Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanisms and energetic implications of osmoregulation in embryos and larvae of Chum and Coho salmon Groot, Erick P. 1998

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

Item Metadata

Download

Media
831-ubc_1998-345270.pdf [ 13.25MB ]
Metadata
JSON: 831-1.0089121.json
JSON-LD: 831-1.0089121-ld.json
RDF/XML (Pretty): 831-1.0089121-rdf.xml
RDF/JSON: 831-1.0089121-rdf.json
Turtle: 831-1.0089121-turtle.txt
N-Triples: 831-1.0089121-rdf-ntriples.txt
Original Record: 831-1.0089121-source.json
Full Text
831-1.0089121-fulltext.txt
Citation
831-1.0089121.ris

Full Text

M E C H A N I S M S A N D E N E R G E T I C IMPLICATIONS OF  OSMOREGULATION  I N E M B R Y O S A N D L A R V A E OF C H U M A N D C O H O S A L M O N by ERICK PETER GROOT  B . S c , The University of Victoria, 1982 M . S c , The University of British Columbia, 1989  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY  IN  T H E FACULTY OF GRADUATE STUDIES (Department of Animal Science) _We accept this thesis as conforming to the require^ stancjartp  T H E UNIVERSITY OF BRITISH COLUMBIA August 1998 © Erick Peter Groot, 1998  In  presenting  this  thesis  in  degree at the University of  partial  fulfilment  of  the  requirements  for  an advanced  British Columbia, I agree that the Library shall make it  freely available for reference and study. I further agree that permission for extensive copying of this thesis for department  or  by  his  or  scholarly purposes may be granted her  representatives.  It  is  by the head of  understood  that  copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department of The University of British Columbia Vancouver, Canada  DE-6 (2/88)  ABSTRACT Mechanisms and the energetic implications of osmoregulation were investigated in the early life stages of chum (Oncorhynchus keta) and coho salmon (O. kisutch) at three different developmental stages: eyed embryo, prehatch embryo, and yolk sac larva. Embryos and larvae were acclimated (>7 d) to selected ranges of saltwater (SW) concentrations (0 to 30 %oS) in order to examine whole-animal (Chapter 1), cellular (Chapter 2), and biochemical (Chapter 3) aspects of early life osmoregulation Chum embryos and larvae survived exposure to SW at higher salinities than coho embryos and larvae. Saltwater-challenged chum fry were able to successfully hypo-osmoregulate in 35 %oS for 24 h (plasma osmolality ~ 350 mOsm) whereas coho fry were only able to tolerate 24 %oS and were less capable at effective hypo-osmoregulation (plasma osmolality ~ 440 mOsm). Measurements of whole-animal metabolism and saltwater tolerance showed that chum salmon were markedly more tolerant to SW than coho salmon and that SW acclimation had markedly different effects at each of the three developmental stages tested. Saltwater acclimation had the most marked effect on chum salmon larvae with increased oxygen consumption rates (Mo ) of 2  up to 53% in 30 %oS. This increased energy expenditure did not appear to be the result of increased swimming activity or changes in energy allocation to growth processes. It was postulated that evidence for increased energy expenditure might be revealed by examining cellular (Chapter 2) and biochemical (Chapter 3) aspects of early life osmoregulation. Examination of cutaneous and branchial epithelia using fluorescent microscopy (DASPEI) showed an extensive distribution of cutaneous chloride cells (CCs) with relatively high densities (-500 to -1350 CC-mrn ) on all major embryonic and larval cutaneous surfaces. Examination of -2  CCs in prehatch chum and coho embryos, using transmission electron microscopy, showed cellular fine structure that was typical of CCs in adult fish acclimated to similar conditions.  However, SW-acclimated coho embryos appeared to have fewer CCs with typical SWacclimated fine structure than chum embryos. These data along with estimated and modelled changes in total whole-body numbers of cutaneous and branchial CCs, support the theory that cutaneous epithelia are the site of early life osmoregulatory processes until the gills are fully developed in the fry stage. Some preliminary data and speculation are provided that support a recent suggestion in the literature that the larval gill may function as an ion regulatory organ before it assumes its primary role as a respiratory structure in juvenile and adult fish. This study is the first to examine Na ,K -ATPase and H -ATPase activity in embryos and +  +  +  larvae and shows that these enzymes function at activity levels that are similar to adult fish. Ontogenic changes in Na ,K -ATPase and H -ATPase activity in chum salmon incubated in F W +  +  +  increased by 16 times progressively throughout development, from the eyed embryo stage to the fry stage. Acclimation to SW appeared to increase Na ,K -ATPase activity in the branchial +  +  epithelia of chum larvae but not coho. No obvious trends related to SW acclimation were observed in the cutaneous tissues at any of the three developmental stages. H -ATPase activity +  was highest in the branchial epithelium of yolk sac larvae in FW, specifically coho salmon. Based on H -ATPase activity changes in SW, it is suggested that this enzyme also functions as +  part of the ion uptake mechanism in F W CCs of embryos and larvae. Yolk sac epithelium is proposed for use as a model F W ion regulatory epithelium that is flat and contains CCs. Chloride cell density and Na ,K -ATPase and H -ATPase activity in general did not correlate +  +  +  strongly with changes in osmotic concentration and thus did not provide any additional explanation for the observed changes in Mo reported in Chapter 1. Collectively the evidence 2  supports the theory that cutaneous osmoregulation is significant during the early life stages of salmonid development but that the energy consumed by ion transport processes is not high in relation to whole-animal metabolism.  T A B L E OF C O N T E N T S Abstract Table of Contents List of Tables List of Figures List of Abbreviations List of Species Acknowledgements Dedication General Introduction  :  Chapter 1 Energetic Implications of Osmoregulation in Embryos and Larvae of Chum and Coho Salmon Introduction Methods Egg collection, incubation and rearing Acute salinity tests 1994/95 - experimental design Acclimated salinity tests 1993/94 - experimental design Mortality rates Saltwater challenge tests - fry Oxygen consumption measurements (M0 ) Dry weight and growth measurement Data analysis Results Development and mortality rates Plasma osmolality measurements - saltwater challenge tests Oxygen consumption rates in different salinities - acute exposure Oxygen consumption rates in different salinities - acclimated exposure Growth in different salinities Discussion Salt Water Tolerance and Osmoregulatory Ability of Embryos, Larvae, and Fry Interspecific and intraspecific comparisons of oxygen consumption rates Effect of SW exposure on M0 Energetic implications of osmoregulation - embryos Energetic implications of osmoregulation - larvae Figures - Chapter 1 2  2  ii iv vi vii x xi xii xiii 1 16 16 16 20 20 21 22 24 25 26 28 29 30 30 32 33 .35 37 39 39 41 44 48 57 .64  Chapter 2 71 Distribution and Ontogeny of Cutaneous and Branchial Chloride Cells in Chum and Coho Salmon Embryos and Larvae 71 Introduction 71 Methods 77 Fish 77 Fluorescent microscopy - tissue preparation 78 Fluorescent microscopy - analysis 79  iv  Cutaneous versus branchial surface comparisons - total numbers of chloride cells.... 81 Electron Microscopy - tissue preparation and analysis 82 Results 84 Chloride cell distribution - fluorescent microscopy 84 DASPEI staining - tissue and developmental stage differences 89 Branchial chloride cells 90 Chloride cell structure - T E M 91 Discussion 92 Intra- and interspecific comparisons in salmonids 94 Relative numbers of cutaneous compared with branchial CCs 98 Ventilation of cutaneous surfaces 104 Allometric considerations ...106 Variations in Staining effectiveness and C C visibility 109 Figures - Chapter 2 113 Chapter 3 Na ,K -ATPase and H -ATPase Activity in Gill, Skin, and Yolk Sac Epithelium of Chum and Coho Salmon Embryos and Larvae in Relation to Osmoregulation Introduction Methods Fish Tissue preparation Tissue analysis Statistical analysis Results Ontogeny of gill ATPase activity in F W - chum salmon Na ,K -ATPase activity in different salinities H -ATPase activity Discussion Na ,K -ATPase activity in embryos and larvae H -ATPase activity in embryos and larvae Ion regulation in embryos and larvae Figures - Chapter 3 General Discussion and Conclusions References  124 124 128 128 128 129 132 134 134 134 136 138 138 145 147 151 158 168  Appendix A A Microrespirometer Designed for Energetic Studies of Fish Embryos and Larvae Introduction Some General Considerations for Energetics Studies Respirometer Design and Operation Respirometer design Respirometer operation Computerized data acquisition Data analysis  182 182 182 186 189 189 192 196 196  +  +  124  +  +  +  +  +  +  +  v  L I S T OF T A B L E S  Table 1.1. Test salinities used for acclimation and respirometer testing of chum and coho embryos and larvae during 1993/94 and 1994/95 experiments Table 1.2. Median resistance times (LT ) for chum and coho salmon embryos, and larvae exposed to a range of salinities  23  50  ....31  Table 1.3. Oxygen consumption rates of chum and coho salmon embryos and larvae exposed to various test salinity regimes 36 Table 1.4. Summary of oxygen consumption rates of salmon embryos and larvae from the literature  43  Table 2.1. Chloride cell densities on the skin of chum salmon embryos and larvae pooled across salinity treatments for each developmental stage  85  Table 2.2. Chloride cell densities on the skin of coho salmon embryos and larvae pooled across salinity treatments for each developmental stage  86  Table 2.3. Surface area estimates, mean whole-body C C density, and total C C number per individual for cutaneous and branchial surfaces of chum salmon embryos, larvae, and fry. 102 Table 2.4. Estimates of chloride cell diameter during the early life stages offish as measured from published studies in the literature 108  Table 3.1. Sampling details of tissues collected for Na ,K -ATPase and H -ATPase analyses. 130 +  +  +  vi  LIST OF FIGURES Fig. G I 1 . Diagram of active and passive exchange of ion and water in freshwater and marine teleost fish  3  Fig. GI 2. Diagrams of current models of chloride cell function in a marine and freshwater teleost fish  4  Fig. 1.1. Mortality rates of eyed and prehatch embryos and yolk sac larvae of chum and coho salmon exposed to a selected range of salinities 64 Fig. 1.2. Cumulative hatching rate of chum and coho salmon eggs acclimated as prehatch embryos to a selected range of salinities  65  Fig. 1.3. Mean plasma osmolalities of chum and coho salmon fry exposed to a 24-h SW challenge of 35 %oS and 24 %oS, respectively  66  Fig. 1.4. Mean plasma osmolalities of long term acclimated chum and coho salmon fry exposed to a 24-h SW challenge of 35 and 24 % o S , respectively  67  Fig. 1.5. Mean mass-specific oxygen consumption rates (M0 ) and mean tissue dry weights (TDW) of chum salmon eyed embryos, prehatch embryos, and 50% yolk sac larvae acclimated for 7 to 10 d to a selected range of salinities  68  Fig. 1.6. Mean mass-specific oxygen consumption rates (M0 ) and mean tissue dry weights (TDW) of coho salmon eyed embryos, prehatch embryos, and 50% yolk sac larvae acclimated for 7 to 10 d to a selected range salinities  69  2  2  Fig. 1.7. Mean percent tissue (Pj) for chum and coho salmon embryos and larvae acclimated to a selected range of salinities 70  Fig. 2.1. A generalized diagram of a salmon yolk sac embryo showing the various cutaneous tissues sampled for C C distribution and density in chum and coho embryos and larvae using DASPEI fluorescent microscopy 113 Fig. 2.2. Reference diagrams showing specific locations of DASPEI stained chloride cell photographs  114  Fig. 2.3. Y S E tissue of a eyed coho embryo 5 %oS showing change in C C density from Y S E to Y S E zone 115 m  V  Fig. 2.4. Y S E of chum prehatch embryo 0 %oS showing well rounded and relatively bright CCs with discrete edges  115  vii  LIST OF FIGURES  (CONT'D)  Fig. 2.5. Y S E tissue of a coho yolk sac larva 6 %oS showing the amorphous and irregular shaped and faint CCs  116  Fig. 2.6. TRE tissue showing numerous, relatively evenly distributed CCs on the main trunk in the lower neck region of an eyed coho embryo 10 %oS 116 Fig. 2.7. External OPE tissue of a prehatch chum embryo 0 %oS showing high density of evenly spaced CCs  117  Fig. 2.8. H D E tissue located directly above the eye in an eyed coho embryo 15 %oS showing evenly spaced, bright CCs 117 Fig. 2.9. Internal OPE tissue from a prehatch chum embryo 18 %oS showing specific location of CCs in the external most layer of the integument 118 Fig. 2.10. TRE tissue of coho yolk sac larva; similar CC densities as seen in earlier stages but increased variation in CC size compare with Fig. 2.6. Inset: higher magnification of same tissue but different specimen, YS larva, 0 %oS  118  Fig. 2.11. Chloride cell located in the gill epithelium of a prehatch chum embryo incubated in F W  119  Fig. 2.12. Chloride cell located in the gill epithelium of a prehatch chum embryo acclimated (7 d) to 24 %oS SW 119 Fig. 2.13 Internal OPE tissue of a coho yolk larva 6 %oS showing irregularly shaped and variably sized CCs  120  Fig. 2.14. Dorsal end of gill arch from newly eyed chum embryo 5 %oS showing what appears to be relatively close association of CCs with rudimentary gill blood circulation. ..120 Fig. 2.15. Gill arch from eyed chum embryo 10 %oS: numerous CCs are located along the developing filaments as well as the gill arch dorsal end  121  Fig. 2.16. Gill filaments of a prehatch chum embryo 12%oS showing CCs grouped at the base of presumptive secondary lamellae 121 Fig. 2.17. Chloride cells on the gill filaments of yolk sac larvae; top view of filament showing CCs lined up along presumptive secondary lamellae, chum 12 %oS and coho 0%oS  ;  122  Fig. 2.18. Total numbers of chloride cells (CC) per individual in cutaneous and branchial epithelia, and the relative relationships of total cutaneous and branchial surface areas and total CC numbers during the early life stages of chum salmon 123  viii  LIST OF FIGURES  (CONT'D)  Fig. 3.1. Ontogeny of Na ,K -ATPase and H -ATPase activity in branchial, trunk, and yolk sac epithelium tissue of chum salmon embryos and larvae  151  Fig. 3.2. Branchial epithelium Na ,K -ATPase activity in chum and coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities  152  Fig. 3.3. Trunk epithelium Na ,K -ATPase activity in chum and coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities  153  Fig. 3.4. Yolk sac epithelium Na ,K -ATPase activity in chum and coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities  154  Fig. 3.5. Branchial epithelium FT-ATPase activity in chum and coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities  155  +  +  +  +  +  +  +  +  +  Fig. 3.6. Trunk epithelium H -ATPase activity in chum and coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities 156 +  Fig. 3.7. Yolk sac epithelium FT-ATPase activity in chum and coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities  157  ix  L I S T OF A B B R E V I A T I O N S %oS ATP ATU  = parts per thousand salinity (g/L) = adenosine triphosphate = accumulated thermal unit; also equal to 'degree-day' — average daily incubation temperature above 0 °C multiplied by number of days (°C-d) Ca = calcium ion Cl~ = chloride ion dH 0 = distilled water dPOH = days posthatch dPRH = days prehatch FFE = finfold epithelium (only in two embryonic development stages) F N E = fin epithelium (only in larval development stage) FW = freshwater and fresh water 2 +  2  HDE  = head epithelium [subdivided into 2 zones of different CCs densities (HDE : general and H D E : cranium, see text)] K = potassium ion L D H = lactate dehydrogenase M0 = mass specific oxygen consumption rate (e.g., nmolmg T D W - h ) Na = sodium ion N A D H = nicotinamide adenine dinucleotide - reduced form N E M = A^-ethylmaleimide OPE = opercular epithelium [separated into external (OPE ) and internal (OPEj) surfaces] P s critical oxygen concentration PEP = phosphoenolpyruvate PK = pyruvate kinase p0 = partial pressure of oxygen P R H = prehatch P = percent embryonic or larval tissue SD = standard deviation S E M = scanning electron microscope or microscopy SW = saltwater and salt water T D W = tissue dry weight (i.e., formalin-fixed embryo or larva dried in 60°C oven for 48 h; includes the yolk sac epithelium but not the contents) T E M s transmission electron microscope or microscopy TMS = tricaine methane sulfonate "MS222" (Sigma Chemical Company, St. Louis, M O , USA) TRE = trunk (body) epithelium; examined for CCs using fluorescent microscopy T W W = tissue wet weight; calculated as 6 times TDW, assuming an 83% moisture content Y D W = yolk dry weight (formalin-fixed yolk material processed the same way as TDW) Y S E = yolk sac epithelium; [subdivided into 3 zones of different CC densities ( Y S E : dorsal, Y S E : middle, and Y S E : ventral) see text] g  C  +  -1  _1  2  +  e  c  2  r  d  m  V  x  LIST OF SPECIES Ambassis (Ambassis interrupta, Bleeker) American eel (Anguilla rostrata, Lesueur) American shad (Alosa sapidissima) Atlantic herring (Clupea harengus, Linnaeus) ayu (Plecoglossus altivelis, Temminck and Schlegel) carp (Cyprinus carpio, Linnaeus) milkfish (Chanos chanos, Forsskal) flounder, marine (Kareius bicoloratus, Basilewsky) goby (Chaenogobius urotaenia, [Chaenogobius annularis]) grubby (Myoxocephalus aenaeus) guppy (Poecilia reticulata, Peters) longhorn sculpin (Myoxocephalus octodecemspinosus) mudsucker (Gillichthys mirabilis) mummichog (killifish) (Fundulus heteroclitus, Linnaeus) Pacific sardine (Sardinops caerulea, Girard) plaice (Pleuronectes platessa, Linnaeus) salmon, Atlantic (Salmo salar, Linnaeus) salmon, chinook (Oncorhynchus tshawytscha, Walbaum) salmon, chum (Oncorhynchus keta, Walbaum) salmon, coho (Oncorhynchus kisutch, Walbaum) salmon, pink (Oncorhynchus gorbuscha, Walbaum) salmon, sockeye (Oncorhynchus nerka, Walbaum) sheepshead minnow (Cyprinodon variegatus, Lacepede) snook (Centropomus undecimalis, Bloch) (Mugil cephalus, Linnaeus) striped mullet tilapia, Florida red hybrid (Oreochromis mossambicus x O. hornorum) (Oreochromis mossambicus, Peters) tilapia, Mozambique (Oncorhynchus mykiss, Walbaum) trout, rainbow (Oncorhynchus mykiss, Walbaum) - sea-run rainbow trout trout, steelhead (Stizostedion vitreum, Mitchill) walleye  ACKNOWLEDGEMENTS A long term project such as a thesis always involves a wide array of people whose contributions have been essential at one time or another, each in their own specific way. For some it is only a one time affair, whereas for others it becomes a regular part of their existence. These people range from: the contributors that are directly involved with the scientific and educational aspects of the thesis, such as supervisors, committee members, colleagues, university staff, agency staff, and institute staff; to those who provide moral and logistical support, both professional and personal, such as family, friends, and acquaintances. When one tries to make a list such as this, to acknowledge all of these contributions after the years have come and gone, undoubtedly there will be omissions; some more significant than others. For all those people who have assisted me throughout the years but whom I do not include specifically in this acknowledgement, pleasefirstaccept a heartfelt "thank you" for your assistance and support, and secondly, please accept a sheepish apology for my failing memory. Specifically, I would like to thank my supervisor, Dr. George Iwama, for his continued support from the day I walked into his office to request doing a PhD thesis under his direction. He possesses a high degree of professional dedication and scientific rigor, combined with a view of the "bigger picture" that provided me with a much sought after balance between the pursuit of a career in science and the "other" things in life. I would also like to thank my research committee members, Drs. Craig Clarke, Dave Randall, and Peter Rombough for their guidance, assistance, and support over the years as well as their helpful comments in thefinalpreparation of this thesis. Financial support for this thesis was provided by an NSERC operating grant to Dr. Iwama and by BC Science Council G.R.E.A.T. Award in conjunction with the Department of Fisheries and Oceans, Pacific Biological Station, Nanaimo, BC, and BC Packers, to me for salary support. Working in and around the 'Iwama Lab' and the Department of Animals Science, was a pleasure and I would like to acknowledge some of the many people who assisted me there over the years. In particular, thanks to Paige Ackerman, Shannon Balfry, Laura Brown, Jim McGeer, Kira Salonius, and Jonathan Wilson. A special thanks to John Morgan for his interest to collaborate, interesting discussions, and support, both logistical and moral. Also I would like to thank Ellen Teng for her numerous hours of assistance and willingness to help, including sampling many of the little fish that provided some of the data for this thesis. Of course there are always those people whose contributions cannot be forgotten since they have supported me through thick and thin, always asked the right questions, or not, and willingly made sacrifices in order that I could reach this once in a lifetime goal. A very special thanks to my family for their support and interest over the years. A lifelong thanks to my mother and father for sharing with me their curiosity of the world and openness to new ideas. A special thanks also to Hanneke and Job for their encouragement over the years. Thanks also to Job for all of the great discussions about science as well as just about everything else -"Cheers Bro!". I would also like to acknowledge the support received from numerous friends and colleagues; for assisting me, feeding me, rooming me, humouring me, etc. For this and much more, many thanks to Lome and Christine Collicutt, Carol Elder and John Candy, Nina von Keyserlingk and Nick Schmaling, Bronwen Lewis, Kelly Malange, and Andrew Simons. Thanks also go the many great people at the Pacific Biological Station, friends and colleagues alike, who have supported me over the years with willing assistance and good friendship. Special thanks go to John Jensen for his friendship, endless generosity, patience, and support, both logistical and moral. Thanks also to the other member of the Fish Culture Research Section, John Blackburn, Bill Damon, Henrik Kreiberg, and John Shelbourn for their assistance. More recently I have had the pleasure of working with the faculty and staff in the Department of Fishereies and Aquaculture and the Faculty of Science and Technology at Malaspina University-College. I would also like to extend a very appreciative 'thank you' to these people for their continued support and patience during this past year (or was it two?) Finally, a very special expression of appreciation, that extends well beyond words, to Joanne Rankin for her patience and support over the past few years and especially this last one when writing took up so much of my spare time. "Soon, soon!".  DEDICATION  This thesis is dedicated to my parents: In loving memory of my mother 'I'runs Augusta Maria (Hartong) Groot,  better known as "Mom", my beloved mother, dear friend, and lifelong teacher;  and to my father Cornelis (Kees) Groot, beloved father, who continues to be an inspiration in so many different ways.  G E N E R A L INTRODUCTION  Osmoregulation is one of the major physiological processes that ensures a stable internal environment for proper cell function and ultimate maintenance of an animal's health. Failure of this process quickly results in disruption of osmotic homeostasis and death ensues shortly thereafter. Fish face a unique set of physiological challenges compared to terrestrial animals in that their entire bodies, and perhaps more significantly their thin-walled breathing structures, are bathed in an aqueous environment. Both saltwater and freshwater teleost fish generally maintain a blood osmotic concentration that is about one-third that of seawater (Schmidt-Nielsen 1975). Therefore, marine or saltwater (SW) teleost fish are hypo-osmotic to their environment, whereas freshwater (FW) teleost fish are hyperosmotic to theirs. In both cases diffusive losses or gains of ions and water must be restricted or controlled in order to maintain homeostasis. Compensatory processes that are used to regulate the osmotic and ionic balances involve active transport of ions and thus require energy. Although the regulatory processes involved have been studied extensively and are relatively well understood in juvenile and adult fishes at least, the energetic consequence of these processes remains ambiguous. Our lack of understanding of the energetic implications of osmoregulation becomes even more apparent when we consider the early life stages of fish. Embryos and larvae do not possess the same organ systems early on in development as adult fish, yet they must cope with similar physico-chemical constraints using much more limited regulatory systems. This thesis examines osmoregulation and its energetic  1 The term 'osmoregulation' and its derivatives will be used here in general reference to the regulation of both water and ions. The term 'ion regulation' and its derivatives will be used in specific reference to the regulation of ions - most commonly Na , K , Ca , M g and Cl". The two terms are closely associated since it is the ions, especially N a and Cl~, that are very significant in producing osmotic interactions in biological systems (Withers 1992). +  +  2+  2+  +  1  implications during the 'early life' stages of fish, using two species of Pacific salmon, chum (Oncorhynchus keta) and coho (O. kisutch), as models.  Osmoregulation in adult fish Saltwater teleost fish are faced with a continuous passive loss of water and passive gain of ions, which occurs largely across the relatively permeable gill epithelium and the intestine (Evans 1993). Adult fish compensate for water loss by drinking water. However, the ingestion of SW results in absorption of both monovalent and divalent ions across the intestine. N a and Cl~ also +  diffuse into the blood along an electrochemical gradient across the gill epithelium. In order to compensate for this increased salt load and maintain a blood osmotic concentration below that of the surrounding water (hypo-osmoregulation), the fish must excrete the excess ions. Excess monovalent ions are excreted by the gill via specialized 'chloride' cells (CC), first described by Keys and Willmer (1932), whereas excess divalent ions are excreted via the kidney (Fig. GI 1) (Moyle and Cech 1996). Because the teleost kidney is only capable of producing urine that is isoosmotic to the blood (Schmidt-Nielsen 1975), SW fish rely on the CCs in the gill to excrete the majority of excess ions (Evans 1993). The current model for ion excretion by the CC, originally proposed by Silva et al. (1977) and reviewed by Zadunaisky (1984), Karnaky (1986), and Marshall (1995), involves the secretion of C F as a result of the N a electrochemical gradient +  established by basolaterally localized Na ,K -ATPase (Fig. GI 2). The enzyme units are located +  +  in the basolateral membrane which is highly convoluted and continuous with the tubular system of the cell. Chloride cells typically contain numerous mitochondria that provide energy in the form ATP to actively transport N a out of the cell via Na ,K -ATPase. The resultant N a +  +  +  +  2  Freshwater teleost  Fig. GI 1. Diagram of active and passive exchange of ion and water in freshwater and marine teleost fish. Modified from Moyle and Cech (1996); illustration redrawn from C. Groot, unpublished drawing.  3  Leaky junction  Ji  accessory cell Na+ .-•>3Na+_ ATP  / Saltwater  -—  ( K + )  *"  2K+  Blood  1  \ \  0 mV  imr  ii  -60 mV ?  Na+, K+ 2CI+40 mV  pavement cell  Tight junction  pavement cell  J  - mu r  Na+L  t  ^  ^-^ H+(NH4+) T  Na+—A »  Freshwater  x  V__!_/ ° ca.y HC0 C  .3Na+ I ATP (<v *~\  s*  2 + H 2  3  0 mV  °  ^Z\-^  CI-" " -60 mV ?  t  2K+  ^  Blood  ci+10 mV  pavement cell Fig. GI 2. Diagrams of current models of chloride cell function in a marine (a) and freshwater (b) teleost. Passive ion movement is indicated by dashed lines whereas active transport and cotransport are indicated by solid lines. The abbreviation ' c a . ' refers to carbonic anhydrase. Note that N a movement across the apical membrane in FW may occur by either one or both of the depicted mechanisms (1, 2) and may occur in chloride cells and/or pavement cells; Modified Marshall 1995, and Lin and Randall 1995 +  4  concentration gradient allows N a to re-enter the cell and appears to drive a N a / K / C l +  +  +  cotransport system that accumulates Cl~ intracellularly. C P then exits the cell to the external environment down its electrochemical gradient across anion channels in the apical membrane (Foskett et al. 1983, Marshall 1995), which typically forms an apical crypt in marine fish (Karnaky 1986). N a follows a slightly different pathway to exit the cell. It is secreted from the +  CC down its electrochemical gradient via cation-selective channels into the paracellular space, and diffuses from the blood into the SW via "leaky" intercellular junctions (Fig. GI 2). These junctions connect to adjacent "accessory cells" cells that resemble smaller and less differentiated CCs (Karnaky 1986, Marshall 1995). Freshwater teleost fish need to deal with the opposite set of problems compared to marine fish; they are faced with compensating for the passive gain of water and the passive loss of ions across permeable epithelia. Adult fish compensate for water gain by producing large volumes of dilute urine via the kidney. Ion loss is counteracted by actively taking up ions from an often ionpoor environment (Fig. GI 1), using a combination of gill CCs and gill pavement cells (Marshall 1995). Fish exposed to ion-deficient water have been shown to respond by increasing the numbers of gill CCs (Perry and Laurent 1989). The current model for ion uptake in F W involves both of these cells types in the gill epithelium but it is not entirely clear which regulatory processes occur in which cells. Either way, ion uptake has been shown to be closely linked with acid-base balance since N a and C F appear to be either directly or indirectly exchanged for acid +  (H ) or base equivalents (HC0 ~), respectively, at the apical membrane. Proton excretion at the +  3  apical membrane establishes a favourable electrical potential that drives N a across the apical +  membrane into the cell. Two mechanisms have been suggested for this process and in fact they  5  may operate in parallel (Marshall 1995). Initially a N a - H exchanger, that also accepted N H , +  +  +  4  was proposed (Maetz 1970), whereas more recently an apically located H -ATPase has been +  located that functions in combination with an apical N a channel (Lin et al. 1994) (Fig. GI 2). +  Once N a has entered the cell, Na ,K -ATPase at the basolateral membrane then actively +  +  +  transports it into the blood (Fig. GI 2). H -ATPase has been identified in both CCs and pavement +  cells (Lin et al. 1994), whereas Na ,K -ATPase appears to be located mostly in the CCs (Witters +  +  et al. 1996). C F is currently believed to enter the gill epithelium via an electro-neutral 0 7 H C 0 ~ 3  exchanger and is then passively transported into the body fluid down a favourable electrochemical gradient through a C F channel on the basolateral membrane (Fig. GI 2). As in marine fish, CCs in F W fish also contain numerous mitochondria to supply ATP to the ion translocating enzymes, however, they differ in that F W CCs lack an apical crypt, usually have a less extensive tubular network, and are not associated with adjacent accessory cells (Zadunaisky 1984). It is important to remember that the mechanisms of ion regulation outlined above for adult SW and F W fish are based on current models only and that many of the details surrounding ion exchange are still not fully understood (Marshall 1995, Lin and Randall 1995). This limited state of knowledge is even more evident when we consider the processes of osmoregulation in the early life stages of SW and F W fish.  Osmoregulation in fish embryos and larvae Fish Embryos and larvae in FW or SW must cope with the same physico-chemical challenges as adults. However, it is not until they reach metamorphosis that they seem to be capable of  6  osmoregulating in a manner that is similar to the adult (Holliday 1963, Alderdice 1988). Prior to this time they have not yet fully developed the necessary adult-type organs to deal with these environmental challenges. Our understanding of early life osmoregulation prior to metamorphosis is limited (Alderdice 1988, Brown and Tytler 1993). The current model is that the developing blastula relies initially upon 'tight' cell junctions and low cell membrane permeabilities to minimize diffusive losses and gains (Alderdice 1988). As development progresses and the energetic demands of the embryo increase, the embryo requires a greater degree of exchange with its external environment. This transition, from a relatively impermeable state to a more open one, is thought to occur near the time of completion of epiboly (gastrulation) (Alderdice 1988). Although the embryo is functionally osmoregulating by this developmental stage, the gills, kidney and gut have not yet fully developed. Instead, osmoregulation by embryos and early larvae appears to be accomplished by a combination of low epithelial permeability and cutaneous ion exchange (Holliday and Jones 1965, Alderdice 1988, Tytler and Bell 1989, Brown and Tytler 1993). In embryos and larvae of SW fish at least, the skin appears to function as an organ for ion secretion since cutaneous CCs have been located in a number of SW species (Shelbourne 1957, Lasker and Threadgold 1968, Depeche 1973, Guggino 1980a, Hwang and Hirano 1985). It is very likely that these cells are functioning to excrete excess ions since a number of embryos and larvae of SW species are known to drink SW; presumably to compensate for diffusive water loss (Guggino 1980b, Mangor-Jensen and Adoff 1987, Tytler and Blaxter 1988). The situation is much less clear for the embryos and larvae of F W species, since far fewer species have been examined, and some of the earlier studies reported no or few cutaneous CCs ((Leatherland and Lin 1975, Shen and Leatherland 1978a). However, more recently a number of studies have located cutaneous CCs in the early life stages of several F W fish (Hwang and  Hirano 1985, Kaneko et al. 1995). As with SW species, organogenesis progresses throughout the larval stage of development and the gut, kidney, and gills differentiate. Branchial CCs then become visible and presumably functional for the purposes of ion excretion in SW and ion uptake in FW, as the adult pattern of osmoregulation develops (Hwang and Hirano 1985, Alderdice 1988). Details are limited on the specific mechanisms of early life osmoregulation, including the actual functionality of CCs both in the cutaneous and branchial epithelia of SW and F W fish embryos and larvae. However, examinations using scanning (SEM) and transmission electron microscopy (TEM) indicate that the cellular fine structure of CCs during the early life stages are similar to those of the adult fish. As a result it has been suggested that embryonic and larval CCs function similarly to the adults cells (Alderdice 1988, Hwang 1989, 1990). However, most of the information relating to early life C C distribution and fine structure has focussed on a limited number of species, few of which inhabit FW. Thus, we still know little about how fish embryos and larvae osmoregulate. A wide variety of questions remain. For example, do all teleost fish embryos and larvae use similar osmoregulatory mechanisms, i.e., low epithelial permeability and cutaneous ion exchange? If maintenance of low epithelial permeability is common, how do embryos and larvae establish and maintain it? Can they control permeability on the short term and use it as a regulatory mechanism? How common and significant are cutaneous CCs? Do cutaneous C C density or fine structure change in response to environmental changes the same way they do in adult fish? Of additional interest are the energetic implications of osmoregulation. Although low epithelial permeability is a largely passive form of regulation, ion excretion and uptake using CCs, cutaneous or branchial, requires active ion transport and thus energy. Because embryos and  8  larvae function as relatively simple and more or less closed systems until first-feeding, is it possible to use them as models to estimate the energetic cost of osmoregulation? Energetic implications of osmoregulation Fish that are exposed to osmotic or ionic environments other than their natural or preferred one, presumably incur a change in metabolic cost as a result of increased or decreased compensatory or regulatory ion transport processes. These processes are necessary to re-establish or maintain osmotic homeostasis which generally includes both ion and water balance. Researchers have commonly used this principle in efforts to estimate the energetic cost of osmoregulation (Kirschner 1995). However, the picture still remains unclear for juvenile and adult fishes (Kirschner 1995), and is even more uncertain for the earlier life stages. In theory, a change in osmotic and ionic gradients across the external epithelia will induce a change in the requirements placed on the osmotic and ionic regulatory mechanisms and result in a concomitant change in the energy dedicated to these processes. Calculations of the energy requirements based on such theoretical models of ion regulation in relation to whole-body metabolism in adult fish have estimated changes from as low as <1% (Potts 1954, Eddy 1982) to as high as 6.5% and 15% (Kirschner 1993). Experimental measurements have been even more variable and have reported not only a wide range of different values in oxygen consumption rates, but also a wide range of fundamentally different responses. For example, experimental estimates of the energetic cost of osmoregulation for juvenile and adult fishes range from no measurable change in routine oxygen consumption rate in euryhaline teleosts across a salinity range of 10 to 100 %oS (Nordlie 1978 Ambassis interrupta, Bleeker; Haney and Nordlie 1997, sheepshead minnow, Cyprinodon variegatus), to increases above routine oxygen consumption rate measured in iso-osmotic conditions, of as much as 27% for rainbow trout (Oncorhynchus mykiss) in S W (~ 33 %oS) (Rao 1968) and 65% for juvenile snook (Centropomus undecimalis) in 9  FW (0 %oS) (Perez-Pinzon and Lutz 1991). Febry and Lutz (1987) reported similar results in Florida red hybrid tilapia (Oreochromis mossambicus x O. hornorum) with an estimated additional energy allocation of 16% and 12% in F W and SW, respectively above that in isoosmotic conditions. Many of these researchers concluded that the observed differences were the result of additional energy allocations related to the transport and regulation of ions and water. In many cases the type of response is related to the degree of variability in the physicochemical environment that a species experiences naturally. If the animal is commonly exposed to these conditions then it is probably capable of dealing with the associated environmental changes in an energetically efficient manner. Nordlie (1978) categorized some of these patterns from a metabolic perspective, into 4 groups (Type I, II, III, IV), and Morgan and Iwama (1991) suggested some modifications including the addition of a fifth group (Type V). Briefly, the patterns are as follows: (I) oxygen consumption rate does not change significantly over a wide range of salinities (FW to SW), (II) oxygen consumption rate is minimal at a salinity that is isoosmotic with the blood and increases at lower and higher salinities, (III) oxygen consumption rate is minimal in F W and increases at higher salinities, (IV) oxygen consumption rate is highest in F W and decreases in iso-osmotic salinities (high salinity is not tolerated), and (V) oxygen consumption rate is lowest in SW and increases in lower salinities. One of the reasons for such a wide range of responses is that fish may cope with varying environmental conditions using fundamentally different regulatory mechanisms. Some of these mechanisms appear to include: (1) increased or decreased energy allocation associated with effective functioning of osmoregulatory mechanisms (Rao 1968, Febry and Lutz 1987); (2) possible changes in permeability of epithelia such as the gill (Bath and Eddy 1979) and the integument (Skadhauge and Lotan 1974, Stiffler 1988, Lillywhite and Maderson 1988, Tytler and Bell 1989, Nordlie et al. 1991, Brown and Tytler 1993); (3) behavioural changes such as increased or decreased 10  activity (Holliday et al. 1964, Almatar 1984, Swanson 1996); and (4) stress response to a new environment (Morgan 1997). The categories defined by Nordlie (1978) and Morgan and Iwama (1991) provide a useful framework from which to view changes in oxygen consumption rates of a wide variety of different fishes in relation to changing environmental salinities. Unfortunately, many of the categories for a given species are based only upon a few studies and almost exclusively deal with the juvenile and adult life stages. Both Nordlie (1978) and Morgan and Iwama (1991) identified differences between life stages and body size that would place a given species in a different category depending on whether it was a juvenile or an adult. Undoubtedly such differences also exist for the earlier life stages. However, information regarding early life osmoregulation and its energetic implications is even more limited than for adults, especially when considering F W species (Morgan and Iwama 1991). It was this minimal state of knowledge of early life energetics in conjunction with the limited knowledge of early life osmoregulation as a whole, that motivated me to investigate some of these questions more closely. This paucity of information on osmoregulation of F W embryos and larvae became very evident during my MSc in the mid-1980's, when I examined the ecological physiology of intertidally spawning chum salmon (Oncorhynchus keta) (Groot 1989). These salmon spawn in the intertidal zone of streams and rivers and as a result the embryos and larvae experience and survive regular (daily) SW exposure. Similar to other Pacific salmon, chum salmon are a F W species during the early phase of their life cycle. However, unlike most of the other Pacific salmon, they are able to survive and apparently osmoregulate, at least partially, in near full strength SW (Weisbart 1968). Interesting as this observation was, there existed no definitive explanation or mechanism to account for this ability. The involvement of cutaneous CCs had been suggested (Weisbart 1968) but was not supported by any convincing data for chum 11  or any other species of salmonid (Leatherland and Lin 1975, Shen and Leatherland 1978a). As for the energetic implications of osmoregulation during the early life stages of salmonids, only one relatively recent paper has examined this question (Morgan et al. 1992, chinook salmon, O. tshawytscha, and steelhead trout, O. mykiss) and the results remained inconclusive. Thus, I chose to examine the early life stages two species of salmon, one which was relatively well-adapted to early life SW exposure (i.e., chum salmon), and one that was not (i.e., coho salmon, O. Msutch). Coho salmon are certainly more representative of the 'typical' Pacific salmon than chum, and presumably are also more representative of the 'average' F W fish.  