Seed Dispersal, Germination and Fine-Scale Genetic Structure in the Stream Lily, Helmholtzia Glaberrima (Philydraceae)

Queensland University of Technology

School of Natural Resource Sciences

Seed dispersal, germination and fine-scale

genetic structure in the stream lily,

Helmholtzia glaberrima (Philydraceae)

Peter Prentis B.Sc. (Hons).

Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

July 2005

Abstract

Seed dispersal in aquatic habitats is often considered to be a complex multistage process, where initial seed shadows are redistributed by water (hydrochory). The roles of hydrochory in seed dispersal and influencing population genetic structure were examined in Helmholtzia glaberrima using both ecological and genetic techniques.

Ecological experiments showed that water can redistribute seeds and seedlings over local scales and that hydrochory can provide the potential for very long distance seed and seedling dispersal. Patterns of seedling genetic structure were affected by micro- drainages that direct water flow within populations and influence water-borne seed dispersal on a local scale. Strong non-equilibrium dynamics and persistent founder effects were responsible for the patterns of genetic structure observed among established populations of H. glaberrima. Classical metapopulation models best described dispersal patterns, while water-borne seed dispersal could potentially explain patterns of genetic differentiation within a stream system, it could not explain the distribution of genetic variation among stream systems. The current study found that although hydrochory influenced seed dispersal and seedling genetic structure within a population, it had little effect on the spatial pattern of genetic variation among established populations of H. glaberrima. Moreover, even though prolonged buoyancy and viability in water provide the potential for long-distance hydrochory, results presented here do not support the hypothesis that flowing water is an effective long distance seed dispersal vector for H. glaberrima. Taken together, these results suggest that the relative importance of gene flow via water-born seed dispersal in H. glaberrima may be low compared with that of some other riparian species.

Keywords: genetic diversity; Helmholtzia; hydrochory; seed dispersal

List of manuscripts

1. Prentis, P. J., A. Vesey, N. M. Meyers, and P. B. Mather 2004. Genetic structuring

of the stream lily Helmholtzia glaberrima (Philydraceae) within Toolona Creek,

south-eastern Queensland. Australian Journal of Botany 52: 201—207.

2. Prentis, P. J., N. M. Meyers, and P. B. Mather. In press. The significance of post-

germination buoyancy in Helmholtzia glaberrima and Philydrum lanuginosum

(Philydraceae). Australian Journal of Botany.

3. Prentis, P. J., N. M. Meyers, and P. B. Mather. (In Prep) Seed dispersal and

seedling establishment in the riparian plant Helmholtzia glaberrima. Freshwater

Biology.

4. Prentis, P. J., N. M. Meyers, and P. B. Mather. (In Review) Micro-geographic

landscape features demarcate seedling genetic structure in the stream lily,

Helmholtzia glaberrima (Philydraceae)1. American Journal of Botany.

5. Prentis, P. J., N. M. Meyers, AND P. B. Mather. (In Review) Fine-scale patterns of

genetic diversity and population structure in the stream lily Helmholtzia

glaberrima (Philydraceae) along rainforest streams, south-east Queensland.

Freshwater Biology.

Table of Contents

Abstract 1

List of Manuscripts 3

Table of Contents 4

Statement of Original Authorship 7

Acknowledgments 8

Chapter 1: GENERAL INTRODUCTION

Introduction 9

Spatial and temporal dynamics of aquatic habitats 10

Seed dispersal in freshwater habitats 12

Colonisation of isolated habitat patches via long-distance hydrochory 16

Other mechanisms affecting long-distance seed transport 17

Study species and system 22

Account of research progress linking manuscripts 24

References 26 Chapter 2: GENETIC STRUCTURING OF THE STREAM LILY Helmholtzia glaberrima (PHILYDRACEAE) WITHIN TOOLONA CREEK, SOUTH-EAST

QUEENSLAND

Statement of Joint Authorship 34

Manuscript 1 35

Introduction 37

Materials and Methods 40

Results 43

Discussion 45

References 50

Tables and Figures 54

Chapter 3: THE SIGNIFICANCE OF POST-GERMINATION BUOYANCY IN

Helmholtzia glaberrima AND Philydrum lanuginosum (PHILYDRACEAE)

Statement of Joint Authorship 61

Manuscript 2 62

Introduction 64

Materials and Methods 67

Results 72

Discussion 74

References 78

Tables and Figures 81

Chapter 4: SEED DISPERSAL AND SEEDLING ESTABLISHMENT IN THE RIPARIAN

PLANT Helmholtzia glaberrima

Statement of Joint Authorship 84

Manuscript 3 85

Introduction 88

Materials and Methods 91

Results 94

Discussion 96

References 100

Tables and Figures 102

Chapter 5: MICRO-GEOGRAPHIC LANDSCAPE FEATURES DEMARCATE

SEEDLING GENETIC STRUCTURE IN THE STREAM LILY, Helmholtzia glaberrima

(PHILYDRACEAE)

Statement of Joint Authorship 105

Manuscript 4 106

Introduction 108

Materials and Methods 110

Results 112

Discussion 114

References 118

Tables and Figures 121

Chapter 6: FINE-SCALE PATTERNS OF GENETIC DIVERSITY AND POPULATION

STRUCTURE IN THE STREAM LILY Helmholtzia glaberrima (PHILYDRACEAE)

ALONG RAINFOREST STREAMS, SOUTH-EAST QUEENSLAND

Statement of Joint Authorship 126

Manuscript 5 127

Introduction 129

Materials and Methods 132

Results 134

Discussion 136

References 140

Tables and Figures 143

Chapter 7: GENERAL DISCUSSION

General Discussion 147

References 153

Statement of Original Authorship

This work contains no material which has been accepted for the award of any other degree or diploma in any university or other tertiary institution and, to the best of my knowledge and belief, contains no material previously published or written by another person, except where due reference has been made in the text.

Signed

Date

Acknowledgments

Firstly, I would like to thank my supervisors for your guidance and encouragement in helping me complete this project. Many thanks must go both to the ecology and genetics discussion groups for fruitful discussions on sampling designs and data. My parents deserve many thanks for their support of my chosen path and providing guidance during the gloomy PhD blues.

A special thankyou to all the people who volunteered to help me with field and lab work, particularly Ana Pavasovic, Cameron Schulz, Mark Schutze, Doug Harding, Alex Wilson and Grant Hamilton. Big thanks must go to Dr Graham Kelly the man whose lectures and infectious enthusiasm are responsible for my interest in plant biology. Lastly and mostly importantly, I have to thank my partner Ana for her support, encouragement and cooking without which I would be a lonely, thin man.

CHAPTER 1. General introduction

Introduction

How seeds disperse among isolated habitat patches, establish and contribute to the genetic pool of new plant populations has long fascinated botanists (Darwin, 1859). This is because dispersal is one of the primary processes that influence population dynamics and evolution of plants (Nathan &

Muller-Landau, 2000). In particular the spatial and temporal dynamics of plant populations will be determined chiefly by the movement of seeds within and among populations (Husband & Barrett, 1998). At a broader geographic scale, the range at which seed dispersal is effective will influence the possibility that extirpated populations are recolonised and the probability that new isolated habitat patches are colonised (Cain et al., 2000).

Plant species are rarely distributed uniformly in space but often occur as isolated local populations where favourable conditions for successful establishment exist (Edwards & Sharitz, 2000; Ellison & Parker, 2002). Dispersal patterns within and among local populations will determine the extent to which local populations are interconnected via gene flow (Ouborg et al., 1999). Gene flow will only occur however, if the seeds dispersing among populations establish and contribute to future reproduction in the new population (Ouborg et al., 1999).

The level of gene flow among patches will determine the distribution of genetic variation within and among local populations and whether they function collectively, or as isolated units (Tero et al., 2003). In situations where dispersal

rates among populations are high, gene flow will tend to homogenise gene frequencies among local populations, so they form a single panmictic unit

(Ouborg et al., 1999). Alternatively, if gene flow is very low or absent, local populations are likely to diverge and evolve independently due to the diversifying forces of drift and selection (Tero et al., 2003). A number of factors can influence the level of seed dispersal and gene flow that occurs among isolated local populations, including temporal and spatial distribution of favourable habitat patches, the type of vectors that disperse seeds and the individual life-history characteristics of a plant species (Pannell and Charlesworth, 1999, 2000). This review will focus primarily factors acting on aquatic plants restricted to freshwater habitats.

Spatial and temporal dynamics of aquatic habitats

Many studies have reported that habitat patches are often dispersed heterogeneously across the landscape (Husband & Barrett, 1998; Levins, 1969;

McCauley, 1989). Spatial discontinuities, including mountains, valleys or rivers can influence significantly the distribution of many organisms (Lawton, 1993).

Most organisms do not exhibit uniform distributions in space (Lawton, 1993).

Instead organisms frequently occur as isolated local populations confined to suitable habitat surrounded by a matrix of less favourable habitat (Andrewartha &

Birch, 1954; Dejong, 1995; Dupre & Ehrlen, 2002). This is particularly true for organisms with narrow habitat requirements such as freshwater aquatic plants, whose preferred habitat is often distributed like naturally occurring islands in a

sea of terrestrial habitat (Barrett, 1985; Husband & Barrett, 1998; Santamaria,

2002).

Freshwater aquatic plant species are restricted to permanent, seasonal or ephemeral aquatic environments, such as rivers, lakes or ponds (Edwards &

Sharitz, 2000; Ellison & Parker, 2002). Most freshwater aquatic habitats occur as discrete habitat patches often separated by hundreds or thousands of metres of unfavourable areas (Barrett et al., 1993). As a result, populations of many aquatic plant species have spatially disjunct distributions. Freshwater aquatic habitats in most landscapes are also highly heterogeneous at several spatial and temporal scales, meaning that the distribution of many aquatic plant species can also be temporally unstable (Husband & Barrett, 1995, 1998).

Most freshwater aquatic habitats are not static as they are created and replenished by precipitation or runoff (Brock & Rogers, 1998; Hampe, 2004;

Lopez, 2001; Scarano et al., 2003; Van der Valk, 1981). As rainfall is not usually evenly distributed in space or time, temporal and spatial fluctuations in water level occur in freshwater habitats particularly in areas with seasonal climates

(Hampe, 2004; Husband & Barrett, 1998). Consequently, many aquatic habitats are ephemeral or experience dramatic water level fluctuations among seasons depending on the frequency and duration of rain or runoff from catchments

(Brock & Rogers, 1998). In periods of heavy rain, freshwater habitats may flood and connect to other isolated wetland patches; however during drought many aquatic habitats may disappear. As freshwater aquatic habitats can be

ephemeral, aquatic plants inhabiting them may experience alternating wet and dry conditions (Lopez, 2001).

