USF Libraries
USF Digital Collections

The biocomplexity of benthic communities associated with a shallow-water hydrothermal system in papua new guinea

MISSING IMAGE

Material Information

Title:
The biocomplexity of benthic communities associated with a shallow-water hydrothermal system in papua new guinea
Physical Description:
Book
Language:
English
Creator:
Karlen, David
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Biocomplexity
Shallow-water hydrothermal vents
Environmental gradients
Benthic infaunal communities
Eukaryotic diversity
Dissertations, Academic -- Biology -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Shallow-water hydrothermal vents occur world-wide in regions of volcanic activity. The vents located at Tutum Bay, Ambitle Island, Papua New Guinea are unique in that the vent fluids and surrounding sediments contain some of the highest concentrations of arsenic in a natural system. This study addresses the effects of the vent system on the benthic communities, focusing on the eukaryotes, macrofauna, meiofauna and bacteria. Samples were collected in November 2003 and May/June 2005. Analysis of the 2003 macrofaunal samples indicated that pH, rather than arsenic was influencing the benthic community, and that the hydrothermal influence occurred at a greater distance than expected. Results of more intensive sampling carried out in 2005 are the primary focus of this dissertation. The pore water and sediment characteristics revealed distinct physical habitats corresponding with distance from the vent. There was a trend of decreasing temperature and arsenic concentration and increasing salinity and pH with distance from the vent. The vent sediment was poorly sorted volcanic gravel, while sediments along the transect showed a gradient from fine, well sorted volcanic sands to coarser carbonate sands farther away. The macrofauna showed a trend of increasing diversity with distance from the vent and similar taxa were present in both the 2003 and 2005 samples. The vent community was dominated by the polychaete Capitella cf. capitata. The inner transect from 30 m to 140 m had low diversity. Dominant taxa included thalassinid shrimp and the amphipod Platyischnopus sp.A. The 180 m to 300 m sites had significantly higher diversity. The Danlum Bay reference site had relatively higher diversity than the nearshore transect sites and was dominated by deposit feeding polychaetes. Macrofaunal community structure was influenced by the sediment characteristics, notably by CaCO3 content, sorting and median grain size. The meiofaunal community also showed changes with distance from the vent. Chromadorid nematodes were dominant at the vent site and were a major component of the meiofauna at most sites, along with copepods. The meiofaunal community at the reference site showed greater similarity to the vent community and both sites had low abundances. Nematodes were more abundant than copepods near the vent, but copepods were more abundant farther offshore and at the reference site. Meiofaunal community structure was influenced primarily by the pore water temperature and salinity. Biological interactions with the macrofaunal community through physical disturbance and predation may also influence the meiofaunal community. The molecular analysis of eukaryotic and bacterial diversity also revealed changes with distance from the vent. The 0 m and reference sites grouped together due to the presence of fungal sequences and the 140 m and 300 m sites grouped together due to a common molluscan sequence. Metazoans and fungi dominated the eukaryote sequences. The most abundant eukaryotic OTUs included fungi matching Paecilomyces sp. and Cladosporium cladosporioides and metazoans matching Viscosia viscosa (Nematoda) and Astarte castanea (Mollusca). The eukaryotic community structure correlated with sediment sorting, pore water arsenic and median grain size. The bacterial community was represented by 24 phyla and was dominated by Actinobacteria and γ-Proteobacteria. More bacterial phyla were present near the vent, while more overall OTUs were found at the intermediate sites along the transect. The most distant site had much lower diversity dominated by Firmicutes. The macrofaunal community had the strongest correlation with environmental variables. Comparison between the meiofauna and the metazoan sequences showed the proportion of nematodes found in both datasets were comparable, but the meiofauna analysis found a higher proportion of arthropods, while the molecular results were disproportionally high for platyhelminthes. Overall, the vents increased the complexity of the system by creating unique habitats. The extreme environment created by the hydrothermal activity maintained the surrounding habitat at an early successional stage colonized by a few opportunistic species. There was a gradation in the benthic communities away from the vent towards a more carbonate based climax community. The low pH environment had an effect on the sediment composition, which in turn influenced the benthic community. These findings can serve as a model for studying the potential effects of ocean acidification and climate change on benthic communities and marine biocomplexity.
Thesis:
Dissertation (PHD)--University of South Florida, 2010.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by David Karlen.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0004736
usfldc handle - e14.4736
System ID:
SFS0028035:00001


This item is only available as the following downloads:


Full Text

PAGE 1

The B iocomplexity of Benthic Communities Associated with a Shallow water Hydrothermal System in Papua New Guinea by David J. Karlen A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Cell Biology, Microbiology and Molecular Biology College of Arts and Sciences University of South Florida Major Professor: James R. Garey, Ph.D Susan S. Bell, Ph.D Pamela Hallock Muller, Ph.D Kathleen M. Scott, Ph.D Date of Appro val: October 14, 2010 Keywords: biocomplexity, shallow water hydrothermal vents, environmental gradients, benthic infaunal communities, eukaryotic diversity Copyright 2010, David J. Karlen

PAGE 2

Dedication In memory of my father, Rev. Marloe H. Karlen September 30, 1933 April 25, 2008

PAGE 3

Acknowledgments There are so many people to whom I owe my gratitude for their help and support during my time working on this dissertation. First I would like to thank Dr. James R. Garey for his g uidance and encouragement these past nine years as my major professor, and for providing me with the opportunity to work on this project. I would also like to thank my committee members Dr. Susan Bell, Dr. Pame la Hallock Muller, and Dr. Katie Scott for al l of their help and advice. I wish to also thank Dr. Richard Turner for agreeing to serve as chairman for my defense. T his dissertation was part of a National Science Foundation Biocomplexity grant (BE: CBC# 0221834) awarded to Dr. Thomas Pichler, and I a m grateful to him for having had the opportunity to participate in this project and to travel to Papua New Guinea. I also would like to thank my other fellow researchers on the grant, Roy Price, Pamela Hallock Muller, Brian McCloskey, Jan Amend, and Darcy Meyer Dombard for their field assistance in collecting s amples, sharing data and for their friendship. Thanks also goes to the crew of the M/V Star Dancer for two wonderful trips in Papua New Guinea and for their help in the field and to Dr. Steve Sanders for hosting our stay in PNG and serving as our tour guide in Rabaul. This research project was a very large undertaking, and I could have never finished without the help of many people. I owe a lot to Terry Campbell, Stefi Depovic and Tiehang Wu from the Garey lab for their assistance with the 18S rDNA molecular work that was part of this research. I would still be trying to get my PCR to work if not

PAGE 4

for their help. Haydn Rubelmann in the Garey Lab provided the bacterial 16S rRNA data for the bacterial com munity analysis. Sarah Mike sorted macrofauna samples and Staci e Villanueva spent many hours sorting meiofauna samples and making slide mounts of specimens for me. Dr. Ping Wang and Dr. John Lawrence allowed me to use their lab facilities for sediment anal ysis a nd Roy Price provided sediment and pore water analysis for arsenic. I would also like to thank my bosses and fellow workers at the Environmental Protection Commission of Hillsborough County for their support, encouragement and understanding while I pursued my research. I am also greatly indebted to my employer for allowing me the use of our lab facilities to conduc t my research and supporting me in my academic endeavors. And finally I owe a lifetime of gratitude to my parents for the love and support they have always given me.

PAGE 5

i Table of Contents List of Tables ................................................................................................................. iii List of Figures ............................................................................................................... vi Abstract. ........................................................................................................................ xi Chapter One. Introduction ...............................................................................................1 1.1 Background ...................................................................................................1 1.2 Dissertation o utline ........................................................................................2 1.3 Ambitle Island hydrothermal vent system .......................................................3 1.4 Other shallow water hydrothermal vent systems .............................................4 1.5 Arsenic in the marine environment ................................................................. 6 1.6 The effects of pH ............................................................................................7 1.7 Animal sediment relationships and biogeochemistry ......................................9 1.8 T he concept of biocomplexity ...................................................................... 10 Chapter Two. Site Description and Physical Characteristics ........................................... 12 2.1 Introduction .................................................................................................. 12 2.2 Material and m ethods ........................................................................ 12 2.2.1 Sampling design and field collection .............................................. 12 2.2.2 Sediment analysis .......................................................................... 16 2.3 Results ......................................................................................................... 17 2.4 Discussion .................................................................................................... 32 2.5 Summary and c onclusions ............................................................................ 36 Chapter Three. Changes in Benthic Macrofauna associated with a Shallow water Hydrothermal Vent Gradient in Papua New Guinea ................................................. 38 3.1 Introduction .................................................................................................. 38 3.2 Material and methods ................................................................................... 41 3.3 Results ......................................................................................................... 46 3.4 Discussion .................................................................................................... 57 Chapter Four. Macrofaunal communities associated with the shallow water hydrothermal vent system at Ambitle Island, Papua New Guinea ............................. 61 4.1 Introduction .................................................................................................. 61 4.2 Mat erial and methods ................................................................................... 62 4.2.1 Sampling design and field collection .............................................. 62 4.2.2 Sample processing and identification ............................................. 62

PAGE 6

ii 4.2.3 Data analysis .................................................................................. 63 4.3 Results ......................................................................................................... 64 4.4 Discussion .................................................................................................... 81 4.5 Summary and conclusions ............................................................................ 85 Chapter Five. Meiofaunal communities associated with the shallow water hydrothermal vent system at Ambitle Island, Papua New Guinea ............................. 87 5.1 Introduction .................................................................................................. 87 5.2 Material and methods ................................................................................... 90 5.2.1 Sampling design and field collection .............................................. 90 5.2.2 Sample processing and identification ............................................. 91 5.2.3 Data analysis .................................................................................. 92 5.3 Results ......................................................................................................... 92 5.4 Discussion .................................................................................................. 116 5.5 Summary and conclusions .......................................................................... 118 Chapter Six. Molecular diversity of eukaryotic and bacterial communities associated with the shallow water hydrothermal vent at Ambitle Island, Papua New Guinea ........................................................................................................... 119 6.1 Introduction ................................................................................................ 119 6.2 Material and methods ................................................................................. 120 6.2.1 Sampling design and field collection ............................................ 120 6.2.2 DNA extraction and amplification ................................................ 121 6.3 Results ....................................................................................................... 125 6.4 Discussion .................................................................................................. 161 6.5 Summary and conclusions .......................................................................... 164 Chapter Seven. Biocomplexity of the communities associat ed with the shallow water hydrothermal system at Tutum Bay, Ambitle Island, Papua New Guinea. ............................................................................................................. 166 7.1 Comparison of different biological communities and physical parameters. ................................................................................................. 166 7.2 Summary of findings and conclusions: the biocomplexity of the Tutum B ay hydrothermal system: .......................................................................... 171 7.2.1 The Vent Community (0 m) ......................................................... 171 7.2.2 Diffuse V enting Zone C ommunity (30 m 140 m). ..................... 172 7.2.3 The Low hydrothermal Activity Z one C ommunity (180 m 300 m) ......................................................................................... 174 7.2.4 Danlum Bay Reference Site ......................................................... 176 7.3 Conclusions ................................................................................................ 177 Lit erature Cited ........................................................................................................... 182 About the Author ................................................................................................ End Page

PAGE 7

iii List of Tables Table 2.1 Site depths and median, minimum and maximum values for pore water variables by site (n = 5 measurements per site). .............................................. 19 Table 2.2 Median, minimum and maximum dry weight percentage of sediment by Wentworth size class. ................................................................................... 24 Table 2.3 Median, minimum and maximum measurements for sediment parameters. .................................................................................................... 25 Table 2.4 Principal component eigenvalues and percent variation .................................. 29 Table 2.5 Principal component e igenvectors. ................................................................. 29 Table 2.6 Spearman correlation coefficients between pore water and sediment characteristics. .............................................................................................. 31 Table 3.1 Summary of pore water characteristics at Tutum Bay, Ambitle Island, Papua New Guinea November 2003 .............................................................. 47 Table 3.2 Sediment grain size class percentages, percent organic carbon and carbonates (mean 1 standard deviation); arsenic concentrations from Price and Pichler (2005) ................................................................................. 48 Table 3.3 Benthic community index values cumulated for all five replicate samples at each site ........................................................................................ 49 Table 3.4 SIMPER analysis results by distanc e (Bray Curtis Similarity among five replicate samples) and percent contribution of top five taxa. ................... 53 Table 3.5 BIO ENV results for single and multiple parameters ...................................... 56 Table 4.1 Median, minimum and maximum values for macrofaunal community indices by site ................................................................................................ 65 Table 4.2 Top ranked taxa comprising >50% of the relative abundance at each site ................................................................................................................ 70 Table 4.3 SIMPER results from macrofaunal community groups defined by B ray Curtis similarity groupings ............................................................................. 79

PAGE 8

iv Table 4.4 a BIO ENV correlations between the benthic macrof aunal community struct ure and physical parameters.. ............................................................... 80 Table 4.4b BIO ENV correlations between the benthic macrofaunal community structure and physical parameters ; reference site omitted. ............................. 80 Table 5.1 Meiofaunal total taxonomic richness (S), abundance (N ) and relative abundance (%) by site. ................................................................................... 99 Table 5.2 Median, minimum and maximum meiofaunal community index values for taxonomic richness raw abundance counts, Shannon diversity index (base e), and evenness by site. ..................................................................... 103 Table 5.3 Relative abundance of do minant taxa representing >50% relative abundance at the 0 m site (excluding crustacean larvae). .............................. 103 Table 5.4 Median, minimum and maximum values for nematode and copepod abundances and the nematode:copepod ratio at each site. ............................. 107 Table 5.5. SIMPER result s for meiofaunal community groupings. ............................... 114 Table 5.6 a BIO ENV Spearman correlations and best fit of physical parameters with the meiofaunal co mmunity structure. .................................................. 115 Table 5.6b BIO ENV Spearman correlations and best fit of physical parameters with the meio faunal community structure ; r eference site omitted. ............... 115 Table 6.1 Ranked OTUs representing 50% of the 2,987 total sequences. OTUs ranked by percent abundance and then by number of sites at which they occur. ........................................................................................................... 134 Table 6.2 Relative abundance the top five r anked OTUs (including ties) at each site ............................................................................................................... 135 Table 6.3 a BIO ENV correlations between eukaryotic community structure and physical parameters. ................................................................................... 141 Table 6.3b BIO ENV correlations between eukaryotic community structure and physical paramet ers; reference si te omitted. ................................................. 141 Table 6.4 Top ten ranked metazoans comprising >50% of the 1,059 metazoan sequences. ................................................................................................... 144

PAGE 9

v Table 6.5 a BIO ENV correlations between metazoan community structure and physical parameters. ................................................................................... 152 Table 6.5b BIO ENV correla tions between metazoan community st ructure and physical parameters; r eference site omitted ................................................ 153 Table 6. 6 Number of OTUs (S) and sequences (N) for bacterial phyla for all sites (Total) and individual sites. .......................................................................... 156 Tabl e 7.1 Seco nd Stage MDS Spearman correlations P hysical parameters, m acrofauna and m eiofauna relative abundance data averaged by site for comparison .................................................................................................. 169 Table 7.2 Second Stage MDS Spearman correlations for physical parameters, macrofauna, and m eiofauna relative abun dance for all replicate samples ...... 169

PAGE 10

vi List of Figures Figure 2.1 Ambitle Island, Papua New Guinea sampling locations ................................. 13 Figure 2.2 Focused venting at Tutum Bay ...................................................................... 13 Figure 2.3 Diffuse venti ng (background) at Tutum Bay. ................................................ 14 Figure 2.4 Sampling quadrat and hand corer for macrofaunal samples .......................... 15 Figure 2.5 Ambitle Island May/June 2005: mean pore water temperature 1 standard deviation. ....................................................................................... 20 Figure 2.6 Ambitle Island May/June 2005: mean pore water salinity 1 standard deviati on. ..................................................................................................... 20 Figure 2.7 Ambitle Island May/June 2005: mean pore water pH 1 standard deviation ...................................................................................................... 21 Figure 2.8 Ambitle Island May/June 2005: mean pore water oxidation reduction potential 1 standard deviation. ................................................................... 21 Figure 2.9 Ambitle Island May/June 2005: mean pore water arsenic concentra tions 1 stand ard deviation. .......................................................... 22 Figure 2.10 Ambitle Island May/June 2005: mean median grain size 1 standard deviation. .................................................................................................... 26 Figure 2.11 Ambitle Island May/June 2005: mean sorting coefficient 1 standard deviation. .................................................................................................... 26 Figur e 2.12 Ambitle Island May/June 2005: mean sediment organic carbon content 1 standard deviation. .................................................................... 27 Figure 2.13 Ambitle Island May/June 2005: mean sediment carbonate content 1 standard deviation. ...................................................................................... 27 Figure 2.14 Ambitle Island May/June 2005: Principal components analysis of pore water and sediment characteristics. ...................................................... 28

PAGE 11

vii Figure 3.1 A: Map of Papua New Guinea and its geographic relation to Australia. ......... 42 Figure 3.1 B: Map of Ambitle Island with the study sites marked with arrows .............. 42 Figure 3.2 Photograph of focused hydrothermal vent at Tutum Bay, Ambitle Island, Papua New Guinea with surrounding diffuse venting (Photo credit: DJK). ................................................................................................. 43 Figure 3.3A: The mean number of taxa (S) and abundance (N) of benthic macrofauna/sample at the transect and reference sites ............................... 50 Figure 3.3B: Mean ShannonW iene r Diversity (H) and Pielous Evenness (J) values at transect and reference sites. ......................................................... 50 Figure 3.4 A: Mean number of taxa by major taxonomic groups at transect and reference sites. ......................................................................................... 52 Figure 3.4 B: Mean abundance of major taxonomic groups a t transect and reference sites .......................................................................................... 52 Figure 3.5 Cluster analysis: Bray Curtis Similarity based on square root transformed macrofaunal abundance data. .................................................... 55 Figure 4.1 Ambitle Island May/June 2005: Mean number of macrofaunal taxa 1 st andard deviation. ....................................................................................... 72 Figure 4.2 Ambitle Island May/June 2005: Mean macrofaunal abu ndance 1 standard deviation ........................................................................................ 72 Figure 4.3 Ambitle Island May/June 2005: Mean macrofaunal div ersity 1 standard deviation ........................................................................................ 73 Figure 4.4 Ambitle Island May/June 2005: Mea n macrofaunal evenness 1 standard deviation ........................................................................................ 74 Figure 4.5 Ambitle Island May/June 2005: Benthic macrofauna multi dimensional scaling plot based on Bray Curtis simil arity between site replicates .............. 76 Figure 4.6a Ambitle Island May/June 2005: Benthic macro fauna multidimensional scaling plot based on Bray Curtis similarity averaged by site .............................................................................................................. .7 6

PAGE 12

viii Figure 4.6b Ambitle Island May/June 2005: Benthic macrofauna multi dimensional scaling plot based on Bray Curtis similarity averaged by site ; reference site omitted. .......................................................................... 77 Figure 4.7 LINKTREE showing physical parameters characterizing site groups based on the benthic macrofauna similarity .................................................. 81 Figure 5.1 Relative taxonomic composition of total meiofaunal abundance (N = 17346 individuals) by phylum (top graph) and subse ts for Nematoda (bottom left; N = 6,935 individuals) and Arthropoda (botto m right; N = 8,019 individuals) ..................................................................................... 97 Figure 5.2 Meiofaunal taxonomic composition by site. .................................................. 98 Figure 5.3 Ambitle Island May/June 2005: Mean meiofaunal taxonomic richness 1 standar d deviation. ............................................................................... 104 Figure 5.4 Ambitle Island May/June 2005: Mean meiofaunal abundance 1 standard deviation. ..................................................................................... 105 Figure 5.5 Ambitle Island May/June 2005: Mean meiofaunal Shannon diversity index 1 standard deviation. ...................................................................... 105 Figure 5.6 Ambitle Island May/Ju ne 2005: Mean meiofaunal evenness 1 standard deviation. ..................................................................................... 106 Figure 5.7 Ambitle Island May/June 2005: Mean nematode abundance 1 standard deviation. ..................................................................................... 108 Figure 5.8 Ambitle Island May/June 2005: Mean copepod abundance 1 standard deviation .................................................................................................... 108 Figure 5.9 Ambitle Island May/June 2005: Mean nematode:copepod ratio 1 standard deviation ...................................................................................... 109 Figure 5.10 Ambitle Island May/June 2005: Meiofaunal Bray Curtis similarity cluster analysis. ......................................................................................... 112 Figure 5.11a Ambitle Island May/June 2005: Meiofaunal Bray Curtis s imilarity cluste r analysis averaged by site. ............................................................. 113 Figure 5.11b Ambitle Island May/June 2005: Meiofaunal Bray Curtis similarity cluster analysis averag ed by site; reference site omitted ........................... 113

PAGE 13

ix Figure 5.12 LINKTREE diagram of meiofaunal community site grouping s and correspondi ng physical characteristics. ..................................................... 116 Figure 6.1 Cyclesequenceing reaction for eukary otic 18S gene amplification. .............. 122 Figure 6.2 Cyclesequenceing reaction for bacterial 16S gene amplification. ................. 124 Figure 6.3 Ambitle Island May/June 2005: Num ber of eukaryotic sequences obtained per site. ........................................................................................ 126 Figure 6.4 Ambitle Island May/June 2005: Number of eukaryote operational taxonomic uni ts (OTUs) obtained per site. .................................................. 126 Figure 6.5 Linear regression analysis of OTUs vs. number of sequences. ..................... 127 Figure 6.6 Ambitle Island May/June 2005: Percentage of eukaryote sequences by kingdom level taxonomic cat egory for all sites combined ........................... 128 Figure 6.7 Ambitle Island May/June 2005: Percentage of eukaryote sequences by kingdom level taxo nomic category at each site. .......................................... 128 Figure 6.8 Ambitle Island May/June 2005: Percentage of eukaryote OTUs by kingdom level taxonomic category for all sites combined. .......................... 129 Figure 6.9 Ambitle Island May/June 2005: Percentage of eukaryote sequences by kingdom level taxo nomic category at each site. .......................................... 129 Figure 6.10a Ambitle Island May/June 2005: Multi Dimensional Scaling (MDS) of eukaryotic community structure based on Bray Cur tis similarity among sites.. ............................................................................................ 139 Figure 6.10b Ambitle Island May/June 2005: Multi Dimensional Scaling (MDS) of euka ryotic community structure based on Bray Curtis similarity among sites; r eference site omitted .......................................................... 140 Figure 6.11 LINKTREE results showing physical characteristics between eukaryotic community site groups. ............................................................ 143 Figure 6.12 Proportion of metazoan sequences b y phyla. ............................................. 145 Figure 6.13 Proportio n of metazoan OTUs by phyla. ................................................... 145 Figure 6.14 Percentage of metazoan sequences by phyla at each site and all sites combined (All). ......................................................................................... 149

PAGE 14

x Figure 6.15 Percentage of metazoan OTUs by phyla at each site and f or all sites combi ned (All). ......................................................................................... 149 Figure 6.16a Non metric MultiDimensional Scaling (MDS) plot of metazoan sequence Bray Curtis similarity between sites. ........................................ 150 Figure 6.16b Non metric Multi Dimensional Scaling (MDS) plot of metazoan sequence Bray Curtis similarity between si tes; reference site omitted ...... 151 Figure 6.17 LINKTREE results showing physical characteristics between metazoan sequence community site groups ............................................... 154 Figure 6.18 Percentage of total bacteri al OTUs (S = 1,081) by phyla ........................... 158 Figure 6.19 Percentage of total bacterial sequ ences (N = 1,973) by phyla.. .................. 158 Figure 6.20 Total number of bacterial phyla represented at each transect site.. ............. 159 Figure 6.21 Total number of bacterial OTUs repres ented at each transect site. ............. 159 Figure 6.22 Percentage of bacterial OTUs by phyla at each transect site. ..................... 160 Figure 6.23 Percentage of bacterial sequences by phyla at each transect site. ............... 160 Figure 6.24 Cluster analysis of bacterial sequence Bray Curtis similarity between sites .......................................................................................................... 161 Figure 7.1 Second Stage Multi Dimensiona l Scaling plot showing correlation between the differ ent datasets ..................................................................... 169 Figure 7.2 Percentage of meiofaunal abundance (top graph) and metazoan sequen c es (bottom graph) by phyla ............................................................. 170

PAGE 15

xi A bstract Shallow water hydrothermal vent s occur world wide in reg ions of volcanic activity The vent s located at Tutum Bay, Ambitle Island, Papua New Guinea are unique in that the vent fluids and surrounding sediments contain some of the highest concentrations of arsenic in a natural system This study addresses the eff ects of the vent system on the benthic communities focu s ing on the eukaryotes macrofauna, meiofauna and bacteria. Samples were collected in November 2003 and May/ June 2005. A nalysis of the 2003 macrofauna l samples indicated that pH, rather than arsenic was influencing the benthic community and that the hydrothermal influence occurred at a greater distance than expected. Results of m ore intensive sampling carried out in 2005 a re the primary focus of this dissertation. The pore water and sediment charact eristics revealed distinct physical habitats corresponding with distance from the vent. There was a trend of decreasing temperature and arsenic concentration and increasing s alinity and pH with distance fro m the vent. The vent sediment was poorly sorted vo lcanic gravel while sediments along the transect showed a gradient from fine, well sorted volcanic sands to coarser carbonate sands farther away. The macrofauna showed a trend of increasing diversity with distance from the vent and similar taxa were pres ent in both the 2003 and 2005 samples. The vent community was dominated by the polychaete Capitella cf. capitata. The inner transect

PAGE 16

xii from 30 m to 140 m had low diversity. D ominant taxa included thalassinid shrimp and the amphipod Platyischnopus sp.A The 1 80 m to 300 m sites had significantly higher diversity The Danlu m Bay reference site had relatively higher diversity than the nearshore transect sites and was dominated by deposit feeding polychaetes. Macrofaunal community structure was influenced by the sediment c haracteristics, notably by CaCO3 content sorting and median grain size. The meiofaunal community a lso showed changes with distance from the vent. Chromadorid nematodes were dominant at the vent site and were a maj or component of the meiofauna at most sites along with copepods. The meiofaunal community at the reference site showed greater similarity to the vent community and both sit es had low abundances. Nematodes were more abundant than copepods near the vent, but copepods were more abundant farther offshore and at the reference site. Meiofaunal community structure was influenced primarily by the pore water temperature and salinity. Biological interact ions with the macrofaunal commun ity through physical disturbance and predation may also influ ence the meiofaunal community. The molecular analysis of e ukaryotic and bacterial diversity also revealed changes with distance from the vent. The 0 m and reference sites grouped together due to th e presence of fungal sequences and t he 140 m and 300 m sit es grouped together due to a common molluscan sequence Metazoans and fungi dominated the eukaryote sequences The most abund ant eukaryotic OTUs included fungi matching Paecilomyces sp. and Cladosporium cladosporioides and metazoans matching Viscosia visco sa (Nematoda) and Astarte castanea (Mollusca). The e ukaryotic community structure correlated with sediment sorting, pore water arsenic and median grain size. The bacterial community was

PAGE 17

xiii represented by 24 phyla and was Proteobacteria. More bacterial phyla were present near the vent, while more overall OTUs were found at the intermediate sites along the transect. The most distant site had much lower diversity dominated by Firmicutes. T he macrofaunal community had the strongest correlation with environmental variables. Comparison between the meiofauna and the metazoan sequences s howed the proportion of nematodes found in both datasets were comparable, but the meiofauna analysis found a hig her proportion of arthropods, while the molecular results were disproportionally high for plat yhelminthes. Overall, the vents increased the complexity of the system by creating unique habitats. The extreme environment created by the hydrothermal activity maintained the surrounding habitat at an early successional stage colonized by a few opportunistic species. There was a gradation in the benthic communities away from the vent towards a more carbonate based climax community. The low pH environment had an effect on the sediment composition which in turn influenced the benthic community. These findings can serve as a model for studying the potential effects of ocean acidification and climate change on benthic communities and marine biocomplexity.

PAGE 18

1 Chapt er One Introduction 1.1 B ackground The shallow water hydrothermal vent system at Ambitle Island, Papua New Guinea is a unique marine habitat which is most notably characterized by extremely high concen trations of arsenic in the pore water and sediment s. Earlier studies at this site reported arsenic concentrations which were 275 times that of normal seawater (Pichler et al. 1999b). T he surrounding reef habitat and fauna appear ed to be unaffected by the elevated concentration of arsenic (Pichler and Dix 1996). In order to fu rther study the biodiversity and biogeochemistry of this unique system, a team of scientists and graduate students from the University of South Florida and Washington University in St. Louis were awarded a National Science Foundation B iocomplexity grant (BE: CBC# 0221834) in 2003. Two sampling trips were made to Ambitle Island, the first in November 2003 and the second in May June 2005. The group consisted of four teams, each of which focused on different components of the hydrothermal system including the pore water and sediment chemistry, the archaebacterial community and their biogeochemical processes, the foraminiferan community, and the sediment eubacterial, eukaryotic and metazoan communities. The resul ts from the sediment and pore water analysis and the

PAGE 19

2 foraminiferan community analysis from this project have been published in two doctoral dissertations and a masters thesis ( Price 2008, McCloskey 2009, Engel 2010) and peer reviewed papers (Price and Pichler 2005, Pichler et al. 2006, Price et al 2007 Karlen et al 2010). The primary goal of my study is to investigate the effect of a natural hydrothermal gradient on the benthic infaunal community structure. To do this, I utilized measurements by collaborators for p ore water and sed iment arsenic concentrations, temperature, pH, and characterized the sediment composition and benthic fauna at different distances from a shallow hydrothermal vent. This dissertation wil l focus on the biodiversity and community structure of the eubacteria, eukaryota and the benthic infauna around the hydrothermal vent system and will fu rther look at the biocomplexity among the se different biotic communities and the physica l environment surrounding the hydrothermal vent. 1.2 Dissertation outline This d issertation is organized into seven chapters. The first chapter presents background information on the overall project and a review of other shallow water hydrothermal vent systems around the world and of past work conducted at Ambitle Island, Papua New Gu inea Chapter two present s results from the pore water and sediment analysis and set s the environmental context for the following chapters on the biotic communities. Chapter thr ee presents data on the macrofaunal community collected from the 2003 sampling trip and recently published in Pacific Science (Karlen et al 2010). Chapters four and five present the results from the 2005 macrofaunal and meiofaunal c ommunity analyse s respectively and their relation ships to the environmental

PAGE 20

3 parameters. Chapter six pr esents results from the molecular analysis of the eukaryotic and bacterial communities, and chapter seven summarizes the overall biocomplexity of the hydrothermal system and compare s trends in the community structure at the different levels of biodiversity presented in the previous chapters ( molecular meiofaunal and macrofaunal) and relate s these trends with the environmental conditions 1.3 A mbitle I sland hydrothermal vent system The shallow water hydrothermal vent system at Ambitle Island was first repo rted by Pichler and Dix (1996). Two types of venting occur in this system, focused venting of hydrothermal fluid at discrete openings of 1015 cm diameter in the seafloor and diffuse venting of gas bubbles and hydrothermal fluids through the sand and grave l sediments surrounding the focused vents (Pichler and Dix 1996). The focused vents discharge hydrothermal fluids at a rate of 300400 L/min at temperatures of 89 98C (Pichler and Dix 1996, Pichler et al 1999a). The diffuse venting consists of CO2 gas bubbles released through the sandy sediments as far as 150 m from the focused vent (Pichler et al 1999a) Previous work carried out on Ambitle Island focused on the geochemistry of the hydrothermal fluids and surrounding sediment. Vent and pore waters are enriched in several elements, but most notably, arsenic concentrations are 275 times that of normal seawater making this site one of the highest naturally occurring sources of arsenic on record ( Pichler et al. 1999b). Arsenic is naturally occurring in the volcanic bedrock of Ambitle Island and under heat and pressure is mobilized by ground water which is the source of the hydrothermal fluids at these vents (Pichler et al. 1999b).

PAGE 21

4 Despite the elevated concentration of arsenic, the surrounding reef habitat and fauna appear to be unaffected (Pichler and Dix 1996). This is likely due to the removal of arsenic from the water column via the precipitation of Fe (III) oxyhydroxides (Pichler and Veizer 1999, Pichler et al. 1999a ) which binds the bioavailable arse nic in the sediments (Price and Pichler 2005, Price 2008) In addition to the high arsenic concentrations, other environmental factors characteristic of the vent system are high temperatures low salinity and low pH from the degassing of CO2. The temperatu re and salinity extremes are limited to the immediate area of focused venting, but the reduced pH extends farther away due to the diffuse venting of CO2. The effects of the reduced pH on the foraminiferan communities at Ambitle Island were reported by McCl oskey (2009 ), wh o found low abundances of foraminiferans in the area of diffuse venting and by Engel ( 2010) who documented the dissolution of foram iniferan tests near the vent site. Karlen et al (2010) also noted that molluscan macrofauna were absent near the vent, which was attributed to the low pH. 1.4 Other shallow water hydrothermal vent systems Shallow water hydrothermal vents occur worldwide and there have been several earlier studies focused on the geochemistry and biology of such systems. Vents that have been studied in some detail include sites in the Mediterranean and Aegean Seas ( Varnavas and Cronan 1988, Thiermann et al. 1994 and 1997, Gamenick et al. 1998a Morri et al. 1999, Cocito et al. 2000 and reviews by Dando et al. 1999, 2000), Antar ctica (Bright et al. 2003, Deheyn et al. 2005), the Kurile Islands (Tarasov et al. 1990, Sorokin et al. 2003, Kamenev et al. 2004), New Zealand (Kamenev et al. 1993), Mexico and

PAGE 22

5 southern California (Melwani and Kim 2008) and Rabaul har bor, Papua New Guinea (Tarasov et al. 1999) The geochemistry at vent systems varies depending on the local geology and source water of the vent fluids; elevated arsenic concentrations have been reported at other vents (Varnavas and Cronan 1988) Some vent systems, such as th e one at Milos, Greece, show elevated levels of H2S and host communities of sulfide bacteria and sulfide tolerant infauna (Thiermann et al 1994, 1997, 2000, Gamenick et al 1998a,1998b Wenzhfer et al 2000) Unlike deep sea hydrothermal vents, which hos t a biological community of uniquely adapted species, t he bi ological communities around shallow water vents tend to represent a subset of the more stress tolerant species from the surrounding biota (Gimnez and Marn 1991, Thiermann et al 1997, Melwnai an d Kim 2008). Invertebrate taxa dominant at high sulfide vents do however exhibit adaptations to these environments and may host symbiotic sulphide metabolizing microbes. Gamenick et al (1998a ) studied the Capitella capitata population associated with the hydrothermal vent system at Milos Greece originally found by Thiermann et al (1997). They designated this population as Capitella sp. M and found through protein analysis and crossbreeding experiments with several other C. capitata sibling species popul ations that the Milos population represented a distinct sibling species which was characterized by a tolerance to high sulfide concentrations and anoxic conditions (Gemenick et al 1998a ). Bright et al (2003) sampled the meiofaunal comm unities associated with shallow water fumaroles at Deception Island, Antarctica. This system was also characterized by high sulfide concentrations. The dominant metazoan meiofaunal taxon was an unidentified flatworm which was tolerant of high temperatures and also had presum ably symbiotic

PAGE 23

6 bacteria colonizing the worms outer surface (Bright et al 2003). Similarly, a nematode found associated with bacterial mats around the shallow water vents in the Bay of Plenty, New Zealand, was found to have symbiotic sulfur bacteria which colonized the surface of the nematode s cuticle (Tarasov 2006). 1.5 Arsenic in the marine environment Elemental arsenic is commonly found in two inorganic forms in the marine environment : the oxidized form arsenate [As (V)] and the more toxic reduced fo rm arsenite [As (III)] (Francesconi and Kuehnelt 2002, Oremland and Stolz 2003, Watt and Le 2003). These compounds can be metabolized by several prokaryotes either through the respiration of arsenate or the oxidation of arsenite; particularly in extreme en vironments (Inskeep et al. 2002, Oremland and Stolz 2003, Oremland et al. 2005) including shallow water hydrothermal vents (Hand l ey et al 2009) Additionally, eukaryotic organisms can convert inorganic arsenic into organic compounds via methylation (Andre ae 1979, Kitts et al. 1994, Cutter et al. 2001) Specifically invertebrate organisms have adapted physiological mechanisms to detoxify or eliminate high concentrations of arsenic and other contaminants (Langdon et al 2003). Most metabolic pathways in marine invertebrates involve the reduction of As ( V) to As ( III) which is then methylated through a series of organoarsenical compounds typically to arsenobetaine (AB) as the end product (Langdon et al 2003, Argese et al 2005 Grotti et al 2010). T here is s ome evidence that these can be transferred and accumulated among different trophic levels ( Kirby and Maher 2002, Barwick and Maher 2003).