Life history of Pacific salmon The life history patterns of Pacific salmon vary widely. Although as a group they are anadromous (i.e., spawn in F W but develop to maturity in SW) there is a wide range of variation in the extent to which different species or races rely on F W for completion of their life cycle. On one end of the scale there are species that have only a minimal reliance on FW, such as chum and pink salmon (O. gorbuscha), whereas on the other end there are species that are completely reliant on FW, such as 'land-locked' sockeye (O. nerka), also referred to as kokanee. Unlike the term implies, many populations of kokanee are not truly land-locked, in that they may have unrestricted access to the ocean (Foerster 1968, W ood et al. 1987). }  Chum and pink salmon share a close evolutionary relationship to one another. However, depending on which phylogenetic model one subscribes to, a freshwater or marine origin for the family Salmonidae, reside either near the top or the bottom of the evolutionary tree for the genus Oncorhynchus as a whole (Salo 1991). Currently the arguments favour a freshwater origin, and thus suggest that chum and pink salmon are more specialized by being less reliant on the F W environment than other Pacific salmon (Salo 1991). Examples of these specializations are 12  evident in their respective early life ecology and physiology. Upon emergence from the gravel, the fry of both species generally migrate directly to the ocean, and thus are capable of effective hypo-osmoregulation in full strength SW at the fry stage (Salo 1991, Heard 1991). Chum and pink salmon fry do not undergo the classic smoltification process characteristic of the other Pacific salmon (coho, O. kisutch; chinook; and sockeye). The early life osmoregulatory abilities of these two species are not limited to the fry stage since embryos and larvae can tolerate simulated and actual intertidal conditions (Bailey 1964, Helle et al. 1964, Groot 1989) and also possess at least limited hypo-osmoregulatory abilities (Weisbart 1968). Intertidal spawning chum and pink salmon spawn in sections of stream or river that are exposed to regular tidal influence and as a result the developing eggs experience regular SW exposures. In the lowest parts of the intertidal zone, SW exposure can occur twice daily with surprisingly high intragravel salinities (e.g., 25 to 29 %oS, Groot 1989). Laboratory studies have shown that chum salmon embryos and larvae can survive (> 95% survival) SW exposures (beginning 12 h postfertilization) of 20 %oS twice daily for 4 h each or 15 %oS twice daily for 8 h each (Groot 1989). Bailey (1964) reported similar results for pink salmon embryos and larvae. In general, the majority of chum and pink salmon spawn in F W sections of rivers and streams, but there are geographical regions where large proportions of the populations spawn in the intertidal zone. In these areas intertidal spawners can constitute as much as 70% and in some cases up to 100% of the mature returning fish; in systems where the total number reaches 200,000 fish (Helle et al. 1964). Obviously, intertidal spawning as a life history strategy, is not a minor anomaly. Moreover, intertidal spawning fish may actually constitute discrete populations or races with different migration timing and different habitat selection criteria than freshwater spawning fish (Thornsteinson 1971). Intertidal spawning appears to be more common in northern  13  latitudes of North America, especially southeastern Alaska, U S A but also occurs, albeit less commonly, in areas of southern B C , Canada (Groot 1989, Heard 1991, Salo 1991). Assuming a F W origin for the family Salmonidae, coho salmon are considered to be less specialized than chum salmon since they rely more extensively on F W for spawning and early life rearing (Sandercock 1991). Even though some coho salmon occasionally spawn in the lower reaches of streams and rivers, there are no documented reports of coho salmon spawning in intertidal zones. Upon emergence from the gravel, coho salmon fry usually remain in F W for a year or more and generally undergo 'smolting' before migrating to sea (Sandercock 1991). In some short coastal river systems coho may migrate to the estuary shortly after emergence, but because they are not yet able to tolerate full strength SW, they rely upon the variable salinities of the estuary for acclimatization (Tschaplinski 1987). In systems where coho and chum coexist, coho almost invariably spawn further upstream than the chum salmon (Sandercock 1991).  Thesis objectives and organization In my MSc thesis (Groot 1989), I showed that intertidally spawned embryos and larvae appeared to benefit from exposure to regular tidal fluxes through increased temperatures and thus more rapid development, regular flushing of incubation gravel with well-oxygenated water, and better survival than some F W incubation sites plagued by low intragravel flow and hypoxia. I speculated that intertidal spawning and the potential associated ecological benefits may have contributed to chum salmon being the most widely (naturally) distributed of all the Pacific salmon (Bakkala 1970, Fredin et al. 1977). Two interesting questions arose from that study and remained unanswered. (1) What is the physiological basis that allowed two species offish (chum and pink salmon), within a genus and family that was otherwise heavily dependent on FW, to expand their range by utilizing habitats that would be considered suboptimal or unusable for the 14  other species of this group? [Pink salmon are also widely distributed and are the most abundant of all 7 species of Pacific salmon (Heard 1991).] (2) Does this ability to exploit the intertidal habitat confer any energetic advantage for maintenance of osmotic homeostasis in the eggs and subsequent larvae incubating and rearing in these conditions? Thus, I chose chum salmon as the model for an intertidally-spawning species and coho as the model for a more FW-dependent species to explore these questions. This study examines and compares the physiology of chum and coho salmon during the early life stages with an emphasis on the mechanisms and energetics of osmoregulation. The specific objectives were: to examine the saltwater tolerance and energetic implications of osmoregulation by measuring oxygen consumption rate and growth in various salinities (Chapter 1); to examine the role of cutaneous and branchial CCs in early life osmoregulation (Chapter 2); and to examine the activity of cutaneous and branchial Na ,K -ATPase and H -ATPase, two ion translocating +  +  +  enzymes that are important for osmoregulation (Chapter 3). This thesis relates physiological data from three different levels of investigation (whole-animal, tissue, and cellular) in an effort to reconcile whole-animal metabolism in different osmotic conditions with early life osmoregulatory mechanisms using both qualitative (cutaneous CCs) and quantitative indicators of osmoregulation (Na ,K -ATPase and H -ATPase activity). +  +  +  Appendix A describes a custom designed computerized microrespirometer that I constructed to measure oxygen consumption rates of fish embryos and larvae.  15  Chapter 1 E N E R G E T I C IMPLICATIONS OF O S M O R E G U L A T I O N IN E M B R Y O S A N D L A R V A E OF C H U M A N D C O H O S A L M O N  Introduction Fish embryos and larvae usually develop in direct contact with the water and thus must cope with the associated physico-chemical challenges throughout development. The process of development progresses from a relatively simple mass of cells to a more complex embryo, followed by a free-swimming larva at hatch and finally to a juvenile upon completion of metamorphosis. During most of this period the embryo or larva does not yet have the adult mechanisms for maintaining osmotic and ionic homeostasis and thus relies upon a series of interim measures ranging from passive 'tight' epithelia to cutaneous osmoregulatory mechanisms and finally the adult type mechanisms of branchial, intestinal and renal osmotic and ionic regulation (Alderdice 1988). These stages appear to represent physiological different strategies as the early embryo develops through to an organism that must progressively interact more with its environment as the metabolic demands of rapid growth and development increase. During the egg stage, the embryo is at least partially buffered from its external environment by the surrounding perivitelline fluid (PVF) and external egg membrane, whereas the hatched larva is directly exposed to its environment (Alderdice 1988). For chum salmon eggs spawned in the intertidal zone of streams and rivers this environment includes regular exposure to intragravel tidal saltwater (Groot 1989). Unlike the other species of Pacific salmon, embryos and larvae of chum and pink salmon have been shown to be tolerant of near full strength SW on the short term and lower concentrations on the longer term (i.e., daily 16  4 h exposure to 20 %oS from fertilization to first feeding) (Bailey 1964, Groot 1989, Kaneko et al. 1995). The differences in SW tolerance of salmonid embryos and larvae appears to be related to differential osmoregulatory ability. Weisbart (1968) showed that chum and pink salmon embryos and larvae were capable of at least partial osmotic and ionic regulation for at least 24 h following direct transfer into full strength SW (32 %oS), whereas no other Pacific salmon could. Although Weisbart (1968) suggested that cutaneous regulation was likely, no convincing data has been reported for salmonids until recently when a study using eyed chum salmon embryos, related the presence and size of CCs in the yolk sac epithelium (YSE) to hypo-osmoregulatory ability (Kaneko et al. 1995). If cutaneous CCs do in fact play a role in early life osmoregulation, then increased osmoregulatory activity such as active ion transport necessary for effective hypoosmoregulation, would require energy to fuel ion translocating enzymes such as Na ,K -ATPase +  +  and H -ATPase. +  Assessment of the energetics of osmoregulation in adult fish has interested researchers for many years, however, as mentioned previously the picture still remains unclear. Not surprisingly, our current understanding of these energetic relationships during the embryonic and larval stages for fish is even less clear. A number of studies have examined oxygen consumption rates of embryos and larvae for marine fish acclimated to Varying osmotic or ionic environments and generally have not found significant metabolic responses (Lasker and Theilacker 1962, Holliday et al. 1964, Walsh and Lund 1989, Walsh et al. 1991, Swanson 1996). In general, these studies were motivated by either a desire to obtain a better basic scientific understanding of early life osmoregulation and its energetic implications in fishes, or by attempts to optimize early life culture conditions by looking for osmotic environments that  17  minimize the energy expenditure dedicated to osmoregulation and thus hopefully improve early life growth and survival. Far less bioenergetic research has been done on the early stages of F W species, especially salmonids. Morgan et al. (1992) measured the metabolic responses of the embryos and larvae of 2 species of Pacific salmon (chinook and steelhead trout) acclimated to different salinities but the results were inconclusive. One factor that may have contributed to this outcome is that neither chinook nor steelhead embryos and larvae appear to be capable of osmoregulating in a hyperosmotic environment (Weisbart 1968, Shen and Leatherland 1978b). Although chinook salmon have been shown to be relatively tolerant of saline conditions, perhaps due to their larger body size, ( L T  50  chinook embryo 14 d prehatch (dPRH): 81 h, compare with L T  5 0  chum embryo  14dPRH: 102 h), they apparently do not hypo-osmoregulate (Weisbart 1968). Steelhead trout embryos and larvae survival data from Morgan et al. (1992) are similar to chinook and therefore, indicate that they too are not effective at hypo-osmoregulation during the early life stages. Perhaps one of the reasons that the results from Morgan et al. (1992) were inconclusive is that the species they chose were not especially SW tolerant or competent at osmoregulation Chum salmon on the other hand are known to be markedly more SW tolerant than almost all of the other species of Pacific salmon, the exception being pink salmon (Weisbart 1968, Heard 1991, Salo 1991). Life history patterns, such as intertidal spawning and direct emigration to the ocean upon fry emergence, in addition to field and laboratory studies, indicate that chum salmon are not only relatively SW tolerant but also capable of limited hypo-osmoregulation as embryos and larvae (Helle et al. 1964, Weisbart 1968, Groot 1989). In contrast, coho salmon embryos are not as SW tolerant and do not appear to be capable of hypo-osmoregulation based upon Weisbart's (1968) plasma ion and osmolality measurements and median resistance time data (LT  50  coho embryos: 47 h compared with 102 h for chum salmon embryos). 18  The aim of this study was to elucidate further the energetic implications of osmoregulation in the early life stages of anadromous fishes (i.e., F W embryo and larval stages) using two closely related species; one that is SW tolerant and one that is not. Given these criteria, chum and coho salmon were chosen as model species since they are ecologically (spawning location and fry outmigration behaviour) and physiologically (early life osmoregulatory abilities) quite different and, yet phylogenetically and developmentally relatively similar. Based on these observations, the hypothesis was that as the external salinity of the environment increased, chum salmon would survive longer and increase metabolism more than coho to meet the increased energy dedicated to hypo-osmoregulation. This assumes that chum salmon embryos and larvae have the necessary ion transport mechanisms to compensate for salt influx in hyperosmotic conditions, and that coho do not. Thus hypothetically, coho salmon should not incur an increase in metabolism to supply energy to ion transport processes. Energy allocation at different developmental stages was measured as oxygen consumption and growth rate in embryos and larvae exposed to a range of salinities. Mortality rates were monitored to assess SW tolerance. Ffypo-osmoregulatory ability of fry was assessed by measuring plasma osmolality levels following saltwater challenge tests.  t  19  Methods Embryos and larvae of chum and coho salmon were exposed acutely and acclimated (7 d) to a range of selected salinities at three different developmental stages (eyed embryo, prehatch embryo and yolk sac larva). Mortality, oxygen consumption, and growth rates were monitored. Saltwater challenge tests including sampling for plasma osmolality were conducted on fry of both species once they had reached 100% yolk sac absorption.  Egg collection, incubation and rearing Chum salmon gametes were obtained from ripe adults caught in the Goldstream River [ca. 15 km North of Victoria, British Columbia (BC)] using a beach seine. Gametes were stripped from five females and three males in 1993/94 and from three females and three males in 1994/95. Coho salmon gametes were obtained from fish returning to the Cowichan River Hatchery (Duncan, BC) in 1993/94 and Chase River Hatchery (Nanaimo, BC) in 1994/95. Gametes were stripped from three females and three males in both years. Unfertilized eggs and milt were transported to the Pacific Biological Station in Nanaimo, B C in a cooler (2°C to 5°C). Once the milt had been checked for normal motility (observations under a compound microscope at lOOx magnification), eggs were fertilized within 6 h of collection by combining milt and eggs (1 c? :1 9 matings) for 2 min before adding F W water. Coho and chum eggs were sampled for fertilization success at the 4 cell stage of development, 12 and 16 h, respectively after fertilization (at 10 °C), by gently removing 20 eggs from each batch (mating) of eggs and fixing them in Stockard solution [5% buffered formalin, 4% glacial acetic acid, 6% glycerin in distilled water (dH 0), Velsen 1980]. Examinations for development were made the following 2  day. In all cases fertilization success rates were > 95% or more for both species.  20  Eggs were incubated in 40 L, covered, temperature controlled tanks maintained at 10 ± 0.1 °C in a darkened room. Tanks were aerated continuously and water was changed every third day. Saltwater concentrations below 30 % S were produced by diluting full strength SW 0  (28 to 29 %oS) from the Pacific Biological Station, Nanaimo, B C , Canada (drawn in from the ocean at about 18m and filtered to 1 um nominal particle size) with dechlorinated Nanaimo city FW. Salinities above full strength SW were made by adding commercial aquarium sea salt to SW (Instant Ocean, Aquarium Systems, Mentor, OH, USA). Salinities were checked every second day with a temperature-compensated refractometer (American Optical Instruments, Buffalo, N Y , USA) The refractometer was calibrated weekly with Standard Sea Water (35 %oS, "Eau de Mer"). At the eyed stage of embryonic development (stage 20, Jensen 1988), eggs were mechanically 'shocked' by siphoning into a bucket. The unfertilized eggs turned white and were removed, and the viable ones remained clear and were pooled within each species.  Acute salinity tests 1994/95 - experimental design In 1994/95 embryos and larvae of chum and coho salmon were exposed acutely to a range of salinities at three predetermined developmental stages; eyed embryo, prehatch embryo, and 50% yolk sac absorption larva (developmental stages 20, 23, and half-way between 24 and 25, respectively, Jensen 1988)]. Eyed embryos were exposed to salinities of 0, 5, 10, and 15 %oS, whereas prehatch embryos and yolk sac larvae were exposed to 0, 6, 12, 18, 24, and 30 %oS (Table 1.1). Two acute exposure regimens were tested: (1) an ascending exposure, where test animals were incubated and reared in 0 %oS and then tested sequentially at increasing salinities during the respirometer run as indicated in Table 1.1; and (2) a descending exposure, where test animals were incubated and reared in 0 %oS, but before testing, first acclimated (see definition in next section) for at least 7 d to the maximum tolerated salinity (as determined in the 1993/94 21  mortality tests), and then tested sequentially at decreasing salinities during the respirometer run as indicated in Table 1.1. Prior to each acute exposure oxygen consumption measurement  (Mo ; 2  see Appendix A. for definition, p. 186), test animals were acclimated to the respirometer environment and respective test salinity for 30 min, to ensure complete flushing of the respirometer water with the next test salinity. This was especially important during testing of the embryonic stages since the embryo, which is surrounded by the external egg membrane and bathed in the perivitelline fluid, is not subjected immediately to changes in external conditions. The external egg membrane of salmonid eggs is highly permeable to ions and water, and the perivitelline fluid reaches a 95% osmotic and ionic equilibrium with the external environment within about 20 min after transfer (Kalman 1959, Potts and Rudy 1969, Eddy and Talbot 1983, Groot 1989). The 30 min equilibration period ensured that the embryo was actually being exposed to the ambient test salinity and that the water quality of the small volume of water within the respirometer was not significantly altered by diffusion from the perivitelline fluid.  Acclimated salinity tests 1993/94 - experimental design In 1993/94 embryos and larvae of chum and coho salmon were tested at the same three developmental stages as the acute experiments; eyed embryo, prehatch embryo, and yolk sac larva. However, prior to testing embryos and larvae were 'acclimated' to the test salinity for a period of at least 7 d and no more than 10 d (acclimated = a gradual exposure, see below). Eyed embryos were acclimated to salinities of 0, 5, 10, and 15 %oS, and prehatch embryos and yolk sac larvae were acclimated to 0, 6, 12, 18, 24, and 30%oS (Table 1.1). 'Acclimation' was accomplished by transferring embryos and larvae to progressively higher salinities and included a 24 h adjustment period in each subsequent salinity until the target acclimation salinity was reached. Embryos and larvae were maintained in the target salinity for 7 d before sampling or I  22  Table 1.1. Test salinities used for acclimation and respirometer testing of chum and coho embryos and larvae during 1993/94 and 1994/95 experiments. Acclimation and respirometer test salinities (%oS) Eyed embryo Test year & Treatment  Prehatch embryo  Yolk s a c larva  Chum  Coho  Chum  Coho  Chum  Coho  0  0  0  0  0  0  0, 5,10,15  0, 5, 10, 15  0, 6,12,18, 24  0, 6, 12, 18, 24, 30  0, 6, 12, 18, 24, 30  0, 6, 12, 18, 24, 30  15  15  18  18  24  18  15, 10, 5, 0  15, 10, 5, 0  18, 12, 6,0  18, 12, 6,0  24, 18, 12,6, 0  18, 12, 6,0  0, 5, 10, 15  0, 5, 10, 15  0, 6, 12, 18, 24  0, 6, 12, 18, 24  0, 6, 12,18, 24, 30  0, 6, 12  1994/95 Acclimated at: a  Acutely exposed to and tested at: 1994/95 Acclimated at: b  Acutely exposed to and tested at: 1993/94 Acclimated and tested at: ascending exposure  descending exposure  23  testing for routine oxygen consumption rate and growth, (e.g., prehatch embryos acclimated to 18 %oS were treated as follows: removed from FW, placed into 6 %oS for 24 h, then into 12 %oS for another 24 h and finally into 18 %oS, the target salinity, for 7 d before testing). Development rate was measured in 'accumulated thermal units' [ATUs: also known as 'degree-days' (°C-d), average daily incubation temperature (above 0 °C) multiplied by number of days of development (Velsen 1980)]. Sampling times for chum salmon eyed embryo, prehatch embryo, and yolk sac larva stages occurred at 349 to 365 ATUs, 489 to 502 ATUs, and 719 to 730 ATUs, respectively, whereas sampling for coho occurred at 327 to 332 ATUs, 397 to 402 ATUs, and 607 to 611 ATUs, respectively. The different time periods are due to the inherent interspecific differences in development rates of the two species (Velsen 1987). Predictions of development times to specific developmental stages were calculated using a microcomputer program called Tncub6' (Jensen 1988) to ensure consistent sampling times.  Mortality rates At the start of each acclimation period (1993/94 only), for a given developmental stage and species, 6 subsamples of 10 individuals each were placed into subdivided containers for evaluation of SW tolerance. These embryos and larvae were held in the same incubators as the embryos and larvae that were sampled for oxygen consumption rates and growth. However, in all treatments where they survived beyond the 7-d acclimation period, cumulative mortality rates were monitored for up to 40 d postexposure. Median resistance times ( L T ) were estimated by 50  recording the time (d) to 50% mortality.  24  Saltwater challenge tests - fry Chum and coho fry (100% yolk sac absorption, Stage 25, Jensen 1988) were exposed to 24-h SW challenge tests and then sampled for plasma osmolality (Clarke et al. 1989). Preliminary tests showed that chum fry were able to tolerate a 24 h exposure to 35 %oS, whereas coho fry were able to tolerate 24 % S without any mortalities but not full strength SW (29 %oS); thus the 0  tests were conducted at 35 and 24 % S, respectively. Sixty chum and coho fry each were 0  transferred directly from 0 %oS to the challenge salinities in the same tanks as the previous acclimation tests (10.0°C, > 95% air saturation) and maintained in a darkened environment to minimize stress. Fish were killed 24 h after transfer with an overdose of TMS (100 mg/1) and sampled immediately. Blood was collected from the caudal artery by cutting off the tail and drawing blood into 5 uL Microcap® tubes (nonheparinized, Microcap®, Drummond Scientific Co. Broomall, PA, USA). These tubes were sealed and kept on ice until they were centrifuged in a hematocrit centrifuge for 2 min at 5000 x g (the small 5 uL tubes were slid into larger 20 uL tubes to prevent breakage during centrifugation). Due to the small volume of blood available from each fish, plasma samples were pooled from 7 to 10 fish to produce a final volume of at least 10 uL. This was accomplished by inserting the 5 uL tube into a 20 uL tube and dispensing the contents into the larger tube using a Microcap® 'eye-dropper' style dispenser. Depending on fish availability and volume of blood from each fish, 4 to 7, ten uX sample tubes were obtained for each salinity treatment. Pooled samples were sealed on both ends and frozen at -80°C. A second series of SW challenges of chum and coho fry was conducted on fish that were left over from the main acclimation experiment and thus were already acclimated to a variety of salinities. This included chum and coho fry that were acclimated to: (1) 5 and 10 %oS at the eyed embryo stage, (2) 6 and 12 %oS at the prehatch embryo stage, and (3) 6, 12, 18, and 24 %oS at  25  the yolk sac larval stage. These fry were submitted to the same SW challenge protocol as outlined previously. The SW challenge test was repeated for coho in 1994/95 since a different stock was used (Chase River instead of Cowichan River) for the acute exposure series of Mo tests and there 2  was evidence that this stock was more SW tolerant at the yolk sac larva stage of development than those of the previous year. The Chase River larvae were able to tolerate 18 %oS for more than 10 d unlike the Cowichan River larvae which were able to tolerate only 12 % o S . The SW challenge test was used to determine whether this difference persisted through to the fry stage of development. For measurement of plasma osmolality, frozen tubes were thawed on ice and re-centrifuged to eliminate gas bubbles. Both sealed ends of the centrifuged tubes were scored and broken off and the contents were transferred to an 8 uL glass microcapillary tube using the Microcap® dispenser. Plasma samples were processed within 2 min of transfer to the 8 uL tubes in a Wescor 5100C Vapour Pressure Osmometer (Wescor Inc., Logan, UT, USA) calibrated to 290 and 1000 mOsm/kg (mOsm) standards (Wescor).  Oxygen consumption measurements (Mo ) 2  Oxygen consumption rate measurements (Mo ) were conducted using a custom built 2  2  microrespirometer system comprised of four independently operated respirometer loops (Groot and Iwama 1994, Appendix A). The total volume of each loop ranged from 19 to 21 mL. Tests were run in quadruplicate with 10 eggs (eyed and prehatch embryos) or 3 yolk sac larvae per  2 The term 'oxygen consumption' rate will be used instead of 'metabolic' rate since the contribution of anaerobic metabolism during the early life stages of fish is not clear. For details see Appendix A Some General Considerations for Energetic Studies (p. 186)  26  chamber. Test specimens in both acute and acclimated experiments were allowed to adjust to the respirometer environment for 60 min prior to testing (reduced to 30 min between tests for the acute exposure experiments), while each loop was flushed with the 100% air-saturated test water (i.e., 'open' mode). After this adjustment period each loop was switched to the 'closed' mode to initiate the test. Recirculating flows were maintained at 20 mL-min , approximately equal to -1  complete turnover of the respirometer loop volume each minute. The partial pressure of oxygen (p0 ) was monitored in each loop until it dropped to about 80% of saturation. Temperature and 2  p0  2  values were monitored and recorded at 1 min and 0.25 min intervals for embryos and  larvae, respectively, using a computerized data acquisition system (Groot and Iwama 1994). Upon completion of a test, specimens were removed, killed in a lethal dose of tricaine methane sulfonate (TMS, 500 mg/L) and preserved in 10% buffered formalin (Velsen 1980) for dry weight determinations. A l l M0 tests were conducted at 10.0 ± 0.1 °C in complete darkness and 2  randomized for the various test salinities. Blank tests (i.e., no animals) were conducted for each respirometer loop at the start and finish of each series of tests (e.g., chum salmon, eyed embryo) to correct for background changes in p0  2  levels due to microbial respiration and electrode drift. To minimize the effects of  exponentially increasing growth of micro-organisms during the actual respirometer test, the respirometer loops were washed but not disinfected between tests (Dalla V i a 1983). This procedure minimized differences between the initial and final blank test slopes resulting from changes in microbial respiration and allowed calculation of an average of the two blank measurements. Oxygen consumption rate for a given test (e.g., chum salmon eyed stage, 5 %oS) was determined by calculating the slope (least squares regression) of the decreasing p0  2  level  27  (usually 20 to 30 min. into the test, only linear portions of the p0  2  trace were used) and  subtracting the mean slope of the blank tests. This net slope, equal to the instantaneous rate of decrease in p0 , was converted to the molar-equivalent oxygen consumption rate 2  (nmol 0 respirometer -h~ ) using oxygen solubility equations from Green and Carritt (1967) 1  1  2  with modifications by Forstner and Gnaiger (1983), to compensate for temperature, salinity, and atmospheric pressure. Mass-specific Mo [i.e., Mo /Mass of organism or tissue dry weight 2  2  (TDW); nmol Cymg TDW" -IT ] was calculated by dividing the respirometer Mo by the mean 1  1  2  T D W for each respirometer. For comparisons with other studies that used Mo values based on 2  tissue wet weight (Mo /TWW), conversions were made assuming a 6:1 ratio of T W W compared 2  to T D W (Rombough 1988b, unpublished data this study)  Dry weight and growth measurement Embryos and larvae preserved from the respirometer tests were processed more than 21 d after fixation to ensure a constant tissue water content. Tissue and yolk (YDW) dry weights were determined by separating the hardened yolk material from the body tissue and drying at 60°C for 48 h (Rombough 1986) followed by weighing to the nearest 0.01 mg. The total number of eyed and prehatch embryos weighed for each test salinity from the respirometer tests was 40 (10 indiv./respirometer loop), whereas for yolk sac larvae it was 12 (3 indiv./respirometer loop). To increase the sample sizes of larval samples, an additional 10 yolk sac larvae per salinity were preserved for supplemental dry weight measurements, thus increasing the total measured to 22 individuals. . Weights of all eggs and larvae within a developmental stage test group were assumed to be similar and collected as random samples. Percent embryonic or larval tissue ( P ) r  28  was calculated by dividing TDW by the total embryonic or larval weight (TDW + Y D W ) multiplied by 100 (Rombough 1985).  Data analysis Two way analysis of variance (ANOVA) was used to test for effects of species and salinity on Mo  2  at each developmental stage. Similar analyses were conducted for T D W and salinity.  Where within-factor differences were statistically significant one-way A N O V A was used to test for significant differences between treatments using the Tukey multiple comparison test (SigmaStat version. 2.0, Jandel Scientific 1995, Zar 1984). Prior to statistical testing, P values r  were normalized using the arcsine transformation (Zar 1984). In a number of cases where the data did not meet the normality and equal variance assumptions for A N O V A (e.g., some Mo  2  and all P data), one-way A N O V A on ranks (Kruskal-Wallis test) was used with the Dunn r  multiple comparison test to determine treatment differences. One-way A N O V A was performed on the SW challenge data followed by Tukey multiple comparison tests. Data averages are reported and graphically presented as the arithmetic mean ± 1 standard deviation (SD), unless otherwise noted, to provide an indication of within sample variability. Where possible, graphical data are presented using similar scales to allow for comparisons between developmental stages and species.  29  Results Development and mortality rates  The basic mortality response to S W exposure in all treatments was typical of a critical threshold response; <10% mortality at or below the threshold salinity and >80% (usually 100%) mortality above the threshold salinity (Fig. 1.1). Interspecific comparisons showed that chum salmon embryos and larvae were more tolerant to diluted SW exposure since they had higher L T ' s 5 0  than coho at all three sampled stages of development (Table 1.2). Neither chum nor coho eyed embryos exposed to 15 %oS survived much beyond hatch. Chum exposed from the eyed stage, had a threshold salinity of 10 to 15 % S (Fig. 1.1a). The 0  threshold salinity for coho was 5 to 10 %oS (Fig. l.ld). Median resistance times in 10 %oS were 21 d for coho eyed embryos and > 50 d for chum (Table 1.2). Similar interspecific differences were observed for embryos exposed at the prehatch stage of development. Once again neither species was able to tolerate long term exposure to higher salinities (18 and 24 % S) but chum 0  embryos, and later the hatched larvae, survived (< 5% mortality) for 33 d at the threshold salinity of 12 % S (Fig. 1.1b). B y comparison, coho had a lower threshold salinity of 6 %oS 0  (Fig. l.ld). The most dramatic interspecific differences were seen in larvae exposed to SW at the 50% yolk sac larva stage. The threshold salinity for chum larvae was 24 to 30 %oS (Fig. 1.1c), whereas for coho it was much lower at 12 to 18 %oS (Fig. l.lf). In almost all of the SW treatments no survival occurred above the threshold salinity (i.e., 100% mortality), however, in 30 %oS (see Fig. 1.1c) 20% of the chum larvae survived and remained active and vigorous until the end of the experiment.  30  Table 1.2. Median resistance times ( L T ) for chum and coho salmon embryos, and larvae exposed 50  to a range of salinities. LT50 (days)  T e s t salinity (%oS)  Eyed embryo  Prehatch embryo  Y o l k s a c larva  Chum  Coho  chum  coho  chum  Coho  > 50  > 35  >40  > 27  > 19  > 10  > 50  > 35  >40  > 27  > 19  > 10  > 50  21  >40  14  > 19  > 10  25  18  14  14  > 19  2  24  -  -  14  12  > 19  <1  30  -  -  -  3  <1  0 5or6  a  10 or 12  a  15 or 1 8  a  Eyed embryos were exposed to 5, 10, and 15 %oS, whereas prehatch embryos and yolksac larvae were exposed to 6, 12, and 18 %oS.  31  Eggs exposed to saltwater hatched earlier in the higher salinities. This was most apparent in salinity-acclimated prehatch embryos. Chum embryos hatched earlier in 12, 18 and 24 %oS than those in 0 and 6 %oS, and survived for progressively shorter durations with increasing salinity (Fig. 1.2). Coho embryos acclimated to 18 and 24 % S all died as eggs whereas those acclimated to 12 %oS 0  died shortly after hatching (Fig. 1.2). It should be noted that some differences were observed in coho S W tolerance, during the 1994/95 acute exposure Mo experiments compared to the detailed mortality records collected 2  during the 1993/94 experiments. A different stock of coho was used for the acute experiments (see Methods section). Although no specific mortality experiment was conducted in this second series of tests, qualitative mortality observations revealed that the threshold salinity for 50% yolk sac larvae was 18 to 24 %oS instead of the 12 to 18 %oS recorded in the previous year. These differences were no longer apparent at the fry stage since results from the S W challenge tests (see below) were very similar between the two stocks.  Plasma osmolality measurements - saltwater challenge tests Chum fry were markedly better hypo-osmoregulators than coho. This difference was initially evident in the preliminary tests where chum fry were able to tolerate direct transfers from 0 to 35 %oS, and coho were only able to tolerate direct transfers from 0 to 24 %oS for 24 h. Plasma osmolalities of chum and coho fry exposed to 24-h S W challenges provided clear physiological evidence for this difference. Chum fry saltwater challenged in 35 %oS (~1080mOsm) had a relatively small but significant increase in plasma osmolality of 7%; from 328.6 to 352 mOsm (p < 0.01). In contrast, the 2 stocks of coho challenged in 24 %oS had much larger increases of up to 34%; from 329.1 to 439.6 mOsm for 1994/95 coho (Fig. 1.3). Although coho fry were not  nearly as effective at hypo-osmoregulation as chum, they were still capable of some degree of regulation since the plasma osmolalities (~440 mOsm) were only about 60% that of the external osmolality (-740 mOsm). However, it is uncertain how long these fry would have survived beyond the 24 h of the saltwater challenge test with such high plasma osmolalities. The second series of S W challenge tests with acclimated chum and coho fry indicated that for both chum and coho, acclimation had a positive effect on osmoregulatory ability. The data show clearly how this response became progressively more pronounced as development proceeded and acclimation salinity increased, since fry exposed to 6 %oS as prehatch embryos were more effective hypo-osmoregulators (i.e., had significantly lower plasma osmolalities [p < 0.05]) than fry exposed to 5 %oS as eyed embryos (Fig. 1.4)  Oxygen consumption rates in different salinities - acute exposure Embryos and larvae of chum and coho salmon exposed acutely to a range of salinities did not show any notable changes i n M 0 as a function of salinity (Table 1.3). This response was the 2  same in both of the experimental regimes where test animals were either (1) acutely exposed to progressively increasing salinities, starting at 0 %oS, or (2) first acclimated (7 d) to their highest tolerated salinity (as determined in 1993/94) and then acutely exposed to progressively decreasing salinities ending with 0 %oS (see Table 1.1). There were no significant progressive reductions or increases in Mo values (p > 0.05) at any of the developmental stages as seen in 2  the acclimated experiments. Even though there were no significant relative Mo changes within a given developmental 2  stage of the acutely exposed test animals, the magnitude of M0 values differed noticeably in 2  comparison to the previous year's acclimated experiments. This was despite maintaining similar  33  incubation and rearing conditions, and sampling periods (i.e., developmental stages) for each of the two series of experiments; some of the experimental treatments were even identical. These differences were most evident at the eyed stage of development. Oxygen consumption rates of eyed chum embryos exposed acutely to SW, ascending and descending regimes, ranged from 22.1 to 24.0 nmol-mg TDW~'-h , whereas acclimated embryos ranged from 44.0 to 55.1 _1  nmolmg T D W - ! ] . Similar differences were seen in coho eyed embryos; ranging from 36.6 to -1  -1  41.4 nmol-mg T D W - I f for acutely exposed embryos compared to 56.9 to 60.2 nmol-mg -1  1  T D W •h~ for acclimated ones (see Table 1.3). These 'between year' differences in Mo were - 1  1  2  of a similar magnitude at the prehatch embryo stage and much less at the yolk sac larva stage (Table 1.3). Despite the differences observed in Mo values of acclimated embryos and larvae compared 2  to acute exposure test animals, it is possible nevertheless to compare the fundamental effect of SW acclimation on these life stages. At each of the three developmental stages, test animals exposed to the descending acute exposure regime were initially acclimated for 7 d to the maximum tolerated salinity for that developmental stage. Following acclimation, they were exposed acutely and tested for M o in descending salinities. The acclimation conditions were 2  essentially the same as those from the acclimated exposure experiment of the previous year at the same salinities (see Table 1.1). Thus, it is reasonable to compare the Mo values from the 2  highest tested salinity from each acute exposure test with the Mo measurements made at the 2  same salinity in the acclimated exposure experiments of the previous year. The basic response was the same. In both experiments, salinity had no effect on Mo in eyed embryos, produced a 2  reduction in Mo in prehatch embryos, and caused an increase in Mo in yolk sac larvae, 2  2  34  compared to test animals of the same developmental stage maintained in 0 % S (compare 0  Fig. 1.5 and Fig. 1.6 with Table 1.3). Although these results were not unexpected, since the acclimation conditions of the selected treatments were essentially the same, they do show that the basic metabolic responses of chum and coho salmon embryos and larvae acclimated to SW were similar between two batches of eggs obtained in separate years.  Oxygen consumption rates in different salinities - acclimated exposure Chum and coho salmon embryos and larvae had very different Mo and growth responses in 2  relation to SW exposure at the three developmental stages. In general, acclimation to various salinities resulted in no effect on Mo at the eyed stage of development, decreased M0 at the 2  2  prehatch stage and increased Mo at the 50% yolk sac larva stage. Within a given developmental 2  stage, interspecific differences were only minor. Changes in growth (TDW) paralleled the changes in Mo closely in both species for all stages except chum yolk sac larvae. 2  Oxygen consumption of eyed stage embryos tended to decrease slightly as salinity increased from 0 to 15 %oS (Fig. 1.5a and Fig. 1.6a). Some interspecific variation in the respective responses of chum and coho salmon was observed but the absolute differences were relatively small; ranges of 44.0 to 55.1 nmol-mg TDW" -IT and 56.9 to 65.8 nmol-mg T D W ' - h , 1  1  - 1  respectively. Significantly lower M0 values were observed in chum salmon compared to coho 2  in all salinities (p < 0.05) (Fig. 1.5a and Fig. 1.6a). This difference was most significant (p < 0.001) in 6 %oS where chum embryos showed a decrease in Mo whereas coho embryos 2  showed an increase.  35  q * -*t co i n co  q i n 00 co cvi co  o  oo I  I  iri •t  I  co  TJ"  cd  I  I  I  I  CO co c o C O C T ) - * C O CM •<t •  inh-co i n CN  C O CD -<t  co  o  •<t  co co  -<t  (D -*t i n co •t  CJ) CD -<t r-- c o  CO C N  co  CD •<t  cd cd •<t  CN  o co to o  CO  CM  C O <J> -<t  CM in •<t  co  00 i -  LO O ) TjCM  0  tn  CO  q  co I  co co co  in  o CT>  o< co  •<- C O  CO  •r- O i -Jf  h - CN ">*•  0 1 CN  O)  -=t  m -*t • cd  CN  CO CO C O  T t  iri Tt  I  T t  in  CO  m o * oo  cn co -«t cd cd  co T J " co co in  h - r^• m  CO  o  c c _o  -4—' CL  E  I  o  'CM  E  CD  CT>  oo -<t co  CO C O  E  O  CD  T t  t- i n o> i n •<t  CN  •*  co  •"=r  co q  I— CN CO  "t  r*-  -=t  m  I  I  I  CO •  I  m  CO I - CO S  o XJ  E CD  -st  CN  CM  oq  csi  00  1^ CN  CM  cd c\i co  co  -<t  ow  ^  CD CO  T-  •i- CN 1  CD  >^  CN q  X  -*t  N - CN 1 ^ CN  in  i n CN  CO  CN  m o>  T  iri  •<t C N  o  •*  co •<t  co  Tt  I  I  I  co  co  h-; T t C O  in i h iri  co cd m  co CD  ca  CT)co co CM  oo C M co N - co  i^- i n • * • * CN CO  co i n co  ^t i n co  in  i^- co  Tt N - CN CO  co co co in • in  oir^-co co CN •<t  T3  O) CN "st  co * co  CN i n  co  3  co m co  co  CO  Tt  in  CO  CN • * C O T-^  C D i n •<*•  CO  o  -<t  CN  CN CN  q  00 O W CN CN -<t  CN  C O CN CO  CN  i n CN  C N -<-  CN CN •*  SZ  zi o to co c JZ o CD o t— c Q_  q  TJ-  CT) CN  CN ' CN  I  I o  <n  -*-» o C L  CO  ci  ecj  • CO CO  oo co  o JO  E  -3-  CN  00 O CO o co  co  T t  m  0  "O  >» LU  m  c=S  £  "3- C  c o  c— I  E •o to CD  CO CO  to I  E o o ® Q-l, 3 JZ < co o O  CJ) O)  D C  CM • CD  co • CM  O  CO  •S CD » Zi  o  <  "t  CD  co  0) •  co ci  c o i n *  oo m c o  *t co co  i n cd CD  co  •<- C O 1*  C Q  '  COfft  cn c co  C  CO CO  I  ;o CD • O <  co  co o  o* co  .  XJ O CO  oi cd co  Q) "<t in  CN co ^ i -  Q U  c\i  CN  o ca <u l-l <D  cu  q oo T i ed cd  I X3  "3"  CN  h-; C O Tf co CN CN  CD  in  co co  co CM  ; "<t *  CO 0) CD  CT) ^  CT)-<t  CN  in  C Q  CN  CM  • co •=t co •<t  CD  •<- o * CN co  O CT)  c  I  O  XJ  E 2  co to E  00  CO  co C I O  C  CD  Q  cm  o to  C  CD CO  o  (fl o  Q  CO C O / n  c  O  o  E o  I XJ  CM  co  G> CO//-)  ^— co E  T— C O  CT)  CD CM  C3  XJ  o to  0  0  o  o  <  <  1 CD O CCJ  JZ  u  CD  fa 3 CO  o &, U M  .g  c  CD O  CO  <  O  ex CD M S3  T3  C CD o CO  CD  Q  Chum and coho salmon prehatch embryos showed very similar Mo responses when 2  acclimated and tested in salinities ranging from 0 to 24 % S; range 22.7 to 33.7 nmol-mg T D W  -  0  ' • I f and 25.2 to 37.7 nmol-mg TDW^'-h , respectively. The M o values were almost identical 1  -1  2  for all salinities except 0 %oS, where coho showed a significantly higher mean Mo than chum 2  (p < 0.01); 37.7 compared to 31.2 nmol-mg TDW" I T (Fig. 1.5b and Fig. 1.6b). 1  1  On the whole, coho yolk sac larvae had significantly higher M o values than chum 2  (p < 0.01) (Fig. 1.5c and Fig. 1.6c). Intraspecifically, chum larvae had significantly higher Mo  2  values in 6 %oS compared to 0 %oS (p < 0.01), however, the apparently elevated Mo values 2  measured in 12 and 18 %oS were not significantly different than those at 0 and 6 %oS (p > 0.05) (Fig. 1.5c). Oxygen consumption rates of chum larvae in 24 and 30 %oS were significantly higher than those at lower salinities (p < 0.01). No significant differences in M o values were 2  observed in coho larvae tested at 0, 6, and 12 %oS (p > 0.05) (Fig. 1.6c).  Growth in different salinities Tissue dry weights of chum and coho eyed embryos did not change much in relation to acclimation salinity (Fig. 1.5a and Fig. 1.6a). Although chum embryos acclimated to 5 and 10 %oS weighed slightly and significantly more (p < 0.05) compared to 0 %oS, it is unlikely that these difference are biologically significant. Percent tissue values for chum eyed embryos were also slightly and significantly higher in 5 and 10 %oS than 0 and 15 %oS (p < 0.05, Fig. 1.7a). There were no marked differences in P values for coho embryos. 7  In both species, TDW was highest at 6 %oS and decreased with increasing salinity (Fig. 1.5b and Fig. 1.6b). These weight changes were similar to the patterns seen in the Mo data. In chum, 2  37  the mean T D W value at 6 %oS was significantly higher than both 0 and 12 %oS (p < 0.05), whereas in coho it was significantly higher than 12 % S but not 0 % S (p < 0.05). In both species 0  0  P decreased above 6 %oS and was significantly higher in 6 %oS than in 12, 18, and 24 %oS but r  notO%oS(Fig. 1.7b). Chum and coho larval weights were more variable than the embryonic weights and consequently were not significantly different between species or acclimation salinities; range 24.87 to 27.68 mg and 26.96 to 29.98 mg, respectively (Fig. 1.5c and Fig. 1.6c). Percent tissue values of chum larvae reflected the changes in T D W with no significant trends in relation to exposure salinity. Percent tissue values of coho larvae in 12 %oS were significantly lower than those in 0 and 6 % S (Fig. 1.7c). 