Citation
Assessment of the Milletseed Butterflyfish, Chaetodon Miliaris, as a Model Species for Marine Ornamental Aquaculture and an Evaluation of Early Culture Parameters

Material Information

Title:
Assessment of the Milletseed Butterflyfish, Chaetodon Miliaris, as a Model Species for Marine Ornamental Aquaculture and an Evaluation of Early Culture Parameters
Creator:
Degidio, Jon-Michael L
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (193 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Fisheries and Aquatic Sciences
Forest Resources and Conservation
Committee Chair:
YANONG,ROY P
Committee Co-Chair:
OHS,CORTNEY L
Committee Members:
WATSON,CRAIG A
Graduation Date:
12/19/2014

Subjects

Subjects / Keywords:
Aquaculture ( jstor )
Diameters ( jstor )
Eggs ( jstor )
Embryonic structures ( jstor )
Exchange rates ( jstor )
Female animals ( jstor )
Fish ( jstor )
Larvae ( jstor )
Nauplii ( jstor )
Photoperiod ( jstor )
Forest Resources and Conservation -- Dissertations, Academic -- UF
aquaculture -- butterflyfish -- larvae
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Fisheries and Aquatic Sciences thesis, M.S.

Notes

Abstract:
As part of the Rising Tide Conservation effort to advance marine ornamental aquaculture, the milletseed butterfly was chosen as a candidate species for the development of butterflyfish aquaculture methodologies. Spawning of Chaetodon miliaris was not inhibited in broodstock tanks and resembled that of previously documented chaetodontid species. An assessment of practical fecundity established that a population of 3 males and 8 females produced the largest number of viable eggs and that relationships between photoperiod and spawning frequency were present in all three broodstock populations. Larvae hatched after approximately 28 hours, were capable of feeding at 4 dph, and survived to 44 dph in preliminary culture trials. An examination of the first feeding requirements of C. miliaris established that the calanoid copepod Parvocalanus crassirostris at densities between 1-15 nauplii ml-1, algal induced turbidity greater than ~190,000 cells ml-1, water exchange rates below 300% day-1, and stocking densities of 15-20 larvae l-1 enhanced the proportion of larvae feeding and larval feeding intensity, while tank size and light intensity did not influence first feeding. Growth and survival trials indicated that a photoperiod of 18 L:6 D, an algal cell density of 500,000-530,000 cells ml-1 and a stocking density of 15 larvae l-1, and a water exchange rate of 100% day-1 and prey density of 10 nauplii ml-1 were optimal culture conditions for C. miliaris larvae to 10 dph. Further investigation into the culture requirements of C. miliaris will supply a crucial base for developing aquaculture protocols for other pelagic spawning marine ornamental species. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (M.S.)--University of Florida, 2014.
Local:
Adviser: YANONG,ROY P.
Local:
Co-adviser: OHS,CORTNEY L.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2015-06-30
Statement of Responsibility:
by Jon-Michael L Degidio.

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright Degidio, Jon-Michael L. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
6/30/2015
Resource Identifier:
974373063 ( OCLC )
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

ASSESSMENT OF THE MILLETSEED BUTTERFLYFISH, Chaetodon miliaris , AS A MODEL SPECIES FOR MARINE ORNAMENTAL AQUACULTURE AND AN EVALUATION OF EARLY CULTURE PARAMETERS By JON MICHAEL DEGIDIO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2014

PAGE 2

© 2014 Jon Michael Degidio

PAGE 3

To all those who have supported my journey

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my mother, Deborah Barone, and father, Michael Degidio, for their unconditional love and support. Their teachings have guided me to be the man I am today. I would like to thank Paige Miller for her patience, support, love, and reminding me that life is can always be fun. Also, thank you to my sister, Nicole Barone, and her three daughters for their love and support. I would like to acknowledge my major professor, Roy Yanong, for his experience, guidance, and confidence in my abilities to perform my duties. I would also like to acknowledge my committee members, Cortney Ohs and Craig Watson, for their guidance and expertise in the aquaculture industry. Thank you to Eric Cassiano and Kevin Barden for their advice, leadership, knowledge, and friendship when I was in need. Additionally, I would like to thank all staff and students at University of Tropical Aquaculture Lab for their support, knowledge, guidance, and willingness to help whenever it was needed. I would like to thank Jane Davis, Eric Curtis, and Marjorie Awai for providing milletseed butterflyfish broodstock for these trials. Lastly, I would like to thank Judy St. Le ger, Rising Tide Conservation, and the Sea World and Busch Gardens Conservation Fund for providing the funding for my schooling and research.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 15 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 17 Marine Ornamental Trade ................................ ................................ ....................... 17 Marine Ornamental Aquaculture ................................ ................................ ............. 20 Reproductive Metabolics ................................ ................................ .................. 21 Envir onment ................................ ................................ ................................ ..... 24 Gametogenesis and Embryology ................................ ................................ ...... 26 Larval Development and Culture ................................ ................................ ...... 26 Chaetodontidae ................................ ................................ ................................ 29 Objectives ................................ ................................ ................................ ............... 30 2 SPAWNING, EMBRYOLOGY, AND LARVAL DEVELOPMENT ............................. 33 Foreword ................................ ................................ ................................ ................. 33 Methods ................................ ................................ ................................ .................. 36 Acquisition of Broodstock and Broodstock Management ................................ .. 36 Spawning Behavior, Conditions, and Fecundity ................................ ................ 38 Hydrated Egg Characteristics and Larvae Evaluation ................................ ....... 40 Water Quality ................................ ................................ ................................ .... 40 Statistical Methods ................................ ................................ ........................... 41 Embryology ................................ ................................ ................................ ....... 42 Larval Development ................................ ................................ .......................... 42 Results ................................ ................................ ................................ .................... 44 Spawning Behavior ................................ ................................ ........................... 44 Timeline ................................ ................................ ................................ ..... 44 Spawning Conditions ................................ ................................ ........................ 46 Fecundity ................................ ................................ ................................ .......... 47 Hydrated Egg Characteristics and Larvae Evaluation ................................ ....... 48 Embryology ................................ ................................ ................................ ....... 49 Larval Development ................................ ................................ .......................... 49 Water Quality ................................ ................................ ................................ .... 53 Discussion ................................ ................................ ................................ ............... 53

PAGE 6

6 Spawning Behavior ................................ ................................ ........................... 54 Spawning Conditions ................................ ................................ ........................ 55 Fecundity ................................ ................................ ................................ .......... 56 Hydrated Egg Characteristics and Larvae Evaluation ................................ ....... 58 Embryology ................................ ................................ ................................ ....... 59 Larval Development ................................ ................................ .......................... 59 3 FIRST FEEDING PARAMETERS OF Chaetodon miliaris ................................ ....... 74 Foreword ................................ ................................ ................................ ................. 74 Methods ................................ ................................ ................................ .................. 76 Embryo Stocking and Culture ................................ ................................ ........... 77 Assessment of Feeding Response ................................ ................................ ... 78 Feeding Experiments ................................ ................................ ........................ 79 3 1 1 Prey selectivity ................................ ................................ .................. 79 3 2 and 3 3 Algal cell density ................................ ................................ ..... 79 3 4 4 Prey density ................................ ................................ ...................... 80 3 5 Larval stocking density ................................ ................................ ......... 80 3 6 Tank size ................................ ................................ .............................. 81 3 7 Water exchange rates ................................ ................................ .......... 81 3 8 Light intensity ................................ ................................ ....................... 82 Statistical Methods ................................ ................................ ........................... 82 R esults ................................ ................................ ................................ .................... 83 Water Quality ................................ ................................ ................................ .... 83 Larval Measurements ................................ ................................ ....................... 83 3 2 2 Algal cell density ................................ ................................ ............... 84 3 3 Lower limit algal cell density ................................ ................................ . 84 3 4 4 Prey density ................................ ................................ ...................... 85 3 5 Larval stocking density ................................ ................................ ......... 86 3 6 Tank size ................................ ................................ .............................. 86 3 7 Water exchange rate ................................ ................................ ............ 87 3 8 Light intensity ................................ ................................ ....................... 88 Discussion ................................ ................................ ................................ ............... 88 First Feeding Response ................................ ................................ .................... 88 3 1 1 Prey selectivity ................................ ................................ .................. 88 3 2 2 Algal cell density ................................ ................................ ............... 89 3 3 Lower limits algal cell density ................................ ............................... 90 3 4 4 Prey density ................................ ................................ ...................... 91 3 5 Larval stocking density ................................ ................................ ......... 92 3 6 Tank size ................................ ................................ .............................. 93 3 7 Water exchange rate ................................ ................................ ............ 94 3 8 Light intensity ................................ ................................ ....................... 95 4 GROWTH AND SURVIVAL OF Chaetodon miliaris TO 10 DAYS POST HATCH 116 Foreword ................................ ................................ ................................ ............... 116 Methods ................................ ................................ ................................ ................ 118

PAGE 7

7 Embryo Stocking and Culture ................................ ................................ ......... 119 Assessment of Larval Size, Growth, Condition, and Survival .......................... 120 Larval Size, Growth, Condition, and Survival Experiments ............................. 121 4 1 1 Algal cel l density and stocking density ................................ ............ 121 4 2 Water exchange rates and prey density ................................ ............. 122 4 3 3 Photoperiod ................................ ................................ ..................... 123 Statistical Methods ................................ ................................ ......................... 124 Water Quality ................................ ................................ ................................ .. 125 Results ................................ ................................ ................................ .................. 125 4 1 1 Algal cell density and stocking density ................................ ............ 125 4 2 Water exchange rates and prey density ................................ ............. 128 4 3 3 Photoperiod ................................ ................................ ..................... 13 1 Discussion ................................ ................................ ................................ ............. 133 Larval Size, Growth, Condition, and Survival ................................ .................. 133 4 1 1 Algal cell density and stocking density ................................ ............ 134 4 2 Water exchange rates and prey density ................................ ............. 136 4 3 3 Photoperiod ................................ ................................ ..................... 138 5 CONCLUSION ................................ ................................ ................................ ...... 156 APPENDIX FATTY ACID ANALYSIS AND GUT HISTOLOGY OF Chaetodon miliaris 161 Larval Fatty Acid Composition ................................ ................................ .............. 161 Methods ................................ ................................ ................................ .......... 161 Results ................................ ................................ ................................ ........... 162 Ontogeny of Visual and Digestive Systems ................................ ........................... 162 Methods ................................ ................................ ................................ .......... 162 Results ................................ ................................ ................................ ........... 163 LIST OF REFERENCES ................................ ................................ ............................. 177 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 193

PAGE 8

8 LIST OF TABLES Table page 2 1 Effects of lunar phase and photoperiod on the likelihood of a spawn .................. 63 2 2 Practical fecundity measures of C. miliaris populations. ................................ ..... 63 2 3 Spearman rank correlation coefficients for all compared morphological characteristics ................................ ................................ ................................ ..... 64 2 4 Water quality variables measured within the culture trial. ................................ ... 65 2 5 Comparison of eggs and larvae of chaetodontids ................................ ............... 65 3 1 Culture parameters for C. miliaris during all experiments. ................................ ... 97 3 2 Optimized first feeding parameters for the milletseed butterflyfish C . miliaris ...... 98 4 1 Culture parameters for Experiments 4 1, 4 2, and 4 3 ................................ ..... 141 4 2 Growth rate of notochord length (SL), body depth (BD), and eye diameter (Ed) for Experiment 4 1. ................................ ................................ ................... 142 4 3 Condition index of Experiment 4 1. ................................ ................................ ... 143 4 4 Growth rate of notochord length (SL), body depth (BD), and eye diameter (Ed) for Experiment 4 2 ................................ ................................ .................... 144 4 5 Condition index for Experiment 4 2. ................................ ................................ .. 145 4 6 Growth rate of notochord length (SL), body depth (BD), and eye diameter (Ed) for Experiment 4 3. ................................ ................................ ................... 145 4 7 Condition index for Experiment 4 3 ................................ ................................ ... 146 A 1 Description of histologically observed ontogeny of C. miliaris larvae throughout development. ................................ ................................ .................. 164

PAGE 9

9 LIST OF FIGURES Figure page 2 1 Temperature, salinity, and pH throughout the spawning period .......................... 66 2 2 A diagrammatic view of the spawning behaviors of C. miliaris in the broodstock tank ................................ ................................ ................................ .. 67 2 3 Lunar phases and photoperiods overlapped with spawning data for ................... 68 2 4 Embryonic development of C. miliaris eggs at 25.5 °C. ................................ ...... 69 2 5 Larval development of Chaetodon miliaris ................................ .......................... 70 2 6 Thiolichthys stages of Chaetodontidae ................................ ............................... 72 2 7 Larval growth of C. miliaris ................................ ................................ .................. 73 3 1 First feeding responses of C. miliaris larvae with and without the addition of algae and different prey items. ................................ ................................ ............ 99 3 2 First feeding responses of C. miliaris larvae with and without the addition of algae and different prey items. ................................ ................................ .......... 100 3 3 First feeding responses of C. miliaris larvae in different algal cell densities. ..... 101 3 4 First feeding responses of C. miliaris larvae in different algal cell densities. ..... 102 3 5 First feeding responses of C. miliaris larvae in different algal cell densities. ..... 103 3 6 First feeding responses of C. miliaris larvae in different algal cell densities ...... 104 3 7 First feeding responses of C. miliaris larvae at different prey densities. ............ 105 3 8 First feeding responses of C. miliaris larvae at different prey densities ............. 106 3 9 First fe eding responses of C. miliaris larvae at different larval stocking densities ................................ ................................ ................................ ........... 107 3 10 First feeding responses of C. miliaris larvae at different larval stocking densities ................................ ................................ ................................ ........... 108 3 11 First feeding responses of C. miliaris larvae in different tank sizes ................... 109 3 12 First feeding responses of C. miliaris larvae in different tank sizes ................... 110 3 13 First feeding responses of C. miliaris larvae in different water exchange rates 111

PAGE 10

10 3 14 First feeding responses of C. miliaris larvae in different water exchange rates 112 3 15 First feeding responses of C. miliaris larvae under different light intensities. .... 113 3 16 First feeding responses of C. miliaris larvae under different light intensities ..... 114 3 17 Parvocalanus crassirostris nauplii exoskeleton present in the gut of a C. miliaris larva compressed between two microscope slides. .............................. 115 4 1 Survival (%) of C. miliaris larvae in Experiment 4 1 ................................ .......... 147 4 2 Measurements of (A) notochord length (SL), (B) body depth (BD), and (C) eye diameter (Ed) in Experiment 4 1 ................................ ................................ 148 4 3 Survival (%) of C. miliaris larvae Experiment 4 2 ................................ ............. 150 4 4 Measurements of (A) notochord length (SL), (B) body depth (BD), and (C) eye diameter (Ed) in Experiment 4 2 ................................ ................................ 151 4 5 Survival (%) of C. miliaris larvae in Experiment 4 3 ................................ .......... 153 4 6 Measurements of (A) notochord length (SL), (B) body depth (BD), and (C) eye diameter (Ed) in Experiment 4 3 ................................ ................................ 154 A 1 Total fatty acid composition of C. miliaris larvae at 0, 5, 11, 13 and 17 dph ...... 169 A 2 Ratios of DHA, EPA, and ARA in C. miliaris larvae at 0, 5, 11, 13, and 17 dph 169 A 3 Histological analysis of 0 dph C. miliaris larvae ................................ ................ 170 A 4 Histological analysis of 7 dph C. miliaris larvae ................................ ................ 170 A 5 Histological analysis of 7 dph C. miliaris larvae ................................ ................ 171 A 6 Histological analysis of 19 dph C. miliaris larvae ................................ .............. 171 A 7 Histological analysis of 26 dph C. miliaris larvae ................................ .............. 172 A 8 Histological analysis of 32 dph C. miliaris larvae ................................ .............. 173 A 9 Histological analysis of 43 dph C. miliaris larvae ................................ .............. 174 A 10 Eye development of C. miliaris at 0, 2, 4, 7, and 43 dph ................................ ... 175

PAGE 11

11 LIST OF ABBREVIATIONS % hatch hatch success 3L 3 dph notochord length AI Anterior intestine ARA arachidonic acid B brain BD body depth Bpx Buccopharynx CI condition index cm centimeter D dark D.O. dissolved oxygen DHA docosahexaenoic acid dph days post hatch E eye Ed eye diameter ED egg diameter EPA eicosapentaenoic acid F female g gram G gill filaments Ga gill anlage Gl ganglion cell layer h hour H heart

PAGE 12

12 hph hours post hatch HUFA highly unsaturated fatty acid IN inner nuclear layer IP inner plexiform layer K kidney L liter L light L liver Ln lens lx lux m meter M male MAD median absolute deviation mg micrograms min minute ml milliliter mm millimeter n sample size Nl nerve cell layer NHL new hatched notochord length NO 2 N nitrate nitrogen NO 3 nitrate OD oil globule diameter OD:ED oil globule diameter to egg diameter ratio Oe esophagus

PAGE 13

13 OG oil globule ON outer nuclear layer OP outer plexiform layer P pancreas pf post fertilization PI posterior intestine Pe pigmented epithelial layer PLD pelagic larval duration R retina S stomach Sb swim bladder SD standard deviation SE standard error SGR specific growth rate SL notochord length sp. species spp. species sqrt squareroot T thymus TAN total ammonia nitrate Temp temperature Tiso Tisochrysis lutea TL total length Ud urinary duct UV ultra violet

PAGE 14

14 V egg volume Vc Visual cell layer Vtg vitellogenin YS yolk sac YV yolk volume YV:V yolk volume to volume ratio micrometer

PAGE 15

15 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science ASSESSMENT OF THE MILLETSEED BUTTERFLYFISH, Chaetodon miliaris , AS A MODEL SPECIES FOR MARINE ORNAMENTAL AQUACULTURE AND AN EVALUATION OF EARLY CULTURE PARAMETERS By Jon Michael Degidio Chair: Roy Yanong December 2014 Major: Fisheries and Aquatic Sciences As part of the Rising Tide Conservation effort to advance marine ornamental aquaculture, the milletseed butterfly was chosen as a candidate species for the development of butterflyfish aquaculture methodologies. Spawning of Chaetodon miliaris was not inhibited in broodstock tanks and resembled that of previously documented chaetodontid species. An assessment of practical fecundity established that a population of 3 males and 8 females produced the largest number of viable eggs and that relationships between photoperiod and spawning frequency were present in all three broodstock populations. Larvae hatched after approximately 28 hours, were capable of feeding at 4 dph, and survived to 44 dph in preliminary culture trials. An examination of the first feeding requirements of C. miliaris established that the calanoid copepod Parvocalanus crassirostris at densities between 1 15 nauplii mL 1 , algal induced turbidity greater than ~190,000 cells mL 1 , water exchange rates below 300% day 1 , and stocking densities of 15 20 larvae L 1 enhanced the proportion of larvae feeding and larval feeding intensity, while tank size and light intensity did not influence first feeding. Growth and survival trials indicated that a photoperiod of 18 L:6 D, an algal cell density of 500,000 530,000 cells mL 1 and a stocking density of 15 larvae L 1 , and a

PAGE 16

16 water exchange rate of 100% day 1 and prey density of 10 nauplii mL 1 were optimal culture conditions for C. miliaris larvae to 10 dph. Further investigation into the culture requirements of C. miliaris will supply a crucial base for developing aquaculture protocols for other pelagic spawning marine ornamental species.

PAGE 17

17 CHAPTER 1 INTRODUCTION Aquaculture is the fastest growing sector of agriculture in the world. Progression of this industry is reliant on the advancement and refinement of culture methods. In 1999, approximately 25% of total food fish was supplied by aquaculture with an estimated value of $56.47 U.S. billion; however, by 2011 approximately 42% of total food fish was supplied by aquaculture with an estimated value of $119 U.S. billion (FAO, 2000; FAO, 2012). In the ornamental industry approximately 90% of freshwater ornamentals were supplied by aquaculture in 2003 while an estimated 1 10% of marine ornamentals were supplied by aquaculture in 2012 (Sadovy and Vincent, 2002; Wabnitz et al . , 2003; Rhyne et al., 2012). Marine ornamental aquaculture development has lagged significantly behind that of food fish and freshwater ornamental aquaculture. Substantial obstacles within marine ornamental aquaculture such as spawning, nutrition, and pelagic larval stages have inhibited the advancement of the industry (Olivotto et al., 2011). Many challenges are encountered in marine ornamental aquaculture. Assessing developmental parameters in a logical stepwise manner will highlight crucial components for advancing marine ornamental aquaculture techniques. Marine Ornamental Trade The trade of marine ornamental species has grown rapidly in the past several decades. There are approximately 1800 marine fish species, hundreds of species of invertebrates, and over 150 species of stony corals that are currently being traded throughout the world (Rhyne et al., 2012). Advances in the aquaculture industry and growing concerns about the sustainability of harvesting these species from the wild has

PAGE 18

18 sparked interest in developing methods for production of these species (Tlusty, 2002; Pomeroy et al., 2006; Wittenrich, 2007; Olivotto et al., 2011). Eight major families account for approximately 75% of all marine ornamental fish traded with in the United States (Rhyne et al., 2012). Pomacentridae (clownfish and damselfish), Labridae (wrasses), Pomacanthidae (angelish), Gobiidae (gobies), Acanthuridae (surgeon fish, tangs, and unicornfishes), Apogonidae (cardinalfishes), Microdesmidae (wormfishes), and Chaetodontidae (butterflyfish) represent the top 8 traded fish families respectively (Wabnitz et al., 2003; Rhyne et al., 2012). Exports mainly come from the Coral Triangle in the South Pacific, which includes the three highest export countries the Phili ppines, Indonesia, and Sri Lanka. The value of the marine ornamental trade was estimated to be between $200 330 U.S. million wholesale per year in 2003 (Wabnitz et al., 2003). The sustainability of the marine ornamental trade has come under question because of collection methods, regulation, ability to incur losses in biodiversity, over harvesting, vulnerability for spread of diseases and invasive species, and high mortality during transport (Wabnitz et al., 2003; Bell et al., 2009; Smith et al., 2009; Rhyne et al., 2012); however increased regulation and outreach initiatives are creating a more sustainable trade (Wood, 2001). While the marine ornamental trade is a high value industry there is often a large disconnect between the price point in an aquarium store and the price when collected (Wabnitz et al., 2003). In some areas the regulations in place create a sustainable harvest for the marine ornamental trade, but this is not always the case. When illegal harvesting methods are used, they can cause damage that goes beyond the extent of just

PAGE 19

19 removing fish from the ecosystem. Cyanide fishing is the most destructive method used for the collection of fish from the wild, causing damage to the coral reef, non target species, and mortalities in targeted species (Barber and Pratt, 1998; Tissot and Hallacher, 2003; Bell et al., 2009). In addition, mortalities experienced by long transport times from the South Pacific to the U.S. and Europe creates a need to increase collection number s to compensate for the loss of fish during transport (Sadovy and Vincent, 2002). The import and export of these species can also lead to biological invasions in non native waters such as that of the lionfish along the southeastern seaboard of the United States and the spread of diseases (Goodsell, 1996) such as megalocytivirus and betanodavirus (Smith et al., 2009; Yanong, 2010; Yanong and Waltzek, 2010; Betancur R et al., 2011). While there are still some areas of the world that harvest marine ornamental species with unsustainable methods, many regions have imposed regulations for sustainable harvest. Countries such as Australia and Sri Lanka regulate collection net type and size and the use of cyanide has been outlawed in every country that exports marine fish (Wood, 2001). The collection of marine ornamental species can have negative impacts; however, if regulated sustainably the harvest of marine ornamentals can be beneficial for coastal communities, since the value of corals, invertebrates, and fish is approximately 900% higher than that of species harvested for food (Wabnitz et al., 2003).

PAGE 20

20 Marine Ornamental Aquaculture Ornamental aquaculture is prominent in the freshwater industry with approximately 90% of traded fish in production. This is nearly opposite in the marine ornamental industry (Wabnitz et al., 2003; Moorhead and Zeng, 2010). Currently, at least one representative species has been cultured from the families Pomacentridae, Labridae, Pomacanthidae, Gobiidae, Apogonidae, Microdesmidae, and Chaetodontidae; however culturing of an Acanthuridae species has eluded completion (Sweet, 2014 ). Only two of the eight most imported fish families (Pomacentridae and Gobiidae) have species in commercial production and the number of species being produced is 1 5% of the over 1800 species harvested (Moe, 2003; Chapman et al., 1997; Sadovy and Vincent, 2002; Tissot and Hallacher, 2003; Calado, 2006). Limited understanding of the reproductive metabolism, environmental conditions, and social requirements of each species as well as production economics reduce the feasibility of producing the array of species collected from the wild. While research into the aquaculture of new marine species is being conducted at many facilities around the world, these bottlenecks have restricted the expansion of marine ornamental aquaculture. Broodstock Management and Spawning Broodstock management is the accumulation of all factors that pertain to the production of high quality gametes (Moorhead and Zang, 2010). Understanding natural reproductive cycles of the species in question helps define specific parameters, which can optimize production of eggs and sperm. Differing reproductive strategies necessitate different culture methods, presenting a major challenge to marine ornamental aquaculture (Watson and Shireman, 1996; Ostrowski and Laidley, 2001).

PAGE 21

21 Reproduction is marked by high energy demands for adult fish. Spawning requires precise signals from a variety of internal and external cues. A successful spawn is reliant upon metabolic factors such as energy status and sufficient size of brood fish, optimal environmental conditions for reproductive success, and the appropriate social structure (Olivotto et al., 2011). Careful selection of mature and healthy broodstock can reduce conditioning time and influence egg production. Reproductive Metabolics Important biological and metabolic requirements for spawning include mature gonads, sufficient energy for developing gametes, and proper nutrition. In food fish species, Kamler (2005) and Vuthipahandchai and Zohar (1999) found that egg and sperm quality was decreased in newly matured and older fish while middle aged adults produced eggs of higher nutritional content and greater sperm counts with higher spermatocrit. Correspondingly, size can impact fecundity of broodstock (Trippel et al., 1997; Reavis, 1997; Morita and Takashima, 1998; Kolm, 2002). Reavis (1997) found that fecundity of the goby, Valencennea strigata , increased exponentially with length. Additionally, Beyer et al . (2014) correlated maternal length and somatic weight with increasing fecundity of four different rock fish species in California. Selecting broodstock of the ideal size and age incre ases the fecundity, but energetic demand, the amount of energy taken in, as well as nutritional requirements, proper ratios of proteins, amino acids (AA), lipids, essential fatty acids (EFA), carbohydrates, and vitamins, must be supplied as well. Successful reproduction requires excess intake of food because the energy demands for body maintenance and growth of gonadal products are occurring (Lupatsch et al., 2010). Ma et al. (1997) examined the effects of increased energy

PAGE 22

22 consumption on fecundity and somatic growth in Atlantic herring. Decreased rations resulted in decreased somatic growth and impeded gonadal maturation. Analysis of freshwater species reproductive seasons exhibited similar trends shifting energy allocation from somatic growth towards ovarian growth (Henderson et al . , 1996; Henderson et al., 2000; Lupatsch et al., 2010). Specifically, in tilapia, Lupatsch et al. (2010) found that fish fed a maximum amount (2.13 ± 0.44 g day 1 ), a medium amount (0.95 ± 0.20 g day 1 ) and zero (0.01 ± 0.00 g day 1 ) exhibited positive, negative, and negative total energy deposition as gonadal and somatic growth, respectively. Only fish being fed to satiation three times per day, the maximum feeding regime, exhibited positive weight gain during the reproductive period. Energetic demands during reproduction are extremely high and broodstock feeding regimes must account for this to optimize egg production. Broodstock Nutrition Broodstock nutrition is critical to the production of quality eggs, gametes, and larvae (Izquierdo et al., 2001; Lupatsch et al., 2010; Callan et al . , 2012a; Callan et al., 2012b). Complete artificial diets supply all ingredients (protein, carbohydrates, lipids, vitamins, and minerals) in the proper amounts to optimize the health and growth of fish, while supplemental diets are used in addition to wild foraging (Craig and Helfirch, 2002). Fish fed a diet with the appropriate protein and lipid levels have improved gamete quality (Tocher, 2003; Kahn et al., 20 05; Lupatsch et al., 2010; Fuiman and Faulk, 2013). Research of protein requirements has focused predominantly on somatic growth in food fish species and established that higher dietary protein levels produce greater

PAGE 23

23 somatic growth (Einen and Roem, 1997; Millamena, 2002; Craig and Helfrich, 2002; Holt et al., 2007). On average, dietary protein levels range between 18 50% with some raw feeds such as mysis shrimp containing greater than 50% protein (Craig and Helfrich, 2002). If protein demands for reproduction are not provided in the diet, catabolism of muscle within the body can occur to meet protein requirements of developing eggs (Lupatsch et al., 2010). Lipids are another important component of the broodstock diet. A majority of freshwater fish have the ability to elongate and desaturate fatty acids while marine fish have evolved to depend on the fatty acids acquired from their diet (Sargent et al., 1997). In aquaculture of marine species, this means that the required essential fatty acids must be supplied through the diet (Craig and Helfrich, 2002). The highly unsaturated fatty acids (HUFAs); docosahexaenoic acid (DHA; 22:6n 3), eicosapentaenoic acid (EPA; 20:5n 3) and arachidonic acid (ARA; 20:4n 6) are essential for the growth and development of marine fish embryos and larvae (Craig and Helfrich, 2002; Holt et al., 2007; Moorhead and Zeng, 2010; Callan et al., 2012a; Callan et al., 2012b; Fuiman and Falk, 2013). Callan et al. (2012b) examined the effects of dietary HUFAs on broodstock flame angelfish reproduction, egg, and larval quality measuring fecundity, fertilization rates, and embryo viability to decipher differences in dietary effects. ARA, EPA, and DHA were combined to calculate total HUFA n 3 percentages with Cen tropyge loriculus . HUFA level of 3.6% n 3, increased fecundity, fertilization rates, and embryo viability significantly compared to lower dietary levels 1.8% and 2.3% n 3. In addition to lipids and protein, incorporation of vitamins, minerals, and carotenoids in diets have been shown to have beneficial effects on production of eggs (Craig and Helfrich, 2002).

PAGE 24

24 Another important molecule is vitellogenin (Vtg), a glycophospholipoprotein, exogenously synthesized by the liver, transported by the blood stream to the ovary, and taken up by the oocyte (Patiño and Sullivan, 2002). Molecules are cleaved into 2 yolk proteins, lipovitellin and phosphovitin (Watts et al., 2003). Vtgs contain up to 20% lipid by weight, with a large portion (60 80%) being phospholipids typically rich in polyunsaturated fatty acids (Patiño and Sullivan, 2002). Oocytes of pelagic spawning species contain oil droplets, which coalesce into one or two large oil globules during maturation and are composed of neutral lipids rich in monounsaturated fatty acids (Patiño and Sullivan, 2002). The neutral lipids provided by oil globules are transported into the oocyte by other means than Vtg, which mainly transports structural lipids, and are used as energy reserves for embryogenes is and larval growth (Patiño and Sullivan, 2002). New research indicates that a different structure for Vtg exists in each species (Watts et al., 2003). Environment Different reproductive strategies necessitate different culture approaches. Pelagic spawning species release eggs into the water column for transport into the pelagic environment, while demersal spawners attach or deposit eggs on benthic substrate (Johannes, 1978; Barlow, 1981; Olivotto et al., 2011; Hoey et al., 2012). Other species, such as the Bangii cardinalfish ( Pterapogon kauderni ) and yellow headed jawfish ( Opistognathus aurifrons ) are mouth brooders, holding clutches of eggs in their mouth. Chemical and physical environmental conditions such as water flow, temperature, salinity, photoperiod, tank size, and structure can induce or prevent spawning (Holt and Riley, 2000; Miller et al., 2009; Donelson et al., 2010; Moorhead and Zang, 2010; Olivotto et al., 2011). While fish exhibit adaptive behavior in a captive

PAGE 25

25 environment, replicating natural spawning conditions can stimulate reproduction (Coward et al., 2002). Holt and Riley (2000) manipulated photoperiod and temperature regimes to induce spawning in the pygmy angelfish ( Centropyge argi ), spotfin hogfish ( Bodia nus puchellus ), bluehead wrasse ( Thalassoma bifasciatum ), and clown wrasse ( Halichores maculipinna ). Representing spring season conditions, a photoperiod of 11 h light and 13 h dark with a temperature of 22 °C induced continuous daily spawning of the pygmy angelfish, spotfin hogfish, and bluehead wrasse. The clown wrasse spawned when temperature (27.5 °C) and photoperiod (13 L:11 D) matched that of the summer season. By controlling key environmental parameters spawning periods can be extended under captive conditions. Additionally, observed spawning behavior was adapted to fit the captive constraints of the 300 L broodstock tanks from all four species (Holt and Riley, 2000). Social cues such as gender ratios, number of fish present in a spawning group, territoriality, and hermaphroditisim pose further obstacles to aquaculturists (Tricas, 1989; Hourigan, 1989; Yabuta and Kawashima, 1997; Holt and Riley, 2000; Moorhead and Zang, 2010; Olivotto et al., 2011). Information pertaining to social dynamics of marine ornamental species is limited in the wild and further lacking for captive environments. Other broodstock considerations should include selecting fish from the same source and implementing a quarantine period. These steps can help decrease the risk of disease transmission and allow for dietary assessment, observations of behavior and condition prior to establishment as broodstock.

