ELEMENTAL SIGNATURES IN BONE TO DETERMINE LIFE HISTORY CHARACTERISTICS IN FISH by Adrian D. Clarke BSc., University o f Northern British Columbia, 2002 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF REQUIREMENTS FOR THE DECREE OF MASTER OF SCIENCE IN NATURAL RESOURCES A N D ENVIRONMENTAL STUDIES (BIOLOGY) THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA October 2004 © Adrian Clarke, 2004 1^1 Library and Archives Canada Bibliothèque et Archives Canada Published Heritage Branch Direction du Patrimoine de l'édition 395 W ellington Street Ottawa ON K 1A 0N 4 Canada 395, rue W ellington Ottawa ON K 1A 0N 4 Canada Your file Votre référence ISBN: 0-494-04650-3 Our file Notre référence ISBN: 0-494-04650-3 NOTICE: The author has granted a non­ exclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distribute and sell theses worldwide, for commercial or non­ commercial purposes, in microform, paper, electronic and/or any other formats. AVIS: L'auteur a accordé une licence non exclusive permettant à la Bibliothèque et Archives Canada de reproduire, publier, archiver, sauvegarder, conserver, transmettre au public par télécommunication ou par l'Internet, prêter, distribuer et vendre des thèses partout dans le monde, à des fins commerciales ou autres, sur support microforme, papier, électronique et/ou autres formats. The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. L'auteur conserve la propriété du droit d'auteur et des droits moraux qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation. In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis. Conformément à la loi canadienne sur la protection de la vie privée, quelques formulaires secondaires ont été enlevés de cette thèse. While these forms may be included in the document page count, their removal does not represent any loss of content from the thesis. Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant. Canada Abstract We^ used Laser-Ablation-lnductively-Coupled-Plasma-Mass-Spectrometry (LA-ICP-MS) to determine if trace metals deposited in calcified structures could be used to infer the life histories o f three different species o f fish. W e were successful in resolving the movements and age at maturity for an anadromous species, the eulachon {Thaleicthys pacificus) using both Strontium (Sr):Calcium (Ca) and Barium (Ba):Calcium ratios deposited in otoliths. Stream of residence was identified for a non-migratory freshwater species, the slimy sculpin {Cottus cognatus) in the W illiston Reservoir by matching the chemical fingerprint measured in the otolith to the water chemistries where they were captured. W e could not reveal, however, movement o f bull trout {Salvelinus confluentus) in the Morice River watershed. Water chemistry was similar throughout the length of the M orice River and movements w ithin the mainstem o f this river, therefore could not be distinguished. We conclude that chemical ratios measured in calcified structures are useful for quantifying life-history in fish providing that sufficient differences exist in the fishes ambient chemical environment. ' Clarke, A.D., Shrimpton, J.M., and Telmer, K. are all contributing authors for: Chapter 2 Life History and Patterns o f Movement Between Fresh- and Seawater in Eulachon {Thaleicthys pacificus)'. Chapter 3 Discrimination o f Habitat use by Slimy Sculpins {Cottus cognatus) in Tributaries o f the Williston Reservoir using Natural Elemental Signatures; Chapter 4 Movement Patterns o f Bull Trout {Salvelinus confluentus) in the Morice Watershed using Chemical Signatures Deposited Spatially in Fin Rays. Table of Contents APPROVAL................................................................................................................................... ii ABSTRACT................................................................................................................................... iii TABLE OF CONTENTS...............................................................................................................iv LIST OE TABEES.........................................................................................................................vii LIST OE FIGURES..................................................................................................................... vlii ACKNOWEEDGEMENTS...........................................................................................................xi PROEOGUE................................................................................................................................. 1 CHAPTER 2: LIEE HISTORY AN D PATTERNS OF MOVEMENT BETWEEN FRESH- AN D SEAWATER IN EULACHON (THALEICTHYS PACIFICUS).................................... ABSTRACT..................................................................................................................................11 INTR O D U CTIO N .......................................................................................................................13 MATERIAES AN D M ETHODS.................................................................................................. 20 RESULTS..................................................................................................................................... 22 DISCUSSION............................................................................................................................. 36 Comparison o f Life History to other smelts.......................................................................36 Ageing.................................................................................................................................... 39 Repeat Spawning Potential.................................................................................................46 CONCLUSIO N...........................................................................................................................49 CHAPTER 3: DISCRIMINATION OE HABITAT USE BY SEIMY SCULPINS (COTTUS COGNATUS) IN TRIBUTARIES OE THE WIELISTON RESERVOIR USING NATURAL ELEMENTAL SIGNATURES...................................................................................................... 51 ABSTRACT..................................................................................................................................51 11 INTR O D U CTIO N .......................................................................................................................53 M ETHODS.................................................................................................................................. 54 Study Location.......................................................................................................................54 W ater C olle ction ..................................................................................................................54 W ater Analyses..................................................................................................................... 57 Fish Collection.......................................................................................................................57 O tolith C hem istry.................................................................................................................58 Statistical Analyses................................................................................................................59 RESULTS..................................................................................................................................... 60 Stability of W ater Chemistry................................................................................................60 Heterogeneity of W ater Chemistry.................................................................................... 64 Otolith C hem istry.................................................................................................................64 DISCUSSION..............................................................................................................................76 Stability of Elements............................................................................................................. 76 Geographic Separation........................................................................................................78 CHAPTER 4 : MOVEMENT PATTERNS OE BULL TROUT (SALVELINUS CONFLUENTUS) IN THE MORICE WATERSHED USING CHEMICAL SIGNATURES DEPOSITED SPATIALLY IN FIN RAYS.................................................................................... 83 ABSTRACT..................................................................................................................................83 INTRO DUCTIO N ...................................................................................................................... 85 METHODS..................................................................................................................................87 Study Location...................................................................................................................... 87 Water C olle ction ..................................................................................................................88 Fish C ollection...................................................................................................................... 88 iv Fin Ray Chemistry.................................................................................................................88 RESULTS..................................................................................................................................... 90 DISCUSSION............................................................................................................................ 102 Geographic Separation..................................................................................................... 102 Fin-ray Chemistry............................................................................................................... 104 CONCLUSION.........................................................................................................................106 CHAPTERS: EPILOGUE......................................................................................................... 108 REFERENCES............................................................................................................................. 113 List of Tables Table 3.1 Average element ratios measured by ICP-OES for each tributary over the sampling period from June - November 2002...........................................................61 Table 3.2 Element concentration (PPB) measured by solution-based ICP-MS for tributaries of the Parsnip River.....................................................................................66 Table 4.1 W ater chemical ratios measured in the Morice River watershed......................91 Table 4.2 Relationship between fin-ray chemistry measured at the outer edge and water chemistry for 10 bull tro u t captured in the Morice River watershed........... 94 Table 4.3 Length and age for the 10 bull trout used in this investigation........................101 VI List of Figures Figure 2.1 Relationship between length and weight for the populations of eulachon exam ined....................................................................................................... 23 Figure 2.2 Condition factor as a function of length for the Columbia, Fraser, Skeena, Kemano, and Area 23-6 fish examined....................................................... 24 Figure 2.3 Representative line scans for Ba:Ca chemical ratios over the lifetime of individual eulachon from the Columbia, Fraser, Skeena, Kemano, and Area 23-6 eulachon.......................................................................................................27 Figure 2.4 Fluctuations in Ba measured in this study for southern populations of eulachon........................................................................................................................ 28 Figure 2.5 Fluctuations in Ba measured in this study for northern populations of eulachon........................................................................................................................ 29 Figure 2.6.1 O tolith images and elemental profiles for an eulachon sampled from the Skeena River (S35)...................................................................................... 30 Figure 2.6.2 O tolith images and elemental profiles for an eulachon sampled from the Skeena River (S45).................................................................................................. 31 Figure 2.6.3 O tolith images and elemental profiles for an eulachon sampled from the Kemano River (K23)................................................................................................ 32 Figure 2.7 Cross sections o f eulachon otoliths showing pseudo annuli that appear to represent yearly growth............................................................................................33 Figure 2.8 The relationship between number of summer periods represented by Ba:Ca elemental ratios and fork lengtho f eulachon used in this study..................... 35 Figure 3.1 Map o f the Williston Reservoir showing main tributaries. The numbered tributaries include: Table River (1), Tacheeda Creek (2), Bill's Creek (3), Hominka River (4), Misinka River (5), W ichika Creek (6), Swannell River (7), Davis River (8), Pelly Creek (9), Factor Ross Creek (10), Colbourne Creek (11 ), Reynolds Creek (12). The other tributaries sampled include the Parsnip River (4 locations), Anzac River, W ooyadilinka Creek, Misinchinka River, Nation River, Manson River, Omineca River, Osilinka River, Mesilinka River, Ingenika River, and the Finlay River................................................ 56 V ll Figure 3.2 A comparison o f Ba (PPM) and Sr (PPM) measured over the duration of the sampling period in the Table River to Ba:Ca and Sr:Ca ratios....................62 Figure 3.3 A comparison o f Ba (PPM) and Sr (PPM) measured over the duration of the sampling period in the Table River to Ba:Ca and Sr:Ca ratios....................63 Figure 3.4 Canonical discriminant function analysis characterising tributaries of the Williston Reservoir according to the multivariate signatures o f Sr:Ca, Ba:Ca Mg:Ca, and M n:C a........................................................................................................ 65 Figure 3.5 Cathodoluminescence image showing the elemental profile for Ba. Note the distinctive variation in the elemental concentration corresponding to the white luminescent region. The portion o f the otolith coloured blue represents the mainstem river habitat which appears to have remained stable for 3 + years...................................................................................................................67 Figure 3.6 Relationship between measured Sr:Ca (top) and Ba:Ca (bottom) values in slimy sculpin otoliths compared to the measured values in the river where the fish was captured....................................................................................................69 Figure 3.7 No significant relationship exists between Mg:Ca (top) and Mn:Ca (bottom) measured in slimy sculpin otoliths compared to the values measured in the rivers the fish were captured in ....................................................................................70 Figure 3.8 Canonical dicrim inant function analysis showing how sculpins from individual populations clustered according to the multivariate combination of Sr:Ca, Ba:Ca, and Mn:Ca (discriminant factors) measured in each otolith..............................................................................................................................72 Figure 3.9 Representative cathodoluminescent images of otoliths from sculpin captured in the Table River (T7), Anzac River (AID), Bill's Creek (B5), Osilinka River (0 5 & O lO ), and Swannell River (51)............................................................... 74 Figure 3.9 Continued............................................................................................................... 75 Figure 4.1 Map of the Morice River watershed showing water sampling locations. Water was collected from Gold Creek, Houston Tommy Creek, Owen Creek, Lamprey Creek, Thautil River, Denys Creek,Gosnell Creek,Crystal Creek, Redslide Creek, Nanika River, Morice Lake, and3 locations in Morice River....... 89 V lll Figure 4.2 Canonical discriminant function analysis according to the chemical signatures of Ba:Ca, Sr:Ca, Mn;Ca (discriminant factors)showing the variation measured in the Morice River watershed. Discriminant function analysis shows that the Morice River, M orice Lake, and the Nanika River group too close together to discriminate movements of bull trout between these systems. Known spawning tributaries separate into distinct locations (Gold Creek, Crystal Creek, Gosnell River, Thautil River, Redslide Creek)...................................92 Figure 4.3 Representative line scans for a bull trout (bull trout # 7 ) that had chemical signatures which were indicative of movements throughout its life-history........... 95 Figure 4.4 Representative line scans for a bull trout (bull trout #10) that had chemical signatures which showed little movement throughout its life-history....96 Figure 4.5 Canonical discriminant function analysis for bull trout # 1 0 . The letters indicate locations throughout the life o f each individual with (a) representing the capture location and (i) representing the approximate first year of growth........... 97 Figure 4.6 Canonical discriminant function analysis for bull trout # 7. The letters indicate locations throughout the life of each individual with (a) representing the capture location and (i) representing the approximate first year of growth. One potential migration (g) to the Gosnell Creek watershed is noted................... 98 Figure 4.7 Zn profile for bull tro u t # 2 . (top) and bull trout # 7 (bottom) There are six clear oscillations that correspond to the 6 annuli present in the fin-ray for bull trout # 2 and 7 clear oscillations that correspond to the seven annuli present in the fin-ray in bull trout # 7 ...........................................................100 IX Acknowledgements I w ould like to thank my supervisor Dr. J. Mark Shrimpton for all of the help and guidance throughout the completion o f this thesis. I express my gratitude to the rest of my research committee, Drs. Lito Arocena, Mike Gillingham, and M ike Rutherford, for their helpful insight and comments throughout my studies at UNBC. In addition. Dr. Kevin Telmer of the Centre for Earth and Ocean Research at the University o f Victoria for reading over sections of my project and providing valuable advice. Adam Lewis, Ecofish Research, assisted in the collection of the eualchon and provided much needed insight on my age data. I acknowledge the contributions o f Dr. David Dick, University of Northern British Columbia, for conducting the water analyses required for this project and Dr. Richard Cox, Centre for Earth and Ocean Research, University o f Victoria, for all of the help using the LA-ICP-MS and CL microscope. In addition many people assisted me in the field with sample collection and helped me complete my lab activities: Marcel Macullo, Kevin Mernickle, Aurora Sentlinger, and Gilgamesh Eamer. I also appreciate the recommendations and advice of the Peace W illiston Eish and W ildlife Technical Committee. Finally this project could not have been completed w ithout the generous funding provided by the Kitsumkalum First Nation for the eulachon study, the Peace Williston Fish and W ildlife Compensation Program for the sculpin study, and the Natural Science and Engineering Research Council o f Canada for the bull trout study. Chapter 1 : Prologue Tracking fish through m ultiple life-history stages with conventional tagging techniques has contributed useful information to identify the tim ing and duration of habitat utilization. Additionally, stocks are usually defined by geographic separation and tagging information has been useful to discriminate between stocks. It has been determined, however, that there is an inherent bias in many tagging techniques as tagging programs often end up concentrating on the re-captured, non-mobile portion of the population (migrating fish often leave the study area), or on the members that are physically large enough to receive tags (Cowan et al. 1994). Moreover, some species that are susceptible to injury from handling may experience higher mortality rates resulting directly from the application of physical tags. Partly, for these reasons, the amount of dispersal and tim ing of migration for various species and different age classes is poorly understood. O ther techniques have to be developed to determine habitat use and movement of small fish species and of individual fish throughout their entire life cycle. One method that has shown promise in providing habitat-utilization data for fish is elemental analysis of bony structures. Bony structures have been used to assess movement o f anadromous