidence of phylogenetic underdispersion in ~60% of published studies.

Our experiment has implications for understanding whether interactions among species from

different regions are fundamentally different from interactions among species from a common

region. Observational studies of plant invasions have produced seemingly contradictory results;

distantly related plants are less likely to establish upon introduction (Diez et al. 2008) but

become noxious invaders more frequently if they do establish (Strauss et al. 2006), compared to

close relatives which tend to naturalize (Diez et al. 2008; Strauss et al. 2006). In our study,

distant relatives from Spain had much lower or much higher mean fitness than competitors from

California, a result that reconciles previous work (Diez et al. 2008; Strauss et al. 2006).

Specifically, our results suggest that at the earliest stage of invasion, divergence in fitness may

generally predict why some species fail to establish (distant relatives of lower fitness (Diez et al.

2008)) while others have spectacularly negative impacts on native diversity (distant relatives of

higher fitness (Strauss et al. 2006)). Although this result does not establish plant characteristics

that make some species noxious invaders (van Kleunen et al. 2010), it supports the general


86

finding that species have a greater potential to become noxious invaders when they are naïve to a

region (Simberloff et al. 2012).

Our patterns of rapid predictable divergence in fitness differences but not stabilizing differences

among species from different floras has been hypothesized (MacDougall et al. 2009), but has

remained untested until now. Future work is needed to identify the traits that underlie stabilizing

and fitness differences among non-native competitors. Although specific traits have been

implicated as contributors to species’ competitive differences (Kraft et al. 2015) and invasion

success (van Kleunen et al. 2010), the potential for traits to differentially contribute to stabilizing

and fitness differences depending on biogeographic history is heretofore unexplored.

The intricate interplay between the ecology and evolution of organisms remains an important

area of research for understanding diversity and its response to global changes, such as species

invasions. In this study, we have tested one of the most long-standing hypotheses about the

relationships between evolutionary relatedness, competition, and coexistence. Our work

highlights the role that historical interactions play in determining the stability of current-day

interactions and the impacts of non-native species.

Data accessibility

The raw data used to fit our annual plant models and our nexus tree file have been made

publically available on Dryad (doi:10.5061/dryad.v2211); the fitted parameter estimates, seed

viability data, and Genbank sequence accessions are in the electronic supplementary materials.

Authors’ contributions

R.M.G., B.G. and J.W. designed the experiment, R.M.G. collected the data, R.M.G. and B.G.

analyzed the data, R.M.G. and J.T.W. constructed the phylogeny, R.M.G., B.G., and J.T.W.

wrote the manuscript.

Funding statement

Funding was provided by NSERC Discovery (research; B.G.) and NSERC-CGSD (personal

funding; R.M.G.).


87

Acknowledgements

We thank M. Cadotte and members of the Gilbert lab for their feedback, as well as two

anonymous reviewers for their comments improving the manuscript. Greenhouse assistance was

provided by B. Hall, A. Petri, and many undergraduate assistants, most notably C. Blackford.

Bees were provided by J. Thomson.

Literature cited

Abrams, P.A. (1986) Character displacement and niche shift analyzed using consumer-resource

models of competition. Theoretical Population Biology, 29, 107–160.

Adler, P.B., HilleRisLambers, J. & Levine, J.M. (2007) A niche for neutrality. Ecology Letters,

10, 95–104.

Anderson, T.M., Shaw, J. & Olff, H. (2011) Ecology’s cruel dilemma, phylogenetic trait

evolution and the assembly of Serengeti plant communities. Journal of Ecology, 99, 797–

806.

Bartolome, J.W. 1979 Germination and seedling establishment in California annual grassland.

Journal of Ecology, 67, 273–281.

Baskin, C.C. & Baskin, J.M. (2001) Seeds: Ecology, Biogeography, and Evolution of Dormancy

and Germination. Academic Press, San Diego.

Bell, C.D., Soltis, D.E. & Soltis, P.S. (2010) The age and diversification of the angiosperms re-

revisited. American Journal of Botany, 97, 1296–1303.

Cahill, J.F., Kembel, S.W., Lamb, E.G. & Keddy, P.A. (2008) Does phylogenetic relatedness

influence the strength of competition among vascular plants? Perspectives on Plant

Ecolology, Evolution, and Systematics, 10, 41–50.

Chesson, P. & Huntly, N. (1989) Short-term instabilities and long-term community dynamics.

Trends in Ecology and Evolution, 4, 293–298.

Chesson, P. (2000) Mechanisms of maintenance of species diversity. Annual Review of Ecology

and Systematics, 31, 343–366.

Connell, J.H. (1980) Diversity and the coevolution of competitors, or the ghost of competition

past. Oikos, 35, 131–138.

Daehler, C.C. (2003) Performance comparisons of co-occurring native and alien invasive plants:

implications for conservation and restoration. Annual Review of Ecology, Evolution, and



88

Darwin, C. (1859) On the Origin of Species by Means of Natural Selection, or the Preservation

of Favoured Races in the Struggle for Life. John Murray, London.

Dayan, T. & Simberloff, D. (2005) Ecological and community-wide character displacement: the

next generation. Ecology Letters, 8, 875–894.

Diez, J.M., Sullivan, J.J., Hulme, P.E., Edwards, G. & Duncan, R.P. (2008) Darwin’s

naturalization conundrum: dissecting taxonomic patterns of species invasions. Ecology

Letters, 11, 674–681.


Ecology, 15, 54–68.

Germain, R.M. & Gilbert, B. (2014) Hidden responses to environmental variation: maternal

effects reveal species niche dimensions. Ecology Letters, 17, 662–669.

Godoy, O. & Levine, J.M. (2014) Phenology effects on invasion success: insights from coupling

field experiments to coexistence theory. Ecology, 95, 726–736.



Gravel, D., Bell, T., Barbera, C., Bouvier, T., Pommier, T., Venail, P. & Mouquet, N. (2011)

Experimental niche evolution alters the strength of the diversity-productivity relationship.

Nature, 469, 89–92.

Harmon, L.J. & Glor, R.E. (2010) Poor statistical performance of the Mantel test in phylogenetic

comparative analyses. Evolution, 64, 2173–2178.

Harper, J.L., Clatworthy, J.N., McNaughton, I.H. & Sagar, G.R. (1961) The evolution and

ecology of closely related species living in the same area. Evolution, 15, 209–227.

Harrison, S. (1999) Local and regional diversity in a patchy landscape: native, alien, and

endemic herbs on serpentine. Ecology, 80, 70–80.

HilleRisLambers, J., Adler, P.B., Harpole, W.S., Levine, J.M. & Mayfield, M.M. (2012)

Rethinking community assembly through the lens of coexistence theory. Annual Review of

Ecology, Evolution, and Systematics, 43, 227-248.

Johnson, M.T. & Stinchcombe, J.R. (2007) An emerging synthesis between community ecology

and evolutionary biology. Trends in Ecolgy and Evolution, 22, 250–257.

Kisdi, E. & Geritz, S.A.H. (2001) Evolutionary disarmament in interspecific competition.

Proceedings of th Royal Society B: Biological Sciences, 268, 2589–2594.

Kraft, N.J.B., Adler, P.B., Godoy, O., James, E.C., Fuller, S. & Levine, J.M. (2014) Community

assembly, coexistence and the environmental filtering metaphor. Functional Ecology, 29,

592-599.


89

Kraft, N.J.B., Cornwell, W.K., Webb, C.O. & Ackerly, D.D. (2007) Trait evolution, community

assembly, and the phylogenetic structure of ecological communities. American Naturalist,

170, 271–283.

Kraft, N.J.B., Godoy, O. & Levine, J.M. (2015) Plant functional traits and the multidimensional

nature of species coexistence. Proceedings of the National Academy of Sciences, 112,

797-802.



