Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Broodstock conditioning and larval rearing of the geoduck clam (Panopea generosa Gould, 1850) Marshall, Robert 2012

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2012_spring_marshall_robert.pdf [ 1.62MB ]
Metadata
JSON: 24-1.0072611.json
JSON-LD: 24-1.0072611-ld.json
RDF/XML (Pretty): 24-1.0072611-rdf.xml
RDF/JSON: 24-1.0072611-rdf.json
Turtle: 24-1.0072611-turtle.txt
N-Triples: 24-1.0072611-rdf-ntriples.txt
Original Record: 24-1.0072611-source.json
Full Text
24-1.0072611-fulltext.txt
Citation
24-1.0072611.ris

Full Text

BROODSTOCK CONDITIONING AND LARVAL REARING OF THE GEODUCK CLAM (Panopea generosa GOULD, 1850)  by Robert Marshall  B.Sc.(hons), Dalhousie University, 1993 M.Aq., Simon Fraser University, 1997 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  The Faculty of Graduate Studies  (Animal Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  March 2012  © Robert Marshall, 2012  Abstract The aim of this thesis was to identify conditions that optimize Panopea generosa broodstock conditioning and larval growth and survival in a hatchery setting. A series of experiments subjected broodstock (adults) to various levels of key factors [i.e. temperature (Ch. 2), salinity (Ch. 3) and nutrition [ration (Ch. 4) and feed type (Ch. 5)]. A larval experiment examined the effects of stocking density and feed level combinations on growth and survival (Ch. 6). Broodstock responses were quantified using gravimetric (condition and gonadosomatic indices) and histological techniques (development classification, volume fractions and oocyte diameter). Survival and spawning rates were also examined. Of the temperatures tested (7, 11, 15 and 19˚C) 11˚C had the highest spawning rates (% individuals) and more oocytes follicle-1, than 15 and 19˚C. At 7˚C gonadosomatic indices were highest but this temperature did not produce spawning clams. Gonads degenerated at 19˚C. Among salinities of 17, 20, 24, and 29 gonad sheath thickness and area occupied by gametes increased at 29 but not at 24. Salinities of 17 and 20 were associated with fungal infection and had high mortality rates after 26 d exposure. With higher ration treatments (up to 7.2 × 109 cells clam-1 d-1 [Isochrysis sp. (TISO) and Chaetoceros muelleri, 50:50 by cell count] clams became more spawned. Very high rations (10.0 × 109 cells clam-1 d-1) increased mortality. Algal type [Dunaliella tertiolecta, (TISO), C. muelleri and TISO + C. muelleri) had no measurable impact on gonad development based on any of the response variables used. The larval experiment showed that there was a significant interaction between stocking density and feed level with respect to growth and survival. The most efficient treatment was 10 larvae ml-1 fed  ii  5,000 cells larva-1 (TISO) as it had among the highest growth (3.75 μm day-1) and survival (42%) rates but low algal requirements. These findings illustrate reproductive responses of P. generosa that can be applied to a hatchery management strategy.  iii  Preface For this thesis I contributed to the development of the experimental protocols, conducted the research, collected the data, performed the statistical analysis, interpreted the data and prepared the various chapters. My committee members (Dr. Abayomi Alabi, Dr. Anthony Farrell, Dr. Claudio DiBacco, Dr. Marina von Keyserlingk, Dr. Christopher M. Pearce, and Dr. Robert Scott McKinley) helped to conceive the design of the original experiments and provided valuable feedback on the results. Drs. McKinley and Pearce provided overall supervision and guidance and ensured that the experiments and the data analyses were correctly performed. My co-authors, Drs. McKinley and Pearce, reviewed and provided critical comments on all of the manuscripts submitted based on this dissertation. A version of Chapter 2 has been accepted for publication [Marshall, R., McKinley, R.S., Pearce, C.M., 2012. Effect of temperature on gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850). Aquaculture, in press]. A version of Table 1.1 has been published as part of a review paper [Marshall, R., McKinley, R.S., Pearce, C.M., 2010. Effects of nutrition on larval growth and survival in bivalves. Reviews in Aquaculture 2, 33-55.  Dr. Pearce provided funding (through an NSERC Discovery grant) and provided laboratory space at the Pacific Biological Station and ensured that the experiments were conducted in with the approval of the DFO Pacific Region Animal Care Committee and the guidelines of the Canadian Council on Animal Care (CCAC). The DFO PRACC Animal Use Protocol approval numbers were 06-016, 07-007a and 08-011.  iv  Table of Contents Abstract ...........................................................................................................................ii Preface ........................................................................................................................... iv Table of Contents ............................................................................................................ v List of Tables ...............................................................................................................viii List of Figures ................................................................................................................ ix List of Abbreviations ....................................................................................................xiii Acknowledgements ...................................................................................................... xiv 1 Introduction ................................................................................................................. 1 1.1 General introduction .............................................................................................. 1 1.2 General biology and ecology of Panopea generosa ................................................ 4 1.3 Geoduck aquaculture and potential ........................................................................ 7 1.4 Factors influencing the gonad development and larval growth and survival of molluscan bivalves ...................................................................................................... 8 1.4.1 Effects of physical and biological factors on broodstock conditioning of bivalve molluscs ...................................................................................................................... 9 1.4.1.1 Introduction ................................................................................................. 9 1.4.1.2 Temperature ................................................................................................. 9 1.4.1.3 Salinity ...................................................................................................... 11 1.4.1.4 Water quality ............................................................................................. 12 1.4.1.5 Nutrition .................................................................................................... 14 1.4.1.6 Conclusions ............................................................................................... 21 1.4.2 Effects of stocking density and ration on larval growth and survival ................. 22 1.5 Thesis objectives and hypotheses ......................................................................... 27 1.5.1 Experiment 1: Effect of temperature on the gonad development of the geoduck clam....................................................................................................................... 28 1.5.2 Experiment 2: Effect of salinity on the gonad development of the geoduck clam....................................................................................................................... 29 1.5.3 Experiment 3: Effect of food ration on the gonad development of the geoduck clam....................................................................................................................... 31 1.5.4 Experiment 4: Effect of food type on the gonad development of the geoduck clam....................................................................................................................... 33 1.5.5 Experiment 5: Effects of stocking density and algal feed ration on growth, survival, and ingestion rate of larval geoduck clams ............................................... 34 2 Effect of temperature on gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850) .................................................................................................. 36 2.1 Introduction ......................................................................................................... 37 2.2 Materials and methods ......................................................................................... 38 2.2.1 Algal culture ................................................................................................. 38 2.2.2 Broodstock collection and initial maintenance ............................................... 38 2.2.3 Broodstock conditioning: experimental design .............................................. 39 2.2.4 Sampling ....................................................................................................... 41 2.2.5 Mortality rate and spawn events .................................................................... 42 2.2.6 Condition and gonadosomatic indices ........................................................... 42 GSI is the gonadosomatic index..................................................................................... 43  v  2.2.7 Ash-free dry weight ...................................................................................... 44 2.2.8 Histological sampling ................................................................................... 44 2.2.9 Statistical analyses ........................................................................................ 46 2.3 Results ................................................................................................................. 47 2.3.1 Mortality rate and spawn events .................................................................... 47 2.3.2 Condition and gonadosomatic indices ........................................................... 48 2.3.3 Ash-free dry weight ...................................................................................... 49 2.3.4 Histological sampling ................................................................................... 49 2.4 Discussion ........................................................................................................... 51 3 Effect of salinity on survival and gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850) .................................................................................. 64 3.1 Introduction ......................................................................................................... 65 3.2 Materials and methods ......................................................................................... 66 3.2.1 Algal culture ................................................................................................. 66 3.2.2 Broodstock collection and initial maintenance ............................................... 67 3.2.3 Broodstock conditioning: experimental set up ............................................... 67 3.2.4 Spawning events and mortality rates ............................................................. 69 3.2.5 Clam sampling .............................................................................................. 70 3.2.6 Condition and gonadosomatic indices ........................................................... 70 3.2.7 Gonad ash-free dry weight ............................................................................ 71 3.2.8 Histological sampling ................................................................................... 72 3.2.9 Oxygen consumption and clearance rates ...................................................... 74 3.2.10 Statistical analyses ...................................................................................... 76 3.3 Results ................................................................................................................. 78 3.3.1 Spawning events and mortality rates ............................................................. 78 3.3.2 Condition and gonadosomatic indices ........................................................... 79 3.3.3 Gonad ash-free dry weight ............................................................................ 80 3.3.4 Gonad histology ............................................................................................ 80 3.3.5 Oxygen consumption and clearance rates ...................................................... 82 3.4 Discussion ........................................................................................................... 83 4 Effect of ration on gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850) ................................................................................................................. 97 4.1 Introduction ......................................................................................................... 98 4.2 Materials and methods ......................................................................................... 99 4.2.1 Algal culture ................................................................................................. 99 4.2.2 Broodstock collection and initial maintenance ............................................... 99 4.2.3 Broodstock conditioning: experimental design ............................................ 100 4.2.4 Clam sampling ............................................................................................ 101 4.2.5 Mortality rate and spawn events .................................................................. 102 4.2.6 Clearance and ingestion rates ...................................................................... 102 4.2.7 Condition and gonadosomatic indices ......................................................... 103 4.2.8 Histological sampling ................................................................................. 105 4.2.9 Statistical analyses ...................................................................................... 107 4.3 Results ............................................................................................................... 108 4.3.1 Mortality rate and spawn events .................................................................. 108 4.3.2 Clearance and ingestion rates ...................................................................... 109  vi  4.3.3 Condition and gonadosomatic indices ......................................................... 110 4.3.4 Histological sampling ................................................................................. 111 4.4 Discussion ......................................................................................................... 113 5 Effect of food type on gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850) ................................................................................................ 129 5.1 Introduction ....................................................................................................... 129 5.2 Materials and methods ....................................................................................... 131 5.2.1 Algal culture ............................................................................................... 131 5.2.2 Broodstock collection, and initial maintenance ............................................ 131 5.2.3 Broodstock conditioning: experimental design ............................................ 132 5.2.4 Sampling ..................................................................................................... 133 5.2.5 Clearance and ingestion rates ...................................................................... 134 5.2.6 Condition and gonadosomatic indices ......................................................... 134 5.2.7 Histological sampling ................................................................................. 135 5.2.8 Statistical analyses ...................................................................................... 137 5.3 Results ............................................................................................................... 138 5.3.1 Mortality rate and spawn events .................................................................. 138 5.3.2 Clearance and ingestion rates ...................................................................... 139 5.3.3 Condition and gonadosomatic indices ......................................................... 139 5.3.4 Histological sampling ................................................................................. 140 5.4 Discussion ......................................................................................................... 141 6 Effects of stocking density and algal feed ration on growth, survival, and ingestion rate of larval geoduck clams (Panopea generosa)............................................................... 153 6.1 Introduction ....................................................................................................... 153 6.2 Materials and methods ....................................................................................... 156 6.2.1 Microalgal culture ....................................................................................... 156 6.2.2 Larval culture and experimental design ....................................................... 156 6.2.3 Data collection and analysis ........................................................................ 159 6.3 Results ............................................................................................................... 162 6.3.1 Growth and survival rates............................................................................ 162 6.3.2 Ingestion rates ............................................................................................. 164 6.4 Discussion ......................................................................................................... 164 7 Conclusions ............................................................................................................. 177 7.1 Thesis background ............................................................................................. 177 7.2 Chapter 2: Effect of temperature on broodstock conditioning ............................. 177 7.3 Chapter 3: Effect of salinity on broodstock conditioning .................................... 180 7.4 Chapter 4: Effect of food ration on broodstock conditioning .............................. 184 7.5 Chapter 5: Effect of food type on broodstock conditioning................................. 187 7.6 Chapter 6: Effect of food ration and stocking density on larval production ......... 190 7.7 Overall conclusions ........................................................................................... 193 References ................................................................................................................... 195  vii  List of Tables Table 1.1 Biochemical components and dry weights of some commonly cultured phytoplankton species ................................................................................................... 25 Table 2.1. Results of ANOVAs on various attributes and ANCOVAs on condition index of experimental Panopea generosa. Values in bold are significant at P < 0.05. NS = not significant. .................................................................................................................... 57 Table 3.1 Results of ANOVAs on various attributes of experimental Panopea generosa. One-way ANOVAs pertain to day 0 and include all salinities (17, 20, 24, 29). Two-way ANOVAs pertain to all sample days (0, 41, 62 d) and include only salinities 24 and 29. Sources of variation are salinity (S, fixed factor), time (T, fixed factor), interactions (S×T), and error. Values in bold are significant at P < 0.05. NS = not significant.......... 89 Table 4.1 Analyses of covariance showing the effects of ration, sampling time, block (i.e. tank), and co-variate (i.e. dry shell weight) on various condition indices (CI) – based on wet (CIw), dry (CId), and submerged (CIs) tissue weights – in Panopea generosa. P values in bold are significant at a level of at least <0.05. .............................................. 121 Table 4.2 Analyses of variance showing the effects of ration, sampling time, and block (i.e. tank) on various gonadosomatic indices (GSI) – based on wet (GSIw), dry (GSId), and submerged (GSIs) tissue weights – and digestive gland index (DGI) in Panopea generosa. P values in bold are significant at a level of at least <0.05. ......................... 122 Table 4.3 Results of ANOVAs from histological examination for experimental Panopea generosa. Sources of variation are ration (R, fixed factor), Time (T, fixed factor) Block (B, random factor). Values in bold are significant at α=0.05. ....................................... 123 Table 5.1 Results of ANOVAs (clearance and ingestion rates), ANCOVAs (Condition index) and 2-way ANOVAs for various attributes of experimental Panopea generosa. Sources of variation are algae treatment (A, fixed factor), C (covariate = shell weight; Condition index only), Time (T, fixed factor), interaction (A × T) and error. Values in bold are significant at α=0.05. ..................................................................................... 147 Table 6.1 Linear growth models for larvae of Panopea generosa for each treatment pairing of larval density and algal ration up to 23 days after reaching D-stage. ............ 171 Table 6.2 Results of separate two-way ANOVAs on various attributes of experimental Panopea generosa larvae. Sources of variation are larval stocking density (D, fixed factor), ration (R, fixed factor) and interaction (D×R). Values in bold are significant at P<0.05 or less. ............................................................................................................. 172  viii  List of Figures Figure 2.1. Schematic showing the Panopea generosa conditioning tube system. Dashed lines depict water levels in the respective containers. Arrows depict water flow. Actual configuration had eight tubes per water-bath tank. Seven tubes contained one clam each and one tube remained empty for water quality testing. There were four waterbath tanks for each temperature treatment. Not to scale. See text for dimensions. ........ 58 Figure 2.2. Mean percentage of Panopea generosa individuals spawning at various temperatures (7, 11, 15, 19°C) over time (weeks of conditioning). Error bars indicate SEM (n=16 in Phase I, n=4 in Phase II). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters in Phase I (below bars) apply to differences among weeks and upper-case letters in Phase II (above bars) apply to differences among temperatures. Temperature was not a significant factor in Phase I and time was not a significant factor in Phase II (ANOVA). ................................................ 59 Figure 2.3. Mean (a) wet and (b) dry condition indices (CI) of Panopea generosa held at various temperatures (7, 11, 15, 19°C) over time (days). Error bars indicate SEM (n=4). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods and upper-case letters (above legends) apply to differences among temperatures. .................................. 60 Figure. 2.4. Mean (a) wet and (b) dry gonadosomatic indices (GSI) of Panopea generosa held at various temperatures (7, 11, 15, 19°C) over time (days). Error bars indicate SEM (n=4). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods and upper-case letters (above legends) apply to differences among temperatures. ...................................................................................................................................... 61 Figure. 2.5. Histograms showing the percentage of various reproductive stages of Panopea generosa held at various temperatures [(a) 7°C, (b) 11°C, (c) 15°C, (d) 19°C] over time (days), as determined by examination of histological sections. Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Upper-case letters (next to temperatures) apply to differences among temperatures. ......................... 62 Figure. 2.6. Mean (a) percentage of mature oocytes, (b) number of oocytes follicle -1, and (c) percentage of connective tissue occupation in the gonads of Panopea generosa held at various temperatures (7, 11, 15, 19°C) over time (days). Mature oocytes are 30.1–40.0 μm in diameter. Error bars indicate SEM (n’s for 6a,b shown in 6a, n=4 in 6c). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods and upper-case letters (above legends or bars) apply to differences among temperatures. ....................... 63 Figure 3.1 Mean cumulative mortalities of Panopea generosa at four salinities during the salinity adjustment phase (16 d) and experimental phase (26 d). Error bars indicate SEM (n = 4). Treatments denoted by different letters differ significantly (P < 0.05, TukeyKramer test). Capital letters (above bars) apply to the effects of salinity within each time period while lower-case letters (below bars) apply to the effects of time within each salinity level. ................................................................................................................. 90 Figure 3.2 Mean condition indices (CIs) of Panopea generosa at four salinities and three sampling times: (a) wet weight condition index (CI w) and (b) dry weight condition index  ix  (CId). Error bars indicate SEM (n = 4). Days 41 and 62 exclude salinities of 17 and 20 due to high mortalities. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). Capital letters above day 0 bars denote treatment differences for all salinities on that day. NS = non-significant (P > 0.05). ....................................... 91 Figure 3.3 Mean gonadosomatic indices (GSIs) of Panopea generosa at four salinities and three sampling times: (a) wet weight gonadosomatic index (GSI w) and (b) dry weight gonadosomatic index (GSId). Error bars indicate SEM (n = 4). Days 41 and 62 exclude salinities of 17 and 20 due to high mortalities. NS = non-significant (P > 0.05). ........... 92 Figure 3.4 Histograms showing the percentage of various reproductive stages of Panopea generosa held at various salinities (given above bars) over time (days), as determined by examination of histological sections. ...................................................... 93 Figure 3.5 Mean gonad development parameters of Panopea generosa at four salinities and three sampling times: (a) oocyte diameter (OD), (b) gamete occupation index (GOI), (c) connective tissue occupation index (COI), and (d) gonad thickness (GT). Error bars indicate SEM (n = 4). Days 41 and 62 exclude salinities of 17 and 20 due to high mortalities. Treatments denoted by different letters differ significantly (P < 0.05, TukeyKramer test). Capital letters above day 0 bars denote treatment differences for all salinities on that day. Capital letters above day 41 and 62 bars for OD and COI apply to the effect of salinity while lower-case letters below bars apply to the effect of time (i.e. no significant interaction). Capital letters above day 41 and 62 bars for GOI and GT apply to the effects of salinity within each time level while lower-case letters below bars apply to the effects of time within each salinity level (i.e. significant interaction). NS = non-significant (P > 0.05). ............................................................................................. 94 Figure 3.6 Mean oxygen consumption of Panopea generosa across a range of salinities. Error bars indicate SEM (n=11). Results standardized to 77 g DW clams. Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). ................ 95 Figure 3.7 Mean clearance rates of Panopea generosa at four salinities and three sampling times. Error bars indicate SEM (n=4). Days 38 and 58 exclude salinities of 17 and 20 due to high mortalities. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). Capital letters above day 2 bars denote treatment differences for all salinities on that day. Lower-case letters below bars denote overall differences over time. NS = non-significant (P > 0.05). ................................................ 96 Figure 4.1 Mean cumulative percent mortality of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. ....................................... 124 Figure 4.2. Mean percentage of Panopea generosa spawning in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). Upper-case letters (above bars) indicate significant differences among ration levels within each time level and lower-case letters (below bars) indicate significant differences between time levels within each ration level (i.e. significant interaction). .................................................... 124 Figure 4.3 Mean (a) clearance rate and (b) ingestion rate of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. Rations  x  denoted by different letters above bars differ significantly (P < 0.05, Tukey-Kramer test). .................................................................................................................................... 125 Figure 4.4 Mean condition indices – based on (a) wet, (b) dry, and (c) submerged tissue weights – of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. There were no significant differences among rations or times for any condition index. All indices are standardized to 250 g dry shell weight. ............... 126 Figure 4.5 Mean gonadosomatic indices – based on (a) wet, (b) dry, and (c) submerged tissue weights – and (d) digestive gland index of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). For all gonadosomatic indices, lower-case letters (below bars) indicate significant differences among time levels across all ration levels (i.e. no significant interaction). There were no significant differences among rations. For digestive gland index, upper-case letters (above legend) indicate significant differences among ration levels across all time levels and lower-case letters (below bars) indicate significant differences among time levels across all ration levels (i.e. no significant interaction). .......................................................................... 127 Figure 4.6 Mean (a) development index, (b) oocyte diameter, (c) oocytes per unit area, (d) spermatic material occupation index, and (e) connective tissue occupation index of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). For development index, oocyte diameter, and oocytes per unit area, upper-case letters (above legend) indicate significant differences among ration levels across all time levels and lower-case letters (below bars) indicate significant differences among time levels across all ration levels (i.e. no significant interaction). For spermatic material occupation index, upper-case letters (above bars) indicate significant differences among ration levels within each time level and lower-case letters (below bars) indicate significant differences among time levels within each ration level (i.e. significant interaction). There were no significant differences among rations or time for connective tissue occupation index. . 128 Figure 5.1 Mean (a) water volume cleared of algae (b) cells ingested and (c) dry wt algae ingested for Panopea generosa fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri). Error bars indicate SEM (n=4). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test)................................. 148 Figure 5.2 Mean (a) wet, (b) dry, and (c) submerged condition indices (CI) of Panopea generosa fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri) over time (days). Error bars indicate SEM (n=4). ....................................................................... 149 Figure 5.3 Mean (a) wet, (b) dry, and (c) submerged gonadosomatic indices (GSI) and (d) ratio of visceral mass surface area covered by digestive gland (DGI) of Panopea generosa fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri) over time (days). Error bars indicate SEM (n=4). Treatments denoted by different letters differ  xi  significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods. .................................................................................. 150 Figure 5.4 Mean (a) developmental index and (b) oocyte diameter of Panopea generosa gonads fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri) over time (days). Error bars indicate SEM (n=4). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods. .................................................................................. 151 Figure 5.5 Mean (a) oocytes mm-2 and percentage of gonad surface areas occupied by (b) oocytes, (c) spermatic material, and (d) connective tissue in Panopea generosa fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri) over time (days). Error bars indicate SEM (n=4). ............................................................................................. 152 Figure 6.1 Growth of Panopea generosa larvae in response to larval stocking density (a: 2 inds ml-1, b: 5 inds ml-1, and c: 10 inds ml-1) and algal ration (5, 10, 20, and 100×103 cells ind-1 d-1). Error bars are SEM and n = 5. ............................................................. 173 Figure 6.2 Growth rates in shell length of Panopea generosa larvae in response to larval stocking density (2, 5, and 10 inds ml-1) and algal ration (5, 10, 20, and 100×103 cells ind1 -1 d ). Different letters represent significant differences among the treatments (Tukey test, P < 0.05). Upper-case letters signify the results of the effect of larval density within each feed ration level and lower-case letters signify the results of feed ration level within each larval density. Error bars are SEM and n = 5....................................................... 174 Figure 6.3 Percent survival 23 d after D-stage of Panopea generosa larvae in response to larval stocking density (2, 5, and 10 inds ml-1) and algal ration (5, 10, 20, and 100×103 cells ind-1 d-1). Different letters represent significant differences among the treatments (Tukey test, P < 0.05). Upper-case letters signify the results of the effect of larval density within each feed ration level and lower-case letters signify the results of feed ration level within each larval density. Error bars are SEM and n = 5. ........................................... 175 Figure 6.4 Ingestion rates of Panopea generosa larvae in response to larval stocking density (2, 5, and 10 inds ml-1) and algal ration (5, 10, 20, and 100×103 cells ind-1 d-1). Different letters represent significant differences among the treatments (Tukey test, P < 0.05). Upper-case letters signify the results of the effect of larval density within each feed ration level and lower-case letters signify the results of feed ration level within each larval density. Error bars are SEM and n = 5............................................................... 176  xii  List of Abbreviations AFDW ash-free dry weight CI  condition index  CM  Chaetoceros muelleri  COI  connective tissue occupation index  CR  clearance rate  DGI  digestive gland index  DHA  docosahexaenoic acid  DI  development index  EFA  essential fatty acids  EPA  eicosapentaenoic acid  GOI  gamete occupation index  GSI  gonadosomatic index  GT  gonad thickness  IR  ingestion rate  OA  oocytes per unit area  OD  oocyte diameter  OF  oocytes per follicle  OM  oocyte maturity  OOI  oocyte occupation index  SL  shell length  SOI  spermatic material occupation index  TISO  Tahitian Isochrysis sp.  %WC  percent water content of soft tissue  xiii  Acknowledgements This research was supported by a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada (NSERC) and funding from Fisheries and Oceans Canada (both to C. Pearce). We thank Laurie Keddy for culturing algae; Sean Williams and Grant Dovey for diving; Kirk Montgomery for supplying geoducks; Bill Bennett for assistance with histology; Ian Anderson, Carol Bob and John Blackburn for technical assistance; and Dr. Yomi Alabi of Seed Science Ltd. and the Under Water Harvesters Association for generously supplying larvae. I would like to thank my wife Tram for her years of support. I would also like to thank my parents - Ron and Maude, my brother - Mark, and my sister - Lori, for inspiring to continue my education. The people of Mac’s Oysters deserve special thanks, especially Gordy, Sally, Gordon, Ron, Dave, John and Tom, who have all been loyal, supportive and willing to pick up the slack when I am not around.  xiv  1 Introduction 1.1 General introduction Pacific geoduck clams (Panopea generosa, Gould 1850) are the world’s largest burrowing clam and a commercially important species in northwestern North America. British Columbia (BC), Canada and Washington (WA), U.S.A. are the largest harvesters of this species (Beattie, 1992; Beattie and Blake, 1999; Bureau et al., 2003) with landings from the BC fishery valued at CAD $32.7 million in 2005 (James, 2008) and those from WA valued at USD $50 million in 2007 (Gordon, 2007). Fisheries landings peaked in the 1980s in BC at over 5.5 million kg annually, but have been steady at approximately 2 million kg annually from the late 1990s to 2005 (James, 2008). Markets for the geoduck clam are chiefly in Asia, including Hong Kong, Japan, and Taiwan (Chew, 1998). Demand is strong for this product, but supply is limited by natural recruitment that is both temporally and spatially variable and unpredictable (Breen and Shields, 1983; Harbo et al., 1983; Campbell et al., 2004; Zhang and Campbell, 2004). Adults (6–7 years old) recruit to the fishery at rates of only 0.01 to 0.18 clams m-2 yr-1 in geoduck beds, the recruitment level depending on location and growth rates (Campbell et al., 2004). Because of these low recruitment rates, geoduck beds can take decades to recover after a commercial harvest. As a result, certain fisheries regulations have been put in place that limit harvest in order to maintain the long-term sustainability of the resource. In BC the harvest is set at a maximum of 1.2–1.8 % of the total biomass on a bed by bed basis (James, 2008) while in WA the quota is 2.8 % of the total biomass (Anonymous, 2008).  1  Geoduck aquaculture is becoming an important component in meeting market demand given these natural and regulatory restrictions on wild geoduck supply. Aquaculture of P. generosa in BC is very much in its infancy, but significant strides have been taken in WA where aquaculture production reached 397,000 kg in 2007 (Gordon, 2007). Securing a consistent seed supply for P. generosa has been an ongoing issue for many geoduck farmers, however, because hatchery production is inconsistent from year to year (pers. obs.). Collection of seed from wild sources is not practiced with P. generosa since collectors (e.g. hanging spat collectors and cultch bags) are more effective for species that have more homogeneously dispersed populations and high densities of larvae in the water column (Thorarinsdóttir, 1991; Bervera and Monteforte, 1995; Tammi et al., 1995; Garcia et al., 2003). The heterogeneous distribution and variable recruitment of P. generosa make hatchery production the only viable option for seed procurement. Also, P. generosa post-larvae rely on pedal palp feeding which requires substrate (King, 1986), otherwise the post-larvae may starve. Hatchery rearing of bivalve larvae has been practiced since at least the 1930s in North America (Loosanoff and Davis, 1952, 1963), but there have been no published studies examining the effects of specific biological or physical factors on the reproductive development of P. generosa as it relates to hatchery production [but see notes on spawning methods in early larval-rearing studies by Goodwin (1973) and Goodwin et al. (1979)]. Hatchery production generally consists of a gametogenesis or broodstockconditioning phase and a larval-rearing phase. Broodstock conditioning techniques are intended to stimulate high fecundity of the parent stock while producing viable eggs and larvae (Utting and Millican, 1997). Stimulation of gonad development may involve  2  physical cues [e.g. temperature, photoperiod (Fabioux et al., 2005)], adequate ration [i.e. energy ingested must exceed maintenance metabolic demands and losses to waste for reproductive growth to occur (Bayne and Newell, 1983)], and adequate nutrition [i.e. biochemical components such as lipids, proteins, and carbohydrates (Gallager and Mann, 1986; Pernet et al., 2003a)]. The overall energy balance must be positive in order for gonad production to occur and this energy balance will be influenced by various abiotic factors that impact metabolism, including temperature (Mann, 1979b, 1979a; Martínez et al., 2000a; Navarro et al., 2000), salinity (Shumway and Youngson, 1979; Stickle and Sabourin, 1979; Baba et al., 1999), dissolved oxygen (Burnett, 1997), dissolved carbon dioxide (Burnett, 1997; Berge et al., 2006), and nitrogenous wastes (Epifanio and Srna, 1975; Aldridge et al., 1995; Heasman et al., 1996; Delgado and Pérez-Camacho, 2007). Biotic factors influencing metabolism are body size, gill surface area, and activity levels (Hamburger et al., 1983; Shumway, 1983; Yukihira et al., 1998). Without specific information on how these types of factors influence broodstock and larval development it is difficult to make the kinds of effective aquaculture management decisions that ensure long-term commercial success. For my thesis I examined the responses of broodstock and larvae of P. generosa to various levels of biological and physical factors to fill some of these knowledge gaps. What follows is a literature review, followed by the statement of research objectives and hypotheses for each of the five results chapters.  3  1.2 General biology and ecology of Panopea generosa The Pacific geoduck clam (Panopea generosa), of the Family Hiatellidae and Order Myoida, is found from the lower intertidal zone to depths of over 60 m (Shaul and Goodwin, 1982) and from Alaska to Baja California (Goodwin, 1976; Goodwin and Pease, 1989; Calderon-Aguilera et al., 2010). Older specimens can exceed 160 years in age (Bureau et al., 2002; 2003) with shell lengths in excess of 200 mm (Goodwin and Pease, 1989; Bureau et al., 2002; 2003) and a body mass (including shell) of up to 3.25 kg (Goodwin and Pease, 1989). Adult clams burrow up to 1 m in depth (Andersen, 1971). Market-sized clams have shell lengths of approximately 130 to 140 mm and weigh 900 to 1,400 g (Harbo et al., 1983; Bureau et al., 2002; 2003). A full anatomical description of P. generosa is available in Bower and Blackbourn (2003). The body of the clam is not completely encompassed by the shell, with the periostracum-covered muscular mantle (commonly known as the “belly”) and the fused siphons left exposed. The clam is oriented in the substratum with its posterior end upward and the siphon tip exposed to the water column for feeding, gas exchange, and waste expulsion (Bower and Blackbourn, 2003; Feldman et al., 2004). At the anterior end there is a dorso-ventral slit known as the pedal aperture through which the foot can be extended. This is the only opening from the mantle cavity to the exterior other than the siphons. The internal structures of P. generosa are typical of most bivalves. Water passes into an infrabranchial chamber from the inhalant siphon, through the gills, into the suprabranchial chamber, and out through the exhalant siphon. Food particles are trapped by the gills and transported to the labial palps for sorting. Subsequently, the food is passed to the esophagus and into the stomach. The digestive  4  structures – comprised of the style sac, stomach, intestine, and digestive gland – are contained in the visceral mass along with the gonad (Bower and Blackbourn, 2003). Panopea generosa are gonochoristic broadcast spawners, males reproductively maturing at 75-100 mm shell length (~ 2–4 years of age) and females maturing at >100 mm shell length (~ 5 years of age) (Andersen, 1971; Breen and Shields, 1983; Campbell and Ming, 2003). Geoducks are reproductively competent for at least 100 years (Sloan and Robinson, 1984). Gametogenesis begins in late summer and continues through winter until gonads mature in February to March in the Pacific northwest. Spawning begins in March–April and ends in mid-summer (July–August) (Andersen, 1971; Goodwin, 1976; Sloan and Robinson, 1984). Embryogenesis in P. generosa is believed to follow the general meiotic development process typical of most bivalves (Feldman et al., 2004). Fertilized eggs (82 μm in diameter) begin to divide within 2 h of fertilization and reach the D-stage within 48 h at 14˚C (Goodwin, 1973). At 165 μm shell-length the umbone appears and at a length of 300 μm the foot becomes visible (Goodwin et al., 1979). Metamorphosis begins at a shell length of about 380 μm after approximately 47 days at 14˚C or 30 days at 18˚C (Goodwin et al., 1979). The cues that induce settlement in geoduck larvae are unknown, but geoduck beds tend to be associated with chaetopterid polychaete mats (Cooper and Pease, (1988), perhaps because they act as a passive depositional force, similar to what is seen with bivalve settlement in eelgrass beds (Bologna and Heck, 2000). However, geoduck larvae metamorphose in the presence of the tubes, supernatant, and seawater extracts of the polychaetes Spiochaetopterus costarum, Phyllochaetopterus prolifica, and Diopatra ornata (Cooper and Pease, 1988). At metamorphosis the swimming function of the  5  velum is lost, spines develop on the growing edge of the shell, and the post-larvae begin to actively crawl or burrow using their foot (Goodwin et al., 1979; Goodwin and Pease, 1989). This dissoconch stage lasts for 2–4 weeks (Goodwin and Pease, 1989) and the post-larvae pedal-palp feed using their foot to transfer detritus to the mouth (King, 1986). Also, the post-larvae can use byssal threads to attach to substratum surfaces or form long threads that act as ‘parachutes’ to help disperse them on ocean currents (Goodwin and Pease, 1989). Post-larval bivalves can disperse up to 15 km using ‘parachute’ threads (Beukema and de Vlas, 1989). Juveniles of 1.5–2.0 mm shell length take on the general morphology of adults and tend to favour burrowing activity over crawling on the substratum surface (King, 1986). Beyond 5 mm shell length it is believed that the juveniles become sessile and stay in one location permanently (Goodwin and Pease, 1989). Mortality rates are highest during the juvenile phase (up to 75 mm shell length) (Sloan and Robinson, 1984; Orensanz et al., 2004), presumably in large part due to predation (Goodwin, 1984). There are anecdotal reports that newly planted juveniles (for aquaculture) are consumed by the sunflower star (Pycnopodia helianthoides), the lean basket-whelk (Nassarius mendicus), the red rock crab (Cancer productus), the graceful crab (C. gracilis), starry flounder (Platichthys stellatus), English sole (Parophrys vetulus), rock sole (Lepidopsetta bilineata), sand sole (Psettichthys melanostictus), and moon snails (Polinices lewisii and Natica sp.) (see Goodwin and Pease, 1989). Controlled experiments showed that C. productus and C. gracilis can consume 40 out of 100 juvenile geoducks within 48 h (Beattie, 1992; Beattie and Blake, 1999). Mortality rates slow within a year of settlement (Sloan and Robinson, 1984) and it is assumed that  6  the clams are relatively safe from predation after digging to a depth of 60 cm when they are about 2 years old (Goodwin and Pease, 1989). There are, however, reports of sea otters (Enhydra lutris) eating adult geoducks in BC, but there is no strong evidence that these otters have a significant impact on geoduck populations (Reidy, 2011). Geoducks follow the von Bertallanffy model, growing rapidly in the first 3 years (Goodwin, 1976) but gaining little in shell length and mass after 10 years of age (Goodwin, 1976; Bureau et al., 2002; 2003). Growth rate correlates to location and depth (Goodwin and Pease, 1991; Bureau et al., 2002; 2003), likely as a function of temperature and food availability. Geoducks tend to be larger in more southern areas and shallower waters while tending to be smaller in muddy substrata compared to mud-sand or sand substrata (Goodwin and Pease, 1991). The density of geoducks (ind m-2) varies with latitude (increasing from 48.2˚ to 47˚), water depth (increasing from <9 m depth to >13.7 m deep), and sediment type (more clams found in mud-sand or sand than mud or peagravel) (Goodwin and Pease, 1991).  