PAGE 26

26 Gametogenesis and Embryology Understanding gamete development and embryology is necessary to develop captive breeding techniques for marine ornamental species (Coward et al., 2002). In many species, gonads are undifferentiated until fish become sexually mature. Once mature, females will begin oocyte development during the spawning season while males produce spermatozoa. Oocyte development is defined by three major stages: primary growth, secondary growth, and maturation with multiple steps occurring within each stage (Grier et al., 2009; Rhody et al., 2013). Careful cannulation, removal of a gonad sam ple by insertion of a tube through the genital poor of the broodstock, can produce valuable information on maturity, gender ratio, oocyte development stage, and broodstock conditioning (Ralston, 1981; Neidig et al., 2000; Grier et al., 2009; Rhody et al., 2013). Embryological development among marine ornamental species varies greatly. In demersal spawning species like clownfish, newly hatched larvae are more developed than pelagic spawning species such as butterflyfish, which produce a prolarvae (Suzuki et al., 1980; Avella et al., 2007; Kumar and Balasubramanian, 2009; Olivotto et al., 2011). Various abiotic and biotic factors during egg incubation effect embryonic development, and ultimately the viability of early larvae (Moorhead and Zeng, 2010). Larval Development and Culture Larviculture poses many challenges to successful production of a species. Larval development varies according to species and spawning strategy. Larvae from demersal spawning species hatch with developed eyes and mouths, whereas, pelagic larvae hatch without developed eyes, mouths, or digestive tracts (Suzuki et al., 1980; Leis, 1989; Wittenrich, 2007; Kumar and Balasubramanian, 2009; Setu et al., 2010; Olivotto

PAGE 27

27 et al., 2011; Callan et al., 2012b; Madhu et al., 2012). Clownfish is a good example of a commercial species with demersal larvae; larvae hatch after approximately 6 8 days and measure 4 5 mm total length depending on species. Rotifers are used as a first feed and are enriched to add essential nutrients. After approximately 10 days, the larvae will transition onto Artemia nauplii and about 5 days later the larvae will metamorphose into juvenile fish (Gordon and Bok, 2001; Ignatius et al., 2001; Avella et al., 2007; Wittenrich, 2007; Kumar and Balasubramanian, 2009; Setu et al., 2010). In contrast, tropical pelagic larvae typically hatch 24 48 h after spawning. Larvae will absorb yolk reserves, develop functional eyes and mouths, and begin feeding between 48 96 h post hatch (Suzuki et al., 1980; Leis, 1989; Holt and Riley, 2000; Holt et al., 2007; Olivotto et al., 2011; Callan et al., 2012b; Leu et al., 2013; Zavala Leal et al., 2013). The transition of larvae from utilizing endogenous yolk reserves to exogenous feeding presents a major bottleneck for many species (Suzuki et al., 1980; Holt and Riley, 2000; Leu et al., 2009; Moorhead and Zeng, 2010; Olivotto et al., 2011; Leu et al., 2013; Zavala Leal et al., 2013). Particularly in the marine ornamental industry, a variety of alternative feeds are being examined to replace more traditional live feeds (Kraul, 1989; Reitan et al., 1997; Holt, 2003; Mckinnon et al., 2003; Sampey et al., 2007; Baensch and Tamaru, 2009; Moorhead and Zeng, 2010; Olivotto et al., 2011; Leu et al., 2013; Zavala Leal et al., 2013). Cassiano et al. (2011) evaluated the copepod, Pseudodiaptomus pelagicus , as an alternative feed for the Florida pompano, Trachinotus carolinus . Survival of pompano larvae to 7 dph was significantly higher in the treatment fed copepod nauplii versus the rotifer, Branchionus sp.

PAGE 28

28 Larval nutrition is especially important for increasing survival and optimizing development. Little information is known about the nutritional requirements of marine larvae. Some research exists on marine larval f ood fish nutrition (Reitan et al., 1997; Ronnestad et al., 1999; Sargent et al., 1999; Garcia Ortega et al., 2002; Faulk and Holt, 2003; Garcia Ortega, 2008), but overall larval nutritional research on marine species is limited (Holt, 2011). Marine fish larvae are thought to require a diet including carbohydrates, proteins, lipids, vitamins, and minerals (Holt, 2011; Hamre et al., 2013). In particular, research has focused on essential amino acids, % protein, and essential fatty acids requirements (Sargent et al., 1999; Garcia Ortega et al., 2002; Cahu et al., 2003; Holt, 2003; Meeren et al., 2007; Garcia Ortega, 2008; Moorhead and Zeng, 2010; Zambonino Infante and Cahu, 2010; Holt, 2011; Olivotto et al., 2011; Hamre et al., 2013). Digestive systems of larval fish, especially those of pelagic spawning marine ornamental species, are inadequately developed at hatch compared to adult fish, consisting of an unopened, undifferentiated alimentary canal. In the wild, primary digestive flora is obtained through passive ingestion; however, microbial communities differ between culture systems and the wild (Grisez et al., 1997; Skjermo et al., 1997). Development of a fully functional digestive system does not occur until metamorphosis into a juvenil e stage in some species (Boulhic and Gabaudan, 1992; Sarasquete et al., 1995; Sanz et al., 2011; Trevino et al., 2011; Cohen et al., 2013). An assessment of digestive system ontogeny can be beneficial for understanding the nutritional requirements of larvae throughout development.

PAGE 29

29 Additional bottlenecks often occur throughout larval development. Mortality is often associated with swim bladder inflation, flexion, and metamorphosis (Battaglene and Talbot, 1990; Battaglene et al., 1994; Moorhead and Zeng, 2010; Olivotto et al., 2011). Innumerable parameters can affect the survival of larvae but feeds, temperature, water flow, water quality, light intensity, photoperiod, and turbidity have been assessed (Houde et al., 1978; Leis, 1989; Naas et al., 1992; Piers and Purser, 1995; Arvedlund et al., 2000; Holt, 2003; Mckinnon et al., 2003; Baensch and Tamaru, 2009; Moorhead and Zeng, 2010; Olivotto et al., 2011; Cobcroft et al., 2012; Leu et al., 2013; Zavala Leal et al., 2013). After metamorphosis, juveniles have similar nutritional needs as adults. Increased feeding will increase growth, however, overfeeding can lead to water quality problems and possible lipid deposition in the organs, especially the liver (Wang et al., 2013). Chaetodontidae Chaetodontid butterflyfishes are highly sought after in the marine aquarium trade for their vivid colors and complex patterns. Butterflyfish species can be susceptible to stress, problems can occur with acclimation, some sp. have corallivor ous feeding behavior, and can be territorial; but once established in an aquarium they exhibit beautiful colors and displays in behavior. Approximately 4% of all fish imported and exported globally are chaetodontid butterflyfish, ranking them as the sixth most traded family in the world in 2003 (Wabnitz et al., 2003), and eighth most traded family in the United States in 2012 (Rhyne et al., 2012). Of the 120 species of chaetodontid butterflyfish, 97 are traded within the United States (Rhyne et al., 2012).

PAGE 30

30 All species have a compressed body form with an ovate body shape and are usually closely associated with coral reefs in the tropics. Very few species have ranges outside tropical waters 21 °C) (Burgess, 1978). Many Chaetodontidae species are considered poor choices for aquariums because their diet consists of coral and other invertebrates. Four different diet preferences of chaetodontid species have been described: (a) obligatory corallivores, (b) facultative corallivores and benthic invertebrate feeders, (c) non coraline benthic invertebrate feeders, and (d) zooplanktivores (Sano, 1998). It is possible that once species are captive, dietary preferences can be changed. Aquaculture of chaetodontid species has been pursued several times with one successful culture attempt (Suzuki et al., 1980; Tanaka et al., 2001; Wittenrich and Cassiano, 2011; Baensch, 2014). Heniochus diphreutes , from wild collected eggs, is the singular representative of the butterflyfish family that has been successfu lly cultured. Frank Baensch of the Hawaiian Larval Fish Project, a privately funded venture, stated during personal communications that culture methods would not be revealed until the project is completed. Eric Cassiano and Matt Wittenrich at the University of Tropical Aquaculture Lab had preliminary success at raising H. diphreutes larvae to 46 dph, but were not able to raise fish through the complete larval duration. Both records suggest that copepods are sufficient as an early feed with unknown wild zooplankton used as feed later in development (Wittenrich and Cassiano, 2011; Baensch, 2014). Objectives The development of the marine ornamental aquaculture industry has come to a pivotal time. As the marine ornamental industry grows and wild collection is further pressured by regulation and closure, there will be an increased demand for cultured

PAGE 31

31 marine ornamental species. Commercial producers capable of efficiently culturing juvenile fish to market size, at a cost below market value, stand to profit. Consequently, aquaculture production of marine species can be used to supplement natural collection, reduce harvesting pressure on wild populations, and support a growing aquaculture commodity. The milletseed butterflyfish, Chaetodon miliaris , was selected as a model Chaetodontidae species for this project. Preliminary success in culturing similar butterflyfish species indicates that culturing of C. miliaris is plausible. The purpose of this thesis is to evaluate the milletseed butterflyfish and its potential for marine ornamental aquaculture. Broodstock fecundity, spawning behavior and conditions, embryology, and larval development were assessed. Furthermore, eight experiments were conducted to evaluate first feeding parameters and three multifactorial experiments were conducted to assess survival and growth to 10 dph. Differences in fecundity measures, spawning behaviors, and spawning conditions were observed to evaluate the effect of spawning population size on egg production. The proportion of larvae feeding and number of prey items in the guts were measured to determine optimal first feeding parameters. Finally, differences in survival and growth (notochord length, body depth, and eye diameter) were measured to determine the effect of algal turbidity and stocking density, prey density and water exchange rate, and photoperiod on C. miliaris larvae. Supplementary research and results beyond those included as part of this thesis are contained within the Appendix A. These objectives include: a) analysis of fatty acid

PAGE 32

32 profiles of eggs and larvae, and b) organogenesis of the digestive track and visual systems.

PAGE 33

33 CHAPTER 2 SPAWNING, EMBRYOLOGY, AND LARVAL DEVELOPMENT Foreword Records of captive Chaetodon reproduction are limited (Suzuki et al., 1980; Hioki, 1997; Tanaka et al., 2001). A comparison of spawning between wild and captive butterflyfish stocks demonstrates similar reproductive behavior (Suzuki et al., 1980; Colin, 19 89; Lobel, 1989; Tricas and Hiramoto, 1989; Londraville, 1990; Yabuta, 1997; Yabuta and Kawashima, 1997; Tanaka et al., 2001; Pratchett et al., 2014). A typical spawning event begins in the afternoon/evening with courtship behavior occurring between pairs, harems, or multiple males and a singular female. Courtship behavior generally consists of aggressive male male behavior, males following females, circular swimming patterns, and gentle nudges of th e snout along the abdomen of the female. The act of spawning will then occur with the male and female dashing in a direction, usually upwards, with their abdomens together and releasing gametes into the water column at the top of their ascent. Additional males may trail behind the dominant male during group spawning events in hopes of fertilizing ova with their own gametes (Suzuki et al., 1980; Colin, 1989; Lobel, 1989; Tricas and Hiramoto, 1989; Londraville, 1990; Yabuta, 1997; Yabuta and Kawashima, 1997; Tanaka et al., 2001; Pratchett et al., 2014). Ralston (1976; 1981) examined the reproductive biology and feeding ecology of C. miliaris in the waters surrounding Oahu, Hawaiian Islands. Sampling uncovered that spawning was seasonal (January May), with no correlation to lunar periodicity. Ralston (1976) assessed the fecundity of C. miliaris as the number of mature oocytes within the ovary of a spawning female. Average fecundity was estimated at 79,100 mature oocytes

PAGE 34

34 per female. At maturity, approximately 9 cm total length (TL), an energetic shift from somatic to reproductive growth was inferred by a quadratic relationship between fecundity and fish weight. The quadratic relationship here implies that the larger the fish the more energy is allotted towards reproductive growth. Unlike natural environments, captive environments can be controlled. Suzuki et al. (1980) observed periodic spawning from C. nippon throughout March, August, September, and October at water temperatures greater than 23 °C and Tanaka et al. (2001) observed spawning of C. modestus over an 18 month period at temperatures between 18 26 °C. Similarly to Ralston (1980), spawning of C. nippon and C. modestus was not correlated with lunar periodicity (Suzuki et al., 1980; Tanaka et al., 2001). Information on the embryology of butterflyfish comes from four sources (Madden and May, 1977; Suzuki et al., 1980; Hioki, 1997; Tanaka et al., 2001). Embryo development was similar in all four instances except for timing of each stage. Embryogenesis of zebrafish ( Danio rerio ) is well studied and serves as a model for teleost development. After fertilization, embryos immediately enter the zygote period and undergo two cell divisions before entering the cleavage period, then the blastula period, the gastrula period, and the segmentation period before hatching occurs (Kimmel et al., 1995). Milletseed butterflyfish embryogenesis parallels that of other butterflyfish and marine angelfish (Madden and May, 1977; Suzuki et al., 1980; Hoiki et al., 1990; Hioki, 1997; Tanaka et al., 2001; Leu et al., 2009). Fertilized eggs of described chaetodontid species are pelagic, range in size from 0.6 0.8 mm in diameter, and hatch 15 30 h post spawning depending on the species and water temperature (Burgess, 1978; Suzuki et al., 1980; Colin and Clavijo, 1988;

PAGE 35

35 Colin, 1989; Tanaka et al., 2001; Wittenrich and Cassiano, 2011). Newly hatched larvae of marine ornamentals are small, typically measuring between 1 3 mm TL. Butterflyfish larvae range from 1.2 1.5 mm at hatching to approaching 2.5 mm at the transition from endogenous to exogenous feeding. Similar to other reef fish, first feeding will occur 72 96 h after spawning (Suzuki et al., 1980; Leis, 1989; Tanaka et al., 2001; Baensch and Tamaru, 2009; Leu et al., 2009; Moorh ead and Zeng, 2010; Olivotto et al., 2011; Callan et al., 2012a; Leu et al., 2013). Information on development of growing larvae past the first feeding stage is extremely limited to several online articles and photographs (Wittenrich and Cassiano, 2011; Baensch, 2014). Suzuki (1980) and Tanka (2001) had no success initiating a feeding response resulting in mortality from starvation. According to Baensch (2014) and Wittenrich and Cassiano (2011) first feeding bannerfish larvae are capable of feeding on copep od nauplii 80 Development of these larvae is not well documented with only two representative samples of development. Larvae begin to increase body depth around 11 dph. At 25 dph flexion has begun and large bony plates on the head, an elongated preopercular, a post temporal, and a supracleithrum plate form (Johnson, 1984; Leis, 1989). Variations among plate shape and head morphology are present between different species (Leis, 1989). It is estimated that the pelagic larval duration (PLD) of butterflyfishes ranges from 30 60 dph and possibly up to 80 dph (Leis, 1989; Booth and Parkinson, 2011; Soeparno et al., 2012). Natural will differ from those exhibited under aquaculture conditions. Baensch (2014) observed a PLD of 85 dph with H. diphreutes . However, as seen with many other aquacultured species, larval durations can be altered by optimizing egg and larval quality, larval nutrition, environmental parameters, feeding conditions, water

PAGE 36

36 quality, and settlement cues (Cowen, 1991; Leis and McCormick, 2002; Bergenius et al., 2005; Lecchini et al., 2005; Gjedrem and Baranski, 2009; Oyarzun and Strathmann, 2011; Leis and Yerman, 2012; Wenger and McCormick, 2013). Methods Acquisition of Broodstock and Broodstock Management Broodstock of the milletseed butterflyfish, Chaetodon miliaris , were obtained from the Rainbow Reef exhibit at Aulani, A Disney Resort and Spa in Ko Olina, on two different occasions in 2013. During May of 2013 a group of 23 broodstock were transported to Segrest Farms in Apollo Beach, FL, where they were pH acclimated before transport to University of Tropical Aquaculture Lab in Ruskin, FL. Initial skin, fin, and gill samples were taken for diagnostic purposes. A 5 minute freshwater dip was used to remove parasites (Noga, 2010). Fish were quarantined for 30 days using a 10 mg L 1 dose of chloroquine phosphate EP (PCAA 9901 South Wilcrest Drive Houston, TX 77099) as a precautionary measure for protist parasites and three 24 h treatments of praziquantil 5 days apart were used as a precautionary measure for treamatode parasites (PCAA 9901 South Wilcrest Drive Houston, TX 77099) (Yanong, 2003; Yanong, 2009; Noga, 2010). Finally, another set of diagnostic samples was taken to verify absence of parasites or other obvious diseases prior to introduction into the broodstock system. Because milletseed butterflyfish lack sexual dimorphism, every fish within the population was cannulated using a 0.76 mm inner diameter and 1.65 mm outer diameter polyethylene tube to determine sex (Ralston, 1981; DiMaggio et al., 2013). Female gonad condition was preliminarily assessed by a wet mount technique to determine spawning condition (Rhody et al., 2013). Presence of milt through palpation

PAGE 37

37 of the coelom anterior to the urogenital opening or cannulation was used to differentiate males (DiMaggio et al., 2013). The first group contained 2 males and 20 females, with one mortality after arrival. Additional fish were collected from Rainbow Reef in October 2013, cannulated prior to transport to determine sex, and 12 confirmed males were transported using the same methodology as the first group. The same quarantine protocols were followed for this group. Male fish were tagged in the caudal fin rays using visible implant elastomer tags (Northwest Marine Technology, Inc. PO Box 427, Ben Nevis Loop Rd. Shaw Island, WA 98286). The butterflyfish were divided into three different populations: 1 male and 1 female, 3 males and 8 females, and 10 males and 11 females. Two 1937 L circular tanks and one 2325 L square tank, each with a 1000 L sump, were used to hold broodstock. In both systems a recirculating design included a bead filter, a fluidized sand bed, and a 50 watt ultraviolet (UV) sterilizer. Fish were fed to satiation 4 5 times daily with California black worms ( Lumbriculus variegatus ), frozen striped mullet ( Mugil cephalus ) roe, opossum shrimp ( Mysis relicta ) (Piscine Energetics, British Colombia, Canada; 69.5 % protein, 8.35% lipids, 2.75% fiber, 5.5% ash), and 1.7 mm pellets (Reed Mariculture, Campbell CA, USA: APBreed TDO EP1; 46% protein, 16% lipid, 2% fiber, 6.5% moisture, 14% ash, 1.7% phosphorus, 250 mg L 1 astaxanthin). Water temperatures and salinity were measured daily. Water temperature in both systems was maintained between 23 30 °C. Salinity was maintained between 30 36 g L 1 . Excess feed was siphoned from the tank daily. Tanks were scrubbed and siphoned once every other week to remove algal growth and water quality parameters were tested weekly

PAGE 38

38 (total ammonia nitrogen (TAN), nitrite Nitrogen, nitrate, alkalinity, pH) using Hach Permachem reagents (Hach Company, Loveland, CO, USA). Two 58 L external egg collectors were designed to collect floating and sinking eggs within a 19 L bucket lined with 500 mesh (Baensch and Tamaru, 2009; Callan et al., 2012b; DiMaggio et al., 2013). In the square tank, a floating airlift surface collector was used to skim surface waters and collect eggs. Egg collectors were set between 18:00 and 20:00 nightly and removed at 08:00 to collect spawn throughout the night. Preliminary assessments of egg quality, egg morphometrics, hatch success, and 3 dph survival found no differences between eggs collected immediately after spawning (~21:00 22:00) and eggs that remained over night in the egg collector. Spawning Behavior, Conditions, and Fecundity Spawning behavior was observed a total of five times for each population. Data from four personal observations and one observation via video recording was compared to documented behavior of wild and captive butterflyfish spawns (Suzuki et al., 1980; Lobel, 1989; Tricas and Hiramoto, 1989; Colin, 1989; Londraville, 1990; Yabuta, 1997; Yabuta and Kawashima, 1997; Tanaka et al., 2001; Pratchett et al., 2013). Visual observations were made in the evening starting around 18:00. The individual would stand a minimum distance of 5 feet from the tank, remaining quite and still until spawning was observed. Video observations were recorded with a Gopro Hero 3 underwater camera attached to the side of the tank. Both video and stop motion pictures were used to analyze spawning behavior. Salinity and temperature measurements were taken each evening and pH was monitored weekly (Suzuki et al., 1980). Salinity, temperature, and pH remained fairly constant throughout the spawning period and were not included in the logistic

PAGE 39

39 regression ( Figure 2 1 ). Records of daily photoperiod lengths and lunar phases were amassed using the United States Naval Observatory complete sun and moon data for one day database ( http://aa.usno.navy.mil/data/docs/RS_OneDay.php ). Daily lunar phases were grouped into 3 day periods around percentages of the moon visible. Lunar phases were grouped into 0, 20, 40, 60, 80, and 100% visible to increase the likelihood that lunar phase woul d have a significant association with spawning. Although fecundity estimates of milletseed butterflyfish were conducted by Ralston (1976), additional fecundity data was collected. Ralston (1976) defined fecundity as the estimated number of mature egg cells within the ovary of a spawning female or absolute fecundity . In this document the term fecundity will refer to the practical fecundity or the quantity of viable eggs obtained from a female post ovulation (Baensch and Tamaru, 2009; Callan et al . , 2012b). Eggs spawned the night before were gently rinsed with broodstock water onto a 400 sieve from egg collectors each morning (Baensch and Tamaru, 2009; DiMaggio et al., 2010; Callan et al . , 2012b; DiMaggio et al ., 2013). Eggs were then rinsed using broodstock water into individual clean 1 L plastic containers containing at least 500 mL of broodstock seawater (Baensch and Tamaru, 2009; Callan et al . , 2012b). Eggs were allowed to float for 2 min and separated into a 100 mL graduated cylinders to volumetrically determine total egg production. A 1 mL sample of eggs from 20 different spawns was enumerated to obtain an accurate count for volumetric readings (DiMaggio et al . , 2010). Sinking eggs were considered non viable. Fertilization rate and development stage were quantified from a sample ( n =100) of floating eggs and ( n =100) of sinking eggs. A sample of 100 viable eggs were placed on a Sedgwick Rafter slide and ensuing photographs (Jenoptik,

PAGE 40

40 ProgRes Capture Pro v2.8.8) were taken for morphometric analysis using ImageJ image analysis software (Baensch and Tamru, 2009; DiMaggio, 2012; DiMaggio et al . , 2013). Hydrated Egg Characteristics and Larvae Evaluation The morphological measurements for egg diameter (ED), oil globule diameter (OD), yolk volume (YV) and egg volume (V) were taken using photographs from 50 different spawns and measured using ImageJ image analysis software. Three replicates of 16 eggs were measured from each spawn. Egg radius and egg volume were calculated using r=d avg /2 and V=4/3 3 respectively. Yolk volume was calculated using the formula for a prolate sphere V y L y H y where L y is yolk length and H y is yolk height (Bonislawska et al . , 2001). Ratios of OD:ED and YV:V were also calculated (Imanpoor and Bagheri 2010; DiMaggio, 2012). To obtain growth and survival for each spawn six 1 L containers with 55 bottom screens wer e stocked with 16 eggs each in a temperature regulated (24 °C) 193 L water bath to ensure equivalent water quality and temperature between replicates (DiMaggio, 2012). Three replicates were removed 30 h after stocking (0 dph), hatch success and larval survival enumerated, and a minimum ( n =10) larvae from each replicate were sampled for notochord length (DiMaggio, 2012). On 3 dph another three replicates were removed using the previous methodology. All larvae surviving to this point were measured. Notochord length was defined as the anterior most point of the head to the posterior tip of the notochord (DiMaggio, 2012). Water Quality Water for all experimental and culture trials was synthetic salt mixed with reverse osmosis filtered water (Spectrum Brands Inc., Madison WI, USA: Instant Ocean Aquarium Sea Salt Mixture). Water was mixed in 14,000 L towers and pumped into

PAGE 41

41 culture tanks. Water quality parameters were evaluated daily in the water bath. Dissolved oxygen (DO) and temperature were evaluated with a YSI 85 multiparameter meter (YSI inc., Yellow Springs, OH, USA). Total ammonia nitrogen (TAN), nitrite nitrogen, nitrate, and pH were measured using Hach Permachem reagents (Hach Company, Loveland, CO, USA). Statistical Methods All statistics were analyzed using SPSS v. 21 (IBM, Armonk, NY). A logistic regression was conducted to determine if lunar phase and photoperiod were associated with an increased likelihood of spawning. The presence of spawn was coded as 1 and no spawn coded as 0. A P value of 0.05 was considered statistically significant for all analyses. Practical fecundity data did not meet parametric assumptions; consequently population measures were analyzed using a Kruskal Wallis one way analysis of variance with a pairwise multiple comparison test. All percentage data was arcsine square root transformed prior to analysis. All non parametric numerical data is represented as median ± MAD. A P value of 0.05 was considered statistically significant for all analyses. Arcsine(sqrt) transformed data was back transformed and represented in original units. The biological characteristics of hydrated eggs from each population were compared using a MANOVA. A P value of 0.05 was considered statistically significant. Relationships between egg diameter, yolk volume, oil globule diameter, egg volume, and ratios of OD:ED and YV:V to hatching success, notochord length, and survival did not meet normality assumptions. Data was analyzed using a nonparametric

PAGE 42

42 Spearman rank correlation. A P va lue of 0.01 was considered statistically significant to protect for type II error. Embryology Eggs were colle cted using a fine mesh net (50 m) immediately after spawning occurred. Eggs were rinsed from the net into a 1 L plastic container with 500 mL of water from the broodstock tank. A sample of 100 ± 10 eggs were then placed in a welled slide and photographed (Jenoptik, ProgRes Capture Pro v2.8.8) every hour until hatching. Development stages were assessed by observation according to Glamuzina et al. (2000; 1998), specifically observing for the zygote period, cleavage period, blastula period, gastrula period, and segmentation period. Larval Development Eggs were stocked into three 128 L (61.0 cm diameter x 54.6 cm height) culture tanks at 15 20 eggs L 1 . Light aeration 0.1 L min 1 ) was supplied for each tank throughout culture trials. At 3 dph Tisochrysis lutea (Tiso), formerly the Tahitian strain of Isochrysis galbana (Bendif et al., 2013), was dripped into each tank. Algal cell density was maintained at 400,000 600,000 cells mL 1 for the duration of larval experiments. At 4 dph the nauplii of the copepod Parvocalanus crassirostris were fed to larvae at a density of 1 5 nauplii mL 1 depending on the number harvested. Copepods cultures were fed Tisochrysis lutea , Chaetoceros gracilis , and Tetraselmis suecica at a predetermined volume each day. Copepod nauplii were harvested using an airlift collector each morning and sieved between 20 Copepod densities were monitored daily and added as necessary to maintain prey densities. Copepod nauplii were fed throu ghout the first 24 dph. At 25 dph, the addition of copepodites to culture tanks occurred once every 8 days to add a larger prey item. Larvae were maintained on

PAGE 43

43 a diet of copepod nauplii and copepodites until all larvae died at 44 dph. Throughout larval culture stages water was exchanged with new seawater daily, progressively increasing from 5% of the tank volume at 4 dph to 100% at 15 dph. At 21 dph 200% of the tank volume was exchanged daily increasing to 300% by 30 dph. Larvae were reared in a flow through system. Constant aeration was provided to ensure adequate DO concentration. Water quality parameters including DO, temperature, salinity, pH, total ammonia nitrogen (TAN), nitrite nitrogen, and nitrate were measured once daily. Larval tanks were maintained at a salinity of 33.2 35.3 g L 1 and a temperature of 23.8 26.9 °C. A mean of 10 larvae were removed from each tank on each of the first 4 days and every 2 days thereafter till 15 dph. After 15 dph a mean of 5 larvae were removed from each tank on 17, 19, 21, 24 and 26 dph. From 28 37 dph survivorship decreased within the culture tanks and sampling numbers were reduced to 3 larvae removed from each tank every 3 days. At 43 and 44 dph only 1 tank had surviving larvae, and remaining larvae were removed as they were dying. After all larvae were removed from each tank subsequent photographs (Jenoptik, ProgRes Capture Pro v2.8.8) and measurements (ImageJ 1.48r) were taken. Butterflyfish larvae were removed and immediately euthanized with buffered MS 222 (Western Chemical, Ferndale, Washington). Images were analyzed for the presence of inflated swim bladders, the initiation of flexion and the formation of the thiolichthys plates. Notochord length as previously defined will be measured for larvae preflexion. Post flexion notochord length is defined as a straight line measurement from the most anterior point of the head to the edge of the hypurals (DiMaggio, 2012). Total length (TL) measurements will be taken

PAGE 44

44 for post flexion larvae as well. Larvae will be characterized throughout development similarly to those of Leu et al. (2013). Results Spawning Behavior Within each group courtship behavior varied slightly based upon tank size and group size but the general behaviors remained constant. The spawning group consisted of a single male and a single female. The female was easily recognized by her swollen abdomen while the male was tagged with orange elastomer dye in the caudal fin. Broodstock in the circular tanks spawned naturally from December 28, 2013 July 2, 2014 when data for this study was no longer collected. Broodstock in the square tanks spawned naturally from March 2, 2014 until July 2, 2014 when data was no longer collected. Timeline ~4 h 00 min (16:04) The dominant male periodically displays aggression towards other males. The abdomen appears to be swollen. ~3 h 00 min (17:04) Group schools together around structure. Generally the gravid female remains lower in the water column than the other fish. The male stays close to female. ~1 h 30 min (18:34 ) The gravid female remains close to bottom of the tank with the male hovering over her. The male quickly chases away any advancing fish. ~1 h 00 min (19:04) The spawning female begins to separate from the remaining fish. The male breaks from the school to follow the female and follows slightly behind and above the female. The pair sometimes separates and the male will return to the group of fish, however, the female remains at a distance from the rest of the school.

PAGE 45

45 ~45 min (19:19) The pair begins to separate from the group. The male makes frequent runs with the female, sometimes chasing her around the tank. The male also begins. ~30 min (19:34) The m ale becomes increasingly aggressive towards other males within the tank, chasing them away from himself and the female. The male now chases the female from behind and below. Swimming speed has increased as fish dart across the tank. ~15 min (19:49) The male defends the female from other males, placing himself between her and them. The male tries to court the female, brushing his snout along her abdomen but must chase intruding males away. ~10 1 min (19:54 20:03) The male and female attempt spawning rises but are broken up by intruding males darting into the act. The male swims behind and below the female as they rise towards the waters surface but breaks away from the rise to chase the pursuing fish away. 0 min (20:04) The pair darts towards the surface with the male following just behind the female. As they reach the surface the pair face abdomens together with the male slightly below the female and release gametes into the water column. A hazy cloud appears near the surface and the other fish in the tank rush into the cloud. Other males release milt to potentially fertilize eggs. After spawning (20:24) There is still a lot of energy within the tank and fish continue to chase and dart throughout the tank for approximately 20 minutes before settling down.

PAGE 46

46 2 Courtship behavior started approximately 4 h before the act of spawning. The female was fairly inactive for the first two h. The male warded off other fish however it was not until ~1 h before spawning that the male became truly aggressive. The male and female were much more active in the later stages of courtship behavior with the male following the female closely and sometimes aggressively chasing her. As spawning approached preemptive spawning rises were attempted, howev er, during most efforts intruding males broke up the rise. The male followed the female very closely, often nudging his snout against her caudal region. The act of the spawning rise started with an extremely fast downward drop then rapid approach towards the surface. The male touched his body to the body of the female as the females back broke the surface of the water. Both fish tilted slightly to allow their urogenital openings to face each other and downward and then gametes were released. The male quickly moved in front of the female to defend the cloud of gametes against intruding fish, however, the remaining fish in the tank rush in and males release there gametes while females appear to eat some of the eggs. As the gamete cloud dissipates the fish reduce their quick, darting swimming and begin to school again ( Figure 2 2 ). Spawning Conditions A logistic regression was performed to ascertain the effects of lunar phase and photoperiod on the likelihood that a spawn will occur in the 1 M:1 F group (Model 1). The logistic regression model was statistically significant 2 =6.72, P =0.035). The model explained 4.8% (Nagelkerke R 2 ) of the variance in spawning and correctly classified 60.4% of cases. Increasing day length was the only significant variable associated with the increased probability of a spawn occurring ( Table 2 1 ).