fish between freshwater and seawater due to the large differences in the chemical composition between these tw o media (Coûtant and Chen 1993; Friedland et al. 1998; Limburg 1998; Veinott et al. 1999). Recently, distinct chemical signatures have been found to vary in bony structures o f saltwater fish as a result o f variable ion content (inshore/offshore), temperature, and sources o f food (Kalish 1991; Elsdon and Gillanders 2002); therefore, seasonal cycles in elemental concentration are often apparent when measured in the structure. Distinctive chemical signatures have also been found among freshwater watersheds. Specific elemental and isotopic stream signatures are dependent on the underlying bedrock geology. Kennedy et al. (2000) were able to discriminate location of origin for Atlantic salmon (Salmo salar) among several streams in Verm ont using stable isotopes of Strontium (Sr). Stable isotopes vary with composition and age of bedrock and Kennedy et al. (2000) argue that isotopic ratios are better suited for determining fish movement with geochemistry than elemental concentrations. These authors indicate that the elemental composition of stream waters may vary too much over tim e and tend to have low spatial variability, lim iting the use of elemental fingerprints in resolving fish movement. This finding is contradicted by Wells et al. (2003), w ho determined that stream chemistries were stable seasonally and over a two-year duration. Wells et al. (2003) were able to use elemental signatures to determine origin for Westslope cutthroat trout {Oncorhynchus clarki lewisi) from three populations w ithin the Couer d'Alene river system. Cutthroat trout could be differentiated using the combination of barium (Ba):calcium (Ca), Sr:Ca, and magnesium (Mg):Ca ratios. Elemental signatures, therefore, may provide an alternative to the examination of isotopic stream signatures as a whole range of elements can be examined (Wells et al. 2003). Furthermore, the cost of analyzing m ultiple elements, as opposed to different isotopes o f specific elements, is 2 much lower. The lower cost associated with analysis should allow researchers the opportunity to increase the number of samples examined in a particular study. Three structures that show promise in tracking the environmental life-histories of fish are otoliths, pectoral fin-rays, and scales. Otoliths are located w ithin the semicircular canals of the inner ear in fish. There are three pairs o f otoliths, the sagittae, lapilli, and asterisci; the sagittae is the largest. The main functions o f the otolith complex in fish are equilibrium and sound detection (Moyle and Cech 2004). Growth offish otoliths is a continual process (Campana and Thorrold 2001), resulting from successive deposits of CaCOg with residual proteins (Campana and Nielson 1985). Campana and Nielson (1985) suggest that otoliths are metabolically inert even under extreme stress. Further, otoliths are unique when compared to other bony structures found in fish as daily increments are often visible. For this reason spatial resolution using a probe-based technique for chemical analysis is very good due to the chronological properties o f the otolith (Pannella 1971 ; Campana and Nielson 1985; Campana and Thorrold 2001 ). A major disadvantage o f using otoliths for chemical analysis is the fact that the animal must be sacrificed. Fin-rays offer a non-lethal alternative for chemical investigations examining lifehistory in fish. These structures support fish fins, but do not necessarily grow continuously like otoliths. Fin-rays grow incrementally and show alternating opaque and translucent zones that correspond to w inter and summer growth respectively. There is the potential that material may reabsorb and be mobilized in bone after deposition; a 3 process that does not occur in otoliths, and may subsequently limit the utility of investigations using fin-rays to trace environmental life histories (Campana and Nielson 1985; Veinott and Evans 1999). Nevertheless, fin-rays are potentially still useful in rebuilding environmental life histories, as Sr, Mg, lead (Pb), bromine (Br), zinc (Zn), Ba, tin (Sn), manganese (Mn), and sodium (Na) have been shown to remain stable in fin-rays of white sturgeon (Acipenser transmontanus) over tim e (Veinott and Evans 1999). Elemental analysis on scales also shows promise for rebuilding the environmental life histories offish using natural chemical markers. For instance, measured Sr values in scales reflect proportional incorporation from the aquatic environment (Eek and Bohlin 1997; Wells et al. 2000b; Wells et al. 2003). Scales offer a similar advantage to fin-rays, as fish do not need to be sacrificed in order to examine life-history. These structures have also been used routinely to age fish in the past and are adequate to resolve a chemical time series. Farrell et al. (2000) successfully resolved the spatial heterogeneity of trace metals in Arctic grayling (Thymallus arcticus) scales. One problem with scales is that controversy still surrounds the potential for resorption of apatite (Bilton and Robbins 1971); however, other studies have determined that trace elements remain metabolically inert in scales and can be used to rebuild environmental life histories (Yamada and Mulligan 1982; W ell et al. 2000b). Thus, otoliths are likely the best choice for an examination of life-history using natural chemical tags; however, when life-history information is required on fish species that cannot be lethally sampled, both fin-rays and scales may offer non-lethal alternatives. 4 Physiological regulation and transport o f ions differs among otoliths, fin-rays, and scales; therefore, the mechanisms of incorporation are outlined below. The processes involved are difficult to track in some instances, as fish incorporate elements from both the water in which they live and the nutrients they acquire. Changes in the elemental composition across bony structures generally reflect ambient water chemistry; however, some trace elemental and isotopic variations are due to the physiology of the animal during a specific time period (Kalish 1990; Kennedy et al. 2000). Most ion exchange occurs from direct contact with water across the gills; however, some occurs across the intestine, particularly in saltwater fish where there is continual intake o f water to compensate for water losses (Fenwick 1989). The incorporation o f ions into otoliths differs from fin-rays and scales and follows from tw o distinctive pathways. Otoliths are immersed in the endolymph fluid where ion concentrations show less variation than in the blood (Campana et al. 2000). Ions crystallize directly onto the otolith from the endolymph fluid (Campana and Nielson 1985), while fin-rays and scales incorporate ions directly into the crystal lattice from the blood (Veinott et al. 1999). One major assumption o f using elemental signatures to infer life-history is that trace element deposition onto the bony structure is mainly a result of branchial uptake directly from the water (Fenwick 1989) and not a result of elemental deposition due to diet. Further to this, the researcher must be certain that the chemical composition of the aquatic environment is influencing the deposition onto the structure, rather than the physiology of the animal. To date, the following studies have determined that non5 essential trace elements are related proportionately to water chemistry and not diet: Sr:Ca ratios in juvenile black bream {Acanthopagrus butcheri) (Elsdon and Gillanders 2 0 02 ;Elsdon and Gillanders 2003); Sr:Ca and Ba:Ca in Westslope cutthroat trout (Oncorhynchus clarki iewisii) (Wells et al. 2003); and Sr:Ca, Ba:Ca, and Cd:Ca in juvenile spot ( Leiostomus xanthurus) (Wells et al. 2000a; Bath et al. 2000). Thus, it appears that trace metal deposition in the otolith must be proportional to the dissolved elements in the ambient aquatic environment in order to link habitat use, and therefore, rebuild fish environmental life histories. The use of elemental signatures shows promise for determining temporal locations offish, both in seawater and freshwater, from the elemental profile in bony structures. The ability to analyze the elemental composition of a sample spatially is key to this technique, as it provides an indication o f which habitats were utilized over the life of the animal. A method that gives good resolution for determining elemental ratios over different life-history stages is Easer-Ablation-lnductively-Coupled-Plasma-MassSpectrometry (LA-ICP-MS). LA-ICP-MS is a micro-analytical tool that enables a spatial analysis of element distribution in a sample (Wang et al. 1994). This technique combines the beam capabilities of a high-energy laser with the analytical capabilities of ICP-MS (Denoyer et al. 1991 ). The laser is directed at a small area on the sample where photon energy is converted to thermal energy and a small portion of the structure is vaporized to a depth of a few micrometers (Fowler et al. 1995). The vaporized material is then carried by a flow of argon gas into the plasma of an ICP and the elemental and isotopic 6 composition is determined by the mass spectrometer (Campana et al. 1997). LA-ICP-MS can discriminate between spatial regions on bony structures that are as small as 30 fjm. Resolution of trace elements can be identified as low in concentrations as 0.1 to 0.01/vg/g in multi-element analyses (Wang et al. 1994). There have been some problems associated with this technique in the past. These problems, however, appear to be linked to specific laboratories performing the analytical procedures. A comparison of techniques was made between numerous laboratories using a standardized sample measured w ith the electron microprobe, X-Ray diffraction, and LA-ICP-MS. Both the electron microprobe and LA-ICP-MS were determined to be valid techniques in reconstructing environmental life histories (Campana et al. 1997). Furthermore, LA-ICPMS has been reliable for determining the environmental life histories of white sturgeon (Veinott et al. 1999), Atlantic cod {Gadus morhua) (Campana et al. 1994), striped bass {Morone saxatilis) (Coûtant and Chen 1993), and several species of salmonids (Kalish 1990; Thorrold and Shuttleworth 2000). Recently, Sanborn and Telmer (2003) have greatly increased the spatial resolution of this technique, while substantially reducing the cost by using continuous line scans with LA-ICP-MS. The authors determined that line scans are at a minimum equivalent to discrete spot analysis and likely provide better information about the distribution of elements in heterogeneous solids. The above methods use a quantitative examination of metal deposition in bone. An alternate approach for examination o f elemental deposition in bony structures is Cathodoluminescence (CL) microscopy. CL microscopy can potentially be used to 7 com plem ent otolith chemistry investigations in a qualitative manner. CL is very sensitive to the presence o f low levels of trace elements that provide a unique visualization of colour changes in the otolith which is attributed to variations in chemical zonation. CL emissions, therefore, are a useful method for understanding the chemical variation experienced by fish (Halden 2001). The variation in luminescence likely results from changes in the chemical environment or physiology experienced by the fish. Typically, Mn causes most o f the luminescence associated w ith carbonates (Marshall 1988); but Sr may have contributed to the luminescence found in whitefish otoliths (Halden 2001). To date, there has been very little research done using this technique on otoliths to determine if it would be useful for rebuilding the environmental life histories of various fish species. This project utilized discrete chemical signatures deposited spatially in bony structures o f three different species of fish over the duration of their life-history. Elemental analysis was used to match specific signatures to the ambient water chemistry in order to examine life histories of an anadromous species, a freshwater non-migratory species, and a freshwater migratory species. In the first part of this study (Chapter 2), large chemical differences in saltwater and freshwater were used to examine the life-history of the eulachon (Thaleicthys pacificus) a small osmerid species. Sr:Ca ratios have been shown to be effective in observing movements in anadromous fish between these two media. As well, age information for eulachon is uncertain to date. Seasonal fluctuations in elemental concentration were examined in eulachon otoliths to attempt to age these animals correctly. The second portion o f this project (Chapter 3) examined the chemical heterogeneity in a freshwater watershed and determined that the different chemicalstream signatures corresponded to the otoliths of fish from each specific area. Slimy sculpins {Cottus cognatus) in the Williston watershed are a non-migratory fish species and were used to determine if chemical signatures are geographically distinct by linking specific water chemistries to the elemental signatures found in otoliths. The final portion o f this examination (Chapter 4), determined if distinct chemical signatures deposited spatially in fin-rays could be used to assess the movements of a migratory fish. Bull trout {Salvelinus confluentus) a provincially blue-listed species from the Morice watershed has been classified as 'movers' and 'non-movers' from a radio­ telemetry study (Bahr 2002). Ten fish from the M orice River telemetry study were examined to see if chemical signatures could be used to assess movements w ithin the watershed. W ater chemistries were determined from tributaries o f the M orice River and w ithin the Morice River itself to determine if migrations highlighted by variations in measured elemental ratios across the fin-ray could be linked to specific regions w ithin the watershed. This project should aid in our ability to reconstruct the movement o f an individual's entire life-history and provide im portant information and a potential tool which can be used towards the conservation of threatened fish species. 10 Chapter 2: Life-history and Patterns of M ovem ent Between Fresh- and Sea-water in Eulachon {Thaleicthys pacificus) ABSTRACT Populations of eulachon (Thaleicthys pacificus) have declined significantly in recent years and it is crucial to further our understanding of their life-history. The main objectives of this study were to determine the age at maturity and repeat spawning potential for eulachon, tw o aspects o f eulachon life-history that are not known, but are im portant for successful management of this species. Trace-element analysis of bone can be used to reconstruct many life-history characteristics because elements are incorporated from the water as fish grow. We used Laser-Ablation-lnductively-Coupled-Plasma-MassSpectrometry (LA-ICP-MS) to reconstruct the Ba;Ca and Sr:Ca molar ratios deposited spatially into the otolith, a bone located in the inner ear. Spawning eulachon examined in this study were at least 160 mm in length and 30 g in weight (suggesting that eulachon spawn after reaching a m inimum size). Age at maturation, however, differed among populations examined. Three full cycles of fluctuations in Ba:Ca molar ratios were observed in the majority of otoliths from spawning fish, indicating that eulachon mature at three years of age. Based on the seasonal fluctuations in Ba:Ca molar ratios, we determined that most Columbia River eulachon spawn after two years, w hile Fraser, Kemano, and Skeena River eulachon generally mature after three years. Two Skeena River eulachon matured after four years. In contrast to the Ba:Ca molar ratios in the otolith, Sr:Ca molar ratios maintained a relatively flat profile over the life of the eulachon. The lack o f a change in Sr:Ca ratios w ithin the otolith, the single size class of spawners 11 across all river systems, and the single age class w ithin most rivers strongly suggest that eulachon from the populations in our study are semelparous. Thus, examination of otolith microstructure was successful in identifying tw o important life-history characteristics, age at maturity and repeat spawning potential, furthering our understanding of the eulachon. 12 INTRODUCTION The eulachon (Thaleicthys pacificus) is a small anadromous smelt that spawns from northern California to the southern Bering Sea. Despite the wide geographic range, there is much unknown regarding the general biology and life-history o f this species. There is a pressing need to acquire more information regarding this species as populations of eulachon have shown a dramatic decline in numbers o f spawners over much of their geographic range in the last 20 years. This trend has become particularly apparent since the mid-1990s and reasons for the decline are unknown. It is likely, however, that many factors are responsible for this trend. Information on the life-history o f eulachon is needed to determine what mechanisms are responsible for the observed declines. The eulachon exhibits an anadromous life-history (McAllister 1963) and has been observed to spawn in many of British Columbia's rivers during March and April (Barraclough 1964). Other observations have shown that eulachon enter the Skeena and Nass Rivers as early as February and some Alaskan rivers as late as May (Hay and McCarter 2000). The spawning habitats of the eulachon range from Northern California in streams o f the Klamath River, to the Nushagak River in Alaska (McAllister 1963). Eulachon are found in 33 rivers in British Columbia; of these, only 14 have regular runs (Hay and McCarter 2000). Barraclough (1964) determined that eulachon begin maturing at age two and spawn at three years of age. 13 Eulachon spawn in rivers that have spring freshets characterized by headwaters w ith large snow packs or glaciers (Hay and McCarter 2000). Spawning occurs a short distance upriver in areas that are covered by coarse sand (Hart and McHugh 1944). In the Nass River, eulachon migrate a maximum o f 24-32 km upstream and spawn after the ice breaks up, when water temperatures are between 4.4-7.8 °C (Scott and Crossman 1973). The Bella Coola River has eulachon spawning in the first 6 km (Pootlace and Siwallace 2000) and in the Kowesis River, eulachon have been found 5 km upstream (Kelson 2000). Eulachon generally spawn at night (Kelson 2000) and remain in some river systems for approximately three weeks (Pootlace and Siwallace 2000). Females produce between 17,000 and 40,000 eggs, each egg having an adhesive outer membrane that sticks to the coarse sand particles (Hart and McHugh 1944). The fertilized eggs take approximately two to three weeks to hatch depending on water temperature (Shepherd and Vroom 1977). It took 41 days for eulachon to hatch at water temperatures between 4-7 °C in the Kitimat River (Farara 2000). After the eggs hatch, the newly emerged larvae are carried down to the ocean by the current (Hart and McHugh 1944). As a consequence, eulachon spend very little tim e in the freshwater environment. The freshwater residence of larval eulachon, however, may be a critical stage for development before the fish enter the ocean and also after they return to spawn and complete their life-cycle. The marine life-history of eulachon begins when the larvae that have been swept downstream by the current enter the ocean. Larval stages may remain in estuarine and 14 marine waters close to natal rivers for several months or longer. This leaves very little tim e for eulachon to im print to the freshwater environment, and it is possible that im printing does not occur at all (McLean et al. 1999). Hay and McCarter (2000) indicate that any im printing is likely to occur in the estuarine environment adjacent to spawning rivers. McLean et al. (1999) found that population subdivision was more likely to occur as a result of the marine environment, rather than factors that affect other anadromous species (e.g. a distinct chemical cue present in natal stream), because there was an apparent high level o f gene flow. O ther anadromous species, such as salmon, im print to chemical signatures from a specific stream or river and generally show high fidelity when returning to natal rivers to spawn. Population structure is weakly developed in eulachon with < 2% total genetic difference found among populations from the Columbia River, Cowlitz River, Fraser River, Franklin and Klinaklina Rivers in Knight Inlet, Kowesas, Kitimat, and Kemano Rivers from Gardner Canal, Mass River, Cook Inlet, and the Bering Sea (McLean et al. 1999). Additional evidence that extensive mixing of stocks may occur arises from observations of abundance. There have been reports that eulachon w ill repeat-spawn in a river system over subsequent years, disappearing from the system and returning years later (Stacey 1998 cited in Hay and McCarter 2000). One aspect o f eulachon life-history that is poorly understood is age at maturity. Commonly, ages of fish are determined from the spatial deposition o f rings on bony 15 structures such as otoliths. Estimates of age at maturity for eulachon are between four and six years (Delacey and Batts 1963), but there is considerable controversy. Hay and McCarter (2000) suggest that age estimates from otoliths may be unreliable from eulachon. This conclusion is based on the high variation in age analysis using otoliths in previous studies. Ricker et al. (1954) compared scales and otoliths from eulachon and showed that age estimates from otoliths were one to tw o years higher than from scales. Hay and McCarter (2000) further note that De Lacy and Batts (1963) obtained much higher ages for eulachon using otoliths than other studies. An interesting observation put forward by Hay and McCarter (2000) is that length and weight do not vary between eulachon aged between tw o and six years; a finding that indicates the determined ages may not be accurate. This observation may not be warranted for all eulachon populations, as repeat spawning potential is still undetermined. Fish often return to normal feeding after spawning, but growth is restricted (McDowall 1987); furthermore, older members of a population typically grow slower than younger members. Confirmation of age for eulachon may provide insight to the variability in age at spawning, with younger members of the population potentially returning to repeat spawn at a later date. Hay and McCarter (2000) felt that using otoliths for ageing w ill not be reliable unless further studies verify otolith age determination with other means of analyses. There are many methods that have been used to verify age estimates, or at least to correlate annuli deposition. Campana (2001) outlines the use and misuse o f these 16 many techniques. The techniques for validating age estimations outlined include: release o f known age and marked fish, mark-recapture, radio-chemical dating, marginal increment analysis, and captive rearing. For the most part, these techniques are not suitable for use in validating eulachon ages, as eulachon cannot be effectively marked and recaptured, they are potentially short lived, and captive rearing would require development of rearing techniques and elaborate holding facilities. One method that may be appropriate for age validation in eulachon is observation o f elemental signals. This procedure may be viewed as only corroborating the frequency o f annuli formation, as elemental signatures may not fluctuate regularly or annually (Campana 2001 ). One way around this is to utilize chemical signatures that are shown to have regular seasonal variation that is represented by fish migration or intra-seasonal variation due to temperature effects. Chemical signatures of otoliths from salmon have shown periodicity as fish migrate between different regions of the ocean. Distinct chemical signatures have been found to vary in bony structures o f saltwater fish as a result of variable ion content (inshore/offshore), temperature, and food sources (Kalish 1991; Elsdon and Gillanders 2002; Bath-Martin et al. 2004); therefore, seasonal cycles are often apparent. Seasonal variation among elements in the marine environment has been reported for Na, Sr, K, S (Kalish 1991), Li, Mg, Sr, and Ba (Elsdon and Gillanders 2002). 