1.3 Geoduck aquaculture and potential Geoduck aquaculture research began at the Point Whitney Laboratory, WA in the early 1970s focusing on larval development (Goodwin, 1973; Goodwin et al., 1979) and growth modeling (Goodwin, 1976; Beattie, 1992). Research later focussed more on outplanting trials but seed spread directly to the bottom with no predator protection had poor survival rates averaging a mere 6.2%. The use of predator exclusion devices such as screen capped PVC tubing have improved juvenile survival rates up to 60-80 % in the first year after outplanting(Beattie, 1992). Tubes are 30 cm long and 10 – 15 cm in  7  diameter with 1.27 cm mesh caps (Beattie and Blake, 1999). The tubes are generally removed after the first year (Beattie and Blake, 1999), Predator exclusion in BC has focused on using netting in sub-tidal plots(Beattie and Blake, 1999). Nursery techniques have also progressed, most notably with the development of raceway nursery systems which contain a sand substrate (Beattie, 1992; Davis and Barenberg, 2000a). These allow post-larval geoducks to pedal palp feed while eliminating the harsh grading and cleaning practices used in traditional upwelling nurseries (Beattie, 1992). Hatchery production of geoduck seed is highly variable from year to year. Stock crashes are often related to disease outbreaks (Kent et al., 1987; Elston et al., 2008) that are potentially linked to climate change (Elston et al., 2008). There is also the distinct possibility that deficiencies lie with hatchery methodology. One of the only references available for geoduck hatchery techniques (Shaul, 1981) concedes that the techniques described were largely adopted from methods developed for other species and therefore may not be optimal for P. generosa.  1.4 Factors influencing the gonad development and larval growth and survival of molluscan bivalves  Hatchery production of bivalves generally consists of two phases, a gametogenesis or broodstock conditioning period and a larval rearing stage (Helm et al., 2004). Key factors determining success in both phases include various physical and biological factors (temperature, salinity, water quality, stocking density and disease) and nutritional factors (ration, feed type, feed quality) (Brenko and Calabrese, 1969; Gallager  8  et al., 1986; Utting, 1986; His et al., 1989; Helm et al., 2004). There is dearth of research concerning the effects of these factors on P. generosa but there is a broad base of information on bivalves in general. The following sections of this chapter review the influence of various physical and biological factors on bivalve gonad development and the effects of various nutritional factors on larval bivalve development. Throughout, the knowledge gaps relevant to P. generosa are identified.  1.4.1 Effects of physical and biological factors on broodstock conditioning of bivalve molluscs 1.4.1.1 Introduction The following sections discuss the effects of temperature, salinity, water quality (i.e. dissolved oxygen, carbon dioxide, and nitrogenous wastes), and nutrition on bivalve reproductive conditioning for commercially-cultured, temperate species in North America and Europe, as these species represent the largest proportion of the literature. These factors are discussed because they are the essential to the general health of bivalves and contribute to the success of gametogenesis. This review will identify those conditions that maximize gamete production and establish the knowledge gaps for P. generosa culture that I will address with my thesis.  1.4.1.2 Temperature Temperature is considered the most important environmental factor controlling bivalve reproduction (Sastry, 1968) and central to successful broodstock conditioning  9  (Loosanoff and Davis, 1963). Bivalves have a minimum temperature for gametogenic activity. For example, the eastern oyster, Crassostrea virginica, needs a temperature of 15˚C or greater (Loosanoff and Davis, 1952) and the brooding Olympia oyster, Ostreola conchaphila, will never reach sexual maturity at 12˚C (Santos et al., 1993). Likewise, bivalve broodstock held at high temperatures may fail to ripen (Heasman et al., 1996), lose condition (Wilson et al., 1996), show declining carbohydrate levels (Mann, 1979b; Delaporte et al., 2006), or produce smaller eggs (Honkoop and van der Meer, 1998). An optimal temperature provides for the best gonad development and gamete production including reduced time to gamete maturation (Mann, 1979b, 1979a; Fabioux et al., 2005), increased numbers of eggs (Honkoop and van der Meer, 1998), more rapid accumulation of lipid and protein in the gonad (Martínez et al., 2000b), and more successful larvae in terms of survival, growth, and metamorphosis (Robinson, 1992a; Utting and Doyou, 1992). The Pacific oyster, Crassostrea gigas, for example, has more rapid gametogenesis between 18 and 24˚C than at 12˚C (Mann, 1979b) and produces more mature oocytes (over 19 days) at 22˚C than at 16, 19, or 25˚C (Chávez-Villalba et al., 2002). Holding broodstock of the Kumamoto oyster, C. gigas kumamoto (= C. sikamea), at 24˚C, as opposed to 20˚C, can reduce the conditioning period by 2–6 weeks (Robinson, 1992a). The Chilean scallop, Argopecten purpuratus, had a higher percentage of mature broodstock at 16˚C than at 20˚C (42.6% compared to 29.1%) (Martínez et al., 2000a). Optimal temperature for broodstock conditioning in P. generosa is unknown. Goodwin et al. (1979) conditioned the Pacific geoduck clam at 9–10ºC (near ambient winter temperatures), but did not determine an optimal temperature. Therefore, one objective of my thesis is to examine geoduck gonad development over a range of  10  temperatures relevant to the BC environment with the goal of identifying the temperature best suited for broodstock conditioning. 1.4.1.3 Salinity Salinity can significantly influence reproductive development in bivalve molluscs (Mackie, 1984). Inappropriate salinity triggers shell closure as a typical response to isolate internal organs from the hyper or hyposaline environment (Shumway, 1977; Strange and Crowe, 1979; Davenport and Wong, 1986; Elston et al., 2003), but cell volume must be regulated if water pumping resumes (Van Winkle, 1968; Deaton and Pierce, 1994; Glémet and Ballantyne, 1995; Neufeld and Wright, 1996, 1998; Deaton, 2001). Osmoconformers, like geoducks, maintain cell volume through the gain or loss of organic and inorganic solutes (Shumway, 1977; Shumway and Youngson, 1979; Somero and Bowlus, 1983), a homeostatic mechanism that has a cost, as indicated by increased oxygen consumption in several studies (Stickle and Sabourin, 1979; Kim et al., 2001; Hamer et al., 2008). An increased metabolic demand can result in a negative energy balance (Aldridge et al., 1995), especially if combined with reduced clearance rates as observed in lowered salinities (Hutchinson and Hawkins, 1992; Rajesh et al., 2001). An energy deficit leads to reduced somatic (Nakamura et al., 2005) or reproductive growth (Davis, 1958; Navarro and Gonzalez, 1998; Navarro et al., 2000). Survival is also influenced by salinity. Intertidal clams such as Mya arenaria are relatively tolerant and can survive at a salinity of 6 for at least 4 wk (Shumway, 1977). Other examples include the Manila clam, Venerupis philippinarum, which can tolerate a salinity of 8 for up to 8 d (Elston et al., 2003) and the New Zealand little neck clam, Austrovenus stutchburyi, which can tolerate sustained (>30 d) exposure at a salinity of  11  <10 before survivorship decreases (McLeod and Wing, 2008). In contrast to intertidal clams, subtidal scallops have much lower tolerance to hyposaline conditions. For example, step-wise acclimated (i.e. salinity change of 3 every 3 d) lion’s paw scallops, Nodipecten nodosus, died rapidly below a salinity of 22 (Roldán-Carrillo et al., 2005) and had a LC50 (48 h) at a salinity of approximately 23 (Rupp and Parsons, 2004). Also, the Japanese scallop, Patinopecten yessoensis, died within 24 h at a salinity of 18 (Izumi et al., 2000). The response of adult P. generosa to salinity changes has not been studied, but this species has poor success rates of embryogenesis at salinities below 25 (Goodwin, 1973). This result, and the fact that P. generosa typically inhabits physically-stable subtidal environments (Sloan and Robinson, 1984), suggests that they may be less tolerant to salinity changes than intertidal clams. I propose to fill this knowledge gap in Chapter 3 by examining survival, oxygen consumption, clearance rate, and reproductive development over a range of salinities relevant to BC estuaries.  1.4.1.4 Water quality Water quality can have a profound influence on broodstock conditioning, especially in a closed, re-circulating water system where there can be accumulation of nitrogenous wastes and dissolved carbon dioxide and a depletion of dissolved oxygen (Lawson, 1995) and especially under elevated temperatures when metabolic rate is increased (Hamburger et al., 1983; Riisgård and Seerup, 2003) and oxygen solubility is lower (Mann, 1978; Aldridge et al., 1995). For example, ammonia levels in tanks  12  holding V. philippinarum were 0 mg l-1 at 10˚C but 20 mg l-1 at 18˚C (Spencer and Edwards, 1995). Epifanio and Srna (1975) studied the 96-h median tolerance limits of nitrite ion (sodium nitrite at 8×10-2 mol l-1), ammonia (ammonium chloride at 5×10-2 mol l-1), nitrate (sodium nitrate at 5×10-2 mol l-1), and orthophosphate (sodium phosphate at 19×10-2 mol l-1) on adult C. virginica and the northern hard clam, Mercenaria mercenaria. They found that these intertidal species had relatively high tolerance to these compounds. Juveniles (30–40 mm shell length) of the deep-sea scallop, Placopecten magellanicus, however, are more sensitive to NH3 [96-h LC50 of 1.8 mg l-1 NH3 at 4˚C and 1.0 mg l-1 NH3 at 10˚C (Abraham et al., 1996)]. Macrobenthic organisms typically tolerate a low dissolved oxygen (DO) level (Diaz and Rosenberg, 1995; Miller et al., 2002) but hypoxia can result in lowered metabolism or increased cardiac output and ventilation rate (Burnett, 1997), all of which may result in an energy balance unfavourable for gonad development. Proper stocking densities and water exchange rates can help maintain favourable DO levels. For example, a Manila clam stocking density of 167 g live weight l-1 will have 50% DO saturation at 1 tank volume exchange rate per hour compared to near full saturation at 5 tank volumes per hour (Spencer and Edwards, 1995). In addition, low DO is typically correlated with hypercapnia that can lower pH levels in the water and cause acidosis in aquatic animals (Burnett, 1997). The effects of water quality on adult P. generosa have not been studied. In a hatchery, however, water can be aerated or oxygenated, carbon dioxide can be readily removed with ‘air stripping’, and ammonia can be reduced with biofiltration (Lawson,  13  1995). For these reasons I do not propose to examine the effects of water quality in my thesis. 1.4.1.5 Nutrition Supplemental feeding of broodstock is beneficial because it increases gametic output and decreases conditioning time (Helm et al., 1973; Chávez-Villalba et al., 2003b), but the benefits depend greatly on both the quantity and quality of feed. Successful embryogenesis and early larval survival are largely dependent on energy and structural components incorporated into the egg by the parent prior to fertilization (Gallager and Mann, 1986; Utting and Millican, 1997), at least until the trochophore or veliger stage when the digestive organs become functional (depending on the species) (Bändel, 1988; Carriker, 2001), Important nutritional factors are ration, algal species, biochemical composition of feed, and feed type (i.e. live phytoplankton, preserved phytoplankton, supplemental and grain-based feeds).  1.4.1.5.1 Feed ration Broodstock can be conditioned in the hatchery without the addition of food, but results depend on the endogenous lipid, carbohydrate, and protein reserves at the time of collection (Utting and Millican, 1997; Okumus and Stirling, 1998; Racotta et al., 1998; Kang et al., 2000; Chávez-Villalba et al., 2002, 2003a, 2003b; Dridi et al., 2007). Natural reserves in bivalves vary with season (Cigarría, 1999; Dridi et al., 2007; Kang et al., 2007), year (Navarro et al., 1989; Lubet et al., 1991; Ren et al., 2003), and location (Barber et al., 1991; Chávez-Villalba et al., 2003a) and unfed wild-collected broodstock  14  cannot be relied on for consistent gamete output. Eggs produced by unfed broodstock can be of high quality [i.e. normal size, normal lipid content (Caers et al., 2002; Nevejan et al., 2003b), normal embryogenesis rates (Robinson, 1992a; Chávez-Villalba et al., 2003b)], but their fecundity is typically low compared to fed broodstock (ChávezVillalba et al., 2003b; Delgado and Pérez-Camacho, 2003). Feeding broodstock an appropriate ration can maintain condition index and shorten the conditioning period (Delgado and Pérez-Camacho, 2003). Most often, ration is simply determined using a percentage of the weight of the broodstock animal. A generalized recommendation is a 3% ration of dry weight algae to dry meat weight animal for bivalves held at temperatures below 20°C and 6% for bivalves held at temperatures greater than 20˚C (Utting and Millican, 1997). For adult clams and oysters, 3.2 billion algal cells per day of mixed algal species is recommended per animal, amounting to approximately 3% ration dry weight algae to dry meat weight animal (Utting and Spencer, 1991). Grooved carpet shells, Ruditapes decussatus, held at 18˚C have a positive energy balance at a ration above 0.1% (based on ash-free dry weight of the feed in relation to live weight of the clams) (Delgado and Pérez-Camacho, 2003). In V. philippinarum it was found that the most favourable condition index and gonad occupation percentage was at a ration of 1,000–1,100 μg organic weight of phytoplankton g-1 clam live weight at 20–21˚C (approximately 0.1% ration) (Delgado and PérezCamacho, 2007). The king scallop, Pecten maximus, has been fed a phytoplankton mixture that consists of Pavlova lutheri (20% by cell count), Isochrysis aff. galbana (20% by cell count), Chaetoceros calcitrans (20% by cell count), and Skeletonema costatum (40% by cell count) at roughly 10 billion cells individual-1 (approximately  15  0.14% ration of dry algae to live weight) d-1 (Robert and Gérard, 1999). In France, C. gigas is typically fed with S. costatum and/or Tetraselmis suecica at a rate of 2 billion cells individual-1 d-1 (approximately 0.14% ration of dry algae to live weight) (Robert and Gérard, 1999). Based on all these results, the recommended ration for bivalve broodstock conditioning of 3% dry algal weight to dry meat weight animal, or similarly 0.10–0.14% organic algal weight to live weight, is generally appropriate. High-ration diets beyond those recommended for conditioning may be useful in the post-spawn phase to help broodstock recover. In C. gigas, for example, a ration of 12% (dry weight of algae to dry weight of oyster) lead to higher restoration rates of carbohydrate and lipid levels in tissue after spawning, compared to a ration of 4% (Delaporte et al., 2006). Over-feeding can, however, result in declining assimilation efficiency (Newell, 1981) and a decrease in egg production (Utting and Millican, 1998). Appropriate food rations for P. generosa conditioning have not been determined and there is no specific information available. Broodstock spawned for larval studies by Goodwin et al. (1979) and Goodwin (1973) were conditioned without food. A goal of my thesis is, therefore, to examine the gonad development of adult P. generosa fed a range of rations.  1.4.1.5.2 Algal species Microalgae are the basic diet for filter-feeding bivalves and in aquaculture species are selected for their ease of culture, cell size, digestibility, and food value (Webb and Chu, 1981; Coutteau and Sorgeloos, 1993). A list of commonly-used algal species and their characteristics are summarized in Table 1.1. In reviewing the effects of various  16  algal species on broodstock, Utting and Millican (1997) found that mixtures of algae containing various species combinations (including I. galbana, C. calcitrans, C. gracilis, and S. costatum) had high value for broodstock conditioning while uni-algal feeds (especially Dunaliella tertiolecta) had low value. Differences among these feeds were suggested to be the result of varying essential fatty acid (EFA) content. Essential fatty acids are important structural components, but their synthesis in bivalves is limited (Laing et al., 1990; Chu and Greaves, 1991). The most important of these, the omega-3 fatty acids eicosapentaenoic acid (EPA, 20:5n-3) and docosahexaenoic acid (DHA, 22:6n-3), are major membrane components in marine animals (Dunstan et al., 1994; Hendriks et al., 2003) and modulators of membrane function (Palacios et al., 2005). EPA may also serve as an energy source for embryogenesis (Hendriks et al., 2003), as noted during development in Ostrea edulis, Crassostrea sp., and V. philippinarum (Utting, 1993). In order to meet nutritional requirements for embryogenesis EFAs are selectively accumulated in gonad tissues from exogenous food sources (Soudant et al., 1996a,b; Palacios et al., 2005). If threshold levels of EFAs are not met during conditioning, egg quality can be compromised. For example, low-DHA diets (Dunaliella spp.) supplied to broodstock of C. gigas resulted in poor larval survival – less than half that of broodstock fed live algae and DHA-emulsion diets (Caers et al., 2002). In general, diatoms have high levels of EPA and flagellates have high levels of DHA (Dunstan et al., 1994). Because of the different levels of DHA and EPA in different classes of algae, combination diets of at least one flagellate species and one diatom species are typically recommended for shellfish broodstock conditioning (Utting and Millican, 1997; Marasigan and Laureta, 2001; Helm et al., 2004).  17  Adult broodstock are capable of ingesting and digesting a large range of algal shapes and sizes (Jones et al., 1993) but some algal species, such as T. suecica, may be rejected by broodstock because of their large size, as has been shown with the commercial scallop, Pecten fumatus (Heasman et al., 1996). Other algae, such as Phaeodactylum tricornutum, a species with good nutritional properties based on EPA levels, is poorly ingested and digested due to its long chains and spines (Helm et al., 2004). Organic content and biochemical composition of the feed may also have an impact on ingestion rates (Ward et al., 1992) and hence broodstock nutrition. Poor quality food (i.e. low organic content) can induce heightened selection behaviour for higher organic matter particles as seen in the clam Mulinia edulis, the Chilean blue mussel, Mytilus chilensis (Velasco and Navarro, 2002), and the common cockle, Cerastoderma edule (Iglesias et al., 1992). Argopecten purpuratus appears to increase filtration rates when essential fatty acids (DHA) are detected, presumably through chemical receptors in the gills or labial palps (Navarro et al., 2000). It is unknown what algal species are appropriate for broodstock conditioning of P. generosa. Since algal types vary in acceptability and nutritional value it is important to identify those most suitable for broodstock conditioning. For this reason in Chapter 5 I propose to treat broodstock with various feed types and combinations. Feed types selected will represent different morphologies and biochemical composition.  1.4.1.5.3 Preserved and supplemental feeds Preserved and supplemental feeds are highly desirable options for bivalve hatcheries as the culture of live algae is expensive, labour intensive, and prone to crashes  18  (Jones et al., 1993; Caers et al., 2003). A failed algal culture system will quickly stop all production in a hatchery and can result in mass losses of current larval or seed stock. The ability to feed broodstock with either a fully or partially supplemental feed could significantly cut costs and minimize risks. The various forms of supplementation include: dried heterotrophically-grown algae, algal concentrates (pastes), yeast-based feeds (Coutteau and Sorgeloos, 1993), lipid-emulsion spheres, and grain-based additives such as starch and wheat germ (Albentosa et al., 1999; Pirini et al., 2007). We need to be cautious about conclusions on these types of feeds because the results obtained in commercial hatcheries are often inferior to those that are obtained in smaller-scale laboratory experiments (Coutteau and Sorgeloos, 1993; Jones et al., 1993). Heterotrophically-grown algae such as T. suecica and Schizochytrium sp. (Coutteau and Sorgeloos, 1993) – although relatively inexpensive to produce [30% the cost of live algae (Utting, 1993)] – have yielded inferior results for broodstock conditioning when compared to live algae. For example, the fecundity of V. philippinarum fed spray-dried T. suecica (6% dry algae/dry weight ration) was 25 to 50% that of clams fed live T. suecica, D. tertiolecta, or S. costatum (Laing and LopezAlvarado, 1994). Despite the limitations of this feed type as a single feed source it may have potential as a supplement to live algae (Utting, 1993; Laing and Lopez-Alvarado, 1994; Utting and Millican, 1997). Algal concentrates are commonly used for the remote setting of larvae (Jones and Jones, 1988; Jones et al., 1993), but applications for conditioning of broodstock are not well represented in the scientific literature, with most of the research kept confidential by commercial hatcheries (Robert and Trintignac, 1997). There has been some research on  19  juveniles and sub-adults suggesting there is potential for algal paste as a feed supplement (Esquivel and Voltolina, 1996; Heasman et al., 2000). For example, algal pastes were found to be effective supplements to live feed for juvenile C. gigas (Brown and Robert, 2002; Ponis et al., 2003a) and juvenile P. fumatus (Knuckey et al., 2006). Concentrates of Chaetoceros sp. and P. tricornutum fed to 4–5 cm shell length Gallo mussels, Mytilus galloprovincialis (daily ration of 6% of dry body weight) showed that live paste provided high growth rates and condition indices while air-dried, freeze-dried, and frozen pastes gave poor results (Cordero and Voltolina, 1997). Poor performance associated with frozen and dried concentrates is the likely result of cell destruction (Robert and Trintignac, 1997; Heasman et al., 2000; Knuckey et al., 2006). Lipid supplements come in microcapsule form (lipid-walled, gelatine-acacia, and micro-gel), lipid microspheres, and lipid emulsions (Robert and Trintignac, 1997). Lipid supplements alone are not appropriate for broodstock nutrition but can increase egg output when combined with live algae (Nevejan et al., 2003b). Using a standard algal diet (i.e. I. galbana and Chaetoceros neogracile), A. purpuratus supplemented with DHA and EPA (Caers et al., 1999b) or saturated fatty acids (Nevejan et al., 2003b) had improved fecundity, reduced conditioning period, and increased levels of EFAs in the gonad in comparison to non-supplemented scallops. Similarly, broodstock of C. gigas conditioned with the EFA-poor algal species D. tertiolecta had nearly double the fecundity and more than double the D-larvae recovery rate when supplemented with EPA- and DHA-emulsion spheres (Caers et al., 2002). Grain-based supplements do not appear to be particularly effective for the enhancement of bivalve conditioning. Wheat germ (high in 18:2n-6) fed to adult M.  20  galloprovincialis caused a drastically reduced tissue content of both DHA and EPA (Pirini et al., 2007). As a supplement to live algae, however, corn starch increased energy intake in R. decussatus (Albentosa et al., 2003) and produced an equivalent percentage of mature animals compared to either microalgae (I. galbana and Chaetoceros gracilis) or microalgae and lipids (DHA emulsion) in adult A. purpuratus (Martínez et al., 2000a). However, the study by Martínez et al. (2000a) found that the quality of the eggs of A. purpuratus was not as high with the starch-supplemented diet and, despite high fertilization rates (88–89%), survival to D-larval stage was poorer (50%) compared to live-algae controls (88%). Despite the drawbacks of being a nutritionally incomplete supplement, there may be potential for the use of carbohydrate feeds in the early stages of conditioning or in post-spawn recovery when accumulation of carbohydrate reserves is particularly important (Gabbott, 1983). Very little information is available on the effects of preserved feeds and supplements on P. generosa broodstock. Unsatisfactory broodstock conditioning results were obtained with a Schizochytrium-based feed (Davis and Barenberg, 2000c) but the details of this study are not available. This is an area of research worthy of examination but is a large subject, beyond the scope of this thesis, and therefore not examined.  1.4.1.6 Conclusions The factors reviewed – temperature, salinity, water quality, ration, and feed type/quality – are important with respect to bivalve broodstock conditioning for aquaculture purposes. Temperature is critical because of its influence as a cue for reproductive development. Operating outside of the optimal temperature range for gonad  21  development in a hatchery could inhibit commencement of gametogenesis (low temperature) or cause the degeneration of gonads (high temperature). Inappropriate salinity levels can stress bivalves and cause them to expend energy on osmoregulation. Since salinity is not easily controlled in a flow-through system (unless intake water is never influenced by fresh water) or an extensive culture system it is important to know the effects of variable salinity on bivalves used as broodstock or at a growout site. The correct feed ration must also be used or otherwise there may be nutritional deficiencies that retard development or reduce fecundity. It is also very important to identify which micro-algal species are acceptable to broodstock and result in quality eggs and larvae. Each of these factors (temperature, salinity, and nutrition) must be maintained and regulated within optimal ranges to facilitate gonad and gamete development in a hatchery. Chapters 2, 3, 4, and 5 examine the effects of temperature, salinity, ration, and algal feed type, respectively, on adult P. generosa. The objective of these chapters is to define a set of conditions that are effective for P. generosa broodstock conditioning.  1.4.2 Effects of stocking density and ration on larval growth and survival At present the information available on P. generosa larval rearing is limited in comparison to other commercially-important bivalve species in BC and WA, such as the Manila clam (V. philippinarum) (Toba et al., 1992; Helm et al., 2004) and the Pacific oyster (C. gigas) (Utting and Spencer, 1991; Thompson and Harrison, 1992; Laing and Earl, 1998). Early work on larval P. generosa development suggested that a stocking density of 3 inds ml-1 is superior to 4–10 inds ml-1 (Goodwin et al., 1979), but no actual data is provided to support the recommendations. Shaul (1981) later provided 22  suggestions for reducing larval geoduck density with advancing larval development and advised an algal cell density of 30,000 to 50,000 cells ml-1. The culture methods used by both Goodwin and Shaul were borrowed from guidelines for other species and the latter conceded that they may not be optimal for P. generosa. Despite being a species with strong aquaculture potential, there has been a dearth of published research on P. generosa larval rearing since these early studies in the late 1970s and early 1980s. Stocking densities (Deming and Russell, 1999; Yan et al., 2006) and rations (Riisgård, 1988; Pechenik et al., 1990; Beiras and Pérez-Camacho, 1994; Pérez-Camacho et al., 1994; Laing, 1995) are two critical factors affecting the growth and survival of hatchery-reared bivalve larvae. There is a wealth of information available on these factors for bivalve species other than P. generosa, but as suggested by Shaul the information may not be directly transferrable. Examples of this are the Manila clam which performs well at relatively high densities of >10 inds ml-1 (Laing et al., 1990; Utting and Doyou, 1992; Laing and Utting, 1994; Yan et al., 2006) and C. gigas which does well at densities ranging from 5 to 9 inds ml-1 (Utting, 1986; Ponis et al., 2006a,b,c). The scallop P. maximus also grows well at 5–9 inds ml-1 (Delaunay et al., 1993; Ponis et al., 2006b). Larger larvae, like those released by the brooding oyster O. edulis are typically reared at lower densities of <3 inds ml-1 (Helm et al., 1973; Holland and Spencer, 1973; Ferreiro et al., 1990; Jonsson et al., 1999). It is essential that the appropriate larval stocking densities are selected otherwise growth and survival could be compromised, especially if the densities are too high. High densities can deplete oxygen and increase water toxicity through the accumulation of metabolic wastes (Yan et al., 2006). Normal development is inhibited by high ammonia (Geffard et al., 2002) and low  23  oxygen levels can result in reduced growth rates and premature settlement (Wang and Widdows, 1991). Stocking at low densities may not reduce performance but it can unnecessarily limit the production of a hatchery system. Feed concentration affects feeding efficiency and growth of bivalve larvae (Gallager, 1988). High algal concentrations can overwhelm the feeding apparatus, leading to increased rejection rates of algae (Gallager, 1988) and inhibited growth (Loosanoff et al., 1953; Loosanoff and Davis, 1963). Another risk of excess algae is the creation of fouling that may lead to bacterial infestations and decreased survival (DiSalvo et al., 1978; Torkildsen et al., 2000; Helm et al., 2004; Torkildsen and Magnesen, 2004). In contrast, if feed concentration is too low the larvae spend an excessive amount of time and energy searching for food (MacKenzie and Leggett, 1991; Beiras and PérezCamacho, 1994). This can result in low growth rates and prolonged onset of metamorphosis (Robert et al., 1988; Pechenik et al., 1990; Tang et al., 2006). Since the optimal conditions for stocking density and ration for larval P. generosa are unknown another task of my thesis (Chapter 6) is to examine growth and survival over a range of paired treatments of larval stocking densities and rations. Of particular interest is the interaction between the factors which may provide information on how to optimize feeding rates if there are changes in stocking density.  24  Table 1.1 Biochemical components and dry weights of some commonly cultured phytoplankton species 1 Class  Species  Haptophyceae (=Prymnesiophyce ae)  Isochrysis galbana (TISO)  Protein (pg cell-1) 6.8  Carbohydrate (pg cell-1) 1.8  Lipid (pg cell-1) 5.9  Energy (cal x10-8 cell-1) 9.326  Dry weight (pg cell-1) 21  Total fatty acid (pg cell-1) 3.07  5.3  0.62  12.3  14.195  Diacronema vlkianum  Pseudoisochrysis paradoxa  Bacillariophyceae  Chaetoceros calcitrans  0.82  0.01  0.09  0.0128  0.25  0.048  0.834  0.77  0.18  0.08  2.5  0.545  0.181  0.61  0  0.231  0.3  0.11  0.061  1.64  3.8  0.68  1.8  3.674  DHA (pg cell-1) 0.26  0.028  16  Pavlova lutheri  EPA ( pg cell-1) 0.028  1.5  2.45  11.3  0.224  0.023  0.4  0.03  0.40365  0.0255  0.165  0.015  10% TFA  1.25% TFA  8.47 Chaetoceros gracilis  9  2  5.2  9.738  Source (Nevejan et al., 2003a) (RiveroRodríguez et al., 2007) (Delaunay et al., 1993) (Patil et al., 2007) (MartínezFernández et al., 2006) (Delaunay et al., 1993) (Ponis et al., 2006b) (Mourente et al., 1990) (Ponis et al., 2006b) (Laing et al., 1990) (Pernet et al., 2003b) (RiveroRodríguez et al., 2007) (Delaunay et al., 1993) (Rico-Villa et al., 2006) (Lavens and Sorgeloos, 1996) (Laing et al., 1990) (Martinéz et al., 1992)  25  Class  Species  Protein (pg cell-1)  Carbohydrate (pg cell-1)  Lipid (pg cell-1)  Energy (cal x10-8 cell-1)  Chaetoceros muelleri Chaetoceros neogracile Phaeodactylum tricornutum  Prasinophyceae  Skeletonema costatum Tetraselmis suecica  Dry weight (pg cell-1) 23.15  70 23  6.4  10.7  22.968  13.1  2.4  5  11.555  52.1  20.2  16.8  47.317  Total fatty acid (pg cell-1)  12.93  EPA ( pg cell-1) 0.241  DHA (pg cell-1) 0.0116  0.3  0  (MartínezFernández et al., 2006) (Nevejan et al., 2003b) (Lavens and Sorgeloos, 1996) (RiveroRodríguez et al., 2007) (Dunstan et al., 1994) (RiveroRodríguez et al., 2007) (Delaunay et al., 1993) (RiveroRodríguez et al., 2007) (Utting, 1993)  0.52  0  (Utting, 1993)  0.44  0  0.9  0.1  1.21  0  (Laing et al., 1990) (Laing et al., 1990) (Laing et al., 1990)  nd  nd  nd  nd  nd  nd  4.13% FAME 0% TFA  0% FAME 0.1% TFA  1.45  0.155  3.79  0.1473  1.38  0.11  1.32  0  76.7  5.6 168.2  15.7  246.82  Chlorophyceae  Dunaliella tertiolecta  Nannochloris atomus  20  6.4  12.2  5  15  4.5  27.98  9.11  10  12.94  Source  (Caers et al., 2002) (Utting, 1993) (Langdon and Waldock, 1981) (Laing et al., 1990) (Mourente et al., 1990)  nd = not determined; FAME = fatty acid methyl esthers; TFA = total fatty acids 1 A version of this table has been published. Marshall, R., McKinley, R.S., Pearce, C.M., 2010. Effects of nutrition on larval growth and survival in bivalves. Reviews in Aquaculture 2, 33-55.  26  1.5 Thesis objectives and hypotheses The overall objective of this dissertation was to identify hatchery conditions and practices to improve the broodstock conditioning (i.e. gonad development) and larval rearing of P. generosa. To achieve the overall objective with respect to broodstock conditioning, adult clams were subjected to various levels of each key factor indentified in the review (temperature, salinity, feed ration, and feed type/quality; Chapters 2, 3, 4, and 5, respectively). Several methods were applied to evaluate responses of the clams to the various treatments. The gonadosomatic index (Chapters 2, 3 ,4, and 5) was applied as a bulk indicator of relative gonad size but for more conclusive evaluations of gonad and gamete development histological analysis was used (Chapters 2, 3, 4, and 5). Since P. generosa have a gonad layer sheathing the digestive gland, estimates of gonad development based on the thickness (Chapter 3) and surface area (digestive gland surface area : total visceral mass surface area) of this layer (Chapters 4 and 5) were developed. Ash-free dry weight was also used as a bulk indicator of organic content in gonads (Chapters 2 and 3). Other response variables analysed were mortality rates (Chapters 2, 3, 4, and 5) and condition index (soft tissue weight : shell weight) (Chapters 2, 3, 4, and 5). Oxygen consumption and clearance rates were used in Chapter 3 to make inferences about physiological and behavioural responses that relate to salinity stresses. Clearance rates were used in Chapters 4 and 5 to make inferences about feeding response to variations in ration and feed type, respectively. Ingestion rates were examined in Chapter 5 to estimate cell numbers and the mass of algae ingested. Chapter 6 examined growth (shell length) and survival of larvae of P. generosa supplied with various feed rations and held at various stocking densities.  27  1.5.1 Experiment 1: Effect of temperature on the gonad development of the geoduck clam It is unknown which temperature range is best for geoduck clam broodstock conditioning in a hatchery setting. Optimal temperature can result in shortened gametogenesis times and higher gamete output. Excessively low or high temperatures can inhibit gametogenesis or spawning. To identify the most appropriate temperature, adult geoducks were subjected to a range of temperatures that reflect the seasonal extremes in BC waters. The effects of temperature were tested over time at levels of 7, 11, 15, and 19ºC by examining spawning percentage and reproductive development (assessed by gonadosomatic index, ash-free dry weight of gonads, development index, oocyte maturity, gamete and connective tissue occupation). My hypothesis is that temperature manipulation will influence spawning activity and reproductive development with the optimal temperature for broodstock conditioning being in the range where geoducks peak during natural seasonal cycles [i.e. 8–12˚C (Sloan and Robinson, 1984)].  H0: Temperature will not significantly affect the proportion of clams spawning. HA: Temperature will significantly affect the proportion of clams spawning. I predict that the proportion of clams spawning will be highest at the temperature closest to natural conditions when spawning occurs. H0: Temperature will not significantly affect reproductive development. HA: Temperature will significantly affect reproductive development. I predict that: -temperatures closest to the natural conditions when gonad maturity peaks will  28  produce higher gonadosomatic index than other temperatures. -higher temperatures will produce lower ash-free dry weight than lower temperatures. -higher temperatures will produce gonads in a more resorbed state (based on histological analysis for development classification) than lower temperatures. -higher temperatures will produce a lower proportion of mature oocytes than lower temperatures -lower temperatures will produce higher gamete occupation (more oocytes per follicle and more oocytes per unit area in the gonad) than higher temperatures. -higher temperatures will produce higher connective tissue occupation levels than lower temperatures.  1.5.2 Experiment 2: Effect of salinity on the gonad development of the geoduck clam Geoduck clams typically inhabit physically-stable subtidal environments and are anecdotally reported to have poor survival at lower salinities. Salinity stress in other bivalves has been shown to increase mortality, increase metabolic rates, inhibit reproduction, and reduce clearance rates. How reduced salinities may impact P. generosa, however, is unknown at this time. Since aquaculture operations in BC are often in estuarine environments with variable salinity levels this is important to know. To address this knowledge gap the effects of four salinities (17, 20, 24, and 29) were tested on the survival, condition index, reproductive development (assessed by gonadosomatic index, ash-free dry weight of gonads, development index, oocyte diameter, connective tissue occupation, and gonad thickness), oxygen consumption, and clearance rates of adults.  29  H0: Salinity will not significantly affect survival. HA: Salinity will significantly affect survival. I predict that lower salinities will produce higher mortality rates than higher salinities. H0: Salinity will not significantly affect condition index and tissue water content. HA: Salinity will significantly affect condition index and tissue water content. I predict that lower salinities will reduce condition index and increase water content in comparison to higher salinities. H0: Salinity will not significantly affect reproductive development. HA: Salinity will significantly affect reproductive development. I predict that lower salinities will result in gonads with: -lower gonadosomatic index -lower ash-free dry weight -less ripened developmental stage -smaller oocyte diameters -lower gamete occupation levels -higher connective tissue occupation levels -a thinner layer of gonad sheathing the visceral mass than higher salinities. H0: Salinity will not significantly affect oxygen consumption. HA: Salinity will significantly affect oxygen consumption. I predict that lower salinities will increase oxygen consumption in comparison to higher salinities. H0: Salinity will not significantly affect clearance rates.  30  HA: Salinity will significantly affect clearance rates. I predict that lower salinities will reduce clearance rates in comparison to higher salinities.  1.5.3 Experiment 3: Effect of food ration on the gonad development of the geoduck clam Endogenous reserves in broodstock at the time of collection can have a significant influence on conditioning success but cannot be solely relied upon for gametogenesis due to variations with season, year, and location. Supplemental feeding of broodstock with an appropriate ration can provide surplus energy that can be channelled to reproductive growth which, in turn, can shorten the conditioning period and increase fecundity. Both under- and over-feeding can, however, result in decreased egg production. The ration levels appropriate for broodstock P. generosa are, however, unknown at this time and must be studied in order to improve hatchery production. To investigate the influence of ration, a range of rations [0.8 × 109, 2.4 × 109, 4.0 × 109, 5.6 × 109, 7.2 × 109, and 10.0 × 109 cells ind-1 d-1 (50:50 cell count of Isochrysis sp. and C. muelleri) were fed to adult P. generosa over time. The impact of ration was evaluated by examining mortality, clearance rates, ingestion rates, condition index, and reproductive development (assessed by spawner percentage, gonadosomatic index, digestive gland index, gonad development classification, oocyte diameter, gamete occupation, and connective tissue occupation).  H0: Ration will not significantly affect survival. HA: Ration will significantly affect survival. I predict that excessive or low rations will increase mortality in comparison to moderate rations.  31  H0: Ration will not significantly affect the proportion of clams spawning. HA: Ration will significantly affect the proportion of clams spawning. I predict that excessive or low rations will reduce spawning in comparison to moderate rations. H0: Ration will not significantly affect clearance/ingestion rates. HA: Ration will significantly affect clearance/ingestion rates. I predict that clearance and ingestion rates will increase with increasing ration up to an optimum ration. Excessive rations will decrease clearance rates. H0: Ration will not significantly affect condition index. HA: Ration will significantly affect condition index. I predict that lower rations will reduce condition index in comparison to higher rations. H0: Ration will not significantly affect reproductive development. HA: Ration will significantly affect reproductive development. I predict that lower rations will result in gonads with: -lower gonadosomatic index -a thinner layer of gonad sheathing the visceral mass (higher digestive gland index) -less ripened developmental stage -smaller oocyte diameters -lower gamete occupation (less area of gonad occupied by eggs or sperm) -higher connective tissue occupation than higher rations.  32  1.5.4 Experiment 4: Effect of food type on the gonad development of the geoduck clam The purpose of this chapter was to determine how various microalgal diets influence gonad and gamete development of P. generosa since it is currently unknown what algal species (or combination of species) are most appropriate for broodstock conditioning. The acceptability of a feed to bivalves can be assessed through filtration rates. The nutritional value of the feed for the broodstock can be reflected in fecundity, spawning rates, and levels of gonad development. To investigate the effects of feed type on broodstock conditioning four feed combinations were tested: Isochrysis sp. singly (high DHA, low EPA), C. muelleri singly (high EPA, low DHA), D. tertiolecta singly (low DHA. Low EPA), and a combination of Isochrysis sp and C. muelleri (balanced DHA and EPA). Responses examined were clearance/ingestion rates and reproductive development (gonadosomatic index, digestive gland index, development stage, oocyte diameter, gamete occupation and connective tissue occupation).  H0: Feed type will not significantly affect clearance/ingestion rates. HA: Feed type will significantly affect clearance/ingestion rates. I predict that clearance and ingestion rates will be lower with D. tertiolecta than with other species. H0: Feed type will not significantly affect reproductive development. HA: Feed type will significantly affect reproductive development. I predict that D. tertiolecta will result in gonads with: -lower gonadosomatic index -a thinner layer of gonad sheathing the visceral mass -a less ripened developmental stage  33  -smaller oocyte diameters -lower gamete occupation -higher connective tissue occupation levels  1.5.5 Experiment 5: Effects of stocking density and algal feed ration on growth,  survival, and ingestion rate of larval geoduck clams The appropriate levels of larval stocking density and algal feed ration have yet to be determined for P. generosa. Larval stocking density and algal feed ration are two factors known to impact larval bivalve growth, development, and survival. Overstocking of larvae can lead to water quality deterioration (e.g. reduced oxygen, increased ammonia) and increased competition for space and food. Over-feeding can overwhelm the larval feeding apparatus and/or cause mortalities due to water quality deterioration. Under-feeding can result in low growth rates and prolong the onset of metamorphosis as can over-feeding. In a review by Marshall et al. (2010), it was found that there was a correlation between nutrient levels fed to oyster larvae (on a per individual basis) and growth rates. This chapter, therefore, takes a novel approach by presenting rations on a cells per larva basis, rather than the typical cells per ml method. A factorial design was used with larval stocking densities of 2, 5, and 10 ind ml-1 and rations of 5000, 20,000, 40,000, and 100,000 cells ind -1. The combined effects of these two factors on larval growth rates, survival rates, and ingestion rates were examined.  H0: Larval stocking density will not significantly affect growth and survival rates.  34  HA: Larval stocking density will significantly affect growth and survival rates. I predict that treatments with the highest stocking density will have slower growth rates and lower survival rates than the other density treatments. H0: Ration will not significantly affect growth and survival rates. HA: Ration will significantly affect growth and survival rates. I predict that low or excessive rations will reduce growth and survival rates in comparison to moderate rations. H0: The interaction of larval stocking density and ration will not affect growth and survival rates. HA: The interaction of larval stocking density and ration will affect growth and survival rates. H0: Larval stocking density will not significantly affect ingestion rates. HA: Larval stocking density will significantly affect ingestion rates. I predict that treatments with the highest stocking density will have lower ingestion rates than the other density treatments. H0: Ration will not significantly affect ingestion rates. HA: Ration will significantly affect ingestion rates. I predict that low or excessive rations will reduce ingestion rates in comparison to moderate rations. H0: The interaction of larval stocking density and ration will not affect ingestion rates. HA: The interaction of larval stocking density and ration will affect ingestion rates.  35  2 Effect of temperature on gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850)2  To determine the appropriate broodstock conditioning temperatures for Panopea generosa, adults were held in the laboratory for 155 days at fixed temperatures of 7, 11, 15, and 19ºC. Clams held at the two lower temperatures (7 and 11ºC) displayed a more advanced state of reproductive development than those at the two higher temperatures (15 and 19ºC). Significantly higher percentages of individuals with mature gonads were seen at 7 and 11ºC than at 19ºC and significantly more oocytes follicle -1 were evident at 7 and 11ºC than at 15 and 19ºC. Clams at 7ºC had a significantly higher gonadosomatic index than at 11ºC (which were not significantly different from clams held at 15 or 19ºC), but this was likely because there was a significantly lower incidence of spawning activity at 7 than at 11ºC. The 11ºC treatment had a significantly higher overall percentage of individuals spawning between weeks 15 and 17 of the experiment than any of the three other temperature treatments. Gonads of individuals in the 19ºC treatment degenerated, with 0% mature oocytes and 90% of the gonad occupied by connective tissue after 113 days. In a hatchery setting, P. generosa should be held at a temperature of 11ºC if reproductive output is to be maximized. Lower temperatures (e.g. 7ºC), however, could potentially be used to hold ripe broodstock for extended periods.  