PAGE 24

7 1.6 The effects of p H Previous studies at the Ambitle Island vents found reduced pH levels in the vent fluids which were attributed to high levels of CO2 in the fluids and gas bubbles discharging from the vents (Pichler et al 1999b). Pore water pH is an important variable in geochemical processes and can influence sediment composition by the dissolution of carbonate sediments (Burdige and Zimmerman 2002) Initial field observations at the Ambitle Island study site indicated a distinctive gradation in the sediment composition in relation to the measured offshore distance from the vent. Field observations also indicated t hat carbonate content of sediment was lower nearer the vent, with carbonates increasing with distance from the vent. Near shore and in the vicinity of the vents the sediments were predominately fine grained volcanic sands, with some gravel sized deposits n ear the vent. A gradual mixture of biogenic carbonates was observed starting around 100 m from the vent, consisting of fragments of calcareous algae and coral rubble from the surrounding reef. Low pH can also impair carbonate incorpor ation in shell bearing organisms such as mollusks and foraminiferans (Green et al. 1993). McClo skey (2009) observed that shellbearing foraminiferans were absent from sediments near the Ambitle Island vent, which was attributed to the acidic conditions of the pore water. A simi lar trend was seen in the macrofaunal mollusks (Karlen et al 2010). There has been a lot of attention recently on global climate change and the effects of increased atmospheric CO2 from anthropogenic sources on lowering ocean pH, a process referred to ocean acidification or hypercapnia (Caldeira and Wickett 2003, 2005, Feely et al 2004, Pelejero et al 2005 ) and the possible effects this may have on marine

PAGE 25

8 organisms Atmospheric levels of CO2 have increased by nearly 100 ppm since the beginning of the Ind ustrial Revolution due mainly to the combustion of fossil fuels (Feely et al 2004), and current models predict fu rther increases during the 21st Century. The oceans act as a sink for around 30% of the CO2 released into the atmosphere (Feely et al 2004). The increased CO2 dissolved in the ocean water reacts with carbonate ions and water molecules forming carbonic acid: CO2 + H2O H2CO3 + + HCO3 The formation of carbonic acid results in a reduction in the pH of the ocean water and decreased saturation of carbonate ions (Caldeira and Wickett 2003). Globally, ocean pH has decreased by 0.1 pH units over the past 200 years and climate models predict the pH could drop by another 0.4 units by the end of the 21st Century (Feely et al 2004). Possible impacts o f lowered pH on marine organism s include bleaching and reduced symbiotic productivity in corals, as well as reduced growth and calcification in reef building corals (Anthony et al 2008); reduced calcification and dissolution in shell bearing organisms ( Fa bry et al 2008, Dupont et al 2010); physiological impacts due to acid base imbalance which can result in metabolic suppression and affect blood oxygen binding (reviewed in Fabry et al 2008) and reduced growth rates (Shirayama and Thornton 2005 W ood et al 2010). Other studies however, suggest that marine organisms are more adaptable to the projected changes in ocean pH and question the significance of long term impacts to the function of marine ecosystems (Hendriks and Duarte 2010, Hendriks et al 2010 Byrne et al 2010). Because of the release of CO2 from the hydrothermal fluids, shallow water vent systems offer a unique opportunity for looking at these effects on marine organism (Hall -

PAGE 26

9 Spencer et al 2008). Acidification of sea water associated with release of CO2 at shallow water vents has also been observed off of Italy by Hall Spencer et al. (2008). In their study they reported reduced densities of calcareous algae and invertebrates at sites with reduced pH and observed signs of shell dissolution o n live gastropods (Hall Spencer et al. 2008). Several other recent studies have taken advantage of shallow vent systems, including Tutum Bay, to look at the impacts of low pH on calcareous organisms such as bryozoans (RodolfoMetalpa et al 2010) and foram iniferans (Engel 2010). 1.7 Animal sediment relationships and biogeochemistry The relationship between sediment characteristics (grain size, organic carbon content) and the benthic infaunal community have long been recognized (Sanders 1958). The interrel ationships of sediment properties and benthic communities are complex (Snelgrove and Butman 1994), but in general, fine grained sediments and high organic carbon content tend to be dominated by deposit feeders while coarse sediments tend to support more suspension feeding organisms (Mancinelli et al 1998) Bioturbation by burrowing organisms can fu rther affect grain size and carbonate conte nt of the sediment (Aller 1982) as well as the geomorphology and the stabilization/destabilization of sediments (Paarlberg et al 2005). Burrowing organisms, furthermore influence the biogeochemical processes in the sediments (Snelgrove et al 1997, MermillidBlondin 2006).

PAGE 27

10 1.8 The concept of biocomplexity Biocomplexity is a fairly recent ter m that was first used in the late 1990s by the National Science Foundation as part of a new funding initiative to look at the complex interrelationships in the environment (Anand and Tucker 2003). Michener et al (2001) defined biocomplexity as properties emerging from the in terplay of behavioral, biological, chemical, physical and social interactions that affect, sustain, or are modified by living organisms and f urther go on to discuss that biocomplexity is an interdisciplinary field that takes a holistic approach to stu dying ecological systems as opposed to traditional reductionist methods. Cadenasso et al (2006) fu rther developed the concept by presenting a framework of biocomplexity which incorporates three dimensions : spatial heterogeneity; organizational connectivit y; and historica l contingency. Spatial heterogeneity is presented in terms of patches or discret e areas that differ in structure, organizational connectivity refers to the interactions among the defined patches, and historical contingency refers to the dev elopment of indirect effects over time (Cadenasso et al 2006). In terms of my study of the biocomplexity of the benthic communities associated with the hydrothermal vent system in Tutum Bay, the spatial heteroge neity component will encompass evaluating the changes in the physical environmental parameters and in the different biological communities along the hydrothermal gradient. Organizational connectivity is addressed by evaluating the interrelationships among the different hierarchal levels of the bio logical communities (macrofauna, meiofauna, molecular eukaryota and eubacteria). The historical contingency component of biocomplexity will

PAGE 28

11 only be briefly addressed in the comparison of the macrofauna results from the 2003 and 2005 sampling periods.

PAGE 29

12 Ch apter Two Site Description and Physical Characteristics 2.1 Introduction Benthic organisms are intimately tied to their physical environment. Factors such as pore water chemistry and the sediment characteristics determine the assemblage of species which are found in a given habitat and are thus the controlling factors structuring the biological community. The primary focus of this chapter is to establish the environmental context of the biocomplexity at Tutum Bay in terms of the pore water and sediment c haracteristics. To do this, sampling sites along the hydrothermal gradient and at a non hydrothermal reference site are classified using Principal Component Analysis (PCA), grouping them into discrete habitats based on their pore water chemistry and sedime nt type. The established habitat groups will serve as the basis of comparison for the different levels of biological communities in Tutum Bay. 2.2 Material and methods 2.2.1 Sampling design and field collection Samples were collected along a transect or iginating at the focused vent and running 300 meters offshore. Sampling sites were located at the vent (0 meters) and at 30 50 meter intervals along the transect for a total of nine transect sites; 0, 30, 60, 90, 120, 140, 180, 250 and 300 meters. Sample s were also collected at an additional reference site ( Danlum Bay, samples designated as Ref) located in a non hydrothermally influenced

PAGE 30

13 area 2 km south of the Tutum Bay vent site (Figure 2.1). The reference site was chosen based on the apparent similari ty of the sediment type to that of the vent transect. Figure 2.1 Ambitle Island, Papua New Guinea sampling locations. 2009 Google Earth; Image 2010 Terra Metrics and Europa Technologies. Figure 2.2 Focused venting at Tutum Bay

PAGE 31

14 Figure 2.3 Diffus e venting (background) at Tutum Bay. At each sampling site a 1m2 PVC quadrat, divided into 100 10x10 cm grid cells was placed on the sediment surface (Figure 2.4) Grid cells were numbered by an x,y coordinate system and random cells were selected for sam pling using a random number table ( Rolf and Sokal 1981). Within selected cells, samples were collected for pore water chemistry, sediment analysis and biology. The 1 m2 quadrat was repositioned an a dditional four times and sampling was repeated for a total of five replicate samples for each measured parameter at a given sampling site.

PAGE 32

15 Figure 2.4 Sampling quadrat and hand corer for macrofaunal samples. Temperature was measured in situ at a sediment depth of 5 cm using a Fisher Scientific Tractable dig ital thermometer (Price 2008). Pore water samples were collected by inserting a plastic probe attached to a 60 cc syringe to a sediment depth of 5 cm. Measurements for pH and oxidation reduction potential ( ORP ) were made on board using a MyronL pH meter. Salinity was measured using an optical refractometer. Procedures for the handling and analysis of the pore water samples for arsenic analysis are detailed in Price (2008). Laboratory analyses for total arsenic and arsenic speciation were conducted by the Center for Water and Environmental Analysis, University of South Florida Geology Department (Price 2008).

PAGE 33

16 2.2.2 Sediment analysis Surface sediments were collected for sediment grain size, organic carbon and carbonate content analysis using a 60 cc syringe modified as a sediment corer by cutting off the end (diameter = 3.0 cm). The sediment samples were dried at 60 C and a 50 60 g subsample was mixed with 50 ml of approximately 0.5% sodium hexametaphosphate then wet sieved through a 63 m mesh screen wi th deionized water to separate the sand and silt+clay fractions. The sand fraction and silt+clay fractions were rinsed into preweighed beakers and dried at 60C. Once dried, the two fractions were weighed and the sand fraction was sieved through a series of 29 sieves by mechanical shaking. The sieve series ranged from 8 mm ( The retained s ediment in each sieve was weigh ed and divided by the total dry weight to derive the dry weight percentage. The weight of any residual sediment falling into the bottom pan (<0.0625mm) was added to the silt+clay fraction for the final dry weight percentage. The percent error was calculated as a quality control check from the difference of the initial dry weight prio r to sieving minus the total recovered dry weight/initial dry weight. The weight percentages were farther grouped into the seven Wentworth size classes (Percival and Lindsay, 1997): Gravel (> 2mm; Very Coarse Sand (>1mm; The m edian grain size was calculated by plotting the cumulative frequency distribution Systat S oftware

PAGE 34

17 San Jose, CA ). The best fit equation was selected from a ranked list of possible curves based on a combination of the r2 value and on the visual inspection of the curves fit to the plotted data points. The median grain size was then calculated by substituting 0.5 for converted to millimeters using the formula: mm = 2The sorting coefficient was calculated using the Inclusive Graphic Standard Deviation method (Gray and Elliot 2009): ting coefficient calculation were derived from the best fit equation in TableCurve 2.0 (Systat Software, San Jose, CA) Sorting classification is as defined in Gray and Elliot (2009) The organic content of the sediment was determined by loss on igniti on (LOI) from 2 g of sample at 550C for 4 hours and sediment carbonate content were calculated after additional LOI at 950C for 2 hours following Heiri et al (2001). 2.3 Results Median, minimum and maximum readings for the pore water variables are show n in T able 2.1. Temperatures were highest at the vent with a median value near 70C ( T able 2. 1). The pore water temperature dropped rapidly away from the vent to near ambient seawater (approximately 30C; but were variable among replicate samples from each site between 0 m and 140 m (Table 2. 1; Figure 2.5).

PAGE 35

18 Pore water salinities were significantly lower at the vent relative to the other sites, but highly variable ranging from 3 25 psu (Table 2. 1 Figure 2.6 ). The salinity increased to near 30 psu at 30 m but was widely variable as far as 140 m along the transect with minimum readings as low as 25.5 psu (Table 2. 1, Figure 2.6 ). Salinities at the 18 0 m 250 m and 300 m sites showed less variability than the other transect sites and had median salinity valu es around 34 psu (Table 2.1). The median salinity at the reference site was slightly lower than at the farthest transect sites. The pore water pH values fell within three general groups along the transect : 0 60 m, 90 140 m, and 180 300 m with the ref erence site. The pH values at the 0 60 m sites were low and highly variable ( Table 2.1 ; Figure 2.7 ). Several individual measurements at the 90 140 m sites were also in the acidic range, with the lowest overall reading (5.81) at the 120 m site ( Table 2. 1). The pH approached normal seawater levels (near 8.0) farther away from the vent and were less variable among replicate samples however the reference site exhibited a slight ly lower pH than the farthest transect locations. The oxidization reduction pote ntial (ORP) was highly variable along the transect, but significantly lower at the vent site ( Table 2.1 ; Figure 2.8). With the exception of a few measurements at the vent, all of the transect sites had ORP readings within the oxidizing (i.e. positive ) ran ge. The ORP was significantly lower at the reference site relative to all of the transect sites. All of the ORP readings at the reference site were in the negative mV range indicating a reducing sediment environment. The pore water total arsenic concentrat ion was highest at the vent site and decreased by two orders of magnitude along the transect ( Table 2.1 ; Figure 2.9). Median

PAGE 36

19 300 m site and individual measurements ran ged widely at each site The pore water arsenic concentration was also slightly elevated at the reference site relative to the farthest transect sites ( Table 2.1 ; Figure 2.9). Table 2. 1 Site depths and median, minimum and maximum values for pore water var iables by site (n = 5 measurements per site). Site n Depth meters Temperature C Salinity psu pH ORP mV Pore water [As] 0 m 5 8.84 69.8 5.5 6.26 1.00 595.44 51.3 81.9 3.0 25.0 6.11 6.99 -32.00 32.00 128.09 868.85 30 m 5 10.06 31.6 30.0 6.34 126.00 49.65 31.4 33.6 29.0 30.5 6.05 7.23 96.00 136.00 21.95 151.93 60 m 5 11.58 34.6 31.5 6.40 108.00 38.61 31.7 35.6 30.0 33.0 6.03 6.69 66.00 112.00 17.63 76.02 90 m 5 12.50 31.1 34.0 7.68 155.00 20.94 30.9 31.3 33.5 35.0 6.65 7.81 148.50 165.00 10.08 33.84 120 m 5 13.11 31.1 33.5 6.13 112.00 34.24 29.8 34.2 25.5 34.0 5.81 7.41 83.00 130.00 11.18 114.37 140 m 5 14.63 30.4 33.0 7.34 42.00 19.72 29.8 35.4 26.5 33.0 5.96 7.57 32.00 87.00 13.20 57.98 180 m 5 17.68 30.7 33.5 7.81 72.00 9.90 30.7 30.7 33.0 33.5 7.73 7.98 69.00 89.00 4.88 17.39 250 m 5 23.77 29.6 34.0 7.80 79.00 3.98 29.6 29.7 33.5 34.5 7.64 7.94 22.00 144.00 3.03 7.72 300 m 5 28.35 29.6 34.0 7.78 78.00 3.77 29.6 29.7 33.5 34.0 7.61 7.83 75.00 132.00 2.38 9.58 Ref 5 12.50 30.0 32.5 7.46 54.00 12.75 29.9 30.0 32.5 33.5 7.32 7.65 -530* -41.00 5.88 15.38 Outlying value ex cluded from statistical analysis

PAGE 37

20 Figure 2.5 Ambitle Island May/June 2005: m ean pore water temperature 1 standard deviation. Figure 2.6 Ambitle Island May/June 2005: mean pore water salinity 1 standard deviation.

PAGE 38

21 Figure 2.7 Ambitle Island May/ June 2005: mean pore water pH 1 standard deviation. Figure 2.8 Ambitle Island May/June 2005: mean pore water oxidation reduction potential 1 standard deviation.

PAGE 39

22 Figure 2.9 Amb itle Island May/June 2005: mean pore water arsenic concentrations 1 standard deviation. The sediment characteristics at each site are summarized in Tables 2. 2 and 2. 3. Sediments at the vent site were composed primarily of gravel sized fragments, and the median grain size was significantly larger than at the other sites (T able s 2. 2 & 2. 3; Figure 2.10). Fine sands composed the largest fraction of the sediments from 30 140 meters along the transect and there was a steady, curvilinear increase in the median grain size moving out towards the 300 m site ( Table 2.2 & 2. 3; Figure 2.10). Fine sands were less prevalent at the 180 300 m sites as median grain sizes increased towards the medium to coarse sands range (Figure 2.10). The gravel size fraction also increased starting at the 180 m site and comprised over 25% of the sedimen ts at the 250 and 300 m sites ( Table 2.2). The reference site sediments were composed of fine sands with a median grain size similar to the 90 meter site (Table s 2. 2 & 2. 3; Figure 2.10 ).

PAGE 40

23 The sediment sorting coefficient at the vent site was significantly higher than at the other sites and indicated the sediments were very poorly sorted (Table 2. 3; Figure 2.11). The sediment sorting coefficients exhibited an increasing trend along the transect with values falling within the well sorted to the moderately well sorted range out to the 14 0 m site and increasing to the poorly sorted range at the 180 300 m sites (Figure 2.11). The sediments at the reference site were moderately to poorly sorted with a median sorting coefficient value falling right at the border between these two sorting classifications (Table 2. 3; Figure 2.11). The percent organic carbon in the sediments generally increased with distance from the vent and was highest at the reference site (Table 2.3; Figure 2.12). The percentage of carbonates in the sediments was low at the vent site and at the transect sites out to the 140 m site (Table 2.3; Figure 2.13). Carbonates increased significantly at the 180 m site and were highest at the 250 m and 300 m sites. There was no significant differ ence in the percent sediment carbonate between the reference site and the 180 m site, both of which were significantly lower than the 250 m and 300 m sites, but significantly higher than the 0 m 140 m sites (Figure 2.13).

PAGE 41

24 Table 2.2 Median, minimum an d maximum dry weight percentage of sediment by Wentworth size class. Site n Gravel % Very Coarse Sand % Coarse Sand % Medium Sand % Fine Sand % Very Fine Sand % Silt + Clay % 0 m 5 65.18 6.30 3.93 6.34 11.44 2.88 3.63 60.12 70.40 6.00 7.31 3.59 5.57 5. 71 8.11 9.39 14.14 1.75 3.50 1.86 4.41 30 m 5 0.00 0.38 0.30 2.71 61.10 33.29 2.10 0.00 0.05 0.22 0.73 0.26 0.33 1.42 2.83 56.82 63.87 30.25 38.87 2.04 2.81 60 m 5 0.00 0.07 0.11 2.91 68.06 26.86 1.57 0.00 0.06 0.04 0.69 0.00 0.28 2.02 11.75 66.99 76.53 9.97 28.98 1.50 1.64 90 m 5 0.07 1.78 0.77 7.93 76.14 11.86 1.02 0.04 0.47 0.71 3.46 0.39 1.23 6.39 9.56 73.37 77.74 10.55 13.68 0.97 1.67 120 m 5 0.00 1.05 1.16 16.83 74.45 4.23 1.17 0.00 0.15 0.48 2.56 0.69 1.86 14.10 19.67 72.92 77.14 3.62 5.86 1.03 1.26 140 m 5 0.04 2.75 5.70 27.28 60.33 2.22 1.27 0.00 0.22 1.70 4.42 4.01 6.67 23.26 37.53 51.76 64.64 1.00 3.32 1.00 1.45 180 m 5 13.25 8.82 11.35 27.61 34.51 3.45 0.88 8.11 19.99 8.65 10.67 9.63 14.80 23.41 30.72 27.14 36.28 1.67 5.11 0.84 2.18 250 m 5 26.52 13.23 14.05 20.79 22.73 4.06 1.15 9.20 27.54 10.89 14.25 12.55 17.25 18.94 26.51 20.66 25.76 3.70 5.61 1.05 1.63 300 m 5 28.74 15.94 14.93 17.72 16.32 2.84 0.99 22.94 33.15 14.01 20.46 13.69 18.09 15.42 19.69 13.08 18.67 2. 07 3.90 0.89 1.45 Ref 5 0.42 2.16 8.56 16.09 43.41 25.21 4.42 0.18 0.90 1.41 2.77 6.87 9.59 15.24 19.75 41.45 45.81 21.39 26.16 4.40 4.73

PAGE 42

25 Table 2. 3 Median, minimum and maximum measurements for sediment parameters. Site n Median Grain Size mm Sorting Coefficient Organics % CaCO3 % 0 m 5 4.33 2.34 2.24 0.73 3.49 5.35 2.12 2.45 1.93 2.45 0.48 0.96 30 m 5 0.14 0.45 1.81 0.32 0.13 0.15 0.44 0.46 1.44 2.06 0.25 0.46 60 m 5 0.15 0.42 1.55 0.39 0.14 0.17 0.41 0.43 1.41 1.82 0.31 0.47 90 m 5 0. 17 0.44 1.48 0.52 0.17 0.18 0.42 0.58 1.39 1.57 0.39 0.70 120 m 5 0.20 0.41 1.46 0.39 0.19 0.20 0.38 0.47 1.42 1.51 0.33 0.60 140 m 5 0.22 0.60 1.60 0.42 0.22 0.24 0.54 0.69 1.52 1.70 0.36 1.18 180 m 5 0.29 1.59 3.91 6.71 0.28 0.34 1.33 1.86 3.41 4.30 6.34 9.71 250 m 5 0.51 1.85 5.79 28.38 0.36 0.60 1.46 1.97 5.49 7.48 24.55 33.48 300 m 5 0.79 1.80 7.11 30.77 0.68 1.12 1.66 1.92 6.28 8.22 28.52 37.04 Ref 5 0.17 1.00 9.18 6.37 0.16 0.18 0.91 1.07 7.52 9.40 5.85 7.40

PAGE 43

26 Figure 2.10 Ambitle Island May/June 2005: mean median grain size 1 standard deviation. Figure 2.11 Ambitle Island May/June 2005: mean sorting coefficient 1 standard deviation.

PAGE 44

27 Figure 2.12 Ambitle Island May/June 2005: mean sediment organic carbon content 1 standard deviation. Figure 2.13 Ambitle Island May/June 2005: mean sediment carbonate content 1 standard deviation.

PAGE 45

28 Principal component analysis (PCA) of the sediment and pore water variables (Figure 2.14) shows fo ur general groupings of the s ampling sites based on their environmental attributes. The five 0 m replicates separate out from the rest of the samples along the first principal component axis (PC1). The wide distance between the five 0 m data points reflects the high variability in the physical parameters measured at that site. The remaining samples generally group together according to their distance along the transect. Three distinct groupings are seen: the 3 0 m 140 m samples; the 180 m + the reference site samples, and the 25 0 m + 300 m samples (Figure 2.14). Figure 2.14 Ambitle Island May/June 2005: Principal components analysis of pore water and sediment characteristics.

PAGE 46

29 The first two principal component axes account for 85% of the variation in the data ( Table 2.4 ). The fi rst principal component (PC1) accounted for nearly half of the variation ( Table 2.4 ) and was weighted primarily by the pore water temperature, salinity and total arsenic concentration ( Table 2.5 ). The second principal component (PC2) accounted for 35% of t he variation in the data ( Table 2.4 ) and was weighted largely by the sediment parameters particularly the sorting coefficient, and percent carbonates and organics ( Table 2.5 ). Table 2.4 Principal component eigenvalues and percent variation. Principal Component Eigenvalues %Variation Cum.%Variation 1 4.48 49.7 49.7 2 3.18 35.3 85.0 3 0.49 5.4 90.5 4 0.382 4.2 94.7 5 0.219 2.4 97.2 Table 2.5 Principal component e igenvectors. Variable PC1 PC2 PC3 PC4 PC5 Temperature 0.450 0.127 0.071 0.014 0 .059 pH 0.319 0.287 0.141 0.823 0.177 ORP 0.290 0.270 0.873 0.195 0.063 Salinity 0.410 0.155 0.355 0.113 0.577 Total pore water [As] 0.438 0.127 0.101 0.049 0.158 % Organic 0.221 0.442 0.092 0.431 0.565 % CaCO3 0.260 0.445 0.096 0.283 0.128 Median grain size 0.360 0.331 0.197 0.055 0.327 Sorting 0.079 0.534 0.143 0.023 0.402

PAGE 47

30 Table 2.6 presents the Spearman correlations between the measured pore water and sediment variables. For the pore water variables, there were strong negati ve correlations between temperature and pH and between temperature and salinity, while salinity and pH showed a strong positive correlation. The pore water pH was also negatively correlated with total pore water arsenic and positively with t he sediment carbonates. The ORP exhibited negative correlations with the sediment organic content and the sorting coefficient The total arsenic concentration was negatively correlated with salinity and sediment carbonates. The sediment organics carbonates and sorting coefficient were positively correlated with each other.

PAGE 48

31 Table 2.6 Spearman correlation coefficients between pore water and sediment characteristics. Significance (p) in parenthesis. pH ORP Salinity Total [As] %Organic %CaCO3 Median Grain Size Sorting Temp 0.78 0.09 0.76 0.88 0.56 0.67 0.25 0.29 ( 0.000 ) ( 0.549 ) ( 0.000 ) ( 0.000 ) ( 0.000 ) ( 0.000 ) ( 0.076 ) ( 0.044 ) pH 0.11 0.74 0.81 0.53 0.74 0.31 0.44 ( 0.451) ( 0.000) ( 0.000) ( 0.000) ( 0.000) ( 0.028) ( 0.001) ORP 0. 35 0.06 0.57 0.35 0.35 0.50 ( 0.013) ( 0.674) ( 0.000) ( 0.013) ( 0.012) ( 0.000) Salinity 0.75 0.13 0.49 0.14 0.06 ( 0.000 ) ( 0.355 ) ( 0.000 ) ( 0.334 ) ( 0.671 ) Total [As] 0.58 0.71 0.25 0.30 ( 0.000 ) ( 0.000 ) ( 0.077 ) ( 0 .034 ) %Organic 0.79 0.43 0.72 ( 0.000) ( 0.002) ( 0.000) %CaCO3 0.63 0.73 ( 0.000) ( 0.000) Median Grain Size 0.80 ( 0.000)

PAGE 49

32 2.4 Discussion The pore water and sediment variabl es show a strong environmental gradient with distance from the vent which can influence the structure of the surrounding sediment community High temperatures associated with tectonic activity are the defining characteristic of hydrothermal system s. Temper atures of the hydrothermal fluids at the Tutum Bay vent are near 100C when they emerge from the focused vent opening (Pichler and Dix 1996, Pichler et al 1999b ) and the temperatures of the interstitial pore waters near the vent were as high as 82C ( Tabl e 2.1 ). The variability in temperatures at several sites along the transect was due to the sporadic diffuse venting that occurred as far away as 15 0 m from the focused vent. This was most evident at the 60 m, 120 m and 140 m sites. McCloskey (2009) deploye d temperature data loggers at selected sites along the transect during the 2005 field sampling trip to Ambitle Island. The data loggers were buried approximately 1cm in the sediment and recorded temperature readings every minute over a s even day period Th ese loggers revealed localized temperature spikes of 10 15C extending as far as 150 m from the vent (McCloskey 2009) further showing the influence of diffuse venting along this section of the transect The reduced salinity observed at the vent is attr ibuted to the source water of the hydrothermal fluids, which is ultimately from meteoric (i.e. rain) derived groundwater from Ambitle Island (Pichler and Dix 1996, Pichler et al 1999b ). The high variability in the salinity measurements at the 0 m site may possibly be due to the entrainment of the ambient seawater during sampling, or could reflect the actual patchiness of the vent fluid percolating through the sediments surro unding the vent opening. The coa rse sediments at

PAGE 50

33 the 0 m site readily allowed mixin g of th e interstitial waters with the ambient seawater. T he median salinity value of 5.5 psu and minimum reading of 3 .0 psu however, agree with the salinities reported from earlier studies at Tutum Bay (Pichler et al 1999b). The variable s alinity measure ments observed along the transect at the 120 m and 140 m sites are due to the diffuse venting of hydrothermal fluids and correspond to the spikes in temperature observed at these two sites. With high temperature and low salinity the hydrothermal fluid is less dense than the surrounding seawater and quickly rises to mix with the surface waters (Pichler et al 1999b). Pichler et al (1999b) used the concentration of silica in the vent fluid as a tracer to model the dissipation of the vent water as it mixes with the surrounding seawater and found that the surface water over the vents showed a 10x higher concentration of vent fluid constituents relative to the bottom water at the vent opening. Those findings suggest that harmful constituents, such as arsenic, are rapidly transported away from the benthic habitat surrounding the vent, mitigating potential harmful effects on the benthic fauna. The pH values at all of the sites were lower than expected for normal marine systems, which are typically above 8.0 pH u nits. The pH measures had an overall range of 2.17 pH units, with the lowest reading of 5.81 and highest measurement of 7.98 pH units. By comparison, pH measured on the Great Barrier Reef by Gagliano et al (2010) only ranged 0.39 pH units across all habit ats with the lowest value being 7.98 from water extracted from a goby burrow. The lowest pH observation was at the 12 0 m site, with low measurements also at 14 0 m which reflects the influence of diffuse venting at these sites.

PAGE 51

34 The persistence of reduced p H out to even the 300 m site illustrates the extent of influence the hydrothermal venting has on the environment at Tutum Bay. Studies at other shallow water vent systems have also reported low pH values comparable to Tutum Bay. For example, Melwani and Kim (2008) reported mean pH values of 6.5 at their Baha Concepcin vent site and 7.2 at their White Point site. At both sites their outside zone pH values were 7.8 and 7.4 respectively (Melwani and Kim 2008), although they did not specify the distance f rom the vent area these samples were taken. HallSpencer et al (2008) reported pH as low as 6.57 at a shal low volcanic vent in Italy. That study focused specifically on the effects of reduced pH on calcareous organisms, using the vent system as a natural laboratory to study the potential ecosystem impacts of ocean acidification due to the projected increase of atmospheric CO2 over the next century (Hall Spencer et al 2008). The pH values observed at Tutum Bay, even at the farthest site from the focused v enting in predominately carbonate sediments, were still at or below the lowest values expected in a natural coral system (Gagliano et al 2010) and less than endof century values predicted by current climate models. The ORP measurements indicated that al l of the sites had oxidizing sediment environments at 5cm with the ex c eption of the Danlum Bay reference site. The low ORP values seen at the reference site were due to the high organic content of these sediments. The Danlum Bay reference site was near sho re and close to a small stream flowing from the island This resulted in the input of a significant amount of terrestrial detritus The ORP values were also low but oxidizing, at the 0 m site, with one of the replicate sample s recording a negative (reduci ng) value. Pichler et al (1999b) found that the vent

PAGE 52

35 fluids are reduced and the lower ORP values at the 0 m site are a result of mixing of the vent fluids with the ambient seawater. Past studies at the Tutum Bay vents discovered that the vent fluids and surrounding sediment deposits had the highest naturally occurring concentrations of arsenic recorded in a marine environment (Pichler and Dix 1996, Pichler and Veizer 1999, Pichler et al 1999a&b). Results presented here show nearly two order s of magnitude difference in the pore water arsenic concentrations along the transect. Arsenic concentrations of seawater near the 300 m site were still higher than normal (~2 was most likely due to the influen ce of diffuse venting. The slightly elevated arsenic concentrations measured at the reference site may reflect the transport of volcanic sediments from Ambitle Island. The sediment characteristics also define specific habitat types along the transect. At the 0 m site, the sediment was composed of coarse, very poorly sorted volcanic gravel. The deposition of coarse sediments is indicative of a high energy environment and may be due to the constant resuspension and transport of finer grained sediments offsho re (Gray and Elliot 2009) This is likely caused by ac tive focused venting a nd the influence of wave action at the shallower depth s. The transect sites from 3 0 m out to 140 m w ere characterized by fine, well sorted volcanic sand that exhibited a steady in crease in the median grain size with distance from the vent. This sediment type indicates a relatively lower energy, stable environment which allows for the deposition of finer grained sediments (Gray and Elliot 2009) The transect was surrounded by high er relief areas of coral reef, particularly towards the shoreline and the area of focused venting. This may have sheltered the inshore part of the transect from the effects of waves and currents.