Alternating wet and dry conditions can influence the persistence of aquatic plant populations in particular survival during the phase from seed release to seedling establishment (Husband & Barrett, 1998; Lopez, 2001). Due to the unpredictable nature of habitats occupied by most aquatic plants, individual species may have only a short window of opportunity for seed dispersal, germination and seedling establishment (Hampe, 2004). Many wetland plants, however, have evolved unusual adaptations to water level changes or are capable of variable responses (plasticity), promoting survival in heterogeneous environments (Arber, 1920; Dorken & Barrett, 2004a; Santamaria et al., 2003;

Sculthorpe, 1967; Wells & Pigliucci, 2000). Thus seed dispersal, germination and seedling establishment in aquatic plants may be less constrained by fluctuations in water levels and alternating wet and dry conditions than had previously been thought (Lopez, 2001). Traits that favour increased seed floatation during flooding, seed survival in water and/or seedling establishment under flooded conditions should be advantageous to aquatic or riparian plants.

Seed dispersal in freshwater habitats

Patterns of seed dispersal and seedling establishment have important effects on the dynamics and persistence of plant populations (Barrett et al.,

1993). Dispersal of seeds provides an opportunity for individual plants or populations to recruit because some seeds are likely to arrive at suitable

microhabitats for seedling establishment, thus increasing their potential for survival (Nathan & Muller-Landau, 2000). Seed dispersal is often a complex multistage process, with seeds known to be dispersed by a variety of vectors

(primary and secondary; sensu Vander Wall & Longland, 2004) over a number of different spatial scales (local and long-distance) (Chambers & McMahon, 1994;

Hampe, 2004). Local and long-distance dispersal in complex seed dispersal systems can often affect different components of the demography and dynamics of individual plants and populations (Cain et al., 2000).

Seed dispersal in aquatic habitats often occurs as a multistage process, complicated by each individual plant’s life history characteristics (Hampe, 2004).

Broadly we can classify aquatic plants into four life-history categories; 1) semi- aquatic/riparian species, 2) emergent aquatic species 3) free-floating aquatic species and 4) submerged aquatic species (Sculthorpe, 1967). For the purposes of this review we will concentrate principally on emergent and semi- aquatic/riparian species. Early life-history traits that affect seed dispersal in emergent and semi-aquatic species can be quite similar (Boedeltje et al., 2003).

For example many emergent and semi-aquatic wetland species exhibit a limited capacity for primary seed dispersal due to their reliance on gravity as a vector

(Waser et al., 1982).

The mean dispersal distance for the herbaceous wetland perennial

Sarracenia purpurea, was 12.8 cm with 78% of seeds caught within five centimetres of focal plants (Ellison & Parker, 2002). Similarly short primary dispersal distances have also been found in many other wetland herbs (Cain et

al., 1998). Reliance solely on short-distance primary seed dispersal in these species should constrain the potential for movement of seeds among patches, yet propagules are known to have established in isolated sites following rare dispersal events across large tracts of unsuitable habitat (Ellisson & Parker,

2002). Understanding dispersal mechanisms that facilitate these rare events is the first step in understanding colonisation, establishment and dynamics of spread in emergent or semi-aquatic plant species that otherwise possess limited dispersal capacity.

Many riparian and emergent aquatic species exhibit prolonged seed buoyancy (Boedeltje et al., 2003). Consequently seeds in wetland species can be redistributed by water (Hampe, 2004). Redistribution of seeds in fresh water may affect the spatial pattern of the seed shadow generated by primary seed dispersal vectors (Hampe, 2004). Prolonged seed buoyancy may enhance the potential range of seed dispersal, possibly conferring the capacity to colonise isolated habitat patches. Redistribution of seeds by fresh water however, depends strongly on the amount of rainfall or runoff received between seed release and seedling establishment (Hampe, 2004). For example, if seed release coincides with dry conditions then there is limited or no opportunity for seed dispersal by water in ephemeral wetlands. Alternatively if rainfall does occur during times of seed release then redistribution of seeds by water is highly likely.

Dispersal via primary and secondary dispersal vectors are poorly understood, particularly systems where initial seed shadows are redistributed by freshwater

(Hampe, 2004).

The relative contribution that hydrochory makes to secondary dispersal and establishment of wetland species remains contentious. Most likely, in many freshwater herbs water-born seed dispersal contributes little to long-distance dispersal and if it does, it represents a rare event (Ellison & Parker, 2002). Some investigators have found evidence that hydrochory can result in effective long- distance seed dispersal at scales of hundreds to thousands of metres (Kinamoto et al., 2005; Kudoh & Whigham, 1997; Lonsdale, 1993; Nilsson et al., 1991;

Ridley, 1930). This indicates that in at least some emergent or riparian species long-distance water-borne seed dispersal is possible. Consequently, rare long- distance hydrochory may explain colonisation of isolated habitat patches in both emergent and semi-aquatic species (Tero et al., 2003).

Redistribution of seeds via water however may affect seed germination and seedling recruitment. Water-borne seed dispersal can affect seed germination or establishment in at least three ways; first, if seeds dispersed by water are not flood tolerant then immersion in water for prolonged periods may reduce seed viability and germination (Edwards et al., 1994; Lopez, 2001;

Middleton, 2000; Scarano et al., 2003). Second, redistribution of seeds by water will determine if seeds are deposited in suitable microhabitat conditions for germination and establishment (Boedeltje et al., 2004; Ozinga et al., 2004; Van der Valk, 1981). Last, if water-born dispersal changes the density of the initial seed shadow, germination and establishment may be negatively or positively affected by density dependent processes such as mortality (Debussche &

Isenmann, 1994; McMurray et al., 1997; Nathan & Casagrandi, 2004; Orth et al.,

2003; Russo & Augspurger, 2004). Therefore, how seeds of emergent or riparian species respond to water-borne dispersal and where they end up (deposition site) will influence potential for colonisation of isolated habitat patches (Nathan &

Muller-Landau, 2000).

Colonisation of isolated habitat patches via long-distance hydrochory

Populations of emergent or riparian plant species often occur in spatially isolated habitat patches (Edwards & Sharitz, 2000). Many hundreds to thousands of metres may separate individual wetland habitat patches (Ellison &

Parker, 2002). Thus colonisation of suitable isolated patches will require effective long-distance dispersal. The capacity of seeds from emergent or riparian species immersed in water to float for long periods of time (Boedeltje et al., 2003), may represent an adaptation to long-distance water-borne dispersal. For example,

Fridriksson (1975) found that species with adaptations that promoted buoyancy were more likely to disperse over distances of greater than 20 km on ocean currents to colonise the newly emerged volcanic island of Surtsey. In fact, over

78% of the vascular plant taxa that arrived on Surtsey between 1963 and 1972 were known to be dispersed by water (Fridriksson, 1975). This demonstrates that water-borne dispersal can be an effective method for long-distance dispersal in plants and the colonisation of isolated habitats. It is unknown however, if secondary seed dispersal by flowing freshwater can disperse seeds over relatively long distances allowing colonisation of isolated habitats.

Experiments have shown that seeds of many wind dispersed species, often float and remain viable in runoff, streams and rivers (Ridley, 1930). It is thought in some of these instances, that seeds can be dispersed much greater distances via water than by wind (Carlquist, 1967). Furthermore, a recent study by Hampe (2004) found that secondary dispersal by water changed the distribution of the bird-mediated seed shadow of the Eurasian tree, Frangula alnus from a negative exponential towards an extended Poisson distribution. This indicates that secondary seed dispersal by water can result in regular long- distance seed dispersal for F. alnus. Secondary water-borne dispersal may therefore represent an important long-distance dispersal vector, aiding the colonisation of isolated habitat in some emergent or riparian species. More than one type of dispersal vector however may result in long-distance secondary seed dispersal.

Other mechanisms affecting long-distance seed transport

There are several ways that long distance seed dispersal events may occur including; dispersal by water (Kudoh & Whigham, 1997; Lonsdale, 1993;

Nilsson et al. 1991; Ridley, 1930), by biotic influences (Figuerola & Green, 2002;

Fragosa, 1997; Holbrook et al. 2002; Shilton et al. 1999), wind vectors (Bullock &

Clarke, 2000; Nathan et al. 2002) or rafting on other dead or live organisms

(Nathan & Muller-Landau, 2000). Some researchers assume a link between seed attributes such as morphology and dispersal (e.g. buoyant seeds imply hydrochory) (e.g. Hughes et al., 1994). A recent review however, recognised that

the morphological attributes a seed may possess does not necessarily determine which vector will disperse it (Higgins et al. 2003). Moreover, this study suggests that although dispersal types based on morphological traits provide a useful framework for describing local dispersal processes, they are often poorly related to the actual mechanisms responsible for long-distance seed dispersal in many plant species (Higgins et al. 2003).

Many wind and water dispersed seeds from wetland species are known to adhere to the fur, feathers or feet of some vertebrate species (Carlquist, 1967;

Darwin, 1859; Figuerola & Green, 2002; Higgins & Richardson, 1999; Ridley,

1930; Vivian-Smith & Siles, 1994). This will confer a potential for long-distance dispersal by mechanisms other than those determined primarily by morphological traits (Higgins et al. 2003). For example, it is estimated that approximately one quarter of the plant species that colonised Easter Island and the Juan Fernandez

Islands were transported in mud attached to the feet of migratory birds (Carlquist,

1967). In addition, the distribution of some wetland plants with buoyant seeds are strongly correlated with the dispersal patterns and preferred habitat of migratory waterfowl (Frith et al. 1977) suggesting that waterfowl may act as dispersal vectors for these species. Further evidence supporting the role that biotic vectors play in dispersing the seeds of aquatic plants over long distances comes from population genetic studies (Tero et al., 2003).

Studies of the genetic structure of many aquatic species suggest that dispersal mechanisms other than water have been important in some long distance seed dispersal events (Dorken & Barrett, 2004b; Godt & Hamrick, 1998;

Tero et al., 2003). A study of the genetic structure of the fennel pondweed

(Potamogeton pectinatus) in Europe found strong evidence that long distance seed dispersal in this species was not related to hydrochory (Mader et al., 1998).

The study established a correlation between genetic and geographic distance among ponds not visited by swans (a potential dispersal agent). When ponds visited by swans were included in the analysis no relationship was evident.

Based on this finding the authors concluded that swan-mediated long-distance dispersal influenced the pattern of genetic structure observed (Mader et al.,

1998). Therefore, molecular markers in combination with well designed experiments can be used to examine patterns of dispersal in plant species

(Abbott et al., 2000; He et al., 2004; Mader et al., 1998).

