Biology and Ecology of Irukandji Jellyfish (Cnidaria: Cubozoa)

CHAPTER ONE

Lisa-ann Gershwin*,1, Anthony J. Richardson†,{, Kenneth D. Winkel}, Peter J. Fenner}, John Lippmann||, Russell Hore#, Griselda Avila- Soria**, David Brewer†, Rudy J. Kloser*, Andy Steven†, Scott Condie* *CSIRO Marine and Atmospheric Research, Castray Esplanade, Hobart, Tasmania, Australia †CSIRO Marine and Atmospheric Research, EcoSciences Precinct, GPO Box 2583, Dutton Park 4001, Qld, Australia { Centre for Applications in Natural Resource Mathematics (CARM), School of Mathematics and Physics, University of Queensland, St Lucia, 4072, Brisbane, Queensland, Australia } Australian Venom Research Unit, Department of Pharmacology and Therapeutics, University of Melbourne, Parkville, Victoria, Australia } Occupational Health Mackay, PO Box 3080, North Mackay, Queensland, Australia ||Divers Alert Network Asia-Pacific, PO Box 384 (49A Karnak Road), Ashburton, Victoria, Australia #Reef Biosearch, Meridian Marina, Port Douglas, Queensland, Australia **James Cook University, 1 James Cook Dr Douglas, Townsville, Queensland, Australia 1Corresponding author: e-mail address: [email protected]

Contents 1. Introduction 2 1.1 History of study 6 2. Biology of Irukandji 11 2.1 Taxonomy 11 2.2 Evolution 22 2.3 Reproduction and life cycle 24 2.4 Eyes and vision 30 2.5 Behaviour 34 3. Ecology of Irukandji 35 3.1 Diet and feeding 35 3.2 Geographic distribution 36 3.3 Vertical distribution 41 3.4 Temporal changes 44 3.5 Movements and aggregations 52 3.6 Environmental variables 57 4. Toxins 58 4.1 Which part of the animal is toxic? 64 4.2 Evolution of Irukandji toxins 65 5. Stinger Management 65 5.1 Prediction 66 5.2 Detection 67 5.3 Prevention 68 5.4 Treatment 68

6. Research Gaps 69 Acknowledgements 71 Appendix A. Notes on Rearing and Life Cycle of Carukia barnesi 72 Appendix B. Notes on Australian Alatina mordens Occurrence 75 References 76

Abstract Irukandji stings are a leading occupational health and safety issue for marine industries in tropical Australia and an emerging problem elsewhere in the Indo-Pacific and Caribbean. Their mild initialsting frequently results indebilitatingillness,involving signsof sympathetic excess including excruciating pain, sweating, nausea and vomiting, hypertension and a feeling of impending doom; some cases also experience acute heart failure and pulmonary oedema.Thesejellyfisharetypicallysmallandnearlyinvisible,andtheir infestationsaregen- erally mysterious, making them scary to the general public, irresistible to the media, and disastrous for tourism. Research into these fascinating species has been largely driven by the medical profession and focused on treatment. Biological and ecological information is surprisingly sparse, and is scattered through grey literature or buried in dispersed pub- lications, hampering understanding. Given that long-term climate forecasts tend toward conditions favourable to jellyfish ecology, that long-term legal forecasts tend toward increasing duty-of-care obligations, and that bioprospecting opportunities exist in the powerfulIrukandjitoxins,thereisaclear needforinformationtohelpinformglobalresearch and robust management solutions. We synthesise and contextualise available information on Irukandji taxonomy, phylogeny, reproduction, vision, behaviour, feeding, distribution, seasonality, toxins, and safety. Despite Australia dominating the research in this area, there are probably well over 25 species worldwide that cause the syndrome and it is an understudied problemin the developingworld. Major gaps inknowledge are identified for future research:our lackof clarityonthe socio-economic impacts,and our need fortimeseries and spatial surveys of the species, make this field particularly enticing. Keywords: Irukandji syndrome, Marine stingers, Envenomation, Jellyfish blooms, Carybdeida, Carukia, Malo, Morbakka, Gerongia, Alatina

1. INTRODUCTION

Seemingly minor stings from certain species of jellyfish can result in a constellation of debilitating symptoms in human victims, which in turn result in high medical costs, closed beaches, negative publicity, fear in the recreational public, and financial impacts for the tourism industry (see Box 1.1). These jellyfish, loosely grouped under the common name Irukandji, are understudied relative to their medical, financial, and social implications. Irukandji syndrome typically manifests as severe lower back and abdominal pain, relentless nausea and vomiting, full-body cramps and spasms, difficulty breathing, profuse sweating, anxiety, muscular restlessness, headaches, and a BOX 1.1 Why Irukandji research matters?

Top: (A) Mild sting to chest resulting in full-blown Irukandji syndrome, note localised sweating (copyright S. Cohen). (B) Beach closed due to Irukandji. (C). Mild bell-shaped sting to bicep resulting in Irukandji syndrome (copyright B. Currie). (D) Five-year-old female, whose Irukandji sting required 3 days in intensive care (copyright J. Margaglione). Bottom: A sample of media headlines about Irukandji (from the collection of K. Moss). Irukandji jellyfish meet all the criteria for a Hollywood horror film: Many are the size of a peanut and invisible in water; their four tentacles are 100 times their body length and as thin as cobwebs; their mild sting is rarely noticed, but within half an hour, the victim’s whole body is in agony and experiencing a bizarre constellation of seemingly unrelated symptoms. Many victims require hospitalisation, some require life support and some die. And Irukandji occasionally infest the most popular tropical beaches en masse. But consistent with the very best of Hitchcock, nobody has known when or where (or who) the danger will strike. These features make them downright scary and predictably attractive to the media. 4 Lisa-ann Gershwin et al. feeling of impending doom (Williamson et al., 1996). Many victims also experience coughing and/or involuntary grunting, shivering and teeth chattering, a creepy skin feeling, and, in some cases, priapism (prolonged erection) in males. In some cases hypertension is severe and life-threatening: the highest reading published as part of a case history is 280/180 (Fenner and Carney, 1999), and readings >300 have also been reported (Gershwin et al., 2009). These cases of hypertension may lead to pulmonary oedema (fluid on the lungs) and, rarely, to cerebral haemorrhage (stroke). A small proportion of cases develop some form of acute cardiac failure (Fenner and Carney, 1999; Huynh et al., 2003; Macrokanis et al., 2004; Nickson et al., 2009) and some, as yet undefined number, may have ongoing or recurrent symptoms. Two people died in Australia in 2002 from complications arising from Irukandji syndrome (Fenner and Hadok, 2002; Huynh et al., 2003). Another fatality in 2012 and two more in 2013 are still under investigation for a potential Irukandji basis. However, the actual death toll is likely to be much higher. Often with little or no mark on the body and nothing to test for postmortem, the mechanism of death would be a heart attack, stroke, or drowning or could even mimic decompression illness, and the underlying cause may never be recognised (Gershwin et al., 2009). Some larger Irukandji species, such as Morbakka prevalent in the Gulf of Thailand, can cause immediate severe pain and large weals prior to the onset of systemic symptoms and possible death (DAN AP Case reports; Fenner and Lippmann, 2009; Fenner et al., 2010). Although Irukandji are largely (but not entirely) unknown in temperate regions, in the tropics, the scale of the problem extends far beyond the medical effects. A conservative estimate placed the losses to the tourism industry due to negative publicity at more than $65 million in 2002 (Williams, 2004). Stings from Irukandji are considered the number one occupational health and safety issue for Australia’s tropical lobster fishery, pearling industry, and beˆche-de-mer fishery (Gershwin et al., 2009). Industrial downtime, as either the result of stings or the threat of stings, has impacted the Australian Navy, the oil and gas industry, and quite likely many other industries where personnel come in contact with tropical waters. One of the most unusual features of the Irukandji story, however, is that despite their danger to humans, the problem is poorly acknowledged globally and has received little attention from biologists and ecologists. Indeed, much of what we know about these animals beyond their medical effects is based on conjecture or scant anecdotal evidence, and even the Biology and Ecology of Irukandji Jellyfish 5

World Health Organization failed to recognise the hazard in their 2003 and 2009 Guidelines for Safe Recreational Water Environments (WHO, 2003, 2009). Even in regions with dozens of hospitalisations per year and decades of awareness programmes, the high number of stings suggests that the safety message is often ignored (Gershwin et al., 2009; Sando et al., 2010). Progress in research and management is compromised by our fragmented understanding of these species. For example, most of the literature has focused on envenomation and was published primarily in medical journals. Taxonomic information has largely appeared in taxonomic journals. Biolog- ical and ecological information is scattered through the pages of medical and taxonomic works or has appeared in grey literature. The lack of available information has likely exacerbated the reluctance of authorities in certain countries to acknowledge and provide information about the incidence and management of stings in their waters. The purpose of this chapter is to combine and summarise the disparate and often scant information about the biology and ecology of Irukandji jellyfish. We have tried to summarise global knowledge, but there is bias toward Australia because this is where most of the species are known and where they have been studied the longest. The emphasis on Carukia barnesi in the Great Barrier Reef region underscores the fact that we still know so little about other species from other regions. In many instances in this chapter, we draw on knowledge of other cubozoans to develop hypotheses about what features Irukandji are likely to have. This chapter is likely to be useful for researchers across a wide variety of disciplines including biology, ecology, taxonomy, toxinology, and evolutionary science and to medical researchers and rural and remote medical practitioners, those concerned with marine tourism threats, and managers concerned with provision of safety at beaches, tourism resorts, dive destinations, and marine industrial facilities. Those seeking detailed information on the medical and societal aspects of Irukandji syndrome are directed elsewhere (Fenner and Carney, 2001; Fenner and Hadok, 2002; Gershwin et al., 2009; Williamson et al., 1996), as these subjects are covered only briefly here. We hope that this comparative synthesis will help stimulate future research into these understudied species and, in so doing, contribute to an informed understanding of how to manage the threats they pose to human health. 6 Lisa-ann Gershwin et al.

1.1. History of study The name Irukandji was taken from an anglicized version of the name of the original aboriginal custodians of the lands between Cairns and Port Douglas (the Yirrganydji people), where Irukandji syndrome was first reported (Flecker, 1952). Prior to the first use of this term in 1952, a number of cli- nicians had noted a range of severe constitutional effects following a minor sting without skin wealing (Flecker, 1945, Southcott, 1952). Southcott and Powys (1944) termed these ‘type A stingings’ (formally published in Southcott, 1952). In late 1944, unbeknownst to these investigators, an army doctor in New Guinea had similar cases but had the sting skin lesions and causative agent (a small transparent jellyfish) pointed out by a local witch doctor (Barnes, 1964). Before that, a few scattered reports had emerged around the world noting clusters of stings with severe symptoms in common (Lord and Wilks, 1918; Old, 1908, 1912; Stenning, 1928). It was not until 1961 that an unclassified species of jellyfish was proven to be a causal agent of Irukandji syndrome, when Dr. Jack Barnes of Cairns, used a single specimen to sting himself, his 9-year-old son and a volunteer surf life-saver, all of whom became ill (Barnes, 1964). This jellyfish was later named Carukia barnesi, after its intrepid discoverer, and was long referred to as ‘the Irukandji jellyfish’ (Southcott, 1967). Only some 40 years later were additional species discovered and linked with variations of the syndrome (Gershwin, 2005b,c, 2007, 2008; Gershwin and Alderslade, 2005). The important advances in Irukandji syndrome research were reviewed by Tibballs and his colleagues (2012) and are expanded here in Table 1.1 to include Irukandji biology and ecology. To summarise the global knowledge of Irukandji, we conducted a meta- analysis of published literature using ISI Web of Knowledge Search Table 1.1 Timeline of the history of Irukandji research Date Event and reference 1935–1936 A Cairns medical conference in 1935 recommended the collection of information on injuries, including marine stings, from around North Queensland. This commenced under the care of Dr. Hugo Flecker who began registering stingings, including cases of what was retrospectively recognised as Irukandji syndrome, from December 1935 (reported in Flecker, 1952, and commented on by Barnes, 1960) 1944 Retrospectively reported in 1964, two cases of Irukandji syndrome from Noemfoer Island offshore northwest Papua in late 1944 were attributed to the sting of a “small (3–5 cm) almost colourless ... medusa by a local ‘witch doctor’” (Barnes, 1964) Biology and Ecology of Irukandji Jellyfish 7

Table 1.1 Timeline of the history of Irukandji research—cont'd Date Event and reference 1945 First formal recognition of the basic symptomatology and epidemiology of what is now known as the Irukandji syndrome; grouped as category (e) of “injuries by unknown agents to bathers in north Queensland” (Flecker, 1945a,b) 1952 Southcott retrospectively reports about 90 cases of ‘severe constitutional effects without wealing’ from ‘the beaches around Cairns’ in the summer of 1943–1944 and calls them ‘type A stingings’ (Southcott, 1952) The term ‘Irukandji syndrome’ is first used as a descriptor reflecting the major focus of cases around Cairns and recognising the traditional aboriginal custodians of that locality, the Yirrganydji people (Flecker, 1952). Flecker begins to define the epidemiology of the stingings drawing on his long-term register of North Queensland cases 1960 Barnes details the largest series of Jellyfish stingings, including comparative data from Irukandji cases from around Cairns, reflecting records from Cairns Ambulance Centre and Base Hospital during 1956–1960 (Barnes, 1960) 1961 Small carybdeid jellyfish captured off Cairns beach on 10 December and used in experimental stinging to recapitulate the features of Irukandji syndrome (retrospectively reported in Barnes, 1964) 1967 Carybdeid jellyfish causing Irukandji syndrome described and classified as Carukia barnesi (Southcott, 1967) 1970 Box jellyfish antivenom (CSL Limited) released for clinical use and the first instance of its use for systemic envenoming, by Dr. Jack Barnes, most likely attributable to Irukandji syndrome, recorded by CSL (retrospectively reported in Winkel et al., 2003). Barnes reported the treatment as ‘ineffective’ 1986 First formal published report of Chironex fleckeri box jellyfish antivenom being used for the treatment of two Irukandji syndrome cases—without consistent benefit (Fenner et al., 1986) Irukandji syndrome with severe hypertension hypothesised due to sympathetic overstimulation leading to the suggestion that a- and b-adrenoceptor antagonists be considered for treatment (Fenner et al., 1986) 1987 Acute pulmonary oedema, with left ventricular dysfunction, recognised as a component of Irukandji syndrome (Fenner et al., 1988; Herceg, 1987) Continued 8 Lisa-ann Gershwin et al.

Table 1.1 Timeline of the history of Irukandji research—cont'd Date Event and reference 1988 Jack Barnes’ recognition of the diversity of Irukandji-like carybdeids and the associated variable envenomation syndromes, reported by Barbara Kinsey (1988) 1997 The first report of a sting victim recalling being stung by a small carybdeid jellyfish prior to developing the syndrome (Hadok, 1997) First report of papilloedema and coma associated with Irukandji syndrome (probably due to cerebral oedema) (Fenner and Heazlewood, 1997) 1998 Irukandji-like syndrome presents to Geelong Hospital, Victoria (Cheng et al., 1999) 2000 Carukia barnesi extracts reported as causing massive release of catecholamines in experimental animals (Tibballs et al., 2000) 2001 Life-threatening cardiac failure occurs in a case of Irukandji syndrome (Little et al., 2001) 2002 Possible Irukandji-like syndrome cases reported from Hawaii, USA (Yoshimoto and Yanagihara, 2002) Two fatalities attributed to Irukandji syndrome in Queensland, both due to intracerebral haemorrhage secondary to hypertension (Fenner and Hadok, 2002; Huynh et al., 2003; Pereira et al., 2010) 2003 Huynh et al. (2003) identified nematocyst scrapings from one of the fatalities as not belonging to Carukia barnesi—the first nonanecdotal evidence that other species may cause Irukandji syndrome. These nematocysts were linked by Gershwin (2007) with the new Irukandji species Malo kingi. It was later reargued by Pereira et al. (2010) that the nematocysts originated from hitherto unobserved mature Carukia barnesi. This series reported the variable clinical outcome after Carukia barnesi stingings Irukandji syndrome recognised in Florida, USA (Grady and Burnett, 2003) 2003 Intravenous magnesium infusion first used for treatment of the sympathetic features of Irukandji syndrome (Corkeron, 2003) 2004 Intravenous magnesium reported as being used effectively, in a nonrandomised, unblended case series, to treat the pain and hypertension of Irukandji syndrome (Corkeron et al., 2004) Irukandji syndrome documented in Broome along with ecological conclusions (Macrokanis et al., 2004) Biology and Ecology of Irukandji Jellyfish 9

Table 1.1 Timeline of the history of Irukandji research—cont'd Date Event and reference 2005 New genera and species of Irukandji jellyfish described: Malo maxima, Carukia shinju, Alatina mordens, and Gerongia rifkinae (Gershwin, 2005b,c; Gershwin and Alderslade, 2005) Pharmacological analysis of Carukia barnesi venom extracts confirming release of catecholamines (Ramasamy et al., 2005; Winkel et al., 2005) and modulation of neural sodium channels (Winkel et al., 2005) Probable Irukandji-like syndrome reported from Guadeloupe, Caribbean (Pommier et al., 2005) 2006 Irukandji syndrome described from Thailand (de Pender et al., 2006) 2007 Description of Irukandji species linked with fatal case in 2002: Malo kingi (Gershwin, 2007) 2008 New genus and species of Irukandji described: Morbakka fenneri (Gershwin, 2008) Catecholamine release demonstrated experimentally as being caused by the offshore North Queensland Irukandji Alatina mordens (Winter et al., 2008) 2009 Unpublished Ph.D. thesis reports the establishment of the first Irukandji cDNA libraries (A´ vila-Soria, 2009) First case series of Irukandji syndrome cases reported from the Northern Territory, reflecting data collected from 1990 to 2007 (Nickson et al., 2009) 2010 Irukandji syndrome cases reported from Malaysia (Lippmann et al., 2011) 2011 Experimental studies of Malo maxima venom effects reported, confirming Irukandji-type effects on sympathetic neurotransmitter release (Li et al., 2011) 2012 CSIRO developed a preliminary model for forecasting weather conditions linked to Irukandji infestations (Gershwin et al., 2012) 2013 Anomalous high-latitude clusters of Irukandji stings at Fraser Island, southern Queensland (7 stings), and around Ningaloo Reef, Western Australia (23 stings plus 2 potential fatalities); both clusters occurred with unusual climatic conditions

Adapted and updated from Tibballs et al. (2012). 10 Lisa-ann Gershwin et al.

A 120 200,000 Irukandji Global marine 100 150,000

80 ) WOS ( 60 100,000 y loball

40 g 50,000 20 Cumulative number ofCumulative marine papers 0 0 Cumulative number ofCumulative Irukandji papers (WOS) 1980 1985 1990 1995 2000 2005 2010 Year

BC USA

Biology Medical SOUTH KOREA ENGLAND CHINA CANADA AUSTRALIA BRAZIL

Toxicology

Molecular Pharmacology Figure 1.1 ISI Web of Science search conducted on 30 June 2013. (A) The cumulative number of Irukandji papers (left axis: search term¼IrukandjiþCarukiaþCarybdeidaþ MorbakkaþGerongiaþAlatina) and global marine research (right axis: search term¼ ‘marine’). (B) Research areas addressed in the Irukandji papers, based on ISI WOS categories (classification Medical¼ANESTHESIOLOGYþCARDIOVASCULAR SYSTEM CARDIOLOGYþEMERGENCY MEDICINEþGENERAL INTERNAL MEDICINEþLIFE SCIENCES BIOMEDICINE OTHER TOPICSþNEUROSCIENCES NEUROLOGYþPHYSIOLOGYþPUBLIC ENVIRONMENTAL OCCUPATIONAL HEALTHþRESEARCH EXPERIMENTAL MEDICINEþ SPORT SCIENCES; Biology¼ENVIRONMENTAL SCIENCES ECOLOGYþEVOLUTIONARY BIOLOGYþMARINE FRESHWATER BIOLOGYþZOOLOGY; Toxicology¼TOXICOLOGY; Pharmacology¼PHARMACOLOGY PHARMACY; Molecular¼BIOCHEMISTRY MOLECULAR BIOLOGYþGENETICS HEREDITY). (C) The global distribution of knowledge on Irukandji, based on ISI Web of Knowledge search. (For colour version of this figure, the reader is referred to the online version of this chapter.) Biology and Ecology of Irukandji Jellyfish 11

(Figure 1.1). Since 1980, there have been 119 scientific articles on Irukandji. Based on the cumulative number of papers, there was relatively very little Irukandji research throughout the 1980s and 1990s compared with global marine research (using the search term ‘marine’ as an index). Irukandji research accelerated during the 2000s and is now keeping pace with the ongoing increase in global marine research. We classified the scientific articles on Irukandji into five research areas provided in the Web of Science. Medical is by far the largest research area, followed by toxicology and biology (including taxonomy) (Figure 1.1B). Irukandji research is rare in countries other than Australia and the United States, which recorded 76% and 16% of all papers respectively (Figure 1.1C). Based on an extensive online search of newspapers in Australia, we found 1670 articles about Irukandji since 2001 (Figure 1.2). There has been an increase in newspaper reports about Irukandji over the past 10 years, with considerable interannual variation. The spike in newspaper articles in 2002 followed the death of the two tourists on the Great Barrier Reef. There were relatively few articles about Irukandji in 2011, and this dip corresponded to few stings in that year.

2. BIOLOGY OF IRUKANDJI 2.1. Taxonomy It is suspected that up to 25 species may be able to cause Irukandji syndrome (Table 1.2). The ‘box jellyfish’ species that cause Irukandji syndrome are all in the order Carybdeida and closely resemble the more familiar species in the genus Carybdea, that is, they have unforked pedalia with only a single tentacle attached to each corner of the bell (Figure 1.3). However, not all mem- bers of the order cause Irukandji syndrome, and there are even species that produce the syndrome in other classes of jellyfish. Additional species are likely to exist and may explain regional variations in the syndrome. When originally identified, the first Irukandji jellyfish, namely, Carukia barnesi, was distinguished from the more familiar genus Carybdea based primarily on its unusual tentacles. All cubozoans have their tentacular nematocysts concentrated in numerous transverse bands. In most species, the bands are simple, smoothly rounded, and alike (Figure 1.4C). Some species of cubozoans may have repeating patterns of broader bands alternating with narrower ones, but the bands are still simple and smoothly rounded. In some Irukandji, however, the bands may be decorated. In Carukia barnesi, alternating bands are ‘tailed’ on one side (Figure 1.4A and B). The later-named 12 Lisa-ann Gershwin et al.

120

100

20 Number of newspaper articles

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Year Figure 1.2 Irukandji in the media. The annual number of articles in Australian newspapers since 2001.