PAGE 53

36 Diffuse venting occurred mostly in this inshore portion of t he transect resulting in low pore water pH which likely caused the dissolution of larger grained biogenic carbonate sediments. The sediments appeared to be extensively rework ed by burrowing shrimp ( Thalassinid ea spp.) at th is site which may have influenc ed the sediment distribution and geochemistry. This will be discussed fu rther in the chapters on the macrofaunal communities. The 18 0 m site represe nted a mixing zone of the finer grained volcanic sands and biogenic carbonate sediments. The 250 m and 300 m where characterized by coarsegrained carbonate sediments comprised of fragments of shells, calcareous algae and coral rubble. The larger sediment grain size at these sites reflects the increas ed influence of offshore currents, which were noticeable espe cially at the 30 0 m site. The Danlum Bay reference site also represented a transitional sediment zone between the nearshore, volcanic sands originating from Ambitle Island, and biogenic calcareous sands from the offshore reef environment This site had sim ilar grain size characteristics as the inner transect sites, but was more similar to the 18 0 m site in terms of its mixed volcanic and carbonates sediments. 2.5 Summary and conclusions The study sites consisted of four distinct habitat types characterized by pore water and sediment types The 0 m site represented a unique habitat characterized by high temperature and pore water arsenic concentrations, low pore water salinity and pH, and co a rse, very poorly sorted volcanic gravel with low organic carbon an d carbonate content The transect sites from 3 0 m to 140 m formed a distinct habitat type characterized by the zone of diffuse venting. The shared attributes of these sites were

PAGE 54

37 near ambient temperature and salinity, low pH, elevated arsenic concentration and very wellsorted, fine grained volcanic sands with low organic carbon and carbonate content. The 18 0 m and reference site formed a third habitat type and shared similar temperature, salinity, pore water arsenic concentrations and sediment carbonate com position. These sites represented a transitional zone between the near shore volcanic sediments and the offshore carbonate sediments. The reference site stood out from all the other sites along the transect due to its high organic carbon content The 25 0 m and 300 m sites represented a fourth distinct habitat type, fu rther removed from the effects of the hydrothermal activity These two offshore sites were characterized by near ambient temperature and salinity values, slightly elevated arsenic concentratio ns higher pH and poorly sorted, coarse carbonate sediments with relatively high organic carbon. Each of these four habitats or zones exhibit distinct environmental characteristic s which in turn influence the biological community structure found there. Th e succeeding chapters will focus on t hese biological communities at the macrofaunal, meiofaunal, and molecular eukaryotic levels an d the influence of the physical environment in structuring these communities.

PAGE 55

38 Chapter Three Changes in Benthic Macrofauna associated with a Shallow water Hydrothermal Vent Gradient in Papua New Guinea ( This chapter has been published in Karlen et al 2010) 3.1 Introduction Shallow water hydrothermal vents occur worldwide, but relatively few studies have been carried out on the benthic community structure of these systems (Thiermann et al 1997). Several such systems that have been studied in some detail include sites in the Mediterranean and Aegean Seas (Thiermann et al 1997, Gamenick et al 1998a Morri et al 1999, Cocit o et al 2000), the Kurile Islands (Tarasov et al 1990, Sorokin et al 2003, Kamenev et al 2004), New Zealand (Kamenev et al 1993), the Gulf of California and southern California (Melwani and Kim 2008) and several sites in Papua New Guinea including Rab aul harbor (Tarasov et al 1999), and Ambitle Island (Pichler and Dix 1996, Pichler and Veizer 1999, Pichler et al 1999a,b). The last site was the focus of the current study. Previous work carried out on Ambitle Island focused on the geochemistry of the hydrothermal fluids and surrounding sediment. Vent and pore waters were enriched in several elements, but most notably, arsenic concentrations were 275 times that of normal seawater (Pichler et al 1999b). Despite the elevated concentration of arsenic, the surrounding reef habitat and fauna appear to be unaffected (Pichler and Dix 1996). This

PAGE 56

39 is likely due to the removal of arsenic from the water column via the precipitation of Fe (III) oxyhydroxides (Pichler and Veizer 1999, Pichler et al 1999a). Because o f this previous work, we focused this study on the influence of arsenic, along with other vent related factors such as temperature, pH and sediment characteristics on the benthic macrofauna near the Ambitle Island vent. Arsenic is naturally occurring in A mbitle Island ground water which is the source of the hydrothermal fluids at these vents (Pichler et al 1999b). Elemental arsenic is commonly found in two inorganic forms: arsenate [As (V)] and arsenite [As (III)] (Francesconi and Kuehnelt 2002, Oremland and Stolz 2003, Watt and Le 2003). These compounds can be metabolized by several prokaryotes either through the respiration of arsenate or the oxidation of arsenite, particularly in extreme environments (Inskeep et al 2002, Oremland and Stolz 2003, Oreml and et al 2005). Additionally, eukaryotic organisms can convert inorganic arsenic into organic compounds via methylation (Andreae 1979, Kitts et al 1994, Cutter et al 2001) and there is some evidence that these can be transferred among different trophic levels (Barwick and Maher 2003). Pore water pH is an important variable in geochemical processes. Previous studies at the Ambitle Island vents found reduced pH levels in the vent fluids which were attributed to high levels of CO2 in the fluids and gas bu bbles discharging from the vents (Pichler et al 1999b). Additionally, low pH can influence sediment composition by the dissolution of carbonate sediments (Burdige and Zimmerman 2002) and impair carbonate incorporation in shell bearing organisms (Green et al 1993). Acidification of sea water associated with release of CO2 at shallow water vents has been observed off of Italy by HallSpencer et al (2008). In their study they reported reduced densities of calcareous

PAGE 57

40 algae and invertebrates at sites with red uced pH and observed signs of shell dissolution on gastropods (Hall Spencer et al 2008). Initial field observations at the study site indicated a distinctive gradation in the sediment composition in relation to the measured offshore distance from the vent The relationship between sediment characteristics (grain size, organic carbon content) and the benthic infaunal community have long been recognized (Sanders 1958). The interrelationships of sediment properties and benthic communities are complex (Snelgr ove and Butman 1994), but in general, fine grained sediments and high organic carbon content tend to be dominated by deposit feeders while coarse sediments tend to support more suspension feeding organisms. Field observations also indicated that carbonate content of sediment was lower nearer the vent, with the carbonate fraction increasing with distance from the vent. Near shore and in the vicinity of the vents the sediments were predominately fine grained volcanic sands, with some gravel sized deposits near the vent s A gradual mixture of biogenic carbonates was observed starting around 100 m from the vent, consisting of fragments of calcareous algae and coral rubble from the surrounding reef. Bioturbation by burrowing organisms can farther affect grain size and carbonate content of the sediment (Aller 1982). The primary goal of this study is to investigate the effect of a natural hydrothermal gradient on the benthic macrofauna. To do this, we measured pore water and sediment arsenic concentrations, temperat ure, pH, and sediment composition, and characterized the benthic fauna at different distances from a shallow hydrothermal vent. It was expected that elevated arsenic concentrations would be a significant factor in

PAGE 58

41 structuring the benthic community. Studies of similar systems focused on depth rather than distance from the shallow vents (Tarasov et al 1999), sampled at only select sites (Kamenev et al 1993) or short transects of 5 m (Thiermann et al 1997). This study is unique in that sampling was carried out over a 150 m transect and significant effects were found even at that distance. 3.2 Material and methods The study site was located in Tutum Bay, Ambitle Island, Papua New Guinea (4 05S Latitude, 153 40 W Longitude) (Figure 3.1). The island is an active volcano, with several large hot springs on shore in addition to the submarine hydrothermal vents. The hydrothermal vents in this study were located at a water depth of 10 m and were surrounded by an extensive coral reef system with intermittent pat ches of sand. The main vent had a focused discharge of 300400 L/min hydrothermal fluids through a narrow, 15 cm diameter opening in the sea floor (Pichler et al 1999b). In the sandy areas immediately surrounding the vent, a diffuse discharge of hydrother mal fluids and gases was observed as a steady stream of bubbles rising from the sediment. This diffuse venting was strongest near the main vent, but appeared sporadically as far as 60 m away (Figure 3.2).

PAGE 59

42 Figure 3.1 A: Map of Papua New Guinea and its geo graphic relation to Australia. Ambitle Island is marked with an arrow and is located east of New Ireland. Latitude and longitude are marked in degrees. B: Map of Ambitle Island with the study sites marked with arrows. Latitude and longitude are indicated in degrees and minutes.

PAGE 60

43 Figure 3.2 Photograph of focused hydrothermal vent at Tutum Bay, Ambitle Island, Papua New Guinea with surrounding diffuse venting (Photo credit: DJK). Benthic macrofauna samples were collected at three sites along a transect ru nning from a focused hydrothermal discharge. The transect ran southwest from the vent along a sandy bottom for 30 m, and then continued offshore to the west to avoid the reef. The maximum depth along the transect was 15 m. Sample sites were at 7.5, 60 and 150 m along the transect line, representing a subset of sites chosen for geochemical analysis. These sites were selected based on sediment temperature profiles conducted along the transect which indicated that each site would represent a decreasing trend in the hydrothermal influence with distance from the vent, with the 15 0 m site presumably outside the hydrothermal zone. Additional samples were collected at a single reference

PAGE 61

44 site located approximately 1.6 km north of Tutum Bay in an area not affected by hydrothermal activity and at a similar depth as the transect sites. Sampling was limited by logistical problems encountered in shipping supplies to the site and in shipping samples back to the laboratory in the U.S At each site along the transect a 1 m2 grid was placed next to the transect line. A single grid was sampled at the reference site. The grid was subdivided into 100 10x10 cm cells and five cells were selected for sampling using a random number table ( Rolf and Sokal 1981). Five core samples (diam eter = 7.62 cm; area = 45.6 cm2) were taken from random grid cells within the 1m2 quadrat to a depth of 15 cm. The core samples were sieved through a 500 m mesh sieve and the retained animals were fixed in 10% formalin with Rose Bengal stain for a minimum of 72 hours then transferred into 70% methanol for preservation. Organisms were sorted from the sediment under a dissecting microscope, identified to the lowest practical taxonomic level and enumerated. Three smaller cores (diameter = 3.0 cm) were taken at randomly selected grid cells for sediment grain size, organic carbon and carbonate content analysis at each site. The sediment samples (50 g) were dry sieved through a stacked series of sieves and divided into six Wentworth Size Classes (Percival and L indsay, 1997): Gravel (> 2000 m), Coarse Sand (> 500 m), Medium Sand (>250 m), Fine Sand (> 125 m), Very Fine Sand (> 63 m), Silt + Clay (< 63 m). The organic content of the sediment was determined by loss on ignition (LOI) from 2 g of sample at 550C for 4 hours and sediment carbonate content was calculated after additional LOI at 950C for 2 hours following Heiri et al (2001). Methods used for the field measurements of temperature and pH and the collection and analysis of pore water and sediment chemistry samples

PAGE 62

45 are described in Price and Pichler (2005). Pore water measurements were taken at a depth of 10 cm except where noted in Table 3. 1. Sediment arsenic content indicated in Table 3. 2 is from surface sediment samples. Parametric statistica l analysis was conducted using SigmaStat 3.5 (Systat Software, Richmond CA). One Way Analysis of Variance (ANOVA) with a Holm Sidak multiple comparison procedure test was used to test for between site differences in sediment characteristics and benthic co mmunity metrics (Species Richness, Abundance, Diversity). A nonparametric Kruskal Wallis ANOVA on ranks was used to test for the presence of between site differences in the Pielous evenness index because index values failed normality testing. All other me trics passed for normality without data transformations. Sediment grain size, organic carbon and carbonate content data were arcsine transformed for normality prior to analysis. Between site statistical testing for the pore water temperature, pH and sedime nt/ pore water arsenic concentrations were not performed since only one measurement or sample was collected per site. Basic community indices (Species Richness, Abundance, ShannonWien er Diversity and Pielous Evenness) and multivariate analysis was done us ing PRIMER v.6 (Clarke and Gorley 2006). The Bray Curtis similarity index was calculated on square root transformed abundance data to measure the similarity of the benthic species composition among replicates within a site and between sites. Analysis of S imilarity (ANOSIM) was used to test for differences in the species composition between sites and The Similarity Percentage (SIMPER) procedure was used to determine which taxa contributed to the within site similarity or between site differences. The BIO E NV analysis procedure in PRIMER v.6 was used to calculate Spearman ranked correlations between the

PAGE 63

46 physical/sediment measurements and the benthic macrofauna composition. All physical and sediment data were log transformed and normalized prior to analysis. The BIO ENV procedure compares the underlying similarity/distance matrices for the biological and physical data to find the combination of physical variables that best explains the ordination pattern among the samples based on their species similarity ( Clarke and Ainsworth 1993). This method has several advantages in that it incorporates the full range of biotic and physical data collected and is not dependent on a normal distribution of the data because it utilizes rank based correlations. 3.3 Resul ts There were observed differences between sites for all physical parameters (Table 3. 1). Temperature showed the highest value at 7.5 m, but decreased to near ambient levels at 60 m. The pH was lowest at 7.5 m and increased with distance from the vent. Po re water arsenic was highest near the vent, decreased substantially at 60 m and increased again at 150 m. All three transect sites had greatly elevated arsenic concentrations relative to the reference site.

PAGE 64

47 Table 3.1 Summary of pore water characteristic s at Tutum Bay, Ambitle Island, Papua New Guinea November 2003. All pore water measurements are from Price and Pichler (2005) and represent values measured at a depth of 10cm except = 0 cm. Distance (meters) Temp. ( C) pH [As] 7.5 45.5 6.2 81* 60.0 33.3 6.8 16 150.0 31.0 7.2 63 REF 30.2* 7.6 3.3 Sediment characteristics also exhibited differences between sites along the transect and relative to the reference site (Table 3. 2). The sediment at 7.5 m was predominantly gravel and very fine s ands with low organic carbon and carbonates. The 60 m site was dominated by very fine sands low in organic carbon and carbonate. The 150 m site was also predominately very fine sand, but had a greater proportion of medium and fine sized sediments. The reference site had a high percentage of gravel largely composed of coral rubble, as well as a large amount of medium sands. There was a significant difference in the percent organic carbon between sites ( F3,8 = 95.5; P < 0.001), with the 150 m site having high er values than at the 7.5 and 60 m sites and the reference site having higher values than all three transect sites. The percent carbonates were also significantly higher at the reference site ( F3,8 = 2011; P < 0.001). Total arsenic in the surface sediments exhibited a decreasing trend along the transect away from the vent but all three transect sites were higher than the reference site.

PAGE 65

48 Table 3.2 Sediment grain size class percentages, percent organic carbon and carbonates (mean 1 standard deviation); ar senic concentrations from Price and Pichler (2005). Wentworth Size Class 7.5 m 60 m 150 m REF Gravel 32.99 10.39 0.17 0.06 0.92 0.22 24.83 10.26 Coarse Sand 0.71 0.33 0.23 0.17 0.19 0.02 1.32 0.14 Medium Sand 4.38 2.27 4.30 0.32 17.8 7 2.14 36.24 0.08 Fine Sand 7.36 1.53 3.50 0.43 36.15 1.32 12.54 2.25 Very Fine Sand 53.10 8.61 89.60 1.15 42.58 1.04 18.10 5.76 Silt + Clay 1.46 0.11 2.20 1.18 2.30 0.68 6.97 2.61 % Organic Carbon 1.66 0.08 1.33 0.29 3. 57 0.43 7.59 1.04 % Carbonates 0.51 0.11 0.29 0.18 3.75 0.63 50.37 0.99 [As] (ppm) 783 614 402 2.2 Both the number of taxa ( S) and overall abundance ( N ) increased with distance away from the vent (Table 3 .3 ) and all transect sites were s ignificantly lower than the reference site for both measures (Figure 3. 3A; F3,16 = 44.1 and 27. 6 for S and N respectively; P < 0.001). The 150 m site also was significantly higher in both numbers of taxa and abundance than the other two transect sites; ho wever there was no significant difference between 7.5 and 60 m for either measure. The Shannon Wiener diversity index ( H ) also showed an increasing trend with distance from the vents (Figure 3. 3B). All sites

PAGE 66

49 were significantly different from each other ( F3,16 = 77.7, P < 0.001). There was no significant difference in evenness between sites ( df = 3, H = 0.766 P = 0.86). Most taxa were represented by a single individual resulting in high evenness values. Table 3.3 Benthic community index values cumulated fo r all five replicate samples at each site. Mean values 1 standard deviation are shown in parenthesis. Abundance ( N ) expressed as count/sample; sample area = 0.00456m2. Distance (meters) Number of Taxa (S) Abundance (N) Diversity (H ) Evenness (J ) 7.5 6 (2 0.71) 14 (2.8 1.3) 1.48 (0.59 0.38) 0.82 (0.92 0.08) 60.0 20 (5.8 3.03) 38 (7.8 5.4) 2.82 (1.60 0.36) 0.94 (0.98 0.02) 150.0 55 (17 3.87) 111 (22.2 7.69) 3.71 (2.73 0.21) 0.92 (0.97 0.02) REF 116 (31.6 7.64) 201 (40.2 10.92) 4.50 (3.32 0.25) 0.95 (0.97 0.01)

PAGE 67

50 Figure 3.3 A: The mean number of taxa ( S ) and abundance ( N ) of benthic macrofauna/sample at the transect and reference sites. Error bar = 1 standard deviation, sample area = 0.00456 m2. All pairwise compariso ns between the REF, 150 m and 6 0 m sites were significantly different for both S and N The 6 0 m and 7.5 m sites were not significantly different. B: Mean ShannonWiener Diversity ( H ) and Pielous Evenness ( J ) values at transect and reference sites. Erro r bar = 1 standard deviation. All pairwise comparisons between sites were significantly different for H There was no significant difference in J between sites.

PAGE 68

51 Annelids comprised the largest percentage of taxa at all sites, (Fig. 3. 4A) and were represented by 92 taxa overall or > 50% of the taxa identified in this study. The 7.5 m site had only three annelid taxa, which represented half of the taxa found due to the low species richness at that site. The 6 0 m and 150 m sites had 10 and 24 annelid taxa respectively, while the reference site had 60 annelid taxa present. Annelids also were the most abundant invertebrates, comprising close to 50% of the overall abundance for all sites, followed by arthropods which represented close to 25% of the total abun dance overall (Fig ure s. 3. 4A and B). Mollusks were also prevalent at 150 m and at the reference site (12 and 11 taxa respectively), but were absent at 7.5 m and were represented by only one species at 60 m (Fig ure s. 3. 4A and B). Analysis of Similarity ( ANOSIM) indicated that the species composition was significantly different between sites (Global R = 0.764, P = 0.001). The SIMPER Analysis by distance (Table 3.4) showed that the five replicates from 7.5 m had an average similarity of 29% due to the presence of the polychaetes Malacoceros sp. A and Capitella cf. capitata which accounted for 52% and 48% of the similarity respectively. The 60 m replicates had an average similarity of 21% with the unidentified cumacean (designated as Cumacea sp. A) contributi ng 50% to the similarity. The 150 m samples had an average similarity of 26% mostly due to an unidentified sipunculan worm (sipunculan sp. B), an isopod, (Paranthuridae? sp. A) and the polychaete Heteropodarke sp. A. The five replicates from the reference site had an average similarity of 15.6%, with the polychaete Typosyllis sp. A accounting for 11% of the similarity between the samples.

PAGE 69

52 Figure 3.4 A: Mean number of taxa by major taxonomic groups at transect and reference sites. B: Mean abundance of major taxonomic groups at transect and reference sites.

PAGE 70

53 Table 3.4 SIMPER analysis results by distance (Bray Curtis Similarity among five replicate samples) and percent contribution of top five taxa. Rank 7.5 meters Avg. Similarity = 27.40 60 meters Avg. Si milarity = 21.36 150 meters Avg. Similarity = 26.48 REF Avg. Similarity = 15.59 Taxon % Contribution Taxon % Contribution Taxon % Contribution Taxon % Contribution 1 Malacoceros sp. A 52.02 Cumacea sp. A 49.88 Paranthuridae? sp. A 20.39 Typosyllis sp. A 11.32 2 Capitella cf. capitata 47.98 Platyischopidae? sp. A 21.73 sipunculan sp. B 20.39 Lucinidae sp. A 7.52 3 Glycera sp. B 11.71 Heteropodarke sp. A 13.36 Thalassinid ea? sp. A 6.10 4 Sthenelais sp. A 8.65 Cumacea sp. A 7.41 Eusamytha? pacifica? 5.90 5 Capitella cf. capitata 8.03 holothuroidean? sp. A 6.38 Ehlersia sp. A 5.81 Cumulative 100.00 100.00 67.93 36.66

PAGE 71

54 The cluster analysis of abundance data showed the sites clustering in three main groups designated as A, B, and C (Figure 3. 5) Group A consisted of all five replicates from 7.5 m along with two replicates from 60 m. SIMPER analysis of the cluster groups showed that the Group A sites had an average similarity of 22%, with the polychaete Capitella cf. capitata accounting for 64% o f the similarity. Group A was divided into two subgroups: A1 and A2. The A1 subgroup consisted of three of the 7.5 m samples and had an average similarity of 48% due to the presence of the polychaete Malacoceros sp. A. The A2 subgroup was composed of two 7.5 m samples and two 60 m samples with an average similarity of 32% due to the polychaete Capitella cf. capitata. Group B was composed of the remaining three 60 m samples and all five 150 m samples and had an average similarity of 15%. The similarity amon g the group B samples was due to the abundance of Cumacea sp. A which accounted for 33% of the similarity. Group B was divided into two subgroups; B1 and B2. The B1 subgroup included three replicates from 60 m and had an average similarity of 25% due to th e presence of Cumacea sp. A. The B2 subgroup consisted of the five 150 m replicates and had an average similarity of 26.5% due to the isopod Paranthuridae? sp. A and the sipunculan sp. B. Group C consisted of the five reference site samples and had an aver age similarity of 15.6% due in part to the polychaete Typosyllis sp. A.

PAGE 72

55 Figure 3.5 Cluster analysis: Bray Curtis Similarity based on square root transformed macrofaunal abundance data. The BIO ENV analysis indicated that for single environmental variabl es, pH best explained the b enthic species assemblage data [ Spearman correlation (rho) = 0.703] ; followed by temperature ( = 0.682) (Table 3. 5). The combination of pH and temperature with various sediment grain size classes resulted in a slightly higher correlation ( = 0.709; Table 5). Conversely, the total arsenic concentration in the s ediment and pore water showed relatively weaker correlations with the benthic macrofauna assemblage (Table 5; = 0.568 and 0.429 respectively).

PAGE 73

56 Table 3.5 BIO ENV results for single and multiple parameters. Variable Spearman Correlation Coefficient ( s ) % CS; Temperature; pH 0.709 % G; %CS; % FS; Temperature; pH 0.709 %G; %FS; Sediment As; Temperature; pH 0.709 pH 0.703 %CO 3 ; Temperature; pH 0.703 %G; %FS; %CO 3 ; Temperature; pH 0.703 Temperature 0.682 % Silt + Clay 0.603 % Fe 0.568 Sediment As 0.568 Pore water As 0.429 %G = % Gravel; %CS = % Coarse Sand; %FS = % Fine Sand; %CO3 = % Carbonates; %Fe = % iron .

PAGE 74

57 3.4 Discussion Despite the small sample size and limited area sampled at each site, there was a statistically significant trend of in creasing species richness, abundance and diversity with distance from the vent out to 15 0 m These measures were still depressed relative to the reference site, suggesting that the hydrothermal system still influenced the infaunal structure as far away as 150 m. The benthic macrofauna composition at the reference site showed little similarity to any of the transect sites as revealed by the cluster analysis (Figure 3. 5). Among the three transect locations the 7.5 and 150 m species assemblages showed little similarity to each other while the benthic community at 60 m appeared to be transitional between the other two transect sites. Several physical variables could be responsible for the observed trends in the benthic macrofauna assemblage, including temperature, pH, arsenic content and sediment composition. Other factors such as currents, tides, wave action, and changes in the volume of vent flow could also affect the vents influence on the infaunal community but these variables were not observed in this stu dy. There is a correlation between increasing pH and increasing presence of carbonates in the sediment. These changes in sediment composition and benthic infaunal abundance and diversity may be influenced by both geochemistry and bioturbation due to burro wing organisms (Aller 1982, Dhalgren et al 1999, Widdicombe et al 2000). Numerous active callianassid shrimp burrows were observed along the transect. In our core samples a single individual was found at 60 m, two at 150 m and four at the reference sit e. However, callianassid shrimp are likely underrepresented in our samples due to the depth of their burrows. Additionally,

PAGE 75

58 observations made during a night dive along the transect revealed high densities of sea pens (Octocorallia: Pennatulacea) starting approximately 100 m out from the vent in fine grain, volcanic sand. These apparently remain in deep burrows during the day and thus were not represented in our core samples. The change in macrofaunal composition along the transect is most strongly correl ated with pH although temperature and grain size were also important as shown in the BIO ENV results (Table 3. 5). Mollusks were completely absent at the 7.5 m site, a single opisthobranch gastropod (Cylichnidae genus undet ermined ) was found at the 60 m site while the 150 m site contained 12 molluscan species or 22% of the species present at that site. The increased presence of shelled mollusks may be caused by the pH gradient along the transect as solid carbonates are more difficult to precipitate and maintain at low pH (Milliman 1974). The apparent lack of molluscan fauna near the vent in this study differs from studies conducted at three other shallow hydrothermal systems (Kamenev et al 1993, Thiermann et al 1997, Kamenev et al 2004), where moll usks were found in association with hydrothermal vents. These sites were different from Ambitle Island in that hydrogen sulfide concentrations seemed to be an important factor influencing infaunal benthic communities. Those studies did not report pH value s, so the association between pH and molluscan communities was not established. One recent study (Hall Spencer et al 2008), did show that reduced pH from the release of CO2 in a shallow volcanic vent caused reductions in the abundance of coralline algae and other calcareous organisms. One species of note in this study is the capitellid polychaete Capitella cf. capitata which was found at both the 7.5 and 60 m sites. This cosmopolitan taxon has long been recognized as a complex of several morphologically s imilar species and historically has

PAGE 76

59 been associated with degraded or disturbed habitats (Grassle and Grassle 1976). Thiermann et al (1997) found a sibling species of Capitella capitata associated with the transition zone near a shallow water hydrothermal vent in the Aegean Sea. This population was designated as Capitella sp. M and was characterized by its tolerance of high sulfide levels (Gamenick et al 1998a ). High sulfide concentrations have not been associated with the Ambitle Island vents (Pichler et al 1999b), but the C. capitata specimens found in this study were associated with high temperatures and high arsenic levels as well as low pH. Our results suggest that the infauna near the vents are represented by few taxa and low abundances and primarily cons ist of opportunistic taxa such as Capitella capitata These taxa are most likely a subset of the surrounding benthic community rather than specialized organisms adapted to a chemosynthetic metabolism as found in deep sea hydrothermal vents. A result s imilar to ours was found in shallow hydrothermal vent systems near southern California and the Gulf of California (Melwani and Kim 2008). These results suggest that the arsenic from the vent fluids is effectively sequestered in the sediments as found in earlier studies (Pichler et al 1999a) and does not significantly influence either benthic community structure or the observed surrounding reef habitat. Instead, the reduced pH appears to affect the benthic species composition around the vent which is evide nced by the absence of mollusks near the vent, the reduced molluscan fauna at 60 m and the absence of carbonate sediments at both of these sites. The influence of pH on the benthic community structure is farther supported by the BIO ENV analysis results. T his finding farther supports the results found at a similar site in the Mediterranean Sea (Hall Spencer et al 2008) that showed that reduced pH due to the

PAGE 77

60 venting of CO2 at shallow water volcanic vents affected the abundance of calcareous organisms in the surrounding benthic community. This study of the Ambitle Island vent system demonstrates that the effects of a small, shallow hydrothermal vent can extend much farther from the point source than previous studies have shown, presumably due to the effects o f diffuse venting of CO2 gas and hydrothermal fluids in the surrounding sediments. We found effects as far as 150 m from the vent, suggesting that future studies at shallow hydrothermal vents incorporate much longer transects.

PAGE 78

61 Chapter Four Macrofau nal c ommunities associated with the shallow water hydrothermal vent system at Ambitle Island, Papua New Guinea 4.1 Introduction The background for this chapter can be found in Chapters 2 and 3 which report the environmental factors at the site and macro fauna findings from the 2003 sampling. Chapter three describes an increasing trend in abundance, species richness and diversity with distance away from the vent. These community measures were still reduced as far out as the 150 m transect site, which was t he farthest sample collected along the transect The 2003 results fu rther suggested that pH, rather than the arsenic concentration, was a controlling factor in structuring the macrofaunal community ( Chapter 3, Karlen et al 2010). The 2003 results were bas ed on a limited sample size due to difficulties with shipping sampling suppli es to Papua New Guinea. The sampling effort was greatly expanded o n the second sampling trip in May 2005. In order to confirm the preliminary results from the 2003 data, several c hanges were made to the sampling design. These included exte nding the transect to 300 m, adding six more sites, collecting samples closer to the focused vent (0 m site), and collecting replicate samples at each site over a wider spatial area instead of wit hin a single 1 m2 quadrat. A new reference site was selected (Danlum Bay) that more closely resembled the sediment characteristics found at the transect sites. These modifications to the sampling design allowed for better statistical

PAGE 79

62 analysis of the benthi c community trends along the hydrothermal gradi ent and better correlations to the p ore water and sediment parameters. 4.2 Material and methods 4.2.1 Sampling design and field collection Sampling and analysis methods for pore water and sediment samples co llected at the transect sites and the Danlum Bay reference site are given in Chapter 2. Sediment samples for macrofauna analysis were collected at the nine transect sites and at the Danlum Bay reference site as described in Chapter 2. At each site, m acrof auna samples were collected from a random grid in each of the five quadrats at each site using a PVC corer (diameter = 7.62 cm; area = 45.6 cm2). Sediment cores were collected to a depth of 15 cm. The five cores from each site were treated as separate repl icate samples. organisms and sediment were rinsed into HDPE sample bottles and fixed in 10% borax buffered formalin with Rose Bengal stain for a minimum of 72 hours. The samp les were then transferred into 70% ETOH with Rose Bengal stain for preservation. 4.2.2 Sample processing and identification The macrofauna samples were sorted in the lab by first decanting the ETOH rinsing the sediment several times with tap water and decanting the water into the sieve to float off the more delicate organisms. The sediment was then rinsed into a stacked series of 2 3 sieves ranging from 4 mm to 500 nt was rinsed through the

PAGE 80

63 sieve series to separate the sample into various size fractions to aid in finding and sorting the macrofaunal invertebrates. Each sediment size fraction was rinsed into a separate contai ner and macrofauna were picked out and sorted into general taxonomic groups for identification and counting under a dissecting microscope. Macrofauna was identified to the lowest practical taxonomic level and counted. For the macrofauna analysis, lowest practical taxonomic levels gener ally was gen us or species identification. In cases w h ere more precise identifications were not possible, specimens were identified to the lowest level and assigned a letter designation based on their distinct morphological characteristics. All specimens were retained and archived for f urther taxonomic analysis and voucher specimens were photographed and a photographic log was kept to aid in maintaining consistency in identifications between samples. 4.2.3 Data analysis Parametric statistical analysis was conducted usi ng SigmaStat 3.5 (Systat Software, Richmond CA). One Way Analysis of Variance (ANOVA) with a Holm Sid ak multiple comparison post hoc test was used to test for between site differences in pore water, sediment, and benthic community metrics Data were log ( n=1) or square root transformed if it failed tests for normality and equal variances If the transformed data failed to meet the criteria for ANOVA, a nonparametric Kruskal Wallis ANOVA on ranks was used. Biological community indices (Species Richness, Abu ndance, ShannonW ien er Diversity and Pielous Evenness) and multivariate analysis was done using PRIMER v.6 (Clarke and Gorley 2006). The Bray Curtis similarity index was calculated on square root transformed abundance data to measure the similarity of th e benthic

PAGE 81

64 species composition among replicates within a site and between sites. Results were further analyzed using cluster analysis and/or non metric MultiDimensional Scaling (MDS). The Similarity Percentage (SIMPER) procedure was used to determine which taxa contributed to the within site similari ty or between site differences. Principal components analysis (PCA) was used on log transformed and normalized pore water and sediment data to group sites based on their physical characteristic. The BIO ENV ana lysis procedure in PRIMER v.6 was used to calculate Spearman ranked correlations between the environmental measurements and the biological community similarity and the LINKTREE procedure was used to evaluate the environmental characteristics defining the s ites comprising defined biological communities. 4.3 Results A total of 368 taxa in ten phyla were identified from all sites with a total abundance of 1,430 individuals. The annelids, predominately polychaetes, were the most speciose and abundant phylum w ith 185 taxa and 664 specimens representing 50% and 46% of the species richness and abundance respectively. The arthropods were represented by 92 taxa (25%) and 560 specimens (39%), and mollusks had 48 taxa and 122 specimens (13% and 8.5% respec tively). T wenty three taxa (6 %) accounted for >50% of the relative abundance while 199 taxa (54%) were represented by a single specimen. The polychaete Capitella cf. capitata was the most abundant species found and was dominant at the 0 m site. Other abundant taxa i ncluded the amphipod Platyischnopus sp. A, the burrowing shrimp Thalassinid ea sp A, and the amphipods Cyrtophium? sp. A and Ampithoe sp. A. No single taxon was present at al l 10 sites while 253 taxa (69%) were

PAGE 82

65 only present at a single site. The most widely occurring taxon was the burrowing shrimp Thalassinid ea sp A, which was found at 9 of the sites, and was only absent at 180 m. The amphipods Platyischnopus sp. A and Ampelisca sp A were present at 7 sites, but both were absent at 0 m, 250 m and 300 m. The bivalve mollusk Diplodonta sp. A was also found at 7 sites and was absent close to the vent at the 0 m 3 0 m and 6 0 m sites. The median, minimum and maximum values for the macrofauna taxonomic richness abundance, Shannon diversity index, Pielous evenn ess index and the Simpson diversity index at each site are presented in T able 4.1 Table 4.1 Median, minimum and maximum values for macrofaunal community indices by site. Site n Taxa Richness Abundance Shannon Diversity Evenness Simpson Index 0 m 5 10 2 2 0.94 0.68 0.53 2 16 10 104 0.20 2.58 0.29 0.93 0.10 0.94 30 m 5 6 8 1.57 0.92 0.81 2 6 3 14 0.50 1.73 0.72 0.97 0.40 0.93 60 m 5 3 7 1.04 0.95 0.83 3 9 4 13 0.80 2.14 0.73 0.97 0.52 0.95 90 m 5 11 19 2.26 0.92 0.89 8 16 17 34 1.90 2.45 0.8 6 0.94 0.87 0.94 120 m 5 11 17 2.17 0.91 0.91 7 12 11 26 1.77 2.36 0.81 0.95 0.82 0.95 140 m 5 13 25 2.25 0.85 0.87 6 15 12 52 1.39 2.37 0.71 0.99 0.72 0.99 180 m 5 24 38 2.93 0.93 0.96 19 37 28 57 2.79 3.32 0.92 0.95 0.95 0.96 250 m 5 42 62 3 .58 0.96 0.98 29 46 40 68 3.16 3.66 0.94 0.98 0.97 0.99 300 m 5 30 39 3.21 0.97 0.98 15 41 22 55 2.54 3.61 0.90 0.98 0.94 0.99 Ref 5 17 30 2.62 0.92 0.94 14 19 27 32 2.22 2.78 0.84 0.95 0.87 0.96 The 0 m site had a total of 39 taxa and 184 spe cimens, with a median of 10 taxa and 22 individuals per sample ( Table 4.1). Twentyfour taxa (62 %) were represented by a