PAGE 47

47 2 2 The logistic regression model for the 3 M:8 F (Model 2) group was statistically significant 2 = 7.28, P =0.026). The model explained 5.2% (Nagelkerke R 2 ) of the variance in spawning and correctly classified 60.1% of cases. Likeliness of a spawn occurring was significantly associated with increasing day length. Increasing Lunar phase was not significantly associated the increasing likelihood of a spawn occurring ( Table 2 1 ). A third logistic regression was performed to establish the effects of lunar phase and photoperiod on the likelihood of a spawn in the 10 M:11 F group (Model 3). Statistical significance was detected in the logistic regression 2 =15.36, P <0.001). Model 3 correctly categorized 65.5% of cases and explained 16.6% (Nagelkerke R 2 ) of spawning variance. A significant association between increasing day length and decreasing likelihood of a spawn was indicated from the model. Lunar phase was not significantly associated with increased likelihood of a spawn ( Table 2 1 ). Spawn data overlapped with lunar phases and photoperiod for each population (1 M:1 F, 3 M:8 F, 10 M:11 F) are represented in Figure 2 3 . Fecundity A 1 mL sample of eggs contained 1419 1869 eggs mL 1 (mean ± SD = 1628 ± 123 eggs mL 1 , n =20). Practical fecundity measures of the three populations were assessed using a Kruskal Wallis one way analysis of variance with a pairwise multiple comparison to compare medians of the three populations. The medians of nonviable eggs were established to be significantly different (H 2 =130.26, P <0.001) between populations. The pairwise multiple comparison test determined that the median number of nonviable eggs from the 1 M:1 F group (90) was significantly less compared to the 3 M:8 F (3,245) and 10 M:11 F (4,329) populations. Significant differences (H 2 =152.239,

PAGE 48

48 P <0.001) were identified among viable egg medians between populations. The 1 M:1 F group (57) had significantly lower number of viable eggs than the other two populations (10 M:11 F 5,435 and 3 M:8 F 7,425 eggs). Total egg production was found to be significantly different (H 2 =149.29, P <0.001) among groups. The 1 M:1 F group (150 eggs) had significantly lower total egg production than the 3 M:8 F (11,726 eggs) and 10 M:11 F (10,275 eggs) populations. Fertilization rates were statistically different (H 2 =6.45, P =0.040) among populations. The number of eggs produced per fish of the total population was significantly greater (H 2 =105.84, P <0.001) in the 3 M:8 F (1,066 eggs) when compared with the number of eggs produced by the 10 M:11 F (489 eggs) and 1 M:1 F (75 eggs) populations ( Table 2 2 ). Hydrated Egg Characteristics and Larvae Evaluation Morphological measures of eggs were not significantly different between populations of milletseed butterflyfish (F 1 =2.58, P =0.0973). Since there was no statistical difference between populations, samples were combined to increase the likelihood of an association between egg morphological c haracteristics and hatch success, survival, and notochord length. The Spearman rank correlation was run to assess the relationship between egg diameter (ED), oil globule diameter (OD), yolk volume (YV), ratio of oil globule diameter to egg diameter (OD:ED), egg volume (V), ratio of yolk volume to egg volume (YV:V), and new hatch notochord length (NHL), 3 dph notochord length (3L), hatch success (% hatch), 0 dph survival and 3 dph survival. All statistical results are reported in Table 2 3 . Egg diameter and egg volume were significantly positively correlated with 3 dph notochord length (r 133 =0.24, P <0.001) and (r 133 =0.24, P <0.001), respectively. A significant positive cor relation existed between OD and 0 dph survival (r 148 =0.54,

PAGE 49

49 P <0.001) and between OD and 3 dph survival (r 148 =0.38, P <0.001). Similarly yolk volume was significantly positively correlated to 0 dph survival (r 148 =0.53, P <0.001) and 3 dph survival (r 148 =0.40, P <0.001). Ratios of OD:ED and YV:V exhibited significant positive correlations between 0 dph survival (OD:ED r 148 =0.58, p <0.001 and YV:V r 148 =0.63, P <0.001) and 3 dph survival (OD:ED r 133 =0.41, P <0.001 and YV:V r 133 =0.49, P <0.001). Hatch success was positively correlated to 0 dph survival (r 148 =0.33, P <0.001). A strong significant positive correlation was observed between 0 dph survival and 3 dph survival (r 148 =0.80, P <0.001). Embryology C. miliaris produced small, pelagic, transparent, and spherical fertilized eggs measuring 0.68 0.73 mm (mean ± SD = 0.707 ± 0.008 mm, n =2400) in diameter. Eggs had a clear chorion with a single oil globule measuring (mean ± SD = 0.166 ± 0.006 mm, n =2400) at the vegetal pole, a homogenous and unsegmented yolk (mean ± SD = 0.117 ± 0.009 mm 3 , n =2400), and a narrow perivitelline space. At 25.5 ° C The egg remained in the zygote period (1 cell stage) for 37 min before the first cleavage occured dividing the cell in two. Eggs reached the 4 cell stage 61 min post fertilization (pf) ( Figure 2 4 A ), the 16 cell stage in 93 min pf, and the blastula stage 1 h 55 min pf. Eggs reached the gastrula stage approximately 7 h 30 min pf ( Figure 2 4 B ) and the segmentation stage 14 h 30 min pf ( Figure 2 4 C ). Larvae developed 27 somites with leucophores present on the body and head and were ready to hatch at approximately 28 h pf ( Figure 2 4 D ). Incubated at 25.5 °C most eggs hatched within 27 30 h pf. Larval Development At 0 h post hatch (hph) larvae measured 1.20 1.24 mm (mean ± SD = 1.221 ± 0.015 mm, n =30) notochord length, with an elliptical yolk sac ( Figure 2 5 A ). The oil

PAGE 50

50 globule varied in location from the ventroanterior portion of the yolk sac to the ventroposterior portion of the yolk sac and measured (mean ± SD = 0.166 ± 0.006 mm, n =30). The mouth remained closed with a thin, unopened, undifferentiated digestive tract running along the ventral side of the notochord. Majorities of branched leucophores were present along the ventral margin of the notochord, with small numbers present along the dorsoposterior portion of the body and very few present on the head. Larvae had 14~15+12=26~27 myomeres. At 24 hph larvae had grown rapidly measuring 2.171 2.198 mm (mean ± SD = 2.191 ± 0.011 mm, n =30) but are still incompletely developed. The mouth was still not open, eyes remained unpigmented, and the digestive tract remained closed. The yolk was approximately 2/3 its original size, with leucophores present on the ventral side of the yolk sac, the head, and ventroposterior portion of the notochord and surrounding the oil globule ( Figure 2 5 B ). At 2 dph larvae measured (mean ± SD = 2.217 ± 0.095 mm, n =30). Larvae did not exhibit much notochord growth, however, the expansions of the dorsal and ventral fin folds made larvae appear deeper. The mouth remained closed but the digestive tract was open at the anus. Eyes were still unpigmented and the yolk sac had reduced in size. Melanophores developed along the ventral margin of the notochord in alignment with the digestive tract, along the base of the yolk sac, and on the anterior portion of the head. Xanthophores developed on the head and along the dorsal and ventral axis of the notochord. At 3 dph nearly all of the yolk sac was depleted with just the oil globule remaining. Larvae measured 2.321 2.487 mm (mean ± SD = 2.403 ± 0.074 mm, n =30)

PAGE 51

51 but had not developed a functional mouth, or pigmented eyes yet. Pigmentation was similar to that of larvae 2 dph. At 4 dph larvae measured 2.384 2.539 mm (mean ± SD = 2.412 ± 0.031 mm, n =30), had fully pigmented eyes, and a functional mouth. Increased presence of melanophores amplified pigmentation along the ventral region of the notochord and posterior dorsal region of the notochord. Retinas were pigmented completely black while the cornea had a greyish blue hue. Gape height measured between 174 184 (mean ± SD = 178 ± 3.62 n =10) ( Figure 2 5 C ). At 7 dph larvae measured 2.420 2.631 mm (mean ± SD = 2.502 ± 0.077 mm, n =30), had completely absorbed the yolk, and began developing xanthophores around they eyes, at the tip of the mouth, and along the edges of the notochord ( Figure 2 5 D ). A clear separation between the posterior intestine and anterior intestine was present as well. At 9 dph larvae notochord length averaged (mean ± SD = 2.671 ± 0.043 mm, n =10) and larvae were similar in appearance to 7 dph larvae. Ten percent of larvae had inflated swim bladders at 9 dph ( n =30). Between 9 dph and 11 dph (mean ± SD = 3.58 ± 0.094 mm, n =30) larvae exhibited increased notochord growth, body depth growth, swim bladder inflation, and xanthophore formation. Branched xanthophores covered most of the body, with melanophores outlining the edges of the notochord ( Figure 2 5 E ). At 11 dph 70% of larvae had inflated swim bladders ( n =30). By 13 dph 90% of larvae had inflated swim bladders and larvae measured 3.309 3.680 mm (mean ± SD = 3.588 ± 0.15 mm, n =30) ( Figure 2 5 F ).

PAGE 52

52 At 15 dph larvae measured 3.754 3.889 mm (mean ± SD = 3.849 ± 0.054 mm, n =30 ). Finfold size decreased substantially and hugged the contours of the body. From 17 21 dph the body increased in depth and the dorsal portion of the body darkened. At 24 dph larvae measured 5.172 5.407 mm (mean ± SD = 5.319 ± 0.088 mm, n =15). Formation of the hypural plates began and the finfold collapsed inward forming a shape of a fin around the posterior end of the notochord ( Figure 2 5 G ). At 26 dph the body became deeper, especially in the area of the gut, and the ventral hypural plate was further developed than the dorsal hypural plate. Larvae measured 5.472 5.691 mm (mean ± SD = 5.648 ± 0.045 mm, n =15). Between 24 and 26 dph the thiolichthys plates begin to form. Three boney plates formed: 1) an elongated preopercular, 2) a post temporal, and 3) a supracleithrum plate ( Figure 2 5 H ). At 28 dph larval notochord length began to decrease because flexion of the notochord tip began. Larvae exhibited decreased growth (mean ± SD = 5.558 ± 0.038 mm, n =9) but increased body depth. The tholichthys plates increased in size ( Figure 2 5 I ). At 31 dph larvae measured 5.383 5.686 mm (mean ± SD = 5.464 ± 0.166 mm, n =9) from the tip of the snout to the edge of the hypural plates. Body depth increased in the gut region, eye diameter increased, and larvae were oval in shape. Fin rays became visible in the dorsal, caudal, and anal fins. Larvae had massive bony plates on their head. The supracleithrum plate appears as a bulbous bulge above the eyes, the post

PAGE 53

53 temporal plate is flat extending posteriorly from above the eye, and a large preopercular spine extended posteriorly to the edge of the pelvic fin girdle. Additionally, small spines began to develop in the anterior region of the dorsal fin ( Figure 2 5 J ). Between 31 44 dph larvae maintained a general ovoid shape. At 34 dph the dorsal and anal fins begin to appear lobular and the caudal fin was nearly as deep as the body. At 37 dph larvae measured 6.055 6.132 mm (mean ± SD = 6.082 ± 0.042 mm, n =9) and had slightly larger dorsal spines. At 44 dph larvae measured 6.495 6.563 mm (mean ± SD = 6.53 ± 0.025 mm, n =2) and were similar in appearance to 37 dph larvae ( Figure 2 5 K ). Subsequent culture trials of C. miliaris were attempted, however complete data was lacking. Additional observations demonstrated that in subsequent culture attempts larvae showed increased development measuring 6.438 mm from the tip of the snout to the end of the hypural plates and 8.390 mm total length at 35 dph. Larvae had increased dorsal spine growth with the most anterior spine reaching almost 1 mm in length ( Figure 2 5 L ). Larvae did not eat newly hatched Artemia nauplii or Brachionus sp. rotifers but selectively preyed on the copepod ( P. crassirostris ) throughout development, and were capable of capturing copepodites by the end of the larval culture attempt. Water Quality Water quality parameters remained within acceptable normal limits ( Table 2 4 ). Discussion This is the first known attempt at spawning and larval culturing of Chaetodon miliaris . Spawning was voluntary and handling was limited to a few occasions,

PAGE 54

54 minimizing stress. Additionally, C. miliaris adults quickly adapted to their captive diet. From this study, reproduction of C. miliaris seems to be achievable in captivity with limited conditioning and environmental manipulation. Spawning Behavior Spawning behavior was similar between all three groups within this study. Observations of C. miliaris spawning were similar to spawning described by Lobel (1989) for Chaetodon multicinctus and Suzuki et al. (1980) for Chaetodon nippon . Pairing between the dominant male and a female seemed to occur with C. miliaris , however, during all observed spawning events additional males were present ( Figure 2 2 ) except for the 1 M:1 F group. Depending on the species, it appears that some butterflyfish form mating groups while others form individual pairs (Suzuki et al., 1980; Lobel, 1988; Colin and Clavijo, 1988; Colin, 1989; Yabuta and Kawashima, 1997). Spawning observations suggest that reproduction of C. miliaris , at least in aquarium conditions, involved one direct male and several ind irect males. Comparisons of spawning behavior between captive and wild butterflyfish do not appear to substantially differ. Descriptions of wild spawning behavior generally consisted of pairs or groups of fish swimming over the reef with occasional courtship behavior where the male followed the female closely often nudging the females caudal region with his snout. Then several preemptive spawning rises occurred before the final rise and gamete release at the top of the ascent (Colin and Clavijo, 1988; Colin, 1989; Lobel, 1989; Yabuta and Kawashima, 1997). Captive spawning behavior of C. miliaris , as described in this study, C. modestus (Tanaka et al., 2001), and C. nippon (Suzuki et al., 1980) generally began with the female resting on the bottom of the tank, followed by the dominant male and female

PAGE 55

55 rising and swimming circularly around the aquarium. The male frequently attempted to deter following males. Several practice ascents occurred before the male and female ascended upwards and at the top of the ascent released their gametes. Intruding males rushed in and released their gametes into the spawn cloud ( Figure 2 2 ). Spawning Conditions Spawning events were observed from the 1 M:1 F group on 83 days out of the 184 days of this study ( Figure 2 3 A ), on 86 days out of the 184 days from the 3 M:8 F group ( Figure 2 3 B ), and on 62 days out of the 184 study days from the 10 M:11 F group ( Figure 2 3 C ). Similar to Ralston (1981) lunar periodicity was shown to have no influence on spawning for all three spawning groups ( Table 2 1 ). In the 1 M:1 F and 3M:8F increased photoperiod increased the likelihood of a spawning event occurring, while to the contrary the 10 M:11 F population showed that increasing photoperiod decreased the likelihood of a spawn ( Table 2 1 ). These results are contradictory, however, this is most likely due to the fact that the 10 M:11 F group did not begin spawning until March 2014, whereas, the 1 M:1 F and 3 M:8 F groups began spawning in January 2014. Spawning of the 1 M:1 F and 3 M:8 F populations was infrequent during the first two months when photoperiods were below 11 h light with only 19 spawns produced from each population. Collection of data over a longer period of time would better address the lack of spawning in January and February observed in the 10 M:11 F group, as well as, determine if lower spawning incidence by the 3 M:8 F and 1 M:1 F group was based upon photoperiod or acclimation to tank conditions. In this study, increasing photoperiod may be a trigger for C. miliaris to begin spawning, however, without further data

PAGE 56

56 collection, analyses only indicate partial trends and are not indicative of yearly reproductive cycles of C. miliaris . It appears that all three populations followed similar trends during the months of March, April, May, and June . Frequency of spawning began to increase in March for all three groups with 15 spawns occurring in the 1 M:1 F and 3M:8F groups and 18 spawns from the 10 M:11 F group. In April the 1 M:1 F, 3 M:8 F, 10 M:11 F groups produced 14, 16, and 17 spawns respectively. In May the 1 M:1 F and 3 M:8 F groups produced over 20 spawns each while the 10 M:11 F group spawned 17 times and in June the number of spawns were decreased in all three groups (1 M:1 F 13 spawns, 3 M:8 F 13 spawns, and 10 M:11 F 10 spawns) ( Figure 2 3 ). Spawning season during this study matched that of the wild spawning season measured by Ralston (1981). He found that C. miliaris gondasomatic indices were highest from January June and decreased thereafter matching observed spawning during this study. Comparable spawning seasons to Ralston (1981) and this study were observed in 5 other chaetodontid species in Hawaii and the Marshall Islands (Lobel, 1978). It is possible that a combination of factors such as temperature, photoperiod, and adequate food resources contribute to C. miliaris spawning in the wild. With sufficient food resources it is likely that manipulation of photoperiod and temperature could increase the spawning season in captivity. Fecundity Fecundity data indicated that spawn size seemed to be reliant on additional males intruding on the spawn. The 1 M:1 F group produced the lowest total egg count and had a higher median number of non viable eggs than viable eggs ( Table 2 2 ). The 1 M:1 F group should not be considered for any scale of production for culturing

PAGE 57

57 milletseed butterflyfish. The 3 M:8 F group produced the most viable eggs and the most eggs per fish but had the low est fertilization rate ( Table 2 2 ). The 1 M:1 F group produced 12,566 viable eggs over the 184 day spawning season while the 3 M:8 F group produced a total of 747,009 viable eggs during the 184 day spawning season with only 68,060 viable eggs produced during January and February. The 10 M:11 F group produced 475,907 viable eggs during the 184 day spawning season. Further replicated evaluation of egg production of different sized C. miliaris broodstock groups would be beneficial for a more accurate analysis. The use of two different egg collection methods in this study due to the broodstock tank restrictions may have influenced the number of eggs collected from each population. Additionally, the increased number of non spawning fish in the 10 M:11 F group may have consumed more eggs during spawning than did the non spawning fish of the 3 M:8 F population. The fertilization rate 64.08 100% of all C. miliaris groups was comparable to other marine ornamental species: Pomacanthus semicirculatus mean = 68.1% (Leu et al., 2009), Centropyge debelius mean = 19% (Baensch and Tamaru, 2009), Orthopristis chrysoptera mean = 98% (DiMaggio et al., 2013), and Lutjanus campechanus mean = 83.9% (Papanikos et al., 2008). Ralston (1981) thought that females produced one batch of eggs per spawning season, however, this study indicates that C. miliaris females undergo asynchronous oogenesis producing multiple batches of eggs throughout the spawning season. Continual spawning and high fecundity of a C. miliaris are ideal characteristics of a candidate species for commercial production.

PAGE 58

58 Hydrated Egg Characteristics and Larvae Evaluation There was no significant difference between egg morphological characteristics, early larvae survival, or notochord of the 3 M:8 F and 10 M:11 F groups. Since all fish came from the same original population, were approximately the same size, and were fed the same diet eggs were consistent between groups. The data collected from C. miliaris eggs and early larvae showed that many associations were present between the two ( Table 2 3 ). Relationships between egg characteristics and 3 dph survival were more powerful compared to associations between egg characteristics and 0 dph survival ( Table 2 3 ). All significant associations indicate that increasing size of egg characteristics were correlated with increased growth and survival of larvae. Literature showed that increased size of egg characteristics resulted in increased early larval size, however, increased development cannot be predicted beyond early larval stages (Springate and Bromage, 2003; Kennedy et al., 2007; Imanpoor and Bagheri, 2010). The relationships between egg characteristics and 3 dph survival were strongest because larvae with increased size have more nutritional reserves compared to smaller larvae making them capable of withstanding harsher conditions (Imanpoor and Bahher, 2010). Hatch success was not sig nificantly associated with any egg morphological characteristics. The strongest positive association was between 0 dph survival and 3 dph survival. From a production standpoint this is crucial information as the number of larvae surviving to first feeding can be estimated from the number of larvae surviving on 0 dph.

PAGE 59

59 Embryology Characteristics of fertilized eggs of C. miliaris bear resemblance to other Chaetodon species previously studied. The diameters of fertilized eggs (0.684 0.729 mm) were comparable in size to other butterflyfish species ( Table 2 5 ). Time to hatching, TL at hatching, and TL at first feeding were consistent among all species with collected data ( Table 2 5 ). C. miliaris embryonic development was the same as other chaetodontids with variations in developmental timing (Suzuki et al., 1980; Tanaka et al., 2001). Larval Development Larvae of C. miliaris exhibited similar early larval development for described species of butterflyfish (Suzuki et al., 1980; Tanaka et al., 2001). Culturing of the schooling bannerfish, Heniochus diphreutes , to metamorphosis has been completed but a description of larval development has not been released, so comparisons of C. miliaris to other Chaetodon species are only to 8 dph. Early larval stages of the two other described species of butterflyfish ( C. modestus and C. nippon ) resemble early larval development of C. miliaris (Suzuki et al., 1980; Tanaka et al., 2001). Newly hatched larvae had a large yolk sac extending beyond the anterior tip of the head, unpigmented eyes, an undifferentiated digestive tract, and an unopened mouth. At hatch, C. miliaris had 26~27 myomeres, while C. nippon and C. modestus had 25 myomeres. Larval growth was consistent between species with length almost doubling by first feeding, however, the opening of the mouth, pigmentation of the eyes, and capability of feeding did not occur until 4 dph for C. miliaris while it occurred on day 3 for other species ( Table 2 5 , Figure 2 5 C ). Interestingly, newly hatched C. miliaris had varying positions of the oil globule on the

PAGE 60

60 yolk sac, while other references of marine butterflyfish and angelfish state that the oil globule was present in the ventroposterior position of the yolk sac (Suzuki et al., 1980; Tanaka et al., 2001; Leu et al., 2009; Leu et al., 2013). Between 4 and 9 dph larvae looked similar. At 7 dph, xanthophores began to develop on the tip of the snout and along the edges of the notochord. Melanophores also were present along the edges of the notochord. The yolk sac and oil globule were completely absorbed at 7 dph matching descriptions of other chaetodontids (Tanaka et al., 2001). There were clear signs of differentiation within the gut with the pyloric valve forming between the mid gut and hind gut. Differentiation of the gut has not been documented in chaetodontid larvae previously and timing varies compared to other species (DiMaggio et al., 2013; Leu et al., 2013). At 9 dph swim bladder inflation had begun with approximately 10% of larvae having inflated swim bladders. At 11 dph larvae exhibited increased growth, swim bladder inflation (70%), and xanthophore formation. Xanthophores covered most of the body by 11 dph. At the time of swim bladder inflation a large mortality (10 30%) was observed within the culture tanks. Observation of overinflated swim bladders in larvae looked like a possible bacterial infection, likely from bio film accumulation at the air water interface, may be responsible for increased mortality at this time (DiMaggio et al., 2013). The timing of the first mortality of C. miliaris was similar to the timing of the first mortality experienced by Wittenrich and Cassiano (2011) with Heniocus diphreutes . At 13 dph, 90% of larvae sampled had inflated swim bladders. Larvae were similar in appearance from 15 21 dph. Body depth increased noticeably in the gut region and the dorsal region of the larvae was darkened by branched melanophores.

PAGE 61

61 At 24 dph larvae measured 5.17 5.40 mm and began to develop hypural plates along the posterior tip of the notochord. Leu et al. (2013) described similar development in the angelfish Chaetodontoplus septentrionalis . Larvae increased body depth prior to the formation of the hypural plates and beginning of flexion. Between 24 26 dph larvae developed the thiolichthys plates on the head. Leis (1989) described the thiolichthys sta ge as a round, deep, and compressed body with fused head plates. The thiolichthys plates cover the head with posteriorly extending broad, flat, more or less blunt, rugose plates originating from the post temporal and supracleithrum dorsally and the preoperculum ventrally (Johnson, 1984). Leis (1989) described two other variations in head shape and spination ( Figure 2 6 ). Chaetodon miliaris appears to form a modified thiolichthys stage with a large post temporal plate extending posteriorly reaching the base of the dorsal fin, a bulbous supracleithrum plate between the eyes, and elongated spiny preopercular plate ( Figure 2 6 C ). Another large mortality was observed during the flexion period. At 28 dph the notochord tip rose to a 45° angle and the hypural plates, especially the ventral plate, formed. Larvae had undergone flexion and significantly increased body depth obtaining an ovoid shape by 31 dph. Eye diameter had also increased and fin rays were present in the dorsal, caudal, and anal fins ( Figure 2 5 J ). Larvae exhibited some dorsal spine growth between 31 44 dph, however, there was very little other development ( Figure 2 5 K ). In ensuing larval trials increased growth was observed at 35 dph ( Figure 2 5 L ). In this trial, a micro diet (Skretting, Gemma Micro 150, Tooele, Utah), was added to the culture trial starting at day 15, however, larvae were never

PAGE 62

62 observed to feed on the micro diet particles. Growth curves of notochord and total length of C. miliaris are represented in Figure 2 7 . Spawning behavior indicates that population size does affect spawning success and that groups greater than 1 male and 1 female increased egg production. A ratio of 3 M:8 F produced the most consistent spawning and largest number of eggs. Data on fecundity and spawning conditions show that C. miliaris has a strong aquaculture potential for marine ornamental aquaculture since egg production over the six month trial period was fairly consistent and only required temperature manipulation. It is possible that with temperature and photoperiod control, that natural spawning could be continuous for longer periods of time. Embryology of C. miliaris was similar to that of other butterflyfish species. In the present study C. miliaris larvae did not survive longer than 44 days. According to the only successful culturing of a butterflyfish species (Baensch, 2014), C. miliaris was close to reaching metamorphosis. The growth of the dorsal spine was the last stage of the larval phase before metamorphosis in Heniochus diphreutes and sizable growth of the dorsal spine was present in C. miliaris before the larvae died.

PAGE 63

63 Table 2 1. Effects of lunar phase and photoperiod on the likelihood of a spawn occurring based upon a logistic regression. The predictive power and model equation are displayed for each overall model. The odds ratio, parameter estimate, Wald statistic and significance are portrayed for each individual parameters. * Denotes statistical significance within the model at Model Equation Predictive % Parameter Odds ratio Estimate Wald P 1 ____1____ 60.4 Lunar phase 1.259 0.231 0.212 0.645 1+e 4.268+.231X +.321X 1 2 Photoperiod 1.379 0.321 6.356 0.012* Constant 0.014 4.268 6.983 0.008* 2 ____1____ 60.1 Lunar 1.840 0.610 1.483 0.223 1+e 4.199+.610X +.305X 1 2 Photoperiod 1.356 0.305 5.766 0.016* Constant 0.015 4.199 6.769 0.009* 3 ____1____ 65.5 Lunar 3.142 1.145 2.949 0.086 1+e 12.823+1.145X + 1 Photoperiod 0.366 1.005 9.892 0.002* 1.005X 2 Constant 370,781.100 12.823 9.100 0.003* Table 2 2. Practical fecundity measures of C. miliaris populations presented as median ± MAD per spawn. Upper case letters represent statistical differences between populations. Population Nonviable Eggs Viable Eggs Total Eggs Fertilization (%) Eggs Fish 1 (total population) 1M:1F 90 ± 47 B 57.00 ± 27 B 150 ± 77 B 100.00 ± 0.00 A 75 ± 38 C 3M:8F 4,329 ± 1,953 A 7,425.00 ± 4,040 A 11,726 ± 4,533 A 98.05 ± 1.89 B 1,066 ± 412 A 10M:11F 3,245 ± 2,452 A 5,435.00 ± 3,335 A 10,275 ± 4,834 A 99.00 ± 0.93 AB 489 ± 230 B

PAGE 64

64 Table 2 3. Spearman rank correlation coefficients for all compared morphological characteristics. ** Denotes statistical significance at P 0.01 Variables NHL 3L % Hatch 0 dph survival 3 dph survival ED 0.073 0.244** 0.860 0.108 0.160 OD 0.017 0.104 0.129 0.388** 0.548** YV 0.102 0.162 0.041 0.405** 0.534** V 0.071 0.242** 0.091 0.100 0.157 OD:ED 0.006 0.038 0.150 0.413** 0.582** YV:V 0.111 0.066 0.119 0.497** 0.630** NHL 0.007 0.045 0.098 0.076 3L 0.046 0.070 0.100 % Hatch 0.333** 0.150 0 dph survival 0.807** 3 dph survival

PAGE 65

65 Table 2 4. Water quality variables measured within the culture trial. Salinity, temperature, dissolved oxygen (DO), pH, total ammonia nitrogen (TAN), nitrite nitrogen (NO 2 N), and nitrate (NO 3 ) were measured daily. Values are given as the mean ± standard deviation and the range. The number of samples ( n =) for the trial is given. Salinity (g L 1 ) Temperature (°C) DO (% Saturation) pH TAN (mg L 1 ) NO 2 N (mg L 1 ) NO (mg L 1 ) 3 Culture Trial n =81 34.26 ± 0.62 33.20 35.30 25.36 ± 0.81 23.80 26.90 97.32 ± 1.14 94.70 99.33 8.15± 0.05 8.10 8.20 0.069 ± 0.03 0.030 0.11 0.002 ± 0.01 0.000 0.03 0.007 ± 0.004 0.000 0.010 Table 2 5. Comparison of eggs and larvae of chaetodontids Species Egg diameter (mm) Incubation temperature (°C) Hatching time (h) TL at hatch (mm) TL at first feeding (mm) Time to first feeding (days) Resource C. nippon 0.70 0.74 22.2 23.7 28 30 1.43 1.53 2.46 2.48 3 Suzuki et al., 1980 C. modestus 0.75 0.80 25.8 28.8 16 1.43 1.58 2.39 2.52 3 Tanaka et al., 2001 C. aculeatus 0.74 0.76 25.0 26 36 Colin, 1989 C. capistratus 0.76 0.77 Colin, 1989 C. ocellatus 0.60 0.70 Colin, 1989 C. striatus 26.0 30 3 Colin, 1989 C. miliaris 0.69 0.73 23.8 26.9 26 30 1.38 1.45 2.41 2.56 4 This study

PAGE 66

66 A 40.00 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00 1/1/14 2/1/14 3/1/14 4/1/14 5/1/14 6/1/14 7/1/14 Date Temp °C Salinity (ppt) pH B 40.00 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00 1/1/14 2/1/14 3/1/14 4/1/14 5/1/14 6/1/14 7/1/14 Date Temp °C Salinity (ppt) pH C 40.00 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00 3/9/14 4/9/14 5/9/14 6/9/14 Date Temp °C Salinity (ppt) pH Figure 2 1. Temperature (°C), salinity (g L 1 ), and pH throughout the spawning period for the (A) 1 M:1 F, (B) 3 M:8 F, (C) 10 M:11 F populations. pH/g l 1 /°C pH/g l 1 /°C pH/g l 1 /°C

PAGE 67

67 Figure 2 2. A diagrammatic view of the spawning behaviors of C. miliaris in the broodstock tank; (1) Approximately 3 hours before spawning the spawning pair began to separate from the schooling fish. The dominant male (M) hovers over the gravid female (shaded fish) whil e the female remains close to the tank bottom; (2) About 1 hour prior to spawning the male and female begin to separate from the school, swimming throughout the tank as a pair. The male will return to the school briefly before rejoining the female; (3) Fifteen minutes prior to spawning the pair begins to make preemptive spawning rises towards the surface of the water. They move in a rapid, darting motions. The dominant male (M) often breaks away from the spawning rise to chase off intruding males following the pair before returning to the female to continue the attempted rises; and (4) Just before spawning the pair makes a swooping rise towards the surface. The male presses his body against the females directing her towards the surface. Once at the surface the pair immediately release gametes. The other males in the tank instantaneously race into the cloud releasing their own sperm and females appear to ingest some eggs. The dominant male (M) tries to defend against the intruders but is unsuccessful in deterring the males.

PAGE 68

68 Photoperiod (L) Eggs Spawned Lunar Cycle Photoperiod (L) Eggs Spawned Lunar Cycle Photoperiod (L) Eggs Spawned Lunar Cycle A 14000 16 12000 14 10000 12 8000 10 8 6000 6 4000 4 2000 2 0 0 1/1/14 2/1/14 3/1/14 4/1/14 5/1/14 6/1/14 7/1/14 Date B 100000 16 14 80000 12 60000 10 8 40000 6 20000 4 2 0 0 1/1/14 2/1/14 3/1/14 4/1/14 5/1/14 6/1/14 7/1/14 Date C 50000 16 14 40000 12 30000 10 8 20000 6 10000 4 2 0 0 3/9/14 4/9/14 5/9/14 6/9/14 Date Figure 2 3. Lunar phases and photoperiods overlapped with spawning data for (A) 1 M:1 F, (B) 3 M:8 F, and (C) 10 M:11 F populations. Eggs Spawned Eggs Spawned Eggs Spawned Hours Hours Hours

PAGE 69

69 Figure 2 4. Embryonic development of C. miliaris eggs at 25.5 °C, (A) 4 cell stage 61 min post fertilization (pf), (B) formation of the germ ring in the gastrula stage 7 h 30 min pf, (C) formation of the first somite 14 h 30 min pf, and (D) 28 h pf embryo prior to hatching with 27 somites developed. A B C D

PAGE 70

70 Figure 2 5. Larval development of Chaetodon miliaris ; (A) a newly hatched larvae, 1.22 mm notochord length (SL); (B) 1 day post hatch (dph) larvae with unpigmented eyes and an unopened mouth, 2.19 mm (SL); (C) 4 dph larvae, eyes are pigmented, mouth is opened, and larvae are capable of feeding, 2.41 mm (SL); (D) 7 dph with absorbed yolk sac, 2.50 mm (SL); (E) 11 dph with inflated swim bladder and increased pigmentation, 3.58 mm (SL); (F) 13 dph with 90% of larvae with inflated swim bladders and differentiation between the mid and hind gut, 3.58 mm (SL); (G) 24 dph with hypural plate formation, 5.31 mm (SL); (H) 26 dph with further development of the hypural plates, flexion beginning, and larvae entering the thiolichthys stage, 5.64 mm (SL) and 5.80 mm total length (TL); (I) 28 dph with notochord tip at a 45° angle and decreasing notochord length due to flexion of the notochord tip, 5.55 mm (SL), 5.95 mm (TL); (J) 31 dph with increased body depth, eye diameter, an oval shaped larvae, and large thiolichthys plates . Fin rays are visible, 5.46 mm (SL), 6.52 mm (TL); (K) 44 dph larvae with lobular fins and dorsal spine growth, 6.530 mm (SL), 8.17 mm (TL); and (L) 35 dph with increased dorsal spine growth. A B C D E F

PAGE 71

71 Figure 2 5. Continued. G H I J K L

PAGE 72

72 Figure 2 6. Thiolichthys stages from Leis, J.M., 1989. Larval biology of butterflyfishes (pisces, chaetodontidae): What do we really know? In: The butterflyfishes: Success on the coral reef. Envi. Bio. Fishes. 125, 87 100; (A) Coradion sp., with a conventional percoid head spination including serrate spines along the preopercular border, a very large, sharp serrate spine at the preopercular angle, serrate infraorbital, dentary, supraocular and supraoccipital ridges, and reduced post temporal and supracleithral spination; (B) The classic thiolichthys stage with the head covered by fused plates which extend posteriorly over the trunk in the form of broad, flat, more or less blunt, rugose plates originating from the post temporal and supracleithrum dorsally and the preoperculum ventrally; (C) A Chaetodon sp., variation with an elongate dorsal and pelvic fin spine, a rounded supraoccipital crest, and an elongated sharp preopercular sp ine.

PAGE 73

73 9 8 7 6 5 4 3 2 1 0 0 10 20 30 40 50 Days Post Hatch (dph) Notochord Length Total Length Figure 2 7. Larval growth of C. miliaris . Notochord length was measured from the tip of the snout to the posterior tip of the notochord prior to flexion and from the tip of the snout to the edge of the hypural plates after flexion. Total length was measured from the tip of the snout to the end of the caudal fin. All values are represented as mean ± SD. For observations on 0 15 dph ( n =30), for observations on 16 26 dph ( n =15), for observation on 27 37 dph ( n =9), and for observations on 38 44 dph ( n =2). Larval Length (mm)

PAGE 74

74 CHAPTER 3 FIRST FEEDING PARAMETERS OF Chaetodon miliaris Foreword First feeding in larval fishes represents a major bottleneck for many species. Marine ornamental larvae generally develop functional eyes and mouths, and begin feeding 48 96 hours post hatch (hph) (Suzuki et al., 1980; Leis, 1989; Holt and Riley, 2000; Holt et al., 2007; Baensch and Tamaru, 2009; Olivotto et al., 2011; Callan et al., 2012a; Leu et al., 2013; Zavala Leal et al., 2013). For all butterflyfish species currently examined, first feeding occurred between 72 96 hph (Suzuki et al., 1980; Tanaka et al., 2001; Wittenrich and Cassiano, 2011; Baensch 2014). A variety of parameters affect the success of feeding in early larvae. Marine food fish aquaculture is subsistent on traditional live feeds Brachionus spp. rotifers and Artemia spp. nauplii, however the small size of first feeding pelagic marine ornamental species limits the use of traditional feeds and necessitates smaller, natural feeds (Holt, 2003; Mckinnon et al., 2003; Sampey et al., 2007; Baensch and Tamaru, 2009; Leu et al., 2013; Zavala Leal et al., 2013). Alternative feeds include copepod nauplii, ciliates, dinoflagellates, tunicate larvae, algae, and other various wild planktons which have been studied primarily in research scale experiments (Kraul, 1989; Reitan et al., 1997; Moorhead and Zeng, 2010; Olivotto et al., 2011). Particularly in the marine ornamental industry a variety of alternative feeds are being examined to replace more traditional live feeds. Perhaps the largest impediment to understanding first feeding in marine ornamental larvae is the lack of data from wild larvae. While some work has been published (Mckinnon et al. , 2003; Sampey et al., 2007) examining wild larval diets, data is limited to only a few species and also limited

PAGE 75

75 by sampling methods and sizes, prey identification, and the magnitude of the pelagic environment. Furthermore, the isolation and culturing of zooplankters is poorly understood. Many similar complications to those observed with marine ornamental larvae culture exist in the culture of zooplankton species (Drillet et al., 2011). Copepods are known to be a significant nutrient pathway from primary producers to fish larvae and as a feed item of many pelagic larvae (Hunter, 1981; Stottrup, 2003; Cassiano et al., 2011). Constraints such as size, locomotion, swimming speed and escape response further limit the number of prey species that coincide with larvae capture competence (Buskey, 1993; Stottrup and Norsker, 1997; Buskey et al., 2002; Conceicao et al., 2010; Cassiano et al., 2011). Optimizing capture success during early larval stages translates into a multitude of benefits as larval development occurs. Experiments to enhance first feeding in marine species are well documented for a variety of food fish species, but remain limited in marine ornamental species. Focus areas for amplifying first feeding success include prey density, larval stocking density, turbidity, light intensity, tank size, nutrition, and water exchange rates (Brownell, 1980; McGurk, 1984; Boehlert and Morgan, 1985; Schmitt, 1986; Duray and Kohno, 1988; Coughlin, 1991; Naas et al., 1992; Cook, 1996; Dower et al., 1997; Faulk and Holt, 2003; Carton, 2005; Battaglene et al., 2006; Pekcan Hekim and Lappalainen, 2006; Sanchez Hernandez et al., 2011). Yin and Blaxter (1987) examined the effects of delayed first feeding in herring, cod, and flounder through the endogenous to exogenous feeding transition. Survival of larvae during progressive starvation trials showed that larvae, those withheld from first feeding, reached a point where they did not have the strength to feed even after being introduced to feed. These

PAGE 76

76 conclusions emphasize the importance of first feeding in marine fish larvae as early as possible after the transition to exogenous feeding. Nutritional content is another obstacle in live feeds. Traditional feeds such as rotifers and Artemia are often enriched to meet the nutritional requirements of larvae. In marin e fish larvae essential fatty acids are crucial for development. Marine fish are not capable of elongating fatty acid chains and must acquire from the environment. In the wild the variety of food items and passively ingested particles supply these however, in culture, the feed is the sole method of EFA procurement for larvae. Proper ratios of highly unsaturated fatty acids specifically docosahexaenoic acid (DHA), eicosapentaenoic acid (EPA), and arachidonic acid (ARA), increase growth, survival, and stress resistance in larvae (Sargent et al., 1999). Methods To assess first feeding parameters of the milletseed butterflyfish eight single factor replicated experiments with 2 separate trials each were run to determine the proportion of larvae feeding and larval feeding intensity ( Table 3 1 ). Since there is limited information available on first feeding in butterflyfishes starting parameters were estimated based upon preliminary trials and current available knowledge (Suzuki et al., 1980; Tanaka et al., 2001; Wittenrich and Cassiano, 2011; Baensch, 2014). A review of first feeding literature was used to select specific parameters for the milletseed butterflyfish first feeding investigation including prey selectivity, algal cell density (high and low), prey density, larval stocking density, tank size, water exchange rate, and light intensity ( Table 3 1 ) (Brownell, 1980; McGurk, 1984; Boehlert and Morgan, 1985; Schmitt, 1986; Duray and Kohno, 1988; Coughlin, 1991; Naas et al., 1992; Cook, 1996;

PAGE 77

77 Dower et al., 1997; Faulk and Holt, 2003; Carton, 2005; Battaglene et al., 2006; Pekcan Hekim and Lappalainen, 2006; Sanchez Hernandez et al., 2011). For each experiment two trials were run with each variable having three to six treatment levels and a minimum of three replicates. Embryo Stocking and Culture C. miliaris embryos were obtained from broodstock tanks at the University of Florida Tropical Aquaculture Laboratory (Ruskin, FL). Broodstock spawned daily during this period allowing for experiments to be run consecutively. Sixteen spawns were collected during March April 2014. Embryos were assessed for viability. Fertilization was greater than 95% and hatching success greater than 97% for all cohorts. All culture parameters for C. miliaris are summarized in Table 3 1 . For Trials 1 and 2, eggs were stocked at 15 20 eggs L 1 in the experimental system 12 h post fertilization. Light aeration (<0.1 L min 1 ) supplied by a single cylindrical air stone (2.5 cm height x 2.5 cm circumference) in 14 L (30.5 cm diameter x 33.0 cm height) tanks and two rectangular (5.0 cm height x 2.5 cm width ) air stones in 128 L (61.0 cm diameter x 54.6 cm height) and 210 L (81.3 cm diameter x 54.6 cm height) tanks, resting on the bottom of the tanks throughout all experiments. All experimental tanks were fiberglass with black sides and a white bottom. Larvae were observed for abnormalities such as large mortalities , and larvae lying on the tank bottom during the first 3 dph before a predetermined value of Tisochrysis lutea (Tiso) was added to each tank at 3 dph. At 4 dph larvae had fully functional mouths, digestive tracts, and pigmented eyes indicating they were capable of feeding. Feed was added to each tank individually with a 5 minute interval between feedings to allow a 5 minute time frame to collect larvae (Hilder et al., 2014).

