HISTAMINE – a NATURALLY OCCURRING SETTLEMENT CUE for LARVAE of the AUSTRALIAN SEA URCHIN Holopneustes Purpurascens

HISTAMINE – A NATURALLY OCCURRING SETTLEMENT CUE FOR LARVAE OF THE AUSTRALIAN SEA URCHIN Holopneustes purpurascens

Rebecca L. Swanson

A thesis submitted to the University of New South Wales for the degree of Doctor of Philosophy

2006

ORIGINALITY STATEMENT

‘I hereby declare that this submission is my own and to the best of my knowledge contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement has been made to the research of others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in this thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the projects design and conception or in style, presentation and linguistic is acknowledged’

Signed……………………………………

Date……………………………………...

‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signed ……………………………………………......

Date ……………………………………………......

AUTHENTICITY STATEMENT

‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’

Signed ……………………………………………......

Date ……………………………………………......

THE UNIVERSITY OF NEW SOUTH WALES

Thesis/Dissertation Sheet

Surname or Family name: Swanson

First name: Rebecca Other name/s: Lyn

Abbreviation for degree as given in the University calendar: PhD

School: School of Biological, Earth and Environmental Science Faculty: Science

Title: A naturally occurring settlement cue for larvae of the Australian sea urchin Holopneustes purpurascens

Abstract 350 words maximum: (PLEASE TYPE) The importance of chemical cues in triggering the settlement of marine invertebrate larvae has long been recognised but very few such cues have been definitively identified. Larvae of the Australian sea urchin Holopneustes purpurascens, which lives enmeshed in the fronds of macroalgae, are induced to settle by a water-soluble cue produced by the host alga Delisea pulchra. This cue was previously identified as a floridoside-isethionic acid complex. I present evidence in this thesis which supports histamine as the true settlement cue for larval H. purpurascens. The settlement cue was isolated from the polar extract of D. pulchra by bioassay-guided cation-exchange chromatography and identified as histamine using nuclear magnetic resonance spectroscopy. Algal derived and synthetic histamine at ~5 μM induced rapid settlement in 80–100 % of larval H. purpurascens. In the first study of its kind for any marine invertebrate, variation in the distribution of new recruits was compared with quantitative variation in the distribution of histamine in the habitat. More than 90 % of new recruits were found on either the foliose red alga D. pulchra or on coralline turf algae. These algae induced >90 % settlement of larvae in laboratory assays after 24 h. D. pulchra contained far higher levels of histamine than all other algae, however, the coralline algae lacked measurable histamine. Seawater collected in situ adjacent to D. pulchra induced up to 16 % settlement of older larvae and contained the highest concentration of histamine (~5 nM). With the exception of coralline algae, variation in settlement and recruitment was consistent with the variation among species histamine contents. Initial results supported a biofilm derived settlement cue from coralline algae. I also showed that older H. purpurascens larvae settle in response to lower concentrations of histamine than younger larvae and required less exposure to histamine (10 μM) in order to initiate irreversible metamorphosis. Histamine induced settlement of two other echinoids with non-feeding larvae. Histamine did not induce settlement of feeding larvae of two echinoids or settlement of non-feeding larvae of asteroids. Histamine may be a general settlement cue for echinoids with direct development.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

…………………………………………… …………………………………….. ……….……………………...… Signature Witness Date

FOR OFFICE USE ONLY Date of completion of requirements for Award:

Acknowledgements

I thank my supervisor Professor Peter Steinberg for his wise supervision over the years and for giving me the opportunity to work with such a wonderful group of people. I thank my Co-Supervisor Professor Rocky de Nys for his enthusiasm for science and life in general which is inspiring, and for his encouraging words particularly in the early years of this project (when confusion and doubt were supreme) and in the final weeks of writing this thesis. I would like to give special thanks to Jane Williamson for laying the groundwork for this project which I have thoroughly enjoyed researching. You have each helped me to discover a true interest in marine chemical ecology, in particular, in marine invertebrate larvae which are truly amazing creatures (not to mention very, very cute)! I should also thank my Yr 11 Biology teacher, Justine Waters, who first inspired my love of Biology.

I have been very lucky to share my research experience with many fellow students and research staff in the School of Biological (Earth and Environmental) Science and the Centre for Marine Biofouling and Bio-Innovation. I sincerely hope that the friendships made during this time will last a lifetime. Fellow students, especially Megan Huggett, Sharon Longford, Nicholas Paul, Niina Tujula, Mike Taylor, Tim Charlton, Jacinta Green, Keyne Monro, Adriana Verges and Alex Campbell, have made my PhD experience all the more special. Research staff past and present; Odette Ison, Sophia McCloy, Neda Shakibaee, Louise McKenzie, Kirsty Collard, Peter Schupp, Dustin Marshall, Paul Gribben and Lachlan Yee, have been great fun to work with. I apologise for being bossy and grumpy at times, and for writing all those notes, but someone had to try and organise you lot!!

Special thanks go to Tim Charlton for always happily providing advice and direction on GC-MS (the bane of this project!). Special thanks to Professor Maria Byrne, Paula Cisternas and Tom Prowse of the Evolutionary Development Laboratory at the University of Sydney for providing other larval species for histamine tests and for sharing your facilities. I thank Maria for providing helpful comments on Chapter Five and for her enthusiasm for larval biology which is inspiring. Big thanks to Jacinta for

helping me format and ‘master’ this thesis. This was stressful enough with your expert help. I can imagine how awful the last few days would have been without you!

My family have always provided great support for my academic endeavours, both emotional and financial, from the senior years of high school through to this everlasting PhD. Special thanks to my beautiful mum, Marlene, for many years of love, support and encouragement, for minding Nicholas, and for all those delicious home cooked meals which have sustained me, even many years after moving out of home!! Thanks to my dad John, and my sisters Jacqui and Karen, for their continued love and support. Thanks Aunty ‘Goo’ for your support, especially for your cooking in the last few months. Special thanks to my husband Neil for the emotional and moral support over the years. Neil has lived my PhD experience and endured the lows of such a protracted affair (I hope I am not so grumpy now this is over)! I also thank Neil’s family who have all been very interested in my work. Nicholas, my gorgeous little boy, has brought me great love and happiness. Although the arrival of Nicholas put the PhD on hold, he also put my life into perspective which helped me to finish writing this thesis in the right frame of mind. To Baby Swanson on the way, now I can think about you!!

“There is one more thing…………….it’s been emotional.”

iii

TABLE OF CONTENTS

Abstract i Acknowledgements ii Table of Contents iii List of Figures vii List of Tables ix

CHAPTER ONE

General Introduction 1

1.1 The study organism - Holopneustes purpurascens 8 1.2 Aims and structure of this thesis 9 1.3 A note on definitions 111

CHAPTER TWO

Isolation and characterisation of a settlement cue for Holopneustes purpurascens 12

2.1 Introduction 12 2.2 Materials & Methods 14 2.2.1 Study site 14 2.2.2 Preparation of the polar extract of Delisea pulchra 14 2.2.3 Isolation of the settlement cue in Delisea pulchra 15 2.2.4 Identification of settlement cue 16 2.2.5 Larval culture 18 2.2.6 Settlement assays 19 2.2.7 Quantitative analysis of histamine in various algae 20 2.2.8 The source of the settlement cue from Delisea pulchra 22 2.2.9 Reanalysis of samples from Williamson et al. (2000) 23 2.3 Results 23 2.3.1 Isolation of the settlement cue in Delisea pulchra 23 2.3.2 Identification of settlement cue 24 2.3.3 The response of larvae to natural and synthetic histamine 26 2.3.4 Quantitative analysis of histamine content in algae 26 2.3.5 The source of the settlement cue from Delisea pulchra 27 2.3.6 Reanalysis of samples from Williamson et al. (2000) 27 2.4 Discussion Error! Bookmark not defined. 40

CHAPTER THREE

In situ quantification of histamine in the habitat and recruitment of Holopneustes purpurascens 43

3.1 Introduction 43 3.2 Materials and Methods 44 3.2.1 Study site 44 iv

3.2.2 Recruitment Survey 44 3.2.3 Larval culture 45 3.2.4 Settlement assays testing host algae 46 3.2.5 Settlement assays testing in situ seawater 47 3.2.6 Temporal analysis of the histamine content of algae 48 3.2.7 Variation of histamine within algal thalli 49 3.2.8 The histamine concentration of seawater in the habitat 50 3.2.9 Settlement cue from Amphiroa anceps 51 3.2.10 Statistical treatment 55 3.3 Results 56 3.3.1 Recruitment Survey 56 3.3.2 Settlement assays testing host algae 56 3.3.3 Settlement assays testing in situ seawater 57 3.3.4 Temporal analysis of the histamine content of algae 58 3.3.5 Variation of histamine within algal thalli 58 3.3.6 The histamine concentration of seawater in the habitat 59 3.3.7 Settlement cue from Amphiroa anceps 59 3.4 Discussion 76

CHAPTER FOUR

The sensitivity and specificity of the larval response to histamine 81

4.1 Introduction 81 4.2 Methods 84 4.2.1 Larval culture 84 4.2.2 Settlement Assays 84 4.3 Results 88 4.3.1 Dose-response of larvae of different ages 88 4.3.2 Larval response to 'pulse’ histamine exposure 89 4.3.3 Larval response to neuroactives and histamine-analogues 90 4.3.4 Larval response to amino acids 90 4.3.5 Larval response to K+ and GGR 91 4.4 Discussion Error! Bookmark not defined. 101 4.4.1 Older larvae become more sensitive to histamine 101 4.4.2 Older larvae metamorphose faster than younger larvae 102 4.4.3 Specificity of the larval settlement response 104

CHAPTER 5

Is histamine a general inducer of settlement in echinoderm larvae? 108

5.1 Introduction 108 5.2 Materials and Methods 110 5.2.1 Larval culture and settlement assays testing histamine 110 5.2.2 Class Echinoidea, Order Temnopleuroida, Family Temnopluridae Holopneustes inflatus and Holopneustes purpurascens 111 5.2.3 Class Echinoidea, Order Temnopleuroida, Family Toxopneustidae Tripneustes gratilla 112 v

5.2.4 Class Echinoidea, Order Echinoida, Family Echinometridae Heliocidaris erythrogramma 114 5.2.5 Heliocidaris tuberculata 114 5.2.6 Class Asteroidea, Order Valvatida, Family Asterinidae Meridiastra calcar 115 5.2.7 Meridiastra hybrids – Meridiastra oriens & Meridiastra occidens 116 5.2.8 Class Ophiuroidea, OrderOphiurida, Family Ophiocomidae Clarkcoma canaliculata Error! Bookmark not defined.117 5.3 Results 118 5.3.1 Class Echinoidea, Order Temnopleurida, Family Temnopluridae Holopneustes inflatus and Holopneustes purpurascens 118 5.3.2 Class Echinoidea, Order Temnopleuroida, Family Toxopneustidae Tripneustes gratilla 119 5.3.3 Class Echinoidea, Order Echinoida, Family Echinometridae Heliocidaris erythrogramma 119 5.3.4 Heliocidaris tuberculata 119 5.3.5 Class Asteroidea, Order Valvatida, Family Asterinidae Meridiastra calcar 120 5.3.6 Meridiastra hybrids - Meridiastra oriens & Meridiastra occidens 120 5.3.7 Class Ophiuroidea, Order Ophiurida, Family Ophiocomidae Clarkcoma canaliculata 121 5.4 Discussion 129

CHAPTER 6

General Discussion 132

6.1 Histamine is an ecologically relevant settlement cue for Holopneustes purpurascens 132 6.1.1 Histamine as a naturally occurring settlement cue 132 6.1.2 Settlement and recruitment of Holopneustes purpurascens 132 6.1.3 Dissolved histamine as a waterborne settlement cue 134 6.1.4 Older larvae become more sensitive to histamine as a settlement cue 135 6.1.5 Histamine as a general settlement cue for non-feeding echinoid larvae 136 6.1.6 Settlement cue for Holopneustes purpurascens from coralline algae 138 6.1.7 The effect of flow on efficacy of histamine as a settlement cue 139 6.2 Metamorphosis and chemical signals 140 6.2.1 Metamorphosis via specific histamine receptors? 140 6.2.2 Chemical signals and biological responses 142

Literature cited 144

Appendix 1 166 Induction of settlement of larvae of the sea urchin Holopneustes purpurascens by histamine from a host alga. Biological Bulletin 206:161-172. R. L. Swanson, J. E. Williamson, R. de Nys, N. Kumar, M.Bucknall & P. D. Steinberg Appendix 2 178

In situ quantification of a natural settlement cue and recruitment of the Australian sea urchin Holopneustes purpurascens. Feature Article, Marine Ecology Progress Series 314:1-14. R. L. Swanson, R. de Nys, M. J. Huggett, J. K. Green & P. D. Steinberg

vii

LIST OF FIGURES

CHAPTER TWO

Figure 2.1 A schematic diagram illustrating bioassay-guided fractionation of the polar extract of Delisea pulchra 28 Figure 2.2 Settlement of larvae in response to HPLC fractions 29 Figure 2.3 Settlement of larvae in response to cation-exchange fractions 30 1 Figure 2.4A H-NMR (D2O) spectrum of active fraction F5 31 1 Figure 2.4B H-NMR (D2O) spectrum of synthetic histamine 32 1 Figure 2.4C H-NMR (D2O) spectrum of F5 spiked with histamine 33 Figure 2.5 The mass spectrum of isolated histamine and synthetic histamine 34 Figure 2.6 Settlement of larvae in response to isolated and synthetic histamine 35 Figure 2.7 Settlement of larvae in response to antibiotic-treated Delisea pulchra 36 Figure 2.8 Gas-chromatography–mass spectrometric analysis of ‘F-I complex’ 37

CHAPTER THREE

Figure 3.1 Photographs of reared and recruited Holopneustes purpurascens 61 Figure 3.2 Recruits found each month on each alga 62 Figure 3.3 Settlement of larvae in response to host algae 63 Figure 3.4 Settlement of older larvae in response to in situ seawater 64 Figure 3.5 Short term variation in histamine content of Delisea pulchra 65 Figure 3.6 Settlement of larvae in response to ‘no-contact’ Amphiroa anceps 66 Figure 3.7 Settlement of larvae in response to antibiotic-treated Amphiroa anceps 67 Figure 3.8A Settlement of larvae in response to bacterial isolates from biofilm 68 Figure 3.8B Histamine content of broth cultures of bacterial isolates 68

CHAPTER FOUR

Figure 4.1 Settlement of larvae of different ages in response to histamine 92 Figure 4.2 Induction of irreversible metamorphosis in larvae of different ages after exposure to histamine for different periods 93 Figure 4.3 Δ-Metamorphosis in larvae of different ages after exposure to histamine for different periods 94 Figure 4.4A Chemical structures of compounds which induced normal metamorphosis of larvae 95 Figure 4.4B Chemical structures of compounds which did not induce settlement of larvae 96 Figure 4.5 Settlement of larvae in response to amino acids 97 Figure 4.6 Settlement of larvae in response to K+ 98

CHAPTER FIVE

Figure 5.1A Cladogram of relationships among Class Echinoidea 122 Figure 5.1B Cladogram of relationships among classes of Echinodermata 122 Figure 5.2 Settlement of Holopneustes inflatus and Holopneustes purpurascens larvae in response to histamine 123 Figure 5.3 Settlement of Holopneustes inflatus and Holopneustes purpurascens larvae in response to host plants 124 viii

Figure 5.4 Settlement of Heliocidaris erythrogramma larvae in response to histamine 125 Figure 5.5 Settlement of Heliocidaris tuberculata larvae in response to histamine 125 Figure 5.6 Settlement of Meridiastra hybrid larvae in response to histamine 126 Figure 5.7 Settlement of Clarkcoma canaliculata larvae in response to histamine 126

LIST OF TABLES

CHAPTER TWO

Table 2.1 The histamine content of co-occurring algae 38

CHAPTER THREE

Table 3.1 BLAST analysis for bacterial isolates from coralline algae 69 Table 3.2 ANOVA of effect of algal on the settlement response of larvae and planned comparisons 70 Table 3.3 ANOVA of effect of in situ seawater samples (Nov-03)on the settlement response of older larvae and planned comparisons 71 Table 3.4 ANOVA of effect of in situ seawater samples (Jul-02) on the settlement response of older larvae and planned comparisons 72 Table 3.5 ANOVA effect of in situ seawater samples (Jan-03) on the settlement response of older larvae and planned comparisons 73 Table 3.6 The histamine content of algae over four seasons and ANOVA 74 Table 3.7 Within plant variation of histamine content of Delisea pulchra and Ecklonia radiata and ANOVA 75

CHAPTER FOUR

Table 4.1 ANOVA of the effects of larval age and histamine concentration on the settlement response of larvae (batch A, B, C) 99 Table 4.2 ANOVA of the effects of larval age and histamine concentration on the settlement response of larvae (batch A only) 100 Table 4.3 ANOVA of the effects of larval age and exposure to histamine on the induction of irreversible metamorphosis of larvae 100

CHAPTER FIVE

Table 5.1 ANOVA of the effects of species and histamine concentration on the settlement response of Holopneustes inflatus and Holopneustes purpurascens larvae 127 Table 5.2 ANOVA of the effects of species and histamine concentration on the settlement response of Holopneustes inflatus and Holopneustes purpurascens larvae 128

CHAPTER ONE

General Introduction

Most marine invertebrates have complex life histories in which a planktonic larval phase alternates with a benthic juvenile/adult phase. The planktonic larval phase can last from minutes to months depending on the mode of development (Strathmann 1985). Planktotrophic or feeding larvae can spend up to several months in the plankton due to a dependency on exogenous food sources for growth and development to a competent state capable of metamorphosis. Lecithotrophic or non-feeding larvae typically have a shortened period of development in which yolk reserves provisioned in the egg nourish larval development to competency. Regardless of the length of larval life, the transition from a planktonic to benthic existence through larval settlement and metamorphosis is a crucial stage in the life history of such organisms. For sessile or low-mobility organisms, the fate of the adult is particularly dependent on larvae settling in an appropriate habitat. The question of how planktonic larvae return to and ‘choose’ an appropriate habitat in which to settle has been a major focus of marine ecology for over fifty years (Thorson 1950, Meadows & Campbell 1972, Pawlik 1992, Hadfield & Paul 2001).

The factors that govern larval settlement and subsequent recruitment are among the most important determinants of population dynamics in benthic communities (Underwood & Keough 2000). Larval supply and initial settlement rates contribute to the observed distribution and abundance of recruits in benthic habitats (Connell 1985, Gaines et al. 1985, López et al. 1998, Olivier et al. 2000). Larval supply is dependant on three processes; the initial production of larvae (Levin 1984), the extent of larval dispersal by oceanic currents (Young 1995) and the level of mortality suffered during the planktonic phase (Morgan 1995). Settlement rates of larvae are influenced by passive (i.e., purely hydrodynamic) and active processes (i.e., larval behaviour) which operate on different spatial scales (Butman 1

1987). Early mortality rates of new recruits through the post-settlement processes of predation and competition, and movement of mobile species may also contribute to observed distributions (Cameron & Schoeter 1980, Keough & Downes 1982, Connell 1985, Rowley 1989, Wilson 1991, Menge 2000, Mercier et al. 2000, Ebert 2001).

There are very few reports of direct observations of larval behaviour in the natural environment due to the intrinsic difficulties in tracking such minute organisms. Direct observations of larval behaviour in the natural habitat are largely restricted to the relatively large tadpole larvae of colonial ascidians (Davis & Butler 1989, Davis et al. 1991, Stoner 1992, 1994) and larvae of decapod crustaceans (Shanks 1985). The majority of observations of larval behaviour have occurred in the laboratory under tightly controlled conditions, which often include no water movement. While such studies may highlight species-specific differences which affect larval distribution and dispersal, larvae often behave differently in still and flowing water (Snelgrove et al. 1998, Altieri 2003). Laboratory investigations in recent times have attempted to emulate more closely the natural conditions experienced by larvae, for example, by observing larval behaviour in flumes under realistic flow conditions (Turner et al. 1994, Tamburri et al. 1996, Altieri 2003, Finelli & Wethey 2003, Hadfield & Koehl 2004). These studies may reflect how two factors such as flow and substrate interact to affect larval behaviour, however, they still do not reflect the natural habitat where the interplay of many factors affect larval behaviour (Forward 1988). Methodological breakthroughs are required to allow direct observations of smaller larvae in the natural habitat. Until such time, investigations of larval distributions in the water column and of recruitment patterns in the field on natural or experimental substrata (Keough & Downes 1982, Harvey & Bourget 1997, Thompson et al. 1998, Wright & Boxshall 1999, Incze et al. 2000) provide the best insight into larval behaviour under natural conditions.

The literature suggests that the relative contribution of hydrodynamic factors and active larval behaviour in larval settlement is highly species-specific and scale-dependant (Butman 1987). Hydrodynamic factors such as oceanic and tidal currents operate on larger spatial scales (km, m, cm) transporting larvae back to inshore areas, even to the initial point

of contact with the substratum (Hannan 1984, Mullineax & Butman 1991, Harvey & Bourget 1997). At smaller spatial scales (cm, mm, μm) active larval behaviour can interact with hydrodynamic conditions to influence where they settle. Settlement and recruitment patterns of a wide range of species (including polychaetes, bivalves, gastropods, echinoderms and corals) can be explained by passive transport and deposition, based on observations that the distributions of larvae in these studies were similar to observed/predicted distributions for passive, neutrally buoyant bodies (Eckman 1983, Eckman 1987, Sammarco & Andrews 1989, Black & Moran 1991, Snelgrove 1994). However, it is also clear that many larvae control (to varying degrees) their horizontal distribution and dispersal by vertical migration through the water column (Young 1995), one aspect of larval behaviour that has been well investigated in the field (Forward 1988, Tremblay & Sinclair 1990). Larvae alter their vertical position in the water column on diel and ontogenic cycles by responding to vector (e.g., light, gravity, current) and/or scalar (e.g., salinity, pressure, temperature) cues (Young 1995). For example, some species of crustacean larvae show complex migratory and depth regulation behaviours which involve the interplay of multiple vector and scalar cues (Tankersley et al. 1995). The phototactic and geotactic response of many larval species also changes with ontogeny (Thorson 1964, Sulkin et al. 1980). Ontogenetic regulation of phototaxis and geotaxis in larvae is likely to facilitate dispersal early in the planktonic period (photopositive, geonegative) and bring competent larvae to the near-bed region at the time of settlement (photonegative, geopositive) (Pires & Woollacott 1997).

All marine invertebrate larvae swim actively using cilia, however, the majority swim at speeds less than 1 mm·s-1 (Chia et al. 1984). Horizontal flow velocities in benthic boundary layers considerably exceed larval swimming speeds, even at only several body lengths above seabed (Butman 1986). Therefore many researchers have assumed that larvae would have little ability to navigate in turbulent benthic environments let alone search and discriminate between settlement sites (Crisp 1974, Butman 1987, Pawlik 1992). However, larvae do exhibit active behaviour in the physically demanding habitats of the intertidal and subtidal zones (Strathmann et al. 1981, Raimondi 1988, Koehl & Hadfield 2004). Barnacle larvae, for example, settle on rock surfaces in spite of strong shear forces encountered in

the surf zone of rocky shores but such behaviour is dependant on the remarkably strong attachment of cyprids to the substratum provided by an adhesive secreted by the antennules (Eckman et al. 1990). In subtidal areas of medium to high flows, larval exploration of substrata during settlement probably involves swimming down (or sinking) and back up into the water column repeatedly until a favourable site is encountered rather than manoeuvring in the horizontal (Butman 1987). However, there are areas of low flow in most hydrodynamic environments such as in boundary layers and areas in the lee of benthic objects (Eckman 1983) which may be quite common in subtidal areas and in intertidal zones of sheltered shores (Eckman 1996). Such regions of low flow can provide larvae with the opportunity to swim and explore the substratum in search of a favourable site in which to settle.

Habitat cues which influence larval settlement range from simple cues to complex cues. Larvae may simply respond to the physical factors of a habitat such as light intensity (Maida et al. 1994) or surface texture (Berntsson et al. 2000, Berntsson et al. 2004), or they can use complex cues such as responding to successional changes that occur during development of biofilms (Keough & Raimondi 1995). However, it would appear that the most common settlement cues used by marine invertebrate larvae are chemical cues (Hadfield & Paul 2001). The source of chemical settlement cues can be host organisms (Krug & Manzi 1999, Williamson et al. 2000, Swanson et al. 2004), prey species (Hadfield & Scheuer 1985), conspecifics for gregarious settlers (Burke 1984, Toonen & Pawlik 2001) or the surface-associated microbial communities (biofilms) on substrata (Johnson & Sutton 1994, Wieczorek & Todd 1998, Huggett 2006, Huggett et al. 2006). Larvae of invertebrate species which are specialist feeders are likely to be induced by chemical cues (Hadfield & Paul 2001) as survival of the juvenile is completely dependant on the presence of the species on which they feed. Chemical settlement cues can be surface-bound and detected by larvae upon contact with the substrate (Matsumura et al. 1998b), or dissolved in seawater and detected in the water column nearby the source (Zimmer-Faust & Tamburri 1994). While most surface-bound cues are non-polar, some surface-bound cues have water-soluble components (Morse et al. 1984, Morse & Morse 1984, Boettcher & Targett

1996) or may release water-soluble components into surrounding seawater (Rittschof 1985).

A large number of chemical settlement cues have been partially characterised, including many classes of compounds ranging from low molecular weight (LMW) peptides (Zimmer- Faust & Tamburri 1994, Lambert et al. 1997, Fleck & Fitt 1999) to high molecular weight carbohydrates (Krug and Manzi 1999). However, despite over sixty years of research effort very few chemical cues have been fully characterised for a number of reasons. Settlement cues appear to be potent molecules that are presented to larvae at very low concentrations hence it is difficult to extract sufficient quantities for isolation and structural characterisation. Water-soluble compounds are particularly difficult to isolate and purify from seawater and biological tissues due to the interaction of salts with chromatographic procedures.

Despite the paucity of known chemical settlement cues, intensive study of some species due to their economic or ecological importance has illustrated the complexity of the underlying processes involved in larval settlement. The following examples are from species that highlight a common theme, for example, many settlement cues appear to be peptides. Chemically specific gregarious settlement of barnacles and oysters is also discussed as these are among the best studied species for purified or partially purified inducers.

Abalone larvae (Haliotis rufescens) settle upon contact with crustose coralline algae such as Lithothamnium californicum (Morse et al. 1984). Settlement activity in this study was initially confined to the macromolecular fraction of the crude extract but the activity could be separated from proteins into a water-soluble LMW fraction. The settlement cues have been partially characterised as small peptides (640-1250 Da) with unusual composition, which are thought to be complexed to phycobiliproteins on the algal surface (Morse et al. 1984, Morse & Morse 1984). The authors proposed that the surface-bound peptide may be structurally similar to γ–aminobutyric acid (GABA) or γ–aminolevulinic acid because these compounds ‘mimicked’ the effect of the natural cue (Morse et al. 1979). Furthermore, 5

partially purified inducer molecules were shown to compete for mammalian GABA receptors indicating that they were truly GABA-mimetic (Morse & Morse 1988). The close association of the inducer with phycobiliproteins suggests that they may be related to the synthesis, function or degradation of phycobiliproteins (Morse et al. 1984, Morse & Morse 1984); a theory supported by the settlement inducing activity of γ–aminolevulinic acid which is a direct precursor in the biosynthesis of the tetrapyrrole chromophores of algal phycobiliproteins (Troxler & Lester 1967).

Barnacles, because of their importance as fouling organisms and ubiquitous presence in the intertidal worldwide, have received considerable attention with respect to settlement cues, and arguably the common fouling organism Balanus amphitrite has been studied the most. B. amphitrite larvae settle in response to multiple cues produced by conspecifics (Clare & Matsumura 2000). A surface-bound adult glycoprotein, known as the settlement inducing protein complex (SIPC), induces cypris larvae to attach temporarily to the substratum via an antennular adhesive (Walker & Yule 1984). If larvae choose to accept the site then they secrete a permanent adhesive, a cement from the paired glands (Walker 1971) and subsequently metamorphose. SIPC consists of 3 major subunits with estimated molecular masses of 76 (often as a dimer), 88 and 98 kDa (Matsumura et al. 1998b). The temporary adhesive secreted by cyprids of Balanus balanoides and B. amphitrite during substrate exploration leaves behind a path of ‘footprints’ which also contain a settlement cue (Walker & Yule 1984). This cue in B. amphitrite ‘footprints’ may be related to the 76 kDa subunit of SIPC, as the anti-76 kDa antibody cross reacted with the cypris temporary adhesive (Matsumura et al. 1998c). Adults also produce a waterborne cue that is more potent than SIPC (Rittschof 1985). It is proposed to be a 3-5 kDa peptide with arginine or lysine at the carboxy terminus (Tegtmeyer & Rittschof 1989). A synthetic peptide analogue to barnacle settlement pheromone, glycine-glycine-arginine (GGR), was found to be a potent inducer of settlement in barnacle larvae (Tegtmeyer & Rittschof 1989), however, the activity of GGR has been disputed (Clare and Yamazaki 2000).

Small peptides also appear to function in the settlement of the oyster Crassostrea virginica which settle gregariously in response to conspecifics (Crisp 1967). The biofilm on oyster 6

shells (Tamburri et al. 1992) produce the chemical cue(s) which evoke settlement behaviour (downward swimming, surface exploration, attachment) in oyster larvae. The behavioural cue appears to be LMW peptides (500-1000 Da) with arginine at the C- terminus (Zimmer-Faust & Tamburri 1994) and its effects are mimicked by GGR (Tamburri et al. 1996). Thus LMW peptides evoke settlement behaviour in C. virginica larvae and settlement of Balanus amphitrite larvae. Interestingly, these same LMW peptides seem to also attract oyster drills (borers) to oysters and barnacles suggesting that the same peptide cue is used by different organisms within the same habitat (Rittschof 1985).

These investigations of oyster larvae provided the first experimental evidence demonstrating that waterborne cues could evoke behavioural responses in larvae in flow conditions (Turner et al. 1994, Tamburri et al. 1996), eliminating a substantial conceptual argument against settlement induction by waterborne cues: that waterborne cues were unlikely to act as settlement cues because stable chemical gradients could not be maintained above substrata in turbulent benthic boundary layers. However, Turner et al (1994) showed that in fact gradients were unnecessary, because oyster larvae settled (i.e., attached) in response to any low concentration of peptide and did not move toward a higher concentration in a chemical gradient. That is, they responded in a binary way to any concentration above a certain threshold. These findings were supported by a recent study which found that stable chemical gradients were not formed in turbulent flow, instead, cue is distributed in fine-scale filaments of high concentration interspersed within cue-less water that larvae experienced as rapid on/off encounters (Zimmer et al. 1999, Hadfield & Koehl 2004).

Oyster larvae have provided further evidence of the potentially remarkable behaviour of larvae in boundary layers in the study of Finelli and Wethey (2003). Observations of the behaviour of oyster larvae in the bottom 1 cm of a flume boundary layer discovered that a small percentage (~5 %), regardless of age or size, were capable of remarkable behaviour termed ‘dive bombing’ (Finelli & Wethey 2003). The vertical acceleration at the initiation of dive-bombing was 30-400 times that measured for other live or dead larvae 7

demonstrating that some larvae are capable of rapid acceleration and thus may have considerable control over their approach to the bottom (Finelli & Wethey 2003). Rapid downward acceleration upon detection of a waterborne cue would bring larvae into close proximity of the source, be it a host plant or prey species or suitable substrata for settlement. Rapid responses of larvae to the detection of waterborne cues are likely to be essential for larvae to target specific settlement sites in turbulent marine habitats.

The literature from over half a century of research attests that larvae of many species are influenced by specific chemical cues to guide their settlement into appropriate habitats for juvenile growth to reproductive age (Hadfield & Paul 2001). Yet very few such chemical cues have been definitively identified. Identifying settlement cues which have fundamental consequences for population structure and community dynamics of benthic ecosystems (Underwood & Keough 2000) is thus critical to our understanding of the ecology of these systems. Due to the cross-disciplinary nature of the field, the identification of naturally occurring chemical cues for larval settlement will require the collaboration of ecologists, biologists and chemists to succeed.

1.1 The study organism - Holopneustes purpurascens

Holopneustes purpurascens (Temnopluridae: Echinodermata) typically occurs in shallow, rocky, subtidal habitats in south-eastern Australia (Miskelly 2002). H. purpurascens is an unusual echinoid which shares some characteristics with much smaller herbivores or mesograzers (Brawley 1992). Although similar in size to regular bottom-dwelling echinoids, H. purpurascens live enmeshed in the canopy of macroalgae/kelp using the host plant as a source of food and habitat, similar to mesograzers (Brawley 1992). At Bare Island, Sydney, H. purpurascens are predominantly found on the foliose alga Delisea pulchra (Bonnemaisonales: Rhodophyta) or the kelp Ecklonia radiata (Laminariales: Phaeophyta) (Steinberg 1995, Williamson et al. 2000, Williamson et al. 2004). Although abundant on both host plants, the smaller size classes of H. purpurascens are most abundant on D. pulchra, with the smallest size class (test diameter ≤ 5 mm) found only on D. pulchra. This distribution suggested that D. pulchra produces a settlement cue for the 8

lecithotrophic larvae of H. purpurascens (Williamson et al. 2000). Fresh pieces of D. pulchra (but not E. radiata), the polar extract of D. pulchra and seawater collected in situ nearby D. pulchra plants induced settlement in larvae of H. purpurascens. The water- soluble cue from D. pulchra was subsequently isolated and characterised as a ‘complex’ between the sugar floridoside and isethionic acid, or the F-I complex (Williamson et al. 2000). The chemical identity of the settlement cue was confirmed when a synthetic F-I complex induced the settlement of H. purpurascens larvae (Williamson et al. 2000).

1.2 Aims and structure of this thesis

During initial further research on this system, the characterisation of the F-I complex as a chemical cue for settlement of this urchin was called into question. In particular, I obtained active fractions from Delisea pulchra which contained isethionate but not floridoside and I was unable to make an active synthetic F-I complex which induced settlement. It became apparent that the F-I complex was not a settlement cue for Holopneustes purpurascens. Subsequently, the overall aim of this research became the identification of the true settlement cue from D. pulchra for H. purpurascens. In Chapter Two, I describe the bioassay guided fractionation procedures which led to the isolation of the settlement cue from D. pulchra, and the chemical analyses which identified the settlement cue as histamine. The histamine content of the primary hosts of H. purpurascens, D. pulchra and Ecklonia radiata, and other algae in the habitat, was quantified by Gas Chromatography- Mass Spectrometry. D. pulchra had the highest histamine content, the source of which (i.e., algal-derived or biofilm-derived) is investigated in this chapter.

The aim of Chapter Three was to demonstrate that histamine is an ecologically relevant settlement cue for Holopneustes purpurascens, by showing that variation in the distribution of new recruits in the habitat was related to variation in the distribution of histamine in the habitat. To achieve this, recruitment of H. purpurascens onto algal substrata was monitored each month for two years, a range of algae (host and non-host species) were tested in settlement assays for their ability to induce metamorphosis of H. purpurascens larvae, and the histamine content of selected algae was quantified over four seasons to 9

determine the temporal variation in histamine production. Further, seawater samples were collected in situ near-by algae and tested for their ability to induce metamorphosis of newly competent and older H. purpurascens larvae. The histamine concentration of these seawater samples is reported. The localisation (i.e., surface-bound or waterborne) and source (i.e., algal-derived or biofilm-derived) of the settlement cue for H. purpurascens larvae produced by coralline algae is also investigated in this chapter.

The sensitivity of larvae of different ages to histamine is assessed in Chapter Four. These experiments were prompted by the fact that some seawater samples (collected in situ nearby Delisea pulchra) induced the metamorphosis of older larvae with no effect on newly competent larvae. Competent (7-d-old) and older (up to 28-d-old) larvae were tested against a range of histamine concentrations in settlement assays. Another set of experiments investigated whether larval age affects the duration of exposure to histamine required to induce irreversible metamorphosis of larvae. The specificity of induction of metamorphosis of H. purpurascens by histamine is also explored in this chapter. Various compounds with either similar structure or function (i.e., other biogenic amines) to histamine were tested for their ability to induce metamorphosis of H. purpurascens.

In Chapter Five, the question of whether histamine is a general inducer of metamorphosis in echinoids or other echinoderms is addressed. Histamine was tested as a possible inducer of metamorphosis of larvae of regular echinoids (Class Echinoidea: Holopneustes inflatus, Heliocidaris erythrogramma, Heliocidaris tuberculata and Tripneustses gratilla), sea stars (Class Asteroidea: Meridiastra spp.) and a brittle star (Class Ophiuroidea: Clarkoma canaliculata).

In Chapter Six, the findings presented in each chapter are brought together and discussed in a broader context of chemical signalling in the marine environment. Directions for future research are proposed.

1.3 A note on definitions

There is no consistency in the literature when referring to the processes of settlement and metamorphosis hence it is appropriate to clearly define the use of these terms in this thesis. The term settlement (and settle) is used in reference to the whole process of larvae transiting from a planktonic existence to a benthic one; including the behavioural aspects of settlement such as surface exploration and attachment, as well as the metamorphosis of larvae to the juvenile form. The term metamorphosis is used when specifically referring to the morphological transformation to the juvenile form and was the end point scored in all settlement assays (rather than settlement behaviour i.e., exploration, attachment).

11 CHAPTER TWO

Isolation and characterisation of the settlement cue for

Holopneustes purpurascens*

* This chapter is published as Swanson et al. (2004) Biological Bulletin 206: 161-172 (Appendix 1)

2.1 Introduction

The apparent diversity of chemical cues for larval settlement isolated from natural sources within the habitat is considerable. However, most of these cues have only been partially characterised. These include; small peptides (500-1000 Da) which induce attachment of oyster (Crassostrea virginica) larvae (Zimmer-Faust & Tamburri 1994), and settlement of sand-dollar (Dendraster excentricus) larvae (Burke 1984) and abalone (Haliotis rufescens) larvae (Morse et al. 1984); low molecular weight (LMW, <1000 Da) water-soluble compounds which induce settlement of the coralivorous nudibranch Phestilla sibogae (Hadfield & Pennington 1990) and the opisthobranch mollusc Haminaea callidegenita (Gibson & Chia 1994); surface-bound, cell-wall associated polysaccharides on crustose coralline algae which induce settlement of many species of coral larvae including Agaricia spp. and Acropora spp. (Morse & Morse 1991, Morse & Morse 1996); high molecular weight (HMW, >100 000 Da) surface-bound carbohydrates and LMW (<2000 Da) water-soluble carbohydrates produced by the algal host, Vaucheria longicaulis, which induce settlement of larvae of the specialist ascoglossan Alderia modesta (Krug & Manzi 1999); HMW glycoproteins and a more active water soluble cue produced by adult barnacles Balanus amphitrite which induce settlement of conspecific larvae (Rittschof 1985, Clare & Matsumura 2000).

In contrast to the many partially characterised inducers in the literature there are only a few examples in which the chemical structure of a settlement cue isolated from a natural source has been determined. Even in these cases, however, the ecological relevance of

12 the putative settlement cue in situ is not clear. Kato et al. (1975b) reported that δ- tocopherol epoxides produced by Sargassum tortile induced larvae of the hydroid Coryne uchidai to settle on this alga. Synthetic forms of these compounds also induced settlement of C. uchidai larvae (Kato et al. 1975a). However, these compounds are lipophilic whereas the original study suggested that the cue from S. tortile was water- soluble (Nishihira 1968). Jacarone isolated from the red alga Delesseria sanguinea induced settlement of the scallop Pecten maximus (Yvin et al. 1985), however, these larvae are not known to settle preferentially on the red alga. Narains and anthosamines A and B isolated from marine sponges induced settlement of ascidian larvae (Tsukamoto et al. 1994, 1995), however, it is not clear whether Ciona savignyi larvae settle on these sponges. Ascidians often settle gregariously (Svane & Young 1989), which lends support to a conspecific-derived settlement cue in these species. Urochordamine A, isolated from the tunic of C. savignyi induced settlement of conspecific larvae and may function in gregarious settlement (Tsukamoto et al. 1993). Likewise, lumichrome, isolated from eggs, gonads and the tunic tissue of adult conspecifics, and from seawater conditioned by large numbers of larvae, induced settlement of larvae of the ascidian Halocynthia roretzi (Tsukamoto et al. 1999). However, none of these putative settlement cues for ascidian larvae have been confirmed as natural inducers in the habitat: further biological, ecological and chemical data are required (Hadfield & Paul 2001).

A naturally occurring characterised settlement cue that appears to strongly affect the demography of the sea urchin Holopneustes purpurascens (Temnopluridae: Echinodermata) was recently reported by Williamson et al. (2000). The water-soluble cue from the alga Delisea pulchra was isolated and characterised as a floridoside and isethionic acid complex – the F-I complex (Williamson et al. 2000). During further research on this system I obtained inductive fractions of D. pulchra extract which contained isethionic acid but not floridoside and I was unable to reproduce a synthetic F-I complex which induced settlement of larval H. purpurascens. Subsequently, I hypothesised that the F-I complex was not a natural settlement cue for H. purpurascens. This chapter identifies the true nature of the chemical cue from D. pulchra that induces the settlement of H. purpurascens larvae, correcting the previous finding of Williamson et al. (2000). In addition, the settlement cue was quantified in host and non-host algae

13 of H. purpurascens which is the first time that a natural settlement cue has been quantified in the habitat of a marine organism. The source of the settlement cue in D. pulchra was investigated.

2.2 Materials & Methods

2.2.1 Study site

All urchins and algae used in this study were collected from sub-littoral habitats (1–3 m depth) at Bare Island (33° 59' 38" S, 151° 14' 00" E) at the north head of Botany Bay, Sydney, Australia. At this site adult Holopneustes purpurascens are primarily found wrapped in the laminae of the brown kelp Ecklonia radiata. A more detailed description of this habitat and the ecology of this system are found in Wright and Steinberg (2001) and Williamson et al. (2004).

2.2.2 Preparation of the polar extract of Delisea pulchra

The results of Williamson et al. (2000) indicated that any settlement cues were contained within the polar fraction of the crude extract of Delisea pulchra. A polar extract of D. pulchra was thus prepared from 1.0 kg (wet weight - ww) of algae collected from Bare Island. Epibiota were removed, the plants blotted dry and the thallus exhaustively extracted in methanol (OmniSolv, EM Science). The methanol extract was filtered (Whatman #1), dried by rotary evaporation in vacuo at 40°C and partitioned between dichloromethane (OmniSolv) and Milli-Q. The Milli-Q phase was filtered (Whatman #1) and dried in vacuo at 40°C. The dried crude polar extract was dissolved in absolute ethanol three times, pooling each extract, and dried in vacuo at 40°C to yield the polar extract.

