Distribution, population characteristics and trophic ecology of a sulphophilic hydrothermal vent tonguefish (Pleuronectiformes: Cynoglossidae)

(1)

By Jennifer Tyler

B.Sc, McGill University, 2005

A thesis submitted in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE

in the Department of Biology

 Jennifer Tyler, 2008 University of Victoria

(2)

Distribution, Population Characteristics and Trophic Ecology of a Sulphophilic Hydrothermal Vent Tonguefish (Pleuronectiformes: Cynoglossidae)

By Jennifer Tyler

B.Sc, McGill University, 2005

Supervisory Committee

Dr. Verena Tunnicliffe, Supervisor (Department of Biology)

Dr. John Dower, Co-supervisor (Department of Biology)

Dr. Kim Juniper, Departmental Member (Department of Biology)

(3)

iii

Supervisory Committee

Dr. Verena Tunnicliffe, Supervisor (Department of Biology)

Dr. John Dower, Co-supervisor (Department of Biology)

Dr. Kim Juniper, Departmental Member (Department of Biology)

Abstract

Fish are not abundant at hydrothermal vents due to the toxicity of venting fluids. Those that are present usually roam the periphery of the vent field or visit occasionally to feed on the abundance of life supported by chemosynthesis. In the past decade, dense aggregations of a newly described flatfish, Symphurus n.sp, have been observed in association with hydrothermal vents in the western Pacific hydrothermal vent

biogeographic province. In this thesis I provide evidence that Symphurus n.sp is a vent obligate and consider the ramifications that this association with hydrothermal vents may have for its distribution, population characteristics, behaviour and diet.

Symphurus n.sp has a widespread but disjunct distribution throughout the western Pacific hydrothermal vent biogeographic province. Symphurus n.sp appears to be

restricted to hydrothermally active, shallow, sulphur rich seamounts. Symphurus n.sp occurs on unconsolidated volcanoclastic ash and solid sulphur crusts and in close association with molten elemental sulphur. The obvious affinity that this species has for native sulphur is unusual and remains unexplained. Unlike most vent-associated fish, Symphurus n.sp occurs in close contact with point source venting and its distribution extends to the periphery of vent fields but not beyond. The density of flatfish on these seamounts surpasses density estimates of flatfish nursery grounds on the continental shelf. On Daikoku Seamount (Mariana Volcanic Arc), mean flatfish abundances were 100 and 66 individuals m-2 in 2005 and 2006 respectively. The prey items that support

(4)

such high densities of flatfish vary over spatial scales. Differing prey, in turn, results in differing foraging modes. On Nikko Seamount (Mariana Volcanic Arc), Symphurus n.sp is a “sit and wait” predator that feeds exclusively on a vent endemic shrimp, Opaepele loihi. On other seamounts, Symphurus n.sp is an opportunistic forager that preys mostly on polychaetes and small crustaceans. By counting annuli on otoliths I constructed growth curves and determined that growth rates differ between seamounts. This

difference in growth rates is likely due to differences in their diet and foraging strategies. Symphurus n.sp may be allocating more energy to growth when less energy is required to forage. Furthermore, size distributions also differ between populations, likely due to variability in growth rates as well as differences in strong recruitment years.

(5)

v

Table 2.3.1: Location, date and dive information of the Symphurus n.sp populations in the western Pacific examined in this study. (*) Indicates verified populations for which no data is available for this study...13 Table 2.3.2: Summary of seamount characteristics on which Symphurus n.sp are known to occur ...17 Table 2.3.3: Summit depth of other explored volcanoes on the Mariana, Kermadec and Tonga arcs

including the depth range of Symphurus n.sp where present. Seamounts in bold indicate a summit depth within the observed depth range that Symphurus n.sp inhabits. * indicates seamounts with possible

populations based on the observation of only a single individual. ...18 Table 2.3.4: Length of dive tracks, length of Symphurus n.sp observation along dive tracks and maximum distance from point source venting ...21 Table 2.3.5: Communities and characteristic substrata associated with Symphurus n.sp. Species in bold were observed to be very common. * indicates known vent obligate species. ** indicates undescribed species that have only been observed in venting environments. ...21 Table 2.3.6: Average behavioural events per minute and standard error of Symphurus n.sp inside

depression of loose sediments and outside depressions in 2006. ...24 Table 2.3.7: Average behavioural events and standard error of Symphurus n.sp observations. ...24 Table 2.3.8: Biomass estimates associated with mean density estimates from Daikoku seamount and point estimates from Nikko, Kasuga-2 and Volcano-1...28 Table 2.4.1: Examples of published high-density flatfish estimates ...35 Table 3.2.1: Preserved samples of Symphurus n.sp obtained from all dives. All specimens were preserved in 95% EtOH ...58 Table 3.2.2: Number of Symphurus n.sp specimens dissected for gut content analysis per seamount ...60 Table 3.3.1: Size range of all preserved specimens and mean length (with standard error) per year and seamount. Collection techniques differed between years and location. ...61 Table 3.3.2: Mean standard length and standard error of Symphurus n.sp obtained from still images and preserved specimens obtained from Daikoku Seamount in 2005 and 2006 and Nikko Seamount in 2006. ..62 Table 3.3.3: Sex ratio of Symphurus n.sp. * indicates a sex ratio that differs significantly from 1:1 ...71 Table 3.3.4: Occurrence of gravid females per seamount. ...72 Table 3.3.5: Range of standard lengths and associated otolith-based of Symphurus n.sp on Daikoku, Kasuga-2 and Nikko seamounts ...73 Table 3.3.6: Diet items per size class of Symphurus n.sp on Daikoku, Nikko, Kasuga-2 seamounts and Volcano-1. ...78

(8)

Table 3.3.7: Range in carbon and nitrogen isotope values obtained from all organisms sampled from Daikoku Seamount. ...81 Table 3.4.1: Comparison of growth parameters and maximum lengths and ages of small flatfish species. From the von Bertalanffy growth function, L∞ = Asymptotic length, k = rate of growth and t = theoretical time at length 0. The growth parameters of Symphurus n.sp are L∞ = 107.19, k = 0.199, t = -1.29,

(9)

ix

List of Figures

Figure 2.1.1: Map of volcanic arcs in the West Pacific biogeographic province. Map created using the online Marine Geoscience Data System (Carbotte et al. 2004). Names in red indicated volcanic arcs

explored during this study...8

Figure 2.2.1: Location of transects on Daikoku seamount in 2005 and 2006. Red stars indicate high temperature point source venting ...10

Figure 2.3.1: Known distribution of Symphurus n.sp based on exploratory cruises to western Pacific seamounts between 2004-2007. The distribution of Symphurus n.sp on Kaikata Seamount are based on observations made by Hashimoto et al. (1995), Munroe and Hashimoto (in press) however no data are available from this seamount for this study. ...15

Figure 2.3.2: Explored seamounts and back-arc spreading sites of the Mariana Arc. Sites in red indicate seamounts where Symphurus n.sp populations occur...16

Figure 2.3.3: Average behavioral events per minute of Symphurus n.sp on three seamounts on the Mariana Arc. n=10 for each seamount. Error bars indicate standard error. ...24

Figure 2.3.4: Density of Symphurus n.sp along a transect on Daikoku in 2005. “Sulphur crust” indicates substrata comprised of over half hard sulphur crust. The surrounding substrata was always volcanoclastic ash. n=112 ...27

Figure 2.3.5: Density of Symphurus n.sp along a transect on Daikoku in 2006. n= 44 ...27

Figure 2.3.6: Density frequency distribution of Symphurus n.sp on Daikoku seamount in 2005 and 2006 .28 Figure 2.3.7: Mean density of flatfish on sulphur crust (n=16) and volcanoclastic sediments (n=96) on the transect on Daikoku seamount in 2005. Error bars indicate standard error. ...28