2  A version of this chapter has been accepted for publication. Marshall, R., McKinley, R.S., Pearce,  C.M., 2012. Effect of temperature on gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850). Aquaculture, in press.  36  2.1 Introduction Temperature is a factor known to influence the reproductive biology of bivalves (Loosanoff, 1937; Loosanoff and Davis, 1952, 1963; Ropes and Stickney, 1965; Bayne, 1976). Manipulation of temperature under controlled conditions (e.g. laboratory or hatchery) can be a powerful tool for increasing the rate of gametogenesis – as seen in Tapes philippinarum (Mann, 1979a), Crassostrea gigas (Robinson, 1992a; ChávezVillalba et al., 2002; Fabioux et al., 2005), and Ostrea edulis (Mann, 1979b). It can also increase larval production [e.g. O. lurida (Santos et al., 1993)] and subsequent larval survival and spat fall [e.g. C. gigas (Robinson, 1992a)]. Specific studies on the effects of temperature on P. generosa gonad development, however, have not been conducted. Seasonal reproductive cycles do show strong associations of reproductive state with season (Andersen, 1971; Goodwin, 1976; Sloan and Robinson, 1984; Campbell and Ming, 2003) suggesting a likely correlation with temperature. Similar reproductive-cycle studies on other species of geoduck [P. zelandica (Gribben et al., 2004), P. abbreviata (van der Molen et al., 2007), and P. globosa (Aragón-Noriega et al., 2007)] have suggested that temperature is a major factor in reproductive development. The objective of the present study was to investigate the reproductive responses of P. generosa to being held at fixed temperatures in a controlled environment for prolonged periods. The response variables investigated were: timing of spawn events, condition index, gonadosomatic index, gametogenic (development) stage, oocyte maturity, oocytes per follicle, oocytes per unit area, and connective tissue occupation index. The ultimate goal of the study was to identify temperatures that are suitable for the maintenance and conditioning of P. generosa broodstock in a hatchery setting.  37  2.2 Materials and methods 2.2.1 Algal culture Live algal cultures of Isochrysis sp. (TISO clone, CCMP 1324) and Chaetoceros muelleri (CCMP 1316) were used as feed. Specific levels of fatty acids required for successful gametogenesis in P. generosa are unknown, but these two phytoplankton species were selected to provide a balanced complement of polyunsaturated fats known to be typically required for successful bivalve broodstock conditioning; C. muelleri has high levels of eicosapentaenoic acid (20:5n-3) (Helm et al., 2004) and moderate levels of arachidonic acid (20:4n-6) (Soudant et al., 1996c) while TISO has high levels of docosahexaenoic acid (22:6n-3) (Soudant et al., 1996b). The algae were grown semicontinuously in 300-L fibreglass columns and 500-L polyethylene bags at a temperature of 18.7 ± 0.1ºC (mean ± SD, n = 588) under full-spectrum fluorescent bulbs (Philips DayLight Deluxe®, Philips Electronics Ltd., Markham, Canada). Seawater for algal culturing was filtered to 0.2 μm, sterilized with sodium hypochlorite, neutralized with sodium thiosulfate, and fertilized with a modified Harrison’s formula (Harrison et al., 1980). That modification was the partial substitution of organic phosphates by inorganic phosphates.  2.2.2 Broodstock collection and initial maintenance Broodstock were collected from a natural geoduck bed on December 2, 4, and 7, 2006 in Comox Harbour, BC (49° 39.07´ N latitude, 124° 52.93´ W longitude). The clams were collected from a single genetic population (VanKoeveringe, 1998) as close to the same date and location as possible to minimize the influences of genetic variability  38  (Barber et al., 1991), season, and location (Chávez-Villalba et al., 2002, 2003a) on reproductive condition. Clams were collected by SCUBA diving at a depth of 9–12 m using a hydraulic harvester (consisting of a water jet supplied by a surface pump). Water temperature at depth ranged from 7.3 to 8.9°C and salinity was 28. The substratum was a cobble and mud mixture. After harvest, the clams were isolated from the elements, temperature controlled (5 to 7ºC), and delivered within 7 h to the Pacific Biological Station in Nanaimo, BC. Clams were subsequently held in seawater tables at 11.1 ± 0.8°C (mean ± SD, n = 14,400) with sand-filtered seawater until the start of the acclimation phase. Shell length (SL) (measured to the nearest 1.0 mm on the anterior-posterior axis of the right valve using vernier callipers) of the clams averaged 151 ± 18 mm (mean ± SD, n = 112). Live weight [to the nearest 0.1 g using a Mettler PM4800 Delta Range balance (Mettler Toledo, Mississauga, Canada)] averaged 1,412 ± 443 g (mean ± SD, n = 112). The minimum size of clam used was 110 mm SL to ensure an even sex ratio and sexual maturity, as reported by (Andersen, 1971; Sloan and Robinson, 1984; Campbell and Ming, 2003; Gribben and Creese, 2005). Reproductive senility has not been reported in P. generosa (Sloan and Robinson, 1984), but larger (over 2 kg and presumably older) individuals were avoided as reproductive success can diminish with age in bivalves (Sukhotin and Flyachinskaya, 2009).  2.2.3 Broodstock conditioning: experimental design On December 12, 2006 the clams were set vertically – to provide a more natural orientation – inside PVC tubes (H: 55 cm, Diameter: 12.5 cm) with one individual per  39  tube (Fig. 2.1). A U-shaped, PVC-coated, wire-mesh “girdle” (2.5 x 2.5 cm mesh size, 3 mm gauge wire, cut into 8 x 34-cm strips) was moulded around the shell of each clam. Without external pressure on a geoduck (normally provided by the natural substratum) the shell can gape, causing tears in the periostracum (which covers the shell, siphon, and muscular mantle) making the clam susceptible to infection. Necrotic tissue always developed in the immediate area where the shell separated from the periostracum. The tubes had an outflow 2.5 cm from the top, giving a functional volume of 6,443 ml. Tubes were placed vertically in Plexiglas water-bath tanks (L x W x H: 58 x 58 x 30 cm, Volume: 101 L) (Fig. 2.1). Eight tubes were placed in each tank (total of 16 tanks) and seven clams were randomly assigned to a tube with one tube per tank kept clear as a control for testing water quality parameters. Water was fed to each tube from a plastic header tank (L x W x H: 30 x 25 x 64 cm, Volume: 47.3 L) and entered the geoduck tubes approximately 10 cm from the bottom (Fig. 2.1). The average seawater flow rate in the tubes was 25.4 ± 6.8 L h-1 (mean ± SD, n = 336). Flow rate to each tube was set to exceed the average clearance rate of the clams which was determined to be 13.2 L h-1 clam-1 at 11 to 19ºC. Average clearance rate was determined through a preliminary study using the method of Quayle (1948) [as described by Coughlan (1969)]. Water flowed to waste except during feeding when it was recirculated. Salinity was at ambient levels (28–30) throughout the experiment. Photoperiod was maintained at 16 h light:8 h dark to replicate late spring/early summer conditions when natural spawning events occur (Sloan and Robinson, 1984). Lighting was provided solely by overhead fluorescent lights. After being introduced to the tubes, the clams were starved at ambient temperature [9.7 ± 0.14ºC (mean ± SD, n = 386)] for 3 weeks to promote resorption of  40  the gametes. Clams that died during the starvation period were replaced with healthy animals that were kept at ambient temperature in a reserve tank. Targeted experimental temperatures were 7, 11, 15, and 19°C which reflect the range of average surface water temperatures throughout the year in the Comox Harbour area (Mac’s Oysters Ltd., unpublished data). Clams were adjusted to final temperatures in increments of 1ºC d -1 beginning January 3, 2007. Four randomly-selected replicate tanks (experimental unit) at each temperature were maintained at each temperature by mixing heated and chilled seawater (sand filtered) which flowed into the header tanks. Conditioning at final experimental temperatures began January 17 (0 d). Actual experimental temperatures recorded were 7.1 ± 0.50ºC, 11.2 ± 0.48ºC, 14.9 ± 0.31ºC, and 19.1 ± 0.29ºC (mean ± SD, n = 2,419 per temperature). Target temperatures (7, 11, 15 and 19 ºC) are referred to in the remainder of the chapter. During the temperature-adjustment phase and the experimental phase the clams were fed at a rate of 4 × 10 9 cells animal-1 d-1 (C. muelleri:TISO; 50:50 by cell count; 0.1% dry weight algae: dry weight clam soft tissue) for 4 h. This was the maximum amount that adult P. generosa could ingest in a 4-h feeding, as determined in preliminary tests (see above). Since each clam was in its own conditioning tube, removal of an individual from the experiment had no effect on overall food ration received by each clam in a replicate tube.  2.2.4 Sampling One randomly-selected clam from each tank (i.e. 4 per temperature treatment) was selected at each sample date. Clams were sampled for condition and reproductive stage at 5 to 10 d post-harvest (December 12, 2006) and after conditioning for 0, 113,  41  135, and 155 d. Sampled clams were rinsed in sand-filtered seawater, blotted dry with paper towel, and weighed. The clams were then cut open and drained of internal water. The visceral mass (alimentary canal, digestive diverticula, sheathing gonad tissue, and attached foot), somatic tissue (all tissue other than the visceral mass), and shells were separated, blotted dry, and weighed to the nearest 0.1 g. Sub-samples of the visceral mass were collected for histological analysis (see below). Dry weights of soft tissues and shells were determined by drying them at 58.5 to 59.5ºC to constant weight (72 h minimum). Since small portions of the visceral mass were removed for histological analyses, a correction factor was necessary for determining the final visceral mass dry weight. This was done by multiplying the wet weight of the removed piece by the ratio of wet to dry weight of the remaining visceral mass tissue. The resulting value was added to the dry visceral mass weight.  2.2.5 Mortality rate and spawn events Health was assessed daily by examining for: lesions in the periostracum, a flaccid or distended siphon, cracks in the shell, excessive mucus, and slow or absent siphon aperture closure in reaction physical stimuli. Non-responsive clams were classified as dead. Checks for spawning (i.e. presence of eggs and/or sperm in each tube) were made throughout each day. 2.2.6 Condition and gonadosomatic indices Condition index (CI), based on the method of Walne and Mann (1975), was adapted for geoduck clams and calculated as: CI = (T × 100) / S  42  CI is the condition index, T is the total soft-tissue weight S is the shell weight.  This was calculated for wet (CIw) and dry (CId) weights to determine changes in the relative amounts of soft tissue. Changes in shell weight were assumed to be negligible (<0.7%) over the period of the experiment based on the allometric relationship models of age to shell weight in Sloan and Robinson (1984) and Bureau et al. (2003). The gonadosomatic index (GSI), which has been correlated with reproductive state in wild populations of P. generosa (Sloan and Robinson, 1984) and P. zelandica (Gribben et al., 2004), was calculated after Sloan and Robinson (1984): GSI = V/T GSI is the gonadosomatic index V is the visceral mass weight T is the total soft-tissue weight without valves.  This was calculated for wet (GSIw) and dry (GSId) tissue. The proportions were converted to percentages. It must be noted that although referred to as a ‘gonad’ index, the entire visceral mass is used in the numerator of the GSI since the gonad and visceral mass are not readily separated. This is the standard used in other geoduck studies (Sloan and Robinson, 1984; Gribben et al., 2004) .  43  2.2.7 Ash-free dry weight Visceral mass samples were analyzed for ash-free dry weight (AFWD) as an estimate of organic carbon (Moreno et al., 2001). Samples of approximately 300–400 mg were collected from each clam, rinsed in distilled water to remove excess salts, and blotted dry (Moreno et al., 2001). Samples were vacuum packed in plastic bags and stored at -80ºC for no more than 12 months before processing. This followed the procedures for storage of fish gill tissue for protein analysis (Weisbart et al., 1994). Dry weights of gonad tissue were determined by freeze drying the samples (Labconco FreeZone 6, Labconco Corporation, Kansas City, Missouri, USA) with a vacuum setting of 0.030 mBar and the collector at -48˚C for 24 h. Samples were held in a desiccating chamber for up to 2 days before being placed in a drying oven at 60˚C for 24 h and dry weights recorded. Samples were combusted using a muffle furnace (Sybron Thermodyne, Molyne 30400, Milwaukee, Wisconsin, USA) at 465˚C for 12 h. Postcombustion samples (ash) were weighed on a Mettler Toledo scale to 10 -5 g. The AFDW was calculated as: AFDW = (DW – AW)/DW × 100 where AFDW is the percentage of tissue that is organic, DW is dry weight, and AW is ash weight.  2.2.8 Histological sampling A small section (~1 g) was taken from the right side of the posterior end of the visceral mass of each sampled clam (Sloan and Robinson, 1984). The samples were fixed in Davidson’s solution for 72 h and transferred to 70% isopropanol for storage. Within 30 d the samples were cut into 4–5 sub-sections, dehydrated using a graded ethanol series, and embedded in paraffin wax. Embedded tissue samples were sectioned  44  to 5 μm, stained with hematoxylin-eosin, and mounted on slides for examination. Slides were examined using a Motic B5 compound microscope with Motic Images Advanced 3.2 software (Motic Electric Group Co., Ltd., Richmond, Canada). Sex was determined by the presence of oocytes or spermatocytes under microscopic examination. The tissue samples were classified as early active, late active, ripe, partially spent, or spent/resorbed, based on the classifications used for P. generosa (Andersen, 1971; Goodwin, 1976; Sloan and Robinson, 1984) and P. zelandica (Gribben et al., 2004). The developmental stage was determined by the most dominant stage of at least 10 randomly-chosen follicles in each gonad sample (Gribben et al., 2004). For data analysis, the stages were grouped into two categories: sexually mature (late active, ripe, partially spent) or sexually immature (spent/resorbed, early active). Diameters of 75 randomly-selected oocytes were measured to the nearest 0.2 μm for each female clam. Only oocytes with visible nuclei were measured (Hesselman et al., 1989; Villalejo-Fuerte et al., 1996; Gribben et al., 2004). The oocyte diameters were used to determine oocyte maturity (OM) (as a percentage of mature oocytes relative to total oocytes) using the scale developed for C. gigas: early gametogenesis = 3.0–16.0 µm, growing = 16.1–30.0 µm, mature = 30.1–40.0 µm, and degenerating = 40.1–60.0 µm (Chávez-Villalba et al., 2002). The scale classifications were appropriate except that the primary oocytes were slightly larger in P. generosa (10.3–16.0 µm). Gonads with no traces of oocytes were classified as degenerating. For oocytes per follicle (OF) the mean count from 5 randomly-chosen follicles from each female clam’s gonad sub-samples (20–25 follicles per clam) were used. The mean oocyte counts from 10 randomly-selected fields (2.5–3.5 mm2 each) from each  45  gonad sub-sample (40–50 readings per clam) were used to calculate the oocytes per unit area (OA) [after (Heffernan et al., 1989b)]. Fields excluded digestive tubule area. Connective tissue occupation index (COI), the ratio of connective tissue surface area to total gonadal surface area (including connective tissue, follicles, and vacuoles) was used as an indicator of gonad degeneration. A preliminary test comparing COI of males (n = 7) and females (n = 6) showed no significant difference between the sexes within the same temperature range (10–12ºC) (t-test: t = 0.64, df = 11, P > 0.5). This technique, therefore, allowed for the comparison of COI among treatments regardless of sex. The COI was determined colorimetrically using the Motic microscope and software. Images were analyzed in 8-bit greyscale (0 = black, 255 = white). Within that colour scale, connective tissue ranged from 113 to 233. Ten fields (each field 2.5–3.5 mm2) were measured from each gonad sub-sample (totalling 40–50 readings per clam) and the mean calculated. Fields excluded digestive tubule area. The COI is similar to the volume fractions developed by Lowe et al. (1982).  2.2.9 Statistical analyses To test the influence of temperature on mortality, a chi-square test for independence was used. For spawn events, the proportions of spawners to non-spawners in each replicate were analyzed using repeated measures ANOVA (repeated measures to account for multiple spawns from individuals) with temperature as a fixed variable, time as a random factor, and replicate tank as the subject. The CI data were analyzed using ANCOVA with time and temperature as fixed factors and dry shell weight as the covariate. A 2-way ANOVA was used to analyze GSI, developmental classification  46  (proportion of clams with mature gonads), OM, OF, OA, and COI with temperature and time as fixed factors. For all parametric tests, normality of data was tested using the Shapiro-Wilk W test and homogeneity of variances was tested using the modified Levene’s equal variance test. For ANCOVA, the F-test for equality of slopes was also used. To correct for violations of normality and homogeneity of variances the arcsine transformation was used for spawner proportions, development classification, OM, and COI while the natural logarithm transformation was used for OF. No transformations were necessary for CI, GSI, or OA. Significant differences among treatment levels were tested using a post-hoc multiple comparisons test (Tukey-Kramer test).  2.3 Results 2.3.1 Mortality rate and spawn events The overall mortality during the experimental period in all temperature treatments was 28% with all mortalities occurring between 0 and 112 d. An independence test for mortality indicated that temperature did not significantly influence mortality rate (χ2-test: χ2 = 1.708, df = 3, P = 0.635). In almost every case, mortalities were associated with damage to the periostracum. Although obviously-damaged clams were excluded from the experiment it was later noted that small (1–2 mm), difficult-to-see tears in the periostracum could become infected and ultimately result in death of the clam. Only 2 of the 31 individuals that died during the experiment showed no visible signs of infection due to periostracum damage. No spawn events were observed until the end of the acclimation period, 6 weeks after harvest (i.e. week 0 of the experiment). Spawning occurred in 2 distinct phases  47  which were separated by a 10-week quiescent period: Phase I (weeks 1–5) and Phase II (weeks 15–17). These two phases were analysed as unique data sets. In Phase I there were regular spawn events in all temperature treatments and temperature was not a significant factor affecting spawner percentage (Fig. 2.2, Table 2.1). During Phase I, the percentage of spawners was significantly affected by time and tank (subject), but not by the interaction between temperature and time (Table 2.1). Spawner percentage was significantly greater during weeks 2, 3, and 5 than during weeks 1 and 4 (Fig. 2.2). During Phase II, temperature was a significant factor affecting spawner percentage, while time, tank (subject), and the interaction between temperature and time were not (Table 2.1). There was a significantly higher percentage of spawners at 11ºC than at 7, 15, and 19º with no significant differences among the latter three treatments (Fig. 2.2).  2.3.2 Condition and gonadosomatic indices Both CIw and CId were significantly affected by time but not the interaction between temperature and time, while temperature significantly influenced CIw but not CId (Table 2.1). Both CIw and CId showed a decrease from day 0 to day 155 (Fig. 2.3a,b), indicating soft-tissue loss. Clams in the 7ºC treatment had a significantly higher CIw than those in the 19ºC one, while individuals held at 11 and 15ºC did not differ significantly from those at any other temperature (Fig. 2.3a). The lack of a significant effect of temperature on CId suggests that the differences seen with CIw were due to water content. The covariate (i.e. shell weight) was significant for both CIw and CId (Table 2.1). Both GSIw and GSId were significantly affected by temperature but not the interaction between temperature and time, while time significantly influenced GSIw but  48  not GSId (Table 2.1). The GSIw on day 135 was significantly higher than on day 113, but there were no other significant pair-wise comparisons among days (GSIw on days 0 and 155 both averaged 19%) (Fig. 2.4a). The wet and dry GSIs were both significantly higher at 7ºC than at all other temperatures with no other significant pair-wise comparisons among temperatures (Fig. 2.4a,b).  2.3.3 Ash-free dry weight For the analysis of visceral mass AFDW, day 0 samples were excluded because poor preservation methods resulted in the loss of several samples. AFDW was significantly affected by temperature, but not by time or the interaction between temperature and time (ANOVA: F3, 33 = 3.80, P < 0.05; F2,33 = 0.94, P > 0.2; F6,33 = 1.29, P > 0.2, respectively). Gonad organic content was significantly lower at 7˚C (85%) than 19˚C (90%). Levels at 11˚C (86%) and 15˚C (89%) were intermediate, but not significantly different from any other temperature treatment.  2.3.4 Histological sampling The percentage of geoducks with mature gonads was influenced by temperature but not time (Table 2.1). Interaction significance could not be determined since this was a non-replicated 2-way ANOVA. Clams held at 7 and 11ºC did not differ significantly from each other but had significantly higher percentages of individuals with sexually mature gonads (i.e. late active, ripe, partially spent) than those at 19ºC while clams in the 15ºC treatment did not significantly differ from those in any other temperature treatment (Fig. 2.5).  49  The percentage of mature oocytes was significantly influenced by temperature and the interaction between temperature and time, but not by time alone (Table 2.1). Only one significant pair-wise comparison among days within temperatures, however, was observed; in the 7ºC treatment there was a significantly higher proportion of mature oocytes on day 155 than on day 113 (Fig. 2.6a). On days 0 and 113, OM was not significantly affected by temperature treatment, but by day 135 differences among temperatures began to emerge with clams at 7ºC having significantly higher OM than those at 15 or 19ºC on day 135 and significantly higher OM than those at 19ºC on day 155 (Fig. 2.6a). Both temperature and time significantly affected OF, with no significant interaction between the two factors (Table 2.1). The number of OF remained fairly stable over time; the only detectable pair-wise time difference in OF was a significant decline from day 0 to day 113 (Fig. 2.6b). Clams held at 7 and 11ºC did not differ significantly in the number of OF, but they had significantly higher numbers of OF than those at 15 or 19ºC (Fig. 2.6b). Results for OA were similar to OF and data are not shown. Both temperature and time significantly affected OA, but the interaction between these two factors was not significant (Table 2.1). The average OA declined significantly from day 0 to day 113, but remained stable after that. Clams held at 7ºC had significantly more oocytes mm-2 than those at 15 and 19ºC. Individuals in the 11ºC treatment did not differ significantly from those in the 7 or 15ºC treatments, but had a significantly higher OA than individuals at 19ºC.  50  The COI was significantly affected by temperature, time, and the interaction between these two factors (Table 2.1). The COI did not significantly change in the 7, 11, or 15ºC treatments over the duration of the experiment, while the COIs for 19ºC significantly increased from day 0 to day 113 (and remained at similarly high levels until the end of the experiment). On day 0 there were no significant differences among any of the temperature treatments, but by day 113 clams held at 7ºC had a significantly lower COI than those held at 15 and 19ºC (Fig. 2.6c). The COI of clams at 7ºC remained significantly lower than those at 15 and 19ºC on day 135 and significantly lower than those at 19ºC on day 155 (Fig. 2.6c). At no point did the COI of clams at 7ºC differ significantly from the COI of those at 11ºC. The wide variance in the COI of clams at 15ºC on days 135 and 155 was likely due to the development of large empty follicles as a result of gonad degeneration.  2.4 Discussion Results of this study indicated that 11ºC was the most appropriate temperature for P. generosa broodstock conditioning. Supporting evidence included the 11ºC treatment having the highest percentage of clams spawning during weeks 15 and 17 (means: 70 and 80%, respectively) and having reproductive development indicators (i.e. OM, OA, OF) that were similar to the other temperature treatments or significantly better. The overall GSI (both wet and dry) was significantly lower at 11ºC than at 7ºC, but this was likely due to loss of mass from spawning (Heffernan et al., 1989a, 1989b; Hesselman et al., 1989; Kanti et al., 1993; Gribben et al., 2004). A temperature of 15ºC appeared to be less appropriate for broodstock conditioning than 11ºC because it triggered fewer spawn  51  events and showed more signs of degeneration with significantly fewer oocytes per follicle and high variation in the levels of connective tissue in the latter stages of the experiment (days 135 and 155). All temperature treatments had spawning individuals in Phase I (weeks 1–5) with all clams subsequently entering a quiescent phase (10 weeks) independent of temperature. Only the 11ºC treatment had a polymodal spawning pattern, having significantly higher percentages of spawning clams than any other temperature treatment during Phase II (weeks 15–17). In general, marine invertebrates are expected to show protracted or polymodal spawning behaviour when they experience conditions conducive to reproductive growth (e.g. optimal temperature and food production) for a prolonged period (Giese, 1959). Even within the same species, however, plasticity in reproductive strategy can occur, as has been demonstrated in relation to latitude (Keck et al., 1975). For example, Mercenaria mercenaria typically displays a bimodal reproductive cycle in warmer southern regions of the United States (Florida, Georgia, North Carolina, and South Carolina) and only one reproductive peak in northern regions (Connecticut, Delaware, and New York) (Heffernan et al., 1989a; Hesselman et al., 1989). The present study indicates that, under optimal laboratory conditions, P. generosa is able to extend the period of gamete production. Maintenance of optimal temperatures in the laboratory has also been shown to prolong reproductive activity in other bivalve species such as C. gigas, which were able to spawn and maintain ripe gonads for over 10 weeks (until trial termination) at 18 and 21ºC (Mann, 1979b), and T. philippinarum which could spawn and maintain ripe gonads at 15 to 21ºC for 4 to 8 weeks (until trial termination) (Mann, 1979a). Crassostrea gigas also maintained a high proportion (>60%) of mature oocytes  52  for over a month when held between 19 and 25ºC (Chávez-Villalba et al., 2002). The temperatures of 11 and 15ºC in the present study were closest to the temperature range (12–14ºC) observed during the natural spawning cycle of P. generosa (Andersen, 1971; Goodwin, 1976; Sloan and Robinson, 1984) further suggesting that this is the optimal range for broodstock conditioning. The 7ºC treatment proved to be appropriate for ripening of gonads, showing a relatively high GSI, OM, and OF with low COI, but despite the high levels of gonad maturity there was relatively little spawning at this temperature. This is not surprising as P. generosa in natural beds have been reported to ripen from late December to February (Goodwin, 1976; Sloan and Robinson, 1984; Campbell and Ming, 2003) when temperatures typically range from 7 to 9ºC (Sloan and Robinson, 1984). At low temperatures, though, bivalves typically maintain a high level of gonad development until spawning is triggered by an increase to a critical temperature (Hesselman et al., 1989; Kanti et al., 1993). The 7ºC treatment was likely below this critical temperature threshold and the low intensity spawns during Phase I were most likely triggered by the addition of food. Goodwin (1976) reported that P. generosa could spawn at 8.5ºC in the laboratory, but the best spawning results occurred at 12 to 14ºC. In a related geoduck species, P. zelandica, this trigger temperature is between 15 and 17ºC (Gribben et al., 2004). The relatively low percentage of spawners at 19ºC in Phase II was likely due to gonad degeneration. By day 113 oocytes were virtually absent at 19ºC and gonads predominantly consisted of connective tissue. This is the same spent state typical of midsummer (July and August) at the end of the reproductive season for P. generosa  53  (Andersen, 1971; Goodwin, 1976; Sloan and Robinson, 1984) when temperatures peak. The stresses of prolonged high temperatures can inhibit gametogenesis, as has been seen in the cockle Laevicardium elatum (Villalejo-Fuerte et al., 1996) and most temperate bivalve species need declining temperatures to cue a new gametogenic cycle (Newell and Bayne, 1980; Gabbott, 1983; Beninger and Lucas, 1984). This temperature was not, however, high enough to significantly decrease survival and mortality results were consistent with those of natural populations of P. generosa that can seasonally experience temperatures up to 21–22ºC in the low intertidal and shallow subtidal without harm to the animals (Goodwin and Pease, 1989). The ability to acclimate to a wide range of temperatures is typical of temperate marine bivalves (Mann, 1979a; Bayne and Newell, 1983; Navarro et al., 2000; Chávez-Villalba et al., 2003b). Condition index (both CIw and CId) decreased in all temperature treatments over the duration of the experiment. Condition index typically increases to the peak of gonad maturity and decreases after spawning – as described in P. staminea (Feder et al., 1979), C. gigas , T. philippinarum (Robert et al., 1993), and Argopecten purpuratus (Navarro et al., 1989) – which may explain the CI decline from day 0 to 113. The gradual decline of CI between days 113 and 155, with the lack of spawning in 7 and 19ºC, implies that tissue reserves were being consumed (MacDonald et al., 1998; Navarro et al., 2000) and the health of the animals was declining (Feder et al., 1979). Similarly, hatchery held oysters O. edulis showed declining CIs compared to wild stocks, presumably due to less food and higher temperatures in the hatchery (Gabbott and Walker, 1971). Ash-free dry weight of the visceral mass was marginally lower at 7˚C than 19˚C. Differences between temperatures may have been expected since soft tissue ash content  54  in Cerastoderma edule fluctuated seasonally (and therefore with temperature), peaking in spring (25%) and declining in winter (15%) (Navarro et al., 1989). Increased ash content correlates to spawn events as seen in T. philippinarum (Robert et al., 1993) but the lower organic content at 7˚C (which had low spawn rates) contrasts this pattern. The present study, however, unlike the studies of Navarro et al. (1989) and Robert et al. (1993), examined organic content in just the visceral mass and not total soft tissue. Since the visceral mass samples included various proportions of gonad and digestive material, residual algae in the digestive tissue may have influenced the results. The organic content of bivalve gonads in particular has not been thoroughly studied. In another marine invertebrate, the stinging hydromedusa Olindias sambaquiensis (Limnomedusae, Olindiidae), it was reported that percent organics (and hence AFDW) of gonads did not change with gonad size or gonad developmental stage (Chiaverano et al., 2004). This indicates that organic content (and AFDW) could be relatively stable in invertebrate gonads, regardless of changes in reproductive condition (as is suggested in the current study). Despite the low magnitude of differences in percent organics, proportions of various biochemical components in bivalve gonads – including glycogen, lipid, and free glucose – may vary with season (Gabbott, 1983; Beninger and Lucas, 1984; Okumus and Stirling, 1998; Kang et al., 2000, 2007; Ngo et al., 2006; Dridi et al., 2007). Proportions of biochemical components were not explored in the present study, but should be the subject of future research. One potential source of variability in the present study is the possibility of difference in AFDW between males and females (Deslous-Paoli and Heral, 1988).  Unfortunately, we could not test this hypothesis as low sample sizes did not  55  allow for the sexes to be analyzed separately; however we encourage future work in this area. In conclusion, the results of the present study suggest that, in a hatchery setting, P. generosa should be held at a temperature of 11ºC if reproductive output is to be maximized. Lower temperatures (e.g. 7ºC) may be used to hold ripe broodstock for prolonged periods of time, although future research would need to examine the viability of gametes in clams that are held long-term. The results were in agreement with the alternative hypotheses (section 1.5.1) that temperature would significantly affect spawner percentage and reproductive development as indicated by gonadsomatic index, ocyte maturity, connective tissue occupation and gamete occupation.  56  Table 2.1. Results of ANOVAs on various attributes and ANCOVAs on condition index of experimental Panopea generosa. Values in bold are significant at P < 0.05. NS = not significant. Source  T W S (T)  df SS F ratio Percentage of spawning clams (Phase I = weeks 1–5) 3 1674.82 0.48* 4 32884.16 12.86 12 18502.48 2.41  T×W W × S (T)  12 48  C T D T×D Error  Condition index (wet tissue) 1 7.48 13.87 3 5.82 3.21 3 11.26 6.95 9 8.25 1.70 46 24.84  < 0.001 < 0.05 < 0.001 > 0.10 NS  Condition index (dry tissue) 1 0.51 11.23 3 0.23 1.71 3 0.67 4.89 9 0.68 1.66 46 2.09  < 0.005 > 0.15 NS < 0.005 > 0.10 NS  T D T×D Error  Gonadosomatic index (wet tissue) 3 1016.19 18.02 3 207.85 3.68 9 229.60 1.36 47 883.68  < 0.001 < 0.05 > 0.20 NS  Gonadosomatic index (dry tissue) 3 426.66 9.66 3 80.89 1.83 9 174.81 1.31 47 691.76  < 0.001 > 0.15 NS > 0.15 NS  T D T×D Error  Percentage with mature gonads 3 11404.69 6.89 3 6229.69 3.76 N/A N/A N/A 9 4964.06  < 0.05 > 0.05 NS N/A  Oocyte maturity 3 0.57 3 0.06 9 0.50 24 0.36  12.67 1.78 3.70  < 0.001 > 0.2 NS < 0.05  T D T×D Error  Oocytes per follicle 3 13.10 3 5.54 9 5.11 24 9.40  < 0.001 = 0.01 > 0.20 NS  Oocytes per area 3 28449.06 3 24362.04 9 7992.55 24 21377.53  10.65 9.12 1.00  < 0.001 < 0.001 > 0.2 NS  T D T×D Error  Connective tissue occupation 3 7323.00 23.56 3 1330.33 4.28 9 2107.38 2.26 47 4869.57  6041.11 30677.12  0.79  11.14 4.72 1.45  P value  > 0.5 NS < 0.001 < 0.05 > 0.5 NS  df SS F ratio Percentage of spawning clams (Phase II = weeks 15–17) 3 20077.65 6.25* 1 277.68 0.41 12 3940.64 0.49 3 12  1598.93 8047.77  0.79  P value  < 0.05 > 0.5 NS > 0.5 NS > 0.5 NS  < 0.001 < 0.01 < 0.05  Sources of variation are temperature (T, fixed factor), weeks of conditioning (W, random factor only applicable to percentage of spawning clams), days of conditioning (D, fixed factor), tank (S, subject factor; only applicable to percentage of spawning clams), shell weight (C, covariate; only applicable to condition index), interactions, and error. One tank from 19ºC on final sampling day had no clams. * = approximate F-test. N/A = not applicable (no interactions, un-replicated 2-way ANOVA)  57  mix of heated and chilled seawater header tank  overflow outlet  conditioning tube tube outflow  water-bath  stand-pipe  geoduck  flow to waste  Figure 2.1. Schematic showing the Panopea generosa conditioning tube system. Dashed lines depict water levels in the respective containers. Arrows depict water flow. Actual configuration had eight tubes per water-bath tank. Seven tubes contained one clam each and one tube remained empty for water quality testing. There were four waterbath tanks for each temperature treatment. Not to scale. See text for dimensions.  58  Phase I  Phase II  100 B  90  B  Spawner percentage  80 Temperature 7˚C  70  11˚C  60  15˚C A  19˚C  50 40 30 20 10 0  A  a  1  b  b  a  b  2  3  4  5  9  A  15  A  A A  17  Time (weeks of conditioning)  Figure 2.2. Mean percentage of Panopea generosa individuals spawning at various temperatures (7, 11, 15, 19°C) over time (weeks of conditioning). Error bars indicate SEM (n=16 in Phase I, n=4 in Phase II). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters in Phase I (below bars) apply to differences among weeks and upper-case letters in Phase II (above bars) apply to differences among temperatures. Temperature was not a significant factor in Phase I and time was not a significant factor in Phase II (ANOVA).  59  a. A 7ºC  600  AB 11ºC  AB 15ºC  B 19ºC  Condition index (wet) (%)  500  400  300  200  100  0 a  b  ab  b  b. 140 A 7ºC  Condition index (dry) (%)  120  A 11ºC  A 15ºC  A 19ºC  100 80 60 40 20 0  a  ab  ab  b  0  113  135  155  Time (days)  Figure 2.3. Mean (a) wet and (b) dry condition indices (CI) of Panopea generosa held at various temperatures (7, 11, 15, 19°C) over time (days). Error bars indicate SEM (n=4). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods and upper-case letters (above legends) apply to differences among temperatures.  60  a. 40  Gonadosomatic index (wet) (%)  A  35  7˚C  B  11˚C  B  15˚C  B  19˚C  30 25 20 15 10 5 0  ab  a  b  ab  b.  Gonadosomatic index (dry) (%)  35 30  A  7˚C  B  11˚C  B  15˚C  B  19˚C  25 20 15 10 5 0  a  a  a  a  0  113  135  155  Time (days)  Figure. 2.4. Mean (a) wet and (b) dry gonadosomatic indices (GSI) of Panopea generosa held at various temperatures (7, 11, 15, 19°C) over time (days). Error bars indicate SEM (n=4). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods and upper-case letters (above legends) apply to differences among temperatures.  61  (a) 7ºC  B 100%  Stage percentage  80% 60% 40% 20% 0%  (b) 11ºC  B  Stage percentage  100% 80% 60% 40% 20% 0%  (c) 15ºC  AB  100% Stage percentage  80% 60% 40% 20% 0%  (d) 19ºC  A  Stage percentage  100% 80% 60% 40% 20% 0% Harvest  Spent  Early active  0  113 Time (days)  Late active  Ripe  135  155  Partially spawned  Figure. 2.5. Histograms showing the percentage of various reproductive stages of Panopea generosa held at various temperatures [(a) 7°C, (b) 11°C, (c) 15°C, (d) 19°C] over time (days), as determined by examination of histological sections. Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Upper-case letters (next to temperatures) apply to differences among temperatures.  62  a.  90 A  Mature oocytes (%)  80  7ºC  11ºC  15ºC  19ºC  70 A  60  AB AB  50 40 30  AB  20  B  10 0  B  ab  a  a  a  a  a  a  a  ab  a  a  B  a  b  a  a  a  b. 25 A  Oocytes per follicle  20  B  A  7ºC  11ºC  B  15ºC  19ºC  15 10 5  c.  Connective tissue occupation (%)  0  a  ab  b  7ºC  100  ab  11ºC  15ºC  19ºC  C  90  C BC  BC  80  AB  B  70 60  AB  50 AB  40 A  30  A  A  A  20 10 0  a  a  a  0  a  a  a  a  b  a  113  a  135  a  b  a  a  a  b  155  Time (days)  Figure. 2.6. Mean (a) percentage of mature oocytes, (b) number of oocytes follicle -1, and (c) percentage of connective tissue occupation in the gonads of Panopea generosa held at various temperatures (7, 11, 15, 19°C) over time (days). Mature oocytes are 30.1–40.0 μm in diameter. Error bars indicate SEM (n’s for 6a,b shown in 6a, n=4 in 6c). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods and upper-case letters (above legends or bars) apply to differences among temperatures.  63  3 Effect of salinity on survival and gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850)  To investigate low salinity limits for survival and reproduction in Pacific geoduck clams, adult Panopea generosa were step-wise acclimated to salinities of 17, 20, 24, and 29 at a temperature of 12.5˚C. The clams were fed Isochrysis sp. (Tahitian strain: TISO clone) and Chaetoceros muelleri (50:50 by cell count) at a ration of 4 x 109 cells animal-1 d-1 for 62 d, with animals being sampled across time to assess survival, gonad development, and various condition indices (i.e. condition index, gonadosomatic index, gametogenic development stage, oocyte diameter, gamete occupation index, connective tissue occupation index, and gonad thickness). By day 26 of the 62-d experiment 90% of the clams at salinity 17 and 75% at salinity 20 had contracted fungal infections and died. No traces of fungus were seen in clams held at salinities of 24 and 29 with only 5% mortality in these treatments during the same period. At the start of the experiment the clams had no gamete development, but by day 62 the clams at a salinity of 29 were maturing, with gametes occupying an average of 29.9% of the gonad tissue area, while maturation was inhibited at a salinity of 24 with only 4.9% gamete occupation. The wet weight gonadosomatic index was significantly higher in clams held at a salinity of 29 (mean: 14%) than in those held at a salinity of 24 (mean: 11%). At a salinity of 29 the gonad sheath thickness of the visceral mass increased (from 2.8 mm at day 0 to 8.8 mm at day 62), but did not significantly increase at a salinity of 24. Geoduck clams cannot withstand prolonged periods at low salinities (≤20) and broodstock conditioning of this species for gamete production is best done at near full-strength seawater (salinities of at least 29). 64  3.1 Introduction For both logistical and economic reasons there is considerable interest in BC in expanding existing intertidal bivalve farms and hatcheries to include geoduck clams as well as developing new sites for geoduck culture. Many existing shellfish culture operations are, however, located in estuarine areas where winter and summer freshets can range in salinity from 0.5 to 17 and can be between 1 and 10 m in depth (Thompson, 1981). Euryhaline commercial species commonly cultured in these waters (Manila clams, Venerupis philippinarum, and Pacific oysters, Crassostrea gigas) are tolerant to such shifts in salinity but P. generosa, unlike these intertidal bivalves, typically populates physically-stable subtidal environments (Sloan and Robinson, 1984) that are not subject to rapid or long-term salinity changes. How adult P. generosa react to salinity changes has not been studied, but this species has poor success rates of embryogenesis (Goodwin, 1973) and an inhibited juvenile burrowing response (Davis and Barenberg, 2000b) at salinities below 25. Intertidal bivalve species typically respond to acute salinity stress by sealing off soft tissues with water-tight shells (Shumway, 1977; Elston et al., 2003), a response that is impossible in geoducks due to the relatively small shells that leave much of the mantle and the entire siphon exposed. Under chronic exposure to salinity stress, bivalves regulate cell volume (Van Winkle, 1968; Deaton, 1994; Deaton and Pierce, 1994; Glémet and Ballantyne, 1995; Neufeld and Wright, 1996, 1998; Deaton, 2001) through the gain or loss of intra- and extra-cellular organic and inorganic solutes (Shumway, 1977; Shumway and Youngson, 1979; Somero and Bowlus, 1983). There are possible  65  metabolic costs associated with osmoregulation (Stickle and Sabourin, 1979; Kim et al., 2001; Hamer et al., 2008) and potential long-term outcomes to low salinity exposure are reduced growth rates (Bataller et al., 1999; Shurova, 2001; Jansen et al., 2009), inhibited reproduction (Butler, 1949; Watts and Lawrence, 1990) (Baba et al., 1999), increased mortality (Elston et al., 2003; Tomiyama et al., 2008) and structural abnormalities such as altered mineral content of shells (Strasser et al., 2008). The objective of this study was to investigate the responses of adult P. generosa to prolonged hyposalinity (replicating seasonal lows in the Strait of Georgia, BC) with an emphasis on reproductive development. The parameters investigated were: mortality rate, condition index, gonadosomatic index, gametogenic development stage, oocyte diameter, gamete occupation index, connective tissue occupation index, and gonad thickness. Effects of salinity on oxygen consumption and clam clearance rates were also investigated. The ultimate goal of the study was to identify salinity levels that are suitable for P. generosa broodstock conditioning and field-based grow out.  3.2 Materials and methods 3.2.1 Algal culture Clams were fed live algal cultures of Isochrysis sp. (Tahitian strain, TISO clone, CCMP 1324) and Chaetoceros muelleri (CCMP 1316) as per Chapter 2 (section 2.2.1).  66  3.2.2 Broodstock collection and initial maintenance Broodstock were collected off of the western shore of Thormanby Island, BC (49˚ 30.6' N latitude, 124˚ 1.7' W longitude) on August 14, 2008. The clams were collected at the same location and on the same date to minimize potential variability in reproductive condition due to season, location (Chávez-Villalba et al., 2002, 2003a), and genetics (Barber et al., 1991; VanKoeveringe, 1998). Clams were collected at approximately 7– 12 m depth and 12–13˚C from a cobble/mud substratum as part of a commercial harvest. The clams were transferred to the Pacific Biological Station (Nanaimo, BC) inside of temperature-controlled (14–16˚C) plastic totes. After delivery, the clams were held in seawater tables at 13 ± 0.5˚C (mean ± SD, n = 720) for 5 days until they were put into experimental tubes. The clams had an average shell-length (SL) of 130.7 ± 7.4 mm (mean ± SD, n = 80) and an average wet weight of 987.3 ± 148.6 g (mean ± SD, n = 80). Clams greater than 110 mm SL were used because geoducks smaller than this are biased toward males or are sexually immature (Andersen, 1971; Sloan and Robinson, 1984; Campbell and Ming, 2003). Gonads of eight randomly-selected clams were sampled immediately after harvest to determine initial reproductive state (as per Section 2..2.8). All sampled clams were in a resorbed state, indicating recent spawning.  3.2.3 Broodstock conditioning: experimental set up Seawater salinities and temperatures were maintained by mixing heated or chilled water (fresh and salt) in mixing chambers that flowed into header tanks. Each salinity treatment had four replicate tanks randomly distributed within the laboratory. Daily monitoring and adjustments maintained relatively consistent temperatures across replicate  67  tanks and time (78 d) (mean ± SD: 12.4 ± 0.6˚C, n = 2976). The experimental temperature chosen is one at which gonads ripen in the natural environment (Sloan and Robinson, 1984) and at which there is successful spawning under laboratory conditions (Goodwin, 1976; R. Marshall, pers. obs.). The salinity levels tested were (mean ± SD) 17.0 ± 0.1 (n = 108), 20.0 ± 0.1 (n = 108), 24.0 ± 0.1 (n = 252), and 29.0 ± 0.2 (n = 252), which are typical of the salinity range within the Strait of Georgia, BC (Thompson, 1981). Preliminary tests indicated that salinities below 17 resulted in severe swelling and death within 3–4 d. Salinities were adjusted to final levels in step-wise increments of 2 every second day. The clams were held for 17 d at the final salinities before the start of the experiment (day 0). Salinity was measured with a YSI 30 salinity meter (YSI, Yellow Springs, Ohio, USA). To prevent possible shell gaping and tearing of flesh, PVC-coated wire mesh (mesh size: 25 x 25 mm; wire diameter: 3 mm) was cut into strips (length x width: 34 x 8 cm), fashioned into a U-shape, and moulded around the shell of each clam. The clams were set vertically inside of PVC tubes with one clam per tube and five tubes per tank. These tubes were 55 cm high and 12.5 cm in diameter, capped at the bottom, and equipped with an outflow 2.5 cm from the top of the tube (giving a functional volume of 6,443 ml). Tubes were placed vertically into plexiglass tanks (length x width x height: 58 x 58 x 30 cm; volume: 101 l) with four replicate tanks per salinity. The size of the clams was not significantly different among the tanks (ANOVA: F15,64 = 0.71, P > 0.50). Water was fed to each tube from a plastic header tank (length x width x height: 40.6 x 29.8 x 64.1 cm; volume 47.3 l) via clear vinyl siphon tubes (inner diameter: 7.5 mm). Water entered the geoduck tubes approximately 10 cm from the bottom and flowed up past the  68  geoduck and out through the outflow. Seawater flow rates within the tubes ranged from 48.6 to 66.0 l h-1, a level set to exceed the average clearance rate of the clams (13.2 l h -1) which was determined through a preliminary study (as per section 2.8). Clams were fed 4 x 109 cells animal-1 d-1 – or ~ 0.1% dry weight algae : dry weight geoduck tissues – of a 50:50 (by cell count) combination of C. muelleri and TISO. Clams were not able to consume higher levels of algae on a daily basis, as determined by preliminary trials. The effluent from each tube flowed to waste except during feeding when the water was recirculated. Photoperiod during the acclimation and experiment was 16 h light per day, to simulate natural late-spring spawning conditions.  