MacDougall, A.S., Gilbert, B. & Levine, J.M. (2009) Plant invasions and the niche. Journal of

Ecology, 97, 609–615.

Mahler, D.L., Ingram, T., Revell, L.J. & Losos, J.B. (2013) Exceptional convergence on the

macroevolutionary landscape in island lizard radiations. Science, 341, 292–295.


phylogenetic structure of communities. Ecology Letters, 13, 1085–1093.

Meyer, J.R. & Kassen, R. (2007) The effects of competition and predation on diversification in a

model adaptive radiation. Nature, 446, 432–435.

Narwani, A., Alexandrou, M.A., Oakley, T.H., Carroll, I.T. & Cardinale, B.J. (2013)



Peay, K.G., Belisle, M. & Fukami, T. (2012) Phylogenetic relatedness predicts priority effects in

nectar yeast communities. Proceedings of the Royal Society B: Biological Sciences, 279,

749–758.

Schoener, T.W. (2011) The newest synthesis: understanding the interplay of evolutionary and

ecological dynamics. Science, 331, 426–429.

Simberloff, D., Souza, L., Nunez, M.A., Barrios-Garcia, M.N. & Bunn, W. (2012) The natives

are restless, but not often and mostly when disturbed. Ecology, 93, 598-607.

Stenseth, N.C. (1979) Where have all the species gone? On the nature of extinction and the Red

Queen Hypothesis. Oikos, 33, 196–227.

Stevens, P.F. (in press) Angiosperm Phylogeny Website. Version 12, July 2012 [and more or less

continuously updated since].


more invasive. Proceedings of National Academy of Science, 103, 5841–5845.

Thompson, J.N. (1999) The evolution of species interactions. Science, 284, 2116–2118.


90



592.

van Kleunen, M., Dawson, W., Schlaepfer, D., Jeschke, J.M. & Fischer, M. (2010) Are invaders

different? A conceptual framework of comparative approaches for assessing determinants

of invasiveness. Ecology Letters, 13, 947-958.

van Kleunen, M., Weber, E. & Fischer, M. (2010) A meta-analysis of trait differences between

invasive and non-invasive plant species. Ecology Letters, 13, 235–245.

Violle, C., Nemergut, D.R., Pu, Z. & Jiang, L. (2011) Phylogenetic limiting similarity and

competitive exclusion. Ecology Letters, 14, 782–787.

Walker, J.B., Sytsma, K.J., Treutlein, J. & Wink, M. (2004) Salvia (Lamiaceae) is not

monophyletic: implications for the systematics, radiation, and ecological specializations of

Salvia and tribe Mentheae. American Journal of Botany, 91, 1115–1125.



Wikström, N., Savolainen, V. & Chase, M.W. (2001) Evolution of the angiosperms: calibrating

the family tree. Proceedings of the Royal Society B: Biological Sciences, 268, 2211–2220.

Zuppinger-Dingley, D., Schmid, B., Petermann, J.S., Yadav, V., De Deyn, G.B. & Flynn, D.F.B.

(2014) Selection for niche differentiation in plant communities increases biodiversity

effects. Nature, 515, 108–111.

CHAPTER 6: GENERAL CONCLUSIONS

91

Chapter 6

General Conclusions

My doctoral research is motivated by understanding the role of historical contingencies as

mechanisms that maintain diverse species assemblages, as well as their potential to obscure

meaningful ecological phenomena if overlooked. I have explored three ways in which historical

contingencies manifest, and where they fall along a time continuum of effects that persist from

one (chs. 2 and 3) to tens (ch. 4) to tens of millions (ch. 5) of generations (Fig. 1.1). As I will

discuss, each chapter has addressed contemporary issues at the interface of ecology and

evolution, and serves to provide a broader demonstration of the spatial and temporal dimensions

of species’ niches.

Maternal effects of past environments

In chapters 2 and 3, I present the first step in exploring the ecological significance of maternal

effects by quantifying their overall prevalence among assemblages of species in controlled

environments. This work led to three main findings: (i) species differ in magnitude and direction

of maternal effects in response to the same sets of abiotic conditions, (ii) these maternal effects

can reinforce or counteract responses through seed number, depending on the species, and (iii)

maternal effects are exacerbated in competitive environments, indicating that existing estimates

of the prevalence of maternal effects are likely conservative. My research demonstrates that by

ignoring maternal effects, ecologists effectively miss over half of the ways in which species

respond to the environment; ecologists should work to develop a more synthetic understanding

of when maternal effects are likely to occur and the ways in which they influence ecological

dynamics.

There are several next steps for this work towards the goal of understanding how maternal

effects might contribute to ecological dynamics in real communities and under different maternal

effects scenarios. First, to test their importance in real communities, ecologists could either parse

out maternal effects from demographic data or test how different maternal environmental

histories affect the ecologies of offspring. As an example of the latter, I am involved in ongoing


92

work in the Gilbert lab that tests competitive differences between offspring of two Vulpia

species, reciprocally transplanted so that offspring originated from maternal plants that either

experienced the same or different soil moisture conditions as the other species. Second, the

maternal effects observed across species in my experiments suggest that they will have large

influences on parameters in population models, and alter coexistence and diversity under

different biotic and abiotic environments. This could be tested by estimating parameters that

modify demographic rates, such as finite rates of increase, competition coefficients, and seed

bank viability, based on biotic and abiotic conditions experienced in the previous generation.

Incorporating these parameter estimates into population models would improve our

understanding the relative importance of maternal effects in promoting or limiting diversity in

ecological communities. Finally, the intricacies of how different types of maternal effects (e.g.,

seed size, dormancy, seedling performance) are distributed among species warrants the use of

comparative methods that link maternal effects expression to species’ ecologies.

Dispersal histories in a metacommunity

The interplay between local and regional processes underlies contemporary perspectives on the

mechanisms that maintain diverse communities. Chapter 4 uses a novel method of manipulating

species pools across natural landscapes, and in doing so, offers definitive evidence of scale-

specific processes and their roles in structuring the distributions of whole communities of

species. The key findings are: (i) species’ distributions are strongly influenced by dispersal

histories that manifest at distinct spatial scales, (ii) the removal of dispersal limitation doubles

the strength of species-environment relationships, and (iii) that dispersal histories are constrained

the natural spatial structure of environmental conditions. Together, these results indicate that

ecologists should consider moving away from questions of whether species’ distributions are

more constrained by environments or dispersal histories, and instead explicitly test and model

their interdependence in future studies. Additionally, the scale-specific transitions that we

observed shed light on the potential shortcomings of the “local” vs. “regional” dichotomy in

ecology.

Given that our community surveys were performed a year after the seed pooling manipulations

were carried out, future surveys will ascertain their longer-term effects. Specifically, we will test


93

whether our seed pooling effects are maintained or swamped out through time with incoming

propagules from surrounding communities, and whether they affect the temporal dynamics of

local communities. Though preliminary, a follow-up survey in 2016 points towards the

persistence of increasing species richness with the spatial scale of seed pooling, even in face of

extreme natural events, including record-setting drought (2014-2015), wildfire (2015), and heavy

la Niña rains (2016). Formal analyses will be performed to confirm this result, as well as the

persistence of effects on other components of diversity reported in chapter 4.

Historical interactions and macroevolution

In chapter 5, I presented empirical mechanistic evidence that challenges a fundamental

assumption of community phylogenetics, and informs our understanding of how competitive

interactions evolve on macroevolutionary timescales. Contrary to classic ecological theory

(Darwin 1856; Elton 1946), and Webb et al.’s (2002) initial predictions, my work shows there is

no evidence that local coexistence is less likely to occur amongst closely related species. This

result was especially interesting given that the competitive components that underlie coexistence

did correlate with relatedness. Although it is not known how broadly my results might apply

across diverse communities, the fact that a key assumption of community phylogenetics is not

supported in the few definitive tests that are available (e.g., Narwani et al. 2013; Godoy et al.