Figure 2.3.8:Close up of Symphurus n.sp on Daikoku seamount. B) Symphurus species 2 (undescribed) that coexists with Symphurus n.sp on Volcano-19. Note the distinguishing eyespots on the caudal end. C) Symphurus n.sp and the crab Austinograea yunohana on consolidated sulphur crust on Nikko seamount in 2006. D) Symphurus n.sp aggregations on Nikko seamount, with A. yunohana and the tube worm, Lamellibrachia satsushima. E) Symphurus n.sp, A. yunohana and Opaepele loihi (shrimp) on newly congealed sulphur beside a molten sulphur rivulet on Nikko Seamount. F) Tonguefish on a sulphur chimney on Volcano-19 in 2007. G) Two Symphurus n.sp individuals in direct contact on Kasuga-2 Seamount in 2005. H) Dense aggregations of Symphurus n.sp in an unconsolidated depression on sulphur rich, volcanoclastic ash on Daikoku seamount in 2005 ...29

Figure 2.5.1: Daikoku Seamount, Mariana Arc ...42

Figure 2.5.2: Kasuga-2 Seamount, Mariana Arc ...43

Figure 2.5.3: Nikko Seamount, Mariana Arc...43

Figure 2.5.4: Macauley Seamount, Kermadec Arc...44

(10)

Figure 2.5.6: Volcano-19, Tonga Arc...45 Figure 2.5.7: Symphurus n.sp. observations and point source venting on Nikko seamount (Mariana Arc) in 2005 and 2006...46 Figure 2.5.8: Symphurus n.sp. observations and point source venting on Kasuga-2 seamount (Mariana Arc) in 2004 ...47 Figure 2.5.9: Symphurus n.sp. observations and point source venting on Daikoku seamount (Mariana Arc) in 2005 and 2006...48 Figure 2.5.10: Symphurus n.sp. observations and point source venting on Macauley seamount (Kermadec Arc) in 2005 ...49 Figure 2.5.11: Symphurus n.sp. on Volcano-1 (Tonga Arc) in 2005...50 Figure 2.5.12: Coordinates of Symphurus n.sp observations and known point source venting on Volcano-1 in 2007...51 Figure 2.5.13: Symphurus n.sp on Volcano-19 (Tonga Arc)...52 Figure 2.5.14: Coordinates of Symphurus species 1 and 2 observations and locations of known point source venting on Volcano-1...53 Figure 3.2.1: Seamounts where Symphurus n.sp populations were sampled for this study. ...57 Figure 3.3.1: Size structure of Symphurus n.sp obtained from video imagery (n=94) and preserved

specimens (n=24) from Daikoku Seamount in 2005. ...64 Figure 3.3.2: Size structure of Symphurus n.sp obtained from video imagery (n=52) and preserved

specimens (n=50) from Daikoku Seamount in 2006 ...64 Figure 3.3.3: Size structure of Symphurus n.sp obtained from video imagery (n=35) and preserved

specimens (n=15) from Nikko Seamount in 2006 ...65 Figure 3.3.4: Size frequency distribution of Symphurus n.sp specimens obtained on Kasuga-2 (n=24), Nikko (n=39) and Daikoku (n=69) and Volcano-1 (n=29) in 2005, 2006 and 2007 ...66 Figure 3.3.5: Mean SL of Symphurus n.sp males and females on Nikko, Daikoku, Kasuga-2 and Volcano-1. Error bars indicate standard error. ...67 Figure 3.3.6: Size structure of Symphurus n.sp based on sex. Size structure is significantly different

between males and females ...67 Figure 3.3.7: Proportion of males and females of Symphurus n.sp based on five equal size classes on A. Nikko, B. Kasuga-2, C. Daikoku and D. Volcano-1.  indicates males and £ indicates females...68 Figure 3.3.8: Length weight relationship from all preserved specimens. ...69

(11)

xi

Figure 3.3.9: Length-weight relationship. This plot shows weights obtained by converting preserved wet-weights using the equation from Thornstad et al. 2007 and compares them to the wet-weights of frozen

individuals. ...70 Figure 3.3.10: Mean weight of preserved Symphurus n.sp per seamount. Error bars indicate standard error. ...70 Figure 3.3.11: Mean weight of Symphurus n.sp males and females of on Daikoku, Kausga-2, Nikko and Volcano-1. Error bars indicate standard error. ...71 Figure 3.3.12: Length at age data for specimens from Nikko, Daikoku and Kasuga-2 seamounts ...73 Figure 3.3.13: Mean standard length at age fit to the von Bertalanffy growth equation for Symphurus n.sp on Daikoku and Nikko Seamounts. Error bars are 95% confidence intervals. R2 _{= 0.956. L∞ = 107.19, k =}

0.199, t = -1.29...74 Figure 3.3.14: Mean residuals from length at age data from specimens on Nikko and Daikoku seamounts. Error bars indicate standard error. ...74 Figure 3.3.15: Size frequency and associated ages in years of all aged individuals from Daikoku, Kasuga-2 and Nikko Seamount...75 Figure 3.3.16: Sagittal otolith from Symphurus n.sp on Nikko seamount. Age 10. White circles indicate annuli. ...75 Figure 3.3.17: Food items in gut contents of Symphurus n.sp per seamount ...77 Figure 3.3.18: Frequency of occurrence of food items in the gut contents of Symphurus n.sp on Daikoku Seamount in 2005 and 2006. Frequency of occurrence excludes individuals with empty guts. ...79 Figure 3.3.19: Symphurus n.sp feeding on a myctophid that fell to the sea floor likely as a result of contact with volcanic plumes ...80 Figure 3.3.20: Stable isotope graph of δ13C and δ15N from samples taken from Daikoku Seamount in 2006.

...81 Figure 4.1.1: Summary of thesis findings depicting why this species is considered a vent obligate and the implications of this species association with hydrothermal vents ...93

(12)

Acknowledgements

First I would like to thank my supervisors, Dr. Verena Tunnicliffe and Dr. John Dower for their support, ideas and insight and for providing me with the opportunity to work in such a fascinating field of science. I would also like to thank my committee member Dr. Kim Juniper for his support with my stable isotope work and for providing me with collections from the Tonga Arc from the MANGO Cruise with help from Cat Stevens. I especially would like to thank my lab mates: Heidi Gartner, Candice St. Germain, Ian Beveridge and Mathis Stoeckle and honorary lab mates Damian Grundle and Will Duguid and other members of the Page and Juniper labs for their endless encouragement and excellent sense of humor. I am very thankful for Le Fonds québécois de la recherche sur la nature et les technologies (FQRNT), Maureen de Burgh Memorial Scholarship and Gordon Fields Memorial Fellowship for the financial support has allowed me to complete my studies. Dr. Thomas Munroe from the Smithsonian for his help with tonguefish morphology, Dr. Peter Stoffers and Lonny Lundsten (MBARI) for providing me with imagery from the 2005 SITKAP cruise, Joanne Groot at the Pacific Biological Station for her excellent training in reading otoliths, Susan Merle (NOAA) and Dr. Mark Hannington (University of Ottawa) for providing maps and dive tracks from Submarine Ring of Fire cruises and Dr. Malcolm Clark (NIWA) for providing me with dive videos from the Kermadec Arc. Last and most importantly I would like to thank my family and friends for their love and encouragement and Dan for his unfaltering support and for reminding me what is important in life.

(13)

1

Chapter 1: Introduction

1.1 Hydrothermal Vents

Life is relatively sparse throughout the deep-sea. Light is not available for photosynthesis and most animals rely on food that falls from above. In 1977, scientists were therefore surprised to discover communities with high levels of biomass thriving in some of the most inhospitable deep-sea environments (Corliss et al. 1979).