3.2.4 Spawning events and mortality rates Observations were made 3–4 times daily, between 8 am and 10 pm, for signs of spawning activity (i.e. presence or absence of eggs or sperm in the water). The health of each animal was assessed daily by examining for lesions in the periostracum, flaccid or distended siphons, cracks in the shell, excessive mucus, and slow or absent reactions to a physical stimulus. If responses were abnormally slow or absent the clam was classified as moribund or dead and removed. Since each clam was in its own experimental tube, removal of a dead clam from the experiment had no effect on the food ration received by other clams.  69  3.2.5 Clam sampling Sampling was done at days 0, 41, and 62. On each sampling day one randomlyselected clam was taken from each tank, totalling four individuals per salinity. The clams were dissected and drained of internal water. Visceral mass [i.e. alimentary canal (stomach excluded), digestive diverticulum, sheathing gonad tissue (Sloan and Robinson, 1984), and attached foot (Gribben et al., 2004)] wet weight was recorded to the nearest 0.01 g. Wet somatic tissue (i.e. all tissue other than the visceral mass) and shells were weighed to the nearest 0.1 g. Soft tissues (cut into approx. 1 × 1 cm pieces to facilitate proper drying) and shells were dried at 58–60˚C for 72 h (to constant weight). Dry and wet tissue weights were used to calculate condition indices (CI) and gonadosomatic indices (GSI), as outlined below.  3.2.6 Condition and gonadosomatic indices 3.2.6.1 Condition index A condition index (CI) based on the method of Walne and Mann (1975) was adapted for geoduck: CI = (T × 100)/S where CI is the condition index, T is the total softtissue mass, and S is the shell mass. This was calculated for both wet (CI w) and dry (CId) tissues. A declining CI over time would indicate a loss of soft tissue in relation to shell mass and an increase in CI w independent of CId would indicate increased water levels in the soft tissue. Increases in shell weight were un-likely to be a factor influencing CI as indicated by the allometric relationship of age to shell-weight for P. generosa from Thormanby Island (Bureau et al., 2003). The model from Bureau et al. (2003) estimated  70  a shell weight increase of <0.7% over the duration of the experiment. Percent water content of soft tissue was calculated as follows: %WC = (WW – DW)/WW × 100 where %WC is the percent water content, WW is the total wet tissue weight, and DW is the total dry tissue weight.  3.2.6.2 Gonadosomatic index Gonadosomatic indices (GSIs) were calculated to determine relative visceral mass (including gonad and digestive gland) sizes among salinity treatments. For details see Chapter 2 (Section 2.2.6). Calculation was based on the method of Sloan and Robinson (1984): GSI = V/T × 100 where GSI is the gonadosomatic index, V is the visceral mass weight, and T is the total soft-tissue weight. Calculations were done for both wet (GSI w) and dry (GSId) tissues. Since part of the visceral mass was removed for histological analysis before drying, a correction factor was necessary for the final visceral mass dry weight. This was done by multiplying the ratio of dry to wet visceral mass weight by the wet weight of the sub-sample and adding it to the dry weight of the visceral mass.  3.2.7 Gonad ash-free dry weight Gonad samples (approximately 300–400 mg) were analyzed for ash-free dry weight (AFWD) (see Chapter 2, section 2.2.7). Only gonad tissue, and not digestive diverticula, stomach or foot, was included in the sample.  71  3.2.8 Histological sampling A section, approximately 1–2 cm3, was removed from the same area of the right side (posterior end) of the visceral mass in each sampled clam [as in Sloan and Robinson (1984)]. This sample was then weighed, fixed in Davidson’s solution for 72 h, and subsequently transferred to 70% isopropanol for storage. Within 30 d of sampling, the preserved tissue samples were cut into 4–5 sub-sections, dehydrated using a graded ethanol series, and embedded in paraffin wax. Embedded tissue samples were sectioned to 5 μm, stained with hematoxylin-eosin, and mounted on slides for examination. Slides were analyzed using a Motic B5 compound microscope with Motic Images Advanced 3.2 software (Motic Electric Group Co., Ltd., Richmond, BC, Canada).  3.2.8.1 Development classification The tissue samples were classified as early active, late active, ripe, partially spent, or spent/resorbed, based on the classifications used for P. generosa (Andersen, 1971; Goodwin, 1976; Sloan and Robinson, 1984) and P. zelandica (Gribben et al., 2004). The developmental stage was determined by the most dominant stage (highest proportion) of at least 10 randomly selected follicles in each gonad sample (Gribben et al., 2004). For statistical analysis, they were grouped into two main categories: sexually mature (late active, ripe, partially spent) or sexually immature (spent/resorbed, early active).  72  3.2.8.2 Oocyte diameter Oocyte diameters (OD) of a minimum of 15 randomly-selected oocytes per clam were measured to the nearest 0.2 μm with a mean being calculated for each sampled clam. Only oocytes with visible nuclei were measured (Hesselman et al., 1989; VillalejoFuerte et al., 1996; Gribben et al., 2004).  3.2.8.3 Connective tissue and gamete occupation indices Occupation indices similar to the “volume fractions” of Lowe et al. (1982) were employed. For connective tissue occupation index (COI) and gamete occupation index (GOI) the surface areas of connective tissue and gametes were colourimetrically determined [as in Delgado and Pérez-Camacho (2003)] using Motic Images Advanced 3.2 software. Images were analyzed in 8-bit gray-scale (0 = black, 255 = white). Within the colour scale, spermatic material and oocytes ranged from 33 to 109 whereas connective tissue ranged from 113 to 233. Empty follicles and vacuoles comprised the remainder of the area. The COI was calculated as: COI = connective tissue surface area / area of the field analyzed. The GOI was calculated as: GOI = gamete surface area / area of the field analyzed. Ten randomly selected fields (2.5–3.5 mm2 each) were analyzed from each sub-sampled gonad (40–50 readings per clam) with a mean being calculated for each sampled clam. Fields excluded digestive tubule area. Since the reproductive cycles of males and females are very similar in P. generosa (Andersen, 1971; Goodwin, 1976), P. zelandica (Gribben et al., 2004), and Panopea globosa (Aragón-Noriega et al., 2007) the sexes were pooled for statistical analysis.  73  3.2.8.4 Gonad thickness Panopea generosa gonads, although an integral part of the visceral mass, sheath the outside of the digestive gland. When tissue samples were collected, it was noted that some gonads had a thicker sheath of gonad material around the digestive gland than others. To determine if the variability in the gonad thickness (GT) was related to salinity, the histological samples were measured from the edge of the epithelium to the edge of the digestive gland, the area in between consisting of gonad material. Five random measurements (to the nearest 0.25 mm) were taken from each clam’s visceral mass subsample (20–30 measurements per clam) using a dissecting microscope and stage micrometer. The mean values from each clam were used for statistical analysis.  3.2.9 Oxygen consumption and clearance rates On the last day of the acclimation phase and one day before the start of the experiment (0 d), oxygen consumption was measured. The clams were starved for three days to allow the gut to clear and avoid elevated metabolic rates associated with the specific dynamic action (the energy expended incidental to the ingestion, digestion, absorption, and assimilation of a meal) (Secor, 2009). Twelve clams were measured for each salinity treatment. Clams were within a limited size range (soft-tissue dry weight averaged 77 ± 16 g (mean ± SD, n = 48) to reduce variability associated with the allometric relationship between oxygen uptake and body mass. The clams were also distributed such that there were no differences in masses among treatments (ANOVA F3,44 = 1.05, P > 0.25). Oxygen consumption was determined using a flow-through system which consisted of the holding tube as the chamber and oxygen sensors fitted to  74  the inflow tube at the header tank and the outflow on the holding tube. Preliminary trials paired with a closed containment respirometer showed that results did not significantly differ between the two systems, but the geoducks themselves in the closed system tended to extend and damage their siphons as oxygen levels dropped in the chamber. The flowthrough system was favored for this reason. To prevent oxygen transfer from the air, the tube was capped with no air space between the cap and the water surface. The inflow within the tube was situated at the bottom and flowed parallel to the bottom and the side. The water flowed upward past the clam and eventually exited through the outflow at the top. Initial dye tests in a transparent tube indicated that this configuration minimized stratification. Inflow and outflow tubes were fitted with a vinyl tubing adapter containing a YSI 550A dissolved oxygen sensor. The meters had internal corrections for solubility of oxygen based on Clesceri et al.(eds.) (1998). For salinities of 17, 20, 24 and 29 these corrections were 0.899, 0.884, 0.863 and 0.836 respectively (at 12.5ºC and 760 mmHg). The water flowed past the sensors at >16 cm s -1, minimizing the effects of oxygen consumption by the electrodes. Clam free control trials indicated that there was no significant reduction in O2 saturation from the inflow to the outflow (paired t-test, t = 1.21, df = 119, P > 0.1). The signal to noise ratio (mean measurement/SD) of control trials was 210 indicating very little noise in the measurements. Readings from each clam were averaged over a 15 min trial. To ensure accurate readings in a flow-through respirometer, dilution properties must be considered (Steffensen, 1989). The water exchange rate was 95% of the water volume in 8.7 min and measurements were started a minimum of 15 min after the clams began to filter steadily. This provided sufficient time for the respirometer chamber to reach steady state justifying the use of the following  75  equation (based in the Fick principle) to calculate oxygen consumption (Steffensen, 1989): MO2 = Vw (Cwo2, in – Cwo2, out) / bw Mo2: total oxygen consumption per unit of body mass (mg O2 g h-1) Cwo2, in: concentration of oxygen in water flowing in (mg O 2 l-1) Cwo2, out: concentration of oxygen in water flowing out (mg O 2 l-1) Vw: water flow rate (l h-1) bw: is body weight (dry weight in g)  Clearance rates (CR) were determined on days 2, 38, and 58. On those days the clams were held in the experimental tubes under static conditions (aerated with no light) and fed 2.5 x 105 cells ml-1 of TISO. Water samples of 20 ml were taken from each replicate tube at the beginning and end of a 2-h feeding. Clearance rate was calculated using the method of Quayle (1948) [as described in Coughlan (1969)]: M = m / (n × t) × loge(C0 / Ct), where M = clearance rate (ml of water cleared of cells clam-1 h-1), m = volume of suspension (ml), n = number of animals, t = duration of trial (h), C 0 = initial cell concentration (cells ml-1), and Ct = final cell concentration (cells ml-1).  3.2.10 Statistical analyses High mortality rates in salinities 17 and 20 lead to the termination of these treatments before the second set of samples were taken on day 41. Results were therefore analyzed on day 0 using a one-way ANOVA (with the fixed factor salinity) with the full range of salinity treatments while a two-way ANOVA (with the fixed factors salinity and time) was used for 24 and 29 salinity treatments incorporating all three sampling days.  76  The parameters tested with one-way and two-way ANOVAs were: CIw, CId, GSIw, GSId, AFDW, OD, GOI, COI, GT, %WC, CR and mortality rate. Analyses were always performed on the mean data from each replicate clam in cases where more than one measurement was taken from each individual. There were no oocytes present at the first sampling point, so no 0-d analysis was possible for OD. Percent water content was analyzed only at day 0 using a one-way ANOVA. Cumulative mortality rates were analyzed using a two-way ANOVA examining two phases: 1) the adjustment phase from the time salinity changes began up to day 0 and 2) the experimental phase up to the termination of 17 and 20 salinity treatments (day 26). Development stage was analyzed with a chi-square test for independence. For oxygen consumption, data inspections indicated consumption rates per gram scaled with body mass. To control this concomitant variable, ANCOVA was used with total dry soft-tissue weight as the covariate and salinity as a fixed factor. Clearance rates were analyzed with repeated measures ANOVA. Assumptions of normality and equal variances were tested with Shapiro-Wilk W and modified Levene’s tests, respectively. Departures from these assumptions were corrected with natural-log transformation for mortality, CI w, and CId. The arcsine transformation was used for GOI and COI, while no transformations were necessary for GSIw, GSId, OD, GT, %WC and CR. ANCOVA assumptions for oxygen consumption were examined with plots of residuals (versus fitted values, soft-tissue dry weight and salinity) and the normal probability plot of residuals; no departures were found. Significant differences among factors were tested using a post-hoc multiple comparison Tukey-Kramer test (α = 0.05).  77  3.3 Results 3.3.1 Spawning events and mortality rates Seven days after harvest, 17 (i.e. 4 females, 13 males) of 80 clams had spawned. All spawns were very weak with each female producing only a few thousand eggs. Gametes were viable, however, with a fertilization rate of 97.5 ± 2.4% (mean ± SD, n = 13). There were no more spawning events for the duration of the experiment. As the adjustment phase progressed and the final salinities were reached, clams at salinities of 17 and 20 became visibly swollen with increasingly flaccid siphons. At day 26 of the experiment the 17 and 20 salinity treatments were terminated due to rapidly spreading fungal infections (which began on day 22) and mass mortalities. Figure 3.1 shows the cumulative percent mortality during the adjustment phase (16-d duration) and experimental phase up to the termination of the 17 and 20 salinity treatments (26-d duration). A two-way ANOVA indicated that both time and salinity significantly affected percent mortality with a significant interaction between the two factors (F1,24 = 7.05, P = <0.02; F3,24 = 24.64, P = <0.0001; F3,24 = 4.00, P = <0.02, respectively). During the adjustment phase, cumulative percent mortality in salinity 17 was significantly higher than in salinities 20, 24, and 29 with no significant pair-wise differences among the latter three treatments (Fig. 3.1). During the experimental phase, cumulative percent mortalities at salinities 17 and 20 were not significantly different, but both were significantly higher than at 24 and 29 (Fig. 3.1). Cumulative percent mortality increased significantly from the adjustment to the experimental phase in the 17 and 20 salinity treatments, but there was no significant increase in the 24 and 29 treatments (Fig. 3.1). Histological examination showed that fungal hyphae (unidentified species)  78  penetrated the mantle and muscle tissue of all dead clams at salinities 17 and 20. There were no traces of fungus in clams at salinities 24 or 29.  3.3.2 Condition and gonadosomatic indices 3.3.2.1 Condition index On day 0, CIw in the 17 and 29 treatments differed significantly with no other significant pair-wise comparisons among salinity treatments (Table 3.1, Fig. 3.2a). These significant differences were due to higher water content in the lower salinity treatments (17 and 20) as indicated by %WC on day 0 (Table 3.1); clams held at salinities of 17 and 20 had significantly higher %WC [84.8 ± 1.0% and 83.0 ± 1.0% (mean ± SE, n = 4), respectively] than those held at 24 and 29 [77.8 ± 0.9% and 78.4 ± 1.1% (mean ± SE, n = 4), respectively] (Tukey-Kramer test). Drying eliminated significant differences among salinities in CI on day 0 (Table 3.1, Fig. 3.2b). Two-way ANOVAs (salinities of 24 and 29 only) showed that there was no significant effect of salinity, time, or the interaction of these two factors on CIw, CId, or %WC (Table 3.1, Fig. 3.2a,b). The overall average of %WC was 79.4 ± 0.52% (mean ± SE, n = 24). Salinity did not significantly influence GSI w or GSId on day 0 (Table 3.1, Fig. 3.3a,b). Over the full duration of the experiment (which included salinities 24 and 29 only), salinity significantly affected GSI w, while time and the interaction between salinity and time did not (Table 3.1): GSIw was significantly lower at 24 than at 29 (Fig. 3.3a). The GSId was not significantly affected by salinity, time, or the interaction of these two factors (Table 3.1, Fig. 3.3b).  79  3.3.3 Gonad ash-free dry weight The percentage of organic material (%AFDW) on the first day of the experiment was not significantly different among salinity treatments (F3,12 = 0.44, P > 0.5) and averaged 88.0 ± 0.6% (mean ± SE, n = 16). Organic content at 29 averaged 86.5 ± 1.39% (n = 8) and at 24 averaged 89.4 ± 1.39% (n = 8) over the duration of the experiment, with no significant differences due to salinity (F1,18 = 1.39, P > 0.25), time (F2,18 = 1.04, P > 0.25), or the interaction (F2,18 = 0.35, P > 0.50).  3.3.4 Gonad histology 3.3.4.1 Development classification At the time of collection, all of the clams sampled (n = 8) had gonads in spent or resorbed stages. After salinity adjustment (day 0 of the experiment), all sampled clams were in the resorbed stage. By day 41, no clams were ripe, but 37.5% had entered the early active stage with no difference among salinities (ϰ22 = 0.2, P > 0.5). On day 62, 75% of clams at salinity 29 were in advanced development stages (nearly or fully mature in late active or ripe stages, respectively), but none of the clams at salinity 24 had advanced beyond the resorbed or early active stages (ϰ22 = 8.0, P < 0.02) (Fig. 3.4).  3.3.4.2 Oocyte diameter No oocytes were present on day 0, but they began generating by day 41 (averaging 8.1 μm in diameter) and grew significantly to 16.5 μm in diameter by day 62  80  (Fig. 3.5a). Oocyte diameter was significantly affected by time, but not by salinity or the interaction between these two factors (Table 3.1).  3.3.4.3 Gamete occupation index Gamete occupation index was very low on day 0, averaging 0.23% with no significant differences among salinity treatments (Table 3.1, Fig. 3.5b). A two-way ANOVA (salinities of 24 and 29 only) revealed that GOI was significantly affected by time and the interaction between time and salinity, but not by salinity alone (Table 3.1). There were no significant differences between salinities 24 and 29 on days 0 or 41, but on day 62 GOI at 29 was significantly higher than at 24 (Fig. 3.5b). The GOI in salinity 24 remained unchanged during the experiment, but there was a significant increase in GOI in salinity 29 on day 62 compared to days 0 and 41 (Fig. 3.5b).  3.3.4.4 Connective tissue occupation index On day 0 there was no significant difference in COI among salinity treatments (Table 3.1, Fig. 3.5c). Over the duration of the experiment COI decreased significantly from day 41 to day 62, but there was no significant difference between the salinity treatments and no significant interaction (Table 3.1, Fig. 3.5c). Decreases in COI in salinity 29 coincided with increases in GOI, but in salinity 24 the COI decrease coincided with increases in vacuoles and empty follicles associated with degenerating gonads.  81  3.3.4.5 Gonad thickness Gonad thickness on day 0 averaged 3.9 mm among all salinity treatments with no significant salinity effect (Table 3.1, Fig. 3.5d). A two-way ANOVA (salinities of 24 and 29 only) revealed that both main effects and the interaction term significantly affected GT (Table 3.1). On days 0 and 41 the 24 and 29 salinity treatments were not significantly different, but by day 62 GT was significantly higher in clams held at a salinity of 29 than in those held at 24 (Fig. 3.5d). Gonad thickness of clams in the salinity 29 treatment increased significantly from 2.8 mm on day 0 to 8.2 mm on day 41 and remained statistically unchanged on day 62 (8.8 mm). There was no significant change among sample dates at a salinity of 24 (Fig. 3.5d).  3.3.5 Oxygen consumption and clearance rates At the end of the acclimation phase, ANCOVA results indicated that salinity was a significant factor (F3,43 = 6.65, P < 0.001) in O2 consumption. The covariate (total softtissue dry weight) was also significant (F1,43 = 10.42, P < 0.005), indicating that oxygen consumption scaled significantly with body size. Significantly lower oxygen consumption was found in clams at a salinity of 29 compared to 17, 20, and 24 (Fig. 3.6). There were no significant differences in the weights of the clams among treatments so the differences in oxygen consumption can be attributed to salinity and not animal size. It was observed, but not quantified during the trials, that clams held at salinities of 17 and 20 had erratic pumping behaviour, characterized by frequent starts and stops and closure of siphon tips.  82  On day 2, salinity significantly affected CR (Table 3.1, Fig. 3.7). Clams held at salinities of 17 and 20 had negligible CRs (likely a result of frequent closures of the inhalant and exhalant siphons during the trials), significantly lower than those of clams in salinities of 24 and 29 (Fig. 3.7). A two-way ANOVA revealed that salinity did not significantly affect CR over the three sampling dates, but CR did vary significantly with time with no significant interaction between the two factors (Table 3.1). Clearance rate was significantly higher at day 38 than at days 2 and 58 (Fig. 3.7).  3.4 Discussion Panopea generosa were negatively impacted by chronic exposure to reduced salinities. This was evident in the high mortality rates at salinities of 17 and 20 within 26 d of completing the adjustment phase. The hyposalinity tolerance of P. generosa appears to be relatively low compared to intertidal clams such as Mya arenaria that can survive at a salinity of 6 for at least 4 wk (Shumway, 1977) and V. philippinarum that can tolerate a salinity of 8 for up to 8 d (Elston et al., 2003). The New Zealand little neck clam, Austrovenus stutchburyi, and the pipi, Paphies australis, can tolerate sustained exposure (>30 d) at a salinity of <10 before a significant decrease in survivorship is experienced (McLeod and Wing, 2008). In contrast to intertidal clams, subtidal scallops have lower tolerance to hyposaline conditions. For example, step-wise acclimated (i.e. salinity change of 3 every 3 d) lion’s paw scallops, Nodipecten nodosus, died rapidly below a salinity of 22 (Roldán-Carrillo et al., 2005) and had a LC50 (48 h) at a salinity of approximately 23 (Rupp and Parsons, 2004). Also, the Japanese scallop, Patinopecten yessoensis, died within 24 h at a salinity of 18 (Izumi et al., 2000). The response of P.  83  generosa to hyposaline conditions appears to be intermediate between intertidal clams and deep-water scallops, having died rapidly below 17 and tolerating salinities of 17 to 20 for more than 43 d. This characteristic might be expected since the habitat range of P. generosa spans from physically-stable subtidal depths to very low intertidal levels (<1 m above low tide) (Goodwin and Pease, 1989) where salinity fluctuations are more likely. The thick periostracum of geoducks that covers the siphons and muscular mantle, combined with the ability to close off the inhalant and exhalant siphons (as seen in CR trials at salinities of 17 and 20), likely helps to temporarily isolate internal structures from salinity changes. Siphon-tip closure under low salinity challenge was also observed in the sand gaper clam, Scrobicularia plana (Akberali and Davenport, 1981). Mortalities in the present study were associated with a fungal infection. Although the fungus was not identified, all of the clams that died had similar fungal hyphae penetrating the periostracum several millimetres into the muscle tissue of the mantle and/or siphons. Fungal infection was probably prevalent because cellular swelling (as indicated by a higher percentage of soft-tissue water content at salinities of 17 and 20) structurally compromised the periostracum. Also, the lowered salinity may have been a more favourable environment for fungi which are not typically abundant in the marine biota (Don, 1957; Gareth-Jones and Jennings, 1964). To compound the problem, marine invertebrates under salinity stress can have immune suppression (Li et al., 2010). In bivalves, salinity extremes on either end of the spectrum can impact immune response. For example, V. philippinarum has reduced numbers of haemocytes at a salinity of 20 (Reid et al., 2003) and the clam Chamelea gallina has reduced haemocyte functionality at a salinity of 40 (Monari et al., 2007). Clearance rates at salinities of 17 and 20 were also  84  very low, indicating that nutritional stress was also a likely factor in elevated mortalities at these lower salinities. The inability to survive salinities below 24 is very important to P. generosa culture site selection. In certain areas within the Strait of Georgia, salinity can drop to as low as 0.5 during a freshet (Thompson, 1981) while others are in estuaries that have consistently low salinity. Any farm sites within these locales could be severely impacted. Major shifts in wild populations of bivalves due to low-salinity events have already been recorded. For example, populations of wild Mediterranean mussels, Mytilus galloprovincialis, had a 12% reduction in survival rates in areas with freshwater run off (Shurova, 2001). Freshwater discharge from a hydroelectric power station caused massive population shifts of A. stutchburyi, reducing densities to 0.7 clams m-2 compared to 30 clams m-2 in unaffected areas (McLeod and Wing, 2008). Salinity can also be a major factor in scallop population distribution (Roldán-Carrillo et al., 2005; Rupp et al., 2005). Therefore, selecting farm sites within an unfavourable salinity range is likely to yield poor results. The salinities of 24 and 29 were similar with respect to OD. Oocytes were growing at the same rate in clams held at salinities of 24 and 29, a rate similar to P. zelandica under natural conditions where oocytes increased in diameter from ~10 to ~20 μm within one month during the development season (Gribben et al., 2004). There were, however, signs of reproductive inhibition at salinity 24, based on development classification, GOI, COI, and GT. The animals were spent at the time of collection (August) – as would be expected as they were harvested at the end of the natural spawning season (Andersen, 1971; Goodwin, 1976; Sloan and Robinson, 1984) – and  85  remained in the resorbed or early-active stages at a salinity of 24 throughout the experiment. In contrast, at a salinity of 29, gonads developed mostly to the late-active stage by day 62, similar to what would be seen under natural conditions two to three months after the end of the spawning season (Sloan and Robinson, 1984). Lowered salinity (<6) also reduced gonad development in the American oyster, Crassostrea virginica (Butler, 1949). As gonad development progressed at a salinity of 29 the surface area occupied by gametes (GOI) increased with a concomitant decrease in connective tissue (COI) area. This is to be expected as connective tissue (composed largely of glycogen) is converted to gametes as gonads mature (Bayne and Newell, 1983). The COI also decreased at a salinity of 24, but there was no corresponding increase in GOI. Instead, there were more vacuoles and voided follicles: a likely symptom of gonad degeneration. Phagocytosis of gametes in populations of the brackish-water bivalve Corbicula japonica exposed to low salinity (1.5) resulted in similar patterns of gonad degeneration (Baba et al., 1999). This gonad atrophy would result in both a low COI and low GOI. Salinity did not have an influence on GSI d, but GSIw was significantly (albeit slightly) higher at 29 than 24; these differences were likely due to higher water content in the somatic tissue relative to gonadal tissue at a salinity of 24. The GT results provided clearer evidence of how salinity affected gonad growth. Since, in many bivalves, energy reserves are in the form of connective tissue (Bayne and Newell, 1983; Gabbott, 1983), the thickness of the gonad sheath is reflective of the total amount of reserves. The increasing GT (2.8 to 8.8 mm over 62 days) in the salinity treatment of 29 strongly suggests that these clams were better able to amass reserves than those in the lower  86  salinity of 24 (3.4 to 4.3 mm over 62 days). Inhibited GT growth in the 24 salinity treatment was possibly due to the metabolic costs associated with cell-volume regulation via osmolyte transport (Neufeld and Wright, 1996). Metabolism (as estimated by oxygen consumption) increases under chronic exposure to lowered salinity (Stickle and Sabourin, 1979; Kim et al., 2001) and remains elevated while hyposaline conditions persist (Shumway and Koehn, 1982; Hamer et al., 2008). The Chilean scallop (Argopecten purpuratus), for example, when held at salinities of 24 and below had increased oxygen consumption which resulted in a negative scope for growth (Navarro and Gonzalez, 1998). A negative scope for growth can result in the metabolism of stored reserves [seen in both V. philippinarum (Delgado and Pérez-Camacho, 2007) and the carpet shell clam, Ruditapes decussatus (Delgado and Pérez-Camacho, 2003)] at the expense of gonad development (Navarro et al., 2000) or somatic growth (Bataller et al., 1999; Jansen et al., 2009). Research investigating the effects of salinity on oxygen consumption and scope for growth in P. generosa is warranted. Trials conducted during the present study showed increased oxygen consumption of clams held at a salinity of 24 compared to those held at 29. In summary, salinities of 17 and 20 resulted in heavy mortalities (75–90%) of adult geoduck clams, all of which were associated with a fungal infection. There were reproductive penalties to living at lower salinity as no clams at 24 developed gonads beyond the early-active stage while 75% of clams at 29 developed to the late-active stage. At a salinity of 24 there was also a reduced volume of gametes (lower GOI) compared with 29 and a lower GT suggesting that there may be increased metabolic costs associated with chronic exposure to low salinity and presumably reduced fecundity.  87  These results have major implications for geoduck aquaculture and culturists would be well advised to choose nursery and grow-out locations where freshwater input is minimal and annual salinity fluctuations are minor. The results were in agreement with the alternative hypothesis (section 1.5.2) that salinity would significantly affect mortality, tissue water content, gonadosomatic index, development stage, gonad thickness and oxygen consumption.  88  Table 3.1 Results of ANOVAs on various attributes of experimental Panopea generosa. One-way ANOVAs pertain to day 0 and include all salinities (17, 20, 24, 29). Two-way ANOVAs pertain to all sample days (0, 41, 62 d) and include only salinities 24 and 29. Sources of variation are salinity (S, fixed factor), time (T, fixed factor), interactions (S×T), and error. Values in bold are significant at P < 0.05. NS = not significant. One-way ANOVAs Source  df  SS  S Error  Condition index, wet 3 4.36 12 2.52  S Error  Condition index, dry 3 0.06 12 0.15  Two-way ANOVAs F ratio 6.92  P value <0.001  Source  df  S T S×T  Condition index, wet 1 0.0019 2 0.0043 2 0.25  Error 1.67  >0.2 NS  S T S×T Error  S Error  Water content (%) 3 117.88 12 38.87  12.13  <0.0001  S T S×T Error  S Error  Gonadosomatic index, wet 3 0.0021 1.34 12 0.0063  >0.2 NS  S T S×T Error  S Error  Gonadosomatic index, dry 3 0.0025 1.60 12 0.0068  S Error  Gamete occupation index 3 27.79 2.70 12 41.13  S Error  Connective tissue occupation index 3 233.79 2.17 >0.1 NS 12 431.67  >0.2 NS  S T S×T Error  >0.05 NS  S T S×T Error S T S×T Error  S Error  Gonad thickness 3 10.87 12 18.83  2.31  >0.1 NS  S T S×T Error  S Error  Clearance rateA 3 4518.79 12 398.57  45.35  <0.0001  S T S×T Error  SS  18  P value  0.02 0.03 1.53  >0.5 NS >0.5 NS >0.2 NS  0.07 0.07  >0.5 NS >0.5 NS  1.44  >0.2 NS  1.11 0.25  >0.2 NS >0.5 NS  2.42  >0.1 NS  11.40 2.00 2.10  <0.005 >0.1 NS >0.1 NS  3.75 2.10 2.03  >0.05 NS >0.1 NS >0.1 NS  2.06 5.38 3.61  >0.15 NS <0.02 <0.05  1.47  Condition index, dry 1 0.01 2 0.02 2  0.40  18  2.50  Water content (%) 1 6.79 2 3.03 2  29.58  18  109.97  Gonadosomatic index, wet 1 0.0057 2 0.0020 2 0.0021 18  0.0090  Gonadosomatic index, dry 1 0.0025 2 0.0028 2 0.0027 18  0.0119  Gamete occupation index 1 233.21 2 1220.58 2 819.76 18  F ratio  2041.36  Connective tissue occupation index 1 100.51 3.24 2 496.67 8.00 2  38.51  18  558.42  Gonad thickness 1 21.07 2 78.95 2  25.76  18  63.89  Clearance rateB 1 3.11 2 1872.39 2 18.64 18  0.62  >0.05 NS <0.005 >0.5 NS  5.94 11.12  <0.05 <0.001  3.63  <0.05  0.07 21.10 0.21  >0.5 NS <0.0001 >0.5 NS  798.65 C  A B C  S  Oocyte diameter 1  0.01  0.0042  >0.5  T  1  24.73  11.21  <0.05  S×T  1  1.18×10-8  0.0001  >0.5  Error  3  6.62  Clearance rates measured on day 2. Clearance rates measured on days 2, 38, and 58. Oocytes present only on days 41 and 62  89  Mortality (%) (±SE)  100 90 80 70 60 50 40 30 20 10 0  A A  Salinity  17  20  24  29  A  B a  a  B B a  B  a  adjustment  b  b  a  B a  experimental Time period  Figure 3.1 Mean cumulative mortalities of Panopea generosa at four salinities during the salinity adjustment phase (16 d) and experimental phase (26 d). Error bars indicate SEM (n = 4). Treatments denoted by different letters differ significantly (P < 0.05, TukeyKramer test). Capital letters (above bars) apply to the effects of salinity within each time period while lower-case letters (below bars) apply to the effects of time within each salinity level.  90  a.  400 A  17  Salinity 20 24  29  CI w (%) (±SE)  350 300  NS  NS AB  AB  250 B  200 150  ns  ns  ns  b. 80 NS  CId (%) (±SE)  70 60  NS  NS  50 40 30 20 10 0  ns  0  ns  41  ns  62  Time (days)  Figure 3.2 Mean condition indices (CIs) of Panopea generosa at four salinities and three sampling times: (a) wet weight condition index (CI w) and (b) dry weight condition index (CId). Error bars indicate SEM (n = 4). Days 41 and 62 exclude salinities of 17 and 20 due to high mortalities. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). Capital letters above day 0 bars denote treatment differences for all salinities on that day. NS = non-significant (P > 0.05).  91  a. Salinity 17  20  20  24  29  GSIw (%) (±SE)  29 > 24  15  NS  10 5 0  GSId (%) (±SE)  b.  18 16 14 12 10 8 6 4 2 0  ns  ns  ns  NS NS  NS  ns  0  ns  41 Time (days)  ns  62  Figure 3.3 Mean gonadosomatic indices (GSIs) of Panopea generosa at four salinities and three sampling times: (a) wet weight gonadosomatic index (GSIw) and (b) dry weight gonadosomatic index (GSId). Error bars indicate SEM (n = 4). Days 41 and 62 exclude salinities of 17 and 20 due to high mortalities. NS = non-significant (P > 0.05).  92  Stage percentage  100 90 80 70 60 50 40 30 20 10 0  24 29  17 20  24  0  29  24  41  29  62  Time (days) Spent/Resorbed  Early active  Late active  Ripe  Partially spawned  Figure 3.4 Histograms showing the percentage of various reproductive stages of Panopea generosa held at various salinities (given above bars) over time (days), as determined by examination of histological sections.  93  a. 17 20  OD (mm) (± SE)  NS  Salinity  20  24  29  15 NS  10 5 no oocytes present  0  b.  a  b  45  A  40  GOI (%) (± SE)  35 30 25 20  NS  15  B  10 5 0  c.  NS  a a  a  a  a  b  95 90  NS  NS  COI (%) (± SE)  85 80  NS  75 70 65 60 55 50  d.  a  a  12  A  8 6  A  A  10  GT (mm) (± SE)  b  NS  B  4 2 0  a  0  a  a  b  41  a  b  62  Time (days)  Figure 3.5 Mean gonad development parameters of Panopea generosa at four salinities and three sampling times: (a) oocyte diameter (OD), (b) gamete occupation index (GOI), (c) connective tissue occupation index (COI), and (d) gonad thickness (GT). Error bars indicate SEM (n = 4). Days 41 and 62 exclude salinities of 17 and 20 due to high mortalities. Treatments denoted by different letters differ significantly (P < 0.05, TukeyKramer test). Capital letters above day 0 bars denote treatment differences for all salinities on that day. Capital letters above day 41 and 62 bars for OD and COI apply to the effect of salinity while lower-case letters below bars apply to the effect of time (i.e. no significant interaction). Capital letters above day 41 and 62 bars for GOI and GT apply to the effects of salinity within each time level while lower-case letters below bars apply to the effects of time within each salinity level (i.e. significant interaction). NS = non-significant (P > 0.05).  94  0.25  A  A  O2 consumption (mg l-1g-1hr-1)  0.2  A  0.15  B  0.1  0.05  0 17  20  24  29  Salinity  Figure 3.6 Mean oxygen consumption of Panopea generosa across a range of salinities. Error bars indicate SEM (n=11). Results standardized to 77 g DW clams. Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test).  95  Salinity  25  17  20  24  29  Clearance rate -1 (l clam h -1 )(±SE)  NS  20 15  NS  10  A A  5 B  B  0  b  2  a  38  b  58  Time (days) Figure 3.7 Mean clearance rates of Panopea generosa at four salinities and three sampling times. Error bars indicate SEM (n=4). Days 38 and 58 exclude salinities of 17 and 20 due to high mortalities. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). Capital letters above day 2 bars denote treatment differences for all salinities on that day. Lower-case letters below bars denote overall differences over time. NS = non-significant (P > 0.05).  96  4 Effect of ration on gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850)  The effect of food ration on gonad development of the Pacific geoduck clam, Panopea generosa (Gould 1850), was tested over a 52-d period. The clams were fed Isochrysis sp. (Tahitian strain, TISO) and Chaetoceros muelleri (50:50 by cell count) at rations of 0.8 × 109, 2.4 × 109, 4.0 × 109, 5.6 × 109, 7.2 × 109, and 10.0 × 109 cells clam-1 d-1 (R1, R2, R3, R4, R5, and R6, respectively). The highest ration (R6) caused 25% of the clams to die within 3 d and the treatment was subsequently terminated. Ration did not significantly affect condition index, gonadosomatic index, connective tissue occupation index, or oocyte diameter. There was a non-ration related decline in gonadosomatic index (i.e. submerged visceral mass weight: submerged total soft-tissue weight) over time that was likely related to spawning. The clams in the R1 treatment had slightly, but significantly, less gonadal material compared to those fed R4 and R5 rations, based on the ratio of the surface areas of gonad to digestive gland. Clams fed the R 5 ration had significantly lower numbers of oocytes per unit area than those fed the R 1, R2, and R3 rations and significantly lower levels of sperm occupation than clams in any of the other ration treatments on the last day of the experiment. Clearance rates of clams increased significantly from the R1 ration to the R2 ration and subsequently decreased significantly at R4 and R5 rations. Cell ingestion rates also increased significantly from R 1 to R2, but remained statistically constant at the higher rations. Rations above R2 yielded few benefits and resulted in spontaneous and intense spawns at R5 that would reduce the operator’s ability to control spawn events. Rations of 4.0 × 10 9, 5.6 × 109 cells clam-1 d-1 97  are likely adequate for broodstock conditioning, but a higher ration of 7.2 × 109 cells clam-1 d-1 may be necessary if frequent spawns are to be induced.  4.1 Introduction Bivalve reproduction is known to be influenced by several factors, including temperature (Mann, 1979b; Santos et al., 1993; Martínez et al., 2000a; Chávez-Villalba et al., 2002, 2003b; Fabioux et al., 2005), salinity (Baba et al., 1999), water quality (Helm et al., 2004), and nutrition (Robinson, 1992c; Wilson et al., 1996; Utting and Millican, 1997; Nevejan et al., 2008). The latter, specifically ration, can be very important in broodstock conditioning as a positive energy balance is necessary to allow for reproductive growth (i.e. energy ingested must exceed maintenance metabolic demands and losses to waste) (Newell, 1981; Bayne and Newell, 1983; Gabbott, 1983; Bayne, 1998). In other commercially-important bivalve species, ration has been shown to influence egg output (i.e. fecundity) (Utting and Millican, 1997; Caers et al., 2002; Chávez-Villalba et al., 2002; Chávez-Villalba et al., 2003a; Chávez-Villalba et al., 2003b) and rate of gonad development (Delgado and Pérez-Camacho, 2003), possibly leading to increased fertilization rates (Lannan et al., 1980) and higher success rates of embryogenesis (Utting and Millican, 1998). Excess levels of feed can, however, result in declining assimilation efficiency, a negative scope for growth (SFG) (Newell, 1981), and ultimately a decrease in egg production (Utting and Millican, 1998). In this study I fed adult P. generosa a number of different daily rations to evaluate the influence of ration on gonad development in order to identify an optimal level for broodstock conditioning. The impact of ration was evaluated by examining the following  98  parameters: mortality, spawning rate, condition index, gonadosomatic index, digestive gland index, gonad development classification, oocyte diameter, oocytes per area, spermatic material occupation index, connective tissue occupation, and clearance rate.  4.2 Materials and methods 4.2.1 Algal culture Live algal cultures of Isochrysis sp. (Tahitian strain, TISO clone, CCMP 1324) and Chaetoceros muelleri (CCMP 1316) were cultured as per Chapter 2 (section 2.2.1). 4.2.2 Broodstock collection and initial maintenance Adult P. generosa were collected on February 8, 2009 as part of a commercial harvest from Thormanby Island, BC (49˚ 30.6’ N, 124˚ 1.7’ W) using a high-pressure hydraulic pump to extract the clams from the substratum. Water temperature at harvest depth (8–12 m) was 8˚C. After landing, the clams were transported to the Pacific Biological Station (Nanaimo, BC) within 8 h at an air temperature of 7˚C and held in ambient seawater (9.5˚C and salinity 28) in sea tables for 4–5 d until deployment in the replicate tanks. Whole wet weight and shell length (measured to the nearest 1.0 mm on the anterior-posterior axis of the right valve using vernier callipers) were recorded for each clam (mean ± SD: 1,609 ± 271 g and 162 ± 10 mm, respectively, n = 100) . To prevent possible shell gaping and tearing of flesh, PVC-coated wire mesh (mesh size: 25 x 25 mm; wire diameter: 3 mm) was cut into strips (length x width: 34 x 8 cm), fashioned into a U-shape, and moulded around the shell of each clam. Clams were starved within  99  the replicate tanks for 20 d, and then held on a maintenance diet of 4 × 10 9 cells clam-1 wk-1 for 4 wk prior to starting the experiment.  4.2.3 Broodstock conditioning: experimental design The experimental ration treatments were 0.8 × 109 (R1), 2.4 × 109 (R2), 4.0 × 109 (R3), 5.6 × 109 (R4), 7.2 × 109 (R5), and 10.0 × 109 (R6) cells clam-1 d-1 (50:50 cell count of TISO and C. muelleri), which were fed to the clams daily for 52 d. The range of rations was determined by a preliminary experiment using adult geoducks where the maximum ingestion rate was approximately 7.2 × 109 cells clam-1 d-1 with and average of approximately 4.0 × 109 cells clam-1 d-1. An additional ration (R6) was selected to exceed the maximum ingestion rate measured. The experiment was a completely randomized block design consisting of 4 blocks (tanks) with one replicate of each ration treatment per block. Each block consisted of a 6,200-l oval-shaped tank (L × W: 2.3 × 3.2 × 1.0 m) filled with 4,980-l of seawater, with a water flow rate of 1,200 l h-1. Seawater was filtered to 5 μm using bag filters and maintained at 11˚C by mixing heated and chilled seawater. Within each block there were five 220-l totes (L × W × D: 105 x 50 x 42 cm), each unit containing five clams. Clams were placed flat on their right valve with siphons oriented toward the centre of the tote. The totes sat on the bottom of the tanks, each being aerated to help circulate the water. Each treatment tote was static fed daily for 4 h by lowering the water level in the block tank below the lip of the totes, thereby isolating each replicate tote. Water volumes in the totes were adjusted so that initial cell concentrations during feeding were 1.6 × 104, 4.8 × 104, 8.0 × 104, 11.2 × 104, 14.4 × 104, and 20.0 × 104 cells ml-1 for the R1, R2, R3, R4, R5, and R6 rations, respectively. During  100  feeding, aeration kept the totes oxygenated. After feeding, the totes were emptied, to prevent mixing of water and residual feed among treatments, and the tanks were refilled with fresh filtered seawater. The photoperiod was 16 light:8 dark with dim lighting (< 160 lux at water surface at tank center). Salinity was ambient and ranged from 28 to 29 during the experiment.  4.2.4 Clam sampling Samples were collected on days 0, 25, and 52. At each sample point one randomly-selected clam was taken from each tote for a total of four clams per treatment. Sampled clams were rinsed in filtered seawater, blotted dry with paper towel, and weighed with an electronic balance. The clams were then cut open and drained of internal water. The visceral mass (alimentary canal, digestive diverticula, sheathing gonad tissue, and attached foot), somatic tissue (all tissues other than the visceral mass, comprised largely of the siphon and mantle muscle), and shell were all separated, drained of excess water, and blotted dry. Live weights and wet weights of visceral mass, somatic tissue, and shell were recorded to the nearest 0.1 g. Submerged weights of the live animal, shell, gonad, and somatic tissue were also measured to examine the potential of the method as an alternative to dry weight analysis. This was done by placing the scale on a wire rack over a plexiglass tank (L × W × H: 58 × 58 × 30 cm, Volume: 101 L) filled with seawater. A fiberglass mesh screen (L × W: 38 × 52 cm, mesh size: 1.5 × 1.5 mm) was attached to the suspended weight hook on the bottom of the electronic balance and suspended 20 cm under the water line. The clam or isolated tissues were placed on the screen and the submerged weights recorded after the  101  readings became stable. Water temperature in the weighing tank was kept within 1˚C of treatment water via addition of chilled seawater. Preliminary trials between 8 and 13˚C indicated that submerged weight had a significant (F1,20 = 13.11, P = 0.0017, R2 = 0.40) linear relationship to temperature. The slope of this relationship was, however, very shallow and not considered a factor, altering the total weight by 0.12 g for each degree (˚C) of change (<0.04% of the total submerged weight of an adult clam).  