Genetic methods are broadly applicable to investigating both patterns of local and long-distance seed dispersal in plant species (Cain et al., 2000; He et al., 2004; Kinamoto et al. 2005; Ouborg et al., 1999; Sork et al., 1999). Studies of seed dispersal in plant species that have applied genetic markers however, have usually sampled seedlings or older stages of plant development (Cain et al.,

2000). If genetic studies of dispersal are based on samples collected from seedlings or later stages of development, they not only provide estimates of seed dispersal but also plant establishment, thus providing a measure of ‘effective’ seed dispersal (Nathan & Muller-Landau, 2000; but see Godoy & Jordano, 2001).

Effective seed dispersal may be the parameter of greater importance in many ecological and conservation studies, since seed movement to locations in which

establishment fails will not influence the population dynamics and genetic structure of the recipient population (Nathan, 2006).

Several models have been proposed to explain how dispersal can influence population structure in plant populations. The types of models proposed include: island models, fragmented models, stepping stone models, classical metapopulation models and source/sink models (Gaggiotti and Smouse, 1996;

Kimura 1953; Kimura & Weiss 1964; Levins, 1969; Wright, 1931).

In an island model, all populations form a single panmictic unit with free gene flow across a genetically uniform population (Wright, 1931). This model assumes that populations have reached equilibrium, that all are of equal size and that dispersal rate is uniform among populations irrespective of their relative proximity in the landscape (Wright, 1951). Where panmixia is operating in an island model, populations should not show any structure (Wright, 1951). A fragmented model however, assumes that a formerly continuous population has been fragmented into remnant populations where no contemporary gene flow occurs among populations (Slatkin, 1985). Strong genetic structuring is expected therefore among all populations in a fragmented system (Slatkin, 1985).

In contrast to the island and fragmented models, a stepping-stone model assumes dispersal is limited by distance and occurs at higher rates among adjacent local populations (Kimura 1953; Kimura & Weiss 1964). At equilibrium in a stepping-stone model, genetic differentiation among populations should increase linearly with increasing geographical distance among populations

(Kimura & Weiss 1964). Therefore a signature of isolation by distance is expected under a stepping-stone model (Slatkin, 1985).

In both source/sink and classical metapopulation models, populations are not permanent and are subject to extinction/colonization dynamics (Pannell and

Charlesworth, 1999). Source/sink metapopulation models assume that individuals disperse from more permanent, high-quality source subpopulations into low quality ephemeral sink populations and that sink populations are likely to go extinct without ongoing dispersal from source populations (Gaggiotti and

Smouse, 1996). Therefore sink populations are expected to possess lower levels of genetic diversity relative to source populations. Classical metapopulation models assume that all patches are ephemeral and may be founded by different numbers of individuals from single or multiple source patches (Giles and Goudet,

1997). This model also assumes that only limited dispersal occurs among extant populations. Consequently, levels of genetic variation may decrease or increase randomly within and among populations under a metapopulation model (Pannell and Charlesworth, 1999, 2000).

These models provide a number of testable hypotheses about how different modes of dispersal can influence the distribution of genetic variation within and among subpopulations. Therefore this genetic information permits inferences to be made about patterns of dispersal within and among populations.

In conjunction with ecological experiments genetic studies can be used to elucidate patterns of local and long-distance seed dispersal.

Study species and system

The Philydraceae is a small family of chiefly erect herbs that is restricted to Australia and south east Asia (Clifford, 1983; Hamann, 1998). Family

Philydraceae is closely related to the Pontederiaceae and Haemodoraceae in the

O. Commelinales (Graham et al., 2002). Species within this family are only found in saturated areas like permanently waterlogged soil, swamps or soaks (Clifford,

1983). The Family Philydraceae comprises a total of six species distributed in 3 genera (Adams 1989). The 3 genera include the montane rainforest genus

Helmholtzia (3 species: H. glaberrima, H. acorifolia and H. novoguineensis) and

2 lowland genera Philydrella (2 species: P. drummondii and P. pygmaea) and

Philydrum (monotypic: P. lanuginosum) (Adams, 1989). The species examined in this study was H. glaberrima a semi-aquatic riparian plant (Hamann, 1998).

H. glaberrima is a long-lived perennial herb (Philydraceae), (2n =34) that is known to live for at least ten years and which grows to 2.5 m (Adams 1989).

The species has low levels of protracted flowering throughout the year and flowers are produced on either single or multiple large pink panicles which are believed to be pollinated by small insects (Adams 1989). Numerous 4-5 mm long seeds are produced in capsules throughout the infructescence after pollination

(Adams 1989). Currently, no dispersal vectors are known that disperse the seed of this plant, but hydrochory has been suggested as a possible dispersal mechanism (Hamann, 1998).

The stream lily, H. glaberrima (Philydraceae), is a rare but occasionally locally abundant, riparian species confined to high-elevation dendritic streams in

temperate and subtropical rainforest of the McPherson Ranges on the border of

Queensland and New South Wales (Clifford, 1983). Within its range of approximately 100 km, H. glaberrima is restricted to riparian wetlands and moist gullies along high elevation creek banks (Hamann, 1998). As H. glaberrima is restricted to permanently waterlogged areas, it occurs in groups of spatially isolated patches along stream systems. Consequently, populations at the top of stream systems are thought to be less isolated from each other than downstream populations. Although extant populations are spatially isolated from one another they are interconnected by permanent stream systems (Adams 1989).

In the McPherson Ranges amount of rainfall received decreases with altitude. Altitudes above 1000 m receive ≤ 3000 mm in annual rainfall, while areas between 800 – 1000 m receive ≤ 1800 mm annually and areas below

800m receive < 1300 mm of annual rainfall (BOM, 2003). Consequently, the preferred habitat of H. glaberrima declines with altitude as presumably do the number and size of H. glaberrima populations.

Rainfall is also highly seasonal in south east Queensland and northern

New South Wales with most rain falling over the summer period between

December and March (BOM, 2003). Many patches occupied by H. glaberrima are subject to periodic inundation and soil saturation in summer that usually recedes to moist soil conditions over winter, the same time that many species disperse their seed in this area (Prentis, personal observation). Therefore H. glaberrima seeds have a strong possibility of being redistributed by water during periods of heavy rain.

A combination of genetic and ecological approaches were used to explore the role that hydrochory plays in the dispersal of H. glaberrima seeds over local and long distances. Ecological approaches were used to examine the potential for water-borne seed dispersal to transport seeds over local and long distances.

Estimates of genetic diversity among established populations of H. glaberrima and of seedlings within a population were examined to elucidate the actual role that water-borne seed dispersal may have on population genetic structure over local and long-distance scales. These data were used to address the five aims of this project:

1. To define what model of dispersal can best explain the observed

population structure (manuscript 1)

2. To examine potential for water-borne seed dispersal over local and

long-distance scales (manuscript 3)

3. To investigate the role that post-germination buoyancy plays in

seed dispersal and seedling establishment (Manuscript 2)

4. To test whether microtopographic landscape features that direct

water flow can influence seed dispersal and seedling genetic

structure in local populations (manuscript 4)

5. To quantify the influence that water-borne seed dispersal has on

observed population genetic structure (manuscript 5)

Account of research progress linking manuscripts

The first manuscript (chapter 2) examines the model of dispersal which best explains the observed pattern of population genetic structure in Helmholtzia along a single stream system. The pattern of genetic structure observed along this stream system was best explained by classical metapopulation dynamics and a complex seed dispersal strategy which was hypothesised to be the result of hydrochory. The subsequent two manuscripts (chapter 3 & 4) investigated the role of hydrochory to ‘effectively’ disperse seeds and seedlings over local and long distances. The fourth and fifth manuscripts (chapter 5 & 6) then evaluated if the hydrochory actually influenced the pattern of genetic structure in Helmholtzia over a local (seedling genetic structure) and long-distance scale (among stream systems).

References

Abbott, R.J. Smith, L.C. Milne, R.I. Crawford, M. Wolff, K. & Balfour, J. (2000) Molecular

analysis of plant migration and refugia in the Arctic. Science, 289, 1343-1346.

Adams, L.G. (1989) Philydraceae. In: Flora of Australia v45 Hydatellaceae to Liliaceae.

(Ed. George, A.S.) pp. 40-46. AGPS, Canberra.

Andrewartha, H.G. & Birch, L.C. (1954) The distribution and abundance of animals

University of Chicago Press, Chicago.

Arber, A. (1920) Water plants: a study of aquatic angiosperms Cambridge University

Press, Cambridge.

Barrett, S.C.H. (1985) Floral trimorphism and monomorphism in continental and island

populations of Eichhornia paniculata (Spreng.) Solms (Pontederiaceae).

Biological Journal of the Linnean Society, 25, 41-60.

Barrett, S.C.H., Eckert, C.G., & Husband, B.C. (1993) Evolutionary processes in aquatic

plant populations. Aquatic Botany, 44, 105-145.

Boedeltje, G. Bakker, J.P. Bekker, R.M. Van Groenendael, J.M. & Soesbergen, M.

(2003) Plant dispersal in a lowland stream in relation to occurrence and three

specific life-history traits of the species in the species pool. Journal of Ecology,

91, 855-866.

Boedeltje, G. Bakker, J.P. Ten Brinke, A. Van Groenendael, J.M. & Soesbergen, M.

(2004) Dispersal phenology of hydrochorous plants in relation to discharge, seed

release time and buoyancy of seeds: the flood pulse concept supported. Journal

of Ecology, 92, 786-796.

BOM (2003) Bureau of meteorology weather records. Brisbane, Australia.

Brock, M.A. & Rogers, K.H. (1998) The regeneration potential of the seed bank of an

ephemeral floodplain in South Africa. Aquatic Botany, 61, 123-135.

Bullock, J.M. & Clarke, R.T. (2000) Long distance seed dispersal by wind: measuring

and modelling the tail of the curve. Oecologia, 124, 506-521.

Cain, M. L. Damman, H. & Muir, A. (1998) Seed dispersal and the Holocene migration of

woodland herbs. Ecological Monographs, 68, 325-347.

Cain, M. L. Milligan, B. G. & Strand, A. E. (2000) Long-distance seed dispersal in plant

populations. American Journal of Botany, 87, 1217-1227.

Carlquist, S. (1967) The biota of long distance dispersal. V. Plant dispersal to Pacific

islands. Bulletin of the torrey Botanical Club, 94, 129-162.

Chambers, J.C. & MacMahon, J.A. (1994) A day in the life of a seed: movements and

fates of seeds and their implications for natural managed systems. Annual

Review of Ecology and Systematics, 25, 263-292.

Clifford, T.A. (1983) Philydraceae. In: Flowering plants in Australia (eds. Morley, B.D. &

Toelken, H.R.), pp. 345-347. Rigby Publishers, Adelaide.

Darwin, C.R. (1859) The origin of species. John Murray, London.

Debussche, M. & Isenmann, P. (1994) Bird-dispersed seed rain and seedling

establishment in patchy mediterranean vegetation. Oikos, 69, 414-426.

Dejong, J. (1995) Habitat use and species richness of bats in a patchy landscape. Acta

Theriologica, 40, 237-248.