14 2.2.3 Isolation of the settlement cue in Delisea pulchra using bioassay-guided fractionation

i) High performance liquid chromatography

The polar extract of Delisea pulchra was fractionated using reversed-phase high performance liquid chromatography (HPLC — Adsorbosil C18 column, 5 µm particle size, 250 mm x 4.6 mm, Waters R410 RI-detector) (100% Milli-Q at 1 ml·min-1). The polar extract was dissolved in Milli-Q (50 mg·ml-1), filtered (0.22 µm) and manually injected (20 µl). HPLC resolved 2 major peaks, peak 1 (retention time [rt]-2.7 min) and peak 2 (rt-3.4 min) (Figure 2.1A). Each peak fraction was collected from multiple injections and dried by rotary evaporation in vacuo at 40°C. Peak fractions were tested for bioactivity in settlement assays and analysed by 1H- and 13C-nuclear magnetic resonance (NMR) spectroscopy (Bruker DMX 500). ii) Cation-exchange chromatography

The settlement cue could not be isolated as a pure fraction using HPLC so an alternative procedure, cation-exchange (CX) chromatography, was used to fractionate the polar extract of Delisea pulchra. CX resin (AG50W-X2 [H+ form], BioRad) in Milli-Q was poured into a 50 ml burette taking care to exclude air bubbles. The resin (25 ml bed volume) was equilibrated with Milli-Q at 2 ml·min-1 until the eluant was pH 5–6. The polar extract of D. pulchra (1–2 g) was dissolved in 5-ml Milli-Q, filtered (0.22 µm) and gently loaded onto the column. Unbound compounds were collected in 100 ml Milli-Q (fraction 1) and another 100 ml Milli-Q (fraction 2). Retained compounds were eluted using a series of basic solutions: 30 ml diluted NH3 in Milli-Q (pH 10 - fraction

3), 30ml 3%-NH4OH w/w (fraction 4) and 30 ml 30%-NH4OH w/w (fraction 5, Figure 2.1B). ‘Control’ fractions 1–5 were collected using the same method without loading any D. pulchra extract on the column; none of these fractions had any subsequent activity. CX-fractions were dried in a centrifuge in vacuo (Speed-Vac SVC200, Savant) and tested for bioactivity in settlement assays and analysed by 1H-NMR spectroscopy.

15 2.2.4 Identification of isolated settlement cue

i) Nuclear magnetic resonance spectroscopy

Bioassay-guided fractionation of the polar extract of Delisea pulchra by cation- exchange chromatography yielded one active fraction (CX-fraction 5, F5). The 1 13 inducing compound in F5 was identified by H and C-NMR experiments (D2O) and a 1 15 high field two-dimensional H- N HMBC NMR experiment (d4 MeOH, Bruker DMX 500). To confirm the putative structure of F5 as histamine 3 mg of F5 was dissolved in 1 D2O and analysed by H-NMR spectroscopy. Synthetic histamine (3 mg) was analysed by 1H-NMR spectroscopy which was then added to F5 and the sample was re-analysed. The 1H-NMR spectra of the unspiked F5 sample and the spiked F5 sample were then compared. ii) Gas chromatography – mass spectrometry

NMR spectroscopy analyses identified the isolated settlement cue as histamine and this was confirmed by gas chromatography–mass spectrometry (GC-MS). Putative (naturally isolated) histamine (1 mg) and synthetic histamine (1 mg) were derivatised with heptafluorobutyric anhydride (Aldrich, 1st derivative) and then acetic anhydride (Aldrich, 2nd derivative) using the method of (Barancin et al. 1998). Heptafluorobutyric anhydride (100 μl) was added to each sample in 8 ml culture tubes and flushed with N2. Samples were left at room temperature (RT) for 2 min, mixing occasionally. Anhydrous ethyl acetate (anEtoAc, 100 μl) and 0.1M thiethylamine (in EtoAc, 10 μl) were added and samples incubated at 150°C for 13 min, mixing occasionally. Samples were cooled on ice and the excess reagent was dried off under a gentle stream of N2. To the dried residue, 2.5M KHCO3 (pH 8, 1 ml) and 0.5M KOH (0.5 ml) were added, mixing well, followed by 2 ml anEtoAc. Samples were inverted 20 times and then allowed to settle so the two phases separated. The aqueous phase (lower) was discarded. HCl (0.1M, 2 ml) was added and samples were inverted 20 times and then allowed to settle. The organic phase (upper) was discarded. KHCO3 (2.5M, pH 8, 1 ml) was added followed by 1 ml anEtoAc. Samples were inverted 20 times and then allowed to settle. The organic phase was transferred to a fresh culture tube being careful not to transfer any aqueous phase. Samples were dried to 20-50 μl under a very

16 gentle stream of N2 and resuspended in 1 ml anEtoAc (samples were placed in freezer at this point and 2nd derivative done just prior to analysis). Samples were brought to RT. Approximately 10 crystals of anhydrous sodium acetate and 40 μl of acetic anhydride were added to each sample. Samples were mixed and left for 45 min at RT, mixing occasionally. Milli-Q (1 ml) was added and the samples were inverted 20 times and then allowed to settle. The organic phase was transferred to 1.5 ml GC-MS vials and dried to 20-50 μl under a very gentle stream of N2. Samples were resuspended in the appropriate volume of anEtoAc.

Derivatised samples were resuspended in 1 ml anEtoAc and diluted again 100-fold before analysis. A Zebron ZB-5 column (15 m, 0.25 µm x 0.25 mm ID; Phenomenex) was used on a Hewlett Packard (HP) 5980 series II gas chromatograph interfaced to a HP5971A or HP5972 mass selective detector equipped with HP Chemstation G1034C software (Version C.03.00). Injections (2 µl) were in the split-less mode with an inlet pressure of 170 kPA. The injection port was held as 290°C and the interface at 300°C. The gas chromatograph was held at 90°C for 2 min and ramped at 10°C·min-1 to 200°C, then at 50°C·min-1 to 310°C and held for 2 min (17.2 min run). Helium was used as the carrier gas. The mass selective detector was operated in scan mode (m/z 50–550). The average retention times of derivatised putative histamine and derivatised synthetic histamine were recorded from five injections of each sample (mean ± SD, n = 5). The electron impact ion-spectra of derivatised putative histamine and derivatised synthetic histamine were compared. iii) Matrix-assisted laser desorption/ionisation – time-of-flight mass spectrometry

The elemental formula of putative histamine was determined by matrix-assisted laser desorption/ionisation – time-of-flight mass spectrometry (MALDI-TOFMS) (Bucknall et al. 2002). A Perseptive Voyager DE STR (Perseptive Biosystems, Framingham, MA) MALDI-TOFMS was operated in both positive ion linear delayed extraction and reflector delayed extraction modes for accurate mass analysis. The test samples were prepared in acetonitrile:Milli-Q (50:50) and contained either 100 ng·µl-1 of putative histamine or synthetic histamine. ∝-Cyano-4-hydroxycinnamic acid (5 mg·ml-1) prepared in acetonitrile:Milli-Q:trifluoroacetic acid (80:20:0.02) was used as the matrix.

17 -1 15 -1 Glycine (500 ng·µl ) and [sarcosine- N-methyl-d3]creatinine·HCl (5 ng·µl , Cambridge Isotope Laboratories #DNLM-2171) were added as internal mass calibrants for accurate mass determinations. An accurate mass for the putative protonated histamine molecular ion [M+H]+ was determined by 10 repeat analyses of each sample. The mean molecular weight was calculated for these mass spectra and compared with both the theoretical molecular weight for histamine and the molecular weight measured for synthetic histamine using the same analytical technique. The standard deviation for these mass measurements was taken as an estimate of the mass measurement error.

2.2.5 Larval culture

Approximately 20 adult Holopneustes purpurascens (test diameter ≥ 40mm) were collected from Ecklonia radiata laminae using SCUBA and returned to the laboratory in ambient seawater (SW). Urchins were induced to spawn in a constant temperature room (CTR, 19°C) by injecting 3–5 ml of 0.5 M KCl into the intracoelomic cavity through the peristomal membrane (Tyler 1949, King et al. 1994). Sperm was collected by pipette from at least 2 males, transferred to separate sterile 36 mm petri-dishes and covered with alfoil to exclude light. Eggs were collected by pipette and transferred to 2 x 2 l sterile beakers containing autoclaved SW with antibiotics added (22 mg·l-1 of penicillin G and 37 mg·l-1 of streptomycin sulfate, SSW). Each beaker contained eggs from at least 2 females which were rinsed twice with SSW. Sperm (50–250 μl, depending on sperm density) from each male was added to each beaker until SSW containing eggs turned slightly opaque. Eggs and sperm were aerated for 10–15 minutes after which time ~30 eggs were examined for raised vitelline membranes, using a dissecting microscope, indicating fertilisation had occurred. If the majority of eggs were fertilised (as indicated by a fertilisation membrane), the egg/sperm suspension was rinsed 3 times with SSW, to remove sperm. Embryos/larvae were cultured in a CTR at 19°C with a 12-h light/12-h dark regime. Embryos were aerated gently in SSW overnight and checked the next morning. If the majority were spinning (indicating fertilisation success) they were rinsed twice in SSW. Larvae hatched later that day which were held at a density of approximately 1–3 larvae·ml-1. SSW was changed daily and abnormal

18 larvae were removed as necessary. Larvae reached competency (developmentally ready for settlement) within 6 d, recognised by the presence of five well-developed tube-feet.

2.2.6 Settlement assays

All settlement assays were done in a CTR (19°C, 12-h light/12-h dark regime) in 36 mm sterile petri-dishes and 5 ml SSW. Replicates were randomly assigned among treatments with 10–15 replicates per treatment and one competent larva (6-d-old) per replicate dish (settlement is not gregarious, Williamson et al. 2000). I did not use multiple larvae per dish in these assays because this species is a ‘dribble’ spawner for most of the year and generally yields low numbers of larvae (Williamson & Steinberg 2002). Larvae were added once all petri-dishes were prepared and percent settlement (i.e., percent metamorphosed) was recorded at set time intervals. i ) HPLC peak fractions

Peak 1 and peak 2 fractions were tested against larvae to determine the presence of a settlement cue. Peak fractions were dissolved in Milli-Q (10 mg·ml-1 stock solution) and aliquots of each stock solution were added to petri-dishes for final test concentrations of 25 µg·ml-1 of peak 1 and 51 µg·ml-1 of peak 2. A natural ‘F-I complex’ sample from the previous study was also tested in the assay at a final concentration of 76 µg·ml-1. Pieces of fresh Delisea pulchra (~20 mg) were used as a positive control, and Milli-Q and SSW were used as the negative controls. Percent settlement was scored after 18 h (n = 12). ii) Cation-exchange fractions

Each CX-fraction (F) was tested against larvae to determine the presence of a settlement cue. F1, F2, F3, F4 and the polar extract of Delisea pulchra (used as a positive control), were dissolved in Milli-Q at 5 mg·ml-1. Aliquots of the appropriate fraction were added to the Petri-dish to give final test concentrations of 50 µg·ml-1 for each treatment. F5 was dissolved in Milli-Q at 100 µg·ml-1 and aliquots were added to petri-dishes for final test concentrations of 0.1–1.0 µg·ml-1 (much lower concentrations of F5 were tested because of a low yield in F5). Initial settlement assays showed only F5 induced

19 settlement therefore CX-control-fraction 5 (CF5) was tested in future settlement assays as the procedural control. CF5 was dissolved in Milli-Q at 100 µg·ml-1 and tested at 1.0 µg·ml-1. Milli-Q and SSW were used as the negative controls. Percent settlement was scored after 1 h (n = 10). iii) Natural versus synthetic histamine

The response of larvae to: i) natural histamine isolated using CX chromatography, ii) synthetic histamine, and iii) synthetic histamine run through the same procedure used to isolate natural histamine, were compared in settlement assays. Stock solutions of 900 µM of each histamine treatment were prepared in Milli-Q, and aliquots of the appropriate stock solution added to petri-dishes for final test concentrations of 0.9–9.0 µM. Pieces of fresh Delisea pulchra (~20 mg) and 50 µg·ml-1 of the polar extract of D. pulchra were used as the positive controls while Milli-Q and SSW were used as the negative controls. Percent settlement was scored after 1 h (n = 12).

2.2.7 Quantitative analysis of histamine in various algae

If histamine is a natural settlement cue for this urchin, I would expect Delisea pulchra, the primary host plant of new recruits of Holopneustes purpurascens, to have higher levels of histamine than other algae in the habitat. To test this, I quantified the histamine content of six species of algae from the habitat of H. purpurascens. The two- primary host-plants D. pulchra and E. radiata, and four other prominent species of algae; Amphiroa anceps and Corallina officinalis (Corallinales: Rhodophyta), Homeostrichus olsenii (Dictyotales: Phaeophyta) and Sargassum vestitum (Fucales: Phaeophyta); were collected from Bare Island (January, 2003). Five replicates of each alga were analysed, with each replicate consisting of 3 small sections taken from different parts of one thallus, which were then pooled into a single sample for analysis (2–4 g ww). A polar extract of each algal sample was prepared as described above. Polar extracts were dissolved in Milli-Q (200 µl) and acidified with 50 µl of glacial acetic acid. [α, α, β, β-d4]Histamine·2HCl (1 µg, Cambridge Isotope Laboratories, #DLM 2911) was added to each sample as the internal standard (ISTD). This choice of ISTD was based on the notion that the most suitable ISTD is a dueterated form (or

20 another isotopic form) of the analyte which should behave equally to chemical treatment (i.e., extraction and derivatisation) but have slightly different masses allowing for quantification. Strong cation-exchange solid phase extraction cartridges (50 mg, Alltech) were equilibrated with Milli-Q (5 ml) at a flow rate of 1 ml·min-1 and the sample loaded. Unbound compounds were eluted in 2ml Milli-Q (fraction 1) and another 2 ml Milli-Q (fraction 2). All retained compounds were eluted in 1 ml 30%-

NH4OH w/w (fraction 3) and dried in a Speed-Vac. Standards that contained either, 0.1, 0.5, 1.0, 5.0 or 10 µg of synthetic histamine, and 1 µg ISTD were prepared. Standards and fraction 3 samples were derivatised with heptafluorobutyric anhydride and acetic anhydride as described in section 2.2.4ii.

A DB-5MS column (15 m, 0.25 µm x 0.25 mm ID, J & W Scientific) and a packed liner (3% SP-2250, Supelco) (Smythe et al. 2002) were installed on the GC-MS instrument described in section 2.2.4ii using the same run conditions. The Mass Selective Detector was operated in selected ion monitoring mode using ions characteristic of the analyte (derivatised histamine - m/z 94, 307, 349) and the ISTD (m/z 97, 311, 353). Although the analyte and ISTD coelute in the gas chromatogram, the peaks consist of different masses for each compound and form distinct peaks in extracted ion chromatograms. Extracted ion chromatograms were used to manually integrate the area under each ion peak (which is proportional to the amount of analyte in the sample). For each standard and sample the areas of the analyte ions (m/z 94, 307, 349) were added together and the areas of the ISTD ions (m/z 97, 311, 353) were added together. The ratio of the combined areas of analyte:ISTD in standards was used to generate a standard curve. The histamine content of the samples was calculated by reference to the standard curve, and was expressed in terms of µg·g-1 (ww) of algal tissue. The difference between the wet weight (ww) and dry weight (dw) of 10 samples of each alga was used to calculate a ww-dw conversion factor for each algal type, which was used to convert histamine contents from µg·g-1 (ww) to ug·g-1 (dw).

The histamine content of different algae was compared with a 1-factor analysis of variance (ANOVA, data transformed by ln[x + 1]). We excluded Amphiroa anceps and Corallina officinalis from the analysis as no histamine was detected in these species. Bonferroni’s post-hoc test was used to determine which species differed in their

21 histamine contents (SYSTAT® 7.0 for Windows). I was concerned that one high value for Delisea pulchra may have unduly influenced the analysis but when the analysis was repeated with this value omitted the outcome was unchanged. Therefore I report the results of the initial analysis.

2.2.8 The source of the settlement cue from Delisea pulchra – settlement assay with algae treated with antibiotics

Because some marine bacteria produce histamine (Fujii et al. 1997), the identification of histamine as the settlement cue (see Results) raises the possibility that the bacterial biofilm on the surface of Delisea pulchra, rather than the alga itself, may be the source of the cue. To test this, the ability of D. pulchra to induce settlement after being treated with various antibiotics was assayed. Antibiotic treatments were adapted from previous studies where treatments were shown to be effective in reducing the abundance or diversity of surface bacteria (Xue-Wu & Gordon 1987, Aguirre-Lipperheide & Evans 1993, Johnson & Sutton 1994, Huggett et al. 2005, Huggett et al. 2006). D. pulchra (7 plants) were collected and brought back to the laboratory where portions of each plant were allocated to each of 7 treatments.

There were 6 antibiotic treatments and a procedural control. All antibiotic treatments of Delisea pulchra included a 5 min soak in a 10 % betadine-SSW solution, followed by 3 rinses in SSW and a 24 h treatment in either: 1) SSW (termed the “soak” treatment); 2) SSW containing 20 mg·l-1 streptomycin (Aldrich), 10 mg·l-1 penicillin G (Aldrich) and 10 mg·l-1 kanamycin (Aldrich, “SPK” treatment) ; 3) SSW containing 10 mg·l-1 ciprofloxacin (Bayer, “ciprofloxacin” treatment); 4) SSW after pieces of D. pulchra were gently wiped across an agar plate, before and after the 24 h soak, to physically remove bacteria (“wipe” treatment); and the combination treatments 5) “wipe + SPK”, and 6) “wipe + ciprofloxacin”. The procedural control was a 24 h soak in SSW without the initial betadine soak (“soak control” treatment). The next day, subsections of several D. pulchra plants were collected as a “fresh control” treatment and used in the settlement assay on that day. Pieces of D. pulchra (~10 mg) from each treatment were added to sterile petri-dishes and percent settlement scored after 20 h (n = 15 replicates per treatment).

2.2.9 Reanalysis of samples from Williamson et al. (2000)

Samples remaining from Williamson et al. (2000) were analysed by GC-MS for the presence of histamine. Any histamine in the old samples was isolated using cation- exchange solid phase extraction cartridges as outlined for isolating algal histamine (section 2.2.7) and derivatised with heptafluorobutyric anhydride and acetic anhydride as described in section 2.2.4ii.

2.3 Results

2.3.1 Isolation of the settlement cue in Delisea pulchra using bioassay-guided fractionation

i) HPLC fractions - NMR spectroscopy analysis and settlement assays

The polar extract of Delisea pulchra was separated into two fractions using HPLC (peak 1 and peak 2, Figure 2.1A). These were analysed by NMR spectroscopy (1-min 1H- and 30-min 13C-NMR experiments) and tested in settlement assays. Peak 1 displayed the pattern of isethionic acid (Barrow et al. 1993), as determined by 1H- and 13C-NMR spectroscopy, as well as some additional signals that were not characteristic of floridoside (see next paragraph). The 13C-NMR spectrum of peak 2 corresponded to previously published data for floridoside [∝-D-galactopyranosyl-(1-2)-glycerol] (Karsten et al. 1993). Therefore the isethionic acid and floridoside components of the F-I complex eluted separately in peak 1 and peak 2, respectively. Peak 1 induced settlement of Holopneustes purpurascens larvae whilst peak 2 did not (Figure 2.2). Four batches of peak 1 (25 µg·ml-1) induced 80–100 % settlement of larvae in five assays whereas neither of two batches of peak 2 (51 µg·ml-1) induced settlement in two assays (representative data shown in Figure 2.2). These data suggested that the F-I complex is not a settlement cue for H. purpurascens and that peak 1 (which lacked floridoside) contained the settlement cue.

23 Isethionic acid and taurine were the major compounds in peak 1 as determined by 1H- and 13C-NMR spectroscopy and comparison with synthetic samples. When isethionic acid (1–25 µg·ml-1), sodium isethionate (15–30 µg·ml-1) and taurine (1–13 µg·ml-1) were tested in settlement assays with Holopneustes purpurascens larvae, none of these compounds induced settlement (data not shown). Different combinations of these compounds were tested together (e.g., 15 µg·ml-1 of isethionic acid and taurine) in case two cues were required for settlement of larval H. purpurascens. There was no settlement in the combination treatments (data not shown). Following these results, I hypothesised that a trace compound(s) in peak 1, not yet detected by NMR analysis, was inducing settlement. To test this, a larger amount of peak 1 was collected and a much longer 13C-NMR experiment (24 h) was run on the sample. The 13C-NMR spectrum showed approximately 20 additional carbon signals not detected previously by NMR spectroscopy, indicating that additional compounds were present in peak 1 in trace amounts. Since peak 1 induced settlement of larvae of H. purpurascens, but the identified major components (isethionic acid, taurine) in peak 1 did not, this implied that one (or more) of the compounds present in trace amounts was the settlement cue. ii) Cation-exchange fractions – settlement assay

The settlement cue could not be isolated as a pure fraction using HPLC so the polar extract of Delisea pulchra was fractionated using CX chromatography (Figure 2.1B). Five CX-fractions (F) were obtained which were tested in settlement assays; only F5 induced settlement of larvae of Holopneustes purpurascens (Figure 2.3). F5 at a concentration of 1.0 µg·ml-1 induced 100 % settlement in larvae after 1 h, 0.5 µg·ml-1 induced 70 % settlement, and 0.1–0.25 µg·ml-1 did not induce settlement. There was no settlement in the control fraction CF5 (1.0 µg·ml-1) and SSW treatments (Figure 2.3).

2.3.2 Identification of the settlement cue for Holopneustes purpurascens

i) Nuclear magnetic resonance spectroscopy

1 The H-NMR (D2O) spectrum of F5 showed proton signals at δ 2.76 (2H, t, J 7.0 Hz, H2), 3.03 (2H, t, J 8.2 Hz, H1), 6.86 (1H, s 2H, imidazole H) and 7.57 (s, 1H, imidazole 13 H) (Figure 2.4A). The C-NMR (D2O) and DEPT spectra of F5 showed carbon signals at δ 25.9, 39.5 (CH2); 116.4, 136.5 (CH) and 134.0 (quaternary C). These signals 24 supported the assignment of F5 as histamine (2-[1H-imidazol-4-yl]-ethylamine, MW 111.15, structure shown in Figure 4.4A) as confirmed by the 1H-NMR spectrum of synthetic histamine (Figure 2.4B). The structure of F5 was further confirmed by a high- field two-dimensional 1H-15N HMBC NMR experiment in which the methylene triplet at 2.76 ppm showed two three-bond correlations to the ethylamine NH2 group and the imidazole nitrogen. The identity of F5 was further confirmed by a spiking experiment (Figure 2.4C). All F5 signals increased in intensity and no additional signals were detected, confirming the identity of F5 as histamine. ii) Gas chromatography – mass spectrometry

The identity of putative histamine (F5) isolated from Delisea pulchra was confirmed using GC-MS. The retention times (rt) of the heptafluorobutyrlacyl derivative of putative histamine (rt = 9.728 ± 0.0045, mean ± SD, n = 5) and synthetic histamine (rt = 9.732 ± 0.0045, mean ± SD, n = 5) were nearly identical suggesting it was the same compound. The electron impact ion-spectra of both derivatised compounds displayed the same major fragment ions (m/z 54, 69, 81, 94, 138, 169, 226, 307, 349; Figure 2.5A- B) and overall fragmentation pattern confirming they were the same compound. The electron impact ion-spectra for derivatised histamine matched that which is reported in the literature (Barancin et al. 1998). iii) Matrix-assisted laser desorption/ionisation – time-of-flight mass spectrometry

The elemental formula of putative histamine isolated from Delisea pulchra was confirmed by accurate mass measurements using MALDI-TOFMS. The measured accurate mass of the putative protonated histamine molecular ion [M+H]+ was 112.08878 ± 0.0026 (n = 10, mean ± SD) and the measured mass for synthetic histamine was 112.08853 ± 0.0025 (n = 10, mean ± SD). The measured masses of the two samples were only different by 2.2 ppm. These values were different from the calculated monoisotopic mass for protonated histamine (112.08692 - elemental formula

C5H10N3) by only 15 ppm for synthetic protonated histamine and 17 ppm for putative protonated histamine. This is most likely due to measurement bias introduced by the very different chemical properties of histamine, glycine and creatinine (the internal calibrants).

25 An elemental calculator was used to generate all possible elemental formulae with a mass of approximately 112.08878. The nearest other candidate was C6H10NO at 112.07569 with a difference of 117 ppm from the measured mass of putative protonated histamine. This difference was much higher than 17 ppm (difference of measured mass for putative histamine relative to calculated mass for C5H10N3) confirming that the putative protonated histamine had the elemental formula of C5H10N3.

2.3.3 The response of Holopneustes purpurascens larvae to natural and synthetic histamine

Natural histamine isolated from Delisea pulchra using CX chromatography, synthetic histamine and synthetic histamine eluted from CX resin all resulted in very similar responses in larvae when assayed concurrently (Figure 2.6). Over 80 % of larval Holopneustes purpurascens had settled within an hour of incubation in 4.5 and 9 µM natural and synthetic histamine. The lowest test concentration of synthetic histamine that consistently induced rapid settlement of all larvae was 4.5 µM (in ten separate assays). Larvae exhibited a more variable response to 0.9 and 2.3 µM histamine both within and across different batches (Figure 2.6). Up to 80 % of larvae settled in response to 0.09–0.45 µM synthetic histamine but only after long incubation times (up to 96 h) and/or as larval age increased to 13–21 d (data not shown but see Figure 4.1 for similar experiments).

2.3.4 Quantitative analysis of histamine content in algae

The histamine content of six algal species was determined by GC-MS (Table 2.1). Delisea pulchra, the alga on which new recruits of Holopneustes purpurascens are found (Williamson et al. 2000), had the highest histamine content of all algae surveyed. Histamine was not detected in any samples of Amphiroa anceps or Corallina officinalis. The histamine content of D. pulchra, Ecklonia radiata, Homeostrichus olsenii and

Sargassum vestitum differed significantly from each other (F3, 16 = 9.903, p < 0.001). Pairwise comparisons showed that the histamine content of D. pulchra (76.83 ± 42.64 µg·g-1 dw) was significantly higher than the histamine content of E. radiata (8.70 ± 6.87

26 µg·g-1 dw, p = 0.009), S. vestitum (1.75 ± 1.60 µg·g-1 dw, p = 0.002) and H. olsenii (1.20 ± 0.43 µg·g-1 dw, p = 0.002). The amount of histamine in different D. pulchra plants was highly variable ranging from 12–222 µg·g-1 dw of algal tissue. The histamine content of E. radiata was variable with undetectable levels in 2 plants but up to 32.2 µg·g-1 dw of algal tissue in another. Likewise, histamine was not detected in 3 of the S. vestitum samples however another sample contained 7.4 µg·g-1 dw of algal tissue. The H. olsenii plants analysed showed consistently low levels of histamine ranging from 0.2–2.2 µg·g-1 dw of algal tissue.

2.3.5 The source of the settlement cue from Delisea pulchra - settlement assay with algae treated with antibiotics

Delisea pulchra that had been treated with antibiotics induced over 60 % settlement of Holopneustes purpurascens larvae which was equivalent (or greater) than levels of settlement in control D. pulchra treatments (Figure 2.7). SSW did not induce settlement of H. purpurascens larvae.

2.3.6 Reanalysis of samples from Williamson et al. (2000)

Several samples remaining from the previous study were analysed by GC-MS for the presence of histamine. Histamine was detected in F-I complex fractions from Delisea pulchra (1.5–46 µg·mg [sample]-1) (Figure 2.8) in a synthetic F-I complex sample (0.35 µg·mg [sample]-1) and in a batch of floridoside used to make the synthetic complexes (0.45 µg·mg [sample]-1).

A – HPLC Polar extract 50 mg·ml-1 (MQ)

Adsorbosil C18 column FR - 1 ml·min-1 (MQ)

Peak 1 Peak 2 (rt-2.7 min) (rt-3.4 min)

NMR spectroscopy analysis & settlement assays

B – CX Chromatography Polar extract 200-400 mg·ml-1 (MQ)

AG50W-X2 resin FR - 2 ml·min-1 (MQ)

MQ MQ MQ NH3 NH3 pH-10 3% 30% F1 F2 F3 F4 F5

NMR spectroscopy analysis & settlement assays

Figure 2.1 A schematic diagram illustrating bioassay-guided fractionation of the polar extract of Delisea pulchra using either A. reversed-phase HPLC or B. cation-exchange (CX) chromatography. MQ = Milli-Q, FR = flow rate, rt = retention time, F = fraction.

28 100

40 Percent settlement Percent 20

0 " ea s (A) (B) MQ plex 1 1 SSW Deli ak Peak 2 eak com P Pe "F-I Treatment

Figure 2.2 Percent settlement of larvae of Holopneustes purpurascens after 18 h incubation with fresh Delisea pulchra (~20 mg) or HPLC peak fractions of the polar extract of the alga. Peak 1 (batch A or B) was tested at 25 μg·ml-1, peak 2 was tested at 51 μg·ml-1, and a floridoside-isethionic acid complex sample (“F-I complex”) from Williamson et al. (2000) was tested at 76 μg·ml-1. Sterile seawater (SSW) and Milli-Q were included as the negative controls (n = 10).

100

40 Percent settlement

0 ) ) ) .0 W (1 (0.5 (0.1) (1.0 S 1 (50) 2 (50) 3 (50) 4 (50) S PE (50) F F F F F5 F5 5 F CF5

Fraction

Figure 2.3 Percent settlement of larvae of Holopneustes purpurascens after 1 h incubation with the polar extract of Delisea pulchra (PE) and cation-exchange fractions (F) of the PE. The different test concentrations of each treatment are shown in brackets (μg·ml-1); note the lower concentrations for F5 and the procedural control (CF5). Sterile seawater (SSW) was used as the negative control (n = 10).

30 3.0479 3.0228 3.0002 2.8295 2.8057 2.7831 7.6104 7.6066 6.8950

7.8 7.6 7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6 5.4 5.2 5.0 4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 (ppm)

1 Figure 2.4A The H-NMR (D2O) spectrum of F5 (~3mg), the fraction isolated from the polar extract of Delisea pulchra using cation-exchange chromatography which induced settlement of Holopneustes purpurascens larvae at 0.5-1.0 μg.ml-1, showing proton signals at δ 2.8 (2H, t, J 7.0 Hz, H2), 3.00 (2H, t, J 8.2 Hz, H1), 6.9 (1H, s 2H, imidazole H) and 7.6 (s, 1H, imidazole H). Proton signal at δ 4.8 is D2O.

31 7.5728 7.5690 6.8335 2.9098 2.9061 2.8835 2.8622 2.8596 2.7417 2.7191 2.6952

7.8 7.6 7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6 5.4 5.2 5.0 4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 (ppm)

1 Figure 2.4B The H-NMR (D2O) spectrum of synthetic histamine (~3 mg) showing proton signals showing proton signals at δ 2.8 (2H, t, J 7.0 Hz, H2), 3.00 (2H, t, J 8.2 Hz, H1), 6.9 (1H, s 2H, imidazole H) and 7.6 (s, 1H, imidazole H). Proton signal at δ 4.8 is D2O.

32 2.9977 2.9739 2.9513 2.7994 2.7755 2.7530 7.5941 6.8736

7.8 7.6 7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6 5.4 5.2 5.0 4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 (ppm)

1 Figure 2.4C The H-NMR (D2O) spectrum of F5 (~3mg) spiked with an equal amount of synthetic histamine, showing proton signals at δ 2.8 (2H, t, J 7.0 Hz, H2), 3.00 (2H, t, J 8.2 Hz, H1), 6.9 (1H, s 2H, imidazole H) and 7.6 (s, 1H, imidazole H) with no additional signals. Proton signal at δ 4.8 is D2O.

A - Putative histamine Abundance

Average of 9.742 to 9.753 min.: 0101001.D 94 14000 81

12000

10000

8000

307 6000

138 4000 169 54 2000 226 109 349 122 288 67 268 0 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 m/z-->

B - Synthetic histamine Abundance

Average of 9.732 to 9.754 min.: 0105005.D 94 35000 81

30000

25000

20000

15000 307

10000 138 169 54 5000 226 109 349 68 122 288 197 268 0 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 m/z-->

Figure 2.5 The mass spectrum of A. derivatised putative histamine isolated from Delisea pulchra and B. derivatised synthetic histamine, showing major ions used in quantitative analysis of histamine content of algae (m/z 94, 307, 349)

100

40 Percent settlement

a E 9 3 5 9 9 3 5 9 9 3 5 9 ...... W e - - - P 0 2 4 0 2 4 0 2 4 s S i t l ------n - - - X a S e t t t y n n n C N X X X a a a S D y y y n C C C N N N S S S y n n n S y y y

S S S Treatment

Figure 2.6 Percent settlement of larvae of Holopneustes purpurascens after 1 h incubation with fresh Delisea pulchra (~20 mg), its polar extract (PE, 50 μg·ml-1) and 0.9 – 9.0 μM of natural histamine isolated from D. pulchra (Nat), synthetic histamine (Syn) or synthetic histamine eluted from cation-exchange resin (Syn-CX). Sterile seawater (SSW) was used as a negative control. Data from two experiments using different batches of larvae are shown (n = 12, black and grey bars are different batches of larvae).

100

40 Percent settlement Percent

ol ol k K K in o r r pe W t P P pr i oa ac i S s S S c w on ont S c c ox fl e+ h k ipe+ o p s i e oa w pr w r i f s c treatment

Figure 2.7 The percent settlement of larvae of Holopneustes purpurascens after 20 h incubation with Delisea pulchra subjected to various antibiotic treatments. All antibiotic treatments included a 5 min soak in a 10 % betadine solution, followed by 3 rinses in sterile seawater (SSW) and a 24 h treatment in either SSW (“soak”); SSW containing streptomycin (20 mg·l-1), penicillin G (10 mg·l-1) and kanamycin (10 mg·l-1, “SPK”); or SSW containing ciprofloxacin (10 mg·l-1, “ciprofloxacin”). Other treatments involved wiping pieces of D. pulchra across an agar plate gently, to physically remove bacteria, before and after a 24 h soak in SSW (“wipe”), SSW containing SPK (“wipe + SPK”), or SSW containing ciprofloxacin (10 mg·l-1, “wipe + cipro”). D. pulchra soaked in SSW for 24 h (without betadine soak) was the procedural control (“soak control”), fresh D. pulchra was used as a positive control (“fresh control”) and SSW was used as a negative control (n = 15).

A Abundance

TIC: 1101014.D

8000

7500

7000

6500

6000

5500

5000

4500

4000

3500 9.78 3000

2500

2000

1500

6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 Time--> B Abundance

Ion 94.00 (93.70 to 94.70): 1101014.D I 9.81 450

400

350

300

250

200 9.55 9.60 9.65 9.70 9.75 9.80 9.85 9.90 9.95 Time--> Abundance

Ion 307.00 (306.70 to 307.70): 1101014.D II 9.82 700

600

500

400

300

200 9.55 9.60 9.65 9.70 9.75 9.80 9.85 9.90 9.95 Time--> Abundance

III Ion 349.00 (348.70 to 349.70): 1101014.D 9.81 340 320 300 280 260 240 220 200 180 9.55 9.60 9.65 9.70 9.75 9.80 9.85 9.90 9.95 Time--> Figure 2.8 A. Selected ion chromatogram of derivatised ‘F-I complex’ sample showing derivatised histamine in the chromatogram at rt-9.78 min. B. Extracted ion chromatograms show individual peaks for ions 94 (I), 307 (II), and 349 (III).

Table 2.1 The histamine content of six algal species [μg.g –1 dw (dry weight), mean ± SE, n = 5] was significantly different (ANOVA - F3, 16 = 9.903, p < 0.001). * Indicates species where histamine content differs significantly from D. pulchra (Bonferroni’s pairwise comparisons, p < 0.01). nd - not detected.

Species μg.g -1 dw (mean ± SE, n = 5)

Delisea pulchra 76.83 ± 42.64 Ecklonia radiata 8.70 ± 6.87* Sargassum vestitum 1.75 ± 1.60* Homeostrichus olsenii 1.20 ± 0.43* Corallina officinalis nd Amphiroa anceps nd

2.4 Discussion

Few studies have definitively characterised settlement cues (reviewed by (Hadfield & Paul 2001, Steinberg et al. 2001). Williamson et al. (2000) reported on one such putative characterised cue, the floridoside-isethionic acid (F-I) complex, isolated from the red algal host Delisea pulchra which induced settlement of larval Holopneustes purpurascens. New evidence presented in this chapter has shown that histamine (not the F-I complex) is a natural inducer of settlement in larval H. purpurascens. The settlement cue was isolated from the polar extract of D. pulchra using bioassay-guided fractionation by cation-exchange chromatography. The isolated compound at 0.5 µg·ml-1 induced settlement of 80–100 % of larvae by 1 hr. The settlement cue was identified as histamine using NMR spectroscopy which was confirmed by GC-MS and MALDI-TOFMS. The response of larvae to synthetic histamine in settlement assays mirrored their response to natural histamine isolated from D. pulchra. D. pulchra, the primary plant on which new recruits of H. purpurascens are found (Williamson et al. 2000), had the highest average histamine content approximately an order of magnitude higher than other algae surveyed. Seawater collected in situ adjacent to D. pulchra plants (Delisea-SW) induced rapid settlement of larval H. purpurascens in Williamson et al. (2000). Together, these findings support my proposal that histamine released from D. pulchra is a settlement cue for larval H. purpurascens.

Reanalysis of samples from Williamson et al. (2000) provides an explanation for the incorrect conclusion that the F-I complex is a settlement cue for larvae of Holopneustes purpurascens. The F-I complex was isolated from the polar extract of Delisea pulchra using reversed-phase HPLC and eluted as a single peak when methanol was used as the mobile phase (Williamson et al. 2000). 13C-NMR spectroscopic analysis of this peak showed only 13C-signals for floridoside and isethionic acid (Williamson et al. 2000). However, trace amounts of histamine were also present but not detected, as levels were below the limit of detection for 13C-NMR spectroscopy. Histamine elutes in the first peak from reversed-phase (C18) columns regardless of the mobile phase used so any histamine present in polar extracts of D. pulchra used by Williamson et al. (2000) would

have co-eluted with the F-I complex fraction. Consequently the ‘F-I complex’ samples also contained histamine, detected here using GC-MS, and induced settlement of H. purpurascens larvae. Although a synthetic F-I complex induced rapid settlement of H. purpurascens larvae (Williamson et al. 2000), not all batches induced settlement (R. de Nys, pers. obs.). The synthetic F-I complexes were made using natural floridoside isolated from D. pulchra and synthetic isethionic acid. Floridoside used to make synthetic F-I complex was contaminated by histamine and thus the synthetic F-I complex induced settlement. Confirming this, histamine was detected by GC-MS in a floridoside sample (used for preparation of the synthetic complex) and a synthetic F-I complex sample prepared by Williamson et al. (2000). In summary, histamine was present in trace amounts in the ‘F-I complex’ samples that induced settlement of larval H. purpurascens in Williamson et al. (2000) and histamine was the inductive compound in these samples.

The finding that histamine is a natural settlement cue for Holopneustes purpurascens is of considerable interest in the context of linking ecological patterns with physiological mechanisms. Histamine is a biogenic amine produced by the decarboxylation of the amino acid histidine. It is one of five primary biogenic amines in invertebrates, along with serotonin, octopamine, dopamine and tyramine (Blenau & Baumann 2001). Biogenic amines, monoamine compounds formed by the decarboxylation of amino acids, play critical roles in the physiology and behaviour of invertebrates by acting as classical neurotransmitters, neuromodulators and neurohormones (Katz 1995, Beltz 1999). For example, dopamine activates hunting behaviour in an opistobranch mollusc (Norekyan & Satterlie 1993) and serotonin controls aggressive behaviour in crustaceans (Huber et al. 1997). The photoreceptors in all classes of arthropod eyes are histaminergic, that is, they synthesise histamine and use it as their neurotransmitter (Stuart 1999). Also, histamine is thought to be an inhibitory neurotransmitter in the stomatogastric and cardiac ganglia, as well as in the sensory system of lobsters (Claiborne & Selverston 1984, Bayer et al. 1989, Hashemzadeh-Gargari & Freschi 1992). Importantly, in the context of this study, histamine directly gates a chloride channel in the receptor cells of the olfactory pathway of lobsters (McClintock & Ache 1989). Fast neurotransmitters directly gate ion channels leading to fast behavioural and physiological outcomes. I have observed that histamine

(~10 μM) induces rapid settlement of all H. purpurascens after 1 h. This fast response is consistent with the notion that larval H. purpurascens have specific receptors that bind histamine which act directly on ion channels (i.e., ionotropic receptors), leading to rapid metamorphosis of these larvae.

Neurotransmitters, or their precursors, have been suggested to mimic the function of natural settlement cues (Morse 1985, Bonar et al. 1990), however, this argument has been disputed (Pawlik 1990). The best-known example is the gamma-aminobutyric acid (GABA)-mimetic peptide(s) proposed by Morse and colleagues as a settlement cue for abalone present on the surface of crustose coralline algae (Morse et al. 1979, Morse et al. 1984). Another example comes from oyster larvae, where L-3,4-dihydroxyphenylalanine (L-DOPA) induced stereotypical searching behaviour, while epinephrine and norepinephrine induced settlement (Coon et al. 1985). Endogenous levels of neurotransmitters, and their precursors, also appear to modulate the behavioural and physiological processes accompanying settlement (Coon & Bonar 1987, Pires et al. 2000). Our findings show that a naturally produced neurotransmitter is in fact a settlement cue for Holopneustes purpurascens larvae, a phenomenon that may be widespread in the marine environment.

The finding that histamine, rather than the F-I complex, is a settlement cue for Holopneustes purpurascens potentially complicates the previous interpretations of the relationship between settlement cues and the demography of this sea urchin (Williamson et al. 2000, Williamson et al. 2004). Histamine, a simple breakdown product of the amino acid histidine, may be broadly distributed across the natural habitat of H. purpurascens; for example, in algal and animal tissue, and in bacterial communities living on their surfaces. For histamine to be an ecologically relevant settlement cue its distribution in the natural habitat must relate to the recruitment patterns of H. purpurascens. This was in fact the case. The histamine content of the algae surveyed was consistent with the recruitment patterns of the organism observed by Williamson et al. (2000), with much higher levels of histamine measured in Delisea pulchra, a primary host plant for new recruits.

Delisea pulchra had the highest average histamine content ranging from 12–222 µg.g-1 dw. Levels of histamine also varied for Ecklonia radiata with concentrations ranging from 0–32 µg.g-1 dw. Since only subsections of plants (not whole plants) were extracted, these results may reflect within-plant variation and/or within-species variation. Future histamine analyses will extract whole plants, as well as specific regions of thalli, to directly test this. The low (or absent) levels of histamine typically measured in E. radiata samples may explain why pieces and extracts of E. radiata did not induce settlement in Williamson et al. (2000). However, we have observed that some pieces of E. radiata do induce settlement of H. purpurascens, which is consistent with the variation in levels of histamine in E. radiata measured here. Histamine was not detected in the coralline turf algae, Corallina officinalis and Amphiroa anceps, although they induce settlement of larval H. purpurascens (Williamson et al. 2000, Swanson et al. 2004) and provide a habitat for new recruits (Figure 3.2). Larger samples of A. anceps (up to 180 g wet weight) were extracted and no histamine was detected. The coralline algae may produce a different settlement cue for H. purpurascens, or histamine may only be produced and released in situ, for example, by surface-associated bacteria. The nature of the settlement cue from coralline algae is addressed in Chapter 3.

Finally, possible sources of histamine in Delisea pulchra include the host alga or the surface-associated bacterial community (biofilm), or both. D. pulchra treated with various antibiotics still induced high levels of settlement in Holopneustes purpurascens, suggesting that the host alga produces the histamine. A bacterial source of histamine is also possible, however, as a known histamine-producing bacterium, Photobacterium phosphoreum (Fujii et al. 1997) may be a constituent of the microbial community on local algal species (Huggett et al. 2006). If histamine-producing bacteria are colonising algal surfaces within the habitat, then it is possible that they produce and release histamine to seawater, which could lead to the induction of settlement of larval H. purpurascens.