PAGE 83

66 single specimen and 27 (69%) were only found at the 0 m site. This site was strongly dominated by the polychaete Capitella cf. capitata, which accounted for 63 % of the relative abundance at this site (Table 4.2) The 30 m site had a total of 13 taxa and a total abundance of 40 specimens, with median values of 6 taxa and 8 individuals per sample ( Table 4.1). Seven taxa (54%) were singleton s and two were only present at the 3 0 m site. Dominant taxa at this site included Thalassinid ea sp. A and Platyischnopus sp. A, which comprised 35% and 18% of the relative abundance respectively (Table 4.2) A total of 17 taxa and 37 specimens were identif ied from the 60 m site, with median values of 3 taxa and 7 specimens per sample ( Table 4.1). Four taxa comprised >50% of the abundance at this site and 10 taxa (59%) were singletons. Five taxa were only found at the 60 m site. Platyischnopus sp. A was the most abundant species, representing 27% of the relative abundance, followed by Thalassinid ea sp. A which made up 11% of the relative abundance (Table 4.2) Other dominant taxa included the amphipod Ampelisca sp. A, and the polychaetes Prionospio (Minuspio) sp. A and Sthenelais sp. A, each comprising 8% of the abundance (Table 4.2) The 90 m site had a total of 32 taxa present and 111 individual specimens, with median values of 11 taxa and 19 individuals per sample ( Table 4.1). Four taxa accounted for >50% of the a bundance at this site, 18 (56 %) of the taxa were singletons and 9 taxa (28%) were only found at the 90 m site. The most abundant taxon was the isopod crustacean Flabellifera sp. A, with a relative abundance of 26% (Table 4.2) Other dominant taxa i ncluded Platyischnopus sp A, Thalassinid ea sp. A, and the mysid crustacean Mysidacea sp. B (Table 4.2)

PAGE 84

67 The 120 m site had a total of 30 taxa and 92 individual specimens with a median of 11 taxa and 17 specimens per sample ( Table 4.1). Four taxa accounted for >50% of the total abundance at this site, 18 taxa (60%) were singletons and 7 taxa (23%) were only found at the 120 m site. The amphipods Ampelisca sp. A and Platyischnopus sp. A each comprised 15% of the relative abundance (Table 4.2) The crustaceans Cumacea sp. C and Thalassinid ea s p. A were also among the dominant taxa at this site. The 140 m site had a total of 35 taxa and 135 individual specimens with median values of 13 taxa and 25 individuals per sample ( Table 4.1). Four taxa comprised >50% of t he abundance at this site, 18 taxa (51%) were singletons and 9 taxa (26 % ) were only found at the 140 m site. This site was dominated by the amphipods Ampithoe ? sp. A and Ampelisca sp. A accounting for 22% and 11% of the relative abundance respectively (Tab le 4.2) Other abundant taxa included the polychaete Heteropodarke sp. A and the amphipod Platyischnopus sp. A. The 180 m site had a total of 90 taxa and 206 specimens. Median taxonomic richness and abundance values were 24 taxa and 38 individuals per sam ple (Table 4.1) Thirteen taxa accounted for >50% of the abundance, 58 taxa (64%) were singletons, and 34 taxa (38%) were only found at the 180 m site. The dominant taxa included the amphipod Cyrtophium ? sp. A, the polychaetes Grubeosyllis sp. B and Pholoe sp. A and the bivalve mollusk Codakia sp A (Table 4.2) The 25 0 m site had the highest taxonomic richness and abundance with a total of 141 taxa and 273 individual specimens. The median number of taxa and individuals per sample were 42 taxa and 62 indivi duals ( Table 4.1). Twenty eight taxa comprised >50% of the abundance while 91 taxa (6 5%) were singletons and 75 taxa (53%) were only

PAGE 85

68 present at the 250 m site. The most abundant taxon was the polychaete Grubeosyllis sp. B which had a relative abundance of 4.8% (Table 4.2) Other dominant taxa included the bivalve mollusk Codakia sp. A, the polychaetes Amphinome sp. A, Typosyllis cornuta, and Pholoe sp. A (Table 4.2) The 300 m site had the second highest richness and abundance with a total of 112 taxa iden tified and 202 specimens. Median richness and abundances values were 30 taxa and 39 individuals per sample ( Table 4.1). Twentythree taxa accounted for >50% of the abundance at the site, 78 taxa were singletons and 62 taxa were only found at the 300 m site T he most abundant species was the amphipod Cyrtophium sp. A, w ith a relative abundance of 9.4 % (Table 4.2) Other dominant taxa included the amphipod Melita sp. A and the polychaetes Pholoe sp. A, Typosyllis cornuta, and Protodorvillea sp. B. The Danlum Bay reference site had a total of 56 taxa and 150 individual specimens. Median richness and abundance values were 17 taxa and 30 specimens per sample (T able 4.1 ). Seven taxa accounted for >50% of the abundance, 34 taxa (61 %) were singletons and 23 taxa (41 %) were only present at the reference site. The dominant taxa were primarily polychaetes Litocorsa sp. A, Mediomastus sp. A, Heteropodarke sp. A, and an unidentified capitellid tentatively designated as Capitellidae w/spatulate chaetae based on the uniqu e morphology of its chaetae (Table 4.2) Also among the top ranked taxa was the borrowing shrimp Thalassinid ea sp. A. The overall trend in species richness ( S) showed an increase with distance from the vent site (Figure 4.1 ), with significant differences between sites (One way ANOVA F=22.796, df=9,40, p<0.001 analysis on nontransformed mean values, n= 5). There were no pair wise significant differences in richness between the 0 m 30 m, 60 m 90 m,

PAGE 86

69 120 m and 140 m sites, or between the 0 m and the Danlum Bay r eference site. Richness was significantly higher at the 180 m, 250 m and 300 m sites relative to the other sites, while the 250 m site had higher species richness than the 180 m site, but was not significantly higher than at 300 m. The 18 0 m and 300 m sites were also not significantly different from each other. The species richness at the Danlum Bay reference site was higher relative to the 3 0 m and 6 0 m sites and lower than at the 25 0 m and 300 m sites. The macrofaunal abundance was above the overal l mean value at the vent site but was highly variable (Figure 4.2). There was an increasing trend in abundance from the 30 m site out to the 250 m site, with abundance values generally below the overall mean from 30 m 140 m and above the mean value at the 180 m 300 m sites. The mean abundance at the reference site was nearly equal to the overall mean (Figure 4.2). Abundances were significantly different among sites [ one way ANOVA, df = 9, 40; F = 11.041; p<0.001; analysis on log(n+1) transformed abundan ce data] Most pair wise comparisons between sites were not significantly different. The 30 m and 60 m sites did have significantly lower abundances than all of the other sites except for 120 m. The 120 m site was also significantly lower than the 250 m site.

PAGE 87

70 Table 4.2 Top ranked taxa comprising >50 % of the relative abundance at each site. 0 m 3 0 m 6 0 m 9 0 m 12 0 m S = 39 S = 13 S = 17 S = 32 S = 30 N = 184 N = 40 N = 37 N = 111 N = 92 Taxa Rel. Abund. Taxa Rel. Abund. Taxa Rel. Abund. Taxa Rel. Abund. Taxa Rel. Abund. Capitella cf. capitata 62.5% Thalassinid ea sp. A 35.0% Platyischnopus? sp. A 27.0% Flabellifera (nr. Aegidae) sp. A 26.1% Ampelisca sp. A 15.2% Platyischnopus? sp. A 17.5% Thalassinid ea sp. A 10.8% Platyischnopus? sp. A 14.4% Platyisch nopus? sp. A 15.2% Prionospio (Minuspio) sp. A 8.1% Thalassinid ea sp. A 9.0% Cumacea sp. C 13.0% Sthenelais sp. A 8.1% Mysidacea sp. B 9.0% Thalassinid ea sp. A 9. 8% Ampelisca sp. A 8. 1% Cumulative: 62.5% Cumulative: 52.5% Cumulative: 62.2 % Cumulative: 58.6% Cumulative: 53.3%

PAGE 88

71 Table 4.2 Continued. 14 0 m 18 0 m 25 0 m 30 0 m REF S = 35 S = 90 S = 141 S = 112 S = 56 N = 135 N = 206 N = 273 N = 202 N = 150 Taxa Rel. Abund. Taxa Rel. Abund. Taxa Rel. Abund. Taxa Rel. Abund. Taxa Rel. Abu nd. Ampithoe? sp. A 24.4% Cyrtophium? sp. A 10.2% Grubeosyllis sp. B 4.8% Cyrtophium? sp. A 9.4% Litocorsa sp. A 12. 7% Ampelisca sp. A 11.1% Grubeosyllis sp. B 9.2% Amphinome? sp. A 3.7% Pholoe sp. A 4.5% Mediomastus sp. A 12.0% Heteropodarke? sp A 11.1% Pholoe sp. A 3.9% Codakia? sp. A 3.7% Melita? sp. A 4.5% Heteropodarke? sp A 6. 7% Platyischnopus? sp. A 8. 9% Codakia? sp. A 3.9% Typosyllis cornuta 3.3% Typosyllis cornuta 4.0% Capitellidae w/ spatulate chaetae 6.0% Typosyllis cornuta 3.4% Pholoe sp. A 2.9% Protodorvillea sp. B 3.0% Thalassinid ea sp. A 5.3% Melita? sp. A 3.4% Tanaidacea sp. C 2. 6% Mediomastus sp. A 2.5% Diasterope? sp. A 5.3% Semelina? sp. A 3.4% Aspidosiphon sp. A 2.2% Nephasoma? sp. A 2.0% Sigambra cf. parva 4.0% Exogone cf lourei 2.9% Amphipoda sp. K 2.9% Cumulative: 55. 6% Cumulative: 43.2% Cumulative: 23.1% Cumulative: 29.7% Cumulative: 52.0%

PAGE 89

72 Figure 4.1 Ambitle Island May/June 2005: Mean number of macrofaunal taxa 1 standard deviation. Figure 4. 2 Ambitle Island May/June 2005: Mean macrofaunal abundance 1 standard deviation. Abundance presented as individuals/sample.

PAGE 90

73 The ShannonWiener diversity index also showed an increasing trend with distance from the vent with the lowest mean value at the 30 m site and the highest value at the 250 m site (Fig 4.3 ). There were no significant pair wise differences among the 0 m 140 m sites along the transect The 18 0 m and the reference sites were more diverse than the 0 m, 30 m and 60 m sites but were not significantly different from the other sites. The diversity at the 25 0 m and 300 m sites was significantly higher than the inner transect locations from 0 m to 140 m. Figure 4.3 Ambitle Island May/June 2005: Mean macrofaunal diversity 1 standard devia tion. The Pielou evenness index ( J ) was significantly lower at the vent site relative to all other sites except the 140 m site (Figure 4.4) All other pair wise comparisons between sites did not show significant differences in the evenness index. The arc sine transformed evenness values did not meet the assumptions for ANOVA (normal

PAGE 91

74 distribution or equal variances ), however, the statistical power of the analysis was still high (1 parametric Kruskal Wallis ANOVA on ranks also showed a significant difference among sites, but the corresponding pair wise comparison test only showed a signific ant difference between the 0 m and 250 m sites. Figure 4.4 Ambitle Island May/June 2005: Mean macrofaunal evenness 1 standard deviation. The nonmetric multidimensional scaling (MDS) analysis o f all 50 samples (Figure 4.5) and o f the site averaged da ta (Figure 4.6a ) both suggest that the macrofaunal species composition falls into four distinct faunal communities along the transect and at the reference site. The 0 m and reference sites each comprise their own distinct macrofaunal communities the 30 m 140 m transect sites group together as do the 180 m 300 m sites. Since the environmental characteristics were different at the Danlum Bay reference site relative to the Tutum Bay transect sites, the MDS analysis was also

PAGE 92

75 done omitting the reference sit e (Figure 4.6b). Excluding the reference site resulted in tighter clustering of the transect sites in the previously defined faunal communities and a lower overall 2D stress for the MDS plot (Figure 4.6b). The SIMPER analysis defines the average similarity of the samples and the contribution to the species similarity wi thin each defined group (Table 4.3). The 0 m site is defined primarily by the polychaete Capitella cf. capitata, which accounted for >86% of the simil ar ity among the 0 m samples ( Table 4.3 ). The 30 m 140 m community is defined by the presence of Thalassinid ea sp. A and Platyischnopus sp A which contributed 32% and 26% respectively to the similarity among these samples ( Table 4.3 ). Six taxa cumulatively contributed to >50% of the similarity a mong the samples within th e 18 0 m 300 m community group (Table 4.3). The amphipod Cyrtophium? sp. A and the polychaetes Typosyllis cornuta and Pholoe sp. A each accounted for >10% of the similarity within this group. The Reference site community was defi ned by the polychaetes Litocorsa sp. A, Mediomastus sp. A and Capitellidae w/spatulate chaetae. Cumulatively, these three taxa accounted for >55% of the similarity among the reference site replicates (Table 4.3).

PAGE 93

76 Figure 4.5 Ambitle Island May/June 2005: Benthic macrofauna multi dimensional scaling plot based on BrayCurtis similarity between site replicates Figure 4.6 a Ambitle Island May/June 2005: Benthic macrofauna multi dimensional scaling plot based on BrayCurtis similarity averaged by site.

PAGE 94

77 Figure 4.6b Ambitle Island May/June 2005: Benthic macrofauna multi dimensional scaling plot based on BrayCurtis similarity averaged by site ; reference site omitted. The BIO ENV analysis comparing the macrofaunal community structure to the physical para meter (Table 4.4a) found the strongest correlation with a combination of total arsenic, % CaCO33, and sediment sorting only had a slightly weaker correlation with the community structure ). The % CaCO3 total arsenic had a correlation of 0.50 (Table 4.4a). The BIO ENV analysis was also reevalu ated without the reference site (Table 4.4b). Those results generally had slightly higher correlation coefficients and indicated a stronger correlation with the sediment organic content, rather than the % CaCO3 (Table 4.4b). The LINKTREE analysis (Figure 4.7) separated the macrofaunal communities at the 180 m, 250 m and 300 m from the remaining sites. These three sites were

PAGE 95

78 characterized by relatively higher pore water pH (>7.74 vs. <7.49), higher sediment CaCO3 content (>7.4% vs. <6.5%) and lower total ar senic concentrations (<9.9 ppb vs. >11.6 ppb). The 0 m site separated out from the remaining sites due to larger median grain size, higher temperature, higher arsenic concentrations, lower pore water salinity, higher sediment sorting coefficient and lower pore water pH. The reference site split from the remaining sites because of its higher sediment organic content and lower ORP values, higher sediment sorting coefficients and CaCO3 content, higher pH and lower temperature and arsenic concentrations relati ve to the remaining transect sites. The final split in the LINKTREE separated the 90 m, 120 m and 140 m sites from the 30 m and 60 m sites, based on slightly lower temperature, and slightly higher median grain size and sediment CaCO3 content within the 90 m, 120 m, 140 m site group.

PAGE 96

79 Table 4.3 SIMPER results from macrofaunal community groups defined by Bray Curtis similarity groupings. 0 m 3 0 m 14 0 m 18 0 m 30 0 m Ref Average similarity: 16 Average similarity: 23 Average similarity: 18 Average similarit y: 29 Taxa % Taxa % Taxa % Taxa % Capitella cf. capitata 86.9 Thalassinid ea sp. A 31.7 Cyrtophium? sp. A 10.9 Litocorsa sp. A 23.9 Platyischnopus? sp. A 26.4 Typosyllis cornuta 10. 6 Mediomastus sp. A 19. 5 Pholoe sp. A 10.1 Capitellidae w/ spat ulate chaetae 12.4 Codakia? sp. A 7.6 Grubeosyllis sp. B 6.7 Protodorvillea sp. B 6.7 Cumulative % 86.9 Cumulative % 58.1 Cumulative % 52.5 Cumulative % 55.8

PAGE 97

80 Table 4.4a BIO ENV correlations between the benthic macrofaunal communit y structure and physical parameters. Spearman Correlation Parameters 0.740 Total [As], %CaCO 3, Sorting 0.735 %CaCO 3, Sorting 0.732 Total [As], % Organics, %CaCO 3 Sorting 0.718 ORP, %CaCO 3, Sorting 0.717 ORP, Total [As], %CaCO 3, Sorting 0.71 7 ORP,Total [As], % Organics, %CaCO 3 Sorting 0.712 Total [As], %CaCO 3 0.711 pH, Total [As], % Organics, %CaCO 3 Sorting 0.707 Temperature, %CaCO 3, Sorting 0.705 Temperature, %CaCO 3, Median Grain Size, Sorting 0.671 %CaCO 3 0.626 Sorting 0.604 Median Grain Size 0.551 % Organics 0.496 Total [As] 0.294 pH 0.292 Temperature 0.222 ORP 0.162 Salinity Table 4.4b BIO ENV correlations between the benthic macrofaunal community structure and physical parameters ; reference site omitted. Spearman Correla tion Parameters 0.790 Total [As], % Organics, %CaCO 3 0.789 Total [As], % Organics, %CaCO 3 Sorting 0.780 % Organics, %CaCO 3 Sorting 0.780 pH, Total [As], % Organics, %CaCO 3 Sorting 0.774 pH, % Organics, %CaCO 3 Sorting 0.770 Total [As], %CaCO 3 0.7 68 ORP,Total [As], % Organics, %CaCO 3 0.763 Total [As], %CaCO 3, Sorting 0.762 ORP,Total [As], % Organics, %CaCO3, Sorting 0.758 % Organics, %CaCO 3 0.746 % Organics 0.726 %CaCO 3 0.622 Sorting 0.617 Median Grain Size 0.554 Total [As] 0.332 pH 0.3 20 Temperature 0.249 ORP 0.166 Salinity

PAGE 98

81 Figure 4.7 LINKTREE showing physical parameters characterizing site groups based on the benthic macrofauna similarity. 4.4 Discussion The benthic macrofaunal communities are known to be good indicators of th eir surrounding habitat characteristics and the macrofaunal community structure in Tutum Bay reflects this. Four distinct comm unity assemblages were observed corresponding to different pore water and sediment characters. These communities are defined as t he vent community consisting of the 0 m site the n ear vent community consisting of the transect sites from 30 m t hrough 140 m away from the vent, the offshore community consisting of

PAGE 99

82 the transect s ites from 180 m through 300 m and the Danlum Bay reference site community. The vent community unique to the 0 m site had a low species richness and diversity, but high abundance, dominated by a single species. The dominant species at the 0 m site, Capitella cf. capitata, has often been associated with disturbed habita ts and has been considered to be a pollution indicator species (Grassle and Grassle 1974, 1976) The taxon Capitella capitata has long been recognized as a complex of many morphologically similar sibling species (Grassle and Grassle, 1974, Gamenick e t al 1998b, Blake 2009). Blake (2009) redescribed Capitella capitata based on museum specimens collected from Greenland near the original type locality. According to Blakes redescription, Capitella capitata has an arctic to subarctic geographical distribu tion and other Capitella capitata reported in the literature from other localities globally are most likely different but closely related sibling species. Given this, the specimens identified from Tutum Bay may well represent an undescribed sibling sp ecies within the Capitella capitata species complex that is unique in its tolerance to the extreme environmental conditions near the shallow water hydrothermal vent. The near vent community had low species richness and abundance and was dominated primari ly by the burrowing shrimp Thalassinid ea sp. A and the amphipod Platyischnopus sp. A which seemed to be wel l adapted to living in the fine grained sediments characteristic at these sites. The biotic interacti on of these two species with other benthic infa una may influence the overall community structure. The constant burrowing activity of thalassinid shrimp, for example, physically disturbs the sediment which can prevent the colonization of that area by other species ( Alongi 1986) The same

PAGE 100

83 reworking of th e sediment can also affect the geochemical processes and microbial productivity in the sediment column (Lohrer et al 2004, Mermillod Blondin and Rosenberg, 2006, Jordan et al 2009) In particular the burrowing activity draws oxygenated water deeper into the sediment which can influence redox reactions. In the case of metals, such as arsenic, this can potentially affect the valence state of the metal and its toxicity and bioavailabilty Species in the amphipod genus Platyischnopus are known to be predato rs on recently settled larvae of other infaunal animals and on meiofauna ( J.D. Thomas, personal communication ). The presence of this genus in high densities may exclude the colonization by other species affecting the overall species composition of this co mmunity. The physical characteristics of the pore water and sediment at the near vent sites also influence the benthic community structure at these sites in several possible ways. In addition to the high arsenic concentrations occurring at these sites, the low pH has adverse e ffects on shell bearing species. McCloskey (2009) in his corresponding study of the foraminifera in Tutum Bay also found pH to be a controlling factor with foraminifera shells being absent in the sediments as far away as 150 m along th e same sampling transect. Engel (2010) studied the effect of short term (5 day) exposure to the pH and temperature gradient at Tutum Bay on benthic foraminifera. She observed evidence of shell dissolution as far as 30 m from the vent Most species appeared to be abl e to survive exposure to low pH but were less tolerant of high temperatures. D issolution was more apparent on dead shells (Engle 2010).

PAGE 101

84 The offshore community was characterized by higher species richness, abundance and overall species diversity The sediment characteristics such as CaCO3 content appeared to be the controlling factor in structuring the macrofaunal community at these sites. This was farther related to the higher pore water pH values, reflecting the reduced influence from the hydrothermal vent with distance offshore. The high organic carbon content in the Danlum Bay sediments was correlated with the macrofaunal community found at the reference site. The macrofaunal community was dominated by several polychaete species such as Litocor sa sp. A (Family Pilargiidae) and representatives in the family Capitellidae including Mediomastus sp. A. and a possibly undescribed genus ( Capitellidae w/spatulate chaetae). Both of these polychaete families have been classified as deposit feeders by Fauc h ald and Jumars (1979) and are commonly found in organically enriched environments. Although arsenic concentrations in the sediments and pore water at Tutum Bay are among the highest recorded in a natural marine system ( Pichler and Dix 1996, Pichler et al 1999a ), this parameter had a lesser influence in structuring the benthic macrofaunal community than expected as seen in the BIO ENV results. Several factors could explain this. First, the co precipitation of the arsenic with hydrous ferric oxides (HFOs) bin ds the arsenic in the sediments reducing its bioavailability to benthic infauna (Pichler and Veizer 1999, Pichler et al 1999a Price and Pichler 2005). Price and Pichler (2005) found that as much as 98.6% of the arsenic in the vent fluids is precipitated into the sediments. Only a small percentage of the bound arsenic (mean = 3.6%) was found to be easily extractable and still potentially bioavailable to benthic organisms (Price and Pichler 2005). Secondly, because the low salinity and high temperature vent discharge is less

PAGE 102

85 dense than the surrounding seawater it rises to the surface (Price and Pichler 2005) away from the benthic environment s Many invertebrate organisms have also adapted physiological mechanisms to detoxify or eliminate high concentrations of arsenic and other contaminants (Langdon et al 2003). Most metabolic pathways in marine invertebrates involve the reduction of As(V) to As(III) which is then methylated through a series of organoarsenical compounds typically to arsenobetaine (AB ) as th e end product ( Langdon et al 2003, Argese et al 2005, Grotti et al 2010) Price (2008) looked at the arsenic speciation in tissue samples collected from several sessile organisms living near the Tutum Bay vent and at sites along the transect. He found e levated levels of total arsenic in o rganisms collected near the vent. Price (2008) further found elevated levels of organoarsenical compounds in the tissue samples from the soft coral Clavularia and the ascidian Polycarpa suggesting that these organisms were able to detoxify the arsenic via m ethylation to DMA and AB 4.5 Summary and conclusions T he benthic macrofaunal community structure at Tutum Bay was influenced by the environmental gradient along the transect. The community immediately surrounding the focused venting was represented by a few opportunistic species and most notably dominated by the polychaete Capitella cf. capitata. The transect sites out to 140 m away from the focused vent still showed evidence of hydrothermal influence due to diffuse v enting The macrofaunal community in this area was characterized by low species diversity and abundances and was dominated by Thalassinid ea sp. A and the amphipod Platyischnopus sp. A. This suggests that the community structure at these sites may have

PAGE 103

86 been influenced by a combination of the physical environment and biological interactions due to physical disturbance through bioturbation and by predation of settling larvae. The macrofaunal community farther from the vent, by contrast, had relatively h igh div ersity and abundances. The preliminary results from the 2003 samples had suggested that pH was a controlling factor in structuring the benthic macrofaunal community at Tutum Bay ( Chapter 3; Karlen et al 2010). The results from the 2005 samples, however, do not agree with the earlier findings. Instead, sediment characteristics such as carbonate content, sediment sorting and median grain size appeared to structure the macrofaunal community. These factors in turn were correlated with the pore water pH which suggests that even though pH may not have a direct effect on the community structure it may have an indirect influence by controlling the sediment composition.

PAGE 104

87 Chapter Five Meiofaunal communities associated with the shallow water hydrothermal vent sys tem at Ambitle Island, Papua New Guinea 5.1 Introduction The meiofauna refers to the size range of organisms occupying the intermediate size range between the smaller microfauna (predominately prokaryotes and protozoan eukaryotic groups) and larger m acrofauna ( predominately invertebrate metazoan groups). The size range that defines the meiofauna varies with different authors, with an upper size limit of either 1 mm ) and a lower limit of ranging from 62 to as Higgins and Thiel 1988). For this study, The m eiofaunal taxa included in this analysis included the metazoan gro ups and ciliates. Foraminifera were excluded since this group was analyzed by McCloskey (2009) as part of the larger research effort at Tut um Bay. Several studies over the past 20 years have looked at the meiofaunal com munities associated with shallow water hydroth ermal systems around the world Past studies have included sites in New Zealand (Kamenev et al 1993), the Aegean Sea (Thiermann et al 1994, 1997), Papua New Guinea (Tarasov et al 1999) Antarctica (Bright et al 2003),

PAGE 105

88 the Kuril Islands, Sea of Okhotsk (reviewed in Tarasov 2006) and Indonesi a (Zeppilli and Danovaro 2009). Kamenev et al (1993) documented the meiofaunal distribution at several sites in the Bay of Plenty, New Zealand. They reported higher dens ities of nematodes in two samples from a hydrothermally influenced site ( McEwens Bay site 1) relative to an adjacent nonhydrothermal site Thiermann et al (1994) foun d no meiofauna present in sediments directly impacted by the venting, while the meiofau nal community farther away was dominated by nematodes One nematode species, Oncholaimus camplocercoides occurred in high densities at the border of the sulfidic sediment patch surrounding the vent and the nonsulfidic sediment farther from the vent (Thier mann et al 1994). Tarasov et al (1999 ) studied the distribution of meiofauna near sh allow water hydrothermal vents at several sites within Matupi Harbor in the Rabaul Caldera, Papua New Guinea. Most of the meiofaunal communities they reported were domina ted by nematodes. A t one site near the active T a vurvur volcano the high temperature (85 90C), shallow zone was devoid of infauna, while the area with hydrothermal seeps and slightly cooler temperatures (50 60C) was dominated by n ematodes A second hydrothermal area they sampled was characterized by bacterial mats. The dominant meiofauna living on t he mats was a spionid polychaete identified as Polydora sp. which formed burrows through the mat down to the underlying sediment. At a third site near the Rabalankaia volcano they found a single species of large nematode (Oncholaimidae) at high densities (131 x 103/m2) in hydrothermal sediments (50 60C). The meiofaunal densities at the several non volcanic sites within Matupi Harbor and the Rabaul Caldera

PAGE 106

89 (Blanche Bay) were variable and appeared to be influenced primarily by sediment characteristics and bottom slope (Tarasov et al 1999). Zeppilli and Danovaro (2009) also found that meiofaunal communities associated with shallow water hydrothermal vents in Indonesia were dominated by nematodes with the exception that copepods were dominant 1 00 cm from the vent in an area o f intermediate hydrothermal influence. This they attributed to the alteration of the sediment grain size by hydrothermal fluids. They fo und a general trend of decreasing abundance and diversity closer to the vent site but noted that the meiofauna abundances they reported were as much as 10x higher than those recorded at other shallow water vent sites. Zeppilli and Danovaro (2009) also reco rded the occurrence of several species of Oncholaimid nematodes near the vent, including Oncholaimus the same genus found by Thiermann et al (1994) at Milos, Greece. Some meiofaunal communities have adapted to surviving in anoxic sediments below the leve l of oxygen penetration in the surface sediments. This habitat is characterized by high levels of hydrogen sulfide (H2S) and the associated meiofaunal communities are referred to as the Thiobios (Fenchel and Riedl 1970). These high sulfide conditions of ten occur in extreme environments such as hydrothermal vents and brine seeps (Powell and Bright 1981 Powell et al 1983). Several sh allow water hydrothermal vent systems have high levels of H2S present, and the associated meiofauna show unique adaptations to these sulfide enriched environments The dominant nematodes found near the hydrothermal vents in the Bay of Plenty, New Zealand, were associated with symbiotic sulfur bacteria, which covered the nematodes outer cuticle (Kamenev et al 1993, Tarasov 20 06). Similarly, Bri ght et a l. (2003) discovered that the

PAGE 107

90 outer surface of the dominant meiofaunal flatworm found at Fumarole Bay, Deception Island, Antarctica was associated with symbiotic bacteria which colonized its outer surface. The dominant nematode a ssociated with high sulfidic sediments at Milos, Greece, ( Oncholaimus campylocercoides ) was able to detoxify sulfide by s equestering elemental sulfur as intracellular inclusions in their epidermal cells (Thiermann et al 1994, 2000). At the volcanic site i n Rabaul Caldera studied by Ta rasov et al (1999), one of the dominant nematodes they found was in the same family (Oncholaimidae) as the species found by Thiermann et a l. (1994) at Milos, Greece Additionally, they noted that the dominant meiofaunal polyc haete occurring in the bacterial mats ( Polydora sp.) was associated with filamentous sulfur bacteria that colonized their burrow wal ls. These observations suggest possible symbiotic relationships between sulfur metabolizing bacteria and meiofaunal species. The Amb itle Island vent system was similar to these other sites with regards to its high temperature, but differed in that high levels of hydrogen sulfide were not evident, while arsenic concentrations were high. Because of this, it was expected that th e meiofaunal community surrounding the vent would be influenced by the high arsenic, and possibly be associated with arsenic metabolizing bacteria. 5.2 Material and methods 5.2.1 Sampling design and field collection Sampling and analysis methods for pore water and sediment samples collected at the transect sites and the Danlum Bay reference site are given in Chapter 2.

PAGE 108

91 Sediment samples for meiofaunal community analysis were collected at the nine transect sites and at the Danlum Bay reference site as descri bed in Chapter 2. Meiofauna samples were collected from a random grid in each of the five quadrats at each site using a 60cc syringe coring tube (diameter = 3cm ; area = 7.07 cm2). Sediment cores were collected to a depth of 5 cm. The five cores from each s ite were sized organisms. The retained meiofauna and sediment was rinsed with filtered seawater into 50 ml HDPE sample bottles and fixed in 10% borax buffered formalin with Rose Bengal stain. Seawater used for sieving and rinsing the samples was filtered through a 50 ntamination of the samples. 5.2.2 Sample processing and identification The laboratory processing of the meiofauna samples followed the following procedure: First the formalin was decanted from of plankton netting plac ed in a funnel. The sediment was rinsed into a 100 ml graduated cylinder with filtered tap water. The graduated cylinder was filled up to the 100 ml mark with filtered tap water then stoppered and the sediment was mixed by inverting the cylinder 10 times to suspend and separate the meiofauna from the sediment. The water float off procedure was repeated 5 times. The organisms retained on the sieve were rinsed into a grid ded pet ri dish with filtered tap water and sorted into the lowest practical

PAGE 109

92 taxonomic levels and counted. The sediment fraction was also placed in a gridded petri dish and scanned under the dissecting microscope and any remaining meiofauna were removed For the meiofauna analysis, lowest practical taxonomic levels generally were O rders, although lower taxonomic identifications were made when possible. All specimens were retained and archived for farther taxonomic analysis and voucher specimens were photographed. 5.2.3 Data analysis Data analyses were as described in Chapter 4. 5.3 Results A total of 17, 346 meiofaunal specimens were found at all sites, representing 61 higher taxonomic groups. Arthropods and nematodes represented 46% and 40% of the overall mei ofaunal abundance respectively ( Figure 5.1). Within the a rthropods, copepods (predominantly Harpacticoidea) comprised 61% of the abundance, and unidentified crustacean larvae (naupulii) were 27% of the arthropod abundance. The nematode abundance was repres ent ed primarily by the order Chromadorida which accounted for 76% of the nematode abundance (Figure 5.1). Ar thropods and nematodes were dominant across all sites and accounted for over 80% of the relative abundance at every site (Figure 5.2). Total numbe r of taxa and abundance pooled for the five replicate cores within each site along with ranked relative abundance are shown in Table 5.1. Chromadorid nematodes and copepods were the top two ranked taxa at all sites except 90 m and

PAGE 110

93 accounted for > 50% of th e relative abundance at the most of the sites (Table 5.1) Seven of the 61 identified taxonomic groups were found at all sites. These were Chromadorid, Monhysterid, and Enoplid nematodes; Copepoda, Ostracoda, Platyhelminthes, and unidentified crustacean larvae. Seventeen taxonomic groups were present at a single site, and 12 were r epresented by a single specimen. The 0 m site had a total of 17 higher taxa and 382 total individual specimens with a median richness of 10 taxa/sample and a median abundance of 52 individuals/sample (Tables 5.1 and 5.2) Capitellid polychaetes accounted for nearly 5% of the relative abundance at this site, a much higher contribution than at the other sites where they were present. These were primarily juvenile individuals of Cap itella cf. capitata. Unique at the 0 m site was a single specimen of the interstitial polychaete family Nautiliniellidae. Taxonomic identifications for the 0 m samples were initially done to lower classification levels when possible Table 5.3 presents the top ranked taxa at the 0 m site utilizing the lower level identifications. For these calculations, the larval crustaceans were excluded from the total count resulting in the observed discrepancies with Table 5.1. T he h arpacticoid copepods here are the hi ghest ranking taxon with 24% of the relative abundance. The chromadorid nematode Innocuonema sp. A ranked second and mites (Acari spp.), juvenile Capitella cf. capitata, and the enoplid nematode Trileptium? sp. A were among the top five most abundant taxa at the 0 m site. The 30 m site had a total of 19 higher level taxa and 1,190 individuals, with median taxonomic richness of 11 taxa/sample and median abundance of 210 individual/sample (Tables 5.1 and 5.2). Unique meiofauna taxa at this site included the

PAGE 111

94 interstitial polychaete family Polygordiidae and chaetognaths. Additional ly, the interstitial polychaete, Protodriloides cf. chaetifer was also found at this site. A total of 24 higher level taxa and 1,250 individuals were present at the 60 m site (Tabl e 5.1). The median taxonomic richness was 12 taxa/sample and median abundance was 208 individuals/sample (Table 5.2). This site included a decapod and mysid crustaceans which were not found at the other sites (Table 5.1). The 90 m site had a total of 25 higher level taxa and 2,098 individuals (Table 5.1). Median taxonomic richness was 25 taxa/sample and median abundance was 452 individuals/sample. Crustacean larvae were the second most abundant group at this site, accounting for over 25% of the relative ab undance. The 120 m site had a total of 24 higher level taxa and 2,001 individuals (Table 5.1). Median richness was 15 taxa/sample and median abundance was 419 i ndividuals/sample (Table 5.2). There were no taxa unique to this site. There were a total of 2 9 higher level taxa and 2,197 individuals present at the 140 m site (Table 5.1). The median richness and abundance were 15 taxa/sample and 376 individuals/sample respectively Unique taxa present at this site included an opisthobranch mollusk and a possible chironomid insect larva The 180 m site had a total of 32 taxa and 2,294 individuals with a medium taxonomic richness of 17 taxa/sample and a median abundance of 395 individuals/sample (Tables 5.1 and 5.2). Unique taxa at this site included the polychaete families Nereididae and Ampharetidae (Table 5.1). The 250 m site had a total of 33 higher level taxa and a total abundance of 2,023 individuals (Table 5.1). Median taxonomic richness and median abundance were 18

PAGE 112

95 taxa/sample and 338 individuals/sample r espectively (Table 5.2). Copepods were the dominant taxa group a t this site, accounting for 38 % of the relative abundance (Table 5.1). There were no unique taxa which were found only at the 250 m site, but t he interstitial polychaete Protodriloides cf. cha etifer was recorded (Table 5.1). The 300 m site had the highest overall taxonomic richness and abundance with a total of 39 higher level taxa and 3,180 individuals (Table 5.1). Median taxonomic richness was 21 taxa/sample and median abundance was 582 indi viduals/sample (Table 5.2) Copepods w ere also the most abundant group at this site. Several polychaete families were only found at the 300 m site, including Ctenodrilidae, Fauvelopsidae?, Flabelligeridae? and Maldanidae (Table 5.1). Recently settled troch ophore larvae were also recorded at this site (Table 5.1). The Danlum Bay reference site (Ref) had a total of 24 taxa and a relatively low abundance of 731 individuals (Table 5.1). The median richness was 15 taxa/sample and median abundance was 141 indivi duals/sample. This site was unique in the strong dominance of copepods, which accounted for 47 % of the relative abundance at the reference site. Table 5. 2 presents the summary statistics for the meiofaunal community indices at each site. The higher level t axonomic richness ranged from five taxa/sample at the 0 m site to a maximum of 24 taxa/sample at the 180 m site, with the highest median value at the 300 m site (Table 5. 2). The taxonomic richness increased with distance from the vent site (Figure 5. 3) an d richness at the farthest transect sites (180 m 300 m) were significantly higher than the 0 m and inner transect sites (ANOVA, F=9.45; p<0.001).