Dorken, M.E. & Barrett, S.C.H. (2004a) Phenotypic plasticity of vegetative and

reproductive traits in monoecious and dioecious populations of Sagittaria latifolia

(Alismataceae): a clonal aquatic plant. Journal of Ecology, 92, 32-44.

Dorken, M.E. & Barrett, S.C.H. (2004b) Chloroplast haplotype variation among

monoecious and dioecious populations of Sagittaria latifolia (Alismataceae) in

eastern North America. Molecular Ecology, 13, 2699-2707.

Dupre, C. & Ehrlen, J. (2002) Habitat configuration, species traits and plant distributions.

Journal of Ecology, 90, 796-805.

Edwards, A.L. & Sharitz, R.R. (2000) Population genetics of two rare perennials in

isolated wetlands: Sagittaria isoetiformis and S. teres (Alismataceae). American

Journal of Botany, 87, 1147-1158.

Edwards, A.L. Wyatt, R. & Sharitz, R.R. (1994) Seed buoyancy and viability of the

wetland wilkweed Asclepias perennis and an upland milkweed, Asclepias

exaltata. Bulletin of the Torrey Botanical Club, 121, 160-169.

Ellison, A.M. & Parker, J.N. (2002) Seed dispersal and seedling establishment of

Sarracenia purpurea (Sarraceniaceae). American Journal of Botany, 89, 1024-

1026.

Figuerola, J. & Green, A.J. (2002) Dispersal of aquatic organisms by waterbirds: a

review of past research and priorities for future studies. Freshwater Biology, 47,

483-494.

Frith, H.J. Crome, F.H. & Brown, B.K. (1977) Aspects of the biology of the Japanese

Snipe Gallinago hardwickii. Australian Journal of Ecology, 2, 341-368.

Fragosa, J.M. (1997) Tapir-generated seed shadow: scale-dependent patchiness in the

Amazon rain forest. Journal of Ecology, 85, 519-529.

Fridriksson, S. (1975) Surtsey: evolution of life on a volcanic island. Butterworths,

London.

Gaggiotti, O.E. & Smouse, P.E. (1996) Stochastic migration and maintenance of genetic

variation in sink populations. American Naturalist, 147, 919–945.

Giles, B.E. & Goudet, J. (1997) Genetic differentiation in Silene dioica metapopulations:

estimation of spatiotemporal effects in a successional plant species. American

Naturalist, 149, 507–526.

Godoy, J.A. & Jordano, P. (2001) Seed dispersal by animals: exact identification of

parental trees with endocarp DNA microsatellites. Molecular Ecology, 10, 2275-

2283.

Godt, M.J.W. & Hamrick, J.L. (1998) Genetic divergence among intraspecific taxa of

Sarracenia purpurea. Systematic Botany, 23, 427-438.

Graham, S.W. Olmstead, R.G. & Barrett S.C.H. (2002) Rooting phylogenetic trees with

distant outgroups: a case study from the commelinoid monocots. Molecular

Biology and Evolution, 19, 1769-1781.

Hamann, U. (1998) Philydraceae. In: The families and genera of vascular plants. IV.

Flowering plants-monocotyledons-alismatanae and commelinanae (ed. Kubitski,

K.), pp. 389-394. Springer-Verlag, Berlin.

Hampe, A. (2004) Extensive hydrochory uncouples spatiotemporal patterns of seedfall

and seedling recruitment in a 'bird-dispersed' riparian tree. Journal of Ecology,

92, 797-807.

He, T.H. Krauss, S.L. Lamont, B.B. Miller, B.P. & Enright, N.J. (2004) Long-distance

seed dispersal in a metapopulation of Banksia hookeriana inferred from a

population allocation analysis of amplified fragment length polymorphism data.

Molecular Ecology, 13, 1099-1109.

Higgins, S.I. Nathan, R. & Cain,M.L. (2003) Are long-distance dispersal events in plants

usually caused by nonstandard means of dispersal? Ecology, 84, 1945-1956.

Higgins, S.I. & Richardson, D.M. (1999) Predicting plant migration rates in a changing

world: the role of long-distance dispersal. American Naturalist, 153, 464-475.

Holbrook, K.M. Smith, T.B. & Hardesty, B.D. (2002) Implications of long-distance

movements of frugivorous rain forest hornbills. Ecography, 25, 745-749.

Hughes, L. Dunlop, M. French, K. Leishman, M.R. Rice, B. Rodgerson, L.A. &

Westaboy, M. (1994) Predicting dispersal spectra: a minimal set of hypotheses

based on plant attributes. Journal of Ecology, 82, 933-950.

Husband, B.C. & Barrett, S.C.H. (1995) Estimates of gene flow in Eichhornia paniculata

(Pontederiaceae): Effects of range substructure. Heredity, 75, 549-560.

Husband, B.C. & Barrett, S.C.H. (1998) Spatial and temporal variation in population size

of Eichhornia paniculata in ephemeral habitats: implications for metapopulation

dynamics. Journal of Ecology, 86, 1021-1031.

Kimura, M. (1953) ‘Stepping-stone′ model of population. Annual Report of the National

Institute of Genetics, Japan, 3, 62–63.

Kimura, M. & Weiss, G.H. (1964) The stepping stone model of population structure and

the decrease of genetic correlation with distance. Genetics, 49, 561–576.

Kitamoto, N. Honjo, M. Ueno, S. et al. (2005) Spatial genetic structure within and among

populations of Primula sieboldii growing beside separate streams. Molecular

Ecology, 14, 149-157.

Kudoh, H. & Whigham, D.F. (1997) Microgeographic genetic structure and gene flow in

Hibiscus moscheutos (Malvaceae) populations. American Journal of Botany, 84,

1285-1293.

Lawton J.H. (1993) Range, population abundance and conservation. Trends in Ecology

and Evolution, 8, 409-413.

Levins, R. (1969) Some demographic and genetic consequences of environmental

heterogeneity for biological control. Bulletin of the Entomological Society of

America, 15, 237-240.

Lonsdale, W.M. (1993) Rates of spread of an invading species Mimosa pigra in northern

Australia. Journal of Ecology, 81, 513-521.

Lopez, O.R. (2001) Seed flotation and postflooding germination in tropical terra firme

and seasonally flooded forest species. Functional Ecology, 15, 763-771.

Mader, E. van Vierssen, W. & Schwenk, K. (1998) Clonal diversity in the submerged

aquatic macrophyte Potamogeton pectinatus L. inferred from nuclear and

cytoplasmic variation. Aquatic Botany, 62, 147-160.

McCauley, D.E. (1989) Extinction, colonisation and population structure: a study of a

milkweed beetle. American Naturalist, 134, 365-376.

McMurray, M.H. Jenkins, S.H. & Longland, W.S. (1997) Effects of seed density on

germination and establishment of a native and an introduced grass species

dispersed by granivorous rodents. American Midland Naturalist, 138, 322-330.

Middleton, B. (2000) Hydrochory, seed banks, and regeneration dynamics along the

landscape boundaries of a forested wetland. Plant Ecology, 146, 169-184.

Nathan, R. (2006) Long-distance dispersal of plants. Science, 313, 786-788.

Nathan, R. & Casagrandi, R. (2004) A simple mechanistic model of seed dispersal,

predation and plant establishment: Janzen-Connell and beyond. Journal of

Ecology, 92, 733-746.

Nathan, R. Katul, G.G. Horn, H.S. et al. (2002) Mechanisms of long-distance dispersal of

seed by wind. Nature, 418, 409-413.

Nathan, R. & Muller-Landau, H.C. (2000) Spatial patterns of seed dispersal, their

determinants and consequences for recruitment. Trends in Ecology & Evolution,

15, 278-285.

Nilsson, C. Gardfjell, M. & Grelsson, G. (1991) Importance of hydrochory in structuring

plant communities along rivers. Canadian Journal of Botany, 69, 2631-2633.

Orth, R.J. Fishman, J.R. Harwell, M.C. & Marion, S.R. (2003) Seed-density effects on

germination and initial seedling establishment in eelgrass Zostera marina in the

Chesapeake Bay region. Marine Ecology-Progress Series, 250, 71-79.

Ouborg, N. Piquot, Y. & Goenendael, J. (1999) Population genetics, molecular markers

and the study of dispersal in plants. Journal of Ecology, 87, 551-568.

Ozinga, W.A. Bekker, R.M. Schaminee, J.H.J. & Van Groenendael, J.M. (2004)

Dispersal potential in plant communities depends on environmental conditions.

Journal of Ecology, 92, 767-777.

Pannell, J.R. & Charlesworth, B. (1999) Neutral genetic diversity in a metapopulation

with recurrent local extinction and colonization. Evolution, 53, 664–676.

Pannell, J.R. & Charlesworth, B. (2000) Effects of metapopulation processes on

measures of genetic diversity. Philosophical Transactions of the Royal Society

London, Series B, 355, 1851–1864.

Ridley, H.N. (1930) The dispersal of plants throughout the world. Reeve, London.

Russo, S.E. & Augspurger, C.K. (2004) Aggregated seed dispersal by spider monkeys

limits recruitment to clumped patterns in Virola calophylla. Ecology Letters, 7,

1058-1067.

Santamaria, L. (2002) Why are most aquatic plants widely distributed? Dispersal, clonal

growth and small-scale heterogeneity in a stressful environment. Acta

Oecologica-International Journal of Ecology, 23, 137-154.

Santamaria, L. Figuerola, J. Pilon, J.J. Mjelde, M. Green, A.J. De Boer, T. King, R.A. &

Gornall, R.J. (2003) Plant performance across latitude: The role of plasticity and

local adaptation in an aquatic plant. Ecology, 84, 2454-2461.

Scarano, F.R. Pereira, T.S. & Rocas, G. (2003) Seed germination during floatation and

seedling growth of Carapa guianensis, a tree from flood-prone forests of the

Amazon. Plant Ecology, 168, 291-296.

Sculthorpe, C.D. (1967) The biology of aquatic vascular plants. Edward Arnold, London.

Shilton, L.A. Altringham, J.D. Compton, S.G. & Whittikar, R.J. (1999) Old world fruit bats

can be long-distance seed dispersers through extended retention of viable seeds

in the gut. Proceedings of the Royal Society, Series B: Biological Sciences, 266,

219-223.

Slatkin M. (1985) Gene flow in natural populations. Annual Review of Ecology and

Systematics, 16, 393-430.

Sork, V.L. Nason, J. Campbell, D.R. & Fernandez, J.F. (1999) Landscape approaches to

historical and contemporary gene flow in plants. Trends in Ecology & Evolution,

14, 219-224.

Tero, N. Aspi, J. Siikamaki, P. Jakalaniemi, A. & Tuomi, J. (2003) Genetic structure and

gene flow in a metapopulation of an endangered plant species, Silene tatarica.