Malo kingi has a ‘halo-tentacled’ form, in which each shelf-like band has a ring of nematocysts arranged end on, fanning out like a missile array (Figure 1.4D). The two main families containing species that produce Irukandji syndrome are easily distinguished (Figure 1.5). The family Carukiidae is characterised by the lack of gastric phacellae, that is, the clumps of cirri typically found in cubozoan stomachs. These species also have frown- shaped rhopalial niche openings where both the upper and lower covering scales are undivided, similar to species in the more familiar genus Tamoya, which does not cause illness. In contrast, the family Alatinidae is characterised by having T-shaped rhopalial niche openings, where the lower covering scale is divided at the midline and the gastric cirri are arranged in large crescentic bundles in each corner of the stomach. All Irukandji species in the Carukiidae also possess ‘rhopaliar horns’, peculiar blind canals of unknown function pushing upward from the rhopalial niches. Species in the Alatinidae do not possess these structures. Within these families, the genera can be distinguished using readily observable structural characteristics (Table 1.3); these were treated comprehensively by Gershwin (2005a). Table 1.2 Summary of ecology and syndrome characteristics of species known or believed to cause Irukandji syndrome Mature size Typical Associated Evidence of Main Species (bell height) Habitat Seasonality Swarming depth with salps Syndrome Irukandji Locality refs. Alatina 80 mm Outer reef 8th–10th Yes Surface Unknown Life- Eyewitness Outer Great 1 mordens nights after threatening Barrier Reef, full moon QLD Alatina 85 mm Beach 8th–12th Yes Surface Unknown Moderate Eyewitness Hawaii 2, 3 moseri days after full moon Alatina sp. 1 Unknown Reef Unknown Unknown Unknown Unknown Moderate Expert West Indies, 4–8 statement Puerto Rico, Florida Keys, Guadeloupe, Grand Cayman Alatina sp. 2 500 mm Reef Unknown No Near Unknown Moderate Eyewitness Ningaloo 1 bottom Reef, WA Carukia 9–14 mm Beach Height of Yes Surface Yes Moderate Experimental Tropical, 10, 11 barnesi summer sting QLD Carukia 17 mm Unknown Unknown Unknown Surface Yes Unknown Phylogenetic Broome 12 shinju (caught inference region, WA April) Carybdea 15 mm Beach Height of Yes Surface No Moderate Circumstantial Perth, WA; 13, 14 xaymacana summer Cooktown and Port Douglas, QLD Continued Table 1.2 Summary of ecology and syndrome characteristics of species known or believed to cause Irukandji syndrome—cont'd Mature size Typical Associated Evidence of Main Species (bell height) Habitat Seasonality Swarming depth with salps Syndrome Irukandji Locality refs. Gerongia 60 mm Coastal Year Sometimes Surface Unknown Mild Experimental NT and Gulf 15 rifkinae round, sting of particularly Carpentaria, late QLD summer Malo kingi 30 mm Shelf/ Late Sometimes Not Yes Mild to Nematocyst Primarily 16 coastal summer/ surface severe recovery offshore autumn (lethal) Great Barrier Reef, QLD Malo maxima 50 mm Shelf Late Yes 8–9 m Yes Severe Eyewitness Primarily 12 summer/ offshore autumn Broome, WA Malo 40 mm Unknown Unknown Unknown Unknown Unknown Possible Phylogenetic Philippines 17, 18 philippina fatality inference Malo sp. 20 mm Reefs Unknown Unknown Unknown Unknown Unknown Phylogenetic Montebello 19 inference Islands, WA Morbakka 150 mm Beach Year No Not No Mild to Eyewitness Moreton Bay 20 fenneri round, surface life- to Cape particularly threatening York, QLD autumn Morbakka 150 mm Inland sea Autumn, Unknown Unknown Unknown Unknown Phylogenetic Japan 17, 21 virulenta winter inference Morbakka sp. 65 mm Beach Unknown No Unknown Unknown Mild to Phylogenetic New South 20 severe inference Wales, Australia Morbakka sp. >100 mm Beaches November No Variable Unknown Moderate Phylogenetic Thailand 22, 23 and reefs to February to severe inference and photographs Acromitoides 115 mm Beach All seasons Very Shallows Unknown Mild to Experimental Manila, 24 purpurus (as common in and moderate sting Philippines Catostylus) Manila estuaries Nemopilema Bell to 2 m Coastal June to Often Shallow Unknown Severe Eyewitness China, Japan, 25 nomurai diameter and November occur in Korea and 200 kg oceanic vast blooms Lobonema 50–100 cm In the Summer Blooms Unknown Unknown Moderate Speculative Philippines 26, 27 smithii diameter harbour to severe Physalia sp. Float to Beach No Huge Floats at No Mild to Eyewitness QLD, NSW 28 15 cm apparent armadas surface moderate season Gonionemus 5–15 mm Rocky Mid-June Unknown Among No Moderate Experimental Japan 29 oshoro diameter seashore to end of Sargassum to severe August seaweed Gonionemus Diameter Sandy July to Venomous Shallow No Mild to Eyewitness Russia and 30, 31 vertens typically Zostera August in dense moderate New 2cm beds swarms England Unidentified 10 cm bell, Beach Unknown Unknown Shallow Unknown Severe Eyewitness, North coast 32 tentacles (spring?) light blue of Efate, 20 cm long carybdeid Vanuatu Continued Table 1.2 Summary of ecology and syndrome characteristics of species known or believed to cause Irukandji syndrome—cont'd Mature size Typical Associated Evidence of Main Species (bell height) Habitat Seasonality Swarming depth with salps Syndrome Irukandji Locality refs. At least nine other species are probable but remain almost entirely unknown except for their stings; their disparate locations suggest potential taxonomic distinction when found: a summertime beach sting in Victoria, Australia 33 a cluster of beach stings in North Wales, U.K. 34 stings in Fiji 35 beach stings in Papua New Guinea 36, 37 a reef sting in the Gulf Sea, Qatar 38 numerous stings in Phuket and the Gulf of Thailand that co-occur with salps and onshore breezes 39, 40 beach stings in Malaysia between May and July 41 a reef sting along the mid-east coast of Bali 42 a beach sting in Goa, India 43

Main references are noted numerically in far right column and are summarised in the succeeding text. Australian localities: QLD, Queensland; WA, Western Australia; NT, Northern Territory. References: 1Gershwin (2005c), 2Thomas et al. (2001), 3Yoshimoto and Yanagihara (2002), 4 Kramp (1970), 5Cutress in Williamson et al. (1996), 6 Grady and Burnett (2003), 7 Pommier et al. (2005), 8 Anonymous (2011b), 10 Barnes (1964), 11 Kinsey (1988), 12 Gershwin (2005b), 13 Little et al. (2006), 14 Gershwin (2006b), 15 Gershwin and Alderslade (2005), 16 Gershwin (2007), 17 Bentlage and Lewis (2012), 18 Old (1908, 1912), 19 Gershwin (2005a), 20 Gershwin (2008), 21 Kishinouye (1910), 22 Fenner et al. (2010), 23 Divers Alert Network Asia-Pacific case reports/photos, 24 Light (1921), 25 Mingliang Zhang in Williamson et al. (1996: 215), 26 Mayer (1910), 27 Light (1914), 28 Fenner et al. (1993), 29 Otsuru et al. (1974), 30 Yakovlev and Vaskovsky (1993), 31 Evans (2010), 32DAN Asia-Pacific unpublished records 3 October 2010, 33 Cheng et al. (1999), 34 Lord and Wilks (1918), 35 Flecker (1957a, 1957b), 36 Tyson (1957); 37 Barnes (1964), 38 Salam et al. (2003), 39 Fenner and Lippmann (2009), 40 Fenner et al. (2010), 41 Lippmann et al. (2011), 42DAN Asia-Pacific unpublished records 2003, 43Gershwin unpublished notes 2006. Biology and Ecology of Irukandji Jellyfish 17

Figure 1.3 Species of Australian Irukandji jellyfish. (A) Alatina sp. from Ningaloo Reef, Western Australia (image by P. Baker). (B) Malo kingi from North Queensland. (C) Morbakka fenneri from Central Queensland. (D) Carukia barnesi from North Queensland. (E) Carukia shinju from Broome, Western Australia. (F) Malo maxima from Broome, West- ern Australia (image by M. Alexander, Paspaley Pearling Company). (G) Gerongia rifkinae from Northern Territory. (H) Carukia sp. from the Great Barrier Reef. (I) Alatina mordens from the Outer Great Barrier Reef. (J) Morbakka sp. from New South Wales. Images not otherwise noted are copyright L. Gershwin. (For colour version of this figure, the reader is referred to the online version of this chapter.)

2.1.1 Nematocysts Nematocysts (stinging cells) are essentially a capsule with a harpoon coiled inside and bathed in venom, with a hair trigger on one end. Rapid discharge is accomplished by explosive uncoiling of the harpoon as it everts. When discharged, the nematocyst is clearly divided into three functional portions: the bulbous capsule, the stiff shaft that acts as the penetrative portion, and the long flexible tubule that holds most of the venom. Identification is based on the size and shape of the capsule and the number and position of spines on the shaft (Figure 1.6). Irukandji species have nematocysts of various types and sizes that are useful for species identification and diagnosis of stings (Barnes, 1965; Gershwin, 2006a). One type of nematocyst, the type 4 microbasic mastigophore, 18 Lisa-ann Gershwin et al.

Figure 1.4 Types of tentacles of Irukandji species. (A) Tailed bands of Carukia spp. (B) Close-up of Carukia tailed band. (C) Undecorated bands of Carybdea, Malo, Gerongia, Morbakka, and Alatina. (D) Halo-form bands of Malo kingi. All images copyright L. Gersh- win. (For colour version of this figure, the reader is referred to the online version of this chapter.)

Figure 1.5 Anatomy of carybdeid Irukandji jellyfish. Structures useful for taxonomic distinction are explained in the text. (For colour version of this figure, the reader is referred to the online version of this chapter.) Table 1.3 Comparison of characteristics useful for distinguishing the genera of cubozoans that produce Irukandji syndrome Alatina (12 spp.) Carukia (3 spp.) Gerongia (1 sp.) Malo (4 spp.) Morbakka (4 spp.) Maximum 50 cm 1–2 cm 6 cm 2–5 cm 9–15 cm bell height Bell shape Very tall and Small and Cuboid and robust, with Taller than wide, with flat Taller than wide, with flat slender, with pyramidal, with rounded apex apex apex narrowed flat rounded apex apex Exumbrellar Whitish freckles Red warts Pale freckles Purple freckles Bright pink warts warts Rhopaliar T-shaped Frown-shaped Frown-shaped Frown-shaped Frown-shaped niche ostium Rhopaliar Lacking Narrow, long, Broad, short, curved; Broad, short, curved; Broad, long, straight, horns straight; thread- devil-horn-shaped devil-horn-shaped pointy; rabbit-ear-shaped shaped Number of 6 (2 median, plus 6 (2 median, plus Unknown, possibly 6 2 median lensed eyes 2 median lensed eyes only, eyes per 4 lateral); lower 4 lateral) only, lacking laterals lacking laterals rhopalium lensed eye enormous Pedalial Broadly Narrow Broadly rounded, Narrow Scalpel shape rounded overhanging Pedalial Simple Simple Thorn Knee-shaped Thorn canal bend Tentacles Round in cross Round in cross Round in cross section, Round in cross section, Flat in cross section, section, very section, with heavy, with flared base fine heavy, with flared base fine tailed bands Continued Table 1.3 Comparison of characteristics useful for distinguishing the genera of cubozoans that produce Irukandji syndrome—cont'd Alatina (12 spp.) Carukia (3 spp.) Gerongia (1 sp.) Malo (4 spp.) Morbakka (4 spp.) Gastric Large and Absent Absent Absent Absent phacellae crescentic Mesenteries Extremely Flap-like half Robust, flap-like halfway Flap-like one-third way Robust, flap-like halfway reduced or way; cord-like to to rhopalium, without to rhopalium; cord-like to rhopalium, with fine lacking rhopalium cord-like extension to rhopalium cord-like extension to rhopalium Velarial 3, mostly simple 2, simple or 7, with laminar 1 root, with 3–4 Very complexly canals (per somewhat branching, lacking lateral unbranched fingers, branched; too many to octant) branched, lacking diverticula lacking lateral diverticula easily count, with lateral lateral diverticula diverticula Perradial Lacking lappets Lacking or single 2 rows of 3–6 (typically 5) 2 rows of 1–4 (typically 2) 2 rows of large warts plus lappet warts on one side scattered warts Cnidome Tentacles: Tentacles: egg- or Tentacles: type 4 club- Tentacles: type 4 club- Tentacles: 3 types, type 4 euryteles or lemon-shaped shaped subovate shaped subovate club-shaped microbasic multiple types tumiteles about microbasic microbasic p-mastigophores Bell: spherical 25 mm long p-mastigophores about p-mastigophores about 60–70 mm long and large isorhizas; some Bell: spherical 40–60 mm long 30–50 mm long; oval isorhizas of both tight species also have isorhizas about Bell: spherical isorhizas significant species and loose tubules, both other types 20 mmin about 20–25 mmin differences in shaft 50 mm long Nematocyst diameter diameter spination Bell: 2 types, spherical sizes are quite Bell: spherical isorhizas isorhizas 30 mmin variable about 20–30 mmin diameter and an unusual between species diameter unclassified type with a thin, papillated, oval cuticle 45 mm long

Parenthetical numbers after genera indicate number of combined described and undescribed species known at this time for each group; note that not all species of Alatina are believed to cause Irukandji syndrome, whereas those of the other genera apparently do. Modified from Gershwin and Alderslade (2005). Biology and Ecology of Irukandji Jellyfish 21 appears to be unique to the Carukiidae, though not all species have it (Gershwin, 2006a). Nematocysts are obtained either by skin-scraping or by sticky-tape sample, the latter being more effective, less damaging to the nematocysts and without scarring to the patient (Currie and Wood, 1995). To obtain a sticky-tape sample, gently blot or blow-dry the area, arrange sticky tape over the sting, smooth it once or twice to increase contact, and peel off in one smooth action; sticky-tape samples affixed to a glass slide or folded back on themselves can often be readily identified. In some species with multiple types, the proportions of one nematocyst over another may also be useful and may change through ontogeny; however, the cnidome of Carukia barnesi does not change as the animal grows (Underwood and Seymour, 2007).

2.1.2 Statoliths Another remarkable feature of cubozoans is the statolith, or balance stone. Each rhopalium contains one large, solid statolith below the cluster of eyes. The statolith grows by accretion of daily growth rings (Ueno et al., 1995), similar to the otoliths of fish or like the annual growth rings of trees; these

Figure 1.6 Irukandji nematocysts. (A) Carukia shinju, discharged. (B) Malo maxima, undischarged. (C) Carukia shinju, undischarged. (D) Malo kingi, discharged. (E) Carukia shinju, bell nematocysts. All images copyright L. Gershwin. (For colour version of this figure, the reader is referred to the online version of this chapter.) 22 Lisa-ann Gershwin et al. daily growth rings can be counted for ageing the animal (Gordon et al., 2004; Kawamura et al., 2003; Ueno et al., 1997), and chemical signatures in the statolith rings may some day be analysed to identify where and when the medusa has spent time. Statolith shape is genus-specific, and its use as a taxonomic indicator was reviewed by Gershwin (2005a). Because statoliths are the only hard part in an otherwise soft body, they may be useful for identification of fragmentary specimens, ethanol-preserved or frozen specimens, or even possibly fossil material.

2.2. Evolution 2.2.1 Phylogeny Both morphological and partial 18S genetic evidence suggest that the Irukandji species that lack phacellae and have frown-shaped rhopaliar niche ostia form a monophyletic group (Figure 1.7): this includes the genera Car- ukia, Malo, Gerongia, and Morbakka (Gershwin, 2005a). However, the other Irukandji genus Alatina, which has large crescentic phacellae and T-shaped rhopaliar niche ostia without horns, appears to be only distantly related. These patterns were corroborated by later work using both nuclear and mitochondrial genes (Bentlage et al., 2010). Intriguingly, although species in the genus Alatina were long considered to fall within the genus Carybdea until separated into their own family by Gershwin (2005c), recent genetic analyses suggest that the Alatinidae is in fact ancestral to other single-tentacled cubozoans and only a distant cousin of Carybdea (Bentlage et al., 2010). Identifying which species cause Irukandji syndrome is a compelling quest, and phylogeny provides interesting argument in at least two cases. First, one species in the genus Carybdea, the Australian form of Carybdea xaymacana, has been assumed to cause illness (Kingsford et al., 2012; Little et al., 2006), but this view has been challenged on phylogenetic grounds and remains speculative (Gershwin, 2006b). No other species in the genus are linked with the illness, and vast numbers of non-systemic stings in regions where these species are common would appear to falsify the hypothesis. Second, at least three species in the genus Alatina are credibly linked with the syndrome through eyewitness sting events: the Hawaiian Alatina moseri (Thomas et al., 2001), the east Australian Alatina mordens (Gershwin, 2005c), the west Australian A.sp.1 (Gershwin, 2005c), and the Caribbean A. sp. 2 (Grady and Burnett, 2003; Kramp, 1970). Whether other species in the genus Alatina cause the syndrome is unknown, but seems likely given the known distributions of the species and Biology and Ecology of Irukandji Jellyfish 23

Figure 1.7 Unrooted phylogeny of Cubozoa (from Gershwin, 2005a). Asterisks denote species known to cause Irukandji syndrome. From top centre: Chirodectes maculatus (image by R. Hore), Chironex fleckeri, Chiropsella bart, Chiropsella bronzie, Chiropsalmus quadrumanus (image by A. Migotto), Carybdea branchi, Carybdea rastonii (image by I. Bennett/Australian Museum), Carybdea xaymacana, Tripedalia cystophora, Copula sivickisi, Carukia barnesi, Gerongia rifkinae, Morbakka fenneri (image by G. Cranich/ Queensland Museum), Malo maxima (image by M. Alexander/Paspaley Pearling Com- pany), Malo kingi, Alatina rainensis, Alatina sp. (image by P. Baker/Western Australian Museum), and Alatina mordens. Uncredited images by L. Gershwin. (For colour version of this figure, the reader is referred to the online version of this chapter.) stings. These interesting cases underscore the utility of phylogenetic bases for prediction of unknown features and for testing of hypotheses.

2.2.2 Fossil evidence The evolutionary age of Irukandji is unclear. The oldest undisputed fossil cubozoan is the chirodropid Anthracomedusa turnbulli from the Middle Penn- sylvanian (ca. 300 mya), near Essex, Illinois (Johnson and Richardson, 1966, 1968). The Essex fauna of the Mazon Creek formation is found in the Francis Creek Shale, a member of the Liverpool cyclothem of the Car- bondale group. These spectacularly preserved specimens contain all the structures one would hope to find in a fossil chirodropid: cuboidal body 24 Lisa-ann Gershwin et al. form, many tentacles arising from a pedalium at each of the four lower corners of the bell, and a simple margin. The notable feature of Anthracomedusa is that it is a fully formed chirodropid, that is to say, if it were found alive today, it would be unlikely to raise eyebrows as there is nothing particularly ‘primitive’ about it. Therefore, it seems plausible that the sister group to the Chirodropida, namely, the Carybdeida, branched off well before Anthracomedusa was fossilised. The oldest apparent carybdeids are Bipedalia cerinensis and Paracarybdea lit- hographica from the Cerin Lagersta¨tte (Late Kimmeridgian of the Upper Jurassic, ca. 150 mya) near Ain, Eastern France (Gaillard et al., 2006). Noth- ing in their morphology suggests an Irukandji affinity of any sort, and numerous aspects of their morphology draw even their cubozoan affinity into question. To date, there is no fossil evidence identifiable as Irukandji jellyfish.

2.3. Reproduction and life cycle Cubozoans with life cycles that have been resolved have a complex life history consisting of a primary benthic sedentary polyp, a secondary creeping polyp, complete or near-complete metamorphosis into a juvenile medusa, and dioecious adult medusae (Arneson, 1976; Arneson and Cutress, 1976; Cutress and Studebaker, 1973; Hartwick, 1991a,b; Stangl et al., 2002; Straehler-Pohl and Jarms, 2005, 2011; Studebaker, 1972; Werner et al., 1971; Yamaguchi, 1982; Yamaguchi and Hartwick, 1980). Life cycle and growth data for Irukandji jellyfish are lacking for most species, and, even for Carukia barnesi, there is only sketchy knowledge. Early development of the Caribbean Alatina sp., which is believed to produce Irukandji syndrome, is similar to that of other cubozoans (Arneson, 1976; Arneson and Cutress, 1976). No life cycle or growth data have been published for carukiid species. A detailed account of observations made while rearing Carukia barnesi is given in Appendix A. A summary of cubozoan early life history is provided in Figure 1.8 and Table 1.4; it seems likely that Irukandji species fall somewhere along this spectrum.

2.3.1 Where do they breed? Irukandji typically coincide with periods of sustained onshore breeze, so it is often assumed that they are blown in from offshore (Barnes, 1964; Kinsey, 1988). And because their stings occur on reefs and islands across the shelf, it seems logical that they must be living and breeding well offshore. However, Biology and Ecology of Irukandji Jellyfish 25

Figure 1.8 Hypothesised Carukiidae life cycle. The planula larva and polyp metamorphosis are unknown but are likely to be similar to those of other cubozoans. The polyp pictured here is believed to be that of Carukia barnesi but remains unconfirmed. Images by L. Gershwin and Heather Walling. (For colour version of this figure, the reader is referred to the online version of this chapter.)

collection data appear to falsify this hypothesis, at least in part. Sampling in the Cairns region during coastal infestations has yielded a range of sizes and maturity stages from 2 to 14 mm (Appendix A), suggesting that there are more local polyp nursery areas. Barnes (1964) speculated that Irukandji might be breeding around a small nearshore island or possibly that the islands facilitated retention of drifting individuals, based on the high incidence of stings on the facing coast. While there has been at least one claim of discovering such a breeding ground (Anonymous, 2005), subsequent sampling has been unable to con- firm its existence and the ‘breeding at the island’ hypothesis remains speculative. Kingsford and his colleagues (2012) noted that the greatest numbers of Irukandji were found near granite islands, again suggesting that this may be where they bred or that retention is facilitated by oceanographic and wind Table 1.4 Comparison of the polyp and young medusa characters of the taxa for which the life cycle is known Carybdea Carybdea xaymacana (as mora (as Carybdea morandinii Carybdea Carybdea (possibly¼Carybdea Tripedalia Copula Chironex Carukia marsupialis) rastonii) sivickisi) cystophora sivickisi Alatina sp. fleckeri barnesi 1–4 5 6 7, 8 9 10, 11 12, 13 14 Polyp Polyp height 1.5 mm 2 mm 1.8 mm 1.4 mm ? 2 mm 1.2 mm 1mm No. of 24 ? 16 7–9 ? 16 40–45 To 18 tentacles Nematocysts Single large Multiple? Single large stenotele Numerous Single 2–3 small Single Single large at tip of pseudostenotele bean-shaped large euryteles large stenotele tentacles euryteles stenotele replaced by stenotele around single single large stenotele stenotele Tentacles Solid Solid Solid Solid Solid Solid Solid Solid Juv. medusa Umbilical absorbed ? Within 2 days ? ? 10–13 h ? Within cord within a few (Arneson 2–12 h of hours (Stangl) says 10–30) capture Tentacles ? ? Hollow, brown/ ? ? Hollow Hollow Hollow white No. of 2 at release ? 4 4 at release, ? 4 at release 4 at release 4 with tentacles þ8 next day polyp remnant Nematocysts Adradial rows ? Irregular warts ? ? Spherical Vertical Numerous of 4 large warts holotrichs, rows on haphazard microbasic exumbrella warts euryteles Colour Brown ? Clear, with ? ? Ochre Ochre Red zooxanthellae

Main references are noted numerically after species and are summarised in the succeeding text. References: 1 Studebaker (1972), 2 Cutress and Studebaker (1973), 3 Stangl et al. (2002), 4 Straehler-Pohl and Jarms (2005), 5 Okada (1927), 6 Straehler-Pohl and Jarms (2011), 7 Werner et al. (1971), 8 Werner (1983), 9 Hartwick (1991a), 10 Arneson (1976), 11 Arneson and Cutress (1976), 12 Yamaguchi and Hartwick (1980), 13 Hartwick (1991b), Appendix A. 28 Lisa-ann Gershwin et al. effects. In fact, most of the apparent island hot-spots in the Great Barrier Reef region occur off the southwest portion of islands or rocky outcrops, which are in the lee of the easterly winds (Figure 1.16). Similarly, Alatina moseri in Hawaii aggregates on the leeward side of the island. These observations would appear to favour the retention hypothesis, but nonetheless do not rule out the possibility that these leeward habitats are more favourable for polyps and therefore also act as breeding grounds. Whether these patterns are true for Irukandji at other localities has not been tested.