PAGE 113

96 Taxa richness at the Danlum Bay reference site was similar to the middle transect sites with a median value of 15 taxa/sample (Table 5. 2; Figure 5. 3).

PAGE 114

97 Figure 5.1 Relative taxonomic composition of total meiofaunal abundance (N = 17, 346 individuals) by phylum (top graph) and subsets for Nematoda (bottom left ; N = 6,935 individuals ) and Arthropoda (bottom r ight ; N = 8,019 individuals )

PAGE 115

98 Figure 5.2 Meiofaunal taxonomic composition by site.

PAGE 116

99 Table 5.1 Meiofaunal total taxa richness (S) abundance (N) and relative abundance (%) by site. 0 m Ref S = 24 N = 731 S = 17 N = 382 Taxa % Taxa % Chromadorida 27.23% Copepoda 46.92% Copepoda 23.82% Chromadorida 17.37% Enoplida 10.99% Monhysterida 6.29% Crustacean larvae 6.54% Enoplida 5.47% Monhysterida 6.54% Crustacean larvae 5.20% Acari 6.28% Syllidae 5.06% Capitellidae 4.97% Trefusiida 3.69% Syllidae 3.93% Ciliophora 2.46% Ostracoda 2.88% Hesionidae 1.78% Platyhelminthes 2.36% Platyhelminthes 1.37% Nerillidae 1.31% Ostracoda 0.96% Tanaidacea 1.05% Acari 0.55% Trefusiida 0.79% Sabellidae 0.41% Cumacea 0.52% Spionidae 0.41% Gammeridea 0.26% Tanai dacea 0.41% Isopoda 0.26% Capitellidae 0.27% Nautiliniellidae 0.26% Polychaete (larvae) 0.27% Isopoda 0.27% Dorvilleidae 0.14% Lumbrineridae 0.14% Paraonidae 0.14% Pilargiidae 0.14% Bivalvia 0.14% Gastropoda 0.14%

PAGE 117

100 Ta ble 5.1 Continued 30 m 60 m S = 19 S = 24 N = 1190 N = 1250 Taxa % Taxa % Chromadorida 38.32% Chromadorida 29.12% Copepoda 22.52% Copepoda 26.80% Monhysterida 14.54% Platyhelminthes 11.36% Crustacean larvae 7.82% Monhysterida 11.12% Ciliophora 4.87% Crustacean larvae 10.16% Enoplida 4.03% Ciliophora 4.16% Platyhelminthes 2.69% Enoplida 3.28% Polychaeta (larvae) 1.60% Tardigrada 1.20% Tardigrada 0.76% Gastrotricha 0.88% Ostracoda 0.76% Ostracoda 0.32% METAZOA (Undet.) 0.50% Gammeridea 0.24% Ca pitellidae 0.42% Tubificidae 0.16% Trefusiida 0.34% Capitellidae 0.16% Polygordiidae 0.25% Glyceridae 0.16% Chaetognatha 0.17% Acari 0.16% Tanaidacea 0.17% Hesionidae 0.08% Protodriloides cf. chaetifer 0.08% Nerillidae 0.08% Tubificidae 0.08% Spionid ae 0.08% Caprellea 0.08% Syllidae 0.08% Decapoda 0.08% Isopoda 0.08% Mysidacea 0.08% Tanaidacea 0.08% Trefusiida 0.08%

PAGE 118

101 Table 5.1 Continued 90 m 120 m 140 m S = 25 S = 24 S = 29 N = 2098 N = 2001 N = 2197 Taxa % Taxa % Taxa % Chromadorida 28.41% Chromadorida 34.03% Copepoda 30.22% Crustacean larvae 25.60% Copepoda 21.34% Chromadorida 28.49% Copepoda 17.25% Monhysterida 10.99% Crustacean larvae 12.56% Monhysterida 6.20% Crustacean larvae 7.70% Hesionidae 7.60% Ostracoda 5.86% Platyhelminthes 6.45% Ostracoda 5.37% Platyhelminthes 5.67% Gastrotricha 4.70% Gastrotricha 4.19% Gastrotricha 4.00% Enoplida 3.85% Platyhelminthes 3.46% Enoplida 2.96% Ostracoda 3.45% Monhysterida 2.00% Ciliophora 1.81% Hesionidae 3.20% Syllidae 1. 55% Tardigrada 0.43% Ciliophora 1.40% Ciliophora 1.41% Syllidae 0.38% Syllidae 0.90% Enoplida 1.23% Cumacea 0.19% Tardigrada 0.35% Acari 0.41% Capitellidae 0.19% Acari 0.30% Tardigrada 0.32% Cnidaria 0.19% Tubificidae 0.25% Polychaeta (larvae) 0.23% Trefusiida 0.14% Cumacea 0.20% Nerillidae 0.18% Orbiniidae 0.14% Capitellidae 0.20% Kinorhynchia 0.14% Sipuncula 0.14% Enchytraidae 0.20% Enchytraidae 0.09% Glyceridae 0.10% Nerillidae 0.15% Capitellidae 0.05% Dorvilleidae 0.05% Cnidaria 0.10% Dinophil idae? 0.05% Paraonidae 0.05% Glyceridae 0.05% Glyceridae 0.05% Polychaeta (larvae) 0.05% Polychaeta (larvae) 0.05% Orbiniidae 0.05% Sabellidae 0.05% Kinorhynchia 0.05% Pisionidae 0.05% Enchytraidae 0.05% Gastropoda 0.05% Gammeridea 0.05% Nemertea 0.05 % Nemertea 0.05% Caprellea 0.05% Porifera 0.05% Tanaidacea 0.05% Insect larvae 0.05% Gastropoda 0.05% Opisthobranchia 0.05% Nemertea 0.05%

PAGE 119

102 Table 5.1 Continued 180 m 250 m 300 m S = 32 S = 33 S = 39 N = 2294 N = 20 23 N = 3180 Taxa % Taxa % Taxa % Chromadorida 38.54% Copepoda 37.52% Copepoda 34.37% Copepoda 25.59% Chromadorida 29.81% Chromadorida 26.95% Crustacean larvae 9.94% Crustacean larvae 8.80% Crustacean larvae 16.07% Ostracoda 7.24% Monhysterida 4.35% Ostracoda 5.03% Monhysterida 3.97% Platyhelminthes 2.47% Enoplida 3.43% Syllidae 2.75% Enoplida 2.42% Monhysterida 3.02% Gastrotricha 2.40% Ciliophora 2.37% Ciliophora 1.82% Enoplida 2.01% Syllidae 2.32% Syllidae 1.76% Ciliophora 1.92% Ostracoda 2.17% T ardigrada 1.42% Platyhelminthes 1.48% Acari 1.93% Acari 1.38% Nerillidae 0.92% Polychaeta (larvae) 1.09% Gastrotricha 1.04% Acari 0.92% Tardigrada 1.04% Platyhelminthes 0.79% Tardigrada 0.48% Kinorhynchia 0.99% Kinorhynchia 0.66% Hesionidae 0.22% Gast rotricha 0.54% Hesionidae 0.44% Polychaeta (larvae) 0.22% Hesionidae 0.40% Cumacea 0.31% Sabellidae 0.22% Dorvilleidae 0.35% Dorvilleidae 0.22% Spionidae 0.22% Cumacea 0.35% Trefusiida 0.16% Dorvilleidae 0.13% Capitellidae 0.15% Polychaeta (larvae) 0.1 3% Cumacea 0.13% Cirratulidae` 0.15% Enchytraidae 0.09% Pisionidae 0.09% Spionidae 0.10% Caprellea 0.09% Cnidaria 0.09% Isopoda 0.10% Pisionidae 0.06% Nereididae 0.09% Dinophilidae? 0.05% Sabellidae 0.06% Rotifera 0.09% Nerillidae 0.05% Tanaidacea 0.0 6% Ampharetidae 0.04% Paraonidae 0.05% Cnidaria 0.06% Cirratulidae` 0.04% Pilargiidae 0.05% Rotifera 0.06% Paraonidae 0.04% Pisionidae 0.05% Trochophore larvae? 0.06% Polynoidae 0.04% Polynoidae 0.05% Bivalvia 0.06% Gammeridea 0.04% Protodriloides cf. chaetifer 0.05% Capitellidae 0.03% Kinorhynchia 0.04% Sabellidae 0.05% Ctenodrilidae 0.03% Trefusiida 0.04% Tubificidae 0.05% Dinophilidae? 0.03% Nemertea 0.04% Enchytraidae 0.05% Fauveliopsidae? 0.03% Priapulida 0.04% Cnidaria 0.05% Flabelligeridae? 0.03% Rotifera 0.05% Maldanidae 0.03% Paraonidae 0.03% Pilargiidae 0.03% Polynoidae 0.03% Gastropoda 0.03% Sipuncula 0.03% Priapulida 0.03%

PAGE 120

103 Table 5. 2 Median, minimum and maximum meiofaunal community index values for taxa richness, raw abundance counts, Shannon diversity index (base e), and evenness by site. Site n Taxa Richness Abundance Shannon Diversity Evenness 0 m 5 10 52 1.93 0.87 5 15 11 178 1.41 2.17 0.67 0.94 30 m 5 11 210 1.81 0.74 11 16 96 370 1.58 2.05 0.63 0.76 60 m 5 12 208 1.87 0.71 10 15 168 361 1.63 1.94 0.69 0.78 90 m 5 16 452 1.96 0.71 14 16 244 515 1.72 2.06 0.65 0.76 120 m 5 15 419 1.93 0.70 14 16 229 573 1.67 2.05 0.62 0.76 140 m 5 15 376 1.71 0.61 11 18 158 727 1.43 2.06 0.59 0.80 180 m 5 17 395 1.87 0.61 15 24 332 745 1.60 1.95 0.59 0.70 250 m 5 18 338 1.86 0.62 17 23 253 619 1.73 1.94 0.57 0.66 300 m 5 21 582 1.81 0.60 19 23 468 971 1.64 1.99 0.56 0.65 Ref 5 15 141 1.77 0.64 12 16 118 176 1.52 2.06 0.58 0.76 Table 5.3 Relative abundance of dominant taxa representing >50% relative abundance at the 0 m site (excluding crustacean larvae) Taxon Phylum Relative Abundance (N = 357) Cumulative Abundance Harpacticoidea spp. Arthropoda 24.09% 24 .09% Innocuonema sp. A Nematoda 12.89% 36.97% Acari spp. Arthropoda 6.72% 43.70% Capitella cf. capitata Annelida 5.32% 49.02% Trileptium ? sp. A Nematoda 3.92% 52.94% The meiofaunal abundance ranged from as few as 11 specimens/sample at one of the 0 m replicates to 971 individuals/sample at the 300 m site (Table 5. 2). Meiofaunal abundance increa sed with distance from the vent but was highly variable at each individual site (Figure 5. 4) There was a significant difference in abundance among sites

PAGE 121

104 (ANO VA, F = 4.158; p < 0.001), with abundances at the 300 m, 180 m and 120 m sites being significantly higher than at the 0 m and reference sites (HolmSidak multiple comparison test; p<0.001) Meiofaunal abundance at the reference site was lower than at all t he transect site except 0 m and was less variable among replicate samples (Table 5. 2; Figure 5. 4). Figure 5. 3 Ambitle Island May/June 2005: Mean meiofaunal taxa richness 1 standard deviation. The Shannon diversity index values were highly variable a mong the replicate samples at each site (Figure 5. 5) and there was not a significant difference between sites (Kruskal Wallis ANOVA: H = 6.352; df = 9; p = 0.704). Evenness values showed a decreasing trend with distance from the vent (Table 5. 2; Figure 5. 6) The re were significant differences in the evenness values among sites (ANOVA: F = 6.030; p<0.001), with the 0 m site being significantly higher than the 120 m through 300 m transect sites and the reference site.

PAGE 122

105 Figure 5. 4 Ambitle Island May/June 2005: Mean meiofaunal abundance 1 standard deviation. Abundance presented as individuals/sample. Figure 5. 5 Ambitle Island May/June 2005: Mean meiofaunal Shannon diversity index 1 standard deviation.

PAGE 123

106 Figure 5. 6 Ambitle Island May/June 2005: Mean meiofaunal evenness 1 standard deviation. The ratio of nematode to copepod abundance has been used as a measure of environmental stress (Rafaelli and Mason 1981) The abundance values for nematodes and copepods and the nematode :copepod ratios are presented in Table 5. 4. Both ne matodes and copepod abundances increased with distance from the vent (Figures 5. 7 and 5. 8). Nematode abundances were significantly different among sites (ANOVA: F = 4.158; p < 0.001), with the120 m, 180 m and 300 m sites having si gnificantly higher nematode abundance than at the 0 m and reference sites (Holm Sidak multiple comparison test; p<0.001) The copepod abundance among sites was also significantly different (Kruskal Wallis: H = 28.281; df = 9; p < 0.001). The copepod abunda nce was significantly higher at the 300 m site than the 0 m and 30 m sites and the 250 m site a lso had higher copepod abundance than at the 0 m site (Tukey Test; p<0.05) The nematode:copepod ratio exhibited a decreasing trend with distance fr om the vent s ite

PAGE 124

107 (Figure 5. 9), approaching a 1:1 ratio at the 250 m and 300 m sites. The reference site had a ratio value < 1 (Table 5.2; Figure 5.7). There were significant differences among sites (Kruskal Wallis: H = 25.802; df = 9; p = 0.002), with the 30 m site bei ng significantly higher than the reference site (Tukey Test; p<0.05) Table 5. 4 Median, minimum and maximum values for nematode and copepod abundances (individuals/sample) and the nematode:copepod ratio at each site. Site n TOTAL_NEMATODA COPEPODA N_C_RAT IO 0 m 5 18 10 2.00 2 106 1 43 1.13 2.47 30 m 5 99 45 2.83 60 223 20 106 1.94 3.60 60 m 5 84 60 1.63 71 195 43 119 1.01 2.91 90 m 5 170 75 2.00 65 226 53 85 1.23 3.32 120 m 5 186 81 2.11 86 278 49 121 1.16 5.67 140 m 5 140 134 1.09 43 257 5 260 0.74 8.60 180 m 5 202 119 1.70 140 315 74 185 1.56 2.08 250 m 5 133 120 1.11 100 251 77 255 0.53 1.30 300 m 5 249 205 0.99 101 330 141 333 0.52 1.77 Ref 5 42 68 0.67 30 66 52 84 0.44 1.27

PAGE 125

108 Figure 5. 7 Ambitle Island May/June 2005: Mean nematode abundance 1 standard deviation. Abundance presented as individuals/sample. Figure 5. 8 Ambitle Island May/June 2005: Mean copepod abundance 1 standard deviation. Abundance presented as individuals/sample.

PAGE 126

109 Figure 5. 9 Ambitle Isla nd May/June 2005: Mean nematode:copepod ratio 1 standard deviation. Figures 5. 10 and 5. 11a show the similarity analysis based on all replicates and on site averaged abundance data respectively. The results for both analyses show the sites generally gr ouping together relative to their distance from the vent based on Bray Curtis similarity values >60 and SIMPROF test groupings with more variability seen in the complete dataset The analysis for the complete dataset shows on e replicate from the vent site (0 B) as an outlier. This sample had the fewest taxa and lowest abundance which separated it out from the other samples. Several replicates from the 120 m and 140 m sites also were outliers, but overall replicate samples tended to group together within th eir sites or with adjacent sites. The results for the group averaged data suggest that the meiofauna form four distinct communities relative to t he distance along the transect, with the Danlum Bay reference site grouping with the 0 m site (Figure 5. 11a ). The four meiofaunal communities identified consisted of the 0 m site + the Danlum Bay reference

PAGE 127

110 site (Group A); the 30 m + 60 m sites (Group B); the 90 m, 120 m + 140 m sites (Group C); and the 180 m + 250 m + 300 m sites (Group D). Removing the reference site data from the analysis still resulted in the same site groups (Figure 5.11b). SIMPER analysis of the defined meiofaunal community groups (Table 5. 5) showed that several taxa were common across all communities including harpacticoid copepods, chromado rid nematod es, monhysterid and enoplid nema todes larval crustaceans, and flatworms (Platyhelminthes) The group A community was characterized by the co ntribution of t refusiid nematodes and marine mites (Acari spp.) to the similarity among the 0 m and refe rence sites (Table 5.5 ). Th e group B community also had ciliates (Ciliophora) and tardigrades contributing to the similarity among the 30 m and 60 m sites. The Group C community included gastrotrichs, ostracods and the polychaete families Syllidae and Hesi onidae. The group D community was the most diverse with 15 taxa cumulatively contributing to >90% of the similarity among the 180 m, 250 m and 300 m sites. This community also was characterized by tardigrades, mites, cumaceans and kinorhynchs. The BIO ENV analysis (Table 5.6a) correlating the meiofaunal community structure with the physical parameters measured at each site shows that the strongest salinity, and sediment organic carbon. The combinat ions of temperature, pore water salinity, sediment organic carbon and median grain size had only a slightly lower 96). Reanalysis of the BIO ENV without the reference site (Table 5.6b) showed a stronger correlation with the combined

PAGE 128

111 T he LINKTREE analysis (Figure 5.12) separated the 0 m and reference site meiofaunal communities from the remaining sites. These two sites had lower ORP readings relative to the other sites. Node B of the LINKTREE separated the 30 m and 60 m meiofaunal commu nity from the other sites based on slightly higher temperature, smaller median grain size and lower CaCO3 sediment content at those two sites. Node C of the link tree divided the remaining sites into two groups: the 180 m, 250 m and 300 m group and the 90 m, 120 m and 140 m group. The 180 m, 250 m, 300 m group was characterized by higher sediment sorting coefficients, higher sediment organic content, higher pH, higher CaCO3 content, lower arsenic concentrations, and larger median grain size relative to the other remaining sites.

PAGE 129

112 Figure 5. 10 Ambitle Island May/June 2005: Meiofaunal Bray Curtis similarity cluster analysis.

PAGE 130

113 Figure 5. 11a Ambitle Island May/June 2005: Meiofaunal Bray Curtis similarity cluster analysis averaged by site. Figure 5.11b Amb itle Island May/June 2005: Meiofaunal Bray Curtis similarity cluster analysis averaged by site; referen ce site omitted.

PAGE 131

114 Table 5. 5 SIMPER results for meiofaunal community groupings. Group A Group B Group C Group D 0 m + Ref 3 0 m +6 0 m 9 0 m +12 0 m +14 0 m 18 0 m +25 0 m +30 0 m Average similarity: 70 Average similarity: 79 Avera ge similarity: 81 Average similarity: 83 Taxa Contrib% Taxa Contrib% Taxa Contrib% Taxa Contrib% Copepoda 20. 7 Nematoda/Chromadorida 22.1 Nematoda/Chromadorida 19.9 Copepoda 18.3 Nematoda/Chromadorida 17.7 Copepoda 19.4 Copepoda 16.0 Nematoda/Chromadorida 18.1 Nematoda/Monhysterida 10.6 Nematoda/Monhysterida 13.6 Crustacean larvae 11.3 Crustacean larvae 10.4 Nematoda/Enoplida 9.9 Crustacean larvae 11.4 Platyhelminthes 7.6 Nematoda/Monhyst erida 6.3 Crustacean larvae 9.7 Ciliophora 8.3 Gastrotricha 7.5 Ostracoda 5.9 Syllidae 8.4 Nematoda/Enoplida 7.4 Ostracoda 7.5 Nematoda/Enoplida 5.0 Platyhelminthes 5.0 Platyhelminthes 6.7 Nematoda/Monhysterida 6.6 Syllidae 4. 8 Ostracoda 4.2 Tardigrada 3. 6 Nematoda/Enoplida 4. 9 Ciliophora 4. 7 Nematoda/Trefusiida 3.8 Ciliophora 4.4 Acari 3.5 Acari 3.1 Syllidae 2.7 Platyhelminthes 3.4 Hesionidae 2.2 Gastrotricha 2.9 Tardigrada 2.8 Hesionidae 1.8 Cumacea 1.5 Kinorhynchia 1.4 Cumulative % 92.9 Cumulative % 92.6 Cumulative % 90.6 Cumulative % 90.7

PAGE 132

115 Table 5.6a BIO ENV Spearman c orrelations and best fit of physical parameters with the meiofaunal community structure. Spearman Correlation Parameters 0.565 Temperature, ORP, Salinity, % Organics 0.564 Temperature, Salinity ,% Organics, Median Grain Size 0.563 Temperature, ORP, Salinity, % Organics, Median Grain Size 0.560 Temperature,pH,Salinity,% Organics, Median Grain Size 0.559 Tempera ture,pH,ORP, Salinity, % Organics 0.557 Temperature, Salinity, % Organics 0.557 Salinity, % Organics, Median Grain Size 0.556 pH,ORP, Salinity, % Organics, Median Grain Size 0.556 ORP, Salinity, % Organics, Median Grain Size 0.553 pH, Salinity, % Orga nics, Median Grain Size 0.512 Salinity 0.496 Temperature 0.344 ORP 0.342 Median Grain Size 0.327 Sorting 0.326 Total [As] 0.244 pH 0.239 % Organics 0.100 %CaCO 3 Table 5.6b BIO ENV Spearman correlations and best fit of physical parameters wit h the meiofaunal community structure; r eference site omitted. Spearman Correlation Parameters 0.664 Temperature, Salinity 0.623 Salinity 0.622 Temperature,pH,Salinity, Total [As], Median Grain Size 0.621 Temperature,pH,Salinity, Median Grain Size 0. 619 Temperature,pH, ORP,Salinity, Total [As] 0.618 Temperature 0.617 pH, Salinity, Median Grain Size 0.617 pH, ORP,Salinity, Total [As], Median Grain Size 0.616 pH, Salinity, Total [As],Median Grain Size 0.614 Temperature,pH, ORP,Salinity, Median Grai n Size 0.510 Median Grain Size 0.450 Total [As] 0.396 Sorting 0.350 ORP 0.318 pH 0.112 % Organics 0.105 %CaCO 3

PAGE 133

116 Figure 5.12 LINKTREE diagram of meiofaunal community site groupings and corresponding physical characteristics. 5.4 Discussion Meiofaunal communities at Tutum Bay were dominated by nematodes and copepods, reflecting a global pattern typical of most meiofaunal communities ( Riemann 1988) Nematodes were dominant at the vent and at most of the inner transect sites which were influen ced by diffuse venting. This trend of nematode dominance is similar to other hydrothermal vent systems, both shallow water and deep sea, which also host meiofaunal communities dominated by nematodes (Kamenev et al 1993, Thiermann et al 1994 Vanreusel et al 1997). The trend of low taxonomic richness and abundance at the vent site and increasing richness and abundance away from the focused venting is also

PAGE 134

117 characteristic of other shallow water hydro thermal systems (Kamenev et al 1993, Thiermann et al 1994). P resent among the meiofaunal community were juvenile Capitella cf. capitata, which were present at 9 of the 10 sites and indicate s a continued recruitment of this polychaete throughout the system. Juveniles were most abundant at the 0 m site which accounted for 46% of the specimens found. The dominance of adult Capitella cf. capitata at the vent site, and its relatively low abundance at the other sites in the macrofauna samples indicate the ability of this species to tolerate the extreme conditions at the vent while being out competed at the other sites The meiofaunal communities at the inner transect sites from 30 m to 140 m away from the vent were similar. These sites split into two subgroups. The meiofaunal community at the 30 m and 60 m sites had fewer taxa and lower abundances than the 90 m 140 m sites, which also had relatively fewer taxa but increased abundances. This trend possibly reflects the sensitivity of the meiofaunal taxa to the pore water conditions and reflects the influence of diff use venting along this zone of the transect At all sites, nematodes were proportionally most abundant, but there was an increasing number of arthropods with distance along the transect. This may reflect the gradual increase in sediment grain size observed along the transect. The meiofaunal community in this zone may also have been influenced by the dominant macrofaunal taxa through physical disturbance by the burrowing activities of thalassinid shrimp and direct predation The negative influence of bioturb ation by the thalassinid shrimp Callianassa on nematode communities was reported by Alongi (1986) on the Great Barrier Reef where he found reduced densities of nematodes associated with high densities of Callianassa burrows.

PAGE 135

118 The meiofaunal communities at the 180 m 300 m sites had higher taxonomic richness than the other transect sites but the abundances were similar to the 90 m 140 m sites. The se offshore transect sites had nearly equal proportion of nematodes and copepods. The higher abundance of cop epods and of many typically interstitial phyla might have been due to the larger median grain size and higher sorting coefficient of the sediments, resulting in more interstitial space between sediment grains to support meiofaunal diversity. The meiofaunal community at the Danlum Bay reference site had a relatively low number of taxa and abundance and grouped with the 0 m vent site in terms of its species similarity. This site was unique in the strong dominance of copepods, which accounted for 47% of the re lative abundance. Copepods are typically more abundant in coarser, well oxygenated sediments, or associated with phytal habitats (Hicks and Coul l, 1983) so their dominance at the Danlum Bay site is enigmatic and may be an artifact of the low overall meiof aunal abundance at this site 5.5 Summary and conclusions The meiofaunal communities in Tutum Bay and at the Danlum Bay reference site reflect the complex interactions among the physical and biological components of the system. The meiofauna were most st rongly correlated with pore water characterist ic s such as temperature and salinity as well as with the sediment parameters such as organic content and median grain size. Biological interactions with the macrofaunal community also may play an important role in structuring the meiofaunal community especially at sites where high densities of thalassinid shrimp rework the sediments.

PAGE 136

119 Chapter Six Molecular diversity of eukaryotic and bacterial communit ies associated with the shallow water hydrothermal vent at Ambitle Island, Papua New Guinea 6.1 Introduction Molecular m ethods have been used for a number of years to measure bacterial diversity and over the last decade have been used increasingly more to measure micro eukaryotic diversity in different envir onments such as soils (Wu et al 2009) In the marine environment, molecular surveys of eukaryotic diversity have focused on extreme habitats such as anoxic and sulfide environments (Dawson and Pace 2002, Stoeck and Epstein 2003, Behnke et al 2006, Takish ita et al 2007a ), deep sea hydrothermal vents ( Atkins et al 2000 Edgcomb et al 2002, Lpez Garca et al 2003, 2007, Takishita et al 2005, Le Calvez et al 2009), and deepsea methane cold seeps (Takishita et al 2007b and 2010). One com mon finding in molecular surveys is the unexpectedly hig h diversity of eukaryotic groups and the discovery of new lineages. Examples include early mole cular surveys of marine picoplankto n ( Lpez Garca et al 2001, Moon van der Staay et al 2001) which found high divers it ies of previously unknown taxa in samples collected from Antarctica and the equatorial Pacific respectively and a global survey of the protozoan phylum Cercozoa (Bass and Cavalier Smith 2004). Despite the high diversity discovered in these studies, mole cular methods tend to underestimate the actual diversity believed to be present in the habitat (Potvin and Lovejoy 2009). Molecular analysis of 18S and 16S rDNA was employed in this study in order to

PAGE 137

120 describe the composition of the sediment microeukaryoti c and bacterial communities along the hydrothermal gradient at Tutum Bay. Additionally, I wanted to evaluate the effectiv eness of using molecular methods as a measure of meiofaunal diversity in comparison to traditional techniques by focusing on the metazo an component of the eukaryotic analysis. The results of this comparison will be presented in Chapter 7. 6.2 Material and methods 6.2.1 Sampling design and field collection Sampling and analysis methods for pore water and sediment samples collected at th e transect sites and the Danlum Bay reference site are given in Chapter 2. Sediment samples for meiofaunal community analysis were collected at the nine transect sites and at the Danlum Bay reference site as described in Chapter 2. Sediment samples for DNA extractions were collected from a random grid in each of the five quadrats at each site using a 60cc syringe coring tube (diameter = 3cm). Sediment cores were collected to a depth of 5 cm. The five sediment cores from each site were pooled into a single s mesh plankton netting to remove macrofauna and retain meiofaunal sized organisms. net to remove any planktoni c organisms and prevent possible contamination of the samples. The retained sediment was split into three aliquots, two of which were kept frozen and the third was preserved in 95% ETOH. The frozen samples were hand carried back to the United States and st ored at 80C upon return to USF.

PAGE 138

121 6.2.2 DNA extraction and amplification Sediment DNA was extracted from a 0.5g subsample from each site using a MoBio UltraClean Soil DNA isolation kit following the alternate protocol for maximum yield with the followin g modification: Step # 15 of the protocol was repeated one extra time The 18S rDNA fraction of the extracted sediment DNA was PCR amplified using 18S4 and 18S5 primers following the following PCR mixture: 10x buffer dXTPs Taq 18S4 Primer 18S5 Primer MgCl 2 extracted sediment DNA ultrapure H 2 0 Total PCR reaction mix 18S4: 5 CCGGAATTCAAGCTTGCTTGCTTGTCTCAAAGATTAAGCC 3 18S5: 5 CCGGAATTCAAGCTTACCATACTCCCCCCGGAACC3 The PCR reaction was cycle sequenced using t he following reaction : 95C denaturing for 15 seconds, 50C annealing for 1 minute, 72C polymerization for 1 minute x 45 cycles followed by 72C extension for 7 minutes (Figure 6.1).

PAGE 139

122 4 C Hold Figure 6.1 Cyclesequenc ing reaction for eukaryotic 18S gene amplification. 2 min 95 C 95 C 95 C Hold 15 sec 7 min 72 C 72 C 1 min 50 C 1 min 45 x

PAGE 140

123 cloned into TOPO cloning vector and transformed into chemically competent TOP10 strain E. coli using a n Invitrogen TOPO Cloning Kit. C lones were minipreped in 96 well plates using the Eppendorf Perfectprep Plasmid 96 Vac miniprep kit The final sequencing was done by Polymorphic DNA Technologies Inc. (Alameda, CA). Final 18S sequences were trimmed to a final length of 576 bp around the 18S6 region (150 bp before, and 400 bp after) using ANCHOR software (Garey Lab proprietary software) Operational Taxonomic Units (OTUs) were assembled based on 99% sequence similarity using S EQUENCHER 4.7 (Gene Codes Ann Arbor, MI). The assembled OTU sequences were BLAST searched against the NCBI database (GENBANK) for tentative taxonomic and phylogenetic identifications. Sediment samples for bacterial community analysis using16S r DNA were collected at selected sit es along the transect Results and community analysis are presented here for comparison with the eukaryotic sequence trends along the transect. The DNA was extracted from approximately 1 gram of sediment sample using the MoBio UltraClean Soil DNA isolation kit protocols. PCR amplification of the 16S rDNA was done using primers 27F (5 AGA GTT TGA TCC TGG CTC AG 3) and 1492R (5 GGT TAC CTT CTT ACG ACT T 3 ) and using the following thermocycling procedure: 95C denaturing for 15 seconds, 45C anneal ing for 1 minute, 72C polymerization for 2 minutes x 40 cycles followed by 72C extension for 7 minutes (Figure 6.2).