0  38  Discussion Salt Water Tolerance and Osmoregulatory Ability of Embryos, Larvae, and F r y Central to the issue of estimating varying energy expenditure in response to changes in salinity is the need to establish whether or not the test animals are able to tolerate a given salinity and ultimately maintain homeostasis in that salinity. Salt water tolerance data of embryos and larvae and plasma osmolality of SW challenged fry indicate that chum salmon are more tolerant and more competent at osmoregulation than coho. In general, the threshold salinity above which the chum salmon died compared to the coho was 5 %oS higher for eyed and prehatch embryos and 12 %oS higher for yolk sac larvae. Few studies have compared saltwater tolerance of the prehatch and larval stages of Pacific salmon. One such study, Kashiwagi and Sato (1969) exposed chum salmon to a range of salinities as eyed embryos and yolk save larvae as different time periods after hatch. Median resistance times of posthatch larvae exposed as eyed eggs were 17, 5, 1 and < 1 d in 7.5, 15, 22.5, and 30 %oS, respectively. These results are similar to those found in this study, however, their embryos were slightly less resistant in 7.5 %oS and slightly more resistant in 15 %oS than those in this study (posthatch L T  5 0  > 20 d in 10 %oS and about 2-  3 d in 15 %oS). Chum salmon yolk sac larvae exposed by Kashiwagi and Sato (1969) at 20 d posthatch had essentially the same L T  5 0  values (> 30 for all salinities up to 22.5 %oS and 4 d for  30 %oS) as those in this study (compare with Table 1.2). Weisbart (1968) exposed prehatch embryos (14 d prehatch) and yolk sac larvae (14 d posthatch) directly to 32 % S so they are not 0  directly comparable to the tests in this study. Nonetheless, he showed that chum salmon embryos ( L T  50  = 118 h) and larvae ( L T  embryos ( L T  50  = 47 h) and larvae ( L T  50  50  = 56 h) were much more SW resistant than coho  = 26 h). Plasma ion concentrations and osmolality of  chum embryos and larvae were also markedly lower than coho. Plasma osmolality of eyed  39  embryos of chum salmon compared with coho embryos (24 h postexposure to 32 %oS) was 400 mOsm and 570 mOsm, respectively. A similar measurement for 14 d posthatch larvae was 475 mOsm for chum but was not available for the coho since they did not survive. More recently Kaneko et al. (1995) showed that eyed chum salmon embryos were capable of effective hypoosmoregulation in 15 and 30 %oS but in this case they survived for up to 7 d with steadily decreasing osmolalities (plasma osmolality = 370 and 415 mOsm in 15 and 30 %oS, respectively on day 7). Collectively these data suggest that chum and coho embryos and larvae that tolerated a given salinity treatment for 7 d, were probably osmoregulating to some extent and probably also maintaining lower plasma osmolalities in the case of chum salmon. The differences between chum and coho salmon were even more apparent at the fry stage during the saltwater challenge tests. Chum fry were markedly more competent at hypoosmoregulation than the coho fry as indicated by their inability to tolerate the same SW challenge salinity as chum fry (35 %oS). Even though coho fry were capable of surviving in 24 %oS for 24 h, their plasma osmolality values were about 110 mOsm higher than the F W value (about 330 mOsm), indicating that they were not regulating nearly as effectively as chum fry, which were surviving in a much higher salinity of 35 %oS with markedly lower plasma osmolalities. The mean plasma osmolality for chum fry of 350 mOsm after 24 h after transfer is similar to values reported for more developed SW-adapted fry and juveniles of both chum and coho (Weisbart 1968, Clarke et al. 1989, Hasegawa et al. 1987, Iwata and Komatsu 1984). It is clear from the embryo, larvae, and fry data collected in this study that chum salmon on the whole were osmoregulating relatively well and definitely more effectively than coho. Even though the coho fry were not especially competent regulators, they did exhibit some improved osmoregulatory abilities after acclimation in low and moderate SW concentrations during earlier embryonic and larval stages of development (see Fig. 1.4). This suggests that coho  40  fry do possess some ability to adjust their regulatory mechanisms to improve their osmoregulatory competence. However, neither naively-exposed nor acclimated coho fry were ever able to reach the same levels of osmoregulatory competence as naively-exposed chum fry. This clearly indicates the fundamental difference in osmoregulatory capabilities of these two species, a feature that has been recognized by numerous previous comparative studies (Weisbart 1968, Iwata and Komatsu 1984, Clarke et al. 1989).  Interspecific and intraspecific comparisons of oxygen consumption rates Ontogenic changes of oxygen consumption rates for chum and coho salmon, regardless of salinity, were consistent with previous time course studies using Oncorhynchus embryos and larvae (Alderdice et al. 1958, chum salmon; Rombough 1988b, chinook salmon). Oxygen consumption rates were generally lowest during the prehatch embryo stage of development compared with the less developed eyed stage and the more advanced larval stage (Table 1.4). Oxygen consumption rates of chum and coho embryos and larvae from this study were similar in magnitude to those previously reported in the literature (Table 1.4). Differences in Mo between studies of as much as two fold (compare coho Mo from this study with McLean 2  2  et al. 1991) are not unusual, especially during the early life stages. Generally values from this study were within 1.2 to 1.8 times that of others. Morgan et al. (1992) noted variations in Mo  2  values of steelhead trout and chinook embryos and larvae compared with Rombough (1988b), that were similar in magnitude (2 times) with whom they shared precisely the same respirometer system. Oxygen consumption measurements from Alderdice et al. (1958) allow direct intraspecific comparisons with chum embryos in this study. Although their tests were not related to SW exposure their Mo values at least provide some indication of the accuracy of 0 %oS 2  measurements made in this study. Oxygen consumption rates (converted to common units) for 41  eyed and prehatch embryos were 30.9 and 33.7 nmol 0 • mg TDW" • h" , respectively, 1  1  2  compared to 55.1 and 31.2 nmol-mg TDW-h" , respectively for similar stage embryos in this 1  study (Table 1.4). Even though there is a notable difference between the two measurements made for eyed embryos (1.8 times higher M 0 in this study), there is virtually no difference between the two 2  prehatch embryo measurements. McLean et al. (1991) provided equations to predict Mo values 2  in relation to development time for five species of Pacific salmon, including chum and coho. Predicted values were consistently higher for all three stages of development for both species compared to this study. The largest differences occurred at the eyed stages and became progressively less towards the yolk sac larva stages; ranging from 1.9 times to 1.1 times higher, respectively (Table 1.4). The differences in Mo values noted herein are probably a combination 2  of interspecific and inter-stock variability, slight developmental stage differences, as well as differences associated with technique and apparatus. A l l of these factors are known to affect oxygen consumption measurements (Rombough 1988a). Due to the sensitivity of small scale respirometry measurements, variation between measurements is difficult to control. This is evident even when the exact same respirometer system is used in two different experiments. For example, the higher Mo values obtained for chinook embryos and larvae by Morgan et al. 2  (1992) compared to Rombough (1988b), and the higher Mo values obtained in this study from 2  one year to the next (e.g., M0 values in 0 %oS from 1993/94 to 1994/95 experiments, see 2  Table 1.3). Notwithstanding the various differences among early life metabolic studies  42  Table 1.4. Summary of oxygen consumption rates of salmon embryos and larvae from the literature. A l l values are converted to common units; shaded data are from this study Eyed embryo (nmol 0 • mg TDW" -h- ) [ATUsf  P R H embryo (nmol 0 • mg TDW" -rr ) [ATUs]  Y S larva (nmol O • mg TDW- -IT ) [ATUs]  Nominal Temp.  Chum salmon  55.1 [353]  31.2 [452]  42.1 [722]  10  This study  Chum salmon  30.9  33.7  10  [349]  [489]  Alderdice et al. 1958  Chum salmon  89.5  10  McLean et al. 1991  Species  2  1  1  a  2  1  1  b  ;  z  1  1  b  Source  CO  43.8°  53.1  [353]  [452]  [722]  Coho salmon  60.2 [330]  37.7 [390]  51.3 [609]  10  This study  Coho salmon  118.8°  65.6  56.3°  10  [330]  [390]  McLean et al. 1991  Steelhead trout  39.1 [-246]  33.6 [-444]  9  Morgan et al. 1992  Chinook salmon  13.2 [-436]  18.4 [-646]  10  Morgan et al. 1992  C  C  c  [609]  a  PRH= prehatch; Y S = yolk sac; T D W = tissue dry weight  b  Accumulated thermal units (ATUs) in brackets.  c  Predicted value from McLean et al. 1991, based on ATUs and temperature from this study.  43  [e.g., Alderdice et al. (1958), Rombough (1988b), Morgan et al. (1992) and this study], the comparisons indicate that the measurements and the associated variability obtained within this study are relatively accurate compared to other studies on similar species at similar life stages.  Effect of SW exposure on Mo  2  Acute SW exposure Acute exposure to SW did not significantly affect Mo values of chum and coho salmon 2  embryos and larvae. It appears that acute SW exposure does not elicit an obvious stress response, at least not one that affects whole-animal metabolic rates. Morgan et al. (1992) suggested a stress response as a possible explanation for elevated Mo levels in 7 d posthatch 2  steelhead trout yolk sac larvae acclimated to 12 %oS. Concern over a possible metabolic response to osmotic stress was one of the reasons I measured Mo following an acute SW 2  exposure, in addition to an acclimated one. In none of the acute tests did a short term exposure either depress or elevate Mo values as an indication of an acute stress response (Barton and 2  Schreck 1987). Weisbart (1968) showed that following exposure to SW (-32 % S), changes in 0  plasma osmolality and plasma N a , and C l " concentrations of Pacific salmon embryos and +  larvae occurred rapidly within the first 1 to 4 h. Embryos of the two species that were able to osmoregulate and ionoregulate, namely chum and pink salmon, showed a re-establishment of near pre-exposure levels of these parameters within 4 to 8 h postexposure. Measurement of glucose and lactate, biochemical indicators of aerobic metabolism, in rainbow trout yolk sac larvae show that changes occur within 15 min of exposure to a mild stressor such as 5 %o NaCI and are sustained for at least 120 min (Krumschnabel and Lackner 1993). These authors speculated that the decrease in glucose and increase in lactate concentrations were the result of  44  increased aerobic metabolism associated with increased osmoregulatory activities. Unfortunately, they did not measure whole-animal Mo which would have allowed a direct 2  comparison between indicators of cellular metabolism with those of whole-body metabolism. Krumschnabel and Lackner (1993) suggested that complete metabolic recovery occurred within 4 h of exposure. Therefore, i f an acute stress response did occur during the acute exposure tests in this study, and it influenced oxygen consumption rate, it seems likely that it would have become apparent shortly after exposure; certainly within the 6 to 8 h that it took to complete the tests. However, this assumes that a stress response necessarily would be manifested as a change in oxygen consumption rate (Barton and Schreck 1987) and that changes in salinity result in a stress response. The literature concerning the response of early life stages of fish to acute changes in salinity is limited and variable regarding this question. Acute exposure to higher or lower salinities have been shown to either increase oxygen consumption rates (Atlantic herring, Clupea harengus, Holliday et al. 1964; Atlantic herring and plaice, Pleuronectes platessa, Almatar 1984), or result in no change (grubby, Myoxocephalus aenaeus, and longhorn sculpin, Myoxocephalus octodecemspinosus, Walsh and Lund 1989; striped mullet, Mugil cephalus, Walsh et al. 1991). Studies that have examined potential stressors other than SW, such as temperature, have reported minimal responses during the earlier life stages. Zinichev (1990) exposed chum salmon embryos and larvae to acute temperature changes of 5°C and reported no change in M0 , even 2  after 10 h. Rombough (1994) observed a similar response with chinook embryos and larvae incubated, and therefore acclimated, in a range of temperatures from 5°C to 12.5°C. Above 7.3°C no increase in maximum oxygen consumption rate was observed, only increased developmental rate. Others studies however, have shown increased Mo values following acute 2  temperature exposures (Walsh and Lund 1989 and Walsh et al. 1991). In the case of salmon 45  embryos and larvae it appears that they do not necessarily exhibit a change in oxygen consumption rate when faced with acute stressors such as changes in temperature or salinity. Two possible explanations for this apparent lack of stress response are as follows: either there in fact is no measurable metabolic response to acute exposure to SW, or the response is more physiologically complex and not readily apparent by measuring whole-body metabolism. This might be true i f energy allocation follows a compensatory model rather than a simple additive one (Rombough 1994). This idea will be discussed in more detail below.  Acclimated SW Exposure Acclimation to SW had a variable effect on the oxygen consumption rates of chum and coho embryos and larvae, depending on the developmental stage and to a lesser extent the species. Since the responses were not consistent across developmental stages, it is not possible to invoke a simple energetic explanation to account for the differences. Acclimation to increasing salinity resulted in oxygen consumption rate changes that varied from little or no response in the eyed embryo stage, to one of reduced oxygen consumption rate at the prehatch embryo stage, and finally increased oxygen consumption rate at the yolk sac larval stage. These different responses in Mo of embryos and larvae to SW acclimation suggest that fundamentally different processes 2  or mechanisms underlie the basic physiological response to SW at a given life stage. There are only a few studies with which it is possible to make relatively direct comparisons for the Mo responses here, specifically for chum and coho acclimated to SW. Morgan et al. 2  (1992) used a similar experimental design with the early life stages of steelhead and chinook embryos and yolk sac larvae, two other Oncorhynchus species. Embryos were acclimated to 0, 4, 8, and 12 % S for 13 to 14 d and then tested at an advanced eyed stage (Stage 22b, Ballard 0  1973) and larvae, acclimated for 34 to 36 d, were tested at 7 d posthatch. In general, Morgan et 46  al. (1992) observed no change in M o across the range of salinities tested, except for steelhead 2  larvae which showed a significant reduction in 8 %oS and a significant increase in 12 %oS. They were not able to provide a definitive reason for these responses and speculated that they may have been due to altered activity or possibly a stress response. As discussed above, a stress response does not appear to be a likely explanation. Nagibina (1983) exposed early life stages of coho salmon to a range of salinities up to 20 %oS, and reported that M 0 values of embryos, free 2  embryos, and yolk sac larvae were suppressed in salinities over 12, 14, and 20 %oS, respectively. Oxygen consumption rates in salinities of 4, 6, and about 7 %oS were not suppressed and Mo in 2  6 %oS was the highest (22.3 nmol 0 -mg TDW" -If ) during ontogeny and resulted in the best 1  1  2  growth (Nagibina 1983). Interestingly, a similar result was observed in chum salmon prehatch embryos; T D W and P  r  (i.e., growth), were significantly higher in 6 %oS (see Fig. 1.5 and  Fig. 1.7). The trend was similar for coho prehatch embryos but was not statistically significant. Interspecific differences in the Mo responses of yolk sac larvae to different salinities were 2  most ostensibly the result of the different SW tolerances of the two species. Chum were able to tolerate salinities up to 30 %oS whereas coho did not survive in salinities above 12 %oS (1993/94 only). Oxygen consumption rates of chum salmon larvae increased from 0 to 18 %oS and then again at 24 and 30 %oS. It is noteworthy that larvae at 30 % S might provide biased Mo and 0  2  growth results since they were the survivors (20% of total) that were able to withstand the acclimation procedure into 30 %oS. This 'survivor effect' may bias some of the data in a positive direction since they no longer represent a random sample from the whole test population, (i.e., inadvertently selected for better than average osmoregulatory ability). For coho larvae, M o and 2  T D W were not significantly affected by SW exposure. Percent tissue however, was significantly  47  lower in 12 %oS than 0 and 6 %oS and may have been an indication of reduced growth as a result of morbidity in this higher salinity.  Energetic implications of osmoregulation - embryos The oxygen consumption rate and osmoregulation data together indicate that the two embryonic stages of chum and coho salmon responded in a similar manner despite a difference in their osmoregulatory capabilities. Two important points arise from this. First, it suggests that embryos of both species were either using similar osmoregulatory mechanisms to response to SW acclimation or at least were influenced similarly by a basic effect of SW exposure on Mo . 2  Second, it indicates that acclimated SW exposure either does not require measurable increased energy allocation to cope with increased osmoregulatory activity, or SW acclimation results in more complex interactions that are not readily detectable with whole-body oxygen consumption rate measurements. Such interactions could include energy repartitioning such that energy reallocated to support additional ion transport processes, would be made available through compensatory decreases in energy allocation to other process, and thus not increase the total amount of energy consumed (Wieser et al. 1988, Rombough 1994). The possibility of compensatory energy partitioning underscores the importance of measuring changes in parameters such as neuromuscular activity, and growth in conjunction with changes in oxygen consumption rate.  MO and growth 2  How exactly chum and coho salmon embryos respond on an organistic and cellular level to SW exposure is not clear. The current state of knowledge was discussed above (see General Introduction - Osmoregulation in embryos and larvae). Whatever the mechanisms are, they 48  appear to function more effectively in chum salmon compared with coho, since chum are more SW tolerant and more capable of effective hypo-osmoregulation during the embryonic stages (Weisbart 1968, Kaneko et al. 1995, this study). Presumably active osmoregulation requires active transport of ions across cell membranes and thus energy. However, no change in energy utilization was apparent from the M o data of eyed embryos acclimated to increasing salinities. 2  Eyed embryos of both species showed no marked response to SW acclimation. This result may be indicative of the low epithelial permeabilities that early embryos are thought to use to minimize diffusive losses and gains before the adult-type osmoregulatory organs have fully developed (Alderdice 1988). This relatively simple explanation is further supported by no obvious signs of salinity related changes in growth. This suggests that there is no measurable compensatory re-allocation of energy away from maintenance and growth related processes in favour of osmoregulatory ones. As the embryo develops, an improved ability to tolerate higher salinities indicates that it is more capable of maintaining at least partial osmotic homeostasis in the face of an increasingly hyperosmotic external environment. It seems logical that embryos that are effective osmoregulators, based upon active osmoregulatory processes (e.g., chum salmon, Weisbart 1968, Kaneko et al. 1995), would increase the amount of energy allocated to controlling internal water and ion concentrations when faced with increasing osmotic and ionic gradients. In contrast, embryos that were not capable osmoregulators might either exhibit a stress response, marked by a possible increase in metabolic rate (Bartpn and Schreck 1987) or suffer from more rapid loss of osmotic homeostasis and die sooner. Given an additive model of energy allocation, additional energy allocation to ion transport would increase the total metabolic rate whereas, in a compensatory model, such energy allocation would result in repartitioning of energy from one process to another without changing the total metabolic rate. Contrary to the traditional  49  expectation, oxygen consumption rates of prehatch chum salmon embryos, an apparently competent hypo-osmoregulator, did not increase with increasing salinity. Growth however, did decrease in salinities of 18 and 24 %oS (see Fig. 1.5 and Fig. 1.7). Initially this result might suggest a compensatory scenario for energy allocation. However, this does not explain why both M o and growth of chum and coho prehatch embryos responded in an almost identical fashion, 2  namely both deceased in higher salinities. This was unexpected since coho prehatch embryos presumably are not competent hypo-osmoregulators. A number of possibilities exist to explain these results; three of which will be discussed here. (1) There is no difference in the active osmoregulatory abilities of these two species at this life stage, and thus they do not vary in their energy expenditure on maintaining internal osmotic homeostasis; (2) the similar interspecific results of chum and coho salmon are due to a more fundamental response that is not directly related to relative osmoregulatory abilities; and (3) the metabolic compensation required by increased osmoregulatory activity is not significant or measurable with respect to the magnitude of whole-body metabolism. The first possible explanation, that chum and coho do not differ in their osmoregulatory abilities, is least likely since it simply does not coincide with current data. Chum salmon embryos and larvae are more SW tolerant and more effective at hypo-osmoregulation than coho (Weisbart 1968, Kaneko et al. 1995). The second explanation for depressed M 0 values in higher salinities is that salinity has a 2  negative effect on growth rather than on the energetics of osmoregulation. Morgan et al. (1992) showed reduced growth (TDW and P ) in steelhead trout and chinook salmon yolk sac larvae r  acclimated to 12 %oS. Few other studies have measured indicators of growth in conjunction with M 0 changes related to SW exposure. However, Swanson (1996) measured T D W in her study 2  on the effect of S W on metabolism in the embryos and larvae of a pelagic marine fish, milkfish 50  (Chanos chanos). Prehatch embryos acclimated to 20 %oS showed reduced M o and P values 2  r  (calculated from tabulated data) compared with those tested in the natural occurring salinity of 35 %oS. Because the metabolically active tissue was smaller and the remaining yolk larger (i.e., lower P ) in 20 %oS than 35 %oS, she suggest reduced growth as the reason for the M 0 r  2  changes  rather than a reduced energetic cost of osmoregulation. Percent tissue calculations quantify the relative difference between metabolically active tissue and yolk material regardless of the actual size of the test organism. It is interesting to note that even though the size of chum and coho prehatch embryos (i.e., TDW) differed in magnitude by about 2 mg (-20%), the magnitude of the P values were nearly identical (difference < 5%). Similar to the explanation proposed by 7  Swanson (1996), the reduced TDW and P values found in both chum and coho prehatch r  embryos in this study suggest that the cause of reduced M o in higher salinities is due to a basic 2  effect of salinity on growth rather than a change in the energy allocated to osmoregulation. The third possible explanation is that the additional energy dedicated to osmoregulation in higher salinities is not significant with respect to whole-body metabolism, or measurable using whole-body metabolism measurements. Even though it appears that energy allocated to osmoregulatory processes did not increase based on Mo measurements, it is still possible that 2  energy reallocation did occur. It may simply have been below the resolution of a relatively coarse measurement such as whole-body M o . Theoretical studies (Potts 1954, Eddy 1982) and 2  some tissue-specific ones (Morgan 1997) on the energetic cost of osmoregulation suggest that the energetic cost of osmoregulation is low (2 to 4%). If so, then it may be unreasonable to measure these changes at the level of the whole animal (Morgan 1997). Moreover, any changes that are measured will likely be the result of both additional energy required for osmoregulation, as well as other physiological processes that may be stimulated by SW exposure and  51  subsequently influence Mo levels. Such processes may not be directly related to the regulation 2  of water and ions (Morgan 1997).  Other effects of SW exposure on M0  2  One of the indirect effects of salinity on M o  2  may be related to embryonic activity. Activity is  often an unknown factor in early life respirometry studies especially when dealing with the egg stages. Under normal conditions activity of the embryo such as pectoral fin movement and trunk flexures appear to provide essential mixing of the perivitelline fluid which results in increased diffusive oxygen exchange across the external egg membrane (Peterson and Martin-Robichaud 1983, Rombough 1988a). Changes in salinity have been reported to reduce activity in some embryos (Lasker and Theilacker 1962, Holliday et al. 1964, Swanson 1996). If the higher salinity treatments of 18 and 24 %oS in this study resulted in reduced embryonic activity, that would provide another possible explanation or at least contributing factor for the observed decrease in Mo . Embryos near the hatching stage, are especially susceptible to decreased 2  oxygen diffusion into the egg since they have a very high critical oxygen level (P ). This is the c  level of ambient oxygen concentration at which the embryo switches from oxygen independent metabolism to oxygen dependent metabolism. This switch occurs as a result of insufficient oxygen transfer across the egg membrane and is related to the surface area of the egg and the relatively large volume of metabolically active tissue of the embryo when it nears hatching. Predicted P values for prehatch chum and coho embryos in the present study are 8.31 and 8.09 c  mg/1, respectively (calculated from McLean et al. 1991), which is about 72% of the oxygen content of fully air saturated water (11.29 mg/1 at 10°C, 1 atm, and 0 %oS). Thus any reduction in oxygen availability below that value would cause them to switch to oxygen dependent  52  metabolism. Further, increased salinity results in decreased oxygen solubility and at 18 and 24 %oS the relative values of the P s increase to about 80 and 85% of air saturated water, c  respectively. Even though incubation water was well aerated and always maintained above 95% saturation in all treatments, the lack of internal P V F mixing may result in oxygen tensions inside the egg dropping below the P value. If so, this may explain the lower Mo values measured in c  2  18 and 24 % S. Although there are insufficient data in this study to establish a definitive 0  connection between reduced activity in lowered oxygen consumption rates, such a relationship provides a possible explanation of the results. Swanson (1996) considered a similar explanation for her results in milkfish embryos and noted that the onset of reduced oxygen consumption rates was coincident with the initiation of embryonic movement. A similar relationship is noted in this study since at the eyed stage, before significant embryonic movement is established (Peterson and Martin-Robichaud 1983), there is no salinity related difference whereas at the prehatch stage when the embryo is much more capable of movement, there is a salinity related difference. In addition to providing a possible explanation for the oxygen consumption rate changes, reduced oxygen diffusion into the egg may also provide an explanation for the associated reduction in growth (TDW) found in the 18 and 24 % S treatments for both chum and coho 0  salmon (see Fig. 1.5 and Fig. 1.6). Sublethal effects of reduced ambient oxygen tension include decreased growth and delayed development of embryos and larvae (Silver et al. 1963, Shumway et al. 1964, Rombough 1988a). The P data (Fig. 1.7) indicate that the relative size of the r  embryos also decreased in the higher salinities and further supports the possibility that reduced oxygen availability may have played a role in reducing the oxygen consumption rates. However, since the mass-specific oxygen consumption rates factor out differences in absolute weight this indicates that the tissue that was present was less metabolically active. 53  Buffering capacity of the egg membrane-PVF complex Another mechanism that undoubtedly also is important for embryos in low to moderate salinities is the buffering effect of the external egg membrane and PVF. Even though all indications are that the egg membrane-PVF complex is permeable to ions and water (Kalman 1959, Rudy and Potts 1969, Eddy 1974, Groot 1989, Kaneko et al. 1995), it nonetheless provides the enclosed embryo with a significant degree of protection from external conditions including osmotic and ionic stresses (Lasker and Theilacker 1962, Weisbart 1968, Eddy and Talbot 1983, Shepard and McWilliams 1989). Lasker and Theilacker (1962) reported that the osmolality of P V F of intact Pacific sardine (Sardinops caerulea) eggs transferred from SW to distilled water (dH 0) 2  attained the osmotic pressure of the medium (~0 mOsm) within about 3h but the embryo did not swell up from osmotic inflow of water. In contrast, embryos with the external egg membrane removed swelled up quickly and the yolk sac subsequently burst. Weisbart (1968) reported that the survival ( L T , h) of chum and coho eyed embryos (14 d prehatch) exposed to 32 %oS was 50  20% to 30% shorter, for embryos with the external egg membrane removed than for intact eggs. Similarly, survival of yolk sac larvae (14 d posthatch) exposed to the same salinity was about one half as long as intact eggs prior to hatching (chum, 118 h vs. 45 h; coho, 47 vs. 26 h) (Weisbart 1968). Eddy and Talbot (1983) concluded that the P V F functioned as an 'ion trap' by reducing the diffusion of ions both in and out of the egg. Reduced ion diffusion through the P V F appears to be based on the cation exchange properties of the negatively charged macromolecules (colloids) contained in the P V F and the external egg membrane (Shepard 1987). Shepard and McWilliams (1989) suggested that these two materials effectively excluded anions and essentially provided an ion permeability barrier that was external to the embryo, yet part of the whole egg. Although most of this work has been 54  done in low ion concentration solutions, there is some evidence that the PVF-egg membrane complex may also provide reduced ion permeability in higher ionic conditions [Weisbart 1968 (data presented above), Shepard and McWilliams 1989]. The presence of a permeability barrier would alter the osmotic and ionic gradients that eyed and prehatch embryos in this study were actually exposed to and thus reduce the osmoregulatory activity required to maintain osmotic homeostasis during the SW exposure. This is consistent with the relatively similar Mo values 2  of chum and coho eyed and prehatch embryos acclimated to salinities up to 15 and 12 % o S , respectively. It supports the suggestion that both species use a similar mechanism to cope with the increased osmotic and ionic conditions encountered up to 15 and 12 % o S ; a mechanism that appears to have negligible direct influence on Mo . Based on SW tolerance data in higher 2  salinities it appears that the buffering capacity of the egg membrane-PVF complex diminishes as the external salinity increases.  Factors affecting 0 supply 2  It is possible that some of the observed metabolic changes may be related to changes in cutaneous permeability in response to changes in salinity alterations. Baggerman (1960) considered differences in cutaneous permeability as a possible reason why newly hatched Atlantic salmon (Salmo salar) larvae (1 wk posthatch) had better survival rates in 75% and 100%) SW than older larvae and fry (6 wk and 3 mo, respectively). Skin permeability in marine eggs is low in efforts to minimize diffusive loss of water and gain of ions (Mangor-Jensen 1987, Tytler and Bell 1989). Some studies have suggested that marine larvae are capable of changing cutaneous and branchial permeability (Brown and Tytler 1993). Skadhauge and Lotan (1974) and Nordlie et al. (1991) concluded that reduced oxygen consumption rates with increasing  55  salinities in the adults of two species of euryhaline minnow (Cyprinodon spp.), were probably the result of decreased gill and integument permeabilities. Ktiltz and Onken (1993) showed that opercular membrane permeability of adult Mozambique tilapia (Oreochromis mossambicus) decreased in response to increased salinity along with a subsequent decrease in ion fluxes. Is it possible that changes in cutaneous permeability affect oxygen permeability and ultimately oxygen diffusion? This would be especially significant in embryos and larvae since the cutaneous surfaces are very important in early life respiration (Rombough 1988a). If so, reduced oxygen diffusion might be another indirect effect of SW exposure that contributed to both the reduced Mo values and growth reductions in higher salinities. Unfortunately, there are 2  insufficient data to differentiate between the relative importance of changes in permeability compared with changes in activity and possible growth efficiency. Our current understanding of the changes in and control of membrane permeability in relation to ion and gas exchange remain relatively poor and much of that knowledge pertains specifically to amphibians (Lillywhite and Maderson 1988, Stiffler 1988). There is no doubt that an improved understanding of these relationships in the early life stages of fishes would assist in answering some of the permeability related questions posed here. Another consideration that may affect Mo values of fish in higher salinities is the effect of 2  tissue dehydration. Bath and Eddy (1979) showed that rainbow trout (15 cm) showed a 33% decrease in gill lamellar surface area after 30 min exposure to SW and suggested that this alone was sufficient to explain the initial decrease in Mo they observed during the first 6 h of SW 2  exposure. Dehydration probably results in collapse of branchial blood vessels and thus restricts blood flow and oxygen transfer. However, normal gill function is presumably restored once the fish regains osmotic and ionic balance (Shuttleworth and Freeman 1973). Similar mechanisms have been suggested as important aspects of epithelial permeability in cutaneous respiring 56  vertebrates such as amphibians and fish. The permeability of intercellular epithelial junctions undoubtedly depends on the osmotic state of the epidermal cells which in turn influences the permeability of the skin. Changes in the structure of intercellular spaces can affect diffusion of water and solutes, even independently of cellular swelling or shrinkage (Lillywhite and Maderson 1988). Moreover, perfusion of cutaneous surfaces by the microvascularization of the secondary circulation in fish (Burggren 1988, Randall and Wright 1995) would presumably be affected by dehydration in a manner that is similar to the adult fish gill. Further elucidation of the variability and control of epithelial permeability will undoubtedly contribute to a better understanding of the potential ontogenic and interspecific differences in active and passive water and ion transport across fish epithelia. If in fact S W exposure does result in reduced cell membrane permeability and reduced perfusion of cutaneous tissues then it seems probable that reduced oxygen diffusion into the tissues could reduce growth in embryonic tissues. This may be at least a partial explanation for the reduced growth observed in the prehatch stage of chum and coho.  Energetic implications of osmoregulation - larvae The most striking difference between chum and coho yolk sac larvae was their ability to tolerate high salinity conditions. Almost all of the chum larvae survived in 24 %oS (> 95% survival for 17 d) and some even survived in 30 %oS (-20% survival for 17 d), whereas coho did not survive in salinities above 12 %oS, therefore interspecific oxygen consumption rate comparisons were not possible above this salinity. There appeared to be a step-wise metabolic response on the part of the chum larvae that indicated increased metabolism in relation to increasing salinity. Although no quantitative measurements of larval activity were made during testing in the respirometers, qualitative observations did not indicate any obvious differences in activity rates  57  between the different salinity acclimated groups. Larvae residing in the higher salinities undoubtedly would be more buoyant and as a result may have swum more actively than those animals lying on the bottom of the rearing containers. However, as mentioned above, this was not apparent from observations made on the animals in these studies. Holliday et al. (1964) reported that unanaesthetized larvae of Atlantic herring larvae were less active in lower salinities (< 15 %oS), but that this was largely the result of these otherwise pelagic larvae sinking and resting on the bottom of the container due to their negative buoyancy in lower salinities. Along with differences in activity levels in the different salinities they also recorded highly variable Mo values. Similar results have been reported for other fish larvae (Pacific sardine, Lasker and 2  Theilacker 1962; Nile tilapia, De Silva et al. 1986). In contrast, variability among oxygen consumption rate measurements in this study were very low.  M0 and Larval activity 2  There are a number of additional factors that suggest that the differences in the chum oxygen consumption rate data are not related to a difference in activity levels among salinity treatments. In general, larval fish, especially salmon, are more active in light conditions than they are in dark (MacCrimmon and Kwain 1969, Hamor and Garside 1975, De Silva et al. 1986). Testing procedures were conducted in the dark and minimized physical disturbances during the measurement period in an effort to minimize spontaneous movement (see Appendix A for details). The respirometer system was designed specifically for measuring small changes in oxygen tension typical of embryos and larvae. A high degree of sensitivity was obtained by using low-volume respirometers (-20 mL total volume) that were monitored continuously for changes in oxygen tension (once every 0.25 min). Any unusual levels of movement were immediately apparent on the computer screen of the data logger and were not used for final 58  calculations of Mo values. This ensured that any sporadic activity was apparent without 2  actually needing to view the larvae; however, it should be noted that these methods would not detect differences in continuous activity levels. Simultaneous replicate measurements indicate that the variability was remarkably similar for all of the Mo measurements both within and 2  between developmental stages (Table 1.3, Fig. 1.5 and Fig. 1.6). This consistent variation would be less likely i f activity levels were markedly different between measurements of Mo  2  within  and between treatments and developmental stages.  Energy partitioning If one assumes that the activity levels were not significantly different between the salinity treatments, then changes in oxygen consumption rate of chum salmon larvae acclimated to different salinities represent a very significant change in energy utilization for these larvae. The increase in oxygen consumption rate above the baseline F W value (42.1 nmol-mg TDW^'-h ) -1  was 20%, 39% and 51%, respectively in 6, 24 and 30%oS. This magnitude of increase in M0  2  has not been shown before for F W fish larvae acclimated to varying SW concentrations. Although the oxygen consumption rates at 12 and 18%oS were not statistically higher than 0 %oS, the overall trend did appear to be one of increased oxygen consumption rate in salinities above 0 %oS. If the metabolic response to salinity is indeed a stepwise one with an increase from 0 %oS up to 6, 12, and 18%oS then it represents an increase in oxygen consumption rate of about 14 to 20%. Applying an additive energy allocation model to early life metabolism would suggest that this increase reflects additional energy required to osmoregulate around the isoosmotic point (plasma osmolality = 328.9 mOsm/kg =10.1 % S, data from SW challenge tests, 0  59  conversion from Alderdice et al. 1979) and that the higher oxygen consumption rate increases of 39% and 51% represent the additional energy allocated for osmoregulation at 24 and 30 %oS. It is unlikely that energy allocation is a simple process, especially during the early stages of fish development. Studies examining energetic costs of osmoregulation in juvenile and adult fishes indicate that changes in whole-animal metabolism in different osmotic and ionic conditions cannot be simply attributed solely or even directly to changes in the energy allocated to osmoregulation (Febry and Lutz 1987, Morgan and Iwama 1991, Kirschner 1993, Soengas et al. 1995). For example changes in other physiological variables such as Cortisol can alter oxygen consumption in juvenile and adult fish, via either its osmoregulatory role in S W adaptation of salmonids (Madsen 1990a) or its stimulatory effect on metabolism (Barton and Schreck 1987, Vijayan et al. 1994, Morgan and Iwama 1996). The role of Cortisol during the early life stages is less well understood but indications are that it enhances hypo-osmoregulation in Mozambique tilapia larvae in a manner similar to adults (Hwang and Wu 1993). If this similarity holds true for its stimulatory effect on metabolism, then it presumably also would have a positive effect on larval Mo . Such a response may be a contributing factor in the observed elevated oxygen 2  consumption rates in the higher salinity treatments. In addition to the added effects of SW exposure, such as hormone stimulation, the role of energy partitioning should also be considered. Rombough (1994) showed that the growth rate of chinook salmon embryos and larvae was independent of the oxygen consumption rate and suggested that this was due either to compensatory energy allocation or perhaps changing costs of basic processes such as growth. Current knowledge does not indicate unequivocally which scenario applies to fish embryos and larvae (Rombough 1994). Applying this non-additive model to the increased Mo values for chum larvae might predict that as the energy allocated to 2  osmoregulatory processes increased in higher salinities that energy allocated to other processes  60  such as growth would decrease. However, this was not apparent for chum larvae since no definitive changes in growth, (TDW and P ) were observed between 0 and 30 %oS. In contrast, r  although coho larvae did not show any significant salinity dependent differences in oxygen consumption rates, they did show a trend of decreasing growth, measured as P , in 12 %oS r  compared with 0 and 6 %oS. This response may be indicative of their inability to osmoregulate at this life stage, and be indicative of moribund larvae. Rombough's (1994) second suggestion for the lack of correlation between growth rate and oxygen consumption rate was that the cost of growth or other basic physiological processes may not be constant during these early life stages. The chum larva data from this perspective provokes speculation that salinity dependent increases in oxygen consumption rate may be related to changes, in this case increases, of the cost of some of these processes. It may be possible that even though growth did not change in chum larvae exposed to a range of salinities, that the energy required to maintain this level of growth under high salinity conditions actually increased, in addition to what might have been required for added osmoregulatory activities. Even if no obvious metabolic or growth response is observed this should not be seen as unequivocal evidence that significant energy allocation changes have not occurred, since compensatory changes by definition will be especially difficult to pinpoint. As Rombough (1994) noted, evidence for this more complex energetic budgeting is very limited and at that to date it had only been demonstrated with individual cells and not whole animals (Pannevis and Houlihan 1992, Houlihan and Smith 1993). A n example of more recent research addressing this question is Conceicao et al. (1997) who have shown that the very high growth rates of African catfish yolk sac larvae are the result of the minimization of the cost of protein synthesis in association with high rates of protein synthesis and elevated R N A efficiencies.  