Hydrothermal vents form where cold seawater penetrates into cracks in the young oceanic crust (Corliss et al. 1979). The water is chemically altered by the interaction of heat with the surrounding rock. The heated, chemical laden water then rises back to the seafloor. Upon its return, the hot, anoxic venting fluid contains high levels of hydrogen sulphide, carbon dioxide and various heavy metals (Weber et al. 2003). These substances are toxic to most animals (Fisher 1995). However, it is these same fluids that are

responsible for the abundance of life at hydrothermal vents.

Autotrophic microbes at hydrothermal vents use reduced ions such as hydrogen sulphide in the vent effluent to fix carbon dioxide via chemosynthesis (Corliss et al. 1979, Jaanasch and Wilson 1979). This primary production supports extremely high levels of animal biomass, although biodiversity is relatively low (Tunnicliffe 1991), as few species have developed the extensive physiological adaptations required to live in these toxic environments (Hessler and Kaharl 1995). Those that have adapted are at a significant advantage, as they have access to a plentiful food source with little

competition (Tunnicliffe 1991, Hessler and Kaharl 1995).

Because of the extreme physiological challenges, most of the animals that live around hydrothermal vents occur nowhere else in the deep-sea (Tunnicliffe 1991). Interestingly, the ability to inhabit hydrothermal vents has evolved in many unrelated taxa (Hessler and Lonsdale 1991). Amongst vent animals, considerable regional

variations in species composition occur surrounding vent fields (Tunnicliffe et al. 2003). Individual vent fields are relatively small and are often separated by large distances. Not surprisingly, vent fields located in close proximity to each other generally share species and gene pools, whereas vent fields separated by large distances have distinct faunal

(14)

assemblages and share few species (Tunnicliffe 1988). Currently, six hydrothermal biogeographic provinces are recognized worldwide based on similarity of species composition (Tunnicliffe et al. 1998, Van Dover et al. 2001, Tyler and Young 2003). At any given vent site the species composition thus represents an assemblage of species drawn from a regional pool. This local species assemblage therefore reflects both the regional pool of species as well as the specific habitat availability and physical properties of the vent field (Van Dover 2000).

Unlike invertebrates, fish are not usually found in great abundances around hydrothermal vents (Biscoito et al. 2002, Sancho et al. 2005, Wolff 2005) even though they are quite common elsewhere in the deep-sea. Most that are present roam the periphery of vent fields (Biscoito et al. 2002). Therefore, during exploratory cruises to hydrothermally active seamounts in the west Pacific, scientists were surprised to find extremely high densities of flatfish thriving in vent environments. This newly described species, Symphurus n.sp (Munroe and Hashimoto in press), is the first flatfish known to inhabit hydrothermal vents.

1.2 Tonguefishes, Symphurus (Cynoglossidae:

Pleuronectiformes)

Flatfishes form the order Pleuronectiformes, which is currently divided into 14 families (Desoutter et al. 2001). Flatfishes undergo the migration of an eye, resulting in an ocular and a blind side of the body and an asymmetrical cranium. Flatfishes all have a dorsal fin that extends over the cranium and a recessus orbicularis, a sac-like part of the wall of the orbit (Osse and Van den Boogart 1997, Desoutter et al. 2001). The family Cynoglossidae, also referred to as the “tonguefishes”, has a subterminal mouth and a continuous fin that extends from the skull along the entire ventral and dorsal body area up to the edge of the underside of the gill area (Osse and Van den Boogart 1997). Members of this family have also lost their pectoral fins, lateral line and the pelvic fin on the left side (Munroe 1998). Cynoglossids are sinistral flatfish; meaning their eye always

migrates to the left side of the body when they undergo metamorphosis (Hensley 1997). The family Cynoglossidae is comprised of two subfamilies: Symphurinae and Cynoglossinae (Saldierna-Martinez et al. 2005). The genus Symphurus is part of the subfamily Symphurinae and is the most species-rich member of the family Cynoglossidae

(15)

3 (Munroe 1998). There are currently 73 described species in the genus Symphurus

(Munroe 2006). Symphurine tonguefishes are small to medium sized, reaching a

maximum size of 30cm (Munroe 1998). Members of the genus Symphurus have a small mouth with toothed and strongly curved jaws on the blind side. Most larval stages of species in this genus remain undescribed although they are known to be bilaterally symmetrical and hatch at less that 2.5mm (Saldierna-Martinez et al. 2005). In relation to other Symphurus species, Symphurus n.sp has a notably deep body, a moderately long and bluntly pointed head, and relatively large, round eyes (Munroe and Hashimoto in press).

The egg and larval stages of most flatfishes are pelagic (Leggett and Deblois 1994). The distribution of eggs after spawning is completely dependent upon

hydrodynamic forces (Gibson 1997). After hatching, larvae have limited power of locomotion and are generally unable to swim horizontally against currents that they may encounter. However, many species are able to migrate vertically, thereby altering their distribution through variations in current speed and direction at different depths in the water column (Gibson 1997). Pelagic larval stages experience high mortality rates, usually in excess of 99%. Thus, even small differences in larval mortality rates may translate to large differences in numbers of survivors (Rijnsdorp et al. 1995). The survival of pelagic eggs and larvae is determined mainly by hydrographic features such as turbulence and advection, temperature and the synchronization of egg production with the plankton production cycle (Rijnsdorp et al. 1995). Following the bilaterally

symmetric, pelagic larval stage, flatfishes metamorphose into benthic dwelling

asymmetrical juveniles and congregate in demersal nursery grounds (Shreiber 2001). The duration of the larval period varies widely between species and ranges from 1 week to 1 year (Shreiber 2001).

Metamorphosis involves the flattening of the body, extreme development of the unpaired fins, eye migration and changes in pigmentation on both the ocular and blind sides (Osse and Van den Boogart 1997). The flatfish head skeleton is cartilaginous until after transformation to allow for eye migration. Ossification of the head skeleton occurs only after metamorphosis has been completed (Osse and Van den Boogart 1997).

(16)

commonly occurs at total lengths between 10-25mm and takes several hours to several weeks (Osse and Van den Boogart 1997). Body size at metamorphosis varies within species. Many cynoglossids are completely transformed at total body lengths of less than 5 mm. The duration of transformation depends on the species and environmental

conditions (Osse and Van den Boogart 1997).

After settlement, flatfish temporarily stop feeding until the completion of

metamorphosis (Yamasita et al. 2001). The shift from a pelagic to a benthic life style also corresponds with a shift from feeding on zooplankton to feeding on benthic organisms (Yamasita et al. 2001). The majority of juvenile and adult flatfish feed on benthic

invertebrates (Link et al. 2002). Therefore, following metamorphosis foraging behaviour changes from searching through the water column to ambushing prey on the bottom.

1.3 Research Focus

Unusual abundances of a new species of flatfish occur at vents in the western Pacific. The overall objectives of this thesis were to:

1. Provide evidence that this species has an obligate association with hydrothermal vents, and

2. Determine what ramifications this association has for the distribution, behaviour, population characteristics and diet of Symphurus n.sp.

Addressing the objectives of this thesis was executed in two parts. In Chapter 2: Field observations, distribution and abundance, my questions included the following: i) Is this species restricted to hydrothermally active environments?

ii) Is Symphurus n.sp found on all hydrothermally active seamounts in the western Pacific or is it restricted to seamounts with particular geological, chemical and physical

parameters?

This question was addressed by comparing summit depths, venting products and geological features of seamounts inhabited by Symphurus n.sp.

iii) Does this species have a dispersion pattern similar to most vent-associated fish species that occur mostly around the periphery of vent fields?