2.3.2 Metamorphosis induction Triggers for natural onset of metamorphosis from polyp to medusa are poorly understood and are likely to differ among species. In captivity, metamorphosis is reliable for some cubozoans and unpredictable for others. For example, in the non-Irukandji Caribbean Carybdea xaymacana (often erroneously called Carybdea marsupialis), metamorphosis is reliably induced by shifting the temperature from 20 to 28 C (Stangl et al., 2002), but could not be induced in Carybdea morandinii (Straehler-Pohl and Jarms, 2011). Statolith ring analyses have shown that natural metamorphosis corresponds to semilunar cycles (Ueno et al., 1997). In particular, statolith formation commences in the last stage of medusa formation near the time of liberation. In wild-caught medusae, initiation of statolith formation was back-calculated to occur within 2 days before or after a spring tide in 86% of the statoliths studied. Moreover, one or more of the daily growth rings are darker, better developed, and more conspicuous than others and often occur at 2-week intervals; 70% of these darker rings correspond with a period 3 days before or after neap tide. These darker rings were inferred to potentially correspond to spawning cycles. While metamorphosis has not yet been observed in carukiid Irukandji, in captive Caribbean Alatina sp., metamorphosis commenced spontaneously within 2–3 days after the polyp reached the definitive 16-tentacle stage (Arneson and Cutress, 1976).

2.3.3 Ontogenetic changes in toxicity The larger, more virulent multi-tentacled box jellyfish Chironex fleckeri changes its cnidome ratio as it grows to become more toxic (Carrette et al., 2002; Oba et al., 2004), but it is unclear whether the same is true for Irukandji. Media claims of a similar process with Carukia barnesi (Bateman, 2010b) are unsubstantiated (Underwood and Seymour, 2007). Biology and Ecology of Irukandji Jellyfish 29

In one recent study, gel electrophoresis of specimens pooled into immature and mature size classes found distinct differences in protein banding of tentacular venom proteins (Underwood and Seymour, 2007). The actual potency of the venom at different stages was not assessed, but ontogenetic shifting seems likely, given that the venom profiles correlate strongly with a change in prey type as the animal grows. The study further found that bell venom had a different protein profile than tentacular venom in mature animals; ontogeny of bell venom was not studied. The Hawaiian Irukandji species Alatina moseri presents an interesting question. Vast numbers occur periodically in inshore waters (Thomas et al., 2001), but it appears that only a small percentage actually cause Irukandji syndrome (Yoshimoto and Yanagihara, 2002). Whether this is due to imma- turity of specimens or some other factor is unknown. However, a similarly low percentage of illness relative to stings is anecdotally observed for the Australian Alatina mordens (R. Hore, unpublished data) and for the Japanese Irukandji syndrome-producing hydromedusa Gonionemus oshoro (Otsuru et al., 1974). It also therefore seems plausible that many more carukiid Irukandji stings not resulting in illness occur than are recorded. Irukandji polyps possess completely different cnidomes than their medusa counterparts (Gershwin, 2006a). Whether the polyps also cause Irukandji syndrome is unknown but seems unlikely.

2.3.4 Lifespan and natural mortality Cubozoans are inherently difficult to monitor in nature or raise in captivity; therefore, little reliable information on their lifespan is available. In general, most jellyfish are assumed to die at the end of the summer. However, Carybdea rastonii in southern Australia, the California Carybdea, and Alatina spp. are found throughout the year, suggesting that such a die-off does not always occur (Gershwin, 2005c; Matsumoto, 1995; Thomas et al., 2001). Only scant evidence is available regarding the lifespan and senescence of Carukia barnesi. In the summer of 2003–2004, one of us (LG) raised wild- caught specimens in a laboratory. Small, young specimens generally took about 2 weeks to grow to sexual maturity (LG, unpublished data). In the laboratory, medusae raised from young began to degrade within a few days of reaching full size and gonad maturity. Senescence was observed as ulcers on the bell, loss of tentacles, and refusal to take food. Whether this fast growth and short lifespan is typical of natural cycles is unconfirmed. Larger Irukandji species have proven more difficult to keep in captivity; no information is available about their longevity. 30 Lisa-ann Gershwin et al.

Statolith studies on other species demonstrate the presence of daily growth rings (Gordon and Seymour, 2012; Gordon et al., 2004; Kawamura et al., 2003; Ueno et al., 1995, 1997). These rings have been used to infer the age of the medusae at the time of capture, but have not yet been used to show or estimate maximum ages.

2.4. Eyes and vision Cubozoans have long been a source of intrigue, possessing well-developed eyes but lacking a comparably complex brain. The literature indicates that they are highly visual and capable of sophisticated behaviours and some form of decision making (Table 1.5). Therefore, no discourse on their biology and ecology would be complete without discussion of their vision and eyes. Per- haps more so than for any other feature, most of this information is only available for non-Irukandji cubozoans; however, by understanding the range of visual apparatus and function, hypotheses can be developed to better understand where Irukandji species are likely to fall on this spectrum.

2.4.1 Physical properties of the eyes Cubozoans are extraordinary among the jellyfish, and indeed among most invertebrates, in having complex eyes. Whether these complex eyes have the physical capacity to resolve images has been debated for many years. Arguments supporting image formation are generally based on observed behaviour and are expanded in the succeeding text, whereas some arguments examining the physical properties of the eyes conclude that they are unlikely to form images. That cubozoans use their vision to navigate their environment and find prey and mates is clear; what is less clear is whether the eyes form images in ways that we do not yet understand or whether the animals simply ‘make do’ with blurry or indistinct pictures. Generally all cubozoans have 24 eyes clustered into four groups on each rhopalium (Figure 1.9). Along the midline of the rhopalium are two complex eyes, each with a lens, retina, and cornea. Along the sides of the lensed eyes are two pairs of simple pigment-cup ocelli: one pair of slit eyes between the two lensed eyes and one pair of pit eyes next to the upper lensed eyes. Some Irukandji, however, have modified eyes. Species in the genus Malo have only the median complex eyes and lack the lateral eye spots (Gershwin, 2005b), and in several species of Alatina, the distal lensed eye is greatly enlarged (Gershwin, 2005c). The four types of eyes have different structural features and different func- tions (Nilsson et al., 2005; O’Connor et al., 2009; Yamasu and Yoshida, 1976). Biology and Ecology of Irukandji Jellyfish 31

Table 1.5 Visual capabilities of cubozoans Visual capability Details References Phototaxis Strongly attracted to light 1–9 Colour perception Blue, green, and UV-sensing opsins suggest 8, 10–12 colour perception, as do behavioural studies; colour blindness has also been argued Obstacle avoidance Clearly and consistently moves away from 4, 5, 10, dark objects and toward light-coloured 13–19 objects, possibly using colour or contrast Terrestrial navigation Use visually guided cues above the water to 20, 21 manoeuvre to, around, and through complex habitats Sexual dimorphism Dark spots develop on the female velarium 14, 22, 23 and mate recognition when she is ready to mate, thought to offer a visual signal to males Courtship and Sophisticated mating behaviours appear to be 5, 14, 22, copulation visually driven 24

References: 1 Barnes (1966), 2 Studebaker (1972), 3 Arneson and Cutress (1976), 4 Matsumoto (1995), 5 Stewart (1996), 6 Gershwin (2005b), 7 Gershwin (2005c), 8 Gershwin and Dawes (2008), 9 Kingsford et al. (2012), 10 Martin (2004), 11 Coates et al. (2006), 12 Garm et al. (2007a), 13 Barnes in Kinsey (1986), 14 Hartwick (1991a), 15 Hamner et al. (1995), 16 Stewart (1997), 17 Ueno et al. (2000), 18 Buskey (2003), 19 Garm et al. (2007b), 20 Garm et al. (2011), 21 Garm et al. (2012), 22 Lewis and Long (2005), 23 Lewis et al. (2008), 24 Werner (1973).

The two median eyes are camera-type eyes, each with a spherical or ellipsoid, cellular fisheye-like lens, a retina, and cornea. There are about 11,000 sensory cells in the cubozoan eye (Pearse and Pearse, 1978). The retina is composed of four layers: a sensory layer, a pigmented layer, a nuclear layer, and a layer of nerve fibres (Berger, 1900; Pearse and Pearse, 1978). The lens is separated from the retina by a thin cellular space. The pigment layer covering the outside of the retina forms an iris around the lens. The pupil of the lower eye can respond to changing light intensity by changing the aperture in less than a minute; however, the pupil of the upper eye is immobile. The upper lensed eye is orientated straight up regardless of the position of the jellyfish and has been demonstrated to be used in terrestrial navigation by the non-Irukandji mangrove-inhabiting Tripedalia, as described in the succeeding text. This eye has a nearly circular field of view with a width of 95–100, closely matching Snell’s window (the 97 circular visual field through which the entire 180 of the terrestrial world is visible to an 32 Lisa-ann Gershwin et al.

Figure 1.9 Cubozoan eyes. (A) Anatomy of typical cubozoan eye (from Chiropsalmus). (B) Malo: note the lack of lateral pit and slit eyes. (C) Alatina: note the greatly enlarged lower lensed eye. All images copyright L. Gershwin. (For colour version of this figure, the reader is referred to the online version of this chapter.)

underwater observer, compressed by refraction as light passes through the air/water interface) (Garm et al., 2011). The lower main lensed eye is orientated obliquely downward, with a much broader visual field than the upper eye. The lateral pit and slit eyes lack lenses and have different structural properties (Berger, 1900; Garm et al., 2008; Laska and Hu¨ndgen, 1982; Martin, 2004; Satterlie, 2002). The pit eyes have only a single-cell type, namely, pigmented photoreceptors, and are thought to function only as light metres without any spatial resolution. The slit eyes are made of four cell types, including a canoe-shaped group of vitreous cells forming a lens-like structure over the retina; this vitreous group has a lobed surface and it is thought that it may act as a UV filter. The slit eyes appear to have the potential for spatial resolution and most likely detect vertical movement, but this is not well understood. The pit eyes and upper lensed eyes point directly upward, whereas the slit eyes and main lower lensed eyes point obliquely downward; the heavy crys- talline statolith keeps them orientated vertically, even when the animal is upside down (Garm et al., 2011). In this way, some eyes are orientated for looking up through the water surface at celestial or terrestrial cues, and other eyes are orientated for looking downward at underwater structures and shadows. Biology and Ecology of Irukandji Jellyfish 33

2.4.2 Visual ecology The anatomy and histology of cubozoan eyes were studied in detail more than a century ago by Conant (Berger, 1900; Conant, 1898), who speculated that the eyes might ‘see’. Since that time, numerous studies have added to our knowledge but not to our understanding. Many studies have convinc- ingly described complex visual behaviours; others have argued why they cannot be so based on the animals’ visual hardware. We are left with the odd impression that cubozoans see, but without the physical basis to do so. Experimental studies indicate that the complex eyes can form images and that cubozoans are able to sense various stimuli in their surroundings such as shapes, shades, and colours of light and react to them, often predictably. Moreover, sophisticated behaviours such as hunting, evasion, navigation, courtship, and copulation appear to involve sight and some manner of cog- nitive processing. As noted by Martin (2004), cubozoans are commonly found in nearshore habitats such as sandy beaches, kelp forests, mangrove thickets, and coral reefs, and they use their vision to navigate these tricky habitats, a particularly important survival strategy for soft-bodied animals easily damaged by crashing waves and collisions with barriers. Coates (2003) provides a good literature review on the visual ecology and relevant functional morphology of cubozoans. It has been proposed that different tasks are associated with different eye types (Garm et al., 2008). Garm and his colleagues (2012) erroneously asserted that all cubozoans have the same four eye types and that their visual system varies only marginally. In fact, Irukandji are the exception to the rule, with several species having unusual modifications to their visual apparatus. For example, Malo lacks lateral eyes (Gershwin, 2007), which are thought to aid the lensed eyes in peripheral filtering of information in other species. It is unclear therefore whether Malo simply has less visual ability or has somehow overcome the need for lateral eyes. Moreover, in Alatina, the lower lensed eye is characteristically about twice the size of the corresponding eye of other species and medusae are typically observed to be active at night (Gershwin, 2005c). It seems plausible that their large eyes are an adaptation to a noctur- nal lifestyle, for example, sensing bioluminescence and possibly lunar syn- chronisation of their monthly spawning aggregations.

2.4.3 Visual evolution The Cubozoa offer insight into the early evolution of vision. While the slit and pit eyes may provide visual information that the lensed eyes cannot, the lensed eyes nonetheless receive far more light and provide better spatial perception 34 Lisa-ann Gershwin et al. than the slit and pit eyes (Garm et al., 2008). And even though apparently out of focus, blurry images are better than no images (Nilsson et al., 2005). Garm and his colleagues (2011) proposed that having different eye types specialised for different visual tasks might require less neural processing than if the information for multiple behaviours were to pass through one eye. Cubozoan behaviours (discussed in the succeeding text) suggest that at least some species are able to perceive colour, suggesting that the Cubozoa might represent early development of colour vision. O’Connor and her colleagues (2010) thought that colour vision can eliminate the brightness noise of flickering from surface ripple. Whether colour vision has allowed these animals to move into flickering coastal habitats, or whether living in coastal habitats selected for improved visual perception, is unclear.

2.5. Behaviour 2.5.1 Phototaxis The propensity of cubozoans to be attracted to light has been noted many times. For example, Barnes (1966) reported that in full daylight, surface and sub-surface light intensities seem to have little effect, but in semidarkness, these animals “are very markedly phototaxic. The light of a match is detected at distances up to 5 ft and ...[they] show a remarkable accuracy in turning towards the light source, even though the latter be extinguished before the turning movement is completed” (p. 322). Although we have only a chequered understanding of the biology and ecology of most Irukandji species, the one thing that seems consistent is that they are easily caught by light attraction. For example, Barnes reported that ‘pseudo-Irukandji’ in Queensland (later-named Malo kingi) were ‘irresistibly attracted’ to a submerged car headlight light held at the water surface (Kinsey, 1988). Similarly, Gershwin (2005b) collected two new species of Irukandji (Carukia shinju and Malo maxima) by attracting them to powerful lights at the back of a ship. Gershwin (2005c) noted that Alatina mordens is often encountered by scuba divers at night, where the divers swim up into the light halo at the back of dive boats where the jellyfish are swarming; so too, one of us (RH) has conducted monthly research on this species since 2002, using above-water lights to attract them (see Appendix B). This hardwired attraction of cubozoans to light has potentially strong implications for safe management of night-time activities where lights shine into the water for prolonged periods, including recreational scuba diving, port and marina facilities, fishing and night-snorkelling jetties, marine con- struction work, and tourism resorts. Biology and Ecology of Irukandji Jellyfish 35

3. ECOLOGY OF IRUKANDJI

We know surprisingly little about the ecological patterns of Irukandji species in general. But the scattered information we do have suggests that while some species have similar ecologies, others are quite different, particularly at the levels of genus and family. Species differ not only in bloom strategies, but also in seasonality and cross-shelf distribution.

3.1. Diet and feeding The natural prey preferences of Carukia barnesi were recently quantified for the first time (Underwood and Seymour, 2007). Stomach contents of 37 individuals of four size classes were examined: <4 mm, 4–6 mm, 6–8 mm, and >8 mm (interpedalial distance1). The authors found a significant difference in the relative proportion of crustaceans versus larval fish in stomachs of medusae of different sizes, with a general trend toward an increasing proportion of larval fish with increasing predator body size. 100% of the smallest size class of medusae had only crustacean prey, whereas 100% of the largest size class had only larval fish. The number of medusae in each category was not specified nor was the percentage of individuals without prey in the gut. A similar ontogenetic shift in diet was noted for the larger multitentacled Chironex fleckeri, but not for the smaller multitentacled Chiropsella bronzie (as Chiropsalmus sp.) (Carrette et al., 2002). Therefore, such a shift does not appear to be universal in the Cubozoa, but may well correlate with toxic species or at least with their toxicity. For Chironex and Carukia, the shift in prey type is accompanied by morphological change. Chironex changes its cnidome ratio, whereas Carukia develops peculiar banding on its tentacles (Figure 1.4). Barnes noted that tentacles of Carukia, which he described as like a ‘cobweb with dewdrops on it’, are invisible in water except for the tailed bands (Kinsey, 1988). Two experimental prey-capture events have been observed: in each case, the larval fish was attached head-on to one of the ‘tails’ of the modified nematocyst bands, leading the authors to conclude that the beaded effect of the bands and the jerking motion of the extended tentacles attracted larval fish by mimicking the movements of their prey (Underwood and Seymour,

1 Interpedalial distance is generally not used because it is like measuring the distance between elbows; bell height or interrhopalial distances are considered more reliable. 36 Lisa-ann Gershwin et al.

2007). These authors further remarked on the ecological advantage of minimising energy expenditure in a small oceanic species that is able to attract and envenom highly mobile larval fish. Although it may be tempting to fascinate on the ‘tailed’ bands as a sophisticated means of lure and capture, it must be borne in mind that juvenile Carukia do not have the ‘tailed’ bands nor do other Irukandji species (Figure 1.4). Based on the energetic needs of adult medusae and the ontogenetic diet patterns of cubozoan species, it seems likely that adult Irukandji of other genera also prey on fish, albeit apparently without lures. A particularly interesting question lies in the halo-like bands of some specimens of Malo kingi, that is, whether this is an unrecognised species difference or another example of ontogenetic shift. Very young juveniles, in contrast, have not yet developed their tentacles and appear to rely on the bell for food capture. Newly metamorphosed Alatina medusae (inaccurately identified as the Atlantic Carybdea alata) were observed to envenom prey with the bell nematocysts and to then pass the prey across the bell warts to the manubrium on the underside of the bell (Underwood and Seymour, 2007). Additional notes on the feeding behaviour of laboratory-reared Carukia barnesi are given in Appendix A.

3.2. Geographic distribution Determining the distribution of Irukandji species and Irukandji syndrome is muddled by the lack of resolution in syndrome variation and lack of data from specimen studies. We are currently presented with a riddle consisting of sting events that we do not know how they correspond with species, and species that we do not know how they correspond with stings, and some species known to cause Irukandji syndrome that are not even in the Cubozoa. Clearly, much more work needs to be done on species identification and elucidation of species-syndrome linkages.

3.2.1 Global distribution While Irukandji jellyfish are often associated with tropical Australia, the numerous substantiated reports from far reaches of the globe make it clear that Irukandji syndrome-producing jellyfish occur throughout the oceans and seas of the world from at least 53 Nto38 S (Figure 1.10) and have done so for many decades. In most cases outside tropical Australia, species are not yet identified. There are large numbers of Irukandji stings on the Great Barrier Reef (e.g. Cairns and Whitsundays), Australia’s North West Shelf (e.g. Broome Biology and Ecology of Irukandji Jellyfish 37

Figure 1.10 Worldwide Irukandji sting distribution. Size of circles qualitatively indicates relative numbers of stings. Only two fatalities have been confirmed, with four others unresolved. Irukandji stings usually leave no mark and nothing to test postmortem, so it is widely believed that additional fatalities have occurred. (For colour version of this figure, the reader is referred to the online version of this chapter.) and Exmouth), Hawaii (e.g. Waikiki Beach), Thailand (e.g. Phuket), Malaysia (e.g. Langkawi), and the Caribbean (e.g. Stingray City and Florida Keys) (Anonymous, 2011b; Fenner and Harrison, 2000; Fenner et al., 2010; Grady and Burnett, 2003; Kinsey, 1988; Le May, 2013; Lippmann et al., 2011; Macrokanis et al., 2004; Thomas et al., 2001). Versions of Irukandji syndrome have been reported from many islands throughout the Pacific (Table 1.2), including Fiji (Flecker, 1957a,b) and Papua New Guinea (Barnes, 1964; Tyson, 1957), as well as Vanuatu, Tahiti, Samoa, and New Caledonia (Williamson et al., 1996). These islands coincide with the distribution pattern of Alatina (Gershwin, 2005c; Kramp, 1961), and it is possible that many of these stings are attributable to species in this genus. In Australian waters where Irukandji species are the most well known, it is evident that species often have localised distributions (Figure 1.11). It therefore seems probable that many more species of Irukandji jellyfish remain to be discovered around the world. In some cases, Irukandji syndrome is believed to result from species other than cubozoans (Table 1.2). For example, symptoms consistent with Irukandji syndrome have been reported from the Chinese giant rhizostome jellyfish Nemopilema nomurai (Mingliang Zhang in Williamson et al., 1996, 38 Lisa-ann Gershwin et al.

Figure 1.11 Irukandji species distribution in Australia: confirmed localities are indicated by coloured dots. Data were gathered from original descriptions of these species and museum specimens around Australia. (For colour version of this figure, the reader is referred to the online version of this chapter.) p. 215). So too, Irukandji syndrome in Japan has been experimentally demonstrated to result from the sting of the hydromedusa Gonionemus oshoro (Otsuru et al., 1974). Curiously, the closely related and widely distributed Gonionemus vertens is not known to be toxic throughout most of the world, but gives a version of Irukandji syndrome when in high densities in Russia and Cape Cod, Massachusetts (Evans, 2010; Yakovlev and Vaskovsky, 1993). In the Philippines, at least two different common species have been blamed for systemic syndromes (Chrysaora quinquecirrha and Lobonema smithii) (Light, 1914; Mayer, 1910); it is probable that these stings are more accu- rately attributable to less visible Malo or Morbakka or some other species of Irukandji.