61  The lack of correlation between growth rate and consistent ontogenic patterns of metabolic intensity led Rombough (1994) to ask what the function was fulfilled by these apparently controlled changes in metabolic intensity. In other words where is the energy being expended i f it is not on growth? A similar question arises for the increased oxygen consumption rates of chum larvae in higher salinities. If these larvae are in fact allocating all or some of this additional energy to osmoregulatory processes then the next obvious question is where precisely is it being directed? What organs or tissues are doing this extra work and is it possible to measure some of these changes in relation to salinity? Given the osmoregulatory tissue available to yolk sac larvae the most likely candidates are the gills and the integument. Chloride cells have been found in the yolk sac epithelium of a wide variety of embryos and larvae including chum salmon (Kaneko et al. 1995) as well as on the gills of chum fry (Uchida et al. 1996). Presumably these cells would provide the majority of ion exchange tissue for developing embryos and larvae. Changes in C C structure and number have been reported for a number of different species offish embryos and larvae in response to SW exposure. (Hwang and Hirano 1985, Hwang 1989, Kaneko et al. 1995). These changes suggest that the cutaneous CCs are important in ion extrusion as part of the hypo-osmoregulatory process. Gills in these larvae also would be reasonably well developed with complete filaments and rudimentary lamellae (Rombough and Moroz 1990, Kaneko et al. 1995). Chapters 2 and 3 of this thesis will examine the structure and biochemical activity of these CCs in detail in an effort to provide some evidence for osmoregulatory function. In summary, salinity exposure had a variable effect on oxygen consumption rate in chum and coho salmon embryos and larvae. These differences appear to be at least partly related to different regulatory mechanisms at different developmental stages. The significantly better osmoregulatory abilities of chum salmon compared with coho were evident at the embryonic  62  stages but most apparent at the larval and fry stages of development. Survival data for larvae and fry and plasma osmolality values of SW challenged fry indicated that chum were more competent hypo-osmoregulators than coho. Reduced M o of chum and coho prehatch embryos 2  in 18 and 24 % S appear to be related to changes in growth rather than the energetic cost of 0  osmoregulation. Changes in TDW and P in chum salmon and to a lesser extent coho provide r  some evidence for better growth in 6 %oS. However, concluding that such trends are indicative of reduced energy requirements for osmoregulatory processes at low salinities must remain speculative due to the apparent complex nature of non-additive energy allocation during these early life stages. Metabolic responses of yolk sac larvae were also very different; most ostensibly because coho did not survive above 12 %oS, but also because of the marked, apparent step-wise elevation of chum larval oxygen consumption rates in salinities above 0 %oS. Proportionally these increases amounted to 20%, 39%, and 51% higher oxygen consumption rates at 6, 24, and 30 %oS and no respective changes in growth. Even though increased activity levels were probably not the cause of these increases, I am hesitant to relate these higher oxygen consumption rates directly or solely to increased osmoregulatory activities, once again due to the complexities associated with compensatory energy allocation which appears to be characteristic of early development in fishes. Nevertheless, the question arises whether it is possible to find evidence of expenditure of this additional energy in specific early life osmoregulatory mechanisms? The following chapters of this thesis are an effort to account for at least some of the increased aerobic metabolism measured in the higher salinities, by examining mechanisms such as the cutaneous CCs and their associated biochemical activity.  63  Figures - Chapter 1 CHUM  Hatch  COHO 7—V—V-W  Hatch — v • • wcv  —|,  CO  •d o E CD  > _co  E o  I  0  10  20  30  Days post-exposure  40 0  10  1  20  1  1  1  I  1  1  30  1  1  I  40  Days post-exposure  Fig. 1.1. Mortality rates of eyed and prehatch embryos and yolk sac larvae of chum (a to c) and coho (d to f) salmon exposed to a selected range of salinities. Legend for coho yolk sac larvae is the same as prehatch embryos (chum: filled symbols; coho: open symbols). Sampling for the acclimated exposure experiments described in the text was conducted 10 to 12 d post initial exposure at which time all of the test animals had been exposed to the final acclimation salinity for at least 7 d.  64  CHUM:  0  1  2  3  4  5  Time (d) Fig. 1.2. Cumulative hatching rate (time from first hatch) of chum and coho salmon eggs acclimated (> 7 d) as prehatch embryos to a selected range of salinities. Each line represents the cumulative numbers of successfully hatched and live larvae resulting from 60 eggs at each salinity; 6 replicates of 10 eggs. Note that both chum eggs acclimated to 18 and 24 %oS, and coho eggs acclimated to 12 %oS did not survive long after hatching and that coho eggs acclimated to 18 and 24 %oS died as eggs.  65  0  24  (OmOsm)  (740mOsm)  35 (1080 mOsm)  Salinity (%o)  Fig. 1.3. Mean plasma osmolalities of chum (CU) and coho (CO) salmon fry (100% yolk sac absorption) exposed to a 24-h S W challenge of 35 %oS and 24 %oS, respectively. Each mean is based on 5 to 7, ten (J.L replicate samples of pooled plasma (7 to 10 individuals/sample), except for coho 1995 24 %oS where the mean is based upon 3 samples. Note that coho fry transferred to 35 %oS died in < 24 h. Means (± SD) that share a common letter are not statistically different (one-way A N O V A , Tukey test,p > 0.05).  66  Chum  450 A  0%o (fry) to 35%o  • 420 A  A 11 A  5% (eyed) to 35%o 10% (eyed) to 35%o 6%o (prehatch) to 35%o 12% (prehatch) to 35%o 6% (larva) to 35%o 12% (larva) to 35%o 18%o (larva) to 35%o 24%o (larva) to 35%o 30% (larva) to 35% 0  0  0  0  390 H  0  • •  360  0  0  330 H y  L  35 0%o (fry) to 24%o  450 A  Coho  5%o (eyed) to 24% 6%o (prehatch) to 24%o 6%o (larva) to 24% 12% (larva) to 24% 0  £  420 A  0  0  0  390 A  360 A  330 A y  L  0  24  Salinity (%) 0  Fig. L 4 . Mean plasma osmolalities of long term acclimated chum and coho salmon fry (100% yolk sac absorption) exposed to a 24-h S W challenge of 35 and 24 %oS, respectively. In the legends, acclimation salinity is listed first, followed by the life stage (in brackets) at which acclimation began, and finally the challenge salinity; (eyed) = eyed stage embryo, (prehatch) = prehatch embryo, and (larva) = 50% yolk sac larva. Each mean is based upon 4 to 6, ten oL replicate samples of pooled plasma (7 to 10 individuals/sample). Means (± S D ) that share a common letter are not statistically different (one-way A N O V A , Tukey test, p > 0.05). Note that data for the 0 to 35 %oS and 0 to 24 % S transfers are the same as shown in (Fig. 1.3). 0  67  Fig. 1.5. Mean mass-specific oxygen consumption rates (M0 ) and mean tissue dry weights (TDW) of c h u m salmon eyed embryos (a), prehatch embryos (b), and 50% yolk sac larvae (c) acclimated for 7 to 10 d to a selected range of salinities. Each mean Mo (± SD) is based upon 4 replicates (except as noted by ^ where number of replicates equals 3); n = 10 individuals/replicate for panels 'a' and 'b', whereas n = 3 individuals/replicate for panel 'c'. Each mean T D W (± SD) is based upon a single sample; n = 40 individuals/sample for panels 'a' and 'b', and n = 12 individuals/sample for panel 'c'. Means that share a common letter are not statistically different (p > 0.05, Tukey test). 2  2  68  COHO:  r  (a) Eyed embryo M0 2  1 6  - 12 8 - 4 - 0  r  Q h-  1 6  CT  CT  E  - 12  & o E c  E  CT - 8  Q) ro i—  it TJ  c: o  - 4  CL  0) 3 (/> W  E 3  (fl  o o  T  C  12  70 -|  - 0  i r 18 24 (c) Yolk sac larva  30  60 -  r- 36 - 32  50 - 28  40 30 -  - 24  20 - ,  r 20 n 6  i i i 12 18 24 Salinity (°/oo) [7 d acclimation]  1  <-0  30  Fig. 1.6. Mean mass-specific oxygen consumption rates (Mo ) and mean tissue dry weights (TDW) of coho salmon eyed embryos (a), prehatch embryos (b), and 50% yolk sac larvae (c) acclimated for 7 to 10 d to a selected range salinities. Each mean M o (± SD) is based upon 4 replicates (except as noted b y where number of replicates equals 3 ) ; n = 10 individuals/replicate for panels 'a' and 'b', whereas n = 3 individuals/replicate for panel 'c'. Each mean TDW (± SD) is based upon a single sample; n = 40 individuals/sample for panels 'a' and 'b', and n = 12 individuals/sample for panel 'c'. Means that share a common letter are not statistically different (one-way A N O V A , Tukey test,/? > 0.05). 2  2  ( 1 )  69  8 (a) Eyed embryo  0  0  J  J  i  1  1  1  1  1  0 a'  6  12  18  24  30  I  1  1  1  1  1  0  6  12  18  24  30  Salinity (o/oo) [7 d acclimation]  Fig. 1.7. Mean percent tissue (P ) for chum and coho salmon embryos and larvae acclimated to a selected range of salinities. Means (± SD) that share a common letter are not statistically different. One-way A N O V A on ranked percent data (arcsine transformed), Dunn test, p > 0.05). r  70  Chapter 2 DISTRIBUTION A N D O N T O G E N Y OF C U T A N E O U S A N D B R A N C H I A L CHLORIDE CELLS IN C H U M A N D COHO S A L M O N E M B R Y O S A N D L A R V A E Introduction The osmoregulatory organs of teleost fishes include the gills, kidney, gut, and urinary bladder. Chloride cells on the filaments and lamellae of the gill appear to provide the dominant functional ion exchange unit of this organ. In marine and SW-adapted fishes these cells, first identified by Keys and Willmer (1932), function by extruding ions out of the body through active transport from the blood to the external environment via the gills (Foskett and Scheffey 1982). In F W fishes they appear to be integrally involved in the process of ion uptake, although the evidence for this is more circumstantial (McCormick 1995). Chloride cells of SW-adapted fishes are characterized by numerous mitochondria, an extensive tubular system, an apical pit or crypt, and leaky junctions with adjacent cells that are usually similar in structure, called accessory cells. In F W and FW-adapted fishes, CCs are characterized by numerous mitochondria (sometimes with less dense cristae), much less extensive tubular network, flush or evaginated apical surface, and tight junctions with adjacent epithelial cells (pavements cells) typically comprised of intercellular interdigitations (Zadunaisky 1984). The majority of research related to CCs has been conducted with juvenile and adult fishes and the role of these cells in the early life stages is much less well understood. This study was conducted to obtain a better understanding of CC prevalence, distribution, and ontogeny in the early life stages of chum and coho salmon.  71  Within the juvenile and adult life stages most of the research has focussed on the role CCs play in SW-adapted fishes; as a result less is known about the function of these cells in FWadapted fishes (McCormick 1995). The proposed F W function of CCs is one of ion uptake, specifically N a and Cl~, as well as acid-base balance through active and passive transport of H +  +  and H C 0 ~ (McCormick 1995). Indications are that, at least in adult fish gills, transport of some 3  of these ions may not be restricted to the CCs but also occurs in other epithelial cells such as pavement cells, (Goss et al. 1995). Some of the most direct evidence for CC involvement in ion uptake, in adult fish gills, comes from CC proliferation resulting from exposure to low ion content water (Perry and Laurent 1989, Laurent and Perry 1991). Differences in C C fine structure have been relatively well documented in SW-adapted versus FW-adapted adult fish (Pisam et al. 1987, Laurent and Perry 1991) and some larval stages of fish (Hwang 1989, Hwang 1990). Fresh water CCs generally are smaller and less columnar, have a less pronounced tubular system and often do not have apical and serosal contacts (McCormick 1995). Detailed information on the distribution, and fine structure of CCs in the early life stages of fishes remains limited, especially in FW. The process of fertilization initiates a cascade of ontogenic processes that originate with a single cell and culminate in a complex multicellular animal resembling the adult form. The early embryo consists largely of a mass of undifferentiated blastomeres and until the onset of gastrulation, uses a simple passive form of osmoregulation characterized by 'tight' cell membranes that limit ion and water exchange with the external environment. A transition from passive regulation to a more active system occurs between gastrulation and epiboly as the embryonic complexity and metabolic demands increase and greater transmembrane exchange is required (Alderdice 1988). Increased exchange of solutes and water with the external  72  environment necessitates an interim regulatory system in lieu of the adult osmoregulatory organs such as the gills, kidney, and gut, which remain either undifferentiated or underdeveloped until organogenesis reaches completion. During this period extrabranchial and extrarenal tissues appear to provide the necessary osmoregulatory capacities required for osmotic and ionic homeostasis of the embryo and larva (Alderdice 1988). Mortality data obtained in Chapter 1 are consistent with results that show chum embryos and larvae to be capable of hypo-osmoregulation in moderate salinities using extrabranchial tissues. In chum and coho salmon at least, these regulatory abilities improved throughout the yolk sac larva stage and may be fully functional by the emergent fry stage (100% yolk sac absorption). Chum yolk sac larva in Chapter 1 also showed increased oxygen consumption rates in 6 %oS followed by progressively even greater rates in 24 and 30 % o S . I ended that chapter by asking what mechanisms were responsible for the osmoregulatory abilities in these larvae and whether it was possible to observe any salinity related changes in these mechanisms that might assist in explaining the observed changes in oxygen consumption rate. Specifically I was interested in the role of cutaneous CCs in early life osmoregulation and any salinity related changes in their distribution, structure and function. Shelbourne (1957) identified the larval skin, specifically the Y S E , as the site of chloride exchange in plaice larvae (Pleuronectes platessa) and implicated eosinophilic cells (probably CCs) as the probable sites for this early life ion exchange. A n improved understanding of C C function in adult osmoregulation led subsequent researchers to suggest that chloride cells located in this external embryonic epithelium were the most probable sites of extrabranchial osmoregulation (Weisbart 1968). Although Jones et al. (1966) did not identify any CCs in Atlantic herring, Lasker and Threadgold (1968) and Depeche (1973) found CCs in the Y S E of Pacific sardine and guppy (Poecilia reticulata), respectively, that were structurally similar to  73  CCs found in the gills of SW-adapted adults. However, the presence and function of cutaneous CCs in embryos or larvae of F W fishes remained equivocal until quite recently. Leatherland and Lin (1975) initially reported not finding any cutaneous CCs in coho salmon embryos and larvae, however, in a later study Shen and Leatherland (1978a) found CCs in the Y S E of rainbow trout adapted to F W and dilute SW. Because the cellular structure was not indicative of an osmoregulatory role (i.e., no apical pit, no obvious basolateral tubules), they suggested that these cells were not functional in ion regulation. Guggino (1980a) reported CCs located on the yolk sac and trunk of the euryhaline killifish (Fundulus heteroclitus) before hatching and not in the branchial epithelium until after hatch. Hwang and Hirano (1985) and Hwang (1989, 1990) also found cutaneous CCs in the larvae and juveniles of 2 F W fish (ayu, Plecoglossus altivelis, and carp, Cyprinus carpid), and 1 SW fish (flounder, Kareius bicoloratus). They reported finding lower cutaneous C C densities in stenohaline versus euryhaline species, and C C fine structural changes following SW and F W transfers that led to the conclusion that these cells played a critical role in early life osmoregulation. Many of these changes were similar to those documented for adult fish (Zadunaisky 1984, McCormick 1995). Recent studies on the early life stages of both F W and SW fishes continue to improve our understanding of the significance of CCs in early life osmoregulation. Examination of the Y S E in the euryhaline Mozambique tilapia embryos revealed high densities of CCs on the yolk sac in both embryos and larvae (Ayson et al. 1994). Chloride cell surface areas of SW-reared larvae consistently decreased within 2 d of transfer to FW. The opposite response occurred when the reverse transfer was done. Kaneko et al. (1995) reported on the effective hypo-osmoregulatory abilities of eyed stage chum salmon embryos in 100% SW over a 7-d period and related it to increased surface area of Y S E CCs and development of apical pits characteristic of typical SW CCs.  74  Studies that have focussed on cutaneous CCs in embryos and larvae consistently report CCs located on the yolk sac and in a few cases the trunk, whereas studies that have examined the later developmental stages emphasize CCs in the branchial epithelium (e.g., L i et al. 1995). Only a few studies have reported CCs on cutaneous surfaces other than the gills and yolk sac, and details on whole-body C C distribution and densities are often limited or unavailable. Fluorescent staining techniques have greatly simplified this procedure (Bereiter-Hahn 1976, McCormick 1990a). Using fluorescent stains Wales and Tytler (1996) examined Atlantic herring, but contrary to Jones et al. (1966) who used older methods, found numerous CCs distributed extensively over much of the dorsal yolk sac and along the ventral trunk with some very specific banding patterns associated with the myosepta of the trunk musculature. Hwang (1989) reported high numbers of CCs on the anterior portion of ayu larvae close to the head and on the anterior region of the trunk but reduced CC numbers towards the posterior regions. Newly hatched larval carp, incubated and reared in FW, showed only low densities of small cutaneous CCs on the Y S E and none on the gill or branchial epithelium, trunk epithelium (TRE), or opercular epithelium (OPE) (Hwang 1989). Juvenile carp (30 d posthatch) exposed to 10 %oS SW suffered 80% mortality within 48 h and showed no alteration of branchial C C fine structure (Hwang and Hirano 1985). These fish obviously utilized their CCs only for functions related to a F W existence. However, no reports of whole-body C C distributions exist for other species, specifically not salmonids. Chum salmon embryos and larvae are more SW tolerant and more competent osmoregulators than coho salmon (Chapter 1, Weisbart 1968). These differences are consistent with early life history patterns of these two species (see, General Introduction) but very little is known about the early life osmoregulatory mechanisms in chum salmon and virtually nothing is known in coho salmon. They may use cutaneous CCs like other species but a number of 75  questions still remain. Thus, the objectives of this study were as follows: (1) to establish the presence or absence of cutaneous CCs in chum and coho salmon embryos and larvae, (2) to examine how the distribution of CCs changes during ontogeny, (3) to compare interspecific C C densities and distributions in chum and coho embryos and larvae in an effort to explain different SW tolerance and osmoregulatory abilities, (4) to examine the numbers of cutaneous CCs in relation to of the number of branchial CCs and calculate the relative contributions of each population of CCs to early life osmoregulation and finally (5) to investigate a simple model for partitioning of relative osmoregulatory capacities in the cutaneous versus the branchial tissues (based on estimates of the total number of cutaneous and branchial CCs per individual) and test whether the suggestion by Rombough and Moroz (1997) that gills in larval fish function as an osmoregulatory organ before a respiratory one is correct.  76  Methods Embryos and larvae of chum and coho salmon were acclimated to a range of salinities (0 to 30 %oS) and then sampled for distribution and density of cutaneous and branchial CCs using fluorescent microscopy. A preliminary T E M examination was also conducted on prehatch embryos to look for cellular features characteristic of CCs of fish adapted to SW and FW. Estimations of the surface areas of embryo and larva surface cutaneous and branchial tissues were used to model the relative ontogenic changes of the total number of cutaneous and branchial CCs.  Fish Chum salmon eggs were obtained from Big Qualicum hatchery (near Qualicum, Vancouver Island, B C , Can.) in January of 1995. Approximately equal numbers of eggs at the eyed stage of development (stage 20, Jensen 1988) were obtained from 3 spawnings (1 9 :1 a mating). Coho salmon eggs were obtained from the Chase River Hatchery (Nanaimo, B C , Can.) and were the same group that was used in the acute exposure respirometry study (1994/95) presented in Chapter 1. These eggs also were obtained at the eyed stage of development from 3 spawnings (1 9:1 d mating). Egg incubation, larval rearing and SW exposure conditions were the same as the acclimated SW exposure study (1993/94) outlined in the Methods section of Chapter 1 (i.e., at least 7 d acclimation in final test salinity). Water quality parameters of the F W used for incubation and rearing were as follows: conductivity = 50.9 uS/cm; total alkalinity as C a C 0 = 3  13.6 mg/1; ion content (in mM): [Na ] = 0.01, [K ] = 0.006, [Ca ] = 0.2, [Mg ] = 0.02, and +  +  2+  2+  [CF] = 0.16. These values are similar to values obtained from the rivers from which the chum and coho broodstock were collected. 77  Fluorescent microscopy - tissue preparation Embryos and larvae of chum and coho salmon were surveyed for distribution and density of epithelial chloride cells using fluorescent microscopy (McCormick 1990a). For both chum and coho, two individuals were sampled from each of the 4 to 6 test salinities (Table 1.1) at each of the 3 developmental stages (eyed embryo, prehatch embryo, and 50% yolk sac larva). Each specimen was sampled for epithelial tissues of the yolk sac (YSE), head (HDE), opercula [internal (OPE;) and external (OPE ) surfaces], trunk or main body (TRE), and finfolds/fins e  (FFE/FNE), and gills (Fig. 2.1). Immediately prior to tissue removal, embryos and larvae were killed with an overdose of TMS (500 and 120 mg/L, respectively) followed by severance of the spinal column and a cut into the cranium. Dissected tissues were placed into cooled (10 ± 1°C) low-K salmon Ringer solution +  (LKSR) prior to staining with fluorescent dye (LKSR in mM: 145 NaCI, 3.4 C a C l , 2  1.7 N a H P 0 , 1.3 M g S 0 , 0.2 N a H P 0 , 0.1 KCI, adjusted to pH 7.1 with H S 0 following 2  4  4  2  4  2  4  aeration with 100% oxygen; modified after McCormick 1990a). Yolk sac epithelium was removed using surgical micro-scissors and micro-forceps. With careful dissection it was possible to make a ventral midline cut and avoid puncturing the yolk sac syncytium and thus avoid contaminating the sample. Once the two resulting halves of the yolk sac tissue they were cut away from the side of the body, small expansion cuts were added to make them lie as flat as possible. For dissection of larval stage specimens, the midline cut was eliminated and only the side body wall was cut around the circumference of the dorsal part of the yolk sac to remove the entire Y S E in one piece, along with a small strip of the body wall for reference. The pectoral fin girdle was often also included as a reference point on most yolk sac samples. Opercula and gills  78  were also removed using micro-scissors. For both of the embryonic stages skin samples of the trunk, finfold, and head regions were obtained by simply sampling the respective body parts, whereas for the larval stage, two approximately 2 by 4 mm, pieces of skin were removed from the musculature of the trunk while the head was examined whole. To allow the head to lie as flat as possible a medial cut was often made through the upper jaw in addition to the one made earlier in the nape of the neck. Finfold samples were examined along with the main trunk of the body for the two embryonic stages but had mostly disappeared in the larval stage. Some estimates were obtained for pelvic and caudal fins from Y S E and TRE examinations at this stage. Following dissection, tissues were transferred from the L K S R to 5 mL of 2 u M dimethylaminostyrylethylpyridiniumiodine (DASPEI, Bereiter-Hahn 1976) in L K S R with 5.5mM glucose added. This staining solution was kept at 10°C (± 1°C) and aerated with pure oxygen for 1 h (McCormick 1990a). Good tissue viability was assumed to be maintained for at least 2 h after dissection on the basis that intermittent muscle contractions (trunk and heart) were observed throughout this period. Tissues were rinsed twice in 4 mL of L K S R (10°C) for 3 min, placed on a glass slide (apical surface up) and covered with a glass coverslip. Mounted tissues were examined using a Zeiss Standard microscope set up for epiflourescence with an H B O 50W mercury illuminator, a 450 to 490 nm bandpass excitation filter, a 510 nm chromatic beam splitter and a 520 nm longwave pass filter. Photographs were taken using 100 ISO colour slide film for 30 to 70 s depending on the magnification used.  Fluorescent microscopy - analysis Chloride cell densities (CC-mm ) were measured by counting CCs in 3 to 10 fields -2  (0.56 mm at 400x magnification) per tissue sample. The total number of fields counted for both 79  species at each of the developmental stages and test salinities was 1337. Since these data were collected as an initial survey of CC distribution and density, the counting procedure was conducted to be representative and not strictly random. Variability of density counts was very high due to location-specific differences for a given tissue and differences in how flat a given tissue laid in one preparation compared with another; these data are presented as means and ranges. A number of control samples were processed and examined without DASPEI staining to test for autofluorescence of tissues; no significant fluorescence was observed in any of these samples. In a number of the sampled tissues, localized differences in C C density were observed, i.e., higher numbers of CCs on the dorsal region compared to the ventral. This localized variability was most evident for Y S E and H D E samples from the two embryonic stages. For Y S E samples these differences were categorized by dividing the tissue into 3 separate zones for the purpose of description and counting; the dorsal Y S E zone, the middle Y S E zone, and the ventral Y S E zone (Fig. 2.1). (Note that the division of this tissue into discrete zones is artificial and incorporates the inherent bias of representing a continuous biological variable with a discontinuous numerical scale.) The ability to make accurate and precise CC density estimates of Y S E tissue was confounded by the fact that it was difficult to maintain this tissue in a completely flat state; as a result, wrinkling of tissue preparations sometimes caused problems. The major concerns were overestimation of C C densities and difficulty in obtaining clear photographs, since not all of the CCs were in the same focal plane. Tissues that were obviously wrinkled were not used for counting. For the H D E tissue samples, counts on and around the two cranial mounds on the dorsal surface of the head were either zero or very low compared to all other countable regions. Therefore, C C density counts for the H D E were split into two sub-categories; HDE-general 80  (HDE ) and HDE-cranium (HDE ) (Fig. 2.1). Chloride cell counts on the body trunk were split g  C  into two separate categories; the trunk proper (TRE) and finfolds (FFE), for the embryonic stages or fins (FNE), for the larval stage. In the larval stage the OPEj generally had only a band of CCs near the trailing edge that extended in a dorso-ventro pattern with few i f any CCs on other areas. Although the whole surface was always examined, routine density estimates for this tissue were confined to this band of tissue.  Cutaneous versus branchial surface comparisons - total numbers of chloride cells Surface area calculations were made for both the cutaneous and branchial surfaces and used to estimate the total number of CCs present on the entire cutaneous as well as branchial surfaces of an individual embryo or larva. These values were estimated for all three developmental stages and further extrapolated for the fry stage.  Tissue Surface Area Estimates of surface areas for specific tissues were calculated for chum salmon using a series of equations developed by Rombough and Moroz (1990) for similar size chinook salmon: A = alvt; where, A is the cutaneous surface area (mm -tissue  or mm -individual ), a and b are  constants, and M i s the tissue wet mass or weight in g (tissue only, not including yolk material). Tissue wet weight (TWW) was estimated as 6 times TDW, assuming an 83% moisture content. Values for the fry stage (100% yolk sac absorption) were calculated using mean T W W for chum fry [mean ± SD (»), 0.2671 ± 0.0137 g (8), unpublished data]. Surface areas estimates were calculated for the following tissues: total cutaneous, total branchial, F F E & FNE, T R E & H D E ,  81  and Y S E . Yolk sac epithelium surface areas were obtained by subtraction [A of Y S E = Tot. cutaneous - (FFE & FNE + TRE & HDE)].  Total numbers of chloride cells per individual Whole-body estimates of the total number of CCs per individual were calculated by summing the total estimated number of CCs per tissue type for all of the major cutaneous surfaces for which surface areas were estimated (e.g., Y S E , TRE, HDE, FFE). Total numbers of CCs for a given tissue were estimated by multiplying the mean CC density of that tissue by the calculated surface area. Since some of the tissues had different zones of CC density, weighted means of C C density were calculated as follows: mean Y S E density for the two embryonic stages was weighted assuming proportions of 0.1, 0.5 and 0.4 for Y S E , Y S E , and Y S E , respectively d  m  V  (Fig. 2.1), whereas for the larval stage the mean was weighted with proportions of 0.1, 0.9 for Y S E and Y S E , respectively; a combined mean density for TRE and H D E ( H D E only) d  m  g  (Fig. 2.1) tissues was weighted with proportions of 0.8 TRE and 0.2 HDE; and OPE tissue densities were assumed to be the same as TRE estimates.  Electron Microscopy - tissue preparation and analysis Gills from prehatch chum salmon embryos acclimated to F W and 24 % S SW were sampled (2 0  per treatment) after a 7-d acclimation period. Individuals were killed in the same manner as those in the fluorescent microscopy protocol. Whole embryos were fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) for 4 h at room temperature (22°C) and overnight (12 to 16 h) at 4°C. Fixed samples were rinsed (3x15 min), and stored in 0.1 M phosphate buffer for less than 6 wk before final processing for T E M .  82  For final processing the branchial basket was excised and transferred to 0.1 M sodium cacodylate buffer, and postfixed in reduced osmium tetroxide [1.5% K F e C N and 1% O s 0 , in 4  6  4  0.1 M cacodylate buffer (pH 7.3); Karnovsky 1971] for 1 h at room temperature. Tissue samples were then rinsed in 0.1 M cacodylate buffer (3 xlO min), and stained en bloc in 1% aqueous uranyl acetate for 1 h at room temperature, and rinsed in double distilled water ( 3 x 1 0 min). Tissue were prepared for embedding in Epon 812 (PolyScience, Warrington, M A , USA) by serial dehydration with ethanol and propylene oxide. Ultrathin sections were mounted on copper grids, and counterstained with saturated uranyl acetate and lead citrate (2 min each). Observation were made using a Phillips T E M 300 at 80kV. The purpose of this examination was to obtain representative examples of C C structure typical of the experimental FW and SW treatments and was not intended as an extensive survey of all parts of the gill and or tissues containing cutaneous CCs.  83  Results Chloride cell distribution - fluorescent microscopy Fluorescent microscopy (DASPEI) revealed CCs distributed extensively on almost all external cutaneous surfaces including the yolk sac, head, operculum, trunk, and finfolds (see Fig. 2.1 for location). Gills also contained very high densities of CCs but due to the three-dimensional nature of this tissue and inconsistencies in staining effectiveness, no density estimates were obtained.  Acclimation salinity and ontogenic trends Except for chum yolk sac larvae, no significant trends in CC density were observed as a function of acclimation salinity for any of the three developmental stages, thus these values were pooled to provide mean densities for each developmental stage. In chum larvae however, there was a general pattern of decreased CC density in the higher salinities; the highest densities occurring in 0 and 6%oS and the lowest in 12 to 30%oS (Table 2.1). Therefore, the 0 and 6 %oS were pooled separately from the 12 to 30 %oS treatments. The pattern of ontogenic changes in cutaneous CC density was the same for both chum and coho; densities were high during the embryonic stages and decreased during the yolk sac larval stage (Table 2.1 and Table 2.2). There were no significant salinity-related differences in CC density in either chum or coho embryos and larvae, except for chum yolk sac larvae in higher salinities (12 to 30 %oS), which had significantly lower CC densities. Variability of the CC estimates was high. Mean CC densities and ranges of YSE and TRE tissues were generally very similar in magnitude for both of the embryonic stages in chum and coho salmon, except for the  84  Table 2.1. Chloride cell densities on the skin of chum salmon embryos and larvae pooled across salinity treatments for each developmental stage (see text for details). SPECIES  Chloride cell densities in different tissues  Tissue Dev. stage (sal. range, %oS)  (CC-mm ) -2  CHUM  Yolk sac epithelium  YSE  YSE  r  YSE,  n  Mean  Range  Mean  Range  Mean  Eyed embryo (0-15)  852  577 to 1340  Prehatch embryo (0-18)  802  346 to 1282  431 412  109 to 891 141 to 897  188 221  Yolk sac larva (0-6) Yolk sac larva (12-30)  725  449 to 795 154 to 487  458 238  218 to 949  no data  a  64 to 378  no data  a  281  TRE Trunk & finfold/fin epithelium  FFE VFNE  c  Range  Mean  Range  Eyed embryo (0-15) Prehatch embryo (0-18)  758 738  494 to 1269 295 to 1321  619 640  519 to 769 365 to 1115  22  Yolk sac larva (0-6) Yolk sac larva (12-30)  457 252  179 to 718 71 to 378  481 207  218 to 769 64 to 372  33 68  OPE Opercular epithelium  b  c  c  OPE,  e  Total fields counted  Range  Mean  Eyed embryo (0-15)  572  218 to 1237  23  0to64  36  Prehatch embryo (0-18)  708  58 to 1090  213  0 to 359  65  Yolk sac larva (0-6)  207  0 to 551  155  0 to 673  27  Yolk sac larva (12-30)  86  0 to 410  53  0 to 558  65  HDE  Mean  Range  Mean  Range  Total fields counted  Eyed embryo (0-15)  730  372 to 1064  42  Oto 199  25  Prehatch embryo (0-18)  766  321 to 1795  18  Oto 147  22  Yolk sac larva (0-6)  490  173 to 686  no data  d  —  9  Yolk sac larva (12-30)  156  32 to 276  no data  d  —  11  Head epithelium  g  C  20 48  60  Mean  HDE  Range  64 to 545  149 137  Total fields counted  Mean  b  51 to 506  Total fields  YSE = yolk sac epithelium, TRE = trunk epithel., OPE = opercular epithel., HDE = head epithel. YSE zone had disappeared by this developmental stage. FFE (finfold epithel.) tissue was most prevalent in the embryonic stages. FNE (fin epithel.) tissue was only found in the yolk sac larval stage. No data available for this tissue; probably due to inadequate DASPEI penetration. Note: Samples from different salinity treatments were pooled for a given developmental stage; n equals 8 individuals for each of the 2 embryonic stages, and 4 and 8 individuals, respectively for the yolk sac larvae pooled in 0 to 6 %oS and 12 to 30 %oS. 8  V  b  c  d  85  Table 2.2. Chloride cell densities on the skin of coho salmon embryos and larvae pooled across salinity treatments for each developmental stage (see text for details). SPECIES  Chloride cell densities in different tissues  Tissue Dev. Stage (sal. range, %oS)  (CC-mm ) -2  COHO  Yolk sac epithelium Eyed embryo (0-15) Prehatch embryo (0-12) Yolk sac larva (0-18)  YSE  YSE  d  Mean  Range  Mean  834 855 571  596 to 1365 667 to 1147 436 to 1109  489 465 347  TRE Trunk & finfold/fin epithelium Eyed embryo (0-15) Prehatch embryo (0-12) Yolk sac larva (0-18)  YSE  m  Range 71 to 1000 308 to 641 128 to 929  FFE /FNE b  c  Mean  Range  191 191 no data  13 to 365 26 to 340 a  Mean  Range  Mean  Range  Total fields counted  691 938 520  282 to 1256 449 to 1321 115 to 878  647 880 544  ~436to1096 365 to 1474 295 to 949  51 30 32  OPE  b  b  c  OPE;  Mean  Range  Mean  Range  Total fields counted  154 to 1308  68  0 to 346  55  Prehatch embryo (0-12)  596 722  Oto 1103  168  0 to 577  Oto 679 0 to 577  32  Yolk sac larva (0-18)  163 78  Opercular epithelium Eyed embryo (0-15)  e  HDE  HDE  Mean  Range  Mean  Range  Eyed embryo (0-15)  542  333 to 1051  2  0 to 96  16  Prehatch embryo (0-12) Yolk sac larva (0-18)  734 524  397to 1577 244 to 974  175 16  Oto436 Oto 160  24 20  C  —  Total fields counted 126 60 32  62 Total fields counted  Head epithelium  g  V  YSE = yolk sac epithelium, TRE = trunk epithel., OPE = opercular epithel., HDE = head epithel. YSE zone had disappeared by this developmental stage. FFE (finfold epithel.) tissue was most prevalent in the embryonic stages. FNE (fin epithel.) tissue was only found in the yolk sac larval stage. Note: Samples from different salinity treatments were pooled for a given developmental stage; n equals 8 and 6 individuals, respectively for the 2 embryonic stages, and 8 individuals for the yolk sac larval stage. a  V  b  0  86  significantly higher densities recorded for prehatch coho embryos (Table 2.2), in addition to the noticeably larger YSE zone (Fig. 2.1). Although densities of the OPE and HDE tissues were d  slightly more variable, they were still similar for both species. Chloride cell density did not vary consistently in a dorso-ventral fashion along the length of the trunk of either embryos or larvae, although in some samples density counts were suggestive of a slightly higher number of CCs on the ventral half of the trunk below the lateral line compared to the dorsal half. However, variability of CC density was too high to provide conclusive support for this observation.  CC distributions within different tissues Patterns of CC distribution (Fig. 2.1) and density are summarized for chum and coho embryos and larvae in Table 2.1 and Table 2.2, respectively. Chloride cells were distributed extensively over the YSE with decreasing densities toward the ventral surface of the yolk sac (Fig. 2.3, Fig. 2.4). The YSE zone had the highest CC densities and in chum salmon comprised a more or d  less semicircular pattern around the base of the pectoral fin. In coho embryos and yolk sac larvae this zone was larger since it extended further in a posterior direction, to a point about equidistant between the pectoral fin and the vent (Fig. 2.1). The Y S E was an area of moderate m  to high CC density located below the YSE to about halfway down the side of the yolk sac d  below the pectoral fin and about a third of the way down at the posterior end of the yolk sac. There was not a clear delineation between the YSE and Y S E zones yet by comparing the CC d  m  density estimates it was possible to categorize a given field of examination as one or the other according to the criteria for these zones. The ventral zone (YSE ) was an area of low to V  moderate CC density and comprised the remaining area below the Y S E . For yolk sac larvae the m  87  Y S E zone was no longer distinguishable from the Y S E zone, since much of the yolk sac V  m  proper had been absorbed, therefore these two zones were combined and the total number of zones was reduced to two for this developmental stage, Y S E and Y S E . Chloride cells in the d  m  Y S E were very obviously less fluorescent in the larval stage and appeared to be breaking down m  because their structure was much less discrete (Fig. 2.5). The cell structure was much less discrete at this stage and obtaining accurate counts was not possible due to the irregular cell outlines. Chloride cells in the skin proper, TRE, OPE , and HDE, were also very numerous (Fig. 2.6, e  Fig. 2.7, Fig. 2.8) at all 3 developmental stages. In a few tissue preparations, it was obvious that the CCs were definitely located in the external most layer of the epithelium since they were located above the melanophores (Fig. 2.9), however, this was not necessarily seen along the entire length of the body. Chloride cells in these tissues did not show the same type of apparent breakdown as was seen in larval Y S E tissues (Fig. 2.5). However, compared with the embryonic stages, some increased variability in CC size was observed in the more developed larval TRE samples in both F W and SW treatments (Fig. 2.10); this variation appeared to be the result of larger single cells as well as clumps of cells (Fig. 2.10 inset). Like Y S E tissue, TRE, OPE , and e  H D E tissues in larvae appeared to be generally less fluorescent than in the embryos. During the two embryonic stages, high densities of CCs were also observed in the finfold tissues that ran along the entire dorsal ridge of the body around the caudal end to a point just ahead of the vent on the ventral surface of the body (Fig. 2.1). At these developmental stages C C distributions on the FFE tissues and the FNE tissue of the pectoral fin appeared to be continuous with the TRE and showed similar densities to the adjacent TRE tissue. At the transition from one tissue type to the other there was no obvious difference in C C density  88  between the trunk and finfolds; CC densities on both the FFE and FNE tissues adjacent to TRE and YSE tissues, respectively, were similar for both species (Table 2.1 and Table 2.2). Chloride cell density did however, decrease towards the outer edge of the fin tissues and very few, if any CCs were observed in the distal 1/3 of the finfolds. Chloride cells were easily observed on both sides of finfolds and embryonic pectoral fins by simply focussing through the tissue from the upper surface down to the lower one. In yolk sac larvae, fewer CCs were observed on fins that had developed the characteristic adult fin structure including the fin rays, although in some preparations they appeared in relatively high densities near the base of the fins.  DASPEI staining - tissue and developmental stage differences Chloride cell fluorescence varied in brightness depending on the specific tissue type as well as the developmental stage. The brightest and most discrete CCs were observed in the YSE of eyed and prehatch embryos compared to all of the other tissues, especially in coho (Fig. 2.3). In addition to the ontogenic decreased CC density in larval YSE tissues there were also very pronounced changes in shape and brightness compared to the eyed embryo stage. These changes were most pronounced in yolk sac tissues but were also evident in internal opercular tissues. Chloride cells in the earlier embryonic stages were rounded with discrete edges and usually fluoresced brightly so that individual CCs appeared as solid or semi-solid masses (Fig. 2.4). In comparison, CCs in the larval YSE and OPE tissues were nearly always amorphous with unclear edges and faint fluorescence; it appeared as if the CCs were in the process of breaking apart (Fig. 2.5 and Fig. 2.13). These features were further confounded by the reduced likelihood of obtaining brightly stained tissues in larval tissue samples, especially in chum salmon. Lack of sufficient penetration or diffusion of DASPEI into thicker tissues appeared to be a limiting factor for viewing CCs in TRE and HDE tissues (e.g., unskinned sections of trunk and 89  head as well as skinned sections of TRE with a thin layer of muscle still attached). Fluorescing CCs were commonly observed at sites where cuts had been made (Fig. 2.6) but not in other areas where the subcutaneous tissue had not been openly exposed to DASPEI. As a result, staining was not always consistent throughout the duration of this study. This aspect of DASPEI staining may have produced a bias in CC localization that resulted in visualizing fewer CCs in the thicker tissues at the yolk sac larva stage and thus may have underestimated the actual number of CCs in this later stage.  