This question was addressed by documenting flatfish-vent associations from video imagery and measuring the maximum distance of Symphurus n.sp observations from point source venting.

(17)

5 iv) Last, how is the association with productive hydrothermal vent environments

manifest in terms of population densities?

This question was addressed by estimating densities using still images captured along a transect followed by remotely operated vehicles in two separate years. These densities were then compared to other studies that estimated densities of flatfish in non-vent environments.

In Chapter 3: Population characteristics and trophic ecology I ask two questions: i.) Are population characteristics the same for all populations of Symphurus n.sp?

I addressed this questions using preserved specimens I obtained from four

populations of Symphurus n.sp in the western Pacific. These specimens along with video imagery were used to compare size and weight distributions of populations. I also used preserved specimens to create growth curves and investigate sex ratios and size at maturity.

ii.) Is the diet of Symphurus n.sp similar for all populations?

To address this question I dissected preserved specimens and compared the diet between populations. I also used stable isotope analysis to compliment gut content data in order to gain a better picture of the trophic interactions of Symphurus n.sp on western Pacific seamounts.

(18)

Chapter 2:

Field observations, distribution and abundance

2.1 Introduction

Organisms inhabiting hydrothermal vents may be exposed to hot, anoxic waters that contain high levels of hydrogen sulphide, heavy metals, and carbon dioxide. These conditions are toxic to most organisms. Animals that live in these environments therefore require extensive physiological and life history adaptations to survive. The ability to tolerate such harsh conditions and take advantage of the high levels of production around hydrothermal vents appears to have evolved independently in many unrelated taxa (Hessler and Lonsdale 1991) and has resulted in high levels of endemism (Wolff 2005).

Invertebrates are generally more resistant to H2S and anoxia than are vertebrates

due likely to their less complex nervous systems and lower metabolic oxygen demands (Cohen et al. 1990). Consequently,fish are not usually found in great abundance in vent systems (Sancho et al. 2005). In 2002, fish had been observed in only 40% of known venting sites (Biscoito et al. 2002). When present, fish usually roam the periphery of vent sites only visiting the food-rich environments for brief periods of time (Hourdez and Weber 2005). As of 2002, only twenty-one species of fish from a total of seven families were known to live inside active vent fields (Biscoito et al. 2002). These constitute only 3.5% of currently known vent species (Wolff 2005). Of all fish species recorded in venting environments, only 22% are endemic (Wolff 2005). Fish from the family Zoarcidae account for the highest species numbers and account for the highest fish biomass in venting environments (Biscoito et al. 2002).

Since the discovery of hydrothermal vents in the late 1970’s, regional variations in species composition have been observed surrounding vent fields (Tunnicliffe et al. 2003). Six hydrothermal biogeographic provinces are currently recognized worldwide (Tunnicliffe et al. 1998, Tyler and Young 2003). The West Pacific is one such province (Figure 2.1.1). Approximately 20, 000 km of volcanic arcs with about 600 submarine volcanoes, some of which are hydrothermally active, occur along convergent plate boundaries in the western Pacific Ocean (Embley et al. 2004). Volcanic Arcs form where

(19)

7 a dense oceanic plate is subducted under a lighter adjacent plate (Embley et al. 2004). As the overlying plate subducts, volatiles are released which results in the partial melting of the overlying mantle wedge. The magma that is produced rises towards the earth’s surfaceforming a chain of volcanoes parallel to the oceanic trench (Stern 2002). Sediments on the subducting plate commonly cause the enrichment of some unusual elements in arc lavas (Stern 2002). Hydrothermal vents occur on some submarine

volcanoes (seamounts) at the front of arcs (Stoffers et al. 2006). Hydrothermal activity on volcanic arcs occurs at a wide range of depths and variations in rock composition produce a greater range of hydrothermal fluid chemistry relative to mid-ocean ridge sites (Embley et al. 2004). This creates a variety of habitats for faunal communities (Embley et al. 2004). As a result, the hydrothermal vent communities of the western Pacific are

relatively heterogeneous compared to those of mid-ocean ridges (Tyler and Young 2003). The first flatfish observed around hydrothermal vents was a tonguefish of the genus Symphurus. Hashimoto et al. (1995) first recorded this tonguefish on the hydrothermally active Kaikata Seamount on the Izu-Bonin Arc and the Minami-Ensei Knoll on the Okinawa Trough in the West Pacific. In a synthesis of known

chemosynthetic communities of the Northwestern Pacific, Kojima (2002) also reports that Symphurus had been observed on the hydrothermally active Nikko Seamount on the northern Mariana Arc. This species was originally identified as Symphurus orientalis but was later determined to be Symphurus n.sp, the new species discussed in this paper. A specimen from Kaikata Seamount later served as the holotype for the description of this species (Munroe and Hashimoto in press).

Approximately 600 species of flatfish are known and almost half occur in the western Pacific (Minami and Tanaka 1992). Distributional patterns of flatfishes occur as a result of selection for particular physiochemical characteristics, prey availability and substrata based on particle size, ease of burying or composition of the sediments (Munroe 1998). Symphurine tonguefishes occur in all temperate and tropical oceans (Munroe 1998) and most are deep-water species (Munroe 2006). Until recently none were reported to be associated with hydrothermal venting.

This paper documents the known distribution and observations of Symphurus n.sp on hydrothermally active seamounts in the West Pacific based on exploratory cruises

(20)

over the past four years. In this chapter I provide evidence that Symphurus n.sp has an obligate association with hydrothermal vents. First I determine if they are restricted to hydrothermally active environments, I then investigate the geological, chemical and physical properties of the seamounts that this tonguefish inhabits to gain an

understanding of the large-scale habitat requirements of this species. The substrata with which this tonguefish are commonly associated are described and the proximity to venting is estimated to determine if Symphurus n.sp has a similar distribution to most vent-associated fish that mostly roam the periphery of the vent field. Community compositions were compared and abundances and biomass were estimated.

Figure 2.1.1: Map of volcanic arcs in the West Pacific biogeographic province. Map created using the online Marine Geoscience Data System (Carbotte et al. 2004). Names in red indicated volcanic arcs explored during this study.

(21)

9

2.2 Methods

i. Field Observations

Symphurus n.sp was observed and sampled during exploratory cruises on the Mariana, Tonga and Kermadec Arcs using both manned and unmanned submersibles.

In 2004 the ROPOS submersible obtained opportunistic video and still images from Daikoku and Kasuga-2 seamounts. Sampling efforts on Daikoku, Kasuga-2 and Nikko seamounts on the Mariana Arc in 2005 and 2006 included focused studies of Symphurus n.sp. Suction samplers mounted on the ROVs were used to obtain samples of Symphurus n.sp from the Mariana Arc in 2005 and 2006 which were subsequently preserved in 95% EtOH. Data obtained from Macauley Seamount (Kermadec Arc), Volcano-1 and Volcano-19 (Tonga Arc) in 2005 were entirely opportunistic and consist only of qualitative observations. In 2007 visibility on Volcano-1 and 19 was either poor or the ROV was moving too fast to make quantitative observations. Video imagery and hi-resolution digital still images were recorded on each dive. Visibility during recording ranged from very good to zero visibility.

A transect was sampled in high-density fish areas on Daikoku seamount in 2005 by the ROV HyperDolphin and in 2006 by Jason II. Still images were recorded along a ~500m transect on Daikoku in 2005 and an ~350m transect in 2006. The transect was run upslope along a sedimented slope at a speed of approximately 0.1 knots. In 2005, the transect ran from approximate depths of 410 to 390m. In 2006, the transect spanned depths of 405 to 385m (Figure 2.1). Still images from both transects were used to estimate tonguefish densities. Parallel lasers were mounted on the cameras of the ROVs HyperDolphin for scale in 2005 and Jason II in 2006. Four lasers spaced 10 cm apart at 90° angles to each other were used on the HyperDolphin. Two lasers (also spaced 10 cm apart) were used on Jason II. One hundred and twelve images with good visibility taken close enough to the sea floor to confidently count fish numbers were available from 2005 and 44 images were available from 2006.