17 Another aspect of eulachon life-history that requires further examination is the potential for eulachon to repeat the spawning cycle. Some of the available data suggest that eulachon spawn once and die (semelparous). Hay and McCarter (2000), indicate that most eulachon are semelparous because: (1) they re-absorb their teeth during spawning; (2) spawning eulachon are larger than those seen in marine waters, (3) to date, no toothless eulachon, or eulachon with regenerating teeth have been found in the sea, and (4) substantial post-spawning mortality has been documented in most BC rivers. This information is evidence that most eulachon, and particularly the largest o f spawning eulachon, do not survive to return to the sea. Bailey (2000) concurs with the hypothesis that eulachon are semelparous, as beaches o f the Fraser River are white w ith dead eulachon following spawning. It has been suggested, however, that eulachon are iteroparous (spawn more than once) in some river systems. Pootlace and Siwallace (2000) indicated that carcasses were not observed on the beaches o f rivers in Dean and Burke Channel in northwestern British Columbia after spawning and suggested that eulachon return to the ocean following spawning. Additionally, Barraclough (1964) reported that eulachon were caught off the Fraser River after they had spawned once and were in good condition and he hypothesized that they may spawn a second time. There is morphological data that suggests iteroparity may increase in the more northern latitudes. Eulachon, over much of their distribution, re-absorb their teeth during spawning, whereas fish caught in the marine environment typically have large pronounced teeth (Hay and McCarter 2000). Fish caught in rivers from Alaska retain 18 their teeth w hile in freshwater and during spawning. Hay and McCarter (2000) hypothesized that these more northerly populations could be iteroparous, however much firm er conclusions are necessary. Determination of movement patterns and m ultiple spawning, however, is difficult for eulachon. Length-frequency relationships have been used in some fish species to examine whether fish spawn more than once, but m ultiple age classes may overlap in size (McDowall 1987) and considerable controversy exists for age at maturity in eulachon (Hay and McCarter 2000). An alternate approach to determine m ultiple spawning is to assess movement patterns between fresh- and sea-water from incorporation o f elemental signatures in bony structures. Elemental analysis of bony structures can be used to determine a variety o f life-history traits in fish. Different chemical signatures exist between marine, estuarine, and freshwater environments and these different signatures are incorporated into bony structures as fish grow and move between these environments. Bony structures, therefore, provide a spatial elemental record corresponding to habitat utilization at each life stage of a specific population or individual. For this reason, elemental analysis o f bony structures can be used to detect migration and movement patterns between freshwater and seawater. A number of elements have proven useful in elemental analysis, most commonly Ca, Ba, and Sr. The ratios of Ba:Ca, and Sr:Ca, vary between fresh- and sea-water and these differences are incorporated into bony structures. 19 In this study we investigated the potential for using chemical signatures naturally deposited in otoliths to further our understanding o f eulachon life-history. The objectives of our study were to acquire eulachon otoliths from a wide geographic range including the Skeena, Kemano, Fraser, and Columbia Rivers, to investigate otoliths for seasonal fluctuations in elemental concentration and link seasonal changes in elemental concentration with annual increment formation on the otoliths to assess age, and to ascertain if eulachon are semelparous or iteroparous by examining elemental ratios deposited in eulachon otoliths. MATERIALS AN D METHODS Eulachon were collected from the Skeena, Kemano, Fraser, and Columbia Rivers between late January and May 2003. All samples came from First Nation's harvest fisheries donated after capture. In the Skeena and Fraser River eulachon were captured in gill-nets, purse-seined in the Kemano River, and dip-netted in the Columbia River. Twenty individuals were randomly selected from each sample to be used in the study. O nly male eulachon were available from both the Skeena and Columbia Rivers. Male and female eulachon were examined for the Kemano and Fraser Rivers; however, there were more male fish available for the analysis. In addition, 10 ocean eulachon captured in July 2001 from Area 23-6 (Barkley Sound) were included in the analysis. Eulachon from Area 23-6 were captured by a commercial shrimp trawler. The ocean fish were immature and sex could not be determined. Length (L) and weight (W) were measured on all fish and condition factor (index of plumpness (K)) was calculated using the 20 equation K = W / * 100 000 (Moyle and Cech 2004). Both L and W were analyzed with a one-way Anova using both Tukey's HSD post hoc test (SPSS version 11.5, Sep 2002 Chicago, IL.). All data are presented with the standard error of the mean. Both the right and left sagittal otoliths were removed from the fish and sonicated for five minutes in ultra-pure water. Otoliths were then embedded in epoxy resin (Allied High Tech embedding medium, Rancho Domiguez, CA) and ground in the transverse plane with 1200 jUm silicon carbide paper until the core was exposed. W et grinding using ultra-pure water prevented external contamination of the samples. Otoliths were then sequentially polished with 6 fjm, 1 fjm, and finally 0.05 /Jm diamond suspension to ensure an adequate surface for ablation with the laser. During preparation and transfer of polished otoliths, some otoliths were damaged and could not be used, reducing the sample size. LA-ICP-MS analysis was conducted following the protocol outlined in Sanborn and Telmer (2003). Material was extracted from the otolith with a PQ II S-P high sensitivity ICP-MS (VC Elemental) coupled to a UV laser ablation system (Merchantek). The laser system operated with an output of 266 nm that has a maximum energy output of 4 mj. Optim ization was conducted using Standard Reference Material (SRM) 613 NIST glass, containing —50 /vg/g total trace elements. All analyses were conducted at a frequency of 20 Hz with 75% power and the aperture of the laser set at one. Average energy while operating at these conditions was 0.70 mJ. We measured the w idth of the 21 laser scan after analysis with a microscope mounted micrometer and determined it to be 25-32 jUm. All otoliths scans were completed by us by tracking the laser across the otoliths at 5.3 /um/sec. The isotopes measured in the otoliths included "^^Ca, ®^Sr, ^^Mg, and ®®Zn. Ca was used as the internal standard, due to the otoliths aragonite (CaCOj) composition. Ca is 40% of the molecular weight of aragonite. An internal standard was required to account for variations in aerosol production caused by the variation in the amount of material being extracted from the otolith by the laser. Background intensities were collected for 30 seconds prior to running the laser. Data collection and reduction were completed using VG Thermo Electron PlasmaLab Software 2003 (Version 1.06.007, Burlington, ON). The Fully Quantitative Analysis option was chosen and an SRM 613 NIST glass was selected as the known standard. Two SRM 613 NIST glasses were analyzed, both at the beginning and end of each run. These certified standards were used to complete an external drift correction to compensate for any changes in machine sensitivity. Five otoliths were analyzed between each set o f standards. An SRM 611 NIST glass was also analyzed as an unknown sample during each run of five otoliths to help ensure measurement accuracy and precision. RESULTS Length-weight relationships (Figure 2.1) and condition factor (Figure 2.2) were plotted for the five eulachon populations examined in this study. There was a significant difference in fork length for the populations examined F (4,182) = 44.123, p < 0.05 and 22 80 -, 70 - • ♦ ■ ▲ # i Columbia Fraser Kemano Skeena Area 23-6 ♦ □ □ a D . 60 - ♦ ♦> à 50 - □ O 05 .O) 40 m 30 20 - 10 80 100 120 140 160 180 T 1 200 220 Fork Length (mm) Figure 2.1. Relationship between length and weight for the populations of eulachon examined. Solid symbols are males, open symbols are females. Area 23-6 fish were immature. Tukey's HSD post hoc test (subset for alpha 0.05) revealed significant differences among fish caught from different locations. Mean length of eulachon was 1 5 1 + 5 mm for Area 23-6, 175 + 3 mm for Columbia, 183 + 3 mm for Fraser, 189 + 2 mm for Skeena, and 196 + 3 for Kemano. 23 1.0 • ♦ ■ ▲ Columbia Fraser Kemano Skeena Area 23-6 O □ □ ♦ 0.8 — • □ ■ ' # 0^ a# ^ ■§ (0 c o .•: n'»‘. Î- c Ü □ 0.6 0.4 I— 80 T 100 120 140 T T T 1 160 180 200 220 Fork Length (mm) Figure 2.2. Condition factor (index of plumpness) as a function of length for the Columbia, Fraser, Kemano, Skeena, and Area 23-6 fish examined. Solid symbols are males, open symbols are females. Area 23-6 fish were immature. 24 weight F (4,182) = 48.788, p < 0.05 among the mature samples collected from the four rivers and one ocean group. Tukey's HSD post hoc test (subset for alpha 0.05) revealed that Area 23-6 eulachon were significantly smaller than the other populations by length; mean length was 152 + 3 mm. Spawning eulachon were larger than the fish caught in the ocean, although, there were significant differences among the populations. Columbia River fish were 175 ± 3 mm and did not differ significantly from the Fraser River fish, 183 ± 3 mm. The Skeena and Kemano River fish were significantly larger than the Columbia River fish. Skeena and Kemano eulachon were 189 + 2 mm and 196 ± 3 mm, respectively. Condition factors are shown for each o f the river systems (Figure 2.2). The group of four fish at the bottom left of the figure captured in Area 23-6, off the west coast of Vancouver Island, had the lowest values for condition factor. There was an oscillation in Ba:Ca deposition in the otolith that appeared to correspond with seasonal ocean temperature. During the summer (July-September) when ocean temperature was highest, Ba:Ca deposition in the otolith was also highest (Figure 2.3 A). This was observed in Area 23-6 eulachon (collected in July), which had the highest Ba:Ca values (referred to as peaks) at the outside edge o f the otolith. All of the eulachon captured on their spawning migration during the w inter and spring were characterized by low values o f Ba:Ca at the outside edge o f the otolith. This otolith region represented the chemical environment the fish were exposed to near the tim e of sampling, as the 30 [Jm resolution attained in this study corresponded to approximately tw o weeks of growth. Sea surface temperatures (Figure 2.3 B) represented oscillations in 25 ocean temperature, however these values did not necessarily correlate w ith measured values in the otolith. Figure 2.4 illustrates the number of Ba;Ca peaks measured in southern eulachon populations. Eulachon captured in Area 23-6 (ocean) had 1.5 and 2.5 peaks, Fraser River eulachon were all characterized by three peaks in Ba:Ca, and Columbia River eulachon exhibited tw o or three peaks in Ba:Ca. Figure 2.5 illustrates the number o f Ba:Ca peaks measured in otoliths from northern populations of eulachon. All of the fish in the Kemano and Skeena rivers examined were characterized by three peaks in Ba:Ca with the exception o f two Skeena River fish that had four Ba:Ca peaks. The number of peaks in Ba:Ca observed in eulachon otoliths tends to increase with increasing latitude. Cross-sectional and w hole-m ount views o f otoliths are shown in Figures 2.6.1, 2.6.2, and 2.6.3. Adjacent to the pictures of each otolith are the elemental signatures determined for Ba:Ca, Sr:Ca, Mg:Ca, and Zn:Ca. Ba:Ca ratios correlate well with opaque zones visible after polishing many o f the transversely sectioned otoliths; however, the profile for Sr:Ca remains relatively flat over the life of the animal. Both Mg:Ca and Zn:Ca ratios are higher in the core and then decline with age and growth of the fish. Ratios of Mg and Zn do not return to values observed near the core o f the otolith. Figure 2.7 illustrates some o f the difficulties regarding age estimation o f eulachon by counting the number o f annuli that are either on cross sections or whole otoliths. We would estimate the ages of the fish to be from one to four years old based on the number 26 W3 Columbia Kemano Skeena W3 Fraser Area 23-6 14 -1 p B <0 12 10 - (O Aug Nov Mar Jun Sep Dec Mar Jun Sep Dec Mar Figure 2.3 (A) Representative line scans showing Ba:Ca chemical ratios over the lifetime of individual eulachon from the Columbia River, Kemano River, Skeena River, Fraser River, and Area 23-6. Winter (W) periods are marked on the graphs and correspond to the lowest measured Ba:Ca chemical ratios. Figure 2.3 (B) Sea surface temperature data was adapted from Hay et al. 2003. Ba:Ca measured in the otoliths appears to correlate well with seasonal variations in ocean temperature. Eulachon spawning in the rivers were collected from January to late April while Area 23-6 eulachon were captured in July. 27 0.006 -| 0.004 Area 23-6 #8 0.002 0.000 - 0.004 Area 23-6 #16 0.002 - 0.000 - "ô Fraser #26 0.008 "ô I 0.004 0.000 - Ô3 00 0.006 Columbia #18 0.003 0.000 0.015 Columbia #1 0.010 - 0.005 0.000 20 40 60 80 100 120 140 160 Scan Tim e (s) Figure 2.4 Fluctuations in Ba:Ca ratios measured in this study for southern populations of eulachon. Area 23-6 eulachon are characterized by two and three peaks of Ba:Ca. Fraser River eulachon had three peaks of Ba:Ca while the majority of Columbia River eulachon had two peaks (three fish had three peaks of Ba:Ca). 28 Kemano #54 0.008 0.006 0.004 0.002 O 0.004 c 0.003 03 0.002 ^ 0.001 Skeena #35 u Ô3 Skeena #38 0.008 0.006 0.004 0.002 - 0 20 40 60 80 100 Scan T im e (s) Figure 2.5. Ba:Ca profiles demonstrating seasonal variation in northern eulachon populations. All Kemano River and most of the Skeena River eulachon had three peaks of Ba:Ca, however two Skeena eulachon exhibited four peaks. 29 Figure 2.6-1. Otolith images and elemental profiles for an eulachon sampled from the Skeena River (535). A and B are cross-section and whole m ount views of otoliths showing annuli. For A the black line indicates the location and length (689 /tm) of the laser line scan. C, D, E, and F are scan lines o f elemental signatures for Ba:Ca, Sr;Ca, Mg;Ca and Zn:Ca. ^ 0.004 - "= 0.003 E 0.0 0 2 - ^ 0.001 0.0 0 0 -I 1 0 -1 I ■■ - f liM : 6 0 1 o E & o O) 5 4 3 S & 0.05 o E o E 0.04 0.03 - ^ 0.02 - Ü r5 0.01 - 0.00 , I---- 1 1 20 40 80 -------------- -------- 60 Scan Time (s) 100 120 140 Figure 2.6-2. O tolith images and elemental profiles for an eulachon sampled from the Skeena River (S45). A and B are cross-section and whole m ount views of otoliths showing annuli. For A the black line indicates the location and length (731 fjm) o f the laser line scan. C, D, E, and F are scan lines of elemental signatures for Ba:Ca, Sr:Ca, Mg:Ca and Zn;Ca. 0.007 n o 0.006 - ft 0.005 - 2 0.004 - m 0.003 - 5 0.002 0.001 ^ 10 n 6 0 1 o E £ o d) S i 4 3 E 2 0.04 0 1 o E E j 0.03 - 0.02 - 0.01 - 0 .0 0 -I —I— 20 40 60 80 Scan Time 100 120 140 Figure 2.6-3. O tolith Images and elemental profiles for an eulachon sampled from the Kemano River (K23). A and B are cross-section and whole m ount views of otoliths showing annuli. For A the black line indicates the location and length (684 jum) of the laser line scan C, D, E, and F are scan lines of elemental signatures for Ba:Ca, Sr:Ca, Mg:Ca and Zn:Ca. 0.006 % 0.004 - <0 0.002 - 0.000 10 -1 0 1 o E E ë L 6 CO 4 6 0 1 o E E 5 d) 5 4 3 E S 0.03 n o % 0 .0 2 - E S (C o c 0 .0 1 - N 0.00 J 0 20 40 60 80 Scan Time 100 120 140 Fish Length W e igh t (mm) (g) # o f summers at sea A K31 209 83 5 3 B 013 179 41.4 2.5 C 04 146 21.9 2.5 D 08 137 17.2 1.5 Figure 2.7 Cross sections of eulachon otoliths showing pseudo-annuli (dots) that appear to represent yearly growth. Tabulated information below figure indicates origin and fish number, size and age estimate based on number of peaks in Ba:Ca ratio. White line indicates 1 mm. (K) refers to the Kemano River and (O) refers to Area 23-6 eulachon. 33 of annuli. O ur estimates of age for these fish from oscillations in Ba:Ca differs, as shown in the tabulated values below (Figure 2.7). Polishing time and the thickness of the polished section appeared to affect the number of annuli visible for many of the samples examined. The data obtained in our study suggest that there may be a m inimum size that eulachon must reach prior to the onset of maturity and spawning (Figure 2.8). Area 23-6 fish characterized by 1.5 peaks o f Ba:Ca have likely not yet reached the m inimum size necessary to migrate to the spawning grounds, therefore, we estimate these fish are more than one-year old. Area 23-6 fish that have 2.5 peaks of Ba:Ca are more than two years old and w ill likely reach the m inimum size needed to mature by the following spring at three years of age. Spawning eulachon, therefore, appear to belong to a single size class. A single age class of fish was also observed to spawn in tw o of the systems examined in this study; only three-year old eulachon were observed from the spawning populations in the Fraser and Kemano Rivers (Figures 2.8). The majority o f fish for the Columbia and Skeena Rivers was also composed o f a single age class; two and three year olds from the Columbia and Skeena Rivers, respectively. The dominance o f a single age class and single size class of fish observed spawning strongly suggest that eulachon spawn once. 34 4 • ♦ ■ ▲ # Columbia Fraser Kemano Skeena Area 23-6 AÆ à iÊ Ê à é k A 3 - CO E 3 — CO CÛ o * 80 100 T T T T 120 140 160 180 200 220 Fork Length (mm) Figure 2.8 The relationship between number of peaks in Ba:Ca elemental ratio and fork length of eulachon from the four river systems examined and for ocean fish caught off the coast of Vancouver Island. Solid symbols are males, open symbols are females. Area 23-6 fish were immature. 