PAGE 78

78 Two channel Viavolt 54 watt 6500 K color temperature light banks (Phillips, Andover, MA) provided light. Using a lux meter (Milwaukee Instruments, Rocky Mount, NC), light intensity was measured at the waters surface in the center of each tank (MW 700 portable Lux meter). A photoperiod of 14 light and 10 dark h was used throughout all trials. At 4 dph notochord, body depth, and eye diameter was recorded ( n =20) from each experiment to ensure there was no size bias for an individual trial. Assessment of Feeding Response Randomly, 20 larvae from each replicate were harvested 6 h after feeding. All larvae were euthanized with an overdose of MS 222 (500 mg L 1 ) buffered with Sodium Bicarbonate (Western Chemical, Ferndale Washington) before harvesting each tank to ensure larvae did not regurgitate food. Larvae were removed from the collection beaker, rinsed with clean seawater, and preserved in 10% buffered formalin (Sigma Aldrich, St. Louis, MO) for analysis. Larvae are transparent at 4 dph making it possible to observe prey items inside the gut once larvae are between a c over slip and microscope slide. Additionally, the stomach wall ruptures enabling prey items to be counted (Hilder et al., 2014). The presence of copepods in the stomach was enumerated by counting individual exoskeletons ( Figure 3 17 ), while rotifers were counted by their individual mastax. The proportion of larvae feeding and feeding intensity (prey item larvae 1 6 h 1 ) were recorded as mean ± SE ( n =3 6). Proportion feeding =feeding larvae/sampled larvae (20) *100 The six hour feeding time period was used to ensure statistical differences in feeding intensity would be capable of being detected. Feeding intensity = prey items ingested per feeding larvae

PAGE 79

79 Feeding Experiments 3 1 Prey selectivity To assess first feeding prey selectivity of the milletseed butterflyfish two commercially available live feeds the rotifer, Brachionus plicatilus, and copepod, Parvocalanus crassirostris , were fed to larvae at 4 dph. Eggs were stocked into twenty four 14 L tanks. Tanks were kept static during experimental runs. Lighting was kept at (mean ± SD) 1,314 lx. Two hours prior to feeding on 4 dph Tiso was added to 12 of the 24 experimental tanks to adjust algal density from 400,000 600,000 cells mL 1 . At 4 dph feed (5 7 nauplii mL 1 ) and (5 7 rotifers mL 1 ) ( n =6) were added separately to individual tanks with a 5 minute interval between feedings. The rotifers and copepod nauplii were size sieved between 20 75 coinciding with gape limitations of the larvae. 3 2 and 3 3 Algal cell density Preliminary research has indicated that the addition of algae to culture tanks may enhance feeding of the milletseed butterflyfish larvae. Through two different experiments milletseed butterflyfish larvae were subjected to a variety of Tiso cell densities for examination of optimal feeding conditions. Experiment 3 2 examined 6 treatments ( n =3) with cell densities of (mean ± SD) 0 ± 0; 192,176 ± 4,909; 336,734 ± 2,945; 522,108 ± 15,710; 624,149 ± 7,855; and 809,523 ± 30,438 cells mL 1 for increased feeding incidence and feeding intensity. Eggs were stocked into eighteen 14 L tanks. Tanks were kept static during experimental runs. Lighting was kept at 1,314 lx. Two hours prior to feeding on 4 dph Tiso was randomly allocated to each to each tank. Five algal cell counts were then measured using a hemocytometer (Sigma Aldrich, St. Louis, Missouri) At 4 dph nauplii of the copepod, P.

PAGE 80

80 crassirostris, (5 7 nauplii mL 1 ) were added to each tank individually with a 5 minute interval between feedings. Experiment 3 3 assessed the lower limits at which algal cell densities enhance larval feeding. Following the same protocol as Experiment 3 2, six treatments ( n =3) of (mean ± SD) 0 ± 0; 30,612 ± 3,231; 81,632 ± 4,179; 122,448 ± 2,995; 153,061 ± 7,354; and 194,131 ± 6,589 cells mL 1 were assessed for enhanced larval feeding. From here on treatments will be referred to by the mean density of cell mL 1 . 3 4 4 Prey density To examine the effects of prey density on larval first feeding in the milletseed butterflyfish, 5 treatments ( n =3) of the copepod P. crassirostris were examined. Prey densities of 1, 2, 5, 10, and 15 nauplii mL 1 were examined for enhanced feeding response. Eggs were stocked into fifteen 14 L tanks. Tanks were kept static during experimental runs. Lighting was kept at 1,314 lx. Two hours prior to feeding on 4 dph a Tiso was added to all experimental tanks to adjust algal density to 400,000 600,000 cells mL 1 . At 4 dph P. crassirostris nauplii at densities 1, 2, 5, 10, and 15 nauplii mL 1 were randomly allocated to each tank individually with a 5 minute interval between feedings. 3 5 5 Larval stocking density To examine the effects of larval stocking density on larval first feeding in the milletseed butterflyfish 6, treatments ( n =3) of larval density were examined. In stocking densities of 10, 15, 20, 30, 40, 50 larvae L 1 were examined for enhanced feeding response. Eggs were randomly stocked into eighteen14 L tanks. Due to higher increased biological loads in the higher stocking density tanks, water was exchanged at a rate of 50% day 1 in all tanks to keep experimental parameters consistent between

PAGE 81

81 treatments. Lighting was kept at 1,314 lx. Two hours prior to feeding on 4 dph Tiso was added to all experimental tanks to adjust algal density to 400,000 600,000 cells mL 1 . At 4 dph P. crassirostris nauplii at density of 5 7 nauplii mL 1 were added to each tank individually with a 5 minute interval between feedings. 3 6 6 Tank size To examine the effects of tank size on larval first feeding in the milletseed butterflyfish, 3 treatments ( n =3) of different tank sizes were examined. Eggs were stocked at a density of 15 20 eggs L 1 into three circular 210 L, 128 L, and 14 L tanks. Tanks were kept static during experimental runs. Lighting was kept at 1,314 lx. Two hours prior to feeding on 4 dph Tiso was added to all experimental tanks to adjust algal density to 400,000 600,000 cells mL 1 . At 4 dph P. crassirostris nauplii at density of 5 7 nauplii mL 1 were added to each tank individually with a 5 minute interval between feedings. 3 7 7 Water exchange rates To examine the effects of water exchange on larval first feeding in the milletseed butterflyfish 3 treatments ( n =6) of three different exchange were examined. Water exchange rates of (mean ± SD) 0 ± 0% day 1 , 300 ± 100% day 1 , and 700 ± 100% day 1 were observed for enhanced larval feeding. Eggs were stocked at a density of 20 eggs L 1 into eighteen 14 L tanks. Lighting was kept at 1,314 lx. Two hours prior to feeding on 4 dph Tiso was added to all experimental tanks to adjust algal density to 400,000 600,000 cells mL 1 . Three algal cell density counts were measured at 1, 3, and 5 hours into the feeding trial to determine ensure that algal cell densities remained constant throughout the trial. Additional algae were added when cell counts reached 400 000 cells mL 1 . At 4 dph P. crassirostris nauplii at density of 5 7 nauplii mL 1 were added to

PAGE 82

82 each tank individually with a 5 minute interval between feedings. Samples were taken every hour to count prey density remaining in the tank. If necessary, prey items were added to maintain the prey density at 5 7 nauplii mL 1 . 3 8 8 Light intensity To examine the effects of light intensity on larval first feeding in the milletseed butterflyfish, 3 treatments ( n =6) of three different light intensities were examined. Light intensities of (mean ± SD) 831 ± 10 lx, 1,314 ± 11 lx, and 3,016 ± 14 lx were investigated for enhanced larval feeding. Eggs were stocked at a density of 20 eggs L 1 into eighteen 14 L tanks. Tanks were kept static during experimental runs. Two hours prior to feeding on 4 dph Tiso was added to all experimental tanks to adjust algal density to 400,000 600,000 cells mL 1 . At 4 dph P. crassirostris nauplii at density of 5 7 nauplii mL 1 were added to each tank individually with a 5 minute interval between feedings. Statistical Methods An independent samples T test was used to compare mean body length, body depth, and eye diameter between all experimental runs. A one way ANOVA with a Tukeys HSD means separation test (SPSS v.21) was used to determine if statistical differences existed between feeding incidences of prey types and mean number of prey items in the guts of larvae. Non parametric data was assessed usin g a Kruskal Wallace one way ANOVA with a pairwise multiple comparison. An 0.05 significance level will be used to determine significance. All proportion data was arcsine(sqrt) transformed before analysis. Data for graphic representation were back transformed into original units and were depicted as mean ± SE or median ± MAD for easy interpretation.

PAGE 83

83 Results Water Quality All water quality parameters were within normal, safe limits ( Table 3 1 ) Larval Measurements At 4 dph no significant difference in larval notochord length at (mean ± SD = 2.411 ± 0.004) (t 15 =1.000, P =1.000), body depth (mean ± SD = 0.458 ± 0.008) (t 15 =0.975, P =0.345), or eye diameter (mean ± SD = 0.1702 ± 0.0004) (t 15 =0.000, P =1.000) was detected between C. miliaris spawns. 3 1 1 Prey selectivity In Trial 1 the proportion of larvae feeding was found to be significantly different between all treatments. Feeding P. crassirostris nauplii to larvae significantly increased the proportion of larvae feeding (F 3, 20 = 124.387, P <0.001) ( Figure 3 1 A ). The addition of algae at 400,000 600,000 cells mL 1 also significantly increased the proportion of larvae feeding within Experiment 3 1. To test for statistical differences in feeding intensity a non parametric pairwise Kruskal Wallace test was run. Feeding intensity was significantly different between treatments (H 3 =21.049, P <0.001). A pairwise comparison revealed that feeding intensity of treatment 4 (green water P. crassirostris nauplii) was significantly higher than both B. plicatilis treatments ( Figure 3 1 B ). The proportion of larvae feeding was significantly different between treatments (H 3 =21.251, P <0.001) (analyzed by a non parametric Kruskal Wallace one way ANOVA) in Trial 2. P. crassirostris in green water had a significantly higher proportion of larvae feeding than did all other treatments ( Figure 3 2 A ). Feeding intensity between treatments was analyzed using a non parametric pairwise Kruskal Wallace test as well. A significant difference (H 3 =21.201, P <0.001) between the feeding intensity of larvae on

PAGE 84

84 P. crassirostris nauplii in green water compared to B. plicatilis was observed ( Figure 3 2 B ). 3 2 Algal cell density In Trial 1, higher algal turbidity significantly (F 5, 12 =3.347, P =0.040) increased the proportion of larvae feeding. Larvae exhibited significantly increased feeding in the highest algal cell density (809,523 cells mL 1 ) and the lowest proportion of larvae feeding in water without the addition of algal cells (0 cells mL 1 ) ( Figure 3 3 A ). Examination of gut contents revealed that feeding intensity was significantly higher in the highest algal density (809,523 cells mL 1 ) (F 5, 12 =5.431, P =0.008) compared to 0 and 624,149 cells mL 1 ( Figure 3 3 B ). Feeding intensity in the highest algal cell density (809,523 cells mL 1 ) was three times higher than feeding intensity in the lowest algal cell density (0 cells mL 1 ) ( Figure 3 3 B ). The proportion of larvae feeding in Trial 2 was significantly highe r (F 5, 12 =7.318, P =0.002) at 522,108 and 809,523 cells mL 1 than the lowest cell density, 0 cells mL 1 . The addition of algae to tanks increased the proportion of larvae feeding from 48.31% (0 cells mL 1 ) to over 75% feeding for all treatments with algal cells ( Figure 3 4 A ). Feeding intensity was significantly higher (F 5, 12 =5.445, P =0.008) at 522,108 and 809,523 cells mL 1 than 0 cells mL 1 . The number of prey items ingested increased from 1.2 nauplii larvae 1 6 h 1 at 0 cells mL 1 to 2.52 nauplii larvae 1 6 h 1 at the highest algal density (809,523 cells mL 1 ) ( Figure 3 4 B ). 3 3 3 Lower limit algal cell density Experiment 3 3 assessed the lower limits at which enhanced feeding was observed. In Trial 1, the proportion of larvae feeding was significantly higher in the highest algal cell density treatment (194,131 ± 6,589) (F 5, 12 =6.334, P =0.004) compared

PAGE 85

85 to all other treatments ( Figure 3 5 A ). The proportion of larvae consuming prey items increased from 34.94% ( SE 2.86%, +SE 2.92%) with no addition of algae to 80.68% ( SE 6.21%, +SE 5.52%) at a cell density of 194,131 cells mL 1 . Feeding intensity of first feeding C. miliaris established that ingestion rates were significantly greater in the highest algal cell density (194,131 cells mL 1 ) than all other treatments (F 5, 12 =5.182, P =0.009) ( Figure 3 5 B ). Feeding intensity gradually increased with t he addition of algal cells reaching a maximum of 1.933 nauplii larvae 1 6 h 1 in the highest algal cell density (194,131 cells mL 1 ) treatment. In Trial 2 the proportion of larvae feeding was significantly different (F 5, 12 =7.252, P =0.002) between treatments with the increased algal density (194,131 cells mL 1 ) having a significantly higher proportion of larvae feeding (mean SE, +SE = 73.47%, SE 3.48%, +SE 3.34%) than all other treatments ( Figure 3 6 A ). Additionally feeding intensity was significantly higher (F 5, 12 =6.482, P =0.004) at 194,131 cells mL 1 than all other treatments ( Figure 3 6 B ). Larval ingestion rates of the copepod P. crassirostris nauplii varied from 0.55 copepods larvae 1 6 h 1 at 81,632 cells mL 1 to 1.78 nauplii larvae 1 6 h 1 at a cell density of 194,131 cells mL 1 . 3 4 4 Prey density In Trial 1 no significant difference was detected in the proportion of larvae feeding between prey densities treatments (F 4, 10 =0.484, P =0.748) ( Figure 3 7 A ). At prey densities of 1, 2, 5, 10, and 15 over 90% of larvae fed during the experimental period. Feeding intensity also exhibited no significant differences in the ingestion rates of larvae (F 4, 10 =0.396, P =0.807) ( Figure 3 7 B ). In Trial 2 there was no significant difference (F 4, 10 =0.287, P =0.880) in the proportion of larvae feeding under prey density treatments ( Figure 3 8 A ). No significant

PAGE 86

86 difference (F 4, 10 =0.811, P =0.546) in feeding intensity of C. miliaris larvae was detected within prey density treatments ( Figure 3 8 B ). 3 5 5 Larval stocking density The proportion of C. miliaris larvae feeding was significantly affected (F 5, 12 =46.168, P <0.001) by larval stocking density in Trial 1. Stocking densities of 15 and 20 larvae L 1 had significantly higher proportion of larvae feeding, 95.46% ( SE 5.46%, +SE 3.38%) and 90.40% ( SE 3.13%, +SE 2.73%) than all other stocking densities ( Figure 3 9 A ). Feeding intensity was also significantly affected (F 5, 12 =109.579, P <0.001) by larval stocking density. Larval stocking densities of 15 and 20 larvae L 1 ingested almost twice as many nauplii as any other treatment ( Figure 3 9 B ). The proportion of larvae feeding in Trial 2 was significantly affected (F 5, 12 =26.501, P <0.001) by larvae stocking density. Fifteen and twenty larvae L 1 had significantly higher proportions of larvae feeding than all other treatments ( Figure 3 10 A ). Larval stocking densities of 10, 40, and 50 larvae L 1 significantly decreased the proportion of larvae feeding with mean feeding percentages of less than 50% ( Figure 3 10 A ). A significant difference (F 5, 12 =7.506, P =0.002) in feeding intensity was detected between larval stocking densities in Trial 2. Only 20 larvae L 1 was significantly higher than all other treatments, while 10, 15, 30, and 40 larvae L 1 did not have significantly different feeding intensities ( Figure 3 10 B ). 3 6 6 Tank size In Trial 1 the proportion of larvae feeding was not significantly different (F 2, 6 =0.281, P =0.764) between tank size treatments. Variance was noticeably larger within the 14 L tanks compared to the 128 L and 210 L tanks ( Figure 3 11 A ). Feeding intensity was also found not to be significantly different (F 2, 6 =0.472, P =0.645) between

PAGE 87

87 tank size treatments. Graphical representation of the data showed that variance within the 14 L tanks was larger than those of the 128 L and 210 L tanks ( Figure 3 11 B ). The proportion of larvae feeding in Trial 2 were not significantly affected (F 2,6 =0.436, P =0.665) by tank size. The proportion of larvae feeding with in the 14 L tank varied more than within the 128 L and 210 L tanks but had no effect on the significance of the data ( Figure 3 12 A ). Feeding intensity in Trial 2 was not significantly different (F 2, 6 =0.213, P =0.814) between treatments ( Figure 3 12 B ). 3 7 7 Water exchange rate The proportion of larvae feeding in Trial 1 was found to be significantly (F 2,15 =8.918, P =0.003) affected by water exchange rates. Water exchange rates of 0 and 300% day 1 had significantly higher proportions of larvae feeding than 700% day 1 ( Figure 3 13 A ). Feeding intensity was also found to be significantly (F 2,15 =24.959, P <0.001) different among treatments. Treatments 1 and 2 had significantly higher feeding intensities (mean ± SE = 3.85 ± 0.27) and (mean ± SE = 4.48 ± 0.45 nauplii larvae 1 6 h 1 ), than did treatment three, (mean ± SE = 2.48 ± 0.15 nauplii larvae 1 6 h 1 ) ( Figure 3 13 B ). In Trial 2 the proportion of larvae feeding was significantly different (F 2,15 =20.895, P <0.001) among treatment levels. A water exchange rate of 300% day 1 a significantly higher proportion of fish feeding than did an exchange rate of 700% day 1 ( Figure 3 14 A). Feeding intensity was found to be significantly different (F 2,15 =10.581, P =0.001), with low water exchange rates having a higher feeding intensity than higher exchange rates ( Figure 3 14 B ).

PAGE 88

88 3 8 8 Light intensity In Trial 1 the proportion of C. miliaris larvae feeding was significantly different (F 2, 15 =5.319, P =0.018) between light intensity treatments. A light intensity of 1,314 lx had a significantly higher proportion of larvae feeding than 831 lx and 3,016 lx ( Figure 3 15 A ). Feeding intensity was not significantly different (F 2, 15 =2.202, P =0.145) between t reatments ( Figure 3 15 B ). The proportion of larvae feeding in Trial 2 was not significantly different between treatments (F 2, 15 =3.123, P =0.073) ( Figure 3 16 A ). Feeding intensity was significantly different (F 2, 15 =3.026, P =0.021) between treatments. Treatment 2 (1,314 lx) was significantly higher than treatment 1 (831 lx) but statistically the same as treatment 3 (3,016 lx) ( Figure 3 16 B ). Discussion First Feeding Response This is the first controlled examination of first feeding requirements for Chaetodon miliaris . At temperatures of 24.5 26.5 °C, larvae reached first feeding on 4 dph and actively fed with an elaborate S strike behavior. Larvae were able to efficiently identify and capture prey under the appropriate conditions. 3 1 1 Prey selectivity There was a considerable difference in the proportion of larvae feeding when given B. plicatilis or P. crassirostris nauplii. In Trials 1 and 2, larvae selected P. crassirostris nauplii significantly more than B. plicatilis ( Figure 3 1 A , 3 2 A ). Additionally, feeding intensities of Trials 1 and 2 showed the median number of P. crassirostris nauplii consumed was greater (0.095 4.265 copepods larvae 1 6 h 1 ) than the feeding intensity of B. plicatilis (0 0.075 rotifers larvae 1 6 h 1 ) ( Figure 3 1 B , 3 2 B ).

PAGE 89

89 The proportion of larvae feeding on P. crassirostris nauplii in was significantly greater than all B. plicatilis treatments and in Trial 1 was significantly higher than the clear water P. crassirostris nauplii treatment ( Table 3 2 ). The addition of algal cells to the tank is hypothesized to increase larval visual contrast allowing them to perceive depth and prey items better against the shaded background (Naas et al., 1992). With better visual contrast the larvae likely have the increased capability of detecting prey items creating more prey encounters and increasing strike efficiency. Larvae of C. miliaris actively selected the nauplii of the copepod, P. crassirostris , over the rotifer, B. plicatilis , this was similar to reports with Lutjanus peru larvae (Zavala Leal et al., 2013). Active selection of prey items by larvae have been attributed to prey availability, swimming ability, color, antipredator behavior, nutritional content, and palatability 1979; Houde and Schekter, 1980; Buskey, 1993; Buskey et al., 2002). Wild data on a large number of early marine ornamental species larvae suggests that copepods are a natural prey item (Mckinnon et al., 2003; Sampey et al., 2007), thus P. crassirostris nauplii likely resembles the appropriate natural prey for C. miliaris . Another possible reason larvae are selecting P. crassirostris nauplii could be the increased nutritional value of copepods compared to rotifers (Stottrup, 2000; Hamre et al., 2008; Cassiano et al., 2011). The significant difference in the proportion of larvae feeding and feeding intensity ruled out using rotifers as a feed for C. miliaris . Studies into other alternative live feeds such as new copepod species or soft bodied live feeds may supplement this information on C. miliaris prey items. 3 2 Algal cell density The enhanced feeding response of larvae to th e treatment in prey selectivity experiment 3 1 confirmed that algal turbidity was an important factor to

PAGE 90

90 investigate to optimize first feeding parameters of C. miliaris. In algal cell density Trial 1, the proportion of larvae feeding was statistically the same in all treatments except for cell densities of 0 cells mL 1 and the highest algal density (809,523 cells mL 1 ) ( Figure 3 3 A ). Trial 2 had similar results except that 522,108 cells mL 1 was also significantly greater than the 0 cells mL 1 treatment ( Figure 3 4 A ). Feeding intensity in Experiment 3 2 followed the same trends as the proportion of larvae feeding. In Trial 2, feeding intensity was statistically the same across all treatments except 522,108 and 809,523 cells mL 1 ( Figure 3 4 B ). It is apparent that algal cell density has an impact on the proportion of larvae feeding and feeding intensity of C. miliaris larvae. 3 3 3 Lower limits algal cell density To test for the lower threshold at which the addition of algae enhanced larval feeding a similar value as the lowest algal cell density from Experiment 3 2 (194,131 cells mL 1 ) was selected as the upper limit of the second algal cell density experiment. Determining the lower limits at which algal density enhanced larval feeding is beneficial for producers and researchers, reducing labor and cost during rearing. Trial 1 indicated that enhanced feeding was observed in the 30,612 (53%) and 194,131 (80.61%) cells mL 1 even though there is a large difference in the proportion of larvae feeding between the two treatments ( Figure 3 5 A ). The large variances associated with these two treatments are likely responsible for the statistical similarity. Trial 2 further supported the conclusion that a minimum cell density of 194,131 cells mL 1 is required to enhance larval feeding. Both the proportion of larvae feeding and the feeding intensity in Trial 2 found that an algal cell density of 194,131 cells mL 1 was significantly greater than all other algal cell densities ( Figure 3 6 ). This study indicates that a minimum number of

PAGE 91

91 particles (~190,000 cells mL 1 ) was necessary for an enhanced proportion of larvae feeding and feeding intensity. This supports the hypothesis that visual contrast is increased by adding microalgae to larval culture tanks thus allowing larvae to recognize prey items, gauge distance and motion of prey items better (Naas et al., 1992; Shaw et al., 2006). Another benefit of adding algae into culture tanks is the increased light intensity gradient and light scattering (Naas et al., 1992). Larvae can distribute vertically within the tank to find an optimal light intensity. Ultimately, increased feeding at early larval stages due to the addition of microalgae can transition into increased larval survival as observed in Neopomacentrus cyanomos (Setu et al., 2010). 3 4 4 Prey density Prey density Experiment 3 4 showed no significant difference between treatments in Trial 1 or 2 in the proportions of larvae feeding and feeding intensity. In all treatments for both Trials 1 and 2, more than 89% of larvae fed and over 3.6 nauplii larvae 1 6 h 1 were consumed ( Figure 3 7 , 3 8 ). The results from this study contrast research on the effects of prey density on larval feeding. Both Zavala et al. (2013) and Hilder et al. (2014) observed increased feeding in Lutjanus peru , Thunnus maccoyii, and Seriola lalandi . Additionally, higher feeding intensities and proportions of Scomberomorus niphonius and Thunnus albacares larvae fed with increased prey density (Shoji and Tanaka, 2004; Wexler et al., 2011). In general, greater prey densities resulted in increased predator and prey encounter rates, raising the probability of a successful capture by larvae until a point of satiation is reached (Houde and Schekter, 1980; Temple et al., 2004; Hilder et al., 2014). From this study the point of satiation for C. miliaris appears to be 1 nauplii mL 1 . No observations of undigested P. crassirostris

PAGE 92

92 were found in any prey density treatment. C. miliaris larvae do not appear to have any issues digesting P. crassirostris . 3 5 5 Larval stocking density The proportion of larvae feeding at 15 and 20 larvae L 1 was significantly greater than all other treatments for both Trials 1 and 2. It appears that at stocking densities greater than 30 larvae L 1 feeding is suppressed. Larvae at this density likely crowd within the tank and reduce feeding rates. Reduced proportions of larvae feeding at 10 larvae L 1 propagate additional questions. Feeding intensity in Trial 1 was representative of the proportion of larvae feeding with higher stocking densities of 15 and 20 larvae L 1 feeding significantly more ( Figure 3 9 ). In Trial 2, treatments of 15, 20, 30 and 40 larvae L 1 had the same feeding intensity 2.46, 3.23, 2.45, and 1.63 nauplii larvae 1 6 h 1 respectively ( Figure 3 10 B ). In both Trials 1 and 2 the proportion of larvae and feeding intensity was suppressed at 10 larvae mL 1 . This is inconsistent with the results found in Experiment 3 5 since it seems that lower stocking densities increase the proportion of larvae feeding and feeding intensity, however, previous research on southern bluefin tuna (SBT) and yellowtail kingfish (YTK) showed that larval density has species specific effects. Hilder et al. (2014) observed that the proportion of larvae feeding and feeding intensity in SBT was statistically equivalent across larval densities from 2 65 larvae L 1 . YTK did not follow the same trend as SBT showing increased proportions of larvae fee ding at 2 and 5 larvae L 1 compared to higher larval densities up to 75 larvae L 1 . Feeding intensity was opposite the proportion of larvae feeding trend though, showing increased rotifer consumption at higher larval densities 25, 50, and 75 larvae L 1 . Intraspecific competition could account for decreased proportion of larvae feeding at

PAGE 93

93 higher larval densities if prey abundance was a limiting factor, but, at the prey densities used in Experiment 3 5 there were approximately a minimum 1,000 prey items per larvae stocked into each tank. Welker et al. (1994) observed intraspecific competition in a mesocosum experiment with Dorosoma cepedianum larvae between densities of 35 and 70 fish m 1 . Reduced predation was observed by larvae at 70 fish m 1 , however, decreased feeding of C. miliaris at high larval densities in this study is more likely due to appetite reduction from stress due to overstocking or the inability to initiate feeding due (Wendelaar Bonga, 1997; King et al., 2000; Hilder et al., 2014). Further experiments into lower larval stocking densities seems necessary for C. miliaris to understand the reason behind the decreased feeding rates and intensity of larvae at 10 larvae L 1 , however, for the commercial production this information does not seem pertinent since optimization and efficiency are crucial for an effective production operation. 3 6 6 Tank size No significant differences in the proportion of larvae feeding or feeding intensity was observe d between tanks size treatments. The proportion of larvae feeding in all treatments was greater than 88% and the feeding intensities of larvae in all treatments were above 3.1 nauplii larvae 1 6 h 1 . Tank size within a static system evidently had no effect on the first feeding response of C. miliaris larvae. Wittenrich et al. (2012) observed significantly larger numbers of Abudefduf saxatilis capturing prey in 120 L tanks compared to 60 L tanks; however, the number of prey items ingested at first feeding (1 dph) did not differ. Larval survival was significantly greater throughout the duration of the trial in 120 L (6.6%). Conversely, et al. (2004) found that survival of Sparus aurata decreased with increasing tank size. Additional studies into the survival of larvae different tank sizes may reveal that a certain tank size is more

PAGE 94

94 beneficial for culturing C. miliaris . Additionally, under different conditions such as water flow or increased aeration the first feeding response of larvae in different tank sizes may change. 3 7 7 Water exchange rate In Trial 1, feeding intensity and the proportion of larvae feeding data suggested that larvae have increased capability of feeding under lower flow conditions. Water flows of 0 and 300% day 1 increased the proportion of larvae feeding by 15% compared to 700% day 1 ( Figure 3 13 A ). In the low flow treatments larvae consumed 1.4 nauplii larvae 1 6 h 1 more than in the high flow treatment ( Figure 3 13 B ). The water flows 0, 300%, and 700% day 1 . Trial 2 supports that first feeding is increased at lower water flows. The proportion of larvae feeding was significantly greater at 300% day 1 , however 0 and 700% day 1 had large variances accounting for the interaction ( Figure 3 14 ). Weyers et al. (2003) found that Moxostoma robustum and Moxostoma collapsum subjected to 4 and 12 h of pulsed, high velocity flow exhibited significantly decreased growth and survival compared to 0 h of high velocity flow. At 700% day 1 a majority of food is flushed out necessitating restocking of new copepod nauplii to maintain prey densities. Prey size remains consistent within the tank but the time needed for larvae to identify, align, and strike at a prey item is reduced with fast moving water currents (Dower et al., 1997). MacKenzie et al. (1994) found that increasing turbulence increased the number of predator prey encounters but decreased the pro bability of a successful pursuit. Ingestion rates of larval cod were highest at a turbulent velocity of approximately 5 mm s 1 showing that there was an optimum balance between predator and prey encounter rates and successful pursuits (MacKenzie et al., 1994). C. miliaris larvae appear to benefit from low flow environments ( Figure 3 13 , 3 14 ). In addition to

PAGE 95

95 the benefit of increased prey ingestion at a water exchange rate of 300% day 1 it was observed that most nauplii not being eaten were flushed from the tank. While restocking of P. crassirostris nauplii was necessary to maintain prey densities, it did not allow for nauplii to grow into copepodites and adult copepods,. 3 8 8 Light intensity The proportion of larvae feeding at 1,314 lx was significantly greater in Trial 1, whereas in Trial 2 there was no significant difference in the proportion of larvae feeding between any treatments. Trials 1 and 2 seem to contradict each other, however, it is likely that increased variance within the light intensity trials reduced the significance of results. All light intensities had high proportions of larvae feeding and feeding intensities with over 79% of larvae feeding and 3.2 nauplii larvae 1 6 h 1 consumed. Increased light intensity has been shown to increase the proportion of larvae feeding and feeding intensity in HIppoglossus hippoglossus , Seriola lalandi , and Lutjanus peru (Naas et al., 1992; Carton, 2005; Zavala Leal et al., 2013). Increased light intensity likely contrasts prey items better than lower light intensities. In Experiment 3 8 the addition of algae is likely the reason for reduced significance between light intensity treatments because of the i ncreased light scattering and attenuation due to large number of algal cells in the water column. Future research could address an increased range of light intensities to identify the potential limits of the effective light intensity range for C. miliaris . The eight experiments conducted in this chapter were used to evaluate methods to optimize the first feeding response of C. miliaris in culture conditions. Table 3 2 displays the enhanced culture conditions from each experiment. It is evident throughout this research that Brachionus sp . rotifers are not a suitable feed for C. miliaris . Nauplii of the copepod, P. crassirostris , elicited a higher feeding response from larvae than rotifers

PAGE 96

96 particularly with the addition of algae. Prey density and tank size experiments established that there was no benefit to increased prey densities or larger tank size on first feeding. Low water exchange rates increased the proportion of larvae feeding and feeding intensity of C. miliaris larvae under the examined conditions. Light intensity Trials 1 and 2 showed that lux from 831 3,016 lx sustain high proportions of larvae feeding and increased feeding intensity. Additional trials would be beneficial to uncover the upper and lower threshold at which light intensity significantly effects first feeding response. This information will be valuable for the producers trying to lower cost of production. While these studies establish optimal first feeding parameters it is important to understand that as larvae develop advantageous conditions will change and culture parameters need to be adjusted accordingly to account for this.