(22)

Figure 2.2.1: Location of transects on Daikoku seamount in 2005 and 2006. Red stars indicate high temperature point source venting

ii. Analysis

A map of Symphurus n.sp distribution was created using the online Marine

Geoscience Data System (Carbotte et al. 2004) and Adobe Photoshop. A complete review of the literature was conducted to compare emission features and depth ranges of the seamounts on which Symphurus n.sp has been found.

The known locations of Symphurus n.sp on the seven seamounts were depicted using Adobe Illustrator. GIS files and dive-track layers from seamounts on the Mariana Arc were obtained courtesy of PMEL, NOAA. Maps from Macauley Seamount were obtained courtesy of the National Institute of Water and Atmosphere (NIWA), New Zealand. Times of flatfish sightings from video imagery were associated with latitude and longitude taken from ROV navigation files on Nikko (dives J198 and J199 from 2006 and

(23)

11 HD501 from 2005), Daikoku (dives J195 and J197 from 2006 and HD491 from 2005), Kasuga-2 (dives R794 in 2004 and HD490 in 2005), Macauley (dive PV617 in 2005), 1 (PIV142 from 2005 and R1050, R1051 and R1053 from 2007) and Volcano-19 (PIV138 from 2005 and R1046, R1047 and R1048 from 2007). The length of the dive tracks was estimated, including the length of the dive track containing fish observations. Areas with point source venting were included in the maps to determine the proximity of Symphurus n.sp to known venting sites. The maximum distance between point source venting and flatfish observations was calculated on each seamount.

I recorded macrofauna that was commonly associated with Symphurus n.sp and visible on video imagery. I also documented the type of substrata that tonguefish and associated communities were found on.

Tonguefish behaviour was documented using video imagery captured when the ROV was sitting on the sea floor and not moving. Ten fish from each seamount were observed and behavioural events were recorded until they were no longer in the field of view of the camera either because the fish had swum out of view or the camera had moved. Behavioural events per minute were then estimated and compared between seamounts. The total observation time for all observations was 41.7 minutes. On Kasuga-2, Nikko, Daikoku (2005) and Daikoku (2006) the average observation time per

individual was 71.1, 58.6, 73.9 and 46.7 seconds respectively.

Density estimates from transects on Daikoku Seamount were estimated using images from transects conducted in 2005 and 2006. Fish were counted in all images with good visibility. Images were captured while the ROV was between 1 and 3 meters off the bottom. The sample area in each image corresponded to a 10cm deep band of the sea floor along the full width of each image. The narrow band was used to reduce the effect of perspective on area estimates as the transects took place going up the slope of the seamount and the image was oblique to the camera lens. I calculated the sample area using the 10cm spacing between the lasers for scale. Fish were counted when any part of an individual fell inside the sample area. The width of the seafloor captured in still images ranged between 30cm to almost 2m. Abundance estimates were extrapolated to individuals m-2. Density estimates were plotted against distance along transects on Daikoku seamount for 2005 and 2006. The substrata on which Symphurus n.sp occurred

(24)

were included in the density plots. Density frequency distributions were constructed in Excel and compared using a chi-squared test. Mean fish density was calculated for the 2005 and 2006 data. A t-test was used to determine if mean abundance differed between years.

A point density estimate of Symphurus n.sp on Kausuga-2 seamount was calculated using a single image from 2004 with visible lasers. No lasers for scale were mounted on the camera during dives on Nikko seamount except for on dive J199. However visibility during this dive was often poor. Furthermore, the terrain on Nikko seamount is quite variable due to uneven substratum and presence of large tubeworm bushes. This made it difficult to use the lasers accurately for scale in many images on this dive. Abundances in a high-density area on Nikko seamount in 2005 were estimated by measuring the length of all individuals in an image. The mean length of individuals in the image was assumed to be equal to the mean length of preserved specimens from Nikko seamount in 2005 (Chapter 3). The area of an image was then estimated by converting the length and width of the image into actual length and width by using the conversion factor of mean length from the image vs. mean length of preserved specimens. Individuals that occurred in the image were counted and converted to density estimates. This method was tested on an image from dive J199 in 2006 that was taken on flat terrain with visible lasers. Density estimates were made using the lasers for scale and also by using the average fish length for scale as described above.

Biomass was estimated using mean weight (Chapter 3) and mean density estimates on Daikoku, Nikko, Kasuga-2 and Volcano-1.

2.3 Results

i. Geographic Range

Since the first reported observations by Hashimoto et al. (1995), various exploratory cruises to volcanic arcs in the western Pacific have expanded the known range of this species. The locations and dates of verified Symphurus n.sp observations are listed in Table 2.3.1. Using morphological features Munroe and Hashimoto (in press) has since determined that populations on the Kermadec and Mariana Arcs are both the same Symphurus species. One individual was observed on Ruby seamount on the Mariana Arc

(25)

13 in 2006 on dive J194. However, this observation is insufficient to assume that a

population exists on this seamount. One individual was also observed during dive (R1041) on Monowai Volcano on the Tonga Arc in 2007, however visibility did not allow for further exploration. One individual was collected from Rumble 3 Seamount on the Kermadec Arc for Munroe and Hashimoto (in press) however the collection of one individual on this seamount is insufficient to assume a population exists on this

seamount. Tonguefish have also been observed on the Minami-Ensei Knoll on the Mid-Okinawa trough (Hashimoto et al. 1995) however no specimens have been collected and no video imagery is available.

Volcanic Arc Volcano Date Vehicle Dive Designator

Izu-Bonin Kaikata* 1996 Shinkai 2000 1233-1236

Mariana Arc Daikoku 4/13/04 ROPOS R-795

10/26/05 HyperDolphin HD-491 5/2/06 Jason II J2-195 5/4/06 Jason II J2- 197 Kasuga-2 4/12/04 ROPOS R-794 10/26/05 HyperDolphin HD-490 Nikko 5/11/05 HyperDolphin HD-501 5/7/06 Jason II J2-198 5/8/06 Jason II J2-199

Kermadec Arc Macauley 4/14/05 Pisces V P5-617

Tonga Arc Volcano-1 6/25/05 Pisces IV P4-142

5/10/07 ROPOS R-1050 5/11/07 ROPOS R-1051 5/14/07 ROPOS R-1053 Volcano-19 6/21/05 Pisces IV P4-138 5/5/07 ROPOS R-1046 5/5/07 ROPOS R-1047 5/6/07 ROPOS R-1048

Table 2.3.1: Location, date and dive information of the Symphurus n.sp populations in the western Pacific examined in this study. (*) Indicates verified populations for which no data is available for this study

The known distribution of Symphurus n.sp is shown in Figure 2.3.1. The known distribution within the western Pacific province is disjunct. A distance of approximately 6500 km separates known populations on the Mariana and Tonga Arcs.

The main focus of this study is the Mariana Arc. Figure 2.3.2 presents the hydrothermally active seamounts of the Mariana Arc that have been explored, including

(26)

seamounts where Symphurus n.sp is known to occur. Symphurus n.sp has only been observed in the Northern Seamount Province of the Mariana Arc and populations are located relatively close together: Nikko seamount is located 150km from Kasuga-2 seamount, and Kasuga-2 Seamount is 112 km from Daikoku Seamount.

ii. Seamount Characteristics

The locations and characteristics of the seamounts that Symphurus n.sp is known to inhabit for which data are available for this study are summarized in Table 2.3.2. Common characteristics include a fairly shallow summit (between 65 and 383m), active hydrothermalism and outcrops or sediments rich in near pure sulphur deposits. Volcano-19 on the Tonga Arc is an exception, as its sediments are rich in iron oxides. The summit depths of all explored seamounts on the Mariana, Kermadec and Tonga arcs including the depth ranges of Symphurus n.sp observations are listed inTable 2.3.3.