3.2.2 Cross-shelf distribution For many years, it was widely believed that Australian Irukandji were only a problem on the coast and not on reefs and islands. These notions have been cat- egorically disproven. In fact, the general pattern appears to be that as one travels further offshore the virulence of the Irukandji species found increases, as does distance from medical care (Gershwin, 2005b; Gershwin, 2007). The majority of Irukandji specimens caught along beaches, in approximately waist-deep water, has been Carukia barnesi. Similarly, the majority of Biology and Ecology of Irukandji Jellyfish 39 coastal stings match the pattern for classic Irukandji syndrome, with 20–30 min onset, pain subsiding with morphine, and low or no hypertension or pulmonary oedema. Therefore, from both sting and specimen data, it would appear that the dominant Irukandji along Queensland coastlines is Carukia barnesi. A study of trends among 62 stings over 1 year in the Cairns region found the following: 47 (76%) patients were stung at coastal locations, 7 (11%) were stung on the reef, and 5 (8%) were stung at the nearby islands (Little and Mulcahy, 1998). Of the 34 patients seen in December 1996, 30 (88%) were stung at coastal locations, compared with 17 of 26 (65%) for the period from January to May. Thirty-nine patients (63%) were stung while swimming inside stinger net enclosures on the beaches. As with earlier studies by Barnes (Kinsey, 1988), they found that the most frequent location to be stung was Palm Cove (17/62; 27%), a beach about 25 km north of Cairns. Interestingly, however, stings over the last decade in the Cairns region have shifted to being more common offshore on the reefs and islands compared to the beaches, steadily trending from 88% beach stings in 2001 to 75% offshore stings in 2007 (Sando et al., 2010). The reasons for this shift are not yet clear and beg further investigation. Perhaps the most obvious explana- tion is that management has improved at the beaches but the offshore regions have yet to follow. This is an attractive hypothesis given the amount of effort that lifeguards put into safety management but, if true, would call for significant action by reef and island operators. Another possibility may be greater use of offshore regions compared to beaches, as the reefs and islands have become more accessible and desirable tourism destinations. Still another possibility is that the jellyfish may have actually shifted their centre of distribution. Occasionally, other genera such as Malo and Morbakka are also found in coastal areas (Gershwin, 2007, 2008). In Central Queensland waters, particularly in the Mackay/Sarina region, Morbakka is most often taken either beached or in tidal fishnets. In northern and southern Queensland waters, Morbakka is most often found swimming in marinas or shallow bays, particularly in the Port Douglas (far north) and Redcliffe (far south) regions. Why these regions are particularly favoured by these species is not known, but the patterns are so predictable as to be worth further investigation. Stings are common in midshelf waters throughout the Great Barrier Reef (Fenner and Carney, 2001; Little and Mulcahy, 1998), but specimens are rarely taken, most probably due to lack of sampling. Midshelf stings are most 40 Lisa-ann Gershwin et al. frequently of the Carukia barnesi type, that is, slow syndrome onset and low incidence of hypertension or life-threatening complications. Some midshelf stings are notably more severe, particularly in the popular dive region of the SS Yongala wreck off Townsville (about 450 km south of Cairns) and on the midshelf islands and reefs. These more severe syndromes often onset rapidly (e.g. 5 min), the pain is severe and unresponsive to opi- ates, and hypertension may be severe, leading to pulmonary oedema (Fenner and Carney, 2001; Fenner and Hadok, 2002; Fenner and Lewin, 2003; Fenner et al., 1988). The culprit species responsible for these severe midshelf stings are not well studied but do not appear to be attributable to Carukia. Scant evidence implicates Malo and Morbakka (Gershwin, 2007, 2008; Huynh et al., 2003; Little et al., 2001, 2003). Further offshore, for example, outer reef and Coral Sea localities, the dominant Irukandji appears to be Alatina mordens (Gershwin, 2005c; Appendix B); however, a case may be made that it is the easiest one to see. Most specimens of Alatina mordens have been collected from dive sites such as Moore Reef and Osprey Reef. Species-syndrome linkages have been established on the basis that Alatina mordens was observed in light halos on night dives of some sting events. The syndrome of Alatina mordens appears to be typically fast onset (e.g. 5 min) and severe in both pain and hypertension. The cross-shelf distribution of Irukandji species elsewhere is less clear. In Western Australia, Carukia shinju has been caught only a few times: once offshore in the pearling grounds and the remaining times coastally around Broome (Gershwin, 2005b). Malo maxima is encountered in large numbers by pearl divers who fish many kilometres offshore, but has also been caught occasionally closer to shore (Gershwin, 2005b). Therefore, segregation by depth seems less clear in these species than in their Pacific counterparts. Sim- ilarly, an undescribed species of Malo is known from the islands off Exmouth, and an undescribed species of Alatina has been caught and photographed numerous times off the Kimberley coast and at Ningaloo Reef (Gershwin, 2005b,c). Neither species is sufficiently well known to infer patterns of distribution, but Alatina does seem to occur closer to shore in the west than in the east. A species of Morbakka is known from Japan (Bentlage et al., 2010), but no information exists on its stings other than as indicated by its name, Morbakka virulenta. Similarly, a species of Malo and an unidentified Morbakka are known from the Philippines (Bentlage et al., 2010), without information on their stings or symptoms. Irukandji syndrome is known from both of Biology and Ecology of Irukandji Jellyfish 41 these regions, but specific linkages with these species or other yet-to- be-identified species are unclear. Southeast Asian and Caribbean species remain to be identified, and as yet no information is available on their distribution other than the few stings opportunistically reported. Fenner and his colleagues (2010) reported three Irukandji stings in Thai waters, and the following year, Lippmann and his colleagues (2011) reported three more from Malaysia. Both reports are believed to dramatically underestimate the true scale of the problem. Clus- ters of three Irukandji stings in the Florida Keys (Grady and Burnett, 2003) and 25 in Stingray City in Grand Cayman (Anonymous, 2011a) are both thought to have been caused by Alatina sp., while a single sting in Guade- loupe (Pommier et al., 2005) remains unattributed to species and virtually no information is available on other stings nearby. All these regions are notable for having oligotrophic shelf habitats that are occasionally flooded by oceanic intrusions, triggering vast salp blooms. Many are likewise notable for having Irukandji infestations accompanied by large numbers of salps. In contrast, the other ‘hot-spot’ for Irukandji, namely, Waikiki Beach in Hawaii, is a completely different mid-ocean volcanic island habitat and is home to the more oceanic Alatina moseri (Thomas et al., 2001).

3.3. Vertical distribution It is generally accepted that Australian Irukandji swim mostly near the surface. While this has never been formally tested, at least for Carukia barnesi it does appear to be fairly reliable. When Barnes famously caught the first Irukandji specimen on 10 December 1961, he did so by concentrating his attention on the top 50 cm of the water column. This tendency for Carukia barnesi to swim near the surface has been noted by many workers (Barnes, 1964, 1966; Cleland and Southcott, 1965; Fenner et al., 1988; Gershwin and Dabinett, 2009; Kinsey, 1988; Southcott, 1959; Williamson et al., 1996). Numerous independent lines of evidence also support this surface- swimming tendency. First, Southcott and Powys (1944) observed that the majority of Irukandji stings occur “on the body or arms while swimming or standing in the sea” (Figure 1.12A). Second, Kinsey (1988) provided a detailed account of where stings occurred on the body and the depth at the time of the sting,concludingthattherewasreasonableevidenceforBarnes’sbeliefthatCar- ukiawasmostlikelytobefoundinthetop0.5 mofwater(Figure1.12B).Third, Fenner and Harrison (2000) found that of 377 Irukandji stings with body-site 42 Lisa-ann Gershwin et al. information, almost half were stung on the arms, whereas less than one third were stung on the legs. Fourth, a widely distributed public safety education figure plots stings reputedly from the Cairns region in 2001–2002; more than 70% were on the upper half of the body (Figure 1.12C). Given that upper and lower parts of the body are exposed as people enter and leave the water, and people stand and float at different depths, the strong bias toward the upper body, arm, neck, and facial stings strongly suggests that Carukia barnesi is primarily encountered near the surface. Our understanding of where other species of Irukandji swim is less clear. Malo maxima is most often observed by pearl divers while hanging at nine metres on their decompression stop (Gershwin, 2005b). Whether this depth is common for Malo, or merely an artefact of the divers’ spare time to observe their surroundings, is unknown. Similarly, Alatina mordens is typically encountered in surface swarms at the back of dive boats in the light halo at night (Gershwin, 2005c). Arneson and Cutress (1976) noted for the Caribbean Alatina sp. (as Carybdea alata) that “The medusae are strong swimmers yet avoid choppy surface conditions. Unless the sea is calm, they remain almost motionless

Figure 1.12 Location of Irukandji stings on the body, demonstrating that most stings occur near the top of the water column. All data from Cairns region. (A) 1942–1943, redrawn from Southcott and Powys (unpublished 1944). (B) 1960s–1970s, redrawn from Kinsey (1988). (C) Believed to be from 2001–2002, from public domain safety education materials. In (C), numbers denote stings on different body regions; black clothing indicates the parts of the body that would have been protected by wearing of a full-body lycra ‘stinger suit’ or equivalent. Biology and Ecology of Irukandji Jellyfish 43 near the bottom. With the usual abatement of wind at night, the medusae rise to the surface to feed”. So too, Grady and Burnett (2003) noted that a series of Irukandji stings in divers off Key West, Florida, occurred while they were swimming close to a sandy, grassy bottom at 3–5 m at night without lights, suggesting that this was the normal habitat for this species. A similar pattern has been noted for many non-Irukandji species as well, so living near the bottom does appear to be a common cubozoan behaviour (Berger, 1900; Hartwick, 1991a; Kinsey, 1986; Larson, 1976; Martin, 2004; Matsumoto, 1995; Studebaker, 1972; Yatsu, 1917).

3.3.1 Shallow stings An interesting corollary to the surface-swimming behaviour of Irukandji is that they are often found right up to the water’s edge and sometimes even stranded by the tide on the beach. Many authors have noted that there is a disproportionately high percentage of Irukandji stings inside stinger-resistant enclosures. For example, Fenner (1988) wrote, “On Christmas Day 1985, the casu- alty room of the Cairns Base Hospital looked like a battleground. Approx- imately 40 people, many of whom had severe symptoms, needed treatment following stings by Irukandji. All were stung in the stinger-resistant enclosures!”. Analysis of 30 Cairns region Irukandji stings in December 1996 revealed that 67% of the stings that month had occurred within the enclosures (Mulcahy and Little, 1997). In looking at the whole of 1996, Little and Mulcahy (1998) found that 63% of Cairns region stings occurred inside these enclosures. Notably, they also found that six (10%) stings that year occurred at the water’s edge, suggesting that the coast itself may have some concentrating effect. Whether the higher-than-expected rate of stings inside the nets is because of the clustering effect of swimming in these areas or because of some eddying effect of the enclosures has not been investigated. It is important to note that the nets were designed to protect against the larger and more dangerous Chironex fleckeri and have proven effective in this respect.

3.3.2 Vertical migration Vertical migration in Irukandji has not been investigated. However, vertical migration could be important for Irukandji to maintain their position inshore and form blooms in tidal environments. For example, the distantly related scyphozoan Aurelia typically spends most of its time near the bottom or randomly distributed through the water column and aggregates at the 44 Lisa-ann Gershwin et al. surface once or twice a day (Mackie et al., 1981; Malej et al., 2007; Yasuda, 1973). It is thought that vertical migration in Aurelia is a means of avoiding tidal dispersion and that aggregation enhances survival by keeping the medusae in an environment that facilitates the meeting of gametes, improves the survival of larvae and juveniles, increases the capture rate of motile prey such as copepods, and reduces the effects of predation by medusivores (Albert, 2007; Purcell et al., 2000).

3.4. Temporal changes The Australian ‘stinger season’ is generally regarded as being November to May. Indeed, experience and local lore suggest that stings are far more prevalent during the warmer months, but in fact, stings and specimens are known from all months of the year (Goggin et al., 2004). However, stings occur in brief epidemics, with generally two primary peaks, one almost invariably in late December or early January and the other often around March or April (Figures 1.13 and 1.14). Many have questioned whether the observed pattern of stings might be an artefact of more people in the water during these holiday periods. How- ever, while certainly a logical concern, numerous streams of evidence have demonstrated that jellyfish peak at these times independent of whether people are in the water. In fact, this has been recognised at least since the 1940s, “The occurrence of these stingings in December and January corresponds with the experience of local inhabitants—the lack of cases during the remainder of the year is certainly not entirely due to the smaller numbers of people bathing” (Southcott and Powys, 1944). So too, later workers in Cairns have noted the prevalence of coastal stings in the height of summer. Little and Mulcahy (1998) analysed Cairns sting data from 1996. They found that of 62 total stings that year, 35 (56%) were stung between 30 November and 19 December, and of those, 30 (88%) occurred at coastal locations. In comparison, for the period January to May, only 17 of 26 (65%) of patients were stung at coastal locations. The authors speculated that the observed swarm period may be due to the species breeding or feeding patterns: “There is a higher proportion of people stung on the Reef between January and May (9/26; 33%) compared with October to December (3/37; 8%). We believe the ‘swarm’ occurs because the Irukandji are either breeding or pursuing food, before moving to open water later in the season” (Little and Mulcahy, 1998, p. 640). It is also possible that they overlooked species differences and that their sampling design con- founded their results by using a calendar year. Biology and Ecology of Irukandji Jellyfish 45

Figure 1.13 Australian Irukandji stings by time of day (A) and by east (B)and west (C) coast seasonality. Data from the Australian Irukandji sting database, comprising 1629 Australian Irukandji sting records from January 1893 to June 2013, obtained from Surf Life Saving, hospital and ambulance records, and media, and curated by a succession of researchers since the 1950s; requests for access to the database can be made through the senior author.

The tight clustering of Irukandji infestations is well illustrated by collection data from a single summer. During the summer of 1999–2000, two of us (LG and RH) made standardised daily collections for 80 days at Palm Cove, a popular beach north of Cairns (6 December–24 February) (Fenner, 2000). During this period, Carukia barnesi were only found in two separate infestation events, 3 specimens in one of six samples on 14 December and then 270 specimens in a 4-day cluster as follows: 29 December, 1 specimen; 30 December, 22 specimens; 31 December, 206 specimens; and 1 January, 41 specimens. Routine daily beach monitoring by lifeguards over the last decade also suggests that Irukandji are not present all the time, but rather, they come and go according to conditions. In particular, since 2003, Australian beach safety protocol mandates closure of the beach when Irukandji are found or 46 Lisa-ann Gershwin et al.

Figure 1.14 Queensland Irukandji stings by location and season. (A) Location of stings. (B) Seasonality of beach stings. (C) Seasonality of island stings. (D) Seasonality of reef stings. Data from Australian Irukandji sting database. (For colour version of this figure, the reader is referred to the online version of this chapter.)

when stings occur (Dawes et al., 2006). Given that most of the patrolled swimming beaches are open and visited by tourists most days, even without the ability to analyse data, we can reasonably conclude that Irukandji are not present most days. For other species, even less information is available. In Queensland, Alatina is typically found at outer reef locations during its monthly swarm events. Morbakka appears to be diffusely scattered along the coast throughout the warmer months of the year, but one particularly severe sting event was attributed to Morbakka on the reef (Gershwin, 2008; Little et al., 2006). Malo is more complicated. Specimens have been taken coastally throughout the northern and central regions, but its nematocysts were identified from a fatal reef sting (Gershwin, 2007; Huynh et al., 2003), and numerous similar severe stings at midshelf locations suggest that it is more abundant there. Biology and Ecology of Irukandji Jellyfish 47

Curiously, its coastal occurrence appears to peak around March-April in the central region but around late December to early January in the north. In the NT, stings in different regions correspond to prevailing offshore winds in those regions, that is, October in Gove (East Arnhem Land) and May farther west in Darwin. Different versions of the syndrome are reported, but only one species is so far known (Gershwin and Alderslade, 2005). Elsewhere in the world, even less is known about Irukandji seasonality. In the Gulf of Thailand, especially around the Koh Samui area, Morbakka is most frequently sighted from November to February, as shown by photographs and reports provided to Divers Alert Network Asia-Pacific. Photog- raphers have usually been local who dive the area year-round. In Langkawi, Malaysia, the main reports of Irukandji stings are from May to July. In Grand Cayman, media reports following a cluster of 26 stings during the morning of 27 April 2011 indicated that these jellyfish were typically present in late spring or early summer and believed to be washed in by deep sea currents (Fuller, 2011). While the marked seasonal prevalence of Irukandji stings in different regions is generally accepted by most workers (Figures 1.13 and 1.14), it has yet to be formally studied or explained. Due to the species diversity of Irukandji (Figure 1.3 and Table 1.2), it seems likely that much of what we perceive as seasonal anomalies will eventually be explained by taxonomy.

3.4.1 Bimodal distribution in space and time Several interesting examples exist where nearby regions have different peaks on the calendar correlated with higher incidences of stings. In the far north of Western Australia, the quiet town of Broome is the main location for Irukandji stings in Western Australia, and may become a Rosetta Stone for understanding Irukandji ecology. Here, Irukandji stings generally occur inside the sheltered Roebuck Bay early in the summer (generally October through December), with later season stings (generally March through June) occurring on the more exposed Cable Beach facing the Timor Sea (Macrokanis et al., 2004). The pearl divers operating offshore somewhat to the south of Broome are also plagued by Irukandji during this late season period (Gershwin, 2005b). So too, the Irukandji season in the Northern Territory (NT) has two peaks. In the eastern-facing region of Gove (East Arnhem Land), the sting season peaks in October, whereas the peak in the more northerly or westerly facing Darwin is in May (Nickson et al., 2009). These peaks coincide with the cusps of the monsoon; the ecological significance of this has not yet been 48 Lisa-ann Gershwin et al. studied. Different versions of the syndrome are reported, but only one species, Gerongia rifkinae, is currently known (Gershwin and Alderslade, 2005). Similarly, in the Great Barrier Reef region of tropical eastern Australia, where sting demographics are better-studied, peak sting times are dissimilar north to south and east to west (Gershwin, 2005c; Kinsey, 1988). The primary peak, generally in late December, occurs in the northern, central, and southern beach regions. The secondary peak appears to be stronger in the central region than in the north, as well as offshore. These cases of bimodality do seem to be genuine jellyfish patterns rather than human swim patterns, because people use the water throughout the year in these localities. The extent to which these geographical shift patterns may represent different species is not yet clear, but late season, more southerly stings inQueenslandtendtohaveahigherrateofseriousillness,lendingsupporttothe ‘different species’ hypothesis (Fenner and Hadok, 2002). The environmental stimuli and biological responses driving these spatio-temporal patterns are not yet clear, but should be a high priority for study.

3.4.2 Lunar periodicity The Hawaiian oceanic Irukandji, Alatina moseri, forms reproductive swarms in nearshore waters along Oahu’s leeward coast about 8–12 days after the full moon every month (Thomas et al., 2001), and subsequently washes ashore in large numbers (Crow et al., 2010). Monthly counts since August 1994 suggest little seasonal pattern, but stronger annual variation. For example, the total annual count for 2001 was more than 10,000, while the total for 2005 was just one-quarter of that. Curiously, anecdotal evidence suggests that these monthly swarms increased in the early 1980s, and by the end of that decade, recreational activity at Waikiki Beach was regularly affected by stings, some of which produced Irukandji syndrome (Yoshimoto and Yanagihara, 2002). The closely related Australian oceanic Irukandji, Alatina mordens, appears to swarm at a similar time of the month (R. Hore, unpublished data; Appen- dix B). However, Alatina moseri presents a health hazard when swarming in shallow waters during the day, whereas Alatina mordens is most commonly encountered over the reef at night.

3.4.3 Diurnal patterns At least some species of cubozoans are found on or near the bottom during daylight hours and are more active in the water column during the dimmer parts of the day or at night. For example, the Caribbean Carybdea xaymacana Biology and Ecology of Irukandji Jellyfish 49 rests on the bottom during the day and is most active in the mornings and evenings (Berger, 1900; Larson, 1976; Studebaker, 1972). The Japanese Carybdea mora was noted to have a similar pattern (Yatsu, 1917), and the Australian Carybdea rastonii and the Californian Carybdea spend most of the day foraging near the bottom and rise to the surface in the early morning or on overcast days (Martin, 2004; Matsumoto, 1995). The Caribbean Alatina sp. spends most of the day almost motionless near the bottom and rises to the surface to feed at night (Arneson and Cutress, 1976). Barnes noted that in the mid- to late afternoon, Chironex would sometimes rest close to the bottom with the bell down and the tentacles retracted into or near the bell (Kinsey, 1986, p. 27). In most cases, the distinction between resting and epibenthic foraging is unclear. Specimen evidence and sting evidence both indicate that Irukandji are more common in the afternoon (Kinsey, 1988; Macrokanis et al., 2004; Nickson et al., 2009; Figure 1.13) raising the question of whether these species respond differentially to different levels of light. Hartwick (1991a) noted that all parts of the life cycle of Copula sivickisi are benthic and that the medusae intriguingly spend the day attached to benthic structures but swim actively in the water column at night. In contrast, Tripedalia cystophora, which is in the same family, has the opposite pattern; it is active near the surface during the day and swims near the muddy bottom at night (Garm et al., 2012). While the physiological and ecological bases for the apparent diurnal rhythms of most cubozoans are not well understood, Garm and his colleagues concluded that these behaviours in Tripedalia and Copula were an adaptation to the activity patterns of their prey. Whether this explains the afternoon activity peak for some Irukandji or the daytime or night-time peak for others is worthy of investigation. It is also possible that for most species in most regions, these morning and evening peaks coincide with periods where low wind turbulence overlaps with adequate daylight for hunting. In general, Irukandji are found in highest abundance during periods of lowest turbulence (Gershwin et al., 2013a; Kinsey, 1988).

3.4.4 El Niño/La Niña influence The question of whether Irukandji occurrence patterns are affected by El Nin˜o–Southern Oscillation (ENSO) cycles is unresolved. On the one hand, the weather conditions that best correspond with Irukandji infestations include above-average temperatures, lack of recent rain, clear skies, and tem- porary subsidence of the alongshore winds (Table 1.6). These hot, dry Table 1.6 Summary of ecological conditions linked with Irukandji stings Wind Wind Water Source Locality speed direction temperature Rainfall Sunlight Tide Time Salps Barnes (1964) Cairns QLD Low Northerly ND ND ND ND ND Swarms Kinsey (1988) Cairns QLD Low NE–N ND ND ND ND Afternoon Swarms Little and Cairns QLD avg avg ND ND ND Mulcahy (1998) NNE Fenner and QLD and NT No or Northerly Air temp None No or little Ebbing, moon ND ND Harrison (2000) low wind mean cloud cover last ¼ 31.2 C Nickson et al. NT Still or Offshore Median None No or little High (also > Noon to ND (2009) slight 29.9 C cloud cover or <) 2:59 pm Gershwin Broome, WA ND ND >26 C ND Clear sky 15 km/ NC >28.3 C NC ND High Afternoon ND et al. (2004) h Fenner et al. Thailand Low Onshore Warm ND ND ND ND Swarms (2010)

NC, no correlation; ND, no data available. Australian states: QLD, Queensland; NT, Northern Territory; WA, Western Australia. Biology and Ecology of Irukandji Jellyfish 51 conditions are also often associated with El Nin˜o in Australia. On the other hand, a direct link has yet to be demonstrated.