4.2.5 Mortality rate and spawn events Daily records were kept of the total number of dead and moribund clams in each ration treatment. These clams were removed from the experiment, but not replaced. Daily observations of the geoducks’ spawning activity (i.e. sperm and eggs in the water column) were also made. The number of clams spawning per event was recorded. Cumulative results were totalled between 0 to 25 d and 26 to 52 d.  4.2.6 Clearance and ingestion rates Clearance rates (CR) and ingestion rates (IR) were determined on day 42 of the experiment. On that day the clams were fed normal cell concentrations, but only TISO. Algal-concentration samples of 20 ml were taken from each replicate tote at the beginning of the feeding cycle and at the end of the 4-h feeding. The water column was thoroughly mixed with a perforated plunger prior to collecting each sample. Clearance rate was calculated using the method of Quayle (1948) [as described in Coughlan (1969)]: M = m / (n × t) × loge(C0 / Ct), where M = clearance rate (ml of water cleared of cells clam-1 h-1). Ingestion rate was calculated using the method described in Khalil  102  (1996): I = [(C0 – Ct) / (n × t)] × m, where I = ingestion rate (cells clam-1 h-1). For both equations: m = volume of suspension (ml), n = number of animals, t = duration of trial (h), C0 = initial cell concentration (cells ml-1), and Ct = final cell concentration (cells ml1  ).  4.2.7 Condition and gonadosomatic indices After recording wet and submerged weights for shell, somatic tissue, and visceral mass the samples were dried at 58-60˚C to constant weight (72 hr) (see Chapter 3, section 3.2.6).  4.2.7.1 Condition index A condition index (CI) based on the method of Walne and Mann (1975) was adapted for geoduck: CI = (T × 100)/S where CI is the condition index, T is the total softtissue mass, and S is the shell mass. This was calculated for wet (CIw), dry (CId) and submerged (CIs) soft tissue and shell masses respectively. For more detail see Chapter 2, section 2.2.6  4.2.7.2 Gonadosomatic index Gonadosomatic indices (GSIs) were calculated using visceral mass (including gonad and digestive gland) and total soft tissue. For details see Chapter 2 (Section 2.2.6). Calculation was based on the method in Sloan and Robinson (1984): GSI = V/T × 100 where GSI is the gonadosomatic index, V is the visceral mass weight, and T is the total  103  soft-tissue weight. Calculations were done using wet wt (GSI w), dry wt (GSId) and submerged wt (GSIs) data respectively. Since part of the visceral mass was removed for histological analysis before drying, a correction factor was necessary for the final visceral mass dry weight. This was done by multiplying the ratio of dry to wet visceral mass weight by the wet weight of the sub-sample and adding it to the dry weight of the visceral mass.  4.2.7.3 Digestive gland index The digestive gland index (DGI) was derived to determine the relative proportion of digestive gland within the visceral mass and is similar to the Gonadal Area Index (Barber et al., 1991). To calculate the DGI, a transverse section was first cut across the widest point of the visceral mass. From the posterior section, two-dimensional surface areas of the total visceral mass and the digestive gland were measured. The total surface area of the visceral mass section was estimated by measuring height and width and calculating the surface area as an oval shape. The digestive gland surface area (roughly rectangular) was calculated using the height and width of the gland. The digestive gland is easily identified within the visceral mass due to its distinctive dark green or brown color (as opposed to the cream or light yellow color of the gonad) and its location at the inner core. The DGI was calculated as a ratio of digestive gland surface area to visceral mass surface area. A high DGI indicates a relatively large digestive gland area and a thin gonad layer.  104  4.2.8 Histological sampling Visceral mass samples (~1 cm3) from each sampled clam were taken from the same area of the right-posterior side, weighed, and fixed in Davidson’s solution for 72 h. Fixed samples were subsequently cut into 4–5 sub-samples before being dehydrated in a graded ethanol series and embedded in paraffin wax. Embedded samples were sectioned to 5-μm, stained with hematoxylin-eosin, and mounted on slides (see Chapter 2, section 2.2.8). Slides were analyzed using a Motic B5 compound microscope with Motic Images Advanced 3.2 software (Motic Electric Group Co., Ltd., Richmond, BC, Canada). For subsequent analysis, 4–5 digital images were captured from each of the clam’s visceral mass sub-samples using the image analysis software at both 40× and 100× magnification.  4.2.8.1 Development classification (index) For analysis of development index (DI) at least two of the 40× magnification digital images from each sub-sample were randomly selected (total of 10 images per clam). Ten randomly-chosen follicles (Gribben et al., 2004) from each image were scored for development which made a total of 100 scores per clam. Follicle development was scored as: early active = 1, late active = 2, ripe = 3, partially spent = 4, and spent/resorbed = 5 [ (see Chapter 2, section 2.2.8) based on the classifications used for P. generosa (Andersen, 1971; Goodwin, 1976; Sloan and Robinson, 1984) and P. zelandica (Gribben et al., 2004)]. Many of the follicles were in intermediate states and if a follicle could not clearly be categorized it was given an intermediate score to the nearest half score. A mean score was calculated for each clam.  105  4.2.8.2 Oocyte diameter For each female clam a minimum of 10 digital images (100× magnification) of the gonad were used (at least two images were randomly selected from each of the visceral mass sub-samples). Using the image analysis software, 5 to 10 randomlyselected oocyte diameters (OD) were measured from each image to the nearest 0.2 μm. A minimum total of 75 oocytes were measured for each clam and only oocytes with visible nuclei were used. The proportion of mature oocytes (diameter: 30.1 – 40.0 μm) in relation to total oocytes was calculated for each female clam. Mean values were calculated for each clam. (see Chapter 2, section 2.2.8).  4.2.8.3 Gamete occupation indices: oocytes per unit area, spermatic material occupation index, and connective tissue occupation index Using the image analysis software, oocytes per unit area (OA), spermatic material occupation index (SOI), and connective tissue occupation index (COI) were quantified. The OA data was collected as per Chapter 2 (Section 2.2.8) while spermatic material and connective tissue occupation were measured colourimetrically as per Chapter 3, (Section 3.3.4). All parameters were determined from a minimum of 10 digital images per clam (minimum of two images per visceral mass sub-sample). Mean values were calculated for each clam.  106  4.2.9 Statistical analyses To test the influence of rations on mortality, a chi-square test for independence was applied. For spawning, cumulative proportions of clams that spawned between sampling dates was analyzed using repeated measures ANOVA (repeated measures to account for multiple spawns by individual clams). The spawning periods were from days 0 to 25 and days 26 to 52. Both CR and IR were analyzed using completely randomized block ANCOVA with dry tissue weight as the covariate since body size influences clearance rates (Khalil, 1996; Riisgård and Seerup, 2003; Zhuang et al., 2004; Zhuang and Wang, 2004). Condition indices (wet, dry, and submerged) were analyzed using completely randomized-block ANCOVA as well with time and temperature as fixed factors and dry shell weight as the covariate. Two-way randomized block ANOVA was used to analyze GSI, DGI, DI, and COI with temperature and time as fixed factors and tank as the blocking factor. The variables OD, OA, and SOI were analyzed without blocking since these were sex-specific parameters and some blocks lacked male or female replicates. These three parameters were analyzed as if completely randomized. Since no block effects were found with any other parameters (chiefly COI), this assumption is valid. Day 0 results masked significant results for some parameters so analysis was reexamined using only the final two sample dates for all parameters sampled over time. It is noted if the reduced analysis changed the significance of the results. For all parametric tests, normality of data was tested using the Shapiro-Wilk W test and homogeneity of variances was tested using the modified Levene’s test. Departures from normality and homogeneity were corrected using the natural-logarithm transformation for CIw, CId, CIs, GSIw, GSId, GSIs, OD, OA, DI, and IR. The arcsine  107  transformation was used for spawn proportions, DGI, SOI, and COI. No transformations were needed for CR. ANCOVA assumptions were examined with plots of residuals and the normal probability plot of residuals; no departures were found. Significant differences among factors were tested using a post-hoc multiple comparison TukeyKramer test (α = 0.05).  4.3 Results 4.3.1 Mortality rate and spawn events Within the first three feedings, 25% of the clams in the R6 ration treatment died while none died in any of the other treatments (χ2 = 21.1 df = 4, P < 0.001) (Fig. 4.1). The remaining live clams in the R6 ration treatment had developed flaccid siphons, a sign that they were moribund. At this point, this particular ration treatment was terminated. None of these clams showed external signs of damage to the periostracum, but necropsies indicated that the stomachs, gills, labial palps, and esophagi had large clumps of algae that were not seen in clams fed lower rations. There were no differences (χ2 = 2.0 df = 4, P > 0.50) in cumulative mortality among the other ration treatments with final cumulative mortality rates for the 52-d experimental period being 15% (R1 ration), 20% (R3 and R4 rations), and 30% (R2 and R5 rations) (Fig. 4.1). Prior to day 0 (holding phase) one random, non-block related spawn event was noted in one replicate of each treatment. ANOVA results revealed that spawning percentage was significantly affected by time (F1,3 = 13.13, P < 0.05) and the interaction between time and ration (F4,12 = 7.59, P < 0.005), but not by ration alone (F4,12 = 1.94, P > 0.10). The cumulative percent of clams spawning from days 0 to 25 ranged from 50% 108  (R1 ration) to 100% (R4 ration), but there was high variance among replicates and no significant differences among ration treatments were detected (Fig. 4.2). Between days 26 and 52 there were significant differences among ration treatments with 85 and 100% of clams spawning in the R5 and R3 ration treatments, respectively. These percentages were significantly higher than with R1, R2, and R4 rations that had spawn rates of 15, 0, and 0%, respectively (Fig. 4.2). The percentage of clams spawning in the R 1, R3, and R5 ration treatments did not change significantly from day 25 to 52, but percent spawning in the R2 and R4 treatments significantly declined between days 26 and 52 (Fig. 4.2).  4.3.2 Clearance and ingestion rates Clearance rate was significantly influenced by ration (F4,11 = 5.72, P < 0.01) and the dry-tissue-weight covariate (F1,11 = 14.25, P < 0.01), but not by block (F3,11 = 0.15, P > 0.50). The highest CR was measured in the R2 ration treatment and this was significantly higher than in the R1, R4, and R5 treatments (Fig. 4.3a). Ingestion rate was also significantly influenced by ration (F4,11 = 11.15, P < 0.001) and the covariate (F1,11 = 12.54, P < 0.01), but not by the block (F3,11 = 2.55, P > 0.1). Clams in the lowest ration (R1) had significantly lower ingestion rates than individuals in any other ration treatment (Fig. 4.3b). This means that increasing rations beyond R2 did not significantly increase the number of cells ingested per clam per unit time.  109  4.3.3 Condition and gonadosomatic indices 4.3.3.1 Condition index ANCOVA results revealed that condition indices – whether based on wet, dry, or submerged weights –were not significantly influenced by ration, time, or the interaction between these two factors (Table 4.1, Fig. 4.4). The covariate (dry shell weight) was a significant factor for all three CIs (Table 4.1).  4.3.3.2 Gonadosomatic index As with CI, both the main factor ration and the interaction between ration and time did not significantly affect wet, dry, or submerged GSIs (Table 4.2, Fig. 4.5). There was a significant time effect, however, for both dry and submerged GSIs (Table 4.2). GSId was significantly higher on day 0 than on day 25 d, but not significantly different from day 52 (Fig. 4.5b). GSIs was significantly higher on day 0 than on day 25, but GSIs on day 52 remained significantly lower than on day 0 (Fig. 4.5c).  4.3.3.3 Digestive gland index For DGI, the main factors of ration and time were significant without a significant interaction (Table 4.2). Clams in the R1 ration treatment had significantly higher DGI than those in the R4 and R5 rations, with no other significant pair-wise comparisons (Fig. 4.5d), suggesting that the R1 clams had relatively less gonad area than R4 or R5 individuals. There was a significant increase in the DGI from day 0 to day 52 (Fig. 4.5d).  110  The relatively large variance in the R3 ration treatment on day 52 was related to one clam that had a particularly heavy spawn prior to sampling; this likely resulted in a high DGI.  4.3.4 Histological sampling 4.3.4.1 Development classification (index) The DI was significantly influenced by both ration and time, but not by the interaction between these two factors or by the block effect (Table 4.3). The DI was significantly higher in the R5 ration treatment than in the R2 or R4 rations when all three sampling days were included in the analysis (Fig. 4.6a). Day 0 results, however, appeared to be obscuring a pattern indicating that the extreme ends of the ration range had higher DIs. To investigate this, a second ANOVA – restricted to days 25 and 52 – was performed. This ANOVA showed that the effect of ration was significant, but that there was no significant effect of time, interaction, or block (Table 4.3). These results showed that clams fed the R5 ration were significantly more spawned out than clams in all other treatments with the exception of the R1 ration. The R1 ration did not differ from any other treatments. Clams were significantly more spawned out on days 25 and 52 than on day 0 (Fig. 4.6a).  4.3.4.2 Oocyte diameter Oocyte diameter was not significantly influenced by ration, time, or the interaction between these two factors (Table 4.3) averaging 24.3 ± 6.3 μm (mean ± SE, n = 30) across all rations and sample dates (Fig. 4.6b). Wide variation in oocyte diameters  111  was expected due to high variability in spawning rates among treatments, whereby the sizes may be skewed due to the expulsion of ripe oocytes or the generation of new ones. Because of this potential data skew, the percentage of mature oocytes was also examined. It was found that the percentage of mature oocytes was also not significantly influenced by ration, time, or the interaction between these two factors (Table 4.3) and averaged 15 ± 0.3% (mean ± SE, n = 30) across all rations and sample dates (data not shown).  4.3.4.3 Gamete occupation indices: oocytes per unit area and spermatic material occupation index The OA was significantly influenced by ration, but not by time or the interaction between ration and time (Table 4.3). Clams fed the R5 ration had the lowest overall OA levels, significantly lower than those fed R1, R2, and R3 rations with no other significant pair-wise differences among rations (Fig. 4.6c). The area of gonad occupied by spermatic material (SOI) in males was significantly influenced by ration, time, and the interaction between these two factors (Table 4.3). There were no significant pair-wise comparisons among ration treatments on days 0 and 25, but males in the R5 treatment had significantly lower SOI than those in any other ration treatment on day 52 (Fig. 4.6 d). Pair-wise multiple comparison tests showed that SOI dropped dramatically between days 25 and 52 in the R 5 ration treatment, but there were no other significant differences among days within any other ration level (Fig. 4.6 d).  112  4.3.4.4 Connective tissue occupation index There were no significant effects of ration, time, interaction, or block on COI (Table 4.3). The COI averaged 32.0 ± 1.7% (mean ± SE, n = 60) across all rations and dates (Fig. 4.6e). One may expect a significant increase in COI concomitant with the decrease in SOI in males and low OA in females, but the post-spawn connective tissue had not yet collapsed into solid parenchyma and contained empty follicles and large vacuoles.  4.4 Discussion The rapid and high incidence of mortalities in the R6 ration led to the termination of that treatment within the first few days of the experiment. High concentrations of particulate organic matter can normally be handled by bivalves through increased production of pseudofeces (Zhuang et al., 2004; Zhuang and Wang, 2004), but pseudofeces production in the R6 treatment was not noted to be higher than in other treatments (not quantified) suggesting that the high ration or cell concentration overwhelmed the labial palps (Velasco and Navarro, 2002) and inhibited the clams’ ability to produce pseudofeces. Necropsies of the dead and moribund clams from R 6 – which revealed packed stomachs and esophagi as well as algal masses on the gills and labial palps – supported this assertion. Mortalities associated with overfeeding have also been reported for spat of the silver-lipped pearl oyster, Pinctada maxima (Mills, 2000). The initial cell concentration in the R6 treatment was 2.0 × 105 cells ml-1 (TISO) and this level should be avoided when feeding P. generosa. There were no adverse effects of feeding at the lower rations.  113  Although ration level from R1 to R5 did not significantly affect clam mortality, there was a significant impact on clearance and ingestion rates. In addition to generating pseudofeces, increases in cell concentrations can be handled through changes in pumping rate (Foster-Smith, 1975; Winter, 1978; Iglesias et al., 1992; Velasco and Navarro, 2002). In the present study, clearance rate increased dramatically from R 1 to R2, but declined significantly from R2 to R4 and R5. Ingestion rate also increased significantly from R1 to R2, but remained essentially constant from R2 to R5. These patterns are consistent with the functional response commonly seen in bivalves (Riisgård and Randløv, 1981; Utting, 1986; Riisgård, 1988; Beiras and Pérez-Camacho, 1994; Pérez-Camacho et al., 1994; Baldwin and Newell, 1995; Brown and Robert, 2002). In the present study, the threshold cell concentration where clearance rate decreased and ingestion rate levelled off was 4.8 × 104 cells ml-1 (starting concentration of TISO in R2). This threshold level is roughly similar to levels reported in other bivalve species such as the mussel Mytilus edulis (3 x 104 cells ml-1 of Phaeodactylum tricornutum) (Riisgård and Randløv, 1981) and the surf clam Paphies donacina (20 × 104 cells ml-1 of Isochrysis galbana) (Marsden, 1999), but slightly higher than Tapes decussatus (103 cells ml-1 of Dunaliella marina) (Khalil, 1996). Based on these results it appears that the feeding apparatus of P. generosa is adapted to handling similar cell concentrations to other bivalve species. It also evident that P. generosa has a maximum daily ingestion capacity that is similar to other, much smaller, bivalve species. In the present study, P. generosa ingested a maximum of approximately 2 × 109 algal cells clam-1 during a 4-h feeding period (extrapolated to 24 h = 12 × 109 algal cells ind-1 d-1). The maximum daily ingestion rates of P. generosa are only marginally higher than other (smaller) species  114  such as Venerupis (= Tapes) philippinarum [2 × 109 algal cells ind-1 d-1 of mixed algal species (Utting and Spencer, 1991)], C. gigas [3 × 109 algal cells ind-1 d-1 of mixed algal species (Utting and Spencer, 1991) and 2 × 109 cells ind-1 d-1 of Skeletonema costatum and/or Tetraselmis suecica (Robert and Gérard, 1999)], and Pecten maximus [10 × 109 cells ind-1 d-1 of Pavlova lutheri, TISO, Chaetoceros calcitrans, S. costatum (20:20:20:40 by cell count) (Robert and Gérard, 1999)]. This is despite P. generosa used in the present study being 35 to 50 times larger (by live weight; unpublished data) than mature V. philippinarum and 4 to 10 times larger (by live weight; unpublished data) than mature C. gigas. The average ingested ration of P. generosa was 0.12% (dry weight of algae:dry soft-tissue weight of animal) which is far lower than recommendations of 3% for bivalves at temperatures below 20˚C and 6% for bivalves at temperatures greater than 20˚C (Utting and Millican, 1997). Despite the proportionally low ration, the amount of cells ingested appeared to be adequate to maintain gross tissue levels based on the CIs for all treatments, even at the lowest ration. If there was a deficiency in ration it is likely that there would have been a measurable reduction in CI over the duration of the experiment. Crassostrea virginica, for example, lost 12% of its whole wet weight after only 42 days of starvation (Wright and Hetzel, 1985). Although there were no significant overall tissue losses (based on CI), there were indications of a non-ration related loss in gonad material, based on a decrease in both GSId and GSIs over time. This could very well be related to spawning which was seen in all ration treatments between days 0 and 25. Decreases in GSI after spawning have been seen in geoducks [P. generosa (Sloan and Robinson, 1984) and P. zelandica (Gribben et al., 2004)] and numerous other bivalve species such as Mercenaria mercenaria, C.  115  virginica (Heffernan et al., 1989a, 1989b), Mercenaria spp. hybrids, (Hesselman et al., 1989), and Spisula solidissima simili (Kanti et al., 1993). Supporting evidence that spawning contributed to GSI reduction over time came from DI (which showed an overall increase from days 0 to 52) and OD (where oocytes did not increase beyond 24 μm in diameter). In data previously collected by individuals in our laboratory, ova from fully-ripe, pre-spawning P. generosa averaged 30.6 ± 8.2 μm (mean ± SD, n = 8) in diameter (unpublished data). No ova of this size were measured during the present study even though one would expect the mean oocyte diameter to have increased over time [as seen in other species (Chávez-Villalba et al., 2003a; Delgado and Pérez-Camacho, 2003; Gribben et al., 2004)]. Based on the high proportion of spawners in all treatments from days 0 to 25, it is likely that as the oocytes did grow and mature, but were expelled before the next scheduled sampling date. (Robinson, 1992a) reported similar results where Crassostrea sikamea (formerly Crassostrea gigas kumamoto) maintained roughly a 50:50 ratio of ova to developing oocytes over an extended period (18 wk) under hatchery conditions. The relative size of the viscera (GSI) was not influenced by ration, but the relative make up of the viscera (DGI) was. This was seen in R1 where clams had a significantly higher DGI (= lower relative gonad surface area) than those in R4 or R5. Since R1 had low levels of spawning between days 25 and 52 this difference was likely attributable to the lower ration. Just as optimal-energy diets can increase the amount of gonad material (Albentosa et al., 2003) sub-optimal rations can lead to decreased gonad growth. Low rations, with a negative energy balance, cause the metabolism of organic reserves and can result in reduced growth (Laing, 1993; Utting and Millican, 1998), decreased somatic  116  tissue (Caers et al., 2002), and fewer mature oocytes (Utting and Millican, 1998; ChávezVillalba et al., 2003b). One would expect that a relative decrease in gonad material would translate to a change in the GSI, which was not observed in the present study, but changes in DGI in this case were caused by differences of only a few mm in gonad thickness. This would translate to a very small change in gonad mass and the GSI method is probably not precise enough to detect such small changes. Based on these results, it appears that the lowest ration (R1) lead to metabolism of gonad reserves and/or a failure to replenish gonad tissue after spawning. The magnitude of either process was relatively small. The highest impact on gonad development was seen in R5 which had an average DI of 4.73 (i.e. an average state between spawned and resorbed), an OA significantly lower than R1, R2, and R3 rations, and a significantly lower SOI at day 52 than any other ration treatment. This may be a result of the high cell concentrations of R 5 causing more intense spawns and more gametes released, since algae are spawn inducers (Goodwin et al., 1979; Utting and Spencer, 1991; Helm et al., 2004) and a high ration diet can increase the number of spawning events (Delgado and Pérez-Camacho, 2003). The R5 ration treatment did have a significantly higher proportion of spawners than the R 1, R2, and R4 treatments from day 25 to day 52, but not the R3 treatment. Despite having similar spawner percentages in the R3 and R5 rations, the former had a significantly lower DI (indicating more ripe gonads). This may be the result of higher spawn intensity in the R 5 ration treatment or a lack of gamete regeneration after spawning; higher algal cell concentrations can result in decreased overall growth (Coutteau et al., 1994a; Coutteau et al., 1994b; Caers et al., 1999a) or inhibited reproduction (Utting and Millican, 1997).  117  Since R3 had a similar spawn percentage to R5 (notably during 25–52 d) and maintained a DI, OA, and SOI similar to R1, R2, and R4 it is the ration most likely to maximize gamete output. As discussed previously, P. generosa ingested few algal cells relative to much smaller bivalve species. Despite this, the ration was adequate to maintain gonad levels with no impact on GSI and only a marginal increase in R1’s DGI (relative to R4 and R5). There are two likely reasons why P. generosa perform well at these relatively low rations. One may be that as a larger animal it has proportionally lower metabolic demands than smaller animals (Hamburger et al., 1983; Brougrier et al., 1995) and therefore requires proportionally less food. Another reason may be that P. generosa is a ‘conservative’ species that relies on endogenous reserves for gamete development, as opposed to an ‘opportunistic’ one that directly transfers food energy to gonad development (Bayne, 1976). Evidence for a ‘conservative’ reproductive strategy may be seen in the natural reproductive cycle of P. generosa where gametogenesis takes place in winter (Andersen, 1971; Goodwin, 1976; Sloan and Robinson, 1984). Primary productivity in BC waters during winter is typically less than 15 μg carbon l-1 (Takahashi et al., 1978) which is a TISO equivalent of approximately 1,000 cells ml -1 [assuming TISO to have 14 pg carbon cell-1 (Lavens and Sorgeloos, 1996)]; far lower than the lowest cell concentration in the present study. Despite the low primary productivity, gamete development of P. generosa nears completion before plankton blooms peak in March and April (Parsons et al., 1969; Takahashi et al., 1978). Presumably much of the reserves necessary for gametogenesis in P. generosa are derived from the mantle, digestive gland (Dridi et al., 2007), and adductor muscle  118  (Devauchelle and Mingant, 1991; Ngo et al., 2006) as in other species of bivalves. Geoducks also have another potential source of reserves in the form of a massive muscular siphon. Evidence for the use of the siphon as a reserve for gonad development was seen in the significant shortening of siphon length [3 cm over 5 months (t = 4.39, df = 10, P < 0.001)] in starved geoducks collected at the same time as those for this experiment. Despite being starved, these clams had thick gonad layers in the visceral mass suggesting that this gonad development came at least in part at the expense of muscle reserves in the siphon. This is an aspect of geoduck conditioning that should be investigated further. In conclusion, under the experimental conditions in the present study, ration did not significantly influence CI, COI, GSI, or OD, but there was a trend toward clams being more spawned-out in R5 (7.2 × 109 cells clam-1 d-1) as indicated by an increased DI, a reduced OA (females), and a reduced SOI (males). The R6 (10.0 × 109 cells clam-1 d-1) ration treatment was unsuccessful as it caused a rapid increase in mortality, which we attributed to an overwhelming of the gill structure and feeding apparatus. Clam clearance rates peaked at the R2 level, subsequently declining with increasing cell concentrations. Ingestion rates also reached a maximum at the R2 ration level but remained constant at cell concentrations above this. This suggests that P. generosa are not well adapted to high seston levels and that attempts to increase ration with high cell concentrations (beyond 5 × 104 cells ml-1 TISO) are not likely to be effective. The lowest ration (R1) did have a significantly higher DGI than the two highest rations (R 4 and R5), indicating a relative loss of gonad material (suggesting a negative energy balance), but overall all rations were sufficient to maintain CI. Future research on the scope for growth of P.  119  generosa is warranted. Rations of 4.0 × 109, 5.6 × 109 cells clam-1 d-1 are likely suitable for broodstock conditioning in a hatchery environment. The results were in agreement with the alternative hypothesis (section 1.5.3) that ration would significantly affect survival, spawner percentage, clearance rate and ingestion rate but the null hypotheses could not be rejected with respect to condition index, gonadosomatic index, oocyte diameter or connective tissue occupation,  120  Table 4.1 Analyses of covariance showing the effects of ration, sampling time, block (i.e. tank), and co-variate (i.e. dry shell weight) on various condition indices (CI) – based on wet (CIw), dry (CId), and submerged (CIs) tissue weights – in Panopea generosa. P values in bold are significant at a level of at least <0.05. Factor CIw Time Ration Covariate Interaction Block Error  df  SS  F  P  2 4 1 8 3 41  0.07 0.27 1.79 0.37 0.03 1.23  1.10 2.25 59.67 1.54 0.30  > 0.25 > 0.05 < 0.00001 > 0.10 > 0.50  CId Time Ration Covariate Interaction Block Error  2 4 1 8 3 41  0.24 0.48 2.60 0.41 0.04 2.11  2.33 2.33 50.52 1.00 0.26  > 0.10 > 0.05 < 0.00001 > 0.25 > 0.50  CIs Time Ration Covariate Interaction Block Error  2 4 1 8 3 41  0.16 0.51 2.46 0.46 0.07 2.1  1.56 2.49 48.03 1.12 0.46  > 0.10 > 0.05 < 0.00001 > 0.25 > 0.50  121  Table 4.2 Analyses of variance showing the effects of ration, sampling time, and block (i.e. tank) on various gonadosomatic indices (GSI) – based on wet (GSIw), dry (GSId), and submerged (GSIs) tissue weights – and digestive gland index (DGI) in Panopea generosa. P values in bold are significant at a level of at least <0.05. Factor GSIw Time Ration Interaction Block Error  df  SS  F  P  2 4 8 3 42  0.18 0.14 0.30 0.40 1.95  1.94 0.75 0.81 2.87  > 0.10 > 0.50 > 0.50 < 0.05  GSId Time Ration Interaction Block Error  2 4 8 3 42  0.011 0.003 0.006 0.004 0.035  6.60 0.90 0.90 1.60  < 0.005 > 0.25 > 0.50 > 0.20  GSIs Time Ration Interaction Block Error  2 4 8 3 42  1.22 0.26 0.46 0.30 2.32  11.04 1.18 1.04 1.81  < 0.001 > 0.25 > 0.25 > 0.10  DGI Time Ration Interaction Block Error  2 4 8 3 42  0.14 0.15 0.14 0.06 0.59  4.98 2.67 1.25 1.42  <0.015 < 0.05 > 0.25 > 0.20  122  Table 4.3 Results of ANOVAs from histological examination for experimental Panopea generosa. Sources of variation are ration (R, fixed factor), Time (T, fixed factor) Block (B, random factor). Values in bold are significant at α=0.05. Source  df F ratio Development index (days 0, 25, 52)  P value  df F ratio Development index (days 25, 52)  P value  R T B R×T Error  4 2 3 8 42  3.15, 6.85 1.00 0.66  < 0.025 < 0.005 > 0.25 > 0.5  4 1 3 4 27  <0.025 >0.25 >0.25 >0.25  R T R×T Error  Oocyte diameter 4 2 8 12  1.86 2.14 0.52  >0.1 >0.1 >0.5  Percentage of mature oocytes 4 1.07 2 1.35 8 1.24 12  >0.25 >0.25 >0.25  R T R×T Error  Oocytes per area 4 3.63 2 0.68 8 1.84 12  <0.05 >0.5 >0.5  Spermatic material 4 2.92 2 6.65 8 2.52 12  <0.05 <0.01 <0.05  R T B R×T Error  Connective tissue occupation 4 1.58 2 0.85 3 1.15 8 1.03 42  3.5 0.66 0.25 0.98  >0.1 >0.25 >0.25 >0.25  123  Cumulative mortality (%)  40 35  R1  30  R2  25  R3  20  R4  15 10  R5  5  R6  0 0  25  52  Time (days)  Spawn percentage (%)  Figure 4.1 Mean cumulative percent mortality of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. B  100 90 80 70 60 50 40 30 20 10 0  B  R1 R2 R3 R4  A  R5 a  25  A b  a  A b  a  52 Time (days)  Figure 4.2. Mean percentage of Panopea generosa spawning in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). Upper-case letters (above bars) indicate significant differences among ration levels within each time level and lower-case letters (below bars) indicate significant differences between time levels within each ration level (i.e. significant interaction).  124  Clearance (litres clamˉ¹hˉ¹)  20  Ingestion (108 cells clamˉ¹hˉ¹)  a  7  C BC  15 10  AB  AB  A  5 0  b B B  6  B  B  5 4 3 2  A  1 0  R1  R2  R5 R4 Ration Figure 4.3 Mean (a) clearance rate and (b) ingestion rate of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. Rations denoted by different letters above bars differ significantly (P < 0.05, Tukey-Kramer test). R3  125  a  R1 R2 R3 R4 R5  Condition index (wet)  4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0  b Condition index (dry)  1.2 1.0 0.8 0.6 0.4 0.2 0.0  c Condition index (sub)  0.30 0.25 0.20 0.15 0.10 0.05 0.00  0  25  52  Time (days)  Figure 4.4 Mean condition indices – based on (a) wet, (b) dry, and (c) submerged tissue weights – of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. There were no significant differences among rations or times for any condition index. All indices are standardized to 250 g dry shell weight.  126  a  Gonadosomatic index (wet)  R1  R2  R3  R4  R5  0.30 0.25 0.20 0.15 0.10 0.05 0.00  b  Gonadosomatic index (dry)  0.30 0.25 0.15 0.10 0.05 0.00  c  a  b  ab  a  b  b  Gonadosomatic index (sub)  0.20  0.15  0.10  0.05  0.00 d  Digestive gland index  B  AB  AB  A  A  R1 R2 R3 R4 R5  0.7 0.6 0.5 0.4 0.3 0.2 0.1 0  a  ab  b  0  25  52  Time (days)  Figure 4.5 Mean gonadosomatic indices – based on (a) wet, (b) dry, and (c) submerged tissue weights – and (d) digestive gland index of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM and n = 4. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). For all gonadosomatic indices, lower-case letters (below bars) indicate significant differences among time levels across all ration levels (i.e. no significant interaction). There were no significant differences among rations. For digestive gland index, upper-case letters (above legend) indicate significant differences among ration levels across all time levels and lower-case letters (below bars) indicate significant differences among time levels across all ration levels (i.e. no significant interaction).  127  a  b AB  A  B  R2  R3  R4  R5  R1  5.0  Development index  A  R1  Oocyte diameter (µm)  AB  4.5 4.0 3.5 3.0  b  a  A  A  AB  25 23 21 19 17  R1 R2 R3 R4 R5  Spermatic material (%)  R1 R2 R3 R4 R5  (OA) -2  Oocytes mm  715 615 515 415 315 215 115 15  90 80  A A  A A A  70 60  R2  R3  R4  R5  A  A  A  A A A  A A  A  50 40 30 20 10 0  B  a a  a  0 R1  R5  B  815  e  R4  15  a  A  R3  29 27  d c  R2  31  a a  a a  a  a  a  25  a  a a  a  b  52  Time (days)  Connective tissue (%)  70 60 50 40 30 20 10 0 0  25  52  Time (days)  Figure 4.6 Mean (a) development index, (b) oocyte diameter, (c) oocytes per unit area, (d) spermatic material occupation index, and (e) connective tissue occupation index of Panopea generosa in various ration treatments. R1 – R5 = 0.8 x 109, 2.4 x 109, 4.0 x 109, 5.6 x 109, 7.2 x 109, and 10.0 x 10 9 cells clam-1 d-1, respectively. Error bars are SEM. Treatments denoted by different letters differ significantly (P < 0.05, Tukey-Kramer test). For development index, oocyte diameter, and oocytes per unit area, upper-case letters (above legend) indicate significant differences among ration levels across all time levels and lower-case letters (below bars) indicate significant differences among time levels across all ration levels (i.e. no significant interaction). For spermatic material occupation index, upper-case letters (above bars) indicate significant differences among ration levels within each time level and lower-case letters (below bars) indicate significant differences among time levels within each ration level (i.e. significant interaction). There were no significant differences among rations or time for connective tissue occupation index.  128  5 Effect of food type on gonad development of the Pacific geoduck clam (Panopea generosa Gould, 1850)  In this chapter, adult Panopea generosa were fed different phytoplankton diets – Isochrysis sp., Dunaliella tertiolecta, Chaetoceros muelleri, and Isochrysis sp. + C. muelleri) – at a ration of 4 × 109 cells clam-1 d-1 (Isochrysis sp. dry wt equivalent) for 47 days. Gonad development was examined using gonadosomatic index, digestive gland index, gonad development index, oocyte diameter, gamete occupation indices (i.e. oocytes per unit area, oocyte area occupation, and spermatic material occupation], and connective tissue occupation index. The acceptability of the different diets was also tested using clearance and ingestion rates. The food type did not significantly affect reproductive development but Dunaliella tertiolecta had the highest clearance and ingestion rates.  5.1 Introduction In Chapter 4 the importance of nutrition with respect to ration was investigated, but ration alone does not ensure successful reproduction. The reason for this is that embryogenesis relies on endogenous reserves, stored in the eggs prior to fertilization, for energy and structural components (Gallager and Mann, 1986; Whyte et al., 1990; Utting and Millican, 1997; Pernet et al., 2003a). The main fuels for embryogenesis in bivalves are neutral lipids, predominately triacylglycerides (Gallager and Mann, 1986; Whyte et al., 1990), that are largely supplied to the eggs via lipogenesis in the mother (Gabbott, 1983). Lipogenesis often depends on stored reserves of glycogen in the reproductive  129  female (Mann, 1979b; Gabbott, 1983; Racotta et al., 1998; Cigarría, 1999), but can also be enhanced through supplemental feeding (Utting, 1992, 1993). Essential fatty acids (EFAs), critical to membrane structure and function (Kraffe et al., 2004; Palacios et al., 2005), are not synthesized de novo in bivalves (Laing et al., 1990; Chu and Greaves, 1991) and therefore reproductive females must selectively accumulate EFAs into gonad tissue from exogenous food sources (Soudant et al., 1996a,b; Utting and Millican, 1997, 1998; Palacios et al., 2005). The most important EFAs are the omega-3 fatty acids eicosapentaenoic acid [20:5n-3 (EPA)] and docosahexaenoic acid [22:6n-3 (DHA)] (Dunstan et al., 1994; Hendriks et al., 2003; Kraffe et al., 2004; Palacios et al., 2005). Proper selection of feed (generally live algae) for broodstock can directly influence the relative EFA content in eggs (Caers et al., 1999a,b, 2002, 2003). In general, diatoms have high levels of EPA and flagellates have high levels of DHA (Dunstan et al., 1994; Brown et al., 1997). It is, therefore, recommended that broodstock conditioning use a mix of algal species which include those components (Utting and Millican, 1997; Marasigan and Laureta, 2001; Helm et al., 2004). In addition to the chemical content, algal species are also selected for their ease of culture, cell size, and digestibility (Webb and Chu, 1981; Coutteau and Sorgeloos, 1993). The purpose of this chapter was to determine how different phytoplankton diets (both mono- and bi-species diets) influence gonad and gamete development of P. generosa. A number of parameters were examined including gonadosomatic index, digestive gland index, gonad development index, oocyte diameter, gamete occupation indices (i.e. oocytes per unit area, oocyte area occupation, and spermatic material  130  occupation], and connective tissue occupation index. Also, the acceptability of the different diets by the animals was tested using clearance and ingestion rates.  5.2 Materials and methods 5.2.1 Algal culture Live algal cultures of Isochrysis sp. (CCMP 1324, TISO clone), Chaetoceros muelleri (CCMP 1316), and Dunaliella tertiolecta (CCMP 1320) were used as feeds. For culture methods see Chapter 2 (section 2.2.1). Three 50 ml aliquots per algal species were collected each week, rinsed with ammonium formate and processed for dry-wt as per Zhu and Lee (1997).  5.2.2 Broodstock collection, and initial maintenance Adult P. generosa were collected on February 8, 2009 as part of a commercial harvest from Thormanby Island, BC (49˚ 30.7’ N, 124˚ 1.73’ W) using a high-pressure hydraulic pump to extract the clams from the substratum. Water temperature at harvest depth (8–12 m) was 8˚C. After landing, the clams were transported to the Pacific Biological Station (Nanaimo, BC) within 8 h at an air temperature of 7˚C and held in ambient seawater (9.5˚C and salinity 28) in sea tables for 4–5 d until deployment to the replicate tanks. Whole wet weight and shell length were recorded for each clam (mean ± SD: 1525 ± 288 g and 160 ± 10 mm, respectively, n = 80). Clams were fitted with Ushaped, PVC-coated, wire-mesh girdles (mesh size: 25 × 25 mm; wire diameter: 3 mm; length × width: 34 × 8 cm) around the shells to prevent gaping and tears in the flesh.  131  After deployment to the replicate tanks the clams were starved for 20 d and then held on a maintenance diet of 4 × 109 cells clam-1 wk-1 (TISO:CM by cell count) for an additional 4 wk before the start of the experiment.  5.2.3 Broodstock conditioning: experimental design The experiment used a fully randomized design consisting of three uni-algal feed treatments [i.e. Isochrysis sp. (TISO), C. muelleri, and D. tertiolecta] and one bi-algal feed treatment (i.e. Isochrysis sp. and C. muelleri). All treatments were fed with a TISO dry weight equivalent of 4 × 109 cells clam-1 d-1. Clams were static (i.e. water flow turned off) fed for 4 h daily. There were four replicates of each treatment. Each replicate consisted of a clear plexiglass tank (L x W x H: 58 x 58 x 30 cm, Volume: 101 L) with a total of 16 tanks for all treatments. Each replicate tank contained five clams, placed on their side, and oriented with siphons toward the centre of the tank so that filtration would not be inhibited. Temperature for each tank was maintained at 11.0˚C by mixing heated and chilled seawater with a mixing chamber [overall average for the whole experiment was 11.0 ± 0.3˚C (mean ± SD, n = 3200). The seawater was sand filtered at the source and with a 5-µm bag filter as it entered each tank. Seawater flow rate to each tank was 127 ± 8 L h-1 (mean ± SD, n = 280). The photoperiod in the experiment was 16 light:8 dark, which replicated late spring/early summer conditions when natural spawning events occur (Sloan and Robinson, 1984). The tanks were covered with corrugated plastic lids to limit direct light and minimize algal fouling in the tanks. Light was still able to enter the tanks through the clear plexiglass sides (<100 lux at the exterior tank walls).  132  The health of each animal was assessed daily by examining for: lesions in the periostracum, flaccid or distended siphon, cracks in the shell, excessive mucus, and slow or absent reaction to a stimulus (see Chapter 2, Section 2.2.5). Clams with flaccid siphons that failed to show siphon contraction after stimulus were considered to be dead or moribund. These clams were removed from the experiment, but not replaced. Tank volumes were adjusted at feedings to maintain consistent cell concentrations among tanks and equal access to food for all clams in all treatments. Daily observations of the geoducks’ spawning activity were also made based on the presence or absence of eggs or sperm in the water. Individuals were observed for the release of gametes. The number of clams spawning per event (in each replicate) was recorded.  5.2.4 Sampling Clams were sampled after the acclimatization period (53 days) on days 0 (first day of full feeding), 25, and 47. On each sampling day one clam was taken from each tank for a total of four clams per treatment. For full sampling procedures see Chapter 2 (section 2.2.4). Live whole wet weights were recorded and then the visceral mass, somatic tissue, and shell were all separated, drained of excess water, blotted dry, and weighed. All weights were recorded to the nearest 0.1 g. Submerged weights of the live animal, shell, visceral mass, and somatic tissue were also measured using the methods described in Chapter 4 (section 4.2.7.1).  133  5.2.5 Clearance and ingestion rates Clearance rates (CR), ingestion rates by cell count (IRc) and ingestion rates by dry wt of algae (IRd) were determined at day 42 of the experiment. Algal-concentration samples of 20 ml were taken from the centre of each replicate tank (10cm depth; water column thoroughly mixed with a perforated plunger) at the beginning of the feeding cycle and at the end of the 4-h feeding. Clearance rate was calculated using the method of Quayle (1948) [as described in Coughlan (1969)] (see Chapter 3, section 3.2.9). Ingestion rate was calculated using the method described in Khalil (1996) (see Chapter 4, section 4.2.6). Ingestion rate of dry wt of algae was determined by multiplying the IR c by average algae dry wt of algae at the time of the trial.  5.2.6 Condition and gonadosomatic indices Shell, somatic tissue, and visceral mass were dried in an oven at 58–60˚C for 72 h (until constant weight). Soft tissues were cut into small pieces (1 × 1 cm) to facilitate drying. Since small portions of the visceral mass were removed for histological analysis a correction factor was necessary for the final visceral mass dry weight. This was done by multiplying the wet weight of the removed piece by the ratio of wet to dry weight of the remaining tissue. This value was added to the dry visceral mass weight to give the total dry weight.  5.2.6.1 Condition index Condition indices were determined as per Chapter 2, section 2.2.6. The condition index (CI) is based on the method of Walne and Mann (1975): CI = (T × 100)/S where CI  134  is the condition index, T is the total soft-tissue mass, and S is the shell mass. This was calculated for wet (CIw), dry (CId), and submerged (CIs) weights.  5.2.6.2 Gonadosomatic index For details see Chapter 2 (Section 2.2.6). Calculations were based on the method used in Sloan and Robinson (1984): GSI = V / T, where GSI = gonadosomatic index, V = visceral mass weight, and T = total soft-tissue weight without valves. This was calculated for wet (GSIw), dry (GSId), and submerged (GSIs) weights.  5.2.6.3 Digestive gland index A digestive gland index (DGI), similar to the gonadal area index (Barber et al., 1991), estimated the relative surface areas of digestive gland to total visceral mass. A transverse section was cut across the widest point of the visceral mass to expose the digestive gland (inner core) and gonad material (exterior sheath). From the posterior section the total surface area of the visceral mass and the surface area of the digestive gland were estimated (See Chapter 4, Section 4.2.7.3 for details). The DGI was calculated as a ratio of digestive gland surface area to visceral mass surface area. A high DGI indicates a relatively thin gonad layer.  5.2.7 Histological sampling Histological samples were collected as per Chapter 2, Section 2.2.8. A section of wet gonad tissue (i.e. posterior, right area of visceral mass) was removed, weighed, and  135  fixed in Davidson’s solution for 72 h. The fixed tissue samples were cut into 4 to 5 smaller sub-samples, dehydrated using a graded ethanol series, and embedded in paraffin wax. Embedded tissue sub-samples were sectioned to 5-μm thick, stained with hematoxylin-eosin, and mounted on a slide for examination. A minimum of 20 digital images (at both 40× and 100×) were captured from each slide (5–6 images from each tissue sub-section) using a Motic B5 compound microscope with Motic Images Advanced 3.2 software (Motic Electric Group Co., Ltd., Richmond, BC, Canada).  