Branchial chloride cells Chloride cells on the gills were numerous and widely distributed but as previously mentioned, they were not quantified due to the three-dimensional nature of this tissue and the inconsistency with which DASPEI staining occurred. Some gill sample preparations did however, provide useful insights into C C distribution and arrangement at the three different developmental stages. Even though gills in early eyed coho embryos (between stage 20 and 21, Vernier 1969) were comprised only of small filament buds extending from the gill arches (Fig. 2.14 inset), CCs were observed along the filament bud surfaces and to a lesser extent, the gill arch epithelium (Fig. 2.14). Similar observations were found on the slightly more developed gills of well-eyed chum embryos; the filaments had differentiated more and were covered with numerous evenly spaced CCs (Fig. 2.15a and b). Gills of prehatch embryos had well differentiated filaments that contained numerous, widely distributed CCs. However, the CCs were no longer randomly arranged; instead, rows of CCs were paired together at what was presumably the base of developing gill lamellae (Fig. 2.16). It is not completely clear whether the lamellae had not yet developed at that point or whether they simply had not absorbed any stain and thus remained invisible. Based upon other gill sample preparations it is most likely that they had not yet 90  developed since in other cases non-fluorescing tissues were still visible to some degree. Similar but more discrete rows of CCs were evident in the gills of yolk sac larvae of chum (Fig. 2.17a) and coho (Fig. 2.17b and c), respectively, even though the lamellae proper were still not evident.  Chloride cell structure - T E M Although no differences in CC density were observed in relation to SW acclimation, there were differences in CC structure in response to SW exposure. T E M examination of gills from prehatch chum salmon embryos acclimated to F W and SW showed filamental (lamellae remained undeveloped) CCs with numerous mitochondria and other structural features characteristic of fish acclimated to F W and SW, respectively. Chloride cells from FWacclimated embryos had an apical surface that was flush with or evaginated from the adjacent epithelial cells and did not have an obvious network of basolateral tubules (Fig. 2.11), Chloride cells from SW-acclimated embryos had a much more noticeable basolateral tubule network as well as an apical crypt (Fig. 2.12). In embryos incubated in F W all CCs resembled those in Fig. 2.11, whereas in the SW treatment some CCs had a fine structure that was more characteristic of F W cells than SW ones. Although these data are only qualitative, more CCs with a characteristic F W appearance occurred in coho larvae acclimated to SW than chum acclimated to similar conditions.  91  Discussion Cutaneous CCs have been identified in embryos and larvae of fishes as the most likely site of ion regulation during the developmental period when the gills are still undergoing differentiation (Shelbourne 1957, Alderdice 1988, Hwang 1990). Chloride cells have been located in the yolk sac epithelium (Pacific sardine, Lasker and Threadgold 1968; Killifish, Guggino 1980a; goby, Chaenogobius urotaenia, Katsura and Hamada 1986; Mozambique tilapia, Ayson et al. 1994; chum salmon, Kaneko et al. 1995; Atlantic herring, Wales and Tytler 1996); the trunk (goby, Katsura and Hamada 1986; carp and ayu, Hwang 1989, 1990; Atlantic herring, Wales and Tytler 1996); and the opercular epithelium (marine flounder, Hwang 1989). The results obtained for salmonids in this study further support these previous studies of other species in that CCs were located on all of the listed epithelia as well as others (e.g., fmfolds, head region). The observed distribution of CCs on the cutaneous surfaces appear to be more wide spread than those reported previously on any one species. Significant proportions of all major external epithelial surfaces were covered with moderate to high densities of CCs (Fig. 2.1). Although other studies investigating early life osmoregulatory abilities and mechanisms have localized CCs on some of those epithelia, none have shown or reported such extensive distributions with such high densities. In general, the presence or absence of cutaneous CCs as well as the ability to modify the cellular fine structure appears to be related to the ability of larvae to acclimatize to osmotic and ionic conditions that are different from their normal environment. For example, Hwang and Hirano (1985) and Hwang (1989) examined the SW tolerance and associated cutaneous CCs in the larvae of two freshwater fish species, the euryhaline ayu and the stenohaline carp as well as one saltwater species, a stenohaline marine flounder. They noted that the carp had the poorest  92  SW tolerance and very few cutaneous CCs. Moreover, unlike the ayu, larval carp had small CCs distributed sporadically over the body and were unable to modify the cellular fine structure of these cells from the typical F W CC fine structure (tight interdigitated intercellular junctions and a less dense basolateral tubular system) to that of the SW fine structure (leaky intercellular junctions and a dense tubular system). The larval ayu and juvenile flounder on the other hand were able to make these modifications rapidly, often within 3 h of exposure and thus survive transfers to SW and FW, respectively. Although chum embryos and fry in this study (Chapter 1) were more tolerant to SW exposure than coho salmon, no obvious interspecific differences in CC density were observed in embryos or larvae that were acclimated to F W compared with SW. The C C densities (means and ranges) were surprisingly similar between these species. A preliminary T E M examination of chum Y S E CCs showed at least some structural features typical of SW-adapted CCs (Fig. 2.12), suggesting that SW-acclimated chum embryos possessed CCs capable of ion extrusion. Although in coho embryos acclimated to SW, some CCs were observed with a characteristic SW fine structure, many of the them still had fine structures characteristic of FW-acclimated fish. A more thorough examination of C C structural changes in relation to SW exposure is needed to unequivocally determine if indeed there are consistent differences between chum and coho embryos and larvae, and whether these differences substantiate the significant differences in SW tolerance from an osmoregulatory perspective. It seems reasonable to postulate that the limited SW tolerance of coho embryos and larvae is related to an inability to successfully modify C C fine structure from the F W type to the SW one.  93  Intra- and interspecific comparisons in salmonids Until recently the majority of early life stage studies have been conducted on species other than salmonids, such as plaice, herring, ayu, and tilapia. Some earlier studies that examined cutaneous CCs in salmonids acclimated to F W and SW have reported contradictory information. Light microscope and T E M examinations of the Y S E of F W coho salmon prehatch embryos (Leatherland and Lin 1975) did not show any cutaneous CCs, whereas similar examinations of Y S E and branchial epithelia in prehatch embryos and yolk sac larvae of the same species (14 d posthatch) (Shen and Leatherland 1978a) revealed small numbers of CCs in both tissues. The latter study also exposed embryos and larvae to dilute SW (13%oS) but did not observe any of the fine structural changes typical of CCs in fish adapted to SW. In contrast, recent results from Kaneko et al. (1995) for prehatch chum salmon embryos, and this study for chum and coho embryos and larvae, showed numerous cutaneous CCs using DASPEI staining. Some of the typical structural features associated with SW and F W acclimation were also observed using T E M examination of the branchial epithelium of chum salmon fry (Uchida et al. 1996) and prehatch embryos (this study). These cellular changes were less common in the CCs of coho acclimated to SW in this study. As speculated earlier, since the limited SW tolerance of coho does not appear to be related to the total number of cutaneous CCs, it may instead be related to their inability to modify C C fine structure. Presumably structural changes of the CCs would be necessary in response to the increased need for ion extrusion associated with SW exposure, in order to accomplish effective hypo-osmoregulation. If coho are not able to modify their CCs for hypo-osmoregulation then it suggests that these CCs are functioning largely in ion uptake and acid-base balance. This leads to the consideration of whether the early life stages of coho salmon are better suited to a life in an ion-poor environment (e.g., F W environments with low total  94  dissolved solids) as opposed to an ion-rich one (brackish water environments with moderate to high total dissolved solids), and would excel in such a situation where chum salmon may not. The reasons for the differences between Leatherland and Lin (1975) and this study with respect to absence or presence of cutaneous CCs are unclear. However, it may be noteworthy that the fish used by the earlier study were coho salmon that had been transplanted to Lake Erie, Ontario, Canada, from the west coast of North America. The precise history of these fish is unknown, except that they were captured in 1973 and historical records of these transplants suggest that that they would have resided as landlocked populations since the mid- to late-1960s (Sandercock 1991). Thus they would not have undergone any anadromous migrations for 2 to 4 generations. It seems unlikely that this short period of time would produce significant genetic adaptations in a physiological process as fundamental as osmoregulation. However, even if they did not occur, it is possible that differences between the founder stocks of Leatherland and L i n (1975) and this study may be a contributing factor in the differences observed between these two studies. Environmental conditions specific to the original habitats of founder stocks presumably would result in specific selective pressures. One potentially important environmental characteristic would be water quality, including ionic content, of the original source location compared to that of the transplant location. The coho used in this experiment originated from a river with relatively low conductivity and total hardness (see Methods section, this Chapter for details) which is consistent with the extensive CC distributions observed in this species and past studies that have correlated high numbers of CC with low ion content water (Laurent and Perry 1991). Unfortunately, the original source of the founder stock used by Leatherland and L i n (1975) was not stated so these considerations cannot be explored further. How potential environmental differences such as these relate to intraspecific density and function of CCs in F W as ion uptake cells, in both early and later life stages, remains to be examined.  95  A commonly documented change in response to SW exposure of embryos and larvae is a 1.5 to 2 times increase in C C surface area and in general, no change in CC density (Ayson et al. 1994, Kaneko et al. 1995, Uchida et al. 1996). This contrasts with other studies that have reported density increases in extrabranchial epithelia of adult fish as a result of S W exposure (Richman et al. 1987, McCormick 1995). Although no quantitative examination of C C surface area was conducted in this study, no obvious changes in CC surface area were noticed. A relatively large range of C C size was observed in different tissues throughout the survey of eyed and prehatch embryos and yolk sac larvae in this study, yet no consistent changes in cell surface area or size appeared to be related to the degree of SW exposure. It was noted that CCs located in certain tissues were more likely to have larger surface areas (wider fluorescent profiles, e.g., OPE; and OPE ) than other tissues (e.g., TRE). Why no consistent changes in C C surface area e  were observed remains unclear - perhaps there simply were none. As noted earlier however, past C C studies conducted on similar or identical species have produced contradictory results. For example, with coho salmon embryos Leatherland and Lin (1975) did not find any CCs whereas this study found numerous ones; also for Atlantic herring, Jones et al. (1966) did not find any CCs in Y S E whereas Wales and Tytler (1996) did. There is no doubt that fluorescent techniques have greatly simplified the process of locating CCs but both of the previous studies that obtained negative results used rather extensive sampling protocols. The negative results in the previous studies are especially surprising given the high densities of CCs observed in both this study and that of Wales and Tytler (1996). It is possible that intraspecific differences are more significant than is currently understood or appreciated. In general, for a given early life stage the density of cutaneous CCs does not appear to change in response to SW exposure. This generalization held true in this study for all of the treatments except the chum yolk sac larvae. Density changes do however, occur as part of the 96  developmental process. Once cutaneous CCs have developed in the embryo, there appears to be a general decrease in CC density as development proceeds from the embryonic and larval forms towards the juvenile form (Hwang 1989, this study). Although only a few studies have followed these measurements throughout early ontogeny, it appears that the cutaneous CCs decrease as the branchial CCs increase (Katsura and Hamada 1986, Hwang 1990, Ayson et al. 1994). Ayson et al. (1994) showed an increase in tilapia Y S E from about 200 CC-mm" prior to epiboly (2 2  days postfertilization) to a density of about 400 to 450 C C - m m for an 8-d period, spanning 2 d -2  prehatch to 6 d posthatch, after which the density dropped back down to about 200 CC-mm" by 2  8 d posthatch. Hwang (1989) also reported a decrease in cutaneous C C density after hatch in larval ayu but noted that the decrease occurred more slowly in SW-acclimated larvae than FWacclimated ones. Whereas no obvious differences between FW- and SW-acclimated chum and coho salmon embryos or larvae were observed in this study, there was a trend towards reduced CC density during the later developmental stages compared with the embryonic ones. This pattern was most evident in the Y S E tissue and appeared to be more pronounced in chum salmon larvae than coho. The density changes in salmon however, were not as pronounced as those of tilapia larvae. This may be because Ayson et al. (1994) monitored the densities over a broader range of developmental stages. Branchial CCs were obvious at all three developmental stages in this study and showed obvious signs of increased differentiation as development progressed. A number of other studies have reported a general pattern of decreasing cutaneous CCs as the number of branchial CCs increases as the larva develops and eventually metamorphoses into a juvenile (Shen and Leatherland 1978a, Katsura and Hamada 1986, Hwang 1989). Unfortunately, I was not able to quantify the density changes accurately in the branchial epithelium and therefore was unable to  97  make direct comparisons between cutaneous and branchial C C densities. Even though gill structure in eyed chum and coho embryos was simple, with only small filament buds protruding from the gill arch, there were already numerous CCs on the arch and filament epithelia. There also appeared to be a close association of some of the CCs with blood vessels in the arch (Fig. 2.14). These vascular associations were not apparent in the prehatch embryo and yolk sac larva stages but nonetheless may be evidence that these early CCs are functional for osmoregulatory control. The higher degree of cellular organization of the CCs on the filaments of prehatch embryos and yolk sac larva suggested these CCs were aligned along the sites where the lamellae would develop at the later stages. It is unclear whether the lamellae had already developed in the larval stage since the DASPEI stain only targets cells containing high number of mitochondria. Shen and Leatherland (1978a) showed that two weeks after hatching, rainbow trout larvae had developed rudimentary lamellae and it appears from a few of their figures that some of the branchial were CCs aligned in a similar pattern to that observed in this study. This pattern, CCs located on the filament near the base of the lamellae, has been reported in juvenile and adult fish, including chum salmon (Uchida and Kaneko 1996, Uchida et al. 1996, Uchida et al. 1997). Filamental CCs are clearly lined up along the base of the lamellae in light micrographs of sagittal sections of gills in both Uchida and Kaneko (1996) and Uchida et al. (1997). Chum salmon transferred to F W from SW or acclimated for 2 weeks or more in SW, showed high numbers of filamental CCs and greatly reduced numbers of lamellar CCs.  Relative numbers of cutaneous compared with branchial CCs As mentioned above, cutaneous CCs appear to provide the developing teleost embryo and larva with an interim osmoregulatory system for the period between gastrulation and completion of organogenesis, by which time the gills have developed and are functional for both 98  osmoregulatory and respiratory purposes. Although the ontogeny of the relative contributions of cutaneous versus branchial surfaces has not been examined for early life osmoregulation, it has been investigated for early life respiration. Rombough and Ure (1991) measured the partitioning of oxygen uptake between cutaneous and branchial surfaces of chinook larvae and fry. They recognized that there should be a transition from the cutaneous mode of respiration, typical of embryos and larvae, to the predominantly branchial mode, typical of juveniles and adults. Initially cutaneous gas exchange was high, up to 84% for newly hatched chinook larvae (-100 mg), however, it dropped rapidly to about 40% by the time of first feeding (-400 mg fry). Changes in tissue-specific oxygen uptake of the gill were not only the result of direct changes in surface areas available for exchange, but also the result of increased oxygen uptake efficiency in the gills (measured as area-specific oxygen uptake). Based on partitioning and efficiency measurements such as these, Rombough and Moroz (1997) estimated an approximate threshold (body mass) at which the fry should make the switch from cutaneous to branchial respiration, since presumably the cutaneous exchange would become limiting at some point in light of the growing fish's increased total oxygen requirements. Their calculations indicated that larvae and fry appeared to have an excess of exchange surface area in the gills prior to this transition. From this Rombough and Moroz (1997) suggested that during these early stages of development the gills might be functioning primarily as an ionoregulatory organ rather than a respiratory one. Given the large changes in cutaneous and branchial surface areas and the relatively high densities of CCs on both these surfaces it is appealing to examine a similar model with regards to the process of osmoregulation. How do the relative numbers of cutaneous CCs compare to those of the branchial CCs during the early life stages of salmonids, and are the ontogenic changes in relative CC numbers consistent with the hypothesis that the gills (branchial CCs) assume a greater osmoregulatory role as early development progresses? 99  In order to test this model, cutaneous and branchial surface areas were estimated using chum salmon as an example. Surface area estimates were calculated for each of the three developmental stages examined in the CC survey (Table 2.3) using mass-dependent models supplied by Rombough and Moroz (1990). Extrapolations were made to estimate comparable values for the fry stage (100% yolk sac absorption). Estimates of 500, 1000, and 1500 C C - m m  -2  were used for the gill C C density estimates for embryo, yolk sac larva, and fry stages, respectively, based on personal observation and the expectation that C C densities would be similar to values reported for juvenile and adult salmonids (Perry and Laurent 1989, Perry et al. 1992, Goss et al. 1994, Uchida et al. 1996, Uchida et al. 1997). Ratios of cutaneous to branchial surface areas and CC numbers were used to estimate the relative contributions or capacities for osmoregulatory function. Similar to Rombough and Moroz's (1990) findings, the skin comprised a much larger surface area compared to the gill during the embryonic stages of development. The surface area ratios of these two tissues showed that the cutaneous surfaces of eyed and prehatch chum embryos were 13.2 and 7.0 times greater than the branchial ones. This is not surprising given the physical appearance of these tissues, i.e., the lack of extensive gill filament (Fig. 2.14, Fig. 2.15; Kaneko et al. 1995) and the relatively large proportion of total cutaneous surface contributed by the Y S E (Fig. 2.1). Although surface area ratios continued to decrease as development progressed (Fig. 2.18), cutaneous areas of chum salmon fry was estimated to still constitute 2.4 times more area than the gill area (Table 2.3). Compared to adult fish it appears that those fry will continue to develop the gills at a faster rate than the cutaneous surfaces since adults typically have ratios of 0.05 to 0.5 (Gray 1954). In general, the results presented here agree with those of Rombough and Moroz (1990). Although some differences in absolute values  100  were noted, they may well be the result of interspecific variation since chinook salmon embryos and larvae tend to be larger than chum. High cutaneous C C densities and extensive distributions of CCs over most parts of the body resulted in high total numbers of cutaneous CCs per individual, with a range from 114,500 CC-indiv. for eyed embryos to 212,900 CC-indiv. for yolk sac larvae. Ratios of cutaneous to -1  -1  branchial CCs indicated similar relationships to the surface areas (Fig. 2.18); cutaneous CCs were 12.3 and 3.6 times more numerous than branchial CCs, for eyed and prehatch embryos, respectively (Table 2.3). Chloride cell ratios of larvae and fry decreased more rapidly than the surface area ratios since both the CC densities and the surface areas were changing. As the gills differentiated into more complex structures with greatly increased surface areas and C C densities, one of the largest cutaneous surface areas, the Y S E , decreased rapidly as part of the yolk sac absorption process, while the average CC densities dropped slightly. The C C ratio of 0.7 for fry is the result of disproportionate changes in surface areas and particularly C C densities, since it is markedly lower than the surface area ratio of 2.4 at this same stage. This fry stage appears to mark a transition in development where for the first time the total number of cutaneous CCs has decreased below that of the branchial CCs, even though the total number of cutaneous CCs has not changed significantly (Fig. 2.18). This ontogenic trend supports the model that cutaneous CCs initially provide most of the ion exchange surfaces, but that as the gills continue to develop and become functional, the cutaneous contribution to ion exchange becomes progressively less significant. Additional support for this model stems from a recent speculation by Rombough and Moroz (1997) that larval gill development does not appear to be fully co-ordinated with the apparent specific respiratory needs of larval teleost fishes. They observed that for a variety of larval fishes, the transition from cutaneous to branchial respiration occurred at a relatively narrow size 101  Table 2.3. Surface area estimates, mean whole-body C C density, and total C C number per individual for cutaneous and branchial surfaces of chum salmon embryos, larvae, and fry (see text for details). SPECIES  Tissue surface area  CC density  C C number  (mm indiv. )  (CCmtrf )  (CCindiv. )  2  Dev. stage (sal. range, %<>S)  Tissue Wet Wt. (9)  Total Cutan.  _1  Total 3  Branc.  2  -1  Ratio b  Total  (Cut:Branc)  Cutan.  c  Branc.  d  Cutan.  Total  Ratio  Branc.  (Cut: Bran  CHUM  Eyed embryo 0-15  0.014  249.1  18.8  13.2  459  500  114,500  9,400  12.2  P H embryo 0-18  0.052  349.8  50.1  7.0  516  1000  180,400  50,100  3.6  Y S larva 0-6  0.162  471.2  118.4  4.0  450  1000  212,900  118,400  1.8  Fry (extrapolated)  0.267  408.9  172.0  2.4  443  1500  181,400  258,000  0.7  d  e  a  Cutan. = cutaneous.  b  Branc. = branchial.  Cutaneous C C density: a weighted average of the C C densities of all major cutaneous surfaces, see text and Table 2.1.  0  Branchial C C density: estimated from personal observations and literature values for larvae, juveniles, and adults ,see text. d  e  f  PRH= prehatch. YS = yolk sac; 0 to 6 %oS values were used to represent this stage, see Table 2.1.  102  range 100 to 500 mg. However, in a number of the species examined this result appeared to be asynchronous with the ontogeny of gill development, since in some species this occurred at much smaller sizes (e.g., 2.5 mg for walleye (Stizostedion vitreum), 1.6 mg for carp, 1.2 mg for plaice). Based on this apparent disparity the authors suggested that the seemingly premature development of gills for respiratory purposes was perhaps indicative of a role other than respiration, namely ionoregulation. This suggestion coincides well with the observation that the cutaneous surface area of a chum salmon fry (267 mg) represents 2.4 times that of the branchial area, whereas the total number of cutaneous CCs represents only 0.7 times that of the total branchial CCs (Table 2.3). Thus at the fry stage of development, the relative importance of the gills as an ionoregulatory structure may be close to 3 times (2.4/0.7) more significant than as a respiratory structure. However, it is important to consider that this modelled estimate is probably an overly simplistic one since it does not take into account any differences in areaspecific efficiencies of these tissues. Shortcomings of the current model include assuming that all of the observed CCs, cutaneous and branchial, are fully functional, and not accounting for the possibility that branchial CCs may be more efficient in ionoregulation due to better ventilation and blood perfusion compared to the cutaneous ones. Rombough and Ure (1991) noted such a difference in efficiency with oxygen uptake rates in cutaneous compared with branchial epithelia of chinook larvae and fry. They reported that although these epithelia were about equally efficient at extracting oxygen in a newly hatched larvae (0.1 g), gills were about 6 times more efficient by the time the fish had reached the fry stage (1 g). There is insufficient evidence to comment on the differences in osmoregulatory efficiency of specific epithelia in this study (e.g., FFE, OPE, TRE, and YSE) but undoubtedly such differences would alter the proposed model significantly.  103  Ventilation of cutaneous surfaces Theoretically the cutaneous osmoregulation would require good ventilation and perfusion of the skin to minimize boundary layers for the same reasons that these processes are important in branchial and cutaneous gas exchange. Water flow over gills and perfusion of tissues with blood are critical factors in ensuring that the oxygen replenishment and carbon dioxide removal occur effectively. The flow of water over gill surfaces ensures that diffusive boundary layers limiting gas exchange are minimized. This would apply both to respiratory as well as ion exchange surfaces, and thus movement of water in respiration and osmoregulation systems of adult fishes is critical to ensure proper function. The embryo is no exception, especially when it is contained within the egg or ovary. Rombough (1988a) noted that stirring of the perivitelline fluid was necessary to meet the metabolic demands of the embryo. B y stimulating fluid movement diffusive boundary layers are minimized, diffusion rates are maximized, and net oxygen transport is increased. This is especially true of more developmentally advanced and larger embryos; including Pacific salmon since they are among the largest of the teleost embryos (Heming and Buddington 1988). Movement of the perivitelline fluid occurs as early as Stage 17 (Jensen 1988, heart beating, but head still attached to yolk sac; 160 ATUs, chum salmon) when the trunk is flexed into a semi-circle and twitched periodically. More regular movements begin occurring once the pectoral fins (without fins rays) have developed at Stage 21 (eyed embryo; 230 ATUs, chum salmon) and are twitched along with flexion of the trunk, which begins at Stage 23 (operculum covers all branchial arches, beginning of hatching; 420 ATUs, chum salmon). Dye studies in Atlantic salmon have shown that this sort of embryonic activity (e.g., trunk movements of 60 to 200 flexures-h" ) results in effective stirring of the perivitelline fluid 1  (Peterson and Martin-Robichaud 1983). These movements appear to fulfil a respiratory function during the later embryonic stages since accidental exposure to hypoxic conditions resulted in an 104  increased frequency of trunk movements. Fundamental changes in body movement also occur as development proceeds. Peterson and Martin-Robichaud (1983) reported a decrease in trunk movements by 350 to 400 ATUs (e.g., 1 every 2 to 4 h) followed by an increase in pectoral fin movements (40 to 150-min" ). Rombough (1988a) reviewed the use of pectoral fins movements 1  offish larvae in relation to cutaneous respiration and provided numerous examples of larval fish using this strategy to improve respiration efficiency, including evidence for a countercurrent system in larval air-breathing fish (Monopterus albus, Liem 1981). Pectoral fin movements in Atlantic salmon larvae did not appear to produce marked water movement over the body surface but rather appeared to function in drawing water over the gills (Peterson 1975). It is clear that these movements play a significant role in early life respiration of embryos and larvae. It seems probable that many of the same water movement requirements that are necessary for early life cutaneous respiration would also be beneficial, if not required, for early life osmoregulation, since both of these basic physiological processes are highly dependent on the maintenance of diffusion gradients. Water movement induced by trunk and pectoral fin flexures in order to ventilate cutaneous respiratory surfaces, may well fulfil a similar function for the ionoregulatory surfaces of embryos and larvae. Although CCs in embryos and larvae may occur on many surfaces of the body, they are most commonly reported in specific areas; areas that appear to be similar in a wide range of species. In this study, the highest concentrations of CCs were found around the base of the pectoral fin and along the dorsal half of the yolk sac (Fig. 2.1). In the few other species where whole-body distributions have been established (ayu, Hwang 1989; Atlantic herring, Wales and Tytler 1996) it seems clear that C C densities are generally highest on the trunk and dorsal portions of the yolk sac, especially around the pectoral fin and pericardial regions. In contrast, C C densities are usually comparatively low on the ventral surfaces of the Y S E , and in the case of herring and ayu also on the posterior region of the 105  trunk. These interspecific similarities in CC distribution suggest that cutaneous CCs are localized in areas that are well ventilated by fin and body movements to minimize boundary layer gradients, and thus optimize the effectiveness of cutaneous CCs in their role as osmoregulatory cells.  Allometric considerations Qualitative comparisons of the cutaneous CC densities and total larval size of other species with my observations indicate that in all three cases, the markedly smaller larvae of tilapia, ayu, and herring all have drastically fewer total cutaneous CCs than chum and coho salmon larvae. If cutaneous CCs do in fact provide much of the osmoregulatory function for developing embryos and larvae then it seems probable that there would be an allometric relationship between the total number of cutaneous CCs per individual and the volume or mass of that individual. To the best of my knowledge there are no other studies that have reported C C density estimates for all of the major cutaneous surfaces of the body or quantified the total number of CCs per individual. However, there are a few studies that have provided either C C densities (Tilapia, Ayson et al. 1994) or illustrations of cutaneous CCs in embryos and larvae ayu, (Hwang 1989; Atlantic herring, Wales and Tytler 1996). Wales and Tytler (1996) reported on how the total number of CCs did not change markedly between early (1 d posthatch) and later larvae (11 to 13 d posthatch). Based on preliminary observations, they suggested that this lack of change in numbers appeared to be at least partially the result of the migration of CCs from the yolk sac up to the trunk during this phase of yolk sac absorption. They speculated that the maintenance of more or less constant numbers of CCs was related to the relatively constant osmoregulatory needs of the larvae throughout that developmental period. It is noteworthy that in this study, the total number of cutaneous CCs did not change drastically either between the 3 developmental  106  stages sampled (Table 2.3), even though the examination period spanned about 35 d. Without additional quantitative data it is difficult to discuss this further, but in general, it appears that not so surprisingly, smaller larvae have fewer total cutaneous CCs than larger larvae, not only because their total surface areas are smaller but also because their C C densities are markedly lower. Yet in relative terms, a small larvae such as a herring appears to have a higher proportion of cutaneous area covered by CCs than a large larva such as a salmon. This observation is consistent with surface area to volume relationships for diffusive processes in animals of different sizes; given similar epithelial permeabilities, the larger surface area to volume ratio of a small larvae (e.g., herring) would presumably result in greater diffusive losses and gains than in a larger larvae (e.g., salmon). In addition to differences in total C C numbers, there are indications that there is also interspecific variation in C C size may also warrant consideration. On examination of illustrations and photographs of herring larvae presented by Wales and Tytler (1996) it appeared that those CCs that were ostensibly larger than salmon CCs from this study (10 to 11 um). A summary of cutaneous and branchial C C sizes from the literature (Table 2.4) indicates that in general, C C sizes are relatively similar including increases in surface area as a result of SW exposure. The majority of CCs, including both F W and SW, ranged from 6 to 17 urn with a few unusually large ones, 23, 26, and 31 um, found in SW acclimated larvae of tilapia, ayu, and herring, respectively (Table 2.4). The largest CCs (31 um) were in fact observed in Wales and Tytler's (1996) study using Atlantic herring larvae. These summarized C C sizes provide an example.of the possible interspecific variation in CC sizes and demonstrates the added value of providing measures of total cutaneous CCs in relation to not only the individual but also the whole-animal volume or mass. Relating CC numbers in such a way will allow more direct comparisons between species and life stages.  107  Table 2.4. Estimates of chloride cell diameter during the early life stages of fish as measured from published studies in the literature. Species  Tissueacclim. b  CC Diameter (um)  YSE-FW  9.1  YSE-SW  13.5  YSE-FW YSE-SW  10.5 11.2  Fluor.  fry  gill-FW  7.4  Light  —  gill-SW  10.6  fry  gill-FW gill-SW  5.6 7.5  Graph  gill-FW gill-SW  12.8 -8.9  Light  gill-not specified  7.2  Light  YS larva  YSE-FW  13.6  Fluor.  Ayson et al. 1994  larva  gill-FW gill-SW  9.4 22.7  Fluor.  Li etal. 1995  YS larva (newly hatch)  OPE-FW  12.9  Light  Hwang 1989  YS larva (1 dPOH) larva (40 dPOH) larva (40 dPOH)  TRE-FW  11.7  Table  Hwang 1990  TRE-FW  17.2  —  —  TRE-SW  26.2  —  —  YS larva (2 dPOH)  TRE-SW  31.3  Life stage  chum salmon (Oncorhynchus  a  eyed embryo  Data source d  Reference  Fluor.  Kaneko etal. 1995  keta)  —  —  chum salmon  PRHembryo  —  —  chum salmon —  chum salmon —  —  chum salmon  juven.  —  —  rainbow trout (Oncorhynchus  c  PRHembryo mykiss)  Mozambique tilapia  —  —  —  —  —  —  This study —  Uchida and Kaneko 1996 —  Uchida etal. 1996 —  Uchida etal. 1997 —  Shen and Leatherland 1978  a  (Oreochromis mossambicus)  Mozambique tilapia —  —  ayu (Plecoglossus  altivelis)  ayu — —  Atlantic herring (Clupea  harengus)  —  Fluor.  —  Wales and Tytler 1996  PRtt= prehatch; YS = yolk sac; dPOH = days posthatch b Tissue type-acclimation environment: OPE = opercular epithelium; TRE = trunk epithelium; YSE = yolk sac epithelium; FW = freshwater; SW = saltwater. Diameter was estimated by measuring the narrow axis of CCs from published photomicrographs, tables and graphs; mean diameters were calculated from up to 5 CCs where possible. d Fluor. = fluorescent microscopy photomicrograph; Light = regular light microscopy photomicrograph; Graph = data graph; Table = data table. a  c  108  Variations in Staining effectiveness and C C visibility Changes in CC fine structure and position within the embryonic and larval epithelia may have contributed to observed differences in DASPEI staining effectiveness that are related to specific environmental conditions (e.g., F W and SW exposure) or ontogenic changes. Although faint cells were observed at all three developmental stages they became progressively more prevalent in the prehatch embryo and yolk sac larva stages. In certain tissues CCs were obviously located in the most external layer of the epithelium (e.g., Fig. 2.9), however, such observations were not made for all tissues at each stage. One possible explanation for faint cells might be the withdrawal or covering of these cells as development progresses. If CCs located deeper in the epithelium do not absorb DASPEI as readily as CCs with exposed apical surfaces, then this may also explain why CCs located in the TRE, and to a lesser extent the H D E and OPE, did not always readily stain with DASPEI. This was most apparent in TRE tissue samples where CCs exhibited the brightest fluorescence when they were near a cut through the skin (Fig. 2.6). Goss and Perry (1994) discussed the location of CCs in rainbow trout and American eel (Anguilla rostratd) gill epithelium and suggested that CCs lying under the surface of the gill and thus not in contact with the water, were non-functional. A number of studies have reported differences or changes in C C location within cutaneous epithelia in response to environmental changes (Richman et al. 1987, Herndon et al. 1991). Wales and Tytler (1996) reported that cutaneous CCs in the TRE of Atlantic herring larvae became progressively fainter as development proceeded from the newly hatched larva into the later larval stages and suggested that one of the contributing factors may have been that these cells were located deeper in the epithelium. One of the common characteristics of FW-adapted CCs is their limited or absence of contact with both the serosal and apical sides of the epithelium (McCormick 1995). Observations using 109  S E M in chum eyed embryos showed the presence of CCs located in between epithelial cells of Y S E (Kaneko et al. 1995) that were similar to arrangements of pavement cells and CCs in adult gill filamental tissue (Perry and Laurent 1989, Goss and Perry 1994). As mentioned above, changes in epithelial location or degree of apical surface exposed may influence the ability for CCs to pick up DASPEI stain and therefore result in faint or invisible CCs under these circumstances. Depth of the cells may also be functionally important (Goss et al. 1995). Removal of CCs as secretory cells may not necessarily require actual relocation but may simply involve overlapping of the CCs by adjacent pavements cells. In juvenile and adult fish pavement cells appear to cover and uncover CCs in response to changes in ambient environmental conditions ( C 0 , pH, ions) (Goss et al. 1995). Depeche (1973) reported this process in Y S E of 2  the guppy in response to osmotic stress. A similar mechanism was suggested to account for the low incidence of C C openings on the Y S E of larval Atlantic herring (Somasundaram 1985) in response to zinc exposure. Overlapping of CCs by pavement cells may also have been observed in branchial epithelium of 2 wk posthatch rainbow trout (Shen and Leatherland 1978a). Although manipulation of epithelial cell structure (e.g., covering and uncovering of CCs) has been observed in a variety of species offish the majority of work has focussed on the branchial epithelium of juvenile and adult fishes. Even so, the mechanisms involved in controlling short term changes in epithelial structure remain largely unresolved (Goss et al. 1995). This paucity of information is even more apparent for the early life stages of fishes. A more detailed examination is required to determine the specific characteristics of cutaneous and branchial C C location and reaction of embryos and larvae to various environmental conditions. Such examinations would allow comparison to related juvenile and adult mechanisms. It may also provide much needed information on how location within the epithelium affects or dictates function (e.g., bright cells versus faint cells).  110  Increased numbers of faint CCs were not the only changes seen in C C fine structure in chum and coho embryos and larvae. In a number of instances larval epithelia had CCs that appeared to be amorphous and irregular in shape (Fig. 2.10). Without detailed electron microscopic examination it is difficult to state with certainty but it appeared as if these cells were physically breaking apart; instead of a single mass of fluorescence, they appeared as separate blebs of fluorescence. Seemingly similar observations have been reported for juvenile and adult fish as well, where CCs appeared to be in the process of degeneration (Richman et al. 1987) or apoptosis (programmed cell death) (Wendelaar-Bonga and van der Meij 1989). In summary, the presence of cutaneous CCs supports the growing body of evidence indicating that early life osmoregulation occurs in these extrabranchial tissues during the early phase of development when the gills are not yet fully differentiated. The early life stages of chum and coho salmon were unusual compared to other fishes investigated to date, in that they had very high densities of CCs on almost all major cutaneous surfaces from the eyed embryo stage through to 50% yolk sac absorption (Table 2.1 and Table 2.2). These densities were as high or higher than other reports in the literature; estimates of total numbers of cutaneous CCs per individual ranged from 100,000 to more than 200,000 (Table 2.3). Density was not noticeably affected by SW acclimation, unlike in a number of previous studies. Interspecific comparisons of C C distribution and numbers between the more SW tolerant chum salmon and less tolerant coho salmon showed no obvious differences that would correlate with their markedly different responses to increased osmotic and ionic concentrations. I speculated that the limited SW tolerance of coho embryos and larvae is due to an inability to modify C C structure for ion extrusion required for effective hypo-osmoregulation. Early life osmoregulation appears to share a number of functional similarities with cutaneous respiration. Both processes require ventilation and perfusion of the cutaneous surfaces 111  to minimize boundary layers and maintain diffusion gradients in order to allow for effective passive or active cutaneous transport of molecules and ions. Speculation about relative temporal and allometric changes in cutaneous and branchial surface areas, and their respective numbers of CCs, appears to support a recent suggestion in the literature that gills fulfil a major ionoregulatory role before a respiratory one. Initially, during the embryonic period and early larval stages, cutaneous CCs are much more numerous than branchial ones and probably provide the developing salmon with the majority of its ionoregulatory surfaces and functionality. However, as the larva nears complete yolk sac absorption or emergence, the combination of rapidly increasing branchial surface areas and increasing branchial C C densities, contrasted against more slowly increasing cutaneous surface areas and decreasing cutaneous C C densities, marks a transition where more potential sites for ion exchange occur in the gills than on the skin (Fig. 2.18). Yet, at this same developmental stage, the ratio between cutaneous and branchial surface areas indicates that cutaneous surfaces provide about 2.4 times that of the gills (Table 2.4). Other studies indicate that these late larvae or early fry still derive a significant proportion of their respiratory function from cutaneous surfaces. These comparisons support the speculation that gills in fish embryos and larvae potentially function as a dominant ionoregulatory organ before a respiratory one. In young larvae, gills appear to be relatively more important as an ion regulatory structure than a respiratory one.  112  Figures - Chapter 2  Fig. 2.1. A generalized diagram of a salmon yolk sac embryo showing the various cutaneous tissues sampled for CC distribution and density in chum and coho embryos and larvae using DASPEI fluorescent microscopy: yolk sac (YSE), head (HDE), opercula [internal (OPEj) and external (OPE ) surfaces], main body or trunk (TRE) and finfolds/fms (FFE/FNE), and gills. Note that Y S E was larger in prehatch coho embryos (area delineated by solid and dashed line) than in chum embryos (area delineated by solid line only) and extended further in a posterior direction to a point about equidistant between the pectoral fin and the vent. Pectoral fin in the embryo stages was counted as finfold tissue (FFE) whereas at the larval stage it had differentiated into fin tissue proper (FNE). e  d  1  113  (a) Fig. 2.6  ~ 3 mm  Fig. 2.2. Reference diagrams showing specific locations of DASPEI stained chloride cell photographs; (a) eyed and prehatch embryos of chum and coho salmon: Fig. 2.3, Fig. 2.4, Fig. 2.6, Fig. 2.7, Fig. 2.8, Fig. 2.9; (b) yolk sac larvae of chum and coho salmon: Fig. 2.5, Fig. 2.10 and Fig. 2.13. Large and small photographic areas represent magnifications of lOOx and 400x, respectively. Scale is only approximate.  