35 DISCUSSION Spawning eulachon (Area 23-6 fish are not included as they were captured in the ocean) in our study were at least 160 mm in length and greater than 30 g in weight; it appears that eulachon spawn after reaching a m inim um size. O ur analysis indicates that the age when fish reach this 160 mm threshold and mature varies w ith latitude (as shown by Columbia River fish (the most southerly in latitude) spawning at the earliest ages and the Skeena River fish (most northerly river in latitude) spawning at the oldest ages. Eulachon spawn after two or three years in the Columbia River, three years in the Fraser and Kemano populations, and three or four years in the Skeena population. The size threshold and lack of an appreciable change in elemental signatures associated with freshwater movement strongly suggest that eulachon are also semelparous. We will discuss our findings in relation to what is known regarding other species o f smelts, validate our method to assess age at maturity, and assess semelparity w ithin this species. Comparison o f Life-history to other smelts The results of our study indicate that eulachon have similar life-history characteristics to other smelts. Most smelts are short lived, demonstrate a mainly semelparous life-history, and utilize distinct habitats for spawning. Examination o f some of these life-history patterns may allow for an increased understanding of eulachon lifehistory. W ithin the Family Osmeridae (or the "true" smelts), there are only 13 species (Moyle and Cech 2004). The true smelts are abundant in the coastal areas o f the 36 northern hemisphere. Some species are entirely marine, some live entirely in freshwater or brackish-water, and some are anadromous. In addition to the eulachon, there are three species o f true smelt that have been extensively studied, although knowledge gaps still exist for these species. An examination o f life-history traits w ithin these species will aid our understanding of potential life-history patterns that may exist for the eulachon. These three species are the rainbow smelt {Osmerus mordax), the ayu (Plecoglossus altivelis), and the delta smelt (Hypomesus transpacificus). The rainbow smelt shows considerable variation in life-history traits. Some populations of rainbow smelt are anadromous, although other populations are lacustrine (live in lakes) (Taylor and Bentzen 1993). Anadromous rainbow smelt are generally 150 - 300 mm in length at maturity (Copeman 1977). Anadromous forms of this species enter coastal streams and deposit adhesive eggs onto shallow riffle areas. Eggs hatch in eight days at a mean water temperature of 15 °C (Cooper 1978). After the eggs hatch, the larvae are carried downstream to the estuary or the open ocean (Akielazek et al. 1985). The distribution of rainbow smelt expands from Vancouver Island to the Canadian Arctic (McPhail and Lindsey 1970). Prey of rainbow smelt consists of mysids and amphipods, however, they also demonstrate piscivory (Haldorson and Craig 1984). Ages determined for mature fish are extremely variable in rainbow smelt. Rainbow smelt have been reported to mature at tw o to four years of age in Lake Huron (Erie and Spangler 1985) and six to seven years (with a maximum age of 15 years) in the Beaufort Sea (Haldorson and Craig 1984). The variation in age at maturity reported may reflect 37 difficulties in ageing smelts, similar to what has been argued by Hay and McCarter (2000) and what we have shown in eulachon. In Lake Huron male rainbow smelt demonstrated higher natural mortality (M =1.3) than females (M =0.9) during spawning (Frie and Spangler 1985). This shows that females have a higher capability to spawn again, which is important as fish generally demonstrate higher fecundity as they get larger. The similarity that exists between eulachon and anadromous rainbow smelt life-history traits is considerable. Both species spawn in coastal streams, deposit adhesive eggs, and larvae are carried to the ocean immediately after hatching. The ayu, is another osmerid smelt; however, it displays an amphidromous lifehistory. These fish move between fresh and salt water for purposes other than spawning. The ayu is common in eastern Asia and demonstrates a one-year life cycle. These smelts are similar to eulachon in that they also spawn in the lower reaches o f rivers and the larvae are then carried to the ocean after hatching. Ayu generally spawn in the autumn after migrating downstream to the lower river reaches (Nishida, 1978, cited in Katano and Iguchi 1996). During their ocean residence, ayu mainly feed on zooplankton (Katano and Iguchi 1996). Throughout the spring, the young ayu migrate back into the river and forage on algae. Ayu are semelparous spawners that mature after one year regardless o f body size (McDowall 1992). This species is different from the rainbow smelt that show repeat spawning and more plasticity in tim ing of maturation. O ur data suggest that eulachon are most likely semelparous (like ayu), but show some plasticity in tim ing of maturation similar to rainbow smelt. 38 Delta smelt only occur In Sulsan Bay located in the Sacremento-San Joaquin estuary (Moyle et al. 1992). Spawning takes place in fresh water from late February to May, when the water temperature is between 7-15 °C. Eggs are adhesive (similar to rainbow smelt and eulachon) (Moyle 1976) and larvae are carried to the estuary after hatching. Delta smelt mainly feed on zooplankton in the estuary, staying in the estuarine environment until the following w inter when they migrate from 10-100 km upstream as maturing adults (Swanson et al. 1998). Delta smelt mature at 55-80 mm of length and most adults die after spawning, having completed a one-year life cycle (Moyle et al. 1992). Delta smelt are listed as a threatened species; habitat disturbance is likely the main factor for the decline of the delta smelt. Sacremento-San Joaquin River system is extremely disturbed and most of the anadromous and resident fish in this system have declined severely in recent years. Entrainment o f water resulting from diversion appears to be the most significant cause of this decline (Moyle et al. 1992). Delta smelt lifehistory is remarkably similar to that of the eulachon. All of the above examples suggest that the life-history we have proposed for eulachon is consistent with the previously documented characteristics of related smelts. Age/ng The seasonal fluctuations in Ba:Ga observed in this study suggest that, to date, eulachon may have been aged incorrectly. An examination of the zonation in whole and transversely polished eulachon otoliths provides an example of the problems 39 encountered when interpreting age (Figure 2.6). W hole otoliths possess numerous dark bands that have been interpreted as w inter growth zones in past ageing attempts. Conversely, some sectioned otoliths viewed under transmitted light reveal fewer zones. Most sectioned otoliths were difficult to interpret, suggesting that ageing by this method is also problematic. Polishing tim e and thickness greatly affected the readability of the structures and seemed to vary among samples. The use o f seasonal variation in elemental signature, therefore, represents an attractive alternative when ageing eulachon otoliths. Campana (2001) discussed the application of using elemental and isotopic signatures for confirming the ages of growth structures as simply corroborating the periodicity of growth increments. This speculation developed from the assumption that any changes in the deposition resulting in noticeable growth increments would, in theory, also result in changes in chemical signatures. The hypothesis appears valid; however, the problem with ageing eulachon in the past has come from identifying the specific increments in otoliths that correspond to annual zones. Eulachon otoliths possess frequent pseudo-annuli (visible increments formed by an unknown process), making ageing extremely difficult. The goal of this study, therefore, was to determine if chemical signatures could help elucidate the presence of annual increments. Ba:Ca profiles examined in this study likely represent oscillations that correspond to seasonal variations. Capture dates were reported for the eulachon examined in the 40 present study. Fish captured in late July in the ocean had peaks of Ba:Ca at the outer edge of the otolith, characteristic o f the maximum values measured in the otoliths. Conversely, fish captured between February and March demonstrated Ba:Ca levels in the outer edge of the otolith that represented the m inim um values measured. The relationship suggests that eulachon are incorporating higher concentrations o f Ba:Ca during the summer and lower concentrations o f Ba:Ca during the winter. There are two possible explanations for the seasonal fluctuations observed: that the actual concentration of Ba:Ca fluctuates on a seasonal basis or that BaiCa incorporation is regulated by temperature dependent processes. Bath et al. (2000) demonstrated that Ba:Ca uptake into the aragonitic matrix of the otolith is proportional to the concentration of Ba:Ca in the am bient environment. Fluctuations in this ratio observed in the otoliths, therefore, should correspond to fluctuating Ba levels, presumably on a seasonal basis. Barium in offshore seawater is very low with concentrations between 10 and 45 nM, while riverine Ba inputs can be tw o to 10 times higher (Chan et al. 1977; Guay and Falkner 1998). Additionally, estuaries can be augmented by Ba discharge (freshwater inputs) and experience high concentrations (Li and Chan 1979). The Ba:Ca profile in this study started high potentially corresponding to a freshwater signal, progressively declined to a magnitude representative o f the marine environment, and then w ent through variable cycles that may represent inshore versus offshore movements. It is also possible that the peaks in Ba:Ca found in the otoliths correspond to times when riverine output is high, such as spring freshet. This could 41 explain the variation in Ba:Ca observed, as the highest levels of Ba:Ca measured in the otoliths do correspond to times when river output is the highest (May-July). An alternate explanation for the variation observed in Ba:Ca incorporation in eulachon otoliths could be that branchial (gill) uptake is mediated by temperature dependent processes. Elsdon and Gillanders (2002) determined that the concentration ratio of Ba;Ca increased significantly in juvenile black bream (Acanthopagrus butcheri) otoliths with increasing ambient water temperature. Annual sea surface temperatures recorded on the west coast of British Columbia show a seasonal oscillation that corresponds to lower water temperatures in the w inter and higher water temperatures in the summer. Data from the Pacific Region State o f the Ocean Report (2003) by Fisheries and Oceans Canada shows that this trend has been stable for many years in both southern and northern sampling locations. The oscillating Ba:Ca ratios in eulachon otoliths may be a reconstruction of life-history information based on the temperature to which the animals were exposed. The seasonal change in Ba:Ca levels measured in eulachon otoliths could be an additive effect of both increased freshwater inputs in the spring and seasonal ocean temperature; however we feel that the seasonal oscillation in temperature is the most likely cause of the Ba:Ca changes measured in the eulachon otoliths. Strontium xalcium ratios did not appear to vary w ith temperature in eulachon otoliths as the profile was flat profile throughout their life-history. The finding that Sr:Ca 42 did not vary w ith temperature in a similar manner to Ba:Ca is surprising. A possible explanation is that Ba uptake through Ca channels in the gill epithelium may be more sensitive to changes in temperature. Bath-Martin et al. (2004) determined that the Sr:Ca partition coefficient increased linearly w ith temperatures from 17-26 °C. The water temperatures that eulachon are exposed to are much lower and range from approximately 8-13.5 °C. It is not clear whether water temperature affects Sr uptake in eulachon at these temperatures. Elsdon and Gillanders (2002) found that the ratio of Sr:Ca was greater at both low and high temperatures, but lower in moderate temperatures. Additionally, Sr incorporation into the otoliths is likely mediated by more than just temperature. Strontium concentration is much higher in seawater (~ 8 ppm) than Ba concentration (—0.015 ppm). Abundance o f Sr in the marine environment may be at a level where branchial uptake is maximal, regardless of small changes in physiological uptake mechanisms that may result from temperature fluctuations. In addition, there were constraints on the resolution of specific elements due to the way that elements from otoliths were measured in our study. Subtle fluctuations in Sr due to temperature change may not have been detectable using our analytical methods. O ur data relied on reference material that does not match the matrix o f a fish otolith, as no such material is currently available for Laser Ablation investigations. Bath-Martin et al. (2004) utilized a sector-field ICPMS w ith dissolved otoliths and could rely on more appropriate reference material. Dissolving samples was not an option in our analysis, as we wanted the spatial resolution that is available when using a probe-based instrument. 43 The results of our study suggest that there is a specific size that eulachon must reach prior to the onset o f maturity. If the maturity size threshold is correct, it could explain the variable age classes observed during spawning. The Ba:Ca profile for the Skeena, Kemano, and Fraser Rivers suggest that most eulachon spawn at age three, with some members spawning at age four. Ricker (1954) and McHugh (1939) aged Fraser River eulachon and determined that the most common age at spawning was tw o years with some fish spawning at age three. Ricker felt that ageing otoliths and scales from eulachon was unreliable due to the difficulties in interpretation o f annuli. Ricker's conclusion on a two-year life cycle for Fraser River eulachon was determined through an examination of reproduction success for odd and even years. Eulachon runs spawning in odd years were, on average, higher in density than in even years. The observation correlates with the odd-even alternation in Fraser River pink salmon (Oncorhynchus gorbuscha). Ricker (1954) speculated that eulachon emerging in the spring o f even years had to compete with the strong odd-year-returning pink salmon fry, which also emerge in the spring of even years. Ricker (1954) also examined a large flood event in the spring o f 1948, where he hypothesized that excessive shifting of the river bottom likely damaged eulachon eggs and lowered survival rates. Interestingly, the lowest return for the period between 1939 and 1953 was the spring of 1950, further suggesting a two-year life-cycle. Potentially, the discrepancy between Ricker's (1954) results and our study, where we found most eulachon to be three years at maturity in the Fraser, is related to ocean productivity. Some of the fish in Ricker's study matured at three years of age 44 suggesting that most o f the fish in the Fraser during 1939-1954 reached the m inimum size needed to mature sooner than the fish examined by us. Columbia River eulachon are an exception in age at spawning as it appears that most fish spawn at age two, with only some members spawning at age three. A possible explanation for this is that most eulachon generally spawn in the Columbia River in late January (Hay and McCarter 2000). Progeny of these fish would, therefore, spend more tim e in the marine environm ent when productivity is the highest (Ware and Thompson 1991). Utilization o f a w inter spawning period may allow Columbia River eulachon the ability to reach the m inimum size required to trigger the onset of gonad maturation earlier than eulachon populations that spawn during the spring. Hay and McCarter (2000) suggest most eulachon spawn at age three. Their finding is a result of an examination of offshore eulachon size classes captured during May 1997 and May 1998, where there is a notable bi-modal age distribution. Hay and McCarter (2000) feel the modes represent age one fish and age tw o fish; consistent with our findings that most eulachon spawn at age three. One potential problem with their analysis is that the modal ages for 1997 and 1998 are not consistent and show both inter­ annual variation and overlap. The authors attribute this to different rearing conditions, population differences, and geographic differences. Hay and McCarter (2000) also indicate that there are significant differences in eulachon growth rates between populations, further complicating length frequency data. Additionally, it is also suggested 45 that eulachon larvae from m ultiple populations mix in different locations. Size classes, therefore, would overlap throughout the marine residency tim e for eulachon. Interestingly, we have estimated that the ocean fish from Area 23-6 provided to us by Dr. Hay to be 1.5 and 2.5 years old. There is a noticeable difference in size for the tw o age groups of ocean fish, suggesting that tw o modal sizes of eulachon do exist simultaneously offshore. Repeat Spawning Potential Models developed to explain life-history patterns in fish indicate that there are a number of substantive differences between semelparous and iteroparous species. Optimal life-history strategy is a function of both fecundity and the relative survival rate of both adults and juveniles (Charnov and Schaffer 1973). It is generally accepted that there is a trade-off between future growth and reproductive effort. Each individual has a finite energy budget, lim iting the amount of resources that can be directed to reproduction if the animal has a chance of survival in the future. In general, iteroparous life-histories are characterized by a long adult life, small clutch size, delayed maturity, large eggs or young, and some form of parental care. Animals with semelparous life histories have a short adult life, large clutch size, small eggs, and no parental care. As well, fish populations with lower densities tend to be iteroparous, although dense populations are semelparous. O ur observations suggest that eulachon fit the criteria 46 describing a semelparous life-history because eulachon are short lived, have high numbers of small eggs, no parental care, and spawn in very high densities. The Sr:Ca profiles determined in our study also indicate that eulachon are most likely semelparous, as there are no corresponding declines in magnitude that would be due to freshwater movements. Strontium:Calcium ratios are significantly higher in the marine environment than the freshwater environment (Kalish 1990). There have been many studies indicating that migrations from freshwater to estuaries or the marine environment can be detected by high Sr:Ca ratios in bone, otoliths, scales, and pectoral fin-rays (Kalish 1990; Coûtant and Chen 1993; Veinott et al. 1999). Factors other than salinity also influence Sr:Ca ratios in some calcified structures. Seasonal changes in reproductive physiology have resulted in variations in SnCa ratios (Kalish 1991). In addition, temperature changes in the ambient environment have resulted in changes to the Sr:Ca ratios (Townsend et al. 1989; Townsend et al. 1995). However, the magnitude of the changes in Sr:Ca ratios when fish migrate from freshwater to seawater are so much greater than variations due to physiology change or temperature. Veinott et al. (1999) examined Sr concentrations in Fraser River white sturgeon (Acipenser transmontanus) pectoral fin-ray annuli using LA-ICPMS and determined that some individuals do make marine migrations into the estuary. Additionally, Howland and Tonn (2001) investigated the Sr profile in inconnu {Stenodus leudchthys) from the Mackenzie River drainage and were able to distinguish between fish that were represented by entirely freshwater, partially anadromous, and completely anadromous life histories. For these reasons, we 47 expected to see a change in Sr:Ca ratios when eulachon migrated from seawater and into freshwater environments. There was little change, however, in the SnCa ratios incorporated into the otoliths of eulachon over tim e (Figure 2.6). Despite the short residence tim e in freshwater, migration to spawning areas should reflect a decline in Sr:Ca measured in the otolith, particularly in areas such as the Fraser and Columbia Rivers where there are heavy influences o f freshwater discharge far into the estuary. Furthermore, northern populations of eulachon make extensive migrations through coastal inlets where a decrease in the magnitude of Sr exists from high freshwater inputs. It appears that eulachon spend too short a time in the river during spawning for a distinct freshwater signature to be incorporated into bony structures. In addition, there is anecdotal evidence that much o f the tim e spent in the river before the fish spawn is within the saltwater wedge. The short duration in freshwater and preference for the deeper saltwater wedge appears to preclude a measurable freshwater Sr signal using our probe based technique; as the signal incorporated during the short spawning period would be at the outer edge o f the otolith and therefore difficult to detect w ith our spatial constraints. Previous observations have shown that some eulachon return to the ocean soon after spawning in the Dean and Burke channels (Pootlace and Siwallace 2000) and in the Gardner Channel from the Kemano River (Adam Lewis, pers. comm). These observations, however, do not rule out a semelparous life-history. There may be evolutionary advantages to moving back into coastal inlets prior to senescence, as most of 48 the eulachon larval growth occurs in this environment. Eulachon do not rear in the freshwater environment but are immediately swept downstream into the ocean after hatching (Hay and McCarter 2000). The im portant nutrients contributed by the large biomass of dead eulachon in the inlets may be im portant in sustaining early growth at this important developmental stage. CONCLUSION Otolith microchemistry has provided valuable information about the life histories o f many fish species in recent years. The spatial resolution of LA-ICP-MS allows for a fine-scale analysis of the chemical environment experienced by fish. Analysis using otolith microchemistry allowed us to address some aspects of eulachon life-history that have been difficult for other researchers to answer to date. The results o f our study suggest that most eulachon spawn at age three and are semelparous. Future investigations into eulachon life-history should involve more samples with females to confirm our findings. Regardless, it seems highly unlikely that one sex would be iteroparous, as there would be a highly skewed sex ratio which has not been observed in eulachon populations. The finding that eulachon are approximately three years of age at spawning contradicts some previous ageing investigations. Difficulties in estimating age for eulachon have been identified. W e propose to conduct further analyses on eulachon captured from the Copper River and one of its tributaries in Alaska. These two 49 populations spawn in January and May, respectively. If Copper River eulachon that spawn in January are one year younger than the cohort that spawns in May, our hypothesis that eulachon need to attain a m inimum size prior to maturation w ill be supported. Independent age analyses of these Copper River eulachon suggests that January spawners are four years old and May spawners are five years old. Perhaps these ages reflect the same errors observed for British Columbia eulachon and the fish are two and three years old respectively. 