CHAPTER THREE

In situ quantification of histamine in the habitat and recruitment of Holopneustes purpurascens*

* This chapter is published as a Feature Article Swanson et al. (2006) Marine Ecology Progress Series 314:1-14 (Appendix 2)

3.1 Introduction

The importance of chemical cues in triggering the settlement of marine invertebrate larvae has been recognised for a wide range of species. However, after 60 yr of research, there is not a single settlement cue for invertebrate larvae which has unequivocally been: (1) structurally characterised, (2) quantified in situ, and (3) related to recruitment patterns of the organism by quantifying variation in the distribution of the settlement cue in the habitat. We recently reported that the low-molecular-weight (LMW) water-soluble amino-compound histamine is a naturally occurring settlement cue for the Australian sea urchin Holopneustes purpurascens (Chapter 2) (Swanson et al. 2004). Histamine was isolated from the preferred host plant of new recruits Delisea pulchra (Williamson et al. 2000) using cation-exchange chromatography, and its identity confirmed by spectroscopy and mass spectrometry (Chapter 2) (Swanson et al. 2004). An initial analysis of the histamine content of algae in the habitat, representing the first field-based measurements of a naturally occurring settlement cue, found the histamine content of D. pulchra to be 10-fold greater than in co-occurring algae (Swanson et al. 2004).

In the first study of its kind, variation in the distribution of new recruits of Holopneustes purpurascens was recorded and compared with quantitative variation in the distribution of its settlement cue (histamine) in the habitat. A range of algae, including known host algae (Williamson et al. 2000) and co-occurring species, were collected each month for

2 yr and searched for new recruits (test diameter ≤ 5mm). The histamine content of these algae (Delisea pulchra, Eckonia radiata, Sargassum vestitum and Homeostrichus olsenii) was quantified quarterly to assess for any temporal variation. The coralline algae Amphiroa anceps and Corallina officinalis were included in the recruitment survey but were excluded from the quantitative analyses as histamine was not detected in these algae previously (Chapter 2) (Swanson et al. 2004). Algae known to host H. purpurascens were tested against larvae in settlement assays in the laboratory. D. pulchra, which produced far more histamine than other algae, was sampled more frequently to assess short-term variation. We confirmed that histamine was present in seawater collected in situ adjacent to D. pulchra by quantitative analysis and by assaying for settlement-inducing activity against H. purpurascens larvae. The localisation (i.e, surface-bound or waterborne) and source of the (unknown) settlement cue from A. anceps was investigated.

3.2 Materials and Methods

3.2.1 Study site

All urchins, algae and seawater used in this study were collected from sub-littoral habitats (1–3 m depth) at Bare Island (33° 59' 38" S, 151° 14' 00" E) at the north head of Botany Bay, Sydney, Australia. A detailed description of this habitat is reported elsewhere (Wright & Steinberg 2001, Williamson et al. 2004).

3.2.2 Recruitment Survey

Holopneustes purpurascens is an “arboreal” sea urchin, typically living off the benthos enmeshed in the fronds/laminae of seaweeds. A range of algae were collected from Bare Island each month for two years from December 2002 – November 2004 (excluding June 03/04 and February 04) and searched for new recruits (test diameter ≤5mm) to determine recruitment patterns of H. purpurascens. A test diameter of ≤5mm was chosen arbitrarily even though recruits of this size are probably several months old.

However, the size of each new recruit (to nearest 0.1 mm) was recorded which allowed monitoring of new recruits of different size classes (≤2 mm, 2.1–3.0mm, 3.1–4.0mm, 4.1–5.0mm). Algae sampled included the main host plants of H. purpurascens at Bare Is., Ecklonia radiata (Laminariales: Phaeophyta) and Delisea pulchra (Bonnemaisonales: Rhodophyta), as well as the common and co-occurring species, Sargassum vestitum (Fucales: Phaeophyta), Homeostrichus olsenii (Dictyotales: Phaeophyta), Corallina officinalis and Amphiroa anceps (Corallinales: Rhodophyta).

Five plants (or parts thereof) of Ecklonia radiata, Delisea pulchra, Sargassum vestitum, Homeostrichus olsenii were collected, or 10 x 10 cm2 (n = 5) of the turf algae Corallina officinalis and Amphiroa anceps (in an attempt to sample approximately equal amounts of each algal type). Algae were rinsed in fresh water and the epifauna collected. The algae were blotted dry with paper towel and the wet weight (ww) recorded. New recruits were measured using a dissecting microscope and identified by comparison to juveniles of Holopneustes purpurascens reared from larvae in the laboratory (Figure 3.1). The total number of new recruits found each month on each alga (pooling data from the coralline turf algae) was standardised to number found per 100 g of algae sampled (# per 100 g ww). Recruitment was analysed by a 1-factor ANOVA (data transformed by x+1) using months as replicates (n = 21). Bonferroni’s post-hoc test was used to determine which treatments differed significantly at p = 0.05.

3.2.3 Larval culture

Larvae for the algal assay were cultured from gametes obtained using a different method to that outlined in section 2.2.5 as I was obtaining low numbers of larvae and sought to improve yields. Instead of injecting urchins with 0.5M KCl to induce the release of spawn, ovaries and testes were removed gently from opened tests of at least 2 females and 2 males. Ovaries were swirled around with a pipette in autoclaved SW with antibiotics added (22 mg·l-1 penicillin G and 37 mg·l-1 streptomycin sulfate, SSW) in a glass dish causing ripe eggs to shed into seawater (Dr. Valerie Morris, University of Sydney, pers. commun.). Testes were placed in petri-dishes with SSW, which became

milky with sperm. Ripe eggs were fertilised with sperm as described in section 2.2.5. Embryos and larvae were cultured as described in section 2.2.5.

Larvae for the seawater assays were obtained from adults that spawned naturally in buckets. Adult Holopneustes purpurascens (30-40) were collected on SCUBA and returned to the laboratory in ambient SW. Urchins were held in a CTR (19°C, 12-h light/12-h dark regime) in 2 x 20 l buckets containing ambient SW with continuous air. Urchins were transferred to clean buckets with fresh ambient SW at least every other day and were maintained until they spawned and died. Buckets were checked each morning for spawn. Air was stopped to allow embryos to float to the surface where they were gently collected in a 200 ml glass dish and poured into 3 x 2 l sterile beakers. Embryos were rinsed 3 times in SSW and hatched as larvae later that day. Larvae were cultured as described in section 2.2.5.

3.2.4 Settlement assays testing host algae

All settlement assays with competent larvae were done with 6-d-old larvae in a CTR (19°C, 12 h light/12 h dark regime) in 36 mm sterile petri-dishes and 4 ml SSW under static conditions. Dishes containing 10 μM histamine or SSW were included in all assays as positive and negative controls, respectively. Replicates were randomly assigned among treatments and larvae were added once all petri-dishes were prepared.

A range of host algae were assayed against larvae of Holopneustes purpurascens to test for a settlement response. We tested Delisea pulchra, Ecklonia radiata, Homeostrichus olsenii, and Amphiroa anceps as these species host recruits, juveniles or adults of H. purpurascens (Williamson et al. 2000). Plants of D. pulchra, fouled and unfouled E. radiata, H. olsenii, and A. anceps were collected and a piece of alga, or fouling (epiphytes) from the surface of E. radiata fronds (20-30 mg ww), were placed in dishes (n = 10). Five larvae were added to each dish and percent settlement was recorded at 1 and 24 h. The response of larvae (proportions) to different algae in the settlement assay was analysed by univariate repeated measures ANOVA. I compared the effect of treatments within each level of time using planned comparisons. First I compared the

effect of inductive algae (D. pulchra, A. anceps and H. olsenii), as I believed a priori that these treatments were unlikely to be different. Treatments in which there was no significant difference were pooled and tested against the non-inductive kelp (E. radiata). Similarly, the effect of fouled E. radiata and epiphytes were compared, which did not differ significantly at p = 0.05, therefore pooled treatments were tested against unfouled E. radiata.

3.2.5 Settlement assays testing seawater collected in situ adjacent to algae

In order to test whether settlement cues for Holopneustes purpurascens were present in seawater as leachate from algae, we collected seawater adjacent to and away from algae in Jul-02, Jan-03 and Nov-03, to test against competent and aged larvae of H. purpurascens in settlement assays. Seawater was collected by sterile syringe, placing the tip within 5 mm of algal fronds and drawing 10 ml of seawater into the barrel. One sample was collected from 10 individual plants of Delisea pulchra (Delisea-SW) and Ecklonia radiata (Ecklonia-SW), or from 10 patches of Amphiroa anceps (Amphiroa- SW). Control seawater was collected at the sea surface approximately 2 m away from any macroalgae (Surface-SW). Samples were stored on ice and filtered (0.22 μm) upon return to the laboratory. Delisea-SW, Ecklonia-SW, Amphiroa-SW and Surface-SW samples (n = 10, 4 ml) were tested in settlement assays with competent larvae on the day of collection (except for samples from Jan-03 which were frozen for 1 month before testing as larvae were not available). In the Jul-02 and Jan-03 assays I added 1 competent larva per dish (n = 10) and in Nov-03 we added 3 competent larvae per dish (n = 5), due to low yields of larvae at all of these times. Percent settlement was scored after 24 h.

The same seawater samples from Jul-02, Jan-03 and Nov-03 were thawed and tested against older Holopneustes purpurascens larvae (3-4 wk post-fertilisation) with 10 larvae added per dish (n = 10). Percent settlement was scored at varying times; 24 and 44 h in the assay of Nov-03 samples, 24 h in the assay of Jul-02 samples, and at 72 h in the assay of Jan-03 samples; with older larvae. The response of older larvae (proportions) to in situ seawater samples from Jul-02, Jan-03 and Nov-03 were analysed

by univariate repeated measures ANOVA. Planned comparisons were used to compare the effects of in situ seawater on larval settlement. For all assays, I first compared the effect of Surface-SW and SSW as I believed a priori that these treatments were unlikely to be different. There was no significant difference hence these treatments were pooled (Control-SW) and tested against each seawater treatment collected in situ adjacent to algae (Delisea-SW, Ecklonia-SW or Amphiroa-SW).

3.2.6 Temporal analysis of the histamine content of algae

I quantified the histamine content of Delisea pulchra, Ecklonia radiata, Homeostrichus olsenii and Sargassum vestitum over four seasons (Sep-03, Jan-04, Apr-04 and Jul-04) using a modified GC-MS analysis which had greater sensitivity than the previous analysis (Chapter 2) (Swanson et al. 2004). The coralline turf algae were excluded from these analyses because histamine was not detected in several previous extractions and analyses (Chapter 2) (Swanson et al. 2004). D. pulchra, E. radiata, H. olsenii and S. vestitum (n = 5) collected in Sep-03, Jan-04, Apr-04 and Jul-04 as part of the recruitment surveys were freeze-dried for histamine analysis. D. pulchra (5 plants) was also collected in successive months Sep-Oct-Nov03 and Jul-Aug-Sep04 in order to assess the short-term variation in the histamine content. A polar extract of each algal sample was prepared as described in section 2.2.7 except that freeze-dried algae were extracted instead of wet algae, and the internal standard (ISTD) for quantitative histamine analysis was added to the methanol extract of each algal replicate before drying it. The addition of a known amount of ISTD at the beginning of sample work-up allows for the most accurate quantification of the analyte (histamine). The ISTD was

[α, α, β, β-d4]Histamine·2HCl (Cambridge Isotope Laboratories #DLM 2911) and 100, 10, 3 and 3 μg of ISTD was added to the methanol extracts of D. pulchra, H. olsenii, E. radiata and S. vestitum, respectively.

Polar extracts of algal samples were dissolved in Milli-Q (200 µl) and acidified with 50 µl of glacial acetic acid. Strong cation-exchange solid phase extraction cartridges (50 mg, Alltech) were equilibrated with Milli-Q (5 ml) at a flow rate of 1 ml·min-1 and the sample loaded. Unbound compounds were eluted in 5 ml Milli-Q and discarded, then

all retained compounds were eluted in 200 µl of 30 %-NH4OH and dried in a Speed- Vac. Two sets of standards were prepared that contained synthetic histamine and ISTD (0.1–10 µg histamine and 3 µg ISTD, or 10–250 µg histamine and 100 µg ISTD). Standards were run through the cation-exchange cartridges as per the algal samples. Extracted standards and algal samples were derivatised with heptafluorobutyric anhydride and acetic anhydride as described in section 2.2.4ii.

A DB-5MS column (15 m, 0.25 µm x 0.25 mm ID; J & W Scientific) and a packed liner (3 % SP-2250, Supelco) were installed on the (EI) GC-MS instrument described on p.24 using the same run conditions. Histamine was quantified in samples as described in section 2.2.7 except that only 2 ions (instead of 3) were monitored by the mass selective detector in selected ion monitoring mode (derivatised histamine - m/z 307, 349 and derivatised ISTD - m/z 311, 353). Better detection of histamine and ISTD in these samples (compared to samples analysed in Chapter 2) allowed automatic (instead of manual) integration of peaks using software (HP Chemstation G1034C Version C.03.00). The statistical analysis of temporal variation of algal histamine levels used a 2-factor (alga=fixed factor, month=random factor) ANOVA (data transformed by ln[x + 1]) excluding data on Sargassum vestitum from the analysis (this alga was not present in the habitat in April). The short term temporal variation in histamine content of Delisea pulchra was analysed by a 1-factor ANOVA (data transformed by ln[x + 1]).

3.2.7 Variation in histamine within the thallus of Delisea pulchra and Ecklonia radiata

Variable results were often obtained in settlement assays when testing fresh pieces of Delisea pulchra and Ecklonia radiata against larvae, suggesting that there may be some within-plant variation of histamine content. To test this hypothesis, different regions of the thallus of D. pulchra and E. radiata were extracted for histamine analysis. D. pulchra (n = 5) were collected in Sep-04 and 1 branch of each plant was sectioned into tip (upper 1 cm), mid (top half of remaining branch), and base (lower half of remaining branch). The histamine content of the tip, mid and base (n = 5) was determined using the methods outlined in section 3.2.6 except that 50 μg ISTD was added to the methanol extracts of tip, mid and base sections (100 μg ISTD for whole plant). Unfouled and

fouled E. radiata (n = 3) were collected in Sep-04 and the unfouled plants were divided into primary (1°) and secondary (2°) laminae. Fouling epiphytes were scraped from the surface of the 2° laminae of the fouled plants and blotted with tissues. The histamine content of each region was determined as described in section 3.2.6, except that 5, 10 and 100 μg ISTD was added to the methanol extracts of the 1° and 2° laminae and the fouling epiphytes, respectively. The within plant variation of histamine content of D. pulchra and E. radiata was analysed by a 2-factor randomised block (algal part = fixed factor, plant = blocking factor) ANOVA (data transformed by ln[x + 1]).

3.2.8 The histamine concentration of seawater in the habitat

To test for the presence of histamine in seawater as leachate from algae we collected seawater in Jul-02, Jan-03 and Nov-03, adjacent to and away from algae for testing in settlement assays and histamine analysis. The seawater samples from Jan-03 and Nov- 03 were unfortunately used up completely in settlement assays and method development therefore only samples from Jul-02 were available for analysis of histamine content. Microcon-SCX adsorptive micro-concentrators (Mic-SCX, Millipore) were used to extract histamine from seawater samples. Approximately 2 ml of each seawater sample (n = 10 collected adjacent to each plant) remained after retesting in settlement assays, to which 100 ng ISTD was added. A duplicate set of standards were prepared in sterile seawater containing either, 25, 10, 5, 1 or 0 ng of synthetic histamine and 500 ng ISTD.

Samples and standards were acidified with 10 μl glacial acetic acid. Mic-SCX membranes were washed with methanol (500 μl) and then Milli-Q (500 μl) by centrifuging for 15 s at 7000 g and discarding the filtrate. Samples (500 μl) were applied to Mic-SCX and centrifuged at 1200 g for 1 min, discarding the filtrate. This binding step was repeated 4 times until the total sample/standard was applied. Mic- SCX were washed with 10 mM HCl by centrifuging at 1200 g for 1 min and discarding the filtrate. A new vial was used to collect bound compounds, which were eluted with two 50 μl applications of freshly prepared desorption reagent (500 μl methanol, 400 μl

Milli-Q and 100 μl NH4OH) and centrifuging at 14 000 g for 15 s. Eluted compounds from 5 samples from each plant were pooled together to form 2 samples for each plant.

Pooled samples and standards were dried in a Speed-Vac in labelled culture tubes and were derivatised with heptafluorobutyric anhydride and acetic anhydride as described in section 2.2.4ii.

The histamine content of seawater samples and standards were analysed by electron- capture negative-ionisation (ECNI) GC-MS, which detects much lower concentrations of histamine (~1000-fold) than EI GC-MS. The ECNI GC-MS instrument used was an Agilent Technologies 6890 gas chromatograph interfaced to an Agilent 5973 mass selective detector equipped with Enhanced Chemstation G1701BA software (version B.01.00). Gas chromatographic separations were performed in the splitless mode using a HP-5MS capillary column (30 m x 0.25 mm i.d.) with 0.25 μm thickness). The gas chromatograph oven temperature was held at 70°C for 2 min, ramped at 30°C·min-1 to 250°C and held for 2 min (10 min run). The Mass Selective Detector was operated in selected ion monitoring mode (derivatised histamine - m/z 306 and derivatised ISTD - m/z 310) and extracted ion chromatograms were integrated using software. The histamine content of the samples (ng) was calculated by reference to the standard curve and converted to ng·ml-1 (nM). The histamine concentration of seawater was analysed by a 1-factor ANOVA (data transformed by ln[x + 1]) and Bonferroni’s post-hoc test was used to determine which treatments differed significantly at p = 0.05.

3.2.9 Settlement cue for Holopneustes purpurascens from Amphiroa anceps

i) Localisation of settlement cue from Amphiroa anceps

The localisation of the settlement cue from Amphiroa anceps, i.e., whether it is surface- bound or waterborne, was explored in an assay testing seawater conditioned by A. anceps and a no-contact treatment. Three 2 l beakers were set up in the CTR filled with A. anceps and ambient seawater, which were aerated for 2 d (CSW1, CSW2, CSW3). Fresh A. anceps was collected on the day of the assay. A. anceps used to condition seawater was removed from CSW and pieces were included in the assay. CSW in 2 l beakers was stirred and 4 ml transferred to petri-dishes. Mesh pieces (80 μm) were rinsed in SW and placed over a piece of fresh A. anceps, excluding the alga from larval contact, for the no-contact treatment. Mesh alone was included as a negative control.

Three to six competent larvae were added to each dish and percent settlement was scored after 24 h. The response of larvae (proportions) to A. anceps treatments was analysed by a 1-factor ANOVA. The effect of each batch of CSW on larval settlement were compared, pooling treatments which did not differ significantly at p = 0.05, as I believed a priori that these treatments were unlikely to be different. The effect of the no–contact treatment on larval settlement was then compared to the CSW treatments (pooled), pooling treatments which did not differ significantly at p = 0.05. The effect of treatments in which larvae did not contact A. anceps was then compared with the control A. anceps treatments in which the alga was accessible to larvae. ii) The source of the settlement cue from Amphiroa anceps – settlement assay with algae treated with antibiotics

Larvae of Holopneustes purpurascens metamorphose rapidly in response to fronds of Amphiroa anceps in laboratory assays and the majority of new recruits of H. purpurascens were found on this alga. However, extracts of A. anceps lacked detectable histamine (Swanson et al. 2004) and various fractions of polar extracts only induced low levels of settlement at high concentrations (unpublished data). The absence of histamine, or another rapid acting settlement cue, in extracts of A. anceps raises the possibility that the biofilm on A. anceps only produces a settlement cue in situ for H. purpurascens and hence is not detectable in extracts of the algae (Swanson et al. 2004).

To test this hypothesis, Amphiroa anceps was subjected to antibiotic treatments and then tested against H. purpurascens larvae in settlement assays, to determine whether the settlement cue/s from A. anceps is produced by the alga or the biofilm (PhD student Megan Huggett did this assay). Antibiotic treatments were adapted from previous studies where treatments were shown to be effective in reducing the abundance or diversity of surface-associated bacterial communities (Aguirre-Lipperheide & Evans 1993, Johnson & Sutton 1994, Huggett et al. 2005, Huggett et al. 2006). The same experiment was run previously with Delisea pulchra in which D. pulchra that had been treated with antibiotics induced equivalent levels of settlement to control plants, implying an algal derived cue (Chapter 2) (Swanson et al. 2004).

Amphiroa anceps (n = 4) were collected and brought back to the laboratory where portions of each plant were allocated to each of 4 treatments. All antibiotic treatments of A. anceps included a 5 min soak in a 10 % betadine-SSW solution, followed by 3 rinses in SSW, and a 48 h treatment in either: (1) SSW containing 20 mg·l-1 streptomycin, 10 mg·l-1 penicillin G and 10 mg·l-1 kanamycin (“AB” treatment); (2) SSW after pieces of A. anceps were gently wiped across an agar plate, before and after the 48 h soak, to physically remove bacteria (“wipe” treatment); (3) the combination of “AB+wipe” treatment. The procedural control was a 48 h soak in SSW (“soak” treatment). Subsections of several A. anceps plants were collected on the day of the assay as a positive control (“fresh” treatment) and SSW was used as a negative control. Pieces of A. anceps (~10 mg ww) from each treatment were added to dishes (n = 10) along with 5 competent larvae. Percent settlement was scored after 24 h. The response of larvae (proportions) to treated A. anceps in the settlement assay was analysed by a 1- factor ANOVA and Bonferroni’s post-hoc test was used to determine which treatments differed significantly at p = 0.05. iii) Bacterial biofilms from coralline algae: settlement assay and histamine analysis

Some species of marine bacteria are known to produce histamine (Fujii et al. 1997) hence we further explored the possibility that the settlement cue from coralline turfing algae may be bacterially derived histamine. Five bacterial isolates from the surface of Amphiroa anceps or Corallina officinalis (which also induces settlement of Holopneustes purpurascens) were screened against larval H. purpurascens in settlement assays (Honours student Jacinta Green did this assay using bacterial strains which were isolated and sequenced by PhD student Megan Huggett). Surface bacteria from A. anceps and C. officinalis were cultured by adding pieces of each alga (n = 3) to 1 ml SSW, vortexing and diluting to 10-1, 10-2, 10-3 and 10-4 and 100 μl of each dilution was spread onto Marine Agar plates. Plates were observed daily for two weeks and each bacterial morphotype was re-streaked until purity was attained, and stored in glycerol at -70°C until sequencing/screening. Single isolates were grown overnight, spun down and extracted as reported elsewhere (Dahllöf et al. 2000).

PCR of 16s rDNA was then performed using primers F27 (5’GAGTTTGATCCTGGC

TCAG-3’) and R1492 (5’-ACGGTTACCTTGTTACGACTT-3’). Purified PCR product (50–100 ng) was sequenced unidirectionally using BigDyeTM terminator cycle sequencing reaction mix (Applied Biosystems), using the F27 and R1492 16S rDNA primers and a 530F primer (5’-GTGCCAGCMGCCGCGG-3’). Sequences were analysed on an ABI 310 DNA system at the Sydney University Prince Alfred Molecular Analysis Centre, aligned and analysed using the BLAST search algorithm (http://www.ncbi.nlm.nih.gov).

Five bacterial isolates from the surface of Amphiroa anceps (A) or Corallina officinalis (C) were selected based on initial sequence data (all gammaproteobacteria, Table 3.1), for screening against larval Holopneustes purpurascens in settlement assays using a method adapted from (Negri et al. 2001). I selected isolate C1 because its closest match using the BLAST search algorithm (at the time) was a known histamine producer, Photobacterium phosphoreum (Fujii et al. 1997). I selected two other isolates (A1, C2) that closely matched Photobacterium spp. and two additional isolates (A2, A3) from A. anceps for comparison (Table 3.1). Isolates were inoculated onto plates of Marine Agar and colonies were transferred 24 h later to suspended Marine Broth culture media (20 ml, n = 5). Sterile coverslips were added to tubes containing inoculated or control Marine Broth (containing no bacteria), for the formation of biofilms including bacteria- free biofilms, which were agitated at room temp (20°C) for 24 h. Coverslips with a visible biofilm were selected from each treatment (n = 5) and rinsed gently with SSW before the settlement assay. Coverslips (biofilms) were placed in sterile petri-dishes with 4 ml of SSW and 5 competent larvae. Percent settlement was scored at 48 and 94 h.

Marine Broth cultures from each treatment were kept for histamine analysis (n = 2, I did the histamine analysis). Broth cultures were centrifuged at 10 000 g for 20 min to separate cells (pellet) from the culture media which contains any exogenous compounds released by bacteria (supernatant). The supernatants were filtered (0.2 μm) into sterile tubes and frozen until analysis. The supernatants were transferred to large vials containing1 μg ISTD and were extracted three times with equal volumes of dichloromethane (DCM). The DCM extract was left to evaporate and the vials rinsed with 5 ml ethanol which was dried in a Speed-Vac. A set of standards containing,

either, 10, 1, 0.1 or 0 μg of histamine and 1 μg ISTD were extracted with DCM as per samples. Extracted compounds and standards were derivatised (section 2.2.4ii) and analysed by GC-MS (ECNI) as described in section 3.2.8.

The response of larvae (proportions) at 94 h in the biofilm assay and the histamine concentration of bacterial supernatants were both analysed by a 1-factor ANOVA and planned comparisons. Following the ANOVA, I compared the effect of isolates which induced >5 % settlement (i.e., inductive treatments), as I believed a priori that these treatments were unlikely to be different. There was no significant difference between the two treatments so they were pooled. Next, the effects of isolates which induced <5 % settlement (i.e., non-inductive treatments) were compared with the control as I a priori believed that these treatments were unlikely to be different. Non-inductive treatments did not differ significantly from the control and were therefore pooled. Inductive treatments (pooled) were then tested against non-inductive treatments (pooled). The histamine content of bacterial supernatants from the respective inductive and non-inductive treatments were analysed using the same planned comparisons outlined for the settlement assay.

3.2.10 Statistical treatment

All experiments were balanced designs which were analysed by factorial analysis of variance (ANOVA) models using SYSTAT ® 10.0 for Windows. Data were transformed when necessary to meet the assumptions of the test and were used only when the transformation improved the data spread or the homogeneity of variance. Details of each analysis are in the relevant sections above.

3.3 Results

3.3.1 Recruitment Survey

Recruitment of Holopneustes purpurascens at Bare Island from December 2002 – November 2004 onto the seaweeds Delisea pulchra, Corallina officinalis, Amphiroa anceps and Homeostrichus olsenii was generally low (1-7 total new recruits/month) but new recruits were found in all months except for February-03 and September-04. The variation between years in the number of new recruits found on each alga (pooling recruits from coralline turf algae) in each month is presented (mean ± SE, Figure 3.2). The highest number of new recruits was recorded in July-04 on coralline turf algae (7 new recruits/100 g alga ww), however, it was more common to find 1-4 new recruits per month (per 100g alga-ww). No new recruits were found on E. radiata or S. vestitum in any month. New recruits with test diameters ≤2.0mm were found in 13 of the 21 months sampled and represented about 20 % of all new recruits found. Another 20 % of new recruits had test diameters between 2.1–3.0mm while the remaining 60% were split equally between the two larger size classes of test diameters between 3.1–4.0 mm or 4.1–5.0 mm. In total 38 % of all new recruits were found on A. anceps, 32 % on D. pulchra, 22 % on C. officinalis and 8 % on H. olsenii. Significantly more new recruits were found on D. pulchra and coralline turf algae than on E. radiata or S. vestitum (1- factor ANOVA, Falga (4, 100) = 9.061, p < 0.001; Bonferroni’s pairwise comparisons, all p < 0.02).

3.3.2 Settlement assays testing host algae

A range of host algae were assayed against larvae of Holopneustes purpurascens to test for a settlement response. The algae tested induced varying levels of settlement of larvae after 1 and 24 h with no settlement observed in SSW (Figure 3.3). Fresh pieces of Delisea pulchra and Amphiroa anceps induced approximately 50 % settlement of larvae after 1 h whereas Homeostrichus olsenii and unfouled Ecklonia radiata induced

significantly lower responses of 27 % and 4 % settlement respectively. Thus, the algae on which we found the most new recruits in the field induced the highest rate of settlement of H. purpurascens larvae in laboratory assays. Fouled E. radiata and epiphytes alone induced 32 % and 49 % settlement respectively after 1 h, significantly higher than the 4 % settlement observed with unfouled E. radiata. After 24 h, all algal treatments had induced over 90 % settlement of larvae except for unfouled E. radiata which had induced a significantly lower response of 18 % settlement (Figure 3.3, Table 3.2). There was a significant alga x time interaction which was explored further by a series of planned comparisons of algal effects within each level of time (Table 3.2). There were no major qualitative differences in the statistical outcome of comparisons at 1 and 24 h hence we concluded that there were algal effects over and above any interaction.

3.3.3 Settlement assays testing seawater collected in situ adjacent to algae

We collected seawater samples in situ, adjacent to and away from algae, to test whether the settlement cue for Holopneustes purpurascens leaches from these seaweeds. None of these seawater samples induced settlement of competent larvae of H. purpurascens (6 d post-fertilisation). However, the same seawater samples were subsequently tested with older larvae (3-4 wk post-fertilisation) as I suspected that they would respond to lower concentrations of histamine (Chapter 4). Seawater collected in situ, adjacent to and away from algae, in Nov-03 induced different levels of settlement of older larvae (Figure 3.4A, Table 3.3). Nine to sixteen percent of older larvae settled in response to Delisea-SW collected Nov-03 after 24 and 44 h, respectively (Figure 3.4A), which was greater than larval settlement in response to Surface-SW and SSW (Control-SW, Table 3.3). Significantly more older larvae settled in response to Delisea-SW than in Control- SW collected in Jul-02 after 24 h (Figure 3.4C, Table 3.4), and a similar trend was found for Delisea-SW and Control-SW collected in Jan-03 after 72h (Figure 3.4B, Table 3.5). Ecklonia-SW and Amphiroa-SW collected in July-02 and Jan-03 were more inductive than those collected in Nov-03 (Figure 3.4A-C). Interestingly, significantly more aged larvae settled in response to Amphiroa-SW than in Control-SW collected in Jan-03 after 72 h (Figure 3.4B, Table 3.5). Comparisons between the larval response to

other in-situ seawater treatments and Control-SW (Jul-02, Jan-03, Nov-03) found no significant differences (Tables 3.3, 3.4, 3.5).

3.3.4 Temporal analysis of the histamine content of algae

Levels of histamine in Delisea pulchra (108–288 μg·g-1 dw) were 2–3 orders of magnitude higher than that of the other algae at all times (Table 3.6). Histamine levels in D. pulchra were 100 times greater than levels in Homeostrichus olsenii (0.41–2.85 μg·g-1dw), and approximately 1000 times greater than Ecklonia radiata (0.05–0.28 μg·g-1dw) and Sargassum vestitum (0.07–1.12 μg·g-1dw). The histamine content of H. olsenii, on which we found 8 % of new recruits of Holopneustes purpurascens, was always higher than that of E. radiata and S. vestitum, on which no new recruits were found. The histamine content of D. pulchra was higher in July and September and this trend was explored further. D. pulchra plants sampled in Jul-Aug-Sep-04 had the highest average histamine contents (262, 507 and 1041 μg·g-1 dw, respectively) while those sampled in Jan-03/04 and Apr-04 contained the least histamine (80–130 μg·g-1 dw, Figure 3.5). The histamine content of D. pulchra varied considerably from month to month (1-factor ANOVA, F8,36 = 12.649, p < 0.001), as reflected by the lowest (12 μg·g-1 dw, Jan-03) and highest (1470 μg·g-1 dw, Sep-04) content recorded for individual plants.

3.3.5 Variation in histamine within the thallus of Delisea pulchra and Ecklonia radiata

The histamine content varied significantly within the thallus of Delisea pulchra, with levels in the base of the plant approximately 2-fold greater than the mid section and 4- fold greater than the tips (Table 3.7). The primary and secondary laminae of Ecklonia radiata both contained very low levels of histamine. The histamine content of the fouling epiphytes (6.0 μg·g-1 dw) was 100-fold greater than the laminae and comparable to the lowest levels recorded in D. pulchra (Table 2.1).

3.3.6 The histamine concentration of seawater in the habitat

Significantly more aged larvae settled in response to Delisea-SW from Jul-02 than to Control-SW after 24 h (Figure 3.4C, Table 3.4, p = 0.013). This Delisea-SW contained the highest concentration of histamine (4.20 ± 0.75 nM) of all seawater samples, with significantly lower levels in Ecklonia-SW (1.54 ± 0.15 nM), Amphiroa-SW (0.58 ± 0.16 nM) and Surface-SW (0.42 ± 0.17 nM; 1-factor ANOVA, F3, 4 = 56.256, p = 0.001; Bonferroni’s pairwise comparisons, all p < 0.02).

3.3.7 Settlement cue for Holopneustes purpurascens from Amphiroa anceps

i) Localisation of settlement cue from Amphiroa anceps

Over 50 % of Holopneustes purpurascens larvae settled in dishes containing Amphiroa anceps covered by mesh, i.e., without contacting the alga (Figure 3.6). All larvae settled in dishes containing A. anceps with no mesh. Three batches of CSW (A. anceps- conditioned seawater) induced between 20 and 50 % settlement after 24 h (Figure 3.5). Larvae showed variable responses to A. anceps, CSW and the no contact treatment (1- factor ANOVA, F4,20 = 6.421, p = 0.002). The larval response to three batches of CSW and the non contact treatment did not differ from one another (planned comparisons, all p > 0.141). The larval response in these treatments (pooled) differed from the larval response to A. anceps that was accessible to larvae (planned comparison, p < 0.001). A. anceps which was used to condition seawater induced 100 % settlement of larvae (not shown). There was no settlement of larvae in the mesh or SSW treatments. ii) The source of the settlement cue from Amphiroa anceps – settlement assay with algae treated with antibiotics

Amphiroa anceps treated with antibiotics (AB), and A. anceps treated with antibiotics and an agar-wipe (AB+wipe), induced minimal settlement (<5 %) relative to fresh and soaked (procedural control) A. anceps, which induced 52–64 % settlement of

Holopneustes purpurascens larvae (Figure 3.7, 1-factor ANOVA, F4,45 = 15.985, p <0.001; Bonferroni’s pairwise comparisons, p < 0.001).

iii) Bacterial biofilms from coralline algae: settlement assay and histamine analysis

There was minimal settlement of Holopneustes purpurascens larvae in bacterial biofilm treatments after 48 h, however, 43 % and 20 % of larvae had settled in response to C1- biofilms and A2-biofilms, respectively, after 94 h (Figure 3.8A). Settlement of larvae in replicate dishes containing these inductive biofilms ranged from 0-100 %. Larval settlement in response to C1-biofilms and A2-biofilms was, however, significantly higher than settlement in other biofilm treatments and control dishes (no biofilm; 1- factor ANOVA, F5,24 = 1.922, p = 0.128; planned comparisons C1-A2 vs C2-A1-A3- control, p = 0.01). The histamine content of the supernatants of broth cultures of isolates in which biofilms were cultured correlated with the level of settlement induced by these biofilms in the assay (Figure 3.8A, B). That is, supernatants from the most inductive isolates in the assay (C1 and A2) contained higher concentrations of histamine (0.091 and 0.068 μg·ml-1, Figure 3.8B). In contrast, supernatants from the non- inductive isolates including A3 (≤5 % settlement) had similar concentrations of histamine (0.017–0.033 μg·ml-1) as the control supernatant (0.021 μg·ml-1) (1-factor

ANOVA, F5,6 = 1.71, p = 0.265; planned comparisons, C2 or A1 or A3 vs control, p > 0.9 and C1-A2 vs C2-A1-A3-control, p = 0.032).

60 Figure 3.1A Photograph of Holopneustes purpurascens reared in the laboratory showing 2 characteristics used to identify urchin recruits found in survey, a) brown pigmented genital plates on the aboral surface (outlined by dashed box), b) white and pink/purple spines which were wider at the tip than at the centre of spine (white arrows). Scale bar = 1 mm.

Figure 3.1B Photograph of new recruit of Holopneustes purpurascens found in survey showing 2 characteristics used to identify it as such a) brown pigmented genital plates (outlined by dashed box), b) two-tone spines which were wider at the tip than at the centre of spine (white arrows). Scale bar = 1 mm.

Delisea pulchra Coralline Turf 6 Homeostrichus olsenii

2 # New recruits (per 100 g alga-ww) # New recruits * 0 * * Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month

Figure 3.2 The number of Holopneustes purpurascen new recruits (test diameter ≤ 5 mm) found each month (n = 2 yr) on Delisea pulchra, the coralline turf algae (Amphiroa anceps and Corallina officinalis) and Homeostrichus olsenii, (total from n = 5 plants of each alga per month standardised to number of recruits per 100 g ww-alga sampled; graph shows mean ± SE of n = 2 yr). New recruits were not found on Ecklonia radiata or Sargassum vestitum at any time. * not sampled in yr 2, * * not sampled in yr 1 or 2

100

1h 24h 80

40 Percent settlement 20

0 SSW Delisea Amphiroa Epiphytes Histamine Eck-fouled Eck-unfouled Homeostrichus

Figure 3.3 Percent settlement of Holopneustes purpurascens larvae (mean ± SE, n = 10, 5 larvae/dish) after 1 and 24 h in response to a range of host algae; Delisea pulchra, Amphiroa anceps, Homeostrichus olsenii, fouled Ecklonia radiata (Eck-fouled), unfouled E. radiata (Eck-unfouled) and the fouling epiphytes removed from laminae of E. radiata (epiphytes). Histamine (10 μM) and SSW were included as the positive and negative controls.

A 25

24h 20 44h

10 Percent settlement Percent 5

0 Del-SW Eck-SW Amp-SW Surface-SW SSW

B 25

10 Percent settlement

0 Del-SW Eck-SW Amp-SW Surface-SW SW

C 14

Percent settlement Percent 4

0 Del-SW Eck-SW Amp-SW Surface -SW SW

Figure 3.4 Percent settlement (mean ± SE, n = 10, 10 larvae/dish) of older Holopneustes purpurascens larvae in response to seawater collected in situ nearby Delisea pulchra (Del-SW), Ecklonia radiata (Eck-SW) and Amphiroa anceps (Amp- SW), or at the sea surface (Surface-SW). Histamine (10 μM - not shown, 100% settlement at 24 h) and SSW were included as the positive and negative controls. A. Nov-03 seawater samples (larvae, 3 wk post-fertilisation) after 24 h and 44 h, B. Jan-03 seawater samples (larvae, 4 wk post-fertilisation) after 72 h, C. July-02 seawater samples (larvae, 3 wk post-fertilisation) after 24 h.

1.6

1.4

1.2

1.0 dw

-1 0.8

mg.g 0.6

0.4

0.2

0.0

Jan Sep Oct Nov Jan Apr Jul Aug Sep 2003 2004

Figure 3.5 Box plots showing the histamine content of Delisea pulchra over time including three successive months in 2003 and 2004 (mg·g-1 dw, mean ± SE).

100

40 Percent settlement Percent

0 a W iro tact SS ph CSW1 CSW2 CSW3 Mesh Con Am No Treatment

Figure 3.6 Percent settlement of Holopneustes purpurascens larvae (n = 5 dishes, 3–6 larvae per dish) in response to fresh Amphiroa anceps, A. anceps covered with mesh (no contact) and 3 batches of seawater conditioned by A. anceps (CSW). Mesh and SSW were included as the negative controls. A. anceps used to condition SW induced 100 % settlement (not shown).

40 Percent settlement 20

0 Fresh Soak Wipe AB AB+wipe SSW Treatment

Figure 3.7 Percent settlement of Holopneustes purpurascens larvae (mean ± SE, n = 10, 5 larvae/dish) after 24 h in response to Amphiroa anceps that had been treated with antibiotics. The antibiotic treatment of A. anceps included “AB” (streptomycin, penicillin G and kanamycin), “wipe” (across agar) and “AB+wipe” treatments. Control treatments were “soak” (in SSW), “fresh” A. anceps and SSW. Treatments that share a line induced a similar larval response (ANOVA, p < 0.001; Bonferroni’s pairwise comparisons, p < 0.001).

70 A 60

Percent settlement Percent 10

0 B 0.15 C1 A2 A3 A1 C2 no bacteria ) -1 g ml μ 0.10

0.05

Supernatant histamine ( histamine Supernatant 0.00 C1 A2 A3 A1 C2 no bacteria Bacterial isolate

Figure 3.8 A. Percent settlement of Holopneustes purpurascens larvae (mean ± SE, n = 5 dishes, 5 larvae/dish) after 94 h in response to biofilms of bacterial isolates from the surface of Amphiroa anceps (A1, A2, A3) or Corallina officinalis (C1, C2; Table 3.1). Biofilms containing no bacteria were used as the negative control and histamine (10 μM) was used as the positive control. B. The histamine concentration (μg·ml-1 ± SE) of the supernatants from the culture of the bacterial isolates (n = 2).

Table 3.1 BLAST analysis and designated Accession Numbers for bacterial isolates from the surface of Corallina officinalis (C) and Amphiroa anceps (A) which were tested in settlement assays with larval Holopneustes purpurascens. The nearest matching bacterial strain (with designated Accession Number) and percent similarity (%) from two search dates are shown.

Isolate Accession Nearest match % Nearest match % ID Number (initial search Aug-03) (recent search Apr-05)

C1 DQ005883 Photobacterium 99 Photobacterium 95 phosphoreum AJ746360 damselae AY147861 C2 DQ005897 Photobacterium sp. 97 Photobacterium 98 AY781193 eurosenbergii AJ842344 A1 DQ005851 Photobacterium sp. 96 Pseudoalteromonas sp. 99 AY582934 AY626830 A2 DQ005882 Thalassomonas viridans 99 Thalassomonas viridans 99 AJ294748 AJ294748 A3 DQ005869 Vibrio mediterranei 99 Vibrio medeterranei 98 X74710 X74710

Table 3.2 Univariate repeated measure ANOVA of the settlement response of Holopneustes purpurascens larvae to algae after 1 and 24 h. Planned comparisons at 24 h are also shown. E.r = Ecklonia radiata (f. = fouled, unf. = unfouled), D.p = Delisea pulchra, A.a = Amphiroa anceps, H.o = Homeostrichus olsenii, epi. = epiphytes

Source df MS F p

Between Subjects Alga 5 1.142 32.421 <0.001 Error 54 0.035 Within Subjects Time 1 7.016 274 <0.001 Time x Alga 5 0.183 7.167 <0.001 Error 54 0.026 Planned Comparisons D.p vs A.a 1 0.018 1.455 0.233 D.p-A.a vs H.o 1 0.008 0.660 0.420 D.p-A.a-H.o vs unf.E.r 1 4.563 369 <0.001 Epiphytes vs f.E.r 1 0.002 0.162 0.689 Epi.-f.E.r vs unf.E.r 1 3.970 321 <0.001

Table 3.3 Univariate repeated measure ANOVA of the settlement response (at 24 and 44 h) of older Holopneustes purpurascens larvae (3 wk post-fertilisation) to Nov-03 in situ seawater samples. Planned comparisons at 24 h are also shown. Seawater collected adjacent to algae; Delisea pulchra (Delisea-SW), Ecklonia radiata (Ecklonia-SW) and Amphiroa anceps (Amphiroa-SW), or at the sea surface (Surface-SW). SSW = sterile seawater, Surface-SW–SSW (pooled) = Control-SW

Source df MS F p

Between Subjects Seawater 4 0.059 5.836 <0.001 Error 45 0.010 Within Subjects Time 1 0.023 9.118 0.004 Time x Seawater 4 0.004 1.586 0.194 Error 45 0.003 Planned Comparisons Surface-SW vs SSW 1 0.000 0.000 1.000 Delisea-SW vs Control-SW 1 0.062 21.66 <0.001 Amphiroa-SW vs Control-SW 1 0.002 0.779 0.382 Ecklonia-SW vs Control-SW 1 0.001 0.286 0.595

Table 3.4 One-factor ANOVA and planned comparisons of the settlement response (at 24 h) of older Holopneustes purpurascens larvae (3 wk post-fertilisation) to Jul-02 in situ seawater samples. Seawater collected adjacent to algae; Delisea pulchra (Delisea- SW), Ecklonia radiata (Ecklonia-SW) and Amphiroa anceps (Amphiroa-SW), or at the sea surface (Surface-SW). SSW = sterile seawater, Surface-SW–SSW (pooled) = Control-SW

Source df MS F p

Main analysis Seawater 4 0.007 2.151 0.091 Error 43 0.003

Planned comparisons Surface-SW vs SSW 1 0.004 1.210 0.277 Delisea-SW vs Control-SW 1 0.021 6.775 0.013 Amphiroa-SW vs Control-SW 1 0.004 1.426 0.239 Ecklonia-SW vs Control-SW 1 0.006 2.063 0.158

Table 3.5 One-factor ANOVA and planned comparisons of the settlement response (at 72 h) of older Holopneustes purpurascens larvae (4 wk post-fertilisation) to Jan-03 in situ seawater samples. Seawater collected adjacent to algae; Delisea pulchra (Delisea- SW), Ecklonia radiata (Ecklonia-SW) and Amphiroa anceps (Amphiroa-SW), or at the sea surface (Surface-SW). SSW = sterile seawater, Surface-SW–SSW (pooled) = Control-SW

Source df MS F p

Main analysis Seawater 4 0.028 1.605 0.190 Error 45 0.018

Planned comparisons Surface-SW vs SSW 1 <0.001 0.003 0.954 Delisea-SW vs Control-SW 1 0.067 3.790 0.058 Amphiroa-SW vs Control-SW 1 0.082 4.699 0.036 Ecklonia-SW vs Control-SW 1 0.025 1.411 0.241

Table 3.6 The histamine content of Delisea pulchra, Homeostrichus olsenii, Ecklonia radiata and Sargassum vestitum over time (μg·g-1 dw, mean ± SE, n = 5 [except n = 4 for Ecklonia radiata in April]). The histamine contents of the four species of algae over time were significantly different (2-factor ANOVA results shown).