114  Fig. 2.3. Y S E tissue of an eyed coho embryo in 5 %oS showing change in C C density from Y S E (left side, dorsal) to Y S E zone (right side, ventral). D A S P E I ; bar = 200 urn. m  v  Fig. 2.4. Y S E of a chum prehatch embryo in 0 %cS showing well rounded and relatively bright CCs with discrete edges. D A S P E I ; bar = 50 um. 115  Fig. 2.5. Y S E tissue of a coho yolk sac larva in 6 %oS showing the amorphous and irregular shaped and faint CCs. D A S P E I ; bar = 50 um.  Fig. 2.6. T R E tissue showing numerous, relatively evenly distributed CCs on the main trunk the lower neck region of an eyed coho embryo in 10 %cS. Note that D A S P E I penetration does not extend much beyond the cut edge. D A S P E I ; bar = 200 um.  Fig. 2.7. External OPE tissue of a prehatch chum embryo in 0 %oS showing high density of evenly spaced CCs. DASPEI; bar = 200 urn.  Fig. 2.8. HDE tissue located directly above the eye (arrow) in an eyed coho embryo in 15 %cS showing evenly spaced, bright CCs DASPEI; bar = 200 um. 117  Fig. 2.9. Internal OPE tissue from a prehatch chum embryo in 18 %cS showing specific location of CCs in the external most layer of the integument. DASPEI; bar = 50 um  Fig. 2.10. T R E tissue of coho yolk sac larva; similar C C densities as seen in earlier stages but increased variation in C C size compare with Fig. 2.6. Inset: higher magnification of same tissue but different specimen, Y S larva in 0 %oS; note that increased size appears to be the resultof groupings of CCs. D A S P E I ; bars = 200 urn and 50 um, respectively. 118  Fig. 2.11. Chloride cell located in the gill epithelium of a prehatch chum embryo incubated in FW. Note the numerous dark mitochondria (arrows), evaginated apical surface (*), and lack of extensive basolateral tubule network. T E M ; bar = 2 um.  Fig. 2.12. Chloride cell located in the gill epithelium of a prehatch chum embryo acclimated (7 d) to 24 %oS SW. Note the numerous darker mitochondria (arrows), invaginated apical crypt (*), and moderately convoluted basolateral tubule network (arrowhead). T E M ; bar = 2 um. 119  Fig. 2.13. Internal OPE tissue of a coho yolk larva in 6 %cS showing irregularly shaped and variably sized CCs commonly found in this development stage but in not earlier embryonic stages. DASPEI; bar = 40 um.  Fig. 2.14. Dorsal end of gill arch from newly eyed chum embryo in 5 %cS showing what appears to be relatively close association of CCs with rudimentary gill blood circulation, as revealed by background fluorescence. Inset: note that filament buds on the middle section of arch have begun to differentiate. DASPEI; bar = 50 um, inset bar = 200 um 120  r  *V  >*  i  Fig. 2.15. Gill arch from well-eyed chum embryo in 10 %oS: numerous CCs are located along the developing filaments (arrows) as well as the dorsal end of the gill arch (arrowhead); note variable staining pattern indicative of limited D A S P E I penetration (distal to right arrow). Inset: higher magnification showing detail of CCs along the filaments; note that secondary lamellae are not yet developed. D A S P E I ; bar = 2 0 0 um, inset bar = 50 um.  Fig. 2.16. Gill filaments of a prehatch chum embryo in 12%oS showing CCs grouped (arrowheads) at the base of presumptive secondary lamellae. D A S P E I ; bar = 50 um. 121  *  * mm  0  Fig. 2.17. Chloride cells on the gill filaments of yolk sac larvae; top view of filament showing CCs (arrows) lined up along either side of presumptive secondary lamellae (*). (a) Chum salmon in 12 %cS and (b) coho in 0 %cS; note variable lamellar staining with most effective DASPEI penetration occurring at the ends of the rows, (c) Sagittal view of same tissue showing a different perspective of the CCs at base of presumptive secondary lamellae; coho in 0 %cS. DASPEI; bar = 50 um, same scale and symbols for all three panels. 122  Development (ATUs) Fig. 2.18. Total numbers of chloride cells (CC) per individual in cutaneous and branchial ' epithelia, and the relative relationships of total cutaneous and branchial surface areas and total C C numbers during the early life stages of chum salmon. The curves are modelled from data (CC density) and calculations (surface areas, Rombough and Moroz 1990) at 4 stages during development (measured in accumulated thermal units, ATUs); eyed embryo, prehatch (PH) embryo, yolk sac (YS) larva and fry (100% yolk sac absorption). ^ Branchial C C numbers are based on qualitative observation and CC densities obtained from published studies for juvenile and adult salmonids (see text). 1  123  Chapter 3 N A , K - A T P A S E A N D H - A T P A S E ACTIVITY IN GILL, SKIN, A N D Y O L K +  +  +  S A C E P I T H E L I U M OF C H U M A N D C O H O S A L M O N E M B R Y O S A N D L A R V A E IN R E L A T I O N TO O S M O R E G U L A T I O N  Introduction Branchial CCs serve a central role in providing the required salt extrusion mechanism for SW resident fishes and also appear to be integrally involved in ion uptake in FW resident fishes (McCormick 1995). A number of studies have demonstrated the importance of extrabranchial sites for ion regulation in a variety of species such as the opercular epithelium of tilapia and killifish, and the jaw skin of the mudsucker (Gillichthys mirabilis), and in all cases CCs within these epithelia are integrally involved in the process of ion regulation (Marshall 1995). Ion excretion and uptake by epithelial CCs was reviewed in the General Introduction; a brief recapitulation follows. In SW fish, the current model for ion extrusion in the C C (Fig. GI 2) is ultimately driven by the basolaterally located Na ,K -ATPase that uses ATP to extrude N a into +  +  +  the paracellular spaces from which it diffuses out across cation-selective leaky junctions. N a re+  enters the cell at the basolateral membrane of the C C down its electrochemical gradient and by doing so, drives Na -dependent C F cotransport into the CC. Ultimately C F diffuses out of the +  cell down its electrochemical gradient via apically located C F channels (Marshall 1995). In FW, the situation is less clear but ion exchange and acid-base regulation appear to be integrally linked (Marshall 1995) and most likely include the involvement of other cell types such as the pavement cell at least, working in concert with the CC. The current FW model  124  (Fig. GI 2) appears to be based on the excretion of protons by an apically localized H -ATPase, +  which consumes ATP and drives the uptake of Na and possibly Ca  via ion conductive  channels in the apical membrane of CCs and pavements cells (Lin and Randall 1995). Another possibility that appears to function in lieu of or conjunction with H -ATPase, is an apically +  located N a - H exchanger that also excretes H and thus creates a favourable gradient for N a +  +  +  +  entry into the cell. At the basolateral membrane, N a is actively exchanged for K and possibly +  NFL via Na ,K -ATPase. C a +  Na /Ca +  2+  +  2 +  +  exits across this membrane via one of 3 possible pathways -  exchange, Ca -ATPase, or another as yet unidentified pathway (Li et al. 1997). 2+  Proton supply for H -ATPase comes from deprotonation of N H +  + 4  and hydration of C 0  2  which enters via diffusion from the blood. C F uptake occurs against an electrochemical gradient and is thought to be via active exchange for H C 0 ~ at the apical membrane (Marshall 1995, Lin 3  and Randall 1995). The precise locations of the proton pump and ion conductive channels for N a and C a +  2 +  remain equivocal, but current evidence suggests that in addition to the C C they  may also be located on the apical membrane of pavement cell (Perry et al. 1992, Goss et al.1994, Lin et al. 1994). It is clear from these models of ion regulation that Na ,K -ATPase in +  +  both SW and FW, and H -ATPase in FW, are involved in maintaining osmotic and ionic homeostasis in juvenile and adult fish. The development of biochemical assays for these enzymes has allowed workers to quantify and correlate enzyme activity with osmoregulatory capacity and thus gain a better understanding of the physiological responses of fish in different osmotic environments (Zaugg 1982, McCormick 1993, Lin and Randall 1993, Schrock et al. 1994).  125  During ontogeny, and prior to the development of the adult osmoregulatory mechanisms, embryos and larvae must rely on extrabranchial tissues to meet their needs for osmotic and ionic regulation during early life (Alderdice 1988, Chapter 2). Numerous studies have located CCs in the cutaneous epithelia of fish embryos and larvae and implicated these cells as the sites for osmoregulation in both SW (Lasker and Threadgold 1968, Guggino 1980a, Hwang and Hirano 1985, Wales and Tytler 1996) and F W fishes (Ayson et al. 1994, Kaneko et al. 1995, Uchida et al. 1996, Chapter 2). The cellular structure of these cells is similar to their adult counterparts and to date there are no reasons to assume they function any differently (Alderdice 1988, Hwang 1989). Studies using fluorescent localization methods (DASPEI) for CCs in embryos and larvae have identified increased C C surface area (Ayson et al. 1994, Kaneko et al. 1995, L i et al. 1995) and increased CC density (Li et al. 1995) in response to osmotic and ionic challenges. McCormick (1990a) used anthroylouabain, a fluorescent derivative of ouabain that specifically binds to the extracellular component of the Na ,K -ATPase molecule, and DASPEI +  +  counterstaining, to verify the high concentration of this enzyme in CCs of adult fish. McCormick (1990a) showed that increased CC fluorescence, density, and surface area in adult tilapia opercular membrane were associated with increased Na ,K -ATPase activity. In adult +  +  fish, salinity is the most widely recognized environmental determinant of Na ,K -ATPase +  +  activity and C C development (McCormick 1995). In Chapter 2 of this study, embryos and larvae of chum and coho salmon acclimated to various salinities were examined for cutaneous CC distribution and density using DASPEI. Chloride cells were distributed more widely over almost all of the major cutaneous surfaces than previously had been reported for embryos or larvae of other species. These observations directly support the proposed importance of cutaneous ion exchange in early life osmoregulation  126  (Alderdice 1988). However, other than the presence of cutaneous CCs there is very little information that details the functional significance of these cutaneous tissues in relation to osmoregulation. The aim of this study was to examine one aspect of this functionality by measuring Na ,K -ATPase and H -ATPase activity levels specific to the TRE, Y S E , and gills of +  +  +  embryos and larvae acclimated to various salinities. To date there are only a few reports on Na ,K -ATPase activity in fish larvae, all of which are limited to gill tissue, and two of which +  +  are very recent (Sullivan et al.1983, pink salmon; Uchida et al. 1996, chum salmon; Zydlewski and McCormick 1997, shad). Even less information exists for the embryonic stages. There do not appear to be any Na ,K -ATPase or H -ATPase activity measurements for embryonic +  +  +  stages, much less for cutaneous tissues of either embryos or larvae.  127  Methods Embryos and larvae of chum and coho salmon were acclimated (>7 d) to a range of salinities (0 to 30 %oS) and sampled for Na ,K -ATPase and H -ATPase activity in gill, trunk, and yolk sac +  +  +  epithelium. Additional gill tissue samples were collected from the F W control treatment of chum salmon to follow the ontogeny of enzyme activities from eyed embryos through to the emergent fry.  Fish Chum and coho salmon eggs sampled for ATPase activity measurements were the same as those used for the acute exposure study (1994/95) presented in Chapter 1; obtained from Goldstream River (near Victoria, BC) and Chase River (Nanaimo, BC), respectively. However, these eggs were exposed to the SW-exposure protocol outlined for the acclimated SW-exposure study (1993/94) in the Methods section of Chapter 1 (see Table 1.1).  Tissue preparation Sampling periods and techniques were generally the same as those described in Chapter 2 for the fluorescent microscopy tissue samples: Eyed and prehatch embryos and yolk sac larvae were sampled for branchial, TRE, and Y S E , but not H D E or FFE. Tissue samples were removed whole and were not cut, because there was no need to make them lie flat. Tissues samples from multiple individuals were pooled to provide sufficient tissue for analysis of a total of 6 replicate samples per salinity treatment at each developmental stage. Number of individuals and approximate wet weights of replicate samples for each developmental stage/tissue/salinity combination are shown in Table 3.1. At the eyed embryo stage only Y S E tissue was collected since the gills were small and largely undeveloped; consisting only of the supporting gill arch 128  and small filament buds. At the prehatch embryo and yolk sac larva stages TRE samples were removed only from 1 'side' of each individual. Care was taken to skin the animal as closely as possible and minimize the amount of underlying muscle tissue collected. Dissected tissues were placed into 1.5 mL microcentrifuge tubes containing 100 uL of ice-cold SEI buffer (150 m M sucrose, 10 m M E D T A , 50 m M imidazole, pH 7.3) and stored at -80°C for less than 4 months before analysis. Additional gill tissue samples were collected for chum salmon in 0 %oS at 3 intermediate developmental stages [25% (620 ATU), 75% (900 ATU), and 100% (1040 A T U ) yolk sac absorption] to allow for a more complete evaluation of ontogenic changes in gill N a , K +  +  ATPase and H -ATPase activity levels in FW. +  Tissue analysis Na ,K -ATPase and H -ATPase activities in crude tissue homogenates were determined using +  +  +  the coupled-enzyme assay described by Penefsky and Bruist (1984), and incorporated modifications presented by Lin and Randall (1993) and Kultz and Somero (1995) for H -ATPase, and by McCormick (1993) for Na ,K -ATPase measurements in a microplate +  +  +  reader (i.e., multi-sample micro-spectrophotometer). In this kinetic assay, the ouabain-sensitive and Af-ethylmaleimide (NEM)-sensitive hydrolysis of ATP is coupled in an equimolar ratio to the oxidation of nicotinamide adenine dinucleotide (reduced form, N A D H ) , utilizing the conversion of phosphoenolpyruvate (PEP) to lactate as catalyzed by pyruvate kinase (PK) and lactate dehydrogenase (LDH).  129  Table 3.1. Sampling details of tissues collected for Na ,K -ATPase and H -ATPase analyses. +  +  +  Number of tissue samples and approximate wet weights  Species Tissue  Eyed embryo  P H embryo  Y S larva  a  a  n  Weight,  n  Weight,  n  Weight,  (indiv.)  (mg/indiv.)  (indiv.)  (mg/indiv.)  (indiv.)  (mg/indiv.)  Chum Gill basket  b  b  3  -1.5  2  -6  Trunk epithel.  b  b  3  ~3  1  -17  3  -3.5  2  -3.5  Yolk sac epithel.  Coho Gill basket  b  3  -1.0  2  -6  Trunk epithel.  b  3  -1.5  1  -9  3  -3  2  Yolk s a c epithel.  -3  a  PRH= Prehatch; Y S = Yolk sac.  b  Tissue not readily available due to developmental stage.  Yolk sac epithelium proper not available due body wall overgrowth Note: each sample size shown (number of individuals) was pooled to provide the tissue for one replicate. Six replicate samples were collected for each developmental stage/tissue/salinity combination indicated (e.g., chum eyed embryo at 0 %oS: yolk sac epithelium (YSE), 3 YSEs x 6 replicates =18 individuals sampled at this salinity). c  130  A n assay mixture (solution A ) containing final concentrations of 45 m M NaCl, 10 m M KC1, 2.5 m M M g C l , 2 m M PEP, 0.5 m M ATP, 0.16 m M N A D H , 3.6 U/mL PK, 2.8 U/mL L D H , 2  1 m M N a N , 1 m M E G T A , and 50 m M imidazole (pH 7.5) was prepared just prior to the assay 3  and kept at 4°C. Assay solution B was prepared as above but also contained 0.5 m M ouabain. Assay solution C was prepared by adding 1 m M N E M to solution B . The presence of sodium azide (NaN ) and E G T A in the assay mixtures ensured the inhibition of F-type (mitochondrial) 3  2_j_  ATPase and Ca -ATPase, respectively. A few minutes before use, the assay mixtures were placed in a 25°C water bath. Prior to thawing the samples, assay mixture A was checked by running an A D P standard curve. The slope of the standard curve was generally between 18 and 20 mOD units/nmol A D P (McCormick 1993). The tissue samples were thawed immediately prior to assay and kept on ice throughout homogenization. Sodium deoxycholate (0.1% in SEI buffer) was added to each microcentrifuge tube and the tissues were homogenized in the tubes using a motorized polypropylene pestle (Kontes, Vineland, New Jersey). The tissue was ground for 10 to 15 s and then centrifuged at 7500 x g for 30 s to remove the insoluble connective tissue material. With the 96-well microplate sitting on a paper towel-covered ice pack, 10 uL of homogenate was added to each of six wells for every sample. The remaining homogenate was removed for later analysis of protein content, which was determined using the bicinchoninic acid procedure (Smith et al. 1985). To begin the ATPase assay, 200 uL of solution A was added to the first two wells per sample, 200 uL of solution B to the next two wells per sample, and 200 uL of solution C to the remaining two wells per sample. The plate was then placed in a temperature-controlled microplate reader (Thermomax, Molecular Devices Corp., Menlo Park, C A , USA) set to 30°C,  131  and the linear rate of N A D H oxidation (i.e., disappearance) was measured at 340 nm for 10 min. For each assay (16 samples per plate) activity measurements were completed within 0.5 h of tissue homogenization. ATPase activity in the three assay mixtures was calculated from the slope of the linear portion of the reaction, the slope of the standard curve, and the protein content of the crude tissue homogenates:  , . . ATPase activity =  slope of sample 1 , _i x x 60 mm • h slope of std. curve protein content  Where the units are: ATPase activity, umol A D P m g protein h" ; slope of sample, -1  1  mOD 10 uL" m i n (OD = optical density); slope of standard curve, mODnmol A D P ; 1  -1  - 1  protein content, u.g-10 u.L . -1  The difference in the ATPase activity between solutions A and B represented the ouabainsensitive Na ,K -ATPase activity, and the difference in the ATPase activity between solutions B +  +  and C represented the NEM-sensitive H -ATPase activity. Data were analyzed using microplate +  software (Softmax, Molecular Devices Corp., Menlo Park, C A , USA) and duplicate subsample readings were averaged for the final value per sample.  Statistical analysis Interspecific differences in Na ,K -ATPase and H -ATPase enzyme activity within each +  +  +  developmental stage were tested using two-way A N O V A and significant differences between species were tested using the Tukey test for multiple comparisons. Within species differences in  132  enzyme activity were analyzed using one-way A N O V A and the Tukey test, or Kruskal-Wallis A N O V A on ranks with the Dunn multiple comparison test i f data did not meet A N O V A requirements of normality and homogeneity of variance (p < 0.05). Data are presented as means ± 1 SD and graph axis scales are standardized where possible.  133  Results Ontogeny of gill ATPase activity in FW - chum salmon Gill Na ,K -ATPase activity of chum salmon increased significantly (p < 0.05) at each +  +  progressive developmental stage sampled from the prehatch embryo to fry stage (100 % yolk sac absorption) (Fig. 3.2a). A large increase in Na ,K -ATPase activity of nearly 4-times occurred +  +  between the 75% yolk sac larva stage and the final fry stage. The pattern of ontogenic changes in H -ATPase activity was similar to the Na ,K -ATPase measurements except that the +  +  +  differences were less dramatic and therefore not always statistically higher between progressive developmental stages Fig. 3.2b.  Na ,K -ATPase activity in different salinities +  +  Gill epithelium Na ,K -ATPase activity was highest in the gills of yolk sac larva compared to gill tissue of +  +  prehatch embryos as well as TRE and Y S E at any of the 3 developmental stages. Gill enzyme activity levels were much higher (p < 0.05) in chum salmon larvae than coho (range: 2.4 to 5.8 umol ADP-mg protein -h compared to 2.5 to 2.9 umol ADP-mg protein -h ), respectively, -1  _1  -1  -1  whereas they were only slightly higher (p < 0.05) in prehatch chum embryos compared with coho (Fig. 3.2a and b). Yolk sac larvae of both chum and coho had enzyme activity levels that were 2 to 4 times higher than prehatch embryos. There were no consistent trends in gill Na ,K -ATPase activity in relation to acclimation +  +  salinity for either chum or coho salmon embryos or larvae, however, there were a few significantly different points. Chum yolk sac larvae acclimated to 24 % S and 6 %oS had N a , K +  +  0  134  ATPase activity levels that were significantly higher and lower (p < 0.05), respectively, than levels recorded in 0, 12, and 18 %oS (Fig. 3.2a).  Trunk epithelium Na ,K -ATPase activity in TRE of chum and coho yolk sac larvae was 4 to 5 times lower than +  +  respective gill enzyme activity. In contrast, enzyme activities in TRE of prehatch embryos were similar to gill tissue enzyme activity of coho and slightly lower (p < 0.05) than chum [Fig. 3.3a and b, compared with Fig. 3.2a and b (note differences in axis scales)]. Unlike the gill measurements, the highest TRE enzyme activities were recorded in coho prehatch embryos instead of chum. Skin Na ,K -ATPase activity levels were significantly higher (p < 0.05) in +  +  prehatch embryos of coho compared with yolk sac larvae of the same species in all salinities except 0 % o S . Similar to the pattern seen in the gills, TRE enzyme activity levels of chum yolk sac larvae were higher than those of prehatch embryos. N a , K - ATPase activity levels in TRE +  +  of coho yolk sac larvae were significantly lower (p < 0.05) in 6, 12, and 18 %oS compared with 0 %oS (Fig. 3.3b). No consistent salinity related differences in enzyme activity levels were observed in either species at the other developmental stages.  Yolk sac epithelium Na ,K -ATPase activity levels in Y S E did not show any clear trends in relation to acclimation +  +  salinity but enzyme activity of chum salmon larvae was consistently higher at all salinities compared with the earlier stages Fig. 3.4a. Unfortunately, no Y S E samples were obtained for the coho yolk sac larval stage. Due to rapid growth of body wall tissue over the yolk sac, this tissue  135  was no longer accessible at the time of sampling (approximately 10 d postexposure). Enzyme activities were significantly higher (p < 0.05) in coho eyed and prehatch embryos compared with respective chum embryos (Fig. 3.4a and b). Chum prehatch embryos had significantly lower (p < 0.05) enzyme activity levels than eyed embryos in all salinities except 18 %oS (Fig. 3.4a and b).  H - A T P a s e activity +  Gill epithelium Proton ATPase activity of yolk sac larvae decreased significantly (p < 0.01) in salinity treatments compared with F W in both chum and coho salmon (Fig. 3.5a and b), however, no differences (p > 0.05) were observed between different salinity treatments above 0 %oS. No salinity-dependent response was observed in gill tissues of prehatch embryos of either species. Interspecific comparisons showed that gill H - ATPase activities of coho yolk sac larvae +  generally were higher than chum larvae with significant differences (p < 0.05) occurring between 0 and 12 %oS. In contrast gill H -ATPase of coho prehatch embryos was significantly +  less active (p > 0.05) at all salinities compared with chum salmon.  Trunk epithelium No significant trends in H -ATPase activity of TRE were observed in relation to salinity for +  either chum or coho embryos or larvae. Chum salmon had markedly higher enzyme activity levels (p < 0.05) in prehatch embryos compared with yolk sac larvae (Fig. 3.6a), whereas no significant ontogenic differences were observed in coho (Fig. 3.6b). Although there were no  136  general interspecific differences between chum and coho, higher H -ATPase activities were +  observed in TRE tissue of coho yolk sac larvae acclimated to 0 and 6 %oS (p < 0.05) (Fig. 3.6a and b).  Yolk sac epithelium Similar to the skin measurements, no significant trends in H -ATPase activity of Y S E were +  observed in relation to salinity in either species at any of the 3 developmental stages. Interspecific differences were observed only in prehatch embryos where H -ATPase activities of +  coho Y S E were higher than chum salmon (Fig. 3.7).  137  Discussion N a , K - A T P a s e activity in embryos and larvae +  +  This study is the first to examine Na ,K -ATPase activity in extrabranchial tissues of teleost +  +  embryos and larvae and the gills of embryos. Cutaneous epithelia appear to play an important role in early life ion regulation and many embryos and larvae have extrabranchial CCs located in the skin and yolk sac epithelium (see Chapter 2 this study). Measurements of Na ,K -ATPase +  +  activity in the gills, skin, and yolk sac epithelium of FW-acclimated chum and coho salmon in this study were reasonably similar for all 3 tissues at the prehatch embryo stage, but by the yolk sac larva stage (50% yolk sac absorption), enzyme activity levels in the gills were as much as 4 to 5 times higher than the cutaneous tissues. By that stage, gills appeared to be a dominant site of Na ,K -ATPase activity in relation to the cutaneous tissues. Gill ATPase activity continued to +  +  increase through to the fry stage where, in chum salmon, it reached a maximum of about 16 umol ADP-mg protein -h . These activity levels are similar to values commonly associated -1  -1  with ocean migrating juvenile salmon. Enzyme activities of presmolt and smolt Atlantic salmon were 3.7 and 15.2 umol ADP-mg protein h , respectively (McCormick et al. 1987). Coho -1  - 1  salmon adapted to SW for 21 d had enzyme levels of 4.2 umol ADP-mg protein -h (Morgan -1  -1  1997). To date only a few studies have reported Na ,K -ATPase activity in teleost larvae and +  +  only in gill tissue (pink salmon, Sullivan et al.1983; Mozambique tilapia, L i et al. 1995; chum salmon, Uchida et al. 1996; American shad, Zydlewski and McCormick 1997). It has been shown that increases in N a , K -ATPase occur largely in the CCs of ion transporting epithelia such as gill and opercular membrane tissue (Sargent et al. 1975, McCormick 1990a and b).  138  Gills The ontogenic increase of gill Na ,K -ATPase activity in larvae and fry acclimated to F W +  +  coincides well with documented innate SW tolerance and hypo-osmoregulatory ability of chum larvae and fry (Weisbart 1968, Hasegawa et al. 1987, Groot 1989, Kaneko et al. 1995, Uchida et al. 1996, this study) and presumably prepares them for migration to SW following emergence (Salo 1991). In contrast, the markedly lower gill Na ,K -ATPase activity observed in coho yolk +  +  sac larvae is consistent with their reduced ability to tolerate SW and to hypo-osmoregulate compared with chum salmon (Weisbart 1968, Blackburn and Clarke 1986, Clarke et al. 1989, this study Chapter 1). These interspecific differences in enzyme activity were already apparent at the prehatch embryo stage of development. Comparison of Na ,K -ATPase activity measurements between studies is often difficult due to +  +  procedural, developmental, interspecific, and inter-stock differences. Two recent studies however, examined enzyme activity in chum salmon using the same basic method as this study. Uchida et al. (1996) measured Na ,K -ATPase activities in three early life stages of chum +  +  salmon: "late" yolk sac embryo (0.3 to 0.4 g), "early" fry (0.6 to 1 g), and fry (2 to 3 g) reared in F W had enzyme activities of about 4, 9.5, and 9 umol A D P m g protein •h , respectively. -1  -1  Uchida and Kaneko (1996) measured slightly higher levels in more advanced chum fry (4 to 7 g) that were either reared in FW or acclimated to SW for 14 d, ~ 13 and 15 umol A D P m g protein  -  ' • h . For ease of comparison I plotted the values from those two studies adjacent to my data -1  (see Fig. 3.1, compare data from literature plotted at stages 7 and 8 with data from this study, stages 5 and 6). On the whole their values were somewhat lower than those from this study. Given the potential for significant procedural differences between studies [presumably Uchida et  139  al. (1996) used 25°C (McCormick 1993) whereas I used 30°C], it may be more meaningful to compare studies by examining the relative increase in enzyme activity from one developmental stage to the next (sampling was conducted at similar developmental stages in these two studies), rather than the absolute values between studies. The relative changes in enzyme activity level were 2.1 times in Uchida et al. (1996) compared with 2.9 times in this study. Enzyme activity measured in larger chum fry (4 to 7 g) by Uchida and Kaneko (1996) was similar to activity of smaller chum fry (0.4 to 0.5 g) in this study, 13 umol ADP-mg protein -h" compared with 16 -1  1  umol ADP-mg protein -IT , respectively. -1  1  Pink salmon, a Pacific salmon that is closely related to chum salmon, also migrates directly to sea following emergence (Heard 1991). Sullivan et al. (1983) measured Na ,K -ATPase +  +  activity in early life stages of this fish starting with yolk sac larvae. However, because those authors used a different method (Zaugg 1982) to assay gill Na ,K -ATPase activity, only the +  +  patterns of enzyme activity are comparable and not the absolute values (5.5 to 7.5 u M inorganic phosphate • mg protein -h" ; 40 to 70 d posthatch). The basic ontogenic pattern of increasing -1  1  enzyme activity during the yolk sac larva stage [50% yolk sac larvae to fry stage (100% yolk sac absorption)] was similar to the pattern observed in this study but the proportional differences were not as great (i.e., increase of 1.3 times compared with 4 times in this study). [According to McCormick (1993) the proportional differences between these two assay methods are consistent.] Sullivan et al. (1983) reported that Na ,K -ATPase activity in pink salmon +  +  decreased rapidly (within 10 d) after 100% yolk sac absorption when they were maintained in FW. In contrast, chum salmon fry in F W maintained relatively high Na ,K -ATPase activity +  +  levels as 21-d old fry (Uchida et al. 1996) and 7 month old fry (Uchida and Kaneko 1996). These measurements are supported by Hasegawa et al. (1987) who reported that chum salmon 140  were capable of effective hypo-osmoregulation after up to a year of rearing in F W and noted that similar information for pink salmon was lacking. The data from Sullivan et al.(1983) suggest that there may be a physiological 'window' within which pink salmon need to migrate to sea or they will reduce or even lose their ability to adapt successfully to SW. Gill Na ,K -ATPase activity in American shad, an anadromous clupeid indigenous to +  +  eastern North America, showed a similar ontogenic pattern to chum larvae and fry but the overall levels were about 1/3 those of chum, even though the assay methods were similar (Zydlewski and McCormick 1997). Increased Na ,K -ATPase activity with development has +  +  also been observed in the euryhaline Mozambique tilapia between 10-d old larvae and 24-d old fry (Li et al. 1995), but due to markedly different assay procedures, including the use of saponin to permeabilize the membranes, absolute values are not comparable to this study. Those authors also reported their values as specific Na ,K -ATPase activity (tissue mass-specific) and +  +  observed that the specific activity of 10-d old larval gills was as much as 400 times higher than 180-d old adults. Specific activity is a useful measure for making comparisons between different life stages and tissue types; unfortunately most studies, including this one, do not report this value. For example, to examine the relative contributions of the cutaneous and branchial enzymes would require specific activity measurements of Na ,K -ATPase and H -ATPase, based on +  +  +  either mass or surface area of the respective tissues; data that unfortunately was not collected as part of this study. However, preliminary estimates based on the generalized tissue weights presented in the Methods section of Chapter 2 (Table 3.1), indicate that to make the measurements comparable on a per individual basis, the gill Na ,K -ATPase activity would +  +  need to be divided by 2 (2 branchial baskets were used) and the trunk epithelial values would  141  need to be multiplied by about 4 (1 side of A of the trunk skin was used). Applying these l  adjustments to 0 %oS data for chum and coho larvae, to estimate whole-body activity levels, indicates that the total activity of each of the tissues would be much closer in total contribution; if anything, the skin would be higher than the others including the gills, especially for coho larvae. More detailed measurements of this type would undoubtedly provide useful information to assist in our understanding of early life osmoregulation. Salinity is the most widely recognized and studied environmental factor affecting N a , K +  +  ATPase activity and C C development in the gills of juvenile and adult fishes (McCormick 1995). Enzyme activity usually rises by 3 to 5 times in juvenile and adult euryhaline teleosts following acclimation to SW (Kirschner 1980). Because this enzyme is localized in the CCs these increases usually are associated with a concomitant increase in number or size of the CCs (McCormick 1990a and b). In this study no obvious salinity-related Na ,K -ATPase activity changes were observed in +  +  gills until the yolk sac larva stage. Even then, the changes were evident only in chum salmon and were equivocal because only chum larvae acclimated to 24 % S had significantly higher 0  enzyme activity levels than in FW. Larvae acclimated to 6 %oS had significantly lower levels than in 0, 12, and 18 %oS (see Fig. 3.2a). In contrast, Uchida et al.(1996) observed a significant increase in gill Na ,K -ATPase activity of more advanced chum fry (2 to 3 g) within 5 d of +  +  transfer directly to SW, and a maximum after 21 d (approx. 14 umol ADP-mg protein -h , 1.4 -1  -1  times higher than initial level). Uchida et al. (1996) also reported an increase in C C surface area but no change in C C density in response to SW exposure, although there was a pronounced reduction of lamellar CCs in SW compared with FW. Together these data suggest that chum larvae and fry are capable of responding to increased osmoregulatory demands by increasing gill  142  Na , K -ATPase activity, which is probably related to changes in gill C C structure and size. Unfortunately, it is not possible to compare the enzyme activity levels of the gills to C C distributions and densities of the gills examined in Chapter 2 because due to inconsistent DASPEI staining (see Fig. 2.15) and the three-dimensional structure of the gills no quantitative measurements were obtained. In addition to the increased enzyme activity in chum larvae acclimated to 24 % o S , there was also a significant decrease in Na ,K -ATPase activity in larvae acclimated to 6 % o S . It is not +  +  clear what caused this reduction, but it is interesting to note that a number of the other tissues showed similar trends of decreasing enzyme activity levels at 6 or 12 % S. A similar result was 0  noted by Morgan (1997) in juvenile coho salmon and Kiiltz et al. (1992) in adult Mozambique tilapia. After 42 d of acclimation Morgan (1997) recorded no change in Na ,K -ATPase activity +  +  in FW, a decrease in iso-osmotic (12 % S) water and an increase, which was higher than the F W 0  water activity level, in SW. These responses are consistent with the idea that there is a reduced requirement for ion transport in iso-osmotic conditions, however, measurements of oxygen consumption rate often do not reflect a reduced energy expenditure (Morgan 1997, Chapter 1 this study). This may be further evidence that the energetic cost of the Na ,K -ATPase pump in +  +  gills constitutes only a small portion of whole-animal metabolism (Kirschner 1995, Morgan 1997).  Skin and yolk sac epithelium Ontogenic changes in Na ,K -ATPase activity of cutaneous tissues were less dramatic than +  +  those observed in the gills. Salinity acclimation did not appear to significantly influence N a , K +  +  143  ATPase activity in either of the cutaneous tissues (trunk or yolk sac epithelium). Although there are no other early life stage studies with which to directly compare cutaneous enzyme activities, the lack of change in enzyme activity in relation to SW exposure appears to be in contrast to measurements made in epithelial tissues of in adults fish. McCormick (1990a) measured Na ,K -ATPase activity, using the same basic enzyme assay method as this study (McCormick +  +  1993), and C C density in the opercular membrane of Mozambique tilapia and jaw skin of mudsucker. Enzyme activities in FW- and SW-adapted tilapia opercular membranes were 0.5 and 1.3 pmol ADP-mg protein h , respectively, and corresponding C C densities were 0 (not -1  - 1  detectable) and 95 C C m m . Na ,K -ATPase activity and CC density were both markedly - 2  +  +  higher in the mudsucker jaw skin adapted to SW; 3.5 umol ADP-mg protein -h and 293 -1  -1  C C m m , respectively. Comparable cutaneous values in chum and coho embryos range from - 2  0.6 to 1.1 umol ADP-mg protein -h and 855 to 252 C C - m m (Table 2.1 and Table 2.2, -1  -1  -2  Chapter 2). Unlike adult tilapia opercular epithelium, Na ,K -ATPase activity in cutaneous +  +  tissues (TRE or YSE) was not stimulated by salinity, and cutaneous C C density actually decreased in chum and coho embryos and larvae exposed to higher salinities (> 12 % o S ) . This result exemplifies how in general there was not a very convincing correlation between cutaneous Na ,K -ATPase activity and cutaneous C C densities during the early life stages of chum and +  +  coho. Very little information exists in the literature to indicate whether this lack of correlation is the result of life stage or interspecific differences. A comparison of the Na ,K -ATPase activity of the two species used suggests that there was +  +  a difference in cutaneous enzyme activity that may have been related to different levels of fluorescence noted in Chapter 2. Trunk and yolk sac epithelium of coho embryos fluoresced  144  more brightly than chum salmon, especially at the prehatch embryo stage. In light of these observations it is interesting to note that the Na ,K -ATPase activity in both the T R E and Y S E +  +  were higher for coho in the two embryonic stages. There is even an exceptional increase in C C density in T R E of coho prehatch embryos that is correlated with a respective elevation in Na ,K -ATPase activity. These observations are intriguing since they appear to add support to a +  +  postulation presented in Chapter 2. In an attempt to explain why coho were less effective hypoosmoregulators than chum despite the relatively high CC numbers in coho, I speculated that this apparent disparity might be related to an inability of coho to adjust their C C structure to effectively extrude ions, which in SW is an essential component of maintaining osmotic and ionic homeostasis in hyperosmotic conditions. Is it possible that the high number of cutaneous CCs and the associated relatively high enzyme activity of these cells is an indication that they are most effective at ion uptake as a requirement of hyperosmoregulation in an ion-poor environment, and unlike chum salmon cannot be altered for hypo-osmoregulation? Further investigation into the structural changes of cutaneous CCs in coho compared with chum salmon would be required to provide more definitive evidence for these speculations.  H -ATPase activity in embryos and larvae +  Gills Gill H -ATPase activity in chum salmon, measured as NEM-sensitive activity, increased about +  5-fold throughout development from the prehatch embryo to the fry stage (see Fig. 3.1). Enzyme activity levels in gill tissue of chum fry in F W (3.3 umol A D P • mg protein -h ) were higher -1  -1  than levels reported for adult rainbow trout gills in F W (1.5 umol ADP-mg protein" -h ) and 1  -1  145  were similar to levels reported for experimentally elevated Ca  and Cortisol injected treatments  (~ 3 and 2.5 umol ADP-mg protein h , respectively) (Lin and Randall 1993). Acclimation of -1  - 1  chum and coho larvae to SW produced a significant decrease in gill H -ATPase activity, and +  was similar to the reduced activity in higher salinities reported by Lin and Randall (1993) for adult rainbow trout gills (0.4 umol A D P m g protein" h ). Unlike the larval stage, neither chum 1  -1  nor coho prehatch embryos showed altered gill H -ATPase activity in response to increased +  salinity. Nevertheless, the decrease in H -ATPase activity seen in the larvae stage in SW was +  consistent with the current model for the role of a proton pump in N a and possibly C a +  2 +  exchange across the apical membrane of adult trout gills. This suggests that larval gills are functioning similar to adult gills with respect to ion translocating enzymes but that embryonic gills are not. When examining the impact of external changes on embryos it is important to consider the buffering effect of the PVF-egg membrane complex (Eddy and Talbot 1983, Shepard and McWilliams 1989). Because this complex appears to provide significant alteration of ion gradients for the embryo it may therefore functionally reduce the gradients and thus reduce the need for the embryo to actively ion regulate during the egg stage. If the embryo ultimately is exposed to a lesser ion gradient, then the usual stimulus that increases Na ,K -ATPase activity +  +  may be diminished. Consequently, SW exposure may not stimulate enzyme activities in embryonic ion exchange tissues in the same manner as larval, juvenile, or adult tissues, which are exposed directly the environment. This may be a factor in the lower embryonic N a , K +  +  ATPase activity levels compared to the larval stages.  146  Skin and yolk sac epithelium H -ATPase activity was markedly lower in the skin and yolk sac epithelium than the gills but +  was similar in magnitude to F W control levels reported for adult trout gills (Lin and Randall 1993). Although the response to SW acclimation was not a clear one, there did appear to be a general pattern of decreased H - ATPase activity in TRE and Y S E of both chum and coho +  prehatch embryos and yolk sac larvae in salinities of either 6 and/or 12 %oS (see Fig. 3.6 and Fig. 3.7). In a number of tissues the low enzyme activities at 6 or 12 %oS appear to be minima since at 18 %oS activity actually increased again. Not all of these responses were statistically significant but they did suggest a generalized, albeit variable, response to salinity. It is interesting to note that similar to the Na ,K -ATPase activity levels in the yolksac of prehatch +  +  coho embryos, Ff - ATPase activity of the yolk sac at the same stage was also higher than in +  chum salmon. This agrees with the observation from Chapter 2 of noticeably brighter fluorescence of CCs in the Y S E of prehatch embryos, especially at the lower salinities. There is not sufficient evidence to draw unequivocal conclusions from these data but they do suggest that cutaneous epithelia are involved in ion uptake in F W during the early life stages. This speculation is consistent with the model for early life ion regulation in fishes where the cutaneous epithelia provide sites for ion exchange during organogenesis when the embryo is actively exchanging nutrients and ions with the environment but the gills have not yet fully developed (Guggino 1980a, Alderdice 1988, Hwang 1989).  