2014; Germain et al. 2016) should call into question the inferential power of existing research in

this field (Vamosi et al. 2009). Furthermore, by additionally contrasting among competitors that

have historically occurred in sympatry or allopatry, I found macroevolutionary patterns of

stabilizing niche differences that are consistent with the presence of character displacement, or

“ghosts of competition past” (Connell 1980). These ghosts have remained elusive, and may have

masked evolutionary patterns of stabilizing differences in previous research due to the inclusion

of species of multiple provenances and thus mixed potentials for coevolutionary interactions

(Godoy et al. 2014).

My work tested competitive divergence among species, and intriguingly, showed rapid increases

in stabilizing differences over a relatively small evolutionary window (<25 mya since

divergence; Fig. 5.1a). This observation has motivated my future work as a postdoctoral fellow

to explore competitive divergences on much shorter timescales, among populations within


94

species. Recent work by Siepielski et al. (2016) showed that the strength of density dependence

among populations of damselflies negatively covaries with the strength of local adaptation.

Because density dependence underlies both stabilizing and fitness differences, disentangling

their relative influences would bring us closer to understanding the ecological mechanisms that

shape the evolutionary divergence of populations. The findings of my PhD work offer a clear

predictive framework for population divergence in competitive differences under different

scenarios of sympatry or allopatry.

More broadly, my work demonstrates the need for syntheses of ecology and evolutionary biology

that are conceptually robust from the perspective of both fields. For example, it is now apparent

that initial expectations for how evolutionary relatedness might impact ecological interactions

were overly simplistic, and can be better resolved via the incorporation of biogeographic history

(Mittelbach & Schemske 2015; Johnson & Stinchcombe 2007). It is likely that similar

improvements to theory could be made elsewhere, for example, in studies that assume Brownian

motion models of trait evolution (Freckleton & Harvey 2006) or ecological equivalence of

species immediately following speciation (Funk et al. 2006). From an evolutionary standpoint,

ecological theory has yet to be incorporated beyond some of the simplest of principles, such as

divergence along single resource or environmental axes (e.g., Funk et al. 2006; Anacker &

Strauss 2014). Because ecological mechanisms are thought to play a primary role during

speciation (Schluter 2001), a more explicit integration of ecological theory to evolutionary

hypotheses would better link the forces that generate and maintain diverse species assemblages

on evolutionary and ecological timescales. Calls for a more integrative synthesis between

ecology and evolution have described it as “one of the most important frontiers for the

conceptual unification of the biological sciences” (Matthews et al. 2011).

Concluding remarks

It is clear that species’ niches are far more complex and difficult to define than initially thought;

where species are now does not necessarily reflect where they could be (ch. 4), and the niche

itself can respond plastically to current and historical conditions (ch. 2 and 3). All of these

complexities add to the number of ways in which species might partition their environments, not


95

just through differences in local resource use, but also in when or where resource use takes place,

and how resource use varies in response to extrinsic factors. The net effect of all of these

possible axes of differentiation is the potential for large numbers of species to coexist across

landscapes. Most biological communities are highly speciose, an observation that early

ecologists struggled to explain under a more static and local perspective of ecological niches. G.

E. Hutchinson’s (1957) own answer to his question of why there are so many species in nature

was that “[diversity is limited] by physical factors, by space, by the fineness of possible

subdivision of niches”. The work presented in this thesis provides demonstrations of each of

these ideas, but also extends them to show how historical contingencies can cause time lags in

each that sometimes promote diversity and sometimes constrain it. Further exploration into how

the past and present interact to structure biodiversity will provide a basis for predicting future

responses to climate change (Alexander et al. 2015), introduced species (Capinha et al. 2015),

habitat fragmentation (Haddad et al. 2015), and other challenges that characterize the

Anthropocene.

Literature cited

Alexander, J.M., Diez, J.M. & Levine, J.M. (2015) Novel competitors shape species’ responses

to climate change. Nature, 525, 515-518.

Anacker, B.L. & Strauss, S.Y. (2014) The geography and ecology of plant speciation: range

overlap and niche divergence in sister species. Proceedings of the Royal Society B:

Biological Sciences, 281, 20132980.

Capinha, C., Essl, F., Seebens, H., Moser, D. & Pereira, H.M. (2015) The dispersal of alien

species redefines biogeography in the Anthropocene. Science, 348, 1248-1251.

Connell, J.H. (1980) Diversity and the coevolution of competitors, or the ghost of competition

past. Oikos, 35, 131-138.




Ecology, 15, 54–68.

Freckleton, R.P. & Harvey, P.H. (2006) Detecting non-Brownian trait evolution in adaptive

radiations. PLOS Biology, 4, e373.


96

Funk, D.J. (2006) Ecological divergence exhibits consistently positive associations with

reproductive isolation across disparate taxa. Proceedings of the National Academy of

Sciences, 103, 3209-3213.

Germain, R.M., Weir, J.T. & B. Gilbert. (2016) Species coexistence: macroevolutionary

relationships and the contingency of historical interactions. Proceedings of the Royal

Society B: Biological Sciences, 283, 20160047.



Haddad, N.M., Brudvig, L.A., Clobert, J., Davies, K.F., Gonzalez, A., Holt, R.C., et al. (2015)

Habitat fragmentation and its lasting impact on Earth’s ecosystems. Science Advances, 1,

e1500052.

Hutchinson, G.E. (1957) Concluding remarks. Cold Spring Harbor Symposium of Quantitative

Biology, 22, 415-427.

Johnson, M.T. & Stinchcombe, J.R. (2007) An emerging synthesis between community ecology

and evolutionary biology. Trends in Ecolgy and Evolution, 22, 250–257.

Matthews, B., Narwani, A., Hausch, S., Nonaka, E., Peterm H., Yamamichi, M., et al. (2011)

Toward an integration of evolutionary biology and ecosystem science. Ecology Letters,

14, 690-701.

Mittelbach, G.G. & Schemske, D.W. (2015) Ecological and evolutionary perspectives on

community assembly. Trends in Ecology and Evolution, 30, 241-247.

Narwani, A., Alexandrou, M.A., Oakley, T. H., Carroll, I.T. & Cardinale, B.J. (2013)



Schluter, D. (2001) Ecology and the origin of species. Trends in Ecology and Evolution, 16, 372-

380.

Siepielski, A. M., Nemirov, A., Cattivera, M. & Nickerson, A. (2016) Experimental evidence for

an eco-evolutionary coupling between local adaptation and intraspecific competition.




592.



97

Appendix A: Supplementary Information to Chapter 2

Supplementary tables and figures

Table S2.1. Taxonomic, sequence accession, and seed supplier information for the 29 species included in this study.