Molecular Ecology, 12, 2073-2085.

Van der Valk, A.G. (1981) Succession in wetlands: a gleasonian approach. Ecology, 62,

688-696.

Vander Wall, S.B. & Longland, W.S. (2004) Diplochory are two dispersers better than

one? Trends in Ecology & Evolution, 19, 155-161.

Vivian-Smith, G. & Stiles, E.W. (1994) Dispersal of salt marsh seeds on the feet and

feathers of waterfowl. Wetlands, 14, 316-319.

Waser, N. M. Vickery, R.K. & Price, M. V. (1982) Patterns of seed dispersal and

population differentiation in Mimulus guttatus. Evolution, 36, 753-761.

Wells, C.L. & Pigliucci, M. (2000) Adaptive phenotypic plasticity: the case of hetrophylly

in aquatic plants. Plant Ecology, Evolution and Systematics, 3, 1-18.

Wright, S.A. (1931) Evolution in mendelian populations. Genetics, 28, 114-138.

Wright, S.A. (1951) The genetical structure of populations. Annals of Eugenics, 15, 323-

354.

Statement of Joint Authorship

Manuscript 1

Prentis, P. J., A. Vesey, N. M. Meyers, and P. B. Mather 2004. Genetic structuring of the stream lily Helmholtzia glaberrima (Philydraceae) within Toolona Creek, south-eastern

Queensland. Australian Journal of Botany 52: 201—207.

A re-print of the journal article is presented in Appendix 1.

Peter Prentis (Candidate)

• Wrote the manuscript and acted as corresponding author

• Designed and formulated sampling design and experimental protocols

• Undertook all field and laboratory work, analysis and interpretation of data

Amanda Vesey

• Undertook all field work and some lab work

• Provided editing and contributed to the structure of the manuscript

Noel Meyers

• Co-supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

Peter Mather

• Co-supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

CHAPTER 2. Genetic structuring of the stream lily Helmholtzia glaberrima

(Philydraceae) within Toolona Creek, south-east Queensland

Peter J. PrentisA,B, Vesey A.A, Meyers N.M.A and Mather P.B.A

A. School of Natural Resource Sciences, Queensland University of Technology, GPO

Box 2434, Brisbane, Qld 4001, Australia.

B. Corresponding author; email: [email protected]

Suggested running title: Population structure in Helmholtzia glaberrima.

Abstract. The distribution of genetic variation among five isolated sites of the riparian species Helmholtzia glaberrima (J.D. Hook) was examined in Toloona Creek (280 13’ S,

153007’E) using dominant AFLP markers. From the 137 fragments assessed, analysis of molecular variance (AMOVA) showed that most genetic variability occurred within sites

(68%), although high (32%) variation also occurred among sites. Highly significant pairwise θ estimates among all sampled sites suggests gene flow is restricted in H. glaberrima. Levels of within site diversity were intermediate and differed significantly across the sampled sites. Significant levels of linkage disequilibrium were detected at all sites except TC3. Differences in linkage disequilibrium and genetic diversity among the sites suggest that sites may be founded by different numbers of colonists. Mantel tests found no correlation between geographic and genetic distance and significant levels of linkage disequilibrium were detected at the total site level. This result supports a non- equilibrium model of population structure. The observed pattern of non-equilibrium population structure and genetic diversity in H. glaberrima are best explained by a classical metapopulation model.

Introduction

Tero et al. (2003) propose several models to explain how dispersal may influence population structure in linearly arranged riparian plant subpopulations. These models include: an island model, a completely fragmented model, a stepping stone model, a classical metapopulation model and a source/sink model. The island model predicts populations will exhibit genetic uniformity over all spatial scales, while the fragmented system model suggests all populations should show significant differentiation due to a lack of recurrent gene flow. In the stepping stone model, populations in close proximity should be more genetically similar than populations at greater geographical distance.

While, the classical metapopulation model recognises that patches are ephemeral and can be founded by differing numbers of individuals from single or multiple source patches. Consequently, genetic variation may decrease or increase randomly within patches while high levels of linkage disequilibrium may also be present. The source/sink model predicts that genetic diversity will be lower in the source populations relative to the founder patch.

Relatively few studies have examined how dispersal may influence the genetic structure of riparian angiosperm species. Those studies that address this question report results that do not always conform to the proposed models. For example, Hibiscus moscheutos did not conform to the models because the moderate differentiation found in a stream system was related to the distance from the stream channel and did not conform to the proposed models (Kudoh and Whigham 1997). Alternatively, Keller (2000) found extremely low levels of differentiation among sampled populations of Phragmites australis in the Charles River watershed attributed to gene flow among populations. An island model of population structure best explains the pattern of genetic structure found in P. australis. A recent study on Silene tatarica along the Oulankajoki River found high

levels of genetic differentiation among sites, which did not conform to an isolation by distance model (Tero et al. 2003). The authors suggest that the observed population genetic structure resulted primarily from “classical” metapopulation dynamics and that populations were not in drift/gene flow equilibrium.

In this paper we studied the genetic structure of a rare but locally abundant, riparian species H. glaberrima along a single stream to evaluate which models of population structure best fit this species. H. glaberrima is confined to high-elevation dendritic streams in the subtropical rainforest of the McPherson Ranges on the border of

Queensland and New South Wales. H. glaberrima is restricted to permanently waterlogged soil along creek banks, resulting in spatially structured patches that become increasingly more isolated with distance downstream. Patches of H. glaberrima comprise dense aggregations of up to 1.6 individuals/m2 (Prentis, Unpublished data). An added complication is that H. glaberrima is rhizomatous over short distances, so ramets may contribute to the expansion of patches after colonisation of new sites by propagules

(Harden 1993). Seed is presumably first dispersed via gravity and secondarily by water.

These factors may disperse seeds over very different spatial scales. Gravity dispersal should increase genetic structure among populations as dispersal by this mechanism is likely to occur over very small spatial scales. While hydrochoric dispersal may spread seed over much greater distances, seed movement will be unidirectional resulting in either increasing or decreasing genetic diversity downstream depending on how patches are colonised.

The study aimed to test the suitability of various population models (Tero et al. 2003) to explain the patterns of genetic variation among riparian populations of H. glaberrima because previous work indicates exceptions to these models (Kudoh and Whigham

1997). H. glaberrima was chosen as a model species since populations occur in discrete patches and are restricted to specific habitat along creek systems. We also sought to determine whether vegetative growth in H. glaberrima could influence patterns of genetic structure and which model best suits population structure in this species.

Materials and methods

Study species

The stream lily, H. glaberrima (Philydraceae), (2n =34), is a perennial herb, which is endemic to the understorey of subtropical, warm and cool-temperate rainforest in the

McPherson Ranges with a range of less than 100 km. Within this range the species exhibits a restricted distribution to high-elevation creek banks. H. glaberrima is a long- lived (at least 10 years) species that grows to 2.5 m and has low levels of protracted flowering throughout the year with most individuals flowering between August and

February. Flowers are produced on either single or multiple large, many-branched panicles which Adams (1989) believes are pollinated by small insects. Numerous 4-5 mm long seeds are produced in capsules throughout the infructescence after pollination.

No dispersal vectors are known to disperse the seed of this species, although Hamann

(1998) suggests hydrochory as a possible dispersal mechanism.

Study site and plant collections

Samples were collected from five populations of H. glaberrima located in the headwaters of Toolona Creek (Fig. 1) in Lamington National Park (280 13’ S, 153007’E). This single creek system allowed sampling of both high and low altitude sites, not possible in other adjacent stream systems. Sampled populations ranged in size from 118 to 651 adult and juvenile plants (Table 1). In all populations, sampling was based on a concentric circle design with distances varying from a known reference individual. The 20 m diameter of this circle was large enough to encompass all individual plants within sites. Samples were collected from adult plants at intervals of 1, 2, 4, 7, and 10 m in each quadrant of the circle yielding a total of 21 individuals from each population for genetic analysis.

DNA extraction and AFLP analysis

Genomic DNA was extracted from 1.2 g of finely ground plant leaf tissue according to the protocol of Doyle and Doyle (1988) with slight modifications.

AFLP procedures were performed according to the restriction ligation, PCR reactions and gel analysis protocols of Adjome-Mardsen et al. (1997) with slight modifications.

Modifications included using the Tru91 (Roche) restriction enzyme (isoschizomer for

Mse1) in place of Taq1 as the dominant cutter. The selective PCR was conducted using the selective nucleotides combinations: (1) E-AAG/M-AG, (2) E-AAG/M-GA and (3) E-

ATG/M-AG. The AFLP procedure was trialled twice on four samples initially, using the same primer sets and reproducible loci were detected for this species.

Data analysis

Statistical analyses using variation in AFLP phenotypes were based on the assumptions that AFLP markers are dominant diploid markers, which conform to Hardy-Weinberg equilibrium and where no co-migration of fragments occur. A Dice similarity matrix

(Sneath and Sokal 1973) was used to compare the similarity and clonality among individuals within sites using SPSS (2002). Mean genetic diversity estimates within populations were calculated three ways: (i) as the percentage of polymorphic loci (P%),

(ii) Shannon’s index of phenotypic diversity (IS) (Lewontin 1972) and (iii) Nei’s (1978) unbiased expected heterozygosity (HE) using POPGENE Version 1.32 (Yeh et al. 1997). To lend support to measures of genetic diversity found within sites average Dice similarity within sites was also calculated and subtracted from one. A Kruskal-Wallace test was used to examine whether levels of genetic diversity were consistent among sites. To investigate the influence that patch size and isolation from nearest neighbouring patch had on diversity indices, a Spearman rank correlation was used. A Monte Carlo

simulation process was used to examine linkage disequilibrium between AFLP phenotypes within and among sampling sites using Lian 3.1(Haubold and Hudson 2000).

The significance of departures from linkage equilibrium in these tests were generated over 5000 iterations. Linkage disequilibrium was examined to asses if populations had gone through bottlenecks or were formed by different levels of admixture.

Principal Coordinates Analysis (PCA) was used to visualise how individuals from sampling sites clustered using Genalex (Peakall and Smouse 2001). A UPGMA dendogram was constructed using Nei’s unbiased genetic distance (Nei 1978) to visualise how sampling sites clustered together with TFPGA (Miller 1997). To evaluate the extent of among population genetic differentiation, data were first analysed using analysis of molecular variance (AMOVA) in Arlequin 2.000 (Schneider et al. 2000).

Estimates of θ were obtained for each pair of sites using coancestry distance (Reynolds et al. 1983) in TFPGA. Using the general framework of Hutchinson and Templeton

(1999) we inferred the importance of genetic drift and migration among sampling sites by constructing scatter plots using pairwise θ comparisons against both in-stream and straight geographic distance. A Mantel test (Mantel 1967) was then used to evaluate whether genetic and in-stream or linear distance matrices were significantly correlated in

TFPGA using 1000 perturbations.