PAGE 97

97 Table 3 1. Culture parameters for C. miliaris during all experiments. Values for salinity, water temperature, dissolved oxygen, light intensity, total ammonia nitrogen, nitrite nitrogen, and nitrate nitrogen are represented as mean ± SD. Values for algal cell density, prey density, and stocki ng density are represented as ranges. Prey selectivity Algal turbidity Lower limits algal turbidity Prey density Larval stocking density Tank size Water exchange Light intensity Time to hatch 28 h @ 25.5 °C 28 h @ 25.5 °C 28 h @ 25.5 °C 28 h @ 25.5 °C 28 h @ 25.5 °C 28 h @ 25.5 °C 28 h @ 25.5 °C 28 h @ 25.5 °C Tank color Black sides /white bottom Black sides /white bottom Black sides /white bottom Black sides /white bottom Black sides /white bottom Black sides /white bottom Black sides /white bottom Black sides /white bottom Water treatment 1 filtered 1 filtered 1 filtered 1 filtered 1 filtered 1 filtered 1 filtered 1 filtered Salinity (g L 1 ) 33.0 ± 0.45 33.0 ± 0.45 33.0 ± 0.45 33.0 ± 0.45 33.0 ± 0.45 33.0 ± 0.45 33.0 ± 0.45 33.0 ± 0.45 Water temperature (°C) 25.5 ± 1.00 25.5 ± 1.00 25.5 ± 1.00 25.5 ± 1.00 25.5 ± 1.00 25.5 ± 1.00 25.5 ± 1.00 25.5 ± 1.00 pH 8.2 ± 1.00 8.2 ± 1.00 8.2 ± 1.00 8.2 ± 1.00 8.2 ± 1.00 8.2 ± 1.00 8.2 ± 1.00 8.2 ± 1.00 Dissolved oxygen (% saturation) 97.8 ± 1.30 97.8 ± 1.30 97.8 ± 1.30 97.8 ± 1.30 97.8 ± 1.30 97.8 ± 1.30 97.8 ± 1.30 97.8 ± 1.30 Photoperiod (h) 14 L:10 D 14 L:10 D 14 L:10 D 14 L:10 D 14 L:10 D 14 L:10 D 14 L:10 D 14 L:10 D Prey item EXP. 3 1 P. crassirostris P. crassirostris P. crassirostris P. crassirostris P. crassirostris P. crassirostris P. crassirostris Algal cell density (cells mL 1 ) 400,000.00 600,000.00 EXP. 3 2 EXP. 3 3 400,000.00 600,000.00 400,000.00 600,000.00 400,000.00 600,000.00 400,000.00 600,000.00 400,000.00 600,000.00 Prey density (prey mL 1 ) 5.0 7.00 5.0 7.00 5.0 7.00 EXP. 3 4 5.0 7.00 5.0 7.00 5.0 7.00 5.0 7.00 Stocking density (larvae L 1 ) 15.0 20.00 15.0 20.00 15.0 20.00 15.0 20.00 EXP. 3 5 15.0 20.00 15.0 20.00 15.0 20.00 Tank size (L) 14.00 14.00 14.00 14.0 14.00 EXP. 3 6 14.00 14.00 Exchange rate (% day 1 ) 0.00 0.00 0.00 0.00 0.00 0.00 EXP. 3 7 0.00 Light intensity (lx) 1,314.0± 11.00 1,314.0± 11.00 1,314.0± 11.00 1,314.0± 11.00 1,314.0± 11.00 1,314.0± 11.00 1,314.0± 11.00 EXP. 3 8

PAGE 98

98 Table 3 1. Continued. Prey selectivity Algal turbidity Lower limits algal turbidity Prey density Larval stocking density Tank size Water exchange Light intensity Total Ammonia Nitrogen (mg L 1 ) 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 Nitrite nitrogen (mg L 1 ) 0 .0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 Nitrate nitrogen (mg L 1 ) 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0.0 ± 0.00 0 .0 ± 0.00 0.0 ± 0.00 Table 3 2. Optimized first feeding parameters for the milletseed butterflyfish C . miliaris . Experiment Optimal parameters Prey selectivity P. crassirostris nauplii (20 75 Algal cell density 809,523.0 cells mL 1 Lower limits algal cell density 194,131.0 cells mL 1 Prey density 1.0 nauplii mL 1 Larval stocking density 15.0 larvae L 1 Tank size 210.0 L Water exchange rate 300.0 % day 1 Light intensity 1314.0 lx

PAGE 99

99 A 120 100 80 60 40 20 0 Green Water Rotifers Clear Water Rotifers Clear Water Copepods Green Water Copepods Treatment B A 5 4 3 2 1 0 Green Water Rotifers Clear Water Rotifers Clear Water Copepods Green Water Copepods Treatment Figure 3 1. First feeding responses of C. miliaris larvae with and without the addition of algae and different prey items, (A) the proportion of larvae feeding (%) represented as mean ± SE ( n =6) and (B) feeding intensity (nauplii larvae 1 6 h 1 ) represented as median ± MAD ( n =6). Capitol letters indicate significant differences P 0.05. A B D C B C C Proprotion of Larvae Feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 100

100 A 120 100 80 60 40 20 0 Green Water Rotifers Clear Water Rotifers Clear Water Copepods Green Water Copepods Treatment B 6 5 4 3 2 1 0 Green Water Rotifers Clear Water Rotifers Clear Water Copepods Green Water Copepods Treatment Figure 3 2. First feeding responses of C. miliaris larvae with and without the addition of algae and different prey items, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ) represented. All data is represented as median ± MAD ( n=6) . Capital letters indicate significant difference P 0.05. A AB B B A AB B B Proportion of Larvae Feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 101

101 A 120 100 80 60 40 20 0 0 ± 0 192,176 ± 4,909 336,734 ± 2,945 522,108 ± 15,710 624,149 ± 7,855 809,523 ± 30,438 Algal Cell Density (cells mL 1 ) B 4 3 2 1 0 0 ± 0 1 Algal Cell Density (cells mL 1 ) Figure 3 3. First feeding responses of C. miliaris larvae in different algal cell densities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant difference P 0.05. AB AB A AB AB B A AB AB AB B B Feeding Intensity (nauplii larvae 1 6 h 1 ) Proportion of Larvae Feeding (%) 92,176 ± 336,734 ± 522,108 ± 624,149 ± 809,523 ± 4,909 2,945 15,710 7,855 30,438

PAGE 102

102 A 120 100 80 60 40 20 0 0 ± 0 192,176 ± 4,909 336,734 ± 2,945 522,108 ± 15,710 624,149 ± 7,855 809,523 ± 30,438 Algal Cell Density (cells mL 1 ) B 4 3 2 1 0 0 ± 0 1 Algal Cell Density (cells mL 1 ) Figure 3 4. First feeding responses of C. miliaris larvae in different algal cell densities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant difference P 0.05. A A AB AB AB B A A AB B AB AB Proportion of Larvae Feeding (%) Feeding Intensity (nauplii arvae 1 6 h 1 ) 92,176 ± 336,734 ± 522,108 ± 624,149 ± 809,523 ± 4,909 2,945 15,710 7,855 30,438

PAGE 103

103 A 120 100 80 60 40 20 0 0 ± 0 30,612 ± 3,231 81,632 ± 4,179 122,448 ± 2,995 153,061 ± 7,354 194,131 ± 6,589 Algal Cell Density (cells mL 1 ) (B) 3 2 1 0 0 ± 0 30,612 ± 3,231 81,632 ± 4,179 122,448 ± 2,995 153,061 ± 7,354 194,131 ± 6,589 Algal Cell Density (cells mL 1 ) Figure 3 5. First feeding responses of C. miliaris larvae in different algal cell densities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A AB B B B B A B B B B AB Feeding Intensity (nauplii larvae 1 6 h 1 ) Proportion of larvae feeding (%)

PAGE 104

104 A 120 100 80 60 40 20 0 0 ± 0 30,612 ± 3,231 81,632 ± 4,179 122,448 ± 2,995 153,061 ± 7,354 194,131 ± 6,589 Algal Cell Density (cells mL 1 ) B 3 2 1 0 0 ± 0 30,612 ± 3,231 81,632 ± 4,179 122,448 ± 2,995 153,061 ± 7,354 194,131 ± 6,589 Algal Cell Density (cells mL 1 ) Figure 3 6. First feeding responses of C. miliaris larvae in different algal cell densities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A B B B B B A B B B B B Proportion of Larvae feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 105

105 A 120 100 80 60 40 20 0 1 2 5 10 15 Prey Density (nauplii mL 1 ) B A A 7 6 5 4 3 2 1 0 1 2 5 10 15 Prey Density (nauplii mL 1 ) Figure 3 7. First feeding responses of C. miliaris larvae at different prey densities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A A A A A A A A Proportion of Larvae Feeding (%) Feeding Intensity (nauplii larve 1 6 h 1

PAGE 106

106 A 120 100 80 60 40 20 0 1 2 5 10 15 Prey Density (nauplii mL 1 ) B 8 7 6 5 4 3 2 1 0 1 2 5 10 15 Prey Density (nauplii mL 1 ) Figure 3 8. First feeding responses of C. miliaris larvae at different prey densities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A A A A A A A A A A Proportion of Larvae Feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 107

107 A 120 100 80 60 40 20 0 10 15 20 30 40 50 Stocking Density (larvae L 1 ) B 5 4 3 2 1 0 10 15 20 30 40 50 Stocking Density (larvae L 1 ) Figure 3 9. First feeding responses of C. miliaris larvae at different larval stocking densities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 . All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A A B C C C A A B C C C Proprtion of Larvae Feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 108

108 A 120 100 80 60 40 20 0 10 15 20 30 40 50 Stocking Density (larvae L 1 ) B 4 3 2 1 0 10 15 20 30 40 50 Stocking Density (larvae L 1 ) Figure 3 10. First feeding responses of C. miliaris larvae at different larval stocking densities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A A AB CD BC D A AB ABC ABC BC C Proportion of Larvae Feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 109

109 120 100 80 60 40 20 0 14 128 210 Tank Size (L) B 5 4 3 2 1 0 14 128 210 Tank Size (L) Figure 3 11. First feeding responses of C. miliaris larvae in different tank sizes, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A A A A A A A Proportion of Larvae Feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 110

110 A 120 100 80 60 40 20 0 14 128 210 Tank Size (L) B 5 4 3 2 1 0 14 128 210 Tank Size (L) Figure 3 12. First feeding responses of C. miliaris larvae in different tank sizes, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A A A A A A Feeding Intensity (nauplii larvae 1 6 h 1 ) Proportion of Larvae Feeding (%)

PAGE 111

111 A 120 100 80 60 40 20 0 0 ± 0 300 ± 100 700 ± 100 Water Exchange (% day 1 ) B 6 5 4 3 2 1 0 0 ± 0 300 ± 100 700 ± 100 Water Exchange (% day 1 ) Figure 3 13. First feeding responses of C. miliaris larvae in different water exchange rates, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A A B A A B Proportion of Larvae Feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 112

112 A 120 100 80 60 40 20 0 0 ± 0 300 ± 100 700 ± 100 Water Exchange (% day 1 ) B 6 5 4 3 2 1 0 0 ± 0 300 ± 100 700 ± 100 Water Exchange (% day 1 ) Figure 3 14. First feeding responses of C. miliaris larvae in different water exchange rates, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. AB A B A A B Proportion of Larvae Feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 113

113 A 120 100 80 60 40 20 0 831 ± 10 1314 ± 11 3016 ± 14 Light Intensity (lx) B 5 4 3 2 1 0 831 ± 10 1314 ± 11 3016 ± 14 Light Intensity (lx) Figure 3 15. First feeding responses of C. miliaris larvae under different light intensities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate significant differences P 0.05. A B B A A A Proportion of Larvae Feeding (%) Feeding Intensity (nauplii larvae 1 6 h 1 )

PAGE 114

114 A 120 100 80 60 40 20 0 831 ± 10 1314 ± 11 3016 ± 14 Light Intensity (lx) B 6 5 4 3 2 1 0 831 ± 10 1314 ± 11 3016 ± 14 Light Intensity (lx) Figure 3 16. First feeding responses of C. miliaris larvae under different light intensities, (A) the proportion of larvae feeding (%) and (B) feeding intensity (nauplii larvae 1 6 h 1 ). All data is represented as mean ± SE ( n =3). Capital letters indicate statistical significant differences P A A A A AB B Feeding Intensity (nauplii larvae 1 6 h 1) Proportion of Larvae Feeding (%)

PAGE 115

115 Figure 3 17. Parvocalanus crassirostris nauplii exoskeleton present in the gut of a C. miliaris larva compressed between two microscope slides.

PAGE 116

116 CHAPTER 4 GROWTH AND SURVIVAL OF Chaetodon miliaris TO 10 DAYS POST HATCH Foreword Early larval survival and growth are key constituents of increased culture performance. Increased energy expenditure, suppressed growth, deformities, and increased larval size variability can result in decreased product yields if early culture conditions are not optimal (Appelbaum and Kamler, 2000). In pelagic marine ornamental species the prolarvae stage, small size, and long larval duration amplify the severity of sub standard culture conditions. Insight into early larvae requirements can improve survival and growth, effectively, increasing the chances of larvae reaching metamorphosis. A variety of abiotic factors such as tank type, tank size, photoperiod, temperature, salinity, water quality, and light intensity, as well as biotic factors such as prey type, prey density, and nutr itional content of feed are known to affect early larval growth and survival (Ostrowski, 1989; Sargent et al., 1999; Appelbaum and Kamler, 2000; Galmuzina et al., 2000; Leu et al., 2009; Cobcroft et al., 2012). In addition, broodstock nutrition, maturity, and age also influence early larval survival and growth (Morehead et al., 2001; Callan et al., 2012a; Callan et al., 2012b). Improved larval survival is the result of high quality eggs, appropriate culture conditions, and the correct larval nutrition. Research on early larvae of pelagic marine ornamental species demonstrates that most bottlenecks encountered in larval culture start at the transition from endogenous to exogenous feeding (Suzuki et al., 1980; McGurk, 1984; Glamuzina et al., 1998; Glamuzin a et al., 2000; Glamuzina et al., 2001; Tanaka et al., 2001; Leu et al., 2009; Leu et al., 2013). Often large mortality events coincide with the point of

PAGE 117

117 irreversible starvation (McGurk, 1984). While irreversible starvation varies specifically, in pelagic marine ornamental larvae it occurs around the endogenous to exogenous transition (3 5 dph) and frequently is marked by large mortalities (Suzuki et al., 1980; Tanaka et al., 2001; Leu et al., 2009). Leu et al. (2009) established that Pomacanthus semicirculatus larvae experienced large mortalities at 5 dph when fed 100% Nanochloropsis sp., 100% B . rotundiformis , 100% Gonyaulax sp., and 50% Nannochloropsis sp. plus 50% B . rotundiformis , because most larvae were not selecting the prey items supplied. However, larvae supplied with Nannochloropsis sp. plus B . rotundiformis and Gonyaulax sp. showed significantly greater survival than all other treatments. Establishment of an appropriate first feed reduces early mortalities, but adjustment of culture parameters can likely further optimize survival. Parameters similar to those examined in first feeding experiments (Hilder et al., 2014) can additionally be addressed to determine if factors impact early larval survival. Photoperiod is known to have specific effects on larval survival (Hart et al., 1996; Arvedlund et al., 2000; Puvanendran and Brown, 2002). Duray and Kohno (1988) assessed the effects of 24 h and natural daylight (10 h) photoperiods on Siganus gattatus larvae revealing that continuous lighting had a higher mean survival and growth than natural photoperiod. A photoperiod of 24 h light allowed larvae more time to encounter food items, increasing food consumption and growth of larvae but can cause endocrine issues. The addition of algae to culture tanks often enhances larval survival. Setu et al. (2010) observed increased survival rates in Neopomacentrus cyanomos with the addition of algae to culture tanks. In tanks without algae 0% of larvae survived where

PAGE 118

118 as, 20% of larvae survived with the addition of algae to culture tanks. Prey density and stocking density can also affect early larval survival. Duray et al. (1996) showed that Epinephelus suillus larval survival to 14 dph was enhanced by high rotifer densities in culture tanks, however, Temple et al. (2004) found that a low density of rotifers increased survival of Centopomus parallelus during larviculture. Optimal culture parameters for early larvae are species specific and require individual assessments. Similar to prey density, stocking density of larvae for optimal culture varies with species. Some species are capable of growing and surviving under high larval densities, while others are not. Baskerville Bridges and Kling (2000) found that decreased Gadus morhua survival at increased larval stocking densities could be nullified with the addition of excess prey items. Water quality is another factor that can impact larval growth and survival. Metabolite (ammonia, nitrite, nitrate) accumulation in the culture tanks is associated with low water exchanges and can incur possibly harmful effects on larval feeding and development (Tandler and Helps, 1985). Under mass scale culture condition, low water exchange rates are beneficial for conserving algae, prey items, and reducing larval exposure to high water flows, however, water quality can deteriorate to harmful levels quickly if not regularly monitored (Tandler and Helps, 1985). Methods In chaetodontids, larval survival has been low with only two instances of a larvae being cultured past 8 dph (Wittenrich and Cassiano, 2011; Baensch, 2014). In Chapter 3 of this thesis an assessment of the first feeding parameters was conducted to determine optimal conditions for C. miliaris larvae. Based upon the optimized first feeding parameters found in Chapter 3 ( Table 3 2 ) two multifactorial experiments on the

PAGE 119

119 combined effects of algal cell density/stocking density and water exchange rates/prey density were designed to test survival and growth of early larvae (Naas et al., 1992; Duray et al., 1996; Baskerville Bridges and Kling, 2000; Temple et al., 2004; Setu et al., 2010). A single factorial experiment addressing three different photoperiods was conducted to see if there we re effects on early larval growth and survival (Duray and Kohno, 1988; Hart et al., 1996; Arvedlund et al., 2001; Puvanendran and Brown, 2002). Embryo Stocking and Culture C. miliaris embryos were obtained from broodstock tanks at the University of Florida Tropical Aquaculture Laboratory (Ruskin, FL). Broodstock spawned daily during this period allowing for experiments to be conducted consecutively. Three spawns were collected during May June 2014. Embryos were assessed for viability. Fertilization w as greater than 96% and hatching success greater than 95% for all cohorts. All culture parameters for C. miliaris are summarized in Table 4 1 . For each trial eggs were stocked in the experimental system 12 h post fertilization. Light aeration supplied by a single cylindrical air stone (2.5 cm height x 2.5 cm circumference) resting on the bottom of the tanks throughout all experiments. All experimental tanks had a volume of 14 L (30.5 cm diameter x 33.0 cm height) and were fiberglass with black sides and a white bottom. Larvae were observed for abnormalities such as large mortalities, and larvae lying on the tank bottom during the first 3 dph before Tisochrysis lutea (Tiso) was added to each tank on 3 dph. At 4 dph larvae had fully functional mouths, digestive tracts, and pigmented eyes indicating they were capable of feeding. Feed was added to each tank individually with a 5 minute interval between feedings to allow a 5 minute time frame to collect larvae (Hilder et al., 2014). Two channel Viavolt 54 watt 6500 K color

PAGE 120

120 temperature light banks (Phillips, Andover, MA) provided light. Using a lux meter, (Milwaukee Instruments, Rocky Mount, NC) light intensity was measured at the waters surface in the center of each tank (MW 700 portable lux meter). At 4 dph, notochord length, body depth, and eye diameter were recorded ( n =5) from each tank to ensure there was no size bias for an individual treatment. Notochord length (SL) was measured as in Chapter 2 and 3. Body depth (BD) was measured as the distance between the dorsal ridge of the notochord and the ventral margin of the coelom in line with the pectoral fin. Eye diameter (Ed) was calculated using Ed = E H + E L /2 where E H is eye height and E L is eye length. Assessment of Larval Size, Growth, Condition, and Survival Randomly, 5 larvae from each replicate were harvested in the morning on 4 dph, 6 dph, and 8 dph. Larvae were immediately euthanized after harvesting with an overdose of buffered MS 222 (500 mg L 1 ) (Western Chemical, Ferndale, Washington) to ensure larvae did not damage themselves in the collection beaker. Larvae were placed on a sedgewick rafter slide with a 1 mm grid (Sigma Aldrich, St. Louis, Missouri) and subsequent photographs (Jenoptik, ProgRes Capture Pro v2.8.8) and measurements (ImageJ 1.48r) were taken. At 10 dph the tank was euthanized, and the remaining larvae in the tank were siphoned from the tank and counted under the dissecting scope for survivorship. A random sample of 20 larvae were selected from each tank and subsequent photographs (Jenoptik, ProgRes Capture Pro v2.8.8) and measurements (ImageJ 1.48r) were taken. If 20 larvae were not present in the tank, the remaining larvae were euthanized, removed, photographed and measured. Replicates with less than 5 larvae present at 10 dph were not included in size, growth, or condition analyses.

PAGE 121

121 Specific growth rates of larvae for SL, BD, and Ed were ca lculated using the formula ([ln(M t ) ln(M 0 )]/t)*100 = SGR (% day 1 ) where M t is the parameter measurement at the end of the time period, M 0 is the parameter measurement at the beginning of the time period, and t is the length of time (Temple et al., 2004). There are many condition indices that can be measured, however, ratios of body depth to length are sensitive to environmental conditions affecting feeding of larvae (Koslow et al., 1985). To estimate condition of larvae, a condition index of the rat io of body depth to body length, which is a function of larval volume, was used for C. miliaris larvae (Koslow et al., 1985; Temple et al., 2004). The condition index was calculated using the formula CI=BD/SL, where CI is condition index, BD is body depth, and SL is notochord length. Larval Size, Growth, Condition, and Survival Experiments 4 1 Algal cell density and stocking density To assess the effects of Tiso cell density and larval stocking density on growth and survival of C. miliaris larvae, nine treatments ( n =3) were evaluated: treatment 1 (180,000 200,000 cells mL 1 and 15 larvae L 1 ), treatment 2 (500,000 530,000 cells mL 1 and 15 larvae L 1 ), treatment 3 (800,000 830,000 cells mL 1 and 15 larvae L 1 ), treatment 4 (180,000 200,000 cells mL 1 and 20 larvae L 1 ), treatment 5 (500,000 530,000 cells mL 1 and 20 larvae L 1 ), treatment 6 (800,000 830,000 cells mL 1 and 20 larvae L 1 ), treatment 7 (180,000 200,000 cells mL 1 and 25 larvae L 1 ), treatment 8 (500,000 530,000 cells mL 1 and 25 larvae L 1 ), and treatment 9 (800,000 830,000 cells mL 1 and 25 larvae L 1 ). Eggs were stocked into twenty seven, 14 L tanks, in a flow through system at each respective treatments larval stocking density. A randomized block design was used to block treatments accor ding to larval stocking density. Tanks were kept static during the first 3 dph. At 4 dph water was exchanged at a rate of 100%

PAGE 122

122 day 1 until 10 dph when Experiment 4 1 ended. Two hours prior to feeding on 4 10 dph Tiso was added to all experimental tanks, with 9 tanks receiving 180,000 200,000 cells mL 1 , another 9 tanks receiving 500,000 530,000 cells mL 1 , and the remaining 9 tanks receiving 800,000 830,000 cells mL 1 . Algal cell densities were monitored three times daily and Tiso was added as necessary to maintain algal cell densities during the 14 h light period. At 4 dph P. crassirostis nauplii were added at a density between 1 2 nauplii mL 1 to each tank individually with a 5 minute interval between feedings. Prey density counts were taken each morning at 08:00, 13:00, and 18:00 on 5 10 dph to maintain prey densities during the 14 h light period. When necessary, prey items were added to maintain the prey density between 1 2 nauplii mL 1 . A photoperiod of 14 L:10 D was used throughout Experiment 4 1. 4 2 2 Water exchange rates and prey density To assess the effects of water exchange and prey density on the early growth and larval survival of C. miliaris , 6 treatments ( n =3) were evaluated: treatment 1 (100% day 1 and 1 nauplii mL 1 ), treatment 2 (100% day 1 and 10 nauplii mL 1 ), treatment 3 (300% day 1 and 1 nauplii mL 1 ), treatment 4 (300% day 1 and 10 nauplii mL 1 ), treatment 5 (700% day 1 and 1 nauplii mL 1 ), and treatment 6 (700% day 1 and 10 nauplii mL 1 ). Eggs were stocked into eighteen, 14 L tanks at 15 eggs L 1 in a recirculating system. Water was circulated from a 2325 L sump through a fluidized sand filter into a 1937 L sump, then pumped through 50 and 10 filter bags and an 80 watt ultra violet (UV) sterilizer before reaching the header tank for supply to culture tanks. A randomized block design was used to block treatments according to water exchange rates. Tanks were kept static during the first 3 dph. At 4 dph water exchange rates were adjusted to 100% day 1 in 6 tanks, 300% day 1 in another 6 tanks, and 700%

PAGE 123

123 day 1 in the remaining 6 tanks. On 4 10 dph Tiso was added two hours prior to feeding at a density of 500,000 530,000 cells mL 1 to each tank. Algal cell densities were monitored three times daily in treatment 1and 2, and 5 times daily in treatments 3,4,5, and 6. Tiso was added as necessary to maintain algal cell densities during the 14 h light period. At 4 dph P. crassirostis nauplii were added at a density of 1 or 10 nauplii mL 1 to 3 tanks with a water exchange rate of 100% day 1 , 3 tanks with a water exchange rate of 300% day 1 , and 3 tanks with a water exchange rate of 700% day 1 . Prey density counts were taken each morning at 08:00, 13:00, and 18:00 on 4 10 dph to maintain prey densities during the 14 h light period. When necessary, prey items were added to maintain the prey densities in each treatment. A photoperiod of 14 L:10 D was used throughout Experiment 4 2. 4 3 3 Photoperiod To examine the effects of photoperiod on early larval growth and survival of C . miliaris 3 treatments ( n =6) of three different photoperiods were examined. Photoperiods of 14 L:10 D, 18 L:6 D, and 24 L:0 D were observed for increased larval condition, growth, size, and survival. Eggs were stocked into eighteen, 14 L tanks at 15 larvae L 1 in a recirculating system. Water was circulated from a 2325 L sump through a fluidized sand filter into a 1937 L sump, then pumped through 50 25 and 10 filter bags, and an 80 watt ultra violet (UV) sterilizer before reaching the header tank for supply to culture tanks. A randomized block design was used to block treatments according to photoperiod. Six tanks were grouped together in each treatment, separated by cardboard dividers to make sure that external light sources did not effect treatments. Tanks were kept static during the first 3 dph. At 4 dph, water was exchanged at a rate of 100% day 1 until 10 dph when Experiment 4 3 ended. At 4 dph, Tiso was added two

PAGE 124

124 hours prior to feeding at a density of 500,000 530,000 cells mL 1 to each tank. Algal cell densities were monitored three times daily and Tiso was added as necessary to maintain algal cell densities during the light period of each treatment. At 4 dph, P. crassirostis nauplii were added at a density between 1 2 nauplii mL 1 to each tank individually with a 5 minute interval between feedings. Samples were taken three times daily to count prey density remaining in the tank during the light period. When necessary, prey items were added to maintain the prey density between 1 2 nauplii mL 1 . Statistical Methods All statistics were conducted using SPSS v. 21 (IBM, Armonk, NY). Three one way with Tukeys HSD means separation tests were conducted to determine if statistical differences existed between survivorship of algal cell density and larval stocking density treatments, water exchange and prey density treatments, and photoperiod treatments. Non parametric data was assessed using a Kruskal Wallace one way ANOVA with a pairwise multiple comparison. A significance level will be used to determine significance. All proportion data was arcsine(sqrt) transformed before analysis. Data for graphic representation were back transformed into original units and were depicted as mean ± SE or median ± MAD for easy interpretation. A repeated measures MANOVA, used to reduce the type 1 error in follow up ANOVAs, and univariate ANOVAs were run to determine if statistical differences existed within sampling days, within the day and treatment interaction, and between treatments for algal cell density and stocking density treatments in Experiment 4 1, water exchange rates and prey density treatments in Experiment 4 2, and pho toperiods in Experiment 4 3. Trace multivariate statistic was used as the test statistic. A P value of 0.05

PAGE 125

125 was considered statistically significant for and An additional series of One way with Tukeys HSD means separation test were conducted for SL, BD, and Ed at each time period to determine when differences in treatments occurred. A Bonferroni correction of n , where n is sample size, was applied to multiple comparison tests to reduce type I error. Differences in specific growth rates and condition of larvae in treatments were assessed using a one way ANOVA with Tukeys HSD means separation test. Non parametric data was assessed using a Kruskal Wallace one way ANOVA with a pairwise multiple comparison. A significance level will be used to determine significance. Water Quality All water quality parameters were recorded within reported safe, normal limits ( Table 4 1 ). Results 4 1 Algal cell density and stocking density Survivorship was found to be significantly different between treatments (F 8,18 =45.117, P <0.001). A Tukeys HSD means separation test found that treatments 3 (mean, SE, +SE = 0.52, SE 0.39, +SE 0.66 %, n =3), 6 (mean, SE, +SE = 0.05, SE 0.05, +SE 0.15 %, n =3), and 9 (mean, SE, +SE = 0.73, SE 0.26, +SE 0.31 %, n =3) had significantly lower survival than all other treatments. Treatments 4 (mean, SE, +SE = 15.30, SE 1.79, +SE 1.88 %, n =3), 1 (mean, SE, +SE = 23.89, SE 5.05, +SE 5.44 %, n =3), and 7 (mean, SE, +SE = 31.92, SE 5.70, +SE 5.98 %, n =3) had significantly greater survival than treatments 3, 6, and 9. Mean survival of treatment 5 (mean, SE, +SE = 36.49, SE 1.96, +SE 1.99 %, n =3), 8 (mean, SE, +SE = 38.87, SE 1.50, +SE

PAGE 126

126 1.51 %, n =3), and 2 (mean, SE, +SE = 40.64, SE 6.29, +SE 6.44, n =3) had significantly greater survival than treatment 4 but were statistically the same as treatment 1 and 7 ( Figure 4 1 ). In Experiment 4 1, several treatments (3, 6, 9) had less than 5 larvae present at the 10 dph sampling period, therefore, treatments were not included in growth data analyses because sample size was to small to be representative of the population. The repeated measures MANOVA analyses detected significant multivariate effects for treatments Trace F 15, 252 =2.876, P <0.001), day Trace F 9, 76 =696.513, P <0.001), and the interaction among day and treatment Trace F 45, 400 =2.277, P <0.001). Univariate between subject analyses indicated that SL (F 5, 84 =1.966, P =0.092) was not significantly different between treatments, however, BD (F 5, 84 =6.172, P <0.001) and Ed (F 5, 84 =2.697, P =0.026) were. Sphericity assumptions were met for SL but not for BD and Ed so the Greenhouse Geisser correction was used in within subject univariate analyses on BD and Ed. Within subject univariate analyses indicated that SL (F 3. 252 =521.482, P <0.001), Ed (Greenhouse Geisser F 2.319, 194.811 =1,030.899, P <0.001), and BD (Greenhouse Geisser F 2.630, 220.918 =526.114, P <0.001) were significantly different at each sampling period. There was a significant interaction of day and treatment for SL (F 15, 252 =2.360, P =0.003), BD (Greenhouse Geisser F 13.150, 220.918 =3.533, P <0.001), and Ed (Greenhouse Geisser F 11.596, 194.811 =2.396, P =0.004). assessing the day by treatment interaction showed that no significant size differences between treatments in SL (F 5, 84 =0.736, P =0.599) ( Figure 4 2 A ), BD (F 5, 84 =1.428, P =0.223) ( Figure 4 2 B ), or Ed (F 5, 84 =0.152, P =0.979) ( Figure 4 2 C ) were

PAGE 127

127 present at 4 dph. Di fferences were detected between treatments in SL at 6 dph (F 5, 84 =4.116, P =0.002), 8 dph (F 5, 84 =2.440, P =0.041), but not at 10 dph (F 5, 84 =1.895, P =0.104). The post hoc analyses with a Bonferroni correction, assessed at revealed that SL of 6 dph larvae in treatment 1 (mean ± SE = 2.569 ± 0.017 mm, n =15) was significantly less than SL of treatment 5 (mean ± SE = 2.674 ± 0.012 mm, n =15). At 8 and 10 dph there were no significant differences between SL in treatments ( Figure 4 2 A ). At 6 and 8 and 10 dph significant differences in BD were detected between treatments (F 5, 84 =4.581, P =0.001), (F 5, 84 =3.424, P =0.007), and (F 5, 84 =5.951, P <0.001). The post hoc analyses with a Bonferroni correction, assessed at revealed that body depth at 6 dph was significantly greater in treatment 5 (mean ± SE = 0.401 ± 0.022 mm, n =15) than in treatments 1 (mean ± SE = 0.359 ± 0.008 mm, n =15) but, at 8 dph there were no significant differences detected between BD in treatments. BD at 10 dph of treatment 2 (mean ± SE = 0.504 ± 0.009 mm, n =15) was significantly greater than treatments 8 (mean ± SE = 0.448 ± 0.005 mm, n =15) and 5 (mean ± SE = 0.448 ± 0.006 mm, n =15) ( Figure 4 2 B ). Significant differences in Ed between treatments were detected at 6 dph Ed (F 5, 84 =5.501, P <0.001), 8 dph Ed (F 5, 84 =3.395, P =0.008), but not at 10 dph Ed (F 5, 84 =0.905, P =0.482). The post hoc analyses with a Bonferroni correction, assessed at revealed that eye diameter of treatment 5 (mean ± SE = 0.210 ± 0.002 mm, n =15) was significantly greater than eye diameter treatment 1 (mean ± SE = 0.192 ± 0.002 mm, n =15) at 6 dph, but at 8 and 10 dph there were no significant differences between Ed in treatments ( Figure 4 2 C ). The specific growth rate of SL did not meet the parametric assumption, therefore was analyzed using a Kruskal Wallis one way ANOVA with a pairwise comp arison.