5.2.7.1 Development index Analysis of development index (DI) used the methods described in Chapter 2 (Section 2.2.8). At least two of the 40× magnification digital images from each subsample were randomly selected (i.e. a minimum of 10 images per clam). Ten randomlychosen follicles from each image were scored for development stage which made a minimum of 100 scores per clam. Follicle development was scored as: early active = 1, late active = 2, ripe = 3, partially spent = 4, and spent/resorbed = 5 (see Chapter 3, section 3.2.8.1). Follicles that were in intermediate states and could not be clearly categorized were given an intermediate score (to the nearest half score). A mean score was calculated for each clam.  5.2.7.2 Oocyte diameter Oocyte diameters were measured as per Chapter 2, Section 2.2.8. A minimum of 10 digital images (100× magnification) of each female gonad were used (i.e. at least two  136  images from each visceral mass sub-sample). Using the image analysis software, 5 to 10 randomly-selected oocyte diameters (OD) were measured from each image to the nearest 0.2 μm. A minimum total of 75 oocytes were measured for each clam and only oocytes with visible nuclei were used. An oocyte maturity index (OM), based on the proportion of mature (in the range of 30.1 to 40.0 μm in diameter) oocytes to total oocytes, was calculated for each female clam (see Chapter 2, section 2.2.8). A mean oocyte diameter was calculated for each clam.  5.2.7.3 Gamete occupation indices: oocytes per unit area, spermatic material occupation index, and connective tissue occupation index Counts of oocytes per unit area (OA), the surface area oocyte occupation index (OOI), spermatic material occupation index (SOI), and connective tissue occupation index (COI) were quantified using the image analysis software. The OA data was collected as per Chapter 2 (Section 2.2.8.). Oocytes per unit area, spermatic material occupation, and connective tissue occupation were measured colourimetrically as per Chapter 4, (Section 4.2.8). All parameters were determined from a minimum of 10 digital images per clam (minimum of two images per visceral mass sub-sample). Mean values were calculated for each clam.  5.2.8 Statistical analyses To test the influence of algal treatment on total cumulative mortality, a chi-square test for independence was used. For spawn events, the proportions of spawned to non-  137  spawned clams were analyzed using two-way repeated measures ANOVA (repeated measures to account for multiple spawns from individuals within tanks). The two fixed factors were algal treatment and time. Time included three periods: acclimatization phase (53 days), 0 to 25 d, and 26 to 47 d. Results are discussed as a percentage. Clearance rates and ingestion rates were analyzed with one-way ANOVA with the fixed factor algal treatment. The CI data were analyzed using ANCOVA with algal treatment and time as fixed factors and dry shell weight as the covariate. Two-way ANOVA was used to analyze GSI, DGI, DI, OD, OM, OA, OOI, SOI and COI with algal treatment and time as fixed factors. Normality of data was tested using the Shapiro Wilk W test and homogeneity of variances was tested using the modified Levene’s equal variance test. Deviations from assumptions were corrected using natural-logarithm (IRd, GSIw, GSId, GSIs DI, and COI) and arcsin (spawn proportions and OOI) transformations. No data transformations were necessary for CR, IRc, CIw, CId, CIs, DGI, SOI, OD, OA, or OM. Differences between treatments were determined with Tukey-Kramer multiple comparisons post-hoc tests (P<0.05). ANCOVA assumptions were examined with plots of residuals and the normal probability plot of residuals; no departures were found.  5.3 Results 5.3.1 Mortality rate and spawn events The cumulative mortality at the end of 47-d experiment did not differ significantly among algal treatments (χ2 = 0.536, df = 3, P > 0.75) with the overall mortality for all treatments for the 47 days being 25%. Clam deaths were generally the result of small lesions in the periostracum that became infected. Spawn events were sporadic and the 138  percentage of clams spawning did not differ significantly among algal treatments, time phases, or with the interaction between these two factors (F3,24 = 1.47, P > 0.10; F2,24 = 0.16, P > 0.50; F6,24 = 1.84, P > 0.10 respectively). Spawn events were characterized by all clams in a tank spawning simultaneously, but there was never more than one spawn event per tank over the duration of the entire experiment.  5.3.2 Clearance and ingestion rates Clearance rates were significantly influenced by algal treatment (Table 5.1). The CR of clams fed C. muelleri was significantly lower than that of clams given D. tertiolecta, but there were no other significant pair-wise comparisons among algal treatments (Fig. 5.1a). Ingestion rates were also significantly influenced by algal treatment (Table 5.1) with the TISO diet being ingested at significantly higher rates than the other three diets (Fig 5.1b). When adjusted for dry weight of algae, D. tertiolecta had by far the highest ingestion rates of any treatment (Table 5.1, Fig 5.1c). The mean dry weights of C. muelleri, D. tertiolecta and TISO were; (± SD, n = 3) 23.2 ± 1.8, 223.1 ±7.9 and 21.0 ± 1.6 pg cellˉ¹ respectively.  5.3.3 Condition and gonadosomatic indices There were no significant results for any of the CI parameters (wet, dry, or submerged) with respect to algal treatment, time, or the interaction between these two factors (Table 5.1, Fig. 5.2). The overall results over the duration of the experiment were (mean ± SE, n = 48): 288.6 ± 14.7%, 63.1 ± 3.6%, and 19.2 ± 1.1% for CI w, CId, and CIs, respectively (Fig. 5.2).  139  The GSIw was not significantly influenced by algal treatment, but did significantly change over time (with no significant interaction effect) (Table 5.1). The overall GSI w increased significantly from day 25 to day 47 with no other significant pair-wise comparisons between dates (Fig. 5.3a). Algal treatment, time, and the interaction between these two factors did not significantly affect GSId or GSIw (Table 5.1, Fig. 5.3b,c), indicating that the GSIw differences were the result of varying water content in the gonads. The DGI was not significantly influenced by algal treatment or the interaction between algal treatment and time, but was significantly affected by time (Table 5.1). There was a significant decrease in DGI from day 0 to day 47 with no other pair-wise differences between dates (Fig. 5.3d). This result indicates that the clams developed relatively more gonad material in the visceral mass over time.  5.3.4 Histological sampling 5.3.4.1 Development index The DI was not significantly influenced by algal treatment or the interaction between algal treatment and time, but was significantly affected by time (Table 5.1). The DI decreased significantly from day 0 to day 47 with no other pair-wise differences among dates (Fig. 5.4a). This indicates that there was a general trend toward a more ripe condition over time.  140  5.3.4.2 Oocyte diameter Diameters of oocytes remained stable throughout the experiment at 23.1 ± 0.6 μm (mean ± SE, n = 19) with no significant influence of algal treatment or time (Table 5.1, Fig. 5.4a). Likewise, the proportion of mature oocytes was not significantly influenced by algal treatment, time or interaction (F2,8 = 2.71, P > 0.10; F3,8 = 0.39, P > 0.10; F6,8 = 0.32, P > 0.50) with the average OM being 28 ± 1.7% (mean ± SE, n = 19) (data not shown).  5.3.4.3 Gamete occupation indices; oocytes per unit area, oocyte spermatic material and connective tissue occupation index The number of oocytes, as quantified by OA, was not significantly influenced by algal treatment or time (Table 5.1) with the overall average being 352 ± 36 oocytes mm -2 (mean ± SE, n = 19) (Fig 5.5a). Similarly, the relative areas within the gonads occupied by oocytes (OOI) or sperm (SOI) in females and males were not significantly influenced by algal treatment or time (Table 5.1) with the overall averages being 15.3 ± 1.9% (mean ± SE, n = 19) and 59.6 ± 6.0% (mean ± SE, n = 29), respectively (Fig 5.5b,c). The COI was also not significantly influenced by algal treatment, time, or the interaction between these two factors (Table 5.1) with the overall average being 30.6 ± 5.1% (mean ± SE, n = 48) (Fig 5.5d).  5.4 Discussion Overall, the parameters measured in this study showed no significant results with respect to algal treatment, except for CR and IRc; clams having significantly higher CR  141  and IRd with D. tertiolecta and significantly higher IRc with TISO compared to the other three phytoplankton treatments. The significantly higher CR with D. tertiolecta may have been related to the size of the various algal species. Cells of C. muelleri and TISO (≈ 4–5 μm) are much smaller than D. tertiolecta (≈ 12–15 μm) and efficiency of retention often increases with cell size [see review by Ward and Shumway (2004)]. Ruditapes decussatus for example retains approximately 80% of 4-5 μm particles but retains nearly 100% of particles between 6 and 10μm (no particles larger than 10 μm were tested) (Sobral and Widdows, 2000). The difference in IRc was more likely the result of feeding rates being standardized by TISO dry wt, meaning that the TISO treatments –having smaller cells- started with a proportionally higher cell concentration. The chain forming diatom, C. muelleri, was not ingested differently than TISO. Chain forming diatoms are often poorly ingested (Helm et al., 2004) as demonstrated in Pecten fumatus which cleared TISO, Chroomonas salina, and Pavlova lutheri (all flagellates) faster than Chaetoceros gracilis (a chain-forming diatom) (Heasman et al., 1996). Spawn events in the present study were consistent with the trickle spawning (low gamete release) behaviour previously noted in P. generosa (Goodwin and Pease, 1989) and did not correlate to any feed treatment. The lack of this correlation was not expected as increased spawning activity has been linked to more balanced feed combinations as seen in the angelwing clam, Pholas orientalis, which spawned two months earlier with a mixed feed of Chaetoceros calcitrans and Tetraselmis suecica (294 million cells animal-1 d-1) than when fed either species singly (Marasigan and Laureta, 2001). Presumably, differences in spawning activity are linked to EFA levels as demonstrated in Macoma balthica (Hendriks et al., 2003). In that study, the clams were fed a single species  142  (preserved Isochrysis sp.) but supplemented with EPA and DHA. The authors found that there was a higher percentage of spawners and higher fecundity when fed supplements (46.25% and 22,220 eggs clam-1, respectively) than without (16.25% and 963 eggs clam1  , respectively) over 24 days (Hendriks et al., 2003). The results in the current study were  more similar to those for C. gigas kumamoto where there was no difference in the rate of gametogenesis due to supplemental feeding (Robinson, 1992b). The advanced state of gonad development at the start of the experiment is likely a major factor in this outcome as the clams may have been relying more on endogenous reserves than exogenous food. Time was a significant factor with GSIw, DGI, and DI all indicating a general ripening of gonads over 47 d. Other experiments of comparable duration with broodstock bivalves had similar results with examples including Crassostrea virginica, Mercenaria mercenaria, and Ostrea edulis which were conditioned in 6 weeks (fed only TISO) (Gallager et al., 1986); Argopecten purpuratus with a GSI that peaked within 40–50 d (fed 50:50 Isochrysis galbana and C. gracilis) (Martínez et al., 2000b); C gigas that reached maximum egg diameter within 30–40 days of conditioning (fed TISO and C. calcitrans) (Chávez-Villalba et al., 2002); Venerupis philippinarum (formerly Tapes philippinarum) which went from an immature state to spawning within 6 weeks (fed S. costatum, D. tertiolecta, or T. suecica) (Laing and Lopez-Alvarado, 1994); and the Catarina scallop Argopecten ventricosus which showed gonad regeneration two weeks after spawning (increased GSI when fed TISO and C. muelleri) (Racotta et al., 1998). The response variables of CI, GSI, DGI, OD, OA, OOI, SOI, and COI did not vary significantly among algal treatments. Algal type can, however, affect CI as noted in another study where V. philippinarum had a higher CI when fed D. tertiolecta for 6  143  weeks compared to T. suecica (Laing and Lopez-Alvarado, 1994). The absence of differences in OD, OA, and OOI (averaging 23 µm, 352 oocytes mm-2, and 15.3%, respectively) among feed types in the present study was not surprising as egg diameters and numbers of eggs released in C. gigas fed D. tertiolecta (singly) for 7 or 8 weeks did not differ significantly compared to those fed a mixed algal diet of D. tertiolecta, Rhodomonas sp., and T. suecica (Caers et al., 2002). Feed type can influence egg size however, as Ostrea chilensis fed Isochrysis aff galbana had larger egg diameters than those fed C. gracilis or Pseudoisochrysis sp. (Wilson et al., 1996). This lack of differences among algal treatments may be at least partially related to the conversion of stored reserves from the time of collection into gonad material (Racotta et al., 1998; Cigarría, 1999). The broodstock were collected in February, which is a point in the year where GSI nears peak levels [approximately 17% wet weight; (Sloan and Robinson, 1984)]. Clams in the present study exceeded 18% at collection. Not only were the gonads at or near maximum size, but P. generosa are a massive bivalve species with large potential reserves in the muscular mantle and the siphon. Scallops are well known to use carbohydrate reserves in the muscle tissue for gametogenesis (Devauchelle and Mingant, 1991; Paon and Kenchington, 1995; Utting and Millican, 1998), but it is unknown if P. generosa have this capability. The somatic tissue could also act as a reserve of EFAs as well. In the bivalve Mactra chinensis, for example, the hepatopancrease, mantle, gill, and adductor muscle have DHA levels of 10.5, 18.7, 16.6, and 19.9% and EPA levels of 12.5, 6.8, 6.6, and 7.8% ,respectively (by percentage of total fatty acid) (Teshima et al., 1988). Although P. generosa may contain similar EFA levels in somatic tissues, it is unknown if they are capable of utilizing them as a reserve  144  for gametogenesis and therefore masking any potential differences between feed treatments. If gonad development came at the expense of stored reserves, however, we would expect lower condition index over time, a longer conditioning period, and an overall low percentage of ripe oocytes (Delgado and Pérez-Camacho, 2003). None of these patterns were seen. The present study tested D. tertiolecta to determine if the overall lack of EFAs in that species might negatively impact gonad development. Based on the response variables used in the study there were no differences in tissue maintenance or gonad development. (Laing and Lopez-Alvarado, 1994) also showed that D. tertiolecta is an adequate feed for tissue maintenance. It is not recommended, however, that D. tertiolecta be used as a single-feed species for broodstock conditioning as this could result in lower embryogenesis rates due to a deficiency of EFAs (Caers et al., 2002). Adding DHA and EPA to the diet of broodstock of A. purpuratus, via lipid emulsions, increased fatty acid levels in the gonad, reduced the period to the first spawn, and increased fecundity (Caers et al., 1999b, 2003). Similarly, broodstock of C. gigas conditioned with the EFA-poor algal species Dunaliella tertiolecta had nearly double the fecundity and more than double the D-larvae recovery rate when D. tertiolecta was supplemented with EPA and DHA emulsion spheres (Caers et al., 2002). It may, however, have a role in broodstock conditioning as a supplement to other feeds to increase the overall ingestion of organic material. Dunaliella tertiolecta contains 20, 15, and 12 pg cell-1 of protein, lipid, and carbohydrate respectively (see Chapter 1 review). High energy supplements can reduce recovery time after spawning (Wikfors et al., 2004) and increase the overall amount of gonad material (Albentosa et al., 2003).  145  In conclusion, algal treatments used in the present study – Isochrysis sp., D. tertiolecta, C. muelleri, and Isochrysis sp. + C. muelleri) – did not significantly differ in their effects on a variety of gonad and gamete attributes measured. Based on development index (DI) and the relative surface area of digestive gland (DGI) the gonads did ripen over time, independent of algal treatment. The fatty acid profiles of the gonads were not examined, however, and should be examined in future research. The results were in agreement with the alternative hypotheses (section 1.5.4) that feed type would influence clearance and ingestion rates, but the null hypothesis that ration significantly affects reproductive development could not be rejected.  146  Table 5.1 Results of ANOVAs (clearance and ingestion rates), ANCOVAs (Condition index) and 2-way ANOVAs for various attributes of experimental Panopea generosa. Sources of variation are algae treatment (A, fixed factor), C (covariate = shell weight; Condition index only), Time (T, fixed factor), interaction (A × T) and error. Values in bold are significant at α=0.05. Source A  df SS Clearance rate 3 1362.24  F ratio  P value  21.69  <0.001 (power = 0.99)  Error  8  A Error  Ingestion rate (dry wt algae) 3 51.73 43.30 8 3.18  C A T A× T Error  Condition index (wet tissue) 1 26.19 57.25 3 0.49 0.36 2 0.54 0.59 6 2.49 0.91 35 16.00  <0.0001 >0.5 (0.11) >0.5 (0.14) >0.5 (0.30)  C A T A× T Error  Condition index (submerged) 1 0.106 34.65 3 0.006 0.65 2 0.007 1.09 6 0.013 0.73 35 0.107  <0.0001 >0.5 (0.17) >0.2 (0.22) >0.5 (0.25)  A T A× T Error  Gonadosomatic index (wet tissue) 3 0.095 0.53 2 0.488 4.08 6 0.178 0.50 36 2.151  A T A× T Error  Gonadosomatic index (submerged) 3 0.03 0.13 2 0.33 2.32 6 0.38 0.92 36 2.54  A T A× T Error  Development index 3 0.01 2 0.56 6 0.22 36 2.65  0.05 3.80 0.50  >0.5 (0.05) <0.05 (0.65) >0.5 (0.18)  A T A× T Error  Oocytes per area 3 3523.7 2 35069.3 6 195024.9 8 104842.0  0.09 1.34 2.48  A T A× T Error  Sperm surface area 3 28.63 2 479.08 6 547.39 17 1085.63  0.15 3.75 1.43  137.49  df SS Ingestion rate (cells) 3 14.63 8  F ratio  P value  4.69  P<0.05 (power (0.7)  36.18 0.36 0.49 0.69  <0.0001 >0.5 (0.11) >0.5 (0.12) >0.5 (0.24)  8.33  <0.0001  Condition index (dry tissue) 1 1.30 3 0.04 2 0.04 6 0.15 35 1.26  Gonadosomatic index (dry tissue) >0.5 (0.15) <0.05 (0.68) >0.5 (0.18)  3 2 6 36  0.019 0.346 0.272 1.615  0.07 3.25 0.85  >0.5 (0.06) >0.05 (0.58) >0.5 (0.29)  Oocyte diameter 3 90.63 2 81.95 6 47.82 8 246.72  0.98 1.33 0.26  >0.2 (0.2) >0.2 (0.2) >0.5 (0.1)  >0.5 (0.06) >0.2 (0.21) >0.10 (0.50)  Oocyte surface area 3 12.33 2 44.91 6 91.33 8 180.75  0.18 0.99 0.67  >0.5 (0.07) >0.1 (0.16) >0.5 (0.15)  >0.5 (0.07) =0.05 (0.6) >0.2 (0.4)  Connective tissue occupation 3 0.071 0.22 2 0.250 1.15 6 0.153 0.23 36 3.923  >0.5 (0.09) >0.2 (0.23) >0.5 (0.10)  >0.5 (0.14) >0.10 (0. 61) >0.5 (0.18)  147  a. Clearance rate (l clamˉ¹hˉ¹)  35  A  30 25  B  B  T  TC  20 15 10  B  5 0  CM  D  (105 cells clamˉ¹hˉ¹)  Ingestion rate  b. 10 9 8 7 6 5 4 3  B  A A A  2 1 0  CM (106 pg dw algae clamˉ¹hrˉ¹)  Ingestion rate  c.  70  D  T  TC  A  60 50 40 30 B  20 B  B  10 0 CM  D  T  TC  Algal species  Figure 5.1 Mean (a) water volume cleared of algae (b) cells ingested and (c) dry wt algae ingested for Panopea generosa fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri). Error bars indicate SEM (n=4). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test).  148  a.  CM  D  T  TC  600  CI wet (%)  500 400 300 200 100  NS  0  b. 100  CI dry (%)  800 600 400 200 0  c. 350  CI sub (%)  300 250 200 150 100 50 0 0  25  47  Time (days)  Figure 5.2 Mean (a) wet, (b) dry, and (c) submerged condition indices (CI) of Panopea generosa fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri) over time (days). Error bars indicate SEM (n=4).  149  a.  CM  D  T  TC  35  GSI wet (%)  30 25 20 15 10 5 0  ab  a  b  b. 25  GSI dry (%)  20 15 10 5 0  c. 20  GSI sub (%)  15  10  5  0 d. 0.6 0.5  DGI  0.4 0.3 0.2 0.1 0.0  a  ab  b  0  25  47  Time (days)  Figure 5.3 Mean (a) wet, (b) dry, and (c) submerged gonadosomatic indices (GSI) and (d) ratio of visceral mass surface area covered by digestive gland (DGI) of Panopea generosa fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri) over time (days). Error bars indicate SEM (n=4). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods.  150  a.  CM  D  T  TC  5.0 4.5 Development index  4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0  a  ab  b  0  25  47  b. Oocyte diameter (μm)  35 30 25 20 15 10 5 0  Time (days)  Figure 5.4 Mean (a) developmental index and (b) oocyte diameter of Panopea generosa gonads fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri) over time (days). Error bars indicate SEM (n=4). Treatments denoted by different letters differ significantly (P<0.05, Tukey-Kramer test). Lower-case letters (below bars) apply to differences among time periods.  151  a. Oocyte area (oocytes mmˉ²)  CM  D  T  TC  600 500 400 300 200 100 0  b.  Oocyte occupation (%)  30 25 20 15 10 5 0  c. Sperm occupation (%)  90 80 70 60 50 40 30 20 10 0  d.  Connective tissue (%)  50 45 40 35 30 25 20 15 10 5 0  0  25  47  Time (days)  Figure 5.5 Mean (a) oocytes mm-2 and percentage of gonad surface areas occupied by (b) oocytes, (c) spermatic material, and (d) connective tissue in Panopea generosa fed different phytoplankton diets (CM = Chaetoceros muelleri, D = Dunaliella tertiolecta, T = Isochrysis sp. TISO clone, and TC = TISO and C. muelleri) over time (days). Error bars indicate SEM (n=4). 152  6 Effects of stocking density and algal feed ration on growth, survival, and ingestion rate of larval geoduck clams (Panopea generosa Gould, 1850)  The combined effects of initial stocking density (2, 5, 10 inds ml-1) and feed ration (5×103, 20×103, 40×103, 100×103 cells ind-1 d-1) on growth, survival, and ingestion rate of larval geoduck clams (Panopea generosa) fed Isochrysis sp. (Tahitian strain, TISO clone) were assessed. All three variables were significantly affected by stocking density, ration, and the interaction of these two factors. Growth rates ranged from 2.15 to 3.85 μm d-1 in the various treatments. Growth rate increased with stocking density at 5×10 3 cells ind-1 d1  , was significantly higher at 5 than at 2 inds ml-1 at 20×103 cells ind-1 d-1, and was  significantly higher at 2 and 5 than at 10 inds ml-1 at 40×103 and 100×103 cells ind-1 d-1. Growth rate increased with increasing ration up to 40×10 3 cells ind-1 d-1 for 2 and 5 inds ml-1, but decreased with increasing ration at 10 inds ml-1. Percent survival over the experimental duration (23 d) ranged from 7 to 56% in the various treatments. There was no significant affect of ration on percent survival at 2 and 5 inds ml-1, but survival decreased with increasing ration at 10 inds ml-1. Ingestion rate ranged from 2×103 to 29×103 cells ind-1 d-1 in the various treatments. Ingestion rate increased with ration at 2 and 5 inds ml-1, but was not significantly affected by ration at 10 inds ml-1.  6.1 Introduction Limited information is available on P. generosa larval rearing techniques in comparison to other commercially-important species in BC and Wa, such as the Manila clam (Venerupis philippinarum) (Toba et al., 1992; Helm et al., 2004) and Pacific oyster  153  (Crassostrea gigas) (Utting and Spencer, 1991; Thompson and Harrison, 1992; Laing and Earl, 1998). Some information is available through the pioneering work of Goodwin, who examined the stages of larval development (Goodwin et al., 1979) and the effects of temperature and salinity on embryogenesis (Goodwin, 1973). In the former study, there are anecdotal suggestions for appropriate stocking densities (i.e. 3 inds ml-1 is superior to 4–10 inds ml-1), but no actual data is provided to support the recommendations. These suggestions were expanded upon by Shaul (1981) who provided guidelines for reducing larval geoduck density with advancing larval development as well as advising an algal cell density of 30,000 to 50,000 cells ml-1. Shaul conceded, however, that the methods in the report were borrowed from guidelines for other species and may not be optimal for P. generosa. Since these early studies in the 1970s and 1980s there has been a dearth of published research with respect to geoduck larval rearing, despite the well established industry and the obvious need for such research. Bivalve hatchery success is characterized by high growth and survival rates, both of which depend on appropriate larval stocking densities (Deming and Russell, 1999; Yan et al., 2006) and rations (Riisgård, 1988; Pechenik et al., 1990; Beiras and PérezCamacho, 1994; Pérez-Camacho et al., 1994; Laing, 1995). As noted, this type of information is readily available for a number of bivalve species, but may not be directly transferable to P. generosa. For example, Manila clams perform well at relatively high densities of >10 inds ml-1 (Laing et al., 1990; Utting and Doyou, 1992; Laing and Utting, 1994; Yan et al., 2006) while C. gigas does well at densities ranging from 5–9 inds ml-1 (Utting, 1986; Ponis et al., 2006c; Ponis et al., 2006b; Ponis et al., 2006a) as does the scallop Pecten maximus (Delaunay et al., 1993; Ponis et al., 2006b). Released larvae of  154  the brooding oyster Ostrea edulis are typically reared at lower densities of <3 inds ml-1 (Helm et al., 1973; Holland and Spencer, 1973; Ferreiro et al., 1990; Jonsson et al., 1999). If inappropriate larval stocking densities are selected, sub-optimal performance can be expected. High densities, for example, may diminish water quality through the accumulation of metabolic wastes and reduced oxygen (Yan et al., 2006). High ammonia levels can inhibit normal development (Geffard et al., 2002) and low oxygen levels can result in reduced growth rates and premature settlement (Wang and Widdows, 1991). Unnecessarily low stocking densities, however, can limit potential production of a hatchery and be more costly. In addition to larval density, feed concentration can also affect feeding efficiency and growth rate of larval bivalves (Gallager, 1988). In a low feed environment, feeding efficiency is reduced due to low encounter rates between larvae and algal cells (MacKenzie and Leggett, 1991; Beiras and Pérez-Camacho, 1994) resulting in low growth rates and prolonged onset of metamorphosis (Robert et al., 1988; Pechenik et al., 1990; Tang et al., 2006). High algal concentrations, in contrast, can inhibit feeding by satiating the feeding apparatus and increasing rejection rates of algae (Gallager, 1988). Over-feeding can result in lower rates of growth (Loosanoff et al., 1953; Loosanoff and Davis, 1963). Excess algae can also create fouling that may lead to bacterial infestations and decreased survival (DiSalvo et al., 1978; Torkildsen et al., 2000; Helm et al., 2004; Torkildsen and Magnesen, 2004). The objective of the current study was to examine the combined effects of stocking density and algal ration on growth, survival, and ingestion rates of larval P. generosa, the ultimate goal being to determine an optimal combination (or combinations)  155  of stocking density and feed ration that optimizes growth and survival while maximizing potential larval/juvenile production  6.2 Materials and methods 6.2.1 Microalgal culture Larvae were fed Isochrysis sp. (Tahitian strain, TISO, CCMP 1324) which was grown semi-continuously in 20-l Nalgene® carboys at a temperature of 18.7±0.1˚C (mean±SD, n= 3200). Other algal species were unavailable for combination feeds due to culture crashes. Light was continuous and supplied by Philips DayLight Deluxe® fullspectrum fluorescent bulbs. Light levels were 2,600 lux immediately adjacent to the carboys. Seawater for algal culture was filtered to 0.2 μm, sterilized with sodium hypochlorite, neutralized with sodium thiosulfate, and fertilized with a formula modified from Harrison et al. (1980). The modification of the formula was a partial substitution of organic phosphates with inorganic phosphates. Cultures were aerated with a CO2-air mixture and salinity was 28–30. Algae in the exponential phase of growth were used for larval feeding  6.2.2 Larval culture and experimental design 6.2.2.1 Larval source Two experiments, one for growth and survival rates and one for ingestion rate, were run. Larvae for the experiments were spawned on March 12, 2008 at the Seed Science Ltd. hatchery facility in Campbell River, BC. Trochophore larvae were collected  156  at 12.8˚C and transported in oxygen-infused seawater within 2 h to the Pacific Biological Station in Nanaimo, BC on March 14, 2008.  6.2.2.2 Growth and survival rates experiment The experiment began on March 17, 2008 after all of the larvae reached D-stage. Mean (±SD) shell length (SL, anterior posterior axis) was 106.9±16.5 μm (n=105) at the start of the experiment. A sub-set of larvae was reared separately for the ingestion rate experiment (see Section 6.2.2.3). The experimental design for the growth and survival experiment was a fully factorial, completely randomized design which consisted of three larval stocking densities (2, 5, and 10 inds ml-1) and four feed rations (5×103, 20×103, 40×103, and 100×103 cells ind-1 d-1). Algal cell levels were presented on a per larva basis (throughout experiment) to ensure that each larva in each treatment had equal amounts of food available. Larval densities were determined at the beginning of the experiment by counting samples in a 1-ml Sedgewick Rafter cell and the larvae put into 1-l glass jars filled with 750 ml of seawater. Seawater for larval culture was filtered to 0.2 μm with cartridge filters and sterilized with ultraviolet light (Coralife Turbo-Twist® UV sterilizer, Oceanic Systems, Inc., Dallas, Texas, USA). Treated seawater was brought to experiment temperature in an incubator before use. Five replicate jars of each treatment pairing were randomly placed into a temperature-controlled water bath and rerandomized daily to minimize any potential location effects. Temperature was monitored with HOBO Water Temp Pro-v2® temperature recorders (Onset Computer Corporation, Pocasset, Massachusetts, USA) and maintained at 13.7±0.24˚C (mean±SD, n=11,000) throughout the experiment. Salinity was ambient and remained constant throughout the  157  experiment at 29, as measured with a Vee Gee STX-3® refractometer (Vee Gee Scientific Inc., Kirkland, Washington, USA). Both temperature and salinity levels were in the ranges determined by Goodwin (1973) to be necessary for successful embryogenesis (27.5 to 32.5 salinity and 10 to 14˚C). Light was on a 12 hour light:12 hour dark photoperiod and was kept dim (<10 lux at culture unit height) to inhibit algal growth during the experiment. Larvae were kept in the 1-l jars under static conditions (i.e. no flow and no aeration) and fed once a day with 24 h between feedings. Prior to feeding, the algae were put into a glass beaker and cooled to the experimental temperature in the water bath. This process did not appear to impact algal motility under microscopic examination. Before the algae were added to each jar, an amount of water equal to the volume of algal culture added was removed to maintain the correct water volume. After each 24-h feeding, the larvae were transferred to clean culture jars with fresh filtered/UV sterilized seawater. To maintain larval stocking densities through time, the numbers of live larvae remaining in each jar were estimated daily (using a Sedgewick Rafter cell) and the water volumes adjusted accordingly. Oxygen and un-ionized ammonia levels in each replicate jar were tested prior to each water change. Oxygen was tested with a YSI 550A meter (YSI, Yellow Springs, Ohio, USA) and ammonia with a Hanna HI93700 colorimeter (Hanna, Woonsocket, Rhode Island, USA) using the Nessler method. Daily observations were made at each water exchange for signs of mortalities (empty shells), predators (ciliates and nematodes), shell deformities, velum deformities, and fouling. If observed, these points were noted, but not quantified.  158  6.2.2.3 Ingestion rate experiment Ingestion rate was evaluated April 2, 2008 (16 days after receiving the larvae) using a completely randomized block design (four replicate blocks) using treatments and feeding procedures identical to those in the growth and survival experiment (see Section 2.2.2). Larvae from the original batch were set aside at the time of receipt and reared in a 19-l glass Pyrex® basin with light aeration. They were fed TISO daily at 60×103 cells ml1  with a water exchange every 2–3 d until the time of the experiment. Larvae were  starved for 24 h before the experiment began. Seawater for rearing the larvae was filtered and UV sterilized as in the previous experiment and temperature was maintained at 13.7°C by using a water bath. One replicate jar of each treatment pairing was placed in each of four areas (blocks) within the water bath. Average SL (±SD) at the beginning of the experiment was 152±18.6 μm (n=115). Phytoplankton sampling in each replicate jar occurred at 0 and 24 h. Prior to sampling, the water in the jars was homogenized with a perforated plunger and three 0.333 ml samples were taken with a pipette (through a 20 μm filter to remove larvae). Samples were preserved with Lugol’s iodine. Control jars, with identical cell concentrations to the treatment jars but with no larvae, were used to account for algal growth and death.  6.2.3 Data collection and analysis 6.2.3.1 Larval densities At the time of daily larval transfer, digital images were captured from each replicate jar at 2, 4, 10, 17, and 23 d after the onset of D-stage. To capture images, the  159  entire contents of each replicate jar were screened onto a 20 μm Nitex screen and the larvae rinsed with filtered/UV treated seawater into a glass dish (diameter: 65 mm, depth: 50 mm, surface area: 3,318 mm2). Larvae were evenly distributed across the bottom of the beaker using a perforated plunger. After being disturbed by the plunger, there was a 30–40 s period before the larvae began swimming again. During this inactive period images were captured from ten randomly pre-selected areas in the dish using a dissecting microscope and Motic Images Advanced 3.2® software (Motic Electric Group Co., Ltd., Richmond, British Columbia, Canada). The captured image field was 9.78 mm2 and water depth covering the larvae was standardized to 3 cm.  6.2.3.2 Growth and survival rates Shell length (anterior posterior axis) was measured using the Motic software described above. Twenty-five randomly selected living larvae (obvious gut and no shell deformities) were measured from each of the ten images per replicate per sampling period and these were averaged for each replicate jar (i.e. n=250 per replicate). On the last day of the experiment, after some treatments had suffered high mortalities, it was not possible to measure 25 larvae from every image. In those high-mortality treatments all of the living larvae in the image were measured (no replicate had less than 35 measurements in total). Growth rates for each replicate were calculated as the slope of the least squares linear regression, with shell length as a function of time. This method was used because there was no evidence for non-linearity of growth over time (2 to 23 d) (lack of fit: P>0.1) in any of the treatment pairs (see Fig. 6.1 and Table 6.1).  160  Survival rates were calculated by counting all of the living larvae in each of the ten images specific to a treatment replicate and sample day. Results from the ten images were averaged for each replicate jar. Only larvae with an obvious gut and no shell deformities were classified as living while those with deformities and empty shells were classified as dead. The total number of surviving larvae in each treatment replicate was calculated as: S = N/A × T S is the total estimated number of living larvae in the treatment replicate N is the average number of living larvae in each image field A is the surface area of the image field T is the total surface area of the dish bottom containing the larvae.  The result of S was used to calculate percent survival for each replicate on each sample day as a percentage of the living larvae at the start of the experiment.  6.2.3.3 Ingestion rates The number of cells ingested per larva was calculated as described in PerezCamacho et al., (1994): IR = V/nt [(c0–c1) – (c'0–c'1)] IR is the ingestion rate (cells ind-1 d-1) V is the volume (ml) n is the total number of larvae (inds replicate-1) t is time (d), c0 is the initial algal concentration  161  c1 is the final algal concentration c'0 is the initial algal concentration of the control c'1 is the final algal concentration of the control (all concentrations in cells ml-1)  6.2.3.4 Statistical analysis Normality and homogeneity of the data were tested with the Shapiro-Wilk W test and modified Levene’s test, respectively, using NCSS 2007 (NCSS, Kaysville, Utah, USA). No data transformations were necessary for growth rate data. Survival percentages were transformed with the arcsine transformation. Deviations from normality in ingestion rate data were corrected with natural-log transformation. Growth and survival rates were analyzed with 2-way ANOVA while ingestion rates were analyzed with 2-way randomized block ANOVA. Differences among treatments were tested using the Tukey post-hoc multiple comparisons test.  6.3 Results 6.3.1 Growth and survival rates Growth rates across 23 d of larval life ranged from 2.15 (10 inds ml-1, 100×103 cells ind-1 d-1) to 3.85 (5 inds ml-1, 40×103 cells ind-1 d-1) μm d-1 (Table 6.1, Fig. 6.1). Two-way ANOVA results indicated that growth rates were significantly influenced by larval density and feed ration with a strong interaction between the two factors (Table 6.2). The strong interaction resulted from the decrease in growth rates concomitant with increasing ration in the 10 inds ml-1 density treatment [this was confirmed by re-  162  analyzing the dataset with the 10 inds ml-1 data removed, which showed that there was no significant interaction between larval density and ration (F6,32 = 0.27, P > 0.75)]. In contrast to the 10 inds ml-1 treatment that showed decreasing growth rate with increasing algal concentration, the 2 and 5 inds ml-1 density treatments had increasing growth rate with increasing ration up to 40×103 cells ind-1 d-1, with a decrease (significant for 5 inds ml-1) at 100×103 cells ind-1 d-1 (Fig. 6.2). Growth rate was significantly higher at 10 than at 2 or 5 inds ml-1 at 5x103 cells ind-1 d-1, statistically the same at 20x103 cells ind-1 d-1, and significantly lower at 40 and 100x103 cells ind-1 d-1 (Fig. 6.2). Up to day 10, percent survival was very high (in excess of 98% in all treatments, data not shown) and not significantly affected by ration or larval density. By day 17, the effects of larval density and ration had become significant with a significant interaction between the two factors (Table 6.2). At this time, all treatments had >97% survival except the highest larval density and algal cell level pairing (10 inds ml-1, 100×103 cells ind-1 d-1) that averaged only 87% survival and was significantly lower than all other treatments (Tukey’s test, data not shown). By day 23, mortalities began to accumulate in all treatments with percent survival ranging from 7% (10 inds ml-1, 100×103 cells ind-1 d1  ) to 56% (2 inds ml-1, 20×103 cells ind-1 d-1) (Fig. 6.3). A two-way ANOVA on the day-  23 data revealed that percent survival was significantly affected by both larval density and ration with a significant interaction between the two factors (Table 6.2). Percent survival in the 2 and 5 inds ml-1 treatments was not significantly impacted by changes in ration (Fig. 6.3). In contrast, percent survival in the 10 inds ml-1 treatment declined significantly with increasing ration (Fig. 6.3). Percent survival generally decreased with increasing stocking density at all rations (Fig. 6.3). Oxygen and ammonia levels were not  163  significantly affected by density or ration with oxygen concentration always in excess of 80% saturation and ammonia concentration never higher than 0.08 ppm (3-way ANOVAs, P > 0.25 for factors of density, ration, time, and all interactions).  6.3.2 Ingestion rates Ingestion rate ranged from 2×103 (5 inds ml-1, 5×103 cells ind-1 d-1) to 29×103 (2 inds ml-1, 100×103 cells ind-1 d-1) cells ind-1 d-1 (Fig. 6.4). Two-way ANOVA results showed that ingestion rate was significantly affected by both larval density and ration with a significant interaction between the two factors (Table 6.2). Ingestion rates increased with increasing ration at stocking densities of 2 and 5 inds ml-1 while ingestion rates remained unchanged in relation to ration at 10 inds ml-1 (Fig. 6.4). At the two lower rations of 5 and 20x103 cells ind-1 d-1 there were no significant differences among the three larval densities in terms of ingestion rate (Fig. 6.4). At the two higher rations of 40 and 100×103 cells ind-1 d-1, however, 2 and 5 inds ml-1were not significantly different from each other but both were significantly higher than 10 inds ml-1 (Fig. 6.4)  6.4 Discussion Results of the current study showed that growth, survival, and ingestion rates of P. generosa larvae were significantly affected by larval density, ration, and the interaction between these two factors. Maximum growth rate in the present study (3.85 μm d-1 at 13.7˚C) was lower than that reported by Goodwin et al. (1979) for larvae of the same species (5.7 μm d-1 at 14˚C) and far lower than that reported by Gribben and Hay  164  (2003) for larvae of P. zelandica (8.9 μm d-1 at 17˚C). The growth rate difference with P. zelandica may be due to species and temperature differences, but the discrepancy with the Goodwin et al. (1979) study may relate to the nutritional value of the feed; Goodwin et al. (1979) used Monochrysis lutheri, Isochrysis galbana, Pseudoisochrysis paradoxa, Phaeodactylum tricornutum (singly or mixed at 50×103 cells-1 ml-1) while the current study used TISO alone. Although TISO is generally considered to be a nutritious species and an effective monospecific feed for the scallop Pecten maximus (Delaunay et al., 1993) and the clams Meretrix meretrix (Tang et al., 2006), Mercenaria mercenaria, and Tapes semidecussata (Helm and Laing, 1987), it has shown poor results when fed alone to the oysters Crassostrea gigas and C. rhizophorae (Helm and Laing, 1987). This is likely because TISO is low in the essential fatty acid eicosapentaenoic acid (EPA) (Nevejan et al., 2003a; Martínez-Fernández et al., 2006; Patil et al., 2007; RiveroRodríguez et al., 2007), which may be necessary for maximum growth and survival of larvae of certain species of bivalves (Thompson and Harrison, 1992; Berntsson et al., 1997; Pernet et al., 2003b; Ponis et al., 2006b). The monospecific diet may also have contributed to the rapid decrease in survival by day 23, as high mortalities after day 21 were also reported in M. mercenaria which was fed monospecifically with Isochrysis sp. (TISO and CISO strains) (Deming and Russell, 1999). Future research on larval geoducks should examine the effects of various algal species combinations on growth and survival. The interactions noted in this experiment are very important. Based on the results from other clam species we might expect that low density treatments always have better growth rates. Mercenaria mercenaria larvae, for example, grow better at 4 inds ml-1  165  compared to 20 inds ml-1 when fed 70,000 cells ml-1 of TISO twice daily (Deming and Russell, 1999) and larvae of Ruditapes philippinarum grow better at 5–10 inds ml-1 compared to 15 and 20 inds ml-1 when fed daily with a maximum of 24,000 cells ml-1 of Isochrysis spp. (Yan et al., 2006). The present study, however, showed that higher larval density treatments can sometimes have growth and survival rates equal to (or better than) lower density treatments, provided that the appropriate ration is provided. For example, larvae in the 10 inds ml-1, 5×103 cells ind-1 d-1 treatment grew significantly better and survived as well as those in the 2 and 5 inds ml-1 densities at the same ration level. Pechenick et al. (1990) suggested that this interaction phenomenon could account for the disparity in larval M. edulis growth rate results from various studies and the modelling work in Powell et al. (2002) indicated that matching a stocking density to its correct ration level could dramatically improve larval growth rates. A likely factor explaining why larval growth rates depend so strongly on ration is encounter rate between larvae and algae. Low food availability causes the larvae to expend more energy searching for food, while at the same time ingesting less. This can lead to a depletion of endogenous biochemical reserves (Millar and Scott, 1967; Holland and Spencer, 1973) and ultimately reduced growth (His et al., 1989; Ponis et al., 2003b; Tang et al., 2006). It is important to note that in the design of the present study, the number of cells per larva was fixed so as the density of larvae increased, the concentration of algae (cells ml-1) increased proportionally. Treatments with higher cell concentrations would necessarily have higher encounter rates than those with lower cell concentrations. Evidence for this comes from a comparison of the three densities at 5×103 cells ind-1 d-1. Treatments of 2 and 5 inds ml-1 at this ration had the lowest algal  166  concentrations of 10,000 and 25,000 cells ml-1, respectively, and among the lowest growth rates (2.27 and 2.80 μm d -1, respectively). In contrast, the 10 inds ml-1 density, despite having an equal number of cells ind -1 available, had a much higher cell concentration (50,000 cells ml-1) and among the highest growth rates of all treatments (3.75 μm d-1). Moreover, at 5×103 cells ind-1 d-1, the 10 inds ml-1 density had among the lowest ingestion rates while sustaining among the highest growth rates, an indication of energetically efficient grazing. The top five treatments in terms of growth rate all had algal concentrations in the range of 50,000 to 200,000 cells ml-1. This is comparable to the cell concentrations identified for high growth rate of larvae of O. edulis (100,000 to 200,000 cells ml-1 of I. galbana) (Beiras and Pérez-Camacho, 1994), P. maximus (100,000 cells ml-1 of Chaetoceros calcitrans f. pumilum and Pavlova sp. AC 250) (Ponis et al., 2006b), V. philippinarum (200,000 cells ml-1 of C. calcitrans) (Utting and Doyou, 1992), and C. gigas (100,000 cells ml-1 of TISO and C. calcitrans f. pumilum) (Brown and Robert, 2002). Treatments with 400,000 cells ml-1 or more (i.e. 5 inds ml-1 at 100×103 cells ind-1 d-1, 10 inds ml-1 at 40×103 cells ind-1 d-1, and 10 inds ml-1 at 100×103 cells ind-1 d-1) all had reduced growth rate. Reduced growth rates can result from excessive feed levels (Loosanoff et al., 1953) likely because high encounter rates with algae result in rapid saturation of the gut and increased particle rejection (Gallager, 1988). Increases in larval density (beyond 1 ind ml-1) have also been shown to reduce ingestion and filtration rates in Japanese scallop (Patinopecten yessoensis) larvae (MacDonald, 1988). The overall higher survival rates at 2 compared to 5 inds ml-1 supports the assertion by both Goodwin et al. (1979) and Shaul (1981) that a reduction in stocking  167  density with increasing body size is necessary. Shaul (1981) provided anecdotal guidelines for stocking density reductions of 10 inds ml-1 at the straight-hinge stage to 3– 5 inds ml-1 at 200–300 μm shell length to 0.3 inds ml-1 at setting. The mean (±SE) shell length of the fastest growing treatments (i.e. 2 inds ml-1 at 40×103 cells ind-1 d-1, 5 inds ml-1 at 20×103 cells ind-1 d-1, 5 inds ml-1 at 40×103 cells ind-1 d-1, and 10 inds ml-1 at 5×103 cells ind-1 d-1) at day 17 in the present study was 160 ± 2.1 μm (n = 20), indicating that this is the appropriate time to begin density reductions if initial larval density is 5 inds ml-1 or higher. One of the more pronounced trends in survival was its significant decrease with increasing ration level at 10 inds ml-1. It has long been known that overfeeding of larval bivalve cultures can result in high mortality (Loosanoff and Davis, 1963). Opportunistic bacteria and/or protozoans may have played a role in this survival trend. Larvae of Pecten maximus grown in stagnant batch cultures without antibiotic treatments, as in the present study, were subject to mass mortalities, presumably due to bacterial proliferation associated with algal accumulation (Torkildsen and Magnesen, 2004). This was also reported anecdotally with P. generosa larvae (Shaul, 1981). Although daily water changes with filtered, UV-sterilized seawater were done to minimize bacterial colonization, an unidentified filamentous alga – which began to grow after day 18 in the treatments with 10 inds ml-1 at 40 and 100×103 cells ind-1 d-1 (which had the lowest survival rates) – potentially acted as a substrate for bacteria. This alga acted as a harbourage for protozoans. Neither this alga nor the protozoans were evident in any other treatments. The algal filaments could be removed by hand, but segments often broke off and it could not be effectively separated from the larvae. The protozoans were  168  unidentified, but they did not appear to be the Isonema-like flagellate reported to infect P. generosa larvae (Kent et al., 1987). The larvae themselves were not examined histologically for parasitic infestation. The highest ingestion rates were 28,961 and 27,212 cells ind -1 d-1 at 2 and 5 inds ml-1, respectively, at 100×103 cells ind-1 d-1, similar to those of C. gigas larvae which removed up to 26,000 cells of I. galbana ind-1 d-1 (Gerdes, 1983). Ingestion rates showed a linear increase with ration level at 2 and 5 inds m-1 (linear regression results: F1,14 = 40.6, p <0.0001, r2 = 0.74 and F1,14 = 41.5, p <0.0001, r2 = 0.75, respectively), but remained constant across ration level at 10 inds ml-1 (F1,14 = 0.90, p > 0.25, r2 = 0.06). These results suggest a functional response; increased ingestion rates in direct response to increased feed concentrations (Baldwin and Newell, 1995). This is a common phenomenon and widely reported among various bivalve species including C. gigas (Utting, 1986; Brown and Robert, 2002), Crassostrea virginica (Baldwin and Newell, 1995), Ruditapes decussatus (Pérez-Camacho et al., 1994), M. mercenaria (Gallager, 1988; Riisgård, 1988), and O. edulis (Beiras and Pérez-Camacho, 1994). The low and stable ingestion rates within all 10 inds ml-1 treatments was likely the result of gut saturation (Baldwin and Newell, 1995) due to the high cell concentrations (i.e. 400,000 cells ml-1 at 40×103 cells ind-1 d-1 and 1,000,000 cells ml-1 at 100×103 cells ind-1 d-1). It must be noted that ingestion rates change allometrically with body size and decline near metamorphosis (Gerdes, 1983), so the ingestion rates shown in the present study cannot be considered accurate over the entire larval period. In conclusion, growth and survival rates of P. generosa larvae are dependent on both stocking density and ration. Due to the strong interaction between these two factors  169  care must be taken not to over feed larvae at high stocking densities and cell concentrations (TISO equivalent) should not exceed 200×103 cells ml-1. Overall, the best treatment with respect to efficient use of space and feed was 10 inds ml-1 fed 5×103 cells ind-1 d-1, which had the best combination of growth and survival with the lowest required amount of algae per larva. Survival rates in general were highest at a larval density of 2 inds ml-1, which indicates that reducing larval density as the larvae exceed 160 μm in shell length (point at which reduced survival rate began in the present study) may be advisable. Future research on larval geoducks should include efforts to further refine density specific rations. It must also be considered that a monospecific feed was used in the present study and further research should investigate the effects of various algal species monospecifically and in combination on the growth and survival of P. generosa larvae. The results were in agreement with the alternative hypotheses (section 1.5.5) that growth, survival and ingestion would be significantly influenced by the interaction of larval stocking density and ration.  