ID codes Species names Order Family Accession # Seed supplier Collection details*

CMELI Centaurea melitensis Asterales Asteraceae HQ540425.1 Herbiseed NP in 2009

CCAPI Crepis capillaris Asterales Asteraceae AJ633353 Herbiseed NP in 2008

LCALI Lasthenia californica Asterales Asteraceae AF467195.1 S & S seed AF <20 years

LGLAB Lasthenia glabrata Asterales Asteraceae AF391593.1 S & S seed AF <20 years

MELEG Madia elegans Asterales Asteraceae AF413612.1 S & S seed AF <10 years

ULIND Uropappus lindleyi Asterales Asteraceae AF386495.1 Theodore Payne NP in 2011

SGALL Silene gallica Caryophyllales Caryophyllaceae U30959/U30985 Herbiseed AF for 1 year

CBERL Chenopodium berlandieri Caryophyllales Chenopodiaceae HE577429.1 Wild Garden Seed NA†

MNUTT Monolepis nuttalliana Caryophyllales Chenopodiaceae HE577375.1 Independent donor NP in 2011

CPERF Claytonia perfoliata Caryophyllales Montiaceae AY764040.1 S & S seed NP in 2011

POLER Portulaca oleracea Caryophyllales Portulacaceae JF508578.1 Herbiseed NP in 2010

LPURS Lotus purshianus Fabales Fabaceae AF467067.1 S & S seed AF <10 years

LBICO Lupinus bicolor Fabales Fabaceae DQ524209.1 S& S seed AF <20 years

TINCA Trifolium incarnatum Fabales Fabaceae AF053160.1 S & S seed NP in 2011

CHETE Collinsia heterophylla Lamiales Lamiaceae AF385337.1 S & S seed NP in 2011

SCOLU Salvia columbariae Lamiales Lamiaceae NA S & S seed NP in 2011

SVIRI Salvia viridis Lamiales Lamiaceae NA B & T World seed NA†

BCUCU Bromus cucamonga Poales Poaceae AY367948.1 S & S seed AF <10 years

HDEPR Hordeum depressum Poales Poaceae AJ607894.1 S & S seed AF <10 years

HVULG Hordeum vulgare Poales Poaceae FJ593180.1 S & S seed NP in 2011

VMICR Vulpia microstachys Poales Poaceae EF584981.1 S & S seed AF <10 years

VMYUR Vulpia myuros Poales Poaceae AY118092.1 S & S seed NP in 2011

VOCTO Vulpia octoflora Poales Poaceae EF584982.1 S & S seed AF <10 years

ECAES Eschscholzia caespitosa Ranunculales Papaveraceae JF892592.1 S & S seed AF <20 years

98

EMEXI Eschscholzia mexicana Ranunculales Papaveraceae DQ912884.1 S & S seed AF <10 years

PDUBI Papaver dubium Ranunculales Papaveraceae DQ250322.1 Herbiseed AF for 1 year

MARVE Myosotis arvensis Boraginales Boraginaceae AY092908.1 Herbiseed NP in 2011

NMENZ Nemophila menziesii Boraginales Boraginaceae AF091183.1 S & S seed NP in 2011

PCAMP Phacelia campanularia Boraginales Boraginaceae AF091188.1 S & S seed AF <20 years

Notes: *The collection details summarize the origins of the seed used in this experiment, based on information provided by the seed

suppliers. NP = collected directly from natural populations; AF = grown in agricultural fields for x years since source seed was

collected from natural populations. In all but two cases†, populations cultivated in agricultural fields are regularly supplemented with

seed from natural populations to maintain genetic diversity. †C. berlandieri and S. viridis populations were raised in agricultural fields

for 8-10 generations and the original source is unknown.

99

Figure S2.1. The timing of shifts in (A) watering intervals in number of days by treatment

[wet/dry], (B) day/night temperature regime (°C), and key methodological events with respect to

days since the maternal generation experiment was initiated. Flowering dates varied by species,

such that the seed collection and biomass harvest was ongoing in the timespan bounded by the

grey arrow.

100

Figure S2.2. Effect size of seedling height responses to our watering treatments across a

phylogeny of 29 species; see Table S2.1 for species codes. Effect sizes were calculated using

Cohen’s d (i.e., (µdry - µwet)/pooled σ; Hartung et al. 2011). Positive effect sizes indicate that trait

values were higher in the dry compared to wet environment, and vice versa. The numbers on the

phylogeny correspond to taxonomic orders: (1) Poales, (2) Ranunculales, (3) Fabales, (4)

Caryophyllales, (5) Boraginales, (6) Lamiales, and (7) Asterales. Dashed branches represent

unknown relationships due to missing sequence data [SCOLU and SVIRI], and ‘×’ indicates

unavailable trait data rather than a true zero. * P < 0.05, º P < 0.10.

101

Figure S2.3. Effect of seed size on seedling height after three weeks of growth. Each data point

represents the average height of the two seedlings per pot from the offspring generation

experiment. The best-fit line indicates a significant positive relationship (R2 = 0.01, P = 0.047).

102

Appendix B: Supplementary Information to Chapter 3

Species Selection

We selected 25 annual Mediterranean plant species that span six orders in the angiosperms

(Boraginales, Caryophyllales, Fabales, Lamiales, Poales and Ranunculales). These species are

native to the Mediterranean-climate regions of California (n = 19) or Spain (n = 6), overlap

geographically with other species from the same native range (CalFlora [http://www.calflora.org]

and Euro+Med [http://www.emplantbase.org]), and all associate with grassland meadow habitat.

For these reasons, species from the same region have potential to compete in the wild. We

obtained the initial seeds for our experiment from commercial suppliers and independent donors,

and tested them for percent germinability prior to the experiment to correct seeding densities

based on germination rates.

Species pairs for the competitive trials were chosen to include comparisons at four taxonomic

levels (within the same genus, tribe, family, or order) and to include comparisons of both

California-California pairs (n = 9) and California-Spain pairs (n = 6). Although we initially

sought out a balanced design with 10 species pairs in each of the biogeographic contrasts, we

were restricted to the 15 pairs that were not so limited by competition that they produced seeds

across a range of relative conspecific frequencies. Because five California species were

competed against two other species of the same (native to California) and different (native to

Spain) origin, a total of 25 species were used to create our 15 species pairs (Tables S3.1 and

S3.2).

Greenhouse Experiment

Seeds for the greenhouse experiment were sown into 12.7 cm diameter, 23 cm deep treepots

filled with a 3:2 mixture of sand and screened topsoil, to mimic the sandy loam soils

characteristic of annual grasslands. The greenhouse was maintained at day/night temperatures of

14/7°C, which was set to gradually increase to 29/17°C on average by the end of the experiment.

High intensity discharge (HID) lighting was provided to create a 12 h photoperiod. Each pot was



103

watered daily to saturation using a drip irrigation system during a three-week establishment

period, after which pots were randomly assigned to either a wet or dry soil moisture treatment.

Using a drip irrigation system, pots in the wet treatment received 175 mL water twice as often as

those in the dry treatment. Pots were watered at one and two day intervals initially, which was

extended to seven and 14 days (for wet and dry treatments respectively) as the growing season

progressed [see Fig. S3.1 in Germain and Gilbert (2014) for a schematic of the watering and

temperature conditions throughout the experiment]. These two soil moisture regimes were

selected to simulate realistic inter-annual differences in precipitation, or equally, the 30-year

average differences between wet (mesic) [662 mm] and dry [312 mm] sites across the species

ranges (http://www.climate-charts.com). We confirmed that the wet (11.1 ± 0.56%; mean ± SE

percent soil moisture content) and dry (5.8 ± 0.42%) treatments were effective using a

volumetric water probe (HydroSense™, Campbell Scientific Australia). Prior to flowering (~60

days after planting), we added 350 mL of 1500 ppm 20-20-20 NPK fertilizer (Plant Products,

Inc., Brampton, Ontario) to each pot. Pollination was provided by commercial colonies of the

generalist pollinator Bombus impatiens, which were active throughout the flowering period

(Biobest Canada, Leamington, Ontario).

At the end of the experiment, subsampled seeds were cleaned of debris, weighed, and

enumerated to estimate the size and number of seeds produced. Seed size was estimated as the

mass of seeds in the subsample divided by number of seeds in the subsample, and seed number

was calculated as the average number of seeds produced per plant in each pot.

Statistical Analyses

In all of our mixed effects models, we used backwards selection of the fixed factors, starting with

the most complex model and dropping higher-order interactions that did not improve model fit

(i.e., that did not minimize the AIC scores). All random effects were kept in the models, as these

were necessary to account for non-independence between some species and samples due to

constraints of the experimental design. Backwards selection of the fixed factors was performed

using the ‘step’ function in the ‘lmerTest’ R package. All analyses were conducted in R v. 3.0.2

(R Core Team 2014).