PAGE 128

128 SGR of SL was significantly different between treatments (H 5 =12.506, P =0.028). Distributions of notochord length SGR in treatment 1 (median, MAD, +MAD = 3.23, MAD 0.28, +MAD 0.29 %, n =15) was significantly greater than SGR of treatment 2 (median, MAD, +MAD = 2.51, MAD 0.41, +MAD 0.45 %, n =15) ( Table 4 2 ). Growth rates in all other treatments were not statistically different. SGR of BD was significantly different between treatments (F 5, 85 =4.023, P =0.003) with treatment 2 (mean, SE, +SE = 9.72, SE 0.67, +SE 0.69 %, n =15) larvae having a significantly greater body depth SGR than treatment 5 (mean, SE, +SE = 6.53, SE 0.50, ±SE 0.52 %, n =15) ( Table 4 2 ). Body depths of all other treatments were statistically similar. SGR of Ed was not found to be significantly different between treatments (F 5, 85 =0.606, P =0.696) ( Table 4 2 ). A Kruskal Wallis one way ANOVA with a multiple comparison test analyzed CI data. CI of larvae was found to be significantly different (H 5 =24.267, P <0.001) between treatment 2 (median ± MAD = 0.182 ± 0.006, n =15), and treatments 4 (median ± MAD = 0.152 ± 0.005 mm 3 , n =15), treatment 5 (median ± MAD = 0.156 ± 0.003 mm 3 , n =15), and treatment 8 (median ± MAD = 0.156 ± 0.003 mm 3 , n =15) ( Table 4 3 ). 4 2 2 Water exchange rates and prey density No significant differences in survivorship (F 5, 12 =3.065, P =0.052) were detected between water exchange and prey density treatments. Treatments 5 and 6 (700% day 1 ) had lower larval survival, 29.45% and 29.76% respectively, than all other treatments 1 (45.67%), 2 (47.13%), 3 (42.13%), and 4 (39.66%) ( Figure 4 3 ). The repeated measures MANOVA analyses detected significant multivariate effects for treatments Trace F 15, 252 =3.584, P <0.001), sampling days Trace F 9, 76 =402.637, P <0.001), and the interaction between day and treatment

PAGE 129

129 Trace F 45, 400 =3.885, P <0.001). Univariate between subject analyses indicated that SL (F 5, 84 =1.563, P =0.179) and Ed (F 5, 84 =2.059, P =0.790) were not significantly different between treatments, however, BD was (F 5, 84 =10.161, P <0.001). Sphericity assumptions were met for SL and Ed, but not for BD so the Greenhouse Geisser correction was used in within subject univariate analyses on BD. Within subject univariate analyses indicated that SL (F 3.252 =521.165, P <0.001), BD (Greenhouse Geisser F 2.481,208.373 =387.569, P <0.001), and Ed (F 3, 252 =788.698, P <0.001) were significantly different at each sampling period. There was a significant interaction between da y and treatment for SL (F 15, 252 =3.595, P <0.001), BD (Greenhouse Geisser F 12.403, 208.373 =4.997, P <0.001), and Ed (F 15, 252 =5.687, P <0.001). ANOVAs assessing the day by treatment interaction showed that there were no significant differences between trea tments in SL (F 5 =1.325, P =0.262) ( Figure 4 4 A ), BD (F 5 =0.857, P =0.514) ( Figure 4 4 B ), or Ed (F 5 =1.331, P =0.259) ( Figure 4 4 C ) were present at 4 dph. At 6, 8, and 10 dph, significant differences were detected among treatments in SL (F 5 =4.742, P =0.001), (F 5 =4.856, P =0.001), and (F 5 =2.519, P =0.036). Post hoc analyses with a Bonferroni correction wer e assessed at Standard length at 6 dph was significantly lower in treatment 1 (mean ± SE = 2.569 ± 0.017 mm, n =15) than in treatments 5 (mean ± SE = 2.674 mm, n =15) and 6 (mean ± SE = 2.675 ± 0.016 mm, n =15). At 8 dph SL was significantly greater in treatments 2 (mean ± SE = 2.771 ± 0.018 mm, n =15) and 4 (mean ± SE = 2. 780 ± 0.024 mm, n =15) compared to treatment 3 (mean ± SE = 2.664 ± 0.009 mm, n =15) and at 10 dph no differences in SL were detected ( Figure 4 4 A ). BD at 6, 8, 10 dph was significantly different (F 5 =4.414, P =0.001), (F 5 =4.002, P =0.003), (F 5 =14.098, P <0.001). Post hoc analyses were

PAGE 130

130 assessed at At 6 dph body depth of treatment 1 (mean ± SE = 0.360 ± 0.009 mm, n =15) was significantly lower than body depth of treatment 4 (mean ± SE = 0.392 ± 0.004 mm, n =15). At 8 dph BD was not significantly different among treatments. At 10 dph BD of treatment 1 (mean ± SE = 0.487 ± 0.006 mm, n =15) and treatment 2 (mean ± SE = 0.494 ± 0.009 mm, n =15) were significantly greater than treatments 3 (mean ± SE = 0.441 ± 0.008 mm, n =15), 5 (mean ± SE = 0.448 ± 0.006 mm, n =15), and 6 (mean ± SE = 0.423 ± 0.007 mm, n =15) ( Figure 4 4 B ). Ed at 6, 8 and 10 dph were significantly different between treatments (F 5 =8.374, P < 0.001), (F 5 =4.811, P =0.001), and (F 5 =4.994, P <0.001) ( Figure 4 4 C ). Post hoc analyses with a Bonferroni correction were assessed at At 6 dph , eye diameter of treatment 1 (mean ± SE = 0.187 ± 0.002 mm, n =15) was significantly lower than eye diameter of treatments 3 (mean ± SE = 2.04 ± 0.004 mm, n =15), 5 (mean ± SE = 0.210 ± 0.003 mm, n =15), and 6 (mean ± SE = 0.207 ± 0.003 mm, n =15). At 8 dph Ed was not significantly different among treatments. At 10 dph mean Ed of treatment 5 (mean ± SE = 0.232 ± 0.004 mm, n =15) was significantly lower than mean Ed of treatments 4 (mean ± SE = 0.251 ± 0.004 mm, n =15) and 1 (mean ± SE = 0.251 ± 0.002 mm, n =15) ( Figure 4 4 C ). Specific growth rates of SL were not found to be significantly different among treatments (F 5, 84 =2.211, P =0.061), however, SGR of body depth was found to be significantly different (F 5, 84 =8.188, P <0.001) ( Table 4 4 ). Treatments 1 (mean, SE, +SE = 6.80, SE 0.249, +SE 0.253 %, n =15) and 2 (mean, SE, +SE = 7.11, SE 0.336, +SE 0.343 %, n =15) had significantly greater SGR than treatment 3 (mean, SE, +SE = 5.91, SE 0.292, +SE 0.299 %, n =15), treatment 5 (mean, SE, +SE = 5.27, SE 0.379, +SE 0.392 %, n =15), and treatment 6 (mean, SE, +SE = 4.87, SE 0.289, +SE 0.297 %,

PAGE 131

131 n =15) ( Table 4 4 ). A Kruskal Wallis one way ANOVA found no significant differences in eye diameter SGR between treatments (H 5 =7.867, P =0.164) ( Table 4 4 ). The condition index of larvae was assessed using a nonparametric Kruskal Wallis one way ANOVA. Significant differences were detected between CI of treatments (H 5 =42.674, P <0.001). Condition indexes of treatment 1 (median ± MAD = 0.165 ± 0.003, n =15) and treatment 2 (median ± MAD = 0.174 ± 0.011, n =15) were significantly greater than treatments 3 (median ± MAD = 0.151 ± 0.003, n =15) and 6 (median ± MAD = 0.149 ± 0.004, n =15) ( Table 4 5 ). Additionally, treatment 4 (median ± MAD = 0.159 ± 0.006, n =15) was significantly greater than treatment 6 ( Table 4 5 ). 4 3 3 Photoperiod Analysis of survivorship data revealed that there were no significant differences in the number of larvae surviving to 10 dph among photoperiod treatments (F 2, 15 =2.053, P =0.163). In all treatments between 37 47% of larvae survived ( Figure 4 5 ). The repeated measures MANOVA analyses detected significant multivariate effects for treatments Trace F 6, 172 =12.290, P <0.001), sampling days Trace F 9, 79 =394.028, P <0.001) , and the interaction between day and treatment Trace F 18, 160 =3.954, P < 0.001). Univariate between subject analyses indicated that SL (F 2, 87 =38.092, P <0.001) and Ed (F 2, 87 =8.013, P =0.001) were significantly different between treatments but that BD was not (F 2, 87 =1.262, P =0.288). Sphericity assumptions were met for SL and Ed, but not for BD so the Greenhouse Geisser correction was used in within subject univariate analyses on BD. Within subject univariate analyses indicated that SL (F 3, 261 =584.822, P <0.001), BD (Greenhouse Geisser F 2.572, 223.785 =444.231, P <0.001), and Ed (F 3, 261 =479.698, P <0.001) were significantly different at each sampling period. There was a significant interaction of day

PAGE 132

132 and treatment for SL (F 6, 261 =7.307, P <0.001), BD (Greenhouse Geisser F 5.144, 223.785 =3.072, P =0.01), and Ed (F 6, 261 =4.573, P <0.001). ANOVAs assessing the day by treatment interaction showed that no significant differences among treatments in SL (F 2, 87 =0.005, P =0.995) ( Figure 4 6 A ), BD (F 2, 87 =0.584, P =0.560) ( Figure 4 6 B ), or Ed (F 2, 87 =0.663, P=0.518) ( Figure 4 6 C ) were present at 4 dph. At 6, 8, and 10 dph significant differences were detected among treatments in SL (F 2, 87 =8.034, P =0.001), (F 2, 87 =4.972, P =0.009), and (F 2, 87 =68.385, P <0.001). Post hoc analyses were assessed at Standard length at 6 dph was significantly greater in treatment 2 (mean ± SE = 2.562 ± 0.021 mm, n =30) than in treatments 1 (mean ± SE = 2.454 ± 0.017 mm, n =30) and 3 (mean ± SE = 2.488 ± 0.020 mm, n =30). At 8 dph, SL was not significantly different between treatments. At 10 dph, SL of treatment 1 (mean ± SE = 2.894 ± 0.014 mm, n =30) was significantly less than SL of treatments 2 (mean ± SE = 3.142 ±0.020 mm, n =30) and 3 (mean ± SE = 3.113 ± 0.015 mm, n =30) ( Figure 4 6 A ). BD at 6 dph was significantly different between treatments (F 2, 87 =3.741, P =0.028) but no significant differences were present at 8 dph (F 2, 87 =2.408, P =0.096) or 10 dph (F 2, 87 =1.861, P =0.162). Post hoc analyses were assessed at Body depth of all treatments was statistically the same at 6, 8 and 10 dph assessed at the Bonferroni corrected value ( Figure 4 6 B ). At 6 dph, Ed was significantly different between treatments (F 2, 87 =3.255, P =0.043) but not at 8 dph (F 2, 87 =0.067, P =0.935). At 10 dph, Ed was significantly different between treatments Ed (F 2, 87 =17.673, P <0.001). Post hoc analyses were assess ed at Treatments of Ed were statistically the same at 6 and 8 dph. At 10 dph, Ed of treatment 1 (mean ± SE = 0.258 ± 0.003 mm, n =30) was significantly lower than Ed of treatments 2 (mean ± SE

PAGE 133

133 = 0.274 ± 0.004 mm, n =30) and 3 (mean ± SE = 0.283 ± 0.003 mm, n =30) ( Figure 4 6 C ). The specific growth rate of SL was significantly different (F 2 87 =16.980, P <0.001) between treatments. Treatment 1 (mean, SE, +SE = 3.32, SE 0.177, +SE 0.182 %, n =30) exhibited significantly reduced SGR compared to treatments 2 (mean, SE, +SE = 4.69, SE 0.173, +SE 0.176 %, n =30) and 3 (mean, SE, +SE = 4.51, SE 0.193, +SE 0.197 %, n =30) ( Table 4 6 ). Data for SGR of body depth was nonparametric and a Kruskal Wallis one way ANOVA found no significant differences among treatments (H 2 =3.350, P =0.187) ( Table 4 6 ). Specific growth rates of Ed were found to be significantly different (F 2, 87 =5.388, P =0.006) between treatments. SGR of treatment 1 (mean, SE, +SE = 6.55, SE 0.302, +SE 0.308, n =30) was significantly lower than SGR of treatment 3 (mean, SE, +SE = 8.04, SE 0.201, +SE 0.204, n =30) ( Table 4 6 ). The condition index of larvae varied significantly between treatments (H 2 =21.197, P <0.001). Treatment 1 (median ± MAD = 0.164 ± 0.007, n =30) had a significantly greater CI than did treatment 2 (median ± MAD = 0.157 ± 0.003, n =30). Treatment 3 was statistically the same as treatments 1 and 2 ( Table 4 7 ). Discussion Larval Size, Growth, Condition, and Survival Survival of aquacultured marine ornamental species to metamorphosis is low. The highest survival rates of aquacultured marine ornamental species occur in clownfish species with 65 95% survival (Kumar et al., 2012; Ignatius et al., 2001). Survival rates of most other species are low compared to clown fish, Abudefduf saxitilis (6.6%) (Wittenrich et al., 2012), Neopomacentrus cyanomos (20%) (Setu et al., 2010), Chaetodontoplus septentrionalis (5.5%) (Leu et al., 2013), and Centropyge debelius

PAGE 134

134 (1.3%) (Baensch and Tamaru, 2009). Many of the bottlenecks in marine larval culture occur during the early larval period. By optimizing early larval growth, survival can be increased. In Experiment 4 1, algal cell density and stocking density parameters from Experiments 3 2, 3 3, and 3 5 in Chapter 3 of this thesis were selected to assess size, growth, condition, and survival of larvae. Experiment 4 2 assessed prey density and water exchange rates from Experiments 3 4 and 3 7 in Chapter 3 of this thesis for size, growth, condition, and survival of C. miliaris larvae to 10 dph. 4 1 Algal cell density and stoc king density It did not appear that stocking densities of 15 25 larvae L 1 had any significant effect on survival of C. miliaris larvae; however, algal cell density influenced survival of larvae. Survival of C. miliaris to 10 dph larvae revealed that mean survival was less than 1% in treatments 3, 6, and 9, which had algal cell densities of 800,000 830,000 cells mL 1 ( Figure 4 1 ) No measurements for larval size or growth were obtained from treatments 3, 6, and 9, because less than 5 larvae were surviving in all replicates. Mean survival in treatments 2, 5, and 8 was greater than all other treatments with 36 41% survival in all replicates, but due to large variances in treatments 7 and 1, statistical analysis indicated that only treatment 4 was significantly different from treatments 2, 5, and 8 ( Figure 4 1 ). Algal cell densities of 500,000 530,000 cells mL 1 increased survival of C. miliaris larvae over algal cell densities of 180,000 200,000 (treatments 1, 4, and 7) and 800,000 830,000 (treatments 3, 6, and 9) cells mL 1 . No trends were observed between larval size and survival. Size differences between treatments were present in SL, BD, and Ed at 6 dph but measurements at 6 dph were not indicative of larval size at 8 or 10 dph, growth, or survival ( Figure 4 2 ). No differences in SL or Ed were present among treatments at 8 and 10 dph. Body depth of

PAGE 135

135 C. miliaris larvae was greatest in treatment 2 at 8 and 10 dph, however, neither SL or Ed of treatment 2 was increased. Additionally, body depth of treatments 5 and 8 were lower than all other treatments indicating that there was no interaction between larval size and survival ( Figure 4 2 ). Notochord length SGR was highest in treatment 1, which had the largest SL of all treatments. Similarly body depth SGR was greatest for treatment 2 which had the largest body depth of all treatments. No differences in eye diameter SGR were detected ( Table 4 2 ). Condition index of larvae indicated that treatments 2, 1, and 7 had the highest index scores ( Table 4 3 ). Survival was highest in treatment 2, which also had the highest CI, and body depth SGR. In Experiment 4 1 it appears that body depth SGR, CI, and survival are related, specifically, increased CI appears to be the most sensitive to culture parameter changes. In Chapter 3, results of algal cell density revealed that the proportion of larvae feeding and feeding intensity at 4 dph was enhanced at algal cell densities of ~810,000 cells mL 1 in both Trials 1 and 2 and at ~520,000 cells mL 1 in Trial 1. The results of Experiment 4 1 indicate that prolonged feeding at ~810,000 cells mL 1 decreased survival to 10 dph. This result seems contrary to the expected results; however, research has shown that turbidity preferences of larvae change over time (Shaw et al., 2006). Research on Rhombosolea tapirina found that feeding was enhanced at different larval stages by different algal turbidities. Nine dph larvae exhibited consistent feeding levels across all turbidities (0 40 NTU), while 16 dph larvae exhibited enhanced feeding at 10 NTU, and 20 dph larvae fed the most at 40 NTU (Shaw et al., 2006). Carton (2005) studied the effects of algal induced turbidity on early larvae of Seriola lalandi and concluded that suppressed feeding occurred at algal densities of 320,000 cells mL 1 ,

PAGE 136

136 however, this does not appear to be the case with C. miliaris because the proportion of larvae feeding and feeding intensity were enhanced at ~810,000 cells mL 1 in Experiment 3 2. Another possible influence on larval survival in the 800,000 830,000 cell mL 1 treatments may have been dissolved oxygen levels. Dissolved Oxygen levels recorded in treatments 3, 6, and 9 (82.3 98.4%) were lower than those recorded in other treatments (93.1 99.2%) but research on Danio rerio and Acanthopargus butcheri found that survival and growth were normal at oxygen levels above 60% saturation (Barrionuevo et al., 2010; Hassell et al., 2008). The body height to body length condition index of Gadus morhua was significantly affected by nauplii d ensity and total zooplankter density but not mean or maximum chlorophyll concentration in the wild (Koslow et al., 1985). C. miliaris larvae appear to follow a similar trend to Gadus morhua larvae in that the condition index of larvae were not separated by different algal cell density treatments. Stocking density seems to have a slight effect on condition of larvae as larvae stocked at 15 larvae L 1 exhibited the two highest condition indexes. Larvae of treatment 2 exhibited the best condition, which aligns with survival to 10 dph. Further investigation using a dry weight:head length ratio may provide more insight into the sensitivity of larvae to other environmental parameters (Koslow et al., 1985). A stocking density of 15 larvae L 1 and algal cell density of 500,000 530,000 cells mL 1 had the highest survival, larvae condition, BD specific growth rate, and BD size. 4 2 2 Water exchange rates and prey density Water exchange rates of 100% day 1 , 300% day 1 , and 700% day 1 and prey densities of 1 and 10 nauplii mL 1 did not significantly effect survival of C. miliaris larvae.

PAGE 137

137 Survival was lower (~29%) at higher water exchange rates (700% day 1 ) than all other exchange rates, however, statistical analyses did not detect any significant differences between treatments ( Figure 4 3 ). Similar to Experiment 4 1, larval size at 6 dph was not an indicator of increased size at later sampling periods. At 10 dph, SL and Ed of treatments 1, 2, 3, and 4 were greater than SL and Ed of treatments 5 and 6. Body depth of treatments 1, 2, and 4 was increased, but was suppressed in treatment 3 ( Figure 4 4 ). Larval size and survival were increased in treatments with water exchange rates of 300% day 1 over w ater exchange rates of 700% day 1 . Increased larval size appears to coincide with increased survival except for in treatment 3. No differences in SL specific growth rates or Ed specific growth rates were observed ( Table 4 4 ). BD size ( Figure 4 4 C ), body depth SGR ( Table 4 4 ), ( Table 4 5 ) were highest in treatments 2, 1, and 4 and coincided with larval survival ( Figure 4 3 ). Increased water exchange rates have suppressed early growth in larvae of some species (Tandler and Helps, 1985) and had negligible effects on others (Chesney, 1989). Increased exchange rates can result in excess flow and turbulence in the tank which is known to over power larvae, reducing prey capture and even damaging larvae (MacKenzie et al., 1994; Dower et al., 1997; MacKenzie and Kiøroboe, 2000), however, it does not appear that this threshold was reached in Experiment 4 2. Tandler and Helps (1985) reported that a water exchange rate of 0% day 1 had a positive growth effect on Sparus aurata larvae over an exchange rate of 25% day 1 for the first 10 dph aligning with the results of this study. Alessio (1975) reported similar findings as well and suggested that the negative effects of increased water exchange were likely due to the amplification of physical perturbation in the culture tank. Larger water exchanges can be

PAGE 138

138 beneficial for flushing exc ess prey items and keeping water quality parameters within safe limits. Further studies into prey retention times and pulsed water exchanges may be advantageous to creating an optimal feeding environment for certain time periods and management of water quality parameters and prey items at others. In Experiment 3 4 in Chapter 3, no difference in the proportion of larvae feeding or feeding intensity were observed between prey densities of 1 15 nauplii mL 1 , however, Experiment 3 4 was conducted under static water exchange conditions. Duray et al., (1996) observed increased survival of Epinephelus suillus at rotifer densities of 30 rotifers mL 1 with water exchanges of 30 50% day 1 . Although C. miliaris survival was highest in treatment 2, which had a prey density of 10 nauplii mL 1 , other treatments of 10 nauplii mL 1 did not exhibit increased survival over low prey density treatments (1 nauplii mL 1 ) ( Figure 4 3 ). Water exchange rate appeared to be more influential on l arval survival than did prey density. 4 3 3 Photoperiod Survivorship of larvae was not significantly different among photoperiod treatments. In all treatments greater than 37% of larvae survived to 10 dph, with the greatest survival (47%) in 18 L:6 D and the lowest survival measured in continuous lighting (37%) ( Figure 4 5 ). No trends in larval size were present between sampling days. Larval SL, BD, and Ed of treatment 1 (14 L:10 D) were less at 10 dph than the other photoperiod treatments. Body depth and eye diameter were greater in the 24 L:0 D treatment but SL was greatest in the 18 L:6 D treatment ( Figure 4 6 ). Growth rates of SL and Ed were highest in treatment 3 but no differences between BD specific growth rates were found ( Table 4 6 ). Other species such Siganus gattatus (Duray and Kohno, 1988) and Gadus morhua (Puvanendran and Brown, 2002) exhibited increased survival

PAGE 139

139 and size in continuous light compared to reduced photoperiods, however, Melanogrammus aeglefinus ( Downing and Litvak, 2000) did not exhibit increased survival in continuous light compared to a 15 L photoperiod. C. miliaris had the greatest survival (47%) at a photoperiod of 18 L:6 D. CI was greatest in treatment 1 ( Table 4 7 ) which had the lowest size and SGR of SL and Ed. Survivorship of larvae did not match CI as was found in Experiment 4 1 and 4 2. Unlike Experiments 4 1 and 4 2, no differences in body depth size or SGR were observed in Experiment 4 3. Increased photoperiod gives larvae the opportunity to ingest more prey items and absorb more energy for development (Kiyono and Hirano, 1981; Duray and Kohno, 1988) explaining the increased SL and Ed of larvae, but failing to explain the lack of difference in BD between treatments. The description of larval growth described in Chapter 2 of this thesis stated that larvae, specifically around 9 11 dph exhibited a large increase in growth. Experiments to 11 or 13 dph would likely explain the absence of BD size and SGR differences between treatments. The impacts of factors such as larvae size, condition, SGR, and their influences throughout development are still poorly understood. Larval condition at one point in time likely influences survival and growth at another, creating a complex series of interactions throughout larval development. Optimization and enhancement of early larval culture conditions may reduce acute and chronic mortality throughout devel opment. An apparent trend was present in Experiments 4 1 and 4 2 establishing that increased BD size and more so growth rate was indicative of increased larval condition and survival. Experiment 4 3 did not come to a similar conclusion; however, there were no significant differences in body depth between treatments. Optimal culture

PAGE 140

140 conditions for C. miliaris larvae to 10 dph as observed in Experiments 4 1, 4 2, and 4 3 were a stocking density of 15 larvae L 1 and an algal cell density of 500,000 530,000 cell mL 1 , under low exchange rates (300% day 1 ) at prey densities between 1 or 10 nauplii mL 1 , and at a photoperiod of 18 L:6 D. There are still additional factors such as nutrition and prey switches that may be the cause of mortalities throughout development; however, by optimizing early larvae condition larvae are more likely to be resilient to substandard culture conditions later in development. Ultimately, measurement of SL, BD, and Ed through metamorphosis would be ideal for analysis to assess the impacts of environmental effects on larval condition, growth, and survival throughout development.

PAGE 141

141 Table 4 1. Culture parameters for Experiments 4 1, 4 2, and 4 3. Salinity, temperature, dissolved oxygen (D.O.), pH, total ammon ia nitrogen (TAN), nitrite nitrogen (NO 2 N), and nitrate (NO 3 ) were measured daily. Values are given as the mean ± standard deviation and the range. The number of samples ( n =) for the trial is given. Experiment 4 1 n =270.00 Experiment 4 2 n =180.00 Experiment 4 3 n =180.00 Tank color Black sides /white bottom Black sides /white bottom Black sides /white bottom Water treatment 1 1 1 Salinity (g L 1 ) 33.00 ± 1.00 33.50 ± 1.00 33.0 ± 1.50 Water temperature °C 25.16 ± 0.82 25.41 ± 1.10 25.68 ± 0.82 pH 8.10 ± 1.00 8.10 ± 1.00 8.1 ± 1.00 Dissolved oxygen (% Saturation) 93.05 ± 4.07 96.91 ± 1.51 96.96 ± 2.19 Photoperiod 14 L:10 D 14 L:10 D 14 L:10 D, 18 L:6 D, 24 L:0 D Prey item Parvocalanus crassirostris Parvocalanus crassirostris Parvocalanus crassirostris Algal cell density (cells mL 1 ) 180,000 200,000, 500 530,000, 800,000 830,000 500,000. 530,000.00 500,000. 530,000.00 Prey density (nauplii mL 1 ) 1.00 1, 10.00 1.00 Stocking density (larvae L 1 ) 15, 20, 25 15.00 15.00 Tank size (L) 14.00 14.00 14.00 Exchange rate (% day 1 ) 100 100, 300, 700 100 Light intensity (lx) 1,314.0 ± 11. 00 1,314.0 ± 11. 00 1,314.0 ± 11. 00 Total Ammonia Nitrogen (mg L 1 ) 0.041 ± 0.025 0.065 ± 0.027 0.057 ± 0.033 Nitrite nitrogen (mg L 1 ) 0.0015 ± 0.0009 0.006 ± 0.008 0.008 ± 0.008 Nitrate nitrogen (mg L 1 ) 0.006 ± 0.004 0.006 ± 0.003 0.007 ± 0.003

PAGE 142

142 Table 4 2. Growth rate of notochord length (SL), body depth (BD), and eye diameter (Ed) for treatment 1 (180,000 200,000 cells mL 1 and 15 larvae L 1 ), treatment 2 (500,000 530,000 cells mL 1 and 15 larvae L 1 ), treatment 4 (180,000 200,000 cells mL 1 and 20 larvae L 1 ), treatment 5 (500,000 530,000 cells mL 1 and 20 larvae 1 ), treatment 7 (180,000 200,000 cells mL 1 and 25 larvae L 1 ), and treatment 8 (500,000 530,000 cells mL 1 and 25 larvae L 1 ) in Experiment 4 2. Values are represented as mean, SE, +SE or median, MAD, +MAD. Median values are marked by a *. Capital superscript letters denote statistical significance between treatments evaluated at P 0.05. Parameter Growth rate (% day 1 ) SL BD Ed mean SE +SE Treatment 1 *3.23, 0.285, +0.297 A A Treatment 2 *2.51, 0.413, +0.450 B Treatment 4 *2.75, 0.265, +0.278 AB Treatment 5 *2.75, 0.150, +0.154 AB Treatment 7 *2.56, 0.488, +0.538 AB Treatment 8 *2.87, 0.476, +0.518 AB Treatment 1 8.56, 0.354, +0.361 AB Treatment 2 9.72, 0.670, +0.692 A A Treatment 4 7.86, 0.594, +0.615 AB Treatment 5 6.53, 0.502, +0.521 B Treatment 7 7.88, 0.483, +0.497 AB Treatment 8 7.74, 0.412, +0.422 AB Treatment 1 6.52, 0.209, +0.212 A A Treatment 2 6.28, 0.250, +0.255 A A Treatment 4 6.25, 0.237, +0.242 A A Treatment 5 6.13, 0.261, +0.266 A A Treatment 7 6.56, 0.337, +0.345 A A Treatment 8 6.68, 0.333, +0.341 A A

PAGE 143

143 Table 4 3. Condition index (body depth (BD)/notochord length (SL)) of treatment 1 (180,000 200,000 cells mL 1 and 15 larvae L 1 ), treatment 2 (500,000 530,000 cells mL 1 and 15 Larvae l 1 ), treatment 4 (180,000 200,000 cells mL 1 and 20 larvae L 1 ), treatment 5 (500,000 530,000 cells mL 1 and 20 larvae L 1 ), treatment 7 (180,000 200,000 cells mL 1 and 25 larvae L 1 ), and treatment 8 (500,000 530,000 cells mL 1 and 25 larvae L 1 ) in Experiment 4 1. Values are represented as median ± MAD. Capital superscript letters denote significant differences between treatments evaluated at P 0.05. Parameter Condition index (BD/SL=CI) median ± MAD Treatment 1 0.165 ± 0.003 AB Treatment 2 0.182 ± 0.006 A A Treatment 4 0.152 ± 0.005 B CI Treatment 5 0.156 ± 0.003 B Treatment 7 0.160 ± 0.003 AB Treatment 8 0.156 ± 0.003 B

PAGE 144

144 Table 4 4. Growth rate of notochord length (SL), body depth (BD), and eye diameter (Ed) for treatment 1 (100% day 1 and 1 nauplii mL 1 ), treatment 2 (100% day 1 and 10 nauplii mL 1 ), treatment 3 (300% day 1 and 1 nauplii mL 1 ), treatment 4 (300% day 1 and 10 nauplii mL 1 ), treatment 5 (700% day 1 and 1 nauplii mL 1 ), and treatment 6 (700% day 1 and 10 nauplii mL 1 ) in Experiment 4 2. Values are represented as mean, SE, +SE or median, MAD, +MAD. Median values are marked by a *. Capital superscript letters denote statistical significance between treatments evaluated at P 0.05. Parameter Growth rate mean/*median, SE/*MAD, +SE/*MAD Treatment 1 3.58, 0.787, +0.881 A A Treatment 2 2.80, 0.852, +1.000 A A Treatment 3 2.83, 0.513, +0.563 A A SL Treatment 4 2.64, 0.779, +0.910 A A Treatment 5 2.92, 0.859, +1.000 A A Treatment 6 2.70, 0.898, +1.070 A A Treatment 1 6.80, 0.249, +0.253 A A Treatment 2 7.11, 0.336, +0.343 A A BD Treatment 3 5.03, 0.339, +0.350 B Treatment 4 5.91, 0.292, +0.299 AB Treatment 5 5.27, 0.379, +0.392 B Treatment 6 4.87, 0.289, +0.297 B Treatment 1 *6.35, 0.809, +0.860 A A Treatment 2 *7.05, 0.908, +0.966 A A Ed Treatment 3 *6.09, 3.280, +4.420 A A Treatment 4 *6.89, 4.120, +5.780 A A Treatment 5 *6.21, 3.340, +4.510 A A Treatment 6 *6.51, 3.670, +5.030 A A

PAGE 145

145 Table 4 5. Condition index (body depth (BD)/notochord length (SL) for treatment 1 (100% day 1 and 1 nauplii mL 1 ), treatment 2 (100% day 1 and 10 nauplii mL 1 ), treatment 3 (300% day 1 and 1 nauplii mL 1 ), treatment 4 (300% day 1 and 10 nauplii mL 1 ), treatment 5 (700% day 1 and 1 nauplii mL 1 ), and treatment 6 (700% day 1 and 10 nauplii mL 1 ) in Experiment 4 2. Values are represented as median ± MAD. Capital superscript letters denote significant differences between treatments evaluated at P 0.05. Parameter Condition index (BD/SL=CI) median ± MAD Treatment 1 0.165 ± 0.003 A AA Treatment 2 0.174 ± 0.011 A AA Treatment 3 0.151 ± 0.003 BC CI Treatment 4 0.159 ± 0.006 AB Treatment 5 0.158 ± 0.003 ABC Treatment 6 0.149 ± 0.004 C Table 4 6. Growth rate of notochord length (SL), body depth (BD), and eye diameter (Ed) for treatment 1 (12 L:12 D), treatment 2 (18 L:6 D), and treatment 3 (24 L:0 D) in Experiment 4 3. Values are represented as mean, SE, +SE or median, MAD, +MAD. Median values are marked by a *. Capital superscript letters denote statistical significance between treatments evaluated at P 0.05. Parameter Growth rate (% day 1 ) SL BD Ed *median * MAD *+MAD Treatment 1 3.32, 0.177, +0.182 B Treatment 2 4.69, 0.173, +0.176 A A Treatment 3 4.51, 0.193, +0.197 A A Treatment 1 *5.62, 5.31, +11.200 A A Treatment 2 *6.26, 5.86, +12.120 A A Treatment 3 *6.39, 5.98, +12.310 A A Treatment 1 6.55, 0.302, +0.308 B Treatment 2 7.26, 0.238, +0.241 AB Treatment 3 8.04, 0.201, +0.204 A A

PAGE 146

146 Table 4 7. Condition index (body depth (BD)/notochord length (SL)) for treatment 1 (12 L:12 D), treatment 2 (18 L:6 D), and treatment 3 (24 L:0 D) in Experiment 4 3. Values are represented as median ± MAD. Capital superscript letters denote significant differences between treatments evaluated at P 0.05. Parameter Condition index (BD/SL=CI) median ± MAD Treatment 1 0.164 ± 0.007 A A CI Treatment 2 0.157 ± 0.003 B Treatment 3 0.161 ± 0.005 AB

PAGE 147

147 100 90 80 70 60 50 A 40 AB 30 20 A AB A B 10 C C C 0 1 2 3 4 5 6 7 8 9 Treatments Figure 4 1. Survival (%) of C. miliaris larvae in treatment 1 (180,000 200,000 cells mL 1 and 15 larvae L 1 ), treatment 2 (500,000 530,000 cells mL 1 and 15 larvae L 1 ), treatment 3 (800,000 830,000 cells mL 1 and 15 larvae L 1 ), treatment 4 (180,000 200,000 cells mL 1 and 20 larvae L 1 ), treatment 5 (500,000 530,000 cells mL 1 and 20 larvae L 1 ), treatment 6 (800,000 830,000 cells mL 1 and 20 larvae L 1 ), treatment 7 (180,000 200,000 cells mL 1 and 25 larvae L 1 ), treatment 8 (500,000 530,000 cells mL 1 and 25 larvae L 1 ), and treatment 9 (800,000 830,000 cells mL 1 and 25 larvae L 1 ). All data was arcsine(sqrt) transformed for statistical analysis and back transformed to original uni ts. Data are represented as mean, SE, +SE. Capital letters indicate significant differences. Suvival (%)