Algal species μg·g -1 (dw) mean ± SE Sep-03 Jan-04 Apr-04 Jul-04 Delisea pulchra 288 ± 77 129 ± 92 108 ± 39 262 ± 52 Homeostrichus 2.62 ± 1.06 0.41 ± 0.14 0.64 ± 0.15 2.85 ± 0.91 olsenii Ecklonia radiata 0.27 ± 0.08 0.06 ± 0.03 0.28 ± 0.11 0.05 ± 0.01 Sargassum 0.10 ± 0.02 0.07 ± 0.01 – 1.12 ± 0.52 vestitum

Source df MS F p Alga 2 133.9 271.2 <0.001 Month 3 1.936 6.172 0.001 Alga x Month 6 0.494 1.575 0.175 Error 48 0.313

Table 3.7 The histamine content (μg·g -1 dw, mean ± SE) of different regions of the thallus of Delisea pulchra (n = 5) and Ecklonia radiata (n = 3). Randomised block ANOVA results are shown.

Algal species μg.g -1 dw - thallus part mean ± SE ANOVA

Delisea pulchra Source base 2657 ± 286 Part F2,8 = 14.92 p = 0.002 mid 1231 ± 194 Block (plant) F4,8 = 2.131 p = 0.168 tip 716 ± 285

Ecklonia radiata 1° 0.059 ± 0.029 Part F2,4 = 175.6 p < 0.001 2° 0.044 ± 0.006 Block (plant) F2,4 = 0.569 p = 0.606 Epiphytes 6.01 ± 1.17

3.4 Discussion

Despite over 60 years of research effort no settlement cue for a marine invertebrate has been unequivocally identified and quantified in situ. In this chapter I present the first investigation of the settlement and recruitment of a marine invertebrate with respect to the quantitative distribution of a chemical settlement cue in the organism’s habitat. We have shown that the distribution of new recruits of Holopneustes purpurascens at the study site can be partially explained by the distribution of histamine in the habitat. We found recruits of H. purpurascens (≤5 mm test diameter) in almost every month of two years of sampling and recruits less than 2 mm were found in all months except February (sampled 1 yr), June (not sampled) and August. The distribution of recruits through time suggests that recruitment of H. purpurascens occurs throughout the year at this site. These findings are consistent with the continuous production of the settlement cue by D. pulchra and the reproductive cycle reported for H. purpurascens (Williamson & Steinberg 2002). The actual breeding period for H. purpurascens is throughout spring and early summer however, vitellogenesis appears to be continuous in this species meaning they are capable of reproducing sporadically throughout the year, which is rare among urchins (Williamson & Steinberg 2002). This ability of H. purpurascens to reproduce throughout the year is reflected in my experience with larval culture. Larvae were obtained in all seasons over a 5 yr period, albeit with low yields of larvae (50-100) in some months.

With the exception of the coralline algae for which histamine could not be detected, variation in settlement and recruitment was consistent with the variation among species’ histamine contents, supporting the link between histamine production and settlement on these algae. Thirty-two percent of new recruits were found on Delisea pulchra, which contained far greater levels of histamine (up to 1000-fold), than the other algae surveyed (Figure 3.2, Table 3.6). The algal histamine data reported here is consistent with a previous study (Swanson et al. 2004) although differences between D. pulchra and Ecklonia radiata / Sargassum vestitum in this paper are 100-fold greater due to greater accuracy and sensitivity of the analysis. Sixty percent of new recruits of Holopneustes purpurascens were collected from the coralline turf algae, Amphiroa anceps (38 %) and Corallina officinalis (22 %), which are far more abundant at this site

than D. pulchra, forming dense tufts in sub-canopies below foliose algae (including D. pulchra) and kelp. Another study at Bare Island found 32 new recruits (≤2 mm test diameter) of H. purpurascens over a 14 month period (Williamson 2001). Half of the new recruits were found on D. pulchra and half were found on coralline turf algae (Williamson 2001) which is consistent with my findings.

The coralline algae also induced rapid settlement of Holopneustes purpurascens larvae (Figure 3.3). The settlement cue from Amphiroa anceps appears to be waterborne as contact with the alga is not required for settlement, and seawater conditioned by A. anceps induced settlement (Figure 3.6). Levels of settlement were not as high in the treatments in which larvae were not in contact with the alga, perhaps because larvae were unable to swim close to the alga and higher concentrations of the cue. Amphiroa- SW collected in Jan-03 induced significantly more larvae to metamorphose than Control –SW which is further support for a waterborne cue from A. anceps. Treatment of A. anceps with antibiotics greatly reduced the settlement in response to this alga, which is in contrast to Delisea pulchra which induced a high settlement response in H. purpurascens larvae even after being treated with these antibacterial agents (Swanson et al. 2004). This treatment of A. anceps with antibiotics reduced the diversity of the surface-associated bacterial community (biofilm) by ~75 % but does not remove all bacterial strains ((Huggett et al. 2005, Huggett et al. 2006). Despite the persistence of some surface bacterial strains after treatment, the antibiotics drastically reduced the capacity of A. anceps to induce settlement of H. purpurascens, consistent with the presence of a biofilm derived settlement cue from A. anceps for H. purpurascens.

Biofilms on the surface of animate and inanimate marine surfaces produce and release settlement cues for a range of invertebrate larvae (Wieczorek & Todd 1998). Biofilms of C1 and A2, bacterial components of the natural biofilm from the surface of Corallina officinalis and Amphiroa anceps, induced 43 % and 20 % settlement of larval Holopneustes purpurascens, respectively (Figure 3.8A). Settlement in dishes containing inductive biofilms ranged from 0–100 % and the rate of settlement was slow, which was possibly due to differences in the density and condition of the biofilm on the coverslips. The rate of settlement induction was too slow (96 h) to be conclusive, given that the composition of C1- and A2-biofilms may have been altered by the introduction

of bacteria along with the larvae, however, there was no settlement in the control (bacteria-free biofilm) by this time. The broth media used to culture isolates contained low concentrations of histamine, however, significantly higher concentrations of histamine were detected in supernatants of media in which the inductive isolates, C1 and A2, were cultured and biofilms formed (Figure 3.8B). The higher concentrations of histamine in these supernatants suggests bacterial production and release of histamine in vitro by C1, as reported previously for this bacterium (Yoguchi et al. 1990). Together, these initial findings support a settlement cue of bacterial origin from coralline algae for Holopneustes purpurascens which may be bacterial derived histamine.

Histamine was not detected in extracts of the coralline algae nor in in situ seawater samples collected adjacent to Amphiroa anceps, however, the biofilm on the algal surface may only produce histamine in situ at certain times releasing histamine into surrounding seawater. If the bacterial settlement cue proves to be histamine, this would be the first system described for which a larval settlement cue for a marine invertebrate is produced by both eukaryotes and prokaryotes in the natural habitat. However, evidence for a bacterial source of histamine is at a much earlier stage than our case for algal derived histamine acting as a settlement cue for larval H. purpurascens.

I have reported that coralline algae are potent inducers of larval settlement, with over 95 % settlement of Holopneustes purpurascens larvae by 24 h, however, Williamson et al. (2000) reported that these algae only induced settlement in 0–20 % of larvae by 24 h. The differences in the larval response may have resulted from testing different sized pieces. However, Williamson et al. (2000) always used the tips of algae in settlement assays whereas I have tested pieces from all regions of the thallus. Perhaps the tips of coralline algae contain less inductive bacteria than other regions of the thallus and consequently do not always induce settlement of larval H. purpurascens

The majority of the literature on chemical settlement cues for larvae focuses on surface- bound, non-polar chemicals that induce larval settlement because for many years the only effective settlement cues were considered to be surface-bound (Morse 1990, Pawlik 1992). It was argued that dissolved settlement cues would be ineffective due to rapid dilution in seawater and, even if detected by larvae, the weak swimming abilities

of most would prevent them from swimming towards the source of the cue, particularly in turbulent environments (Crisp 1974, Butman 1987). However, it is now clear that a number of species of marine invertebrate settle in response to dissolved chemical cues (Hadfield & Paul 2001) in still water assays (Williamson et al. 2000, Swanson et al. 2004), laboratory flumes under realistic flow conditions (Turner et al. 1994, Tamburri et al. 1996), and in the field (Browne & Zimmer 2001).

This chapter reports the first in situ quantitative measurements of a characterised chemical cue in seawater. Very low concentrations of dissolved histamine (~5 nM) were measured in seawater collected in situ adjacent to Delisea pulchra, with significantly lower levels detected in seawater collected adjacent to Ecklonia radiata, Amphiroa anceps or at the sea surface. For a settlement cue to be ecologically relevant it must be present in the habitat at inductive concentrations. None of the Delisea-SW tested here induced settlement of newly competent Holopneustes purpurascens larvae, however, 9–16 % of older larvae settled in response to Delisea-SW (compared to <5 % settlement in SSW), indicating that a dissolved settlement cue was present in situ in the seawater surrounding D. pulchra plants (Figure 3.4A). Ecklonia-SW (Jul-02) contained more histamine than Surface-SW, and Amphiroa-SW (Jan-03) induced more larvae to settle than Control-SW hence histamine also appears to leach from these algae into surrounding seawater. In other experiments (Figure 4.1), 10 nM of histamine induced 10–40 % settlement of older larvae (3-4 wk post-fertilisation) after 24–72 h with less than 5 % settlement in SSW at 72 h. Newly competent larvae do not settle in response to 10 nM histamine which suggests that older larvae show an increased sensitivity to histamine while at the same time maintaining their selectivity for histamine as a settlement cue for at least 4 wk post-fertilisation (Chapter 4). Thus levels of histamine measured in the habitat induce settlement of larvae, though at low rates and only for older larvae; such larvae may be more typical in the natural habitat than newly competent larvae.

An earlier study of this system collected Delisea-SW which induced 100 % settlement of newly competent Holopneustes purpurascens larvae after 4 h (Williamson et al. 2000). The histamine concentration of these samples was not measured but may have contained higher concentrations of histamine than reported here. This earlier result

suggests that inductive levels of histamine do occur in situ in seawater surrounding Delisea pulchra plants in the habitat. We found considerable variation in the histamine content of D. pulchra plants sampled simultaneously and through time (Table 4, 5). Thus, higher concentrations of histamine may occur in seawater surrounding D. pulchra plants during times when the histamine content of the plants are high (for e.g., in Sep- 04), or in the boundary layer at the surface of D. pulchra plants. It is also possible that low concentrations of histamine are more effective at inducing settlement in the natural habitat, when the cue is detected in unison with other chemical cues or physical factors such as flow (Altieri 2003).

Unfouled laminae of Ecklonia radiata contained very low amounts of histamine relative to the epiphytic fouling community growing on the surface of E. radiata laminae. The fouling epiphytes on E. radiata, predominantly filamentous brown and red algae (Sphacelaria sp., Polysiphonia blandii - (Jennings & Steinberg 1997), contained levels of histamine that were comparable to the lowest amounts recorded in Delisea pulchra (Swanson et al. 2004) and consequently fouled E. radiata and epiphytes alone induced a high rate of settlement of Holopneustes purpurascens larvae. Given the biomass of fouled E. radiata in the habitat of H. purpurascens fouled kelp may leach histamine into surrounding seawater, particularly in dense beds of E. radiata. The seawater analysis supports this claim as higher concentrations of histamine were detected in Ecklonia-SW than in Surface-SW (p = 0.038).

In previous studies, Ecklonia radiata and Sargassum vestitum did not induce settlement of newly competent (6-day-old) Holopneustes purpurascens larvae (Williamson et al. 2000, Williamson et al. 2004). However, in this study, ~20 % of 6-day-old larvae settled in response to unfouled E. radiata after 24 h. The different larval response to E. radiata may be explained by the variation in histamine levels within the thallus of E. radiata and by the size of algal pieces being tested. I used larger pieces of E. radiata (20–30 mg) from both the primary and secondary laminae, whereas Williamson et al. (2004) used smaller pieces (15 mg) always from the tip of secondary laminae. The histamine content of the primary laminae was higher and more variable than the secondary laminae (Table 3.7) thus pieces of primary laminae may have induced settlement while tips of secondary laminae did not.

CHAPTER FOUR

The sensitivity and specificity of the larval response to histamine

4.1 Introduction

Non-feeding larvae have limited energetic reserves to support their development during the planktonic phase and through settlement, and so have a limited time to locate a favourable habitat to resume the benthic phase. Larval species from a range of phyla settle in response to specific settlement cues and are able to postpone settlement until such cues are encountered (Pechenik 1990). The ability of larvae to delay settlement in the absence of obligatory cues, despite being developmentally ready (or competent) for metamorphosis, is believed to increase the chance of settling in habitats which can support survival to adulthood (Thorson 1950, Pechenik 1990, Morgan 1995). However, it was proposed over half a century ago (Knight-Jones 1953, Wilson 1953) that larvae may become less- discriminatory in habitat selection with extended time in the plankton, with settlement occurring in sub-optimal sites when finite reserves are depleted, rather than dying without metamorphosing (subsequently known as the ‘desperate larva hypothesis’ – or DLH, Toonen & Pawlik 2001). Since then, DLH appears to explain the decreased substratum- specificity observed in a range of older non-feeding larvae (Miron et al. 2000, Marshall & Keough 2003, Gribben et al. 2006).

More recently, the DLH has been modified to incorporate the effects of larval feeding and the consequences of habitat specialisation (Botello & Krug 2006). Larvae of the opisthobranch Alderia sp., which feeds exclusively on the alga Vaucheria longicaulis (Krug & Manzi 1999, Krug & Zimmer 2000), do not become indiscriminate with regards to settlement as they age. Rather, older, starved larvae of Alderia sp. became more sensitive

to settlement cues from the host alga but older, fed larvae did not exhibit increased sensitivity to cues (Botello & Krug 2006). Botello and Krug suggest that the DLH should not apply to host-dependant species, because settlement in the absence of such hosts/prey would likely be fatal. A theoretical model supports this notion and also suggests that for species that feed as larvae, the DLH is less likely to apply (Elkin and Marshall in press).

The mechanism underlying the decreased selectivity of some larvae with respect to settlement cues (i.e., ‘desperation’) is not known, however, desperate larvae are rarely indiscriminate. For example, the opisthobranch Haminaea callidegenita reacts to a greater variety of substrata as they age; young larvae only settle in response to egg mass jelly but older larvae settle in response to Zostera marina and a green alga but not to biofilmed sediment (Gibson 1995). Older Holopneustes purpurascens larvae also settle in response to a greater range of host algae (including red and brown algae) whereas newly competent larvae settle preferentially on red algae (Delisea pulchra, Amphiroa anceps, Corallina officinalis; (Williamson et al. 2004). Older larvae which accept a wider range of substrata at settlement may be responding to different chemical cues (inducers) in these substrata or, they may become more sensitive to the same chemical cue which is present at lower concentrations in such substrata. S. vestitum and E. radiata contain histamine, albeit at much lower levels than the preferred host of new recruits, D. pulchra. Thus, the mechanism underlying the decreased selectivity of older H. purpuracens larvae toward host algae may be based on an increased sensitivity to lower concentrations of histamine present in these algae.

Because the chemical identities of naturally occurring settlement cues remain elusive in most cases, researchers have relied upon artificial inducers of settlement to gain insight into the underlying mechanisms of signal transduction and metamorphosis. A plethora of compounds can artificially induce characteristic settlement behaviours and/or metamorphosis by non-specific interference with the normal morphogenetic pathway, either at the true receptor or somewhat farther downstream. The use of artificial inducers such as; excess ions (K+ and Ca2+, Yool et al. 1986), neurotransmitters or their precursors (e.g., epinephrine, serotonin, dopamine, L-DOPA; Coon et al. 1985, Bonar et al. 1990, Pearce & 82

Scheibling 1990), amino acids and peptides (Trapido-Rosenthal & Morse 1985, Rittschof et al. 1989, Tegtmeyer & Rittschof 1989, Zimmer-Faust & Tamburri 1994, Kon-ya et al. 1995, Naidenko 1996), and pharmacological agents (e.g., inhibitors, ligand-antagonists, membrane transport blockers; Baxter & Morse 1987, Coon & Bonar 1987), has revealed some key components of the signal transduction pathways of settlement. For example, transmembrane ion fluxes both at the receptor cell and in cells farther downstream, are necessary components for settlement (Yool et al. 1986, Biggers & Laufer 1999, Hadfield et al. 2000).

By investigating the sensitivity and specificity of the larval settlement response to histamine I aim to gain further understanding of the underlying mechanisms of larval settlement in Holopneustes purpurascens. In this chapter, I test the hypothesis that older H. purpurascens larvae settle in response to a greater range of host algae by reducing the threshold concentration of histamine required to trigger settlement. The dose-response of 3 batches of larvae was investigated by testing larvae of different ages (7-d-old, 14-d-old, 21- d-old and 28-d-old) against a range of concentrations of dissolved histamine in settlement assays, to see if older larvae 1) become more sensitive (responsive) to low concentrations of histamine, 2) show decreased specificity for histamine as a settlement cue. Recent findings also suggest that older larvae may require less exposure than younger larvae to settlement cues to initiate settlement (Jackson et al. 2005). To determine whether older H. purpurascens require less exposure to an inducer to initiate settlement than newly competent larvae; 7-d-old, 14-d-old and 21-d-old larvae from the same batch were exposed to10 μM histamine for different time periods and levels of settlement compared.

As part of the investigation into the specificity of the larval settlement response to histamine, other compounds known to be important neuroactive molecules in animals were tested as possible inducers of larval settlement in Holopneustes purpurascens. Biogenic amines are monoamine compounds synthesised from amino acids which play important roles as neurotransmitters, neuromodulators and neurohormones. The five major biogenic amines in invertebrates are serotonin, dopamine, octopamine, tyramine and histamine (Blenau & Baumann 2001). Given that histamine induces larval settlement in H. 83

purpurascens, the 4 other major biogenic amines in invertebrates were tested for settlement activity. The precursor of serotonin (5-hydroxy-tryptophan); histamine-analogues (imidazole, 1-3-aminopropylimidazole and GABA); a range of amino acids and K+ were also tested against H. purpurascens larvae as possible inducers of settlement. The synthetic peptide glycine-glycine-arginine (GGR) is a potent inducer of settlement of barnacle larvae (Tegtmeyer & Rittschof 1989) and of attachment of oyster larvae (Zimmer-Faust & Tamburri 1994). Thus, the effect of GGR on H. purpurascens larvae was investigated.

4.2 Methods

4.2.1 Larval culture

Larvae for the dose-response and exposure assays, as well as larvae for the assays of neuroactive compounds and histamine-analogues, were obtained from adult Holopneustes purpurascens that self spawned in buckets as described in section 3.2.3. Larvae for the amino acid assay were obtained using the spawning technique described in section 3.2.3. Larvae for the assays testing excess K+ and GGR were obtained using the technique described in section 2.2.5.

4.2.2 Settlement Assays

All settlement assays were done in a CTR (19°C, 12-h light/12-h dark regime) under static conditions. Replicates were randomly assigned among treatments. Aliquots of stock solutions of the test compounds were added to assigned sterile petri dishes (36 mm) followed by 4-5 ml SSW. Histamine (10 μM) was used as the positive control in assays testing the dose-response of Holopneustes purpurascens larvae, neuroactive and histamine- analogue compounds and amino acids. The polar extract of Delisea pulchra (50 μg·ml-1) was used as the positive control in assays testing K+ and GGR. SSW and Milli-Q were used as negative controls in all assays. Larvae were added once all dishes were prepared.

Variable numbers of larvae were used per treatment in the different assays depending on availability of larvae (see each section below). Larvae were sparse in the earlier stages of the research project at which times only 10 larvae per treatment were used (assays of amino acids, K+ and GGR). Replication was increased in assays performed later in the study (all other assays) as a revised spawning technique (self spawning of urchins in buckets) yielded many more larvae. Percent settlement was scored at variable times (see each section). Larvae were deemed to have settled (i.e., metamorphosed) when (1) tube-feet were extended and/or attached (2) the larval epidermis had retracted 3) oral side of juvenile was extruded from vestibule showing initial podia and juvenile spines. i) Dose-response of larvae of different ages

The dose-response of Holopneustes purpurascens larvae was investigated to see if older larvae become more sensitive to lower concentrations of histamine. Assays were done with 3 batches of larvae (A, B and C) on day 7 (i.e., newly competent larvae), day 14 (~1 wk post initial-competence), day 21 (~2 wk post initial-competence) and day 28 (~3 wk post initial-competence, batch A only). Histamine was tested at 1 μM, 100 nM and 10 nM in each assay. Treatment and control dishes were prepared as described in section 4.2.2. I used 25 larvae per treatment from batches A and B (5 dishes with 5 larvae), and 100 larvae per treatment from batch C (10 dishes with 10 larvae). Percent settlement was scored at 1, 24, 48 and 72 h in all assays.

A repeated measures ANOVA was used to examine the effects of larval age on the sensitivity of larvae to lower concentrations of histamine (SYSTAT ® 10.0 for Windows). Only histamine concentrations of 1 μM, 100 nM and 10 nM were included in the analysis because 10 μM resulted in 100 % settlement and SSW induced minimal settlement, regardless of larval age. Histamine was treated as a categorical factor rather than a co- variate because of the small range of values but the outcome of the analysis was the same in either case. Finally, values were pooled across all three experimental runs (i.e., batch A, B, and C) with 7-d-old, 14-d-old and 21-d-old larvae for the analysis. Data from batch A were analysed separately to include the response of 28-d-old larvae. For both analyses, Greenhouse-Geisser adjusted p-values were used as ε < 0.75 (Quinn & Keough 2002).

ii) Effect of different exposure times on inducing irreversible metamorphosis

Holopneustes purpurascens larvae were exposed to 10 μM histamine for varying time periods at 7-d-old, 14-d-old and 21-d-old, to determine the exposure time required to induce irreversible metamorphosis. The term metamorphosis is used when referring to these experiments as I am specifically interested in recording the irreversible (and reversible) morphological transformation of larvae to juveniles (rather than other aspects of settlement) which occurs in response to histamine. Histamine was tested at 10 μM in these experiments as this concentration induces 100 % of 7-d-old larvae to commence metamorphosis within 1 h. Newly competent larvae (7-d-old) were exposed to 10 μM histamine for 15 min, 20 min, 30 min, 45 min, 1 h, 2 h, 3 h, 4 h and 5 h; and then were transferred to SSW. The proportion of larvae that had initiated metamorphosis after exposure to histamine (for varying time periods) was scored prior to transferring to SSW. The proportion of larvae that remained metamorphosed after transfer to SSW was recorded 24 h after commencing the experiment. From these data a ‘reversion score’ was calculated which is the difference between the proportion metamorphosed at 24 h (in SSW) and the proportion that appeared metamorphosed after exposure to 10 μM histamine, before transfer to SSW. Thus a positive score indicates that all individuals that had commenced metamorphosis after the exposure period to histamine continued to metamorphose after transfer to SSW, and that additional individuals also commenced and completed metamorphosis in the 24 h period. Conversely, a negative score indicates that a proportion of individuals that had commenced metamorphosis after the exposure period to histamine discontinued metamorphosis after transfer to SSW and reverted back to the larval form.

Larvae were exposed to 10 μM histamine or SSW continuously for 24 h for positive and negative control treatments, respectively. A further set of procedural control dishes was included, transferring larvae from 10 μM histamine dishes to another set of 10 μM histamine dishes at each exposure time, to determine if the transfer process alone affected the metamorphic state of larvae. The assay was repeated with 14-d-old larvae and 21-d-old larvae, however, not all exposure times could be tested due to limited numbers of larvae. Larvae that were 14-d-old were exposed to 10 μM histamine for 1 h, 2 h, 3 h, 4 h and 5 h;

while 21-d-old larvae were exposed to histamine for 30 min, 45 min, 1 h and 2 h (shorter times were selected as older larvae appeared to be metamorphosing completely after 1 h exposure). Fifty larvae were used for each treatment (5 dishes with 10 larvae). A 2-factor (age = fixed factor, exposure time = fixed factor) ANOVA was used to examine the effects of larval age and exposure time on the initiation of irreversible metamorphosis in Holopneustes purpurascens. The ‘reversion scores’ of newly competent larvae (7-d-old) and older larvae (21-d-old) in the 30 min, 45 min, 1 h and 2 h exposure treatments were compared; only these treatments were analysed as not all exposure times were tested with larvae of different ages. iii) Testing neuroactive compounds and histamine-analogues

A range of neuroactive compounds and histamine-analogues were tested in assays with Holopneustes purpurascens larvae, to determine whether neuroactive compounds are general inducers of settlement, or if compounds with a similar structure to histamine induced settlement. Serotonin (5-hydroxy-tyrptamine) and its precursor 5-hydroxy- tryptophan, dopamine (3-hydroxy-tyramine), tyramine, octopamine, imidazole, 1-3- aminopropylimidazole and GABA (γ-aminobutyric acid) were tested at 10 μM. Dopamine and imidazole stock solutions were prepared in ethanol, the serotonin stock solution was prepared in methanol and all other stock solutions were prepared in Milli-Q. Treatment and control dishes were prepared as described in section 4.2.2, however, methanol and ethanol controls were also included. Twenty-five larvae were used per treatment (5 dishes with 5 larvae). Percent settlement was scored at 1, 29, 49 and 75 h. The response of H. purpurascens larvae in these assays was not analysed as only one compound was inductive. iv) Testing amino acids

Larvae of Holopneustes purpurascens were tested against a range of amino acids with different chemical properties. Two neutral amino acids (glycine, alanine); two acidic amino acids (glutamate, aspartate); two polar amino acids (asparagine, glutamine); three basic amino acids (arginine, lysine, histidine); three aromatic amino acids (phenylalanine, tyrosine, tryptophan) and proline, which contains a secondary amino group, were tested at 1 μM and 100 μM. Treatment and control dishes were prepared as described in section 4.2.2 87

with 5 larvae per treatment (5 dishes with 1 larva). Percent settlement was scored at 1 and 24 h. The response of H. purpurascens larvae to amino acids was not analysed as only a single larva was used per treatment replicate. v) Testing K+ and GGR

Larvae of Holopneustes purpurascens were tested against K+ and GGR as these compounds are known to induce attachment or settlement of invertebrate larvae (Yool et al. 1986, Tegtmeyer & Rittschof 1989, Zimmer-Faust & Tamburri 1994). K+ was tested at 100 μM and 200 μM. Assay dishes were prepared as described in section 4.2.2 with 10 larvae per treatment (10 dishes with 1 larva). Percent settlement was scored at 1, 7, 30, 50 and 98 h.

The synthetic peptide-analogue GGR, a potent inducer of settlement of barnacle larvae and settlement behaviour in oyster larvae (Tegtmeyer & Rittschof 1989, Zimmer-Faust & Tamburri 1994), was tested at 10 μM, 100 nM, 1 nM and 10 pM. Assay dishes were prepared as described on p.99 using 10 larvae per treatment (10 dishes with 1 larva). Percent settlement was scored at 2, 24 and 48 h. The response of Holopneustes purpurascens larvae to K+ and GGR was not analysed as only a single larva was used per treatment replicate.

4.3 Results

4.3.1 Dose-response of larvae of different ages

The percentage of larvae that settled in response to the low-range concentrations of 10 and 100 nM histamine increased proportionally with larval age. That is, more 28-d-old larvae settled in response to 10 and 100 nM histamine than 21-d-old larvae after 24, 48 and 72 h; more 21-d-old larvae settled in response to 10 and 100 nM histamine than 14-d-old larvae after 24, 48 and 72 h; and more 14-d-old larvae settled in response to 10 and 100 nM histamine than 7-d-old larvae (Figure 4.1A-C). Only the oldest larvae (28-d-old) settled in

response to 10 nM histamine after 1 h (Figure 4.1A). There was minimal spontaneous settlement in SSW during the assays (Figure 4.1A–C). Between thirty and forty percent of 7-d-old and 28-d-old larvae settled within 1 h in response to 1 μM histamine, however, less than 10 % of 14-d-old and 21-d-old larvae settled rapidly to this concentration (Figure 4.1A). After 24 h, over 70 % of larvae of all ages had settled in response in response to 1 μM (Figure 4.1B), with over 80 % settlement after 48 h. Most larvae (>85 %) of all ages had settled in response to 10 μM histamine after 1 h which increased to 100 % settlement after 24 h.

Larval age strongly affected the minimum concentration of histamine that induced settlement in Holopneustes purpurascens with older larvae responding to lower concentrations of histamine than newly competent larvae (Table 4.1, Figure 4.1A-C). Older larvae settled in greater numbers than newly competent larvae in response to low concentrations of histamine (Table 4.1, Figure 4.1A-C). The dose-response of larvae from batch A at 7- d-old, 14- d-old, 21- d-old and 28-d-old was analysed separately (Table 4.2). Older larvae from batch A also metamorphosed at lower concentrations than newly competent larvae (Figure 4.1A-C), however, in the analysis this interaction was obscured by a significant time x age x conc interaction (Table 4.2).

4.3.2 Effect of different exposure times on inducing irreversible metamorphosis

Larval age strongly affected the induction of irreversible metamorphosis of Holopneustes purpurascens larvae with older larvae requiring less exposure to 10 μM histamine than younger larvae to induce irreversible metamorphosis (Table 4.3). Over 60 % of newly competent larvae (7-d-old) initiated metamorphosis after 20 min exposure to 10 μM histamine (Figure 4.2A) as indicated by 1) the extension and/or attachment of tube-feet, (2) partial retraction of the larval epidermis, 3) extrusion of the juvenile rudiment from vestibule showing initial podia and juvenile spines. However, most of these ‘metamorphs’ reverted back to swimming larvae once removed to SSW when exposed to histamine for 1 h or less, as indicated by the negative reversion scores (approximately -0.6, Figure 4.3A). Following 2 h exposure to 10 μM histamine, approximately half of the ‘metamorphs’ 89

continued to develop into juveniles once transferred to SSW. Newly competent larvae required 3 h continuous exposure to 10 μM histamine in order to maintain the metamorphic state in > 90% of ‘metamorphs’ after transfer to SSW (Figure 4.2A, 4.3A).

Older larvae required less exposure to 10 μM histamine to initiate irreversible metamorphosis. All 14-d-old larvae which had commenced metamorphosis continued to develop into juveniles after 1 h (or longer) exposure to 10 μM histamine, generating reversion scores close to zero (Figure 4.2B, 4.3B). While fewer 21-d-old larvae overall initiated metamorphosis after 30–45 min exposure to 10 μM histamine, most larvae that had commenced metamorphosis continued to develop into juveniles after transfer to SSW (Figure 4.2C, 4.3C). All larvae that were exposed to 10 μM histamine continuously were metamorphosed after 1 and 24 h. None of the 7-d-old or 14-d-old larvae metamorphosed in SSW after 24 h, however, approximately 10 % of 21-d-old larvae metamorphosed in SSW after 24 h. All larvae in the procedural control (transferred from 10 μM histamine to 10 μM histamine) remained metamorphosed.

4.3.3 Larval response neuroactive and histamine-analogue compounds

The catecholamine, dopamine, was the only compound tested which induced settlement of Holopneustes purpurascens larvae (Figure 4.4A-B). After 1 h, 33 ± 16 % of larvae had metamorphosed in response to dopamine (10 μM) and 58 ± 10 % had metamorphosed after 28 h. Dopamine induced normal settlement and all metamorphs were alive after 8 d. 5- Hydroxy-tryptophan was toxic to larvae, all of which had died after 75 h. All larvae settled in response to 10 μM histamine after 1 h and no larvae had settled in SSW nor in the EtOH, MeOH and Milli-Q controls after 24 h.

4.3.4 Larval response to amino acids

None of the amino acids induced rapid settlement (i.e., after 1 h), however, several amino acids induced settlement after 24 h (Figure 4.5A-B). Lysine was the most active amino

acid inducing 80–100 % settlement after 24 h. Histidine, from which histamine is formed, induced up to 40 % settlement of larvae, however, these metamorphs were dead after 48 h. Glutamate, aspartate, glutamine, phenylalanine and tyrosine at 1 μM or 100 μM induced low levels of settlement after 24 h. Phenylalanine and tryptophan at 100 μM were toxic to larvae which died. All larvae settled in response to 10 μM histamine after 1 h, and no larvae settled in SSW or the Milli-Q control after 24 h.

4.3.5 Larval response to K+ and GGR

K+ at 200 μM induced a slow settlement response in Holopneustes purpurascens larvae with no settlement after 1 or 7 h, and approximately 30 % settlement after 30 h exposure (Figure 4.6). K+ at 100 μM induced only 10 % settlement of larvae after 50 h (Figure 4.6). The polar extract of Delisea pulchra (50 μg·ml-1) induced 100 % settlement of larvae after 7 h. There was no settlement in SSW or the Milli-Q control after 98 h.

The synthetic peptide-analogue GGR did not induce settlement of H. purpurascens larvae at any concentration. The polar extract of D. pulchra (50 μg·ml-1) induced 100% settlement of larvae after 2 h with no settlement in SSW or the Milli-Q control after 48h.

50 A - 1 h 7-d-old 40 14-d-old 21-d-old 28-d-old 30

0 100 B - 24 h

20 Percent settlement Percent

0 100 C - 72 h 80

0 SSW 10 nM 100 nM 1 uM

Histamine concentration Figure 4.1 The settlement response of Holopneustes purpurascens larvae (mean ± SE) of different ages to 10 nM, 100 nM and 1 μM of dissolved histamine and SSW at A. 1 h, B. 24 h and C. 72 h. Data from 7-d-old, 14-d-old and 21-d-old larvae were pooled from 3 batches (n = 150 larvae per treatment), however data shown for 28-d-old larvae is from batch A only (n = 25 larvae per treatment). Note the different scales of y-axis.

After Exposure After 24 h

100 A 7-d-old larvae

100 B 14-d-old larvae

20 Percent metamorphosis ** ** 0

100 C 21-d-old larvae

20 ** * ** 0 15min 20min 30min 45min 1 h 2 h 3 h 4 h 5 h

Exposure time Figure 4.2 The percentage of Holopneustes purpurascens larvae (mean ± SE, n = 50 larvae per treatment) that initiated metamorphosis after exposure to 10 μM histamine for varying time periods (black bars), and the percentage of larvae which completed metamorphosis in SSW, 24 h after commencing the experiment (grey bars). The response of larvae at A. 7-d-old, B. 14-d-old and C. 21-d-old are shown. * - exposure time not tested.

0.2 A 7-d-old

0.0

-0.2

-0.4

-0.6

-0.8

0.15 B 14-d-old 0.10

0.05 *** 0 0.00

-0.05

-0.10 -Metamorphosis

Δ -0.15

-0.20

0.15 C 21-d-old 0.10

0.05 * 0.00 * **

-0.05

-0.10 20min30min45min1h2h3h4h5h

Duration of exposure

Figure 4.3 The reversion scores (Δ-metamorphosis) of A. 7-d-old, B. 14-d-old and C. 21- d-old Holopneustes purpurascens larvae after exposure to 10 μM histamine for varying time periods. The reversion score was calculated as the difference between, the proportion of individuals that were metamorphosed (in SSW) 24 h after commencing the exposure to histamine, and the proportion that were metamorphosed after exposure to 10 μM histamine before transfer to SSW. Thus a positive score indicates that all ‘metamorphs’ (plus additional individuals) continued metamorphosis after transfer to SSW while a negative score indicates that a proportion of ‘metamorphs’ reverted back to the larval form after transfer to SSW. * - not tested, 0 – no change in metamorphosis

H N N 2 H

Histamine

H2N OH

Dopamine

H2NOH2N

Lysine

Figure 4.4A The chemical structures of compounds which induced normal settlement of Holopneustes purpurascens larvae.

NH N

H N N H2N 2

Serotonin 1-[3-aminopropyl]imidazole OH

NH H2N N

O N H OH Imidazole

OH 5-hydroxy-tryptophan

H2N OH OH

NH HO 2

Tyramine

Octopamine O

H2N HO

GABA (γ-aminobutyric acid)

Figure 4.4B Structures of neuroactive compounds and histamine-analogues which were tested against Holopneustes purpurascens larvae and did not induce settlement.

100 A.

100 B. 80 Percent settlement Percent

0 e te e e e e e ne n at ine n n n an ine in ni m n MQ amine a art a ys SSW Glyci Al p aragi Proline L ist s lut ylalani Tyrosi Argi Histidine H A G n Glutama Asg Tryptoph Phe Amino acid

Figure 4.5 The settlement response of Holopneustes purpurascens larvae after 24 h incubation with various amino acids at A. 100 μM and B. 1 μM (n = 5 larvae per treatment). Histamine at 100 μM was included as a positive control and SSW and Milli-Q (MQ) were included as negative controls.

100 30 h 50 h

40 Percent settlement Percent

0 PE K+ (200) K+ (100) MQ SSW Treatment

Figure 4.6 The settlement response of Holopneustes purpurascens larvae after 30 and 50 h exposure to K+ tested at 200 μM (200) and 100 μM (100). The polar extract of Delisea pulchra (50 μg·ml-1) was included as a positive control and SSW and Milli-Q (MQ) were included as negative controls.

Table 4.1 Repeated measures ANOVA of the effects of larval age (7-d-old, 14-d-old and 21-d-old), histamine concentration and duration of exposure to histamine on percentage settlement of Holopneustes purpurascens. Data are pooled across Batch A, B, and C. Note that Greenhouse-Geisser adjusted p values are used for the within subjects test as Greenhouse-Geisser ε = 0.62.

Source df MS F p Between Subjects Concentration 2 304506 778.754 <0.001 Age 2 3340 8.542 <0.001 Concentration x Age 4 4249 10.868 <0.001 Error 171 391 Within Subjects G-G Exposure Time 3 40126 396.1 <0.001 Exposure Time x Concentration 6 19146 189.0 <0.001 Exposure Time x Age 6 982 9.7 <0.001 Exposure Time x Age x Conc. 12 57 0.6 0.791 Error 513 101

Table 4.1 Repeated measures ANOVA of the effects of larval age, histamine concentration and duration of exposure to histamine on percentage settlement of Holopneustes purpurascens. Data are for Batch A only at 7-d-old, 14-d-old, 21-d-old and 28-d-old. Note that Greenhouse-Geisser adjusted p values are used for the within subjects test as Greenhouse-Geisser ε = 0.52.

Source df MS F P Between Subjects Concentration 1 220164 321.2 0.000 Age 3 3963 5.8 0.002 Concentration x Age 3 1225 1.8 0.161 Error 51 685 Within Subjects G-G Exposure Time 3 1590 13.6 0.000 Exposure Time x Concentration 3 11714 100.4 0.000 Exposure Time x Age 9 420 3.6 0.007 Exposure Time x Age x Conc. 9 367 3.2 0.014 Error 153 116

Table 4.3 Two-factor ANOVA of the effects of larval age (7-d-old and 21-d-old) and duration of exposure to histamine (30–120 min) on the induction of irreversible metamorphosis of Holopneustes purpurascens larvae.

Source df MS F p Age 1 3.691 146.2 0.000 Exposure time 3 0.017 0.663 0.581 Age x Exposure time 3 0.054 2.122 0.117 Error 32 0.025

100

4.4 Discussion

4.4.1 Older larvae become more sensitive to histamine as a settlement cue

Larval age strongly affected the minimum concentration of histamine that induced settlement in Holopneustes purpurascens, with older larvae settling to much lower concentrations of histamine than younger larvae. This study is the first of its kind to directly examine the effect of a pure form of a naturally occurring settlement cue on the dose-response curve of settlement in young versus older H. purpurascens larvae. This study supports the observations of Botello and Krug (2006) who also found that older larvae were more sensitive to lower concentrations of settlement cue (present in extracts of the alga Vaucheria longicaulis) than younger larvae. H. purpurascens joins a growing list of species that show shifts in the settlement response of larvae as they age (reviewed in Elkin and Marshall, in press). Given that H. purpurascens recruits are found on a number of different substrata (Swanson et al. 2006), it seems reasonable to assume that this species, whilst having settlement preferences, is not a strict specialist but will settle and survive on several different host algae. Thus, whilst larvae settling on preferred host algae may have higher fitness (Williamson et al. 2004), larvae that settle out onto lower ranked hosts may still survive. For non-feeding larvae, this shift to a broader range of host algae in order to settle is predicted by the theoretical model of Elkin and Marshall (in press). As larvae accumulate direct (planktonic mortality) and indirect (reduced energetic stores) costs in the plankton, the benefits of settling in a sub-optimal habitat increase.

Despite older larvae reacting to lower concentrations of histamine, only a very small proportion (<5 %) of very old larvae (28-d-old) settled spontaneously in the absence of any algal cues. Very few older larvae settle spontaneously in seawater during culture, instead requiring exposure to cue/s present in algae for settlement for at least 3 wk after attaining competence (R. Swanson pers. obs.). Thus Holopneustes purpurascens does not become indiscriminate, merely more sensitive to the settlement cue. However, given that histamine is found at lower concentrations in brown algae, this increased sensitivity is likely to result

101

in older larvae accepting a broader range of host algae. My results appear therefore to be a novel example of expanding settlement preferences by becoming more responsive to a single settlement cue which can occur here because of quantitative differences in the concentration of histamine among algae in the field (Swanson et al. 2006). By using histamine concentration as a proxy for a general habitat cue, complex changes in settlement behaviour (initially rejecting brown algae as hosts but later accepting them) derive from very simple changes in the response to a single cue.