104


Table S3.1. ID codes and taxonomic information for the 25 species included in this study.

ID codes Species names Order Family

CCAPI Crepis capillaris Asterales Asteraceae

LCALI Lasthenia californica Asterales Asteraceae

LGLAB Lasthenia glabrata Asterales Asteraceae

MELEG Madia elegans Asterales Asteraceae

ULIND Uropappus lindleyi Asterales Asteraceae

SGALL Silene gallica Caryophyllales Caryophyllaceae

CBERL Chenopodium berlandieri Caryophyllales Chenopodiaceae

MNUTT Monolepis nuttalliana Caryophyllales Chenopodiaceae

CPERF Claytonia perfoliata Caryophyllales Montiaceae

LPURS Lotus purshianus Fabales Fabaceae

LBICO Lupinus bicolor Fabales Fabaceae

CHETE Collinsia heterophylla Lamiales Lamiaceae

SCOLU Salvia columbariae Lamiales Lamiaceae

SVIRI Salvia viridis Lamiales Lamiaceae

BCUCU Bromus cucamonga Poales Poaceae

HDEPR Hordeum depressum Poales Poaceae

HVULG Hordeum vulgare Poales Poaceae

VMICR Vulpia microstachys Poales Poaceae

VMYUR Vulpia myuros Poales Poaceae

VOCTO Vulpia octoflora Poales Poaceae

ECAES Eschscholzia caespitosa Ranunculales Papaveraceae

EMEXI Eschscholzia mexicana Ranunculales Papaveraceae

MARVE Myosotis arvensis Boraginales Boraginaceae

NMENZ Nemophila menziesii Boraginales Boraginaceae

PCAMP Phacelia campanularia Boraginales Boraginaceae

105

Table S3.2. Summary of species’ seed size and seed number responses to conspecific frequency, soil moisture environment, and their

interaction.

Seed size Seed number

In competition Not in

competition

In competition Not in

competition

species origin Freq Env Freq × Env Env Freq Env Freq × Env Env

Ulind California 0 0 0 0 - 0 0 0

Ccapi Spain - - 0 0 - 0 0 -

Ulind California 0 0 0 - 0 ×

Mele California 0 0 0 0 - - 0 0

Lglab California 0 0 0 0 0 - 0 0

Lcali California 0 + 0 0 0 + 0 -

Mnutt California + 0 0 0 + 0 0 +

Sgall Spain - 0 0 0 - - 0 0

Cberl California 0 - 0 0 - 0 0 -

Cperf California 0 0 0 0 0 0 0 0

Lbico California 0 0 0 0 + - 0 0

Lpurs California - - 0 0 - 0 0 0

Scolu California 0 0 0 + 0 0 0 0

Chete California + + × + - 0 0 0

Scolu California - + 0 0 - 0

Sviri Spain 0 + × 0 - + × 0

Bcucu California 0 0 0 0 - 0 × 0

Hdepr California + 0 0 0 0 0 0 0

Bcucu California 0 0 0 0 0 0

Hvulg Spain 0 0 0 + - - 0 0

Vmicr California 0 - 0 - 0 0 0 0

Vmyur Spain 0 - 0 - - - 0 0

Vmicr California 0 - 0 0 0 0

Vocto California 0 0 0 + 0 0 0 0

Ecaes California + 0 0 0 + 0 0 0

Emexi California - - 0 0 0 0 × -

106

Pcamp California + 0 0 0 0 0 0 0

Nmen California - 0 0 0 0 - 0 0

Pcamp California 0 0 × - 0 0

Marv Spain 0 0 0 0 - + 0 -

Total significant

responses

11 11 3 6 17 11 4 6

Notes: Values in red and yellow cells are significant (red cells; P < 0.05) and marginally significant (yellow cells; P < 0.10) responses.

Because some focal species were competed twice against different species (5 cases), species pairs separated by a dashed line have the

same focal species; in these cases, we shade cells in grey to avoid redundancy in environmental responses when not in competition.

Positive symbols indicate that seed size or seed number increased in response to high conspecific frequency or dry soil moisture

conditions; negative symbols are the opposite.

107

Table S3.3. Summary of a reduced backward-selected linear mixed effects model testing for

effects of species identity, soil moisture, and conspecific frequency on the seed size in 25 annual

plant species.

F P

Species 2.26 <0.001 Soil moisture 0.94 0.334

Frequency 0.48 0.490

Species*soil moisture 3.19 <0.001 Species*frequency 2.84 <0.001


effects of species identity, soil moisture, and conspecific frequency on seed number in 25 annual

plant species.

F P

Species 5.05 <0.001 Soil moisture 0.51 0.477

Frequency 53.17 <0.001

Species*soil moisture 1.62 0.022

Species*frequency 6.63 <0.001

Soil moisture*frequency 4.49 0.034


effects of phylogenetic distance, biogeographic history (California-California vs. California-

Spain), soil moisture, and biogeographic origin (California vs. Spain) nested within

biogeographic history on the slope of the relationship between seed size and conspecific

frequency in 25 annual plant species.

F P

Phylogenetic distance (PD) 2.21 0.147

Biogeographic history (BH) 0.50 0.485

Soil moisture (SM) 0.73 0.398

PD x SM 1.74 0.196

PD x BH 4.55 0.041

BH x SM 0.90 0.350

PD x BH x SM 2.29 0.141

PD x BH x biogeographic origin x SM 4.55 0.019

108

Figure S3.1. Maximum clade credibility tree of the 25 annual plant species with median node

heights. The tree was generated in BEAST using ITS1/5.8S/ITS2 and rbcL sequence data, and

calibrated to real-time based on fossil records. The scale bar is in millions of years ago, and

nodal support values are the posterior probabilities; * indicates nodes that were fixed based on a

priori information. The original tree of 30 species (Germain et al. 2016) was pruned for analysis

to the 25 used in the current experiment; pruned species are shown in grey. See Table S3.1 for

species codes.

109

Appendix C: Supplementary Information to Chapter 4


Table S4.1. Information for 73 annual species observed in the unmanipulated control plots (C2

in Fig. S4.1).