3.4.5 Climate change influence Recent media reports have suggested that the Irukandji sting season has length- ened from about 1 month 40 years ago to about 6 months now (McKechnie, 2010) and have used this as evidence of global warming effects. These observations appear to be skewed. While there is no doubt that sting records have increased over the last century (Figure 1.15), there could be multiple reasons for it. Beach tourism has increased several-fold since that time (Harriott, 2002). Moreover, data from the 1960s to 1970s were primarily from the coastal Cairns region because the reef tourism industry was nearly non-existent compared with today and Cairns was the centre of Irukandji awareness and reporting (Barnes, 1964). However, we now know that stings generally occur later in the summer and autumn off Central Queensland and out on the reefs and islands than along the Cairns beaches (Figure 1.14) and that these later sea- sonstingsareoftenmorevirulent(FennerandHadok,2002).Ofcourse,itisalso possible that some fundamental change to the season or species has occurred

Figure 1.15 Interannual variability of Irukandji stings in Australia, 1920–2012. Low numbers of records pre-1960s and 1970s–1980s are believed to be due to low reporting effort; low numbers in late 2000s are believed to be due to improved management at the beaches. Data from Australian Irukandji sting database. 52 Lisa-ann Gershwin et al. sinceBarnes’ time, butcannot be determinedsimply bytaking the available data at face value. Curiously, two anomalous and remarkable high-latitude infestation events occurred in the summer/autumn of 2013. Irukandji stings only rarely occur at Fraser Island in southern Queensland, but in late December and early January, a cluster of seven stings in 8 days occurred (Fraser Coast Chronicle, 2013). Similarly, Irukandji stings rarely occur at Ningaloo Reef off central Western Australia, but from April to June, at least 23 people were taken to hospital with Irukandji syndrome, with nine occurring in less than a week (Le May, 2013). In both of these high-latitude infestations, the number of stings was far higher than for the more typical low-latitude stings for the season, suggesting some sort of productivity shift. The physical and biological context for these infestations remains unclear but is a high priority for urgent study. It might be tempting to look at these events and conclude, as some have done, that Irukandji must be moving toward the more populated temperate regions. However, occasional cases of Irukandji syndrome have occurred in higher latitudes for many decades (Cleland and Southcott, 1965; Gershwin et al., 2009; Williamson et al., 1996). In truth, there is currently no evidence to inform us as to the ecological circumstances that would be required for tropical Irukandji species to migrate and flourish in southern coastal waters. But one might imagine that it would involve ecosystem migration, not just a medusa or two. 3.5. Movements and aggregations Carukia barnesi blooms in large swarms, whereas Morbakka and Malo appear to be more solitary. However, occasionally, clusters of Morbakka have been found (Gershwin, 2008), and large numbers of Malo have been captured on both the east and west coasts of Australia (Gershwin, 2005b, 2007; Kinsey, 1988). Similarly, the type series of Gerongia comprises many specimens caught together, but subsequent findings have been sparse (Gershwin and Alderslade, 2005; Williamson et al., 1996). In contrast, Carukia does seem to be a genuinely blooming species, with more than a thousand specimens taken in one bloom event (Anonymous, 2002) and hundreds taken several other times (AAP, 2005; Bateman, 2010a; Bester, 2012; Fenner, 2000). Alatina mordens is only rarely found, but like its Hawaiian cousin Alatina moseri, it appears to be more prevalent on the 8th, 9th, or 10th nights after the full moon (Gershwin, 2005c; Appendix B). Studies have examined a range of ecological and behavioural variables, but most have associated these variables anecdotally and without statistical rigour. Biology and Ecology of Irukandji Jellyfish 53

3.5.1 Effects of wind Onshore winds have long been regarded as the primary indicator heralding the arrival of an Irukandji infestation. Indeed, the coincidence of Irukandji stings and onshore breezes is so strongly linked that it has been noted by almost every major worker on Irukandji (Barnes, 1964; Fenner and Harrison, 2000; Flecker, 1957a; Gershwin, 2005a; Gershwin et al., 2009; Kinsey, 1988; Little and Mulcahy, 1998; Southcott and Powys, 1944; Williamson et al., 1996). In popular local lore, a gentle onshore breeze (5–10 knots) sustained over several days produces the highest risk conditions in tropical Australia (Parsons, 2013; Roberts, 2002; http://www.cairnsvisitorcentre.com/faq). Local lore also holds that the first couple of days after the breeze stops are also high risk, supposedly because the Irukandji are thought to be ‘heading back the other way’. Paradoxically, however, despite the large number of these observations, this suggestion has rarely been tested rigorously. The first effort at quantifying the link between wind and stings was made by Kinsey (1988), who mapped Barnes’ Cairns region sting records that contained wind information. Kinsey found that 25 stings occurred on winds from the north, 23 on winds from the northeast, and 6 on south- easterlies. Later, Little and Mulcahy (1998) analysed 62 Irukandji stings from around Cairns in 1996 and found that 47 (76%) patients were stung on days when the wind blew from the north, despite these winds only being prevailent on 27% of days in 1996. Fenner and Harrison (2000) examined 544 Irukandji stings in Queensland and the Northern Territory and found that 73% occurred on 0-knot or light-wind days. Most recently, Gershwin et al. (2013a), using a time series of stings over the past 30 years, found that the subsidence of the southeasterly trade winds offered a mechanism for early prediction: as the dominant alongshore winds subside, the onshore northeasterly sea breeze becomes more obvious and, along with sub-surface intrusions, serves to drive the Irukandji conditions shoreward. The phenomenon of Irukandji influx on an onshore breeze gives rise to a long-standing hypothesis that Irukandji swarms are most prevalent on exposed beaches facing into the breeze. However, the opposite seems to be true more often. In particular, of 14 common island sting ‘hot spots’ in the Great Barrier Reef, 12 are near the southwestern portion of the island or southwest of a rocky headland, whereas only two are along the open northeast (Figure 1.16). In contrast to the gentle onshore breezes heralding the arrival of Irukandji on the east coast of Australia, Macrokanis et al. (2004) found that 54 Lisa-ann Gershwin et al.

Figure 1.16 Island Irukandji sting hot-spots in the Great Barrier Reef. Red dots indicate hot-spots leeward of islands or headlands during northeasterly winds; yellow dots indicate hot-spots in the direct path of northeasterlies. Data from Australian Irukandji sting database. Maps from Google Earth. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) higher sting probability on the west coast correlated with wind speeds greater than 15 km/h. However, this does not appear to be universally true. The large cluster of stings in the autumn of 2013 near Exmouth, Western Australia, was accompanied by low wind speeds (Peter Barnes, Dept Envi- ronment and Conservation, pers. comm., May 2013). In Australia’s Northern Territory, it is the offshore wind rather than the onshore wind that corresponds with Irukandji stings (Nickson et al., 2009). Specifically, the other Australian regions discussed in the preceding text (see bimodal distribution, section 3.4.1) peak during prevailing local offshore wind periods. The ecological reason for this discrepancy is not understood.

3.5.2 Infestations For as long as Irukandji stings have been known, they have typically occurred in occasional brief epidemics. For example, 36 stings occurred on Christmas Day 1985 (Martin and Audley, 1990) and 50–70 in 1 day when life-savers entered the water to haul in an enclosure net that was deployed to Biology and Ecology of Irukandji Jellyfish 55 keep swimmers safe (Kinsey, 1988). Barnes described seeing three truckloads of ‘writhing carcasses’ driving past his house one day; they were taking 40–50 sting victims to the hospital (Kinsey, 1988). Barnes (1966) described well the ephemeral nature of these infestations, “Irukandji stings characteristically occur in localised outbreaks, claiming as many as forty victims on a single beach within a period of a few hours. At the same time, neighbouring beaches may remain stinger-free. Each visitation is brief, rarely lasting more than 2 days, but the invasion may be repeated several times within one season” (p. 309). These mass sting events capture the imagination of the public and the media, particularly in decades past or still today in regions where the cause is poorly understood. Newspaper reports from Western Australia such as Forty Swimmers Stung by Sea Snakes are thought to refer to Irukandji stings; even though people did not see any snakes, these were the only thing imag- inable that could cause such discomfort, so people assumed there were a lot of invisible snakes doing the stinging (Kinsey, 1988). Curiously, even though Carukia and Malo both occur coastally and offshore on the eastern and western Australian coasts, Carukia is well understood to be the primary constituent of blooms in the east (Kinsey, 1988), whereas Malo appears to be the primary bloomer in the west (Gershwin, 2005b). Whether these infestation events represent true swarming in the active sense or are merely passive aggregations is not yet clear. However, species in the Irukandji syndrome-causing genus Alatina are believed to actually swarm in the true ethological sense. Thomas et al. (2001) suggested that the monthly influxes of Alatina moseri in Hawaii were spawning aggregations, with mature spawning males arriving at the shore approximately 1 h before the high tide and mature spawning females arriving 1 h later. Hundreds of people are stung during these influxes. For example, more than 800 people were stung on 29 July 1997 (Kreifels, 1997), some developing Irukandji syndrome (Yoshimoto and Yanagihara, 2002). Why this monthly Irukandji swarming phenomenon occurs only on the leeward shore of Oahu remains mysterious, but may be simply a matter of the jellyfish preferring calm conditions or habitats. So too, Arneson and Cutress (1976) reported night-time spawning aggregations of several hundred Alatina (as Carybdea alata) from Puerto Rico. These aggregations occurred on 23 July 1973 and 12 August 1974, but did not occur at other times; both of these dates coincide with the ninth night after the full moon. The authors noted that the water was calm on both nights, but no other environmental factors could be identified. 56 Lisa-ann Gershwin et al.

Elsewhere, we have less information on the species, but the epidemic nature of the stings is similar. In Malaysia, for example, in addition to several confirmed reports (Lippmann et al., 2011), in 2010, Divers Alert Network Asia-Pacific received a report of around 150 people receiving treatment in hospital for Irukandji-like symptoms sustained during aquatic activities in Langkawi, Malaysia, between May and July that year. Similarly in Grand Cayman, an area known as the Sandbar near the popular Stingray City was shut down following a cluster of 26 stings on the morning of 27 April 2011, with eight being hospitalised (Fuller, 2011); this bloom was said to be ‘fairly typical’ for this time of year in this area. The tendency for Irukandji to swarm and only occur periodically makes safety management easier than if they were always present or problematic at low densities. However, early forecasting of when and where the different species are likely to occur is still in many ways the Holy Grail of Irukandji research.

3.5.3 Swimming behaviour Swimming studies have not been performed on Irukandji species, but the California Carybdea has been recorded at 80–100 pulsations per minute (Satterlie, 1979), and large Chironex fleckeri has been clocked at speeds of 4–5 knots (Barnes, 1960; Kinsey, 1986). Large Irukandji species such as Morbakka and Alatina are very agile and powerful swimmers, and it is probable that these species have similar swimming speeds and are able to swim against a current. Even medium-sized species such as Gerongia and Malo are quite powerful. While larger species are quite strong swimmers, the more diminutive Carukia is able to orient itself in the water column but cannot fight a current. Even in a low-flow aquarium environment, it spends most of its time drifting motionless with its tentacles streaming out, presumably fishing (LG pers. obs.).

3.5.4 Effect of temperature The associations between warm temperature and Irukandji infestations seems fairly clear; however, whether thermal increase is a trigger for blooms has not yet been investigated. In a year-long study of northern Great Barrier Reef stings, 92% (57 patients) were stung on days hotter than the average for the month when the sting occurred (Little and Mulcahy, 1998). Similarly, a study over 2.5 years in northern Western Australia found a higher probability of beach Biology and Ecology of Irukandji Jellyfish 57 stings when the water was greater than the yearly median of 28.3 C (Macrokanis et al., 2004). Offshore in the same region, stings are anecdotally believed to be most virulent when the water is above 26 C (Gershwin, 2005b). Moreover, in an 18-year study of 87 cases in Australia’s Northern Territory, water temperature was known for 77 cases; the median was 29.9 C (range 25–32.3 C) (Nickson et al., 2009), most well above the yearly average of 27 C.

3.6. Environmental variables Many workers have found a suite of anomalous conditions associated with higher incidence of Irukandji stings (Table 1.6). In general, periods with low wind, unusually hot weather, less-than-average rainfall, clear skies, and thick blooms of salps tend to coincide with Irukandji infestations.

3.6.1 Effect of tide Tides are often invoked as a causal agent in coastal Irukandji infestations but have yet to be demonstrated. For example, Great Barrier Reef stings are anecdotally thought to be more prevalent on days with an afternoon high tide, and this was also observed by Macrokanis et al. (2004). Malo maxima was observed over the course of several nights around the neap tides off the coast of northern Western Australia in autumn of 2004 (Gershwin, 2005a). The species was most often caught during the penulti- mate hour before slack tide, with greater abundance on a falling tide than on a rising tide. Approximately 1.5–2 h before slack tide, medusae began arriving in the halo under powerful flood lights. The medusae steadily increased in numbers, with more than 10 simultaneously visible much of the time despite being caught and removed when spotted; their attraction to the light seemed undeterred by fish and squids that were also swarming under the light. About 30–60 min before slack tide, the arrival of medusae ended abruptly. Some nights, they came for both slack tides but were more abundant on the earlier slack, whereas other nights, they only came for the first slack tide. These results may have been biased by the collecting trip being limited to the neap period, during which the earlier evening tide was pre- dominantly the low tide. In contrast to the large number of medusae observed at night, divers caught approximately five specimens and saw some several dozen during the day throughout the 9-day neap period. Nickson and his colleagues (2009) found that more stings occurred at high tide than any other time and that together high, incoming and outgoing 58 Lisa-ann Gershwin et al. tides accounted for 86% of stings. Fenner and Harrison (2000) found that more than 70% of the stings for which they had tidal information occurred on an ebbing tide and that half of those with moon phase data occurred during the moon’s last quarter. The majority of stings studied by Fenner and Harrison were attributable to Carukia barnesi. Therefore, this peak of stings during the last quarter moon is particularly intriguing, given that this is the same time period as the monthly influx of Alatina spp. in Australia and Hawaii (Gershwin, 2005c; Thomas et al., 2001).

3.6.2 Effect of sunlight Little and Mulcahy (1998) noted that 69% of patients in their study were stung on days with more hours of sunshine than average. Similarly, Fenner and Harrison (2000) noted that more than 40% of Irukandji stings occurred when cloud cover was less than or equal to two-eighths, and Nickson et al. (2009) found that 73% of stings in their study occurred when the weather was fine.

3.6.3 Effect of rainfall Little and Mulcahy (1998) found a strong link between low rainfall and Queensland Irukandji stings, with 87% of stings occurring on days with 5 mm or less of rain and 76% occurring when less than the average amount of rain had fallen in the past week. Similar conclusions were drawn by two other studies in Queensland and the Northern Territory (Fenner and Harrison, 2000; Nickson et al., 2009). However, no correlation was observed between rainfall and stings in Western Australia (Macrokanis et al., 2004).

4. TOXINS

Venoms and toxins occur abundantly across the animal kingdom facil- itating not only prey capture and digestion but also to avoid predation and for some sessile marine organisms, as a defense against infection. As cnidarians have succeeded in persisting in highly competitive habitats for hundreds of millions of years, it is not surprising to find that they exhibit numerous sophisticated cellular inventions and innovations. These include cnidocyst and venom composition variation and specialisation. Bioactive components have been identified in all cnidarian classes throughout the entire organism and not simply confined to the nematocyst (Aneiros and Garateix, 2004; Ovchinnikova et al., 2006). Nevertheless this review will focus on Irukandji nematocyst venoms. Their chemical arsenal is Biology and Ecology of Irukandji Jellyfish 59 an intricate mixture of pharmacologically active substances. These include cytolysins, neurotoxins and lipases (Talvinen and Nevalainen, 2002), peptidases (Gusmani et al., 1997), protease inhibitors (Delfı´n et al., 1994), and antimicrobials (Morales-Landa et al., 2007). The nature of the ‘Irukandji’ venom, its evolution and ecological significance, remains poorly described. Progress was minimal until, in the early 2000s, multidisciplinary Australian teams began a long-term effort to map the species involved, the nature and action of their nematocyst toxins, and to clone the DNA encoding the key molecules. This was driven, in large part, by the need to address a growing public health problem that lacked a specific treatment. In parallel, these and other investigators have also begun to address the broader phylogenetic questions through complementary cnidarian genome studies. Additionally, progress has recently been made in mapping the nematocyst proteomes from other cnidarian species, providing data that will ultimately assist in accelerating future understanding of the toxins involved in this enigmatic syndrome. Although some preliminary biochemical data and the initial results of pharmacological studies using Carukia barnesi venom extract were reported in 2000 (Tibballs et al., 2000; Wiltshire et al., 2000; Winkel et al., 2000), the first major papers addressing the pathophysiology of Irukandji syndrome were published more recently (summarised in Table 1.7; Li et al., 2011; Ramasamy et al., 2005; Winkel et al., 2005; Winter et al., 2008). Major reviews in the late 1990s (Burnett et al., 1996, 1998) had speculated on the resemblance of the syndrome to cases of adrenal medullary or catecholamine excess, such as seen in cases of phaeochromocytoma and scorpion or funnel-web spider envenomation. As initially suspected from its clinical features, experimental studies confirmed that the syndrome is essentially one of excessive circulating catecholamines, notably noradrenaline and its methylation product, adrenaline (Table 1.7). This appears to be secondary to venom-induced modulation of Tetrodotoxin-sensitive prejunctional neuronal sodium channels in peripheral postganglionic sympathetic sites and, possibly, splanchnic nerve innervations and the adrenal medulla (Winkel et al., 2005). This results in the systemic and pulmonary hypertension and increased cardiac output. Although some variation in venom potency has been described, with Car- ukia barnesi being apparently more potent than either Malo maxima or Alatina mordens, all three species exhibited evidence of sympathetic activation in experimental studies. Note that no intrinsic sympathomimetic activity, such as from endogenous catecholamines, has been yet identified in these 60 Lisa-ann Gershwin et al.

Table 1.7 Summary of key studies and findings related to Irukandji jellyfish venom, toxins, and genomics Irukandji Summary of findings venom study Winkel et al. Investigated the cardiovascular pharmacology of the crude venom (2005) extract (CVE) from Carukia barnesi, in rat, guinea pig, and human isolated tissues and anaesthetised piglets. It was concluded that venom may contain a neural sodium channel modulator (blocked by TTX) that, in isolated atrial tissue (and in vivo), causes the release of transmitter (and circulating) catecholamines. Both sympathetic and parasympathetic nervous system effects observed. Venom may also contain a ‘direct’ vasoconstrictor component. No biochemical data provided. No studies of sensory nerve contributions Ramasamy Investigated in vivo cardiovascular effects of Carukia barnesi venom et al. (2005) and a tentacle extract (devoid of nematocysts). Findings consistent with effects of catecholamine release. Also showed, for the first time, that tentacle extract, free of nematocyst material, produces cardiovascular effects distinct from those caused by venom derived from isolated nematocysts Winter et al. This study characterised the in vitro and in vivo effects of Alatina (2008) mordens venom and indicated cardiovascular effects are at least partially mediated by endogenous catecholamine release. Reported a lower potency of venom compared with Carukia barnesi. Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) profile of Alatina mordens venom showed that the venom is composed of multiple protein bands ranging from 10 to 200 kDa. Western blot analysis using CSL box jellyfish antivenom indicated several antigenic proteins in Alatina mordens venom; however, it did not detect all proteins present in the venom Underwood Venom ontogeny, diet, and morphology in Carukia barnesi were and Seymour assessed. SDS gel electrophoresis revealed differences in protein (2007) banding of tentacular venom between immature and mature animals. This was associated with a change in diet from invertebrate prey in immature Carukia barnesi medusae to vertebrate prey in mature medusae. Unlike other cubozoan studies, a change in venom did not equate to a change in nematocyst types or their relative frequencies A´ vila-Soria This Ph.D. thesis reported success with the development of both Malo (2009) kingi and Carukia barnesi cDNA libraries. This allowed the establishment of an EST resource from which were identified novel transcripts, several serine and zinc proteinases and their inhibitors, two neurotoxin-like genes, and two apparent cytolysins. RNA in situ hybridisation studies revealed restricted expression of these putative neurotoxins, in adult Carukia barnesi, to tentacular nematocyst batteries Biology and Ecology of Irukandji Jellyfish 61

Table 1.7 Summary of key studies and findings related to Irukandji jellyfish venom, toxins, and genomics—cont'd Irukandji Summary of findings venom study Li et al. The in vitro cardiac and vascular pharmacology of Malo maxima was (2011) investigated in rat tissues. Malo maxima CVE appeared to activate the sympathetic, but not parasympathetic, nervous system and to stimulate sensory nerve CGRP release in the left atria and resistance arteries. Effects are consistent with the catecholamine excess thought to cause Irukandji syndrome, with additional actions of CGRP release. Reported a lower potency of venom compared with Carukia barnesi. SDS-PAGE profile of Malo maxima venom showed most toxins to reside between 20 and 100 kDa molecular weight

venoms, although DOPA (dihydroxyphenylalanine), a precursor to norandrenaline that has no pressor action, was detected in Malo maxima venom extracts (Li et al., 2011). This pattern of venom action is quite distinct compared to that reported from non-Irukandji cubozoans such as Chironex fleckeri (Hughes et al., 2012). This archetypal chirodropid venom does not appear to involve autonomic nerves, postsynaptic adrenoceptors, or muscarinic or sensory neural peptide calcitonin gene-related peptide (CGRP) receptors, but may occur through direct effects on the cardiac and vascular muscle (Hughes et al., 2012). Further experimental variation in venom action has been identified between these species. Specifically, minimal parasympathetic effects were observed with Malo maxima venom in vitro, whereas Carukia barnesi venom exhibited definite parasympathetic effects in isolated atrial tissues (Li et al., 2011; Winkel et al., 2005). Hence, the relative significance of sympathetic versus parasympathetic contributions to the dysautonomia (nervous system dysfunction) manifest in the syndrome, as caused by different species, remains to be determined. In addition, recent work on Malo maxima venom has revealed a direct action of this venom on sensory nerves. The role of sensory nerve effects has not been examined for the two other species. While Malo venom action involves the release of the CGRP from sensory nerves, it might also involve neuropeptide Y (Li et al., 2011). This possibility, and the possible role of other afferent pain pathways in the syndrome, requires further investigation. For example, although some cnidarian venoms (Chironex fleckeri, Aiptasia pulchella, Cyanea capillata, and Physalia physalis) appear to activate TRPV-1 (Cuypers et al., 2006), a 62 Lisa-ann Gershwin et al. nonselective cation channel expressed in nociceptive neurones, it is unclear whether these channels are implicated in Irukandji syndrome. Although Irukandji stings may cause a variable degree of local pain, the characteristic severe muscular pain of delayed onset cannot be explained by the involve- ment of local nociceptive effects alone. Indeed, compared to the pro- nounced local and immediate pain associated with those jellyfish venoms tested (Cuypers et al., 2006), Irukandji syndrome pain is very different. Hence, a distinct mechanism would be predicted (Tibballs et al., 2012). As these pharmacological studies progressed, the first details of the biochemistry of ‘Irukandji’ toxins were discovered/found (Table 1.7). Although the first description of the nature of cubozoan toxins began with Wiener’s studies of Chironex fleckeri in the late 1950s (Southcott and Kingston, 1959), no Irukandji species were subject to such assessment until significant numbers of specimens became available in the late 1990s. It appears that these venoms contain a minimum of tens of proteins, most of which reside between 20 and 100 kDa molecular weight, but with some higher weight entities noted in all three species examined. Based on extrap- olation from the Chironex fleckeri proteome, many of these proteins are likely to be posttranslationally modified by glycosylation (Brinkman et al., 2012). Further, ontogenetic studies of Carukia barnesi venom profiles suggest that the venom protein complement changes as the organism matures, coincident with a transition from crustacean-targeting immature forms to adult stages preferring larval fish (Underwood and Seymour, 2007). Note that this change in venom proteome did not equate to a change in nematocyst types or their relative frequencies. Whether the venom of mature adults is actually more toxic to vertebrates, including humans, requires further research. This emergent pattern is consistent with general trends evident in a series of newly published medusozoan nematocyst proteomes. For example, the screening, using high-throughput protein analysis, of nematocysts from two hydrozoans, Hydra magnipapillata and Olindias sambaquiensis, and one cubozoan, Chironex fleckeri, has begun to reveal the evolutionary history of this organelle, one characteristic of the cnidarians, a group that is arguably the most ancient of all venomous animals (Balasubramanian et al., 2012; Brinkman et al., 2012; Weston et al., 2013). The former study reported a complex secretome of 410 proteins with venomous and lytic but also adhesive or fibrous properties (Balasubramanian et al., 2012). The authors concluded that extracellular matrix motif proteins may have provided the evolutionary origin of nematocyst venoms. Biology and Ecology of Irukandji Jellyfish 63

The second hydrozoan proteome (O. sambaquiensis) contained tens of potential toxins homologous to most of the important superfamilies of venom peptides reported from higher organisms. This includes cytolysins, neurotoxins, phospholipases, and toxic peptidases (Weston et al., 2013). These findings were consistent with an earlier study of an anthozoan met- aproteome that revealed a complex mix of diverse toxins, including ion- channel-modulating peptides and cytolytic enzymes (Weston et al., 2012). Such findings argue that the development of these toxin types may represent very early and basal eumetazoan innovations (Weston et al., 2013), consistent with the antiquity of this group of venomous animals. The closest published nematocyst proteome to that of the Irukandji group, representing the first from a cubozoan, is that of Chironex fleckeri (Brinkman et al., 2012). This recent analysis identified 61 proteins included both toxins and proteins important for both nematocyte development and nematocyst formation. The most abundant of these putative toxins identified were isoforms of potent cnidarian cytolysins. Due to the shortage of animals to provide sufficient material for conven- tional bioassay-guided protein purification, it would seem logical to prioritise genomic strategies to accelerate understanding of Irukandji toxins. However, few genomic studies of cnidarian, let alone cubozoan or Irukandji, toxin families have been published. A draft genome of the sea anemone Nematostella vectensis revealed surprising complexity and unexpected affinities with higher-order bilaterians, notably vertebrates (Putnam et al., 2007). Neverthe- less, the first complete mitochondrial genome sequence from a cubozoan, Alatina moseri, confirmed the significant deviation of medusozoan mitochondrial DNAs from that of other animals (Smith et al., 2012). Subsequently, a comprehensive cubozoan phylogeny, based on ribosomal genes, was published (Bentlage et al., 2010). This latter paper proposed that the last common ancestor of Carybdeida probably possessed the mechanism(s) underlying Irukandji syndrome. This has strengthened the case for applying molecular cloning strategies to elucidate the nature of toxin genes. Preliminary success has been reported with the development of both Malo kingi and Carukia barnesi cDNA libraries (A´ vila-Soria, 2009). The former allowed the establishment of an EST resource from which were identified novel transcripts, several serine and zinc proteinases and their inhibitors, two neurotoxin-like genes, and two apparent cytolysins homologous to those previously reported from other cnidarians. Further, RNA in situ hybridisation studies revealed restricted expression of these putative neurotoxins, in adult Carukia barnesi, to tentacular nematocyst batteries. 64 Lisa-ann Gershwin et al.