PAGE 148

148 A 3.500 3.000 2.500 2.000 1.500 1.000 0.500 0.000 4 dph 6 dph 8 dph 10 dph Treatment 1 Treatment 2 Treatment 4 Treatment 5 Treatment 7 Treatment 8 Figure 4 2. Measurements of (A) notochord length (SL), (B) body depth (BD), and (C) eye diameter (Ed) in treatment 1 (180,000 200,000 cells mL 1 and 15 larvae L 1 ), treatment 2 (500,000 530,000 cells mL 1 and 15 larvae L 1 ), treatment 3 (800,000 830,000 cells mL 1 and 15 larvae L 1 ), treatment 4 (180,000 200,000 cells mL 1 and 20 larvae L 1 ), treatment 5 (500,000 530,000 cells mL 1 and 20 larvae L 1 ), treatment 6 (800,000 830,000 cells mL 1 and 20 larvae L 1 ), treatment 7 (180,000 200,000 cells mL 1 and 25 larvae L 1 ), treatment 8 (500,000 530,000 cells mL 1 and 25 larvae L 1 ), and treatment 9 (800,000 830,000 cells mL 1 and 25 larvae L 1 ) at 4, 6, 8, and 10 days post hatch (dph). Treatments 3, 6, and 9 were not included because less than 5 larvae were present in each replicate so no representative sample of the population was attainable. All values are represented as mean ± SE. Capital letters indicate significant diff erences assessed at 0.0033. A A A A A A B AB AB A AB AB A A A A A A A A A A A A Notochord Length (mm)

PAGE 149

149 AB A AB AB AB B A A A A A A B AB AB A AB AB A A A A A A Treatment 1 Treatment 2 Treatment 4 Treatment 5 Treatment 7 Treatment 8 4 dph 6 dph 8 dph 10 dph C 0.300 0.250 0.200 0.150 0.100 0.050 0.000 4 dph 6 dph 8 dph 10 dph Treatment 1 Treatment 2 Treatment 4 Treatment 5 Treatment 7 Treatment 8 Figure 4 2. Continued. A A A A A A A A A A A A A A A A A A B AB AB A AB AB Body Depth (mm) Eye Diameter (mm) B 0.600 0.500 0.400 0.300 0.200 0.100 0.000

PAGE 150

100 90 80 70 60 50 40 30 20 10 0 1 2 3 4 5 6 Treatments Figure 4 3. Survival (%) of C. miliaris larvae in treatment 1 (100 % day 1 and 1 nauplii mL 1 ), treatment 2 (100 % day 1 and 10 nauplii mL 1 ), treatment 3 (300 % day 1 and 1 nauplii mL 1 ), treatment 4 (300 % day 1 and 10 nauplii mL 1 ), treatment 5 (700% day 1 and 1 nauplii mL 1 ), and treatment 6 (700% day 1 and 10 nauplii mL 1 ). All data was arcsine(sqrt) transformed for statistical analysis and back transformed to original units. Data are represented as mean, SE, +SE. No significant differences were detected between treatments. 150 A A A A A A Survival (%)

PAGE 151

151 Treatment 1 Treatment 2 Treatment 3 Treatment 4 Treatment 5 Treatment 6 4 dph 6 dph 8 dph 10 dph Figure 4 4. Measurements of (A) notochord length (SL), (B) body depth (BD), and (C) eye diameter (Ed) in treatment 1 (100 % day 1 and 1 nauplii mL 1 ), treatment 2 (100 % day 1 and 10 nauplii mL 1 ), treatment 3 (300 % day 1 and 1 nauplii mL 1 ), treatment 4 (300 % day 1 and 10 nauplii mL 1 ), treatment 5 (700% day 1 and 1 nauplii mL 1 ), and treatment 6 (700% day 1 and 10 nauplii mL 1 ) at 4, 6, 8, and 10 days post hatch (dph). All values are represented as mean ± SE. Capital letters indicate significant d ifferences assessed at 0.0033. AB A B A A B AB A AB AB AB B B A A A A A A B AB AB AB A A Notochord Length (mm) A 3.500 3.000 2.500 2.000 1.500 1.000 0.500 0.000

PAGE 152

152 B 0.600 0.500 0.400 0.300 0.200 0.100 0.000 C 0.300 0.250 0.200 0.150 0.100 0.050 0.000 4 dph 6 dph 8 dph 10 dph 4 dph 6 dph 8 dph 10 dph Treatment 1 Treatment 2 Treatment 3 Treatment 4 Treatment 5 Treatment 6 Treatment 1 Treatment 2 Treatment 3 Treatment 4 Treatment 5 Treatment 6 Figure 4 4. Continued. AB A C ABC BC C A A A A A A B A AB A A A AB A B AB AB AB AB A B A AB A A A AB A B AB A A A A A A B AB A AB A A Body Depth (mm) Eye Diameter (mm)

PAGE 153

100 90 80 70 60 50 40 30 20 10 0 1 2 3 Treatments Figure 4 5. Survival (%) of C. miliaris larvae in under different photoperiod treatments of 12 L:12 D, 18 L:6 D, and 24 L:0 D. All data was arcsine(sqrt) transformed for statistical analysis and back transformed to original units. Data are represented as mean, SE, +SE. No significant differences were detected between treatments. 153 A A A Survival (%)

PAGE 154

154 Treatment 1 Treatment 2 Treatment 3 4 dph 6 dph 8 dph 10 dph Treatment 1 Treatment 2 Treatment 3 4 dph 6 dph 8 dph 10 dph Figure 4 6. Measurements of (A) notochord length (SL), (B) body depth (BD), and (C) eye diameter (Ed) in treatment 1 (12 L:12 D), treatment 2 (18 L:6 D), treatment 3 (24 L:0 D) at 4, 6, 8, and 10 days post hatch (dph). All values are represented as mean ± SE. Capital letters indicate significant dif ferences assessed 0.0016. A B A B A A A B A A A A A A A A A A A A A A A A Standard Length (mm) Body Depth (mm) A 3.500 3.000 2.500 2.000 1.500 1.000 0.500 0.000 B 0.600 0.500 0.400 0.300 0.200 0.100 0.000

PAGE 155

155 C 0.350 0.300 0.250 0.200 0.150 0.100 0.050 0.000 4 dph 6 dph 8 dph 10 dph Treatment 1 Treatment 2 Treatment 3 Figure 4 6. Continued. B A A A A A A A A A A A Eye Diameter (mm)

PAGE 156

156 CHAPTER 5 CONCLUSION Chaetodon miliaris were evaluated as a model chaetodontidae species for marine ornamental aquaculture by assessing broodstock spawning behavior, spawning conditions, and fecundity, as well as, hydrated egg characteristics, embryology, first feeding parameters, early larval survival, and several preliminary culture trials. Methods for optimizing larval growth and survival to 10 dph were defined and larval development to 44 dph is described from preliminary culture trials. Previous attempts at culturing members of family Chaetodontidae are limited (Suzuki et al., 1980; Tanaka et al., 2001; Wittenrich and Cassiano, 2011; Baensch, 2014), provide little data on culture parameters, and have only resulted in the successful culturing of one individual of Heniochus diphreutes (Baensch, 2014). Madden and May (1977) attempted to induce spawning in two Hawaiian butterflyfish species ( C. miliaris and F. flavissimus ); however, they only acquired viable eggs from the latter and were unable to culture larvae past 6 dph. Similarly, Suzuki et al. (1980) obtained viable eggs from C. nippon in an aquarium but was unable to keep larvae alive past 8 dph. Hioki (1997) experienced similar results with C. nippon and C. modestus larvae not feeding. Tanaka et al. (2001) obtained viable eggs from natural spawns of C. modestus in an aquarium, but larvae perished at 8 dph due to starvation. In 2011, a major breakthrough in butterflyfish aquaculture was made with H. diphreutes at University of Tropical Aquaculture Laboratory, with a larva reaching 46 dph (Wittenrich and Cassiano, 2011). Larvae were fed wild zooplankton (20 collected from Tampa Bay. Gut content analysis showed that larvae selected copepod nauplii as their primary

PAGE 157

157 feed item and continued to eat nauplii until larvae perished. The first report of a cultured butterflyfish ( H. diphreutes ) surfaced in 2014, from the Hawaiian Larvae Rearing Project. The report indicates that wild eggs were collected, and development was similar to that observed by Wittenrich and Cassiano (2011) to 46 dph. Larvae fed on copepod nauplii, and transitioned to a j uvenile in approximately 100 days (Baensch, 2014). This report suggests that H. diphreutes may not be an ideal candidate for marine ornamental aquaculture since these initial trials showed points of large mortality and it had a long pelagic larval duration. However, if culture conditions could be optimized these pitfalls may be reduced. Over a period of 8 months this study established three spawning populations of C. miliaris that consistently produced viable eggs from January 1, 2014 through the termination of this study on June 30, 2014. Additionally, as noted by biologists at University of Tropical Aquaculture Laboratory, spawning continued after the termination of this study. Observed spawning behavior was similar to other species of butterflyfish in both the wild and in captivity and was not affected by the tank constraints in this study (Suzuki et al., 1980; Colin, 1989; Lobel, 1989; Tricas and Hiramoto, 1989; Londraville, 1990; Yabuta, 1997; Yabuta and Kawashima, 1997; Tanaka et al., 2001; Pratchett et al., 2014). A positive relationship between photoperiod length and spawn frequency was observed for the 3 M:8 F and 1 M:1 F groups, while a negative relationship between photoperiod length and spawn frequency was present for the 10 M:11 F group. This discrepancy is likely due to the start of spawning of the 10 M:11 F population lagging 2

PAGE 158

158 months behind that of the other two populations by two months. Photoperiod (10 h 24 min 13 h 54 min) and temperature (23.33 27.78 °C) were similar to the wild spawning season of C. miliaris and other Hawaiian chaetodontids (Lobel, 1978; Ralston, 1981) and if maintained then spawning would likely continue in captivity. Fecundity measures established that 1 M:1 F was not a efficient o ption for production. The 3 M:8 F population was the most productive male to female ratio of the three populations and produced the most spawns, most viable eggs, and most eggs per fish (Table 2 2). Fertilization rates were comparable to other marine ornamental species (Baensch and Tamaru, 2009; Callan et al., 2012a; Leu et al., 2013) and asynchronous oogenesis was confirmed in gonad samples from repeated spawning C. miliaris females. This study represents the first investigation and documentation of captive reproduction of Chaetodon miliaris . Reproductive requirements of adult C. miliaris were minimal and broodstock were capable of consistently producing viable eggs over extended periods of time, making them a suitable species for commercial production. Natural spawning eliminated the need for hormone injections. Further studies into tank size requirements, gender ratios, and broodstock diet may optimize egg production of C. miliaris . Significant correlations between increasing size of egg characteristics and increased early growth and survival of C. miliaris larvae were present. The strongest correlation was present between 0 dph survival and 3 dph survival. Similar associations between egg size and early larval growth and survival were present in Salmo gairdneri ,

PAGE 159

159 Pleuronectes platessa , and Rutilis frisii (Springate and Bromage, 2003; Kennedy et al., 2007; Imanpoor and Bagheri, 2010). Embryology and larval development of C. miliaris at 25.5 °C was similar to that of other described chaetodontids (Suzuki et al., 1980; Tanaka et al., 2001; Wittenrich and Cassiano, 2011; Baensch, 2014). Larvae grew from 1.221 mm notochord length at 0 dph to 6.530 mm notochord length and 8.173 mm total length at 44 dph. Between 9 13 dph swim bl adder inflation occurred and resulted in the first major mortality event. Flexion and the formation of the thiolichthys plates simultaneously began to occur by 24 dph, concluded by 31 dph, and resulted in another mortality event. Preliminary trials indicated P. crassirostris nauplii were an effective feeding regime for larvae. Algal induced turbidity enhanced survival, growth, and feeding of larvae. Described culture methods characterize the first culture attempts of C. miliaris . An evaluation of first feeding of C. miliaris revealed that larvae successfully fed under a variety of conditions but that certain parameters significantly increased the proportion of larvae feeding and the feeding intensity of larvae. A combination of algal induced turbidity and P. crassirostris nauplii significantly enhanced the proportion of larvae feeding and feeding intensity of C. miliaris . Further examination showed that a minimum of ~190,000 Tisochrysis lutea cells mL 1 was needed to enhance feeding of larvae and that cell densities between ~520,000 and ~810,000 cells mL 1 had the highest feeding rates. Stocking densities of 15 20 larvae l 1 and water exchange rates less than 300 % day 1 increased the proportion of larvae feeding and feeding intensity. A light intensity of 1,314 lx had the highest feeding rates. The tank sizes and prey

PAGE 160

160 densities examined in this study did not have any negative effect on first feeding of C. miliaris larvae. Results of larval size, growth, condition, and survival trials showed that larval notochord length and eye diameter were not indicators of survival, however, body depth size and growth rate appeared to relate to survival. Condition index aligned with survival, but only if significant differences were detected b etween body depth size and growth rates likely because condition index was a factor of body depth. Based upon the combined results of these parameters, the best survival larvae was observed at a stocking density of 15 larvae L 1 and algal cell density of 500,000 530,000 cells mL 1 , by a water exchange rate of 100 % day 1 at 1 or 10 nauplii mL 1 , and a photoperiod of 18 L:6 D. Aspects of spawning and larvae of C. miliaris described and tested in Chapters 2, 3, and 4 support the notion that C. miliaris can be successfully cultured. Broodstock required minimal manipulation to consistently produce usable numbers of viable eggs. Preliminary culture trials identified optimal methods for larval culture of C. miliaris . Additionally, the ability to raise larvae to 40 dph indicates that with further research and culture method development of C. miliaris is plausible.

PAGE 161

161 APPENDIX FATTY ACID ANALYSIS AND GUT HISTOLOGY OF Chaetodon miliaris Included within this appendix are additional findings from preliminary, nonessential, or unfinished trials. These include preliminary fatty acid composition of larvae, histological descriptions of eye and gut ontogeny, time to first feeding, and the transfer of larvae at 1, 4, 9, and 13 dph. Larval Fatty Acid Composition Fatty acids are an integral part of larval development. Most research has focused on required amounts of DHA (22:6n 3), EPA (20:5n 3), and ARA(20:4n 6) for optimal egg production and larval development (Tocher, 2003). Marine species are more susceptible to dietary lipid deficiencies than freshwater species due to evolutionary constraints (Sargent et al., 1997). Lack of appropriate fatty acid ratios, specifically DHA, EPA, and ARA, can result in visual impairment and hormone deficiencies (Sarge nt et al., 1997). General observed ratios of DHA:EPA range between 1.5 2:1, and ratios of EPA:ARA approximate 1.5:1 (Sargent et al., 1997; Tocher, 2003). Methods Samples of C. miliaris eggs and larvae were obtained on days 0, 5, 11, 13, and 17 dph. Eggs were collected 3 different times from broodstock tanks, homogenized, and then rinsed 3 times with reverse osmosis filtered water on a 400 screen prior to being placed in a microcentrifuge tube and frozen at 80 °C. Additional eggs from each sampling were stocked in 128 L tanks following larval rearing protocols. A sample of 20 50 larvae were harvested on each sampling day, rinsed with R.O. water, and frozen at 80 °C. Samples were shipped to the Fisheries and Mariculture Laboratory at Unviersity

PAGE 162

162 of Texas Marine Science Institute (Port Aransas, Tx) for analysis as percentage of total fatty acid. Results Fatty acid profiles over the 17 day sampling period exhibited an increase in DHA while all other fatty acids remained relatively stable over the sampling period ( Figure A 1 ). Palmitic acid (16:0) composed ~18% of all fatty acids and remained stable over the 17 day sampling period. DHA increased from 13 35% total compos ition over the sampling period, likely supplied by the diet of copepod nauplii of P. crassirostris (Trocher, 2003). No other fatty acids composed 10% or greater total composition besides oleic acid on 13 dph ( Figure A 1 ). Ratios of DHA:EPA ranged from approximately 2:1 in C. miliaris eggs to ~6:1 at 5 dph, with a ratio of ~5:1 on11, 13, and 17 dph (Figure A 2). Ratios of EPA:ARA ranged between ~1:1 in C. miliaris eggs to ~2.5:1 in 17 dph larvae ( Figure A 2 ). Ontogeny of Visual and Digestive Systems Visual and digestive systems of marine fish larvae are morphologically, histologically, and physiologically different than those of adult fishes (Govoni et al., 1986). Understanding development of these systems can be beneficial for understanding the feed requirements larvae need for efficient prey capture and digestion. Methods Larvae were cultured using protocols form Chapter 1 of this thesis. Throughout development l arvae were sampled every day for the first 4 dph, then every other day until 37 dph, and finally dying larvae were collected at 43 and 44 dph. Larvae were randomly harvested using a large pipette, moved to a beaker with 500 ppm buffered

PAGE 163

163 MS 222, and rinsed with R.O. water before being preserved in McDowell Trumps fixative (BBC Biochemical Atlanta, GA) . Larvae were then sent to Histopathology Laboratory (St. Petersburg, FL), imbedded in JB4 plastic, processed, and stained with hemato xylin and eosin stain. Due to time restraints, a preliminary group of 2 larvae at 0, 1, 2, 3, 4, 7, 13, 19, 26, 32, and 43 dph were submitted for histology and analyzed. Development of the visual and digestive systems were documented at each submitted stage. Results Results are presented in Table A 1 , defining the different ontogenetic processes occurring at each sampling period similar to that of Yufera et al. (2014). Additional figures accompany descriptions of 0 dph ( Figure A 3 ), 7 dph ( Figure A 4 , A 5 ), 19 dph ( Figure A 6 ), 26 dph ( Figure A 7 ), 32 dph ( Figure A 8 ), and 43 dph larvae ( Figure A 9 ). Ontogeny of the visual system is documented in Figure A 10 . Due to limited sample size and the small size of larvae, some development may not be cataloged within table.

PAGE 164

164 Table A 1. Description of histologically observed ontogeny of C. miliaris larvae throughout development. Due to limited sample size and time restrictions, observations are limited to 9 days throughout development and only visible development at each stage is described. 0 1 dph Observations 2 3 dph Observations Endogenous reserves Ovoid acidophilic with a single oil droplet New hatch larvae measured 1.20 1.24 mm (SL) reaching 2.171 2.198 mm (SL), with a large yolk sac and oil droplet present. Very little locomotion was observed in early larvae. Yolk sac still present but reduced in size Yolk reserves are reduced. Larvae reach 2.487 mm notochord length at 3 dph and have increased finfold size. Larvae begin to actively move. Digestive system Mouth and anus closed. Thin, unopened, undifferentiated, tubular gut Anus opens but mouth still closed. Undifferentiated tubular gut, with simple columnar and cuboidal epithelium. Accessory glands Swim bladder Eye Lens and developing retina of undifferentiated cells Formation of the ganglion cell layer, inner plexiform layer, inner nuclear layer, outer plexiform layer, outer nuclear layer, and visual cells Gills and pseudobranch Heart Kidney and urinary bladder Endocrine elements Spleen

PAGE 165

165 Table A 1. Continued 4 dph Observations 7 dph Observations Endogenous reserves Small amount of yolk and oil globule still present Larvae measured 2.384 2.539 mm (SL), and are capable of feeding with fully pigmented eyes and an open mouth. Active swimming, as well as, s shaped striking behavior was observed as larvae searched for food. Yolk sac and oil globule exhausted Larvae measured 2.420 2.631 mm (SL) and were actively searching for prey items. Larvae did not exhibit much body depth growth. Digestive system Mouth is open. Digestive tract is still tubular with a simple columnar and cuboidal epithelium Buccopharynx connected by a short, narrow, oesophagus to the intestine. Anterior intestine (AI) lined with ciliated columnar epithelium, separated by intestinal valve, posterior intestine (PI) lined with brush border, and supranuclear vacuoles with eosinophilic inclusions in enterocytes, and goblet cells present. Circular muscular layer surrounding intestine. Accessory glands Liver and pancreas a mass of undifferentiated basophilic cells Swim bladder Primordial swimbladder visible Eye Pigmented epithelium formation, and nerve fiber layer visible Thickening of the retina. Nerve cell layer visible Gills and pseudobranch Gill anlage visible Heart . Heart, no differentiation of chambers visible Kidney and urinary bladder Endocrine elements Spleen No observed spleen No observed spleen

PAGE 166

166 Table A 1. Continued 13 dph Observations 19 dph Observations Endogenous reserves Larvae measured 3.309 3.680 mm (SL), body was yellow colored with black coloration along the margins of the body. Body depth was increased and larvae had fully inflated swim bladders Larvae measured 4.201 4.531 mm (SL) and exhibited some body depth growth. Larvae began to select larger copepidites but still mainly feed on nauplii. Digestive system No stomach visible. AI still has a columnar epithelium, while PI has large number of lipid vacuoles and more intestinal folds are present Goblet cells present Buccopharynx, oesphagus, AI and PI. No stomach visible. Increased intestinal folding especially in AI. Lipid vacuoles present in both AI and PI. Accessory glands Hepatocytes and pancreatic tissue differentiated. Sinusoids present with granules and lipid vacuoles present in hepatic cytoplasm. Pancreatic islets formed Swim bladder Swim bladder is inflated. Development of rete mirabile Swim bladder is inflated. Development of rete mirabile Eye Increase in thickness of retina Increase in thickness of inner plexiform layer Gills and pseudobranch Development of primordial gill filaments. No lamellae present Primordial gill filaments. Few lamellae present Heart Atrium and ventricle present, blood circulating. Kidney and urinary bladder Renal tubules and hematopoetic tissue present. Urinary bladder present Endocrine elements Spleen No observed spleen No observed spleen

PAGE 167

167 Table A 1. Continued 26 dph Observations 32 dph Observations Endogenous reserves Larvae measured 5.472 5.691 mm (SL) and began developing hard bony plates over the head as part of the thiolichthys stage. Larvae selected large copepods and still used the s strike to capture prey. Coiling of the gut starting. Larvae measured ~6.5 mm (TL) and had 3 large bony plates on the head. The supracleithrum plate appears as a bulbous bulge above the eyes, the post temporal plate is flat extending posteriorly from above the eye, and a large preopercular spine extended posteriorly to the edge of the pelvic fin girdle. The body was an ovate shape and fins were lobular with visible fin rays. Swimming behavior began to change to undulating the posterior half of the body, similar to that of adults. The s striking behavior was diminished and prey capture was more reliant on propulsion from the caudal fin. Digestive system No stomach visible. AI with large number of lipid vacuoles present. Increased length of microvilli in PI. No apparent inclusion in PI Folds present in the esophagus. Stomach is present with gastric glands. Continued proliferation of intestinal folds in AI and PI Accessory glands Increasing size of liver and pancreas. Increasing size of liver and pancreas Swim bladder Complete development of the gas gland and rete mirabile Fully developed swim bladder Eye Development of the visual cells No distinction between cone a rod cells visible. Gills and pseudobranch Formation of gill filaments, vascular structures. Cartilaginous framework forming, chloride cells present Chondrocytes present and further proliferation of gill filaments Heart Atrio ventricular valve formed Increasing size Kidney and urinary bladder No visible posterior kidney. Anterior kidney proliferation of renal tubules and hematopoietic tissue No visible posterior kidney. Anterior kidney proliferation of renal tubules and hematopoetic tissue Endocrine elements Spleen No observed spleen No observed spleen

PAGE 168

168 Table A 1. Continued 43 dph Observations Endogenous reserves Larvae measured ~8.1 mm (TL), had large bony plates covering the head, visible fin rays in the dorsal, anal, and caudal fins, and dorsal spines forming. Proliferation of gut coiling. Digestive system Further development of gastric glands in stomach. Distinct areas of the stomach were not visible. Continued coiling of intestine. No observed tooth development. Accessory glands Increasing size of liver and pancreas Swim bladder Fully developed swim bladder Eye No distinction between cone and rod cells Gills and pseudobranch Further proliferation of gill filaments Heart Fully formed heart Kidney and urinary bladder No visible posterior kidney. Anterior kidney proliferation of renal tubules and hematopoetic tissue Endocrine elements Thymus present Spleen No observed spleen

PAGE 169

169 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 0 5 11 13 17 Days post hatch (dph) 14:0 15:0 15:1 16:0 16:1n7 16:2n4 17:0 16:3n4 18:0 18:1n9 18:1n7 18:2n6 18:3n6 18:3n4 18:3n3 18:4n3 20:1n9 20:2n6 20:3n6 20:4n6 20:3n3 20:4n3 20:5n3 22:5n6 22:5n3 22:6n3 Figure A 1. Total fatty acid composition of C. miliaris larvae at 0, 5, 11, 13 and 17 dph, represented as mean values ( n =2). 14 12 10 8 6 4 2 0 0 5 11 13 17 Days post hatch (dph) DHA/EPA DHA/ARA EPA/ARA Figure A 2. Ratios of DHA, EPA, and ARA in C. miliaris larvae at 0, 5, 11, 13, and 17 dph, represented as mean values ( n =2). % Total Composition Ratio

PAGE 170

170 Figure A 3. Histological analysis of 0 dph C. miliaris larvae with developing eyes (E), a large elliptical yolk sac (YS), and a single oil globule (OG). Figure A 4. Histological analysis of 7 dph C. miliaris larvae with no remaining yolk sac and oil globule, a short, narrow esophagus (Oe) connecting to the anterior intestine (AI). The liver (L) remains a mass of basophilic undifferentiated cells and no chambers are visible in the heart (H). The primordial swim bladder (Sb) has begun to form, and gill anlage (Ga) is present but no filaments or lamellae are visible.

PAGE 171

171 Figure A 5. Histological analysis of 7 dph C. miliaris larvae with the intestinal valve formed between the anterior intestine (AI) and posterior intestine (PI). Figure A 6. Histological analysis of 19 dph C. miliaris larvae. Differentiation of liver (L) and pancreatic (P) tissue, over inflated swim bladder (Sb), increased folding and lipid vacuoles present in both the anterior intestine (AI) and posterior intestine (PI), and differentiation of the atrium and ventricle apparent.

PAGE 172

172 Figure A 7. Histological analysis of 26 dph C. miliaris larvae. Formation of the gill filaments occurring (G), distinction between the atrium and ventricle of the heart (H) by atrio ventricular valve, increased size of liver (L) and pancreatic (P) tissues, complete development of rete mirabile and gas gland in swim bladder (Sb), no visible stomach, lipid vacuoles are present in anterior intestine (AI), and proliferation of renal tubules and hematopoetic tissue in anterior kidney (K)

PAGE 173

173 Figure A 8. Histological analysis of 32 dph C. miliaris larvae. A large stomach (S) has formed within the body cavity and gastric glands are present, gill filaments (G) are proliferating, and larval body depth has increased.

PAGE 174

174 Figure A 9. Histological analysis of 43 dph C. miliaris larvae with further coiling of the intestines present, a small portion of the thymus (T) visible. Other organelle systems are similar to that of 32 dph larvae.

PAGE 175

175 Figure A 10. Eye development of C. miliaris at 0, 2, 4, 7, and 43 dph. A) At 0 dph the lens (Ln) and an undifferentiated retina (R) were forming, B) At 2 dph the retina had differentiated into the ganglion cell layer (Gl), the inner plexiform layer (IP), the inner nuclear layer (IN), the outer plexiform layer (OP), the outer nuclear layer (ON), and the visual cell layer (Vc), C) At 4 dph the pigmented epithelium (Pe) had formed, D) at 7 dph the nerve cell layer was visible (Nl), and E) At 43 dph the visual cell layer was developing but no differentiation between rod and cone cells was visible. B D A C

PAGE 176

176 Figure A 10. Continued. E

PAGE 177

177 LIST OF REFERENCES Alessio, G., 1975. Riproduzione artificiale di orata, Sparus aurata (L.) (Osteichthyes, Sparidae). 5. Primi risultati alimentazione delle larve e degli ava notti. Boll. Pesca, Piscic. Idrobiol. 30, 71 92. Appelbaum, S., Kamler, E., 2000. Survival, growth, metabolism and behaviour of Clarias gariepinus (Burchell 1822) early stages under different light conditions. Aquac. Eng. 22, 269 287. Arvedlund, M., McCormick, M.I., Ainsworth, T., 2000. Effects of photoperiod on growth of larvae and juveniles of the anemonefish Amphiprion melanopus . Naga, The ICLARM Quat. 23, 18 23. Avella, M.A., Olivotto, I., Gioacchini, G., Maradonna, F., Carnevali, O., 2007. The role of fatty acids enrichments in the larviculture of false percula clownfish Amphiprion ocellaris . Aquaculture 273, 87 95. F., G., Ilgaz, S., H., 2004. The effect of tank volumes on survival of gilthead seabream ( Sparus aurata L., 1758) from hatching to the first grading in intensive culture systems. EUJ Fisher. Aquat. Sci. 21, 69 72. Baensch, F., March 01, 2014. Larval rearing of the schooling banner fish, Heniochus diphreutes . Retrieved from http://www.bluereefphoto.org/blog/2014/3/larval rearing of the pennant bannerfish heniochus diphreutes Baensch, F., Tamaru, C.S., 2009, Spawning and development of larvae and juveniles of the rare blue Mauritius angelfish, Centropyge debelius (1988), in the hatchery. J. World Aquac. Soc. 40, 425 39. Barber, C.V., Pratt, V.R., 1998. Poison and profits: cyanide fishing in the Indo Pacific. Environment: Science and Policy for Sustainable Development 40, 4 9. Barlow, G.W., 1981. Patterns of parental investment, dispersal and size among coral reef fishes. Environmental Biology of Fishes 6, 65 85. Barrionuevo, W.R., Fernandes, M.N., Rocha, O., 2010. Aerobic and anaerobic metabolism for the zebrafish, Danio rerio , reared under normoxic and hypoxic conditions and exposed to acute hypoxia during development. Brazilian Journal of Biology 70, 425 34. Baskerville Bridges, B., Kling, L.J., 2000. Larval culture of Atlantic cod ( Gadus morhua ) at high stocking densities. Aquaculture 181, 61 69. Battaglene, S.C., McBride, S., Talbot, R.B., 1994. Swim bladder inflation in larvae of cultured sand whiting, Sillago ciliata Cuvier (Sillaginidae). Aquaculture 128, 177 192.

PAGE 178

178 Battaglene, C., Morehead, T., Cobcroft, M., Nichols, D., Brown, R., Carson, J., 2006. Combined ef fects of feeding enriched rotifers and antibiotic addition on performance of striped trumpeter ( Latris lineata ) larvae. Aquaculture 251, 456 471. Battaglene, S.C., Talbot, R.B., 1990. Initial swim bladder inflation in intensively reared Australian bass larvae, Macquaria novemaculeata (Steindachner) (Perciformes: Percichthyidae). Aquaculture 86, 431 442. Bell, J.D., Clua, E., Hair, C.A., Galzin, R., Doherty, P.J., 2009. The capture and culture of post larval fish and invertebrates for the marine ornamental trade. Reviews in Fisheries Science 17, 223 240. Bendif, E.M., Probert, I., Schroeder, D.C., de Vargas, C., 2013. On the description of Tisochrysis lutea gen. nov. sp. nov. and I sochrysis nuda sp. nov. in the Isochrysidales, and the transfer of Dicrateria to the Prymnesiales (Haptophyta). Journal of Applied Phycology 25, 1763 1776. Bergenius, M.A.J., McCormick, M.I., Meekan, M.G., Robertson, D.R., 2005. Environmental influences on larval duration, growth and magnitude of settlement of a coral reef fish. Marine Biology 147, 291 300. Betancur R, R., Hines, A., Acero P., A., Ortí, G., Wilbur, A.E., Freshwater, D.W., 2011. Reconstructing the lionfish invasion: insights into Greater Caribbean biogeography. Journal of Biogeography 38, 1281 1293. Be yer, S.G., Sogard, S.M., Harvey, C.J., Field, J.C., 2014. Variability in rockfish ( Sebastes spp.) fecundity: species contrasts, maternal size effects, and spatial differences. Environmental Biology of Fishes, 1 20. Boehlert, G.W., Morgan, J.B., 1985. Turbidity enhances feeding abilities of larval Pacific herring, Clupea harengus pallasi. Hydrobiologia 123, 161 170. Bonislawska, M., Formicki, K., Korzelecka Orkisz, A., Winnicki, A., 2001. Fish egg size variability: biological significance. Electronic Journal of Polish Agricultural Universities 4, 14. Booth, .J., Parkinson, K., 2011. Pelagic larval duration is similar across 23° of latitude for two species of butterflyfish (Chaetodontidae) in eastern Australia. Coral Reefs 30, 1071 1075. Boulhic, M., Gabaudan, J., 1992. Histological study of the organogenesis of the digestive system and swim bladder of the Dover sole, Solea solea (Linnaeus 1758). Aquaculture 102, 373 396. Brownell, C.L., 1980. Water quality requirements for first feeding in marine fis h larvae. I. Ammonia, nitrite, and nitrate. Journal of experimental marine Biology and Ecology 44, 269 283.

PAGE 179

179 Burgess, W.E., 1978. Butterflyfishes of the World. T.F.H. Publ., Neptune City. Buskey, E.J., 1994. Factors affecting feeding selectivity of visual predators on the copepod Acartia tonsa: locomotion, visibility and escape responses. Hydrobiologia 292, 447 453. Buskey, E.J., Coulter, C., Strom, S., 1993. Locomotory patterns of microzooplankton: potential effects on food selectivity of larval fish. Bulletin of Marine Science 53, 29 43. Buskey, E.J., Lenz, P.H., Hartline, D.K., 2002. Escape behavior of planktonic copepods in response to hydrodynamic disturbances: high speed video analysis. Marine Ecology Progress Series 235, 135 146. Calado, R., 2006. Marine ornamental species from European waters: a valuable overlooked resource or a future threat for the conservation of marine ecosystems? Scientia Marina 70, 389 398. Callan, C.K., Laidley, C.W., Forster, I.P., Liu, K.M., Kling, L.J., Place, A.R., 2012a. Examination of broodstock diet effects on egg production and egg quality in flame angelfish ( Centropyge loriculus ). Aquaculture Research 43, 696 705. Callan, C.K., Laidley, C.W., Kling, L.J., Breen, N.E., Rhyne, A.L. , 2012b. The effects of dietary HUFA level on flame angelfish ( Centropyge loriculus ) spawning, egg quality and early larval characteristics. Aquaculture Research, 1 11. Carton, A.G., 2005. The impact of light intensity and algal induced turbidity on first feeding Seriola lalandi larvae. Aquaculture Research 36, 1588 1594. Cassiano, E.J., Ohs, C.L., Weirich, C.R., Breen, N.E., Rhyne, A.L., 2011. Performance of larval florida pompano fed nauplii of the calanoid copepod Pseudodiaptomus pelagicus . North American Journal of Aquaculture 73, 114 123. Chapman, F.A., Fitz Coy, S.A., Thunberg, E.M., Adams, C.M., 1997. United States of America trade in ornamental fish. Journal of the World Aquaculture Society 28, 1 10. Chesney Jr, E.J., 1989. Estimating the food requirements of striped bass larvae Morone saxatilis : Effects of light, turbidity and turbulence. Marine ecology progress series 53, 191 200. Cobcroft, J.M., Shu Chien, A.C., Kuah, M. K., Jaya Ram, A., Battaglene, S.C., 2012. The effects of tank colour, live food enrichment and greenwater on the early onset of jaw malformation in striped trumpeter larvae. Aquaculture 356 357, 61 72.