Scientific name Family Status Site occupancy

Achyrachaena mollis Asteraceae Native 3

Acmispon americanus Fabaceae Native 1

Acmispon brachycarpus Fabaceae Native 5

Acmispon wrangelianus Fabaceae Native 10

Agoseris heterophylla Asteraceae Native 9

Amsinckia menziesii Boraginaceae Native 2

Anagallis arvensis Primulaceae Naturalized 3

Ancistrocarphus filagineus Asteraceae Native 14

Astragalus gambelianus Fabaceae Native 2

Athysanus pusillus Brassicaceae Native 8

Avena spp.* Poaceae Invasive 7

Bromus diandrus Poaceae Invasive 2

Bromus hordeaceus Poaceae Invasive 14

Bromus madritensis Poaceae Naturalized 11

Calandrinia ciliata Portulacaceae Native 5

Calycadenia pauciflora Asteraceae Endemic 8

Camissonia graciliflora Onagraceae Native 2

Cardamine oligosperma Brassicaceae Native 1

Castilleja attenuata Orobanchaceae Native 0

Castilleja rubicunda Orobanchaceae Endemic 1

Centaurea solstitialis Asteraceae Invasive 2

Clarkia gracilis Onagraceae Native 6

Clarkia purpurea Onagraceae Native 1

Collinsia sparsiflora Plantaginaceae Native 9

Croton setigerus Euphorbiaceae Native 16

Cuscuta californica Convolvulaceae Native 6

Daucus pusillu Apiaceae Native 1

Epilobium brachycarpum Onagraceae Native 7

Eriogonum covilleanum Polygonaceae Native 3

Eriogonum vimineum Polygonaceae Native 0

Erodium cicutarium Geraniaceae Invasive 2

Euphorbia crenulata Euphorbiaceae Native 6

Galium aparine Gentianales Native 1

Gilia tricolor Polemonaceae Endemic 2

Githopsis specularioides Campanulaceae Endemic 7

110

Hemizonia congesta Asteraceae Native 24

Hesperolinon spp.* Linaceae Endemic 9

Holocarpha virgata Asteraceae Endemic 3

Hypochaeris glabra Asteraceae Invasive 2

Juncus bufonius Juncaceae Native 1

Lactuca spp.* Asteraceae Naturalized 10

Lagophylla minor Asteraceae Endemic 3

Lasthenia californica Asteraceae Native 16

Lepidium nitidum Brassicaceae Native 4

Lessingia ramulosa Asteraceae Endemic 7

Linanthus bicolor Polemoniaceae Native 3

Linanthus dichotomus Polemoniaceae Native 1

Lolium multiflorum Poaceae Naturalized 11

Lupinus bicolor Fabaceae Native 2

Lupinus succulentus Fabaceae Native 1

Micropus californicus Asteraceae Native 10

Microseris douglasii Asteraceae Native 12

Mimulus douglasii Phrymaceae Native 5

Mimulus guttatus Phrymaceae Native 1

Minuartia douglasii Caryophyllaceae Native 1

Navarretia jepsonii Polemoniaceae Rare/Endemic 4

Navarretia pubescens Polemoniaceae Native 1

Nemophila heterophylla Boraginaceae Native 2

Nemophila pedunculata Boraginaceae Native 1

Phlox gracilis Polemoniaceae Native 3

Plantago erecta Plantaginaceae Native 24

Riggiopappus leptocladus Asteraceae Native 6

Sidalcea diploscypha Malvaceae Endemic 5

Stellaria nitens Caryophyllaceae Native 3

Taeniatherum caput-medusae Poaceae Invasive 8

Torilis arvensis Apiaceae Invasive 1

Trifolium albopurpureum Fabaceae Native 5

Trifolium bifidum Fabaceae Native 5

Trifolium depauperatum Fabaceae Native 1

Trifolium fucatum Fabaceae Native 2

Trifolium gracilentum Fabaceae Native 4

Velezia rigida Caryphyllaceae Naturalized 1

Vulpia microstachys Poaceae Native 26

Notes: Species’ statuses were cross-checked with CalFlora Plant Database (www.calflora.org);

species that occurred in 0 sites were observed in one of the other treatment or control plots.

*Avena fatua/A. barbata and Lactuca saligna/L. serriola were indistinguishable, and

Hesperolinon could not be identified to species.


111

Table S4.2. Fine-scale estimates of soil moisture, light availability, and slope inclination among

plots at the same site, among groups of sites occurring within 100 m, and among sites occurring

within the same half of the reserve.

Nested plot design

spatial

extent of

observation

soil moisture

(%)

light availability

(PAR)

slope inclination

(º)

F-

value P

F-

value P

F-

value P

halves w/i reserve 5 km 161.2 <0.001 17.2 <0.001 18.5 <0.001

groups w/i halves 100 m 3.4 0.068 64.9 <0.001 80.7 <0.001

sites w/i groups 5 m 14.7 <0.001 0.5 0.502 65.3 <0.001

plots w/i sites 1 m 0.44 0.505 0.8 0.382 0.3 0.566

residual df 711 475 474

Notes: Analyses are nested analysis of variance using ‘aov’ R function; data are from two (light

availability, slope inclination) or three (soil moisture) subsamples per plot for all eight plots per

site. The statistical annotation of the independent factors is ~halves/groups/sites/plots (Fig. S4.1).

Table S4.3. Summary of a reduced backward-selected linear mixed effects model testing the

effects of composite environmental conditions (PCA1 and PCA2), seed pooling treatment, and

species on the occurrence patterns of 73 annual plant species.

Χ2 P

PCA1 2.31 0.129

PCA2 7.75 <0.005 treatment 338.59 <0.001 species 1295.72 <0.001 PCA1:PCA2 0.266 0.606

PCA1:treatment 3.07 0.546

PCA2:treatment 6.20 0.184

PCA1:species 204.73 <0.001 PCA2:species 211.16 <0.001 PCA1:PCA1:treatment 15.08 0.005

112

Figure S4.1. Map of field site with the six groups of five sites (a) and the seed pooling treatment

design for a block of plots at a single site (b). Dashed grey arrows indicate seed material source

plots, and solid black arrows indicate recipient plots. Plots C1 to C3 are control plots: material

removed without replacement (C1), no manipulation (C2), and material collected and transferred

to the same plot (C3). Colored plots are the treatment plots, receiving material mixed from plots

at increasing spatial scales: a single plot from the same site (1 m), a mix of plots within a site (5

m, mix i), a mix of plots from the same cluster of five sites (100 m, mix j), a mix of plots within

the same half of the reserve (east vs. west side; 5 km, mix k), and a mix of plots across the entire

reserve (10 km, mix l). Source material from the recipient plot of the 1 m treatment was

discarded.

113

Figure S4.2. Community composition biplot of PCoA using Jaccard’s dissimilarity. Points in

grey are control treatments (Fig. S4.1): vacuumed without replacement (C1; triangles),

unmanipulated plots (C2; circles), and vacuumed with replacement but not pooled (C3; squares).

Species composition becomes more distinct and less variable with increasing spatial scale.

Figure S4.3. Effect of spatial scale of seed pooling on the percentage of species that are

significantly associated with the environment. The grey region is the percentage of significant

results expected due to type I error with multiple comparisons.

114

Figure S4.4. Site occupancy (based on number of sites each species occurs in) increases with

spatial scale (also note upwards shift in y-intercept) and becomes more even (decrease in

regional rarity).

Figure S4.5. Photos of McLaughlin reserve showing (a) a serpentine meadow in the foreground

of non-serpentine oak hillside, and (b) an abrupt transition zone between two abutting serpentine

outcrops.

115

Appendix D: Supplementary Information to Chapter 5

Phylogenetic tree construction

We constructed a phylogenetic tree for the 30 species in our experiment using Bayesian methods

on ITS1/5.8S/ITS2 and rbcL nucleotide sequence regions. These regions were chosen because

they are relatively fast and slow evolving, respectively, and are thus good complements for

estimating divergence times at different depths in the phylogenetic tree. Separately for each

region, sequences were retrieved from GenBank, aligned using the MUSCLE algorithm in

MEGA (version 4.0), and combined into a single sequence matrix for analysis. We then loaded

the aligned sequence data from both sequence regions into BEAST (version 2.3.1) and generated

a Bayesian tree using a GTR substitution model selected by ModelTest (version 3.7) and an

uncorrelated lognormal relaxed molecular clock for branch length estimation; the site and clock

models of the two sequence regions were unlinked. We set lognormal priors on four internal

nodes (Poales = 1.8 [mean], 0.5 [standard deviation], 68.1 [minimum age in millions of years];

Fabales = 1.5, 0.5, 59.9; Caryophyllales = 1.5, 0.5, 83.5; Lamiales = 1.5, 0.5, 44.3) based on

published node calibrations for these groups, to calibrate the tree to real-time (in millions of

years) with the minimum age constraints based on the fossil record (Wikström et al. 2001). We

constrained the tree topology at nodes above the family-level based on well-resolved

relationships in the angiosperms (Bell et al. 2010; Stevens 2012) and thus did not require an

outgroup to root the tree because BEAST automatically performs molecular clock rooting above

the constrained nodes. We ran the MCMC chain for 100,000,000 generations, with a 10,000,000

burn-in and data logging every 9,000 generations. This analysis generated 10,000 posterior trees,

which were summarized into a single ultrametric maximum clade credibility tree with median

node heights. The resulting tree (Fig. S5.1) is fully resolved and consistent with Angiosperm

Phylogeny Group III classifications and estimated divergence times of major groups (Stevens

2012). We used the ‘cophenetic.phylo’ function in the R package ‘ape’ to calculate all pairwise

phylogenetic distances based on the branch length information of the 30 study species.