4.4.2 Older larvae metamorphose faster than younger larvae

A number of studies have found that older polychaete, gastropod and abalone larvae metamorphose faster than younger larvae (Knight-Jones 1953, Barlow 1990, Pechenik & Cerulli 1991, Botello & Krug 2006). Similarly, in order to metamorphose older larvae of Haliotis asinina required shorter exposure periods to inductive algae than younger larvae (Jackson et al. 2005). Holopneustes purpurascens larvae initiate metamorphosis rapidly in the presence of 10 μM histamine, undergoing massive tissue reorganisation within 20 min, regardless of age (R. L. Swanson, pers.obs). Over 90 % of 7-d-old larvae appeared to have metamorphosed after 45 min exposure to 10 μM histamine, however, most ‘metamorphs’ reverted back to larvae when placed in SSW (i.e., when removed from the cue). Despite drastic changes in morphology, apparently other components of the morphogenetic (metamorphic) pathway were not activated sufficiently for metamorphosis to complete. Newly competent larvae required 3 h of continuous exposure to 10 μM histamine, for all individuals which had commenced settlement to complete it. On the other hand, 21-d-old larvae required less than 1 h of exposure to 10 μM histamine to complete settlement. Thus older larvae are apparently committed to metamorphose once the transformation is initiated, whereas younger larvae appear to be more flexible at the time of settlement, having the potential to halt the settlement process for some period after its initiation if they can no longer detect the settlement cue.

In Chapter 2, I proposed that it was likely that an ionotropic histamine receptor was involved in the metamorphic pathway of larval Holopneustes purpurascens as ionotropic 102

receptors trigger rapid responses. This was based on the belief that 100 % of H. purpurascens larvae were metamorphosing completely after 1 h exposure to ~10 μM histamine, which appeared to be the case from the drastic changes in morphology that had occurred in these ‘metamorphs’. However, from the exposure experiments reported in this chapter, it is now clear that H. purpurascens larvae require 3 h of continuous exposure to 10 μM histamine for the induction of irreversible metamorphosis in all individuals (which have commenced metamorphosis). This opens the possibility that a metabotropic histamine receptor is involved in the metamorphic pathway of larval H. purpurascens as these receptors trigger slower cellular responses via second messenger pathways. A metabotropic histamine receptor has recently been identified on eggs of the sea urchin Strongylocentrotus purpuratus (Leguia & Wessel 2006) which may be the receptor involved in the metamorphosis of larval H. purpurascens (see section 6.2.1)

The increased sensitivity of Holopneustes purpurascens larvae to histamine occurred gradually with age, suggesting that there is a progressive decrease in the stimulus-threshold required to induce an aging larva to metamorphose, as noted for other invertebrate larvae (Knight-Jones 1953, Crisp 1974, Pechenik 1980, Rittschof et al. 1984, Coon et al. 1990, Gibson 1995). The mechanism in older larvae that leads to increased sensitivity to histamine and a reduction in exposure time required for settlement are unclear but are likely to have the same basis. Jackson et al. (2005) refer to ‘competence factors’ reaching a critical level in order for larvae to attain competency. These factors may include chemoreceptors (Trapido-Rosenthal & Morse 1986b), components of internal signalling pathways (Clare et al. 1995, Knight et al. 2000) or transcription factors (Jackson et al. 2005). These competence factors may continue to accumulate in older larvae meaning they are ‘primed’ to respond to settlement cues, both to lower concentrations and more quickly than younger larvae (Jackson et al. 2005).

Changes in endogenous levels of neurotransmitters, for example, may affect sensitivity to cues. Catecholamines, such as dopamine and its precursor L-DOPA, appear to modulate competency and control settlement in the gastropods Crepidula fornicata (Pires et al. 2000b, Pechenik et al. 2002) and Phestilla sibogae (Pires et al. 2000a). Increasing the

103

endogenous dopamine concentrations of competent P. sibogae larvae made them more sensitive to the settlement cue present in coral extract (Porites compressa, (Pires et al. 2000a). In the gastropod Ilyanassa obsolete, settlement only occurs following the removal of endogenous levels of nitric oxide (NO), a specific inhibitor of settlement (Leise et al. 2001). Bath application of serotonin induced settlement of I. obsolete but not in the presence of two NO donors, and settlement of I. obsolete was induced, in the absence of serotonin or any other inducer, when endogenous nitric oxide production was inhibited (Leise et al. 2001). Other studies suggest that internal energy reserves also affect the sensitivity to settlement cues. Botello and Krug (2006) found that unfed larvae were more sensitive to settlement cues than fed larvae and Marshall and Keough (2003a) found that larger larvae (with more nutritional reserves) can delay settlement in the absence of cues for longer than smaller larvae.

4.4.3 Specificity of the larval settlement response

Most models for the metamorphosis of invertebrate larvae propose a stimulatory pathway, in which an externally excitable cell is depolarised by a threshold level of inducer-receptor binding, which sets off a cascade of stimulatory events that alter gene expression and bring about metamorphosis (Yool et al. 1986, Baxter & Morse 1987, Bonar et al. 1990, Hadfield & Pennington 1990, Degnan & Morse 1995). Larval chemoreceptors involved in settlement have not been isolated but the receptor which binds the settlement cue in abalone Haliotis rufescens has been characterised extensively through biochemical/ pharmacological studies (Trapido-Rosenthal & Morse 1985, 1986b, a). Receptor organs, chemosensory neurons and neurons farther downstream in the settlement pathway have been identified in just a few cases (Burke 1983, Arkett et al. 1989, Hadfield et al. 2000, Leise & Hadfield 2000, Pasternak et al. 2005).

Of all the synthetic compounds tested against Holopneustes purpurascens, only dopamine induced rapid settlement of larvae (in addition to synthetic histamine). K+ induced a slow rate of settlement in larvae which suggests that ion fluxes are a component of the morphogenetic pathway in H. purpurascens, as is thought to be the case for most larvae 104

(Trapido-Rosenthal 2001). Lysine, a basic amino acid, induced a slow rate of settlement in most larvae. Lysine facilitated GABA induction of settlement of abalone larvae, but did not induce settlement when tested alone (Trapido-Rosenthal & Morse 1985). Several other amino acids induced low levels of settlement of H. purpurascens larvae but at a slow rate which suggests non-specific interference with the morphogenetic pathway. Some invertebrate larvae can take up dissolved amino acids from seawater (Jaeckle & Manahan 1989, Manahan et al. 1989, Shilling & Manahan 1994) and therefore inductive amino acids may act on the nervous system internally (Coon et al. 1990).

Dopamine, like histamine, contains an ethylamine chain (-[CH2]2NH2) in its structure; however, tryamine, serotonin and GABA also contain this moiety (Figure 4.4A-B) but none of these compounds induced settlement. Since the induction of settlement by dopamine does not appear to be due to non-specific binding of the ethylamine chain, dopamine may be involved in the true morphogenetic pathway in Holopneustes purpurascens. Dopamine induced rapid and complete metamorphosis of H. purpurascens larvae hence it is likely that dopamine is the inducing compound, rather than hydrogen peroxide which is produced by catechol autoxidation which occurs rapidly in seawater (Pires & Hadfield 1991). Hydrogen peroxide in oxidised solutions of catecholamines induced the partial metamorphosis of Phestilla sibogae larvae (Pires & Hadfield 1991). While it is difficult to interpret findings from experiments in which whole larvae are bathed in test solutions with any certainty corroboration of such findings with immunocytochemical evidence means that such studies can provide valuable information (Pires et al. 2000a, Pennati et al. 2001, Zega et al. 2005). Indeed, dopamine and serotonin appear to modulate competency and control settlement in gastropod larvae (Pires et al. 2000b, Leise et al. 2001, Pechenik et al. 2002).

Using immunofluorescence, L-DOPA, dopamine and serotonin have been localised to sensory structures thought to be involved during settlement. Manipulations of catecholamine biosynthesis in Phestilla sibogae larvae, increased concentrations of catecholamines in specific populations of cells located near, perhaps within, the apical sensory organ (Pires et al. 2000a) which has been implicated as the chemosensory organ responsible for transduction of the coral inducer stimulus (Hadfield et al. 2000). In larvae

105

of the ascidian Phallusia mammillata, serotonin-like immunoreactivity was localised in primary neurons of the adhesive papillae and tail (Pennati et al. 2001) while dopamine-like immunoreactivity was localised in some neurons of the adhesive papillae (Zega et al. 2005). The effects of dopamine, serotonin and agonists/antagonists, were explored further using biochemical manipulations: the authors concluded that dopamine signalling delayed settlement whilst serotonin signalling triggered it (Zega et al. 2005). Serotonin-like immunoreactivity was conspicuous in the brachial arms of the brachiolaria larvae of the sea star Patiriella regularis, the structures which comprise the brachiolar attachment complex by which larvae attach to substrata during settlement (Chee & Byrne 1999).

Similarly, serotonin-like immunoreactivity was localised to regions connected to eye spots (used for phototaxis) in larvae of the bryozoan Bugula neritina, the equatorial nerve-muscle ring, and in two tracts extending from the apical disc to this ring (Pires & Woollacott 1997). Further investigation found that serotonin and dopamine have opposite effects on phototaxis in these larvae (Pires & Woollacott 1997). Shimizu et al. (2000) confirmed the findings of Pires and Woollacott (1997) but also found serotonin-like immunoreactivity in the neural plexus, and in tracts connecting the neural plexus to ciliated cells bordering the pyriform organ. The ciliated cells of the pyriform organ and lining the pyriform groove are thought to act as a sense organ during searching behaviour (Woollacott & Zimmer 1971) which are held in close proximity with settlement substrata during searching behaviour (Loeb & Walker 1977).

A common theme has emerged from almost three decades of research, that many behaviours in invertebrates are initiated, maintained, altered or terminated by the action of monoamine or peptide modulators on neural networks (Hardie 1989, Rittschof et al. 1989, Hashemzadeh-Gargari & Freschi 1992, Zimmer-Faust & Tamburri 1994, Huber et al. 1997, Beltz 1999, Stuart 1999, Blenau & Baumann 2001, Pierobon et al. 2004). Monoamines seem to be particularly important in chemoreception in invertebrates (McClintock & Ache 1989, Ono & Yoshikawa 2004). While much less is known about their role in invertebrate larvae, recent findings would suggest that monoamines are major regulators of larval invertebrate behaviour (Pires & Woollacott 1997, Shimizu et al. 2000, Leise et al. 2001,

106

Python & Stocker 2002, Swanson et al. 2004, Zega et al. 2005). Holopneustes purpurascens larvae metamorphose specifically in response to its settlement cue, the monamine histamine, which is produced in high quantities by a preferred host alga of new recruits, Delisea pulchra. Another monoamine, dopamine, also induced rapid settlement hence it is probable that dopamine is a neurotransmitter in the metamorphic pathway in H. purpurascens larvae.

107 CHAPTER 5

Is histamine a general inducer of settlement in echinoderm larvae?

5.1 Introduction

Considering that histamine induces settlement of Holopneustes purpurascens larvae and given the broad functions of histamine as a neurotransmitter in invertebrate nervous systems (Elias & Evans 1983, Claiborne & Selverston 1984, Hashemzadeh-Gargari & Freschi 1992, Stuart 1999), it is possible that histamine may induce larval settlement in other sea urchins or echinoderms. Because so few settlement cues have been unequivocally identified, we do not know if settlement cues are generally highly species-specific or broadly active across species within a particular class or phylum.

Echinoidea are among the best studied marine invertebrates in the fields of reproduction, developmental biology and larval biology (Emlet et al. 2002). Relationships among the regular and irregular urchins (Class Echinoidea) are presented in a highly robust topology supported by four independent data sets (Figure 5.1A, Smith 1997). However, the interrelationships of the echinoderm classes (Figure 5.1B) are still actively debated (Littlewood et al. 1997, Smith 1997) although there is strong support for placing crinoids as a sister group to the other 4 classes, and for pairing holothurians and echinoids as sister taxa (Littlewood et al. 1997, Smith 1997). The relationships between the asteroids, ophiuroids and the holothurian + echinoid pairing have proved difficult to disentangle which suggests that the three groups may have arisen very close in time (Littlewood et al. 1997, Smith 1997). This phylogeny provides a unique opportunity to test histamine as a general cue for settlement of larvae from the Echinoidea and from each echinoderm class. This allows for a direct test of a current hypothesis in the evolution of development, that developmental

108 pathways are convergent in species that have independently evolved direct development (Raff et al. 2003, Byrne & Voltzow 2004).

Indirect development through a feeding echinopluteus larva, or planktotrophy, is viewed as the ancestral condition for sea urchins (Strathmann 1985). Over the course of their evolution, numerous species of sea urchins (and other echinoderms) have independently evolved a non-feeding, directly developing larva, or lecithotrophy , including the Australian sea urchins Holopneustes purpurascens and Heliocidaris erythrogramma. In sea urchins, this shortened form of development has produced a diverse array of simplified larval forms resulting from the reduction or complete loss of pluteal features (i.e., larval arms and ciliary bands) (Byrne & Voltzow 2004). The larval phenotype of H. purpurascens is among the most modified seen in direct developing echinoids without any vestigial pluteal features (Byrne & Voltzow 2004). The simple larvae of H. erythrogramma are atypical of modified echinoid larvae in having retained a prominent ciliary band (Byrne et al. 2001).

A switch from feeding to non-feeding larvae in sea urchins illustrates how convergence in development can be associated with convergence in anatomy (i.e, similar morphology of non-feeding larvae, Wray 2002). Wray (2002) suggested that convergent developmental mechanisms might underlie such convergent phenotypes, which is the basis of the hypothesis of convergence in evolution of direct development (Raff et al. 2003, Byrne & Voltzow 2004). Hence similar morphogenetic pathways in direct developers may reflect the convergence of underlying molecular pathways (Raff et al. 2003). It follows that echinoderm species which have independently evolved direct development may metamorphose in response to the same inducer.

Histamine was tested as an inducer of settlement on a range of larvae subject to availability. Non-feeding and feeding sea urchin larvae were tested against histamine, including the direct developers Holopneustes inflatus (Order Temnopleuroida) and Heliocidaris erythrogramma (Order Echinoida), and the indirect developers Tripneustes gratilla (Order Temnopleuroida) and Heliocidaris tuberculata (Order Echinoida). Histamine was also tested against two other echinoderm species with lecithotrophic development, the sea stars

109 Meridiastra spp. from Class Asteroidea and the brittle star Clarkcoma canaliculata from Class Ophiuroidea. Crinoid or holothuroid larvae were not available for testing.

5.2 Materials and Methods

5.2.1 Larval culture and settlement assays testing histamine

Larvae of Holopneustes inflatus, Holopneustes purpurascens and Heliocidaris erythrogramma were cultured and assayed at the School of Biological, Earth and Environmental Science UNSW. Larvae of Heliocidaris tuberculata, Meridiastra spp. and Clarkcoma canniculata were cultured (by students and staff of Professor Maria Byrne) in the Evolutionary Development Laboratory at the University of Sydney. I conducted the histamine assays on these species at this laboratory, except for the H. tuberculata assay which was performed by Tom Prowse (PhD student). Tripneustes gratilla larvae were cultured at NSW Department of Primary Industries, Abalone Research Facility (Tomaree Headland, Port Stephens) by Dr Symon Dworjanyn who conducted the histamine assay on site.

All settlement assays testing histamine were done under static conditions. At UNSW assays were conducted in a CTR (19°C under a 12-h light/12-h dark cycle). All assays were set up as follows unless otherwise noted. Replicates were randomly assigned among treatments. Aliquots of stock solutions of histamine were added to assigned sterile petri- dishes (36 mm) followed by 4–5 ml SSW (containing antibiotics, section 2.2.5). SSW was used as a negative control. Larvae were added once all dishes were prepared. Variable numbers of larvae were used per treatment in the different assays depending on availability (see each section). Holopneustes inflatus and Holopneustes purpurascens were tested against histamine and a range of host plants (algae and seagrasses) used by these urchins.

110 5.2.2 Class Echinoidea, Order Temnopleuroida, Family Temnopluridae

Holopneustes inflatus and Holopneustes purpurascens

i) Larval Culture

Holopneustes inflatus produce simple ovoid non-feeding larvae. H. inflatus (20 adults) were collected on SCUBA from seagrass beds off Silver Beach, Kurnell (longitude, latitude) (Botany Bay, Sydney). H. inflatus were collected from beds of Halophila ovalis, however, Zostera capricornia and Posidonia australis were also present. Silver Beach is in the sheltered area of Botany Bay, approximately 1 km to the south-west of Bare Island, where Holopneustes purpurascens (20 adults) were collected for this study. Adult H. inflatus and H. purpurascens were transported back to the laboratory in ambient seawater and were held in a CTR (19°C) as described in section 3.2.3. H. inflatus and H. purpurascens embryos were obtained from the natural spawning of urchins in buckets and were cultured as described in section 3.2.3. H. inflatus and H. purpurascens larvae were cultured as described in section 2.2.5. H. inflatus larvae were indistinguishable from H. purpurascens larvae. H. inflatus larvae became competent at a similar age to H. purpurascens larvae (i.e., 6 d post-fertilisation) as shown by the spontaneous settlement of larvae in culture vessels at this time. Approximately one-quarter of H. inflatus larvae settled spontaneously in culture which was considerably more than ever occurred for H. purpurascens larvae. Different batches of H. inflatus larvae were used for the histamine and algal/seagrass assays. ii) Settlement assay with histamine

A pilot study indicated that histamine (10 μM) induced rapid settlement of larval Holopneustes inflatus. The response of H. inflatus larvae to a range of histamine concentrations (10 μM, 1 μM, 100 nM and 10 nM) was tested in a settlement assay. Treatment and control dishes were prepared as outlined in section 5.2.1 with 100 competent (6-d-old) larvae per treatment (10 dishes with 10 larvae). Percent settlement was scored at 1, 24 and 48h. The dose-response of competent H. inflatus larvae was compared to the dose-response of competent Holopneustes purpurascens larvae from the experiment

111 reported in Chapter 4 (section 4.3.1, response of larval batch C at competency). Statistical analysis of the larval response to histamine used a repeated measures 2-factor ANOVA (species=fixed factor, histamine concentration=fixed factor). iii) Settlement assay with host plants

Common host plants of Holopneustes inflatus and Holopneustes purpurascens were tested as possible inducers of larval settlement of H. inflatus. Delisea pulchra and Amphiroa anceps were selected as they are the most potent inducers of settlement of larval H. purpurascens (Swanson et al. 2006). Ecklonia radiata was tested because it is the primary host plant of adult H. purpurascens at Bare Island (Williamson et al. 2004), however, adult H. inflatus were also found on E. radiata at Kurnell. Three common species of seagrass, Halophila ovalis, Zostera capricornia and Posidonia australis, were tested as H. inflatus occurs in seagrass beds containing these species at Kurnell. Fresh algae and seagrass were collected from Bare Island and Kurnell, respectively. Small pieces of alga or seagrass (10- 30 mg) were placed in dishes with 4 ml SSW. Histamine (10 μM) was included as a positive control. Thirty competent H. inflatus larvae were used per treatment (10 dishes with 3 larvae), however, 50 competent H. purpurascens larvae were used per treatment (10 dishes with 5 larvae). Percent settlement was scored at 1 and 24 h. The response of H. purpurascens and H. inflatus to algae and seagrass was analysed by a repeated measures 2- factor ANOVA (species=fixed factor, plant=fixed factor).

5.2.3 Class Echinoidea, Order Temnopleuroida, Family Toxopneustidae

Tripneustes gratilla

i) Larval culture

Tripneustes gratilla produce feeding echinopluteas (8-arm) larvae. Adults were collected from the shallow sub-tidal habitat near the Abalone Research Facility at Tomaree Headland, Port Stephens (S 32°42.815’, E 152°11.053’). Urchins were maintained in outdoor flow-through seawater tanks at the facility and were fed a diet of mixed macroalgae. Urchins were induced to spawn by intra-coelomic injection of 1 ml of 2M

112 KCl. Inverted males and females shed gametes into 200 ml containers filled with ambient seawater. The eggs of three females were pooled and fertilised with the sperm of three males. The fertilised eggs were placed into a 125 l conical bottomed rearing container filled with filtered (1 μm), UV-sterilised seawater (SFSW). The system was aerated and maintained at 25°C.

Embryos hatched after 24 h and larvae were initially stocked at 10 larvae·ml-1. Two tank volumes (250 L) of SFSW were exchanged in the rearing tank each day at a low flow rate. Larvae were retained within the tank using a banjo screen (80 μm). The larval cultures were transferred into clean tanks weekly. Larval density was reduced to 4 larvae·ml-1 toward the end of the rearing period, approximately 5 wk post-fertilisation. Larvae from 3- d-old were fed Chaetoceros muelleri at 4 000 cells·larva.day-1 and this was gradually increased to 10 000 cells·larva.day-1 by the eight-arm pluteus stage. ii) Settlement assay with histamine

Larvae were approximately 5-wk-old at the time of the assay when at least 50 % of larvae had a rudiment larger than their gut. The large spread in the development of larvae meant that a percentage of larvae that were used in the settlement assay were not competent to settle. The similarity in sizes of competent and non-competent larvae made sorting them by size unpractical. The settlement assay was done at 25ºC in a controlled temperature room at the Abalone Research Facility, Tomaree. Histamine was tested as a possible inducer of settlement of Tripneustes gratilla larvae at 100 μM, 10 μM, 1 μM and 100 nM. Treatment and control dishes were set up as described in section 5.2.1 except that SFSW was used. Corallina officinalis (~30 mg) was included in the assay as T. gratilla larvae have been observed to settle in response to this alga (S. Dworjanyn pers. obs). Twenty larvae were used per treatment (5 dishes with 4 larvae). Percent settlement was recorded after 48 h. Larvae were recorded as settled if they were attached to the dish and had developed tube feet and spines. The assay was repeated with a different batch of larvae.

113 5.2.4 Class Echinoidea, Order Echinoida, Family Echinometridae

Heliocidaris erythrogramma

i) Larval culture

Heliocidaris erythrogramma produce simple barrel-shaped non-feeding larvae. H. erythrogramma (15 adults) were collected from Bare Isl. (Botany Bay, Sydney) and transported back to the laboratory in ambient seawater. At least 3 females and 3 males were spawned by intracoelomic injection of 3-5 ml of 0.5M KCl. Spawned egg strings were collected and transferred to a beaker containing aerated SSW. Sperm was collected in pasteur pipettes and transferred to a dish on ice. Pooled sperm was mixed together and approximately 0.1 ml of sperm concentrate was added to the egg suspension for 10 min. Fertilised eggs were rinsed 3 times in SSW to remove excess sperm, and were cultured in 3 x 2 l beakers at a maximum density of 5 embryos·ml-1. SSW was changed daily until larvae reached competency (4 d post-fertilisation). ii) Settlement assay with histamine

Histamine was tested as an inducer of settlement of Heliocidaris erythrogramma larvae at 200, 100, 50, 10 and 1 μM. Amphiroa anceps was included as a positive control (Huggett et al. 2006). Treatment and control dishes were set up as described in section 5.2.1 with 50 competent larvae used per treatment (10 dishes with 5 larvae). Percent settlement was scored at 1, 24 and 48 h. The larval response to histamine at 48 h was analysed by a 1- factor ANOVA and Bonferroni’s pairwise comparisons to determine which treatments differed from one another.

5.2.5 Heliocidaris tuberculata

i) Larval culture

Heliocidaris tuberculata produce feeding echinopluteas (8-arm) larvae. H. tuberculata were collected from Chowder Bay, Sydney Harbour and were cultured at the Evolutionary

114 Development Laboratory, University of Sydney. Three females and 3 males were spawned by intracoelomic injection of 2 ml of 0.5M KCl. Eggs from each female were fertilised separately with a dilute solution of sperm (pooled from 3 males). After 10 min, eggs were rinsed with filtered (1 μm) ambient seawater (FSW) to remove excess sperm. Embryos from each female were cultured separately in 600 ml beakers for 2 d and then larvae (early 4-arm stage) were transferred to 2 l beakers (3 beakers per female) and stirred gently by a paddle system. FSW was changed every 3–4 d. Cultures were stocked at an initial density of 10 larvae·ml-1 which quickly decreased to 2 larvae·ml-1 at 1-wk-old, and 1 larva·2ml-1 by 7-wk-old. Larvae were fed Chaetoceros muelleri (cultured in modified F2 medium plus Si -1 as Na2O3Si) at 20 000 cells·ml every other day commencing at 3-d-old until reaching competency around 7-wk-old. ii) Settlement assay with histamine

Histamine was tested as an inducer of settlement of Heliocidaris tuberculata larvae at 10 μM only due to low numbers of larvae reaching competency (8-arm stage). The assay was conducted at 19–23°C. Corallina officinalis (~30 mg) was included in the assay as competent H. tuberculata larvae generally settle in the presence of this alga (M. Byrne pers. obs.). Treatment and control dishes were set up as outlined in section 5.2.1 except that FSW was used instead of SSW. Fifteen competent larvae (~ 7-wk-old) were used per treatment (5 dishes with 3 larvae). Percent settlement was scored at 24, 48 and 85 h.

5.2.6 Class Asteroidea, Order Valvatida, Family Asterinidae

Meridiastra calcar

i) Larval culture

The sea-star Meridiastra calcar produces non-feeding braciolaria larvae. M. calcar were collected from Clovelly (Sydney, Australia) and transported back to the laboratory. Eggs from 1 female were obtained by placing dissected ovaries into separate culture dishes which contained 10 μM 1-methyladenine in FSW. Ripe testes were removed from 1 male

115 and stored dry at 4°C until sperm was used for fertilisation. Spawned eggs were added to several culture dishes containing FSW and were fertilised with diluted sperm. After 10 min, eggs were rinsed with FSW to remove excess sperm. Embryos were cultured at 19– 23°C in FSW in multiple 500 ml beakers. FSW was changed after hatching and then every 3–4 d. M. calcar are competent to metamorphose at approximately 10-d-old, however, the first histamine assay was done with 24-d-old larvae (i.e., 2 weeks beyond initial competence). A second batch of M. calcar larvae were tested at competence (i.e., 10-d- old). These larvae were obtained using the same method described above except that eggs were obtained from a female collected from Ulladulla (long, lat) (NSW, Australia). Sperm used to fertilise eggs were from 1 male M. calcar collected from Clovelly. ii) Settlement assay with histamine

Histamine was tested as an inducer of settlement of Meridiastra calcar in two separate assays. Histamine was tested at 100, 10 and 1 μM. Corallina officinalis (~30 mg) was included in each assay as competent M. calcar larvae generally settle in the presence of this alga (M. Byrne pers. obs). An additional histamine treatment of 100 nM was included in the second assay. Each assay was conducted at 19–23°C in tissue culture plates (12-wells). Aliquots of stock solutions of histamine were added to wells followed by 3 ml SSW. Larvae were added once all dishes were prepared. Fifteen aged (24-d-old) larvae were used per treatment (5 wells with 3 larvae) in the first assay, and 25 competent (10-d-old) larvae were used per treatment (5 wells with 5 larvae) in the second assay. Percent settlement was scored approximately every 24 h for 1 wk.

5.2.7 Meridiastra hybrids - female Meridiastra oriens with male Meridiastra occidens

i) Larval culture

Meridiastra hybrid larvae were available and were included in the study. Meridiastra oriens were collected from Clovelly (Sydney, Australia) and transported back to the laboratory in buckets. Meridiastra occidens adults were sent from Perth (Western Australia). Eggs from one female M. oriens were obtained by placing dissected ovaries

116 into separate culture dishes which contained 10 μM 1-methyladenine in FSW. Ripe testes were removed from one male M. occidens and stored dry at 4°C until sperm was used for fertilisation. Spawned eggs from each female were transferred to culture dishes with FSW and were fertilised with diluted sperm from separate males. After 10 min, eggs were rinsed with FSW to remove excess sperm. Embryos were cultured at 19–23°C in FSW in multiple 500 ml beakers. Larvae were kept in low densities (1 larva·ml-1) in 200 ml culture dishes aerating occasionally. FSW was changed after hatching and then every 3–4 d. M. oriens/occidens are competent to metamorphose after approximately 10 d. ii) Settlement assay with histamine

Histamine was tested as a possible inducer of settlement of Meridiastra oriens/occidens hybrid larvae as described for Meridiastra calcar in section 5.2.6. Histamine was tested at 100, 10 and 1 μM, and 100 nM. Corallina officinalis (~30 mg) was included in the assay as competent Meridiastra sp. larvae generally settle in the presence of this alga (M. Byrne pers. obs.). Twenty-five competent larvae (5 wells with 5 larvae) were used per treatment. Percent settlement was scored after 24, 48 and 85 h.

5.2.8 Class Ophiuroidea, OrderOphiurida, Family Ophiocomidae

Clarkcoma canaliculata

i) Larval culture

The brittle-star Clarkcoma canaliculata produces non-feeding vitellaria larvae. C. canaliculata (~10) were collected from Clovelly (Sydney, Australia) and transported back to the laboratory. C. canaliculata (10) were induced to spawn by the temperature shock method (Selvakumaraswamy & Byrne 2000) as they failed to spawn spontaneously within 3 h of returning to the laboratory. A temperature of 34°C was chosen as this temperature induced immediate release of sperm and a few seconds at this temperature does not cause mortality to ophiuroids (Selvakumaraswamy & Byrne 2000). Individuals were transferred repeatedly between 34°C and ambient temperature (19- 23°C) seawater, until they released spawn, and were then kept in the dark for 30 min in a container with aerated seawater at

117 ambient temperature. Embryos were transferred to 500 ml beakers containing filtered FSW and aerated until larvae were hatched. Larvae were transferred to culture dishes (200 ml) containing FSW at a maximum density of 1 larva·ml-1 and were cultured at 19–23°C for 10 days until competent. Seawater was changed every fourth day. ii) Settlement assay testing histamine

In the laboratory Clarkcoma canaliculata larvae settle in glass dishes without adding a biological substrate. Nevertheless, histamine was tested at 10 and 1 μM against larvae of C. canaliculata in a settlement assay. Corallina officinalis (~30 mg) was also included in the assay. Treatment dishes were set up as outlined in section 5.2.1. Twenty-five competent larvae were used for each treatment (5 dishes with 5 larvae). Percent settlement was scored after 12 and 36 h.

5.3 Results

5.3.1 Class Echinoidea, Order Temnopleurida, Family Temnopluridae

Holopneustes inflatus and Holopneustes purpurascens

i) Larval response to histamine

Histamine (10 μM) induced rapid settlement of 100 % of Holopneustes inflatus larvae. The response of competent H. inflatus larvae to a range of histamine concentrations was essentially identical to that of H. purpurascens larvae (Figure 5.2) as indicated by no effect of species in the analysis (Table 5.1). The only difference between species was in their response to 100 nM after 24 h (Figure 5.2B), as suggested by the borderline interaction of time x species x histamine-concentration (Table 5.1, Greenhouse-Geisser adjusted p = 0.072).

118 ii) Larval response to host plants

Although there was no effect of (larval) species in the analysis, there was a significant species x (host) plant interaction (Table 5.2) suggesting that Holopneustes inflatus and Holopneustes purpurascens larvae were responding differently to host plants in the assay. The different effects of the host plants on the response of H. inflatus and H. purpurascens larvae are apparent after 1 h (Figure 5.3A). Amphiroa anceps and 3 species of seagrass; Halophila ovalis, Posidonia australis and Zostera capricornia, all induced rapid settlement in 40–80 % of H. inflatus larvae after 1 h (Figure 5.3A). However, Delisea pulchra, which is not present in seagrass beds, induced less than 20 % settlement of H. inflatus after 1 h. Two differences were apparent in the response of H. purpurascens larvae after 1 h; more H. purpurascens larvae settled in response to D. pulchra than H. inflatus larvae, and less H. purpurascens larvae settled in response to all three seagrasses than H. inflatus larvae (Figure 5.3A). After 24 h, most host plants had induced over 80 % settlement of both larval species (Figure 5.3B). The differences between species observed at 1 h were no longer apparent which is indicated by the time x species x plant interaction ( p = 0.034, Table 5.2).

5.3.2 Class Echinoidea, Order Temnopleuroida, Family Toxopneustidae

Tripneustes gratilla

Histamine did not induce the settlement of Tripneustes gratilla larvae at any concentration tested (100 μM, 10 μM, 1 μM, 100 nM) after 48 h exposure. After 48 h, approximately 30 % of T. gratilla larvae settled in response to Corallina officinalis and none had settled in response to SFSW.

5.3.3 Class Echinoidea, Order Echinoida, Family Echinometridae

Heliocidaris erythrogramma

Histamine induced the settlement of Heliocidaris erythrogramma larvae with a maximum response of approximately 50 % settlement after 48 h exposure to 10 and 100 μM histamine

119 (Figure 5.4). Histamine concentrations of 200, 100, 50 and 10 μM were equally effective at inducing settlement of H. erythrogramma larvae after 48 h (1-factor ANOVA - F5, 54 = 9.709 p < 0.001, Bonferroni’s pairwise comparisons all p > 0.5). However, the larval response to 1 μM histamine was no different to the larval response to SSW (Bonferroni’s pairwise comparison, p = 1.0). Settlement of larvae was not observed in any treatment after 1 h.

5.3.4 Heliocidaris tuberculata

Histamine (10 μM) did not induce settlement of Heliocidaris tuberculata larvae after 7 d exposure (Figure 5.5). Approximately 50 % of H. tuberculata larvae settled in response to Corallina officinalis after 72 h and less than 5 % of larvae settled in FSW.

5.3.5 Class Asteroidea, Order Valvatida, Family Asterinidae

Meridiastra calcar

Histamine did not induce settlement of sea-star larvae, Meridiastra calcar, at any concentration tested (100 μM, 10 μM, 1 μM, 100 nM). M. calcar larvae settled in response to Corallina officinalis after 7 d exposure. There was no larval settlement in FSW.

5.3.6 Meridiastra hybrids - female Meridiastra oriens with male Meridiastra occidens

Histamine did not induce settlement of Meridiastra oriens/occidens hybrid larvae at any concentration tested (100 μM, 10 μM, 1 μM, 100 nM, Figure 5.6). Hybrid larvae settled in response to Corallina officinalis at a faster rate than M. calcar with over 90 % settlement after 48 h. There was no larval settlement in FSW.

120 5.3.7 Class Ophiuroidea, Order Ophiurida, Family Ophiocomidae

Clarkcoma canaliculata

All treatments tested induced the settlement of Clarkcoma canaliculata larvae including FSW (Figure 5.7). Corallina officinalis induced a higher rate of settlement than other treatments. At least half of all larvae had settled in each treatment after 36 h.

121 Echinoida Temnopleuroida Phymosomatoida Arbacoida Cassiduloida Clypeasteroida Spatangoida Diadematoida Echinothurioida Cidaroida

Holothuridea Figure 5.1A Cladogram of relationships among Class Echinoidea derived from the combined analysis of two molecular and two morphological data sets proposed by Smith 1997 (Littlewood & Smith 1995, Smith et al. 1996). Only part of the original cladogram presented in Smith (1997) is reproduced here, i.e., only the Class Echinoidea from a phylogeny of all echinoderm classes.

Crinoidea Asteroidea 98 Ophiuroidea 62 Echinoidea 98 Holothuridea

Figure 5.1B Cladogram of relationships among the echinoderm classes derived from the combined analysis of larval and adult morphological characters and molecular data, proposed by Littlewood et al. (1997). Bootstrap percentages (1000 replicates) are shown, tree length = 446, CI = 0.81, RI = 0.48.

122 100 H. inflatus A H. purpurascens 80

100 H. inflatus H. purpurascens B 80

20 Percent settlement

100 H. inflatus C H. purpurascens 80

0 SSW 10 nM 100 nM 1 μμ M 10 M Histamine concentration Figure 5.2 The settlement response of Holopneustes inflatus larvae (n = 100 larvae per treatment, mean ± SE) to a range of histamine concentrations at A. 1 h, B. 24 h and C. 48 h (black circles). The settlement response of Holopneustes purpurascens larvae (data from Chapter 4, Batch C, n = 100 larvae per treatment, mean ± SE) are overlayed for direct comparison (white circles).

123 A 100 H. inflatus H. purpurascens

0 B 100 Percent settlement 80

e a a ia ia a la n o e r i W i r n n e h i is t S m l lo o p h d s S a p e k i o o t c s l s D Z a i m E o H A P H Treatment

Figure 5.3 The settlement response of Holopneustes inflatus larvae (black bars, n = 30 larvae per treatment, mean ± SE) and Holopneustes purpurascens larvae (grey bars, n = 50 larvae per treatment, mean ± SE) to different host plants after A. 1 h and B. 24 h . Histamine (10 μM) and SSW were positive and negative controls, respectively. Amphiroa = A. anceps, Delisea = D. pulchra, Ecklonia = E. radiata, Posidonia = P. australis, Zostera = Z. capricornia, Halophila = H. ovalis.

124 100 24 h 48 h

40 Percent settlement

0 1 50 10 W 100 S 200 S phiroa m Histamine concentration (μM) A

Figure 5.4 The settlement response of Heliocidaris erythrogramma larvae to a range of histamine concentrations (200, 100, 50, 10 and 1 μM) after 24 h (black bars) and 48 h (grey bars, n = 50 larvae per treatment, mean ± SE). Amphiroa anceps (Amphiroa) and SSW were included as positive and negative controls, respectively.

40 Percent settlement Percent 20

0 Corallina His (10 μ M) FSW Treatment

Figure 5.5 The settlement response of Heliocidaris tuberculata larvae (n = 15 larvae per treatment, mean ± SE) to histamine (his) tested at 10 μM. Corallina = Corallina officinalis.

125 100 48 h 85 h 80

40 Percent settlement Percent

0 Corallina 100 μμμ M 10 M 1 M 100 nM FSW Histamine concentration

Figure 5.6 The settlement response of Meridiastra occidens/oriens hybrid larvae (n = 25 larvae per treatment, mean ± SE) to various concentrations of histamine at 48 h (black bars) and 85 h (grey bars). Corallina = Corallina officinalis

100 12 h 36 h

40 Percent settlement

0 Corallina His (10 μμ M) His (1 M) FSW Treatment Figure 5.7 The settlement response of Clarkcoma canaliculata larvae (n = 25 larvae per treatment, mean ± SE) to histamine (10 μM, 1 μM) at 12 h (black bars) and 36 h (grey bars). Corallina = Corallina officinalis

126

Table 5.1 Repeated measures ANOVA of the effects of species, histamine concentration and exposure time to histamine on the percent settlement of Holopneustes inflatus and Holopneustes purpurascens larvae. Note that Greenhouse-Geisser (G-G) adjusted p values are used for the within subjects test as Greenhouse-Geisser ε = 0.57

Source df MS F p Between Subjects Species 1 0.004 0.662 0.418 Concentration 4 13.456 2152 <0.001 Species x Concentration 4 0.001 0.227 0.923 Error 90 0.006 Within Subjects G-G Exposure Time 2 0.492 158 <0.001 Exposure Time x Species 2 0.010 3.160 0.073 Exposure Time x Concentration 8 0.341 109 <0.001 Exp.Time x Species x Conc. 8 0.007 2.145 0.072 Error 180 0.003

127

Table 5.2 Repeated measures ANOVA of the effects of species, plant type (algae and seagrasses) and exposure time to plants on the percent settlement of Holopneustes inflatus and Holopneustes purpurascens larvae. Source df MS F p Between Subjects Species 1 0.078 1.776 0.186 Plant 5 2.379 54 <0.001 Species x Plant 5 0.130 2.953 0.015 Error 108 0.044 Within Subjects Exposure Time 1 9.323 304 <0.001 Exposure Time x Species 1 0.008 0.258 0.612 Exposure Time x Plant 5 0.298 9.724 <0.001 Exp.Time x Species x Plant 5 0.077 2.519 0.034 Error 108 0.031

128 5.4 Discussion

Histamine is not a general settlement cue for echinoderm larvae, however, histamine may be a general settlement cue for echinoids with direct development (i.e., those with non- feeding larvae). Histamine induced settlement of two other sea urchins with non-feeding larvae, Holopneustes inflatus and Heliocidaris erythrogramma. Interestingly, these urchins are from families (Temnopleuridae and Echinometridae) which diverged between 60–70 million years ago (Smith 1988). The induction of settlement of non-feeding larvae of two phylogenetically disparate echinoids by the same settlement cue provides further support for a current hypothesis in the evolution of development, that developmental pathways are convergent in species that have independently evolved direct development (Wray 2002, Raff et al. 2003, Byrne & Voltzow 2004).

That histamine induces larval settlement in Holopneustes inflatus, a closely related species of Holopneustes purpurascens (Jeffery et al. 2003), is perhaps not that remarkable. H. inflatus and H. purpurascens are placed as sister taxa in a phylogenetic analysis of the temnopleurid echinoids derived from 3 sets of molecular data and morphological characters (Jeffery et al. 2003). Adult H. purpurascens and H. inflatus are morphological similar and are distinguished only by subtle differences in the tests and spines (Miskelly 2002). Taxonomy within the family Temnopleuridae is somewhat problematic as species level classification has often been based on test and spine colouration boundaries (Jeffery et al. 2003). Consequently, species boundaries have proved difficult to define (Jeffery et al. 2003). Indeed, there is some debate over whether H. purpurascens and H. inflatus are truly different species (M. Byrne pers. comm., R. B. Emlet pers. comm.).

Holopneustes inflatus and Holopneustes purpurascens differ primarily in habitat and host- plant use (Miskelly 2002) which was reflected by the initial settlement response of larvae to the different host plants. More H. inflatus larvae settled to seagrasses than H. purpurascens, which is consistent with adult H. inflatus living in seagrass beds, while less H. inflatus settled in response to D. pulchra than H. purpurascens larvae, which is consistent with D. pulchra not being present in seagrass beds. A semi-quantitative analysis

129 of the histamine content of seagrasses Posidonia australis and Halophila ovalis detected low levels of histamine, probably similar to levels measured in brown algae and kelp (~0.5 μg.g-1 dw) (Swanson et al. 2006). While histamine is clearly an inducer of settlement of H. inflatus larvae, it is yet to be determined whether histamine is the settlement cue produced by seagrasses in the natural habitat of H. inflatus.

The induction of settlement of Heliocidaris erythrograma larvae by histamine is particularly interesting in an evolutionary sense and provides support for the hypothesis of convergence in evolution of direct development (Wray 2002, Raff et al. 2003, Byrne & Voltzow 2004). Support for this hypothesis was provided by Raff et al. (2003) when reciprocal crosses between Holopneustes purpurascens and H. erythrogramma gametes produced simple larvae (Raff et al. 2003). A few hybrids completed development and metamorphosed which demonstrated that the development pathways in these species were compatible (Raff et al. 2003). This indicated that convergence in morphological evolution was associated with the shift to direct development and a likely convergence in the underlying genetic regulatory mechanisms (Byrne & Voltzow 2004).

Holopneustes purpurascens and Heliocidaris erythrogramma commonly occur together in sub-tidal rocky reefs habitats although they live on different substrata (H. purpurascens are canopy-dwelling whereas H. erythrogramma are bottom-dwelling). Considering that they share habitats it is perhaps not surprising that they are both induced to settle by histamine which leaches from foliose and marcoalgae within the habitat. They are both induced to settle by the surface-biofilms on coralline algae, however, the chemical cue produced by bacteria in these biofilms is not yet identified. I have proposed in Chapter 3 that the biofilm derived settlement cue for H. purpurascens may be bacterial histamine. Many bacterial strains isolated from biofilms on coralline algae induced settlement of H. erythrogramma larvae (Huggett et al. 2006). Given that histamine induces settlement of H. erythrogramma larvae, it is possible that one of the settlement cues produced by the inductive bacterial strains may be histamine.

130 Histamine did not induce the settlement of feeding larvae of Tripneustes gratilla or Heliocidaris tuberculata, nor did it induce settlement of the non-feeding larvae of the sea star Meridiastra spp.. Settlement may be under complete endogenous control in larvae of the brittle star Clarkcoma canaliculata as larvae settled in all treatments, however, it is equally possible that C. canaliculata larvae settled in response to biofilms forming in dishes. Even though sterile dishes and seawater were used in this assay, bacteria would have been introduced to dishes along with larvae and algae. Compared to echinoids, a higher proportion of asteroids species whose larval development is known produce non- feeding larvae (Emlet et al. 1987) possibly implying a more ancient origin for direct development in asteroids (Smith 1997). Although only a few species were tested in this study, it appears that the settlement response of direct-developing echinoids to histamine differs from that of direct developing asteroids and ophiuroids. This is not surprising considering the ancient divergence of these echinoderm classes about 500 million years ago (Paul & Smith 1984). However, it appears that there is some convergence in the morphogenetic pathways of direct developing echinoids which produce highly modified larvae which provides an interesting avenue for further research.