170  Table 6.1 Linear growth models for larvae of Panopea generosa for each treatment pairing of larval density and algal ration up to 23 days after reaching D-stage. Larval density / Algal ration R2 F1,23 Intercept Slope (inds ml-1 / cells ind-1 d-1) 2 / 5×103  0.81  100.3  106.7  2.27  2 / 20×103  0.87  158.4  111.2  2.92  2 / 40×103  0.90  212.7  108.3  3.37  2 / 100 ×103  0.89  185.9  109.7  2.80  5 / 5×103  0.91  232.0  108.5  2.80  5 / 20×103  0.95  418.8  108.4  3.55  5 / 40×103  0.93  330.5  104.9  3.85  5 / 100×103  0.87  154.5  102.3  3.10  10 / 5×103  0.95  479.6  102.8  3.75  10 / 20×103  0.93  285.2  104.4  3.37  10 / 40×103  0.87  150.5  102.1  2.63  10 / 100×103  0.93  112.9  101.9  2.15  All slopes were significant at P < 0.005. No regressions had evidence for lack of fit (P > 0.1).  171  Table 6.2 Results of separate two-way ANOVAs on various attributes of experimental Panopea generosa larvae. Sources of variation are larval stocking density (D, fixed factor), ration (R, fixed factor) and interaction (D×R). Values in bold are significant at P<0.05 or less. Source df F ratio P value Growth rate D 2,48 7.94 < 0.0025 R 3,48 7.97 < 0.0005 D ×R 6,48 10.94 < 0.0005  D R D ×R  Survival (day 17) 2,48 14.21 3,48 7.85 6,48 7.72  < 0.0001 < 0.0005 < 0.0001  D R D ×R  Survival (day 23) 2,48 51.55 3,48 9.34 6,48 4.95  <0.0001 <0.0001 <0.001  D R D ×R  Ingestion rate 2,18 3,18 6,18  < 0.005 < 0.0001 < 0.0001  16.51 41.15 15.71  172  a. 2 inds mlˉ¹  Shell length (μm)  200  5×10³ cells indˉ¹dˉ¹  10×10³ cells indˉ¹dˉ¹  40×10³ cells indˉ¹dˉ¹  100×10³ cells indˉ¹dˉ¹  180 160 140 120 100  b. 5 inds mlˉ¹  Shell length (μm)  200 180 160 140 120 100  c. 10 c. 10 inds inds mlˉ¹ mlˉ¹  Shell length (μm)  200 180 160 140 120 100 2  6  10  14  18  22  Time (days)  Figure 6.1 Growth of Panopea generosa larvae in response to larval stocking density (a: 2 inds ml-1, b: 5 inds ml-1, and c: 10 inds ml-1) and algal ration (5, 10, 20, and 100×103 cells ind-1 d-1). Error bars are SEM and n = 5.  173  4.3  B  2 inds mlˉ¹ 5 inds mlˉ¹  B B  10 inds mlˉ¹ B  AB  -1  Growth rate (μm d )  3.8  3.3  B  A A  B A  2.8 A  A  2.3  1.8  a  a  5  b  ab bc  b  b  20  c  40 3  a  ab  ab  a  100  -1 -1  Ration (10 cells ind d )  Figure 6.2 Growth rates in shell length of Panopea generosa larvae in response to larval stocking density (2, 5, and 10 inds ml-1) and algal ration (5, 10, 20, and 100×103 cells ind1 -1 d ). Different letters represent significant differences among the treatments (Tukey test, P < 0.05). Upper-case letters signify the results of the effect of larval density within each feed ration level and lower-case letters signify the results of feed ration level within each larval density. Error bars are SEM and n = 5.  174  72  2 inds mlˉ¹ 5 inds mlˉ¹  B 62  10 inds mlˉ¹ B AB  Survival (%)  52  B  42  C B  A A  B  32 A 22 A 12 2  A a  a  5  c  a  a  b  a  20  a  ab  40 3  a  a  a  100  -1 -1  Ration (10 cells ind d )  Figure 6.3 Percent survival 23 d after D-stage of Panopea generosa larvae in response to larval stocking density (2, 5, and 10 inds ml-1) and algal ration (5, 10, 20, and 100×103 cells ind-1 d-1). Different letters represent significant differences among the treatments (Tukey test, P < 0.05). Upper-case letters signify the results of the effect of larval density within each feed ration level and lower-case letters signify the results of feed ration level within each larval density. Error bars are SEM and n = 5.  175  40  Ingestion (103 cells ind -1 d-1)  35  B 2 inds mlˉ¹  30  B  5 inds mlˉ¹ 10 inds mlˉ¹  25  B  20 B  A 15 A  A 10 5 0  A  A  A  A  A a  a  5  a  a  a  a  a  20  b  a  40 3  b  b  a  100  -1 -1  Ration (10 cells ind d )  Figure 6.4 Ingestion rates of Panopea generosa larvae in response to larval stocking density (2, 5, and 10 inds ml-1) and algal ration (5, 10, 20, and 100×103 cells ind-1 d-1). Different letters represent significant differences among the treatments (Tukey test, P < 0.05). Upper-case letters signify the results of the effect of larval density within each feed ration level and lower-case letters signify the results of feed ration level within each larval density. Error bars are SEM and n = 5.  176  7 Conclusions 7.1 Thesis background The responses of adult Panopea generosa to various levels of physical (i.e. temperature and salinity) and biological (i.e. ration and food type) factors were examined with the objective of determining appropriate levels for broodstock conditioning (i.e. gonad development). Proportions of spawning clams were monitored and reproductive development was assessed using a series of response variables including: gonadosomatic index, organic content [or ash-free dry weight (AFDW)], development index, oocyte maturity, gamete occupation, and connective tissue occupation. Another objective was to identify appropriate rations and stocking densities for larval rearing. The effects of these factors were assessed by examining larval growth and survival. The following chapter summarizes the results of my PhD work, places them into context with existing literature, addresses strengths and limitations of the research, provides recommendations for P. generosa aquaculture, and suggests potential areas for future experimentation.  7.2 Chapter 2: Effect of temperature on broodstock conditioning Chapter 2 examined the effects of temperature (7, 11, 15, and 19˚C) on reproductive development of adult P. generosa over time (155 days). Temperatures were selected to span the range typically encountered by geoducks in British Columbia (BC) coastal waters and gonad development was monitored over a time frame relevant to hatchery operations. The null hypothesis for this experiment – that temperature would have no effect on spawning and reproductive development – was rejected. The highest  177  proportion of spawning clams occurred at the temperature closest to that typically experienced under natural spawning conditions (11˚C) in agreement with the alternative hypothesis. Also, as predicted, gonadosomatic index was highest at the lowest tested temperature (7ºC) and gonads degenerated at higher temperatures. Degeneration was evident at 19˚C with fewer mature individuals (compared to 7 and 11˚C), a lower percentage of mature oocytes (compared to 7˚C at 155d), and increased connective tissue occupation (compared to 7 and 11ºC at 155d). Both 15 and 19˚C produced fewer oocytes per follicle than 7 and 11˚C, also in agreement with the alternative hypothesis. In the case of AFDW, however, the result was the opposite of what was stated in the alternative hypothesis with clams at 7˚C having a lower AFDW than those at 19˚C. Prior to this research, appropriate temperatures for broodstock conditioning for P. generosa had not been examined. Goodwin et al. (1979) held geoducks at 9–10ºC prior to spawning, but never determined an optimal temperature for broodstock conditioning. The present results provide much needed information for the successful hatchery conditioning of geoduck clams. The most important result was the identification of 11ºC as the temperature that produced the most spawns while maintaining relatively high numbers of oocytes. This temperature is lower than for other commercial species currently produced in British Columbia, such as Crassostrea gigas and Venerupis philippinarum which have optimal gonad development at 18–24 (Mann, 1979b) and 18– 21˚C (Mann, 1979a), respectively. Conditioning P. generosa using standard hatchery temperatures for those species would result in gonad degeneration. The result at 7ºC – clams with prolonged mature gonads and reduced spawning – was consistent with other bivalve species when held below the specific trigger temperature for spawning (Mann,  178  1979a; Santos et al., 1993). The high temperature results showing gonad degeneration were also consistent with other bivalves (Heasman et al., 1996; Martínez et al., 2000a; Chávez-Villalba et al., 2002) (Villalejo-Fuerte et al., 1996). The various histological techniques used in the present study allowed for a variety of responses to be examined, lending strength to the conclusions about gonad development. A limitation of the study was that it assessed the proportion of spawners and not the numbers of gametes released, information that is also useful. This information was difficult to collect, however, because the flow-through seawater system flushed the majority of gametes to waste before they could be quantified. Future studies could benefit from holding individuals in larger containers that can be more effectively isolated when spawning begins, thus allowing for gamete quantification. Larval viability was also not confirmed in the present study, but would be useful information to collect in future experimentation. Findings of this chapter have immediate application to hatchery operations. If a hatchery operator wants to prolong the gamete production period and limit gonad degeneration, 11ºC is the most appropriate temperature. However, 7ºC can be used to ripen and maintain gonads while limiting spawning. Temperatures above 15ºC should be avoided for broodstock conditioning. For research purposes, holding P. generosa for >100 d at 19ºC, prior to the start of an experiment, can synchronize the animals in a resorbed state, allowing for gametogenesis to be tracked from the earliest stages. Future directions of research based on temperature might involve refinement of the optimal temperature for broodstock conditioning. As the first study on this topic I used fairly wide increments of 4ºC between treatment levels to capture upper and lower  179  limits. Future research should narrow the range (e.g. 9–14ºC) and increments between treatments (2ºC or less). Further research should also address the scope for growth of P. generosa under various temperature regimes. In a hatchery it is unlikely that temperatures can be maintained within the desired range at all times. Recent work by Delgado and Pérez-Camacho (2007) with V. philippinarum, however, suggested that changes in ingestion rates due to temperature can be compensated by adjusting feed concentrations, thereby maintaining a positive scope for growth. This may also be true for P. generosa, but research is required to examine this possibility. Research on scope for growth will require data on energy ingested, energy excreted (i.e. faeces, pseudofaeces, and ammonia) and oxygen consumed at various temperatures. Salinity can also affect oxygen consumption (as determined in Chapter 3) so examining the interactive effects of salinity and temperature on scope for growth may be warranted. Another potential area of future research is the use of low temperatures (e.g. 7˚C) to hold ripened broodstock. Such a technique has the potential to allow for mature broodstock to be held until needed. This technique is already used in commercial hatcheries with species other than geoducks (R. Marshall, pers. obs.) and, at least in the past, broodstock have been moved from warm water regions to cool water regions to delay spawning (Loosanoff and Davis, 1963). Gamete and larval viability resulting from this technique is unknown and should be investigated.  7.3 Chapter 3: Effect of salinity on broodstock conditioning Chapter 3 examined the effect of salinity (17, 20, 24, and 29) on survival, reproductive development, oxygen consumption, and clearance rate of adult P. generosa  180  over time (62 d). Salinities typical of BC estuaries were selected. The null hypotheses were that salinity would have no effect on survival, condition index, tissue water content, reproductive development, oxygen consumption or clearance rates. The null hypotheses that salinity would have no effect on survival or tissue water content were rejected. Lower salinities increased mortality rate, as hypothesized, with salinities of 17 and 20 producing higher mortality rates than 24 and 29. Survival was not significantly different, however, between the treatments of 24 and 29. Also, as predicted, there was an increase in water content in the tissues at salinities of 17 and 20. Reproductive development was affected by salinity, in agreement with the alternative hypothesis (it should be noted, however, that this analysis excluded salinities of 17 and 20 due to high mortality rates). Clams at a salinity of 24 had a lower gonadosomatic index (based on wet weight), no gonad development beyond the early-active stage, reduced gamete occupation, and a thinner gonad sheath in comparison to clams held at a salinity of 29. The null hypotheses with respect to organic content and connective tissue occupation could not be rejected. Oxygen consumption rates at salinities of 17, 20, and 24 were higher than at 29 and clearance rates were lower at salinities of 17 and 20 compared to 24 and 29. Both of these findings were in agreement with the alternative hypotheses. Prior to this research the effects of salinity on adult geoduck clams were unexplored. The only salinity related information available was one report concerning embryogenesis (Goodwin, 1973) and a conference abstract concerning juvenile burrowing behaviour (Davis and Barenberg, 2000b). The results of Chapter 3 provide important new information relevant to aquaculture hatchery management and site selection. A salinity of 24, a level that might be expected in a hatchery during periods of  181  freshwater run-off, inhibited gonad development. A possible explanation is that osmoregulation in response to hyposaline conditions has metabolic costs (Stickle and Sabourin, 1979; Kim et al., 2001; Hamer et al., 2008). Elevated oxygen consumption at the lower salinities supports this explanation. Hyposalinity tolerance of P. generosa proved to be relatively low in comparison to other clam species found in the low intertidal waters of BC such as Mya arenaria (Shumway, 1977) and V. philippinarum (Elston et al., 2003). This finding is extremely important since introducing P. generosa to existing shellfish farms based on the success of other species, without consideration for local salinity regimes, could prove ineffective as prolonged exposure to low salinity can severely reduce long-term bivalve survival (Shurova, 2001)(McLeod and Wing, 2008) (Roldán-Carrillo et al., 2005; Rupp et al., 2005). Panopea generosa does appear to have a higher tolerance to low salinity than subtidal scallops which die within days at salinities below 23 (Roldán-Carrillo et al., 2005)(Rupp and Parsons, 2004)(Izumi et al., 2000). This study had a number of strengths including the testing of a wide range of salinities typical of the seasonal fluctuations in BC estuaries, histological techniques that allowed for a variety of responses to be measured, and a flow-through conditioning tube system that allowed for oxygen consumption data to be collected from a large number of individuals with relative ease. This study also began at the end of the spawning season when gonads are resorbed, allowing for development to be tracked from the beginning of gametogenesis. A limitation of this study was that increments between treatments did not fully capture the lower levels of tolerance. Osmoregulation of the clams broke down somewhere between salinities of 20 and 24 and reproduction became inhibited between salinities of 24 and 29. Examination of responses within these ranges using narrower  182  increments could produce more precise results. The present study also did not examine the clams through to spawning and, hence, gamete (and subsequent larval) viability were not assessed. The results of this study have immediate application to aquaculture operations. Sites selected for hatchery or growout should be somewhat buffered from salinity fluctuations and be restricted to areas where salinities normally exceed 24. Areas where salinities may periodically drop below 17, such as near the mouths of rivers, should be avoided entirely. If a hatchery location is susceptible to low salinity then the seawater intakes should be situated at sufficient depth (i.e. >10 m) to avoid freshets or the hatchery should employ a re-circulating seawater system that can maintain stringent salinity levels. Future studies on the effects of salinity on geoduck reproductive development should examine the quality of gametes and subsequent early larvae (assessed through biochemical analysis, embryogenesis rates, and larval growth/survival rates) in response to salinity. Also, the tolerance limits of P. generosa should be studied more thoroughly, especially in relation to various exposure times and body sizes (pediveligers up to adults). The effects of periodic exposure to low salinity (e.g. representing tidal fluctuations or seasonal cycles) should also be examined. This information is very important for predicting the long-term success of juvenile planting and growout. Temperature is a factor that may interact with salinity and exacerbate mortalities, as well as affect scope for growth, and the interaction between temperature and salinity should be examined in future research.  183  7.4 Chapter 4: Effect of food ration on broodstock conditioning Chapter 4 examined the effect of food ration [0.8 × 10 9, 2.4 × 109, 4.0 × 109, 5.6 × 109, 7.2 × 109, and 10.0 × 109 cells clam-1 d-1 (50:50 cell count of Isochrysis sp. and Chaetoceros muelleri)] on reproductive development, clearance rate, and ingestion rate of adult P. generosa over time (52 d). The range of rations was determined by a preliminary experiment where the maximum ingestion rate was approximately 7.2 × 109 cells clam-1 d-1 with and average of approximately 4.0 × 109 cells clam-1 d-1. An additional excessive ration was added (10.0 × 109 cells clam-1 d-1). The null hypotheses were that ration would not affect survival, spawner percentage, clearance/ingestion rates, condition index and reproductive development). The null hypotheses that ration would not affect survival, spawner percentage, or clearance and ingestion rates were all rejected. I predicted, in the alternate hypotheses, that excessive and low rations would both increase mortality, but only the former prediction was supported by the results with high mortality rates at 10.0 × 109 cells clam-1 d-1. I also predicted that high and low rations would both reduce spawning, but a clear pattern did not emerge with. 4.0 × 109 and of 7.2 × 109 cells clam-1 d-1 spawning more than the other treatments (the 10.0 × 109 cells clam-1 d-1 ration could not be compared due to high mortalities). The clearance rates and ingestion rates were significantly influenced by ration in agreement with the postulated alternative hypothesis. Clearance rate increased with ration up to 2.4 × 109 cells clam-1 d-1 and decreased with higher ration. The ingestion rate pattern was consistent with a functional response, increasing with ration from 0.8 × 109 to 2.4 × 109 cells ind-1 d-1 and remaining statistically unchanged at the higher rations. The null hypotheses could not be rejected with respect to condition index, gonadosomatic index, oocyte diameter or connective  184  tissue occupation with no significant differences detected among ration treatments. The digestive gland index was significantly higher (i.e. relatively less gonad surface area) in the lowest ration compared to the 5.6 × 109 and 7.2 × 109 cells ind-1 d-1 treatments, in agreement with the alternative hypothesis. It was predicted that the reproductive development stage would be more ripe and gamete occupation would increase with increasing ration, but the opposite was found with 7.2 × 109 cells ind-1 d-1 producing more spawned-out gonads. Previously, no information was available concerning appropriate rations for adult P. generosa. Particularly noteworthy among these results are the negative impacts associated with over-feeding including increased mortality rate and spawned-out gonads. Mortalities associated with overfeeding are not commonly reported in the literature, but have been noted for spat of the silver-lipped pearl oyster, Pinctada maxima (Mills, 2000). Necropsies of clams fed the highest ration revealed that the feeding apparatus may have been overwhelmed by the cell concentration offered (2.0 × 105 cells ml-1). Clams in the 7.2 × 109 cells ind-1 d-1 treatment may have been spawned out because high algal cell concentrations are known to induce bivalve spawning (Goodwin et al., 1979; Utting and Spencer, 1991; Helm et al., 2004), but it is unknown if the concentration of algae influences the number of gametes released. It has been reported that over-feeding can inhibit reproduction (Utting and Millican, 1997) which is consistent with the present results. The patterns for clearance and ingestion rates in the current study are consistent with the functional response commonly seen in bivalves (Riisgård and Randløv, 1981; Utting, 1986; Riisgård, 1988; Beiras and Pérez-Camacho, 1994; Pérez-Camacho et al.,  185  1994; Baldwin and Newell, 1995; Brown and Robert, 2002). The threshold cell concentration, where clearance rate decreased and ingestion rate levelled off, was 4.8 × 104 cells ml-1 (starting concentration of Isochrysis sp.), roughly similar to other bivalve species such as the mussel Mytilus edulis (Riisgård and Randløv, 1981), the surf clam Paphies donacina (Marsden, 1999), and Tapes decussatus ((Khalil, 1996). Likewise, the maximum daily ingestion capacity for P. generosa, estimated at 12 × 109 algal cells ind-1 d-1 if fed for 24 h, was only marginally higher than for smaller bivalve species such as V. philippinarum (Utting and Spencer, 1991)], C. gigas (Utting and Spencer, 1991) (Robert and Gérard, 1999)], and Pecten maximus (Robert and Gérard, 1999)]. Underfeeding does not necessarily inhibit gametogenesis, but it can reduce the amount of gonad material (Albentosa et al., 2003), a result consistent with the 0.8 × 109 cells clam-1d-1 treatment which had increased digestive gland index but similar levels of gamete occupation and development levels to the 2.4 × 109, 4.0 × 109, and 5.6 × 109 cells clam-1 d-1 treatments. Underfeeding can also decrease somatic tissue (Caers et al., 2002), but this was not evident in the present study based on various condition indices examined. Strengths of this experiment were the use of randomized blocking and histological techniques (as in previous chapters) that allowed for direct assessment of gonad development. This experiment also used a wide range of treatments sufficient to identify high and low ration levels that should be avoided. The main limitation of the present ration study was that it was conducted at a fixed temperature of 11ºC (optimum identified in Chapter 2). This means that the ration treatments may not be relevant at higher or lower temperatures and the effect of temperature on preferred ration levels should be considered in future research (see above). Also, batch feeding was used in the present  186  study, but continuous (Racotta et al., 1998; Caers et al., 2003) and multiple, daily, small batch feedings (Wikfors et al., 2004) can increase gamete production. It is unknown if batch or continuous feeding is more appropriate for P. generosa and should be considered in future research. The findings of this chapter are readily applicable to hatchery operations. Rations of 4.0 × 109 to 5.6 × 109 cells clam-1 d-1 are likely suitable for broodstock conditioning while rations above 7.2 × 109 cells clam-1 d-1 may be too high. Cell concentrations above 2.0 × 105 cells ml-1 must be avoided. Future research should examine ration in relation to temperature (see section 7.2) Also the merits of batch feeding compared to continuous feeding should be examined. Future research on these factors should include fecundity, egg quality, fertilization rates, embryogenesis rates larval quality (survival and biochemical content).  7.5 Chapter 5: Effect of food type on broodstock conditioning Chapter 5 examined the effects of various phytoplankton diets [three uni-algal feeds (Isochrysis sp., C. muelleri, and Dunaliella tertiolecta) and one combination feed (Isochrysis sp. plus C. muelleri)] on clearance rate, ingestion rate, and reproductive development of adult P. generosa over time (47 d). All were fed at an Isochrysis sp. equivalent of 4 × 109 cells ind-1 d-1,(dry wt equivalent). Algal species were selected to reflect a range of morphologies and fatty acid profiles. The null hypotheses were that clearance/ingestion rates and reproductive development would not be influenced by feed type. The null hypothesis on clearance/ingestion rates was rejected as Dunaliella tertiolecta produced a significantly higher clearance rate and ingestion rate (based on dry  187  weight of algae) than any other treatment. The null hypothesis that algal type would have no influence on reproductive development could not be rejected as none of the indicators examined (i.e. gonadosomatic index, digestive gland index, development stage, oocyte diameter, gamete or connective tissue occupation) showed any significant differences among treatments. All algal species were readily ingested, but the higher clearance and ingestion rates with D. tertiolecta were possibly due to its larger size (~12–15 μm compared to ≈ 4–5 μm for C. muelleri and Isochrysis sp.) which may have allowed for more efficient cell retention [see review by Ward and Shumway (2004)(Sobral and Widdows, 2000)]. The dry weight of D. tertiolecta ingested was more than three times higher than the other species tested which makes it a promising feed for adult P. generosa. In Chapter 4 of this thesis, P. generosa was shown to ingest relatively small amounts of food in relation to its body size and using D. tertiolecta may be one way to increase the amount of energy ingested while avoiding the problems associated with high cell concentrations of C. muelleri and Isochrysis sp. (see Chapter 4). Increasing ingested energy can reduce recovery time after spawning (Wikfors et al., 2004) and increase the overall amount of gonad material (Albentosa et al., 2003). Because the low levels of essential fatty acids (EFA) in D. tertiolecta can yield poor egg viability when used as a single feed (Caers et al., 2002) it should probably be fed in conjunction with high-EFA species. The strength of this experiment was that it included several algal species and a combination of species that represent various cell morphologies and EFA compositions [Isochrysis sp. singly (high DHA, low EPA); C. muelleri singly (high EPA, low DHA), D. tertiolecta singly (low DHA and EPA) and a combination of Isochrysis sp and C.  188  muelleri (balanced DHA and EPA)]. A limitation was that the assessment of reproductive development relied on histological examination alone and not biochemical content (i.e. lipids, proteins, carbohydrates, and essential fatty acids) of the gonads. Biochemical analysis of the eggs and resulting early larvae may have been better able to discern differences among treatments. A potential application of the findings of this chapter is the use of D. tertiolecta in addition to other standard feeds to increase ingestion. Before adopting this practice however, more research needs to be completed (see below). The research of Chapter 5 represents only a minor first step into the vast area of nutritional effects on reproductive development. Future investigations utilizing various algal species and combinations of species should be done beginning with those already known to be appropriate for bivalve culture based on cell size, digestibility, and food value (Webb and Chu, 1981; Coutteau and Sorgeloos, 1993) (see Table 1.1 for examples). The high ingestion rates of D. tertiolecta suggest that it may warrant further investigation, especially when fed in conjunction with high-EFA algae or supplements (e.g. lipid emulsions). The efficacy of various alternative feed types such as dried heterotrophically-grown algae, algal concentrates (i.e. pastes) (Jones et al., 1993; Caers et al., 2003), yeast-based feeds (Coutteau and Sorgeloos, 1993), lipid-emulsion spheres, and grain-based additives (starch and wheat germ) (Albentosa et al., 1999; Pirini et al., 2007) should also be examined. These feeds are more cost effective than live cultures and there is less risk involved (i.e. no risk of crashes) (Jones et al., 1993; Caers et al., 2003). Future research in this area should include biochemical analysis of feeds and gonads as  189  well as examination of fertilization and embryogenesis rates and biochemical composition of early larvae.  7.6 Chapter 6: Effect of food ration and stocking density on larval production Chapter 6 examined the combined effects of larval stocking density (2, 5, and 10 ind ml-1) and feed ration (5,000, 20,000, 40,000 and 100,000 cells ind-1 d-1) on growth, mortality, and ingestion rates of larval P. generosa fed Isochrysis sp. The null hypotheses were that: there will be no significant effect of stocking density on growth, survival or ingestion; there will be no significant effect of ration on growth, survival or ingestion; there will be no significant interaction effect of stocking density and ration on growth, survival or ingestion. All null hypotheses were rejected. The significant interaction meant that the highest growth rates (3.1 to 3.9 μm d-1) were found with a number of treatment pairs; 2 ind ml-1 fed 40,000 cells ind-1 d-1, 5 ind ml-1 fed 20,000, 40,000, or 100,000 cells ind-1 d-1, and 10 ind ml-1 fed 5,000 or 20,000 cells ind-1 d-1. Survival rate was highest in the 2 ind ml-1 treatments (independent of ration) and lowest in 10 ind ml-1 treatments with feed levels above 20,000 cells ind-1. The highest ingestion rates were 28,961 and 27,212 cells ind -1 d-1 at 2 and 5 inds ml-1, respectively, at 100×103 cells ind-1 d-1. Studies often show that low densities (<5 inds ml-1 ) produce higher growth rates in bivalve larvae (Deming and Russell, 1999)(Yan et al., 2006), but my results show that higher densities can be compensated for with increased rations. This was also indicated by the modelling work of Powell et al. (2002) and Pechenick et al. (1990) suggested that feed and density interaction could account for the disparity in larval Mytilus edulis  190  growth rates in various studies. Growth rates of Ruditapes philippinarum larvae did not show an interaction between stocking density and feeding rates when using algal concentrations of up to 24,000 cells ml-1 of Isochrysis spp. (Yan et al., 2006). The present study, however, tested much higher cell concentrations in some treatments (400,000 cells ml-1 or more for 5 inds ml-1 at 100×103 cells ind-1 d-1, 10 inds ml-1 at 40×103 cells ind-1 d-1, and 10 inds ml-1 at 100×103 cells ind-1 d-1) which likely facilitated the interaction by reducing feeding efficiency (Gallager, 1988) and therefore growth rates in those treatments. All of the highest growth treatments in the present study had algal concentrations of 50,000 to 200,000 cells ml-1, similar to the appropriate concentrations for Ostrea edulis (100,000 to 200,000 cells ml-1 of I. galbana) (Beiras and PérezCamacho, 1994), Pecten maximus (100,000 cells ml-1 of Chaetoceros calcitrans f. pumilum and Pavlova sp. AC 250) (Ponis et al., 2006b), V. philippinarum (200,000 cells ml-1 of C. calcitrans) (Utting and Doyou, 1992), and C. gigas (100,000 cells ml-1 of Isochrysis sp. and C. calcitrans f. pumilum) (Brown and Robert, 2002). The low survival rate of the 10 ind ml-1 treatments with feed levels above 20,000 cells ind-1 was associated with the proliferation of filamentous algae. Overfeeding can result in detritus accumulation which may act as a substrate for bacterial growth (Torkildsen and Magnesen, 2004) (Shaul, 1981). Protozoans were also present in those treatments. The highest ingestion rates were 28,961 and 27,212 cells ind -1 d-1 at 2 and 5 inds ml-1, respectively, at 100×103 cells ind-1 d-1, similar to those of C. gigas larvae which removed up to 26,000 cells of I. galbana ind-1 d-1 (Gerdes, 1983). Ingestion rates increased with ration level at 2 and 5 inds m-1, but remained constant across ration level at 10 inds ml-1 suggesting a functional response as shown in other bivalves (Baldwin and  191  Newell, 1995) (Utting, 1986; Brown and Robert, 2002), (Baldwin and Newell, 1995), (Pérez-Camacho et al., 1994), The major strength of this study was the use of a wide range of rations and stocking densities. Using this I was able to demonstrate the interaction of these two factors which is a critical factor when setting feeding rates. Image analysis allowed for the collection of large numbers of readings efficiently without having to alter the treatments by removing and preserving larvae for later analysis. The use of small cultivation units, however, may have increased mortality rates due to the high surface area to volume ratio of the containers which is known to increase larval mortality (Harboe et al., 1994). Daily exchanges with sterilized units should have minimized this effect however. This experiment used only one algal species (Isochrysis sp.) and results may have been improved with other species or combination of different species. It must be noted that ingestion rates change allometrically with body size and decline near metamorphosis (Gerdes, 1983), so the ingestion rates shown in the present study cannot be considered accurate over the entire larval period. The findings of this study can be directly applied to hatchery operations. High larval growth rates can be achieved with the following larval density and feed ration combinations: 2 ind ml-1 fed 40,000 cells (Isochrysis sp.) ind-1 d-1, 5 ind ml-1 fed 20,000, 40,000, or 100,000 cells ind-1 d-1, and 10 ind ml-1 fed 20,000 cells ind-1 d-1. The most cost effective treatment was 10 ind ml-1 fed 5,000 cells ind-1 d-1 which produced high survival and growth rates yet had the highest stocking density and the lowest ration in terms of cells ind-1 d-1.  192  Future research should examine what algal species and combinations of species are most appropriate for P. generosa larvae. Testing of standard feed species (see Table 1.1) in various combinations is an obvious path for future research. This research should include biochemical analysis of algae and larvae to determine the combinations best for growth and survival. There is also evidence that wild algal species, native to the region of a cultured bivalve species, (Gouda et al., 2006) and some other unconventional algal species (Pernet et al., 2005; Martínez-Fernández et al., 2006; Ponis et al., 2006b; Ponis et al., 2006a) yield better growth and survival than standard laboratory species. Another path to explore is the biochemical modification of algae known to be acceptable to P. generosa. Techniques such as manipulating light intensity and temperature can alter the lipid content and fatty acid profiles in algae and may prove beneficial to larval nutrition [see review by Marshall et al., (2010)]. The efficacy of various alternative feed types such as dried heterotrophically-grown algae, algal concentrates (i.e. pastes), yeast-based feeds, lipid-emulsion spheres, and grain-based additives (starch and wheat germ), picoplankton and bacteria should also be examined [see review by Marshall et al., (2010)].  7.7 Overall conclusions In conclusion, I have shown in this dissertation that P. generosa gonad development responds to variations in both physical (temperature and salinity) and biological (ration) factors. My findings can be applied to commercial hatchery operations for the purpose of improving broodstock conditioning. Temperature may be manipulated to increase gamete output (increased percentage of spawners) (11ºC) or  193  suppress spawning (11ºC) or even induce resorption if desired (19ºC). Salinity must be maintained at high levels (>24) to facilitate gonadal growth and avoid mortality. Ration also influences reproductive development and algal type influences ingestion rate. I also found that the interaction of larval stocking density and ration influences larval growth and survival. Panopea generosa remains under-represented in the scientific literature and given its economic significance, research on reproductive development and hatchery techniques should continue.  194  References Abraham, A., Couturier, C., Parsons, J., 1996. Toxicity of un-ionized ammonia to juvenile giant scallops, Placopecten magellanicus. Bull. Aquac. Assoc. Can. 96, 68-70. Akberali, H.B., Davenport, J., 1981. The responses of the bivalve Scrobicularia plana (Da Costa) to gradual salinity changes. J. Exp. Mar. Biol. Ecol. 53, 251-259. Albentosa, M., Fernández-Reiriz, M.J., Labarta, U., Pérez-Camacho, A., 2003. Absorption of biochemical components and feeding behaviour with natural and carbohydrate-rich diets in Ruditapes decussatus and Venerupis pullastra clams. J. Shellfish Res. 22, 443-449. Albentosa, M., Fernández-Reiriz, M.J., Pérez-Camacho, A., Labarta, U., 1999. Growth performance and biochemical composition of Ruditapes decussatus (L.) spat fed on microalgal and wheatgerm flour diets. J. Exp. Mar. Biol. Ecol. 232, 23-37. Aldridge, D.W., Payne, B.S., Miller, A.C., 1995. Oxygen consumption, nitrogenous excretion, and filtration rates of Dreissena polymorpha at acclimation temperatures between 20 and 32°C. Can. J. Fish. Aquat. Sci. 52, 1761-1767. Andersen, A.M., 1971. Spawning, growth and spatial distribution of the geoduck clam, Panope generosa Gould, in Hood Canal, Washington. PhD Diss., Washington Cooperative Fisheries Unit. University of Washington, Seattle, pp. 133. Anonymous, 2008. Habitat conservation plan for the Washington State Department of Natural Resources' wild geoduck fishery. Washington State Department of Natural Resources, Aquatic Resources Program, Olympia, 103 pp. Aragón-Noriega, E.A., Chávez-Villalba, J., Gribben, P.E., Alcántara-Razo, E., MaedaMartínez, A.N., Arambula-Pujol, E.M., García-Jurarez, A.R., MaldonadoAmparo, R., 2007. Morphometric relationships, gametogenic development and spawning of the geoduck clam Panopea globosa (bivalvia: hiatellidae) in the central Gulf of California. J. Shellfish Res. 26, 423-431. Baba, K., Tada, M., Kawajiri, T., Kuwahara, Y., 1999. Effects of temperature and salinity on spawning of the brackish water bivalve Corbicula japonica in Lake Abashiri, Hokkaido, Japan. Mar. Ecol. Prog. Ser. 180, 213-221. Baldwin, B.S., Newell, R.I.E., 1995. Feeding rate responses of oyster larvae (Crassostrea virginica) to seston quantity and composition. J. Exp. Mar. Biol. Ecol. 189, 77-91. Bändel, K., 1988. Stages in the ontogeny and a model of the evolution of bivalves (Mollusca). Palaentolog. Z. 62, 217-254. Barber, B.J., Ford, S.E., Wargo, R.N., 1991. Genetic variation in the timing of gonadal maturation and spawning of the eastern oyster, Crassostrea virginica (Gmelin). Biol. Bull. 181, 216-221. Bataller, E.E., Boghen, A.D., Burt, M.D.B., 1999. Comparative growth of the eastern oyster Crassostrea virginica (Gmelin) reared at low and high salinities in New Brunswick, Canada. J. Shellfish Res. 18, 107-114. Bayne, B.L., 1976. Aspects of reproduction in bivalve molluscs, in: Wiley, M.L. (Ed.), Estuarine Processes. Academic Press, New York, pp. 432-448.  195  Bayne, B.L., 1998. The physiology of suspension feeding by bivalve molluscs: an introduction to the Plymouth “TROPHEE” workshop. J. Exp. Mar. Biol. Ecol. 219, 1-19 Bayne, B.L., Newell, R.C., 1983. Physiological energetics of marine molluscs, in: Saleuddin, A.S.M., Wilbur, K.M. (Eds.), The Mollusca, Volume 4, Physiology Part 1. Academic Press, New York, pp. 407-515. Beattie, H., Blake, B., 1999. Development of culture methods for the geoduck clam in the USA (Washington State) and Canada (British Columbia). World Aquac. 30, 5053. Beattie, J.H., 1992. Geoduck enhancement in Washington State. Bull. Aquac. Assoc. Can. 92(4), 18-24. Beiras, R., Pérez-Camacho, A., 1994. Influence of food concentration on the physiological energetics and growth of Ostrea edulis larvae. Mar. Biol. 120, 427435. Beninger, P.G., Lucas, A., 1984. Seasonal variations in condition, reproductive activity, and gross biochemical composition of two species of adult clam reared in a common habitat: Tapes decussatus L. (Jeffreys) and Tapes philippinarum (Adams and Reeve). J. Exp. Mar. Biol. Ecol. 79, 19-37. Berge, J.A., Bjerkeng, B., Pettersen, O., Schaanning, M.T., Øxnevad, S., 2006. Effects of increased sea water concentrations of CO2 on growth of the bivalve Mytilus edulis L. Chemosphere 62, 681-687. Berntsson, K.M., Jonsson, P.R., Wängberg, S.Å., Carlsson, A.S., 1997. Effects of broodstock diets on fatty acid composition, survival and growth rates in larvae of the European flat oyster, Ostrea edulis. Aquaculture 154, 139-153. Bervera, H., Monteforte, M., 1995. Spat collection trials for pearl oyster Pinctada mazatlanica at Bahia de La Paz, South Baja California, Mexico, World Aquaculture Society Book of Abstracts, World Aquaculture 1995, San Diego, California, U.S.A., pp. 62. Beukema, J.J., de Vlas, J., 1989. Tidal-current transport of thread-drifting postlarval juveniles of the bivalve Macoma balthica from the Wadden Sea to the North Sea. Mar. Ecol. Prog. Ser. 52, 193-200. Bologna, P.A.X., Heck, K.L., 2000. Impacts of Seagrass Habitat Architecture on Bivalve Settlement. Estuaries 23, 449-457. Bower, S.M., Blackbourn, J., 2003. Geoduck clam (Panopea abrupta): Anatomy, histology, development, pathology, parasites and symbionts: Anatomy of geoduck clams. Fisheries and Oceans Canada, http://pacrhqiis/geoduck/anatomy_e.htm. Breen, P.A., Shields, T.L., 1983. Age and size structure in five populations of geoduck clams (Panope generosa) in British Columbia. Can. Tech. Rep. Fish. Aquat. Sci. 1169, 62 p. Brenko, M.H., Calabrese, A., 1969. The combined effects of salinity and temperature on larvae of the mussel Mytilus edulis. Mar. Biol. 4, 224-226. Brougrier, S., Geairon, P., Deslous-Paoli, J.M., Bacher, C., Jonquières, G., 1995. Allometric relationships and effects of temperature on clearance and oxygen consumption rates of Crassostrea gigas (Thunberg). Aquaculture 134, 143-154.  196  Brown, M.R., Robert, R., 2002. Preparation and assessment of microalgal concentrates as feeds for larval and juvenile Pacific oyster (Crassostrea gigas). Aquaculture 207, 289-309. Brown, M.R., Jeffrey, S.W., Volkman, J.K., Dunstan, G.A., 1997. Nutritional properties of microalgae for mariculture. Aquaculture 151, 315-331. Bureau, D., Hajas, W., Hand, C.M., Dovey, G., 2003. Age, size structure and growth parameters of geoducks (Panopea abrupta, Conrad 1849) from seven locations in British Columbia sampled in 2001 and 2002. Can. Tech. Rep. Fish. Aquat. Sci. 2494, 29 p. Bureau, D., Hajas, W., Surry, N.W., Hand, C.M., Dovey, G., Campbell, A., 2002. Age, size structure and growth parameters of geoducks (Panopea abrupta, Conrad 1849) from 34 locations in British Columbia samples between 1993 and 2000. Can. Tech. Rep. Fish. Aquat. Sci. 2413, 84 p. Burnett, L.E., 1997. The challenges of living in hypoxic and hypercapnic aquatic environments. Am. Zool. 37, 633-640. Butler, P.A., 1949. Gametogenesis in the oyster under conditions of depressed salinity. Biol. Bull. 96, 263-269. Caers, M., Coutteau, P., Sorgeloos, P., 1999a. Dietary impact of algal and artificial diets, fed at different feeding rations, on the growth and fatty acid composition of Tapes philippinarum (L.) spat. Aquaculture 170, 307-322. Caers, M., Coutteau, P., Sorgeloos, P., Gajardo, G., 2003. Impact of algal diets and emulsions on the fatty acid composition and content of selected tissues of adult broodstock of the Chilean scallop Argopecten purpuratus (Lamarck, 1819). Aquaculture 217, 437-452. Caers, M., Utting, S.D., Coutteau, P., Millican, P.F., Sorgeloos, P., 2002. Impact of the supplementation of a docosahexaenoic acid-rich emulsion on the reproductive output of oyster broodstock, Crassostrea gigas. Mar. Biol. 140, 1157-1166. Caers, M., Coutteau, P., Cure, K., Morales, V., Gajardo, G., Sorgeloos, P., 1999b. The Chilean scallop Argopecten purpuratus (Lamarck, 1819): II. Manipulation of the fatty acid composition and lipid content of the eggs via lipid supplementation of the broodstock diet. Comp. Biochem. Physiol. B 123, 97-103. Calderon-Aguilera, L.E., Aragon-Noriega, E.A., Hand, C.M., Moreno-Rivera, V.M., 2010. Morphometric Relationships, Age, Growth, and Mortality of the Geoduck Clam, Panopea generosa, Along the Pacific Coast of Baja California, Mexico. J. Shellfish Res. 29, 319-326. Campbell, A., Ming, M.D., 2003. Maturity and growth of the Pacific geoduck clam, Panopea abrupta, in Southern British Columbia, Canada. J. Shellfish Res. 22, 8590. Campbell, A., Yeung, C.W., Dovey, G., Zhang, Z., 2004. Population biology of the Pacific geoduck clam, Panopea abrupta, in experimental plots, southern British Columbia, Canada. J. Shellfish Res. 23, 661-673. Carriker, M.R., 2001. Embryogenesis and organogenesis of veligers and early juveniles, in: Kraeuter, J.N., Castagna, M. (Eds.), Biology of the hard clam. Elsevier, Amsterdam, pp. 751. Chávez-Villalba, J., Barret, J., Mingant, C., Cochard, J.