PAGE 141

124 4 C Hold Figure 6.2 Cyclesequenc ing reaction for bacterial 16S gene amplification. 2 min 80C 95 C 95 C 1 min 15 sec 7 min 72 C 72 C 2 min 45 C 1 min 40 x

PAGE 142

125 The purified PCR product was transformed using an Invitrogen TOPO Cloning Kit and clones were minipr eped in 96 well plates using the Eppendorf Perfectprep Plasmid 96 Vac miniprep kit as described for the 18S r DNA methods. The final sequencing was done by Polymorphic DNA Technologies, Inc. (Alameda, CA). Operational Taxonomic Units (OTUs) were assembled based on 97% sequence similarity using S EQUENCHER 4.7 (Gene Codes, Ann Arbor, MI). The assembled OTU sequences were BLAST searched against the NCBI database (GENBANK) for tentative taxonomic and phylogenetic identifications. 6.3 Results A total of 3,840 eukaryotic sequences were initially obtained from all ten sites. After trimming to the final 576 bp sequence length, 3,004 sequences remained which were assembled into 971 OTUs based on 99% sequence similarity. The dataset was fu rther reduced by removing questionable or non target sequences (i.e. cloning vector, terrestrial organisms, vertebrate sequences) after BLAST searching the GENBANK database, leaving a final dataset of 2,987 sequences and 923 OTUs. The number of useable sequences obtained per site was variable and ranged from 224 to 362 at 120 m and 0 m respectively (Figure 6.3) The number of OTUs per site was also variable, ranging from 68 at the 60 m site to 144 at the 250 m site (Figure 6.4). There was no significant correlation between the numbe r of sequences and corresponding number of OTUs (R2 = 0.174) as shown in the linear regression analysis (Figure 6.5).

PAGE 143

126 Figure 6. 3 Ambitle Island May/June 2005: Number of eukaryotic sequences obtained per site. Figure 6.4 Ambitle Island May/June 2005: Number of eukaryote operational taxonomic units (OTUs) obtained per site.

PAGE 144

127 Figure 6.5 Linear regression analysis of OTUs vs. number of sequences. Metazoans were represented by 35% of the overall sequences, followed by Fungi (30%) and Stramenopiles ( 17%). Uncultured Eukaryota from environmental sequences and Alveolata each comprised 7% of the total sequences (Figure 6.6 ). The sequence composition at each site was variable, with Metazoan s comprising over half of the sequences at the 0 m site as well as at the 250 m and 300 m sites (Figure 6.7) while Fungi accounted for the majority of sequences at the 120 m and the reference sites and Stramen o pile sequences were prevalent at 30 m and Alveolates at 18 0 m (Figure 6.7). Metazoans comprised the largest fraction of the OTUs (30%) followed by the Fun gi and Stramenopiles (Figure 6.8). The OTU composition was dominated by metazoans at the 0 m site as well as at 14 0 m 180 m 250 m and 300 m. Fungi OTUs were dominate at 12 0 m and at the reference site and S tramen opiles comprised the largest percentage of OTUs at the 30 m, 60 m and 90 m sites (Figure 6.9).

PAGE 145

128 Figure 6. 6 Ambitle Island May/June 2005: Percentage of eukaryote sequences by kingdom level taxonomic category for all sites combined. Figure 6. 7 Amb itle Island May/June 2005: Percentage of eukaryote sequences by kingdom level taxonomic category at each site.

PAGE 146

129 Figure 6. 8 Ambitle Island May/June 2005: Percentage of eukaryote OTUs by kingdom level taxonomic category for all sites combined. Figure 6. 9 Ambitle Island May/June 2005: Percentage of eukaryote sequences by kingdom level taxonomic category at each site.

PAGE 147

130 Thirty three OTUs (3.6 %) accounted for 50% of the sequences with the top four ranked OTUs making up 25% of the sequences (Table 6.1). The majority of OTUs were found at only one site ( 851; 92.2%) and 72.3% (667) of the OTUs were represented by a single sequence. The most abundant OTU was Contig [ 0002] identified as the fungus Paecilomyces sp. (GENBANK Accession # DQ401104). This OTU com prised over 12% of the total sequences and was present at 8 of the 10 sites. This OTU was absent at 0 m and 9 0 m and most abundant at 6 0 m w h ere it accounted for 50% of the sequences at that site. Other top ranked OTUs were provisionally identified as the fungus Cladosporium cladosporioides ( Contig [0008]; GENBANK Accession # EF114717), the nematode Viscosia viscosa (AY854198; Contig [0006]), and the bivalve mollusk Astarte castanea (AF120551; Contig [0001] ) No OTUs were present at all ten sites. The most frequently occurring OTUs were the fungi Paecilomyces sp. (Contig [0002]) and Penicillium chrysogenum (Contig [0013]), each present at eight sites (Table 6.1) while Cladosporium cladosporioides (Contig [0008]) was present at seven sites (Table 6.1). Cumul atively these three OTUs accounted for over 20% of the sequences. Additionally, seven OTUs were found at five of the ten sites. These included two OTUs (C onti gs[0644] and [0541]) which were also identified as Paecilomyces sp T hree were identified as stame no piles, including two as the same unclassified diatom ( Contigs [0172] and [0235]; GENBANK Accession number AB18359) an d the third (Contig 0184] as the oom y co te s Phytophthora infestans ( Accession number AY742744). The remaining two OTUs were Contig [0042] identified as the nematode

PAGE 148

131 Chromadorita tentabundum ( AY854208) and Contig[0074], identified as the chlorophyte Cladophora glomerata, (AB062706). The 0 m site had a total of 362 sequences in 76 OTUs. Metazoans dominated both the number of sequences (213; 59%) and the OTUs (34; 45%). Six OTUs accounted for >50% of the sequences at the 0 m site (Table 6.2) The top two OTUs were both metazoans: Contig[0058] was identified as the nematode Viscosia viscosa (AY854198) and comprised 14% of the sequences and C ontig[0066] which was identified as the polychaete Capitella capitata (U 67323). These two taxa accounted for > 25% of the eukaryotic sequences at the 0 m site. The re were 307 sequences representing 124 OTUs at the 30 m site Fifteen OTUs accounted for >50 % of the sequences. Stramenopiles were dominant in both the number of seq uences and OTUs comprising 52% and 44% respectively. The top two OTUs were Contig[0068] identified as the diatom Pauliella toeniata (AY485528) and Contig[0002] identified as the fun gus Paecilomyces sp. 080834, (DQ401104) These two OTUs comprised nearly 24% of the total sequences at the 3 0 m site (Table 6.2) The 60 m site had a total of 273 sequences and had the fewest OTUs with 68. Fungi comprised 58% of the sequences, while Stramenopiles had the largest percentage of OTUs (3 5%). A single OTU, the fungus Paecilomyces sp. (Contig[0002] dominated the site, accounting for over 50% of the sequences (Table 6.2) The 90 m site had a total of 277 sequences and 113 OTUs. Metazoans compris ed the largest pr oportion of the sequences (29 %) while Stramenopiles made up the largest percentage of OTUs (29.2 %). A total of 21 OTUs accounted for >50% of the sequences.

PAGE 149

132 The dominant OTU was Contig[0042] identified as the nematode Chromadorita tentabundum (AY854208) which accounted for 8.3 % of the sequences (Table 6.2) The 120 m site had the fewest amplified sequences with 224 and a total of 122 OTUs. Fungi dominated bot h the number of sequences (37 %) and accounted for 45% of the OTUs. Twenty five OTUs made up >50% of the sequences. Two OTUs tied as the top ranked taxon each accounting for 7.6 % of the sequences (Table 6.2) These were Contig[0002] identified as the fungus Paecilomyces sp. ( DQ401104) and Contig[0010] identified as the acoel flat worm Pse udaphanostoma smithrii (AY078375). The 140 m site had a total of 321 sequences and 104 OTUs. Metazoans made up the larges t percentage of sequences (47.7 %) and accounted for half of the OTUs. Three OTUs accounted for over 50% of the sequences. The top rank ed OTU was the fungus Paecilomyces sp. ( Contig [ 0002]) which comprised nearly 30% of the sequences. The other two top ranked OTUs were Contig [0006] identified as the nematode Viscosia viscosa (AY854198) and Contig [ 0001] identified as a bivalve mollusk ( A starte castanea AF120551). These comprised 15.9% and 10.9% of the sequences respectively (Table 6.2) There were 296 sequences and 127 OTUs at the 180 m site The sequences were dominated by Alveolat es (31.8%) while Metazoans comprised 26.8 % of the OTUs. Fourteen OTUs accounted for >50% of the sequences. The top ranked OTU was Contig [ 0054] identified as the ciliate Strombidinopsis acuminata ( AJ877014) and incorporated 14.2% of the sequences (Table 6.2) The 25 0 m site had 279 sequences and the most OTU s of the ten sites, with 144. Metazoans dominated bot h the number of sequences (55.2%) a nd OTUs (36.1 %).

PAGE 150

133 Fourteen OTUs accounted for >50% of the sequences. The top ranked OTU was Contig [ 0006], identified as the nematode Viscosia viscosa (AY854198), which comprised 28. 7% of the sequences (Table 6.2) The 30 0 m site had 307 sequences and 85 OTUs. Meta zoans made up 56.4% of the sequences and 37.7% of the OTUs. Three taxa accounted for >50% of the sequences, with Contig [ 0002] Paecilomyces sp. (DQ401104) be ing the most abundant with 22.80% of the sequences). Contig [ 0001], identified as the bivalve mollusk Astarte castanea (AF120551) and Contig [ 0107], identified as the turbellarian flatworm Schizorhynchoides caniculatus (AY775748) ranked s econd and third co mprising 17.3% and 11.7% of the sequences respectively (Table 6.2) The reference site had 341 sequences and 96 OTUs. Fungi dom inated the site comprising 51.3% of the sequences and 33. 3% of the OTUs. Three OTUs accounted for >50% of the sequences. The mos t abundant OTU was Contig [ 0008] identified as the fungus Cladosporium cladosporioides (EF114717) and c omprising 31.4% of the sequences. Contig [ 0018], identified as a polychaete annelid worm ( Amphicorina mobilis AY611449), ranked second and had 13.2% of the sequences. The third most abundant OTU was Contig [ 0013] identified as the fungus Penicillium chrysogenum (EU203859) and accounted for 8.50% of the sequences (Table 6.2)

PAGE 151

134 Table 6.1 Ranked OTU s representing 50% of the 2,987 t otal sequences. OTUs rank ed by percent abundance and then by number of sites at which they occur. Tied ranks indicated by an asterisk. Rank Sequence BLAST ID Accession # Kingdom Phylum Percent Cumulative Sites 1 Contig[0002] Paecilomyces sp. 080834 DQ401104 Fungi Dikarya 12.06% 12.06% 8 2 Contig[0008] Cladosporium cladosporioides EF114717 Fungi Dikarya 5.76% 17.82% 7 3 Contig[0006] Viscosia viscosa AY854198 Metazoa Nematoda 5.10% 22.92% 3 4 Contig[0001] Astarte castanea AF120551 Metazoa Mollusca 3.20% 26.12% 4 5 Contig[0013] Penicillium chrysogenum EU203859 Fungi Dikarya 2.33% 28.45% 8 6 Contig[0058] Viscosia viscosa AY854198 Metazoa Nematoda 1.67% 30.11% 1 7 Contig[0107] Schizorhynchoides caniculatus AY775748 Metazoa Platyhelminthes 1.57% 31.68% 3 8 Contig[0042] Chromadori ta tentabundum AY854208 Metazoa Nematoda 1.53% 33.21% 5 9 C ontig[0066] Capitella capitata U67323 Metazoa Annelida 1.53% 34.74% 1 10 Contig[0018] Amphicorina mobilis AY611449 Metazoa Annelida 1.50% 36.24% 1 11 Contig[0068] Pauliella toeniata AY485528 St ramenopiles Bacillariophyta 1.43% 37.67% 2 12 Contig[0054] Strombidinopsis acuminata AJ877014 Alveolata Ciliophora 1.40% 39.07% 1 13 Contig[0003] Cheliplana cf. orthocirra AJ012507 Metazoa Platyhelminthes 0.93% 40.01% 1 14 Contig[0059] Cheliplana cf. or thocirra AJ012507 Metazoa Platyhelminthes 0.70% 40.71% 3 15 Contig[0056] Parodontophora sp. PB 2005 AM234630 Metazoa Nematoda 0.70% 41.41% 1 16 Contig[0644] Paecilomyces sp. 080834 DQ401104 Fungi Dikarya 0.63% 42.04% 5 17 Contig[0541] Paecilomyces sp. 0 80834 DQ401104 Fungi Dikarya 0.57% 42.60% 5 18 Contig[0116] uncultured alveolate AF372785 Alveolata Alveolata 0.57% 43.17% 4 19 Contig[0010] Pseudaphanostoma smithrii AY078375 Metazoa Platyhelminthes 0.57% 43.74% 1 20* Contig[0015] uncultured cercozoan AY620357 Cercozoa Cercozoa 0.53% 44.27% 4 20* Contig[0046] Navicula ramosissima AY485512 Stramenopiles Bacillariophyta 0.53% 44.80% 4 21 Contig[0189] Haliphthoros sp. NJM 0034 AB178865 Stramenopiles Oomycetes 0.50% 45.30% 4 22 Contig[0175] Atolla vanho effeni AF100942 Metazoa Cnidaria 0.50% 45.80% 1 23 Contig[0027] Navicula sp. CCMP2746 EF106790 Stramenopiles Bacillariophyta 0.47% 46.27% 4 24 Contig[0016] Strombidium sp. SNB99 2 AY143564 Alveolata Ciliophora 0.47% 46.74% 1 25 Contig[0074] Cladophora g lomerata AB062706 Viridiplantae Chlorophyta 0.43% 47.17% 5 26* Contig[0113] Pseudechiniscus islandicus AY582119 Metazoa Tardigrada 0.43% 47.60% 1 26* Contig[0557] Strombidinopsis acuminata AJ877014 Alveolata Ciliophora 0.43% 48.03% 1 27 Contig[0112] Pse udechiniscus islandicus AY582119 Metazoa Tardigrada 0.40% 48.43% 3 28 Contig[0060] Schizosaccharomyces japonicus AB243296 Fungi Dikarya 0.40% 48.83% 2 29* Contig[0011] Exophiala salmonis EF413608 Fungi Dikarya 0.40% 49.23% 1 29* Contig[0367] Cheliplana cf. orthocirra AJ012507 Metazoa Platyhelminthes 0.40% 49.63% 1 30 Contig[0439] Diascorhynchus rubrus AJ012508 Metazoa Platyhelminthes 0.37% 50.00% 2

PAGE 152

135 Table 6.2 Relative abundance the top five ranked OTUs (including ties) at each site 0 m 30 m OTU BLAST ID Accession # % OTU BLAST ID Accession # % Contig[0058] Viscosia viscosa AY854198 13.81% Contig[0068] Pauliella toeniata AY485528 12.38% Contig[0066] Capitella capitata U67323 12.71% Contig[0002] Paecilomyces sp. 080834 DQ401104 11.40% Contig[0008] Cl adosporium cladosporioides EF114717 10.22% Contig[0215] Dickieia ulvacea AY485462 2.61% Contig[0056] Parodontophora sp. PB 2005 AM234630 5.80% Contig[0081] Protaspis grandis DQ303924 2.61% Contig[0013] Penicillium chrysogenum EU203859 5.80% Contig[0585] Pauliella toeniata AY485528 2.28% Contig[0175] Atolla vanhoeffeni AF100942 4.14% Contig[0686] Pauliella toeniata AY485528 2.28% Contig[0541] Paecilomyces sp. 080834 DQ401104 2.28% Contig[0644] Paecilomyces sp. 080834 DQ401104 2.28% Contig[0 027] Navicula sp. CCMP2746 EF106790 2.28% Contig[0015] uncultured cercozoan AY620357 1.95% Table 6.2 Continued 60 m 90 m OTU BLAST ID Accession # % OTU BLAST ID Accession # % Contig[0002] Paecilomyces sp. 080834 DQ401104 50.92% Contig[0042] Chroma dorita tentabundum AY854208 8.30% Contig[0042] Chromadorita tentabundum AY854208 4.03% Contig[0016] Strombidium sp. SNB99 2 AY143564 5.05% Contig[0189] Haliphthoros sp. NJM 0034 AB178865 2.20% Contig[0060] Schizosaccharomyces japonicus AB243296 3.61% Co ntig[0119] Montastraea annularis AF238267 2.20% Contig[0222] Pseudaphanostoma smithrii AY078375 3.25% Contig[0112] Pseudechiniscus islandicus AY582119 2.20% Contig[0046] Navicula ramosissima AY485512 2.89% Contig[0055] Uncultured stramenopile clone CCI6 AY179998 1.83% Contig[0068] Pauliella toeniata AY485528 1.83% Contig[0203] Amphidinium sp. PEL 2 AB092335 1.47% Contig[0328] Uncultured eukaryotic picoplankton clone VN3 DQ409093 1.47%

PAGE 153

136 Table 6.2 Continued. 120 m 140 m OTU BLAST ID Accession # % OTU BLAST ID Accession # % Contig[0010] Pseudaphanostoma smithrii AY078375 7.59% Contig[0002] Paecilomyces sp. 080834 DQ401104 29.91% Contig[0002] Paecilomyces sp. 080834 DQ401104 7.59% Contig[0006] Viscosia viscosa AY854198 15.89% Contig[0023] Apodachlya brachynema AJ238663 2.68% Contig[0001] Astarte castanea AF120551 10.90% Contig[0046] Navicula ramosissima AY485512 2.68% Contig[0107] Schizorhynchoides caniculatus AY775748 3.12% Contig[0013] Penicillium chrysogenum EU203859 2 .68% Contig[0644] Paecilomyces sp. 080834 DQ401104 2.49% Contig[0541] Paecilomyces sp. 080834 DQ401104 2.23% Contig[0229] Bacillariophyta sp. MBIC10102 AB183593 1.79% Contig[0001] Astarte castanea AF120551 1.79% Contig[0116] uncultu red alveolate AF372785 1.79% Contig[0117] Gammarus duebeni AF419227 1.79% Contig[0416] uncultured alveolate AF530536 1.79% Contig[0112] Pseudechiniscus islandicus AY582119 1.79% Contig[0139] Xenotrichula sp. aff. Velox AY963686 1.79% Contig[0251] Uncultured eukaryote clone D1P02C04 EF100195 1.79% Contig[0189] Haliphthoros sp. NJM 0034 AB178865 1.34% Table 6.2 Continued 180 m 250 m OTU BLAST ID Accession # % OTU BLAST ID Accession # % Contig[0054] Strombi dinopsis acuminata AJ877014 14.19% Contig[0006] Viscosia viscosa AY854198 28.67% Contig[0008] Cladosporium cladosporioides EF114717 5.74% Contig[0115] Thalassarachna basteri AY692342 2.51% Contig[0557] Strombidinopsis acuminata AJ877014 4.39% Contig[0606 ] Neochromadora BHMM 2005 AY854210 2.51% Contig[0113] Pseudechiniscus islandicus AY582119 4.39% Contig[0087] Leptolaimus sp. LeLaSp1 EF591323 2.51% Contig[0011] Exophiala salmonis EF413608 4.05% Contig[0093] Trichosporon porosum AB051045 2.15% Contig[00 12] Bradya sp. Greenland RJH 2004 AY627016 3.04% Contig[0110] Chromadorita tentabundum AY854208 1.79% Contig[0075] Gymnodinium dorsalisulcum DQ837534 1.79% Contig[0711] Psammodictyon panduriforme AY485485 1.43% Contig[0321] uncultured eukaryote EU087264 1.43% Contig[0218] Polytoma oviforme U22936 1.43%

PAGE 154

137 Table 6.2 Continued 300 m Ref OTU BLAST ID Accession # % OTU BLAST ID Accession # % Contig[0002] Paecilomyces sp. 080834 DQ401104 22.80% Contig[0008] Cladosporium cladospor ioides EF114717 31.38% Contig[0001] Astarte castanea AF120551 17.26% Contig[0018] Amphicorina mobilis AY611449 13.20% Contig[0107] Schizorhynchoides caniculatus AY775748 11.73% Contig[0013] Penicillium chrysogenum EU203859 8.50% Contig[0006] Viscosia vi scosa AY854198 7.17% Contig[0003] Cheliplana cf. orthocirra AJ012507 8.21% Contig[0059] Cheliplana cf. orthocirra AJ012507 5.86% Contig[0099] Uncultured marine eukaryote clone SA2_2G5 EF527173 2.64%

PAGE 155

138 The nonmetric multidimensional scaling analysis grouped the sites into three main eukaryotic communities based on a Bray Curtis similarity index value of 5 (Figure 6.10a ). These groups consisted of the 0 m + Reference sites, the 18 0 m + 25 0 m sites, and the six remaining sites (3 0 m 140 m + 30 0 m ). Th is latter grouping form ed two distinct subgroups consisting of the 3 0 m + 6 0 m sites and the 14 0 m + 30 0 m sites. Leaving out the reference site from the analysis did not alter the overall site groupings (Figure 6.10b) The SIMPER analysis revealed that the 0 m + Ref grouping was due primarily to the presence of two fungi, Cladosporium cladosporioides ( Contig [ 0008]) and Penicillium chrysogenum ( Contig [ 0013]). The grouping of the 180 m + 25 0 m sites was due in part to the presence of the fungus Cladosporium cladosporioides (Contig[0008]), the Chlorophytes Polytoma oviforme (Contig[0218]) and Cladophora glomerata (Contig[0074]) and a Metazoan sequence a marine mite, Thalassarachna basteri (Contig[0115]). The 3 0 m 140 m + 30 0 m grouping was due primarily to the presence of the fungus Paecilomyces sp. (Contig[0002]). Two additional contributing OTUs, Contig[0541] and Contig[0644], were also identified in GENBANK as Paecilomyces sp. (DQ401104), Two Metazoan sequences also contributed to the similarity among the sites in this group, the nematode Chromadorita tentabundum (Contig[0042]) and the bivalve mollusk Astarte castanea (Contig[0001]), as well as Contig[0184], a Stramenopile (Oomycetes) which closely matched Phytophthora infestans (AY742744) The subgroup co nsisting of the 3 0 m + 6 0 m sites was defined by the fungus Paecilomyces sp. (Contig[0002]) as well as by several OTUs identified as Stramenopiles T hese included sequences identified as the diatoms Pauliella toeniata (Contig[0068]),

PAGE 156

139 Navicula sp. (Contig[0027]) and Bacillariophyta sp. (Contig[0172]). Other contributing OTUs included an Oomycetes closely matching Haliphthoros sp NJJM0034 (AB178865; Contig[0189]) and an uncultured Stramenopile clone (AY179998; Contig[0055]) and an uncultured cercozoan (AY620357; Contig [0015]). Figure 6. 10a Ambitle Island May/June 2005: Multi Dimen s ional Scaling (MDS) of eukaryotic community structure based on Bray Curtis similarity among sites.

PAGE 157

140 Figure 6.10b Ambitle Island May/June 2005: Multi Dimensional Scaling (MDS) of eukaryotic community structure based on Bray Curtis similarity among sites; r eference site omitted. The subgroup consisting of the 14 0 m + 30 0 m sites was defined by the fungus Paecilomyces sp. (Contig[0002]), the bivalve mollusk Astarte castanea (Cont ig[0001]), and the nematode Viscosia viscosa (Contig[0006]). The BIO ENV results correlating the eukaryotic molecular community structure and the physical variables are presented in Table 6.3a T he overall correlations were low, but the strongest correlat ion was with the combination of pore water salinity, median n grain size and sorting coefficient had a slightly lower coefficient had the highest correlation for single variables ( t otal pore water 3 0.313).

PAGE 158

141 Table 6.3a BIO ENV correlations between eukaryotic community structure and physical parameters. Spearman Correlation Parameters 0.4 47 Salinity, Median Grain Size, Sorting 0.444 Median Grain Size, Sorting 0.442 ORP, Total [As], %CaCO 3, Median Grain Size, Sorting 0.441 Salinity, %CaCO 3, Median Grain Size, Sorting 0.438 Salinity, %CaCO 3, Sorting 0.437 Temperature, Median Grain Size, Sorting 0.437 ORP, Total [As], %CaCO 3, Sorting 0.434 ORP, Salinity Total [As], %CaCO 3, Sorting 0.433 Total [As], Sorting 0.433 Temperature, Salinity, Median Grain Size, Sorting 0.394 Sorting 0.358 Total [As] 0.324 Median Grain Size 0.313 %CaCO 3 0.241 ORP 0.241 % Organics 0.190 Temperature 0.169 Salinity 0.158 pH Table 6.3b BIO ENV correlations between eukaryotic community structure and physical parameters; reference site omitted. Spearman Correlation Parameters 0.510 ORP, Total [As ], %CaCO3, Median Grain Size, Sorting 0.508 Total [As], Sorting 0.506 ORP, Salinity, Total [As], Sorting 0.505 ORP, Total [As], Sorting 0.504 pH, ORP, Salinity, Sorting 0.504 pH, ORP, Sorting 0.504 Temperature, pH, ORP, Salinity, Sorting 0.503 pH, ORP Salinity, Total [As], Sorting 0.502 Total [As], %CaCO3, Median Grain Size, Sorting 0.501 Temperature, ORP, Total [As], Sorting 0.457 Total [As] 0.432 Median Grain Size 0.430 Sorting 0.321 % Organics 0.310 %CaCO 3 0.286 Salinity 0.283 ORP 0. 277 Temperature 0.202 pH

PAGE 159

142 The BIO ENV results excluding the reference site had higher correlations between the community structure and physical parameters (Table 6.3b). The arsenic concentration had the strongest correlation for a single parameter along with The LINKTREE analysis (Figure 6. 11) separated the 18 0 m + 25 0 m site grouping out from the other sites at the first split of the LINKTREE (node A), base d o n differences in pore water pH with the 18 0 m and 250 m sites having pH > 7.8 vs. < 7.74 at the other sites The split at node B separated out the 0 m + Ref grouping from the other sites based on the oxygen reduction potential (ORP) values, with the 0 m and reference sites having ORP values < 5.2 mV while the remaining sites had ORP > 52 mV. Node C separated the 9 0 m site based on the ORP being > 156 mV vs. <118 mV and pore water salinity being > 34.1 psu (vs. < 33.9 psu ). The final division in the LINK TREE separated the 14 0 m + 30 0 m community from the 3 0 m +6 0 m + 12 0 m grouping based on the pH being >7.06 ( vs. <6.59), the sorting coefficient >0.597 ( vs. <0.449). the ORP <91.2 ( vs. >98.4), the pore water total arsenic <25.8 ppb (vs. >43.5 ppb) pore water temperature <31.5 ( vs. >31.9C ) median grain size >0.23mm (vs. <0.198mm), CaCO3 >0.631% (vs. <0.422%) and pore water salinity > 31.6 psu (vs. <31.5 psu)

PAGE 160

143 Figure 6. 11 LINKTREE results showing physical characteristics between eukaryotic community site groups. The analysis was reevaluated looking only at the metazoan sequences in order to better compare the molecular results with the meiofaunal and macrofaunal communi ty analysis. There were a total of 1,059 metazoan sequences and 273 metazoan OTUs from the ten sites. Nematodes dominated both the number of sequences and OTUs followed by Platyhelm i nthes and Annelida (Figures 6.12 and 6.13). Ten OTUs (3.66%) accounted for >50% of the sequences (Table 6.4) while 201 OTUs (73.63%) were represented by a

PAGE 161

144 single sequence. The most abundant metazoan sequence was Contig[0006] identified as the nematode Viscosia viscosa and comprised 14.45% of the sequences. This OTU occurred at three sites: 14 0 m 250 m and 300 m Contig[0058] ranked third overall and was als o matched to Viscosia viscosa This OTU was only present at the 0 m site. The second ranked OTU was Contig[0001] which was identified as the bivalve mollusk Astarte castanea and represented over 9% of the sequences. Table 6.4 Top te n ranked metazoans comprising >50% of the 1,059 metazoan sequences. OTU BLAST ID (Phyla) Accession # Percent Cumulative Freq Contig[0006] Viscosia viscosa (Nematoda) AY854198 14.45% 14.45% 3 Contig[0001] Astarte castanea (Mollusca) AF120551 9.07% 23.51% 4 Contig[0058] Viscosia viscosa (Nematoda) AY854198 4.72% 28.23% 1 Contig[0107] Schizorhynchoides caniculatus ( Platyhelminthes ) AY775748 4.44% 32.67% 3 Contig[0042] Chromadorita tentabundum (Nematoda) AY854208 4.34% 37.02% 5 Contig[0066] Capitella capi tata (Annelida) U67323 4.34% 41.36% 1 Contig[0018] Amphicorina mobilis (Annelida) AY611449 4.25% 45.61% 1 Contig[0003] Cheliplana cf. orthocirra ( Platyhelminthes ) AJ012507 2.64% 48.25% 1 Contig[0059] Cheliplana cf. orthocirra ( Platyhelminthes ) AJ012507 1.98% 50.24% 3 Contig[0056] Parodontophora sp. PB 2005 (Nematoda) AM234630 1.98% 52.22% 1 Most of the OTUs (95.97%) only occurred at a single site. The most frequently occurring OTU was Contig[0042] identified as the nematode Chromadorita tentabundum T his OTU was the only one present at five of the ten sites, with its highest abundance at the 90 m site. The second most frequently occurring OTU was

PAGE 162

145 Contig[ 0001] ( Astarte castanea) which was found at four sites with its highest numbers at the 30 0 m and 140 m sites. Figure 6. 12 Proportion of metazoan sequences by phyla. Figure 6. 13 Proportion of metazoan OTUs by phyla.

PAGE 163

146 The composition of metazoan sequences was variable between s ites (Figures 6. 14 and 6.15) The 0 m site had 213 metazoan sequences and 34 OTUs. Nematodes dominated both the number of seq uences and OTUs, making up 37% and 27% respectively. Annelids comprised 25% of the sequences, while annelids and pl atyhelminthes each made up 23.5 % of the OTUs. Three OTUs accounted for >50% of the seque nces. The most abundant O TUs were Contig [0058], which was close to the nematode Viscosia viscosa and represented 23.5% of the sequences, followed by Contig[0066] identified as the polychaete annelid Capitella capitata accounting for 22 % The 3 0 m site h ad only 24 metazoan sequences and 14 OTUs. Platyhelminthes accounted for 62.5% of the sequences and 57 % of the OTUs at this site. The most abundant OTU was Contig[0365] identified as the acoel platyhelminthes Pseudaphanostoma smithrii (AY078375) which repr esented 17 % of the sequences. This was closely followed by Contig[0523], identified as the acoel Philocelis karlingi (AJ845243) and Contig[0152], and identified as the acoel Pelophila lutheri (AY078366). Each represented 12.5 % of the sequences at the 3 0 m site. There were 30 metazoan sequences and 9 OTUs at the 60 m site Three OTUs accounted for >75% of the sequences at that site. Nematoda represented 40% of the sequences, while Cnidarians comprised 33% of the OTUs. The most abundant OTU was Contig[0042] identified as the nematode Chromadorita tentabundum (AY854208) which represented 3 7% of the sequences. Contig[ 0119], a cnidarian sequence near Montastraea annularis (AF238267) and Contig[0112], a tardigrade near Pseudechiniscus islandicus (AY582119) each comprised 20% of the sequences at the 6 0 m site.

PAGE 164

147 The 9 0 m site had 80 metazoan sequences represented by 25 OTUs. Nematoda and Platyhelminthes comprised 35% and 30% of the sequences respectively. These two phyla also made up 20% and 36% of the OTUs respect ively. Four OTUs accounted for >50% of the sequences. The most abundant OTU was Contig[0042] identified as the nematode Chromadorita tentabundum (AY854208) which comprised 29% of the sequences at that site. The 12 0 m site had 43 metazoan sequences in 12 OTUs. Pl atyhelminthes comprised 49% of the sequences and 25% of the OTUs. Three OTUs accounted for >50% of the sequences. The dominant OTU was Contig[0010] identified as the acoelomate flatworm Pseudaphanostoma smithrii (AY078375) and representing 39.5% of the sequences. There were 153 metazoan sequences and 52 OTUs at the 140 m site Nematodes dominate d the site, accounting for 54% of the sequences and 58 % of the OTUs. Two OTUs made up >50% of the sequences. The most abundant OTU was Contig[0006] identified as the nematode Viscosia viscosa (AY854198) and represented a third of the sequences. The second most abundant OTU was Contig[0001] identified as the bivalve mollusk Astarte castanea (AF120551) and comprising 23% of the sequences. It should also be note d that in addition to Contig[0006], 22 additional OTUs at the 140 m site also had BLAST search results for the same Viscosia viscosa sequence in GenBank. None of these OTUs were the same as Contig[0058] which was also identified as Viscosia viscosa (AY854198) from the 0 m site. The 18 0 m site had 82 metazoan sequences and 34 OTUs. Tardigrades made up 46% of the sequences and 38% of the OTUs. Five OTUs accounted for >50% of the

PAGE 165

148 sequences. The dominant OTU was Contig[0113] identified as the tardigrade Pseude chiniscus islandicus (AY582119), which represented 16 % of the sequences. There were 154 metazoan sequences and 52 OTUs at the 250 m site The site was strongly dominated by nematodes, which comprised 89% of the sequences and 79% of the OTUs. Contig[0006], identified as the nematode Viscosia viscosa (AY854198) accounted for 52 % of the sequences. Additionally, 29 singleton OTUs from this site also had BLAST search results for Viscosia viscosa (AY854198). The 30 0 m site had 173 metazoan sequences and 32 OTUs Pla tyhelminthes accounted for 43% of the sequences and 41 % of the OTUs. Two OTUs comp rised >50% of the sequences. The most abundant OTU was Contig[0001] identified as the bivalve mollusk Astarte castanea (AF120551) and making up 31% of the sequences. Con tig[0107], identified as the Turbellarian flatworm Schizorhynchoides caniculatus (AY775748) represented 21 % of the sequences. The reference site had 107 metazoan sequences and 29 OTUs. Pol ychaete annelids comprised 51% of the sequences and 34.5% of the OT Us. Platyhelminthes additionally accounted for 36.5% of the sequences and 41 % of the OTUs. Two OTUs represented >50% of the sequences. The most abundant OTU was Contig[0018], identified as a polychaete close to Amphicorina mobilis ( AY611449) and accounting for 42% of the sequences. The second most abundant OTU was Contig[0003] identified as a turbellarian flatworm close to Cheliplana cf. orthocirra (AJ012507) and comprising 26% of the sequences. Nine additional singleton OTUs from the reference site also had BLAST results for Amphicorina mobilis (AY611449) and ten singleton OTUs were also identified as Cheliplana cf. orthocirra (AJ012507).