(27)

15

Figure 2.3.1: Known distribution of Symphurus n.sp based on exploratory cruises to western Pacific seamounts between 2004-2007. The distribution of Symphurus n.sp on Kaikata Seamount is based on observations made by Hashimoto et al. (1995), Munroe and Hashimoto (in press) however no data are available from this seamount for this study.

(28)

Figure 2.3.2: Explored seamounts and back-arc spreading sites of the Mariana Arc. Sites in red indicate seamounts where Symphurus n.sp populations occur

(29)

Volcano Arc Latitude Longitude Depth Range Emission Features References

(m below sea level)

Nikko Mariana 23º 05.00' N 142º19.00' E 383-2705 Molten sulphur pool Embley et al. 2006

Excess sulphate

Abundant elemental sulphur present as thick solidified sulphur flows

Kasuga-2 Mariana 21º 36.60' N 143º 39.00' E 200-1600 High dissolved gas content McMurtry et al. 1993

Enriched in sulphate

High levels of dissolved CO2

Diffuse venting

Elevated H2S Fryer et al. 1997

Sulphur-rich volcanoclastic sand

Daikoku Mariana 21º 02.00' N 144º 32.00' E 323-2763 Molten sulphur pool Embley et al. 2006

Low pH values

Excess sulphate

Widespread diffuse venting

Sulphur-rich volcanoclastic sand and Thick sulphur flows

Volcano-1 Tonga 24º 48.00' S 175º 45.00' W 65-1800 Widespread diffuse venting Stoffers et al. 2006

High levels of CO2 gas

High dissolved gas content

Thick beds of sulphur cemented ash

Volcano-19 Tonga 24º 48.00' S 177º 01.00' W 385-1400 Low temperature diffuse venting Stoffers et al. 2006

Low dissolved gas content

Fe-oxyhydroxide crusts

Macauley Kermadec 30º 12.00' S 178º 28.00' W above sea level-1400 Diffuse venting Merle et al. 2005 Abundant elemental sulphur in substrata And sulphur flows

Table 2.3.2: Summary of seamount characteristics on which Symphurus n.sp are known to occur

(30)

Table 2.3.3: Summit depth of other explored volcanoes on the Mariana, Kermadec and Tonga arcs including the depth range of Symphurus n.sp where present. Seamounts in bold indicate a summit depth within the observed depth range that Symphurus n.sp inhabits. * indicates seamounts with possible populations based on the observation of only a single individual.

iii. Field Observations

Symphurus n.sp is a small (25-125 mm standard length), brown tonguefish with mottled colouration and darker cross bands partially spanning the ocular surface of the body (Figure 2.3.9.A). A second, undescribed Symphurus species was observed on Volcano-19. This species coexists with Symphurus n.sp but was larger and far less abundant. It is mottled brown and has thin darker bands crossing the body and is less tapered at the caudal end (Figure 2.3.9.B). Its distinguishing characteristics are two “eye spots” at its caudal end. One eyespot is on the dorsal fin and the other on the anal fin. The “eye spots” have a black centre and white outline.

Arc Seamount Summit depth Symphurus n.sp depth

(where present) Mariana Esmeralda 53 E Diamante 135 Ruby* 178 Kasuga-2 200 375-445 Daikoku 323 360-400 Nikko 383 383-460 NW Rota-1 517 Seamount X 1128 Forecast 1456 NW Eifuku 1548 Tonga Volcano-1 65 83-288 Volcano-18 285 Volcano-19 385 419-562 Kermadec Giggenbach 75 Monowai* 100 Macauley Cone 138 300-360 Rumble V 400 "W" 840 Clark 860 Healy 975 Brothers 1210

(31)

19

a. Habitat Characteristics and Distribution on Seamounts

Aggregations of Symphurus n.sp were observed on Nikko, Daikoku and Kasuga-2 seamounts on the Mariana Arc, Volcano-1 and Volcano-19 on the Tonga Arc and

Macauley seamount on the Kermadec Arc. On Nikko Seamount Symphurus n.sp were abundant throughout the crater as well as on the flanks of the seamount in the vicinity of the “Naraku” sulphur pool (Embley et al. 2006). Abundances on Nikko Seamount were so high that fish were often seen lying on top of neighbouring individuals (Figure 2.3.9.D). On Kasuga-2, fish observations were very localized on the west side of the summit. On Daikoku Seamount, Symphurus n.sp were abundant throughout much of the explored area on the northwest flank of the seamount. On Macauley Seamount

populations were localized on the northwest flank leading up to the summit of the Macauley cone. The sparsest populations occurred on Volcano-1 and Volcano-19. On Volcano-1 tonguefish occur on the volcano summit and on the scoria cone to the southwest of the crater. On Volcano-19 tonguefish occur on the summit of the central cone.

Symphurus n.sp were most often seen on unconsolidated or semi-consolidated volcanoclastic sediments with sulphurous deposits. On Daikoku Seamount in 2006, meter wide depressions with an increase in fish density were noted. These depressions appeared to be comprised of disturbed sediments amongst semi-consolidated volcanoclastic ash. On Nikko and Macauley seamount Symphurus n.sp were often observed on consolidated elemental sulphur crusts (Figure 2.3.9.C). On Daikoku Seamount in 2006, tonguefish were also observed in a second environment. Aggregations of tonguefish occurred surrounding a 15m2 pool of molten sulphur on substrata consisting of consolidated pebbles of elemental sulphur. Although no obvious food source was evident, the flatfish abundances surrounding this pool were quite high. Two point estimates of density

surrounding the sulphur lake yielded a density of 97 and 60 individuals m-2. Proximity to the molten sulphur clearly does not harm these fish since several individuals were seen touching the surface crust of the undulating sulphur lake and swimming away, seemingly unharmed. A probe inserted into the molten sulphur measured temperatures as high as 187°C. However, the thin crust of solidified sulphur that the fish were lying on, occurred at the interface between liquid elemental sulphur and seawater and was therefore

(32)

probably significantly cooler. This tonguefish was also seen in other unusual and unexpected sulphur rich environments: on Nikko seamount they were observed directly beside a molten sulphur rivulet and even lying on top of the newly congealed sulphur (Figure 2.3.9.E). On that occasion shrimp and crabs (apparently grazing on

chemosynthetic bacteria) were also observed in the vicinity. On Macauley Seamount numerous individuals were oriented vertically, most with their heads pointing up, on a sulphur cliff while on Volcano-19, two individuals were observed on a sulphide chimney (Figure 2.3.9.F)

Symphurus n.sp lives in both the direct vicinity of point source venting but it also extends to the periphery of vent influence. Table 2.3.4 summarizes the distances explored by the ROV, the distance along the dive tracks in which fish were observed and the maximum distance that Symphurus n.sp was observed from known locations of point source venting. Symphurus n.sp occurs further from known point source venting sites within the Nikko crater than at other seamounts. However, Nikko Seamount is

characterized by a large central caldera that retains venting fluids. All observations of Symphurus n.sp took place with some indicator of venting such as shimmering water, microbial mat or coexisting vent obligate macrofauna. Although this species is at times the only macrofaunal species present, when they do coexist with other macrofauna, the communities always contain known vent obligate species. Macrofaunal communities and associated substrata are summarized in Table 2.3.5. Macrofaunal communities that coexist with Symphurus n.sp differ substantially between locations.