116

Details of lambda estimation

As described in the Materials and methods, we grew plants at low densities with seven replicate

pots per species × soil moisture environment. These replicates were used to calculate the

distribution of log-transformed finite rates of increase (λ) for each combination of species × soil

moisture environment by fitting separate linear models. The linear models tested the effects of

the number of individuals in each low density pot on finite rate of increase, to identify species for

which competition might affect our estimates of λ. These linear models were possible because,

although we thinned pots to eight individuals, low germination rates or post-thinning

germination/mortality caused some variation among pots in the numbers of individuals that were

present (mean = 6.3, sd = 2.2 individuals). For species with a significant to marginally-

significant slope (P < 0.1), we confirmed that significance was not driven by outlying

observations, and then used the intercept (± standard deviation [sd]) of the population growth

rate as λ. For species with a non-significant slope, we used the mean (± sd) population growth

rate across all replicates as λ. In both cases, these estimates were used as informative priors in the

Bayesian analysis described in the Materials and methods. For those species that have been

studied elsewhere, our estimated λ values were, on average, within one standard deviation (mean

= 0.56 standard deviations, n = 3) of existing estimates in the literature (Levine &

HillRisLambers 2009).

Supplementary equations

Equations for stabilizing differences (Eq. S5.1) and fitness differences (Eq. S5.2, S5.3) as

described in Materials and methods:

1 − 𝜌 = 1 − √𝛼𝑖𝑗

𝛼𝑗𝑗 ∙

𝛼𝑗𝑖

𝛼𝑖𝑖 (S5.1)

𝜅𝑗 𝜅𝑖⁄ =𝜆𝑗−1

𝜆𝑖−1 ∙ √

𝛼𝑖𝑗

𝛼𝑗𝑗 ∙

𝛼𝑖𝑖

𝛼𝑗𝑖 (S5.2)

𝛫 = max (𝜅𝑗 𝜅𝑖 , 𝜅𝑖 𝜅𝑗)⁄ ⁄ (S5.3)

117


Table S5.1. Taxonomic and sequence accession (ITS1/5.8S/ITS2 and rbcL regions) information for the 30 study species, organized to

show each focal species and its sympatric and allopatric competitor

Species names Order Family Origin ITS1/5.8S/ITS2 rbcL

Lasthenia glabrata Asterales Asteraceae California AF391593.1 AIW51855.1*

Lasthenia californica Asterales Asteraceae California AF467195.1 NA

Centaurea melitensis1 Asterales Asteraceae Spain HQ540425.1 EU384954.1

Uropappus lindleyi Asterales Asteraceae California AF386495.1 NA

Madia elegans Asterales Asteraceae California AF413612.1 AY215141.1*

Crepis capillaris1 Asterales Asteraceae Spain AJ633353 KM360738.1

Phacelia campanularia Boraginales Boraginaceae California AF091188.1

KF158107.1

Nemophila menziesii Boraginales Boraginaceae California AF091183.1

KF158108.1

Myosotis arvensis Boraginales Boraginaceae Spain AY092908.1

HM850186.1

Chenopodium berlandieri Caryophyllales Amaranthaceae California HE577429.1 JF941268.1*

)

Claytonia perfoliata2 Caryophyllales Montiaceae California AY764040.1 AF132093.1

Portulaca oleracea Caryophyllales Portulacaceae Spain JF508578.1 HQ621340.1

Monolepis nuttalliana Caryophyllales Amaranthaceae California HE577375.1 AY270108.1

Atriplex patula2 Caryophyllales Amaranthaceae California DQ499332.1 HM849801.1

Silene gallica Caryophyllales Caryophyllaceae Spain U30959/U30985 HM850354.1

Lupinus bicolor Fabales Fabaceae California DQ524209.1 Z70056.1*

Lotus purshianus Fabales Fabaceae California AF467067.1 HM850139.1*

Trifolium incarnatum Fabales Fabaceae Spain AF053160.1 HM850415.1

Salvia columbariae Lamiales Lamiaceae California DQ667219.1* AY570408.1

118

Collinsia heterophylla Lamiales Lamiaceae California AF385337.1 AF026825.1*

Salvia viridis Lamiales Lamiaceae Spain DQ667222.1* AY570450.1

Bromus carinatus Poales Poaceae California AY367948.1 KM360707.1

Hordeum depressum Poales Poaceae California AJ607894.1 NA

Hordeum vulgare Poales Poaceae Spain FJ593180.1 AY137456.1

Vulpia microstachys Poales Poaceae California EF584981.1 NA

Vulpia octoflora Poales Poaceae California EF584982.1 KJ773986

Vulpia myuros Poales Poaceae Spain AY118092.1 KF713076.1

Eschscholzia caespitosa Ranunculales Papaveraceae California JF892592.1 NA

Eschscholzia californica Ranunculales Papaveraceae California DQ912884.1

HM849984.1

Papaver dubium Ranunculales Papaveraceae Spain DQ250322.1

HM850229.1

Notes: Species contrasts are delineated by row shading, with the focal species in bold type. Superscripts represent contrasts with

partial branch length overlap, one of which (M. nuttalliana/A. patula2) was excluded from analysis because the latter did not flower.

*Sequences were unavailable, and were replaced with those of known sister species (S. columbariae = S. clevelandii, S. viridis = S.

sclarea (Walker et al. 2004)) or congeners as appropriate. ‘NA’ means that sequences were unavailable for that species.

119

Table S5.2. Comparing fits of alternative annual plant models

Model AICc scores

1 𝑁𝑖 𝑡+1 = 𝑁𝑖 𝑡 ∙ 𝜆𝑖

1 + 𝛼𝑖𝑖𝑁𝑖 𝑡 + 𝛼𝑖𝑗𝑁𝑗 𝑡 2236

2 𝑁𝑖 𝑡+1 = 𝑁𝑖 𝑡 ∙ 𝜆𝑖𝑒−𝛼𝑖𝑖𝑁𝑖 𝑡−𝛼𝑖𝑗𝑁𝑗 𝑡 2468


1 + 𝑁𝑖 𝑡𝛼𝑖𝑖 + 𝑁𝑗 𝑡

𝛼𝑖𝑗 2556


1 + (𝛼𝑖𝑖𝑁𝑖 𝑡 + 𝛼𝑖𝑗𝑁𝑗 𝑡)𝑏𝑖 2625

5 𝑁𝑖 𝑡+1 = 𝑁𝑖 𝑡 ∙ 𝜆𝑖𝑒−𝛼𝑖𝑖ln (𝑁𝑖 𝑡+1)−𝛼𝑖𝑗ln (𝑁𝑗 𝑡+1) 2748


(1 + 𝛼𝑖𝑖𝑁𝑖 𝑡 + 𝛼𝑖𝑗𝑁𝑗 𝑡)𝑏𝑖 2892

Notes: All models were simultaneously fit to each focal species i vs. sympatric species j and

allopatric species k. The AICc scores were summed across each three-species contrast × soil

moisture environment for model comparison. The parameters are described in the Materials and

methods, except for bi which allows individuals to vary in competitive impacts as their density

increases. Model formulations are taken from Levine and HilleRisLambers (2009).