PAGE 166

149 Figure 6.1 4 Percentage of metazoan sequences by phyla at each site and all sites combined (All). Figure 6.15 Per centage of metazoan OTUs by phyla at each site and for all sites combined (All). 1

PAGE 167

150 The MDS analysis based on the metazoan sequence similarity between sites grouped the 30 m 140 m sites together with the 30 0 m site at a Bray Curtis similarity of 5 (Figure 6.16a) Within this grouping, the 30 m 6 0 m and 9 0 m sites formed a subgroup at a similarity value of 10, as did the 120 m 140 and 300 m site s (Figure 6.16a ). The 0 m 180 m 250 m and Reference sites each fell out as thei r own g rou ps in this analysis (Figure 6.16a ). Excluding the reference site from the analysis did not change the overall site groupings in the MDS analysis (Figure 6.16b) Figure 6.16a Non metric Multi Dimensional Scaling (MDS) plot of metazoan sequenc e Bray Curtis similarity between sites. Sequence data standardized by site total and square root transformed.

PAGE 168

151 Figure 6.16b Non metric Multi Dimensional Scaling (MDS) plot of metazoan sequence Bray Curtis similarity between sites ; reference site omitted. Sequence data standardized by site total and square root transformed. SIMPER results of the site groupings shown in the MDS analysis found that the grouping of the 30 m 6 0 m and 9 0 m sites was due to Contig[0042] identified as the nematode Chromadorita tentabundum (AY854208) which contributed 77.5% to the similarity among these sites. The grouping of the 120, 140 m and 300 m sites was due to Contig[0001] identified as the bivalve mollusk Astarte castanea (AF120551) and Contig [0059] identified as th e turbellarian flatworm Cheliplana cf. orthocirra (AJ012507) These two OTUs contributed 48% and 18% respectively to the among site similarity The BIO ENV analysis between the metazoan sequences and the physical parameters found t hree combinations of ph ysical variables that had equally strong

PAGE 169

152 correlations to the metazoan community structure (Table 6.5a) The combination of pore water salinity + percent organics + median grain size, the combination of pore water temperature + sediment sorting, and the com bination of ORP + percent organics + median grain size each had a correlation of 0.440. Sediment sorting had the highest correlation 0.352). Table 6.5a BIO ENV correlations be tween metazoan community structure and physical parameters. Spearman Correlation Parameters 0.440 Salinity, % Organics, Median Grain Size 0.440 Temperature Sorting 0.440 ORP % Organics, Median Grain Size 0.437 Total [As], Median Grain Size, Sorting 0.436 Salinity, Total [As], Median Grain Size, Sorting 0.435 % Organics, Median Grain Size 0.435 Temperature, pH, % Organics, Median Grain Size, Sorting 0.434 Temperature, Total [As], Median Grain Size, Sorting 0.433 Salinity, Median Grain Size, Sorti ng 0.433 pH, % Organics, Median Grain Size, Sorting 0.413 Sorting 0.352 Total [As] 0.293 Median Grain Size 0.281 ORP 0.274 % Organics 0.215 Temperature 0.212 %CaCO 3 0.201 pH 0.183 Salinity The BIO ENV results excluding the reference si te had stronger correlations overall (Table 6.5b). These results show pH in combination with other parameters correlating more strongly with the community structure while the percent organics had a weaker correlation (Table 6.5b).

PAGE 170

153 Table 6.5b BIO ENV corre lations between metazoan community st ructure and physical parameters; r eference site omitted. Spearman Correlation Parameters 0.538 pH, ORP, Sorting 0.534 Temperature, pH, ORP, Sorting 0.532 pH, ORP, Total [As], Sorting 0.531 pH, ORP, Salinity, Total [As], Sorting 0.530 Total [As], Median Grain Size, Sorting 0.529 pH, ORP, Salinity, Sorting 0.529 ORP, Salinity, Total [As], Sorting 0.529 Salinity, Total [As], Median Grain Size, Sorting 0.528 ORP, Total [As], Sorting 0.528 Temperature, pH, ORP, Sal inity, Sorting 0.491 Sorting 0.458 Total [As] 0.434 Median Grain Size 0.336 ORP 0.319 Salinity 0.307 Temperature 0.247 pH 0.232 %CaCO 3 0.179 % Organics The LINKTREE analysis based on the metazoan only sequence similarity (Figure 6.17) s eparated the 0 m and Reference site out from the other sites at node A based on their lower ORP measurements (ORP < 5.2 vs. >52.6). Node B split out the 18 0 m + 25 0 m sites from the remaining sites based on their higher pH values (>7.8 vs. <7.74). The fina l node separated the 3 0 m + 6 0 m + 90 m sites from the 12 0 m + 14 0 m + 300 m sites base on differences in the median grain size.

PAGE 171

154 Figure 6.17 LINKTREE results showing physical characteristics between metazoan sequence community site groups.

PAGE 172

155 A total of 25 bacterial phyla were found, representing 1,973 sequences in 1,081 OTUs (Table 6.6). The Proteobacteria had the largest percentage of OTUs and sequences followed by the Actinobacteria (Figures 6.1 8 and 6.19). Most phyla ranked similarly in the percentage of OTUs and percentage of sequences the y represented ( Figures 6.17 and 6.18). One nota ble exception were the Firmicutes, which comprised only 4% of the OTUs but represented 14% of the sequences (Figures 6.18 and 6.19). There was a decreasing trend in the number of phyla found at each site with dis tance from the vent (Figure 6.20 ). T he num ber of OTUs per site was higher at the intermediate sit es, peaking at 20 m (Figure 6.21). The farthest site from the vent (180 m) had only 4 phyla represented and also had the lowest number of OTUs (Table 6.6; Figures 6.20 and 6.21). The trends in the dist ribution of OTUs and sequences among phyla at each site are shown in Figures 6.22 and 6.23. Actinobacteria dominated at the 7.5 m site in both the number of OTUs and sequences, while Choroflexi was dominant at the 12 m site (Table 6.6; Figures 6.22 and 6.23). The Proteobacteria together represented a large percentage of the bacterial composition at the 20 m through 120 m sites along the transect (Table 6.6; Figures 6.22 and 6.23). The 180 m site was strongly dominated by Firmicutes which rep resented 89 % of the OTUs and 99% of the sequences at that site (Table 6.6; Figures 6.22 and 6.23). The cluster analysis results indicate the bacteria form three distinct groups in relation to their distance along the transect (Figure 6.24 ). The two sites closest to the focused venting (7.5 m and 12 m) had a distinct bacterial community, the 20 m through 120 m sites formed a distinct community and the 180 m site separated out as a unique community. SIMPER analysis indicated that the similarity among the 7. 5 m and 12 m

PAGE 173

156 sites w as defined by the presence of Chlorobi and Chloroflexi. The similarity among the Proteobacteria phyla, while the bacterial community at the 180 m site was defined by the dominance of Firmicutes at that site. Table 6. 6 Number of OTUs (S) and seque nces (N) for bacterial phyla for all sites (Total) and individual sites. Total 7.5 m 12 m 20 m 30 m S N S N S N S N S N Acidobacteria 140 205 19 32 3 3 30 36 35 42 Actinobacteria 164 311 40 71 10 11 48 65 35 47 Bacteroidetes 43 58 7 11 15 18 4 4 7 9 Chlamydiae 7 7 2 2 0 0 1 1 0 0 Chlorobi 26 106 13 28 15 78 0 0 0 0 Chloroflexi 60 109 11 22 33 65 10 11 3 3 Cyanobacteria 2 2 1 1 0 0 1 1 0 0 Deferribacteres 5 5 3 3 1 1 1 1 0 0 Deinococcus -Thermus 2 3 2 3 0 0 0 0 0 0 Fibrobacteres 2 2 0 0 0 0 1 1 1 1 Firmicutes 46 278 4 5 4 8 3 3 4 4 Gemmatimonadetes 8 15 3 8 1 1 3 3 1 1 Nitrospirae 12 27 1 1 2 3 1 1 2 5 Planctomycetes 104 116 17 22 5 5 19 19 13 13 Proteobacteria 3 3 0 0 0 0 1 1 1 1 Spirochaetes 1 1 0 0 0 0 0 0 0 0 Thermodesulfobacteria 4 4 1 1 1 1 1 1 0 0 Thermotogae 1 1 0 0 1 1 0 0 0 0 Verrucomicrobia 16 24 1 1 1 3 2 2 4 5 -Proteobacteria 67 115 1 1 15 31 7 7 13 21 -Proteobacteria 3 4 0 0 3 4 0 0 0 0 -Proteobacteria 204 358 15 20 3 3 44 56 45 67 -Proteobacteria 153 209 17 21 14 30 33 37 31 33 -Proteobacteria 4 6 2 2 1 3 0 0 1 1 Bacteria Undet. 4 4 1 1 1 1 2 2 0 0 Total 1081 1973 161 256 129 270 212 252 196 253 Number of Phyla 25 20 19 19 15

PAGE 174

157 Table 6.6 Continued. 60 m 90 m 120 m 180 m S N S N S N S N Acidobacteria 25 29 19 25 30 38 0 0 Actinobacteria 31 52 26 40 21 25 0 0 Bacteroidetes 7 7 5 5 4 4 0 0 Chlamydiae 0 0 1 1 3 3 0 0 Chlorobi 0 0 0 0 0 0 0 0 Chloroflexi 1 2 4 4 2 2 0 0 Cyanobacteria 0 0 0 0 0 0 0 0 Deferribacteres 0 0 0 0 0 0 0 0 Deinococcus -Thermus 0 0 0 0 0 0 0 0 Fibrobacteres 0 0 0 0 0 0 0 0 Firmicutes 3 8 5 6 3 5 24 239 Gemmatimonad etes 0 0 2 2 0 0 0 0 Nitrospirae 1 1 4 7 4 9 0 0 Planctomycetes 13 13 10 12 29 32 0 0 Proteobacteria 0 0 1 1 0 0 0 0 Spirochaetes 0 0 0 0 1 1 0 0 Thermodesulfobacteria 1 1 0 0 0 0 0 0 Thermotogae 0 0 0 0 0 0 0 0 Verrucomicrobia 4 4 3 5 3 4 0 0 -Pr oteobacteria 17 19 12 14 16 21 1 1 Proteobacteria 0 0 0 0 0 0 0 0 -Proteobacteria 33 45 38 78 58 88 1 1 -Proteobacteria 31 34 21 26 20 26 1 2 -Proteobacteria 0 0 0 0 0 0 0 0 Bacteria Undet. 0 0 0 0 0 0 0 0 Total 167 215 152 226 193 258 27 243 N umber of Phyla 12 14 13 4

PAGE 175

158 Figure 6.1 8 Percentage of total bacterial OTUs (S = 1,081) by phyla. Figure 6.1 9 Percentage of total bacterial sequences (N = 1,973) by phyla.

PAGE 176

159 Figure 6.20 Total number of bacterial phyla represented at each transect s ite. Figure 6.21 Total number of bacterial OTUs represented at each transect site.

PAGE 177

160 Figure 6.22 Percentage of bacterial OTUs by phyla at each transect site. Figure 6.23 Percentage of bacterial sequences by phyla at each transect site.

PAGE 178

161 Figure 6.24 Cluster analysis of bacterial sequence Bray Curtis similarit y between sites. Sequence data was standardized by site totals and square root transformed. Dashed horizontal line demarcates 7.5 Bray Curtis similarity. Red dendrogram branches indicate sign ificant SIMPROF groupings. 6.4 Discussion The molecular eukaryotic diversity at the vent site was very similar to the macrofaunal and meiofaunal communities at that site There was a relatively low number of OTUs dominated by m etazoans T he dominant OTU s were provisionally identified as the nematode ( Viscosia viscosa ) and the polychaete ( Capitella capitata ) which are close matches to the taxa found in the macrofaunal and meiofaunal samples at this site. Both of

PAGE 179

162 these taxa are relatively large and were a bundant, which may have also contributed to their representation in the 18S rDNA extractions. The eukaryotic community in inner transect zone, which included the 30 m through 140 m sites also encompassed the 300 m site. This community was defined by the p resence of an OTU provisionally identified as the fungus Paecilomyces sp. and two metazoan OTUs provisionally identified as the nematode Chromadorita tentabundum and the bivalve mollusk Astarte castanea, as well as a Stramenopile (Oomycetes) which closely matched Phytophthora infestans The presence of the molluscan sequence in this zone is enigmatic since there were few mollusks found in the corresponding meiofaunal samples. These sequences may represent recently settled larvae, which may not survive to th e adult stage in the low pH environment. The 30 m and 60 m sites formed a subgroup defined by the fungus Paecilomyces sp. and several stramenopile OTUs provisionally identified as the diatoms Pauliella toeniata Navicula sp., and unidentified Bacillarioph yta The presence of diatom sequences suggests that the benthic microalgal community is im portant in primary production at these sites The eukaryotic community at the 180 m and 250 m sites had the most OTUs which suggests the eukaryotic diversity is hi gher farther from the vent, although this pattern was very variable at the other transect sites The dominant eukaryotes included OTUs closely matching the fungus Cladosporium cladosporioides the c hlorophytes (green algae ) Polytoma oviforme and Cladophora glomerata and a marine mite, Thalassarachna basteri The presence of chlorophyte sequences at these sites suggests that macroalgaes fro m the surrounding reef system are important contributors to the primary production at these sites This may be either th rough photos ynthetic or detrital

PAGE 180

163 pathways and fu rther suggests that the fungal OTUs present at that site may be associated with decomposing algal detritus. The eukaryotic community at the Danlum Bay reference site was dominated by fungi, which are probabl y associated with the breakdown of detritus found at that site. Among the most abundant OTUs were the fungi provisionally identified as Cladosporium cladosporioides and Penicillium chrysogenum. The eukaryotic community was most similar to the 0 m site due to the presence of common fungal sequences. This is possibly due to the proximity to shore of both of these sites and the input of terrestrial detrital material from Ambitle Island. One unexpected finding from the eukaryotic diversity was t he large proport ion of fungal sequences Fungi are an often overlooked component of the biodiversity in marine enviro nments (Le Calvez et al 2009), and this finding highlights the utility of using molecular techniques in evaluating marine biodiversity. The most abundant sequence found was provisionally identified as the fungus Paecilomyces, and several OTUs had BLAST matches with this fungus. The genus Paecilomyces is a speciose and wide spread group of fungi which includes many thermophilic species (Sampson 1974). Paecil omyces sp p. have been recorded in other similar studies including in marine sediment samples from the continental slope of the Bay of Bengal in India (Das et al 2009) and sequences of this fungus have been reported to co amplify with targeted nematode 18 S rRNA in marine environmental samples (Bhadury et al 2006). The analysis of the bacterial sequence data also indicated that the environmental gradient along the transect influenced the bacterial community structure. Several s tudies at other similar sha llow water hydrothermal vent systems had found that molecular

PAGE 181

164 microbial diversity in the sediment was highest near the vent, suggesting the hydrothermal influence stimulates microbial activity ( Sievert et al. 2000 Manini et al 2008). An earlier study at the vent system in Milos, Greece by S ievert et al (1999) however, found bacterial cell counts were highest in the intermediate zone between the strongly hydrothermally influenced sediments and the normal marine sediments, while diversity was lowest near t he vent Corresponding work done at Ambitle Island looked at the possibility of microbial metabolism of arsenic in the sediments. Price et al (2007) looked at the distribution of As(V) and As(III) in sediment pore water s collected along the transect at Tu tum Bay T heir findings suggested that the higher concentrations of As(V) in pore water s near the vent could be due to microbial oxidation of As(III) from the hydrothermal fluid (Price et al 2007). More recently, an arsenic metabolizing Proteobacterium Marinobacter santor iniensis was isolated and described from a shallow water hydrothermal vent in Santorini, Greece (Handley et al 2009). The high Proteobacteria sequenced from the Tutum Bay sediments suggests that a similar species could be present here as well. 6.5 Summary and conclusions The molecular survey of eukaryot es and bacteria showed a high diversity for both of these domains and revealed the contribution of often overlooked lineages to the bio diversity of the shallow water vent system. The results for the eukaryotes did show some similarities to the macrofaunal and meiofaunal communities, particularly with the metazoans at the vent site There were s everal incongruencies in the proportions of metazoan sequences recovered which showed a bias in the molecular methods used. In

PAGE 182

165 particular the metazoan sequence dataset appeared to undersample arthropods, while having a disproportionally high number of flatworm sequences compared to the meiofaunal results. The molecular results howe ver did show the contribution of fungi and other eukaryotic groups as well as the bacteria to the overall diversity of the system, a finding that would have otherwise been overlooked by traditional benthic survey methods.

PAGE 183

166 Chapter Seven Biocomplexity of the communities associated with the shallow water hydrothermal system at Tutum Bay, Ambitle Island, Papua New Guinea. 7.1 Comparison of different biological communities and physical parameters. One goal of this project was to evaluate different method s for measuring biotic communities in particular the use of molecular techniques as a measure of the meiofaunal metazoan community vs. morphological identifications. To do this, I used second stage multidimensional scaling analysis (PRIMER v6) to find cor relations between the underlying distance measures of the physical variables and the similarity measures of each community: macrofauna, meiofauna, molecular eukaryotic and molecular metazoan. The bacterial sequence data were excluded since they were not av ailable from all sites, and several sites did not match the other datasets and lacked corresponding environmental data. The physical, macrofauna and meiofauna data were composited by site in to compare with the molecular data sets. The physical, macrofauna l and meiofaunal datasets were also compared against each other using all replicates. For comparison between the meiofauna and the molecular metazoan data the taxonomic resolution of the molecular OTUs was reduced to higher level taxonomic categories to ma tch the level of meiofaunal identifications. The second stage MDS results (Figure 7.1; Table 7.1) show that the macrofauna are more strongly correlated to the environmental data than either the meiofauna or molecular datasets. The meiofauna had the second highest correlation with the

PAGE 184

167 environmental data, while both the eukaryotic and metazoan molecular communities had relatively weak correlations (Table 7.1). Since the molecular metazoan dataset is a subset of the eukaryotic dataset, they both have similar correlations with the physical data and are also highly correlated to each other (Table 7.1). The meiofauna dataset had a weak correlation with the molecular metazoan data (Table 7.1). Table 7.2 shows the second stage MDS correlations between the physical data, macrofauna and meiofauna using all 50 replicate samples. These results mirror Table 7.1 although the correlation values are lower due to the variability among the replicate samples. Overall, the morphologically based macrofauna and meiofauna appear t o be better indicators of environmental changes along the transect than the molecular analysis. The macrofauna were better indicators than the meiofauna which may have been due to the lower level of taxonomic identification of the macrofauna. Conversely, t he identification of OTUs in the molecular dataset was dependent on the coverage of a particular taxonomic group in the GENBANK database. Additionally, setting the criteria for assembling OTUs at 99% similarity for the molecular data set may have underesti mated the number of species in a sample, as has been shown with soil nematodes and mites (Wu et al. 2009). To further compare the meiofauna and the molecular metazoan data, I looked at the higher level taxonomic composition of both datasets, standardizing them by relative abundance for the meiofauna and relative number of sequences for the molecular metazoans (Figure 7.2). Both datasets were proportionally similar for nematodes and some of the lesser known phyla such as gastrotrichs. The meiofauna data sho wed a much higher proportion of arthropods. The molecular data had higher proportions of platyhelminthes, annelids, mollusks (probably larvae) and tardigrades. These results

PAGE 185

168 suggest that the molecular methods used show a bias towards softer bodied organism s such as flatworms. The manual picking and counting of meiofauna may underestimate the number of soft bodied taxa due to problems with preservation and their small size. The difference in sample sizes in the two datasets may also have affected the compar ison. The meiofauna were represented by over 17,000 individual specimens compared to fewer than 3,000 18S rDNA sequences in the final eukaryote dataset. The sediment DNA extractions used only a small amount (< 1 g) of the total sediment samples whereas the meiofauna from the entire core sample were counted. Additionally, the meiofaunal analysis was based on five replicate samples from each site while for the molecular samples, replicate sediment cores were pooled into a single sample at each site. Despite t hese issues, the molecular dataset revealed a much broader phylogenetic diversity than the macrofauna and meiofauna datasets, and reflected the contribution of the nonmetazoan eukaryotes such as fungi, stramenopiles and alveolates to the overall sediment community. These often overlooked levels of biodiversity serve important ecological functions such as the breakdown of organic matter and the recycling of nutrients and primary productivity through photosynthesis by benthic diatoms and dinoflagellates.

PAGE 186

169 Figure 7.1 Second Stage Multi Dimensional Scaling plot showing correlation between the different datasets. Table 7.1 Second Stage MDS Spearman correlations. Physical parameters, m acrofauna and m eiofauna relative abundance data averaged by site for compa rison Physical parameters Eukaryotic sequences Macrofauna Metazoa sequences Physical parameters Eukaryotic sequences 0.39 Macrofauna 0.84 0.39 Metazoa sequences 0.38 0.76 0.31 Meiofauna 0.53 0.24 0.37 0.40 Table 7.2 Second Stage MDS S p earman correlations for physical parameters, macrofauna, and m eiofauna relative abundance for all replicate samples. Physical (all reps) Macrofauna (all reps) Physical (all reps) Macrofauna (all reps) 0.65 Meiofauna (all reps) 0.51 0.22

PAGE 187

170 Figure 7.2 Percentage of meiofaunal abundance (top graph) and metazoan sequences (bottom graph) by phyla.

PAGE 188

171 7.2 Summary of findings and c onclusions : the biocomplexity of the Tutum Bay hydrothermal system: The physical environment and biological communit ies in Tutum Bay and at the Danlum Bay reference site formed four distinct communities based on their pore water and sediment characteristics and the corresponding macrofaunal, meiofaunal, and molecular eukaryotic and bacterial community structure. These communities are farther developed by complex interactions between the biological benthic community and physical environment and among the different levels of the biodiversity. 7.2.1 The Vent C ommunity (0 m) The focused hydrothermal vent at the 0 m site di rectly impacts the sediment and pore water characteristics resulting in extreme conditions of high temperature, high arsenic, low salinity and pH in the acidic range. The combination of low pH and physical disturbance caused by the focused venting and poss ibly the influence of wave action due to the shallow depth at that site farther influence the sediment composition, resulting in a coarsegrained, poorly sorted volcanic gravel and an absence of calcareous sediments due to the low pH. The pore water enviro nment and sediment characteristics together have a direct influence on the biotic communities present at the vent site. Both the macrofaunal and meiofaunal communities are represented by a few species that can tolerate the harsh conditions at that site. Th e macrofaunal community is defined by the polychaete Capitella cf. capitata, which were found in very high abundance possibly due to the lack of competition with other macrofauna or a release from predation. The meiofaunal

PAGE 189

172 community also had a low number of taxa present and a low overall abundance. The meiofaunal taxa were apparently less tolerant of the extreme pore water conditions than the macrofauna, and no single taxon reached high abundances as was seen with the macrofaunal community. The meiofauna at the vent site were represented by a high proportion of nematodes, but other taxa were present including harpacticoid copepods and mites. Also present among the meiofaunal community were juvenile Capitella cf. capitata, which indicates a continued recruitm ent of the dominant polychaete at the vent site. The molecular eukaryotic diversity at the vent site was very similar to the macrofaunal and meiofaunal communities. There was a relatively low number of OTUs despite a large number of sequences that were ob tained from that site. Metazoans dominated both in the number of OTUs and in the number of sequences, with the dominant OTUs identified as a nematode ( Viscosia viscosa ) and a polychaete ( Capitella capitata). There were no bacterial samples collected at th e 0 m site, but the two closest sites (7.5 m and 12 m) were characterized by a high number of bacterial phyla. Actinobacteria were dominant at the 7.5 m site in both the number of OTUs and sequences, while Choroflexi were dominant at the 12 m site. 7.2.2 The D iffuse Venting Z one C ommunity (30 m 140 m). The samples collected along the transect from 30 m to the 140 m site grouped together based on their physical characteristics. This area was influenced by the diffuse venting of hydrothermal fluids and C O2 gas through the sediment, which resulted in low pore water pH and elevated arsenic concentrations. The temperature and salinity were

PAGE 190

173 near that of normal seawater in the surface sediments, but there was a noticeable increase in temperature with depth in the sediment profile. This area was also surrounded by higher relief from the surrounding reef, which may have protected these sites from physical disturbance from wave action. The combination of this and the low pH influenced the sediment environment resu lting in a well sorted, fine grained volcanic sand with low to absent carbonate content. The macrofaunal community at these sites was characterized by low species richness and abundances and was dominated by burrowing crustaceans such as thalassinid shrimp and the amphipod Platyischnopus sp.A The low diversity in this area may have been due to several factors. First, the elevated arsenic and low pH of the pore waters could have affected the survival of infaunal taxa. Secondly, the fine grained, wellsorted and unconsolidated sand may have provided a poor habitat for benthic infauna. And finally biological influences such as bioturbation by the high density of thalassinid shrimp and predation on settling larvae by Platyischnopus sp. A could have significantl y reduced the macrofaunal community at these sites. The meiofaunal community in this zone split into two subgroups. The meiofaunal community at the 30 m and 60 m sites had fewer taxa and lower abundances than the 90 m 140 m sites, which also had relativ ely fewer taxa but increased abundances. This trend reflects the sensitivity of the meiofaunal taxa to the pore water conditions. At all sites, nematodes were proportionally most abundant, but there was an increasing number of arthropods with distance along the transect. This may reflect the gradual increase in sediment grain size along the transect. The meiofaunal community in this zone may also have been influenced by the dominant macrofaunal taxa through physical disturbance by burrowing as well as direc t predation.

PAGE 191

174 The molecular eukaryotic community in this zone also encompassed the 300 m site and was grouped by the presence of a contig provisionally identified as the fungus Paecilomyces sp. and two Metazoan OTUs provisionally identified as the nematode Chromadorita tentabundum and the bivalve mollusk Astarte castanea, as well as a Stramenopile (Oomycetes) which closely matched Phytophthora infestans The presence of the molluscan sequence in this zone is enigmatic since there were few mollusks found in the corresponding meiofaunal samples. These sequences may represent recently settled larvae, which may not survive to the adult stage in the low pH environment. The 30 m and 60 m sites formed a subgrouping defined by the fungus Paecilomyces sp. and several Stramenopile OTUs provisionally identified as the diatoms Pauliella toeniata Navicula sp., and unidentified Bacillariophyta. Sediment samples for bacterial analysis were collected in this zone at sites 20 m, 30 m, 60 m, 90 m and 120 m. The bacterial co mmunity in this zone was characterized by a high number of phyla and unique OTUs and was dominated by Actinobacteria and Proteobacteria. 7.2.3 The Low hydrothermal A ctivity Z one C ommunity (180 m 300 m) The three transect sites farthest from the vent also grouped together based on their physical characteristics. The pore water at these sites had the same temperature and salinity as the ambient seawater. Arsenic concentrations of the pore water were slightly elevated and pH values were slightly depressed from normal seawater than at the other sites. This suggests that there was still a hydrothermal influence at these site s, though to a lesser extent than in the diffuse venting zone. The sediments were predominantly

PAGE 192

175 biogenic carbonate sand and gravel consisting of coral and shell fragments from the surrounding reef environment. The macrofaunal community had a high species richness and abundance. Dominant macrofauna included the amphipod Cyrtophium? sp. A, the bivalve Codakia sp. A and several polychaetes; Typosyllis cornuta, Pholoe sp. A, Grubeosyllis sp. B and Protodorvillea sp. B. The high diversity at these sites may re flect the coarser sediment type and less variable water quality being more favorable to benthic infauna. The meiofaunal community in this zone had a higher taxonomic richness than the other transect sites but the abundances were similar to the 90 m 140 m sites. The offshore transect sites had nearly equal proportion of nematodes and copepods. The higher abundance of copepods and of many typically interstitial phyla was due to the larger median grain size and more poorly sorted sediments, resulting in more interstitial spaces between sediment grains to support meiofaunal diversity. The molecular eukaryotic community for this zone was represented by the 180 m and 250 m sites. These two sites had the most eukaryotic OTUs and dominant eukaryotes included the f ungus Cladosporium cladosporioides (Contig[0008]), the Chlorophytes Polytoma oviforme (Contig [0218]) and Cladophora glomerata (Contig[0074]) and a marine mite, Thalassarachna basteri (Contig[0115]). Bacterial sequences were only obtained from the 180 m si te. The bacterial community here had only four phyla represented and a total of 27 OTUs. The phylum Firmicutes was dominant at that site. Unlike with the other biological communities, the less variable physical environment provides fewer niches for bacteri a, resulting in lower diversity.

PAGE 193

176 7.2.4 Danlum Bay Reference Site The Danlum Bay reference site was located near the mouth of a large stream and as a result was influenced by terrestrial runoff and detritus from the island. The sediment grain size at D anlum Bay was similar to the Tutum Bay transect site, which was the original rationale for choosing this location as a reference site for this study. Also, the sediment carbonate content was similar to the 180 m site and represented a comparable transition al sediment zone of volcanic sands and calcareous sediments. The Danlum Bay site was unique in the high percentage of organic carbon in the sediments which was due to the input of plant detritus washing in from Ambitle Island. This resulted in a reducing s ediment environment as seen by the negative ORP measurements at this site. The high sediment organics were the primary factor structuring the biological communities at this site at all levels of biodiversity. The macrofaunal community at this site was intermediate to the transect sites in terms of its species richness and abundance. The reference site community was characterized by deposit feeding polychaetes which reflects the high organic content of the sediments. The meiofaunal community had a relativel y low number of taxa and abundance and grouped with the 0 m vent site in terms of its species similarity. This site was unique in the strong dominance of copepods, which accounted for 47% of the relative abundance. This was most likely due also to the larg e amount of plant detritus in the sediment.

PAGE 194

177 The molecular eukaryotic community was dominated by fungi, which are probably associated with the breakdown of detritus found at that site. Among the most abundant OTUs were the fungi Cladosporium cladosporioide s and Penicillium chrysogenum. The eukaryotic community was most similar to the 0 m site due to the presence of common fungal sequences. This is possibly due to the proximity to shore of both of these sites and the input of detrital material from Ambitle I sland. Bacterial sequences were not obtained from this site. The effect of the high sediment organic content and negative ORP values on the benthic communities at the Danlum Bay reference site had a confounding impact on the overall data analysis of this s tudy. Reanalysis of the community similarity and environmental datasets without the reference site resulted in stronger correlations in the BIO ENV results for the Tutum Bay transect sites. This indicates that Danlum Bay did not serve as a good reference s ite for this study, despite its apparent similarity in sediment composition to Tutum Bay. The characteristics of the Danlum Bay site suggest that this site should be omitted in future analysis of the Tutum Bay datasets. 7.3 Conclusions Overall, the phy sical environment along the hydrothermal gradient at Tutum Bay correlated with the biological community structure, directly or indirectly. Different levels of the biodiversity (macrofaunal, meiofaunal, molecular) correlated with different environmental var iables. The macrofaunal and eukaryotic community structure were more strongly influenced by the sediment characteristics while the meiofauna were more closely tied to temperature and salinity.

PAGE 195

178 Despite the high arsenic concentrations in the pore water and sediments, this variable appeared to have a lesser influence on the structure of the biological communities. This may be due to the binding of the bioavailable arsenic in the sediments, the dispersion of the lower density hydrothermal effluent in the surf ace waters, or quite possibly to the tolerance and physiological adaptations of the benthic fauna. The results from the 2003 macrofaunal analysis had suggested that pH was a controlling factor in the benthic community structure. The results from the more intensive 2005 sampling however failed to uphold that and instead indicated that the macrofauna community was correlated to sediment characteristics. Most notably, the %CaCO3 was found to have the highest correlation for a single parameter. This in turn is correlated with pH and suggests that though the macrofaunal community structure is not directly affected by pH, there is an indirect effect where pH influences the sediment composition. Overall, the influence of the hydrothermal vents increased the bioco mplexity at Tutum Bay by creating new habitats which added to the overall habitat complexity of the region. This provided unique niches for certain taxa that might otherwise be outcompeted in regular reef habitats and affecting the entire community assembl age. Shallow water hydrothermal vents are unique, local habitats associated with volcanic regions. These systems also tend to be transient in nature; vents can stop flowing when there are changes in the underground flow of the source water of the hydrother mal fluids. The ephemeral nature of these systems also influences the associated benthic community, which is composed primarily of opportunistic species that can quickly colonize these habitats and are preadapted to tolerate extreme environmental conditio ns. The polychaetes in the Capitella capitata species complex are well known for their

PAGE 196

179 ability to occupy sites that have been disturbed either by natural events or anthropogenic pollution. The dominance of Capitella cf. capitata at the Tutum Bay vent refle cts this. Interestingly, the benthic community around this vent is similar to the shallow water vents found in Milos, Greece in the Aegean Sea, which were also dominated by a sibling species of Capitella capitata and had high densities of thalassinid shrim p burrows around the periphery of the hydrothermal area (Thiermann et al 1994, 1997, Gamenick et al. 1998). The constant physical disturbance from the venting and the extreme environmental conditions at the vent additionally keep the biological community at an early stage of succession. This was evident along the transect at Tutum Bay where the benthic communities changed from being dominated by a few opportunistic taxa near the vent to a highly diverse community associated with the calcareous sediments f arther away. The taxa comprising the benthic community around the vent represent a subset of species recruited from the surrounding area. This has also been noted at similar vent systems around the world (Thiermann et al 1997, Tarasov et al 1999, Melwa ni and Kim 2008). This is in contrast to deepsea hydrothermal vent systems, where selective pressures for alternative carbon sources (chemosynthesis) and tolerance for highly toxic conditions result in the evolution of endemic vent communities. At Tutum B ay, the shallow water environment allows for photosynthesis to serve as the primary carbon source, while the potentially toxic concentrations of arsenic are largely bound in the sediments or possibly dispersed in the surface waters.