131 CHAPTER 6

General Discussion

6.1 Histamine is an ecologically relevant settlement cue for Holopneustes purpurascens

6.1.1 Histamine as a natural settlement cue for Holopneustes purpurascens

Patterns of settlement of larvae and recruitment of new individuals into marine habitats have fundamental consequences for population structure and community dynamics in benthic ecosystems (Underwood & Keough 2000). Identifying settlement cues which can determine those patterns for marine invertebrate larvae is thus critical to our understanding of the ecology of benthic systems. In this study, I have isolated and identified a naturally occurring settlement cue for larvae of the echinoid Holopneustes purpurascens. The settlement cue was isolated from the polar extract of Delisea pulchra using cation-exchange chromatography and identified as histamine using spectroscopy and mass spectrometry (Chapter Two) (Swanson et al. 2004). This was a significant achievement as very few naturally occurring settlement cues have been completely characterised (Kato et al. 1975, Yvin et al. 1985, Tsukamoto et al. 1993, Tsukamoto et al. 1994, 1995, 1999) and none have been convincingly related back to settlement patterns in the field.

6.1.2 Settlement and recruitment of Holopneustes purpurascens

Following isolation and identification of the cue, I investigated the role of histamine in the context of the ecology and habitat of Holopneustes purpurascens. If histamine is

132 functioning as a settlement cue for this species then variation in the distribution of new recruits of H. purpurascens in the habitat should relate to variation in the distribution of histamine. In total, only 63 new recruits (≤5 mm test diameter) of H. purpurascens were found on 21 sampling dates over 2 yr, however, new recruits were found in most months. The smallest size class of H. purpurascens (≤2 mm test diameter) were found in all months except February and August (Chapter Three)(Swanson et al. 2006). This distribution of new recruits through time suggests that low levels of recruitment of H. purpurascens occur throughout the year. This is consistent with the reproductive biology of this urchin (Williamson & Steinberg 2002) and findings of Williamson (2001) who also found low numbers H. purpurascens recruits (32 individuals, ≤2 mm test diameter) over a 14 month period. Low but steady recruitment may maintain this population of H. purpurascens, however, strong recruitment events may occur intermittently. High interannual variability in recruitment is common among echinoid populations (Rowley 1989, López et al. 1998).

Approximately sixty percent of Holopneustes purpurascens new recruits were found on coralline turf algae (Amphiroa anceps, Corallina officinalis), thirty percent were found on Delisea pulchra, and the remainder were found on the brown alga Homeostrichus olsenii. No new recruits were found on the other brown algae, Sargassum vestitum and Ecklonia radiata. A similar distribution was reported by Williamson (2001) with approximately half of the recruits found on both coralline turf algae and D. pulchra. Known host plants of H. purpurascens were tested for their ability to induce the settlement of H. purpurascens larvae. D. pulchra and A. anceps induced the highest rate of settlement followed by H. olsenii hence settlement choices of larvae in the laboratory matched recruitment patterns observed in the field (Chapter Three). The histamine content of these algae was quantified and levels were far greater in D. pulchra than in the other algae surveyed (Swanson et al. 2006). Hence D. pulchra appears to be the major source of histamine in the habitat. However, it is likely that lesser quantities of histamine also leach from the more abundant species of brown algae/kelp, particularly from heavily fouled plants. With the exception of coralline algae for which histamine could not be detected, variation in the settlement and recruitment of H. purpurascens was consistent with variation among species’ histamine

133 contents. I have shown that the distribution of new recruits of H. purpurascens can be partially explained by the distribution of histamine in the habitat.

The natural growth rates of new recruits of Holopneustes purpurascens are not known hence it is possible that the ‘new’ recruits (≤5 mm test diameter) found in this study are several months old. The smallest recruits (≤2 mm test diameter) which were found consistently throughout the year were probably only weeks old. Nevertheless, post- settlement mortality or migration may have contributed to the recruitment patterns observed here (Connell 1985, Rowley 1989, Pearce 1997). However, in settlement assays H. purpurascens larvae settled rapidly in response the same algae (Delisea pulchra, Amphiroa anceps, Corallina officinalis and Homeostrichus olsenii) on which new recruits were found in the habitat, which suggests that H. purpurascens settle on the same algae which they recruit to. While the settlement cue in coralline algae has not been identified, the link between histamine production and settlement of H. purpurascens on D. pulchra and H. olsenii is consistent with a role as a settlement cue.

6.1.3 Dissolved histamine as a waterborne settlement cue

If histamine is a natural settlement cue for larvae of Holopneustes purpurascens then inductive levels of histamine must be accessible to larvae in the habitat. Very low concentrations of histamine (~ 5 nM) were detected in seawater collected in situ adjacent to Delisea pulchra (Delisea–SW), with significantly lower levels detected in seawater collected in situ adjacent to Ecklonia radiata, Amphiroa anceps or at the sea surface. This is the first time that a characterised settlement cue that is dissolved in seawater has been quantified in situ. Waterborne cues have been implicated in the settlement of barnacle larvae (Rittschof 1985) and a diverse range of molluscan larvae (Hadfield & Pennington 1990, Tamburri et al. 1992, Gibson & Chia 1994, Krug & Manzi 1999). Delisea–SW induced up to 16 % settlement of older Holopneustes purpurascens larvae (3-4 wk post- fertilisation) but had no settlement effect on competent larvae. Delisea–SW collected in a previous study on this system induced 100 % of competent larvae to settle within 4 h (Williamson et al. 2000), however, the histamine concentration of these samples was not

134 quantified. Taken together, these results suggest that inductive concentrations of histamine are present in the habitat and accessible to larvae. However, to confirm unequivocally that histamine is operating as a waterborne settlement cue for H. purpurascens larvae in the habitat, it needs to be shown that Delisea–SW which induces settlement of competent larvae contains inductive levels of histamine.

To achieve this goal, a more extensive seawater sampling program needs to be carried out as levels of histamine in Delisea pulchra (and consequently in the seawater surrounding D. pulchra) were highly variable through time. To enhance the chances of detecting inductive levels of histamine in seawater, D. pulchra plants should be collected simultaneously and analysed first for histamine content. Only those seawater samples collected near plants which contained high levels of histamine should be quantified. Furthermore, seawater samples should only be collected during calm conditions and preferably at slack tides.

6.1.4 Older larvae become more sensitive to histamine as a settlement cue

The finding that Delisea–SW induced the settlement of older Holopneustes purpurascens larvae while having no effect on newly competent larvae led to the next discovery, namely that H. purpurascens larvae become increasingly sensitive to histamine as they age. An earlier study reported that the brown algae Ecklonia radiata and Sargassum vestitum had no settlement effect on newly competent (6-d-old) H. purpurascens larvae yet these same species induced 20 % of older (18-d-old) larvae to settle (Williamson et al. 2004). Older H. purpurascens larvae appeared to be less discriminatory in habitat selection in accordance with the ‘desperate larva hypothesis’ (Knight-Jones 1953, Pechenik 1999, Toonen & Pawlik 2001b). I have found that older H. purpurascens larvae settle in response to significantly lower concentrations of histamine as they age (Chapter Four). Therefore, the basis of older H. purpurascens larvae becoming less discriminatory in accepting brown algae as hosts lies in their increased sensitivity to the lower concentrations of histamine produced by these algae.

135 The increased sensitivity of older Holopneustes purpurascens larvae to histamine is likely to result in an acceptance of a broader range of host algae in the habitat, which is predicted by the theoretical model of Elkin and Marshall (in press). These findings are a novel example of larvae expanding their range of host plants by becoming more responsive to a single settlement cue. I suggest that older H. purpurascens larvae may use histamine as a general habitat cue, and that with extended time in the plankton they begin to settle in response to lower concentrations of histamine. Lower concentrations of histamine would be present in seawater surrounding brown algae and kelp which are far more abundant in rocky reef habitats than D. pulchra. Older larvae of the opisthobranch Alderia sp. become increasingly sensitive to a settlement cue present in the alga Vaucheria longicaulis, on which they live and feed exclusively (Botello & Krug 2006). As Alderia sp. is a strict specialist the increased sensitivity of older larvae to its settlement cue would not result in an acceptance of wider range of hosts (as per H. purpurascens) but rather acceptance of different patches of V. longicaulis of lesser ‘quality’ i.e., patches which produce lesser quantities of the settlement cue (Botello & Krug 2006). Alternatively, older Alderia sp. may be able to detect patches of V. longicaulis from a greater distance as a result of becoming increasingly sensitive to lower concentrations of the cue.

6.1.5 Histamine as a general settlement cue for echinoids with non-feeding larvae

The possibility that histamine acts as a general settlement cue for echinoderm larvae was tested with a variety of echinoderm larvae from Class Echinoidea, Asteroidea and Ophiuroidea. Histamine is not a general settlement cue for echinoderm larvae, however, it induced the settlement of two other echinoid species with non-feeding larvae, Holopneustes inflatus from the family Temnopleuridae and Heliocidaris erythrogramma from the Family Echinometridae (Chapter Five). Interestingly, the Temnopleuridae separated from the Echinometridae 60-70 millions years ago (Smith 1988) yet there appears to be convergence in the morphogenetic pathways of species in these families that have independently evolved non-feeding larvae (Wray 2002, Raff et al. 2003, Byrne & Voltzow 2004). H. erythogramma and H. purpurascens both occur in shallow subtidal habitats although the

136 adults occupy different substrata (rocky bottom versus algal canopy) hence perhaps it is not surprising that their larvae settle in response to the same settlement cue produced by foliose and macroalgae in their habitat. The coralline turf algae Amphiroa anceps and Corallina officinalis also produce settlement cues for both species (which are unknown) and are a nursery ground for new recruits of H. erythogramma and H. purpurascens (Chapter 3) (Huggett 2006).

The general hypothesis of convergence of developmental regulatory meachanisms among echinoids that have independently evolved non-feeding larvae (Wray 2002, Raff et al. 2003) could be further examined by testing histamine as an inducer of settlement in other direct developers (i.e., those with non-feeding larvae). There are 11 species within the family Temnopluridae that have non-feeding larvae and the majority are endemic to Australia (Jeffery et al. 2003); Holopneustes porissimus which lives enmeshed in the laminae of kelp (Miskelly 2002), and species from the genera Microcyphus and Amblyneustes, most of which live enmeshed in the fronds of algae however some species live in seagrass beds (Miskelly 2002, Jeffery et al. 2003). The 11 instances of direct development in the temnopleurids can be accounted for by a single developmental transition that occurred an estimated 4.4–7.4 million years ago (Jeffery et al. 2003). Considering the evolution of direct development in the temnopleurids and that species inhabit algae/kelp/seagrass (which may contain and release histamine), histamine may also induce larval settlement in these species.

There is good evidence that the switch from feeding to non-feeding development evolved over a very short period of time in eight lineages of echinoid between 60-70 million years ago (Smith & Jeffery 1997). It would be interesting to test larvae with non-feeding development from these other echinoid lineages to determine whether convergence in developmental pathways has occurred in more primitive lineages which have independently evolved direct development (Wray 2002). Phyllacanthus parvispinus and Phyllacanthus imperialis belong to the order Cidaroida which is believed to be the most primitive extant echinoid order (Paul & Smith 1984). The tropical species P. imperialis produces ‘intermediate’ non-feeding echinopluteus larvae which are the least modified non-feeding

137 larvae (Olson et al. 1993). In contrast, the temperate species P. parvispinus produces highly modified non-feeding larvae that lack any pluteal features (Parks et al. 1989). It would be particularly interesting to test histamine against P. parvispinus because they produce highly derived larvae and live in temperate rocky reefs like Holopneustes purpurascens and Heliocidaris erythrogramma. Echinothuroid echinoids are the earliest living branch of the euechinoid lineage (all other extant echinoid orders) and are thus the second oldest lineage after the cidaroids (Smith 1988). The echinothuroid Anthenosoma ijimai produces highly modified non-feeding pluteus larvae which have retained several pluteal features (Amemiya & Emlet 1992). All echinothuroids appear to produce modifed (i.e., non-feeding) larvae (Emlet et al. 1987) hence they would be an interesting group to test against histamine.

6.1.6 Settlement cue for Holopneustes purpurascens from coralline algae

The majority of new recruits of Holopneustes purpurascens were found on coralline algae and these algae were potent inducers of larval settlement (Chapter Three). Scores of bacterial strains isolated from the surface-associated biofilm community on coralline algae induced settlement of larvae of the generalist herbivore Heliocidaris erythrogramma larvae (Huggett et al. 2006) and two out of five of these bacterial strains tested here induced settlement of H. purpurascens larvae. The potent settlement cue (or cues) for H. purpurascens larvae that is produced by turfing coralline algae remains unidentified, however, it appears to be waterborne and biofilm derived (Chapter Three). Bacterial biofilms associated with a range of coralline algae and inanimate surfaces (rocks, shells and detritus) appear to be important inducers of larval settlement for other generalist species in addition to H. erythrogramma (Huggett et al. 2006), such as the urchins Evechinus chloroticus (Lamare & Barker 2001), Paracentrotus lividus (Gosselin & Jangoux 1996) and Strongylocentrotus droebachiensis (Pearce & Scheibling 1990, 1991), as well as abalone (Daume et al. 1999, Daume et al. 2000). Larvae of the filter-feeding polychaete Hydroides elgans also settle in response to a diverse range bacterial strains isolated from biofilms formed on inanimate surfaces (Unabia & Hadfield 1999, Lau et al. 2002). While

138 many bacterial strains from different genera were shown to induce larval settlement of H. erythrogramma and H. elegans, in both cases highly inductive strains belonged to the genus Pseudoalteromonas (Unabia & Hadfield 1999, Lau et al. 2002, Huggett et al. 2006). Different bacterial strains which induce larval settlement of H. erythrogramma and H. elegans may be producing the same chemical cue or different chemical cues.

I have argued the case that the biofilm derived cue from coralline algae which induces settlement of Holopneustes purpurascens larvae may in fact be bacterial histamine; however, research findings are at a preliminary stage. Given that histamine induces settlement of Heliocidaris erythogramma larvae (Chapter Five), that marine bacteria are known to produce histamine (Fujii et al. 1997), and that many bacterial strains induce settlement of H. erythogramma larvae, it is quite possible that some bacterial strains in the surface-associated biofilm community on coralline algae produce histamine as a settlement cue for H. erythrogramma larvae. Similar arguments apply for H. purpurascens larvae. The hypothesis that bacterially derived histamine is a settlement cue from coralline algae for H. purpurascens needs to be further investigated as does the potential release of histamine in situ by coralline algae (Amphiroa anceps, Corallina officinalis).

6.1.7 The effect of flow on efficacy of histamine as a settlement cue

I did not get the opportunity to investigate the effect of flow on the efficacy of histamine as a settlement cue for Holopneustes purpurascens larvae, and this is probably the primary limitation of this study. It is well documented that flow and chemical cues interact to affect the settlement response of invertebrate larvae, producing either similar or different larval behaviours. Oyster larvae (Crassostrea virginica) swim downward in response to chemical cues in both still water (Coon et al. 1990, Tamburri et al. 1992, Zimmer-Faust & Tamburri 1994) and flowing water (Tamburri et al. 1996). Polychaete larvae (Capitella sp.) also behaved similarly in still and flowing water, choosing to settle in organic rich mud over glass beads (Snelgrove et al. 1993). However, surfclam larvae (Spisula solidissima) behaved differently; in flume-flow, S. solidissima larvae consistently chose to settle in sand over mud, however, in still water assays larval settlement in sand and mud was variable

139 (Snelgrove et al. 1998). The cup coral Balanophylla elegans produces ‘crawl away’ larvae which respond strongly to the interactive effects of water flow and substrate (Altieri 2003). Only 11 % of larvae settled in the presence of rock substrate or flowing seawater when either factor was presented alone, however, 90% settled in the presence of rock substrate in flowing water (Altieri 2003). When rock substrate was available, water velocities of less than 25 cm·s-1 triggered a 5-fold increase in settlement rates relative to standing water (Altieri 2003) showing that fast water flow can enhance settlement.

It would be interesting to test the effect of flow and histamine on the settlement of Holopneustes purpurascens larvae in flume-flow. The interaction of hydrodynamics and histamine induction should also be explored in the natural habitat where many factors interact to affect larval behaviour (Thompson et al. 1998, Wright & Boxshall 1999).

6.2 Metamorphosis and chemical signals

6.2.1 Metamorphosis via specific histamine receptors?

The other aspect of the settlement of Holopneustes purpurascens larvae which was investigated in this study was the specificity of the larval response to histamine as a settlement cue. A range of neuroactive compounds and histamine-analogues were tested for their ability to induce settlement of H. purpurascens larvae, of which only dopamine induced rapid settlement. The settlement activity of dopamine suggests that it may be involved in the true metamorphic pathway of H. purpurascens, as has been suggested for other invertebrate larvae (Pires et al. 2000a, Zega et al. 2005). The fact that histamine was the only compound tested which induced rapid settlement in the majority of H. purpurascens larvae implies that a specific histamine receptor is involved in the signal transduction pathway which leads to metamorphosis. Two types of biogenic amine membrane receptors have been characterised in animals; metabotropic receptors, defined as seven trans-membrane G protein-coupled receptors or 7TMGPCRs, and ionotropic receptors, defined as members of the ligand-gated ion channel superfamily (Blenau &

140 Baumann 2001). Metabotropic receptors trigger slower cellular responses involving second messenger pathways whereas ionotropic receptors mediate rapid responses through the opening of the channel pore which causes excitation or inhibition of the target cells. A sea urchin homolog of the metabotropic histamine H1 receptors of vertebrates (suH1R) has recently been identified on the cell surface of eggs of Strongylocentrotus purpuratus (Leguia & Wessel 2006).

2+ The suH1R is an integral part of the pathway leading to Ca release in fertilisation in Strongylocentrotus purpuratus (Leguia & Wessel 2006), Ca2+ release being the hallmark of fertilisation in all species examined (Stricker 1999). This is the first report of a metabotropic histamine receptor in invertebrates (Roeder 2003). Previously, only ionotropic histamine receptors have been identified in invertebrates (McClintock & Ache 1989, Gisselmann et al. 2001). Interestingly, 10 μM histamine was sufficient to elicit a full calcium release response during fertilisation of S. purpuratus eggs (Leguia & Wessel 2006), the same histamine concentration at which maximal induction of settlement occurs in Holopneustes purpurascens larvae. This suggests that the suH1R which is expressed on egg cells of S. purpuratus may be the same receptor acting in the morphogenetic pathway of H. purpurascens larvae. The fact that H. purpurascens larvae require 3 h of continuous exposure to 10 μM histamine to induce ireeversible metamorphosis also suggests that a metabotropic histamine receptor is operating in the morphogenetic pathway of H. purpurascens.

The possible involvement of the suH1R in the morphogenetic pathway of Holopneustes purpurascens (and in other species that are induced by histamine) needs to be investigated.

If suH1R is expressed by these larvae, does the level of expression relate to the attainment of competence? Are more suH1R expressed in older larvae thereby increasing their sensitivity to histamine as an inducer? This research is currently underway in collaboration with Dr Gary M.Wessel of the Providence Institute of Molecular Oogenesis at Brown University, Rhode Island, U.S.A.

141 6.2.2 Chemical signals and biological responses

Haldane (1954) was the first to propose that an organisms’ response to a waterborne chemical signal was a receptor-mediated process. Carr (1988) argued that the types of molecules that serve as important external chemical signals in the aquatic environment are often potent neuroactive agents at the cellular level, and that their neuroactive properties were initiated by specific receptors. Haldane (1954) also postulated that internal receptors for chemical signals may trace their origins to external chemoreceptors of primitive aquatic organisms. Such a notion is supported by the fact that the embryological origins of the nervous systems’ of higher organisms are traced to the ectoderm; the outer cell layer in which outer membrane surfaces are in direct contact with the external environment (Carr 1988). This would help to explain why many chemical signals that are involved in neurotransmission or neuromodulation at the cellular level (e.g., nucleotides, amino acids and peptides) also function as external chemical signals in the aquatic environment (Carr 1988).

Chemical signals mediate many of the biological and ecological interactions between organisms in the aquatic environment including; food detection, predator avoidance, mate selection, social status and habitat selection (Hadfield & Paul 2001, Steinberg et al. 2001, Trapido-Rosenthal 2001, Riffell et al. 2002, Bergman & Moore 2005). The chemicals which act as feeding stimulants are generally common metabolites of low molecular weight; particularly amino acids, quarternary ammonium compounds, nucleosides, nucleotides and organic acids (Mearns et al. 1987, Carr 1988, Mearns 1989, Kasumyan et al. 1998).

In contrast, the exact chemical identities of those substances which mediate other aspects of animal behaviour remain elusive in most cases. However, a common theme is emerging from almost three decades of research: many behaviours in invertebrates are initiated, maintained, altered or terminated by the action of monoamine or peptide modulators on neural networks (Hardie 1989, McClintock & Ache 1989, Rittschof et al. 1989, Hashemzadeh-Gargari & Freschi 1992, Zimmer-Faust & Tamburri 1994, Huber et al. 1997,

142 Beltz 1999, Stuart 1999, Blenau & Baumann 2001, Ono & Yoshikawa 2004, Pierobon et al. 2004, Swanson et al. 2004, Nishida 2005). While much less is known about the role of monoamines in invertebrate larvae, recent findings would suggest that monoamines are major regulators of larval invertebrate behaviour (Pires & Woollacott 1997, Shimizu et al. 2000, Leise et al. 2001, Python & Stocker 2002, Swanson et al. 2004, Zega et al. 2005). It appears that the monoamine histamine is an important signal molecule in the external aquatic environment which can trigger organismal responses, such as the metamorphosis of invertebrate larvae (Swanson et al. 2004), adding to a growing body of literature which suggests that histamine is a ubiquitous signal molecule in the internal aqueous environment, particularly in the nervous and vascular systems of a diverse range of species (Huggins & Woodruff 1968, Reite 1972, Weinreich 1979, Elias & Evans 1983, Hashemzadeh-Gargari & Freschi 1992, Velasquez et al. 1996, Gisselmann et al. 2001, Stuart et al. 2002, Leguia & Wessel 2006).

143

Literature Cited

Aguirre-Lipperheide M, Evans LV (1993) A sterilization protocol for the dictyotales (phaeophyceae). Journal of Phycology 29:243-251 Altieri AH (2003) Settlement cues in the locally dispersing temperate cup coral Balanophyllia elegans. Biological Bulletin 204:241-245 Amemiya S, Emlet RB (1992) The development and larval form of an echinothurioid echinoid Asthenosoma ijimai revisited. Biological Bulletin 182:15-30 Arkett SA, Chia F-S, Goldberg JI, Koss R (1989) Identified settlement receptor cells in a nudibranch veliger respond to specific cue. Biological Bulletin 176:155-160 Barancin CE, Smoot JC, Findlay RH, Actis LA (1998) Plasmid-mediated histamine biosynthesis in the bacterial fish pathogen Vibrio anguillarum. Plasmid 39:235-244 Barlow LA (1990) Electrophysiological and behavioral responses of larvae of the red abalone (Haliotis rufescens) to settlement-inducing substances. Bulletin of Marine Science 46:537-554 Barrow KD, Karsten U, King RJ (1993) Isethionic acid from the marine red alga Ceramium flaccidum. Phytochemistry 35:1429-1430 Baxter G, Morse DE (1987) G-protein and diacylglycerol regulate metamorphosis of planktonic molluscan larvae. Proceedings of the National Academy of Science 84:1867-1870 Bayer TA, McClintock TS, Grunert U, Ache BW (1989) Histamine-induced modulation of olfactory receptor neurones in two species of lobster, Panulirus argus and Homarus americanus. Journal of Experimental Biology 145:133-146 Beltz BS (1999) Distribution and functional anatomy of amine-containing neurons in decapod crustaceans. Microscopy Research and Technique 44:105-120 Bergman DA, Moore PA (2005) The role of chemical signals in the social behaviour of crayfish. Chemical Senses 30 (suppl 1) Berntsson KM, Jonsson PR, Larsson AI, Holdt S (2004) Rejection of unsuitable substrata as a potential driver of aggregated settlement in the barnacle Balanus improvisus. Marine Ecology Progress Series 275:199-210

144

Berntsson KM, Jonsson PR, Lejhall M, Gatenholm P (2000) Analysis of behavioural rejection of micro-textured surfaces and implications for recruitment by the barnacle Balanus improvisus. Journal of Experimental Marine Biology and Ecology 251:59- 83 Biggers WJ, Laufer H (1999) Settlement and metamorphosis of Capitella larvae induced by juvenile hormone-active compounds is mediated by protein kinase C and ion channels. Biological Bulletin 196:187-198 Black KP, Moran PJ (1991) Influence of hydrodynamics on the passive dispersal and initial recruitment of larvae of Acanthaster planci (Echinodermata: Asteroidea) on the Great Barrier Reef. Marine Ecology Progress Series 69:55-65 Blenau W, Baumann A (2001) Molecular and pharmacological properties of insect biogenic amine receptors: lessons from Drosophila melangoster and Apis mellifera. Archives of Insect Biochemistry and Physiology 48:13-38 Boettcher AA, Targett NM (1996) Induction of metamorphosis in queen conch, Strombus gigas Linnaeus, larvae by cues associated with red algae from their nursery grounds. Journal of Experimental Marine Biology and Ecology 196:29-52 Bonar DB, Coon SL, Walch M, Wener RM, Fitt W (1990) Control of oyster settlement and metamorphosis by endogenous and exogenous chemical cues. Bulletin of Marine Science 46:484-498 Botello G, Krug PJ (2006) Desperate larvae revisited: age, energy and experience affect sensitivity to settlement cues in larvae in the gastropod Alderia sp. Marine Ecology Progress Series 312:149-159 Brawley SH (1992) Mesoherbivores. In: John DM, Hawkins SJ, Price JA (eds) Plant- animal interactions. Clarendon Press, Oxford Browne KA, Zimmer RK (2001) Controlled field release of a waterborne chemical signal stimulates planktonic larvae to settle. Biological Bulletin 200:87-91 Bucknall M, Yung KYC, Duncan MW (2002) Practical quantitative biomedical applications of MALDI-TOF mass spectrometry. American Society for Mass Spectrometry 13:1015-1027 Burke RD (1983) Neural control of metamorphosis in Dendraster excentricus. Biological Bulletin 164:176-188 145

Burke RD (1984) Pheromonal control of metamorphosis in the Pacific sand dollar, Dendraster excentricus. Science 225:440-441 Butman CA (1986) Larval sediment of soft-sediment invertebrates: some predictions based on an analysis of near-bottom velocity profiles. In: Nihoul J (ed) Marine interfaces ecohydrodynamics. Elsevier Butman CA (1987) Larval settlement of soft-sediment invertebrates: the spatial scales of pattern explained by active habitat selection and the emerging role of hydrodynamical processes. Oceanography and Marine Biology: an Annual Review 25:113-165 Byrne M, Emlet RB, Cerra A (2001) Ciliated band structure in planktotrophic and lecithotrophic larvae of heliocidaris species (Echinodermata: Echinoidea): a demonstration of conservation and change. Acta Zoologica 82:189-199 Byrne M, Voltzow J (2004) Morphological evolution in sea urchin development: hybrids provide insights into the pace of evolution. BioEssays 26:343-347 Cameron RA, Schoeter SC (1980) Sea urchin recruitment: effect of substrate selection on juvenile distribution. Marine Ecology Progress Series 2:243-247 Carr WES (1988) The molecular nature of chemical stimuli in the aquatic environment. In: Atema J, Fay RR, Popper AN, Tavolga WN (eds) Sensory Biology of Aquatic Animals. Springer-Verlag, New York, p 3 Chee F, Byrne M (1999) Serotonin-like immunoreactivity in the brachiolaria larvae of Patiriella regularis. Invertebrate Reproduction and Development 36:111-115 Chia F-S, Buckland-Nicks J, Young CM (1984) Locomotion of marine invertebrate larvae: a review. Canadian Journal of Zoology 62:1205-1222 Claiborne BJ, Selverston AI (1984) Histamine as a neurotransmitter in the stomatogastric nervous system of the spiny lobster. The Journal of Neuroscience 4:708-721 Clare AS, Matsumura K (2000) Nature and perception of barnacle settlement pheromones. Biofouling 15:57-71 Clare AS, Thomas RF, Rittschof D (1995) Evidence for the involvement of cyclic AMP in the pheromonal modulation of barnacle settlement. Journal of Experimental Biology 198:655-664

146

Clare AS, Yamazaki (2000) Inactivity of glycyl-glycyl-arginine and two putative (QSAR) peptide analogues of barnacle waterborne settlement pheromone. Journal of the Marine Biological Association of U.K. 80: 945-946 Connell JH (1985) The consequences of variation in initial settlement vs. post-settlement mortality in rocky intertidal communities. Journal of Experimental Marine Biology and Ecology 93:11-45 Coon SL, Bonar DB, Weiner RM (1985) Induction of settlement and metamorphosis of the Pacific oyster, Crassostrea gigas (Thunberg) by L-DOPA and catecholamines. Journal of Experimental Marine Biology and Ecology 94:211-221 Coon SL, Bonar SL (1987) Pharmacological evidence that alpha-1 adrenoreceptors mediate metamorphosis of the Pacific oyster Crassostrea gigas. Neuroscience 23:1169-1174 Coon SL, Fitt WK, Bonar DB (1990) Competence and delay of metamorphosis in the Pacific oyster Crassostrea gigas. Marine Biology 106:379-387 Crisp DJ (1967) Chemical factors inducing settlement in Crassostrea virginica (Gmelin). Journal of Animal Ecology 36:329 Crisp DJ (1974) Factors influencing the settlement of marine invertebrate larvae. In: Grant PT, Mackie AM (eds) Chemoreception in marine organisms. Academic Press, London, p 177-265 Dahllöf I, Baillie H, Kjelleberg SK (2000) rpoB-based microbial community analysis avoids limitations inherent in 16S rRNA gene intraspecies heterogeneity. Applied and Environmental Microbiology 66: 3376-3380 Daume S, Brand-Gardner S, Woelkerling W-J (1999) Settlement of abalone larvae (Haliotis laevigata Donovan) in response to non-geniculate coralline red algae (Corallinales, Rhodophyta). Journal of Experimental Marine Biology and Ecology 234:125-143 Daume S, Krinich A, Farrell S, Gervis M (2000) Settlement, early growth and survival of Haliotis rubra in response to different algal species. Journal of Applied Phycology 12:479-488 Davis AR, Butler AJ (1989) Direct observations of larval dispersal in the colonial ascidian Podoclavella moluccensis Sluiter - evidence for closed populations. Journal of Experimental Marine Biology and Ecology 127:189-204 147

Davis AR, Butler AJ, Van Altena I (1991) Settlement behavior of ascidian larvae preliminary evidence for inhibition by sponge allelochemicals. Marine Ecology Progress Series 72:117-123 Degnan BM, Morse DE (1995) Developmental and morphogenetic gene regulation in Haliotis rufescens larvae at metamorphosis. American Zoologist 35:391-398 Ebert TA (2001) Growth and survival of post-settlement sea urchins. In: Lawrence JM (ed) Edible Sea Urchins: Biology and Ecology. Elsevier Science, p 79-102 Eckman JE (1983) Hydrodynamic processes affecting benthic recruitment. Limnology and Oceanography 28:241-257 Eckman JE (1987) The role of hydrodynamics in recruitment, growth, and survival of Agropecten irradius (L.) and Anomia simplex (D'Orbigny) within eelgrass meadows. Journal of Experimental Marine Biology and Ecology 106:165-191 Eckman JE (1996) Closing the larval loop: Linking larval ecology to population dynamics of marine benthic invertebrates. Journal of Experimental Marine Biology and Ecology 200:207-237 Eckman JE, Savidge WB, Gross TF (1990) Relationship between duration of cyprid attachment and drag forces associated with detachment of Balanus amphitrite cyprids. Marine Biology 107:111-118 Elias MS, Evans PD (1983) Histamine in the insect nervous system: distribution, synthesis and metabolism. Journal of Neurochemistry 41:562-568 Elkin C, Marshall DJ (in press) Desperate larvae: the influence of deferred costs and habitat requirements on habitat selection. Marine Ecology Progress Series Emlet RB, McEdward LR, Strathmann RR (1987) Echinoderm larval ecology viewed from the egg. Echinoderm Studies 2:55-136 Emlet RB, Young CM, George SB (2002) Phylum Echidermata: Echinoidea. In: Young CM (ed) Atlas of Marine Invertebrate Larvae. Academic Press, p 531-551 Finelli CM, Wethey DS (2003) Behavior of oyster (Crassostrea virginica) larvae in flume boundary layer flows. Marine Biology 143:703-711 Fleck J, Fitt WK (1999) Degrading mangrove leaves of Rhizophora mangle Linne a natural cue for settlement and metamorphosis of the upside down jellyfish Cassiopea

148

xamachana Bigelow. Journal of Experimental Marine Biology and Ecology 234:83- 94 Forward RB (1988) Diel vertical migration: zooplankton photobiology and behaviour. Oceanography and Marine Biology-Annual Review 26:361-393 Fujii T, Hiraishi A, Kobayashi T, Yoguchi R, Okuzumi M (1997) Identification of the psychrophilic histamine-producing marine bacteria previously referred to as the N- group bacteria. Fisheries Science 63:807-810 Gaines S, Brown S, Roughgarden J (1985) Spatial variation in larval concentrations as a cause of spatial variation in settlement for the barnalcle, Balanus glandula. Oecologia 67:267-272 Gibson G (1995) Why be choosy? Temporal changes in larval sensitivity to several naturally-occurring metamorphic inducers in the opistobranch Haminaea callidegenita. Journal of Experimental Marine Biology and Ecology 194:9-24 Gibson GD, Chia F (1994) A metamorphic inducer in the opistobranch Haminaea callidegenita: partial purification and biological activity. Biological Bulletin 187:133-142 Gisselmann G, Pusch H, Hovemann BT, Hatt H (2001) Two cDNAs coding for histamine- gated ion channels in D. melangoster. Nature Neuroscience 5:11-12 Gosselin P, Jangoux M (1996) Induction of metamorphosis in Paracentrotus lividus larvae (Echinodermata, Echinoidea). Oceanologica Acta 19:293-296 Gribben PE, Marshall DJ, Steinberg PD (2006) Less inhibited with age? Larval age modifies responses to natural settlement inhibitors. Biofouling 22:101-106 Hadfield MG, Koehl MAR (2004) Rapid behavioral responses of an invertebrate larvae to dissolved settlement cue. Biological Bulletin 207:28-43 Hadfield MG, Meleshkevitch EA, Boudka DY (2000) The apical sensory organ of a gastropod veliger is a receptor for settlement cues. Biological Bulletin 198:67-76 Hadfield MG, Paul VJ (2001) Natural chemical cues for settlement and metamorphosis of marine invertebrate larvae. In: McClintock JB, Baker BJ (eds) Marine Chemical Ecology. CRC Press, Florida, p 431-461

149

Hadfield MG, Pennington JT (1990) Nature of the metamorphic signal and its internal transduction in larvae of the nudibranch Phestilla sibogae. Bulletin of Marine Science 46:455-464 Hadfield MG, Scheuer D (1985) Evidence for a soluble metamorphic inducer in Phestilla: ecological, chemical and biological data. Bulletin of Marine Science 37:556-566 Haldane JBS (1954) La signalisation animale. Année Biologique 58:89 Hannan CA (1984) Planktonic larvae may act like passive particles in turbulent near- bottom flows. Limnology and Oceanography 29:1108-1116 Hardie RC (1989) A histamine-activated chloride channel involved in neurotransmission at a photoreceptor synapse. Nature 339:704-706 Harvey M, Bourget E (1997) Recruitment of marine invertebrates onto arborescent epibenthic structures: active and passive processes acting at different spatial scales. Marine Ecology Progress Series 153:203-215 Hashemzadeh-Gargari H, Freschi JE (1992) Histamine activates chloride conductance in motor neurons of the lobster cardiac ganglion. Journal of Neurophysiology 68:9-15 Huber R, Smith K, Delago A, Isaksson K, Kravitz EA (1997) Serotonin and aggressive motivation in crustaceans: altering the decision to retreat. Proceedings of the National Academy of Science USA 94:5939-5942 Huggett MJ (2006) Settlement of generalist marine invertebrate herbivores in response to bacterial biofilms and other cues. Ph.D., University of New South Wales Huggett MJ, de Nys R, Williamson JE, Heasman M, Steinberg PD (2005) Settlement of larval blacklip abalone, Haliotis rubra, in response to green and red macroalgae. Marine Biology 147:1155-1163 Huggett MJ, Williamson JE, de Nys R, Kjelleberg S, Steinberg PD (2006) Larval settlement of the common Australian sea urchin Heliocidaris erythrogramma in response to bacteria from the surface of coralline algae. Oecologia 149: 604-619 Huggins AK, Woodruff GN (1968) Histamine metabolism in invertebrates. Comparative Biochemistry and Physiology 26:1107-1111 Incze LS, Wahle RA, Palma AT (2000) Advection and settlement rates in a benthic invertebrate: recruitment to first benthic stage in Homarus americanus. ICES Journal of Marine Science 57:430-437 150

Jackson DJ, Ellemor N, Degnan BM (2005) Correlating gene expression with larval competence, and the effect of age and parentage on metamorphosis in the tropical abalone Haliotis asinina. Marine Biology 147:681-697 Jaeckle WB, Manahan DT (1989) Feeding by a "non-feeding" larva: uptake of dissolved amino acids from seawater by lecithotrophic larvae of the gastropod Haliotis rufescens. Marine Biology 103:87-94 Jeffery CH, Emlet RB, Littlewood DTJ (2003) Phylogeny and evolution of developmental mode in temnopleurid echinoids. Molecular Phylogenetics and Evolution 28:99-118 Jennings JG, Steinberg PD (1997) Phlorotannins versus other factors affecting epiphyte abundance on the kelp Ecklonia radiata. Oecologia 109:461-473 Johnson CR, Sutton DC (1994) Bacteria on the surface of crustose coralline algae induce metamorphosis of the crown-of-thorns starfish Acanthaster planci. Marine Biology 120:305-310 Karsten U, Barrow KD, King RJ (1993) Floridoside, l-isofloridoside, and D-isofloridoside in the red alga Porphyra columbina. Plant Physiology and Biochemistry 103:485- 491 Kasumyan AO, Ryg M, Døving KB (1998) Effect of amino acids on the swimming activity of newly hatched turbot larvae (Scophthalmus maximus). Marine Biology 131:189- 194 Kato T, Kumanireng AS, Ichinose I, Kitahara Y, kakinuma Y, Nishihira M, Kato M (1975) Active components of Sargassum tortile effecting the settlement of swimming larvae of Coryne uchidai. Experientia 31:433-434 Kato T, Kumanireng AS, Ichinose I, Kitahara Y, Kakinuma Y, Kato Y (1975b) Structure and synthesis of active component from a marine alga, Sargassum tortile, which induces the settling of swimming larvae of Coryne uchidae. Chemistry Letters : 335-338 Katz PS (1995) Intrinsic and extrinsic neuromodulation of motor circuits. Current Opinion in Neurobiology 5:799-808 Keough MJ, Downes BJ (1982) Recruitment of marine invertebrates: the role of active larval choices and early mortality. Oecologia 54:348-352

151

Keough MJ, Raimondi PT (1995) Responses of settling invertebrate larvae to bioorganic films: Effects of different types of films. Jounal of Experimental Marine Biology and Ecology 185:235-253 King CK, Hoegh-Guldberg O, Byrne M (1994) Reproductive cycle of Centrostephanus rodgersii (Echinoidea) with recommendations for the establishment of a sea urchin fishery in New South Wales. Marine Biology 120:95-106 Knight J, Rowley AF, Yamazaki M, Clare AS (2000) Eicosanoids are modulators of larval settlement in the barnacle, Balanus amphitrite. Journal of the Marine Biological Association of U.K. 80:113-117 Knight-Jones EW (1953) Decreased discrimination during settling after a prolonged planktonic life in larvae of Spirorbis borealis (Serpulidae). Journal of Marine Biological Association of U.K 32:337-345 Koehl MAR, Hadfield MG (2004) Soluble settlement cues in slowly moving water within coral reefs induces larval adhesion to surfaces. Journal of Marine Systems 49:75-88 Kon-ya K, Miki W, Endo M (1995) L-Tryptophan and related compounds induce larval settlement of the barnacle Balanus amphitrite Darwin. Fisheries Science 61:800- 803 Krug PJ, Manzi AE (1999) Waterborne and surface-associated carbohydrates as settlement cues for larvae of the specialist marine herbivore Alderia modesta. Biological Bulletin 197:94-103 Krug PJ, Zimmer RK (2000) Developmental dimorphism and expression of chemosensory- mediated behavior: habitat selection by a specialist marine herbivore. The Journal of Experimental Biology 203:1741-1754 Lamare MD, Barker MF (2001) Settlement and recruitment of the New Zealand sea urchin Evechinus chloroticus. Marine Ecology Progress Series 218:153-166 Lambert WJ, Todd CD, Hardege JD (1997) Partial characterisation and biological activity of a metamorphic inducer of the dorid nudibranch Adalaria proxima (Gastropoda: Nudibranchia). Invertebrate Biology 116:71-81 Lau SCK, Mak KKW, Chen F, Qian P-Y (2002) Bioactivity of bacterial strains isolated from marine biofilms in Hong Kong waters for the induction of larval settlement in

152

the marine polychaete Hydroides elegans. Marine Ecology Progress Series 226:301- 310 Leguia M, Wessel GM (2006) The histamine H1 receptor activates the nitric oxide pathway at fertilization. Molecular Reproduction and Development in press Leise EM, Hadfield MG (2000) An inducer of molluscan metamorphosis transforms activity patterns in a larval nervous system. Biological Bulletin 199:241-250 Leise EM, Thavaradhara K, Durham NR, Turner BE (2001) Serotonin and nitric oxide regulate metamorphosis in the marine snail Llyanassa obsoleta. American Zoologist 41:258-267 Levin LA (1984) LIfe history and dispersal patterns in a dense infaunal assemblage: community structure and response to disturbance. Ecology 65:1185-1200 Littlewood DT, Smith AB, Clough KA, Emson RH (1997) The interrelationships of the echinoderm classes: morphological and molecular evidence. Biological Journal of the Linnean Society 61:409-438 Littlewood DTJ, Smith AB (1995) A combined morphological and molecular phylogeny for echinoids. Philosophical Transactions of the Royal Society, London 347 B:213- 234 Loeb MJ, Walker G (1977) Origin, composition, and function of the secretions from pyriform organs and internal sacs of four settling cheilo-ctenostome byrozoan larvae. Marine Biology 42:37-46 López S, Turon X, Montero E, Palacín C, Duarte CM, Tarjuelo I (1998) Larval abundance, recruitment and early mortality in Paracentrotus lividus (Echinoidea). Interannual variability and plankton-benthos coupling. Marine Ecology Progress Series 172:239-251 Maida M, Coll JC, Sammarco PW (1994) Shedding new light on scleractinian coral recruitment. Journal of Experimental Marine Biology and Ecology 180:189-202 Manahan DT, Jaeckle WB, Nourizadeh SD (1989) Ontogenetic changes in the rates of amino acid transport from seawater by marine invertebrate larvae (Echinodermata, Echiura, Mollusca). Biological Bulletin 176:161-168