Finally, after successful expression of one of these putative Carukia barnesi neurotoxin genes, pilot studies confirmed their neurotoxic potential in the form of lethal paralysis after injection into cockroaches (A´ vila- Soria, 2009).

4.1. Which part of the animal is toxic? Although we have yet to resolve whether the tentacles, or the bell, or both carry the toxic fraction that produces Irukandji syndrome, there are logical arguments on both sides. Jellyfish tentacular nematocysts are the most variable and diagnostic between species (Figure 1.6; Gershwin, 2006a), and therefore, logically, special types would provide a substantial functional advantage to their owners. Moreover, the type of nematocyst that carries the lethal fraction in Chironex, called a mastigophore, is confined to the tentacles of Irukandji (Endean and Rifkin, 1975; Gershwin, 2006a). However, on the unusual occasions that a sting mark is observed, it is generally blobular rather than linear, leading to a hypothesis that the bell is responsible for the sting. For example, a bell-shaped sting mark was evident in a case from the Northern Territory in which bell nematocysts were recovered (Williamson et al., 1996, pl. 5.5). But then, it is equally plausible that the tentacles attached to the bell inject venom without leaving a mark, particularly in the smaller Irukandji species with fine tentacles. Perhaps a more rigorous approach to test whether the Irukandji syndrome-producing toxin (ISPT) is carried in the bell nematocysts lies in the nematocysts themselves. Types of tentacular nematocysts are variable among species, including even the presence or absence of mastigophores; however, all species of syndrome-producing cubozoans share the same type of bell nematocyst, called spherical isorhizas, and these are also found in blue bottles (Physalia spp.), at least one of which causes Irukandji syndrome. However, non-syndrome-producing species of cubozoans and Physalia also have isorhizas, leading us to question this hypothesis. Nonetheless, a switch of presence/absence of syndrome-producing venom in one type of nematocyst would be more parsimonious than similar switches in cnidomes peculiar to each species. Even species in the genus Carybdea, long considered safe compared to their more toxic cousins Carukia and Chironex, have spherical isorhizas on the bell. And there is some suggestion that Carybdea may be capable of causing systemic symptoms consistent with Irukandji syndrome (Fenner, 2006; Gershwin, 2006b; Little et al., 2006). Biology and Ecology of Irukandji Jellyfish 65

4.2. Evolution of Irukandji toxins Combined nuclear large subunit, small subunit, and mitochondrial 16S analysis led Bentlage et al. (2010) to conclude that the carybdeid family Alatinidae is ancestral to other carybdeid families, and therefore, the ISPTs were likely to be present in the ancestral carybdeids and subsequently lost in some groups. It may thus be reasonably hypothesised that ISPTs may be at least 300 million years old. A host of interesting questions arise in discussions about ISPTs. Perhaps the most frequently asked is, “Why would a jellyfish need toxins that cause such powerful systemic effects?”. Typical hypotheses include the following: The delicate body must neutralise prey rapidly or risk damage, highly motile prey such as fish must be quickly subdued in order to get a meal, and a soft gelatinous body requires powerful defence. However, the apparent ancient origin of ISPTs may predate fish. The oldest putative bony fish is the approximately 500-million-year-old armoured and jawless Cambrian species Anaspis. Similar to the modern- day jawless hagfish and lampreys, these early ancestral forms were almost certainly bottom dwellers and therefore probably had little contact with jellyfish. Most fish diversification took place in the Silurian and Devonian (ca. 440–360 mya), the latter of which is often referred to as the ‘Age of Fishes’. However, it appears that all extant marine actinopterygians are derived from a freshwater ancestor (Carrete Vega and Wiens, 2012). Irukandji do not survive in freshwater nor is there any reason to believe that they ever did. It therefore seems plausible that ISPTs either evolved inde- pendently of fish or coevolved in the context of fish predator/prey dynamics that are no longer extant. It is possible that the powerful toxic effect on humans and at least some other vertebrates may be purely coincidental.

5. STINGER MANAGEMENT

Today, management of the Irukandji problem primarily falls into four broad categories with a somewhat sequential relationship: • Prediction of infestations to identify when and where they are likely to occur • Detection of the animals before stings occur • Prevention of stings when the animals are present • Treatment of symptoms when stings occur 66 Lisa-ann Gershwin et al.

5.1. Prediction Decades ago, Barnes (1964) recognised the association between Irukandji and onshore winds. However, it was not until 2012 that a plausible mechanism was identified, which now appears to be the subsidence of alongshore winds (Gershwin et al., 2013a). In the coastal Cairns region, occasional prolonged subsidence of the southeast trade winds corresponds with days on which stings have occurred (Figure 1.17), allowing for early forecasting of heightened risk conditions. On these days, subsidence of the alongshore winds reduces the turbulence and turbidity, creating conditions more favourable for these delicate animals. Simultaneously, release from wind- influenced downwelling pressure results in intrusions of oceanic water onto the shelf, bringing in the oceanic hydromedusae and stimulating the salps that are often observed with Irukandji infestations. Sub-surface intrusions and internal waves may further enhance transport of this Irukandji water mass closer to shore. This hypothesised mechanism has not yet been tested in other locations, but the principles may be applicable to numerous other habitats around the world.

Figure 1.17 Alongshore and cross-shore wind components around the time of three stings in the Cairns region (the first sting corresponded to the 9th January 2007). Most stings coincide with a drop in the alongshore wind (red), allowing the onshore sea- breeze (blue) to dominate. Following such events, high sting rates can persist for up to a week. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Biology and Ecology of Irukandji Jellyfish 67

The habitats in which most Irukandji occur globally are similar in several key features, while the specific infestation conditions are anomalous overall. Namely, Irukandji habitats include an oligotrophic shelf system with occasional salp blooms; they also share the feature of dominant alongshore winds, where the sea breezes occasionally appear dominant as the alongshore winds subside. Therefore, because of these similar anomalous conditions, predicting when and where Irukandji infestations will occur should be feasible.

5.2. Detection 5.2.1 Bioindicators A strong association between Irukandji and salps has been used effectively by Surf Life Saving Queensland since 2005 to better estimate the relative risk of Irukandji. In particular, because Irukandji can be hard to see but swarms of salps are hard to miss, the presence of salps can be used as an indicator that Irukandji jellyfish are likely to be present. Typical densities are on the order of 2–4 l of salps in a 5–10 min hand-towed net drag in waist-deep water; typically just a few, but sometimes dozens, of Irukandji are found in each of these salp samples (Gershwin, unpublished data; Surf Life Saving Queens- land, unpublished data). The days with the highest numbers of Irukandji caught generally coincide with a band of salps and hydromedusae washed up at the tideline. Irukandji infestations have been known for decades to co-occur under an anomalous set of conditions, that is, a thick bloom of salps, a variety of hydromedusae including oceanic species such as Narcomedusae and Liriope, and cool, clear, oceanic water. Barnes and others vividly described this unusual set of conditions (Barnes, 1964, 1966; Cleland and Southcott, 1965; Kinsey, 1988). Even early workers such as Stenning (1928) and Southcott and Powys (1944) noted that Irukandji stings occurred when a large amount of gelatinous zooplankton was in the water. For example, Barnes (1964) noted that during prolonged northerly weather, “Under these conditions there is, about a half mile off shore, a south-going stream of clear oceanic water; near the coast the water is murky, warmer, and also moving south, but at a slower rate ...It is interesting that in the past, during periods of Irukandji infestation, life-savers have commented on ‘drops of solid water’ (salps and small hydromedusae) on their skins, and other observers have noted ‘jelly buttons’ (discoid medusae) cast up by the waves. These correlations provided a valuable method of forecasting the likelihood of Irukandji stings, and greatly reduced their incidence. Collection of current-borne marine life was facilitated, but the very 68 Lisa-ann Gershwin et al. plenitude of this life proved an embarrassment. Fine-mesh nets became clogged with compacted jellies within a few minutes, and detailed examination of the catch was impracticable” (pp. 899–900). The strong anecdotal link between Irukandji jellyfish and salps seems persuasive, at least for Carukia barnesi, despite having never been formally quantified or scientifically tested. It seems safe to say that every worker who has studied Carukia barnesi or its stings in the field has observed the prevalence of salps and ‘sea lice’ during times of infestation. While the term ‘sea lice’ can mean different things in different regions, these Australian ‘sea lice’ are not actually lice, or even arthropods; they are a large mass and diversity of small hydromedusae and tentacle fragments that give the skin a feeling of tiny pinpricks all over. These sea lice are not dangerous and they do not cause Irukandji syndrome. The relationship between salps and Irukandji does not appear to be equally applicable across all species. So far, Carukia barnesi and Malo maxima have been confirmed to co-occur with salps (Barnes, 1964; Gershwin, 2005b), whereas Alatina and Morbakka appear not to. No information indicating an association is available for Gerongia or other species of Carukia and Malo. The reason for the association between salps, sea lice, and Irukandji has not yet been resolved. However, dense aggregations of other types of gelatinous zooplankton have been studied and are believed to have adaptive significance (Gershwin, 2013). It is possible that Irukandji need these dense blooms of salps and hydromedusae where competition and predation are minimal but biomass protection is maximised.

5.3. Prevention The most effective method of in-water protection is a full-body lycra ‘stinger suit’ (Dawes et al., 2006; Gershwin et al., 2009; Figure 1.12). The efficacy of different types of fabrics was tested by Gershwin and Dabinett (2009), who found that smooth fabrics with a tight weave provided the best protection. These tests used Carukia barnesi, which has the finest tentacles of any known Irukandji species; stings from other species with heavier tentacles are likely to be even more successfully prevented.

5.4. Treatment Neither a vaccine nor an antivenom currently exists for Irukandji syndrome. Treatment is largely symptom-based, that is, symptoms are treated as they arise. Biology and Ecology of Irukandji Jellyfish 69

In 2003, intravenous magnesium sulphate was first used to treat the hypertension (high blood pressure) associated with some cases of Irukandji syndrome (Corkeron, 2003). Unexpectedly, it relieved all symptoms, not just the hypertension. Since that time, much debate has ensued regarding the efficacy of magnesium, with some workers reporting total resolution of symptoms (Corkeron et al., 2004; Rathbone et al., 2013), while others find that it does not work for all stings (Little, 2005; McCullagh et al., 2012). These studies have typically not taken species differences, phylogeny, distribution, seasonality, or ontogeny into consideration. However, as with snakes and spiders, these biological and ecological factors are highly likely to govern the relative toxicity of the species we are trying to understand. Treating all species and growth stages as a uniform entity is a cumbersome and antiquated approach; far less painful and more elegant outcomes are likely to be achieved through zoological understanding. Conversely to the magnesium treatment, Hawaiian Irukandji stings are routinely treated with hot water showers (Thomas et al., 2001). Two potentially life-threatening problems arise from this. First, freshwater causes discharge of remaining nematocysts by osmotic action (Glaser and Sparrow, 1909; Grosvenor, 1903). Second, heat dilates capillaries (Jaszczak, 1988), theoretically inviting in more venom faster. Regardless of the efficacy of pain relief, hot water treatment requires further research before it can be confidently considered standard safe treatment (Gershwin et al., 2013b).

6. RESEARCH GAPS

Research on jellyfish in general, and Irukandjis in particular, has been stymied by the relatively small amount of money available (Gibbons and Richardson, 2013). Quantifying the magnitude of socio-economic impacts of blooms will provide the impetus for more directed research into Irukandji dynamics and prediction. This should be the major research priority, as it contextualizes the Irukandji problem, encourages industry and government funding and participation in research, and allows for the prioritization of research questions. Obtaining this information requires innovative collabo- rations among ecologists, economists, medical practitioners and social scientists. It also requires the use of unconventional data sources, including questionnaires to key stakeholder groups and meta-analyses of newspaper articles to estimate the scale of the problem, and analysis of hospitalization records to estimate health costs. Cost-benefit analysis of different mitigation options will be needed to identify the best management practices 70 Lisa-ann Gershwin et al. economically and environmentally. Currently, estimates of the cost of Irukandji to coastal economies are sparse and qualitative, although one estimate of losses to the tourism industry in 2002 in North Queensland due to negative publicity is around AU$65 million (Williams, 2004 in Gershwin et al., 2009). The other major gap hampering our understanding of Irukandji blooms is the lack of data. Few time series of Irukandji exist, and none of co-occurrence with indicator organisms such as salps. To better identify the environmental conditions responsible for blooms and when and where they will occur, time series and spatial surveys of Irukandji abundance are needed. Despite more than 70 years of study, our understanding of Irukandji jellyfish is still in its infancy. We have synthesised the available information on their biology and ecology, but in many ways, this raises more questions than it answers: it is certainly an interesting and wide-open field of study for the curious student of marine biology, ecology, toxinology, and taxonomy. In Table 1.8, we summarise the major gaps, questions and issues, and techniques in the hope of stimulating hypotheses for further study into these most remarkable creatures and their dramatic interface with humans worldwide.

Table 1.8 Summary of the major gaps and issues in Irukandji jellyfish studies Major discipline Gaps, questions, and issues Taxonomy Development of regional taxonomic expertise Systematic collecting in regions with unattributed stings Potential for use of statoliths for gut contents and fossil IDs First-aid research Does vinegar inactivate all Irukandji nematocysts? Define the venom dose – syndrome severity relationship Medical research Defining the links between different species and syndromes, for improved management Define the molecular mechanism/s underlying the various features of the syndrome Define and sequence the responsible venom toxins, and clone and express the relevant genes in vitro Develop a specific antivenom to neutralize the relevant toxin/s across the various responsible genera Better define the biomarkers predictive of syndrome severity Biology and Ecology of Irukandji Jellyfish 71

Table 1.8 Summary of the major gaps and issues in Irukandji jellyfish studies—cont'd Major discipline Gaps, questions, and issues Biology and Breeding grounds of polyps in the wild ecology Seasonal conditions that trigger metamorphosis Potential response to climate change Quantify the relationship between Irukandji and salps Coordination of vision without a brain Age and growth Robust studies on growth rates and longevity Ontogenetic changes in morphology and physiology Venom changes with ontogeny Ontogenetic changes in food preference Genetics Population genetics, species boundaries, and connectivity Better understanding of evolutionary history, age of group Development of tools for rapid identification Why are mitochondria linear rather than circular, and how do they duplicate? Trophic Predator/prey behavioural dynamics relationships Fatty acid and stable isotope analyses Bloom prediction Time series of abundance and spatial surveys of key sites Environmental conditions that cause infestations Socioeconomic Quantifying the magnitude of socio-economic impacts of both impacts stings and the public fear of stings

ACKNOWLEDGEMENTS We gratefully acknowledge the many collectors, research assistants, funding bodies, and people and organisations through the years who have given us specimens, notes, data, and literature relating to Irukandji, without whose help, most of the research would not have been possible. In particular, we humbly thank James Angus, Natalia Aponte, Brad Armstrong, the Australian Biological Resources Study, the Australian Institute of Marine Science, Dave Barker, Paul Barker, the family of Jack Barnes, Nick Barnes, Peter Barnes, Broome Shire Council, Machael Carlson, Michael Corkeron, the CRC Reef Research, Bart Currie, Karen Dabinett, Ian Day, Department of Parks and Wildlife (WA), Marty Durkan, Ben Eales, the Great Barrier Reef Research Foundation, Dean Harrison, Bill Horsford, James Cook University, David Kain, Ebony Keating, Mike Kingsford, Deb 72 Lisa-ann Gershwin et al.

Lewis, Ran Li, Lions Foundation, Col McKenzie, Dale Mengel, John Menico, Kim Moss, Paspaley Pearling Company and its divers and skippers, Pearl Producers Association, Robert King Memorial Foundation, Ron Pollard, Kathryn Porch, Victor Hugo Beltran Ramirez, John Rathbone, Mark Ross-Smith, Jamie Seymour, Grant Small, the family of Ron Southcott, Surf Life Saving, James Tibballs, Tim Trew, Heather Walling, Kathryn Walsh, John Williamson, Carolyn Wiltshire, Christine Wright, and Angel Yanagihara. The AVRU also gratefully acknowledges funding support from the Australian Government Department of Health and Ageing as well as from the National Health and Medical Research Council and Sutherland Trust.

APPENDIX A: NOTES ON REARING AND LIFE CYCLE OF CARUKIA BARNESI

Neither the methods nor the early life cycle stages of Irukandji have previously been described. One of us (LG) has had extensive experience collecting all stages of medusae from the wild over 10 summers and rearing Carukia barnesi in the laboratory. The following is a summary of these unpublished findings. The aquaria used as rearing chambers were plastic hamster cages with fine nylon mesh screening off the outflow in one-third of the tank; a slow but steady fall of water from small airline tubing was used to drive the circulation in the other two-thirds.

Life cycle notes The youngest medusa specimens have the appearance of tiny strawberries. They are about 1–2 mm in diameter, pyramidal to globular in shape, and dark red. The pedalia and rhopalial niches are not yet formed; four stubby tentacles mark the corners and the rhopalia are external. The reddish colour is presumed to come from the dense concentration of nematocyst batteries on the bell, which spread out as the animal grows. A few of the smallest specimens each year are caught with their ‘umbilical cord’ still attached: this is a portion of the polyp that remains still attached to the apex of the bell for the first few hours after liberation. By the time these specimens are processed some hours later, they invariably have lost this structure. Laboratory rearing of other cubozoan species has revealed that the umbilical cord is typically resorbed within 2–12 h but may take up to several days (Arneson, 1976; Arneson and Cutress, 1976; Cutress and Studebaker, 1973; Horita, 1992; Stangl et al., 2002; Straehler-Pohl and Jarms, 2005, 2011; Studebaker, 1972). Therefore, assuming that Carukia barnesi is typical within this resorp- tion range, the species must be breeding near the shore and essentially being picked up by currents and transported the short distance shoreward. Biology and Ecology of Irukandji Jellyfish 73

Hand-fed medusae change and grow rapidly. The juvenile strawberry appearance is lost within a few days, and the animals become more evenly dome-shaped and golden; the pedalia form, the tentacles lengthen, and the rhopalial niches are fully formed within a week. In this time, the animals have tripled to quadrupled in body size. Over the next several days, animals continue to grow rapidly, taking on a more transparent and sculpted appearance; the tentacles develop the handkerchief banding and grow at an astounding rate of about 2.5 cm per day (relaxed length). By 2 weeks, the tentacles can relax to about 100 cm in length and the animals have fully mature gonads. Mature specimens range in size from about 8 mm bell height to 14 mm, with the most common specimens in this size range being about 9–11 mm. After a day or two of spawning and senescing, most specimens do not survive longer than 2 weeks. Medusae were not observed to spawn. In field-collected specimens, ripe gonads are full and broad; after spawning, the gonads of both males and females are narrow. Brooding embryos have not been observed; thus, it appears that Carukia barnesi is a broadcast spawner. Planula larvae were not observed in captive cultures. Cubozoan polyps were observed simultaneously in eight aquaria on a closed system and both inflow and outflow UV sterilisation; therefore, the polyps were assumed to come from the Carukia barnesi specimens in the aquaria. The first polyps were observed while medusae were still alive in the aquaria and colonised rapidly. Within a couple of months, polyp density was estimated at 500–1000 per 24 cm2 hanging acrylic plate. Polyps readily colonised all substrates: plastic aquarium sides and bottom, hanging acrylic plates, shells, terracotta chips, glass petri dishes, and PVC connectors. Many polyps were also found adhering to the algal films and flocculants that covered most of the surfaces and fish scales on the bottom. Polyps were whitish and extremely small, to 0.6 mm in diameter. The polyp was divided into three main parts: an aboral stalk with an adhesive basal disc; a short, stocky body; and a huge conical hypostome with a terminal round mouth, with the margin between the body and hypostome bearing up to 20 solid tentacles. The hypostome was quite plastic in form, as often the polyps were observed either with the mouth raised like a smooth bell curve or with the oral disc quite flat. The oral disc was evenly smooth and glistened from a heavy speckling of nematocysts. The number of tentacles increased with body size: The smallest polyps (0.1 mm diameter) had four tentacles, whereas the largest (0.65 mm diameter) had 19–20. The tentacles were evenly spaced around the disc 74 Lisa-ann Gershwin et al. margin, each elongate–triangular in shape with a single large nematocyst embedded in the tip. Most of the time, the polyps lay in ‘fishing mode’ with the disc flat and the tentacles radiating out. When prey particles were caught, they were rapidly ‘slam-dunked’ into the mouth. Then the tentacle was quickly returned to fishing mode and the single terminal nematocyst was replaced in about 2 h by migration from the body. In the non-Irukandji cubopolyp of Carybdea, restoration of the terminal nematocyst took about 4 h (Maniura et al., 2001). Creeping polyps appeared without their development being observed. These polyps were considerably longer and larger than the sedentary polyps, up to 2 mm long. The oral disc was broadly conical with a flattened top. It bore six evenly spaced cylindrical tentacles, which were about twice the diameter of the oral disc in length. Each tentacle had a single large nematocyst in its distal end.