50 Chapter 3: Discrim ination of Habitat use by Slimy Sculpins (Cottus cognatus) in Tributaries of the W illiston Reservoir using Natural Elemental Signatures ABSTRACT Trace-element analysis o f bony structures is a technique that has been used to identify location o f origin in freshwater fish. This approach may provide fisheries managers with an additional tool when trying to understand the complex movement patterns and life histories o f many species. The aim of our study was to correlate chemical signatures deposited in slimy sculpin (Cottus cognatus) otoliths with those measured in the streams where fish were captured. Initially we assessed the chemical stability of water w ithin a river and the differences among rivers for 27 streams w ithin the Williston Watershed, located in northern British Columbia. Stream chemistries remained stable over the duration of the project according to measured values for six sampling events. In addition, Laser-Ablation-lnductively-Coupled-Plasma-Mass-Spectrometry (LA-ICP-MS) and cathodoluminescence (CL) microscopy determined that stream chemistries were consistent from year to year, as the elemental profile was constant for several years of growth in the otoliths of slimy sculpins. Canonical discriminant function analysis demonstrated that streams examined within the watershed were heterogeneous and that each river could be differentiated. Elemental signatures measured in the otoliths of sculpins sampled in the project were highly correlated to the stream chemistries where the fish were captured. The Incorporation Coefficient (1C), molar ratio o f an element to Ca in the otolith over the molar ratio in the water (e.g. [SriCaotoiiJ / [Sr:Ca^ater]) was 51 calculated: Sr:Ca = 0.21 (SE = 0.0041), Ba:Ca = 0.019 (SE = 0.0013), Mg:Ca = 0.00012 (SE = 5.9E-06), Mn:Ca = 0.31 (SE =0.043). Both Sr:Ca and Ba:Ca ratios in otoliths were highly correlated to stream chemistries. Mg:Ca ratios were weakly correlated to stream chemistry, while Mn:Ca ratios showed a marginal correlation. Multivariate analysis of the chemical fingerprints in the otolith determined a significant relationship to water sample sites. O ur study was successful in discriminating among slimy sculpin populations collected from different streams. Probabilities for correct classification of sculpins to their streams of capture were 100% for Bills Creek, Osilinka River, Manson River, and Davis River. Separation o f Anzac and Table Rivers were classified correctly 93% of the time. The extension o f the methodology used in this project to a migratory species in the Williston watershed appears reasonable. Future studies would benefit from the inclusion o f both fin-rays and scales to enhance discrimination of habitat utilization. The inclusion of additional structures may allow this technique to be used w ithout sacrificing the animals. This is of particular importance when life-history information is required on a threatened or endangered species. 52 INTRODUCTION The headwaters of the Peace River in northern British Columbia offer an opportunity to examine how elemental signatures may be used to link habitat use of fish or identify stream of residence. Preliminary examination of the W illiston Watershed shows that bedrock geology differs between the east and west sides o f the Reservoir, as well as between the north and south (Rutter 1976). Bedrock formations in the Williston watershed were formed during different time periods, including the upper and lower Jurassic, Cretaceous, and Triassic, and are composed of materials that have variable chemical composition (Armstrong 1979). A small fish species, the slimy sculpin {Cottus cognatus), was utilized in this study to determine if chemical ratios incorporated into bones are geographically distinct in tributaries o f the W illiston Reservoir. Slimy sculpins are abundant in the Peace River watershed and are considered to be a non-migratory species, making them a suitable candidate for this study. Very little research has been conducted on this species in British Columbia. It has been documented that slimy sculpins reach a maximum size of 12.1 cm and age o f seven years (M orrow 1980). This small fish occupies a variety o f habitats including fast-flowing cold streams and rocky areas of lakes; it is even tolerant o f brackish water. Slimy sculpins are also noted to inhabit areas with a high amount of groundwater influence, as well as small springs (Page and Burr 1991). In this study, we investigated the potential to discriminate locations of capture for fish within the W illiston watershed using natural chemical tags. The objectives of our 53 study were to investigate the stability of the stream chemistry throughout the year w ithin a group of Parsnip River tributaries, to investigate the heterogeneity o f stream-specific chemical signatures and determine if we could discriminate among locations of rivers within the Parsnip River Watershed and tributaries throughout the W illiston Reservoir Watershed, and to ascertain correlation in elemental signatures between stream of capture and the otolith o f slimy sculpins. METHODS Study Location This study was conducted w ithin tributaries of the Williston Reservoir located in north-central British Columbia (Figure 3.1). The Williston Reservoir is the largest body of freshwater in British Columbia and the reservoir was formed following com pletion o f the WAC Bennet Dam on the Peace River in 1967. The Reservoir now drains an area representing 70,000 k m \ Water Collection The stability o f the stream-specific chemical signatures was determined by collecting water samples, on six separate occasions from June 20, 2002 to November 15, 2002, from four locations w ithin the Parsnip River mainstem and 11 Parsnip tributaries. 54 Access and icy river conditions prevented additional sampling during the remaining portion of the winter. To determine the heterogeneity of streams and rivers w ithin the Williston Watershed, water samples were obtained from a total of 27 geographically distinct locations. These streams included the 15 sites sampled multiple times w ithin the Parsnip watershed and 12 streams sampled a single tim e located throughout the reservoir watershed (Figure 3.1 ). The methods chosen for obtaining water samples followed the recommendations outlined by Shiller (2003) for sampling trace elements in remote locations, with some m inor modifications. High density polyethylene bottles (50 ml) (Fisher Brand) were cleaned with ultra-pure water and filled w ith a solution o f 2% high purity nitric acid and left for a m inimum o f tw o weeks. Bottles were then rinsed five times in ultra-pure water. Polyethylene/Polypropylene 50-ml syringes (Sigma Aldrich) were cleaned in the same manner as the 50 ml bottles. Nylon filters (25 mm x 0.45 /t/m, Fisherbrand) were cleaned by passing 40 ml of a solution of 2% high purity nitric acid followed by a rinse of 20 ml of ultra-pure water. All filters were blown dry with high pressure clean air and left under a fume hood until use. For field sampling, all equipment used at each site was placed into an individual clean Ziploc bag (sampling kit). The kit contained tw o filters (replicate samples), two 50 ml bottles, one syringe, polyethylene gloves, and an extra Ziploc bag for collected water samples. W ater samples were acidified to a solution containing 2% high purity nitric acid (600 / j I H N O 3/ 3 O ml water sample) in the field immediately after collection. 55 150 klom#o«m Figure 3.1 Map of the Williston Reservoir showing main tributaries. The numbered tributaries include: Table River (1), Tacheeda Creek (2), Bill's Creek (3), Hominka River (4), Misinka River (5), Wichika Creek (6), Swan ne 11 River (7), Davis River (8), Pelly Creek (9), Factor Ross Creek (10), Co Ibo urne Creek (11 ), Reynolds Creek (12). The other tributaries sampled include the Parsnip River (4 locations), Anzac River, Wooyadilinka Creek, Misinchinka River, Nation River, Manson River, Omineca River, Osilinka River, Mesilinka River, Ingenika River, and the Finlay River. 56 Water Analyses Water analysis was completed with a PS 1000-UV Inductively-Coupled-PlasmaOptical-Emission-Spectrometer (ICP-OES) (Leeman Labs), at the University of Northern British Columbia. The elements measured included Ba, Ca, Sr, Mg, and Mn. Four calibration standards prepared from traceable (NIST) standards were run for every 10 samples analyzed. Laboratory blanks and field procedural blanks were also included in the analysis. Additionally, samples were analyzed at the University of Victoria with a solution based Inductively-Coupled-Plasma-Mass-Spectrometer (ICP-MS) as an additional quality control analysis. Fish Collection To correlate chemical signatures present in the water with those in fish otoliths, several rivers were chosen to be representative of the different geographic regions o f the watershed. W e chose rivers on the east side of the reservoir that originate from the Rocky Mountain range, as well as rivers on the west side that flow from the Coast Mountain Range. The rivers on the west side o f the reservoir included in this portion of the study were the Manson, Swannell, and the Osilinka . Rivers on the east side of the reservoir included the Table, Anzac, Davis, and Bill’s Creek. M innow traps were used to capture sculpins from the Swannell and Manson Rivers. M innow traps proved to be 57 unreliable; therefore, only two sculpins from the Swannell River and one sculpin from the Manson River were collected. A Smith Root 12B backpack electrofisher was used to capture sculpins from the Table River, Anzac River, Bill’s Creek, and Osilinka River during July 2002. One sculpin, found dead on the bank of the Davis River, was donated by the Ministry of Water, Land & Air Protection and included in the analysis. Fish and water samples were collected from the same area of each river system. Fork length (mm) and weight (g) were recorded for each fish (n=41 ) prior to removal o f the sagittal otoliths. Sagittal otoliths were chosen, as they are the largest of the three pairs of otoliths found in teleosts. O tolith Chemistry Otolith preparation, analysis, and data reduction were completed with the same protocol as Chapter 2. All otoliths were photographed in order to determine at what age chemical signatures were variable w ithin the otolith. Cathodoluminescence Microscopy Cathodoluminescent imaging was performed by focusing a high-energy beam of electrons (20 kV) onto a polished otolith placed in a chamber under vacuum. The incident electrons cause bound electrons to rise to higher energy levels. When the 58 electrons return to their original state, they release the energy and luminesce. The wavelength o f the light emitted is specific to each element. Light emitted by the sample is collected with achromate lenses and guided to the entrance slit o f the spectrometer. Images were acquired with a one-minute exposure onto a cooled colour CCD camera (Q-Imaging, Burnaby, BC, Canada). Statistical Analyses Linear regression was used to determine the relationship between otolith chemistry and water chemistry (SPSS v .l 1.5, Sep 2 0 0 2 , Chicago, IL). As well, the tracemetal chemistry in the otoliths was related to the trace-metal chemistry in the water using an ‘incorporation coefficient’ (Wells et al. 2000b). The coefficient was calculated as the molar ratio in the otolith over the molar ratio in the water (e.g. [Sr;CaotoiiJ / [SriCa^^ter])Chemical signatures for both the rivers and the sculpins were differentiated using Canonical Discriminant Analysis (SPSS v .l 1.5, Sep 2002, Chicago, IL). Canonical discriminant analysis was utilized to provide a visualization of geographic separation using only water chemistry data. A multivariate combination of Sr:Ca, Ba:Ca, and Mn:Ca molar ratios was utilized to discriminate the sculpins in this study, w hile a combination of Sr:Ca, Ba;Ca, Mg:Ca, and Mn:Ca was utilized to discriminate tributaries of the Williston Reservoir. The elemental ratios were the factors loaded into the discriminant function (DCF) in order to produce DCF1 and DCF2. Manova (SPSS v .l 1.5, Sep 2002, Chicago, 59 IL ) was also used to test for significant differences in multi-element ratios between sample locations, under the hypothesis that there is no significant difference in chemical fingerprints between populations and locations. RESULTS Stability o f Water Chemistry For the 15 sample sites where water samples were collected from the same locations six times from June 20, 2002 to November 15, 2002, the absolute concentrations of elements found w ithin each system were observed to change over time, however, the ratios of the individual elements to the concentration o f Ca (e.g. Sr:Ca) in the river systems remained stable over the sampling period from June 2003 until November 2003. Elemental ratios for the Parsnip River and tributaries are shown in Table 3.1. Small variations observed for each river reflect the low amount o f deviation in elemental ratios measured over time. Most o f the variation that was observed was due to lower than average values measured during high flow conditions associated w ith spring freshet. Figure 3.2 and 3.3 show the concentration of elements and the ratio o f elements to Ca for the Table and Anzac Rivers, respectively. Fluctuations in mean Ba (ppm) and Sr (ppm) are apparent on a seasonal basis; however, ratios of these elements to Ca remain fairly stable over time. 60 Table 3.1. Average element ratios (mmol/mol ± SE) measured by ICP-OES for each tributary over the sampling period from June - November 2002. Most tributaries of the Parsnip mainstem were sampled six times. Values for tributaries with no error associated were only sampled once. L o c a t io n SnC a B a :C a M g :C a M n :C a Parsnip Glacier 4.37 0.38 127 4.61 Parsnip Upper 3.67 ± 0 .0 8 0.37 ±0.01 163 ± 2 2.15 ± 0 .3 5 Parsnip H/M 3.90 ± 0.06 0.49 ±0.01 181 ± 4 1.07 ± 0.13 Parsnip A/T 3.42 ± 0 .1 0 0.90 ± 0.06 192 ± 1 1.50 ± 0 .1 4 Parsnip Lower 3.39 ± 0 .0 6 0.64 ± 0.04 178 ± 2 0.41 ± 0 .0 4 Hominka River 4.24 ± 0.11 0.32 ± 0.01 287 ± 6 1.75 ± 0 .0 7 Misinka River 2.43 ± 0.09 0.62 ± 0.01 307 ± 4 1.56 ± 0 .2 7 Wichika Creek 4.01 ± 0.03 1.17 ± 0 .09 97.5 i :3 0.08 ± 0.01 Bills Creek 3.08 ± 0.03 3.44 ±0.11 117± 1 0.21 ± 0.01 Wooyadilinka Creek 3.72 ± 0 .0 5 0.29 ±0.01 965 ± 1 0.06 ± 0.01 Tacheeda Creek 3.32 ± 0.13 3.50± 0.11 186 ± 5 0.37 ± 0 .08 Table River 2.95 ± 0.09 0.42 ±0.01 187 ± 3 1.93 ± 0 .0 9 Anzac River 3.47 ± 0 .0 7 0.33 ± 0.02 151 ± 2 0.66 ± 0.03 Reynolds Creek 2.92 ± 0.50 0.46 ±0.01 221 ± 4 0.93 ± 0.05 Colboume Creek 4.04 ± 0 .1 2 0.45 ± 0.02 226 ± 7 1.41 ± 0 .1 0 Misinchinka Creek 3.01 ± 0 .1 0 0.25 ± 0.01 186 ± 3 1.01 ± 0.13 Davis River 223 0.33 316 0.05 Findlay River 4.18 1.28 253 0.08 Pelly River 7.13 0.54 334 0.03 Ingenika River 6.28 0.51 258 0.22 Swannel River 527 0.88 328 0.35 Factor Ross River 328 3.14 190 0.68 Mesilinka River 3.79 0.74 102 0.15 Omineca River 5.79 1.95 182 0.29 Osilinka River 7.13 3.15 190 0.68 Manson River 4.73 1.44 185 0.19 Nation River 4.64 1.63 182 0.14 61 Table River E 0.050 -1 Q. 0.040 c o ro c 0.030 0.020 8 c o O O 0.005 o O O c o o e Ba Sr O • Ba/Ca Sr/Ca (U E O# Oo 8 -6 — -8 - r~ -8 -6 -4 “ I -2 T " “T “ T T " “T T 0 2 4 6 8 10 1 12 Discriminant Function 1 Figure 3.8 Canonical discriminant function analysis showing how sculpins from individual populations clustered according to the multivariate combination of Sr;Ca, Ba;Ca, and Mn:Ca (discriminant factors) measured in each otolith. Sample sizes vary for each river system (Table n = 8, Anzac n = 10, Bill's n = 11, Manson n = 1, Osilinka n=8, Swannell n = 2, Davis n = 1). 72 Multivariate statistical examination revealed a significant difference between the natural-log-transformed molar ratios o f Sr:Ca, Ba:Ca, and Mn:Ca (dependent variables) for the populations o f sculpins, implying that fish from each stream incorporate a specific chemical signature. The following results were obtained: Pillai - Bartlett Trace = 2.3, d f = 18, 105, F = 19.169, and P < 0.001. Independent factors include the Table River, Anzac River, Bill's Creek, Osilinka River, Manson River, Swannell River, and Davis River. CL microscopy revealed an interesting visualization of trace element deposition in slimy sculpin otoliths (Figure 3.9). The Table River fish have a definite w hite luminescent region that corresponds to the first year of the fish’s life. Anzac River fish and Bill’s Creek fish are characterized by a similar pattern as the Table River; however, there appears to be more fluctuation in trace metal deposition early in life. Osilinka River fish have distinct luminescent regions that are likely correlated to seasonal or yearly movements, or possibly areas highly influenced by ground water. Swannell River sculpins appear to live in a much more chemically constant environment, as there is little to no change in luminescence throughout their lifehistory. 73 Figure 3.9 Representative cathodoluminescent images of otoliths from sculpin captured in the Table River (17), Anzac River (A10), Bill's Creek (B5), Osilinka River (05 & O10), and Swannell River (51 ). 74 Figure 3.9 continued 75 DISCUSSION Our study contributes to a growing field of research investigating the benefit of using natural chemical markers to examine life-history and movement patterns offish in freshwater ecosystems. The streams sampled in this study maintained chemical stability over time and exhibited sufficiently different chemical signatures to allow differentiation based on geographic locations. Additionally, the elemental ratios measured in the fish otoliths were highly correlated to the stream ratios determined where sculpins were captured. Stability o f Elements Tributaries of the Parsnip River remained chemically stable over the duration o f sampling in this project. This was particularly true when examining the element to Ca ratio. Fluctuations were observed in the mean element concentration for each stream; however, element ratios remained remarkably constant. Element ratios were slightly lower during spring freshet (high flow conditions) for the tributaries examined, though this is of limited concern, as most growth for bony structures offish has been documented to occur during base flow conditions (Kennedy et al. 2000). W ater stability was an im portant com ponent of this study, as a large amount of temporal variation in stream chemistry would ultimately confound classification o f individuals to their natal stream. This consistency among rivers is well described by Taylor and Hamilton (1994) w ho examined 25 years of water chemistry data on the Saskatchewan River. W ater was not sampled during the winter months in our study. 