153

Marshall DJ, Keough MJ (2003) Variation in the dispersal potential of non-feeding invertebrate larvae: the desperate larva hypothesis and larval size. Marine Ecology Progress Series 255:145-153 Matsumura K, Nagano M, Fusetani N (1998b) Purification of a larval settlement-inducing protein complex (SIPC) of the barnacle, Balanus amphitrite. The Journal of Experimental Zoology 281:12-20 Matsumura K, Nagano M, Kato-Yoshinaga Y, Yamazaki M, Clare A, Fusetani N (1998c) Immunological studies on the settlement-inducing protein complex (SIPC) of the barnacle Balanus amphitrite and its possible involvement in larva-larva interactions. Proceedings of the Royal Society of London (B) 265:1825-1830 McClintock TS, Ache BW (1989) Histamine directly gates a chloride channel in lobster olfactory receptor neurons. Proceedings of National Academy of Sciences 86:8137- 8141 Meadows PS, Campbell JI (1972) Habitat selection by aquatic invertebates. Advances in Marine Biology 10:271-382 Mearns KJ (1989) Behavioural responses of salmonid fry to low amino acid concentrations. Journal of Fish Biology 34:223-232 Mearns kJ, Ellingsen OF, Doving KB, Helmer S (1987) Feeding behaviour in adult rainbow trout and Atlantic salmon parr, elicited by chemical fractions and mixtures of compounds identified in shrimp extract. Aquaculture 64:47-63 Menge BA (2000) Recruitment vs. postrecruitment processes as determinants of barnacle population abundance. Ecological Monographs 70:265-288 Mercier A, Battaglene SC, Hamel J-F (2000) Settlement preferences and early migration of the tropical sea cucumber Holothuria scabra. Journal of Experimental Marine Biology and Ecology 249:89-110 Miron G, Walters LJ, Tremblay R, Bourget E (2000) Physiological condition and barnacle larval behavior: a preliminary look at the relationship between TAG/DNA ratio and larval substratum exploration in Balanus amphitrite. Marine Ecology Progress Series 198:303-310 Miskelly A (2002) Sea urchins of Australia and the Indo-Pacific, Vol. Capricornica, Sydney 154

Morgan SG (1995) Life and death in the plankton: larval mortality and adaptation. In: McEdward LR (ed) Ecology of Marine Invertebrate Larvae. CRC Press, New York Morse ANC, Froyd CA, Morse DE (1984) Molecules from cyanobacteria and red algae that induce larval settlement and metamorphosis in the mollusc Haliotis rufescens. Marine Biology 81:293-298 Morse ANC, Morse DE (1984) Recruitment and metamorphosis of Haliotis larvae induced by molecules uniquely available at the surfaces of crustose red algae. Journal of Experimental Marine Biology and Ecolgy 75:191-215 Morse ANC, Morse DE (1996) Flypapers for coral and other planktonic larvae. Bioscience 46 (4):254-262 Morse DE (1985) Neurotransmitter-mimetic inducers of larval settlement and metamorphosis. Bulletin of Marine Science 37:697-706 Morse DE (1990) Recent progress in larval settlement and metamorphosis: closing the gaps between molecular biology and ecology. Bulletin of Marine Science 46:465-483 Morse DE, Hooker N, Duncan H, Jensen L (1979) Gamma-aminobutyric acid a neurotransmitter, inducer planktonic abalone larvae to settle and begin metamorphosis. Science 204:407-410 Morse DE, Morse ANC (1988) Chemical signals and molecular mechanisms: learning from larvae. Oceanus 31:37-43 Morse DE, Morse ANC (1991) Enzymatic characterization of the morphogen recognised by Agaricia humilis (Scleractinian). Biological Bulletin 181:104-122 Mullineaux LS, Butman CA (1991) Initial contact, exploration and attachment of barnacle (Balnanus amphitrite) cyprids settling in flow. Marine Biology 110: 93-103 Naidenko TK (1996) Induction of metamorphosis of two species of sea urchin from Sea of Japan. Marine Biology 126:685-692 Negri AP, Webster NS, Hill RT, Heyward AJ (2001) Metamorphosis of broadcast spawning corals in response to bacteria isolated from crustose algae. Marine Ecology Progress Series 223:121-131 Nishida R (2005) Chemosensory basis of host recognition in butterflies: multi-component system of oviposition stimulants and deterrents. Chemical Senses 30 (suppl 1):i293- i294 155

Nishihira M (1968) Brief experiments on the effect of algal extracts in promoting the settlement of the larvae of Coryne uchidai Stechow (Hydrozoa). Bulletin of the Marine Biological Station Asamushi 23:91 Norekyan TP, Satterlie R (1993) Neuronal analysis of hunting behaviour of the pteropod mollusc Clione limacina. Neuroscience of Behavioural Physiology 23:11-23 Olivier F, Tremblay R, Bourget E, Rittschof D (2000) Barnacle settlement: field experiments on the influence of larval supply, tidal level, biofilm quality and age on Balanus amphitrite cyprids. Marine Ecology Progress Series 199:185-204 Olson RR, Cameron JL, Young CM (1993) Larval development (with observations on spawning) of the pencil urchin Phyllacanthus imperialis: a new intermediate larval form? Biological Bulletin 185:77-85 Ono H, Yoshikawa H (2004) Identification of amine receptors from a swallowtail butterfly, Papilo xuthus L.: cloning and mRNA localization in foreleg chemosensory organ for recognition of host plants. Insect Biochemistry and Molecular Biology 34:1247- 1256 Parks AL, Bisgrove BW, Wray GA, Raff RA (1989) Direct development in the sea urchin Phyllacanthus parvispinus (Cidaroidea): Phylogenetic history and functional modification. Biological Bulletin 177 Pasternak Z, Garm A, Høeg JT (2005) The morphology of the chemosensory aesthetasc- like setae used during settlement of cypris larvae in the parasitic barnacle Sacculina carcini (Cirripedia: Rhizocephala). Marine Biology 146:1005-1013 Paul CRC, Smith AB (1984) The early radiation and phylogeny of echinoderms. Biological Reviews 59:443-481 Pawlik JR (1990) Natural and artificial induction of metamorphosis of Phragmatopoma lapidosa californica (Polychaeta: Sabellaridiae) with a critical look at the effects of bioactive compounds on marine invertebrate larvae. Bulletin of Marine Science 46:512-536 Pawlik JR (1992) Chemical ecology of the settlement of benthic marine invertebrates. Oceanography and Marine Biology: an Annual Review 30:273-335

156

Pearce CM (1997) Induction of settlement and metamorphosis in echinoderms. In: Fingerman M, Nagabhushanam R, Thompson M-F (eds) Recent Advances in Marine Biotechnology, Vol 1. Science Publishers Inc., p 283-341 Pearce CM, Scheibling RE (1990) Induction of metamorphosis of larvae of the green sea urchin, Strongylocentrotus droebachiensis, by coralline red algae. Biological Bulletin 179:304-311 Pearce CM, Scheibling RE (1991) Effect of macroalgae, microbial films, and conspecifics on the induction of metamorphosis of the green sea urchin, Strongylocentrotus droebachiensis (Muller). Journal of Experimental Marine Biology and Ecology 147:147-162 Pechenik JA (1980) Growth and energy balance during the larval lives of three prosobranch gastropods. Journal of Experimental Marine Biology and Ecology 44:1-28 Pechenik JA (1990) Delayed metamorphosis by larvae of benthic marine invertebrates: Does it occur? Is there a price to pay? Ophelia 32:63-94 Pechenik JA (1999) On the advantages and disadvantages of larval stages in benthic invertebrate life cycles. Marine Ecology Progress Series 177:269-297 Pechenik JA, Cerulli TR (1991) Influence of delayed metamorphosis on survival, growth and reproduction of the marine polychaete, Capitella sp. Journal of Experimental Marine Biology and Ecology 151:17-27 Pechenik JA, Li W, Cochrane DE (2002) Timing is everything: the effects of putative dopamine antagonists on metamorphosis vary with larval age and experimental duration in the prosobranch gastropod Crepidula fornicata. Biological Bulletin 202:137-147 Pennati R, Groppelli S, Sotgia C, Candiani S, Pestarino M, De Bernardi F (2001) Serotonin localisation in Phallusia mammillata larvae and effects of 5-HT antagonist during larval development. Developmental Growth and Differentiation 43:647-656 Pierobon P, Tino A, Minei R, Marino G (2004) Different roles of GABA and glycine in the modulation of chemosensory responses in Hydra vulgaris. Hydrobiologia 530/531:59-66 Pires AR, Croll RP, Hadfield MG (2000a) Catecholamines modulate metamorphosis in the opistobranch gastropod Phestilla sibogae. Biological Bulletin 198:319-331 157

Pires AR, Guilbault TR, Mitten JV, Skiendzielewski JA (2000b) Catecholamines in larvae and juveniles of the prosobranch gastropod, Crepidula fornicata. Comparative Biochemistry and Physiology C 127:37-47 Pires AR, Hadfield MG (1991) Oxidative breakdown products of catecholamines and hydrogen peroxide induce partial metamorphosis in the nudibranch Phestilla sibogae Bergh (Gastropoda: Opisthobranchia). Pires AR, Woollacott RM (1997) Serotonin and dopamine have opposite effects on phototaxis in larvae of the bryozoan Bugula neritina. Biological Bulletin 192:399- 409 Python F, Stocker RH (2002) Immunoreactivity against choline acetyltransferase, gamma- aminobutyric acid, histamine, octopamine, and serotonin in the larval chemosensory system of Drosophila melanogaster. The Journal of Comparative Neurology 453:157-167 Quinn, GP, Keough, MJ (2002) Experimental design and data analysis for biologists. Cambridge University Press, Cambridge Raff EC, Popodi EM, Kaufmann J, Sly BJ, Turner FR, Morris VB, Raff R (2003) Regulatory punctuated equilibrium and convergence in the evolution of developmental pathways in direct-developing sea urchins. Evolution and Development 5:478-493 Raimondi PT (1988) Settlement cues and determination of the vertical limit of an intertidal barnacle. Ecology 69:400-407 Reite OB (1972) Comparative physiology of histamine. Physiological Reviews 52:778-819 Riffell JA, Krug PK, Zimmer RK (2002) Fertilisation in the sea: the chemical identity of an abalone sperm attractant. The Journal of Experimental Biology 205:1439-1450 Rittschof D (1985) Oyster drills and the frontiers of chemical ecology: unsettling ideas. American Malacology Bulletin, special edition 1:111 Rittschof D, Branscomb ES, Costlow JD (1984) Settlement and behaviour in relation to flow and surface in larval barnacles, Balanus amphitrite, Darwin. Journal of Experimental Biology and Ecology 82:131-146

158

Rittschof D, Forward Jr. RB, Simons DA, Reddy PA, Erickson BW (1989) Peptide analogs of the mud crab pumping pheromone: Structure-function studies. Chemical Sense 14:137-148 Roeder T (2003) Metabotropic histamine receptors-nothing for invertebrates? European Journal of Pharmacology 466:85-90 Rowley RJ (1989) Settlement and recruitment of sea urchins (Strongylocentrotus spp.) in a sea-urchin barren ground and a kelp bed: are populations regulated by settlement or post-settlement processes? Marine Biology 100:485-494 Sammarco PW, Andrews JC (1989) The Helix experiment: differential localised dispersal and recruitment patterns in Great Barrier Reef corals. Limnology and Oceanography 34:896-912 Selvakumaraswamy P, Byrne M (2000) Reproduction, spawning and development of 5 ophiuroids from Australia and New Zealand. Invertebrate Biology 119:394-402 Shanks AL (1985) Behavioural basis of internal -wave-induced shoreward transport of megalopae of the crab Pachygrapsus crassipes. Marine Ecology Progress Series 24:289-295 Shilling FM, Manahan DT (1994) Energy metabolism and amino acid transport during early development of Antarctic and temperate echinoderms. Biological Bulletin 187:398-407 Shimizu K, Hunter E, Fusetani N (2000) Localisation of biogenic amines in larvae of Bugula neritina (Bryozoa: Chelistomatida) and their effects on settlement. Marine Biology 136:1-9 Smith AB (1988) Phylogenetic relationship, divergence times, and rates of molecular evolution for camarodont sea urchins. Molecular Biology and Evolution 5:345-365 Smith AB (1997) Echinoderm Larvae and Phylogeny. Annual Review of Ecology and Systematics 28:219-241 Smith AB, Jeffery CH (1997) Changes in the diversity, taxic composition and lifestyle of echinoids over the past 145 million years. In: Culver SJ, Rawson P (eds) Biotic Response to Global Change: The last 145 Million Years. Chapman & Hall, London Smith AB, Littlewood DTJ, Wray GA (1996) Comparative evolution of larval and adult life-history stages and small subunit ribosomal RNA among post-Palaeozoic 159

echinoids. In: H HP, Leigh Brown AJ, Maynard Smith J, Nee S (eds) New uses for New Phylogenies. Oxford University Press, Oxford Smythe GA, Braga O, Brew BJ, Grant RS, Guillemin GJ, Kerr SJ, Walker DW (2002) Concurrent quantification of quinolic, picolinic, and nicotinic acids using electron- capture negative-ion gas chromatography-mass spectrometry. Analytical Biochemistry 301:21-26 Snelgrove PVR (1994) Hydrodynamic enhancement of invertebrate larval settlement in microdepositional environments: colonization tray experiments in a muddy habitat. Journal of Experimental Marine Biology and Ecology 176:149-166 Snelgrove PVR, Butman CA, Grassle JP (1993) Hydrodynamic enhancement of larval settlement in the bivalve Mulinia lateralis (Say) and the polychaete Capitella sp.I in microdepositional environments. Journal of Experimental Marine Biology and Ecology 168:71-109 Snelgrove PVR, Grassle JP, Butman CA (1998) Sediment choice by settling larvae of the bivalve, Spisula solidissima (Dillwyn), in flow and still water. Journal of Experimental Marine Biology and Ecology 231:171-190 Steinberg PD (1995) Interaction between the canopy dwelling echinoid Holopneustes purpurescens and its host kelp Ecklonia radiata. Marine Ecology Progress Series 127:169-181 Steinberg PD, de Nys R, Kjelleberg S (2001) Chemical mediation of surface colonisation. In: McClintock JB, Baker BJ (eds) Marine Chemical Ecology. CRC Press, Florida, p 335-387 Stoner DS (1992) Vertical distribution of a colonial ascidian on a coral reef: the roles of larval dispersal and life-history variation. American Naturalist 139:802-824 Stoner DS (1994) Larvae of a colonial ascidian use a non-contact mode of substratum selection on a coral reef. Marine Biology 121:319-326 Strathmann RR (1985) Feeding and non-feeding larval development and life-history evolution in marine invertebrates. Annual Review of Ecology and Systematics 16:339-361 Strathmann RR, Branscomb ES, Vedder K (1981) Fatal errors in set as a cost of dispersal and the influence of intertidal flora on set of barnacles. Oecologia 48:13-18 160

Stricker SA (1999) Comparative biology of calcium signaling during fertilisation and egg activation in animals. Developmental Biology 211:157-176 Stuart AE (1999) From fruit flies to barnacles, histamine is the neurotransmitter of arthropod photoreceptors. Neuron 22:431-433 Stuart AE, Mekeel HE, Kempter E (2002) Uptake of the neurotransmitter histamine into the eyes of larvae of the barnacle (Balanus amphitrite). Biological Bulletin 202:53-60 Sulkin SD, Van Heukelem W, Kelly P, Van Heukelem L (1980) The behavioural basis of larval recruitment in the crab Callinectes sapidus Rathbun: a laboratory investigation of ontogenetic changes in geotaxis and barokinesis. Biological Bulletin 159:402-417 Svane IB, Young CM (1989) The ecology and behaviour of ascidian larvae. Oceanography and Marine Biology-Annual Review 27:45-90 Swanson RL, de Nys R, Huggett MJ, Green JK, Steinberg P (2006) In situ quantification of a natural settlement cue and recruitment of the Australian sea urchin Holopneustes purpurascens. Marine Ecology Progress Series 314:1-14 Swanson RL, Williamson JE, de Nys R, Kumar N, Bucknall MP, Steinberg PD (2004) Induction of settlement of larvae of the sea urchin Holopneustes purpurascens by histamine from a host alga. Biological Bulletin 206:161-172 Tamburri MN, Finelli CM, Wethey DS, Zimmer-Faust RK (1996) Chemical induction of larval settlement behavior in flow. Biological Bulletin 191:367-373 Tamburri MN, Zimmer-Faust RK, Tamplin ML (1992) Natural sources and properties of chemical inducers mediating settlement of oyster larvae: a re-examination. Biological Bulletin 183:327-338 Tankersley RA, Mc Kelvey LM, Forward RB (1995) Responses of estuarine crab megalopae to pressure, salinity and light: implications for flood tide transport. Marine Biology 122:391-400 Tegtmeyer K, Rittschof D (1989) Synthetic peptide analogs to barnacle settlement pheromone. Peptides 9:1403-1406 Teruya T, Suenaga K, Koyama T, Nakano Y, Uemura D (2001) Arachidonic acid and alpha-linolenic acid, feeding attractants for the crown of thorns sea star Acanthaster

161

planci from the sea urchin Toxopneustes pileolus. Journal of Experimental Marine Biology and Ecology 266:123-134 Thompson RC, Norton TA, Hawkins SJ (1998) The influence of epilithic microbial films on the settlement of Semibalanus balanoides cyprids - a comparison between laboratory and field experiments. Hydrobiologia 375/376:203-216 Thorson G (1950) Reproduction and larval ecology of marine bottom invertebrates. Biological Reviews 25:1-45 Thorson G (1964) Light as an ecological factor in the dispersal and settlement of marine bottom invertebrates. Ophelia 1:167-208 Tomas F, Romero J, Turon X (2004) Settlement and recruitment of the sea urchin Paracentrotus lividus in two contrasting habitats in the Mediterranean. Marine Ecology Progress Series 282:173-184 Toonen RJ, Pawlik JR (2001a) Settlement of the gregarious tube-worm Hydroides dianthus (Polychaeta: Serpulidae). I. Gregarious and nongregarious settlement. Marine Ecology Progress Series 224:103-114 Toonen RJ, Pawlik JR (2001b) Settlement of the gregarious tube-worm Hydroides dianthus (Polychaeta: Serpulidae). II. Testing the desperate larva hypothesis. Marine Ecology Progress Series 224:115-131 Trapido-Rosenthal HG (2001) Contributions of marine chemical ecology to chemosensory neurobiology. In: McClintock JB, Baker BJ (eds) Marine Chemical Ecology. CRC Press, Florida, p 463-479 Trapido-Rosenthal HG, Morse DE (1985) L- Diamino acids facilitate GABA induction of larval metamorphosis in a gastropod mollusc (Haliotis rufescens). Comparative Biochemistry and Physiology B 155:403-414 Trapido-Rosenthal HG, Morse DE (1986a) Availability of chemosensory receptors is down-regulated by habituation of larvae to a morphogenetic signal. Proceedings of the National Academy of Science 83:7658-7662 Trapido-Rosenthal HG, Morse DE (1986b) Regulation of receptor-mediated settlement and metamorphosis in larvae of a gastropod mollusc (Haliotis rufescens). Bulletin of Marine Science 39:383-392

162

Tremblay R, Sinclair M (1990) Diel vertical migration of sea scallop larvae Placopecten magellanicus in a shallow embayment. Marine Ecology Progress Series 67:19-25 Troxler RF, Lester R (1967) Biosynthesis of phycocyanobilin. Biochemistry 6:3840-3846 Tsukamoto S, Hirota H, Kato H, Fusetani N (1993) Urochordamines A and B: larval settlement/metamorphosis-promoting, pteridine-containing physostigmine alkaloids from the tunicate Ciona savignyi. Tetrahedron Letters 34:4819-4822 Tsukamoto S, Kato H, Hirota H, Fusetani N (1994) Narains: N, N -Dimethylquanidium styryl sulfates, metamorphosis inducers of ascidian larvae from a marine sponge Jaspis sp. Tetrahedron Letters 35:5873-5874 Tsukamoto S, Kato H, Hirota H, Fusetani N (1995) Pipecolate derivatives, Anthosamines A and B, inducers of larval metamorphosis in ascidians, from a marine sponge Anthosigmella aff. raromicrosclera. Tetrahedron 51:6687-6694 Tsukamoto S, Kato H, Hirota H, Fusetani N (1999) Lumichrome- A larval metamorphosis- inducing substance in the ascidian Halocynthia roretzi. European Journal of Biochemistry 264:785-789 Turner EJ, Zimmer-Faust RK, Palmer MA, Luckenbach M, Pentcheff ND (1994) Settlement of oyster (Crassostrea virginica) larvae: Effects of water flow and water-soluble chemical cue. Limnology and Oceanography 39:1579-1593 Tyler PA (1949) A simple non-injurious method for inducing repeated spawning of sea urchins and sand dollars. Collecting Net 19:19-20 Unabia CR, Hadfield MG (1999) Role of bacteria in larval settlement and metamorphosis of the polychaete Hydroides elegans. Marine Biology 133:55-64 Underwood AJ, Keough MJ (2000) Supply-side ecology: The nature and consequences of variations in recruitment of intertidal organisms. In: Bertness MD, Gaines SD, Hay ME (eds) Marine Community Ecology. Sinauer Associates Inc., Sunderland, Massachussets, p 183-200 Velasquez RD, Brunner G, Varrentrapp M, Tsikas D, Frolich JC (1996) Heliobacter pylori produces histamine and spermidine. Z Gastroenterol 34:116-122 Walker G (1971) The cement apparatus of the cypris larva of the barnacle, Balanus balanoides. Marine Biology 9:205-212

163

Walker G, Yule AB (1984) Tempory adhesion of the barnacle cyprid: the existence of an antennular adhesive secretion. Journal of the Marine Biological Association of U.K. 64:679-686 Weinreich D (1979) Gamma-Glutamylhistamine: a major product of histamine metabolism in ganglia of the marine mollusc, Aplysia californica. Journal of Neurochemistry 32:363-369 Wieczorek SK, Todd CD (1998) Inhibition and facilitation of settlement of epifaunal marine invertebrate larvae by microbial biofilm cues. Biofouling 12:81-118 Williamson JE (2001) Larval attractants versus adult deterrents: ontogenetic shifts in the response of a sea urchin to host plant chemistry. PhD, University of New South Wales Williamson JE, Carson DG, de Nys R, Steinberg PD (2004) Demographic consequences of an ontogenetic shift by a sea urchin in response to host plant chemistry. Ecology 85:1355-1371 Williamson JE, de Nys R, Kumar N, Carson DG, Steinberg PD (2000) Induction of metamorphosis in the sea urchin Holopneustes purpurascens by a metabolite complex from the algal host Delisea pulchra. Biological Bulletin 198:332-345 Williamson JE, Steinberg PD (2002) Reproductive cycle of the sea urchin Holopneustes purpurascens (Temnopleuridae: Echinodermata). Marine Biology 140:519-532 Wilson DP (1953) The settlement of Ophelia bicornis Savigny larvae. The 1951 experiments. Journal of the Marine Biological Association of U.K. 31:413-438 Wilson WH (1991) Competition and predation. Annual Review of Ecology and Systematics 21:221-241 Woollacott RM, Zimmer RK (1971) Attachment and metamorphosis of the cheilo- ctenostome byrozoan Bugula neritina (Linné). Journal of Morphology 134:351-382 Wray GA (2002) Do convergent developmental mechanisms underlie convergent phenotypes. Brain Behavior and Evolution 59:327-336 Wright JR, Boxshall AJ (1999) The influence of small-scale flow and chemical cues on the settlement of two congeneric barnacle species. Marine Ecology Progress Series 183:179-187

164

Wright JT, Steinberg PD (2001) Effect of variable recruitment and post-recruitment herbivory on local abundance of a marine alga. Ecology 82:2200-2215 Xue-Wu L, Gordon ME (1987) Tissue and cell culture of the New Zealand Pterocladia and Porphyra species. Hydrobiologia 151-152:147-154 Yoguchi R, Okuzumi M, Fujii T (1990) Seasonal variation in number of halophilic histamine-forming bacteria on marine fish. Nippon Suisan Gakkaishi 56:1473-1479 Yool AJ, Grau SM, Hadfield MG, Jensen RA, Markell DA, Morse DE (1986) Excess potassium induces larval metamorphosis in four marine invertebrate species. Biological Bulletin 170:255-266 Young CM (1995) Behavior and locomotion during the dispersal phase of larval life. In: McEdward LR (ed) Ecology of Marine Invertebrate Larvae. CRC press, New York, p 249-277 Yvin JC, Chevolet L, Chevolet-Maguer AM, Cochard JC (1985) First isolation of jacarone from an alga, Delesseria sanguinea: a metamorphosis inducer of Pecten larvae. Journal of Natural Products 48:814-816 Zega G, Pennati R, Groppelli S, Sotgia C, De Bernardi F (2005) Dopamine and serotonin modulate the onset of metamorphosis in the ascidian Phallusia mammillata. Developmental Biology 282:246-256 Zimmer RK, Commins JE, Browne KA (1999) Regulatory effects of environmental chemical signals on search behaviour and foraging success. Ecology 80:1432-1446 Zimmer-Faust RK, Tamburri MN (1994) Chemical identity and ecological implications of a waterborne, larval settlement cue. Limnology and Oceanography 39:1075-1087

Induction of Settlement of Larvae of the Sea Urchin Holopneustes purpurascens by Histamine From a Host Alga

REBECCA L. SWANSON1,4,*, JANE E. WILLIAMSON,1,4,†, ROCKY DE NYS1,4,‡, NARESH KUMAR2,4, MARTIN P. BUCKNALL3, AND PETER D. STEINBERG1,4 1 School of Biological, Earth & Environmental Sciences, 2 School of Chemical Sciences, 3 Bioanalytical Mass Spectrometry Facility, 4 Center for Marine Biofouling & Bio-Innovation, University of New South Wales, Sydney 2052, Australia

Abstract. Larvae of the Australian sea urchin Holopneu- purpurascens—D. pulchra and Ecklonia radiata—and of stes purpurascens are induced to settle and metamorphose four other common species was quantified using GC-MS. D. (termed settlement herein) by a water-soluble compound pulchra had the highest histamine content, which is consis- produced by the red alga Delisea pulchra, the main host tent with H. purpurascens recruiting to this species. Hista- plant of new recruits. The settlement cue for H. purpura- mine was also detected in the seawater surrounding these scens had previously been identified as a floridoside-isethio- host algae. This is the first time that a settlement cue has nic acid complex, and this paper presents new evidence been quantified in the habitat of a marine organism. correcting that finding. The actual settlement cue produced by D. pulchra was isolated from the polar extract by cation- Introduction exchange chromatography and identified as histamine, using one- and two-dimensional nuclear magnetic resonance Most marine invertebrates have complex life histories in spectrometry. The chemical identity of the cue was con- which a dispersive larval phase alternates with benthic firmed by gas chromatography–mass spectrometry (GC- juvenile and adult phases. The demography of such species MS) and matrix-assisted laser desorption/ionization–time- is highly dependent on larval recruitment to a favorable of-flight mass spectrometry. Synthetic histamine and habitat (Pawlik, 1992; Underwood and Keough, 2000), and histamine at 4.5 ␮M isolated from D. pulchra both induced the question of how planktonic larvae locate an appropriate rapid settlement in 80%–100% of the larvae of H. purpura- benthic habitat in which to settle has long been a focus for scens. Lower concentrations of histamine (0.9–2.3 ␮M) marine biologists. The current view is that hydrodynamic induced larval settlement, but this response varied from processes dominate at large spatial scales (meters, kilome- 0%–90%. The histamine content of two host plants of H. ters), with active habitat selection becoming progressively more important at smaller spatial scales (centimeters, mil- limeters, micrometers) (Keough and Downes, 1982; Mul- Received 24 October 2003; accepted 8 April 2004. lineaux and Butman, 1991; Harvey and Bourget, 1997; * To whom correspondence should be addressed. E-mail: Zimmer and Butman, 2000). [email protected] † Current address: Department of Biological Sciences, Macquarie Uni- Active habitat selection requires that larvae discriminate versity, NSW 2109 Australia. among potential settlement sites, which is possible through ‡ Current address: School of Marine Biology & Aquaculture, James the detection of habitat-specific cues. Many laboratory ex- Cook University, QLD 4811 Australia. periments have confirmed that larvae from a diverse range Abbreviations: CX, cation-exchange; F-I, floridoside-isethionic acid; of phyla respond both behaviorally (settlement—sinking to GC-MS, gas chromatography–mass spectrometry; HPLC, high-performance liquid chromatography; ISTD, internal standard; MALDI-TOF MS, the bottom for substrate exploration) and morphologically matrix-assisted laser desorption/ionization–time-of-flight mass spectrome- (metamorphosis—ontogenesis into the benthic form) to try; NMR, nuclear magnetic resonance; SSW, sterile seawater. such physical factors of a habitat as light (Maida et al.,

161 162 R. L. SWANSON ET AL.

1994), surface orientation (Raimondi and Morse, 2000), size classes of H. purpurascens were most abundant on D. flow conditions (Mullineaux and Butman, 1991), crevices pulchra, with the smallest size class (test diameter Յ 5 mm) (Keough and Downes, 1982), and surface texture (Bernts- found only on that species. This suggested that D. pulchra son et al., 2000). Larvae can also be induced to settle and might produce a settlement cue for larval H. purpurascens metamorphose (collectively termed settlement in this paper) (Williamson et al., 2000). Fresh pieces of D. pulchra (but by surface-bound or waterborne chemical cues, which are not E. radiata) and seawater collected in situ near D. pul- thought to indicate a suitable habitat for the benthic stage chra plants induced settlement in larvae of H. purpurascens. (Hadfield and Paul, 2001). The source of such chemical The water-soluble cue was subsequently isolated and char- cues (inducers) may be conspecifics (Burke, 1986), host acterized as a complex between the sugar floridoside and organisms (Williamson et al., 2000), prey (Hadfield and isethionic acid (F-I complex; Williamson et al., 2000). Scheuer, 1985), or biofilms (Wieczorek and Todd, 1998). During further research on this system, we obtained in- The chemical cues for larval settlement that have been ductive fractions that contained isethionic acid but not flori- isolated from natural sources within the habitat appear to be doside, and we were also unable to reproduce a synthetic F-I diverse; however, most have been only partially character- complex that induced settlement of larval H. purpurascens. ized. These include small peptides (the sand dollar Den- Subsequently, we hypothesized that the F-I complex was draster excentricus—Burke, 1984; the oyster Crassostrea not a natural settlement cue for this urchin. This paper virginica—Zimmer-Faust and Tamburri, 1994; the jellyfish identifies the true nature of this chemical cue from D. Cassiopea xamachana—Fleck and Fitt, 1999), uncharacter- pulchra for settlement of H. purpurascens larvae, correcting ized low-molecular-weight water-soluble compounds (the the previous finding of Williamson et al., (2000). In addi- nudibranch Phestilla sibogae—Hadfield and Pennington, tion, we quantify the settlement cue in host and non-host 1990; the cephalaspidean Haminaea callidegenita—Gibson algae of H. purpurascens—the first time that a natural and Chia, 1994), carbohydrates (the coral Agaricia humi- settlement cue has been quantified in the habitat of a marine lis—Morse and Morse, 1996; the ascoglossan Alderia mod- organism. esta—Krug and Manzi, 1999), and glycoproteins (the barnacle Balanus amphitrite—Clare and Matsumura, 2000). Materials and Methods In contrast to the numerous partially characterized inducers, there are only a few examples in which the chemical Study site structure of a settlement cue isolated from a natural source has been determined. Delta-tocopherols from Sargassum All animals and algae used in this study were collected tortile induced settlement of the hydroid Coryne uchidai from sublittoral habitats (1–3 m depth) at Bare Island (33° (Kato et al., 1975), jacarone isolated from the red alga 59' 38" S, 151° 14' 00" E) at the north head of Botany Bay, Delesseria sanguinea induced settlement of the scallop Sydney, Australia. At this site individuals of Holopneustes Pecten maximus (Yvin et al., 1985), narains and antho- purpurascens are primarily found wrapped in the laminae of samines A and B isolated from marine sponges induced the brown kelp Ecklonia radiata (Laminariales: Phaeo- settlement of ascidian larvae (Tsukamoto et al., 1994, phyta) or in the fronds of the red foliose alga Delisea 1995), and lumichrome isolated from conspecifics induced pulchra (Bonnemaisonales: Rhodophyta). A more detailed settlement of larvae of the ascidian Halocynthia roretzi description of this habitat and the ecology of this system are (Tsukamoto et al., 1999). In most cases, the ecological found in Wright and Steinberg (2001) and Williamson et al. relevance of these compounds in situ is not clear, either (2004). because the source of the settlement cue is not necessarily related to the recruitment patterns of the organism (Yvin et Preparation of the polar extract of Delisea pulchra al., 1985; Tsukamoto et al., 1994, 1995), or because the availability of the cue to settling larvae has not been dem- The results of Williamson et al. (2000) indicated that any onstrated (Tsukamoto et al., 1999). settlement cues were contained within the polar fraction of A naturally occurring characterized settlement cue that the crude extract of D. pulchra. A polar extract of D. appears to strongly affect the demography of the sea urchin pulchra was thus prepared from 1.0 kg (wet weight) of algae Holopneustes purpurascens Agassiz 1872 (Temnopluridae: collected from Bare Island. Epibiota were removed, the Echinodermata) was recently reported by Williamson et al. plants blotted dry, and the thallus exhaustively extracted in (2000). H. purpurascens is an endemic Australian echinoid methanol (OmniSolv, EM Science). The methanol extract that lives in shallow subtidal habitats in the canopy of was filtered (Whatman #1), dried by rotary evaporation in macroalgae, particularly Delisea pulchra Greville (Mon- vacuo at 40 °C, and partitioned between dichloromethane tagne) 1844 and Ecklonia radiata (C. Agardh) J. Agardh (OmniSolv) and Milli-Q water. The Milli-Q phase was 1898 at Bare Island, Sydney (Williamson et al., 2000, filtered (Whatman #1) and dried in vacuo at 40 °C. The 2004). Although abundant on both host plants, the smaller dried crude polar extract was dissolved in absolute ethanol SETTLEMENT INDUCTION BY HISTAMINE 163 three times, pooling each extract, and dried in vacuo at 40 settlement assays and analyzed by 1H- and 13C-nuclear °C to yield the polar extract. magnetic resonance (NMR) spectroscopy (Bruker DMX 500). Isolation of the settlement cue in Delisea pulchra by bioassay-guided fractionation Cation-exchange chromatography. The settlement cue could not be isolated as a pure fraction using HPLC, so an High-performance liquid chromatography. The polar ex- alternative procedure, cation-exchange (CX) chromatogra- tract of D. pulchra was fractionated using reversed-phase phy, was used to fractionate the polar extract of D. pulchra. high-performance liquid chromatography (HPLC—Adsor- CX resin (AG50W-X2 [Hϩ form], BioRad) in Milli-Q ␮ ϫ bosil C18 column, 5- m particle size, 250 mm 4.6 mm, water was poured into a 50-ml burette, taking care to Waters R410 RI-detector) (100% Milli-Q water at 1 exclude air bubbles. The resin (25-ml bed volume) was ⅐ Ϫ1 ml min ). The polar extract was dissolved in Milli-Q equilibrated with Milli-Q water at 2 ml ⅐ minϪ1 until the water (50 mg ⅐ mlϪ1), filtered (0.22 ␮m), and manually eluant was pH 5–6. The polar extract of D. pulchra (1–2g) injected (20 ␮l). HPLC resolved two major peaks, peak 1 was dissolved in 5 ml of Milli-Q water, filtered (0.22 ␮m), with a retention time (rt) of 2.7 min, and peak 2 with rt ϭ and gently loaded onto the column. Unbound compounds 3.4 min (Fig. 1A). Each peak fraction was collected from were collected in 100 ml of Milli-Q water (fraction 1) and multiple injections and dried by rotary evaporation in vacuo another 100 ml of Milli-Q water (fraction 2). Retained at 40 °C. Peak fractions were tested for bioactivity in compounds were eluted using a series of basic solutions: 30

ml of dilute NH3 in Milli-Q water (pH 10; fraction 3), 30 ml of 3%-NH4OH w/w (fraction 4), and 30 ml of 30%-NH4OH w/w (fraction 5, Fig. 1B). Fractions 1–5 were collected as controls, using the same method but without loading any D. pulchra extract on the column; none of these fractions had any subsequent activity. CX-fractions were dried in a centrifuge in vacuo (speed-vac SVC200, Savant), tested for bioactivity in settlement assays, and analyzed by 1H-NMR spectroscopy.

Identification of isolated settlement cue Nuclear magnetic resonance spectroscopy. Bioassay- guided fractionation of the polar extract of D. pulchra by cation-exchange chromatography yielded one active fraction (CX-fraction 5, F5). The inducing compound in F5 was 1 13 identified by H and C-NMR experiments (D2O), and a high-field two-dimensional 1H-15N HMBC NMR experi-

ment (d4 MeOH, Bruker DMX 500). To conﬁrm the putative structure of F5 as histamine, 3 mg of F5 was dissolved 1 in D2O and analysed by H-NMR spectroscopy. Synthetic histamine (3 mg) was added to F5 and the sample re- analysed. The 1H-NMR spectra of the unspiked F5 sample and the spiked F5 sample were then compared.

Gas chromatographyÐmass spectrometry. NMR spectroscopy analyses identified the isolated settlement cue as histamine, and this was confirmed by gas chromatography– mass spectrometry (GC-MS). Putative (naturally isolated) histamine (1 mg) and synthetic histamine (1 mg) were derivatized with heptafluorobutyric anhydride (Aldrich) and then acetic anhydride (Aldrich), using the method of Baran- Figure 1. Diagram of the bioassay-guided fractionation of the polar cin et al. (1998). Derivatized samples were diluted 100-fold extract of Delisea pulchra, using either reversed-phase HPLC (A) or cation-exchange (CX) chromatography (B). MQ ϭ Milli-Q water, FR ϭ in ethyl acetate before analysis. A Zebron ZB-5 column (15 flow rate, rt ϭ retention time, F ϭ fraction. m, 0.25 ␮m ϫ 0.25 mm ID; Phenomenex) was used on a 164 R. L. SWANSON ET AL.