Results

AFLP polymorphism and patterns

Three primer pairs were used to screen 137 unambiguous AFLP loci. One hundred and sixteen of these fragments (92%) were polymorphic across the entire survey, with the primer pairs one, two and three producing 33 (80.5%), 41 (89.1%) and 42 (84%) fragments, respectively. Five individuals did not amplify yielding a total of 100 individuals across five sites for the analysis.

Within-site variation

All individuals screened showed unique AFLP banding profiles inferring that individuals possessed a unique genotype and were not clonal. Table 1 shows the mean average dice similarity, Nei’s expected heterozygosity, Shannon index of phenotypic diversity and percentage polymorphic loci at each site. Similarity values among individual plants ranged between 0.039 and 0.269, while percentage polymorphic loci ranged between

46.8% to 68%. Nei and Shannon indices varied from 0.16 to 0.236 and from 0.239 to

0.352 respectively. A Kruskall-Wallace test showed that Nei and Shannon indices were significantly different among sampled sites (P<0.05), but differences in diversity levels among sites were not correlated with either patch size or distance to neighbouring patches (P>0.2).

Significant levels of linkage disequilibrium were found at all sampling sites except for

TC3, where the result was marginally within the 95% Monte-Carlo confidence interval for linkage equilibrium. This indicates that sampled sites show different levels of linkage disequilibrium which is supported by the observed levels of linkage (VD) in Table 2.

Significant linkage disequilibrium was also detected at the total site level (Table 2).

Significant linkage disequilibrium suggests populations may have gone through bottlenecks or have been formed by different levels of admixture.

Among-site distances

PCA analysis showed relationships among populations, with the first two axes accounting for 28.2% of total variation. This analysis revealed minimal overlap among sampling sites as individuals within sites tended to cluster together, with sites 4 and 5 clustering independently from sites 1, 2 and 3 (Fig. 2). The UPGMA dendrogram also illustrated a similar pattern as sites 4 and 5 clustered together and sites 1, 2 and 3 clustered together (Fig. 3). AMOVA results indicate that although the major proportion of genetic variation was found within populations (68%), a significant proportion was also distributed among populations (32%). Pairwise θ estimates also confirmed that all sites were highly different (Table 3). Neither the scatter plots (Fig. 4) or Mantel tests revealed any significant relationships between genetic and in-stream or linear distance among populations within the stream system.

Discussion

Within-site variation

All Helmholtzia glaberrima individuals sampled possessed unique multilocus genotypes.

This result demonstrates that vegetative growth contributes little to within population structure or that the effect occurs over a spatial scale of less than one metre. An alternative explanation is that infrequent mis-scoring of phenotypes led to the false interpretation that no clones were present. This interpretation is unlikely however, as any difficult to interpret or ambiguous loci were excluded from the analysis to eliminate this problem and a minimum of 5 loci were different between any two individuals.

Helmholtzia glaberrima exhibited moderate levels of polymorphic AFLP loci, and diversity compared to other plants (Gaudeul et al. 2000; Despres et al. 2002, Tero et al.

2003). Two other herbaceous perennial species, Eryngium alpinum and Trollius europaeus (European globe flower) exhibited similar diversity levels to H. glaberrima, but unlike these species, no relationship occurred between genetic diversity and population size or degree of isolation (Gaudeul et al. 2000; Despres et al. 2002). Like H. glaberrima, other plant species with non-equilibrium dynamics often exhibit a poor correlation between genetic diversity and population size or isolation (Schmidt and

Jensen 2000; Tero et al. 2003).

Linkage disequilibrium tests showed that four of the five sampled populations of H. glaberrima were not in equilibrium. High levels of linkage disequilibrium within some, but not all sites, suggest that patches may result from different dynamics. Theory predicts that equilibrium dynamics can be disrupted by different levels of admixture and bottlenecks during patch formation (Tero et al. 2003). Also, patches founded by individuals arising from a greater number of source patches will exhibit greater levels of

genetic diversity relative to other patches (Whitlock and McCauley 1990). Differences in the level of population admixture or the severity of bottlenecks during patch formation are two potential causes for observed patterns of linkage disequilibrium and genetic diversity among sampling sites of H. glaberrima.

Among-site distance

Highly significant genetic differentiation occurred among sampled sites for H. glaberrima

(φst =0.32), in spite of the fact that most genetic variation was present within sites.

Therefore, sampled sites do not form a single panmictic unit. This result demonstrates that the species does not conform to an island model of population structure.

Similar patterns of variation over small spatial scales have been reported with dominant markers in other plant species (Hogbin and Peakall 1999; Gaudeul et al. 2000; Despres et al. 2002). Restricted dispersal among geographically isolated patches has often been used to explain this pattern of variation. An isolation by distance pattern of population structuring is likely to develop under low levels of localised gene flow (Hutchison and

Templeton 1999). However, H. glaberrima populations demonstrated strong differentiation, irrespective of their geographical proximity by linear or in-stream distance.

This suggests that most gene flow among H. glaberrima individuals occurs within patches. Moreover, it demonstrates that gene flow which occurs among patches is not related to patch proximity.

Both UPGMA and PCA results indicate that sites TC2 and TC3 and sites TC4 and TC5 respectively cluster more strongly than other sampled sites. The two clusters are also the most divergent from each other, even though sites TC3 and TC4 are closer to each other than is site TC2. This relationship indicates that one and two dimensional stepping

stone models of dispersal cannot explain H. glaberrima’s population structure.

Furthermore, scatter plots of coancestry distance versus geographic distance (Fig. 4) show a random scattering of points, indicating that recurrent gene flow has less influence than drift or non-equilibrium dynamics on the observed pattern of genetic structure. Two likely explanations may account for this pattern of fine-scale structure: (1) genetic drift predominates because dispersal is low or absent among sites and/or (2) different numbers of individuals arising from different numbers of source populations contribute to founding patches.

Short-distance seed dispersal may result in the strong genetic differentiation among sampled sites. Seed dispersal among patches may be rare because H. glaberrima seed lacks specialised structures for long distance dispersal, except potentially by water. In this situation, genetic drift may cause the random fixation or loss of alleles within geographically isolated patches if effective population sizes are low. Consequently, genetic differentiation should increase among patches as private alleles evolve or alleles are lost within patches. No private alleles were detected, however, among the sampled sites. In addition the highest frequency of null alleles in any patch was 5%. Thus genetic drift within sites would seem to be quite weak and the observed population structure cannot be adequately explained by a fragmented model. Non-equilibrium dynamics associated with patch formation or dispersal represent more likely causes for the observed population structuring.

Chance colonisation events of the limited suitable habitat occurring along stream banks represent a likely explanation for potential differences in population founding numbers in

H. glaberrima, causing populations to be out of equilibrium. Because of the low probability that seeds with similar genotypes will found a patch, when different H.

glaberrima sites are compared highly divergent gene frequencies among sites may occur. Also the occurrence of linkage disequilibrium at the total population level indicates that H. glaberrima patches have been colonised by non-equilibrium dynamics. Highly divergent gene frequencies and high levels of linkage disequilibrium are often associated with typical non-equilibrium dynamics of both source/sink and classical metapopulations

(Whitlock and McCauley 1990; Boileau et al. 1992). However as genetic diversity did not decrease downstream as the source/sink model predicted but varied randomly in the stream system it indicates that classical metapopulation models can best explain the observed pattern of population structuring in H. glaberrima.

In conclusion, the present study has shown that patches of H. glaberrima display intermediate levels of genetic diversity even in small patches and that vegetative growth does not influence patch structure. The random pattern of genetic variation among patches and high levels of linkage disequilibrium within some sites indicate populations do not conform to an equilibrium population structure model. Moreover, classical metapopulation models best explain patterns of genetic variation and differentiation.

Future work should investigate whether H. glaberrima population structure follows metapopulation or non-equilibrium models in other stream systems. This information will assist our understanding of the population dynamics in riparian plant species.

Acknowledgments

The authors thank Grant Hamilton and Shaun Meredith for assistance with collections from field sites and Vincent Chand and Natalie Baker for their help with lab work. We are also grateful to Ana Pavasovic and Dr Grahame Kelly and two anonymous reviewers who read earlier drafts of this manuscript.

References

Adams LG (1989) Philydraceae. In ‘Flora of Australia v45 Hydatellaceae to Liliaceae’.

(Ed. AS George) pp. 40-46. (AGPS: Canberra)

Ajmone-Marson P, Valenti A, Casando M, Vecchiotti-Antaldi G, Bertoni, G, Kuiper M

(1997) AFLP markers for DNA fingerprinting in cattle. Animal Genetics 28, 418-426.

Boileau MG, Hebert PDN, Schwartz SS (1992) Non-equilibrium gene frequency divergence persistent founder effects in natural populations. Journal of Evolutionary

Biology 5, 25-40.

Despres L, Loriot S, Gaudeul M (2002) Geographic pattern of genetic variation in the

European globeflower Trollius europaeus L. (Ranunculaceae) inferred from amplified fragment length polymorphism markers. Molecular Ecology 11, 2337-2347.

Doyle JJ, Doyle JL (1988) Isolation of plant DNA from fresh tissue. Focus 12, 13-15.

Gaudeul M, Taberlat P, Till-Bottraud I (2000) Genetic diversity in an endangered alpine plant, Eryngium alpinum L. (Apiaceae), inferred from amplified fragment length polymorphism markers. Molecular Ecology 9, 1625-1637.

Hamann U (1998) Philydraceae. In ‘The Families and Genera of Vascular Plants. IV.

Flowering Plants. Monocotyledons. Alismatanae and Commelinanae (except

Gramineae)’. (Ed. K. Kubitzki) pp389-394, (Springer: Berlin)

Harden GJ (1993) ‘Flora of New South Wales’ (New South Wales University Press:

Sydney)

Haubold B, Hudson RR (2000) Lian 3.0: detecting linkage disequilibrium in multilocus data. Bioinformatics 16, 847-848.

Hogbin PM, Peakall R (1999) Evaluation of the contribution of genetic research to the management of the endangered plant Ziera prostrata. Conservation Biology 13, 514-

522.

Hutchison DW, Templeton AR (1999) Correlation of pairwise genetic and geographic distance measures: inferring the relative influences of gene flow and drift on the distribution of genetic variability. Evolution 53, 1898-1914.

Keller BE (2000) Genetic variation among and within populations of Phragmites australis in the Charles River watershed. Aquatic Botany 66, 195-208.

Kudoh H, Whigham DS (1997) Microgeographic genetic structure and gene flow in

Hibiscus moscheutos (Malvaceae) populations. American Journal of Botany 84, 1285-

1293.

Lewontin RC (1972) The apportionment of human diversity. Evolutionary Biology 6, 381-

394.