76 This should not be problematic as this period only represents a tim e o f hydrograph draw-down. Therefore, element ratios should remain consistent for this period of time, even if element concentrations rise. Consistency in chemical signature is further supported by the line scans tracked across the otoliths of most individuals, suggesting stable water chemistry. Slimy sculpins are considered to be non-migratory, so if the water chemistry remains stable over extended periods o f time, the chemical deposition in the otolith should also remain stable for elements that are incorporated proportionately. W e found this to be true, as the chemical signature maintained a flat profile throughout the life o f the individual. This period varied from tw o to five years, depending on the age of the fish. Variation in stream chemistries on an annual basis would have resulted in fluctuating elemental fingerprints w ithin each otolith; therefore, it appears that chemical signatures have been stable over several years for the tributaries that we sampled in the Williston Watershed. The region surrounding the core of the otolith for some o f the populations examined, however, showed variation in chemical signatures. Surrounding the core was a noticeable increase in trace metal concentration for Ba:Ca (Figure 3.5) and Mn:Ca. The increase in Ba:Ca and Mn:Ca was seen in sculpins from Bill’s Creek, Table River, and the Anzac River. CL microscopy also demonstrated that there was a significant change in trace metal concentration surrounding the core of the otolith; though, the core itself had a similar chemistry to that seen in the later stages o f the sculpin’s life for both methods. It is possible that we are observing a maternal signature incorporated into the core of the otolith o f larval sculpin when they are still 77 dependent on the egg for nutrition. This is supported by the finding that adult salmonids pass on a maternal elemental signature to juveniles through the egg (Weber et al. 2002; Zimmerman and Reeves 2002). It appears that adult sculpins move into a distinct habitat such as a small stream or an area highly influenced by groundwater to spawn. The juvenile sculpins then rear in this habitat for one to tw o years until moving back into a separate habitat. Geographic Separation Streams w ithin the W illiston watershed were sufficiently heterogeneous in water chemistry to discriminate their locations. Not only were the streams heterogeneous, they appear to form tw o main groupings for the east and west side (longitudinal) of the reservoir, as well as a latitudinal distribution from north to south (Figure 3.4). O ur data support a recent study by Wells et al. (2003) that was also successful in discriminating geographic locations using chemical fingerprints measured in a freshwater environment. Variation in the freshwater environment appears to be common within watersheds, which is not surprising as the geology o f most regions varies spatially due to differences in age and composition of bedrock; these differences result in variable stream chemistries. The elemental composition of sculpin otoliths also differed significantly among sample locations. A combination of Sr:Ca, Ba:Ca, and Mn;Ca allowed adequate separation of all of the populations examined; but, the Table and Anzac Rivers show 78 marginally less separation. This is not surprising, as they are parallel streams that are only separated by a few kilometers. A potential element for inclusion into the analysis is Se, which was only detected in the Table River and is incorporated well into fish scales (Farrell et al. 2000); therefore Se may provide additional resolution between these two systems, as it is only present in detectable levels in the Table River. Magnesium was not utilized to classify individuals because the incorporation into the otolith did not correspond well with the water from sample locations. As Mg does not show a definitive relationship with water chemistry, it may be of limited utility in determining life-history with otolith microchemistry. Magnesium has an upper lim it for incorporation into the aragonitic matrix of the otolith; therefore the Mg levels in all the streams measured may be higher in concentration than the fish is able to incorporate, resulting in a poor relationship between stream Mg concentration and otolith Mg concentration. Wells et al. (2003) found a negative relationship between Mg:Ca in the otolith and Mg:Ca in the water w hile we found a slightly positive relationship. The small differences in the findings observed between the two studies may be due to species-specific incorporation o f trace elements (Campana et al. 2000). The methodology used in our study may be further improved with the inclusion of fin-rays and scales in the analysis. Mg has been demonstrated to be incorporated well into bull trout fin-rays (see Chapter 4), and Mn is incorporated well into scales (Wells et al. 2000b). The use of a complementary method may also 79 enhance resolution between streams. CL microscopy revealed differences in luminescence among streams. The results indicate that there is potential to discriminate sculpin stocks simply by examining the luminescent pattern created with CL emissions, particularly when combined with LA-ICP-MS. CL microscopy is a particularly valuable tool for analysis, as it is extremely cost effective. Not all o f the studies conducted to date have agreed with either our results or w ith those obtained by Wells et al. (2003). A study examining barramundi {Lates calcarifer) was unsuccessful in discriminating movement patterns using Cu, Mn, Zn, Cd, and Pb w ithin the Fly River, Papua New Guinea (Milton et al. 2000). Possibly, the elements chosen for the study were not suitable as some elemental species form colloids in the water, lim iting their bio-availability (Benes 1979; Gaillardet et al. 2003; Peltier et al. 2003) or are incorporated into the matrix o f the otolith as some elements attach to Ca binding proteins (Milton et al. 2000; Campana and Thorrold 2001). Sr and Ba move into the otolith through Ca channels and are more likely to be incorporated based on their concentration in the environment (Kalish 1991). Incorporation coefficients we observed were 0.21 for Sr:Ca and 0.019 for Ba:Ca. The coefficient obtained for Sr:Ca was substantially lower than that obtained for Westslope cutthroat trout in the Couer d ’Alene River basin, where the value was 0.40 (Wells et al. 2003). The value we obtained, however, is similar to values measured in tw o experiments looking at the chemical uptake in juvenile spot {Leistomus xanthuras) (Bath et al. 2000; Wells et al. 2000a). The Ba:Ca incorporation coefficient determined by our study was lower than values obtained for both cutthroat trout 80 (Wells et al. 2003) and juvenile spot (Bath et al. 2000; Wells et al. 2000a). Wells et al. (2003) hypothesized that the differences observed between freshwater and marine species were likely a result of the different elemental uptake mechanisms between these two groups of fish. O ur data suggests that differences are more likely to be attributed to species rather than environmental effects, as our values for Sr:Ca reflect the values measured in a saltwater fish. One fish in this study, captured in the Osilinka River, was a notable outlier (Figure 3.7). This fish had a higher incorporation coefficient for both Sr;Ca and Ba:Ca when compared to the other fish in this study. Both the left and right otoliths were examined in this fish on separate occasions to ensure that sample preparation did not influence the results. It is likely that this fish was associated with a different habitat before it was captured in the Osilinka River. Potentially, this fish could have moved into the area from another stream or been associated with a unique groundwater source. Slimy sculpins are a benthic species (Scott and Crossman 1973) living within the substrate of the river system (Clarke AD, personal observation), so they are more likely to be in direct contact with groundwater percolating into the river system; salmonids reside higher in the water column where there should be more mixing of water. The present study demonstrated that otolith microchemistry is a valid method to discriminate the geographic location offish w ithin a number of tributaries of the W illiston watershed. W e believe this approach would be suitable for discriminating movements o f migratory fish, however, there is the potential for increasing the 81 resolution of the technique used. Classification accuracy of sculpins to their natal stream was excellent as indicated by the high predicted probabilities for each river system (ranging from 92.9% to 100%). Most rivers are clearly distinct, but there is the potential for some overlap in the Table and Anzac Rivers. As indicated earlier, the addition of lithium and selenium may increase the resolution o f this application for these two systems. Lithium is known to be incorporated into the otoliths of teleost fish in proportion to the environment (Campana et al. 2000). Lithium w ould be useful as it is only detectable in the Parsnip River. Both the Table and Anzac Rivers flow into the Parsnip making it a corridor for fish movement between these tw o systems. The utility of this technique could also be increased by analyzing otoliths, fin-rays, and scales for each individual fish. Scales and fin-rays incorporate some elements differently than otoliths. Farrell et al. (2000) determined that Ca, Mg, Zn, Hg, and Pb in Arctic grayling scales could be sufficiently resolved with LA-ICP-MS. Their study determined that these elements could potentially be utilized to rebuild past exposure to metals in the environment. Additionally, selenium was only present in the Table River at detectable levels. Farrell et al. (2000) demonstrated that when Arctic grayling are present in a stream that contains selenium, it is incorporated into the scales. 82 Chapter 4: M ovem ent Patterns o f Bull Trout (Salvelinus confluentus) in the M orlce W atershed using Chemical Signatures Deposited Spatially in Fin-rays ABSTRACT The potential to discriminate between fish that show large movements w ithin a watershed versus fish that show no or few migrations using natural chemical markers deposited in bony structures was examined w ithin the Morice River Watershed. Finrays were collected from 10 bull trout. The chemical make-up of streams known to contain bull trout in the M orice River watershed were measured once during this project, determining that the water chemistry of most o f the streams was similar and not useful for discerning movement patterns. The results o f our study suggest that Sr, Ba, Mg, and Mn are deposited in fin-rays proportionately to the concentrations measured in the water where the fish were captured. Incorporation coefficients were determined for the fin-rays: Ba:Ca = 0.020 (SD = 0.003), Mg:Ca = 0.36 (SD = 0.060), Mn;Ca = 0.19 (SD = 0.037), and Sr:Ca = 0.15 (SD = 0.014). Finding that trace elements are deposited in proportion in fin-rays to what is present in the fishes aquatic environment is im portant because it strengthens the assumption that fin-rays can be used to rebuild environmental life histories in a sufficiently chemically distinct aquatic environment. Seasonal oscillations in Zn:Ca deposition in fin-rays were also observed in this study. W e determined that these oscillations most likely correspond to w inter and summer growth and, thus, represent a method to approximate age using chemical signatures deposited spatially in fin-rays. The application o f using natural chemical tags deposited spatially in fin-rays to rebuild environmental life 83 histories offish, therefore, has the capability to complement or replace otoliths in future studies. Otoliths are preferred in chemical studies examining life-history, as they offer much higher spatial resolution; however, where populations offish are either threatened or endangered, fin-rays may provide an acceptable non-lethal alternative for life-history investigations. 84 INTRODUCTION Bull trout are currently declining in British Columbia and have been designated as blue-listed (endangered) by the provincial government. They are sensitive to habitat degradation and are considered to be an indicator of ecosystem health (Cannings and Ptolemy 1998). Bull trout are a cold-water species, generally found in streams with temperatures less than 12°C. This species is sensitive to human activities that change temperature, habitat, substrate composition, or migration patterns. Determining movement patterns o f bull trout will provide information to assess critical habitat requirements and, potentially, population structure. Bull trout movement patterns, population structure, and habitat utilization were assessed in a previous study using radio telemetry and genetic analysis (Bahr 2002). This work showed that bull trout could be spatially clustered by geographic region in the Morice River Watershed. Additionally, fish from each spatial cluster showed variable patterns of movement, some moving extensive distances (> 10 0 km) and some not moving beyond the stream reach they inhabited for the two-year study. The long distances and variability in movement shown by some bull trout w ithin the Morice River watershed offers an opportunity to examine if differences in movement can be detected using chemical markers. Fin-rays record a spatial and temporal scale similar to otoliths; however, the fin-ray matrix is made up of Ca3(P O j 2 whereas the otolith is mainly composed of 85 CaCOj. The use of fin-rays for rebuilding environmental life histories may be problematic, as these structures have the potential for resorption and mobilization o f trace elements after deposition (Veinott and Evans 1999). Veinott and Evans (1999) determined that K is not stable in fin-rays of white sturgeon (Acipenser transmontanus); however, Sr, Mg, Pb, Br, Zn, Ba, Sn, Mn, and Na remained stable for at least six years. Additionally, fin-rays have been used successfully to determine anadromous movements of white sturgeon from the Eraser River using Sr:Ca ratios spatially deposited in fin-rays to differentiate between freshwater and saltwater, two very distinct chemical environments (Veinott et al. 1999). A closely related species to the bull trout, the Arctic char (Salvelinus alpinus) has been examined extensively with otolith microchemistry to determine between migratory char moving between freshwater and marine environments and nonmigratory freshwater char (Halden et al. 1996; Babaluk et al. 1997). Both studies were successful in differentiating between anadromous and non-anadromous fish by examining Sr concentrations measured across the otolith. Arctic char otoliths also demonstrate well-defined seasonal fluctuations in Zn that correlate to the annular structure o f the otolith, thus enabling an age estimate (Halden et al. 2000). Zn uptake in Arctic char was shown to be higher in younger fish than older members of the population. Halden et al. (2000) hypothesized that Zn uptake was higher during the summer due to elevated water temperatures that resulted in increased nutrient production and intake. 86 In this study we examined the chemical signatures deposited in the pectoral fin-rays of bull trout. We compared results o f the chemical analyses to the telemetry data gathered by Bahr (2002) to compare the different techniques for assessing movement and population structure w ithin the watershed. Fin-ray chemical analysis has the potential to provide information over a much longer tim e scale than the telemetry data, as this technique should indicate movement patterns and habitat utilization over the entire life o f the individual, potentially including their stream of origin. The objectives of our study were to investigate the heterogeneity of streamspecific signatures in the Morice River watershed to determine if geographical locations w ithin the watershed were distinct, to examine fin-ray chemistry to determine if the elemental deposition in the outer edge o f the structure correlated to the water chemistry at the capture location, and to correlate seasonal fluctuations in Zn deposition to annuli deposited in the fin-ray to see if age could be assessed using chemical signatures. METHODS Study Location This project was conducted w ithin the Morice River watershed located in northwestern British Columbia (Figure 4.1). The Morice River watershed represents a drainage basin of 4300 km^. Morice Lake is fed by two large river systems: the Atna and Nanika Rivers. The Morice River flows out of the northeastern portion of Morice 87 Lake. Main tributaries to the Morice River are the Thautil River, Gosnell River, Crystal Creek, Lamprey Creek, Owen Creek, Houston Tommy River, and Cold Creek. Water Collection W ater samples were collected from 15 geographically distinct locations (Figure 4.1) in August 2003 along the mainstem of the Morice River, and from major tributaries and areas where bull trout are known to inhabit. The methods for water collection and analysis are outlined in Chapter 3. Fish Collection To correlate fin-ray chemistry to water chemistry, 10 fish were used where capture location was known (i.e. upper Morice River (n=6), lower Morice River (n -3 ) and the Nanika River (n = 1). Fin-rays were obtained from a radio-telemetry study conducted earlier on the Morice River (Bahr 2002). The leading pectoral finray of each fish was surgically removed prior to release o f the animal. Fin-ray Chemistry Fin-rays were sectioned and sonicated in ultra-pure water for five minutes, embedded in epoxy resin, polished with 1200 yUm silicon carbide paper, and 88 British H o u s to n C o lu m b ia Creek M orice R iver 10 k m Figure 4.1. Map of the Morice River watershed showing water sampling locations. Water was collected from Gold Creek, Houston Tommy Creek, Owen Creek, Lamprey Creek, Thautil River, Denys Creek, Cosnell Creek, Crystal Creek, Redslide Creek, Nanika River, Morice Lake, and three locations in Morice River. 89 sequentially polished with 6 jjvr\, 1 jjm , and 0.05 jum diamond suspension polish to obtain an adequate surface for ablation with the laser. LA-ICP-MS analysis, collection and reduction, and statistical analyses were all completed using the same protocol outlined in Chapter 2 and 3. RESULTS Water chemistry values (mmol/mol) for the 15 sampling sites located throughout the Morice River watershed are shown in Table 4.1. The water chemistries for Ba:Ca, Sr;Ca, and Mg:Ca were similar throughout the watershed. Greater differences in Mn:Ca were found throughout the watershed, particularly in tributaries when compared to the mainstem river. The tw o tributaries that showed distinct differences from the rest of the watershed were Lamprey Creek and Owen Creek. Assigning a unique chemical signature to a particular stream, however, was difficult, as there was considerable overlap in signatures among many o f the streams (Table 4.1). Canonical discriminant function analysis was used to assess separation of the various sampling locations using the factor loadings o f Sr:Ca, Ba:Ca, and Mn;Ca (Figure 4.2). These data clearly show that geographical separation using the elemental ratios measured in this study for the Morice River watershed are not adequate to distinguish movement patterns within Morice Lake and both the Morice and Nanika Rivers, where adult bull trout reside for the majority o f their life-history in this watershed. 90 Table 4.1. Water chemical ratios measured In the Morice River watershed. All measured values are expressed as mmol/mol. M ap# Location Ba:Ca Mg:Ca Mn:Ca Sr:Ca 1 Lower Morice 235 87.1 0.181 458.4 2 Middle Morice 2.09 87.0 0.395 439.4 3 Upper Morice 223 75.1 0.098 445.4 4 Morice Lake 224 74.6 0.134 464.0 5 Gold Creek 2.41 126.0 0.101 2923 6 Houston Tommy 269 91.0 0.033 2920 Creek 7 Denys Creek 1.23 106.0 0.008 3223 8 Crystal Creek 226 70.1 0.015 2783 9 Gosnell River 1.31 108.0 0.622 3820 10 Thautil River 233 146.0 0.074 400.0 11 Owen Creek 1.65 234.0 1.001 900.0 12 Lamprey Creek 2.64 219.5 1.637 909.0 13 Lower Nanika 1.93 83.4 0.361 490.0 14 Upper Nanika L83 825 0.506 501.0 15 Redslide Creek 2.51 49.5 0.427 366.6 91 30 n 20 • # e o ♦ O 10 CM C o C ro c A -10 A A V -20 # # □ ■ 0 -30 V Lower Morice Middle Morice Upper Morice Morice Lake H I Creek Gold Creek Dennys Creek Crystal Creek Gosnell Creek Thautil River Owen Creek Lamprey Creek Lower Nanika Upper Nanika Redslide Creek -40 -50 -100 -50 "T " T 0 50 100 150 200 Discrimant Function 1 Figure 4.2. Canonical discriminant function analysis according to the chemical signatures of Ba:Ca, Sr:Ca, MniCa (discriminant factors)showing the variation measured in the Morice River watershed. Discriminant function analysis shows that the Morice River, Morice Lake, and the Nanika River group too close together to discriminate movements of bull trout between these systems. Known spawning tributaries separate into distinct locations (Gold Creek, Crystal Creek, Gosnell River, Thautil River, Redslide Creek). 