Hewlett Packard (HP) 5980 series II gas chromatograph because this species is a “dribble” spawner and generally equipped with an HP5971A or HP5972 mass selective de- yields low numbers of larvae (settlement is not gregarious; tector. Injections (2 ␮l) were in the splitless mode with an Williamson et al., 2000). Larvae were added once all petri inlet pressure of 170 kPa. The injection port was held as 290 dishes were prepared, and percent settlement (i.e., percent °C and the interface at 300 °C. The gas chromatograph was metamorphosed) was recorded at set time intervals. held at 90 °C for 2 min and ramped at 10 °C ⅐ minϪ1 to 200 ⅐ Ϫ1 °C, then at 50 °C min to 310 °C and held for 2 min (17.2 HPLC peak fractions. Peak 1 and peak 2 fractions were min per run). Helium was used as the carrier gas. The mass tested against larvae to determine the presence of a settle- m/z selective detector was operated in scan mode ( 50–550). ment cue. Peak fractions were dissolved in Milli-Q water The average retention times of derivatized putative hista- Ϫ (10 mg ⅐ ml 1 stock solution) and aliquots of each stock mine and derivatized synthetic histamine were recorded solution were added to assigned petri dishes for final test from five injections of each sample (mean Ϯ SD, n ϭ 5). Ϫ Ϫ concentrations of 25 ␮g ⅐ ml 1 of peak 1 and 51 ␮g ⅐ ml 1 The electron impact ion-spectra of derivatized putative his- of peak 2. A floridoside-isethionic complex sample (“F-I tamine and derivatized synthetic histamine were compared. complex”) from the previous study (Williamson et al., 2000) was also tested in the assay at a final concentration of Matrix-assisted laser desorption/ionizationÐtime-of-flight Ϫ 76 ␮g ⅐ ml 1. Pieces of fresh D. pulchra (ϳ10 mg) were mass spectrometry. The elemental formula of putative his- used as a positive control, and Milli-Q water and SSW were tamine was determined by matrix-assisted laser desorption/ used as the negative controls. Percent settlement was scored ionization–time-of-flight mass spectrometry (MALDI-TOF ϭ MS) (Bucknall et al., 2002). A Perseptive Voyager DE STR after 18 h (n 12 replicates per treatment). (Perseptive Biosystems, Framingham, MA) MALDI-TOF MS was operated in both positive-ion linear delayed-extrac- Cation-exchange fractions. Each CX-fraction (F) was tested tion mode and reflector delayed-extraction mode for accu- against larvae to determine the presence of a settlement cue. rate mass analysis. The test samples were prepared in ace- F1, F2, F3, F4 and the polar extract of D. pulchra (used as tonitrile/Milli-Q water (50:50) and contained either 100 a positive control) were dissolved in Milli-Q water at 5 Ϫ ng ⅐ ␮lϪ1 of putative histamine or synthetic histamine. mg ⅐ ml 1. Aliquots of the appropriate fraction were added ␣-Cyano-4-hydroxycinnamic acid (5 mg ⅐ mlϪ1) prepared in to the petri dish to give final test concentrations of 50 acetonitrile/Milli-Q/trifluoroacetic acid (80:20:0.02) was ␮g ⅐ mlϪ1 for each treatment. F5 was dissolved in Milli-Q Ϫ used as the matrix. Glycine (500 ng ⅐ ␮l 1) and [sarcosine- water at 100 ␮g ⅐ mlϪ1, and aliquots were added to assigned 15 ⅐ ⅐ ␮ Ϫ1 N-methyl-d3]creatinine HCl (5 ng l , Cambridge petri dishes for final test concentrations of 0.1–1.0 Isotope Laboratories #DNLM-2171) were added as internal ␮g ⅐ mlϪ1 (much lower concentrations of F5 were tested mass calibrants for accurate mass determinations. An accu- because of a low yield in F5). Initial settlement assays rate mass for the putative protonated histamine molecular showed that only F5 induced settlement; therefore, CX- ϩ ion [MϩH] was determined by 10 repeat analyses of each control-fraction 5 (CF5) was tested in future settlement sample. The mean molecular weight was calculated for assays as the procedural control. CF5 was dissolved in these mass spectra and compared with both the theoretical Milli-Q water at 100 ␮g ⅐ mlϪ1 and tested at 1.0 ␮g ⅐ mlϪ1. molecular weight for histamine and the molecular weight Milli-Q water and SSW were used as the negative controls. measured for synthetic histamine using the same analytical Percent settlement was scored after1h(n ϭ 10 replicates technique. The standard deviation for these mass measure- per treatment). ments was taken as an estimate of the mass measurement error. Natural versus synthetic histamine. Settlement assays were used to compare the responses of larvae to (i) natural Settlement assays histamine isolated using CX chromatography. (ii) synthetic H. purpurascens larvae were cultured as previously de- histamine, and (iii) synthetic histamine run through the scribed (Williamson et al., 2000). Larvae reached compe- same procedure used to isolate natural histamine. Stock ␮ tency (i.e., become developmentally ready for settlement) solutions of 900 M of each histamine treatment were within 6 days, as recognized by the presence of five well- prepared in Milli-Q water, and aliquots of the appropriate developed tube feet. All settlement assays were done at 19 stock solution added to assigned petri dishes for final test °C with a 12-h-light/12-h-dark regime, in 40-mm petri concentrations of 0.9–9.0 ␮M. Pieces of fresh D. pulchra Ϫ dishes and 5 ml of sterile seawater (SSW). Replicates were (ϳ10 mg) and 50 ␮g ⅐ ml 1 of the polar extract D. pulchra randomly assigned among treatments, with 10–15 replicates were used as the positive controls, and Milli-Q water and per treatment and one 6-day larva per replicate dish. We SSW were used as the negative controls. Percent settlement were unable to use multiple larvae per dish in these assays was scored after1h(n ϭ 12 replicates per treatment). SETTLEMENT INDUCTION BY HISTAMINE 165

Delisea pulchra treated with antibacterial agents. Because bridge Isotope Laboratories, #DLM 2911) was added to some marine bacteria produce histamine (Fujii et al., 1997), each sample as the internal standard (ISTD). Strong cation- the identification of histamine as the settlement cue (see exchange solid-phase extraction cartridges (50 mg, Alltech) Results) raises the possibility that the bacterial biofilm on were equilibrated with Milli-Q water (5 ml) at a flow rate of the surface of D. pulchra may be the source of the cue. To 1ml⅐ minϪ1, and the sample was loaded. Unbound com- test this, the ability of D. pulchra to induce settlement after pounds were eluted in 2 ml of Milli-Q water (fraction 1) and various antibacterial treatments was examined in a settle- another 2 ml of Milli-Q water (fraction 2). All retained ment assay. Antibacterial treatments were adapted from compounds were eluted in 1 ml of 30% NH4OH w/w previous studies in which treatments were shown to be (fraction 3) and dried in a speed vac. Standards that con- effective in reducing surface bacteria (Xue-Wu and Gordon, tained 1-␮g ISTD and either 0.1, 0.5, 1.0, 5.0, or 10 ␮gof 1987; Aguirre-Lipperheide and Evans, 1993; Johnson and synthetic histamine were prepared. Standards and fraction 3 Sutton, 1994). Seven plants of D. pulchra were collected samples were derivatized with heptafluorobutyric anhydride from Bare Island and brought back to the laboratory, where and acetic anhydride using the method of Barancin et al. portions of each plant were allocated to each of seven (1998). treatments. There were six antibacterial treatments and a A DB-5MS column (15 m, 0.25 ␮m ϫ 0.25 mm ID, J & procedural control. All antibacterial treatments included a W Scientific) and a packed liner (3% SP-2250, Supelco; 5-min soak in a 10% betadine-SSW solution, followed by Smythe et al., 2002) were installed on the GC-MS instru- three rinses in SSW and a 24-h treatment in either (1) SSW ment previously described, and the same run conditions (the “soak” treatment); (2) SSW containing 20 mg ⅐ lϪ1 were used. The mass selective detector was operated in streptomycin (Aldrich), 10 mg ⅐ lϪ1 penicillin G (Aldrich), selected ion monitoring mode using ions characteristic of and 10 mg ⅐ lϪ1 kanamycin (Aldrich; “SPK” treatment); (3) the analyte (derivatized histamine—m/z 94, 307, 349) and SSW containing 10 mg ⅐ lϪ1 ciprofloxacin (Bayer, “cipro- the ISTD (m/z 97, 311, 353). Extracted ion chromatograms floxacin” treatment); (4) SSW after pieces of D. pulchra were used to manually integrate the area under each ion were gently wiped across an agar plate, before and after the peak (which is proportional to the amount of analyte in the 24-h soak, to physically remove bacteria (“wipe” treat- sample). For each standard and sample, the areas of the ment); and the combination treatments (5) “wipe ϩ SPK”, analyte ions (m/z 94, 307, 349) were added together and the and (6) “wipe ϩ ciprofloxacin.” The procedural control was areas of the ISTD ions (m/z 97, 311, 353) were added a 24-h soak in SSW without the initial betadine soak (“soak together. The ratio of the combined areas of analyte:ISTD in control” treatment). The next day, subsections of several D. standards was used to generate a standard curve. The his- pulchra plants were collected as a “fresh control” treatment tamine content of the samples was calculated by reference and used in the settlement assay on that day. Pieces of D. to the standard curve and expressed in terms of micrograms pulchra (ϳ10 mg) from each treatment were added to per gram (wet weight) of algal tissue (␮g ⅐ gϪ1). assigned sterile petri dishes, and percent settlement was After checking that the data met the assumptions of the scored after 20 h (n ϭ 15 replicates per treatment). test, the histamine content of different algae was transformed [ln(x ϩ 1)] and compared by using a one-factor Quantitative analysis of histamine in various algae analysis of variance. We excluded A. anceps and C. officinalis from the analysis because no histamine was detected If histamine is a natural settlement cue for this urchin, we in these species. Bonferroni’s post hoc test was used to would expect D. pulchra, the primary host plant of new determine which species differed in their histamine contents recruits of H. purpurascens, to have higher levels of hista- (SYSTAT ver. 7.0). We were concerned that one high value mine than other algae in the habitat. To test this, we quan- for D. pulchra might be unduly influencing our analysis, but tified the histamine content of six species of algae from the the outcome was unchanged when we repeated the analysis habitat of H. purpurascens. The two primary host plants (D. with this value omitted. Therefore, we report the results of pulchra and E. radiata) and four other prominent species of the initial analysis. algae (Amphiroa anceps, Corallina officinalis, Homeostri- chus olsenii, and Sargassum vestitum) were collected from Reanalysis of samples from Williamson et al. (2000) Bare Island in January 2003. Five replicates of each alga Samples remaining from the study published in William- were analyzed, with each replicate consisting of three small son et al. (2000) were analyzed by GC-MS for the presence sections taken from different parts of one thallus, which of histamine. Any histamine in the old samples was isolated were then pooled into a single sample for analysis (2–4g using cation-exchange solid phase extraction cartridges, as wet-weight). A polar extract of each algal sample was outlined previously for isolating algal histamine, and then prepared as described above. Polar extracts were dissolved derivatized with heptafluorobutyric anhydride and acetic in Milli-Q water (200 ␮l) and acidified with 50 ␮l of glacial anhydride, using the method of Barancin et al. (1998) for ␣ ␣ ␤ ␤ ⅐ ␮ acetic acid. [ , , , -d4]Histamine 2HCl (1 g, Cam- quantitative GC-MS analysis. 166 R. L. SWANSON ET AL.

Results and comparison with synthetic samples. When isethionic acid (1–25 ␮g ⅐ mlϪ1), sodium isethionate (15–30 Isolation of the settlement cue in by Ϫ Ϫ Delisea pulchra ␮g ⅐ ml 1), and taurine (1–13 ␮g ⅐ ml 1) were tested in bioassay-guided fractionation settlement assays with H. purpurascens larvae, none of HPLC fractions—NMR spectroscopy analysis and settle- these compounds induced settlement (data not shown). Dif- ment assays. The polar extract of Delisea pulchra was ferent combinations of these compounds were tested to- Ϫ separated into two fractions using HPLC (peak 1 and peak gether (e.g., 15 ␮g ⅐ ml 1 of isethionic acid and taurine) in 2, Fig. 1A). These were analyzed by NMR spectroscopy case two cues were required for settlement of H. purpura- (1-min 1H- and 30-min 13C-NMR experiments) and tested scens. There was no settlement in the combination treat- in settlement assays. Peak 1 displayed the pattern of isethio- ments (data not shown). Following these results, we hypoth- nic acid (Barrow et al., 1993), as determined by 1H- and esized that one or more trace compounds in peak 1, not yet 13C-NMR spectroscopy, as well as some additional signals detected by NMR analysis, were inducing settlement. To that were not characteristic of floridoside (see next section). test this, a larger amount of peak 1 was collected and a much The 13C-NMR spectrum of peak 2 corresponded to previ- longer (24-h) 13C-NMR experiment run on the sample. The 13 ously published data for floridoside [␣-D-galactopyranosyl- C-NMR spectrum showed about 20 additional carbon (1-2)-glycerol] (Karsten et al., 1993). Therefore, the isethio- signals not detected previously by NMR spectroscopy, in- nic acid and floridoside components of the F-I complex dicating that additional compounds were present in peak 1 eluted separately, in peak 1 and peak 2, respectively. Peak in trace amounts. The finding that peak 1 induced settlement 1 induced settlement of Holopneustes purpurascens larvae of larvae of H. purpurascens but the identified major com- in settlement assays, but peak 2 did not (Fig. 2). Four ponents (isethionic acid, taurine) in peak 1 did not implied batches of peak 1 (25 ␮g ⅐ mlϪ1) induced 80%–100% set- that one of the compounds present in trace amounts was the tlement in five assays, whereas neither of two batches of settlement cue. peak 2 (51 ␮g ⅐ mlϪ1) induced settlement in two assays (representative data shown in Fig. 2). These data suggested Cation-exchange fractions—settlement assay. The settle- that the F-I complex is not a settlement cue for H. purpura- ment cue could not be isolated as a pure fraction using scens and that peak 1 (which lacked floridoside) contained HPLC, so the polar extract of D. pulchra was fractionated the settlement cue. using CX chromatography (Fig. 1B). Five CX-fractions (F) Isethionic acid and taurine were the major compounds in were obtained and tested in settlement assays; only F5 peak 1, as determined by 1H- and 13C-NMR spectroscopy induced settlement of larvae of H. purpurascens (Fig. 3). F5 at a concentration of 1.0 ␮g ⅐ mlϪ1 induced 100% settlement in larvae after 1 h, 0.5 ␮g ⅐ mlϪ1 induced 70% settlement, and 0.1–0.25 ␮g ⅐ mlϪ1 did not induce settlement. There was no settlement in the control fraction CF5 (1.0 ␮g ⅐ mlϪ1) and SSW treatments (Fig. 3).

Identification of the settlement cue for Holopneustes purpurascens Nuclear magnetic resonance spectroscopy. The 1H-NMR ␦ (D2O) spectrum of F5 showed proton signals at 2.76 (2H, t, J 7.0 Hz, H2), 3.03 (2H, t, J 8.2 Hz, H1), 6.86 (1H, s 2H, imidazole H), and 7.57 (s, 1H, imidazole H). The 13C-NMR ␦ (D2O) and DEPT spectra of F5 showed carbon signals at 25.9, 39.5 (CH2); 116.4, 136.5 (CH) and 134.0 (quaternary C). These signals supported the assignment of F5 as histamine (2-[1H-imidazol-4-yl]-ethylamine, MW 111.15). The structure of F5 was further confirmed by a high-field two- dimensional 1H-15N HMBC NMR experiment in which the Figure 2. The settlement (%) of larvae of Holopneustes purpurascens methylene triplet at 2.76 ppm showed two three-bond cor- after 18 h incubation with fresh Delisea pulchra (ϳ10 mg) or HPLC peak relations to the ethylamine NH2 group and the imidazole fractions of the polar extract of the alga. Peak 1 (batch A or B) was tested nitrogen. The identity of F5 was further confirmed by a at 25 ␮g ⅐ mlϪ1, peak 2 was tested at 51 ␮g ⅐ mlϪ1, and a floridoside- isethionic acid complex sample (“F-I complex”) from Williamson et al. spiking experiment. All F5 signals increased in intensity (2000) was tested at 76 ␮g ⅐ mlϪ1. Milli-Q water and sterile seawater and no additional signals were detected, confirming the (SSW) were included as the negative controls (n ϭ 10). identity of F5 as histamine. SETTLEMENT INDUCTION BY HISTAMINE 167

elemental formulas with a mass of approximately

112.08878. The nearest other candidate was C6H10NO at 112.07569 with a difference of 117 ppm from the measured mass of putative protonated histamine. This difference was much higher than 17 ppm (difference of measured mass for

putative histamine relative to calculated mass for C5H10N3), conﬁrming that the putative protonated histamine had the

elemental formula of C5H10N3.

The response of Holopneustes purpurascens larvae to natural and synthetic histamine Natural histamine isolated from D. pulchra by using CX chromatography, synthetic histamine, and synthetic histamine eluted from CX resin all resulted in very similar responses in larvae when assayed concurrently (Fig. 4). Figure 3. The settlement (%) of larvae of Holopneustes purpurascens More than 80% of the H. purpurascens larvae settled within after 1 h incubation with the polar extract of Delisea pulchra (PE) and ␮ cation-exchange fractions (F) of the PE. The different test concentrations of an hour of incubation in 4.5 and 9 M natural and synthetic each treatment are shown in brackets (␮g ⅐ mlϪ1); note the lower concen- histamine. The lowest test concentration of synthetic hista- trations for F5 and the procedural control (CF5). Sterile seawater (SSW) mine that consistently induced rapid settlement of all larvae was used as the negative control (n ϭ 10). was 4.5 ␮M (in 10 separate assays). Larvae exhibited a more variable response to 0.9 and 2.3 ␮M histamine, both within and across different batches (Fig. 4). Up to 80% of Gas chromatographyÐmass spectrometry. The identity of larvae settled in response to 0.09–0.45 ␮M synthetic hista- putative histamine (F5) isolated from D. pulchra was con- mine, but only after long incubation times (up to 96 h) or as firmed using GC-MS. The retention times (rt) of the hep- larval age increased to 13–21 days (data not shown). tafluorobutyrlacyl derivative of putative histamine (rt ϭ 9.728 Ϯ 0.0045, mean Ϯ SD, n ϭ 5) and synthetic hista- Response of larvae to Delisea pulchra after antibacterial mine (rt ϭ 9.732 Ϯ 0.0045, mean Ϯ SD, n ϭ 5) were nearly treatments identical, suggesting that they were the same compound. The electron-impact ion spectra of both derivatized com- In response to D. pulchra that had received antibacterial pounds displayed the same major fragment ions (m/z—54, treatments, larvae of H. purpurascens settled at levels 69, 81, 94, 138, 169, 226, 307, 349) and overall fragmen- equivalent to (or greater than) those in response to control tation pattern, confirming that they were the same compound. The electron-impact ion spectra for derivatized histamine matched that reported in the literature (Barancin et al., 1998).

Matrix-assisted laser desorption/ionizationÐtime-of-ﬂight mass spectrometry. The elemental formula of putative histamine isolated from D. pulchra was conﬁrmed by accurate mass measurements using MALDI-TOF MS. The measured accurate mass of the putative protonated histamine molecular ion [MϩH]ϩ was 112.08878 Ϯ 0.0026 (n ϭ 10, mean Ϯ SD), and the measured mass for synthetic histamine was 112.08853 Ϯ 0.0025 (n ϭ 10, mean Ϯ SD). The measured masses of the two samples were different by only 2.2 ppm. These values were different from the calculated monoisotopic mass for protonated histamine (112.08692—elemen- Figure 4. The settlement (%) of larvae of Holopneustes purpurascens ϳ tal formula C5H10N3) by only 15 ppm for synthetic proton- after 1 h incubation with fresh Delisea pulchra ( 10 mg), its polar extract ated histamine and 17 ppm for putative protonated (50 ␮g ⅐ mlϪ1) and 0.9–9.0 ␮M of natural histamine isolated from D. histamine. This is most likely due to measurement bias pulchra (Nat), synthetic histamine (Syn), or synthetic histamine eluted from cation-exchange resin (Syn-CX). Sterile seawater (SSW) was used as introduced by the very different chemical properties of a negative control. Data from two experiments using different batches of histamine, glycine, and creatinine (the internal calibrants). larvae are shown (black and white bars). * Indicates no settlement in An elemental calculator was used to generate all possible treatment (n ϭ 12). 168 R. L. SWANSON ET AL.

D. pulchra treatments (Fig. 5). SSW did not induce settle- Table 1 ment. The histamine content of six algal species was significantly different ϭ ϭ (ANOVA, F3,16 9.903, P 0.0006) Quantitative analysis of histamine content in algae Histamine (␮g ⅐ gϪ1 [ww]) The histamine content of six algal species was deter- Species mean Ϯ SE, n ϭ 5 mined by GC-MS (Table 1). D. pulchra, the alga on which Delisea pulchra 11.82 Ϯ 6.56 new recruits of H. purpurascens are found (Williamson et Ecklonia radiata 1.28 Ϯ 1.01* al., 2000), had the highest histamine content of all algae Sargassum vestitum 0.35 Ϯ 0.32* surveyed. Histamine was not detected in any samples of Homeostrichus olsenii 0.25 Ϯ 0.09* Amphiroa anceps or Corallina officinalis. The histamine Corallina officinalis nd Amphiroa anceps nd content of D. pulchra, Ecklonia radiata, Homeostrichus olsenii, and Sargassum vestitum differed significantly from * Indicates species in which histamine content differs significantly from ϭ ϭ D. pulchra (pairwise comparisons, P Ͻ 0.0092); nd, not detected; ww, wet each other (F3,16 9.903, P 0.0006). Pairwise comparisons showed that the histamine content of D. pulchra weight. (11.82 Ϯ 6.56 ␮g ⅐ gϪ1) was significantly higher than the histamine content of E. radiata (1.28 Ϯ 1.01 ␮g ⅐ gϪ1, P ϭ 0.0092), S. vestitum (0.35 Ϯ 0.32 ␮g ⅐ gϪ1, P ϭ 0.0016), of the S. vestitum samples analyzed, but another contained Ϫ1 and H. olsenii (0.25 Ϯ 0.09 ␮g ⅐ gϪ1, P ϭ 0.0015). The 1.48 ␮g ⅐ g wet weight of algal tissue. The H. olsenii amount of histamine in different D. pulchra plants was plants analyzed showed consistently low levels of hista- Ϫ1 highly variable, ranging from 1.88–34.22 ␮g ⅐ gϪ1 wet mine, ranging from 0.05–0.46 ␮g ⅐ g (wet weight) of weight of algal tissue. The variability in histamine levels of algal tissue. E. radiata was also high, with no histamine detected in two samples, yet another contained 4.73 ␮g ⅐ gϪ1 wet weight of Reanalysis of samples from Williamson et al. (2000) algal tissue. Likewise, histamine was not detected in three Several samples remaining from the previous study were analyzed by GC-MS for the presence of histamine. Hista- mine was detected in F-I complex fractions from D. pulchra (1.5–46 ␮g ⅐ mg [sample]Ϫ1) in a synthetic F-I complex sample (0.35 ␮g ⅐ mg [sample]Ϫ1), and in a batch of floridoside used to make the synthetic complexes (0.45 ␮g ⅐ mg [sample]Ϫ1).

Discussion Habitat-specific cues play an important role in the settlement of many benthic marine invertebrates (Pawlik, 1992; Hadfield and Paul, 2001). Larvae presumably maximize their chances of post-settlement survival by responding to habitat-specific cues, as settlement in a preferred habitat should provide shelter and food to the vulnerable juvenile life-history phase (Gosselin and Qian, 1997). Chemical cues for larval settlement are derived from conspecifics (Burke, Figure 5. The settlement (%) of larvae of Holopneustes purpurascens 1986), host organisms (Williamson et al., 2000), prey (Had- after 20 h incubation with Delisea pulchra subjected to antibacterial field and Scheuer, 1985), or biofilms (Wieczorek and Todd, treatments. All antibacterial treatments included a 5-min soak in a 10% 1998); they include a diverse range of compounds from betadine solution, followed by 3 rinses in sterile seawater (SSW) and a small peptides (Zimmer-Faust and Tamburri, 1994) to com- 24-h treatment in either SSW (“soak”); SSW containing streptomycin (20 mg ⅐ 1Ϫ1), penicillin G (10 mg ⅐ lϪ1) and kanamycin (10 mg ⅐ lϪ1, “SPK”); plex macromolecules (Clare and Matsumura, 2000). The or SSW containing ciprofloxacin (10 mg ⅐ lϪ1, “ciprofloxacin”). Other complete characterization of chemical settlement cues has, treatments involved wiping pieces of D. pulchra across an agar plate however, proved difficult because of the low endogenous or gently, to physically remove bacteria, before and after a 24-h soak in SSW environmental levels of such compounds and the rapid ϩ (“wipe”), SSW containing SPK (“wipe SPK”), or SSW containing dilution of water-soluble cues. Few studies have definitively ciprofloxacin (10 mg ⅐ lϪ1, “wipe ϩ cipro”). D. pulchra soaked in SSW for 24 h (without betadine soak) was the procedural control (“soak control”), characterized settlement cues (reviewed by Hadfield and fresh D. pulchra was used as a positive control (“fresh control”), and SSW Paul, 2001; Steinberg et al., 2001). was used as a negative control (n ϭ 15). Williamson et al. (2000) reported on one such putative SETTLEMENT INDUCTION BY HISTAMINE 169 characterized cue, a metabolite complex isolated from the methanol as the mobile phase, and eluted as a single peak red algal host Delisea pulchra that induced settlement in (Williamson et al., 2000). 13C-NMR spectroscopy analysis larvae of the sea urchin Holopneustes purpurascens. At of this peak showed only 13C-signals for floridoside and Bare Island (Sydney, Australia), H. purpurascens is found isethionic acid (Williamson et al., 2000). However, trace primarily on two algal hosts, D. pulchra and Ecklonia amounts of histamine were also present but not detected, radiata, with the smallest size class (test diameter Յ 5 mm) because their levels were below the limit of detection for only found on D. pulchra. Larvae metamorphosed in re- 13C-NMR spectroscopy. Histamine elutes in the first peak sponse to pieces of D. pulchra and the polar extract, but not from reversed-phase (C18) columns regardless of the mo- to pieces or extracts of E. radiata (Williamson et al., 2000). bile phase, so any histamine in the polar extracts of D. A water-soluble cue was implicated when seawater col- pulchra used by Williamson et al. (2000) would have co- lected near D. pulchra plants in situ also induced settlement eluted with the F-I complex fraction. Consequently, the “F-I of larvae. The settlement cue in D. pulchra was isolated and complex” samples contained histamine, detected here using characterized as the floridoside-isethionic acid (F-I) com- GC-MS, and induced settlement of H. purpurascens larvae. plex (Williamson et al., 2000). Although a synthetic F-I complex induced rapid settlement New evidence presented in this paper shows that hista- in H. purpurascens larvae in the previous study (William- mine, not the F-I complex, is a natural inducer of settlement son et al., 2000), not all batches induced settlement (R. de in H. purpurascens. The settlement cue was isolated from Nys, pers. obs.). The synthetic F-I complexes were made the polar extract of D. pulchra by using bioassay-guided using natural floridoside isolated from D. pulchra and syn- fractionation by cation-exchange chromatography. The iso- thetic isethionic acid. The floridoside used to make the lated compound at 0.5 ␮g ⅐ mlϪ1 induced settlement in synthetic F-I complex was contaminated by histamine and 80%–100% of larvae within an hour. The settlement cue thus induced settlement. Confirming this, histamine was was identified as histamine using NMR spectroscopy, and detected by GC-MS in a floridoside sample (used for prep- this was confirmed by GC-MS and MALDI-TOF MS. The aration of the synthetic complex) and a synthetic F-I com- response of larvae to synthetic histamine in settlement as- plex sample prepared by Williamson et al. (2000). In sum- says mirrored their response to natural histamine isolated mary, histamine was present in trace amounts in the “F-I from D. pulchra. D. pulchra, the primary plant on which complex” samples that induced settlement of larval H. pur- new recruits of H. purpurascens are found, had the highest purascens in the previous study, and histamine was the average histamine content (11.82 Ϯ 6.56 ␮g ⅐ gϪ1 wet inductive compound in the “F-I complex” samples. weight), approximately an order of magnitude higher than The finding that histamine is a natural settlement cue for other algae surveyed. Seawater collected near D. pulchra H. purpurascens is of considerable interest in the context of plants in the study by Williamson et al. (2000) induced linking ecological patterns with physiological mechanisms. rapid settlement of larval H. purpurascens; however, those Histamine is a biogenic amine produced by the decarbox- samples were used completely in bioassays and are there- ylation of the amino acid histidine. It is one of five primary fore not available for histamine analysis. We have since biogenic amines in invertebrates, along with serotonin, oc- detected histamine in seawater surrounding D. pulchra and topamine, dopamine, and tyramine (Blenau and Baumann, E. radiata (at concentrations ranging from 20 to 70 nM), but 2001). Biogenic amines, all decarboxylation products of not in samples 2 m away from the macroalgae. A compre- amino acids, play critical roles in initiating and controlling hensive analysis of histamine levels in seawater will be behavior, and in the physiology of invertebrates, by acting reported in another manuscript. Although the histamine as classical neurotransmitters, neuromodulators, and neuro- concentrations measured in seawater do not induce rapid hormones (Katz, 1995; Beltz, 1999). For example, dopa- settlement in larvae that have just attained competence, this mine activates hunting behavior in an opisthobranch mol- concentration can induce settlement of H. purpurascens lusc (Norekyan and Satterlie, 1993), and serotonin controls larvae over longer time periods and in older larvae (data not aggressive behavior in crustaceans (Huber et al., 1997). The shown). In addition, the natural habitat may contain other photoreceptors in all classes of arthropod eyes are histamin- settlement cues that if detected in conjunction with hista- ergic; that is, they synthesize histamine and use it as their mine, may lower the threshold concentration of histamine neurotransmitter (Stuart, 1999). Also, histamine is thought required for rapid induction of settlement. These findings to be an inhibitory neurotransmitter in the stomatogastric support our proposal that histamine released from macroal- and cardiac ganglia and the sensory system of lobsters gae is a natural settlement cue for H. purpurascens. (Claiborne and Selverston, 1984; Bayer et al., 1989; Reanalysis of samples from the study by Williamson et Hashemzadeh-Gargari and Freschi, 1992). Importantly, in al. (2000) provides an explanation for the incorrect conclu- the context of this study, histamine directly gates a chloride sion that the F-I complex is a settlement cue for larvae of H. channel in the receptor cells of the olfactory pathway of purpurascens. The F-I complex was isolated from the polar lobsters (McClintock and Ache, 1989). Fast neurotransmit- extract of D. pulchra, using reversed-phase HPLC and ters directly gate ion channels, which leads to fast behav- 170 R. L. SWANSON ET AL. ioral and physiological outcomes. We have observed that purascens, which is consistent with the variation we mea- the settlement response of H. purpurascens to histamine is sured in levels of histamine in the alga. Given this, and the rapid, with complete metamorphosis within half an hour. large biomass of E. radiata kelp beds in the natural habitat This fast response is consistent with the notion that the of H. purpurascens, E. radiata may contribute to environ- larvae of H. purpurascens have specific receptors that bind mental levels of histamine, inducing the settlement of larvae histamine and act directly on ion channels, leading to rapid in this habitat. Histamine was not detected in the turfing settlement. coralline algae Corallina officinalis and Amphiroa anceps, Neurotransmitters, or their precursors, have been sug- although they induce settlement of larvae of H. purpura- gested to mimic the function of natural settlement cues scens (Williamson et al., 2000; R. Swanson, pers. obs.) and (Morse, 1985; Bonar et al., 1990). The best-known example provide a habitat for new recruits (R. Swanson, pers. obs.). is the gamma-aminobutyric acid (GABA)-mimetic peptide Larger samples of A. anceps (up to 180 g wet weight) were (or peptides), present on the surface of crustose coralline extracted and no histamine was detected. The coralline algae, which Morse and colleagues proposed as a settlement algae may produce a different settlement cue for H. pur- cue for abalone (Morse et al., 1979, 1984). Another exam- purascens, or histamine may only be produced and released ple comes from oyster larvae, where L-3, 4-dihydroxyphe- in situ—for example, by surface-associated bacteria. nylalanine (L-DOPA) induced stereotypical searching be- Finally, the two possible sources of histamine in D. havior, while epinephrine and norepinephrine induced pulchra are the host alga or the surface-associated bacterial metamorphosis (Coon et al., 1985). Endogenous levels of community (or both). D. pulchra treated with various anti- neurotransmitters, and their precursors, also appear to mod- bacterial agents still induced high levels of settlement in H. ulate the behavioral and physiological processes accompa- purpurascens, suggesting that the host alga produces the nying settlement (Coon and Bonar, 1987; Pires et al., 2000). histamine. A bacterial source of histamine is, however, Our findings show that a naturally produced neurotransmit- possible, as a known histamine-producing bacterium, Pho- ter is in fact a settlement cue for larvae, a phenomenon that tobacterium phosphoreum (Fujii et al., 1997) is a constitu- may be widespread in the marine environment. ent of the microbial community on local algal species (M. The finding that histamine, rather than the F-I complex, is Watson, UNSW Australia; pers. comm.). If histamine-pro- a settlement cue for H. purpurascens potentially compli- ducing bacteria are colonizing algal surfaces within the cates the previous interpretations of the relationship be- habitat, then it is possible that they produce and release tween settlement cues and the demography of this sea histamine to seawater, which could lead to the induction of urchin (Williamson et al., 2000, 2004). Histamine, a simple settlement of H. purpurascens. breakdown product of the amino acid histidine, may be broadly distributed across the natural habitat of H. purpura- Conclusion scens; for example, in algal and animal tissue, and in bacterial communities living on their surfaces. For hista- Many larval species have the ability to respond to low- mine to be an ecologically relevant settlement cue, its dis- molecular-weight, water-soluble settlement cues (Hadfield tribution in the natural habitat must relate to the recruitment and Scheuer, 1985; Zimmer-Faust and Tamburri, 1994; patterns of H. purpurascens. This was in fact the case. The Boettcher and Targett, 1996; Lambert et al., 1997; Fleck histamine content of the algae surveyed was consistent with and Fitt, 1999). This paper has presented evidence that the recruitment patterns of the organism, with much higher histamine is a natural settlement cue for the sea urchin, H. levels of histamine measured in D. pulchra, the primary purpurascens, correcting the previous study of Williamson plant on which we find new recruits. et al. (2000). Histamine at 4.5 ␮M induces settlement (meta- D. pulchra had the highest average histamine content morphosis) in 80%–100% of H. purpurascens larvae within (11.82 Ϯ 6.56 ␮g ⅐ gϪ1 wet weight), ranging from 1.88 to half an hour, fulfilling two essential criteria for an effective 34.22 ␮g.gϪ1 wet weight. Similarly, levels of histamine water-soluble settlement cue: (1) larvae must perceive low varied for E. radiata, with concentrations ranging from 0 to concentrations of inducer, and (2) larvae must respond 4.73 ␮g.gϪ1 wet weight. Since only subsections of plants rapidly to the inducer. D. pulchra had the highest histamine (not whole plants) were extracted, these results may reflect content of all the species surveyed, consistent with the within-plant variation, within-species variation, or both. Fu- recruitment patterns of H. purpurascens. In a preliminary ture histamine analyses will extract whole plants, as well as analysis, we detected histamine in seawater near D. pulchra specific regions of thalli, to directly test these possibilities. and E. radiata plants, but not in seawater collected 2 m The low (or absent) levels of histamine typically measured away, supporting our proposal that histamine leaches from in E. radiata samples may explain why pieces and extracts algae that produce this settlement cue. This hypothesis is of E. radiata did not induce settlement in the study by consistent with the finding of Williamson et al. (2000) that Williamson et al. (2000). However, we have observed that seawater collected very near to D. pulchra induced settle- some pieces of E. radiata do induce settlement of H. pur- ment in larvae of H. purpurascens. We have shown that SETTLEMENT INDUCTION BY HISTAMINE 171 histamine, an invertebrate neurotransmitter, is also a natural settlement and metamorphosis of the Pacific oyster Crassostrea gigas settlement cue for larvae of H. purpurascens, linking the (Thunberg) by L-DOPA and catecholamines. J. Exp. Mar. Biol. Ecol. physiology and ecology of the organism. 94: 211–221. Fleck, J., and W. K. Fitt. 1999. Degrading mangrove leaves of Rhizo- phora mangle Linne provide a natural cue for settlement and metamor- Acknowledgments phosis of the upside down jellyfish Cassiopea xamachana Bigelow. J. Exp. Mar. Biol. Ecol. 234: 83–94. This research was supported by an Australian Postgrad- Fujii, T., A. Hiraishi, T. Kobayashi, R. Yoguchi, and M. Okuzumi. uate Award to RLS, an ARC Research Fellowship to RdN, 1997. Identification of the psychrophilic histamine-producing marine an ARC Large Grant to PDS and RdN, and the Center for bacteria previously referred to as the N-group bacteria. Fish. Sci. 63(5): 807–810. Marine Biofouling and Bio-Innovation. We thank the Bio- Gibson, G. D., and F.-S. Chia. 1994. A metamorphic inducer in the analytical Mass Spectrometry Facility at UNSW for access opisthobranch Haminaea callidegenita: partial purification and biolog- to equipment, and Dr. Dustin Marshall for advice on statis- ical activity. Biol. Bull. 187: 133–142. tical analysis. We especially thank Dr. Tim Charlton for his Gosselin, L. A., and P. Y. Qian. 1997. Can bacterivory alone sustain invaluable assistance with analytical methods. larval development in the polychaete Hydriodes elegans and the barnacle Balanus amphitrite? Mar. Ecol. Prog. Ser. 161: 93–101. Hadfield, M. G., and V. J. Paul. 2001. Natural chemical cues for Literature Cited settlement and metamorphosis of marine invertebrate larvae. Pp. 431– 461 in Marine Chemical Ecology, J. B. McClintock and B. J. Baker, Aguirre-Lipperheide, M., and L. V. Evans. 1993. A sterilization pro- eds. CRC Press, Boca Raton, FL. tocol for the Dictyotales (Phaeophyceae). J. Phycol. 29: 243–251. Hadfield, M. G., and J. T. Pennington. 1990. Nature of the metamor- Barancin, C. E., J. C. Smoot, R.H. Findlay, and L. A. Actis. 1998. phic signal and its internal transduction in larvae of the nudibranch Plasmid-mediated histamine biosynthesis in the bacterial fish pathogen Phestilla sibogae. Bull. Mar. Sci. 46: 455–464. Vibrio anguillarum. Plasmid 39: 235–244. Hadfield, M. G., and D. Scheuer. 1985. Evidence for a soluble meta- Barrow, K. D., U. Karsten, and R. J. King. 1993. Isethionic acid from morphic inducer in Phestilla sibogae: ecological, chemical and biolog- the marine red alga Ceramium flaccidum. Phytochemistry 35: 1429– ical data. Bull. Mar. Sci. 37: 556–566. 1430. Harvey, M., and E. Bourget. 1997. Bayer, T. A., T. S. McClintock, U. Grunert, and B. W. Ache. 1989. Recruitment of marine invertebrates Histamine-induced modulation of olfactory receptor neurons in two onto arborescent epibenthic structures: active and passive processes species of lobster, Panulirus argus and Homarus americanus. J. Exp. acting at different spatial scales. Mar. Ecol. Prog. Ser. 153: 203–215. Biol. 145: 133–146. Hashemzadeh-Gargari, R. M., and J. E. Freschi. 1992. Histamine Beltz, B. S. 1999. Distribution and functional anatomy of amine-contain- activates a chloride conductance in motor neurons of the lobster cardiac ing neurons in decapod crustaceans. Microsc. Res. Tech. 44: 105–120. ganglion. J. Neurophysiol. 68: 9–15. Berntsson, K. M., P. R. Jonsson, M. Lejhall, and P. Gatenholm. 2000. Huber, R., K. Smith, A. Delago, K. Isaksson, and E. A. Kravitz. 1997. Analysis of behavioral rejection of micro-textured surfaces and impli- Serotonin and aggressive motivation in crustaceans: altering the deci- cations for recruitment by the barnacle Balanus improvisus. J. Exp. sion to retreat. Proc. Natl. Acad. Sci. USA 94: 5939–5942. Mar. Biol. Ecol. 251: 59–83. Johnson, C. R., and D. C. Sutton. 1994. Bacteria on the surface of Blenau, W., and A. Baumann. 2001. Molecular and pharmacological crustose coralline algae induce metamorphosis of the crown-of-thorns properties of insect biogenic amine receptors: lessons from Drosophila starfish Acanthaster planci. Mar. Biol. 120: 305–310. melangoster and Apis mellifera. Arch. Insect Biochem. Physiol. 48: Karsten, U., K. D. Barrow, and R. J. King. 1993. Floridoside, l- 13–38. isofloridoside, and D-isofloridoside in the red alga Porphyra columbina. Boettcher, A. A., and N. M. Targett. 1996. Induction of metamorphosis Plant Physiol. Biochem. 103: 485–491. in queen conch, Strombus gigas Linnaeus, larvae by cues associated Kato, T., A. S. Kumanireng, I. Ichinose, Y. Kitahara, Y. Kakinuma, M. with red algae from their nursery grounds. J. Exp. Mar. Biol. Ecol. 196: Nishihira, and M. Kato. 1975. Active components of Sargassum 29–52. tortile effecting the settlement of swimming larvae of Coryne uchidai. Bonar, D. B., S. L. Coon, M. Walch, R. M. Weiner, and W. Fitt. 1990. Experientia 31: 433–434. Control of oyster settlement and metamorphosis by endogenous and Katz, P. S. 1995. Intrinsic and extrinsic neuromodulation of motor exogenous chemical cues. Bull. Mar. Sci. 46: 484–498. circuits. Curr. Opin. Neurobiol. 5: 799–808. Bucknall, M., K. Y. C. Yung, and M. W. Duncan. 2002. Practical Keough, M. J., and B. J. Downes. 1982. Recruitment of marine inver- quantitative biomedical applications of MALDI-TOF mass spectrom- tebrates: the role of active larval choices and early motility. Oecologia etry. J. Am. Soc. Mass Spectrom. 13: 1015–1027. 54: 348–352. Burke, R. D. 1984. Pheromonal control of metamorphosis in the Pacific Krug, P. J., and A. E. Manzi. 1999. Waterborne and surface-associated sand dollar, Dendraster excentricus. Science 225: 440–441. carbohydrates as settlement cues for larvae of the specialist marine Burke, R. D. 1986. Pheromones and the gregarious settlement of marine herbivore Alderia modesta. Biol. Bull. 197: 94–103. invertebrate larvae. Bull. Mar. Sci. 39: 323–331. Lambert, W. J., C. D. Todd, and J. D. Hardege. 1997. Partial char- Claiborne, B. J., and A. I. Selverston. 1984. Histamine as a neurotrans- acterization and biological activity of a metamorphic inducer of the mitter in the stomatogastric nervous system of the spiny lobster. J. Neu- dorid nudibranch Adalaria proxima (Gastropoda: Nudibranchia). In- rosci. 4: 708–721. vertebr. Biol. 116: 71–81. Clare, A. S., and K. Matsumura. 2000. Nature and perception of Maida, M., J. C. Coll, and P. W. Sammarco. 1994. Shedding new light barnacle settlement pheromones. Biofouling 15: 57–71. on scleractinian coral recruitment. J. Exp. Mar. Biol. Ecol. 180: 189– Coon, S. L., and D. B. Bonar. 1987. Pharmacological evidence that 202. alpha-1 adrenoceptors mediate metamorphosis of the Pacific oyster McClintock, T. S., and B. W. Ache. 1989. Histamine directly gates a Crassostrea gigas. Neuroscience 23: 1169–1174. chloride channel in lobster olfactory receptor neurons. Proc. Natl. Coon, S.L., D. B. Bonar, and R. M. Weiner. 1985. Induction of Acad. Sci. USA 86: 8137–8141. 172 R. L. SWANSON ET AL.

Morse, A. N. C., and D. E. Morse. 1996. Flypapers for coral and other Tsukamoto, S., H. Kato, H. Hirota, and N. Fusetani. 1995. Pipecolate planktonic larvae. Bioscience 46(4): 254–262. derivatives, anthosamines A and B, inducers of larval metamorphosis Morse, A. N. C., C. A. Froyd, and D. E. Morse. 1984. Molecules from in ascidians, from a marine sponge Anthosigmella aff. raromicrosclera. cyanobacteria and red algae that induce larval settlement and metamor- Tetrahedron 51: 6687–6694. phosis in the mollusc Haliotis rufescens. Mar. Biol. 81: 293–298. Tsukamoto, S., H. Kato, H. Hirota, and N. Fusetani. 1999. Lumi- Morse, D. E. 1985. Neurotransmitter-mimetic inducers of larval settle- chrome—a larval metamorphosis-inducing substance in the ascidian ment and metamorphosis. Bull. Mar. Sci. 37: 697–706. Halocynthia roretzi. Eur. J. Biochem. 264: 785–789. Morse, D. E., N. Hooker, H. Duncan, and L. Jensen. 1979. Gamma- Underwood, A. J., and M. J. Keough. 2000. Supply-side ecology: the aminobutyric acid, a neurotransmitter, induces planktonic abalone lar- nature and consequences of variations in recruitment of intertidal vae to settle and begin metamorphosis. Science 204: 407–410. organisms. Pp. 183–200 in Marine Community Ecology, M. D. Bert- Mullineaux, L. S., and C. A. Butman. 1991. Initial contact, exploration ness, S. D. Gaines, and M. E. Hay, eds. Sinauer Associates, Sunder- and attachment of barnacle (Balanus amphitrite) cyprids settling in land, MA. flow. Mar. Biol. 110: 93–103. Wieczorek, S. K., and C. D. Todd. 1998. Inhibition and facilitation of Norekyan, T. P., and R. Satterlie. 1993. Neuronal analysis of hunting settlement of epifaunal marine invertebrate larvae by microbial biofilm behavior of the pteropod mollusc Clione limacina. Neurosci. Behav. cues. Biofouling 12: 81–118. Physiol. 23: 11–23. Williamson, J. E., R. de Nys, N. Kumar, D. G. Carson, and P. D. Pawlik, J. R. 1992. Chemical ecology of the settlement of benthic Steinberg. 2000. Induction of metamorphosis in the sea urchin Ho- marine invertebrates. Oceanogr. Mar. Biol. Annu. Rev. 30: 273–335. lopneustes purpurascens by a metabolite complex from the algal host Pires, A., R. P. Croll, and M. G. Hadfield. 2000. Catecholamines Delisea pulchra. Biol. Bull. 198: 332–345. modulate metamorphosis in the opisthobranch gastropod Phestilla si- Williamson, J. E., D. G. Carson, R. de Nys, and P. D. Steinberg. 2004. bogae. Biol. Bull. 198: 319–331. Demographic consequences of an ontogenetic shift by a sea urchin in Raimondi, P. T., and A. N. C. Morse. 2000. The consequences of Ecology 85: complex larval behavior in a coral. Ecology 81: 3193–3211. response to host plant chemistry. 1355–1371. Smythe, G. A., O. Braga, B. J. Brew, R. S. Grant, G. J. Guillemin, S. J. Wright, J. T., and P. D. Steinberg. 2001. Effect of variable recruitment Kerr, and D. W. Walker. 2002. Concurrent quantification of quin- and post-recruitment herbivory on local abundance of a marine alga. olic, picolinic, and nicotinic acids using electron-capture negative-ion Ecology 82: 2200–2215. gas chromatography–mass spectrometry. Anal. Biochem. 301: 21–26. Xue-wu, L., and M. E. Gordon. 1987. Tissue and cell culture of the Steinberg, P. D., R. de Nys, and S. Kjelleberg. 2001. Chemical medi- New Zealand Pterocladia and Porphyra species. Hydrobiologia 151: ation of surface colonization. Pp. 335–387 in Marine Chemical Ecol- 147–154. ogy, J. B. McClintock and B. J. Baker, eds. CRC Press, Boca Raton, Yvin, J. C., L. Chevolet, A. M. Chevolet-Maguer, and J. C. Cochard. FL. 1985. First isolation of jacarone from an alga, Delesseria sanguinea: Stuart, A. E. 1999. From fruit flies to barnacles, histamine is the neu- a metamorphosis inducer of Pecten larvae. J. Nat. Prod. 48: 814–816. rotransmitter of arthropod photoreceptors. Neuron 22: 431–433. Zimmer, R. K., and C. A. Butman. 2000. Chemical signaling processes Tsukamoto, S., H. Kato, H. Hirota, and N. Fusetani. 1994. Narains: in the marine environment. Biol. Bull. 198: 168–187. N,N-dimethylguanidinium styryl sulfates, metamorphosis inducers of Zimmer-Faust, R. K., and M. N. Tamburri. 1994. Chemical identity ascidian larvae from a marine sponge Jaspis sp. Tetrahedron Lett. 35: and ecological implications of a waterborne, larval settlement cue. 5873–5874. Limnol. Oceanogr. 39: 1075–1087.