Ion regulation in embryos and larvae Studies on F W ion regulation in fish embryos and larvae have focussed primarily on N a ; the +  net uptake of which begins at the embryo surface sometime between the eyed stage and hatching 147  in salmonid embryos (Rudy and Potts 1969, Eddy and Talbot 1983, Shepard and McWilliams 1989) and even sooner in tilapia embryos (Hwang et al. 1994). Although this uptake may not result in a net ionic increase in intact eggs, since the ions are simply recycled within the PVF, it indicates that ion transport is already occurring at the embryo surface (Eddy and Talbot 1983, Shepard and McWilliams 1989). Following hatching, reports of increased N a content indicate +  net N a uptake and probably ion regulation (Satoh et al. 1987, McWilliams and Shepard 1989, +  Hwang et al. 1994). Hwang et al. (1994) reported a significant increase N a , K , and C a +  +  2 +  uptake between 1 and 10 d posthatch in Mozambique tilapia larvae. Increased ion uptake at hatching is probably related to a combination of increasing gill differentiation as well as removal of the external egg membrane. Measurements of H -ATPase in this study indicate that markedly higher enzyme activities +  occur in the gills of yolk sac larvae compared with prehatch embryos. Although it is likely that the majority of this ion exchange occurs in the branchial epithelium of yolk sac larvae, H -ATPase and Na ,K -ATPase activity in the cutaneous epithelia suggests that cutaneous ion +  +  +  exchange may still provide a significant contribution at the larval stage and perhaps beyond. In adult rainbow trout, Perry and Wood (1985) reported that up to 50% of non-dietary C a  2 +  uptake  occurred through the skin. McCormick et al. (1992) demonstrated that opercular epithelium of Nile tilapia was capable of transporting Ca  against an ionic and electrical gradient and that this  ability was correlated with CC density. In FW, both N a and C a +  2 +  uptake occur across the apical  membrane of either, or both, the CCs (McCormick et al. 1992, Marshall et al. 1995) and/or the branchial pavement cells (Perry et al. 1992, Goss et al.1994). These ions enter the epithelial cells down a steep electrochemical gradient, which is energized by the proton pump (Marshall 1995,  148  Lin and Randall 1995), through ion-conductive channels in the apical membrane and across the basolateral membrane into the extracellular space via a number of possible routes, including Na ,K -ATPase, Ca -ATPase, and N a / C a +  +  for Ca  2+  +  2+  exchanger (Marshall 1995, Flik et al. 1995), and  a currently unidentified pathway (Li et al. 1997).  In his review on transport processes in isolated teleost epithelia Marshall (1995) reiterated ".. .the need for a flat epithelial model to study the operation of freshwater M R [mitochondria rich or chloride] cells ...". In light of this, and much of the information presented in this thesis, it seems that the yolk sac epithelium (YSE) of large embryos like salmon, may be a suitable candidate as a model for permeability studies and ion transport processes. It is appears to be an important exchange and regulatory epithelium (Alderdice 1988) that contains moderately high densities of cutaneous CCs (Kaneko et al. 1995, this study Chapter 2), is relatively large, flat, robust, and readily available. A number of studies have located and implicated cutaneous and branchial CCs as the functional ion exchange cells during the early life stages of F W teleosts (Guggino 1980a, Hwang and Hirano 1985, Ayson et al. 1994, Kaneko et al. 1995, L i et al. 1995, Chapter 2 this study). The presence of high H -ATPase and Na ,K -ATPase activity, which are implicated in +  Na and Ca  +  +  transport in adult epithelia, further supports the role of the cutaneous surfaces in  osmotic and ionic regulation during early development. The relatively higher activity of H -ATPase in the trunk epithelium of chum salmon earlier in development (embryos) compared +  with later in development (yolk sac larvae) (see Fig. 3.6) may be indicative of a 'switching' from predominantly cutaneous ion regulation in the embryo, to the ultimate branchial-dominated ion regulation in the juvenile, and adult. A possible reason why this same 'switching' is not evident in coho salmon embryos and larvae may be because unlike chum salmon, they are not  149  adapted for early life exposure to SW following emergence from the gravel. This speculation is further supported by the lower Na ,K -ATPase activity levels in the coho larvae compared to the +  +  chum. In summary, Na ,K -ATPase and H -ATPase activity levels measured in the branchial and +  +  +  cutaneous epithelia of embryos and larvae were similar, and in some cases higher, than levels reported for branchial tissue of adult fish. Based on Na ,K -ATPase activity in yolk sac larvae +  +  gills, this tissue appears to provide a significant contribution to early life ion regulation. Although there are no other studies with which to compare enzyme activity of embryonic or larval cutaneous tissues, Na ,K -ATPase activity of opercular epithelium of adult fish was found +  +  to be higher. There was no direct relationship between C C density and Na ,K -ATPase activity +  +  in embryonic and larval cutaneous epithelia. The response of Na ,K -ATPase activity to +  +  increasing salinity for all tissues was variable and equivocal; no definitive trends were evident. A strong ontogenic relationship existed with Na ,K -ATPase and development time for chum +  +  salmon embryos, larvae, and fry in FW. Decreased H -ATPase activity in the branchial tissue of +  chum and coho larvae exposed to SW coincides with results from similar experiments using adult trout gills. The similarity of these responses to SW exposure suggests that embryonic and larval branchial CCs respond in a manner that is similar to adults and supports the idea that they function in ion uptake in FW. A similar but much weaker response was seen in cutaneous H -ATPase activity. These results provide further support for the role of cutaneous tissues in +  osmotic and ionic regulation during the early life stages, when the gills have not yet fully developed.  150  Figures - Chapter 3  16 14  •  rz  12  CD CO  prot  'CD  10  TPa  ctiv  •  E  8  CD  <  ri_  <o  +  7 8  Fry (0.6 to 1.0 g) Fry (4 to 7 g)  1,2  1  2  4  E  CD  Eyed embryo Prehatch embryo Larva; 2 5 % Y S A Larva; 50% Y S A Larva; 7 5 % Y S A Fry; 100% Y S A (0.3 g)  6  Q  +  H  1 2 3 4 5 6  Branchial Trunk K V v l Yolk sac •a Branchial data  2  J2.  0  8  6  >  c  I ?o o  CD CD CO CD  CD  4  1_  Q.  O) Q_ E ri_ + Q X < o E 0  Fig. 3.1. Ontogeny of Na ,K -ATPase and H -ATPase activity in branchial, trunk, and yolk sac epithelium tissue of chum salmon embryos and larvae [25% to 100% yolk sac absorbed (YSA)]acclimated to F W (%oS). Data at developmental stages 1, 2 and 4 are the same as those plotted in Fig. 3.2a to Fig. 3.7a for eyed and prehatch embryos and yolk sac larvae acclimated to 0 % o S . Mean (± SD, n = 6) activity levels with similar letters are not statistically different (one-way A N O V A , Tukey test,/? > 0.05). Data from Uchida et al. +  +  +  1  151  (1996); Data from Uchida and Kaneko (1996). 2  (a) C H U M - branchial epithel. I — P r e h a t c h embryo , Yolk sac larva r  6H  4 H c 'CD  o  Q. CD  E cL Q < o E  2 H  0 •S o  12  6 H  18  24  (b) C O H O - branchial epithel.  03 <U CD  D_ +  4 H  + co 2-^  0 0  6  12  18  Salinity (%o) [7-day acclimation] Fig. 3.2. Branchial epithelium Na ,K -ATPase activity in (a) chum and (b) coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities. Mean (± SD, n = 6) activity levels with similar letters are not statistically different (one-way A N O V A , Tukey test, p>0.05). +  +  j  152  2.0 - , (a) C H U M - trunk epithel. ] Prehatch embryo Yolk sac larva  1.5  H  1.0 'CD  O i_  »  Q. CD  E ri  0.5 H  Q < o E  0.0 0  2.0  12  o  CO CD CO CO Q_  h< +  18  24  (b) C O H O - trunk epithel. 1.5  H  1.0  H  i  + CO  T b" b' b'  0.5  H  0.0 6  12  18  Salinity (% ) [7-day acclimation] 0  Fig. 3.3. Trunk epithelium Na ,K -ATPase activity in (a) chum and (b) coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities. Mean (± SD, n-6) activity levels with similar letters are not statistically different (one-way A N O V A , Tukey test, p > 0.05). +  +  153  (a) C H U M - yolk s a c epithel. 'MB—I  Eyed embryo Prehatch embryo Yolk sac larva  O Q. CO  E ri_ Q  < o E  ZZL  o  (b) C O H O - yolk s a c epithel.  CO CD C/>  CO  D_ h<  +  1.5  H  i  +  CO  1:0 A  b'c' a  a'  u  0.5 A  o.o  0  6  12  18  Salinity (% ) [7-day acclimation] 0  Fig. 3.4. Yolk sac epithelium Na ,K -ATPase activity in (a) chum and (b) coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities. Mean (± SD, n = 6) activity levels with similar letters are not statistically different (one-way A N O V A , Tukey test,/? > 0.05). +  +  154  3.0  -i  (a) C H U M - branchial epithel. • B  Prehatch embryo Yolk sac larva  12  24  i«—•*!  2.5 -  2.0 -  1.5 c  'CD  •4—»  1.0  o 1_  CL cn  0.5  E ri_  Q  < o E o CD CD  to  CD D_  0.0 3.0 (b) C O H O - branchial epithel. 2.5  2.0  i  +  X  1.5  1.0  0.5  H  I  0.0 0  6  12  18  Salinity (% ) [7-day acclimation] 0  Fig. 3.5. Branchial epithelium H -ATPase activity in (a) chum and (b) coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities. Mean (± SD, n = 6) activity levels with similar letters are not statistically different (one-way ANOVA, Tukey test, p > 0.05). +  155  2.0 - i  (a) C H U M - trunk epithel. Prehatch embryo Yolk sac larva  1.5  H ab  I  1.0 H  c '  a'b'  O i_  Q.  cn E ri. Q < o E ZL  0.5  H  b  1  1 a'b'  J L L  b'  o.o  18  24  2.0  >^  (b) C O H O - trunk epithel.  \> O  CO CD CO CO  D_  1.5  H  1.0  H  0.5  H  +  I  0.0 0  6  12  18  Salinity (%o) [7-day acclimation]  Fig. 3.6. Trunk epithelium H -ATPase activity in (a) chum and (b) coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities. Mean (± SD, n = 6) activity levels with similar letters are not statistically different (one-way A N O V A , Tukey test, p > 0.05). +  156  2.0  -i  (a) C H U M - yolk s a c epithel. '  1.5  ' Eyed embryo IB—ai Prehatch embryo • H Yolk sac larva  H  1.0 A c CD O i_  -4—"  Q. CO  E  a"b"  1^ 0.5 H  JtJi  CL Q <  "o E  0.0  >.  2.0  ZL  O CD CD (fi CD CL  ab b'a"b" J ,  a  a"  bal  a'  I 24  (b) C O H O - yolk s a c epithel. 1.5 A  +  I 1.0  H b'  0.5 A  0.0 0  6  12  18  Salinity (% ) [7-day acclimation] 0  Fig. 3.7. Yolk sac epithelium Ff -ATPase activity in (a) chum and (b) coho salmon prehatch embryos and yolk sac larvae acclimated to various salinities. Mean (± SD, n = 6) activity levels with similar letters are not statistically different (one-way ANOVA, Tukey test, />>0.05). +  157  G E N E R A L DISCUSSION A N D C O N C L U S I O N S  This thesis examined the oxygen consumption rate, growth, and osmoregulatory mechanisms of embryos and larvae of chum and coho salmon in relation to SW exposure. I hypothesized at the outset of this project that the different early life histories of these two species would be reflected in the energetics and mechanisms of osmoregulation in early life. Chum salmon spawn in close proximity to the ocean (including the intertidal zone) and migrate immediately to the ocean following emergence, compared with coho salmon that spawn further upstream and reside in FW for a period as fry prior to emigrating. Although these fish differ significantly in their ultimate ability to tolerate SW, especially as larvae and fry (Chapter 1), surprisingly little response to SW was observed in the measured physiological variables (Chapters 1 to 3). There was no evidence to suggest that chum salmon eggs and larvae reared in intertidal conditions would obtain any measurable energetic advantage over coho salmon reared in FW conditions, with regards to reduced energy expenditure resulting from exposure to moderate salinities near iso-osmotic. The passive protection provided by the PVF-egg membrane complex, which in FW water at least constitutes a low permeability barrier between the embryo and the external environment (Eddy and Talbot 1983, Shepard and McWilliams 1989), would presumably act similarly in both species. Yet mortality data, which was consistent with Weisbart (1968), indicated that some combination of this barrier and the embryo's inherent osmoregulatory abilities allowed chum salmon to survive higher salinities compared with coho (Chapter 1). Despite these differences in SW tolerance, no clear evidence for improved osmoregulatory function was indicated by salinity-related changes in distribution or density of cutaneous CCs (Chapter 2), or the activity of cutaneous Na ,K -ATPase (Chapter 3). These data suggest that the regulatory +  +  158  mechanisms of relatively advanced embryos may still be largely passive, which is consistent with the oxygen consumption rate data (eyed embryos Chapter 1). Little is known about changes in epithelial permeability in larval fish but some studies on marine fish have shown that it is definitely lower than adults (Mangor-Jensen 1987, Tytler and Bell 1989) and may change in relation to external osmotic and ionic concentrations (Brown and Tytler 1993). Data from cutaneous tissues in adult fish (Mozambique tilapia, Kiiltz and Onken 1993) and amphibians (Lillywhite and Maderson 1988) suggests that some species are able to adjust it in the face of high osmotic stresses. It is possible that as part of their adaptation to living closer to the ocean, chum salmon either maintain a lower epithelial permeability or are able to control it rendering them less susceptible to the lethal effects of SW exposure as embryos and larvae without incurring a significant and measurable energetic cost. Oxygen consumption rates of eyed and prehatch embryos of chum and coho salmon were similar following SW acclimation and probably indicate the use of similar coping mechanisms (Chapter 1). Measurements of growth (TDW and P ) in prehatch embryos indicated that r  depressed Mo values in higher salinities (18 and 24 %oS) may have been the result of 2  detrimental effect of salinity on growth. A number of possible mechanisms were discussed including increased epithelial permeability resulting in reduced oxygen diffusion into the tissues, and reduced embryonic activity resulting in insufficient mixing of oxygen-poor boundary layers in the PVF. Chloride cell fine structure, as revealed by T E M , is a useful indicator of cell functionality (Hwang and Hirano 1985, Hwang 1989, Chapter 2). In general, preliminary observations indicated that CCs in SW acclimated prehatch chum embryos were often similar in structure to 'typical' SW-adapted CCs, with an apical crypt, extensive tubular system and numerous dark  159  mitochondria. Although similar CCs were also observed in SW acclimated coho embryos, more commonly the CCs resembled 'typical' FW-adapted cells, with a flush or protruding apical membrane, limited tubular system, and fewer and lighter mitochondria. Considering these structural characteristics along with the negligible influence of SW on C C density and N a , K +  +  ATPase activity, the data suggests neither species were capable of adjusting their osmoregulatory processes to fully hypo-osmoregulate at the prehatch embryo stage, but that chum were sufficiently more capable at making the necessary adjustments to survive higher salinities. Cellular organization of chum salmon CCs for hypo-osmoregulation, may be more the result of ontogenic changes associated with early life development rather than the result of short term salinity-induced changes. These changes appear to reflect interspecific life history differences since salinity did not seem to influence Na ,K -ATPase activity nor C C density in +  +  either species. The most notable difference between yolk sac larvae of the two species was the inability of coho larvae to survive in salinities above 12 % S, whereas chum salmon survived without 0  significant mortality at 24 %oS and some (20%) even survived in 30 %oS (Chapter 1). (It should be noted here that stock differences may be a significant factor - coho embryos obtained the following year (1994/95) from a different river system were able to tolerate up to 18 %oS.) Oxygen consumption rates of chum salmon yolk sac larvae increased in a step-wise manner in salinities above 0 %oS and were as much as 51% higher in 30 %oS compared with FW. As with the embryos, no obvious changes in cutaneous C C density or distribution were observed between the two species (Chapter 2), however, marked interspecific differences in gill N a , K +  +  ATPase were observed. In general, enzyme activity in chum yolk sac larvae was significantly higher than coho but may have been in part related to an inherent ontogenic increase in Na , K -  160  ATPase activity, that was independent o f salinity and did not seem to be influenced unequivocally significantly by SW acclimation. (Chapter 3). At the end o f Chapter 11 identified the need to examine the mechanisms of osmoregulation in light o f the fact that acclimation to different salinities did not account for the variable oxygen consumption rates observed in the different salinities. Given that Na ,K -ATPase consumes +  +  A T P to actively transport ions one might expect a positive correlation between the activity o f this enzyme and oxygen consumption rate. However, similar to Morgan (1997), no consistent direct relationship was observed. This result further supports the idea that the energetic cost o f the ion transport may be relatively low compared to whole-body metabolism (Eddy 1982, Morgan and Iwama 1991, Kirschner 1993, 1995, Morgan 1997). The general lack o f changes in CC structure o f embryos, and Na ,K -ATPase activity o f +  +  embryos and larvae in response to SW exposure, suggests that these developmental stages do not yet possess the necessary chemical mediators such as hormones to control these regulatory processes. In adult fish, salinity is widely recognized as the most significant environmental determinant  of  Na ,K -ATPase activity (McCormick 1995). Hormones such as +  +  Cortisol  and  prolactin have profound effects on CC fine structure and activity. In juvenile and adult salmonids,  Cortisol  is one of the major hormones responsible for many of the changes associated  with SW adaptability. Increased plasma Cortisol levels, resulting from either endogenous or exogenous sources, are related to a variety of effects including; increased CC numbers (Richman and Zaugg 1987, Madsen 1990a and b), increased cell size (Madsen 1990b), and increased Na ,K -ATPase activity (McCormick and Bern 1989). However, the role of this hormone +  +  appears to be complex and multifunctional, since it also appears to be related to ion uptake in FW-adapted fish. Laurent and Perry (1990) showed that Cortisol was involved in increased N a  +  161  and C T influx and increased the number and apical surface area of gill CCs in rainbow trout. In addition, it also caused increased gill H -ATPase in coho salmon, which as mentioned earlier +  has been linked to N a and possibly C a +  2 +  uptake in F W (Lin and Randall 1995). Prolactin also  appears to play a significant role in F W osmoregulation and CC function by affecting C C fine structure in juvenile and adult fish at least. In general, it affects CCs in a manner that is more or less opposite to that of Cortisol by reducing CC size and density in SW-adapted tilapia (Foskett et al. 1982), as well as changing the proportions of CCs spanning the OPE in SW-adapted tilapia (Herndon et al. 1991) and thus effectively removing them as functional secretory cells (McCormick 1995). Cortisol levels have been measured in chum salmon from fertilization to 3 weeks past yolk sac absorption (fry stage) (de Jesus and Hirano 1992). Cortisol levels were initially high as a result of maternal contribution at the time of fertilization but decreased rapidly within about 3 wk postfertilization (about 230 ATUs) followed by a slight increase at 1 wk prehatch (540 ATUs). Levels remained more or less constant until 1 wk posthatch and then increased to a peak near the time of complete yolk sac absorption, after which time they decreased again. These changes appear to be reasonably well synchronized with normal development of salt water tolerance for chum fry, which emerge from the gravel and almost immediately begin their seaward migration. However, it would be interesting to know whether endogenous Cortisol levels vary in response to SW acclimation in chum and coho embryos and larvae, and whether exogenous Cortisol would improve their SW tolerance. Hwang and Wu (1993), measured endogenous Cortisol in F W and SW incubated and reared Mozambique tilapia larvae and found that levels increased dramatically but that there was no salinity related difference. In the second part of the experiment they administered exogenous Cortisol via feed and found that larvae that  162  received  Cortisol  survived a SW challenge better and had lower plasma osmolalities than the  control group. Ayson et al. 1995 showed that Cortisol stimulated the size and density of CCs in the yolk sac membrane of Mozambique tilapia larvae. It is apparent that larval cells respond to Cortisol but it is still not clear how effectively these animals can alter endogenous levels to affect osmoregulatory competence. It appears likely that the increased hypo-osmoregulatory ability of tilapia larvae resulting from the application of exogenous Cortisol, is due to increased C C size and density (Ayson et al. 1995). Since similar changes are seen in tilapia larvae acclimated to SW conditions (Ayson et al. 1994, L i et al. 1995) it suggests that endogenously produced Cortisol is effective in causing C C fine structural changes in tilapia larvae at least. Leatherland and Lin (1975) examined this same basic question in salmonids by addressing it from a different perspective. They examined pituitary development from the eyed embryo to complete yolk sac absorption in coho salmon incubated in d H 0 , F W (well water), 11% S and 13%oS SW. They concluded, based upon 2  0  pituitary development, that prolactin cells were active in coho larvae and fry and that this hormone fulfilled a role similar to that found in smolts. However, development of adrenocorticotropic hormone (ACTH) cells in the various treatments were not suggestive of any osmoregulatory function. It appears that these life stages are capable of producing at least some of the hormones involved in osmotic and ionic regulation, even though the extent to which they are actually involved in the control of these regulatory processes remains uncertain. Data from this study support the ideas propsed by Leatherland and Lin (1975); that hormonal influence on hypo-osmoregulatory processes, as regulated by SW exposure at least, appears to be minimal before and up to the yolk sac larva stage of development (50% yolk sac absorption) in chum and coho salmon. Oxygen consumption rates, which have been shown to increase in juvenile and adult fish in response to increased plasma Cortisol (Chan and Woo 1978, Barton and Schreck 163  1987, Morgan 1997), did not change except in chum salmon yolk sac larvae. Further research is required to examine the effects of hormones on the early life stages before we will be able to extrapolate with confidence from our knowledge about juveniles and adults salmonids or euryhaline larvae such as tilapia. In Chapters 2 and 3,1 suggested that the prominence of cutaneous CCs, supports the growing body of evidence that emphasizes the importance of cutaneous osmoregulation during early life. I also used a series of simple calculations to model changes in the proportions of cutaneous and branchial CC numbers, and suggested that towards the latter part of the larval stage, the developing gills may initially fulfil the role of an ion regulatory structure before they become important as a respiratory one. This suggestion is consistent with a similar idea set forth by Rombough and Moroz (1997). In further support of this, increased Na ,K -ATPase and +  +  FT-ATPase activity observed in the gills of chum, and to a lesser extent coho yolk sac larvae, suggest that these tissues were providing markedly greater ionoregulatory function to the larva than the embryo.  Conclusions Although many more questions have arisen out of this research than I have answered, there are a number conclusions that contribute to our overall understanding of the early life osmoregulation of salmonids. 1. Measurements of whole-animal oxygen consumption rate of embryos and larvae generally did not reveal any large changes in energy allocation to osmoregulation, except in chum yolk sac larvae. These larvae showed a progressive increase in oxygen consumption rate above 0 %oS to a maximum at 30 %oS of 51% above the rate in F W . Indirect evidence indicates that  164  larval activity was not the cause of the metabolic change, however, none of the additional physiological variables [growth, cutaneous C C density, C C enzyme activity (Na ,K - ATPase or H - ATPase)] were found to account for the increase either. It is unlikely +  that the full 51% of increased energy utilization is the result of increased osmoregulatory processes. 2. On the whole, the relatively small changes in oxygen consumption rates observed in animals exposed to rather large changes in osmotic gradients, supports the idea that the energetic cost of osmotic and ionic regulation is relatively low compared to whole-animal oxygen consumption rate. This also indicates that chum salmon embryos and larvae that are spawned in the intertidal zone of streams and rivers, probably do not acquire any significant energetic benefits from intermittent SW exposure as a result of reduced energy requirements for osmotic homeostasis. 3. Chum salmon embryos and larvae were more SW-tolerant than coho. Higher SW-tolerance of chum salmon at the later fry stage was clearly due to more effective hypo-osmoregulation. 4. Preliminary T E M examinations of cutaneous CCs indicated that chum salmon possessed more CCs with SW-type cellular structure compared with coho. Although coho had similar numbers of cutaneous CCs, it is probable that they were not able to alter the fine structure to cope with ion extrusion but instead used CCs largely for ion uptake. A detailed examination of cutaneous and branchial CC fine structure would undoubtedly further our understanding of those differences. 5. Calculation of cutaneous and branchial epithelium surface areas, and measurement of C C densities was used to estimate the relative contributions of cutaneous and branchial CCs as potential sites of ion exchange. The estimates indicated that the skin is more important for  165  osmoregulation during the embryonic and early larval stages of development, but that sometime near complete yolk sac absorption the gills appear to gain in relative importance. 6. Measurements of Na ,K -ATPase and H -ATPase activities showed an ontogenic shift +  +  +  between an initial reliance on the skin, followed by an increasing importance of the gills for osmoregulation. This supports the suggestion that during early development, gills may initially function as an osmoregulatory organ before a respiratory one. 7. Relatively high H -ATPase activity in F W acclimated yolk sac larvae implicated the +  branchial and cutaneous epithelia in ion uptake. Reduced enzyme activity in SW suggested that this enzyme operates similarly in larvae as it does in adults. 8. The general lack of noticeable changes caused by SW acclimation on oxygen consumption rate, growth, cutaneous CC density and CC enzyme activity, was somewhat surprising since it was in contrast to observations made on juvenile and adult fish. Although no hormone data were collected in this study, it seems likely that hormonal mediation of osmoregulation was not very important in contributing to adjusting osmoregulatory capacity during the embryonic and larval stages of salmonid life.  Directions for future research I have listed below a number of areas that I think would be useful (and interesting) in furthering our limited understanding of osmotic and ionic regulation in the early life stages fishes. A.  Continue to examine the early life osmoregulatory mechanisms and their development, this might include: i)  use of nanoliter osmometer to measure osmoregulatory competence of embryos and larvae;  166  ii)  detailed examination of CC fine structure and placement within the epithelium in various ionic environments, including a comprehensive ontogenic study,  iii)  use of immunocytochemistry to localize sites of transport molecules and cells (e.g., FT-ATPase, Lin et al. 1994; Na ,K -ATPase, Uchida et al. 1997); +  iv)  +  use concanavalin-A-fluorescein (Li et al. 1995) to identify mature and functioning CCs with apical crypts;  v)  examine the development and ontogenic shift to incorporation of gills, gut, and kidney, including drinking rates and cutaneous permeability;  vi)  measure oxygen consumption rates of different epithelia implicated in early life osmoregulation, and apply specific transport blockers to quantify the contribution of osmoregulatory work by these specific tissues.  B.  C.  Measure neuromuscular activity when conducting whole-body metabolic measurements i)  include measurements of protein synthesis.  ii)  include accurate measurements of instantaneous growth  Measure mass- or surface area-specific Na ,K -ATPase and H -ATPase in cutaneous +  +  +  and branchial tissues and relate contribution of specific tissues to whole-body ion regulation. D.  Measure changes in membrane permeability in different species in relation to different ionic conditions; i)  use salmon Y S E as model for generalized F W epithelium? (Marshall 1995).  167  REFERENCES Alderdice, D.F. 1988. Osmotic and ionic regulation in teleost eggs and larvae. In Fish Physiology. Vol. 11 A . Edited by W.S. Hoar and D.J. Randall. Academic Press, New York. p. 163-251. Alderdice, D.F., Rao, T.R., and Rosenthal, H . 1979. Osmotic responses of eggs and larvae of Pacific herring to salinity and cadmium. Helgolander wiss. Meeresunters. 32: 508-538. Alderdice, D.F., Wickett, W.P., and Brett, J.R. 1958. Some effects of temporary exposure to low dissolved oxygen levels on Pacific salmon eggs. J. Fish. Res. Board Can. 15: 229-249. Almatar, S.M. 1984. Effects of acute changes in temperature and salinity on the oxygen uptake of herring (Clupea harengus) and plaice (Pleuronectes platessa). Mar. Biol. 80: 117-124. Ayson, F.G., Kaneko, T., Hasegawa, S., and Hirano, T. 1994. Development of mitochondrionrich cells in the yolk-sac membrane of embryos and larvae of tilapia, Oreochromis mossambicus, in fresh water and seawater. J. Exp. Zool. 270: 129-135. Ayson, F.G., Kaneko, T., Hasegawa, S., and Hirano, T. 1995. Cortisol stimulates the size and number of mitochondrion-rich cells in the yolk-sac membrane of embryos and larvae of tilapia (Oreochromis mossambicus) in vitro and in vivo. J. Exp. Zool. 272: 419-425. Baggerman, B . 1960. Salinity preference, thyroid activity and the seaward migration of four species of Pacific salmon (Oncorhynchus). J. Fish. Res. Board Can. 17: 295-322. Bailey, J.E. 1964. Intertidal spawning of pink and chum salmon at Olsen Bay, Prince William Sound, Alaska. U.S. Fish Wildl. Serv. Biol. Lab. Auke Bay M S Rep. 64-6: 23 p. Bakalla, R.G. 1970. Synopsis of biological data on the chum salmon, Oncorhynchus keta (Walbaum). U.S. Fish. Wildl. Serv. C i r c , F A O Synopsis 41 315: 89 p. Ballard, W.W. 1973. Normal embryonic stages for salmonid fishes, based on Salmo gairdneri Richardson and Salvelinus fontinalis (Mitchill). J. Exp. Zool. 184: 7-26. Barton, B.A., and Schreck, C.B. 1987. Metabolic cost of acute physical stress in juvenile steelhead. Trans. Am. Fish. Soc. 116: 257-263. Bath, R.N., and Eddy, F.B. 1979. Ionic and respiratory regulation in rainbow trout during rapid transfer to seawater. J. Comp. Physiol. 134: 351-357. Bereiter-Hahn, J. 1976. Dimethylaminostyrylmethylpyridiniumiodine (DASPMI) as fluorescent probe for mitochondria in situ. Biochim Biophys. Acta 423: 1-14. Blackburn, J., and Clarke, W.C. 1987. Revised procedure for the 24 hour seawater challenge test to measure seawater adaptability of juvenile salmonids. Can. Tech. Rep. Fish. Aquat. Sci. 1515: 1-35.  168  Blazka, P., Volf, M . , and Ceplea, M . 1960. A new type of respirometer for the determination of the metabolism of fish in an active state. Physiol. Bohemoslov. 9: 553-560. Brett, J.R. 1962. Some considerations in the study of respiratory metabolism in fish, particularly salmon. J. Fish. Res. Board Can. 19: 1025-1038. Brett, J.R. 1964. The respiratory metabolism and swimming performance of young sockeye salmon. J. Fish. Res. Board Can. 21: 1183-1226. Brown, J.A., and Tytler, P. 1993. Hypo-osmoregulation of larvae of the turbot, Scopthalmus maximus: drinking and gut function in relation to environmental salinity. Fish Physiol. Biochem. 10: 475-483. Burggren, W.W. 1988. Role of the central circulation in regulation of cutaneous gas exchange. A m . Zool. 28: 985-998. Cech, J.J., Jr., Campagna, C.G., and Mitchell, S.J. 1979. Respiratory responses of largemouth bass (Micropterus salmoides) to environmental changes in temperature and dissolved oxygen. Trans. Am. Fish. Soc. 108: 166-171. Chan, D.O., and Woo, N.Y.S. 1978. Effect of Cortisol on the metabolism of the eel, Anguilla japonica. Gen. Comp. Endocrinol. 35: 205-215. Clarke, W.C., Shelbourn, J.E., Ogasawara, T., and Hirano, T. 1989. Effect of initial day length on growth, seawater adaptability and plasma growth hormone levels in underyearling coho, chinook, and chum salmon. Aquaculture 82: 51-62. Conceicao, L.E.C., Houlihan, D.F., and Verreth, J.A.J. 1997. Fast growth, protein turnover and costs of protein metabolism in yolk-sac larvae of the African catfish (Clarias gariepinus). Fish Physiol. Biochem. 16: 291-302. [CBE] Council of Biology Editors, C.B.E.S.M.C. (Editors) 1994. The C B E Manual for Authors, Editors, and Publishers. 6 ed. Cambridge University Press, New York. 825 p. Dalla Via, G J . 1983. Bacterial Growth and Antibiotics on Animal Respirometry. In Polarographic Oxygen Sensors. 1 ed. Edited by E. Gnaiger and H . Forstner. Springer Verlag, Berlin, p. 202-218. st  Davies, R.W., Wrona, F.J., and Kalarani, V . 1992. Assessment of activity-specific metabolism of aquatic organisms: A n improved system. Can. J. Fish. Aquat. Sci. 49: 1142-1148. de Jesus, E.G.T., and Hirano, T. 1992. Changes in whole-body concentrations of Cortisol, thyroid hormones, and sex steroids during early development of the chum salmon, Oncorhynchus keta. Gen. Comp. Endocrinol. 85: 55-61. Depeche, J. 1973. Infrastructure superficielle de la vesicule vitelline et du sac pericardique de l'embryon de Poecilia reticulata (poisson teleosteen). Application a l'etude du role des "cellules a chlorure" dans l'osmoregulation embryonnaire. Z. Zellforsch. 141: 235-253.  169  De Silva, C D . , Premawansa, S., and Keembiyahetty, C.N. 1986. Oxygen consumption in Oreochromis niloticus (L.) in relation to development, salinity, temperature and time of day. J. Fish Biol. 29: 267-277. Eddy, F.B. 1974. Osmotic properties of the perivitelline fluid and some properties of the chorion of Atlantic salmon eggs (Salmo salar). J. Zool., Lond. 174: 237-243. Eddy, F.B. 1982. Osmotic and ionic regulation in captive fish with particular reference to salmonids. Comp. Biochem. Physiol. 73B: 125-141. Eddy, F.B., and Talbot, C. 1983. Sodium balance in eggs and dechorionated embryos of the Atlantic salmon Salmo salar L. exposed to zinc, aluminium and acid waters. Comp. Biochem. Physiol. 81C: 259-266. Evans, D.H. 1993. Osmotic and ionic regulation. In The Physiology of Fishes. Edited by D.H. Evans. CRC Press, Boca Raton, p. 315-341. Febry, R., and Lutz, P. 1987. Energy partitioning in fish: the activity-related costs of osmoregulation in a euryhaline cichlid. J. exp. Biol. 128: 63-85. Flik, G., Verbost, P.M., and Wendelaar Bonga, S.E. 1995. Calcium transport processes in fishes. In Cellular and Molecular Approaches to Fish Ionic Regulation. Edited by C M . Wood and T J . Shuttleworth. (Series edited by W.S. Hoar, D J . Randall, and A.P. Farrell. Fish Physiology, 14.) Academic Press, New York. p. 317-342. Foerster, R.E. 1968. The sockeye salmon, Oncorhynchus nerka. Bull. Fish. Res. Board Can. 162: 422 p. Forstner, H . 1983. A n automated multiple-chamber intermittent-flow respirometer. In Polarographic oxygen sensors, aquatic and physiological applications. Edited by E. Gnaiger and H . Forstner. Springer-Verlag, Berlin, p. 111-126. Forstner, H., and Gnaiger, E. 1983. Appendix A : Calculation of equilibrium oxygen concentration. In Polarographic Oxygen Sensors. 1 ed. Edited by E. Gnaiger and H . Forstner. Springer-Verlag, Berlin, p. 322-333. st  Foskett, J.K., and Scheffey, C. 1982. The chloride cell: Definitive identification as the saltsecretory cell in teleosts. Science 215: 164-166. Foskett, J.K., Bern, H.A., Machen, T.E., and Conner, M . 1983. Chloride cells and the hormonal control of teleost fish osmoregulation. J. exp. Biol. 106: 255-281. Fredin, R.A., Major, R.L., Bakkala, R . G , and Tanonaka, G. 1977. Pacific salmon and the high seas salmon fisheries of Japan (Processed report). Northwest and Alaska Fisheries Center, National Marine Fisheries Service, Seattle, W A : 324 p. Fry, F.E.J. 1971. The effect of environmental factors on the physiology of fish. In Fish Physiology. Vol. 6. Edited by W.S. Hoar and D.J. Randall. Academic Press, New York. p. 198. 170  Gehrke, P.C., Fidler, L.E., Mense, D.C., and Randall, D J . 1990. A respirometer with controlled water quality and computerized data acquisition for experiments with swimming fish. Fish Physiol. Biochem. 8: 61-67. Giorgi, A . E . , and Congleton, J.L. 1984. Effect of current velocity on development and survival of lingcod, Ophiodon elongatus, embryos. Env. Biol. Fish. 10: 15-28. Goss, G., Perry, S., and Laurent, P. 1995. Ultrastructural and morphometric studies on and acidbase transport processes in freshwater fish. In Cellular and Molecular Approaches to Fish Ionic Regulation. Edited by C M . Wood and T.J. Shuttleworth. (Series edited by W.S. Hoar, D.J. Randall, and A.P. Farrell. Fish Physiology, 14.) Academic Press, New York. p. 257284. Goss, G.G., and Perry, S.F. 1994. Different mechanisms of acid-base regulation in rainbow trout (Oncorhynchus mykiss) and American eel (Anguilla rostrata) during N a H C 0 infusion. Discuss rapid covering of CCs by adjacent pavement cells. Phys. Zool. 67: 381-406. 3  Goss, G.G., Wood, C M . , Laurent, P., and Perry, S.F. 1994. Morphological responses of the rainbow trout (Oncorhynchus mykiss) gill to hyperoxia, base (NaHC0 ) and acid (HC1) infusions. Fish Physiol. Biochem. 12: 465-477. 3  Gray, I.E. 1954. Comparative study of the gill area of marine fishes. Biol. Bull. mar. Biol. Lab., Woods Hole 107: 219-225. Green, E.J., and Carritt, D.E. 1967. Oxygen solubility in sea water: thermodynamic influence of sea salt. Science 157: 191. Groot, E.P. 1989. Intertidal spawning of chum salmon: saltwater tolerance of the early life stages to actual and simulated intertidal conditions. M.Sc. Thesis, University of British Columbia, Vancouver. 130 p. Groot, E.P., and Iwama, G.K. 1994. A micro-respirometer designed for energetic studies on the early life stages of fish. In High Performance Fish - Proceedings of an International Fish Physiology Symposium, July 16-21, 1994, Vancouver, B C , Canada. Edited by D.D. MacKinlay. Fish Physiology Association, Vancouver, B C . p. 263-268. Guggino, W.B. 1980a. Salt balance in embryos of Fundulus heteroclitus and F. bermudae adapted to seawater. Amer. J. Physiol. 238: R42-R49. Guggino, W.B. 1980b. Water balance in embryos of Fundulus heteroclitus and F. bermudae in seawater. Amer. J. Physiol. 238: R36-R41. Hamor, T., and Garside, E.T. 1975. Regulation of oxygen consumption by incident illumination in embryonated ova of Atlantic salmon Salmo salar. Comp. Biochem. Physiol. 52 A : 277280. Haney, D . C , and Nordlie, F.G. 1997. Influence of environmental salinity on routine metabolic rate and critical oxygen tension of Cyprinodon variegatus. Phys. Zool. 70: 511-518.  171  Hasegawa, S., Hirano, T., Ogasawara, T., Iwata, M . , Akiyama, T., and Arai, S. 1987. Osmoregulatory ability of chum salmon, Oncorhynchus keta, reared in freshwater for prolonged periods. Fish Physiol. Biochem. 4: 101-110. Heard, W.R. 1991. Life History of Pink Salmon {Oncorhynchus gorbuscha). In Pacific Salmon Life Histories. Edited by C. Groot and L. Margolis. U B C Press, Vancouver, p. 119-230. Helle, J.H., Williamson, R.S., and Bailey, J.E. 1964. Intertidal ecology and life history of pink salmon at Olsen Creek, Prince William Sound, Alaska. U.S. Fish Wildl. Serv. Spec. Sci. Rep. Fish. 483: 26 p. Heming, T.A., and Buddington, R.K. 1988. Yolk absorption in embryonic and larval fishes. In Fish Physiology. Vol. 11 A. Edited by W.S. Hoar and D J . Randall. Academic Press, New York. p. 407-446. Herndon, T.M., McCormick, S.D., and Bern, H.A. 1991. Effects of prolactin on chloride cells in opercular membrane of seawater-adapted tilapia. Gen. Comp. Endocrinol. 83: 283-289. Holliday, F.G.T. 1963. Osmoregulation in marine teleost eggs and larvae. Report of California Cooperative Oceanic Fisheries Investigations 10: 89-95. Holliday, F.G.T., and Jones, M.P. 1965. Osmotic regulation in the embryo of the herring (Clupea harengus). J. mar. biol. Ass. U.K. 45: 305-311. Holliday, F.G.T., Blaxter, J.H.S., and Lasker, R. 1964. Oxygen uptake of developing eggs and larvae of the herring (Clupea harengus). J. mar. biol. Ass. U . K . 44: 711-723. Houlihan, D.F., and Smith, R.W. 1993. Estimating energy costs of protein synthesis, from fish cells to whole fish. In Society for Experimental Biology, Canterbury Meeting, March 29 April 2, 1993. Vol. Animal and Cell Abstracts [A10-2]. Society for Experimental Biology, Cambridge, p. 39. Hwang, P.P. 1989. Distribution of chloride cells in teleost larvae. J. Morphol. 200: 1-8. Hwang, P.P. 1990. Salinity effects on development of chloride cells in the larvae of ayu (Plecoglossus altivelis). Mar. Biol. 107: 1-7. Hwang, P.P., and Hirano, R. 1985. Effects of environmental salinity on intercellular organization and junctional structure of chloride cells in early stages of teleost development. J. Exp. Zool. 236:115-126. Hwang, P.P., and Wu, S.M. 1993. Role of Cortisol in hypo-osmoregulation in larvae of the tilapia (Oreochromis mossambicus). Gen. Comp. Endocrinol. 92: 318-324. Hwang, P.P., Tsai, Y . N . , and Tung, Y . C . 1994. Calcium balance in embryos and larvae of the freshwater-adapted teleost, Oreochromis mossambicus. Fish Physiol. Biochem. 13: 325-333. Iwata, M . , and Komatsu, S. 1984. Importance of estuarine residence for adaptation of chum salmon (Oncorhynchus keta) fry to seawater. Can. J. Fish. Aquat. Sci. 41: 744-749. 172  Jandel Scientific 1995. SigmaStat for Windows version 2.0. (San Rafael, C A ) Jensen, J.O.T. 1988. A microcomputer program for predicting embryonic development in Pacific salmon and steelhead trout. J. World Aquaculture Soc. 19: 80-81. Jones, M.P., Holliday, F.G.T., and Dunn, A.E.G. 1966. The ultra-structure of the epidermis of larvae of the herring (Clupea harengus) in relation to the rearing salinity. J. mar. biol. Ass. U . K . 46: 235-239. Kalman, S.M. 1959. Sodium and water exchange in the trout egg. J. Cell, and Comp. Physiol. 54: 155-162. Kaneko, T., Hasegawa, S., Takagi, Y . , Tagawa, M . , and Hirano, T. 1995. Hypoosmoregulatory ability of eyed-stage embryos of chum salmon. Mar. Biol. 122: 165-170. Karnaky, K.J., Jr. 1986. Structure and function of the chloride cell of Fundulus heteroclitus and other teleosts. Am. Zool. 26: 209-224. Karnovsky, M.J. 1971. Use of ferrocyanide-reduced osmium tetroxide in electron microscopy. Proc. Am. Soc. Cell Biol. : 146. (New Orleans, LA). Kashiwagi, M . , and Sato, R. 1969. Studies on the osmoregulation of the chum salmon, Oncorhynchus keta (Walbaum) I. The tolerance of eyed period eggs, alevins, and fry of the chum salmon to sea water. Tohoku J. Agric. Res. 20: 41-47. Katsura, K., and Hamada, K . 1986. Appearance and disappearance of chloride cells throughout the embryonic and postembryonic development of the goby, Chaenogobius urotaenia. Bull. Fac. Fish. Hokkaido Univ. 37: 95-100. Keys, A . B . , and Willmer, E.N. 1932. "Chloride secreting cells" in the gills of fishes, with special reference to the common eel. J. Physiol., Lond. 76: 368-378. Kirschner, L . B . 1980. Comparison of vertebrate salt-excreting organs. American Journal of Physiology 238: R219-R223. Kirschner, L . B . 1993. The energetics of osmotic regulation in ureotelic and hypoosmotic fishes. J. Exp. Zool. 267: 19-26. Kirschner, L . B . 1995. Energetics of osmoregulation in fresh water vertebrates. J. Exp. Zool. 27: 243-252. Krumschnabel, G., and Lackner, R. 1993. Stress responses in rainbow trout Oncorhynchus mykiss alevins. Comp. Biochem. Physiol. 104A: 777-783. Kultz, D., and Onken, H . 