-C., Le Pennec, M., 2003a. Influence of timing of broodstock collection on conditioning, oocyte production,  197  and larval rearing of the oyster, Crassostrea gigas (Thunberg), at six production sites in France. J. Shellfish Res. 22, 465-474. Chávez-Villalba, J., Cochard, J.-C., Le Pennec, M., Barret, J., Enríquez-Díaz, M., Cáceres-Martínez, C., 2003b. Effects of temperature and feeding regimes on gametogenesis and larval production in the oyster Crassostrea gigas. J. Shellfish Res. 22, 721-731. Chávez-Villalba, J., Pommier, J., Andriamiseza, J., Pouvreau, S., Barret, J., Cochard, J.C., Le Pennec, M., 2002. Broodstock conditioning of the oyster Crassostrea gigas: origin and temperature effect. Aquaculture 214, 115-130. Chew, K.K., 1998. Update on geoduck clam culture in the Pacific Northwest. Aquaculture Magazine 24, 79-82. Chiaverano, L., Mianzan, H., Ramìrez, F., 2004. Gonad development and somatic growth patterns of Olindias sambaquiensis (Limnomedusae, Olindiidae). Hydrobiologia 530/531, 373-381. Chu, F.-L., Greaves, J., 1991. Metabolism of palmitic, linoleic, and linolenic acids in adult oysters, Crassostrea virginica. Mar. Biol. 110, 229-236. Cigarría, J., 1999. Effects of age, size, and season on growth of soft tissue in the oyster Crassostrea gigas (Thunberg, 1793). J. Shellfish Res. 18, 127-131. Clesceri, L.S.E., Greenberg, A.E., Eaton, A.D., (Eds.). 1998. Standard methods for the examination of water and wastewater , 20th Edition. American Public Health Association, Washington, D.C. Cooper, K., Pease, B., 1988. A relationship between selective larval settlement and adult distribution patterns of geoduck clams and the presence of chaetopterid polychaete tube mats in Puget Sound, Washington. J. Shellfish Res. 7, 129. Cordero, C., Voltolina, D., 1997. Viability of mass algal cultures preserved by freezing and freeze-drying. Aquac. Eng. 16, 205-211. Coughlan, J., 1969. The estimation of filtering rate from the clearance of suspensions. Mar. Biol. 2, 356-358. Coutteau, P., Sorgeloos, P., 1993. Substitute diets for live algae in the intensive rearing of bivalve mollusks - a state of the art report. World Aquac. 24, 45-52. Coutteau, P., Curé, K., Sorgeloos, P., 1994a. Effect of algal ration on feeding and growth of juvenile Manila clam Tapes philippinarum (Adams and Reeve). J. Shellfish Res. 13, 47-55. Coutteau, P., Hadley, N.H., Manzi, J.J., Sorgeloos, P., 1994b. Effect of algal ration and substitution of algae by the manipulated yeast diets on the growth of juvenile Mercenaria mercenaria. Aquaculture 120, 135-150. Davenport, J., Wong, T.M., 1986. Responses of the blood cockle Anadara granosa (L.) (Bivalvia: Arcidae) to salinity, hypoxia and aerial exposure. Aquaculture 52, 151162. Davis, H.C., 1958. Survival and growth of clam and oyster larvae at different salinities. Biol. Bull. 114, 296-307. Davis, J.P., Barenberg, C.R., 2000a. Post settlement geoduck clam (Panopea abrupta) growth and survivorship in sand and screen based nursery systems. J. Shellfish Res. 19, 689-690. Davis, J.P., Barenberg, C.R., 2000b. Burrowing response of juvenile geoducks (Panopea abrupta) to changes in temperature and salinity. J. Shellfish Res. 19, 689.  198  Davis, J.P., Barenberg, C.R., 2000c. The use of a Schizochytrium based dry feed for juvenile rearing and broodstock conditioning of bivalve molluscs. J. Shellfish Res. 19, 652-653. Deaton, L.E., 1994. Hypoosmotic volume regulation in bivalves - protein-kinase-C and amino-acid release. J. Exp. Zool. 268, 145-150. Deaton, L.E., 2001. Hyperosmotic volume regulation in the gills of the ribbed mussel, Geukensia demissa: rapid accumulation of betaine and alanine. J. Exp. Mar. Biol. Ecol. 260, 185-197. Deaton, L.E., Pierce, S.K., 1994. Introduction: cellular-volume regulation - mechanisms and control. J. Exp. Zool. 268, 77-79. Delaporte, M., Soudant, P., Lambert, C., Moal, J., Pouvreau, S., Samain, J.-F., 2006. Impact of food availability on energy storage and defense related hemocyte parameters of the Pacific oyster Crassostrea gigas during an experimental reproductive cycle. Aquaculture 254, 571-582. Delaunay, F., Marty, Y., Moal, J., Samain, J.-F., 1993. The effect of monospecific algal diets on growth and fatty acid composition of Pecten maximus (L.) larvae. J. Exp. Mar. Biol. Ecol. 173, 163-179. Delgado, M., Pérez-Camacho, A., 2003. A study of gonadal development in Ruditapes decussatus (L.) (Mollusca, Bivalvia), using image analysis techniques: influence of food ration and energy balance. J. Shellfish Res. 22, 435-441. Delgado, M., Pérez-Camacho, A., 2007. Influence of temperature on gonadal development of Ruditapes philippinarum (Adams and Reeve, 1850) with special reference to ingested food and energy balance. Aquaculture 264, 398-407. Deming, C.J., Russell, M., P., 1999. Assessing manipulations of larval density and culling in hatchery production of the hard clam, Mercenaria mercenaria. J. Shellfish Res. 18, 99-105. Deslous-Paoli, J.M., Heral, M., 1988. Biochemical composition and energy value of Crassostrea gigas (Thunberg) cultured in the bay of Marennes-Oléron. Aquat. Living Resour. 1, 239-249. Devauchelle, N., Mingant, C., 1991. Review of the reproductive physiology of the scallop, Pecten maximus, applicable to intensive aquaculture. Aquat. Living Resour. 4, 41-51. Diaz, Rosenberg, 1995. Marine benthic hypoxia: a review of its ecological effects and the behavioural responses of benthic macrofauna. Oceanogr. Mar. Biol., Annu. Rev. 33, 245-303. DiSalvo, L.H., Bleoka, J., Zebral, R., 1978. Vibrio anguillarum and larval mortality in a California coastal shellfish hatchery. Appl. Environ. Microbiol. 35, 219-221. Don, R., 1957. Salinity optima for marine fungi affected by temperature. Am. J. Bot. 44, 870-874. Dridi, S., Romdhane, M.S., Elcafsi, M., 2007. Seasonal variation in weight and biochemical composition of the Pacific oyster, Crassostrea gigas in relation to the gametogenic cycle and environmental conditions of the Bizert lagoon, Tunisia. Aquaculture 263, 238-248. Dunstan, G.A., Volkman, J.K., Barrett, S.M., Leroi, J.-M., Jeffrey, S.W., 1994. Essential polyunsaturated fatty acids from 14 species of diatom (Bacillariophyceae). Phytochemistry 35, 155-161.  199  Elston, R.A., Cheney, D.P., MacDonald, B.F., Suhrbier, A.D., 2003. Tolerance and response of Manila clams, Venerupis philippinarum (A. Adams and Reeve, 1850) to low salinity. J. Shellfish Res. 22, 667-674. Elston, R.A., Hasegawa, H., Humphrey, K.L., Polyak, I.K., Hase, C.C., 2008. Reemergence of Vibrio tubiashii in bivalve shellfish aquaculture : severity, environmental drivers, geographic extent and management. Dis. Aquat. Org. 82, 119-134. Epifanio, C.E., Srna, R.F., 1975. Toxicity of ammonia, nitrite ion, nitrate ion, and orthophosphate to Mercenaria mercenaria and Crassostrea virginica. Mar. Biol. 33, 241-246. Esquivel, B.C., Voltolina, D., 1996. Nutritional value of preserved microalgae for subadult Mytilus galloprovincialis. J. World Aquac. Soc. 27, 113-118. Fabioux, C., Huvet, A., Le Souchu, P., Le Pennec, M., Pouvreau, S., 2005. Temperature and photoperiod drive Crassostrea gigas reproductive internal clock. Aquaculture 250, 458-470. Feder, H.M., Hendee, J.C., Holmes, P., Mueller, G.J., Paul, A.J., 1979. Examination of a reproductive cycle of Protothaca staminea using histology, wet weight-dry weight ratios, and condition indices. Veliger 22, 182-187. Feldman, K., Vadopalas, B., Armstrong, D., Friedman, C., Hilborn, R., Naish, K., Orensanz, J., Valero, J., Ruesink, J., Suhrbier, A., Christy, A., D., C., Davis, J.P., 2004. Comprehensive literature review and sysnopsis of issues relating to geoduck (Panopea abrupta) ecology and aquaculture production, Prepared for Washington State Department of Natural Resources, pp. 134. Ferreiro, M.J., Pérez-Camacho, A., Labarta, U., Beiras, R., Planas, M., Fernández-Reiriz, M.J., 1990. Changes in the biochemical composition of Ostrea edulis larvae fed on different food regimes. Mar. Biol. 106, 395-401. Foster-Smith, R.L., 1975. The effect of concentration of suspension on the filtration rates and pseudofaecal production for Mytilus edulis L., Cerastoderma edule (L.) and Venerupis pullastra (Montagu). J. Exp. Mar. Biol. Ecol. 17, 1-22. Gabbott, P.A., 1983. Developmental and seasonal metabolic activities in marine molluscs, in: Hochachka, P.W. (Ed.), The Mollusca, Vol. 2, Environmental Biochemistry and Physiology. Academic Press, New York, pp. 165-217. Gabbott, P.A., Walker, A.J.M., 1971. Changes in the condition index and biochemical content of adult oysters (Ostrea edulis L.) maintained under hatchery conditions. ICES J. Mar. Sci. 34, 98-105 Gallager, S.M., 1988. Visual observations of particle manipulation during feeding in larvae of a bivalve mollusc. Bull. Mar. Sci. 43, 344-365. Gallager, S.M., Mann, R., 1986. Growth and survival of larvae of Mercenaria mercenaria (L.) and Crassostrea virginica (Gmelin) relative to broodstock conditioning and lipid content of eggs. Aquaculture 56, 105-121. Gallager, S.M., Mann, R., Sasaki, G.C., 1986. Lipid as an index of growth and viability in three species of bivalve larvae. Aquaculture 56, 81-103. Garcia, E., Thorarinsdottir, G., Ragnarsson, S., 2003. Settlement of bivalve spat on artificial collectors in Eyjafjordur, North Iceland. Hydrobiologia 503, 131–141.  200  Gareth-Jones, E.B., Jennings, D.H., 1964. The effect of salinity on the growth of marine fungi in comparison with non-marine species. Trans. Br. Mycol. Soc. 47, 619625. Geffard, O., Budzinski, H., His, E., Seaman, M.N.L., Garrigues, P., 2002. Relationships between contaminant levels in marine sediments and their biological effects on embryos of oysters, Crassostrea gigas. Environ. Toxicol. Chem. 21, 2310–2318. Gerdes, D., 1983. The Pacific oyster Crassostrea gigas Part I. Feeding behaviour of larvae and adults. Aquaculture 31, 195-219. Giese, A.C., 1959. Comparative physiology: annual reproductive cycles of marine invertebrates. Annu. Rev. Physiol. 21, 547-576. Glémet, H.C., Ballantyne, J.S., 1995. Influences of environmental salinity on the structure and function of gill mitochondrial membranes of an osmoconforming invertebrate, Crassostrea virginica. Mar. Biol. 121, 673-683. Goodwin, C.L., Pease, B., 1989. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Pacific Northwest)--Pacific geoduck clam. U.S. Fish. Wildl. Serv. Biol. Rep. 82(11.120). U.S. Army Corps of Engineers, TR EL-82-4. 14 pp. Goodwin, C.L., Pease, B.C., 1991. Geoduck, Panopea abrupta (Conrad, 1849), size, density, and quality as related to various parameters in Puget Sound, Washington. J. Shellfish Res. 10, 65-77. Goodwin, C.L., and Shaul, W., 1984. Age, recruitment and growth of the geoduck clam (Panope generosa Gould) in Puget Sound, Washington. Wash. Dep. Fish. Prog. Rep. No. 215. Goodwin, L., 1973. Effects of salinity and temperature on embryos of the geoduck clam (Panope generosa Gould). Proc. Natl. Shellfish. Assoc. 63, 93-96. Goodwin, L., 1976. Observations on spawning and growth of subtidal geoducks (Panope generosa, Gould). Proc. Natl. Shellfish. Assoc. 65, 49-58. Goodwin, L., Shaul, W., Budd, C., 1979. Larval development of the geoduck clam (Panope generosa, Gould). Proc. Natl. Shellfish. Assoc. 69, 73-76. Gordon, D.G., 2007. Geoduck Studies- Washington Sea Grant coordinates effort to understand what's at stake. http://www.wsg.washington.edu/communications/seastar/stories/s_07.html. Washington Sea Grant. Gouda, R., Kenchington, E., Hatcher, B., Vercaemer, B., 2006. Effects of locally-isolated micro-phytoplankton diets on growth and survival of sea scallop (Placopecten magellanicus) larvae. Aquaculture 259, 169-180. Gribben, P.E., Creese, R.G., 2005. Age, growth, and mortality of the New Zealand geoduck clam, Panopea zelandica (Bivalvia: Hiatellidae) in two North Island populations. Bull. Mar. Sci. 77, 119-135. Gribben, P.E., Hay, B.E., 2003. Larval development of the New Zealand geoduck Panopea zelandica (Bivalvia: Hiatellidae). N. Z. J. Mar. Freshw. Res. 37, 231239. Gribben, P.E., Helson, J., Jeffs, A.G., 2004. Reproductive cycle of the New Zealand geoduck, Panopea zelandica, in two North Island populations. Veliger 47, 53-65. Hamburger, K., Møhlenberg, F., Randløv, A., Riisgård, H.U., 1983. Size, oxygen consumption and growth in the mussel Mytilus edulis. Mar. Biol. 75, 303-306.  201  Hamer, B., Jaksic, Z., Pavicic-Hamer, D., Peric, L., Medakovic, D., Ivankovic, D., Pavicic, J., Zilberberg, C., Schroder, H.C., Muller, W.E.G., Smodlaka, N., Batel, R., 2008. Effect of hypoosmotic stress by low salinity acclimation of Mediterranean mussels Mytilus galloprovincialis on biological parameters used for pollution assessment. Aquat. Toxicol. 89, 137-151. Harbo, R.M., Adkins, B.E., Breen, P.A., Hobbs, K.L., 1983. Age and size in market samples of geoduck clams (Panope generosa). Can. Manuscr. Rep. Fish. Aquat. Sci. 1714, 77p. Harrison, P.J., Waters, R.E., Taylor, F.J.R., 1980. A broad spectrum artificial seawater medium for coastal and open ocean phytoplankton, in: Berg, C.J. (Ed.), Culture of marine invertebrates: selected readings. Hutchinson Ross Publishing, Stroudsburg, pp. 140-147. Heasman, M., Diemar, J., O'Connor, W., Sushames, T., Foulkes, L., 2000. Development of extended shelf-life microalgae concentrate diets harvested by centrifugation for bivalve molluscs- a summary. Aquac. Res. 31, 637-659. Heasman, M.P., O'Connor, W.A., Frazer, A.W., 1996. Temperature and nutrition as factors in conditioning broodstock of the commercial scallop Pecten fumatus Reeve. Aquaculture 143, 75-90. Heffernan, P.B., Walker, R.L., Carr, J.L., 1989a. Gametogenic cycles of three bivalves in Wassaw Sound Georgia:I. Mercenaria mercenaria (Linnaeus, 1758). J. Shellfish Res. 8, 51-60. Heffernan, P.B., Walker, R.L., Carr, J.L., 1989b. Gametogenic cycles of three marine bivalves in Wassaw Sound, Georgia: II. Crassostrea virginica (Gmelin, 1791). J. Shellfish Res. 8, 61-70. Helm, M.M., Bourne, N., Lovatelli, A., 2004. Hatchery culture of bivalves. A practical manual., FAO Fisheries Technical Paper No. 471, Food and Agriculture Organization of the United Nations, Rome, 177 pp. Helm, M.M., Holland, D.L., Stephenson, R.R., 1973. The effect of supplementary algal feeding of a hatchery breeding stock of Ostrea edulis L. on larval vigour. J. Mar. Biol. Ass. U. K. 53, 673-684. Helm, M.M., Laing, I., 1987. Preliminary observations on the nutritional value of 'Tahiti Isochrysis' to bivalve larvae. Aquaculture 62, 281-288. Hendriks, I.E., van Duren, L.A., Herman, P.M.J., 2003. Effect of dietary polyunsaturated fatty acids on reproductive output and larval growth of bivalves. J. Exp. Mar. Biol. Ecol. 296, 199-213. Hesselman, D.M., Barber, B.J., Blake, N.J., 1989. The reproductive cycle of adult hard clams, Mercenaria spp. in the Indian River Lagoon, Florida. J. Shellfish Res. 8, 43-49. His, E., Robert, R., Dinet, A., 1989. Combined effects of temperature and salinity on fed and starved larvae of the Mediterranean mussel Mytilus galloprovincialis and the Japanese oysters Crassostrea gigas. Mar. Biol. 100, 455-463. Holland, D.L., Spencer, B.E., 1973. Biochemical changes in fed and starved oysters, Ostrea edulis L. during larval development, metamorphosis and early spat growth. J. Mar. Biol. Ass. U. K. 53, 287-298.  202  Honkoop, P.J.C., van der Meer, J., 1998. Experimentally induced effects of water temperature and immersion time on reproductive output of bivalves in the Wadden Sea. J. Exp. Mar. Biol. Ecol. 220, 227-246. Hutchinson, S., Hawkins, L.E., 1992. Quantification of the physiological responses of the European flat oyster Ostrea edulis L. to temperature and salinity. J. Mollus. Stud. 58, 215-226. Iglesias, J.I.P., Navarro, E., Jorna, P.A., Armentia, I., 1992. Feeding, particle selection and absorption in cockles Cerastoderma edule (L.) exposed to variable conditions of food concentration and quality. J. Exp. Mar. Biol. Ecol. 162, 177-198. Izumi, S., Masabumi, S., Norihito, M., Takashi, O., 2000. Tolerance of the Japanese scallop, Patinopecten yessoensis, to low concentrations of dissolved oxygen and salinity. Suisan Zoshoku 48, 137-138. James, M., 2008. Co-operative management of the geoduck and horse-clam fishery in British Columbia, in: Townsend, R., Shotton, R., Uchida, H. (Eds.), Case studies in fisheries self-governance, FAO Fisheries Technical Paper No. 504, Food and Agriculture Organization of the United Nations, Rome, pp. 397-406. Jansen, J.M., Koutstaal, A., Bonga, S.W., Hummel, H., 2009. Salinity-related growth rates in populations of the European clam Macoma balthica and in field transplant experiments along the Baltic Sea salinity gradient. Mar. Freshw. Behav. Physiol. 42, 157-166. Jones, D.A., Kamarudin, M.S., Le Vay, L., 1993. The potential for replacement of live feeds in larval culture. J. World Aquac. Soc. 24, 199-210. Jones, G., Jones, B., 1988. Advances in the remote setting of oyster larvae. Report of Aquaculture Association of British Columbia, pp. 88. Jonsson, P.R., Berntsson, K.M., André, C., Wängberg, S.Å., 1999. Larval growth and settlement of the European oysters (Ostrea edulis) as a function of food quality measured as fatty acid composition. Mar. Biol. 134, 559-570. Kang, C.-K., Kang, Y.S., Choy, E.J., Kim, D.-S., Shim, B.-T., Lee, P.-Y., 2007. Condition, reproductive activity and gross biochemical composition of the Manila clam, Tapes philippinarum in natural and newly created sandy habitats of the southern coast of Korea. J. Shellfish Res. 26, 401-412. Kang, C., Park, M., Lee, P., Choi, W., Lee, W., 2000. Seasonal variations in condition, reproductive activity, and biochemical composition of the Pacific oyster, Crassostrea gigas (Thunberg), in suspended culture in two coastal bays of Korea. J. Shellfish Res. 19, 771-778. Kanti, A., Heffernan, P.B., Walker, R.L., 1993. Gametogenic cycle of the southern surfclam, Spisula solidissima similis (Say, 1822), from St. Catherines Sound, Georgia. J. Shellfish Res. 12, 255-261. Keck, R.T., Maurer, D., Lind, H., 1975. A comparative study of the hard clam gonad developmental cycle. Biol. Bull. 148, 243-258. Kent, M.L., Elston, R.A., Nerad, T.A., Sawyer, T.K., 1987. An Isonema-like flagellate (Protozoa: Mastigophora) infection in larval geoduck clams, Panope abrupta. J. Invertebr. Pathol. 50, 221-229. Khalil, A.M., 1996. The influence of algal concentration and body size on filtration and ingestion rates of the clam Tapes decussatus (L.) (Mollusca: Bivalvia). Aquac. Res. 27, 613-621.  203  Kim, W.S., Huh, H.T., Huh, S.H., Lee, T.W., 2001. Effects of salinity on endogenous rhythm of the Manila clam, Ruditapes philippinarum (Bivalvia: Veneridae). Mar. Biol. 138, 157-162. King, J.J., 1986. Juvenile feeding ontogeny of the geoduck Panope abrupta (Bivalvia: Saxicavacea) and comparative ontogeny and evolution of feeding in bivalves. University of Victoria, Victoria, pp. 281 Knuckey, R.M., Brown, M.R., Robert, R., M.F., F., 2006. Production of microalgal concentrates by flocculation and their assessment as aquaculture feeds. Aquac. Eng. 35, 300-313. Kraffe, E., Soudant, P., Marty, Y., 2004. Fatty acids of serine, ethanolamine, and choline plasmalogens in some marine bivalves. Lipids 39, 59-66. Laing, I., 1993. The response of Manila clam, Tapes philippinarum, juveniles to nutritive stress. J. Exp. Mar. Biol. Ecol. 173, 111-121. Laing, I., 1995. Effect of food supply on oyster spatfall. Aquaculture 131, 315-324. Laing, I., Child, A.R., Achim, J., 1990. Nutritional value of dried algae diets for larvae of Manila clam (Tapes philippinarum). J. Mar. Biol. Ass. U. K. 70, 1-12. Laing, I., Earl, N.H., 1998. The lipid content, spatfall and subsequent growth of early and late settling hatchery-reared Pacific oyster, Crassostrea gigas Thunberg, larvae. Aquac. Res. 29, 19-25. Laing, I., Lopez-Alvarado, J., 1994. Effect of dried algae diets on conditioning and fecundity of Manila clam, Tapes philippinarum (Adams and Reeve). Aquac. Fish. Manage. 25, 157-166. Laing, I., Utting, S.D., 1994. The physiology and biochemistry of diploid and triploid Manila clam (Tapes philippinarum Adam & Reeve) larvae and juveniles. J. Exp. Mar. Biol. Ecol. 184, 159-169. Langdon, C., Waldock, M.J., 1981. The effect of algal and artificial diets on the growth and fatty acid composition of Crassostrea gigas spat. J. Mar. Biol. Ass. U. K. 61, 431-448. Lannan, J.E., Robinson, A., Breese, W.P., 1980. Broodstock management of Crassostrea gigas II. Broodstock conditioning to maximize larval survival. Aquaculture 21, 337-345. Lavens, P., Sorgeloos, P., 1996. Manual on the production and use of live food for aquaculture. FAO Fisheries Technical Paper. No. 361, Food and Agriculture Organization of the United Nations, Rome. Lawson, T.B., 1995. Fundamentals of aquaculture engineering, Chapman and Hall, New York. Li, C.C., Yeh, S.T., Chen, J.C., 2010. Innate immunity of the white shrimp Litopenaeus vannamei weakened by the combination of a Vibrio alginolyticus injection and low-salinity stress. Fish Shellfish Immun. 28, 121-127. Loosanoff, V.L., 1937. Seasonal gonadal changes of adult clams, Venus mercenaria (L.) Biol. Bull. 72, 406-416. Loosanoff, V.L., Davis, H.C., 1952. Temperature requirements for the maturation of gonads of the northern oyster. Biol. Bull. 103, 80-96. Loosanoff, V.L., Davis, H.C., 1963. Rearing of Bivalve Mollusks., U.S. Bureau of Commercial Fisheries Biological Laboratory, Milford, Connecticut.  204  Loosanoff, V.L., Davis, H.C., Chanley, P.E., 1953. Behavior of clam larvae in different concentrations of food organisms. Anat. Rec. 117, 586-587. Lowe, D.M., Moore, M.N., Bayne, B.L., 1982. Aspects of gametogenesis in the marine mussel Mytilus edulis (L.). J. Mar. Biol. Ass. U. K. 62, 133-145. Lubet, P., Faveris, R., Besnard, J.Y., Robbins, I., Duval, P., 1991. Annual reproductive cycle and recruitment of the scallop Pecten maximus (Linnaeus, 1758) from the Bay of Seine, in: Shumway, S.E., Sandifer, P.A. (Eds.), Scallop Biology and Culture. The World Aquaculture Society, Baton Rouge, pp. 87-94. MacDonald, B.A., Bacon, G.S., Ward, J.E., 1998. Physiological responses of infaunal (Mya arenaria) and epifaunal (Placopecten magellanicus) bivalves to variations in the concentration and quality of suspended particles II. Absorption efficiency and scope for growth. J. Exp. Mar. Biol. Ecol. 219, 127-141. MacKenzie, B.R., Leggett, W.C., 1991. Quantifying the contribution of small-scale turbulence to the encounter rates between larval fish and their zooplankton prey: effects of wind and tide. Mar. Ecol. Prog. Ser. 73, 149-160. Mackie, G.L., 1984. Bivalves, in: Tompa, A.S., Verdonk, N.H., Biggellaar, J.A.M.v.d. (Eds.), The Mollusca, Vol. 7. Reproduction. Academic Press, New York, pp. 351418. Mann, R., 1979a. The effect of temperature on growth, physiology, and gametogenesis in the Manila clam Tapes philippinarum (Adams & Reeve, 1850). J. Exp. Mar. Biol. Ecol. 38, 121-133. Mann, R., 1979b. Some biochemical and physiological aspects of growth and gametogenesis in Crassostrea gigas and Ostrea edulis grown at sustained elevated temperatures. J. Mar. Biol. Ass. U. K. 59, 95-110. Mann, R.G., SJ, 1978. The effect of temperature on growth and ammonia excretion of the Manila clam Tapes japonica. Estuarine and Coastal Marine Science 6, 335-339. Marasigan, E.T., Laureta, L.V., 2001. Broodstock maintenance and early gonadal maturation of Pholas orientalis (Bivalvia: Pholadidae). J. Shellfish Res. 20, 10951099. Marsden, I.D., 1999. Respiration and feeding of the surf clam Paphies donacina from New Zealand. Hydrobiologia 405, 179-188. Marshall, R., McKinley, S., Pearce, C.M., 2010. Effects of nutrition on larval growth and survival in bivalves. Rev. Aquac. 2, 33-55. Martínez-Fernández, E., Acosta-Salmón, H., Southgate, P.C., 2006. The nutritional value of seven species of tropical microalgae for black-lip pearl oyster (Pinctada margaritifera, L.) larvae. Aquaculture 257 491-503. Martinéz, G., Torres, M., Uribe, E., Díaz, M., Pérez, H., 1992. Biochemical composition of broodstock and early juvenile Chilean scallops, Argopecten purpuratus Lamarck, held in two different environments. J. Shellfish Res. 11, 307-313. Martínez, G., Aguilera, C., Mettifogo, L., 2000a. Interactive effects of diet and temperature on reproductive conditioning of Argopecten purpuratus broodstock. Aquaculture 183, 149-159. Martínez, G., Brokordt, K., Aguilera, C., Soto, V., Guderley, H., 2000b. Effect of diet and temperature upon muscle metabolic capacities and biochemical composition of gonad and muscle in Argopecten purpuratus Lamarck 1819. J. Exp. Mar. Biol. Ecol. 247, 29-49.  205  McLeod, R.J., Wing, T.R., 2008. Influence of an altered salinity regime on the population structure of two infaunal bivalve species. Estuar. Coast. Shelf Sci. 78, 529-540. Millar, R.H., Scott, J.M., 1967. The larva of the oyster Ostrea edulis during starvation. J. Mar. Biol. Ass. U. K. 47, 475-484. Miller, D.C., Poucher, S.L., Coiro, L., 2002. Determination of lethal dissolved oxygen levels for selected marine and estuarine fishes, crustaceans, and a bivalve. Mar. Biol. 140, 287-296. Mills, D., 2000. Combined effects of temperature and algal concentration on survival, growth and feeding physiology of Pinctada maxima (Jameson) spat. J. Shellfish Res. 19, 159-166. Monari, M., Serrazanetti, G.P., Foschi, J., Matozzo, V., Marin, M.G., Cattani, O., 2007. Effects of salinity on the clam Chamelea gallina haemocytes. Part II: Superoxide dismutase response. Mar. Biol. 151, 1059-1068. Moreno, G., Selvakumaraswamy, P., Byrne, M., Hoegh-Guldberg, O., 2001. A test of the ash-free dry weight technique on the developmental stages of Patiriella spp. (Echinodermata: Asteroidea). Limnol. Oceanogr. 46, 1214-1220. Mourente, G., Lubian, L.M., Odriozola, J.M., 1990. Total fatty acid composition as a taxonomic index of some marine microalgae used as food in marine aquaculture. Hydrobiologia 203, 147-154. Nakamura, Y., Hashizume, K., Koyama, K., Tamaki, A., 2005. Effects of salinity on sand burrowing activity, feeding and growth of the clams Mactra veneriformis , Ruditapes philippinarum and Meretrix lusoria. J. Shellfish Res. 24, 1053-1059. Navarro, J.M., Gonzalez, C.M., 1998. Physiological responses of the Chilean scallop Argopecten purpuratus to decreasing salinities. Aquaculture 167, 315-327. Navarro, E., Iglesias, J.I.P., Larrafiagam, A., 1989. Interannual variation in the reproductive cycle and biochemical composition of the cockle Cerastoderma edule from Mundaca Estuary (Biscay, North Spain). Mar. Biol. 101, 503-511. Navarro, J.M., Leiva, G.E., Martinez, G., Aguilera, C., 2000. Interactive effects of diet and temperature on the scope for growth of the scallop Argopecten purpuratus during reproductive conditioning. J. Exp. Mar. Biol. Ecol. 247, 67-83. Neufeld, D.S., Wright, S.H., 1996. Response of cell volume in Mytilus gill to acute salinity change. J. Exp. Biol. 199, 473-484. Neufeld, D.S., Wright, S.H., 1998. Effect of cyclical salinity changes on cell volume and function in Geukensia demissa gills. J. Exp. Biol. 201, 1421-1431. Nevejan, N., Courtens, V., Hauva, M., Gajardo, G., Sorgeloos, P., 2003b. Effect of lipid emulsions on production and fatty acid composition of eggs of the scallop Argopecten purpuratus. Mar. Biol. 143, 327-338. Nevejan, N., Pronker, A.E., Peene, F., 2008. Hatchery broodstock conditioning of the blue mussel Mytilus edulis (Linnaeus, 1758). Part II. New formulated feeds offer new perspectives to commercial hatcheries. Aquac. Int. 16, 483-495. Nevejan, N., Saez, I., Gajardo, G., Sorgeloos, P., 2003a. Supplementation of EPA and DHA emulsions to a Dunaliella tertiolecta diet: effect on growth and lipid composition of scallop larvae, Argopecten purpuratus (Lamarck, 1819). Aquaculture 217, 613-632. Newell, R.I.E., 1981. Molluscan bioenergetics - a synopsis, in: Pruder, G.D., Landgon, C., Conklin, D. (Eds.), Proceedings of the Second International Conference on  206  Aquaculture Nutrition: Biochemical and Physiological Approaches to Shellfish Nutrition. Special Publication No.2. Louisiana State University, Baton Rouge, Louisiana, pp. 252-271. Newell, R.I.E., Bayne, B.L., 1980. Seasonal changes in the physiology, reproductive condition and carbohydrate content of the cockle Cardium (= Cerastoderma) edule (Bivalvia: Cardiidae). Mar. Biol. 56, 11-19. Ngo, T.T.T., Kang, S., Kang, D., Sorgeloos, P., Choi, K., 2006. Effect of culture depth on the proximate composition and reproduction of the Pacific oyster, Crassostrea gigas from Gosung Bay, Korea. Aquaculture 253, 712-720. Okumus, I., Stirling, H.P., 1998. Seasonal variations in the meat weight, condition index and biochemical composition of mussels (Mytilus edulis L.) in suspended culture in two Scottish sea lochs. Aquaculture 159, 249-261. Orensanz, J.M., Hand, C.M., Parma, A.M., Valero, J., Hilborn, R., 2004. Precaution in the harvest of Methuselah’s clams - the difficulty of getting timely feedback from slow-paced dynamics. Can. J. Fish. Aquat. Sci. 61, 1355-1372. Palacios, E., Racotta, I.S., Kraffe, E., Marty, Y., Moal, J., Samain, J.F., 2005. Lipid composition of the giant lion's-paw scallop (Nodipecten subnodosus) in relation to gametogenesis I. Fatty acids. Aquaculture 250, 270-282. Paon, L.A., Kenchington, E.L.R., 1995. Changes in somatic and reproductive tissues during artificial conditioning of the sea scallop, Placopecten magellanicus (Gmelin, 1791). J. Shellfish Res. 14, 53-58. Parsons, T.R., Stephens, K., LeBrasseur, R.J., 1969. Production studies in the strait of Georgia. Part I. Primary production under the Fraser River plume, February to May, 1967. J. Exp. Mar. Biol. Ecol. 3, 27-38. Patil, V., Källqvist, T., Olsen, E., Vogt, G., Gislerød, H.R., 2007. Fatty acid composition of 12 microalgae for possible use in aquaculture feed. Aquac. Int. 26, 1-9. Pechenik, J.A., Eyster, L.S., Widdows, J., Bayne, B.L., 1990. The influence of food concentration and temperature on growth and morphological differentiation of blue mussel Mytilus edulis L. larvae. J. Exp. Mar. Biol. Ecol. 136, 47-64. Pérez-Camacho, A., Beiras, R., Albentosa, M., 1994. Effects of algal food concentration and body size on the ingestion rates of Ruditapes decussatus (Bivalvia) veliger larvae. Mar. Ecol. Prog. Ser. 115, 87-92. Pernet, F., Tremblay, R., Bourget, E., 2003a. Biochemical indicator of sea scallop (Placopecten magellanicus) quality based on lipid class composition. Part I: Broodstock conditioning and young larval performance. J. Shellfish Res. 22, 365375. Pernet, F., Tremblay, R., Bourget, E., 2003b. Biochemical indicator of sea scallop (Placopecten magellanicus) quality based on lipid class composition. Part II: larval growth, competency, and settlement. J. Shellfish Res. 22, 377-388. Pernet, F., Bricelj, V.M., Parrish, C.C., 2005. Effect of varying dietary levels of ω6 polyunsaturated fatty acids during early ontogeny of the sea scallop Placopecten magellanicus. J. Exp. Mar. Biol. Ecol. 327, 115-133. Pirini, M., Manuzzi, M.P., Pagliarani, A., Trombetti, F., Borgatti, A.R., Ventrella, V., 2007. Changes in fatty acid composition of Mytilus galloprovincialis (Lmk) fed on microalgal and wheat germ diets. Comp. Biochem. Physiol. B 147, 616-626.  207  Ponis, E., Parisi, G., Le Coz, J.R., Robert, R., Zittelli, G.C., Tredici, M., 2006c. Effect of the culture system and culture technique on biochemical characteristics of Pavlova lutheri and its nutritional value for Crassostrea gigas larvae. Aquac. Nutr. 12, 322-329. Ponis, E., Probert, I., Véron, B., Le Coz, J.R., Mathieu, M., Robert, R., 2006b. Nutritional value of six Pavlovophyceae for Crassostrea gigas and Pecten maximus larvae. Aquaculture 254, 544-553. Ponis, E., Probert, I., Veron, B., Mathieu, M., Robert, R., 2006a. New microalgae for the Pacific oyster Crassostrea gigas larvae. Aquaculture 253, 618-627. Ponis, E., Robert, R., Parisi, G., 2003a. Nutritional value of fresh and concentrated algal diets for larval and juvenile Pacific oysters (Crassostrea gigas). Aquaculture 221, 491-505. Ponis, E., Robert, R., Parisi, G., Tredici, M., 2003b. Assessment of the performance of Pacific oyster (Crassostrea gigas) larvae fed with fresh and preserved Pavlova lutheri concentrates. Aquac. Int. 11, 69-79. Powell, E.N., Bochenek, E.A., Klinck, J.M., Hofmann, E.E., 2002. Influence of food quality and quantity on the growth and development of Crassostrea gigas larvae: a modeling approach. Aquaculture 210, 89-117. Racotta, I.S., Ramirez, J.L., Avila, S., Ibarra, A.M., 1998. Biochemical composition of gonad and muscle in the catarina scallop, Argopecten ventricosus, after reproductive conditioning under two feeding systems. Aquaculture 163, 111-122. Rajesh, K.V., Mohamed, K.S., Kripa, V., 2001. Influence of algal cell concentration, salinity and body size on the filtration and ingestion rates of cultivable Indian bivalves. Indian J. Mar. Sci. 30, 87-92. Reid, H.I., Soudant, P., Lambert, C., Paillard, C., Birkbeck, T.H., 2003. Salinity effects on immune parameters of Ruditapes philippinarum challenged with Vibrio tapetis. Dis. Aquat. Org. 56, 249-258. Reidy, R., 2011. The potential impact of sea otters (Enhydra lutris) on the British Columbia geoduck clam (Panopea abrupta) fishery, School of Resource and Environmental Management. Simon Fraser University, Burnaby, pp. 118. Ren, J.S., Marsden, I.D., Ross, A.H., Schiel, D.R., 2003. Seasonal variation in the reproductive activity and biochemical composition of the Pacific oyster (Crassostrea gigas) from the Marlborough Sounds, New Zealand. N. Z. J. Mar. Freshw. Res. 37, 171-182. Rico-Villa, B., Le Coz, J.R., Mingant, C., Robert, R., 2006. Influence of phytoplankton diet mixtures on microalgae consumption, larval development and settlement of the Pacific oyster Crassostrea gigas (Thunberg). Aquaculture 256, 377-388. Riisgård, H.U., 1988. Feeding rates in hard clam (Mercenaria mercenaria) veliger larvae as a function of algal (Isochrysis galbana) concentration. J. Shellfish Res. 7, 377380. Riisgård, H.U., Randløv, A., 1981. Energy budgets, growth and filtration rates in Mytilus edulis at different algal concentrations. Mar. Biol. 61, 227-234. Riisgård, H.U., Seerup, D.F., 2003. Filtration rates in the soft clam Mya arenaria: effects of temperature and body size. Sarsia 88, 415-428.  208  Rivero-Rodríguez, S., Beaumont, A.R., Lora-Vilchis, M.C., 2007. The effect of microalgal diets on growth, biochemical composition, and fatty acid profile of Crassostrea corteziensis (Hertlein) juveniles. Aquaculture 263, 199-210. Robert, R., Gérard, A., 1999. Bivalve hatchery technology: the current situation for the Pacific oyster Crassostrea gigas and the scallop Pecten maximus in France. Aquat. Living Resour. 12, 121-130. Robert, R., His, E., Dinet, A., 1988. Combined effects of temperature and salinity on fed and starved larvae of the European flat oyster Ostrea edulis. Mar. Biol. 97, 95100. Robert, R., Trintignac, P., 1997. Substitutes for live microalgae in mariculture: a review. Aquat. Living Resour. 10, 315-327. Robert, R., Trut, G., Laborde, J.L., 1993. Growth, reproduction and gross biochemical composition of the Manila clam Ruditapes philippinarum in the Bay of Arcachon, France. Mar. Biol. 116, 291-299. Robinson, A., 1992a. Gonadal cycle of Crassostrea gigas kumamoto (Thunberg) in Yaquina Bay, Oregon and optimum conditions for broodstock oysters and larval culture. Aquaculture 106, 89-97. Robinson, A., 1992b. Dietary supplements for reproductive conditioning of Crassostrea gigas kumamoto (Thunberg). I. Effects on gonadal development, quality of ova and larvae through metamorphosis. J. Shellfish Res. 11, 437-441. Robinson, A., 1992c. Dietary supplements for the reproductive conditioning of Crassostrea gigas kumamoto (Thunberg): II Effects of glycogen, lipids and fatty acid content of broodstock oysters and eggs. J. Shellfish Res. 11, 443-447. Roldán-Carrillo, L.M., Maeda-Martínez, A.N., Massó-Rojas, A., Sicard-González, M.T., 2005. Salinity tolerance and resistance of the Pacific lion's paw scallop (Nodipecten subnodosus) and the relationships with species distribution and density in a coastal lagoon. J. Shellfish Res. 24, 353-361. Ropes, J.W., Stickney, A.P., 1965. Reproductive cycle of Mya arenaria in New England. Biol. Bull. 128, 315-327. Rupp, G.S., Parsons, G.J., 2004. Effects of salinity and temperature on the survival and byssal attachment of the lion's paw scallop Nodipecten nodosus at its southern distribution limit. J. Exp. Mar. Biol. Ecol. 309, 173-198. Rupp, G.S., Parsons, G.J., Thompson, R.J., de Bem, M.M., 2005. Influence of environmental factors, season and size at deployment on growth and retrieval of postlarval lion's paw scallop Nodipecten nodosus (Linnaeus, 1758) from a subtropical environment. Aquaculture 243, 195-216. Santos, J.M., Downing, S.L., Chew, K.K., 1993. Studies on the effects of water temperature on the sexual development of adult Olympia oysters, Ostrea lurida. World Aquac. 24, 43-46. Secor, S., 2009. Specific dynamic action: a review of the postprandial metabolic response. J. Comp. Physiol., B 179, 1-56. Shaul, W., 1981. Methods for culturing algae and rearing bivalve larvae and juveniles at the Point Whitney Shellfish Laboratory, 1976-1981, Washington Department of Fisheries. Shaul, W., Goodwin, L., 1982. Geoduck (Panope generosa: Bivalvia) age as determined by internal growth lines in the shell. Can. J. Fish. Aquat. Sci. 39, 632-636.  209  Shumway, S.E., 1977. Effect of salinity fluctuation on the osmotic pressure and Na +, Ca2+ and Mg2+ ion concentrations in the hemolymph of bivalve molluscs. Mar. Biol. 41, 153-177. Shumway, S.E., 1983. Factors affecting oxygen consumption in the coot clam Mulinia lateralis (Say). Ophelia 22, 143-171. Shumway, S.E., Koehn, R.K., 1982. Oxygen consumption in the American oyster Crassostrea virginica. Mar. Ecol. Prog. Ser. 9, 59-68. Shumway, S.E., Youngson, A., 1979. The effects of fluctuating salinity on the physiology of Modiolus demissus (Dillwyn). J. Exp. Mar. Biol. Ecol. 40, 167-181. Shurova, N.M., 2001. Influence of salinity on the structure and the state of bivalve Mytilus galloprovincialis populations. Russ. J. Mar. Biol. 27, 151-155. Sloan, N.A., Robinson, S.M.C., 1984. Age and gonad development in the geoduck clam Panope abrupta (Conrad) from Southern British Columbia, Canada. J. Shellfish Res. 4, 131-137. Sobral, P., Widdows, J., 2000. Effects of increasing current velocity, turbidity and particle-size selection on the feeding activity and scope for growth of Ruditapes decussatus from Ria Formosa, southern Portugal. J. Exp. Mar. Biol. Ecol. 245, 111-125. Somero, G.N., Bowlus, R.D., 1983. Osmolytes and metabolic end products of molluscs: the design of compatible solute systems, in: Hochachka, P.W. (Ed.), The Mollusca. Academic Press Inc., New York, pp. 77-100. Soudant, P., Marty, Y., Moal, J., Robert, R., Quéré, C., Le Coz, J.R., Samain, J.F., 1996c. Effect of food fatty acid and sterol quality on Pecten maximus gonad composition and reproduction process. Aquaculture 143, 361-378. Soudant, P., Marty, Y., Moal, J., Samain, J.-F., 1996a. Fatty acids and egg quality in great scallop. Aquac. Int. 4, 191-200. Soudant, P., Moal, J., Marty, Y., Samain, J.F., 1996b. Impact of the quality of dietary fatty acids on metabolism and the composition of polar lipid classes in female gonads of Pecten maximus (L.). J. Exp. Mar. Biol. Ecol. 205, 149-163. Spencer, B.E., Edwards, D.B., 1995. Observations on simulated depuration trials with Manila clam, Tapes philippinarum Adams and Reeve. Aquac. Res. 26, 249-263. Steffensen, J.F., 1989. Some errors in respirometry of aquatic breathers: how to avoid and correct for them. Fish Physiol. Biochem. 6, 49-59. Stickle, W.B., Sabourin, T.D., 1979. Effects of salinity on the respiration and heart rate of the common mussel, Mytilus edulis L., and the black chiton, Katharina tunicata (Wood). J. Exp. Mar. Biol. Ecol. 41, 257-268. Strange, K.B., Crowe, J.H., 1979. Acclimation to successive short term salinity changes by the bivalve Modiolus demissus. I. Changes in hemolymph osmotic concentration, hemolymph ion concentration and tissue water content. J. Exp. Zool. 210, 221-226. Strasser, C.A., Mullineaux, L.S., Thorrold, S.R., 2008. Temperature and salinity effects on elemental uptake in the shells of larval and juvenile softshell clams Mya arenaria. Mar. Ecol. Prog. Ser. 370, 155-169. Sukhotin, A.A., Flyachinskaya, L.P., 2009. Aging reduces reproductive success in mussels Mytilus edulis. Mechanisms of Ageing and Development 130, 754-761.  210  Takahashi, M., Barwell-Clarke, J., Whitney, F., Koeller, P., 1978. Winter condition of marine plankton populations in Saanich inlet, B.C., Canada. I. Phytoplankton and its surrounding environment. J. Exp. Mar. Biol. Ecol. 31, 283-301. Tammi, K.A., Soars, S., Tumor, W., Rice, M., 1995. Spawning and spat collection of the bay scallop, Argopecten irradians, in the Westport River estuary, Massachusetts, World Aquaculture Society Book of Abstracts, World Aquaculture 1995, San Diego, California, U.S.A., pp. 223. Tang, B., Liu, B., Wang, G., Zhang, T., Xiang, J., 2006. Effects of various algal diets and starvation on larval growth and survival of Meretrix meretrix. Aquaculture 254, 526-533. Teshima, S.-I., Kanazawa, A., Shimamoto, R., 1988. Anatomical distribution of sterols and fatty acids in the bivalve Mactra chinensis. Nippon Suisan Gakkaishi 54, 293297. Thompson, P.A., Harrison, P.J., 1992. Effects of monospecific algal diets of varying biochemical composition on the growth and survival of Pacific oyster (Crassostrea gigas) larvae. Mar. Biol. 113, 645-654. Thompson, R.E., 1981. Oceanography of the British Columbia coast. Can. Spec. Publ. Fish. Aquat. Sci. 56, 291 p. Thorarinsdóttir, G.G., 1991. The Iceland scallop, Chlamys islandica (O.F. Müller) in Breidafjördur, west Iceland. I. Spat collection and growth during the first year. Aquaculture 97, 13-23. Toba, D.R., Thompson, D.S., Chew, K.K., Anderson, G.J., Miller, M.B., 1992. Guide to Manila clam culture in Washington, Washington Sea Grant Program, University of Washington. Tomiyama, T., Komizunai, N., Shirase, T., Ito, K., Omori, M., 2008. Spatial intertidal distribution of bivalves and polychaetes in relation to environmental conditions in the Natori River estuary, Japan. Estuar. Coast. Shelf Sci. 80, 243-250. Torkildsen, L., Magnesen, T., 2004. Hatchery production of scallop larvae (Pecten maximus) - survival in different rearing systems. Aquac. Int. 12, 489-507. Torkildsen, L., Samuelsen, O.B., Lunestad, L.D., Bergh, O., 2000. Minimum inhibitory concentrations of chloramphenicol, florfenicol, trimethoprim/sulfadiazine and flumequine in seawater of bacteria associated with scallops (Pecten maximus) larvae. Aquaculture 185, 1-12. Utting, S.D., 1986. A preliminary study on growth of Crassostrea gigas larvae and spat in relation to dietary protein. Aquaculture 56, 123-138. Utting, S.D., 1992. Procedures for the maintenance and hatchery-conditioning of bivalve broodstocks. J. Shellfish Res. 11, 208. Utting, S.D., 1993. Procedures for the maintenance and hatchery-conditioning of bivalve broodstocks. World Aquac. 24, 78-82. Utting, S.D., Doyou, J., 1992. The increased utilization of egg lipid reserves following induction of triploidy in the Manila clam (Tapes philippinarum). Aquaculture 103, 17-28. Utting, S.D., Millican, P.F., 1997. Techniques for the hatchery conditioning of bivalve broodstocks and the subsequent effect on egg quality and larval viability. Aquaculture 155, 45-54.  211  Utting, S.D., Millican, P.F., 1998. The role of diet in hatchery conditioning of Pecten maximus L.: a review. Aquaculture 165, 167-178. Utting, S.D., Spencer, B.E., 1991. The hatchery culture of bivalve mollusc larvae and juveniles, Laboratory Leaflet, Ministry of Agriculture, Fisheries and Food Directorate of Fisheries Research, Lowestoft (68). van der Molen, S., Kroeck, M., Ciocco, N.F., 2007. Reproductive cycle of the southern geoduck clam, Panopea abbreviata (Bivalvia: Hiatellidae), in north Patagonia Argentina. Invertebr. Repr. Dev. 50, 75-84. Van Winkle, W., 1968. Effects of season temperature and salinity on oxygen consumption of bivalve gill tissue. Comp. Biochem. Physiol. 26, 69-80. VanKoeveringe, M.A.H., 1998. Molecular population genetics of British Columbia geoduck clams, Panope abrupta, based on mitochondrial DNA sequences. Graduate Thesis, Biological Sciences. Simon Fraser University, Burnaby, pp. 59. Velasco, L.A., Navarro, J.M., 2002. Feeding physiology of infaunal (Mulinia edulis) and epifaunal (Mytilus chilensis) bivalves under a wide range of concentrations and qualities of seston. Mar. Ecol. Prog. Ser. 240, 143-155. Villalejo-Fuerte, M., Ceballos-Vazquez, B.P., Garcia-Dominguez, F., 1996. Reproductive cycle of Laevicardium elatum (Sowerby, 1833) (Bivalvia: Cardiidae) in Bahia Concepcion, Baja California Sur, Mexico. J. Shellfish Res. 15, 741-745. Walne, P.R., Mann, R., 1975. Growth and biochemical composition in Ostrea edulis and Crassostrea gigas. in: Barnes, H. (Ed.), Ninth European Marine Biology Symposium. Aberdeen University Press, Scotland, pp. 587-607. Wang, W.X., Widdows, J., 1991. Physiological responses of mussel larvae Mytilus edulis to environmental hypoxia and anoxia. Mar. Ecol. Prog. Ser. 70, 23-236. Ward, E.J., Shumway, S.E., 2004. Separating the grain from the chaff: particle selection in suspension- and deposit-feeding bivalves. J. Exp. Mar. Biol. Ecol. 300, 83-130. Ward, J.E., Cassell, H.K., MacDonald, B.A., 1992. Chemoreception in the sea scallop Placopecten magellanicus (Gmelin). I. Stimulatory effects of phytoplankton metabolites on clearance and ingestion rates. J. Exp. Mar. Biol. Ecol. 163, 235250. Watts, S.A., Lawrence, J.M., 1990. The effect of reproductive state, temperature, and salinity on DNA and RNA levels and activities of metabolic enzymes of the pyloric ceca in the sea star Luidia clathrata (Say). Physiol. Zool. 63, 1196-1215. Webb, K.L., Chu, F.-L., 1981. Phytoplankton as a food source for bivalve larvae, in: Pruder, G.D., Landgon, C., Conklin, D. (Eds.), Procedings of the Second International Conference on Aquaculture Nutrition: Biochemical and Physiological Approaches to Shellfish Nutrition. Special Publication No.2. Louisiana State University, Baton Rouge, Louisiana, pp. 272-291. Weisbart, M., Chakraborti, P.K., Chakraborti, A., Huntley, R.M., Maneckjee, A., McLeese, J.M., 1994. Steroid receptors in fish: membrane and intracellular preparations, in: Hochachka, P.W., Mommsen, T.P. (Eds.), Biochemistry and Molecular Biology of Fishes. Elsevier Science, Amsterdam, pp. 457-468. Whyte, J.N.C., Bourne, N., Ginther, N.G., 1990. Biochemical and energy changes during embryogenesis in the rock scallop Crassadoma gigantea. Mar. Biol. 106, 239244.  212  Wikfors, G.H., Alix, J.H., Dixon, M.S., Smith, B.C., 2004. The northern quahog Mercenaria mercenaria has slower growth and poorer conversion efficiency than oysters or scallops fed identical dietary rations and regimes. J. Shellfish Res. 23, 317. Wilson, J.A., Chaparro, O.R., Thompson, R.J., 1996. The importance of broodstock nutrition on the viability of larvae and spat in the Chilean oyster Ostrea chilensis. Aquaculture 139, 63-75. Winter, J.E., 1978. A review on the knowledge of suspension-feeding in lamellibranchiate bivalves, with special reference to artificial aquaculture systems. Aquaculture 13, 1-33. Wright, D.A., Hetzel, E.W., 1985. Use of RNA:DNA ratios as an indicator of nutritional stress in the American Oyster Crassostrea virginica. Mar. Ecol. Prog. Ser. 25, 199-206. Yan, X., Zhang, G., Yang, F., 2006. Effects of diet, stocking density and environmental factors on growth, survival, and metamorphosis of Manila clam Ruditapes philippinarum larvae. Aquaculture 253, 350-358. Yukihira, H., Klumpp, D.W., Lucas, J.S., 1998. Effects of body size on suspension feeding and energy budgets of the pearl oysters Pinctada margaritifera and P. maxima. Mar. Ecol. Prog. Ser. 170, 119-130. Zhang, Z., Campbell, A., 2004. Natural mortality and recruitment rates of the Pacific geoduck clam (Panopea abrupta) in experimental plots. J. Shellfish Res. 23, 675682. Zhuang, S., Wang, Z.Q., 2004. Influence of size, habitat and food concentration on the feeding ecology of the bivalve, Meretrix meretrix Linnaeus. Aquaculture 241, 689-699. Zhuang, S., Zhang, M., Zhang, X., Wang, Z., 2004. The influence of body size, habitat and diet concentration on feeding of Laternula marilina Reeve. Aquac. Res. 35, 622-628.  213  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0072611/manifest

Comment

Related Items