Mantel N (1967) The detection of disease clustering and a generalised regression approach. Cancer Research 27, 209-220.

Miller MP (1997) Tools for population genetic analysis (TFPGA). Windows program for the analysis of allozyme and molecular population data. Department of Biological

Sciences, Northern Arizona University, Arizona.

Nei M (1978) ‘Molecular evolutionary genetics.’ (Columbia University Press: New York)

Peakall R, Smouse PE (2001) Genalex V5: Genetic analysis in excel. Population genetic software for teaching and research. Australian National University, Canberra.

Reynolds J, Weir BS, Cockerman CC (1983) Estimation of the coancestry coefficient: basis for short term genetic distance. Genetics 105, 767-779.

Schmidt K, Jensen K (2000) Genetic structure and AFLP variation of remnant populations in the rare plant Pedicularis palustris and its relation to population size and reproductive components. American Journal of Botany 87, 678-689.

Schneider S, Kuffer JM, Rossli D, Excoffier L (2000) Arlequin, version 2.0: a software for population genetic data analysis. Genetics and Biometry Laboratory, University of

Geneva, Geneva.

Sneath PHA, Sokal RR (1973) ‘The principles and practice of numerical classification’

(W. H. Freeman: San Francisco)

Spss (2002) SPSS for windows, release 11.5.0. (SPSS Inc.: Chicago)

Tero N, Aspi J, Siikamaki P, Jakalaniemi A, Toumi J (2003) Genetic structure and gene flow in a metapopulation of an endangered species, Silene tatarica. Molecular Ecology

12, 2073-2085.

Whitlock MC, McCauley DE (1990) Some population genetic consequences of colony formation and extinction: genetic correlations with founding groups. Evolution 44, 1717-

1724.

Yeh F, Yang RC, Boyle T (1997) POPGENE, the user-friendly shareware for population genetic analysis. Molecular Biology and Biotechnology Centre, University of Alberta,

Edmonton.

Table 1. Population size, isolation of populations from nearest neighbouring population and within population diversity indices from the sampled H. glaberrima populations

Populatio Isolation Populatio Nei’s Shannon Variable Average n (m) n size Index (HE) Index (IS) Loci (%) Similarity

TC1 250 234 0.236 0.352 68 0.174

TC2 68 118 0.177 0.268 54.5 0.087

TC3 40 618 0.159 0.239 46.8 0.105

TC4 32 307 0.195 0.289 54.2 0.131

TC5 21 213 0.178 0.266 49 0.109

Table 2. Analysis of linkage disequilibrium within each of the sampled sites and at the total population level in H. glaberrima

Levels of significance were derived by 5000 iterations using a Monte-carlo simulation where * < 0.05, ** < 0.025, *** < 0.001

Sites VD VE LMC Significance

TC1 39.3 17.1 21.9 ***

TC2 84 12.5 16.4 ***

TC3 15.6 12.2 15.7

TC4 33.9 14.6 19 ***

TC5 20.6 13.5 17.2 **

Total 60.5 19.4 21.3 **

Table 3. Pairwise θ distance values among sampled populations of H. glaberrima

Population TC1 TC2 TC3 TC4 TC5

TC1 0

TC2 0.358 0

TC3 0.312 0.31 0

TC4 0.33 0.381 0.423 0

TC5 0.447 0.539 0.542 0.374 0

Mt Wanungara Mt Bithongabal N Mt Toolona Lamington National Park

TC5 Toolona Creek TC4

TC3

Toolona Gorge

TC2

0km 0.5km 1km Lamington National Park TC1

Fig. 1. Map of sampling locations within Toloona Creek

0 site 1 -20-100 102030site 2 -10 site 3

axis2 site 4 -20 site 5

-30

-40 axis 1

Fig. 2. Principal coordinate analysis illustrating genetic differentiation of the H. glaberrima individuals sampled across the 5 sites. Axis 1 extracted 17.88% of the variance and axis 2 extracted 10.36% of the variance.

Fig. 3. UPGMA dendrogram illustrating sites 1, 2 and 3 clustering with each other and away from the cluster of sites 4 and 5.

0.6

0.5

0.4

0.3

0.2

Coancestry distance 0.1

0 0 500 1000 1500 2000 2500 3000

0.6 In-stream distance (m)

0.5

0.4

0.3

0.2

0.1

0 0 500 1000 1500 2000

Linear distance (m)

Fig. 4. Scatter plots of pairwise θ distance versus in-stream and linear distance

Statement of Joint Authorship

Manuscript 2

Prentis, P. J., N. M. Meyers, and P. B.Mather. In press. The significance of post- germination buoyancy in Helmholtzia glaberrima and Philydrum lanuginosum

(Philydraceae). Australian Journal of Botany.

Peter Prentis (Candidate)

• Wrote the manuscript and acted as corresponding author

• Designed and formulated sampling design and experimental protocols

• Undertook all field and laboratory work, analysis and interpretation of data

Noel Meyers

• Supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

Peter Mather

• Co-supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

CHAPTER 3. The significance of post-germination buoyancy in Helmholtzia glaberrima and Philydrum lanuginosum (Philydraceae)

Peter J. PrentisA, B, Noel M. MeyersA and Peter B. MatherA

A School of Natural Resource Sciences, Queensland University of Technology, GPO Box

2434, Brisbane, Qld 4001, Australia.

B Corresponding author: E-mail: [email protected]

Running title: Seedling floatation promotes establishment

Statement of Joint Authorship

Manuscript 3

Prentis, P. J., N. M. Meyers, and P. B.Mather. (In Prep) Seed dispersal and seedling establishment in the riparian plant Helmholtzia glaberrima. Freshwater Biology.

Peter Prentis (Candidate)

• Wrote the manuscript and acted as corresponding author

• Designed and formulated sampling design and experimental protocols

• Undertook all field and laboratory work, analysis and interpretation of data

Noel Meyers

• Supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

Peter Mather

• Co-supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

CHAPTER 4. Seed dispersal and seedling establishment in the riparian plant

Helmholtzia glaberrima

Peter J. Prentis1, Noel M. Meyers and Peter B. Mather

School of Natural Resource Sciences, Queensland University of Technology, GPO Box 2434,

Brisbane, Qld, 4001 Australia.

1 Corresponding author:

Peter Prentis

School of Natural Resource Sciences,

Queensland University of Technology,

GPO Box 2434,

Brisbane, Qld, 4001 Australia. [email protected]

Fax: +617 3864 1535

Phone: +617 3864 2186

Running title: Seed dispersal in H. glaberrima

Statement of Joint Authorship

Manuscript 4

Prentis, P. J., N. M. Meyers, and P. B. Mather. (In Review) Micro-geographic landscape features demarcate seedling genetic structure in the stream lily, Helmholtzia glaberrima

(Philydraceae)1. American Journal of Botany.

Peter Prentis (Candidate)

• Wrote the manuscript and acted as corresponding author

• Designed and formulated sampling design and experimental protocols

• Undertook all field and laboratory work, analysis and interpretation of data

Noel Meyers

• Supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

Peter Mather

• Co-supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

105

CHAPTER 5. Micro-geographic landscape features demarcate seedling genetic structure in the stream lily, Helmholtzia glaberrima (Philydraceae)

Peter J. Prentis,1,2,* and Peter B. Mather1

1School of Natural Resource Sciences, Queensland University of Technology, GPO Box

2434, Brisbane, Qld 4001, Australia.

2School of Earth and Environmental Sciences, University of Adelaide, North Terrace, SA

5005, Australia.

* Corresponding author: Peter Prentis

Email: [email protected]

Phone: +618 8303 5594

Fax: +618 8303 4364

Postal address: School of Earth and Environmental Sciences, University of Adelaide,

North Terrace, SA 5005, Australia

106

Statement of Joint Authorship

Manuscript 5

Prentis, P. J., N. M. Meyers, and P. B.Mather. (In Review) Fine-scale patterns of genetic diversity and population structure in the stream lily Helmholtzia glaberrima

(Philydraceae) along rainforest streams, south-east Queensland. Freshwater Biology.

Peter Prentis (Candidate)

• Wrote the manuscript and acted as corresponding author

• Designed and formulated sampling design and experimental protocols

• Undertook all field and laboratory work, analysis and interpretation of data

Noel Meyers

• Supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

Peter Mather

• Co-supervised the sampling design and experimental protocols

• Assisted in the interpretation of data

• Provided editing and contributed to the structure of the manuscript

126

CHAPTER 6. Fine-scale patterns of genetic diversity and population structure

in the stream lily Helmholtzia glaberrima (Philydraceae) along rainforest

streams, south-east Queensland

Peter. J. Prentis, Peter. B. Mather, and Noel. M, Meyers

School of Natural Resource Sciences, Queensland University of Technology, GPO

Box 2434, Brisbane, Qld 4001, Australia.

Corresponding author e-mail: [email protected]

Running head: Population structure in a riparian plant

Keywords : AFLP, genetic diversity, landscape, population structure, riparian plant

127

CHAPTER 7. GENERAL DISCUSSION

A number of recent papers have debated the role of genetic and ecological approaches to understanding seed dispersal in plant species (Cain et al., 2000;

Nathan, 2005; Nathan & Muller-Landau, 2000; Nathan et al., 2003). The authors of these papers have suggested that genetic and ecological approaches usually measure different components of seed dispersal. Ecological approaches are suggested to measure actual or potential seed dispersal, while genetic approaches usually measure realised seed dispersal or the phase after seedling establishment

(Cain et al., 2000). Most studies however, only use one approach to examine seed dispersal, thereby overlooking either potential or realised components of seed dispersal (Nathan, 2005). Information on both potential and realised dispersal is needed to understand the dynamics of persistence, spread and distribution of plant populations. For instance seed may often disperse to among sites over very long distances, but never establish and contribute to realised dispersal among sites.

Ecological data alone in this case would suggest that there is high levels of seed mediated gene flow among sites, while genetic data alone would suggest there is no seed dispersal or gene flow occurring among the sites. In combination an ecological and genetic approach would determine that although a large amount of seeds are dispersed among the sites, there is no recruitment of these seeds into the breeding population and therefore no gene flow. Therefore by integrating genetic and ecological approaches to studying seed dispersal we can determine if seeds are actually dispersing among populations and if they are then incorporated into the breeding population at new sites. To this end, the present study investigated seed dispersal in H. glaberrima using both ecological and genetic approaches.

Ecological approaches indicated primary seed dispersal in H. glaberrima apparently occurs over short physical distances, which has also been reported in many other herbaceous species (Cain et al. 1998; Ellison & Parker, 2002). Unlike

147

some herbaceous wetland species, however, the results of the current study support the contention that water could potentially disperse seeds of riparian or wetland specialists over both local and long-distances (Ellison & Parker, 2002; Waser et al.