Feeding notes In another study involving laboratory rearing of Carukia barnesi, specimens of all sizes would accept fish and prawn food particles touched to the tentacles; however, these were eventually discarded and never ingested. Food particles offered straight to the lips on a probe were readily accepted and rapidly ingested. However, younger specimens tended to spit out fish more often than prawn and thus grow more slowly. By contrast, older specimens tended to spit out food less often but grow more rapidly and maintain a healthy appearance on fish (Gershwin, unpublished notes). When fed, laboratory specimens would typically remain fairly inactive for a brief time, either on the bottom or drifting passively. If disturbed, they would usually expel their food. Undisturbed, prawn-meat particles equal in size to about one-quarter of the stomach took about 2 h to digest, and fish sections of the same size took about two to four. After about 20–30 min, the jellyfish would begin expelling the scales through the mouth in a sparse mucous stream. Between feedings, medusae exhibited what was interpreted as foraging behaviour. Medusae hung nearly motionless several centimetres below the surface with the tentacles relaxed as a loose tangle of fine threads through the water column. Pulsation was slow and irregular. A variety of suitably sized prey items were observed to be envenomed upon contact, made apparent by the struggling movements of the prey. However, in this captive environment, food was never observed to be ingested, but rather, the medusae continued in fishing mode and occasionally food items were ensnared by more than one jellyfish simultaneously. In all cases, food was eventually discarded dead. Biology and Ecology of Irukandji Jellyfish 75

APPENDIX B: NOTES ON AUSTRALIAN ALATINA MORDENS OCCURRENCE

Research from long-term monitoring in Hawaii during the 1990s suggests that an aggregation occurs on coral reefs where oceanic Irukandji (Alatina moseri) appear 8–12 days after the full moon each month (Thomas et al., 2001). Some of these jellyfish remain behind in the lagoon and can potentially cause significant stings (Yoshimoto and Yanagihara, 2002). A review of marine sting reports from a marine operator in far north Queensland from 1996 to 2002 suggested that there was a similar, but milder, pattern with Irukandji. A collaborative research programme involving the CRC Reef Research, James Cook University, Australian Venom Research Unit, and Reef Biosearch began in 2002. Metal halide flood lights were positioned over the side of a research vessel at night with plankton nets used to retrieve specimens. Attempts were made to sample in open water in front of reef systems, but conditions were generally untenable, so to maximise catch efforts, a decision was made early to consistently sample on the leeward side of the reef. Specimens caught were identified to at least genus level, bagged, and snap frozen. These specimens were then sent to other research institutes for genetic and venom analysis and taxonomy purposes. Specimens were caught during this lunar period over most months of the year, with a larger prevalence over the summer months (December to April), but with a surprising mild spike during August to September during some years. Several species of Irukandji were collected in all weather conditions (from 0 to 40 knots), but best results were during 5–10 knot southeasterly conditions, and highest yield per night was over 70 specimens. Some specimens were caught during extreme conditions, such as an unnamed new species of Carukia caught during a cyclone, with 40 knot northwesterly winds. Irukandji appeared to be highly photopositive, actively moving into illu- minated areas, regardless of the presence or absence of prey items. Medusae attracted to light sources were nearly always at the surface of the water, especially Alatina mordens (the most common species encountered).

Alatina mordens sting Most stings from Alatina moseri in Hawaii do not produce Irukandji syndrome, but some do (Yoshimoto and Yanagihara, 2002). Similarly, only 76 Lisa-ann Gershwin et al. about 5% of stings from Alatina mordens in Australia produce systemic illness; more often, the sting produced is merely painful and localised (R. Hore, unpublished data). Some cases attributed to this species have required life support (Gershwin, 2005c). Because of the possibility of life-threatening stings in rare instances, this species should be treated with great care.

REFERENCES AAP, 2005. Jellyfish experts strike jellyfish jackpot. AAP General News Wire, Sydney, Australia. Albert, D.J., 2007. Aurelia labiata medusae (Scyphozoa) in Roscoe Bay avoid tidal dispersion by vertical migration. J. Sea Res. 57, 281–287. Aneiros, A., Garateix, A., 2004. Bioactive peptides from marine sources: pharmacological properties and isolation procedures. J. Chromatogr. B 803, 41–53. Anonymous, 2002. Toxic shock from stinger family. ABC News: Science, Available at http://www.abc.net.au/science/articles/2002/10/15/700864.htm. Anonymous, 2005. Jellyfish researchers find irukandji breeding site. ABC News, Available at http://www.abc.net.au/news/2005-12-22/jellyfish-researchers-find-irukandji-breed ing-site/766488. Anonymous, A., 2011a. Jellyfish invade Sand Bar. Cayman News Service, Available at http:// www.caymannewsservice.com/science-and-nature/2011/04/27/jellyfish-invade-sand-bar. Anonymous, 2011b. UPDATED: Jellyfish outbreak closes Sand Bar...people taken to hospital with stings. Cayman 27. Available: http://www.cayman27.com.ky/2011/04/27/ updated-jellyfish-outbreak-closes-sand-bar-people-taken-to-hospital-with-stings. Arneson, A.C., 1976. Life History of Carybdea Alata (Reynaud, 1831) (MS thesis). Univer- sity of Puerto Rico, Mayaguez. Arneson, A.C., Cutress, C.E., 1976. Life history of Carybdea alata Reynaud, 1831 (Cubomedusae). In: Mackie, G.O. (Ed.), Coelenterate Ecology and Behavior. Plenum Press, New York, pp. 227–236. A´ vila-Soria, G., 2009. Molecular Characterization of Carukia barnesi and Malo kingi. Cnidaria; Cubozoa; Carybdeidae (Ph.D. thesis). James Cook University, Townsville. Balasubramanian, P.G., Beckmann, A., Warnken, U., Schno¨lzer, M., Schu¨ler, A., Bornberg- Bauer, E., Holstein, T.W., Ozbek, S., 2012. Proteome of Hydra nematocyst. J. Biol. Chem. 287, 9672–9681. Barnes, J.H., 1960. Observations on jellyfish stingings in North Queensland. Med. J. Aust. 2, 993–999. Barnes, J.H., 1964. Cause and effect in Irukandji stingings. Med. J. Aust. 1, 897–904. Barnes, J.H., 1965. A diagnostic procedure for marine stings. Recovery of stinging capsules from skin of victims. R.A.N. Med. Newsl. 3, 15–16. Barnes, J.H., 1966. Studies on three venomous cubomedusae. In: Rees, W.J. (Ed.), The Cnidaria and their Evolution. Academic Press, London, pp. 307–332. Bateman, D., 2010a. Irukandji plague. 70 jellyfish caught in one hour. The Cairns Post, pp. 1, 8, Cairns, Qld., Australia. Bateman, D., 2010b. Study finds extra sting to irukandji scientists link species with snorkeller’s death. The Cairns Post 7, Cairns, Qld., Australia. Bentlage, B., Lewis, C., 2012. An illustrated key and synopsis of the families and genera of carybdeid box jellyfishes (Cnidaria: Cubozoa: Carybdeida), with emphasis on the “Irukandji family” (Carukiidae). J. Nat. Hist. 46, 2595–2620. Bentlage, B., Cartwright, P., Yanagihara, A.A., Lewis, C., Richards, G.S., Collins, A.G., 2010. Evolution of box jellyfish (Cnidaria: Cubozoa), a group of highly toxic invertebrates. Proc. R. Soc. B 277, 493–501. Biology and Ecology of Irukandji Jellyfish 77

Berger, E.W., 1900. Physiology and histology of the Cubomedusae, including Dr. F.S. Con- ant’s notes on the physiology. Mem. Biol. Lab. Johns Hopkins Univ. 4, 1–84. Bester, C., 2012. Don’t take the risk ‘up to 50’ irukandji caught in nets. Lifeguard anger as people defy warnings. Far North’s beaches likely to stay closed. The Cairns Post 1, Cairns, Qld., Australia. Brinkman, D.L., Aziz, A., Loukas, A., Potriquet, J., Seymour, J., Mulvenna, J., 2012. Venom proteome of the box jellyfish Chironex fleckeri. PLoS One 7, e47866. Burnett, J.W., Currie, B., Fenner, P., Rifkin, J., Williamson, J., 1996. Cubozoans (‘Box jellyfish’). In: Williamson, J., Fenner, P., Burnett, J.W., Rifkin, J. (Eds.), Venomous and Poisonous Marine Animals. University of New South Wales Press, Sydney, pp. 236–283. Burnett, J.W., Weinrich, D., Williamson, J.A., Fenner, P.J., Lutz, L.L., Bloom, D.A., 1998. Autonomic neurotoxicity of jellyfish and marine animal venoms. Clin. Auton. Res. 8, 125–130. Buskey, E.J., 2003. Behavioral adaptations of the cubozoan medusa Tripedalia cystophora for feeding on copepod (Dioithona oculata) swarms. Mar. Biol. 142, 225–232. Carrete Vega, G., Wiens, J.J., 2012. Why are there so few fish in the sea? Proc. R. Soc. B 279, 2323–2329. Carrette, T., Alderslade, P., Seymour, J., 2002. Nematocyst ratio and prey in two Australian cubomedusans, Chironex fleckeri and Chiropsalmus sp. Toxicon 40, 1547–1551. Cheng, A.C., Winkel, K.D., Hawdon, G.M., McDonald, M., 1999. Irukandji-like syndrome in Victoria. Aust. N. Z. J. Med. 29, 835. Cleland, J.B., Southcott, R.V., 1965. Injuries to Man from Marine Invertebrates in the Australian Region. Commonwealth of Australia, Canberra. Coates, M.M., 2003. Visual ecology and functional morphology of Cubozoa (Cnidaria). Integr. Comp. Biol. 43, 542–548. Coates, M.M., Garm, A., Theobald, J.C., Thompson, S.H., Nilsson, D.-E., 2006. The spec- tral sensitivity of the lens eyes of a box jellyfish, Tripedalia cystophora (Conant). J. Exp. Biol. 209, 3758–3765. Conant, F.S., 1898. The Cubomedusae. A Dissertation Presented for the Degree of Doctor of Philosophy, in the Johns Hopkins University, 1897. Johns Hopkins Press, Baltimore. Corkeron, M.A., 2003. Magnesium infusion to treat Irukandji syndrome. Med. J. Aust. 178, 411. Corkeron, M., Pereira, P., Macrokanis, C., 2004. Early experience with magnesium admin- istration in Irukandji syndrome. Anaesth. Intensive Care 32, 666–669. Crow, G.L., Holland, B.S., Blair, L., Kaneshiro-Pineiro, M.Y., Bridges, K.W., Goto, R.S., Yanagihara, A., 2010. Historical records, Waikiki beach counts, and public advisories of box jellyfish Alatina moseri in Hawaii. In: Third International Jellyfish Blooms Sympo- sium, Mar del Plata, Argentina, p. 32. Currie, B., Wood, Y.K., 1995. Identification of Chironex fleckeri envenomation by nematocyst recovery from skin. Med. J. Aust. 162, 478–480. Cutress, C.E., Studebaker, J.P., 1973. Development of the Cubomedusae, Carybdea marsupialis. Proc. Assoc. Isl. Mar. Labs. Carib. 9, 25. Cuypers, E., Cuypers, E., Yanagihara, A., Karlsson, E., Tytgata, J., 2006. Jellyfish and other cnidarian envenomations cause pain by affecting TRPV1 channels. FEBS Lett. 580, 5728–5732. Dawes, P., Fenner, P., Gershwin, L., Gage, G., Drake, L., Small, G., Moss, K., 2006. Marine Stinger Risk Management Guidelines. Surf Life Saving Queensland, Brisbane. de Pender, A.M.G., Winkel, K.D., Ligthelm, R.J., 2006. A probable case of Irukandji syndrome in Thailand. J. Travel Med. 13, 240–243. Delfı´n, J., Gonza´lez, Y., Dı´az, J., Cha´vez, M., 1994. Proteinase inhibitor from Stichodactyla helianthus: purification, characterization and immobilization. Arch. Med. Res. 25 (2), 199–204. 78 Lisa-ann Gershwin et al.

Endean, T.R., Rifkin, J., 1975. Isolation of different types of nematocyst from the cubomedusan Chironex fleckeri. Toxicon 13, 375–376. Evans, N.T., 2010. Krestovik Medusa sting on Cape Cod. RIBIRDS Yahoo Groups Listserve. Available at http://groups.yahoo.com/group/RIBIRDS/message/5528. Fenner, P.J., 1988. Irukandji – a “new” danger. In: Pearn, J., Covacevich, J. (Eds.), Venoms and Victims. The Queensland Museum and Amphion Press, Brisbane, pp. 25–30. Fenner, P.J., 2000. The toxicology and taxonomy of the Irukandji (C. barnesi) jellyfish: report of a study in progress. ACTM Bulletin, 2–3. Fenner, P.J., 2006. Jellyfish responsible for Irukandji syndrome. Q. J. Med. 99, 802–803. Fenner, P.J., Carney, I., 1999. The Irukandji syndrome: a devastating syndrome caused by a north Australian jellyfish. Aust. Fam. Physician 28, 1131–1137. Fenner, P.J., Carney, I., 2001. The Irukandji syndrome: a devastating syndrome caused by a north Australian jellyfish. An updated review with symptoms including cardiac failure. ACTM Bulletin [Suppl.] 10, 3–4. Fenner, P.J., Hadok, J.C., 2002. Fatal envenomation by jellyfish causing Irukandji syndrome. Med. J. Aust. 177, 362–363. Fenner, P.J., Harrison, S.L., 2000. Irukandji and Chironex fleckeri jellyfish envenomation in tropical Australia. Wilderness Environ. Med. 11, 223–240. Fenner, P.J., Heazlewood, R.J., 1997. Papilloedema and coma in a child: undescribed symptoms of the “Irukandji” syndrome. Med. J. Aust. 167, 650. Fenner, P.J., Lewin, M., 2003. Sublingual glyceryl trinitrate as prehospital treatment for hypertension in Irukandji syndrome. Med. J. Aust. 179, 655. Fenner, P.J., Lippmann, J., 2009. Severe Irukandji-like jellyfish stings in Thai waters. Diving Hyperb. Med. 39, 175–177. Fenner, P.J., Rodgers, D., Williamson, J.A., 1986. Box jellyfish antivenom and “Irukandji” stings. Med. J. Aust. 144, 665–666. Fenner, P.J., Williamson, J.A., Burnett, J.W., Colquhoun, D.M., Godfrey, S., Gunawardane, K., Murtha, W., 1988. The “Irukandji syndrome” and acute pulmonary oedema. Med. J. Aust. 149, 150–156. Fenner, P.J., Williamson, J.A., Burnett, J.W., Rifkin, J., 1993. First aid treatment of jellyfish stings in Australia: response to a newly differentiated species. Med. J. Aust. 158, 498–501. Fenner, P., Lippmann, J., Gershwin, L.-a., 2010. Fatal and non-fatal severe jellyfish stings in Thai waters. J. Travel Med. 17, 133–138. Flecker, H., 1945a. Injuries by unknown agents to bathers in north Queensland. Med. J. Aust. 1, 98. Flecker, H., 1945b. Injuries by unknown agents to bathers in north Queensland. Med. J. Aust. 1, 417. Flecker, H., 1952. Irukandji sting to North Queensland bathers without production of weals but with severe general symptoms. Med. J. Aust. 19, 89–91. Flecker, H., 1957a. Further notes on Irukandji stings. Med. J. Aust. 5, 9. Flecker, H., 1957b. Injuries produced by marine organisms in tropical Australia. Med. J. Aust. 1957, 556. Fraser Coast Chronicle, 2013. Two more fall victim to suspected Irukandji stings on Fraser. Available at http://www.frasercoastchronicle.com.au/news/two-more-fall-victim-to- suspected-irukandji-sting/1708613/. Fuller, B., 2011. Sandbar re-opened after eight hospitalised for jellyfish stings: stingray inci- dent reported as well. Caymanian Compass, Available at http://www.compasscayman. com/caycompass/2011/04/27/Sandbar-re-opened-after-eight-hospitalised-for- jellyfish-stings/. Gaillard, C., Goy, J., Bernier, P., Bourseau, J.P., Gall, J.C., Barale, G., Buffetaut, E., Wenz, S., 2006. New jellyfish taxa from the Upper Jurassic lithographic limestones of Cerin (France): taphonomy and ecology. Palaeontology (Oxford) 49, 1287–1302. Biology and Ecology of Irukandji Jellyfish 79

Garm, A., Coates, M.M., Gad, R., Seymour, J., Nilsson, D.-E., 2007a. The lens eyes of the box jellyfish Tripedalia cystophora and Chiropsalmus sp. are slow and color-blind. J. Comp. Physiol. A 193, 547–557. Garm, A., O’Connor, M., Parkefelt, L., Nilsson, D.-E., 2007b. Visually guided obstacle avoidance in the box jellyfish Tripedalia cystophora and Chiropsella bronzie. J. Exp. Biol. 210, 3616–3623. Garm, A., Andersson, F., Nilsson, D.E., 2008. Unique structure and optics of the lesser eyes of the box jellyfish Tripedalia cystophora. Vision Res. 48, 1061–1073. Garm, A., Oskarsson, M., Nilsson, D.-E., 2011. Box jellyfish use terrestrial visual cues for navigation. Curr. Biol. 21, 798–803. Garm, A.L., Bielecki, J., Petie, R., Nilsson, D.-E., 2012. Opposite patterns of diurnal activity in the box jellyfish Tripedalia cystophora and Copula sivickisi. Biol. Bull. 222, 35–45. Gershwin, L., 2005a. Two new species of jellyfishes (Cnidaria: Cubozoa: Carybdeida) from tropical Western Australia, presumed to cause Irukandji syndrome. Zootaxa 1084, 1–30. Gershwin, L., 2005b. Carybdea alata auct. and Manokia stiasnyi, reclassification to a new family with description of a new genus and two new species. Mem. Queensl. Mus. 51, 501–523. Gershwin, L., 2005c. Taxonomy and Phylogeny of Australian Cubozoa (Ph.D. thesis). James Cook University, Townsville, Queensland. Gershwin, L., 2006a. Nematocysts of the Cubozoa. Zootaxa 1232, 1–57. Gershwin, L., 2006b. Jellyfish responsible for Irukandji syndrome. Q. J. Med. 99, 801–802. Gershwin, L., 2007. Malo kingi: a new species of Irukandji jellyfish (Cnidaria: Cubozoa: Carybdeida), possibly lethal to humans. Zootaxa 1659, 55–68. Gershwin, L., 2008. Morbakka fenneri: a new genus and species of Irukandji jellyfish (Cnidaria: Cubozoa). Mem. Queensl. Mus. 54, 23–33. Gershwin, L., 2013. Stung! On Jellyfish Blooms and Changing Oceans. University of Chicago Press, Chicago. Gershwin, L., Alderslade, P., 2005. A new genus and species of box jellyfish (Cubozoa: Carybdeida) from tropical Australian waters. The Beagle 21, 27–36. Gershwin, L., Dabinett, K., 2009. Comparison of eight types of protective clothing against Irukandji jellyfish stings. J. Coast. Res. 25, 117–130. Gershwin, L.-a., Dawes, P., 2008. Preliminary observations on the response of Chironex fleckeri (Cnidaria: Cubozoa: Chirodropida) to different colors of light. Biol. Bull. (Woods Hole) 215, 57–62. Gershwin, L., De Nardi, M., Fenner, P.J., Winkel, K.D., 2009. Marine stingers: review of an under-recognized global coastal management issue. Coast. Manage. 38, 22–41. Gershwin, L.-a., Richardson, A., Strzelecki, J., Condie, S., 2012. Irukandji forecasting work- shop report. CSIRO Marine and Atmospheric Research, Hobart, Tasmania. Gershwin, L.-a., Richardson, A.J., Mansbridge, J.V., Condie, S.A., 2013a. Dangerous jellyfish blooms are predictable. J. R. Soc. Interface. Gershwin, L.-a., Tibballs, J., Bridgewater, F., 2013b. A critical review of the role of hot water in the first aid management of jellyfish stings (Cnidaria: Cubozoa; Scyphozoa; Hydrozoa: Siphonophora). Wilderness Environ. Med. Glaser, O.C., Sparrow, C.M., 1909. The physiology of nematocysts. J. Exp. Zool. 6, 361–382. Goggin, L., Gershwin, L., Fenner, P., Seymour, J., Carrette, T., 2004. Stinging Jellyfish in Tropical Australia. CRC Reef Research Centre, Townsville, Queensland. Gordon, M., Seymour, J., 2012. Growth, development and temporal variation in the onset of six Chironex fleckeri medusae seasons: a contribution to understanding jellyfish ecology. PLoS One 7, e31277. Gordon, M., Hatcher, C., Seymour, J., 2004. Growth and age determination of the tropical Australian cubozoan Chiropsalmus sp. Hydrobiologia 530/531, 339–345. Grady, J.D., Burnett, J.W., 2003. Irukandji-like syndrome in South Florida divers. Ann. Emerg. Med. 42, 763–766. 80 Lisa-ann Gershwin et al.