92 Incorporation coefficients for each element ratio were calculated as well as the sample standard deviations (SD) (Microsoft Excel 2002, Microsoft Corporation) Ba:Ca = 0.020 (SD = 0.003), Mg:Ca = 0.36 (SD = 0.060), Mn:Ca = 0.19 (SD = 0.037), and SnCa = 0.15 (SD = 0.014) (Table 4.2). Figures 4.3 and 4.4 show representative line scans for two bull trout. In Figure 4.4, there is little change in the line scans for any of the elemental ratios for bull trout 10, except near the core for both Ba:Ca and Mn:Ca ratios. In contrast, the line scans for bull trout # 7 in Figure 4.3 show much more variation over the entire region of the pectoral fin-ray. One fish depicts migratory behaviour, while the other fish appears to be non-migratory, as determined by chemical ratios measured across all life-history stages. Due to the similarity in chemical make-up of both the Morice River and the Nanika River, it is difficult to determine if the non-migratory fish show movement. The fish that show high variability in chemical signatures over their life-history appear to be moving into chemically distinct environments. On the other hand, fish that show relatively constant line scans may be either non-migratory or just moving among river systems where elemental ratios are similar. Figures 4.5 and 4.6 also demonstrate with Canonical Discriminant Function analysis that the Morice River watershed is not suited for an examination o f movement patterns using the factors (water chemistry) Ba:Ca, Sr:Ca, and Mn:Ca. Representative bull trout showing both variation in chemical signatures and a relatively consistent line scan across the fin-ray, bull trout # 7 and # 1 0 respectively, are illustrated (Figures 4.3 and 4.4). Nine different areas were utilized for bull trout # 7 across the line scan, where variability in 93 Table 4.2. Relationship between fin-ray chemistry measured at the outer edge and water chemistry for 10 bull trout captured in the Morice watershed. Incorporation coefficients were calculated for each fish and the sample standard deviation (SD) was calculated. Fish # 1 2 3 4 5 6 7 8 9 10 Ba/Ca (m m ol/mol) Water 2.2 2.2 2.2 2.2 2.2 2.2 2.4 2.4 2.4 1.8 Fish 0.043 0.047 0.045 0.050 0.041 0.042 0.040 0.060 0.040 0.030 1C 0.019 0.021 0.020 0.022 0.018 0.019 0.017 0.025 0.017 0.016 Mg/Ca (mmol/mol) Water 74.9 74.9 74.9 74.9 74.9 74.9 87.1 87.1 87.1 87.6 Fish 26 33 28 32 33 30 25 26 26 30 10 0.347 0.440 0.374 0.427 0.440 0.400 0.287 0.299 0.299 0.342 Mn/Ca (mmol/mol) Water 0.10 0.10 0.10 0.10 0.10 0.10 0.18 0.18 0.18 0.51 Fish 0.022 0.020 0.019 0.021 0.024 0.018 0.030 0.020 0.036 0.083 10 0.224 0.204 0.193 0.214 0.244 0.183 0.165 0.110 0.199 0.164 Sr/Ca (m m ol/m ol) Water 4.5 4.5 4.5 4.5 4.5 4.5 4.6 4.6 4.6 5.0 Fish 0.750 0.730 0.600 0.770 0.680 0.650 0.700 0.700 0.600 0.700 10 0.168 0.164 0.135 0.173 0.153 0.146 0.153 0.153 0.131 0.140 94 0.18 0.16 0.16 0.14 0.14 ^ 0.12 J0.10 o £ £ 0.12 0.10 0.08 Core U 0.06 0.06 0.04 0.04 0.02 0.02 0.00 0.00 CO Core 0.08 38 -) Core Core 36 34 - 0 E 0.9 - “Ô 1 0.8 - U 0.7 - #2422 0.6 - - 20 0.5 0 60 100 Scan Time (s) 150 200 250 0 50 100 150 200 250 Scan Time (s) Figure 4.3. Representative line scans for a bull trout (bull trout #7) that had chemical signatures which were indicative of movements throughout its life-history. 95 0.16 -) 0.18 -| 0.14 - 0.16 0.14 - 0.12 - 0.12 0.10 - Core 0 0.10 E E 0.08 Ô 0.06 1 0.04 - Core - 0.08 0.06 CO 0.04 0.02 - 0.02 - 0.00 0.00 38 -| 36 - o A 0.8 - o E E 0 .7 - Core Core U 0.5 20 0.4 0 20 40 60 80 100 120 Scan Time (s) 140 160 180 200 0 20 40 60 80 100 120 140 160 180 200 Scan Time (s) Figure 4.4. Representative line scans for a bull trout (bull trout #10) that had chemical signatures which showed little movement throughout its life-history. 96 40 -1 • 30 - o □ e A 20 0 o CM C g CJ c 3 10 - V O 0 0 - I -10 V E ■ V N o ü g • # V -20 H <$> A 0 -30 Lower Morice Gold Creek HT Creek Dennys Creek Middle Morice Upper Morice Crystal Creek Gosnell River Morice Lake Lower Nanika Redslide Upper Nanika Thautil River Lamprey Creek Owen Creek □ -40 H o -50 -100 -50 0 T T 50 100 150 200 Discriminant Function 1 Figure 4.5. Canonical discriminant function analysis for bull trout # 1 0 . The letters indicate locations throughout the life of each individual with (a) representing the capture location and (i) representing the approximate first year of growth. 97 30 -1 <$> 20 $0 10 - 0 - A • # CM C O V 3 ro o g c C -10 • o □ - 0 ■ V V O c 0 o -20 A 0 -30 H Lower Morice Gold Creek H I Creek Dennys Creek Middle Morice Upper Morice Crystal Creek Gosnell River Morice Lake Lower Nanika Redslide Upper Nanika Thautil River Lamprey Creek Owen Creek V -40 O -50 -100 -50 T " T T 0 50 100 150 200 Discrimant Function 1 Figure 4.6. Canonical discriminant function analysis for bull trout # 7. The letters indicate locations throughout the life of each individual with (a) representing the capture location and (i) representing the approximate first year of growth. One potential migration (g) to the Gosnell Creek watershed is noted. 98 the chemical signature was noted. The same nine areas were analyzed for bull trout # 1 0 to maintain consistency in the analysis. Letters (a-i ) were assigned where (a) represents the capture location and (i) represents what was estimated to be the first year o f growth. Each letter corresponds to a unique chemical signature (Ba:Ca, Sr:Ca, Mn:Ca) that were input into the Canonical Discriminant function. Bull trout # 7 showed very little variation according to the discriminant functions. Most of the variation in the line scan for bull trout # 7 was attributed to Mn:Ca ratios. Where changes are noted, there was no corresponding river system suggesting that the fish might have moved into an area where water chemistries were unknown. Interestingly, elemental ratios suggest that bull trout # 1 0 remained in the M orice and Nanika Rivers, with one migration to the Gosnell Creek watershed. Measured values of Zn showed an interesting pattern o f deposition across the finray. Oscillations were apparent and seemed to correlate to annulus formation in the finray. Bull trout # 1 0 demonstrated six full oscillations, while bull trout # 7 demonstrated seven full oscillations (Figure 4.7). Table 4.3 provides the length, weight, and age of the bull trout utilized for the analysis. Conventional ageing was accomplished for the 10 bull trout in this study by North/South consultants who counted the opaque and translucent zones formed in the fin-ray. The zones typically represent winter and summer growth, respectively. Ages were also assessed by counting the num ber o f full oscillations shown by Zn:Ca ratios. Zn:Ca oscillations were counted by two readers who did not know the ages of the fish prior to examination o f the data. 99 0.40 0.35 0.30 H 0.25 0.20 Core 0.15 0 1 0.10 E E 0.05 ü 0.30 N 0.28 o 0.26 0.24 0.22 Core 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0 50 100 150 200 250 Scan Time Figure 4.7. Zn profile for bull trout #2. (top) and bull trout # 7 (bottom) There are six clear oscillations that correspond to the six annuli present in the fin-ray for bull trout # 2 and seven clear oscillations that correspond to the seven annuli present in the fin-ray in bull trout #7. 100 Table 4.3. Length and age for the 10 bull trout sampled in this investigation. Age assessments (annuli) were performed independently by North/South consultants. Seasonal fluctuation in Zn were determined by the # of peaks of Zn:Ca measured in the fin-ray. Values separated by comma were conducted by two different readers. Fish # Length (mm) # of Annuli (age) Seasonal Zn fluctuations (age) 1 540 6 6, 5.5 2 420 6 6.5, 6.5 3 515 9 8 ,9 4 545 7 7, 7 5 510 8 unreadable, 8 6 465 7 6.5, 6 7 540 7 7 ,7 8 560 8 7 ,8 9 455 6 5.5, 6 10 520 6 6 ,6 101 DISCUSSION Our research is im portant to the growing knowledge on using natural chemical tags in fish as determinants o f life-history. Elemental ratios fluctuate over the life of the fish, as we observed differences among the 10 fish examined in this study. Changes in Ba:Ca and Mn:Ca ratios were evident, but little change was observed for the Sr:Ca and Mg:Ca ratios. It would appear the elemental signatures do not differ sufficiently in all elements to detect movements w ithin the mainstem of the Morice River, however, we did determine that fin-ray chemistries measured at the outer edge were proportional to water chemistries measured at the same location. Additionally it was shown that Zn incorporation into the fin-ray appears to follow a seasonal oscillation with more incorporation occurring in the summer. Geographic Separation The mainstem Morice River showed a remarkably consistent chemical signature throughout its length. Tributaries, however, were sufficiently different to discriminate water sample locations with Canonical Discriminant Function analysis. Two other tributaries, Owen and Lamprey Creeks, had unique water chemistries; however. Bustard and Schell (2002) note that only Dolly Varden are present in these creeks. Owen and Lamprey Creeks are both lake fed tributaries characterized by seasonally warmer temperatures likely unsuitable for bull trout. Movement patterns of bull trout were 102 difficult to discriminate as it appears most movement is confined to the chemically similar systems including: Morice Lake and both the Morice and Nanika Rivers. Bahr (2002) observed both small movements and large scale movements o f bull trout w ithin these same systems with radio telemetry data. O ur study could not resolve the differences between these systems; thus, we could only come to the conclusion that most movements were restricted to the larger systems in this watershed causing discrimination between individuals to be difficult. W e were also unable to classify movements into spawning tributaries, even though water chemistries were distinctive w ithin these systems. Many bull trout in the Morice watershed spend less than a week in these spawning tributaries (Bahr 2002), possibly limiting the amount of material incorporated into the fin-ray during these migrations. The amount of time spent spawning for the 10 fish examined was probably insufficient for us to spatially resolve with probe-based analysis. An examination using otoliths may have revealed these migrations, as otoliths provide much more spatial resolution due the larger size of the structure. O tolith growth is also continual and, in bull trout, provides a greater surface area for ablation than the fin-ray. Another possible explanation is that the bull trout investigated had not yet made their first spawning migration. The latter scenario is unlikely, as bull trout of the same size class have been observed spawning within the Morice River tributaries (Bahr 2002). It appears that in a chemically similar environment (such as the Morice River watershed), that radio-telemetry is a better tool for resolving fish movements. 103 Fin-ray Chemistry To date, there has been very little research conducted using natural chemical markers with fin-rays. For this reason, the process o f incorporation o f elements is poorly understood. Interestingly, the coefficients obtained for Ba:Ca and Sr:Ca were very close to the coefficients determined in slimy sculpin otoliths: 0.019 and 0.21 respectively (Chapter 3). On the other hand, the coefficients determined for Mg:Ca and Mn:Ca were much higher in fin-rays than the coefficients measured for the same elements in otoliths. These data suggest that fin-rays incorporate both Ba:Ca and Sr:Ca in similiar proportion as otoliths; however fin-rays show a much higher affinity for the incorporation of Mg and Mn. Using fin-rays to resolve movements in chemically unique watersheds, therefore, is very promising. Resolution of stream of origin for slimy sculpins in Chapter 3 would have been much higher if both Mg and Mn could have been used to help differentiate populations; however, bull trout movement patterns were very difficult to resolve in the Morice River watershed, due to the similarity of the water in this geographical region. In fact, we were unable to discriminate between fish that were expected to show large movements from fish where small migrations were anticipated. Some of the fish in this study showed variation in measured fin-ray chemical ratios, suggesting movements into chemically distinct habitats. On the other hand, fish that had relatively flat line scans may have exhibited highly migratory behaviour restricted to Morice Lake, and both the Morice and Nanika Rivers. Bahr (2002) supports this with telemetry data where many 104 Morice bull trout made extensive migrations from the lower Morice River to the Upper Nanika River, (a distance greater than 100 km). Our results support the findings of Halden et al. (2000), who determined that seasonal deposition of Zn in Arctic char otoliths correlates to annulus formation. M ilner (1982), as well as, Bradley and Sprague (1985), suggest that metabolic rate influences Zn deposition in fish. Seasonal summer temperatures likely influence the uptake and production of Zn, while colder w inter temperatures would represent times when lower levels o f Zn production and uptake occur (Halden et al. 2000). Bull trout fin-rays also show seasonal fluctuations in the distribution o f Zn. For nine o f the 10 hull trout examined in our study, we were able to interpret the fluctuations of Zn present in finrays as yearly increments because ages corresponded well to independent age estimates provided by North/South Consulting for the same nine fish. Halden et al. (2000) noted that the incorporation of Zn into Arctic char otoliths decreases with age. The results of Halden et al. (2000) are also consistent with other studies examining Zn uptake by fish (Milner 1982; Bradley and Sprague 1985; Campbell and Stokes 1985). O ur results, however, did not show Zn uptake decreasing with age over the lifetime o f the hull trout in our study. The seasonal fluctuations in Zn for hull trout fin-rays determined by our study appear to vary in magnitude along the line scan; however, there is no trend showing an increase or decrease in overall Zn content. 105 CONCLUSION An examination o f fin-ray chemistry determined that these structures incorporate elements in proportion to the recent ambient water chemistry. The chemical ratios measured at the outside edge o f the fin-rays were representative o f the water chemical ratios measured at the capture location. Measuring chemical ratios in bony structures at the outer edge and comparing the values to those of water is not the most effective way of obtaining an incorporation coefficient. Fish in a non-migratory life-history stage should be captured from an area where it has been determined that water chemistry is stable over time periods long enough to obtain a stable signature from the bony structure. Alternately, fish can be raised in an artificial environment where water chemistries are known. For these reasons, our incorporation coefficients may not reflect the exact physiological relationship between the magnitude of elemental concentration in the water to the uptake of trace elements by the fish; however, we feel that the values are representative. The findings o f our study are important, as they show that fin-rays have the potential to replace or com plim ent otoliths in investigations examining fish life histories with chemical markers and can offer a non-lethal alternative, however, water chemical ratios in the Morice River watershed were very similar, so movements of bull trout could not be determined. Results we obtained in Chapter 3 for the W illiston Reservoir determined that the streams and rivers showed distinctively different elemental signatures. Conflicting results for the tw o studies show that not all watersheds are 106 suitable for an examination offish movement patterns or stock identification w ith natural chemical markers. This determination is important, as researchers must be aware o f the limitations that are associated with this technique. The potential to resolve freshwater migrations using natural chemical tags deposited spatially in bony structures is high, provided the streams and rivers in the study area are both chemically stable over tim e and distinct. W e found in Chapter 3 that the W illiston Reservoir meets both of these requirements and is likely a very good candidate for examining movement patterns of migratory fish using natural chemical tags in the future. 107 Chapter 5: Epilogue This thesis examined the life-history of three different species using natural chemical markers deposited spatially in bony structures. We were successful in resolving the movements of the anadromous eulachon and matching the specific geographic location o f slimy sculpins to the water chemistries where they were captured. Unfortunately, we were unable to reveal the movements o f migratory bull trout w ithin the Morice River watershed. Nevertheless, bull tro u t fin-rays appear to proportionately incorporate some trace elements from the ambient water; this result w ill provide researchers with another avenue to explore the developing field of using natural elemental markers as determinants of life-history. Differences in chemical signatures among aquatic media were measured to assess habitat use and movement patterns in an attempt to understand life-history o f these three species. Substantial differences exist in chemical signatures between both freshwater and saltwater environments. The specific element examined in Chapter 2 to infer the life-history o f eulachon moving between the tw o environments was Sr. Movements should have been easy to assess as the difference in the magnitude of Sr concentration is high between fresh and saltwater; nevertheless short residency time in freshwater preand post-spawning limited our ability to detect a freshwater Sr signal for eulachon. Seasonal differences in Ba:Ca uptake were also observed and we felt that the most plausible explanation was variable ocean temperatures. The fluctuations in Ba:Ca uptake were considered a proxy for w inter and summer growth and were subsequently used to 108 assign ages to the individual eulachon. Geochemical differences in freshwater chemical ratios were utilized in Chapter 3 to infer the stream of residency for slimy sculpins. Large differences between streams were observed which allowed good separation o f habitats based on the chemical geology. Thus, we were successful in assigning the capture locations to the stream chemistry using otolith microchemistry. On the other hand, the geochemical differences in the Morice River watershed were not sufficiently distinct to resolve movements using fin-rays. Results from the W illiston and Morice watersheds suggest that natural chemical markers are only suitable for areas where large difference in chemical signatures exists among tributaries and throughout the mainstem river. Where there are distinct differences the use of this technique w ill offer a new tool in order to understand the complex life histories of many fish species. Fisheries managers are always looking to find new ways to assess fish population status and structure, habitat utilization, and movement patterns. Many fish species in northern British Columbia are difficult to track and locate due to many factors including: river turbidity, lake depth, and remote locations. The ability to reveal previously unknown life-history characteristics w ithout physically tracking the animals will undoubtedly help provide information to make informed management decisions. Population (stock) structure has previously been determined for many species offish using either, or, a combination of, genetics, radio telemetry, and conventional tagging techniques. The results of genetic analyses for determining population structure are sometimes confounded by migration as fish often move in and out of a specific area. 109 Even a very small amount o f juvenile or adult mixing between populations w ill lim it the capabilities o f genetic analyses for inferring stock structure (Haiti and Clark 1989). Genetic analyses are better suited to pedigree analysis (family structure) and determining phenotypic evolutionary linkages (Wilson and Ferguson 2002). Genetic analyses are also not suitable for an examination of fish movement patterns and habitat utilization due to the high dispersal rates and mixing of populations common to many fish species (Gampana and Thorrold 2001). Habitat utilization is a key life-history characteristic that fisheries managers need to understand. Radio-telemetry, and to a lesser extent conventional tagging techniques, provide information on both movement patterns and habitat utilization offish large enough to tag; however, these techniques offer little insight into stock structure as only a small portion of the population can be examined. Additionally, physical tags only allow for an understanding o f a small period of the animals life as it is prohibitive to track an animal from the juvenile stage until senescence. In fact, there is likely a bias in study designs where larger fish are utilized (Steingrimsson and Grant 2003). An inherent problem with physically tagging the animal is that the normal physiology and behaviour of the animal may be affected through the application of the external tag. Bridger and Booth (2003) point out that researchers' must be keenly aware of the effects that the application o f a physical tag (radio transmitter) has on fish and how various attachment locations affect fish differently. 