1982). After prolonged periods of immersion in water, nearly all H. glaberrima seeds tested still floated and over half of these seeds were still viable conferring the potential for successful long-distance water-borne dispersal. The duration of seed buoyancy observed here is similar to those observed in other riparian plant species

(Boedeltje et al., 2003; Ridley, 1930).

The capacity of H. glaberrima seeds to refloat after sinking and the prolonged seedling buoyancy observed may also represent an adaptation to local or long- distance dispersal via water. Seedlings of some other riparian plants are positively buoyant (Ridley 1930; Sculthorpe 1967; Van der Valk 1981), which may promote the probability of seed dispersal or the early anchoring and subsequent establishment of seedlings (Sculthorpe 1967; Nicol and Ganf 2000). Seedling dispersal would be likely to occur over a local scale within populations, as long-distance transport in the fast flowing rocky streams of the McPherson Ranges could damage young leaf tissue.

Nicol and Ganf, (2000) also suggested that seedling dispersal may occur over a local scale, but only if seedlings dispersed to the permanently moist edges of wetland habitats.

Seed germination and seedling establishment was strongly reduced by lowered soil moisture in H. glaberrima, with seedling establishment only occurring under permanently wet conditions. Hydrological conditions have long been known to influence seed germination and seedling establishment in aquatic or riparian plant species (Boedeltje et al., 2003; Keddy and Ellis, 1985; Keddy and Constable, 1986;

Nicol and Ganf 2000). Furthermore, hydrology plays an important role in zonation and structuring the composition of aquatic vegetation as water levels rise and fall and soil moisture levels change (Blanch et al., 1999). Therefore, dispersal of seeds in aquatic or riparian species to areas with unfavourable hydrological conditions is

148

unlikely to result in successful seedling recruitment and establishment.

Consequently, the effectiveness of hydrochory is likely to be limited if seeds or seedlings of H. glaberrima are not dispersed to permanently wet habitat, as wet/dry soil conditions inhibit seedling establishment.

Together these ecological data provide evidence that dispersal by water can move seeds successfully over long distances along streams between isolated wetland habitats if seeds are dispersed to microhabitat conditions favourable for seedling establishment. Long-distance seed dispersal along riparian corridors has been suggested to be an effective method for seed dispersal among streamside populations in other obligate riparian species (Kudoh & Whigham 1997; Kitamoto et al. 2005). In fact seeds of some riparian plant species have been dispersed distances of at least six kilometres in flowing waters along streams (Boedeltje et al., 2003). In many plant species however, potential long-distance seed dispersal is not realised as seedling establishment usually does not result. Therefore, genetic data are required to determine if long-distance seed dispersal has been successful.

Genetic studies showed that classical metapopulation models of dispersal best explain the pattern of genetic structure observed here in H. glaberrima. Mixed patterns of local and long-distance dispersal are reported to be common in plant metapopulations (Jakalaniemi et al., in press). In many cases, most dispersal occurs on a local scale, but very rare long-distance colonisation or dispersal events do occur. The patterns of genetic diversity observed in H. glaberrima suggest that nearly all successful dispersal occurs at a local scale, but that extremely rare long-distance dispersal events may be responsible for establishment of new patches. Similar patterns of genetic structure have also been found in the riparian species, Silene tatarica, which were suggested to arise from mixed patterns of seed dispersal (Tero et al., 2003). Differences in the amount of local versus long-distance dispersal may be influenced by the effectiveness of a vector for short or long-distance dispersal or the presence of more than dispersal vector (Ozinga et al., 2003).

149

On a local scale water-born seed dispersal can strongly influence the pattern of genetic structure of seedlings within a population, however it had little affect on the spatial pattern of genetic variation among established populations of H. glaberrima.

Moreover, results presented here do not support the hypothesis that flowing water is an effective long-distance seed dispersal vector in H. glaberrima. Although ecological experiments demonstrated seeds of H. glaberrima have many obvious adaptations for hydrochory including prolonged buoyancy and extended viability in water, water- borne seed dispersal does not provide effective gene flow among established populations of H. glaberrima. Recently it has been suggested that dispersal attributes related to seed characteristics provide a very useful framework for describing local dispersal processes, but are poorly related to the rarer processes that may move seeds over long distances (Higgins et al., 2003). The authors refer to the dispersal vectors responsible for rare long-distance dispersal as non-standard means of dispersal. It is possible then that long-distance dispersal events responsible for population founding in H. glaberrima may occur via non standard means.

The pattern of genetic differentiation among stream systems in H. glaberrima, suggest that rare dispersal events by nonstandard vectors may be responsible for the colonisation of new patches. The greatest level of genetic variation was found at the smallest spatial scale, among populations within a single stream system, while only a small percentage of variation was found among stream systems. If water was responsible for colonisation of new patches rather than a non-standard vector, population genetic structure would likely conform to a hierarchical model of gene flow. As this was not the case, other seed dispersal vectors therefore, must play an important role in shaping the genetic structure of H. glaberrima in this region and may prove to be a significant influence on the spatial pattern of genetic variation in this species as a whole.

More generally, this study has shed new light on the role of hydrochory and seed dispersal in riparian and wetland habitats. Although the potential for long-

150

distance hydrochory exists in H. glaberrima, it would seem to only occur at a local scale. This contrasts with previous findings which show that hydrochory is often an effective long-distance dispersal vector in complex, multistage dispersal processes

(Fridriksson, 1975; Hampe, 2004). These studies however, have only examined potential seed dispersal and have no genetic data to determine if the plants dispersing over long-distances have contributed to the gene pool of new populations.

This highlights the importance for incorporating genetic components into studies on seed dispersal as without this information we gain an insight into potential seed dispersal. Without a genetic component to the current study, it would be assumed that hydrochory should maintain gene flow among streamside populations of H. glaberrima. This is not the case however, as gene flow among established populations would seem to extremely rare and unrelated to hydrochory.

In H. glaberrima, a non-standard vector would appear to be the most likely long-distance seed dispersal vector. This highlights the lack of generality that can be made about seed dispersal in many riparian or wetland systems except to say that dispersal in these systems is very complex and often involve a number of different seed dispersal vectors. Currently no mechanistic models of seed dispersal can explain such complex seed dispersal patterns as those found in many riparian systems. These models need to be developed in order to understand how seed dispersal influences spatio-temporal patterns of seedling and population recruitment in riparian and wetland plants. This information is critically required to gain knowledge into the persistence of populations and dynamics of population extinction and recolonisation in riparian and wetland plants. Without an understanding of how riparian or wetland plants colonise extirpated populations or new patches we cannot adequately develop management strategies to promote the persistence of these species in urbanised landscapes. Therefore, H. glaberrima and other riparian species present a unique opportunity to investigate the processes of population extinction

151

and recolonisation and long-distance seed dispersal by non-standard dispersal vectors extinction/recolonisation dynamics in plants.

152

References

Blanch, S.J. Ganf, G.G. & Walker, K.F. (1999) Tolerance of riverine plants to flooding

and exposure by water regime. Regulated Rivers: Research and

management, 15, 43-62.

Boedeltje, G. Bakker, J.P. Bekker, R.M. Van Groenendael, J.M. & Soesbergen, M.

(2003) Plant dispersal in a lowland stream in relation to occurrence and three

specific life-history traits of the species in the species pool. Journal of

Ecology, 91, 855-866.

Cain, M. L. Damman, H. & Muir, A. (1998) Seed dispersal and the Holocene

migration of woodland herbs. Ecological Monographs, 68, 325-347.

Ellison, A.M. & Parker, J.N. (2002) Seed dispersal and seedling establishment of

Sarracenia purpurea (Sarraceniaceae). American Journal of Botany, 89,

1024-1026.

Fridriksson, S. (1975) Surtsey: evolution of life on a volcanic island. Butterworths,

London.

Hampe, A. (2004) Extensive hydrochory uncouples spatiotemporal patterns of

seedfall and seedling recruitment in a 'bird-dispersed' riparian tree. Journal of

Ecology, 92, 797-807.

Higgins, S.I. Nathan, R. & Cain,M.L. (2003) Are long-distance dispersal events in

plants usually caused by nonstandard means of dispersal? Ecology, 84,

1945-1956.

Jakalaniemi, A. Tuomi, J. Siikimaki, P. & Kilpia, A. (In Press) Colonization-extinction

and patch dynamics of the perennial riparian plant, Silene tatarica.

Journal of Ecology.

Keddy, P.A. & Constable, P. (1986) germination of ten shoreline plants in relation to

seed size, soil and particle size and water level: an experimental study.

Journal of Ecology, 74, 133-141.

153

Keddy, P.A. & Ellis, T.H. (1985) Seedling recruitment of 11 wetland plant species

along a water level gradient: shared or distinct responses? Canadian Journal

of Botany, 63, 1876-1879.

Kitamoto, N. Honjo, M. Ueno, S. et al. (2005) Spatial genetic structure within and

among populations of Primula sieboldii growing beside separate streams.

Molecular Ecology, 14, 149-157.

Kudoh, H. & Whigham, D.F. (1997) Microgeographic genetic structure and gene flow

in Hibiscus moscheutos (Malvaceae) populations. American Journal of

Botany, 84, 1285-1293.

Nathan, R. (2005) Long-distance dispersal research: building a network of yellow

brick roads. Diversity and Distributions, 11, 125-130.

Nathan, R. & Muller-Landau, H.C. (2000) Spatial patterns of seed dispersal, their

determinants and consequences for recruitment. Trends in Ecology &

Evolution, 15, 278-285.

Nathan, R. Perry, G. Cronin, J.T. Strand, A.E. & Cain, M. L. (2003) Methods for

estimating long-distance dispersal. Oikos, 103, 261-273.

Nicol, J.M. & Ganf, G.G. (2000) Water regimes, seedling recruitment and

establishment in three wetland plant species. Marine & Freshwater Research,

51, 305-309.

Ozinga, W.A. Bekker, R.M. Schaminee, J.H.J. & Van Groenendael, J.M. (2004)

Dispersal potential in plant communities depends on environmental

conditions. Journal of Ecology, 92, 767-777.

Ridley, H.N. (1930) The dispersal of plants throughout the world. Reeve, London.

Sculthorpe, C.D. (1967) The biology of aquatic vascular plants. Edward Arnold,

London.

Tero, N. Aspi, J. Siikamaki, P. Jakalaniemi, A. & Tuomi, J. (2003) Genetic structure

and gene flow in a metapopulation of an endangered plant species, Silene

tatarica. Molecular Ecology, 12, 2073-2085.

154

Van der Valk, A.G. (1981) Succession in wetlands: a gleasonian approach. Ecology,

62, 688-696.

Waser, N. M. Vickery, R.K. & Price, M. V. (1982) Patterns of seed dispersal and

population differentiation in Mimulus guttatus. Evolution, 36, 753-761.

155