1993. Long-term acclimation of the teleost Oreochromis mossambicus to various salinities: two different strategies in mastering hypertonic stress. Mar. Biol. 117: 527-533.  173  Kiiltz, D., and Somero, G. 1995. Osmotic and thermal effects on in situ ATPase activity in permeabilized gill epithelial cells of the fish Gillichthys mirabilis. J. exp. Biol. 198: 18831894. Kiiltz, D., Bastrop, R., Jiirss, K., and Siebers, D. 1992. Mitochondria-rich (MR) cells and the activities of Na /K -ATPase and carbonic anhydrase in the gill and opercular membrane of O. mossambicus adapted to various salinities. Comp. Biochem. Physiol. 102B: 137-152. +  +  Lasker, R., and Theilacker, G.H. 1962. Oxygen consumption and osmoregulation by single Pacific sardine eggs and larvae. ICES 27: 25-33. Lasker, R., and Threadgold, L.T. 1968. 'Chloride cells' in the skin of the larval sardine. Exper. Cell Res. 52: 582-590. Laurent, P., and Perry, S.F. 1990. Effects of Cortisol on gill chloride cells morphology and ionic uptake in the freshwater trout, Salmo gairdneri. Cell Tissue Res. 240: 675-692. Laurent, P., and Perry, S.F. 1991. Environmental effects on fish gill morphology. Phys. Zool. 64: 4-25. Leatherland, J.F., and Lin, L. 1975. Activity of pituitary gland in embryo and larval stages of coho salmon, Oncorhynchus kisutch. Can. J. Zool. 53: 297-310. L i , J., Eygensteyn, J., Lock, R.A.C., Verbost, P.M., Van der Heijden, A.J.H., Wendelaar-Bonga, S.E., and Flik, G. 1995. Branchial chloride cells in larvae and juveniles of freshwater tilapia Oreochromis mossambicus. J. exp. Biol. 198: 2177-2184. L i , J., Eygensteyn, J., Lock, R.A.C., Wendelaar-Bonga, S.E., and Flik, G. 1997. N a and C a homeostatic mechanisms in isolated chloride cells of the teleost Oreochromis mossambicus analysed by confocal laser scanning microscopy. J. exp. Biol. 200: 1499-1508. +  2 +  L i em, K.F. 1981. larvae of air-breathing fishes as countercurrent flow devices in hypoxic environments. Science 211: 1177-1179. Lillywhite, H.B., and Maderson, P.F.A. 1988. The structure and permeability of integument. A m . Zool. 28: 945-962. Lin, H . , and Randall, D.J. 1993. H -ATPase activity in crude homogenates offish gill tissue: inhibitor sensitivity and environmental and hormonal regulation. J. exp. Biol. 180: 163-174. +  Lin, H . , and Randall, D.J. 1995. Proton pumps in fish gills. In Cellular and Molecular Approaches to Fish Ionic Regulation. Vol. 14. Edited by C M . Wood and T.J. Shuttleworth. (Series edited by W.S. Hoar and D.J. Randall. Fish Physiology.) Academic Press, New York, p. 229-255. Lin, H., Pfeiffer, D.C., Vogl, A.W., Pan, J., and Randall, D.J. 1994. Immunolocalization of H ATPase in the gill epithelia of rainbow trout. J. exp. Biol. 195: 169-183. +  174  MacCrimmon, H.R., and Kwain, W.H. 1969. Influence of light on early development and meristic characters in the rainbow trout, Salmo gairdneri Richardson. Can. J. Zool. 47: 631637. Madsen, S.S. 1990a. The role of Cortisol and growth hormone in seawater adaptation and development of hypo-osmoregulatory mechanisms in sea trout parr (Salmo trutta trutta). Gen. Comp. Endocrinol. 79: 1-11. Madsen, S.S. 1990b. Effect of repetitive Cortisol and thyroxine injections on chloride cell number and Na /K -ATPase activity in gill of freshwater acclimated rainbow trout, Salmo gairdneri. Comp. Biochem. Physiol. 95A: 171-176. +  +  Maetz, J. 1970. Mechanisms of salt and water transfer across membranes i teleosts in relation to the aquatic environment. Mem. Soc. Endocrinol. 18: 3-29. Mangor-Jensen, A . 1987. Water balance in developing eggs of the cod Gadus morhua L. Fish Physiol. Biochem. 3: 17-24. Mangor-Jensen, A., and Adoff, G.R. 1987. Drinking activity of newly hatched larvae of cod (Gadus morhua L.). Fish Physiol. Biochem. 3: 99-103. Marshall, W.S. 1995. Transport processes in isolated teleost epithelia: opercular epithelium and urinary bladder. In Cellular and Molecular Approaches to Fish Ionic Regulation. V o l . 14. Edited by C M . Wood and T.J. Shuttleworth. (Series edited by W.S. Hoar and D.J. Randall. Fish Physiology.) Academic Press, New York. p. 1-23. Marshall, W.S., Bryson, S.E., Burghardt, J.S., and Verbost, P . M . 1995. C a transport by opercular epithelium of the fresh water adapted euryhaline teleost, Fundulus heteroclitus. J. Comp. Physiol. 165: 268-277. 2 +  McCormick, S.D. 1990a. Fluorescent labelling of Na ,K -ATPase in intact cells by use of a fluorescent derivative of ouabain: Salinity and teleost chloride cells. Cell Tissue Res. 260: 529-533. +  +  McCormick, S.D. 1990b. Cortisol directly stimulates differentiation of chloride cells in tilapia opercular membrane. American Journal of Physiology 259: R857-R863. McCormick, S.D. 1993. Methods for non-lethal gill biopsy and measurement of Na ,K -ATPase activity. Can. J. Fish. Aquat. Sci. 50: 656-658. +  +  McCormick, S.D. 1995. Hormonal Control of Gill Na ,K -ATPase and chloride cell function. In Cellular and Molecular Approaches to Fish Ionic Regulation. Edited by C M . Wood and T.J. Shuttleworth. (Series edited by W.S. Hoar, D.J. Randall, and A.P. Farrell. Fish Physiology, 14.) Academic Press, New York. p. 285-315. +  +  McCormick, S.D., and Bern, H.A. 1989. In vitro stimulation of Na -K -ATPase activity and ouabain binding by Cortisol in coho salmon gill. American Journal of Physiology 256: R707R715. +  +  175  McCormick, S.D., Hasegawa, S., and Hirano, T. 1992. Calcium uptake in the skin of a freshwater teleost. Proc. Natl. Acad. Sci. U S A 89: 3635-3638. McCormick, S.D., Saunders, R.L., Henderson, E.B., and Harmon, P.R. 1987. Photoperiod control of parr-smolt transformation in Atlantic salmon (Salmo salar. changes in salinity tolerance, gill Na ,K -ATPase activity, and plasma thyroid hormones. Can. J. Fish. Aquat. Sci. 44: 1462-1468. +  +  McLean, W.E., Jensen, J.O.T., and Rombough, P.J. 1991. Microcomputer models for salmonid hatcheries. American Fisheries Society Symposium 10: 516-528. Morgan, J.D. 1997. Energetic Aspects of Osmoregulation in Fish. Ph.D. Dissertation, University of British Columbia, Vancouver. 156 p. Morgan, J.D., and Iwama, G.K. 1991. Effects of salinity on growth, metabolism, and ion regulation in juvenile rainbow and steelhead trout (Oncorhynchus mykiss) and fall chinook salmon (Oncorhynchus tshawytscha). Can. J. Fish. Aquat. Sci. 48: 2083-2094. Morgan, J.D., and Iwama, G.K. 1996. Cortisol-induced changes in oxygen consumption and ionic regulation in coastal cutthroat trout parr. Fish Physiol. Biochem. 15: 385-394. Morgan, J.D., Jensen, J.O.T., and Iwama, G.K. 1992. Effects of salinity on aerobic metabolism and development of eggs and alevins of steelhead trout (Oncorhynchus mykiss) and fall chinook salmon (Oncorhynchus tshawytscha). Can. J. Zool. 70: 1341-1346. Moyle, P.B., and Cech, J.J., Jr. 1996. Fishes: A n Introduction to Ichthyology. 3 ed. Prentice Hall, Toronto. 590 p. rd  Nagibina, G.V. 1983. Influence of salinity on the metabolic rate in coho salmon (Oncorhynchus kisutch) in early periods of ontogenesis. In Physiological Basis of Reproduction of Marine and Anadromous Fish. Edited by L A . Shekhanova, L . B . Klyashtorin, and E.A. Kamenskaya. VNIRO, Moscow, Russia, p. 57-61. Nordlie, F.G. 1978. The influence of environmental salinity on respiratory oxygen demands in the euryhaline teleost, Ambassis interrupta Bleeker. Comp. Biochem. Physiol. 59A: 271-274. Nordlie, F . G , Walsh, S.J., Haney, D.C., and Nordlie, T.F. 1991. The influence of ambient salinity on routine metabolism in the teleost Cyprinodon variegatus Lacepede. J. Fish Biol. 38: 115-122. Pannevis, M.P., and Houlihan, D.F. 1992. The energetic cost of protein synthesis in isolated hepatocytes of rainbow trout (Oncorhynchus mykiss). J. Comp. Physiol. 162: 393-400. Penefsky, H.S., and Bruist, M.F. 1984. Adenosinetriphosphatases. In Methods of Enzymatic Analysis. Vol. 4. Edited by H.U. Bergmeyer. Verlag Chemie, Weinham. p. 324-328. Perez-Pinzon, M . A . , and Lutz, P.L. 1991. Activity related cost of osmoregulation in the juvenile snook (Centropomus undecimalis). Bull. Mar. Sci. 48: 58-66.  176  Perry, S.F., and Laurent, P. 1989. Adaptational responses of rainbow trout to lowered external NaCI concentration: contribution of the branchial chloride cell. J. exp. Biol. 147: 147-168. Perry, S.F., and Wood, C M . 1985. Kinetics of branchial calcium uptake in the rainbow trout: Effects of acclimation to various external calcium levels. J. exp. Biol. 116: 411-433. Perry, S.F., Goss, G.G., and Laurent, P. 1992. The interrelationships between gill chloride cell morphology and ionic uptake in four freshwater teleosts. Can. J. Zool. 70: 1775-1786. Peterson, R.H. 1975. Pectoral fin and opercular movements of Atlantic salmon (Salmo salar) alevins. J. Fish. Res. Board Can. 32: 643-648. Peterson, R.H., and Martin-Robichaud, D J . 1983. Embryo movements of Atlantic salmon (Salmo salar) as influenced by pH, temperature, and state of development. Can. J. Fish. Aquat. Sci. 40: 777-782. Pisam, M . , Caroff, A., and Rambourg, A . 1987. Two types of chloride cells in the gill epithelium of a freshwater-adapted euryhaline fish: Lebistes reticulatus; Their modification during adaptation to saltwater. Am. J. Anat. 179: 40-50. Potts, W.T.W. 1954. The energetics of osmotic regulation in brackish- and fresh-water animals. J. exp. Biol. 31:618-630. Potts, W.T.W., and Rudy, P.P., Jr. 1969. Water balance in the eggs of the Atlantic salmon Salmo salar. J. exp. Biol. 50: 223-237. Randall, D.J., and Wright, P.A. 1995. Circulation and Gas Transfer. In Physiological Ecology of Pacific Salmon. Edited by C. Groot, L. Margolis, and W . C Clarke. U B C Press, Vancouver, p. 510. Rao, G . M . M . 1968. Oxygen consumption of rainbow trout (Salmo gairdneri) in relation to activity and salinity. Can. J. Zool. 46: 781-786. Richman, N . H . , and Zaugg, W.S. 1987. Effects on Cortisol and growth hormone on osmoregulation in pre- and desmoltified coho salmon (Oncorhynchus kisutch). Gen. Comp. Endocrinol. 65: 189-198. Richman, N.H., Tai de Diaz, S., Nishioka, R.S., Prunet, P., and Bern, H.A. 1987. Osmoregulatory and endocrine relationships with chloride cell morphology and density during smoltification in coho salmon (Oncorhynchus kisutch). Aquaculture 60: 265-285. Rombough, P.J. 1985. Initial egg weight, time to maximum alevin wet weight, and optimal ponding times for chinook salmon (Oncorhynchus tshawytscha). Can. J. Fish. Aquat. Sci. 42: 287-291. Rombough, P.J. 1986. Mathematical model for predicting the dissolved oxygen requirements of steelhead (Salmo gairdneri) embryos and alevins in hatchery incubators. Aquaculture 59: 119-137.  177  Rombough, P.J. 1988a. Respiratory gas exchange, aerobic metabolism, and effects of hypoxia during early life. In Fish Physiology. Vol. 11 A. Edited by W.S. Hoar and D.J. Randall. Academic Press, New York. p. 59-161. Rombough, P.J. 1988b. Growth, aerobic metabolism, and dissolved oxygen requirements of embryos and alevins of steelhead, Salmo gairdneri. Can. J. Zool. 66: 651-660. Rombough, P.J. 1994. Energy partitioning during fish development: additive of compensatory allocation of energy to support growth? Funct. Ecol. 8: 178-186. Rombough, P.J., and Moroz, B . M . 1990. The scaling and potential importance of cutaneous and branchial surfaces in respiratory gas exchange in young chinook salmon Oncorhynchus tshawytscha. J. exp. Biol. 154: 1-12. Rombough, P.J., and Moroz, B . M . 1997. The scaling and potential importance of cutaneous and branchial surfaces in respiratory gas exchange in larval and juvenile walleye Stizostedion vitreum. J. exp. Biol. 200: 2459-2468. Rombough, P.J., and Ure, D. 1991. Partitioning of oxygen uptake between cutaneous and branchial surfaces in larval and young juvenile chinook salmon Oncorhynchus tshawytscha. Phys. Zool. 64: 717-727. Rudy, P.P., Jr., and Potts, W.T.W. 1969. Sodium balance in the eggs of the Atlantic salmon, Salmo salar. J. exp. Biol. 50: 239-246. Salo, E.O. 1991. Life History of Chum Salmon (Oncorhynchus keta). In Pacific Salmon Life Histories. Edited by C. Groot and L. Margolis. U B C Press, Vancouver, p. 231-309. Sandercock, F.K. 1991. Life History of Coho Salmon (Oncorhynchus kisutch). In Pacific Salmon Life Histories. Edited by C. Groot and L. Margolis. U B C Press, Vancouver, p. 395-445. Sargent, J.R., Thomson, A.J., and Bornancin, M . 1975. Activities and localization of succinic dehydrogenase and Na /K -activated adenosine triphosphatase in the gills of freshwater and sea water eels (Anguilla anguilla). Comp. Biochem. Physiol. 51B: 75-79. +  +  Satoh, S., Takeuchi, T., and Watanabe, T. 1987. Changes of mineral compositions in whole body of rainbow trout during growing stages. Nippon Suisan Gakkaishi 53: 273-279. Schmidt-Nielsen, K . 1975. Animal Physiology: Adaptation and Environment. Cambridge University Press, London. 560 p. Schrock, R.B., Beeman, J.W., Rondorf, D.W., and Haner, P.V. 1994. A microassay for gill sodium, potassium-activated ATPase in juvenile Pacific salmonids. Trans. Am. Fish. Soc. 123: 223-229. Shelbourne, J.E. 1957. Site of chloride regulation in marine fish larvae. Nature (Lond.) 108: 920922.  178  Shen, A . C . Y . , and Leatherland, J.F. 1978a. Structure of the yolksac epithelium and gills in the early developmental stages of rainbow trout (Salmo gairdneri) maintained in different ambient salinities. Env. Biol. Fish. 3: 345-354. Shen, A . C . Y . , and Leatherland, J.F. 1978b. Effect of ambient salinity on ionic and osmotic regulation of eggs, larvae, and alevins of rainbow trout (Salmo gairdneri. Can. J. Zool. 56: 571-577. Shepard, K . L . 1987. Ion-exchange phenomena regulate the environment of embryos in the eggs of freshwater fish. Comp. Biochem. Physiol. 88A: 659-662. Shepard, K . L . , and McWilliams, P.G. 1989. Ionic regulation by the eggs of salmon. J. Comp. Physiol. 159: 249-254. Shumway, D.L., Warren, C.E., and Doudoroff, P. 1964. Influence of oxygen concentration and water movement on the growth of steelhead trout and coho salmon embryos. Trans. A m . Fish. Soc. 93: 342-356. Shuttleworth, T.J., and Freeman, R.F.H. 1973. The role of gill in seawater adaptation in Anguilla dieffenbachii I. Osmotic and ionic composition of the blood and gill tissue. J. Comp. Physiol. 86: 293-313. Silva, P., Solomon, R., Spokes, K., and Epstein, F.H. 1977. Ouabain inhibition of gill Na-KATPase: Relationship to active chloride transport. J. Exp. Zool. 199: 419-426. Silver, S.J., Warren, C.E., and Doudoroff, P. 1963. Dissolved oxygen requirements of developing steelhead trout and chinook salmon embryos at difference water velocities. Trans. Am. Fish. Soc. 92: 327-343. Skadhauge, E., and Lotan, R. 1974. Drinking rate and oxygen consumption in the euryhaline teleost Aphanius dispar in waters of high salinity. J. exp. Biol. 60: 547-556. Smith, L.S., and Newcomb, T.W. 1970. A modified version of the Blazka respirometer and exercise chamber for large fish. J. Fish. Res. Board Can. 27: 1321-1324. Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A . K . , Gartner, F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N . M . , Olson, B.J., and Klenk, D.C. 1985. Measurement of protein using bicinchoninic acid. Anal. Biochem. 150: 76-85. Soengas, J.L., Aldegunde, M . , and Andres, M . D . 1995. Gradual transfer to sea water of rainbow trout: effects on liver carbohydrate metabolism. J. Fish Biol. 47: 466-478. Somasundaram, B . 1985. Effects of zinc on epidermal ultrastructure in the larva of Clupea harengus. Mar. Biol. 85: 199-207. Stiffler, D.F. 1988. Cutaneous exchange of ions in lower vertebrates. Am. Zool. 28: 1019-1029. Sullivan, C V . Dickhoff, W.W., Brewer, S.D., and Johnston, G P . 1983. Plasma thyroid-hormone concentrations and gill (Na+K)-ATPase activities in post-emergent pink salmon. Trans. A m . Fish. Soc. 112: 825-829. 179  Swanson, C. 1996. Early development of milkfish: effects of salinity in embryonic and larval metabolism, yolk absorption, and growth. J. Fish Biol. 48: 405-4221. Thorasteinson, F.V., Helle, J.H., and Birkholz, D.G. 1971. Salmon survival in intertidal zones of Prince William Sound streams in uplifted and subsided area. In The Great Alaska Earthquake of 1964: Biology. National Academy of Sciences, Washington, p. 194-219. Tschaplinski, P J . 1987. The use of estuaries as rearing habitats by juvenile coho salmon. In Proceedings of the workshop: applying 15 years of Carnation Creek results. Edited by T.W. Chamberlin. Carnation Creek Steering Committee, Pacific Biological Station, Dept. of Fisheries and Oceans, Canada, Nanaimo, B C . p. 123-142. Tytler, P., and Bell, M . V . 1989. A study o f diffusional permeability of water sodium and chloride in yolk-sac larvae of cod (Gadus morhua L.). J. exp. Biol. 147: 125-132. Tytler, P., and Blaxter, J.H.S. 1988. Drinking in yolk-sac stage larvae of the halibut, Hippoglossus hippoglossus (L.). J. Fish Biol. 32: 493-494. Uchida, K., and Kaneko, T. 1996. Enhanced chloride cell turnover in the gills of chum salmon fry in seawater. Zool. Sci. 13: 655-660. Uchida, K., Kaneko, T., Yamaguchi, A., Ogasawara, T., and Hirano, T. 1997. Reduced hypoosmoregulatory ability and alteration of gill chloride cell distribution in mature chum salmon (Oncorhynchus keta) migrating upstream for spawning. Mar. Biol. 129: 247-253. Uchida, K., Kaneko, T., Yamauchi, K., and Hirano, T. 1996. Morphometrical analysis of chloride cell activity in the gill filaments and lamellae and changes in Na ,K -ATPase activity during seawater adaptation in chum salmon fry. J. Exp. Zool. 276: 193-200. +  +  Velsen, F.P.J. 1980. Embryonic development in eggs of sockeye salmon, Oncorhynchus nerka. Can. Spec. Publ. Fish. Aquat. Sci. 49: 19 p. Velsen, F.P.J. 1987. Temperature and incubation in Pacific salmon and rainbow trout: compilation o f data on median hatching time, mortality, and embryonic staging. Canadian Data Report of Fisheries and Aquatic Sciences 626: 58 p. Vijayan, M . M . , Reddy, P.K., Leatherland, J.F., and Moon, T.W. 1994. The effects of Cortisol on hepatocyte metabolism in rainbow trout: a study using the steroid analogue RU486. Gen. Comp. Endocrinol. 96: 75-84. Wales, W., and Tytler, P. 1996. Changes in chloride cell distribution during early larval stages of Clupea harengus. J. Fish Biol. 49: 801-814. Walsh, W.A., and Lund Jr., W.A. 1989. Oxygen consumption by eggs of the grubby, Myoxocephalus aenaeus, and the longhorn sculpin, Myoxocephalus octodecemspinosus. Can. J. Zool. 67: 1613-1619.  180  Walsh, W.A., Swanson, C , and Lee, C.-S. 1991. Effects of development, temperature and salinity on metabolism in eggs and yolk-sac larvae of milkfish, Chanos chanos (Forsskal). J. Fish Biol. 39: 115-125. Walsh, W.A., Swanson, C , Lee, C.-S., Banno, J.E., and Eda, H . 1989. Oxygen consumption by eggs and larvae of striped mullet, Mugil cephalus, in relation to development, salinity and temperature. J. Fish Biol. 35: 347-358. Weisbart, M . 1968. Osmotic and ionic regulation in embryos, alevins, and fry of the five species of Pacific salmon. Can. J. Zool. 46: 385-397. Wendelaar Bonga, S.E., and van der Meij, C.J.M. 1989. Degeneration and death, by apoptosis and necrosis, of the pavement and chloride cells in the gills of the teleost Oreochromis mossambicus. Cell Tissue Res. 255: 235-243. Wieser, W., Forstner, H., Medgyesy, N . , and Hinterleitner, S. 1988. To switch or not to switch: partitioning of energy between growth and activity in larval cyprinids (Cyprinidae: Teleostei). Funct. Ecol. 2: 499-507. Withers, P.C. 1992. Comparative Animal Physiology. Saunders College Publishing, Orlando. 949 p. Witters, H., Berckmans, P., and Vangenechten, C. 1996. Immunolocalization of Na ,K -ATPase in the gill epithelium of rainbow trout, Oncorhynchus mykiss. Cell Tissue Res. 283: 461-468. +  +  Wood, C C , Riddell, B.E., and Rutherford, D.T. 1987. Alternative juvenile life histories of sockeye salmon (Oncorhynchus nerka) and their contribution to production in the Stikine River, northern British Columbia. In Sockeye Salmon (Oncorhynchus nerka) population biology and future management. Edited by Ff.D. Smith, L . Margolis, and C C Wood. ( Can. Spec. Publ. Fish. Aquat. Sci, 96.),. p. 12-24. Wrona, F J . , and Davies, R.W. 1984. A n improved flow-through respirometer for aquatic macroinvertebrate bioenergetic research. Can. J. Fish. Aquat. Sci. 41: 380-385. Zadunaisky, J.A. 1984. The chloride cell: the active transport of chloride and the paracellular pathways. In Fish Physiology. Vol. 10B. Edited by W.S. Hoar and D.J. Randall. Academic Press, New York. p. 129-176. Zar, J.H. 1984. Biostatistical Analysis. 2  n d  ed. Prentice-Hall, Toronto. 718 p.  Zaugg, W.S. 1982. A simplified preparation for adenosine triphosphatase determination in gill tissue. Can. J. Fish. Aquat. Sci. 39: 215-217. Zinichev, V . V . 1990. Effect of sharp temperature change on oxygen consumption during early ontogeny in Pacific salmon. J. Ichthyol. 30: 159-164. Zydlewski, J., and McCormick, S.D. 1997. The ontogeny of salinity tolerance in the American shad, Alosa sapidissima. Can. J. Fish. Aquat. Sci. 54: 182-189.  181  APPENDIX A A MLCRORESPIROMETER D E S I G N E D FOR E N E R G E T I C STUDIES OF FISH E M B R Y O S A N D L A R V A E  3  Introduction Accurate and precise measurement of metabolism in aquatic organisms (e.g., fish and invertebrates) is fundamental to bioenergetic research. Most commonly metabolism is measured using respirometry, where oxygen consumption rate is used as an indirect measurement of aerobic energy utilization. Many bioenergetic studies focussing on the juvenile and adult life stages have utilized swim-type respirometers that allow control and standardization of the test animal's activity level and also provide information on the scope for activity (Blazka et al. 1960, Brett 1964, Smith and Newcomb 1970, Gehrke et al. 1990). Measurement of oxygen consumption rates during the early life stages however, requires a different design specific to the small size and various developmental stages. These requirements present a different set of challenges that in addition to the usual considerations common to juvenile and adult respirometry studies (e.g., constant physical and biotic conditions, intra- and inter-individual variation, temporal variation, etc.), include considerations related to the small size of scale (e.g., small size of test organisms, low total oxygen consumption of individual organisms, fast rate of development, gas permeability of respirometer materials, boundary layers, etc.) (Rombough 1988).  3 A version of this appendix has been published as: Groot, E. P. and Iwama, G. K. 1994. High Performance Fish; proceedings of an International Fish Physiology Symposium, July 16-21, 1994, Vancouver, BC, Canada, p.263-268.  182  In view of these specific considerations, oxygen consumption rates of small aquatic organisms such as fish embryos and larvae, and macroinvertebrates, have been measured using numerous system designs that have employed a variety of methods of measurement, including, manometric (Warburg- or Gilson-type respirometers), Cartesian diver, Winkler, and polarographic oxygen electrodes (Rombough 1988). System designs include both 'closed' and 'open' (flow-through) respirometers (Forstner 1983, Rombough 1988, Wrona and Davies 1984). Closed systems tend to be less complex (and less expensive) than flow through ones and as a result are more commonly used. In their simplest configuration closed respirometer systems measure the difference in oxygen content between an initial and final reading over a known period time. Some of the manometric systems (Warburg or Gilson) that use agitation, may result in biased oxygen consumption rates due to variable responses of the test organisms to the disturbance caused by the mixing; both decreased and increased rates have been reported, depending on the species (Rombough 1988). More complex closed systems incorporate an oxygen electrode in-situ and thus provide continuous monitoring of the decreasing oxygen content in the respirometer. One of the past constraints with in-situ electrodes was the inherent oxygen consumption of the electrode itself as part Of the measurement process. As a result it was necessary to ensure a relatively strong and constant flow of water past the electrode membrane to provide accurate readings. This need for a stirring mechanism can cause additional problems related to disturbance of the test animals and water pressure changes that affect electrode stability. To this end, Rombough (1988) recommended low speed gear pumps to avoid this and other pumping related problems. With the advent of reliable and low oxygen consumption polarographic oxygen electrodes, continuously monitored closed systems and flow-through systems have become more common. Although the latter tend to be more complex and sensitive to changes in electrode stability (drift) they can  183  provide valuable long term, 'real-time' measurements (Davies et al. 1992). Since each type of system tends to have its strengths and weaknesses (Forstner 1983), the appropriateness of one system over another depends a lot on the specific objectives of the study and the logistical and economic constraints faced by an investigator. One of the most common criticisms raised against closed respirometry systems is the potential effect of water quality deterioration as the test progresses as result of hypoxia and accumulating metabolic byproducts. Although these are valid concerns and need to be considered, Rombough (1988) pointed out that embryos and larvae tend not to be especially sensitive to these byproducts and small reductions in oxygen content. As result many researchers do not allow oxygen concentrations to decrease below 80% of full air saturation (Forstner 1983, Morgan and Iwama 1991). Further, continuous monitoring of oxygen concentration inside a closed respirometer with in situ oxygen electrodes allows assessment of any changes in oxygen consumption rate that may occur as a result metabolic byproducts or hypoxia induced changes during the measurement period. Once again it depends on the specific purpose of the study to determine the most suitable respirometer design. For example, in his closed respirometer system Rombough (1988b) utilized the change in slope of the oxygen consumption rate of steelhead embryos and larvae as an indicator of the 'critical' oxygen concentration. Critical oxygen concentration is defined as the point at which animals switch from oxygen independent to oxygen dependent aerobic metabolism (Fry 1957, Cech et al. 1979) and in this case was characterized by an inflection point on the chart recorder trace that signified a decrease in the consumption rate. Another logistical consideration that must be controlled for in bioenergetic studies involving the early life stages of animals is the rapid ontogenic changes that occur during these early developmental stages. To avoid biases resulting from temporal variation replicate measurements 184  need to be conducted preferably simultaneously or at least within a few in replicate respirometers. Closely related to this is the need to account for background or control changes in oxygen content. Both microbial respiration and electrode 'drift' can introduce significant biases (Forstner 1983). Dalla Via (1983) reviewed microbial growth and the use of antibiotics in respirometry, and reported that microbial activity could comprise as much as 30% of the total oxygen consumed at higher temperatures (20°C). This effect was greatly reduced at lower temperatures (10°C). He suggested that the most reliable procedure to account for microbial respiration was to allow the microbial growth to reach a constant level (within 12-20 h) in the respirometer system and then simply conduct the appropriate controls to account for the microbial contribution. One other source of microbial respiration that is often overlooked is the epibiotic contamination of the test organisms, especially eggs (Rombough 1988). In the case of lingcod, a species that produces large and very robust egg masses, Giorgi and Congleton (1984) showed that 10-24% of the total measured respiration of lingcod eggs was due to microbial growth on the egg mass itself. This Appendix describes the design, construction and operation of a four chamber, closed microrespirometer that is interfaced with a computerized data acquisition system. Although the system was originally designed for measuring the oxygen consumption of the fish eggs and larvae, the design is applicable to a wide variety of small and delicate aquatic organisms.  185  Some General Considerations for Energetics Studies Energy utilization or metabolism in fish is most commonly estimated using indirect calorimetry by measuring oxygen consumption of the whole animal. Because oxygen is the only element that aerobic animals use for respiration, any measurement of oxygen consumption provides an accurate estimate of oxygen consumption rate, as long no anaerobic respiration occurs (Brett 1962). However, Fry (1971) cautioned that it cannot be assumed that all fish are obligate anaerobes. This is also true for the early life stages, especially if high levels of activity are involved (Rombough 1988a). In many studies the terms 'oxygen consumption rate' and 'metabolic rate' are used interchangeably. However, I will use the term oxygen consumption rate to avoid any potential ambiguity or inaccuracy, especially given the relatively limited understanding of anaerobic metabolism in the early life stages of fish (Rombough 1988a). Bioenergetic studies offish that use measures of oxygen consumption as the basis for determining metabolic rate and the energetic relationships of various biological processes have reported oxygen consumption data in wide a variety of different formats. Some of these formats are confusing and inconsistent with respect to standards or conventions common to the discipline of physiology as a whole. Unfortunately, there is no one location where it is possible to obtain a standardized or definitive reference source on conventions and abbreviations for use in scientific writing. The C B E Manual (1994) and Gnaiger (1983) list a number of useful tables as well as cite additional references for specific areas of study. Oxygen consumption rate is usually reported as either the volume (L, mL, uL), mass (kg, g, mg, ug), or molar amount (mol, mmol, umol, nmol) of 0 2 consumed per unit of tissue (whole organism, kg, g, mg, \ig) per unit of time (d, h, min). Common abbreviations for oxygen consumption data include V02 and M02- Although the use of the 'overdot' symbol was  186  traditionally used to designate a rate (per unit time) measurement, it is often omitted from published manuscripts since it is not easily produced by most word processing programs. In this thesis I will use M02 , without the 'overdot', to designate mass-specific oxygen consumption rate. The V02 abbreviation is common to respiratory physiology and represents the volume of oxygen consumed per unit time (e.g., mL C»2 niin~l) (CBE Manual 1994), whereas M02 -  represents the mass of oxygen consumed per unit time (e.g., mg 0 2 m i n l ) . However, some -  -  confusion appears to exist with this latter abbreviation. The ' M ' in M o refers to the mass of 2  oxygen and not the mass of the test organism, i.e., M02 does not represent a organism massspecific O2 consumption measurement, unless of course it is stated as such (e.g., M02/M = 250 u.g 0 2 g l ' h 1). Mass-specific oxygen consumption can also be represented in units of the -  —  —  volume of O2 consumed (e.g., V02/M = 175 ul C>2 g~^h !). ' M o ' is also used as an -  -  2  abbreviation for oxygen consumption measurements based on molar quantity of oxygen (e.g., umol 02 g~^min 1). A much less commonly used symbol that also represents the amount of -  —  oxygen is 'N0 ', which is based on the SI Unit that relates to the amount of substance (i.e., 2  mole) Gnaiger (1983). Forstner (1983) provides a useful set of conversion factors for converting between mass and volume of oxygen in addition to a very comprehensive treatise on the calculation of oxygen solubilities under a variety of environmental conditions. One final abbreviation that is also sometimes used to report oxygen consumption rate in energetic studies, is Q02- It appears to be used in reference to the organism mass-specific oxygen consumption rate (Almatar 1984, De Silva et al. 1986, Walsh and Lund 1989, Swanson 1996). A number of additional points concerning weight measurements are worthy of mention and apply more specifically to early life history studies. These are largely a reiteration of similar 1  statements made by Rombough (1988a) in his review on aerobic metabolism in the early life of 187  fishes. In order to allow comparison of oxygen consumption rate data between different species, life stages, and experimental treatments it is useful to use mass-specific O2 consumption, or metabolic intensity, as opposed to, or at least in addition to O2 consumption rates based on whole eggs or larvae. For the most part, mass-specific data effectively remove the very significant influence of different body sizes on oxygen consumption rate. Measures of embryonic or larval mass should, when it is feasible, exclude yolk material in order to more accurately represent the metabolically active tissue. Failure to do so can significantly underestimate metabolic rates when non-respiring tissues are still large during the earlier stages of development (Rombough 1988a). Dry weights also provide a less variable measure of actual tissue weight than wet weights do, especially i f experimental treatments may result in osmotic water loss or gain.  188  Respirometer Design and Operation  Respirometer design The whole respirometer system consists of four isolated and separately controlled respirometer loops connected to two dual-channel oxygen meters (Cameron Instruments® O M 200) that direct output to a computerized data acquisition system (Fig. A . l ) . The loops are immersed in a temperature controlled and insulated water bath that is maintained at temperature by a water circulator. Each loop consists of a central glass respirometer chamber (custom made, Canadian Scientific Glass Blowing, Richmond, BC), an oxygen micro-electrode (Microelectrodes® Inc.), a 24 volt direct current (VDC) gear micro-pump (Micro-Pumps®), a 4-way stainless steel valve (Whitey®), interconnecting stainless steel tubing (type 316, 3.2 mm inner diameter (I.D.)) and vinyl tubing (Tygon®, 2.4 mm I.D., 0.8mm wall thickness) and stainless steel fittings (Swagelock®) (Fig. A . l ) . By turning the 4-way valve handle 90° between the two available positions, flow direction is altered between 'calibration' mode (flow-through) and 'measurement' mode (closed). In the latter mode water continues to circulate through the respirometer loop as it travels from the pump, through the 4-way valve, into and through the respirometer chamber, past the oxygen micro-electrode back to the pump to complete the loop. Short pieces of vinyl tubing are used to connect the glass respirometer and the oxygen electrode to the steel tubing of the respirometer loop. These connections provide a simple and reliable system that allows visual confirmation that the stainless steel tubing is flush against the glass tubing of the respirometer chamber or the oxygen electrode fitting (< 1 mm gap between the two types of tubing) and that no air bubbles are present in the connection. Vinyl has a very low gas permeability coefficient for flexible tubing and thus results in negligible gas diffusion.  189  Flow rate in the loop is controlled by adjusting the pump speed via a 24 V D C variable speed controller and power supply (Cole Parmer®). Pump speeds were individually adjusted to equalize the flow rates between respirometer loops. Each glass respirometer chamber consists of three subunits; two funnel-shaped end sections and a cylindrical centre section. One of the end sections and the centre section were permanently sealed together to create the 'bottom section'. The remaining 'top section' is removable and allows access to the centre animal chamber (Fig. A.2). There are two shallow depressions (0.5 mm) ground out of the inside edge of the top and bottom flange of the centre section to provide clearance for the two stainless steel screens (200 x 200 um holes) when the chamber is completely assembled. The screens isolate the test specimens in the middle section. The flanged joining surfaces of the top and bottom sections are coated with a very thin and smooth layer of silicone sealant to act as a gasket. The chamber is closed by inserting the remaining screen between the top and bottom sections, and joining them together. The chamber is finally sealed for testing by submerging it completely under water and fitting the whole unit into a one piece plastic clamp (Keck®), made originally for glass ball-joint connections (Fig. A.2). The chambers are custom designed and constructed to fit into these clamps. Two different sizes of chambers have been constructed to date; 8 and 16 mL volume chambers. The glass chamber is incorporated into the respirometer loop by butt joining each of the glass tube ends to the stainless steel tubing of the respirometer loop. These joints are held in place with two sections of vinyl tubing that serve as sleeves (Fig. A.2). Total volumes of the individual loops were determined gravimetrically and were about 12 mL and 20 mL with the 8 and 16 mL chambers attached respectively.  190  G co  O  H  13  i-i  o O  ,  ca 3  3 cr o ci  o ^4H  l-l O CO  - u  l-i  e  CD  (D CO CD  CD  3  I O  o CD  CO JZ  H  CD  e3g  C -K U  CD o  CD  CD  *a  A  - 2  B  O  I-i  CD  J  bo o  1  a^ CD  O  to  CD  co CD —  S O  CD  S  CO  ii o  » s4-— co CD  JZ  S •<-> «o  ^=1  U2  -1  fi 2 -3 "S to s 8\l a 2 £ s1 2 ^ CD  I&  o  I-I CD  I  a  *  I  co *7  2 .2 g 5 3  o  CD  S3 CD  CD  o  » .9 CD  S  a£  Each respirometer loop is supported on a metal rod (65 cm long, 1.3 cm diameter) that is secured vertically into a pre-drilled hole in a 1.2 m long piece of wood placed behind the tank. A metal laboratory clamp with a swivel, is attached to the supporting rod and is used to clasp the respirometer micro-pump (Fig. A . l ) . The four rods are spaced 20 cm apart. Constant test temperatures are maintained by immersing the four respirometer loops and two test water reservoirs in a temperature controlled, styrofoam insulated glass aquarium (160 L total volume) filled with freshwater (Fig. A . l ) . Temperature control in the water bath is maintained by a 'closed circuit' heat exchange system. A water circulator pumps a mixture of water and ethylene glycol (50/50) through eight stainless steel heat exchange coils (316 stainless, 3 m total uncoiled length, 7.5 mm I.D.) that are interconnected with clear vinyl tubing and placed on the bottom of the tank. Temperature gradients in the tank were minimized by vigorously mixing the water using a 40 L/min pump located externally. Water temperatures are monitored with an immersed thermistor connected to a digital meter (Cole-Parmer®) that is interfaced with the computer. Respirometer trials have been conducted at temperatures ranging from 6.0 to 24.0°C. Water temperatures vary less than 0.1 °C during respirometer tests ranging from one to six hours.  Respirometer operation Prior to conducting a respirometer test run all of the air bubbles were removed from the individual loops and the 0 electrodes were calibrated to 0 and 100% oxygenated water. 2  Bubbles were removed from each loop by increasing pump speed to maximum and flushing with water (approx. 60 mL/min) while rotating the whole loop 180 degrees in a clockwise direction and then a counter-clockwise direction.  192  To Pump Controller To Oxygen Meter  Micro-pump 4-way Valve Supply Tube From Test Water Reservoir  Oxygen Micro-electrode Stainless Steel tubing  Plastic Clamp With Handle  Drain Tube  Respirometer Chamber  Vinyl Tubing Connector  Fig. A.2. Schematic diagram of respirometer loop showing the configuration of the glass respirometer chamber, oxygen micro-electrode, micro-pump and 4-way valve. Water flow rate is regulated by the pump controller and flow direction is altered between 'flow-through' (calibration) and 'closed' (measurement) modes via the 4-way valve.  Top Section (removable) Plastic Clamp With Handle  Screen (removable) Glass Flange With Gasket Bottom Section  y  Vinyl Tubing Connector  2 cm  V Fig. A.3. Schematic close-up view of custom made glass respirometer chamber (approx. 16 mL total volume). The middle and bottom sections are permanently sealed to facilitate loading of the chambers. Test organisms are placed in the middle section and are confined to this section by screens located between the two end subunits. The chamber is incorporated into the respirometer loop as shown in Fig. A . 1.  194  The 0 electrodes were calibrated by drawing calibration water through each of the four 2  respirometer loops from one of two, 4 L reservoirs containing either 0 or 100% oxygenated water. These are maintained at saturation by bubbling with either air or nitrogen gas, respectively. The tubes that supply water from the 0% oxygen water reservoir are constructed of stainless steel. [Use of flexible tubing (vinyl) allows reoxygenation of the water by diffusion through the tubing walls as it passes through the cooling tank before entering the respirometer loop.] Supply tubes for the 100% calibration water are made of vinyl (2.4 mm ID). The electrodes are calibrated to 0% first and then to 100%. In preparation for a respirometer test run individual respirometer chambers were loaded with test organisms in a separate, small, insulated, and temperature controlled loading tank containing water of similar temperature and salinity as the test water circulating through the respirometer loops. The chamber was loaded by adding eggs or larvae to the bottom section of the chamber while it was partially immersed in the loading tank. The screen and top section were carefully placed on top of the bottom section as the whole unit was slowly submerged, ensuring that no air bubbles were trapped inside while closing the chamber. Once submerged the chamber was pressed into the plastic clamp and attached to the respirometer loop. The number of organisms loaded into each chamber varies from 4 to 30 depending on the species, developmental stage and test temperature. After all four chambers were attached to their respective respirometer loops the whole water bath was covered with a styrofoam lid and a black plastic cover. The test organisms were acclimated for 30 minutes by flushing with the test water at a flow rate of 5-20 mL/min depending on the chamber size, test species and current developmental stage. Once acclimated the 4-way valves were turned to the 'closed' mode and the run was initiated. Oxygen tensions were monitored and recorded by the computerized data acquisition system until saturation levels 195  reach about 70-80%. The run was ended by turning the pumps were turned off and removing the chambers. However, it is also possible to open the valves to return the system to 'flow-through' mode and flush with either new fresh 100% oxygenated test water or different water (e.g., different salinity) and initiate another respirometry run. At the end of a run the test organisms were removed for general examination and weighing (TDW) to allow calculation of massspecific oxygen consumption rates. To account for microbial oxygen consumption and electronic 'drift' of the oxygen electrodes a blank run, containing no test organisms, was conducted once the initial run with the test organisms was complete.  Computerized data acquisition Millivolt outputs from the two oxygen meters and the thermistor meter were directed to a 12 bit data acquisition computer board (analog to digital converter, United Electronics® PC74) via the wiring terminal board (Data Translation® DT707) (Fig. A . l ) . The data acquisition board is mounted in an IBM® compatible personal computer (286, 12 MHz) and digitizes the analog signals from the meters. Digital input is logged to the hard disk in ASCII format and displayed as a real-time image on the computer screen using a data acquisition software package (Labtech Notebook®, version 7.0).  Data analysis Linear regressions were performed on both sets of the raw oxygen tension data (with and without test organisms) to determine the instantaneous slopes of the recorded oxygen traces. Since the slope of each trace is equivalent to the rate of oxygen decrease in that chamber, the  196  oxygen consumption rate for a given respirometer loop was calculated by subtracting the slope of the blank run from that of the test run. These raw consumption rates were standardized by correcting for the total volume of the specific respirometer loop (minus the volume of the organisms) and the mass of the specific organisms tested in that chamber. To date this microrespirometer system has been used successfully to measure the oxygen consumption rates of the eggs and larvae of a variety of fish species including: Pacific halibut (Hippoglossus stenolepis), lingcod (Ophiodon elongatus), Mozambique tilapia (Oreochromis mossambicus), and chum (Oncorhynchus keta) and coho salmon (O. kisutch). For examples of data produced using this respirometer system see Chapter 1; all of the oxygen consumption rate data presented in that chapter were collected using this system  197  

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.831.1-0089121/manifest

Comment

Related Items