PAGE 180

180 Cohen, S., Diaz, M.V., Díaz, A.O., 2013. Histological and histochemical study of the digestive system of the Argentine anchovy larvae ( Engraulis anchoita ) at different developmental stages of their ontogenetic development. Acta Zoologica, 1 12. Colin, P.L., 1989. Aspects of the spawning of western Atlantic butterflyfishes (Pisces: Chaetodontidae). Enviro nmental Biology of Fishes 25, 131 141. Colin, P.L., Clavijo, I.E., 1988. Spawning activity of fishes producing pelagic eggs on a shelf edge coral reef, southwestern Puerto Rico. Bulletin of Marine Science 43, 249 279. Conceição, L.E.C., Yúfera, M., Makridis, P., Morais, S., Dinis, M.T., 2010. Live feeds for early stages of fish rearing. Aquaculture Research 41, 613 640. Cook, A., 1996. Ontogeny of feeding morphology and kinematics in juvenile fishes: a case study of the cottid fish Clinocottus analis . Journal of Experimental Biology 199, 1961 1971. Coughlin, D.J., 1991. Ontogeny of feeding behavior of first feeding Atlantic salmon ( Salmo salar ). Canadian Journal of Fisheries and Aquatic Sciences 48, 1896 1904. Coward, K., Bromage, N.R., Hibbitt, O., Parrington, J., 2002. Gamete physiology, fertilization and egg activation in teleost fish. Reviews in Fish Biology and Fisheries 12, 33 58. Cowen, R.K., 1991. Variation in the planktonic larval duration of the temperate wrasse Semicossyphus pulcher . Marine Ecology Progress Series 69, 9 15. Craig, S., Helfrich, L.A., 2002. Understanding fish nutrition, feeds, and feeding. Virginia Cooperative Extension, Virginia Polytechnic Institute and State University, Publication 420 256. DiMaggio, M.A., 2012. Evaluation of Culture Methods for Two Marine Baitfish Species, Pinfish, Lagodon Rhomboides , and Pigfish, Orthopristis Chrysoptera . PhD dissertation, Program in Fisheries and Aquatic Sciences, University of Florida, p. 134. DiMaggio, M.A., Bro ach, J.S., Ohs, C.L., Grabe, S.W., 2013. Captive Volitional Spawning and Larval Rearing of Pigfish. North American Journal of Aquaculture 75, 109 113. DiMaggio, M. A., Grabe, S.W., DeSantis, S.M., Ohs, C.L., 2010, Induced volitional spawning and larval re aring of pinfish, Lagodon rhomboides . North American Journal of Aquaculture 72, 252 257.

PAGE 181

181 Donelson, M., Munday, L., McCormick, I., Pankhurst, W., Pankhurst, M., 2010. Effects of elevated water temperature and food availability on the reproductive performance of a coral reef fish. Marine Ecology Progress Series 401, 233 243. Dower, J.F., Miller, T.J., Leggett, W.C., 1997. The role of microscale turbulence in the feeding ecology of larval fish. Advances in Marine Biology 31, 169 220. Downing, G., Litvak, M.K., 2000. The effect of photoperiod, tank colour and light intensity on growth of larval haddock. Aquaculture International 7, 369 382. Drillet, G., Frouël, S., Sichlau, M.H., Jepsen, P.M., Højgaard, J.K., Joarder, A.K., Hansen, B.W., 2011. Status and recommendations on marine copepod cultivation for use as live feed. Aquaculture 315, 155 166. Duray, M.N., Estudillo, C.B., Alpasan, L.G., 1996. The effect of background color and rotifer density on rotifer intake, growth and surviv al of the grouper ( Epinephelus suillus ) larvae. Aquaculture 146, 217 224. Duray, M., Kohno, H., 1988. Effects of continuous lighting on growth and survival of first feeding larval rabbitfish, Siganus guttatus . Aquaculture 72, 73 79. Einen, O., Roem, A.J., 1997. Dietary protein/energy ratios for Atlantic salmon in relation to fish size: growth, feed utilization and slaughter quality. Aquaculture Nutrition 3, 115 126. FAO, 2000. The state of world fisheries and aquaculture. Document prepared by Wijkstrom, U., Gumy, A., Grainger, R. FAO, Rome. FAO, 2012. The state of world fisheries and aquaculture. Document prepared by Mathiesen, A.M. FAO, Rome. Faulk, C.K., Holt, G.J., 2003. Lipid nutrition and feeding of cobia Rachycentron canadum larvae. Journal of the World Aquaculture Society 34, 368 378. Fuiman, L.A., Faulk, C.K., 2013. Batch spawning facilitates transfer of an essential nutrient from diet to eggs in a marine fish. Biology Letters 9. García Ortega, A., 2009. Nutrition and feeding research in the spotted rose snapper ( Lutjanus guttatus ) and bullseye puffer ( Sphoeroides annulatus ), new species for marine aquaculture. Fish Physiology and Biochemistry 35, 69 80. García Ortega, A., Hernández, C., Abdo delaParra, I., González Rodríguez, B., 2002. Advances in the nutrition and feeding of the bullseye puffer Sphoeroides annulatu . In: Cruz Suárez, L. E., Ricque Marie, D., Tapia Salazar, M., Gaxiola Cortés, M. G., Simoes, N. (Eds.), Avances en Nutrición Acuícola VI. Memorias del VI Simposiu m Internacional de Nutrición Acuícola. 3 al 6 de Septiembre del 2002. Cancún, Quintana Roo, México.

PAGE 182

182 Gjedrem, T., Baranski, M., 2010. Selective Breeding in aquaculture: An Introduction. Springer Science and Business Media. Glamuzina, B., N., Skaramuca, B., Kozul, V., Tutman, P., 2001. Early development of the hybrid Epinephelus costae × E. marginatus . Aquaculture 198, 55 61. Glamuzina, B., Glavic, N., Tutman, P., Kozul, V., Skaramuca, B., 2000. Egg and early larval development of laboratory reared goldblotch grouper, Epinephelus costae (Steindachner, 1878) (Pisces, Serranidae). Scientia Marina 64, 341 345. Glamuzina, B., Skaramuca, B., Glavic, N., Kozvul, V., Dulcic, J., Kraljevic, M., 1998. Egg and early larval development of lab oratory reared dusky grouper, Epinephelus marginatus (Lowe, 1834) (Picies, Serranidae). Scientia Marina 62, 373 378. Goodsell, A., Wikeley, D., Searle, L., 1996. Histological investigation of swim bladder morphology and inflation in cultured larval striped trumpeter ( Latris lineata ) (Teleostei, Latridae). Marine and Freshwater Research 47, 251 254. Gordon, A.K., Bok, A.W., 2001. Frequency and periodicity of spawning in the clownfish Amphiprion akallopisos under aquarium conditions. Aquarium Sciences and Conservation 3, 293 299. Grier, H.J., Uribe Aranza´bal, M.C., Patiño, R., 2009. The ovary, folliculo genesis, and oogenesis in teleosts. In: Jamieson, B.G.M., (Ed.), Reproductive biology and phylogeny of fishes (agnathans and bony fishes). Science Pu blishers, Enfield, New Hampshire, 24 85. Grisez, L., Reyniers, J., Verdonck, L., Swings, J., Ollevier, F., 1997. Dominant intestinal microflora of sea bream and sea bass larvae, from two hatcheries, during larval development. Aquaculture 155, 387 399. Hamre, K., Srivastava, A., Rønnestad, I., Mangor Jensen, A., Stoss, J., 2008. Several micronutrients in the rotifer Brachionus sp. may not fulfill the nutritional requirement of marine fish larvae. Aquaculture Nutrition 14, 51 60. Hamre, K., Yúfera, M., Rønnestad, I., Boglione, C., Conceição, L.E.C., Izquierdo, M., 2013. Fish larval nutrition and feed formulation: knowledge gaps and bottlenecks for advances in larval rearing. Reviews in Aquaculture 5, S26 58. Hart, P.R., Hutchinson, W.G., Purser, G.J., 1996. Effects of photoperiod, temperature and salinity on hatchery reared larvae of the greenback flounder ( Rhombosolea tapirina Günther, 1862). Aquaculture 144, 303 311. Hart, P.R., Purser, G.J., 1995. Effects of salinity and temperature on eggs and yolk sac larvae of the greenback flounder ( Rhombosolea tapirina Günther, 1862). Aquaculture 136, 221 230.

PAGE 183

183 Hassell, K.L., Coutin, P.C., Nugegoda, D., 2008. Hypoxia impairs embryo development and survival in black bream ( Acanthopagrus butcheri ). Ma rine Pollution Bulletin 57, 302 306. Henderson, B.A., Trivedi, T., Collins, N., 2000. Annual cycle of energy allocation to growth and reproduction of yellow perch. Journal of Fish Biology 57, 122 133. Henderson, B.A., Wong, J.L., Nepszy, S.J., 1996. Reproduction of walleye in Lake Erie: allocation of energy. Canadian Journal of Fisheries and Aquatic Sciences 53, 127 133. Hilder, P.I., Cobcroft, J.M., Battaglene, S.C., 2014. The first feeding response of larval southern bluefin tuna, Thunnus maccoyii (Castelnau, 1872), and yellowtail kingfish, Seriola lalandi (Valenciennes, 1833), to prey density, prey size and larval density. Aquaculture Research, 1 14. Hioki, S., 1997. Spawning and larval rearing of marine fishes at the Marine Science Museum. Procee dings of the 4 th annual marine aquarium congress Tokyo, 361 366. Holt, G. J., 2003. Research on culturing the early life history stages of marine ornamental species. In: Cato, J.C., Brown, C. L., (Eds.), Marine ornamental species: collection, culture and conservation. Iowa State Press, Ames, Iowa, USA, 251 254. Holt, G.J., 2011. (Ed.), Larval Fish Nutrition. Wiley Blackwell, John Wiley & Sons, Inc. U.K. Holt, G.J., Faulk, C.K., Schwarz, M.H., 2007. A review of the larviculture of cobia Rachycentron canadum , a warm water marine fish. Aquaculture 268, 181 187. Holt, G.J., Riley, C.M., 2000. Laboratory spawning of coral reef fishes: effects of temperature and photoperiod. UJNR Technical. Report No. 28, 33 38. Houde, E.D., 1978. Critical food concentrations for larvae of three species of subtropical marine fishes. Bulletin of Marine Science 28, 395 411. Houde, E.D., Schekter, R.C., 1980. Feeding by marine fish larvae, developmental and functional responses. Environmental Biology of Fishes 5, 315 334. Hourigan, T.F., 1989. Environmental determinants of butterflyfish social systems. Environmental Biology of Fishes 25, 61 78. Hunter, J., 1981. Feeding ecology and predation of marine fish larvae. In: Lasker, R., Editor. Marine fish larvae: morphology, ecology, and relation to fisheries. University of Washington Press, Seattle, 33 77.

PAGE 184

184 Ignatius, B., Rathore, G., Jagadis, I., Kandasamy, D., Victor, A.C.C., 2001. Spawning and larval rearing technique for tropical clown fish Amphipr ion sebae under captive condition. Journal of Agriculture in Tropics 16, 241 249. Imanpoor, M.R., Bagheri, T., 2010. Relationship Between Biological Characteristics of Egg with Fertility Success, Hatching Rate and Larvae Size in Female Kutum, Rutilis frisii Kutum. World Journal of Fish and Marine Science 2, 404 409. Izquierdo, M.S., Fernandez Palacios, H., Tacon, A.G.J., 2001. Effect of broodstock nutrition on reproductive performance of fish. Aquaculture 197, 25 42. Johannes, R.E., 1978. Rep roductive strategies of coastal marine fishes in the tropics. Environmental Biology of Fishes 3, 65 84. Johnson, G.D., 1984. Percoidei. In: Moser, H.G., Richards, W.J., Cohen, D.M., Fahay, M.P., Kendall Jr., A.W., Richardson, S.L.,(Eds.), Ontogeny and Systematics of Fishes, American Society of Ichthyology and Herpetology Special Publications, 464 498. Khan, M.A., Jafri, A.K., Chadha, N.K., 2005. Effects of varying dietary protein levels on growth, reproductive performance, body and egg composition of rohu, Labeo rohita (Hamilton), Aquaculture Nutrition 11, 11 17. Kamler, E., 2005. Parent egg progeny Relationships in Teleost Fishes: An Energetics Perspective. Reviews in Fish Biology and Fisheries 15, 399 421. Kennedy, J., Geffen, A.J., Nash, R.D.M., 2007. Maternal influences on egg and larval characteristics of plaice ( Pleuronectes platessa L). Journal of Sea Research 58, 65 77. Kimmel, C.B., Ballard, W.W., Kimmel, S.R., Ullmann, B., Schilling, T.F., 1995. Stages of embryonic development of the zebrafish. Developmental Dynamics 203, 253 310. King N.J., Howell W.H., Huber M., Bengtson D.A., 2000, Effects of larval stocking density on laboratory scale and commercial scale production of summer flounder Paralichthys dentatus . Journal of the World Aquaculture Society 31, 436 445. Kiyono, M., Hirano, R., 1981. Effects of light on the feeding and growth of black porgy, Mylio macrocephalus (Basilewsky), postlarvae and juveniles. Journal du Conseil International pour de la Mer 178, 334 336. Kolm, N., 2002. Male size determines reproductive output in a paternal mouthbrooding fish. Animal Behaviour 63, 727 733. Koslow, J.A., Brault, S., Dugas, J., Fournier, R.O., Hughes, P., 1985. Condition of larval cod ( Gadus morhua ) off southwest Nova Scotia in 1983 in relation to plankton abundance and temperature. Marine Biology 86, 113 121.

PAGE 185

185 Kraul, S., 1989. Production of live prey for marine fish larvae. Advances in Tropical Aquaculture, Tahiti (French Polynesia). In: AQUACOP IFREMER Acetes de Colloque 9, 595 607. Kumar, T.T.A., Balasubramanian, T., 2009. Broodstock development, spawning and larval rearing of the false clown fish, Amphiprion ocellaris in captivity using estuarine water. Current Science 97, 1483 1486. Kumar, T .T., Gopi, M., Dhaneesh, K.V., Vinoth, R., Ghosh, S., Balasubramanian, T., Shunmugaraj, T., 2012. Hatchery production of the clownfish Amphiprion nigripes at Agatti island, Lakshadweep, India. Journal of Environmental Biology 33, 623 628. Lecchini, D., Planes, S., Galzin, R., 2005. Experimental assessment of sensory modalities of coral reef fish larvae in the recognition of their settlement habitat. Behavioral Ecology and Sociobiology 58, 18 26. Leis, J.M., 1989. Larval biology of butterflyfishes (Pisces, Chaetodontidae): what do we really know?. Environmental Biology of Fishes 25, 89 100. Leis, J.M., McCormick, M.I., 2002, The Biology, Behavior; and Ecology of the Pelagic, larval stage of coral reef fishes. In: Coral reef fishes: dynamics and diversity in a complex ecosystem . Academic Press, p. 171. Leis, J.M., Yerman, M.N., 2012. Behavior of larval Butterflyfishes (Teleostei: Chaetodontidae) at settlement on coral reefs. Copeia, 211 221. Leu, M.Y., Liou, C.H., Wang, W.H., Yang, S.D., Meng, P.J., 2009. Natural spawning, early development and first feeding of the semicircle angelfish [ Pomacanthus semicirculatus (Cuvier, 1831)] in captivity. Aquaculture Research 40, 1019 1030. Leu, M.Y., Sune, Y.H., Meng, P.J., 2013. First results of larval rearing and development of the bluestriped angelfish Chaetodontoplus septentrionalis (Temminck & Schlegel) from hatching through juvenile stage with notes on its potential for aquaculture. Aquaculture Research, 1 14. Lobel, P.S., 1989. Spawning behavior of Chaetodon Multicintcus (Chaetodontidae); pairs and intruders. Environmental Biology of Fishes 25, 125 130. Londraville, R., 1990. A Ccourtship like behavior pattern that can Be used to sex a butterflyfish (Chaetodontidae). Copeia 1990, 582 584. Lupat sch, I., Deshev, R., Magen, I., 2010. Energy and protein demands for optimal egg production including maintenance requirements of female tilapia Oreochromis niloticus . Aquaculture Research 41, 763 769.

PAGE 186

186 Ma, Y., Kjesbu, O.S., Jørgensen, T., 1998. Effects of ration on the maturation and fecundity in captive Atlantic herring ( Clupea harengus ). Canadian Journal of Fisheries and Aquatic Sciences 55, 900 908. MacKenzie, B.R., Kiørboe, T., 2000. Larval fish feeding and turbulence: a case for the downside. Limnology and Oceanography 45, 1 10. MacKenzie, B.R., Miller, T.J., Cyr, S., Leggett, W.C., 1994. Evidence for a dome shaped relationship between turbulence and larval fish ingestion rates. Limnology and Oceanography 39, 1790 1799. Madden, W.D ., May, R.C., 1977. Ornamental fish culture project. Final Report, Marine Aquarium Council Task Order No. 136. The Oceanic Institute and The Hawaii Institute of Marine Biology. Madhu, K., Madhu, R., Retheesh, T., 2012. Broodstock development, breeding, embryonic development and larviculture of spine cheek anemonefish, Premnas biaculeatus (Bloch, 1790). Indian Journal of Fisheries 59, 65 75. McGurk, M.D., 1984. Effects of delayed feeding and temperature on the age of irreversible starvation and on the rates of growth and mortality of Pacific herring larvae. Marine Biology 84, 13 26. McKinnon, D., Duggan, S., Nichols, D., Rimmer, A., Semmens, G., Robino, B., 2003. The potential of tropical paracalanid copepods as live feeds in aquaculture. Aquaculture 223, 89 106. Millamena, O.M., 2002. Replacement of fish meal by animal by product meals in a practical diet for grow out culture of grouper Epinephelus coioides . Aquaculture 204, 75 84. Miller, K.M., Schulze, A.D., Ginther, N., Li, S., Patterson, D.A., Farrell, A.P., Hinch, S.G., 2009. Salmon spawning migration: metabolic shifts and environmental triggers. Comparative biochemistry and physiology. Part D, Genomics and Proteomics 4, 75 89. Moe, M.A., 2003, Culture of marine ornamentals: For love, for money and for science. In: Cato, J.C., Brown, C.L., (Eds.), Marine Ornamental Species: Collection, Culture and Conservation. Ames, Iowa: Iowa State Press, 251 254. Moorhead, J.A., Zeng, C., 2010. Development of captive breeding techniques for marine ornamental fish: a review. Reviews in Fisheries Science 18, 315 343. Morehead, D.T., Hart, P.R., Dunstan, G.A., Brown, M., Pankhurst, N.W., 2001. Differences in egg quality between wild striped trumpeter ( Latris lineata ) and captive striped trumpeter that were fed different diets. Aquaculture 192, 39 53.

PAGE 187

187 Morita, K., Takashima, Y., 1998. Effect of female size on fecundity and egg size in white spotted charr: comparison between sea run and resident forms. Journal of Fish Biology 53, 1140 1142. Naas, K.E., Næss, T., Harboe, T., 1992. Enhanced first feeding of halibut larvae ( Hippoglossus hippoglossus L.) in green water. Aquaculture 105, 143 156. Neidig, C.L., Skapura, D.P., Grier, H.J., Dennis, C.W., 2000. Techniques for spawning common snook: broodstock handling, oocyte staging, and egg quality. North American Journal of Aquaculture 62, 103 113. Noga, E.J., 2010. Fish Disease: Diagnosis and treatment. 2 nd edition. Wiley Blackwell. O'Brien, W.J., Slade, N.A., Vinyard, G.L., 1976. Apparent size a s the determinant of prey selection by bluegill sunfish ( Lepornis rnacrochinus ). Ecology 57, 1304 1310. Olivotto, I., Planas, M., Simões, N., Holt, G.J., Avella, M.A., Calado, R., 2011. Advances in breeding and rearing marine ornamentals. Journal of the World Aquaculture Society 42, 135 166. Ostrowski, A.C., 1989. Effect of rearing tank background color on early survival of dolphin larvae. The Progressive Fish Culturist 51, 161 163. Ostrowski, A.C., Laidley, C.W., 2001. Application of marine foodfish techniques in marine ornamental aquaculture: Reproduction and larval first feeding. Aquarium Science Conservation 3, 191 204. Oyarzun, F.X., Strathmann, R.R., 2011. Plasticity of hatching and the duration of planktonic development in marine invertebrates. Integrative and Comparative Biology 51, 81 90. Papanikos, N., Phelps, R.P., Davis, D.A., Ferry, A., Maus, D., 2008. Spontaneous spawning of captive red snapper, Lutjanus campechanus , and dietary lipid effect on reproductive performance. Journal of the World Aquaculture Society 39, 324 338. Patiño, R., Sullivan, C.V., 2002. Ovarian follicle growth, maturation, and ovulation in teleost fish. Fish physiology and Biochemistry 26, 57 70. Pekcan Hekim, Z., Lappalainen, J., 2006. Effects of clay turbidity and density of pikeperch ( Sander lucioperca ) larvae on predation by perch ( Perca fluviatilis ), Die Naturwissenschaften 93, 356 359. Pomeroy, R.S., Pido, M.D., Pontillas, J.F.A., Francisco, B.S., White, A.T., Ponce De Leon, E.M.C., Silvestre, G.T., 2008. Evaluation of policy options for the live reef food fish trade in the province of Palawan, Western Philippines. Marine Policy 32, 55 65.

PAGE 188

188 Pratchett M.S, Berumen, M.L., Kapoor, B.G., 2013. Biology of Butterflyfish. CRC press. Taylor and Francis Group. Puvanendran, V., Brown, J.A., 2002. Foraging, growth and survival of Atlantic cod larvae reared in different light intensities and photoperiods. Aquaculture 214, 131 151. Ralston, S., 1976. Anomalous Growth and Reproductive Patterns in Populations of Chaetodon miliaris (Pisces, Chaetodontidae) from Kaneohe Bay, Oahu, Hawaiian Islands. Pacific Science 30, 395 403. Ralston, S., 1981. Aspects of the reproductive biology and feeding ecology of Chaetodon miliaris , a Hawaiian endemic butterflyfish. Environmental Biology of Fishes 6, 167 176. Reavis, R.H., 1997. The natural history of a monogamous coral reef fish, Valenciennea strigata (Gobiidae): 2. behavior, mate fidelity and reproductive success. Environmental Biology of Fishes 49, 247 257. Reitan, K.I., Rainuzzo, J.R., Øie, G., Olsen, Y., 1997. A review of the nutritional effects of algae in marine fish larvae. Aquaculture 155, 207 221. Rhody,.R., Neidig, .L., Grier, .J., Main, .L. & Migaud, H., 2013. Assessing reproductive condition in captive and wild common snook stocks: a comparison between the wet mount technique and histological preparations. Transactions of the American Fisheries Society 142, 979 988. Rhyne, A.L., Tlusty, M.F., Schofield, P.J., Kaufman, L., Morris, J. A., Bruckner, A.W., 2012. Revealing the appetite of the marine aquarium fish trade: the volume and biodiversity of fish imported into the United States. PloS One 7, 1 9. Rønnestad, I., Thorsen, A., Finn, R.N., 1999. Fish larval nutrition: a review of recent advances in the roles of amino acids. Aquaculture 177, 201 216. Sadovy, Y.J., Vincent, A.C.J., 2002. Ecological issues and the trades in live reef fishes. In: Sale, P.F., (Eds.), Coral reef fishes dynamics and diversity in a complex ecosystem. Academic Press, San Diego, 391 420. Sampey, A., McKinnon, A.D., Meekan, M.G., McCormick, M.I., 2007. Glimpse into guts: overview of the feeding of larvae of tropical shorefishes. Marine Ecology Progress Series 339, 243 257. Sánchez Hernández, J., Vieira Lanero, R., Servia, M.J., Cobo, F., 2011. First feeding diet of young brown trout fry in a temperate area: disentangling constraints and food selection. Hydrobiologia 663, 109 119. Sano, M., 1989. Feeding habits of Japanese butterfyfishes (Chaetodontidae). Environmental Biology of Fishes 25, 195 203.

PAGE 189

189 Sanz, A., Llorente, J.I., Furné, M., Ostos Garrido, M.V., Carmona, R., Domezain, A., Hidalgo, M.C., 2011. Digestive enzymes during ontogeny of the sturgeon Acipenser naccarii : intestine and pancreas development. Journal of Applied Ichthyology 27, 1139 1146. Sarasquete, M.C., Polo, A., Yúfera, M., 1995. Histology and histochemistry of the development of the digestive system of larval gilthead seabream, Sparus aurata L. Aquaculture 130, 79 92. Sargent, J.R., McEvoy, L.A., Bell, J.G., 1997. Requirements, presentation and sources of polyunsaturated fatty acids in marine fish larval feeds. Aquaculture 155, 117 127. Sargent, J., McEvoy, L., Estevez, A., Bell, G., Bell, M., Henderson, J., Tocher, D., 1999. Lipid nutrition of marine fish during early development: current status and future directions. Aquaculture 179, 217 229. Schmitt, P.D., 1986. Prey size selectivity and feeding rate of larvae of the northern anchovy, Engraulis mordax Girard. CalCOFI Report 27, 153 161. Setu, S.K., Kumar, T.T.A., Balasubramanian, T., Dabbagh, A.R., Keshavarz, M., 2010. Breeding and rearing of regal damselfish Neopomacentrus cyanomos (Bleeker, 1856): the role of green water in larval survival. World Journal of Fish Mariculture Science 2, 551 557. Shaw, W., Pankhurst, M., Battaglene, C., 2006. Effect of turbidity, prey density and culture history on prey consumption by greenback flounder Rhombosolea tapirina larvae. Aquaculture 253, 447 460. Shoji, J., Tanaka, M., 2004. Effect of prey concentration on growth of piscivorous Japanese Spanish mackerel Scomberomorus niphonius larvae in the Seto Inland Sea, Japan. Journal of Applied Ichthyology 20, 271 27 5. Skjermo, J., Salvesen, I., Øie, G., Olsen, Y., Vadstein, O., 1997. Microbially matured water: a technique for selection of a non opportunistic bacterial flora in water that may improve performance of marine larvae. Aquaculture International 5, 13 28. Smith, K.F., Behrens, M., Schloegel, L.M., Marano, N., Burgiel, S., Daszak, P., 2009. Ecology. Reducing the risks of the wildlife trade. Science 324, 594 595. Soeparno, Nakamura, Y., Shibuno, T., Yamaoka, K., 2012. Relationship between pelagic larval duration and abundance of tropical fishes on temperate coasts of Japan. Journal of Fish Biology 80, 346 357. Springate, J.R.C., Bromage, N.R., 2003. Effects of egg size on early growth and survival in rainbow trout ( Salmo gairdneri Richardson). Aquaculture 47, 163 172.

PAGE 190

190 Stottrup, J. G., 2000. The elusive copepods: their production and suitability in marine aquaculture. Aquaculture Research 31, 703 711. Stottrup, J. G. 2003. Production and nutritional value of copepods. In Stottrup, J.G., McEvoy, L.A., (Eds.), Live feeds in marine aquaculture. Blackwell Scientific Publications, Oxford, UK, 145 205. Stottrup, J.G., Norsker, N.H., 1997. Production and use of copepods in marine fish larviculture. Aquaculture 155, 231 247. Suzuki, K., T anaka, Y., Hioki, S., 1980. Spawning behavior, eggs, and larvae of the butterflyfish, Chaetodon nippon , in an aquarium. Japanese Journal of Ichthyology 26, 334 341. Sweet, T., 2014. 2014 Captive bred marine fish species list. Coral 11. Tanaka, Y., Hioki, S., Suzuki, K., 2001. Spawning behavior, eggs, and larvae of the butterflyfish, Chaetodon modestus , in an aquarium. Journal of the Faculty of Marine Science and Technology Tokai University (Japan) 51, 89 100. Tandler, A., Helps, S., 1985. The effects of photoperiod and water exchange rate on growth and survival of gilthead sea bream ( Sparus aurata , Linnaeus; Sparidae) from hatching to metamorphosis in mass rearing systems. Aquaculture 48, 71 82. Temple, S., Cerqueira, V.R., Brown, J.A., 2004. The effects of lowering prey density on the growth, survival and foraging behaviour of larval fat snook ( Centropomus parallelus poey 1860). Aquaculture 233, 205 217. Tissot, B.N., Hallacher, L.E., 2003. Effects of aquarium collectors on coral reef fishes in Kona, Hawaii. Conservation Biology 17, 1759 1768. Tlusty, M., 2002. The benefits and risks of aquacultural production for the aquarium trade. Aquaculture 205, 203 219. Tocher, D.R., 2003. Metabolism and functions of lipids and fatty acids in teleost fish. Reviews in Fisheries Science 11, 107 184. Treviño, L., Alvarez González, C.A., Perales García, N., Arévalo Galán, L., Uscanga Martínez, A., Márquez Couturier, G., Fernández, I., Gisbert, E., 2011. A histological study of the organogenesis of the di gestive system in bay snook Petenia splendida Günther, 1862 from hatching to the juvenile stage. Journal of Applied Ichthyology 27, 73 82. Tricas, T.C., 1989. Food and competitors as determinants of territory size in the Hawaii butterflyfish, Chaetodon multicinctus . Animal Behaviour 37, 830 841.

PAGE 191

191 Tricas, T.C., Hiramoto, J.T., 1989. Sexual differentiation, gonad development, and spawning seasonality of the Hawaiian butterflyfish, Chaetodon multicinctus . Environmental Biology of Fishes 25, 111 124. Trippel, E.A., Kjesbu, O.S., Solemdal, P., 1997. Effects of adult age and size structure on reproductive output in marine fishes. In: Chambers, R.C. and Trippel, E.A. (Eds.), Early Life History and Recruitment in Fish Populations. Springer, Netherlands, 31 62. van der Meeren, T., Mangor Jensen, A., Pickova, J., 2007. The effect of green water and light intensity on survival, growth and lipid composition in Atlantic cod ( Gadus morhua ) during intensive larval rearing. Aquaculture 265, 206 217. Vuthiphandchai, V., Zohar, Y., 1999. Age related sperm quality of captive striped bass Morone saxatilis . Journal of the World Aquaculture Society 30, 65 72. Wabnitz, C., Taylor, M., Green, E., Razak, T., 2003. From ocean to aquarium: The global trade in marine ornamental species. UNEP WCMC, Cambridge, UK. Wang, X., Li, Y., Hou, C., Gao, Y., Wang, Y., 2013. Physiological and molecular changes in large yellow croaker ( Pseudosciaena crocea R.) with high fat diet induced fatty liver disease. Aquaculture Re search. Watson, C.A., Shireman, J.V., 1996. Production of ornamental aquarium fish. Food and Aquaculture Publication Series. University of Florida. Information document FA 35, 1 4. Available from http://www.aces.edu/dept/fisheries/ education/ras/publications/Update/FL%20ornamental%20production.pdf Watts, M., Pankhurst, N.W., Pryce, A., Sun, B., 2003. Vitellogenin isolation, purification and antigenic cross reactivity in three teleost species. Comp arative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology 134, 467 476. Welker, M.T., Pierce, C.L., Wahl, D.H., 1994. Growth and survival of larval fishes: roles of competition and zooplankton abundance. Transactions of the American Fisheries Society 123, 703 717. Wendelaar Bonga S.E., 1997. The stress response in fish. Physiological Reviews 77, 591 625. Wenger, A.S., McCormick, M.I., 2013. Determining trigger values of suspended sediment for behavioral changes in a coral reef fish. Marine Pollution Bulletin 70, 73 80. Wexler J.B., Margulies D., Scholey, V.P., 2011. Temperature and dissolved oxygen requirements for survival of yellowfin tuna, Thunnus albacares , larvae. Journal of Experimental Marine Biology and Ecology 404, 63 72.

PAGE 192

192 Weyers, R.S., Jennings, C.A., Freeman, M.C., 2003. Effects of pulsed, high velocity water flow on larval robust redhorse and V lip redhorse. Transactions of the American Fisheries Society 132, 84 91. Wittenrich, M.L., 2007. The complete illustrated guide to marine aquarium fishes. TFH Publications, Inc., Neptune City, New Jersey, USA. Wittenrich, M.L., Cassiano, E.J., November 27, 2011. Schooling close! Retrieved from http://risingtideconservation.blogspot.com/2011/11/schooling bannerfishso close.html Wittenrich, M.L., Turingan, R.G., Cassiano, E.J., 2012. Rearing tank size effects feeding performance, growth, and survival of sergeant major, Abudefduf saxatilis , larvae. AACL Bioflux 5, 393 402. Wood, E., 2001. Global advances in conservation and management of marine ornamental resources. Aquarium Sciences and Conservation 3, 65 77. Yabuta, S., 1997. Spawning migrations in the monogamous butterflyfish, Chaetodon trifasciatus . Ichthyological Research 44, 177 182. Yabuta, S., Kawashima, M., 1997. Spawning behavior and haremic mating system in the corallivorous butterflyfish, Chaetodon trifascialis , at Kuroshima Island, Okinawa. Ichthyological Research 44, 183 188. Yanong, R.P.E., 2003. Fish health management considerations in recirculating aquaculture systems Part 1: Introduction and general priciples. UF IFAS Cooperative Extension Service. University of Florida, Information document FA 120. Yanong, R.P.E., 2012. Cryptocaryon irritans infections (marine white spot disease) in fish. UF IFAS Cooperative Extension Service. University of Florida, Information document FA 164. Yin, M.C., Blaxter, J.H.S., 1987. Feeding ability and survival during starvation of marine fish larvae reared in the laboratory. Journal of Experimental Marine Biology and Ecology 105, 73 83. Zambonino, J.L., Cahu, C., 2010. Effect of nutrition on marine fish development and quality. Recent Advances in Aquaculture Research, 103 124. Zavala Leal, I., Dumas, S., Peña, R., Contreras Olguín, M., Hernández Ceballos, D., 2013. Effects of culture conditions on feeding response of larval Pacific red snapper ( Lutjanus peru , Nichols & Murphy) at first feeding. Aquaculture Research 44, 1399 1406.

PAGE 193

193 BIOGRAPHICAL SKETCH Jon Michael L.A. Degidio was born in Providence, Rhode Island in 1990 and spent his youth in Jamestown, Rhode Island. Growing up, Jon Michael spent countless hours on the water fishing, diving, and working. After graduating high school, Jon Michael attended Florida Institute of Technology where he earned a Bachelor of Science in Marine Biology and a Bachelor of Science in Aquaculture. Throughout each summer of his undergraduate career Jon Michael obtained an internship or job within the marine biology field including working for the National Oceanic and Atmospheric Apex Predator Program. After graduating from Florida Institute of Technology in 2012 Jon Michael was awarded a fully funded graduate assistantship at the University of Tropical Aquaculture Laboratory in Ruskin FL, and began his thesis work under VMD. Roy Yanong. In the Fall of 2014, Jon Michael graduated with his Master of Science in Fisheries and Aquatic Sciences.