Although the distribution of Symphurus n.sp is widespread in many areas with active venting, there were also active venting areas on these seamounts where Symphurus n.sp were not observed. Tonguefish were absent around some explored venting areas on Kasuga-2 and Volcano-19. On Volcano-19 they were absent on the deep venting field found between 900-1000m below sea level. This venting site exceeds the known depth range of Symphurus n.sp. On Kasuga-2 Seamount tonguefish were not observed in the vicinity of the Yellow Overlord vent site. This sedimented venting site is situated within the depth range of other Symphurus n.sp observations.

(33)

21

Arc Seamount Length of Length of dive track with Maximum distance

dive track (m) fish observations (m) from point source vent (m)

Mariana Daikoku 892 473 125

Nikko 1090 395 160

Kasuga-2 2237 203 50

Kermadec Macauley 698 124 110

Tonga Volcano-1 n/a n/a 130

Volcano-19 n/a n/a 42

Table 2.3.4: Length of dive tracks, length of Symphurus n.sp observation along dive tracks and maximum distance from point source venting

Seamount Characteristic Substrata Associated Macrofauna

Daikoku Volcanoclastic sediments Anemones (Cerianthidae, Actinaria)**

Crabs (Austinograea yunohana) *

Hermit crabs (Paguroidea)**

Snails (Oenopota ogasawarana) *

Seastar (Asteroidea)

Urchins (Echinoidea)

Vestimentiferan tubeworms (Lamellibrachia satsushima) *

Surrounding the "Sulphur Cauldron" Crabs (Austinograea yunohana) * Kasuga-2 Volcanoclastic sediments Barnacles (Neoverruca sp.)*

Crabs (Austinograea yunohana) *

Shrimp (Alvinocarididae)

Nikko Sulphur crust Barnacles (Neoverruca sp.)*

Volcanoclastic sediments Crabs (Austinograea yunohana)*

Shrimp (Alvinocaridae and Palaeomonidae) *

Vestimentiferan tubeworms (Lamellibrachia

satsushima) *

Macauley Sulphur crust Mussels (Bathymodiolus sp.) * Volcanoclastic sediments Seastar (Asteroidea)

Volcano-1 Volcanoclastic sediments Anemones (Cerianthidae, Actiniaria)**

Brachyuran crabs (Bythograeidae)*

Mussels (Bathymodiolus sp.) *

Volcano-19 Volcanoclastic sediments Brachyuran crabs (Bythograeidae)*

Clams (Thyassiridae) - all dead**

Table 2.3.5: Communities and characteristic substrata associated with Symphurus n.sp. Species in bold were observed to be very common. * indicates known vent obligate species. ** indicates undescribed species that have only been observed in venting environments.

(34)

b. Behaviour

In general, individual Symphurus n.sp lie on the seafloor and move forward or backwards by undulating their bodies. They also frequently shift their axis of orientation. Symphurus n.sp did not react to the presence of the submersible except when sediment cores or suction samples of the benthos were taken by the ROV. In these instances Symphurus n.sp appeared to be attracted to the loose sediments created as a result of sampling. I observed fish from Daikoku, Volcano-1 and Macauley seamounts aggregating in a pit formed by samples taken by the submersible.

On Daikoku Seamount, behavioural observations were made from video imagery of tonguefish on top of unconsolidated sediment. Tonguefish behaviour was documented on Daikoku Seamount in areas where they co-occurred with large numbers of the snail Oenopota ogasawaarna. The behaviour of 20 individuals was observed from Daikoku Seamount, ten from 2005 and ten from 2006. Of the 10 individuals documented from video imagery in 2006, five occurred in a depression of loose, unconsolidated sediments and five individuals where observed situated outside of the depression. These depressions were observed in various locations on Daikoku Seamount and tended to have a higher density of fish within them.

Individuals rotated or moved forward or backwards every few seconds, pausing briefly before the next movement. Pauses usually lasted 3-4 seconds but sometimes individuals would stay still for over a minute. When moving forward or backwards they most often moved less than a body length. This movement usually requires two

undulations of the body. The number of movements per unit time was not significantly different between years (t-test: P = 0.49) on Daikoku Seamount. However, of the 10 fish observed from video obtained in 2005, only one swam further than one body length. In contrast, half the fish observed in 2006 were seen swimming more than a full body length at once. All fish observed in the depression moved a distance greater than one body length when moving whereas fish outside the depression tend to move less than one body length (Table 2.3.6). Behavioural events occur at the same rate between years, however individuals in the unconsolidated depressions observed from video obtained in 2006 were swimming further. Of the 20 fish observed, only one was observed feeding, which was recognizable by the expulsion of sediment through the operculum. This fish was located

(35)

23 in the unconsolidated depression. One fish from each year was observed burying itself in the sand. Numerous other fish were seen burying themselves during dives on Daikoku Seamount. Buried fish are usually no longer visible and tend not to move for long periods of time. In 2006, the burying behaviour was due to an abrupt movement of the

submersible. Six flatfish were seen swimming directly on top of and often resting on top of the individual being observed. In all cases but one case, the individual under

observation did not react to the presence of another fish sitting on top of it. The one individual that did react to contact with a conspecific, rotated its axis of orientation by about 20º, however the individuals were still touching after the shift in orientation.

Behaviour of ten Symphurus n.sp individuals was documented on Nikko

Seamount in 2006. They were observed in an area with sediment, slabs of sulphur crust and larger pieces of gravel. Symphurus n.sp were well camouflaged against the gravel substrata. Symphurus n.sp co-occurred with large numbers of the crab (Austinograea yunohana) and shrimp (Opaepele loihi). Fish were significantly less active on Nikko seamount than on Daikoku seamount (t-test: P < 0.05). The crabs would often walk on top of the tonguefish causing them to swim a short distance usually less than half a body length away. Fish were often seen sitting on top of each other, not moving for long periods of time. No feeding behaviour was observed.

Behaviour of ten Symphurus n.sp individuals was documented on Kasuga-2 Seamount in 2005. Only crabs and bacterial mat were visibly co-occuring with Symphurus n.sp. The activity rate of tonguefish on Kasuga-2 Seamount appears to be lower than on Daikoku seamount but higher than Nikko Seamount. However, differences in mean rates are not significant. Figure 2.3.9.G shows two individuals on Kasuga-2 Seamount in direct contact. Average events per type of behaviour for all seamounts are shown in Table 2.3.7. Total behavioural events per seamount are shown in Figure 2.3.4.

(36)

Average behavioural event per minute Inside depression (n = 5) Outside of depression (n=5)

Shifts orientation axis 1.57 ± 0.78 1.89 ± 0.79

Moves < 1 body length 1.52 ± 0.74 6.50 ± 2.30

Moves > 1 body length 5.06 ± 2.59 0

Encounters another fish 1.41 ± 0.72 0.02 ± 0.02

Buries in sediment 0.24 ± 0.24 1.46 ± 0.51

Total activity 9.81 ± 1.33 9.87 ± 3.49

Table 2.3.6: Average behavioural events per minute and standard error of Symphurus n.sp inside depression of loose sediments and outside depressions in 2006.