110 Elemental analysis o f bony structures provides fisheries managers with a tool that is indicative offish population structure over a very recent time period (the lifetime of individuals w ithin a stock) so periodic migrations are not an issue; genetic analysis examines population structure over a much longer time period. The combination of these two techniques may provide the best option for examining stock structure in fish species, as natural tags may provide information that is more relevant to recent changes that populations encounter; genetic analyses provides information towards parental and evolutionary linkages (Ferguson and Danzmann 1998). Additionally, we believe that chemical markers have the potential to resolve both movement patterns and habitat utilization over the life of individual fish, regardless o f their size or life-history characteristics. In a recent study using a non-invasive tagging and recovery technique (tagged subcutaneously with dye and monitored with snorkeling) 320 Atlantic salmon {Salmo salar) young of the year were monitored successfully for 28-74 days (Steingrimsson and Grant 2003). That study determined that most o f the young of the year Atlantic salmon moved less than 120 m from their original location. One hundred and nineteen tagged young of the year fish were lost due to either emigration or mortality over the m onitoring period. Those authors suggest qualitatively that the fish not recovered, disappeared as a result of mortality and not emigration from the study site. Although Steingrimsson and Grant (2003) did an exceptional job in ensuring that their tagging protocol did not influence the behaviour of the fish, their study outlines one of the main issues surrounding tagging; the location and fate of the missing fish is unknown. Ill Elemental analysis of bony structures has a definite advantage in that the natural lifehistory of the animal is not affected by tagging, the entire life-history can be examined, and fish cannot be lost due to emigration or mortality. Using chemical signatures, we were able to determine a previously undetermined life-history characteristic, semelparity for eulachon, and confirm age at maturity. Additionally, our study was able to use otolith microchemistry to correlate the specific capture location o f slimy sculpins to the water chemistry at each location w ithin the Williston watershed. Moreover, using natural chemical markers to track fish in sufficiently chemically distinct watersheds appears to be a reasonable next step in this rapidly developing field. 112 REFERENCES Akielazek, J.J, Moring, J.R., and Chapman, S.R. 1985. Experimental culture of young rainbow smelt Osmerus mordax. Trans. Am. Fish. Soc. 114: 596-603. Armstrong, A.T. 1979. Report on exploration activities on North Carbon Creek Coal property Supplementary Geological Investigations. Private Utah Mines Ltd. report. 32 pp. Babaluk, J.A., Halden, N.M., Re1st, J.D., Kristofferson, A.M., Campbell, J.L., and Teesdale, W.J. 1997. Evidence for non-anadromous behaviour of arctic charr (Salvelinus alpinus) from Lake Hazen, Ellesmere Island, northwest Territories, Canada, based on scanning microprobe analysis o f otolith strontium distribution. Arctic 50: 224233. Bahr M.A. 2002. Movement patterns, tim ing of migration and genetic population structure of bull trout (Salvelinus confluentus) in the Morice River Watershed, British Columbia, MSc thesis. University of Northern British Columbia. 99 pp. Bailey, M. 2000. Eulachon Research Council, May 2000. Minutes summarizing meetings in New Westminister, Terrace, and Bella Coola, B.C.. Informal jo in t report prepared jointly by BC Eorests and Eisheries and Oceans, Canada. 24 pp. Barraclough, W.E. 1964. Contribution to the marine life-history of the eulachon Thaleichthys pacificus. J. Fish. Res. Bd. Can. 21; 1333-1337. Bath, G.E., Thorrold, S.R., Jones, C.M., Gampana, S.E., McLaren, J.W., and Lam, J.H.W. 2000. Strontium and barium uptake in aragonitic otoliths of marine fish. Ceochemica et Cosmochimica Acta 64: 1705-1714. Bath-Martin, G.E., Thorrold, S.R., and Jones, C.M. 2004. Temperature and salinity effects on strontium incorporation in otoliths o f larval spot (Leiostomus xanthurus). Can. J. Fish. Aquat. Sci. 61: 34-42. Benes, P. 1979. Semicontinuous monitoring of truly dissolved forms o f trace elements in streams using dialysis in situ. Principles and conditions. Water Research 14: 511513. Bilton, H.T. and Robbins, G.E. 1971. Effects of starvation, feeding, and light period on circulus formation on scales o f young sockeye salmon (Oncorhynchus nerka). j. Fish. Res. Board. Can. 28: 1749-1755. 113 Bustard, D., and Schell, C. 2002. Conserving Morice watershed fish populations and their habitat. Stage II Biophysical profile. Technical Report Submitted to Community Futures Development Corporation o f Nadina, Smithers BC. 97pp. Bradley R.W., and Sprague J.B. 1985. The influence of pH,water hardness, and alkalinity on the acute lethalityof zinc to rainbow trout (Salmo gairdneri). Can. J.Fish. Aquat. Sci. 42: 731-736. Bridger, C.J., and Booth, R.K. 2003. The effects of biotelemetry transmitter presence and attachment procedures on fish physiology and behaviour. Reviews in Fisheries Science 1 1 :1 3 -3 4 . Cam pana, S.E., Chouinard, G.A., Hanson, J.M., Freeh et. A., and Brattey, J. 2000. O tolith elemental fingerprints as biological tracers offish stocks. Fisheries Research 46: 343-357. Campana, S.E. 2001. Accuracy, precision and quality control in age determination, including a review of the use and abuse of age validation models, journal o f Fish Biology 59: 197-242. Campana, S.E., and Thorrold, S.R. 2001. Otoliths, increments, and elements: key to a comprehensive understanding o f fish populations? Can.J. Fish. Aquat. Sci. 58: 30-38. Campana,, S.E., Thorrold, S.R., Jones, C.M., Gunther,D., Tubrett, M., Longerich, H., Jackson, S., Halden, N.M., Kalish, J.M. Piccoli, P., de Pontual, H., Troadec, H., Panfili, J., Secor, D.H., Severin, K.P., Sie, S.H., Thresher, R., Teesdale, W.J., and Campbell, J.L. 1997. Comparison of accuracy, precision and sensitivity in elemental assays o ffish otoliths using electron microprobe, proton-induced X-ray emission, and laser ablation inductively coupled mass spectrometry. Can. J. Fish. Aquat. Sci. 54: 2068-2079. Campana, S.E., Eowler, A.J., and Jones, C.M. 1994. Otolith elemental fingerprinting for stock identification of Atlantic cod (Cadus morhua) using laser ablation ICPMS. Can. j. Fish. Aquat. Sci. 51: 1942-1950. Campana, S. E. and J.D. Nielson. 1985. Microstructure offish otoliths. Can. J. Fish. Aquat. Sci. 42:1014-1032. Campbell, P. C. G., and Stokes, P. 1985. Acidification and toxicity of metals to aquactic biota. Can J. Fish. Aquat. Sci. 42: 2034-3049. 114 Cannings, S.G., and Ptolemy, J. 1998. Rare freshwater fish of British Columbia. British Columbia Ministry of Environment, Lands, and Parks, Victoria, BC. 214 pp. Chan, L. H., Drummond, D., Edmond, J. M, and Grant, B. 1977. On the barium data from the Atlantic GEOSECS Expedition, Deep-Sea Res. 24: 613-649. Charnov, E.E., and Schaffer, W .M . 1973. Life-history consequences of natural selection: Cole's result revisted. Am. Nat. 107: 791-793. Cooper, J.E. 1978. Identification of eggs, larvae and juveniles of the rainbow smelt, Osmerus mordax, w ith comparison to larval alewife, Alosa pseudoharengus and gizzard shad, Dorosoma cepedianum. Trans. Am. Fish. Soc. 107: 56-62. Copeman, D.G. 1977. Population differences in rainbow smelt, Osmerus mordax: multivariate analysis o f mensural and meristic data. j. Fish. Res. Bd. Can. 34: 1220-1229. Coûtant, C.C., and Chen, C.H. 1993. Strontium microstructure in scales of freshwater and estuarine striped bass (Morone saxatilis) detected by laser ablation mass spectrometry. Can. J. Fish. Aquat. Sci. 50: 1318-1323. De lacy, A.C., and Batts, B.S. 1963. Possible population heterogeneity in the Columbia River smelt. Circular No. 198. Eisheries Research Institute, College of Fisheries, University of Washington, Seattle. De noyer, E.R., Eredeen, K.J., and Hager, J.W. 1991. Laser solid sampling for inductively coupled plasma mass spectrometry. Anal. Chem. 63: 445-457. DEO 2003. 2002 Pacific Region State o f the Ocean. DEO Science Ocean Status Report, Nanaimo British Columbia 56 pp. Eek, D. and Bohlin, T. 1997. Strontium in scales verifies that sympatric sea-run and stream-resident brown trout can be distinguished by coloration. J. Fish Biol. 51 : 659-661. Elsdon, T.S., and Gillanders, B.M. 2002. Interactive effects of temperature and salinity on otolith chemistry: challenges for determining environmental histories of fish. Can. J. Fish. Aquat. Sci. 59: 1796-1808. 115 Farara, D. 2000. Eulachon Research Council, May 2000. Minutes summarizing meetings in New Westminister, Terrace, and Bella Coo Ia, B.C.. Informal join t report prepared jointly by BC Forests and Fisheries and Oceans, Nanaimo BC, Canada. 24 pp. Farrell, A.P., Hodaly, A.H., and Wang, S. 2000. Metal analysis of scales taken from Arctic grayling. Arch. Environ. Contam. Toxicol. 39: 515-522. Fenwick, J.C. 1989. Calcium exchange across fish gills. Vertebrate Endocrinology: Fundamentals and Biomedical implications (eds. P.K.T. Pang and M.P. Schreibman). Academic Press, San Diego, CA. Volume 3: 319-342. Ferguson, M .M . and Danzmann, R.G. 1998. The role of genetic markers in fisheries and aquaculture: useful tools or stamp collecting? Can. J. Fish. Aquat. Sci. 55: 15531563 Fowler, A.J., Campana, S.E., Jones, C.M., and Thorrold, S.R. 1995. Experimental assessment of the effect of temperature and salinity on elemental composition of otoliths using laser ablation ICPMS. Can. J. Fish. Aquat. Sci. 52: 1431-1441. Erie, R.V. and G.R. Spangler. 1985. Dynamics of rainbow smelt during and after exploitation in South Bay, Lake Huron. Trans. Am. Fish. Soc. 114: 713-724. Eriedland, K.D., Reddin, D.G., Nobumichi, S., Hass, R.E., and Youngson, A.F. 1998. Strontium: Calcium ratios in Atlantic salmon {Salmo saiar) otoliths and observations on growth and maturity. Can. j. Fish. Aquat. Sci. 55: 1158-1168. Gaillardet j., Viers J., and Dupre, B. 2003 Trace Elements in River Waters. . Surface and Ground Water, Weathering, and Soils (ed. J.l. Drever) vol 5 Treatise on Geochemistry (eds. H.D. Holland and K.K. Turekian), Elsevier-Pergamon, Oxford, pp. 225-272 Gowan, G., Young, M.K., Fausch, K.D, and Riley, S.C. 1994. Restricted movement in resident stream salmonids - a paradigm lost? Can. j. Fish. Aquat. Sci. 51: 6262637. Guay, C. K., and Falkner, K.K. 1998. A survey of dissolved barium in the estuaries of major Arctic rivers and adjacent seas. Cont. Shelf. Res. 18: 859-882. Halden, N.M. 2001. Goloured fish ears: Cathodoluminescence as a guide to variation in aqueous environments. Newsletter o f the Mineralogical Association of Canada 64: 1-16. 116 Halden, N.M., Babaluk, J.A., Kristofferson, A.H., Campbell, J.L., Teesdale, W.J., and Reist, J.D. 1996. MIcro-Pixe studies o f strontium zoning in Arctic charr otoliths: migratory behaviour and stock discrimination. Nucl. Instrum. Meth. B. 189: 190195. Marti, D.L., and Clark, A.G. 1989. Principles of Population Genetics. Sinauer Associates, Sunderland, MA. 450 pp. Haldorson, L. and Craig, P. 1984. Life-history and ecology of a Pacific-Arctic population of rainbow smelt in coastal waters o f the Beaufort Sea. Trans. Am. Fish. Soc. 113: 33-38. Hart, J.L. 1974. Pacific Fishes of Canada. Fish. Res. Bd. Canada Bail. 180: 737 pp. Hart, J.L. and McHugh, J.L. 1944. The smelts (Osmeridae) of British Columbia. Res. Bd. Can. Buii. 64: 27 pp. Hay, D.E., Harbo, R., Southey, C.E., Clarke, J.R., Parker, G., and McCarter, P.B. 1998. Catch composition o f British Columbia shrimp trawls and preliminary estimation of bycatch - with emphasis on eulachons. Canadian Stock Assessment Secretariat Research Document. Fisheries and Oceans Canada 1999, Nanaimo, BC. I l l pp. Hay, D.E. and McCarter, P.B. 2000. Status of the eulachon Jhaieicthys pacificus in Canada. Canadian Stock Assessment Secretariat Research Document. Fisheries and Oceans Canada 2000, Nanaimo, BC. 50 pp. Howland, K.L., and Tonn, W .M . 2001. Identification o f freshwater and anadromous inconnu in the Mackenzie River system by analysis o f otoliths strontium. Trans. Amer. Fish. Soc. 130: 725-731. Kalish, J.M. 1990. Use o f otolith microchemistry to distinguish the progeny o f sympatric anadromous and non-anadromous salmonids. Fish. Bull. 88: 657-666. Kalish, J.M. 1991. Determinants of otolith chemistry: seasonal variation in the composition o f blood plasma, endolymph and otoliths o f bearded rock cod Pseudophycis barbatus. Mar. Eco/. Prog. Ser. 74: 137-159. Karnacky, K.J. Jr. 1998. Osmotic and Ionic Regulation. The Physiology of Fishes Second Edition (ed. David H. Evans). CRC Press, Washington DC. 519 pp. 117 Katano, O. and Iguchi, K. 1996. Individual differences in territory and growth of ayu, Plecoglossus altivelis (Osmeridae). Can. J. Zoo/. 74: 2170-2177. Kelson, J. 2000. Eulachon Research Council, May 2000. Minutes summarizing meetings in New Westminister, Terrace, and Bella Coola, B.C.. Informal jo in t report prepared jointly by BC Forests and Fisheries and Oceans, Nanaimo, BC. 24 pp. Kennedy, B.P., Folt, C.L., Blum, J.D., and Nislow, K.H. 2000. Using natural strontium isotopic signatures as fish markers: methodology and application. Can. J. Fish. Aquat. Sci. 57: 2280-2292. Kennedy, B.P., Klaue, A. Blum, j.D ., Folt, C.L., and Nislow, K.H. 2002. Reconstructing the lives offish using Sr isotopes in otoliths. Can. J. Fish. Aquat. Sci. 59: 925-929. Kennedy, B.P., Blum, J.D., and Folt, C.L. 1997. Natural isotope markers in salmon. Nature. 387: 766-767. Li, Y.-H., and Chan, L. H. 1979. Desorption of Ba and ^^®Ra from river borne sediments in the Hudson Estuary. Earth Planet. Sci. Lett. 43: 343-350. Limburg, K.F. 1998. Anomalous migrations of anadromous herrings revealed with natural chemical tracers. Can. j. Fish. Aquat. Sci. 55: 431-437. Marshall, D.J. 1988. Cathodoluminescence of Geological Materials:, Unwin Hyman, Boston, MA. 146 pp. McAllister, D.E. 1963. A revision of the smelt family, Osmeridae. Bull. 191. Biological Series 71. National Museum o f Canada, Ottawa, ON. 53 pp. McDowall, R.M. 1992. Diadromy: origins and definitions of terminology. Copeia 1992: 248-251. McDowall, R.M. 1987. The occurrence and distribution of diadromy among fishes. American Fisheries Society Symposium 1: 1-13. McDowall, R.M. 1972. The taxonomy o f estuarine and brackish-lake Retropinna from New Zealand (Galaxioidei: Retropinnidae). j. R. Soc. N.Z. 2: 501-531. McLean, J.F., Hay, D.F., and Taylor, F.B. 1999. Marine population structure in an anadromous fish: life-history influences patterns of mitochondrial DNA variation in the eulachon, Jhaieicthys pacificus. Molecular Ecology 8: 5143-5158. 118 McPhail, J.D. and Lindsey, C.C. 1970. Freshwater fishes of north-western Canada and Alaska. Fish. Res. Bd. Can. Bull. 173. Milner, N. J. 1982. The accumulation of zinc by O-group plaice, Pleuronectes platessa (L.), from high concentrations in seawater and food. 7. Fish. Biol. 21: 325 -336. M ilton, D.A., Tenakanai, C.D., and Chenery, S.R. 2000. Can the movements of barramundi in the Fly River, Papua New Guinea be traced in their otoliths? Estuarine, Coastal, and Shelf Science 50: 855-868. M orrow, J.E. 1980. The freshwater fishes of Alaska. Alaska Northwest Publishing Company, Anchorage, AK. 248 pp. Moyle, P.B., Herbold, B., Stevens, D.E., and Miller, L.W. 1992. Life-history of delta smelt in the Sacremento-San Joaquin Estuary, California. Trans. Am. Fish. Soc. 121: 67-77. Moyle, P.B. 1976. Inland fishes of California. University o f California Press, Berkely. 405 pp. Moyle, P.B. and Cech Jr, J.J. 2004. Fishes: An introduction to ichthyology. 5*'^ edition. Prentice-Hall Inc. Upper Saddle River, NJ. 612 pp. Page, L.M. and Burr, B.M. 1991. A field guide to freshwater fishes of North America north o f Mexico. Houghton M ifflin Company, Boston. 432 pp. Pannella, G. 1971. Fish otoliths - daily growth layers and periodical patterns. Science 173: 1124. Payne, S.A., Johnson, B.A., and Otto, R.S. 1999. Proximate composition of some northeastern Pacific forage fish species. Fish Oceanogr. 83: 159-177. Peltier, E.F., Webb, S.M., and Gaillard, J.F. 2003. Zinc and lead sequestration in an impacted wetland system. Advances in Environmental Research 8: 103-113. Pootlace, A. and Siwallace, A. 2000. Eulachon Research Council, May 2000. Minutes summarizing meetings in New Westminister, Terrace, and Bella Coola, B.C.. In formal jo in t report prepared jointly by BC Forests and Fisheries and Oceans, Nanaimo, BC, Canada. 24 pp. Ricker, W.E., Manzer, D.F., and Neave, E.A. 1954. The Fraser River eulachon fishery, 1941-1953. Fish. Res. Bd. Canada. No. 583: 35 pp. 119 Rutter, N.W. 1976. M ultiple glaciations in the area of Williston Lake, British Columbia. Geological Survey o f Canada. Bulletin 273. 31 pp. Sanborn, M., and Telmer, K. 2003. The spatial resolution of LA-ICP-MS line scans across heterogeneous materials such as fish otoliths: an experiment on a sandwich of NIST glasses 611, 613, and 615. JAASIB: 1231-1238. Scott, W.B., and Grossman, E.J. 1973. Freshwater fishes o f Canada. Fish. Res. Bd. Canada Bull. 184: 1-966. Shepherd, B.C., and Vroom, P.R. 1977. Biology o f the Mass River eulachon. Technical Report: PAC, /T-, 77-10. Department of fisheries and the environment. Pacific Region. Shiller, A.M. 2003. Syringe filtration methods for examining dissolved and colloidal trace element distributions in remote field locations. Environmental Science & Technology 37: 3953 -3957. Steingrimsson, S.O., and Grant, J.W.A. 2003. Patterns and correlates o f movement and site fidelity in individually tagged young-of-the-year Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 60: 193-202. Swanson, C., Young, P.S., and Cech Jr., J.J. 1998. Swimming performance o f delta smelt: Maximum performance, and behavioral and kinematic limitations on swimming at submaximal velocities. J. Fxp. Biol. 201: 333-345. Taylor, F.B. and Bentzen, P. 1993. Evidence for m ultiple origins and sympatric divergence of trophic ecotypes of smelt (Osmerus) in North-Eastern North America. Evolution 47: 813-832. Taylor B.R. and Hamilton H.R. 1994. Comparison o f methods for determination o f total solutes in flowing waters. Journal o f Flydrology 154: 291 -300. Thorrold, S.R., and Shuttleworth, S. 2000. In situ analysis of trace elements and isotope ratios in fish otoliths using laser ablation sector field inductively coupled mass spectrometry. Can. J. Fish. Aquat. Sci. 57:1232-1242. Townsend, D.W., Radtke, R.L., Malone, D.P., and Wallinga, J.P. 1995. Use o f otolith strontium: calcium ratios for hindcasting larval cod distributions relative to water masses on Georges Bank. Mar. EcoF Prog. Ser. 119:37-44. Townsend, D.W., Radtke, R.L., Morrison, M.A., and Folsom, S.C. 1989. Recruitment implications of larval herring over wintering distributions in the G ulf o f Maine, inferred using a new otolith technique. Mar. Eco/. Prog. Ser. 55: 1-13. 120 Veinott, G., and Evans, R.D. 1999. An examination of elemental stability in the fin-ray of the white sturgeon with laser ablation sampling inductively coupled plasma mass spectrometry (LAS-ICP-MS). Trans. Amer. Fish. Soc. 128: 352-361. Veinott, G., North cote. T., Rosenau, M. and Evans, R.D. 1999. Concentrations of strontium in the pectoral fin-rays of the white sturgeon {Acipenser transmontanus) by laser ablation sampling - inductively coupled plasma- mass spectrometry as an indicator of marine migrations. Can. J. Fish. Aquat. Sci. 56: 1981-1990. Wang, S., Brown, R., and Gray, D.J. 1994. Application of Laser Ablation-ICPMS to the spatially resolved micro-analysis o f biological tissue. Applied Spectroscopy 48: 1321-1325. Ware, D.M., and Thompson, R.E. 1991. Link between long-term variability in upwelling and fish production in the northeast Pacific Ocean. Can j. Fish. Aquat. Sci. 48: 2296-2306. Weber, P.K., Hutcheon, I.D., McKeegan, K.D., and Ingram, B.L. 2002. O tolith sulfur isotope method to reconstruct salmon (Oncorhynchus tshawytscha) life-history. Can. J. Fish. Aquat. Sci. 59: 587-591. Wells, B.K., Rieman, B.E., Clayton, J.L., Horan, D.L., and Jones, C.M. 2003. Relationship between water, otolith, and scale chemistries of westslope cutthroat trout from the Coeur d'Alene River, Idaho: The potential application of hard-part chemistry to describe movements in freshwater. Frans. Amer. Fish. Soc. 132: 409-424. Wells, B.K., Bath, G.E., Thorrold, S.R., and Jones, C.M. 2000(a). Incorporation of strontium, cadmium, and barium in Juvenile spot {Leiostomus xanthurus) scales reflect water chemistry. Can. J. Fish. Aquat. Sci. 57: 2122-2129. Wells, B.K., Thorrold, S.R., and Jones, C.M. 2000(b). Geographic variation in trace element composition of weakfish scales. Frans. Am. Fish. Soc. 129: 889-900. Wilson, A.J. and Eerguson, M .M . 2002. Molecular pedigree analysis in natural populations of fishes: approaches, applications and practical considerations. Can. J. Fish. Aquat. Sci. 59: 1696-1707 Yamada, S.B., and Mulligan, T.J. 1982. Strontium marking of hatchery-reared coho salmon [Oncorhynchus-kisutch Walbaum). }. Fish. Biol. 14: 267-275. 121 Zimmerman, C. and Reeves, G.H. 2000. Population structure of sympatric anadromous and nonadromous Oncorhynchus mykiss: evidence from spawning surveys and otolith microchemistry. Can. j. Fish. Aquat. Sci. 57: 2152-2162. 122