Grosvenor, H., 1903. On the nematocysts of aeolids. P. R. Soc. London 72, 462–486. Gusmani, L., Avian, M., Galil, B., Patriarca, P., Rottini, G., 1997. Biologically active poly- peptides in the venom of the jellyfish Rhopilema nomadica. Toxicon 35, 637–648. Hadok, J.C., 1997. “Irukandji” syndrome: a risk for divers in tropical waters. Med. J. Aust. 167, 649–650. Hamner, W.M., Jones, M.S., Hamner, P.P., 1995. Swimming, feeding, circulation and vision in the Australian box jellyfish, Chironex fleckeri (Cnidaria, Cubozoa). Mar. Freshw. Res. 46, 985–990. Hartwick, R.F., 1991a. Observations on the anatomy, behaviour, reproduction and life cycle of the cubozoan Carybdea sivickisi. Hydrobiologia 216/217, 171–179. Hartwick, R.F., 1991b. Distributional ecology and behaviour of the early life stages of the box-jellyfish Chironex fleckeri. Hydrobiologia 216/217, 181–188. Harriott, V.J., 2002. Marine tourism impacts and their management on the Great Barrier Reef. CRC Reef Research Centre Technical Report No 46. CRC Reef Research Centre, Townsville, 41 pp. Herceg, I., 1987. Pulmonary oedema following an Irukandji sting. SPUMS J. 17, 95–96. Horita, T., 1992. A cubopolyp and its early metamorphosis of cubomedusa in Japan. In: AAZPA Annual Conference Proceedings, pp. 350–355. Hughes, R.J.A., Angus, J.A., Winkel, K.D., Wright, C.E., 2012. A pharmacological investigation of the venom extract of the Australian box jellyfish, Chironex fleckeri, in cardiac and vascular tissues. Toxicol. Lett. 209, 11–20. Huynh, T.T., Seymour, J., Pereira, P., Mulcahy, R., Cullen, P., Carrette, T., Little, M., 2003. Severity of Irukandji syndrome and nematocyst identification from skin scrapings. Med. J. Aust. 178, 38–41. Jaszczak, P., 1988. Blood flow rate, temperature, oxygen tension and consumption in the skin of adults measured by a heated microcathode oxygen electrode. Dan. Med. Bull. 35, 322–334. Johnson, R.G., Richardson Jr., E.S., 1966. A remarkable Pennsylvanian fauna from the Mazon Creek area, Illinois. J. Geol. 74, 626–631. Johnson, R.G., Richardson Jr., E.S., 1968. The Essex fauna and medusae. Fieldiana (Geology) 12, 109–115. Kawamura, M., Ueno, S., Iwanaga, S., Oshiro, N., Kubota, S., 2003. The relationship between fine rings in the statolith and growth of the cubomedusa Chiropsalmus quadrigatus (Cnidaria: Cubozoa) from Okinawa Island, Japan. Plankton Biol. Ecol. 50, 37–42. Kingsford, M.J., Seymour, J.E., O’Callaghan, M.D., 2012. Abundance patterns of cubozoans on and near the Great Barrier Reef. Hydrobiologia 690, 257–268. Kinsey, B.E., 1986. Barnes on Box Jellyfish. Sir George Fisher Centre for Tropical Marine Studies, James Cook University, Townsville. Kinsey, B.E., 1988. More Barnes on Box Jellyfish. Sir George Fisher Centre for Tropical Marine Studies, James Cook University, Townsville. Kishinouye, K., 1910. Some medusae of Japanese waters. J. Coll. Sci. Tokyo 27, 1–35, 5 plates. Kramp, P.L., 1961. Synopsis of the medusae of the world. J. Mar. Biol. Assoc. U.K. 40, 1–469. Kramp, P.L., 1970. Marine biological investigations in the Bahamas. 16. Some medusae from the Bahamas. Sarsia 44, 59–68. Kreifels, S., 1997. Tourists take stings in stride. Honolulu Star-Bulletin. Available at http:// archives.starbulletin.com/97/07/30/news/story3.html. Larson, R.J., 1976. Cubomedusae: feeding – functional morphology, behavior and phylogenetic position. In: Mackie, G.O. (Ed.), Coelenterate Ecology and Behavior. Plenum, New York, pp. 237–245. Laska, V.G., Hu¨ndgen, M., 1982. The morphology and fine structure of the eyes of Tripedalia cystophora Conant (Cnidaria, Cubozoa). Zool. Jahrb. Anat. 108, 107–123, 25 figs. Biology and Ecology of Irukandji Jellyfish 81

LeMay,R.,2013.IrukandjijellyfishstingspateoffWA.AAP.Availableathttp://news.smh.com. au/breaking-news-national/irukandji-jellyfish-sting-spate-off-wa-20130507-2j5ni.html. Lewis, C., Long, T.A.F., 2005. Courtship and reproduction in Carybdea sivickisi (Cnidaria: Cubozoa). Mar. Biol. 147, 477–483. Lewis, C., Kubota, S., Migotto, A.E., Collins, A.G., 2008. Sexually dimorphic cubomedusa Carybdea sivickisi (Cnidaria: Cubozoa) in Seto, Wakayama, Japan. Pub. Seto Mar. Biol. Lab. 40, 1–8. Li, R., Wright, C.E., Winkel, K.D., Gershwin, L.A., Angus, J.A., 2011. The pharmacology of Malo maxima jellyfish venom extract in isolated cardiovascular tissues: a probable cause of the Irukandji syndrome in Western Australia. Toxicol. Lett. 201, 221–229. Light, S.F., 1914. Some Philippine Scyphomedusae, including two new genera, five new species, and one new variety. Philipp. J. Sci. 9, 195–231. Light, S.F., 1921. Further notes on Philippine scyphomedusan jellyfishes. Philipp. J. Sci. 18, 25–32. Lippmann, J.M., Fenner, P.J., Winkel, K., Gershwin, L.-a., 2011. Fatal and severe box jellyfish stings, including Irukandji stings, in Malaysia, 2000–2010. J. Travel Med. 18, 275–281. Little, M., 2005. Failure of magnesium in treatment of Irukandji syndrome. Anaesth. Inten- sive Care 33, 541–542. Little, M., Mulcahy, R.F., 1998. A year’s experience of Irukandji envenomation in far north Queensland. Med. J. Aust. 169, 638–641. Little, M., Mulcahy, R.F., Wenck, D.J., 2001. Life-threatening cardiac failure in a healthy young female with Irukandji syndrome. Anaesth. Intensive Care 29, 178–180. Little, M., Pereira, P., Mulcahy, R., Cullen, P., Carrette, T., Seymour, J., 2003. Severe cardiac failure associated with presumed jellyfish sting. Irukandji syndrome? Anaesth. Inten- sive Care 31, 642–647. Little, M., Pereira, P., Carrette, T., Seymour, J., 2006. Jellyfish responsible for Irukandji syndrome. Q. J. Med. 99, 425–427. Lord, R.E., Wilks, S.L.B., 1918. A case of jelly-fish sting simulating an acute abdomen. The Lancet 192, 390. Mackie, G.O., Larson, R.J., Larson, K.S., Passano, L.M., 1981. Swimming and vertical migration of Aurelia aurita (L) in a deep tank. Mar. Behav. Physiol. 7, 321–329. Macrokanis, C.J., Hall, N.L., Mein, J.K., 2004. Irukandji syndrome in northern Western Australia: an emerging health problem. Med. J. Aust. 181, 699–702. Malej, A., Turk, V., Lucˇic´, D., Benovic´, A., 2007. Direct and indirect trophic interactions of Aurelia sp. (Scyphozoa) in a stratified marine environment (Mljet Lakes, Adriatic Sea). Mar. Biol. 151, 827–841. Maniura, M., Hanke-Buecker, G., Golz, R., 2001. Restoration of the nematocyte/sensory cell-complex in the tentacles of Carybdea marsupialis after nematocyst discharge. Zoology (Jena) 103, 110(Abstract). Martin, V.J., 2004. Photoreceptors of cubozoan jellyfish. Hydrobiologia 530/531, 135–144. Martin, J.C., Audley, I., 1990. Cardiac failure following Irukandji envenomation. Med. J. Aust. 153, 164–166. Matsumoto, G.I., 1995. Observations on the anatomy and behaviour of the cubozoan Carybdea rastonii Haacke. Mar. Freshw. Behav. Phy. 26, 139–148. Mayer, A.G., 1910. The Scyphomedusae. Medusae of the World, vol. 3. Carnegie Institution, Washington, D.C. McCullagh, N., Pereira, P., Cullen, P., Mulcahy, R., Bonin, R., Little, M., Gray, S., Seymour, J., 2012. Randomised trial of magnesium in the treatment of Irukandji syndrome. Emerg. Med. Australas. 24, 560–565. McKechnie, K., 2010. Marine killers heading south. ABC 7:30 report, Sydney, Australia. Available at http://www.abc.net.au/7.30/content/2010/s2790871.htm. 82 Lisa-ann Gershwin et al.

Morales-Landa, J.L., Zapata-Pe´rez, O., Cedillo-Rivera, R., Segura-Puertas, L., Sima´- Alvarez, R., Sańchez-Rodrı´guez, J., 2007. Antimicrobial, antiprotozoal, and toxic activities of cnidarian extracts from the Mexican Caribbean Sea. Pharm. Biol. 45, 37–43. Mulcahy, R., Little, M., 1997. Thirty cases of Irukandji envenomation from far north Queensland. Emerg. Med. 9, 297–299. Nickson, C.P., Waugh, E.B., Jacups, S.P., Currie, B.J., 2009. Irukandji syndrome case series from Australia’s tropical northern territory. Ann. Emerg. Med. 54, 395–403. Nilsson, D.-E., Gisleń, L., Coates, M.M., Skogh, C., Garm, A., 2005. Advanced optics in a jellyfish eye. Nature 435, 201–205. O’Connor, M., Garm, A., Nilsson, D.-E., 2009. Structure and optics of the eyes of the box jellyfish Chiropsella bronzie. J. Comp. Physiol. A Neuroethol. Sens. Neural Behav. Phy- siol. 195, 557–569. O’Connor, M., Garm, A., Marshall, J.N., Hart, N.S., Ekstrom, P., Skogh, C., Nilsson, D.-E., 2010. Visual pigment in the lens eyes of the box jellyfish Chiropsella bronzie. Proc. Roy. Soc. London, Ser. B 277, 1843–1848. Oba, A., Hidaka, M., Iwanaga, S., 2004. Nematocyst composition of the cubomedusan Chiropsalmus quadrigatus changes with growth. Hydrobiologia 530/531, 173–177. Okada, Y.K., 1927. Note sur l’ontogeńie de Carybdea rastonii Haacke. Bulletin Biologique de la France Tome 61, 241–249. Old, H.H., 1908. A report of several cases with unusual symptoms caused by some unknown variety of jellyfish. Philippine J. Sci. B 3, 329–333. Old, H.H., 1912. Additional report of several cases with unusual symptoms caused by some unknown variety of jellyfish. U.S. Nav. Med. Bull. 6, 377–380. Otsuru, M., Sekikawa, H., Hiroii, Y., Suzuki, T., Sato, Y., Shiraki, T., Nagashima, Y., 1974. Observations on the sting occurring among swimmers in the rocky seashore. Jpn. J. Zool. 24, 225–235 [In Japanese]. Aquat. Sci. Fish. Abst. 1974: 4Q10805M [In English]. Ovchinnikova, T.V., Balandin, S.V., Aleshina, G.M., Tagaev, A.A., Leonova, Y.F., Krasnodembsky, E.D., Men’shenin, A.V., Kokryakov, V.N., 2006. Aurelin, a novel antimicrobial peptide from jellyfish Aurelia aurita with structural features of defensins and channel-blocking toxins. Biochem. Biophys. Res. Commun. 348, 514–523. Parsons, L., 2013. Stingers off Cairns are such a drag. The Cairns Post. http://www.cairns. com.au/article/2013/10/01/248590_local-news.html. Pearse, J.S., Pearse, V.B., 1978. Vision in cubomedusan jellyfishes. Science 199, 458. Pereira, P., Barry, J., Corkeron, M., Keir, P., Little, M., Seymour, J., 2010. Intracranial haemorrhage and death after envenoming by the jellyfish Carukia barnesi. Clin. Toxicol. 48, 390–392. Pommier, P., Coulange, M., Haro, L.D., 2005. Systemic envenomation by jellyfish in Guadeloupe: Irukandji-like syndrome. Med. Trop. 65, 367–369. Purcell, J.E., Brown, E.D., Stokesbury, K.D.E., Haldorson, L.H., Shirley, T.C., 2000. Aggregations of the jellyfish Aurelia labiata: abundance, distribution, association with age–0 walleye pollock, and behaviors promoting aggregation in Prince William Sound, Alaska, USA. Mar. Ecol. Prog. Ser. 195, 145–158. Putnam, N.H., Srivastava, M., Hellsten, U., Dirks, B., Chapman, J., Salamov, A., Terry, A., Shapiro, H., Lindquist, E., Kapitonov, V.V., Jurka, J., Genikhovich, G., Grigoriev, I.V., Lucas, S.M., Steele, R.E., Finnerty, J.R., Technau, U., Martindale, M.Q., Rokhsar, D.S., 2007. Sea anemone genome reveals ancestral eumetazoan gene repertoire and genomic organization. Science 317, 86–94. Ramasamy, S., Isbister, G.K., Seymour, J.E., Hodgson, W.C., 2005. The in vivo cardiovascular effects of the Irukandji jellyfish (Carukia barnesi) nematocyst venom and a tentacle extract in rats. Toxicol. Lett. 155, 135–141. Rathbone, J., Quinn, J., Rashford, S., 2013. Response to ‘Randomised trial of magnesium in the treatment of Irukandji syndrome’. Emerg. Med. Australas. 25, 97–98. Biology and Ecology of Irukandji Jellyfish 83

Roberts, G., 2002. Tiny harbinger of death lurking in our seas. Sydney Morning Herald, News and Features. http://www.smh.com.au/articles/2002/04/19/1019020708284. html. Salam, A.M., Albinali, H.A., Gehani, A.A., Suwaidi, J.A., 2003. Acute myocardial infarction in a professional diver after jellyfish sting. Mayo Clin. Proc. 78, 1557–1560. Sando, J.J., Usher, K., Buettner, P., 2010. ‘To swim or not to swim’: the impact of jellyfish stings causing Irukandji syndrome in Tropical Queensland. J. Clin. Nurs. 19, 109–117. Satterlie, R.A., 1979. Central control of swimming in the cubomedusan jellyfish Carybdea rastonii. J. Comp. Physiol. A Neuroethol. Sens. Neural Behav. Physiol. 133, 357–367. Satterlie, R.A., 2002. Neuronal control of swimming in jellyfish: a comparative story. Can. J. Zool. 80, 1654–1669. Smith, D.R., Kayal, E., Yanagihara, A.A., Collins, A.G., Pirro, S., Keeling, P.J., 2012. First complete mitochondrial genome sequence from a box jellyfish reveals a highly fragmented linear architecture and insights into telomere evolution. Genome Biol. Evol. 4, 52–58. Southcott, R.V., 1952. Fatal stings to north Queensland bathers. Med. J. Aust. 272–274. Southcott, R.V., 1959. Tropical jellyfish and other marine stingings. Mil. Med. 124, 569–579. Southcott, R.V., 1967. Revision of some Carybdeidae (Scyphozoa: Cubomedusae), including a description of the jellyfish responsible for the “Irukandji syndrome” Aust. J. Zool. 15, 651–671. Southcott, R.V., Kingston, C.W., 1959. Lethal jellyfish stings: a study in “sea wasps” Med. J. Aust. 1, 443–444. Southcott, R.V., Powys, N.S., 1944. Marine stingings in North Queensland. pp. 1–37 Unpublished manuscript, available through James Cook University Library. Stangl, K., Salvini-Plawen, L.V., Holstein, T.W., 2002. Staging and induction of medusa metamorphosis in Carybdea marsupialis (Cnidaria, Cubozoa). Vie Milieu 52, 131–140. Stenning, A.E., 1928. Poisoning by Trachymedusae. Med. J. Aust. 1, 568–569. Stewart, S.E., 1996. Field behavior of Tripedalia cystophora (Class Cubozoa). Mar. Freshw. Behav. Phy. 27, 175–188. Stewart, S.E., 1997. The role of vision in the behavior of the medusa Tripedalia cystophora Conant (Cnidaria, Cubozoa) (Ph.D. thesis). The University of Texas at Austin. Straehler-Pohl, I., Jarms, G., 2005. Life cycle of Carybdea marsupialis Linnaeus, 1758 (Cubozoa, Carybdeidae) reveals metamorphosis to be a modified strobilation. Mar. Biol. (Berlin) 147, 1271–1277. Straehler-Pohl, I., Jarms, G., 2011. Morphology and life cycle of Carybdea morandinii, sp. nov. (Cnidaria), a cubozoan with zooxanthellae and peculiar polyp anatomy. Zootaxa 2755, 36–56. Studebaker, J.P., 1972. Development of the cubomedusa, Carybdea marsupialis (MSc thesis). University of Puerto Rico, Mayaguez. Talvinen, K.A., Nevalainen, T.J., 2002. Cloning of a novel phospholipase A2 from the cnidarian Adamsia carciniopados. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 132, 571–578. Thomas, C.S., Scott, S.A., Galanis, D.J., Goto, R.S., 2001. Box jellyfish (Carybdea alata)in Waikiki: their influx cycle plus the analgesic effect of hot and cold packs on their stings to swimmers at the beach: a randomized, placebo-controlled, clinical trial. Hawaii Med. J. 60, 100–107. Tibballs, J., Hawdon, G., Winkel, K.D., Wiltshire, C., Lambert, G., Gershwin, L.A., Fenner, P.J., Angus, J.A., 2000. The in vivo cardiovascular effects of Irukandji (Carukia barnesi) venom. In: Proceedings of the XIIIth Congress of the International Society on Toxinology, Paris, FranceP276, (abstract). Tibballs, J., Li, R., Tibballs, H.A., Gershwin, L., Winkel, K.D., 2012. Australian carybdeid jellyfish causing “Irukandji syndrome”. Toxicon 59 (6), 617–625. Tyson, E.B., 1957. Reaction to tropical jellyfish sting. Mil. Med. 120, 216–217. 84 Lisa-ann Gershwin et al.

Ueno, S., Imai, C., Mitsutani, A., 1995. Fine growth rings found in statolith of a cubomedusa Carybdea rastoni. J. Plankton Res. 17, 1381–1384. Ueno, S., Imai, C., Mitsutani, A., 1997. Statolith formation and increment in Carybdea rastoni Haacke, 1886 (Scyphozoa: Cubomedusae): evidence of synchronization with semilunar rhythms. In: Proceedings of the 6th International Conference on Coelenterate Biology, 1995, pp. 491–496. Ueno, S., Mitsumori, S.-I., Noda, M., Ikeda, I., 2000. Effect of comparative lightness of obstacles on swimming behavior of Carybdea rastoni (Cnidaria; Cubozoa). J. Nat. Fish. Univ. 48, 255–258 (in Japanese, with English abstract). Underwood, A.H., Seymour, J.E., 2007. Venom ontogeny, diet and morphology in Carukia barnesi, a species of Australian box jellyfish that causes Irukandji syndrome. Toxicon 49, 1073–1082. Werner, B., 1973. New investigations on systematics and evolution of the class Scyphozoa and the phylum Cnidaria. Pub. Seto Mar. Biol. Lab. 20, 35–61. Werner, B., 1983. Die metamorphose des polypen von Tripedalia cystophora (Cubozoa, Carybdeidae) in die meduse. Metamorphosis of the polyp of Tripedalia cystophora (Cubozoa, Carybdeidae) into the medusa. Helgoland wiss Meer. 36, 257–276 [in German, with English abstract]. Werner, B., Cutress, C.E., Studebaker, J.P., 1971. Life cycle of Tripedalia cystophora Conant (Cubomedusae). Nature, Lond. 232, 582–583. Weston, A.J., Dunlap, W.C., Shick, J.M., Klueter, A., Iglic, K., Vukelic, A., Starcevic, A., Ward, M., Wells, M.L., Trick, C.G., Long, P.F., 2012. A profile of an endosymbiont- enriched fraction of the coral Stylophora pistillata reveals proteins relevant to microbial- host interactions. Mol. Cell. Proteomics 11, M111.015487. Weston, A.J., Chung, R., Dunlap, W.C., Morandini, A.C., Marques, A.C., Moura- da-Silva, A.M., Ward, M., Padilla, G., Lang da Silva, F., Andreakis, N., Long, F.P., 2013. Proteomic characterisation of toxins isolated from nematocysts of the South Atlantic jellyfish Olindias sambaquiensis. Toxicon 71, 11–17. WHO, 2003. Guidelines for Safe Recreational Water Environments. Coastal and Fresh Waters, vol. 1. World Health Organization, Geneva. WHO, 2009. Addendum to the WHO Guidelines for Safe Recreational Water Environ- ments, vol. 1. World Health Organization, Geneva. Williams, R., 2004. Update on Irukandji issues. In: Report Presented to the Queensland Government Irukandji Task Force Meeting, Townsville, Queensland, 24 February. Williamson, J.A., Fenner, P.J., Burnett, J.W., Rifkin, J. (Eds.), 1996. Venomous and Poison- ous Marine Animals: A Medical and Biological Handbook. NSW University Press, Sydney, Australia, pp. 1–504. Wiltshire, C.J., Sutherland, S.K., Fenner, P.J., Young, A.R., 2000. Optimization and preliminary characterization of venom isolated from 3 medically important jellyfish: the box (Chironex fleckeri), Irukandji (Carukia barnesi), and blubber (Catostylus mosaicus) jellyfish. Wilderness Environ. Med. 11, 241–250. Winkel, K.D., Christopoulos, A., Coles, P., Wiltshire, C., Gershwin, L.A., Fenner, P.J., Angus, J.A., 2000. Irukandji (Carukia barnesi) venom contains a potent neuronal sodium channel agonist. In: Proceedings of the XIIIth Congress of the International Society on Toxinology, Paris, FranceP83, (abstract). Winkel, K.D., Hawdon, G.M., Fenner, P.J., Gershwin, L.-a., Collins, A.G., Tibballs, J., 2003. Jellyfish antivenoms: past, present, and future. J. Toxicol. Toxin Rev. 22, 115–127. Winkel, K.D., Tibballs, J., Molenaar, P., Lambert, G., Coles, P., Ross-Smith, M., Wiltshire, C., Fenner, P.J., Gershwin, L.-A., Hawdon, G.M., Wright, C.E., Angus, J.A., 2005. The cardiovascular actions of the venom from the Irukandji (Carukia barnesi) jellyfish: effects in human, rat and guinea pig tissues in vitro, and in pigs in vivo. Clin. Exp. Pharmacol. Physiol. 32, 777–788. Biology and Ecology of Irukandji Jellyfish 85

Winter, K.L., Isbister, G.K., Schneider, J.J., Konstantakopoulos, N., Seymour, J.E., Hodgson, W.C., 2008. An examination of the cardiovascular effects of an ‘Irukandji’ jellyfish, Alatina nr mordens. Toxicol. Lett. 179, 118–123. Yakovlev, Y.M., Vaskovsky, V.E., 1993. The toxic krestovik medusa Gonionemus vertens. Russ. J. Mar. Biol. 19, 287–294. Yamaguchi, M., 1982. Cubozoans and their life histories. Aquabiology (Tokyo) 4, 248–254. Yamaguchi, M., Hartwick, R., 1980. Early life history of the sea wasp, Chironex fleckeri (Class Cubozoa). In: Tardent, P., Tardent, R. (Eds.), Developmental and Cellular Biology of Coelenterates. Elsevier/North Holland Biomedical Press, Amsterdam, pp. 11–16. Yamasu, T., Yoshida, M., 1976. Fine structure of complex ocelli of a cubomedusan, Tamoya bursaria Haeckel. Cell Tissue Res. 170, 325–339. Yasuda, T., 1973. Ecological studies on the jellyfish, Aurelia aurita (Linne), in Urazoko Bay, Fukui Prefecture – VIII. Diel vertical migration of the medusa in early Fall, 1969. Pub. Seto Mar. Biol. Lab. 20, 491–500. Yatsu, N., 1917. Notes on the physiology of Charybdea rastonii. J. Coll. Sci. Tokyo 40, 1–12. Yoshimoto, C.M., Yanagihara, A.A., 2002. Cnidarian (coelenterate) envenomations in Hawai’i improve following heat application. Trans. R. Soc. Trop. Med. Hyg. 96, 300–303.