120

Table S5.3. Parameter estimates used to calculate stabilizing and fitness differences (Eqs. S5.1 and S5.2)

Species Soil

env

Parameter estimates

i j k λi λj λk αii αjj αkk αij αji αik αki

B. carinatus H. depressum H. vulgare wet 79.0 155.1 22.8 0.175 0.138 0.227 0.001 0.795 0.623 0.006

C. berlandieri C. perfoliata P. oleracea wet 69.4 8.6 509.2 0.223 0.007 279.72 <0.001 0.065 <0.001 219.90

E. caespitosa E. californica P. dubium wet 67.0 53.7 947.4 0.004 0.012 0.705 3.573 2.388 15.10 0.008

L. bicolor L. purshianus T. incarnatum wet 4.3 58.4 1136.3 0.563 12.01 20.18 0.796 0.096 7.845 0.945

L. glabrata L. californica C. melitensis wet 389.8 839.0 1.8 0.054 0.098 0.180 0.003 0.219 0.001 0.728

M. nuttalliana A. patula S. gallica wet 587.5 0.1 989.4 0.263 0.001 0.119 0.003 <0.001 8.227 <0.001

P. campanularia N. menziesii M. arvensis wet 63.6 9.6 352.4 0.142 0.140 0.107 0.006 0.004 0.004 <0.001

S. columbariae C. heterophylla S. viridis wet 318.9 727.9 75.6 0.231 2.358 0.126 0.403 0.001 0.128 0.070

U. lindleyi M. elegans C. capillaris wet 35.9 75.5 323.6 0.384 1.271 1.342 0.069 0.001 0.003 0.002

V. microstachys V. octoflora V. myuros wet 235.8 923.8 314.8 0.099 0.456 0.200 0.083 0.302 0.227 0.002

B. carinatus H. depressum H. vulgare dry 563.3 138.0 15.7 2.176 0.036 0.266 0.001 1.291 2.85 0.005

C. berlandieri C. perfoliata P. oleracea dry 44.1 14.5 331.1 0.267 0.022 0.017 0.001 0.072 0.001 221.85

E. caespitosa E. californica P. dubium dry 100.9 52.2 617.4 0.010 3.091 0.829 9.309 0.002 56.02 0.004

L. bicolor L. purshianus T. incarnatum dry 7.1 13.1 392.8 3.983 3.551 13.42 2.158 0.003 8.221 0.009

L. glabrata L. californica C. melitensis dry 468.7 666.3 1.6 0.157 0.214 0.438 0.042 0.105 0.066 0.689

M. nuttalliana A. patula S. gallica dry 1148.4 0.1 923.8 0.045 0.001 0.153 0.611 <0.001 8.303 <0.001

P. campanularia N. menziesii M. arvensis dry 103.9 3.8 303.2 0.361 0.009 0.136 0.006 0.026 0.002 <0.001

S. columbariae C. heterophylla S. viridis dry 305.3 124.3 69.8 0.130 0.518 0.177 0.847 <0.001 0.444 0.038

U. lindleyi M. elegans C. capillaris dry 38.3 120.3 166.9 0.505 3.608 0.855 0.001 0.005 0.001 0.002

V. microstachys V. octoflora V. myuros dry 152.9 1126.9 677.3 0.082 0.515 0.248 0.002 0.574 0.097 0.022

Notes: See Table S5.1 and Fig. S5.1 for full species names and taxonomic information. Focal species i (bold) were competed against

both species j (sympatric) and k (allopatric).

121

Table S5.4. Germination (g) and species-specific seed bank viability (s) rates used for alternative

seed bank viability scenarios

Species Germination Seed bank viability

i j k gi gj gk si sj sk

B. carinatus H. depressum H. vulgare 0.39 0.24 0.49 0.17 0.17 0.00

C. berlandieri C. perfoliata P. oleracea 0.62 0.25 0.35 0.00 0.35 0.11

E. caespitosa E. californica P. dubium 0.17 0.08 0.28 0.18 0.59 0.28

L. bicolor L. purshianus T. incarnatum 0.14 0.17 0.22 0.01 0.08 0.07

L. glabrata L. californica C. melitensis 0.58 0.46 0.48 0.13 0.23 0.00

M. nuttalliana A. patula S. gallica 0.36 0.24 0.80 0.59 NA 0.00

P. campanularia N. menziesii M. arvensis 0.54 0.28 0.61 0.47 0.64 0.23

S. columbariae C. heterophylla S. viridis 0.55 0.40 0.66 0.29 0.56 0.00

U. lindleyi M. elegans C. capillaris 0.32 0.07 0.09 0.61 0.03 0.23

V. microstachys V. octoflora V. myuros 0.54 0.51 0.63 0.88 0.71 0.58

Notes: See Table S5.1 and Fig. S5.1 for full species names and taxonomic information. Focal

species i (bold) were competed against both species j (sympatric) and k (allopatric). Rates g and s

were calculated separately, from the germination of seeds in the low-density λ pots (for g) and

through a germination trial comparing germination rates before and after gibberellic acid

application (for s); the 0% and 100% seed bank viability scenarios were s = 0 and s = 1 for all

species, respectively. The same values of g and s were used in wet and dry environments,

because the watering treatments were imposed post-germination.

122

Table S5.5. Effects of phylogenetic distance (PD), biogeographic history (BH), soil moisture (SM), and their interactions on

stabilizing differences, fitness differences, and coexistence

logit(stabilizing differences) log(fitness differences) log(coexistence metric)

χ2 P-value χ

2 P-value χ

2 P-value

PD 0.02 0.900 17.72 <0.001 2.08 0.149

BH 0.70 0.403 5.82 0.016 1.92 0.166

SM 0.01 0.922 0.26 0.613 1.94 0.164

PD × BH 8.75 0.003 2.60 0.107 3.89 0.049

PD × SM 1.15 0.283 1.21 0.271 2.07 0.151

BH × SM 1.12 0.290 0.38 0.537 0.18 0.675

PD × BH × SM 0.08 0.775 <0.01 0.960 <0.01 0.979

Variance weights none varIdent(form=~PD|BH) varPower(form=~PD)

Notes: All P-values based on a χ2 tests of maximum likelihood ratios with 1 degree of freedom, summarized using the ‘Anova’

function in ‘car’ R package; significant P-values are in bold type.

Table S5.6. Effects of phylogenetic distance (PD) on the fitness components of sympatric and allopatric species pairs

Sympatric Allopatric

χ2 P-value χ

2 P-value

Demographic component 5.91 0.015 0.70 0.402

Competitive component 0.02 0.881 6.67 0.010

Fitness difference 11.39 <0.001 7.35 0.007

Notes: All P-values based on a χ2 tests of maximum likelihood ratios with 1 degree of freedom, summarized using the ‘Anova’

function in ‘car’ R package; significant P-values are in bold type. Variance weights are varIdent(~PD) for allopatric pairs; no weights

were necessary for sympatric pairs.


123

Figure S5.1. Maximum clade credibility tree of the 30 annual plant species with median node

heights. The tree was generated in BEAST using ITS1/5.8S/ITS2 and rbcL sequence data, and

calibrated to real-time based on fossil records. The scale bar is in millions of years ago, and

nodal support values are the posterior probabilities; * indicates nodes that were fixed based on a

priori information. See Table S5.1 for species information.

124

Figure S5.2. Comparison of fitness differences and coexistence under three seed bank viability

scenarios (0%, species-specific (Table S5.4), and 100%); the 0% scenario is presented in the

main text. We took the approach of Godoy & Levine (2014) for incorporating germination rates,

in which non-zero seed bank viability modifies our calculations of fitness differences and

coexistence, but not stabilizing differences; as such, the latter is not shown. Data points represent

sympatric (light green) and allopatric (dark green) species pairs, either averaged across soil

environments (circles; a-c) or in wet (squares) and dry (triangles) soil moisture conditions (d-f);

soil moisture was included in each analysis. + and × indicate significant main effects or a

significant interaction, respectively, of phylogenetic distance and biogeographic history.


125

Copyright Acknowledgements

Chapter 2: Reprint permissions were provided by the publisher, John Wiley and Sons (license #:

3883620546808)

Chapter 5: Reprint permissions are not required by Proceedings of the Royal Society B:

Biological Sciences or their parent publisher

historical contingencies in the ecology and evolution of

Documents