PAGE 197

180 Perhaps the most inter esting finding in this study is the effect of the low pH environment on the sediment carbonates and the interrelationships between the pore water chemistry and biological community in structuring the sediment, and in turn, the influence of the sediment com position on the benthic community structure. In the immediate area around the vent and in the zone of diffuse venting as far away as 140 m from the vent, carbonate sediments were absent. This was farther influenced by the reworking of the sediment by thala ssinid shrimp. Farther from the vent, where pore water pH was closer to normal seawater, the sediments were predominantly carbonate sands and gravel derived from biological sources such as foraminiferan and molluscan shell fragments, calcareous algae and coral rubble. McCloskey (2009) observed increasing abundance and diversity in the foraminiferan assemblages in Tutum Bay with distance along the transect. Foraminiferans were nearly absent in the sediments surrounding the vent and within the area influenced by diffuse venting, presumably due to the reduced pH. These results corroborate the results seen in the macrofauna and meiofauna communities along the transect and further support the hypothesis that pH is an important factor in structuring the benthic co mmunities associated with the hydrothermal system. There has been much recent attention on the effects of increased atmospheric CO2 from anthropogenic sources on lowering ocean pH (Caldeira and Wickett 2003, Feely et al 2004,Caldeira and Wickett 2005) and the possible effects this may have on marine organisms (Anthony et al 2008, Fabry et al 2008, Dupont et al 2010, Wood et al 2010). Because of the release of CO2 from the hydrothermal fluids, shallow water vent systems offer a unique natural laboratory for looking at these effects on marine organism

PAGE 198

181 (Hall Spencer et al 2008). Several recent studies have taken advantage of shallow vent systems, including Tutum Bay, to look at the impacts of low pH on calcareous organisms such as bryozoans (RodolfoMetal pa et al 2010) and foraminiferans (Engle 2010). These systems offer even greater potential for studying the effects of ocean acidification on the entire marine community and the complex interactions among the different levels of biodiversity and the physi cal environment the overall biocomplexity of the system. The results of my study illustrate this potential, showing the link between reduced pH, changes in sediment composition and benthic community structure.

PAGE 199

182 Literature Cited Aller, R.C. 1982. Carb onate dissolution in nearshore terrigenous muds: the role of physical and biological reworking. J ournal of Geology 90: 7995. Alongi, D.M. 1986. Population structure and trophic composition of the free living nematodes inhabiting carbonate sands of Davies Reef, Great Barrier Reef, Australia. Aust ralian J ournal of Mar ine & Freshw ater Res earch 37: 609 619. Anand, M. and B.C. Tucker. 2003. Defining biocomplexity: and ecological perspective. Comments on Theoretical Biology 8: 497 510. Andreae, M.O. 1979. Arsenic speciation in seawater and interstitial waters: the influence of biological chemical interactions on the chemistry of a trace element. Limnol ogy & Oceanogr aphy 24: 440452. Anthony, K.R.N., D.I. Kline, G. Diaz Pulido, S. Dove, and O. Hoegh Guldb erg. 2008. Ocean acidification causes bleaching and productivity loss in coral reef builders. Proceedings of the National Academy of Sciences 105: 17442 17446.

PAGE 200

183 Argese, E., C. Bettiol, C. Rigo, S. Bertini, S. Colomban, and P.F. Ghetti. 2005. Distribution of arsenic compounds in Mytilus galloprovincialis of the Venice lagoon (Italy). Science of the Total Environment 348: 267 277. Atkins, M.S., A.P. Teske and O.R. Anderson. 2000. A survey of flagellate diversity at four deep sea hydrothermal vents in t he Eastern Pacific Ocean using structural and molecular approaches. Journal of Eukaryotic Microbiology 47: 400 411. Barwick, M., and W. Maher. 2003. Biotransference and biomagnification of selenium, copper, cadmium, zinc, arsenic and lead in a temperat e seagrass ecosystem from Lake Macquarie Estuary, NSW, Australia. Marine Environmental Res earch 56: 471 502. Bass, D. and T. Cavalier Smith. 2004. Phylum specific environmental DNA analysis reveals remarkably high global biodiversity of Cercozoa (Protoz oa). International Journal of Systematic and Evolutionary Microbiology 54: 2393 2404. Beh nke, A., J. Bunge, K. Barger, H W. Breiner, B. Alla, and T. Stoeck. 2006. Microeukaryote community patterns along an O2/H2S gradient in a supersulfidic anoxic fjord (Framvaren, Norway). Applied & Environmental Microbiology 72: 3626 3636. Blake, J.A. 2009. Redescription of Capitella capitata (Fabricius) from West Greenland and designation of a neotype (Polychaeta, Capitellidae). Zoosymposia 2: 55 80.

PAGE 201

184 Bhadury, P. M.C. Austen, D.T. Bilton, P.J.D. Lambshead, A.D. Rogers, G.R. Smerdon. 2006. Molecular detection of marine nematodes from environmental samples: overcoming eukaryotic interference. Aquatic Microbial Ecology 44: 97 103. Bright, M. C. Arndt, H. Keckeis, and H. Felbeck. 2003. A temperature tolerant interstitial worm with associated epibiotic bacteria from the shallow water fumaroles of Deception Island, Antarctica. Deep Sea Research Part II: Topical Studies in Oceanography 50: 1859 1871. Burdige, D.J. and R.C. Zimmerman. 2002. Impact of sea grass density on carbonate dissolution in Bahamian sediments. Limnol ogy & Oceanogr aphy 47: 17511763. Byrne, M., N.A. Soars, M.A. Ho, E. Wong, D. McElroy, P. Selvakumaraswamy, S.A. Dworjanyn, and A.R. Davis. 2010. Fertilization in a suite of coastal marine invertebrates from SE Australia is robust to near future ocean warming and acidification. Marine Biology 157: 2061 2069. Cadenasso, M.L., S.T.A. Pickett, and J.M. Grove. 2006. Dimensions of ecosystem complex ity: Heterogeneity, connectivity, and history. Ecological Complexity 3: 1 12. Caldeira, K. and M.E. Wickett. 2003. Anthropogenic carbon and ocean pH. Nature 425: 365.

PAGE 202

185 Caldeira, K. and M.E. Wickett. 2005. Ocean model predictions of chemistry changes fro m carbon dioxide emissions to the atmosphere and ocean. Journal of Geophysical Research 110, C09S04 doi: 10.1029/2004JC002671. Clarke, K.R., and M. Ainsworth. 1993. A method of linking multivariate community structure to environmental variables. Mar ine E col ogy Prog ress Ser ies 92: 205 219. Clarke, K.R., and R.N. Gorley. 2006. Primer v6: User manual/tutorial. PRIMER E: Plymouth, U.K. Cocito, S., C.N. Bianchi, C. Morri, and A. Peirano. 2000. First survey of sessile communities on subtidal rocks in an ar ea with hydrothermal vents: Milos Island, Aegean Sea. Hydrobiologia 426: 113121. Cutter, G.A., L.S. Cutter, A.M. Featherstone, and S.E. Lohrenz. 2001. Antimony and arsenic biogeochemistry in the western Atlantic Ocean. Deep Sea Res earch Part II : Topical Studies in Oceanogr aphy 48: 2895 2915. Dando, P.R., Stben, D., Varnavas, S.P. 1999. Hydrothermalism in the Mediterranean Sea. Progress in Oceanography 44: 333367. Dando, P.R., Aliani, S., Arab, H., Bianchi, C.N., Brehmer, M. Cocito, S. Fowler, S.W., Gu ndersen, J. Hooper, L.E., Klbl, R., Kuever, J., Linke, P., Makropoulos, K.C., Meloni,

PAGE 203

186 R. Miwuel, J. C., Morri, C., Mller, S., Robinson, C., Schlesner, H., Sievert, S. Sthr, D. Thomm, M., Varnavas, S.P., Ziebis, W. 2000. Hydrothermal Studies in the Aegea n Sea. Physics and Chemistry of the Earth (B) 25: 1 8. Das, S., P.S. Lyla, S.A. Khan. 2009. Filamentous fungal population and species diversity from the continental slope of Bay of Bengal, India. Acta Oecologica 35: 269 279. Dawson, S.C. and N.R. Pace. 2002. Novel kingdom level eukaryotic diversity in anoxic environments. Proceedings of the National Academy of Sciences 99: 8324 8329. Deheyn, D.D., P. Gendreau, R.J. Baldwin, and M.I. Latz. 2005. Evidence for enhanced bioavailability of trace elements in the marine ecosystem of Deception Island, a volcano in Antarctica. Marine Environmental Reseach 60: 1 33. Dhalgren, C.P., M.H. Posey, and A.W. Hulbert. 1999. The effects of bioturbation on the infaunal community adjacent to an offshore hardbottom re ef. Bull etin of Marine Sci ence 64: 2134. Dupont, S., N. Dorey, and M. Thorndyke. 2010. What meta analysis can tell us about vulnerability of marine biodiversity to ocean acidification ? Estuarine, Coastal and Shelf Science 89:182 185.

PAGE 204

187 Edgcomb, V.P., D.T. Kysela, A. Teske, A. de Vera Gomez, and M.L. Sogin. 2002. Benthic eukaryotic diversity in the Guaymas Basin hydrothermal vent environment. Proceedings of the National Academy of Sciences 99: 7658 7662. Engel, B.E. 2010 Effects of a shallow water hyd rothermal vent gradient on benthic calcifiers, Tutum Bay, Ambitle Island, Papua New Guinea. M.S. Thesis. University of South Florida, St. Petersburg, FL. Fabry, V.J., B.A. Seibel, R.A. Feely, and J.C. Orr. 2008 Impacts of ocean acidification n marine fau na and ecosystem processes. ICES Journal of Marine Science 65: 414 432. Fauchald, K and P.A. Jumars. The diet of worms: a study of polychaete feeding guilds. Oceanography and Marine Biology an Annual Review 17: 193 284. Feely, R.A., Sabine, C.L., Lee K., Berelson, W. Kleypas, J., Fabry, V.J., and Millero, F.J. 2004. Impact of anthropogenic CO2 on the CaCO3 system in the oceans. Science 304: 362366. Fenchel, T.M. and R.J. Riedl. 1970. The sulfide system: a new biotic community underneath the oxidiz ed layer of marine sand bottoms. Marine Biology 7: 255 268.Da

PAGE 205

188 Francesconi, K.A., and D. Kuehnelt. 2002. Arsenic compounds in the environment. Pp. 5194 in Environmental Chemistry of Arsenic, W.T. Frankenberger, Jr., ed. Marcel Dekker, Inc., New York. Gagliano, M., M.I. McCormick, J.A. Moore, and M. Depczynski. 2010. The basics of acidification: baseline variability of pH on Australian coral reefs. Marine Biology 157: 1849 1856. Gamenick, I., M. Abbiati, and O. Giere. 1998a Field distribution and s ulphide tolerance of Capitella capitata (Annelida: Polychaeta) around shallow water hydrothermal vents off Milos (Aegean Sea). A new sibling species? Mar ine Biol ogy 130: 447453. Gamenick, I., B. Vismann, M.K. Grieshaber, and O. Giere. 1998b. Ecophysiolog ical differentiation of Capitella capitata (Polychaeta). Sibling species from different sulfidic habitats. Mar ine Ecol ogy Prog ress Ser ies 175: 155166. Grassle, J.F. and J.P. Grassle. 1974. Opportunistic life histories and genetic systems in marine benthi c polychaetes. Journal of Marine Research 32: 253 284. Gimnez, F. and A. Marn. 1991. Los anlidos poliquetos de una solfatara submarina en el Golfo de Npoles. Anales de Biologa 17: 143 151.

PAGE 206

189 Grassle, J.P., and J.F. Grassle. 1976. Sibling species i n the marine pollution indicator Capitella (Polychaeta). Science 192: 567 569. Gray, J.S. and M. Elliott. 2009. Ecology of Marine Sediments: From Science to Management. 2nd ed. Oxford University Press. New York. Green, M.A., R.C. Aller, and J.Y. Aller. 1 993. Carbonate dissolution and temporal abundances of Foraminifera in Long Island Sound sediments. Limnol ogy & Oceanogr aphy 38: 331345. Grotti, M. C. Lagomarsino, W. Goessler, and K.A. Francesconi. 2010. Arsenic speciation in marine organisms from Antarc tic coastal environments. Environmental Chemistry 7: 207 214. HallSpencer, J.M., R. Rodolfo Metalpa, S. Martin, E. Ransome, M. Fine, S.M. Turner, S.J. Rowley, D. Tedesco, and M C. Buia. 2008. Volcanic carbon dioxide vents show ecosystem effects of ocean acidification. Nature 454: 9699. Handley, K.M., M. Hry, and J.R. Lloyd. 2009. Redox cycling of arsenic by the hydrothermal marine bacterium Marinobacter santoriniensis Environmental Microbiology 11: 1601 1611.

PAGE 207

190 Heiri, O., A.F. Lotter, and G. Lemcke 2001. Loss on ignition as a method for estimating organic and carbonate content in sediments: reproducibility and comparability of results. J ournal of Paleolimnol ogy 25: 101 110. Hendriks, I.E., C.M. Duarte, M. lvarez. 2010. Vulnerability of marine bio diversity to ocean acidification: A metaanalysis. Estuarine, Coastal, and Shelf Science 86: 157 164. Hendriks, I.E. and C.M. Duarte. 2010. Ocean acidification: Separating evidence from judgment A reply to Dupont et al Estuarine, Coastal and Shelf Sc ience 89: 186 190. Hicks, G.R.F. and B.C. Coull. 1983. The ecology of marine meiobenthic harpacticoid copepods. Oceanography and Marine Biology an Annual Review 21: 67 175. Higgins R.P. and H. Thiel (eds.). 1988. Introduction to the study of meiofau na. Smithsonian Institution Press, Washington, D.C. Inskeep, W.P., T.R. McDermott, and S. Fendorf. 2002. Arsenic (V)/(III) cycling in soils and natural waters: chemical and microbiological processes. Pp. 183 215 in Environmental Chemistry of Arsenic, W.T Frankenberger, Jr., ed. Marcel Dekker, Inc., New York.

PAGE 208

191 Jordan, M.A., D.T. Welsh, R.J.K. Dunn, and P.R. Teasdale. 2009. Influence of Trypaea australiensis density on benthic metabolism and nitrogen dynamics in sandy estuarine sediment: A mesocosm simula tion. Journal of Sea Research 61: 144 152. Kamenev, G.M., V.I. Fadeev, N.I. Selin, V.G. Tarasov, and V.V. Malakhov. 1993. Composition and distribution of macroand meiobenthos around sublittoral hydrothermal vents in the Bay of Plenty, New Zealand. N ew Z ealand J ournal or Mar ine & Freshwater Res earch 27: 407418. Kamenev, G.M., V. Ya. Kavun, V.G. Tarasov, and V.I. Fadeev. 2004. Distribution of bivalve mollusks Macoma golikovi Scarlato and Kafanov, 1998 and Macoma calcarea (Gmelin, 1791) in the shallow water hydrothermal ecosystem of Kraternaya Bight (Yankich Island, Kuril Islands): connection with feeding type and hydrothermal activity of Ushishir Volcano. Cont inental Shelf Res earch 24: 7595. Karlen, D.J. R.E. Price, T. Pichler, a nd J.R. Garey. 2010. Changes in benthic macrofauna associated with a shallow water hydrothermal vent gradient in Papua New Guinea. Pacific Science 64: 391 404. Kirby, J. and W. Maher. 2002. Tissue accumulation and distribution of arsenic compounds in th ree marine fish species: relationship to trophic position. Applied Organometallic Chemistry 16: 108 115.

PAGE 209

192 Kitts, H.J., G.E. Millward, A.W. Morris, and L. Ebdon. 1994. Arsenic biogeochemistry in the Humber Estuary, U.K. Estuarine, Coastal and Shelf Science 39: 157172. Langdon, C.J., T.G. Piearce, A.A. Meharg, and K.T. Semple. 2003. Interactions between earthworms and arsenic in the soil environment: a review. Environmental Pollution 124: 361 373. Le Calvez, T. G. Burgaud, S. Mah, G. Barbier, and P. Vandenkoornhuyse. 2009. Fungal diversity in deep sea hydrothermal ecosystems. Applied and Environmental Microbiology 75: 6415 6421. Lohrer, A.M., S.F. Thrush, and M.M. Gibbs. 2004. Bioturbators enhance ecosystem function through complex biogeochemical i nteractions. Nature 431: 1092 1095. Lpez Garca, P. F. Rodrguez Balera, C. Pedrs All, and D. Morelra. 2001. Unexpected diversity of small eukaryotes in deep sea Antarctic plankton. Nature 409: 603 607. Lpez Garca, P., H. Philippe, F. Gail, and D. Moreira. 2003. Autochthonous eukaryotic diversity in hydrothermal sediment and experimental microcolonizers at the Mid Atlantic Ridge. Proceedings of the National Academy of Sciences 100: 697 702.

PAGE 210

193 Lpez Garca, P., A. Vereshchaka, and D. Moreira. 200 7. Eukaryotic diversity associated with carbonates and fluid seawater interface in Lost City hydrothermal field. Environmental Microbiology 9: 546 554. Man cinelli, G., S. Fazi, and L. Rossi. 1998. Sediment structural properties mediating dominant feedin g types patterns in soft bottom macrobenthos of the Northern Adriatic Sea. Hydrobiologia 367: 211 222. Mannini, E., G.M. Luna, C. Corinaldesi, D. Zeppilli, G. Bortoluzzi, G. Caramanna, F. Raffa, and R. Danovaro. 2008. Prokaryote diversity and virus abun dance in shallow hydrothermal vents of the Mediterranean Sea (Panarea Island) and the Pacific Ocean (North Sulawesi Indonesia). Microb ial Ecol ogy 55: 626 639. McCloskey, B.J. 2009. Foraminiferal responses to arsenic in a shallow water hydrothermal syste m in Papua New Guinea and in the laboratory. Ph.D. Dissertation. University of South Florida, St. Petersburg, FL. Melwani, A.R. and S.L. Kim. 2008. Benthic infaunal distributions in shallow hydrothermal vent sediments. Acta Oecol ogia 33: 162175. Mermil lod Blondin, F. and R. Rosenberg. 2006. Ecosystem engineering: the impact of bioturbation on biogeochemical processes in marine and freshwater benthic habitats. Aquatic Sciences 68: 434 442.

PAGE 211

194 Michener, W.K., T.J. Baerwald, P. Firth, M.A. Palmer, J.L. Ros enberger, E.A. Sandlin, and H. Zimmerman. 2001. Defining and unraveling biocomplexity. BioScience 51: 1018 1023. Milliman. J.D., 1974. Marine Carbonates. SpringerVerlag, Berlin Heidelberg New York. Moon van der Staay, S.Y., R. De Wachter, and D. Vaul ot. 2001. Oceanic 18S rDNA sequences from picoplan kton reveal unsuspected eukaryotic diversity. Nature 409: 607 610. Morri, C., C.N. Bianchi, S. Cocito, A. Peirano, A.M. De Biase, S. Aliani, M. Pansini, M. Boyer, F. Ferdeghini, M. Pestarino, P. Dando.19 99. Biodiversity of marine sessile epifauna at an Aegean island subject to hydrothermal activity: Milos, eastern Mediterranean Sea. Mar ine Biol ogy 135: 729739. Oremland, R.S., and J.F. Stolz. 2003. The ecology of arsenic. Science 300: 939944. Oremland, R.S., T.R. Kulp, J. Switzer Blum, S.E. Hoeft S. Baesman, L.G. Miller, and J.F. Stolz. 2005. A microbial arsenic cycle in a salt saturated extreme environment. Science 308: 13051308.

PAGE 212

195 Paarlberg, A.J., M.A.F. Knaapen, M.B. de Vries, S.J.M.H. Hulscher a nd Z.B. Wang. 2005. Biological influences on morphology and bed composition of an intertidal flat. Estuarine, Coastal and Shelf Science 64: 577 590. Pelejero, C., E. Calvo, M.T. McCulloch, J.F. Marshall, M.K. Gagan, J.M. Lough, and B.N. Opdyke. 2005. Pr eindustrial to modern interdecadal variability in coral reef pH. Science 309: 2204 2207. Percival, J.B., and P.J. Lindsay. 1997. Measurement of physical properties of sediments. Pp.745 in Manual of Physico Chemical Analysis of Aquatic Sediments, A. Mudroch, J.M. Azcue, and P. Mudroch, eds, CRC Press, Boca Raton. Pichler, T., and G.R. Dix. 1996. Hydrothermal venting within a coral reef ecosystem, Ambitle Island, Papua New Guinea. Geology 20: 435438. Pichler, T., and J. Veizer. 1999. Precipitatio n of Fe(III) oxyhydroxide deposits from shallow water hydrothermal fluids in Tutum Bay, Ambitle Island, Papua New Guinea. Chem ical Geol ogy 162: 1531. Pichler, T., J. Veizer, and G.E.M. Hall. 1999a. Natural input of arsenic into a coral reef ecosystem by hydrothermal fluids and its removal by Fe(III) oxyhydroxides. Environmental Sci ence & Technol ogy 33: 13731378.

PAGE 213

196 Pichler, T., J. Veizer, and G.E.M. Hall. 1999b. The chemical composition of shallow water hydrothermal fluids in Tutum Bay, Ambitle Island, Papua New Guinea and their effect on ambient seawater. Mar ine Chem istry 64: 229252. Pichler, T., J. P. Amend, J. Garey, P. Hallock, N. P. Hsia, D. J. Karlen, D. R. Meyer Dombard, B. J. McCloskey, and R. E. Price (2006), A Natural Laboratory to Study Arse nic Geobiocomplexity Eos Trans actions, AGU 87(23), doi:10.1029/2006EO230002. Powell, E.N. and T.J. Bright. 1981. A thiobios does exist Gnathostomulid domination of the canyon community at the East Flower Garden brine seep. Internationale Revue der ge samten Hydrobiologie und Hydrographie 66: 675 683. Powell, E.N., T.J. Bright, A. Woods, and S. Gittings. 1983. Meiofauna and the thiobios in the East Flower Garden brine seep. Marine Biology 73: 269 283. Price. R.E. 2008. Biogeochemical cycling of a rsenic in the marine shallow water hydrothermal system of Tutum Bay, Ambitle Island, Papua New Guinea. Ph.D. Dissertation. University of South Florida, Tampa, FL. Price, R.E., and T. Pichler. 2005. Distribution, speciation and bioavailability of arsenic i n a shallow water submarine hydrothermal system, Tutum Bay, Ambitle Island, PNG. Chem ical Geol ogy 224: 122135.

PAGE 214

197 Price, R.E., Amend, J.P. and Pichler, T. 2007. Enhanced geochemical gradients in a marine shallow water hydrothermal system: unusual arsenic speciation in horizontal and vertical pore water profiles. Applied Geochemistry 22: 25952605. Potvin, M. and C. Lovejoy. 2009. PCR based diversity estimates of artificial and environmental 18S rRNA gene libraries. Journal of Eukaryotic Microbiology 56: 174 181. Raffaelli, D.G. and C.F. Mason. 1981. Pollution monitoring with meiofauna, using the ratio of nematodes to copepods. Marine Pollution Bulletin 12: 158 163. Riemann, F. 1988. Nematoda. Chapter 23 pp. 293 299 in Introduction to the Study of Me iofauna Higgins, R.P. and H. Thiel (eds). Smithsonian Institution Press Washington, D.C. RodolfoMetalpa, R., C. Lombardi, S. Cocito, J.M. Hall Spencer, and M.C. Gambi. 2010. Effects of ocean acidification and high temperatures on the bryozoans Myriapor a truncate at natural CO2 vents. Marine Ecology 31: 447 456. Rolf F.J., and R.R.Sokal. 1981. Statistical Tables. Freeman and Company, New York.

PAGE 215

198 Samson, R.A. 1974. Paecilomyces and som e allied Hyphomycetes Studies in Mycology 6: 1 119 Sanders, H. L., 1958. Benthic studies in Buzzards Bay. I. Animal sediment relationships. Limnol ogy & Oceanogr aphy 3: 245258. Shirayama,Y. and H. Thornton. 2005. Effect of increased atmospheric CO2 on shallow water marine benthos. Journal of Geophysical Research 110 C09S08, doi: 10.1029/2004JC002618. Sievert, S.M., T. Brinkhoff, G. Muyzer, W. Ziebis, and J. Kuever. 1999. Spatial heterogeneity of bacterial populations along an environmental gradient at a shallow submarine hydrothermal vent near Milos Island (Greece). Applied and Environmental Microbiology 65: 3834 3842. Sievert, S.M., W. Ziebis, J. Kuever, and K. Sahm. 2000. Relative abundance of Archaea and Bacteria along a thermal gradient of a shallow water hydrothermal vent quantified by rRNA slot blot hybridiz ation. Microbiology 146: 1287 1293. Snelgrove, P.V.R., and C.A. Butman. 1994. Animal sediment relationships revisited cause versus effect. Oceanogr aphy & Mar ine Biol ogy 32: 111177.

PAGE 216

199 Snelgrove, P.V.R., H.T. Blackburn, P.A. Hutchings, D.M. Alongi, J.F. Grassle, H. Hummel, G. King, I. Koike, P.J.D. Lambshead, N.B. Ramsing, and V. Solis Weiss. 1997. The importance of marine sediment biodiversity in ecosystem processes. Ambio 26: 578 583. Sorokin, Yu.I., P. Yu. Sorokin, and O. Yu. Zakouskina. 2003. Micr oplankton and its function in a zone of shallow hydrothermal activity: the Craternaya Bay, Kurile Islands. Journal of Plankton Res earch 25: 495506. Stoeck, T. and S. Epstein. 2003. Novel eukaryotic lineages inferred from small subunit rRNA analyses of ox ygendepleted marine environments. Applied and Environmental Microbiology 69: 2657 2663. Takishita, K., H. Miyake, M. Kawato, and T. Maruyama. 2005. Genetic diversity of microbial eukaryotes in anoxic sediment around fumaroles on a submarine caldera flo or based on the small subunit rDNA phylogeny. Extremophiles 9: 185 196. Takishita K., M. Tsuchiya, M. Kawato, K. Oguri, H. Kitazato, and T. Maruyama. 2007a. Genetic diversity of microbial eukaryotes in anoxic sediment of the saline meromictic lake Namak o ike (Japan): on the detection of anaerobic or anoxictolerant lineages of eukaryotes. Protist 158: 51 64.

PAGE 217

200 Takishita, K., N. Yubuki, N. Kakizoe, Y. Inagaki, and T. Maruyama. 2007b. Diversity of microbial eukaryotes in sediment at a deep sea methane cold seep: surveys of ribosomal DNA libraries from raw sediment samples and two enrichment cultures. Extremophiles 11: 563 576. Takishita, K., N. Kakizoe, T. Yoshida, and T. Maruyama. 2010. Molecular evidence that phylogenetically diverged ciliates are acti ve in microbial mats of deep sea cold seep sediments. Journal of Eukaryotic Microbiology 57: 76 86. Tarasov, V.G. 2006 Effects of shallow water hydrothermal venting on biological communities of coastal marine ecosystems of the western Pacific. Advance s in Marine Biology 50: 267 421. Tarasov, V.G., M.V. Propp, L.N. Propp, A.V. Zhirmunsky, B.B. Namsaraev, V.M. Gorlenko, and D.A. Starynin. 1990. Shallow water gasohydrothermal vents of Ushishir volcano and the ecosystem of Kraternaya Bight (the Kurile Is lands). Mar ine Ecology 11: 123. Tarasov, V.G., A.V. Gebruk, V.M. Shulkin, G.M. Kamenev, V.I. Fadeev, V.N. Kosmynin, V.V. Malakhov, D.A. Starynin, and A.I. Obxhirov. 1999. Effect of shallow water hydrothermal venting on the biota of Matupi Harbour (Rabaul Caldera, New Britain Island, Papua New Guinea). Continental Shelf Res earch 19: 79116.

PAGE 218

201 Thiermann, F., R. Windoffer, O. Giere. 1994. Selected meiofauna around shallow water hydrothermal vents off Milos (Greece): ecological and ultrastructural aspects. Vie Milieu 44: 215 226. Thiermann, F., I. Akoumianaki, J.A. Hughes, and O. Giere. 1997. Benthic fauna of a shallow water gaseohydrothermal vent area in the Aegean Sea (Milos, Greece). Marine Biol ogy 128: 149159. Thiermann, F., B. Vismann, O. Giere. 2000. Sulphide tolerance of the marine nematode Oncholaimus campylocercoides a result of internal sulphur formation? Marine Ecol ogy Prog ress Ser ies 193: 251 259. Vanreusel, A., I. Van den Bossche, F. Thiermann. 1997. Free living marine nematodes from hydroth ermal sediments: similarities with communities from diverse reduced habitats Marine Ecology Progress Series 157: 207 219. Varnavas, S.P. and D.S. Cronan. 1988. Arsenic, antimony and bismuth in sediments and waters from the Santorini hydrothermal field, Greece. Chemical Geology 67: 295 305. Watt, C., and X.E. Le. 2003. Arsenic speciation in natural waters. Pp. 11 32 in Biogeochemistry of Environmentally Important Trace Elements: ACS symposium series 835. Y. Cai, and O.C. Braids, eds. American Chemical Society, Washington D.C.

PAGE 219

202 Wenzhfer F. O. Holby, R.N. Glud, H.K. Nielsen, and J.K. Gundersen. 2000. In situ microsensor studies of a shallow water hydrothermal vent at Milos, Greece. Marine Chemistry 69: 43 54. Widdicombe, S., M.C. Austen, M.A. Kenda ll, R.M. Warwick, and M.B. Jones. 2000. Bioturbation as a mechanism for setting and maintaining levels of diversity in subtidal macrobenthic communities. Hydrobiologia 440: 369 377. Wood, H.L., J.I. Spicer, D.M. Lowe, and S. Widdicombe. 2010. Interactio n of ocean acidification and temperature; the high cost of survival in the brittlestar Ophiura ophiura. Marine Biology 157: 2001 2013. Wu, T. E. Ayres, G. Li, R.D. Bardgett, D.H. Wall, and J.R. Garey. 2009. Soil Biology and Biochemistry 41: 849 857. Zeppilli, D. and R. Danovaro. 2009. Meiofaunal diversity and assemblage structure in a shallow water hydrothermal vent in the Pacific Ocean. Aquat ic Biol ogy 5: 75 84.

PAGE 220

About the Author David Karlen was born on September 11, 1968 in Wausau, Wisconsi n and grew up in Ohio. He graduated from Lancaster High School in Lancaster, Ohio in 1987 and moved to Melbourne, Florida that fall to attend the Florida Institute of Technology. David completed his B.S. in Biological Oceanography in the spring of 1991 and continued on as a graduate student at F.I.T. studying benthic ecology under Dr. Walter G. Nelson. David co mpleted his M.S. in Biological Oceanography in the fall of 1993. David worked for the Caribbean Marine Research Center during the winter of 1994, w orking at the CMRC field research facility on Lee Stocking Island, Bahamas and at the CMRC lab in Vero Beach, Florida. He later took a job at Mote Marine Lab in the spring of 1994, working in the MML Benthic Ecology department as an invertebrate taxonomist David accepted a position with the Environmental Protection Commission of Hillsborough County in Tampa Florida in October 1994, where he is currently employed as a Manager in the Water Management Division, overseeing the benthic monitoring program for T ampa Bay.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004736
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Karlen, David.
0 245
The biocomplexity of benthic communities associated with a shallow-water hydrothermal system in papua new guinea
h [electronic resource] /
by David Karlen.
260
[Tampa, Fla] :
b University of South Florida,
2010.
500
Title from PDF of title page.
Document formatted into pages; contains X pages.
502
Dissertation (PHD)--University of South Florida, 2010.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: Shallow-water hydrothermal vents occur world-wide in regions of volcanic activity. The vents located at Tutum Bay, Ambitle Island, Papua New Guinea are unique in that the vent fluids and surrounding sediments contain some of the highest concentrations of arsenic in a natural system. This study addresses the effects of the vent system on the benthic communities, focusing on the eukaryotes, macrofauna, meiofauna and bacteria. Samples were collected in November 2003 and May/June 2005. Analysis of the 2003 macrofaunal samples indicated that pH, rather than arsenic was influencing the benthic community, and that the hydrothermal influence occurred at a greater distance than expected. Results of more intensive sampling carried out in 2005 are the primary focus of this dissertation. The pore water and sediment characteristics revealed distinct physical habitats corresponding with distance from the vent. There was a trend of decreasing temperature and arsenic concentration and increasing salinity and pH with distance from the vent. The vent sediment was poorly sorted volcanic gravel, while sediments along the transect showed a gradient from fine, well sorted volcanic sands to coarser carbonate sands farther away. The macrofauna showed a trend of increasing diversity with distance from the vent and similar taxa were present in both the 2003 and 2005 samples. The vent community was dominated by the polychaete Capitella cf. capitata. The inner transect from 30 m to 140 m had low diversity. Dominant taxa included thalassinid shrimp and the amphipod Platyischnopus sp.A. The 180 m to 300 m sites had significantly higher diversity. The Danlum Bay reference site had relatively higher diversity than the nearshore transect sites and was dominated by deposit feeding polychaetes. Macrofaunal community structure was influenced by the sediment characteristics, notably by CaCO3 content, sorting and median grain size. The meiofaunal community also showed changes with distance from the vent. Chromadorid nematodes were dominant at the vent site and were a major component of the meiofauna at most sites, along with copepods. The meiofaunal community at the reference site showed greater similarity to the vent community and both sites had low abundances. Nematodes were more abundant than copepods near the vent, but copepods were more abundant farther offshore and at the reference site. Meiofaunal community structure was influenced primarily by the pore water temperature and salinity. Biological interactions with the macrofaunal community through physical disturbance and predation may also influence the meiofaunal community. The molecular analysis of eukaryotic and bacterial diversity also revealed changes with distance from the vent. The 0 m and reference sites grouped together due to the presence of fungal sequences and the 140 m and 300 m sites grouped together due to a common molluscan sequence. Metazoans and fungi dominated the eukaryote sequences. The most abundant eukaryotic OTUs included fungi matching Paecilomyces sp. and Cladosporium cladosporioides and metazoans matching Viscosia viscosa (Nematoda) and Astarte castanea (Mollusca). The eukaryotic community structure correlated with sediment sorting, pore water arsenic and median grain size. The bacterial community was represented by 24 phyla and was dominated by Actinobacteria and γ-Proteobacteria. More bacterial phyla were present near the vent, while more overall OTUs were found at the intermediate sites along the transect. The most distant site had much lower diversity dominated by Firmicutes. The macrofaunal community had the strongest correlation with environmental variables. Comparison between the meiofauna and the metazoan sequences showed the proportion of nematodes found in both datasets were comparable, but the meiofauna analysis found a higher proportion of arthropods, while the molecular results were disproportionally high for platyhelminthes. Overall, the vents increased the complexity of the system by creating unique habitats. The extreme environment created by the hydrothermal activity maintained the surrounding habitat at an early successional stage colonized by a few opportunistic species. There was a gradation in the benthic communities away from the vent towards a more carbonate based climax community. The low pH environment had an effect on the sediment composition, which in turn influenced the benthic community. These findings can serve as a model for studying the potential effects of ocean acidification and climate change on benthic communities and marine biocomplexity.
590
Advisor: James R. Garey, Ph.D.
653
Biocomplexity
Shallow-water hydrothermal vents
Environmental gradients
Benthic infaunal communities
Eukaryotic diversity
690
Dissertations, Academic
z USF
x Biology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4736