Average behavioral Daikoku Daikoku Nikko Kasuga-2

event/minute (n = 10) 2005 (n=10) 2006 (n=10) 2006 (n=10) 2005

Shifts orientation axis 3.15 ± 0.53 1.72 ± 0.52 0.57 ± 0.29 1.54 ± 0.43

Moves < 1 body length 4.16 ± 0.94 4.02 ± 1.41 1.56 ± 0.59 2.43 ± 0.77

Moves > 1 body length 0.06 ± 0.06 2.66 ± 1.47 0.62 ± 0.22 2.80 ± 0.47

Encounters another fish 0.18 ± 0.13 0.77 ± 0.41 0.00 0.20 ± 0.13

Buries in sediment 0.10 ± 0.10 0.12 ± 0.12 0.00 0.00

Feeding behaviour 0.70 ± 0.70 0.00 0.00 0.00

Encounters a crab n/a n/a 0.59 ± 0.26 0.00

Total events per minute 7.73 ± 1.28 9.17 ± 1.59 3.35 ± 0.80 5.48 ± 1.00

Average duration of behavioural 84.0 ± 6.4 53.7 ± 9.05 59.2 ± 9.75 76.5 ± 9.71

observations per individual (seconds)

Table 2.3.7: Average behavioural events and standard error of Symphurus n.sp observations.

Figure 2.3.3: Average behavioral events per minute of Symphurus n.sp on three seamounts on the Mariana Arc. n=10 for each seamount. Error bars indicate standard error.

(37)

25

c. Density Estimates

Density estimates along the transect on Daikoku Seamount from 2005 ranged from 0 – 392 individuals m-2 (Figure 2.3.5). In 2006 estimates ranged from 22 – 119 individuals m-2 (Figure 2.3.6). Density values are higher and show much greater variation in 2005 than in 2006. Density frequency distributions show a much wider range from the 2005 transect than from 2006 (Figure 2.3.7). I estimated mean flatfish densities on the Daikoku transect to be 100 individuals m-2 in 2005 and 66 individuals m-2 in 2006. Mean tonguefish density was significantly different between years (t-test, P < 0.05). Median values were 96 individuals m-2 in 2005 and 64 individuals m-2 in 2006. Figure 2.3.9.H shows an example of a high-density flatfish area along the transect in 2005.

Patches of sulphur crust occurred at the beginning of the transect on Daikoku Seamount in 2005. The remainder of the transect consisted of volcanoclastic ash. Figure 2.3.8 shows the density of individuals on the sulphur crust compared to those on

sedimented areas. Densities are significantly higher on volcanoclastic sand. Symphurus n.sp are always absent in areas with basalt outcrops.

A point estimate of density from an image taken by the ROPOS submersible on Kasuga-2 Seamount in 2004 was made. Using the lasers for scale, density was estimated at 114 individuals m-2_.

A point estimate of 151 individuals m-2_{on Nikko Seamount was calculated using}

the length of fish (Chapter 3) to estimate the area of the image. This method was tested using an image with visible lasers on flat terrain. It was found that using the length of fish for scale resulted in an underestimate of density. A density estimate of 34 individuals m-2 was estimated using lasers for scale and 28 individuals m-2 was estimated using fish length for scale on an image from Nikko seamount in 2006. Fish present in images are likely to vary from the average fish size but can still be used for reasonable scale estimates.

No lasers were mounted on the ROV for scale on Volcano-1 in 2005. Therefore the above method was used for a point density estimate. Density from a still image taken by the Pisces IV was estimated to be 93 individuals m-2 for a high-density fish image.

(38)

Density, mean weight and biomass associated with density estimates are shown for Daikoku, Kasuga-2, Nikko and Volcano-1 in Table 2.3.8. Nikko has a fish biomass ten times higher than at other sites due to the high abundances and larger fish.

(39)

27

Figure 2.3.4: Density of Symphurus n.sp along a transect on Daikoku in 2005. “Sulphur crust” indicates substrata comprised of over half hard sulphur crust. The surrounding substrata was always volcanoclastic ash. n=112

(40)

Figure 2.3.6: Density frequency distribution of Symphurus n.sp on Daikoku seamount in 2005 and 2006

Figure 2.3.7: Mean density of flatfish on sulphur crust (n=16) and volcanoclastic sediments (n=96) on the transect on Daikoku seamount in 2005. Error bars indicate standard error.

Density (individuals m-2₎ _{Mean weight (g)} _{Biomass (g m-}2₎

Daikoku 2005 100 0.86 85.60

Daikoku 2006 66 1.68 110.88

Nikko 151 7.51 1134.57

Kasuga-2 114 0.64 72.73

Volcano-1 93 1.08 100.44

Table 2.3.8: Biomass estimates associated with mean density estimates from Daikoku seamount and point estimates from Nikko, Kasuga-2 and Volcano-1

(41)

Figure 2.3.8: A) Close up of Symphurus n.sp on Daikoku seamount. B) Symphurus species 2 (undescribed) that coexists with Symphurus n.sp on Volcano-19. Note the distinguishing eyespots on the caudal end. C) Symphurus n.sp and the crab Austinograea yunohana on consolidated sulphur crust on Nikko seamount in 2006. D) Symphurus n.sp aggregations on Nikko seamount, with A. yunohana and the tube worm, Lamellibrachia satsushima. E) Symphurus n.sp, A. yunohana and Opaepele loihi (shrimp) on newly congealed sulphur beside a molten sulphur rivulet on Nikko Seamount. F) Tonguefish on a sulphur chimney on Volcano-19 in 2007. G) Two Symphurus n.sp individuals in direct contact on Kasuga-2 Seamount in 2005. H) Dense aggregations of Symphurus n.sp in an unconsolidated depression on sulphur rich, volcanoclastic ash on Daikoku seamount in 2005

(42)

2.4. Discussion

i. Geographic Range and Seamount Characteristics

Most species of the genus Symphurus occur in deep waters of the western Pacific. However, Symphurus n.sp is the first flatfish species that has been shown to inhabit hydrothermal vent fields. It occurs on hydrothermally active seamounts associated with subduction zones in the western Pacific. These seamounts are part of the western Pacific hydrothermal vent biogeographic province. No other vent associated flatfish species has been recorded in any other hydrothermal biogeographic province. However, further explorations into seamounts in other provinces will be necessary to determine if the distribution of this species is restricted to this province.

Symphurus n.sp has a widespread distribution in the West Pacific and is found between 23ºN and 30ºS. Although, extensive, it is not the most widespread species of the genus Symphurus. The range of the blackcheek tonguefish, Symphurus plagiusa spans between 41ºN and 23ºS in the Western Atlantic Ocean (Munroe 1998). Within the known range of Symphurus n.sp, a distance of approximately 6500 km separates populations in the Mariana and Tonga Arcs. It is likely however that they are found on other unexplored hydrothermally active seamounts in the western Pacific such as those on the New

Hebrides Arc. Further exploration of volcanic arcs will be important in expanding the known range of Symphurus n.sp.

Changes in pressure effects the cardiovascular and nervous system function of fishes (Gibbs 1997). Therefore depth is an important parameter in determining

distributions of marine organisms. Seamount depth seems to be an important

environmental variable for this species as they only appear to inhabit relatively shallow seamounts and are absent on seamounts with summit depths of over 500m. Many Symphurus species inhabit fairly discrete depth zones (Munroe 1998). This species inhabits a comparatively broad depth range, however the highest densities of individuals were found between 300-400m. The deepest reported occurrence for a member of the genus Symphurus is S. woodmasoni at 896m in the Indo-Western Pacific (Krabbenhoft and Munroe 2003). Although a number of important environmental factors are closely related to depth, such as pressure, oxygen, nutrients, amount of sinking organic material, migrating zooplankton layers and light (Stocks and Hart 2007), depth is likely not the

Distribution, population characteristics and trophic ecology of a sulphophilic hydrothermal vent tonguefish (Pleuronectiformes: Cynoglossidae)

Abstract

Table of Contents

List of Tables

List of Figures

Acknowledgements

Chapter 1: Introduction

1.1 Hydrothermal Vents

1.2 Tonguefishes, Symphurus (Cynoglossidae:

Pleuronectiformes)

1.3 Research Focus

Chapter 2:

Field observations, distribution and abundance

2.1 Introduction

2.2 Methods

2.3 Results

2.4. Discussion