Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The role of prey quality in rhinoceros auklet (Cerorhinca monocerata) productivity Beaubier, Jessica 2006

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

Item Metadata

Download

Media
831-ubc_2006-0371.pdf [ 4.39MB ]
Metadata
JSON: 831-1.0092678.json
JSON-LD: 831-1.0092678-ld.json
RDF/XML (Pretty): 831-1.0092678-rdf.xml
RDF/JSON: 831-1.0092678-rdf.json
Turtle: 831-1.0092678-turtle.txt
N-Triples: 831-1.0092678-rdf-ntriples.txt
Original Record: 831-1.0092678-source.json
Full Text
831-1.0092678-fulltext.txt
Citation
831-1.0092678.ris

Full Text

T H E R O L E O F P R E Y Q U A L I T Y IN R H I N O C E R O S A U K L E T (CERORHINCA MONOCERATA) P R O D U C T I V I T Y by Jessica Beaubier B.Sc. University of Victoria 2001 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF MASTER OF SCIENCE in The Faculty of Graduate Studies (Zoology) THE UNIVERSITY OF BRITISH C O L U M B I A June 2006 © Jessica Beaubier, 2006 Abstract While shifts in the species composition of seabird nestling diets are generally well described, their energetic implications for nestlings are poorly understood due to a lack of information on relative prey quality. I investigated how prey quality related to interannual differences in rhinoceros auklet reproductive success at Triangle Island, British Columbia. I first estimated the energy density and proximate composition of the 4 main prey types (Pacific sand lance (Ammodytes hexapterus), Pacific saury (Cololabis saira), juvenile rockfish species (Sebastes spp.), and juvenile salmon (Oncorhynchus spp.) fed to rhinoceros auklet nestlings at Triangle Island. I then developed models of prey energy content and used them to estimate average energy content and quality of rhinoceros auklet nestling meals in two years of contrasting diets. I then tested to see i f diet differences correlated with differences in chick growth and fledging characteristics. Prey types varied in energy density, both within and between species, which discounts using relative meal biomass as a proxy for relative meal energy content. Both mass and energy of nestling meals were lower in 2003 than in 2004 and varied non-linearly over the course of each breeding season. Average diet quality (kJ/g) did not vary between years and interannual differences in energy were driven by differences in average meal mass. Differences in prey availability may have led to large differences in productivity between years, but primarily by affecting hatching success rather than chick growth. O f nestling variables measured, only tarsus length varied between years: 2003 tarsi at fledge were shorter than those of 2004. Prey availability was likely more important than prey quality in driving annual production during these two years of study at Triangle Island. ii Table of Contents Abstract ii Table of Contents iii List of Tables iv List of Figures . v Acknowledgments vi Co-Authorship Statement .....vii Chapter One: General Introduction . 1 Study Organism '. 1 The Problem 2 Objectives 5 References 6 Chapter Two: Energy Density and Proximate Composition of Rhinoceros Auklet Fish Prey 8 Methods : 11 Results 18 Discussion 26 References 32 Chapter Three: Correlations of Prey Quality and Reproductive Success in Rhinoceros Auklets 37 Introduction 37 Methods 41 Results 46 Discussion 53 References 59 Chapter 4: Conclusions 64 Appendix A. Lipid extraction method testing 66 Appendix B: Dehydration Correction.... 69 Appendix C. Sizes and masses of forage fish species sampled from Rhinoceros Auklet diets at Triangle Island, British Columbia 71 Appendix D. Animal Care Permits < 72 iii List of Tables Chapter Two Table 2.1. Sampling dates and number of bi l l loads sampled from rhinoceros auklets provisioning chicks at Triangle Island, B C 12 Table 2.2. Water content (percent wet mass), proximate composition (percent dry mass), and energy density (percent dry mass) of forage fish fed to rhinoceros auklet nestlings at Triangle Island, British Columbia 22 Table 2.3. Factors accounting for intra-specific variation in energy density in species of forage fish preyed upon by rhinoceros auklets breeding at Triangle Island, B C ...23 Table 2.4. Prediction of total energy (kJ) in forage fish delivered to rhinoceros auklet nestlings on Triangle Island, British Columbia 25 Table 2.5. Prediction of total energy content (in kJ) by dry mass (in g) of forage fish species taken by rhinoceros auklets breeding at Triangle Island, British Columbia 25 Chapter Three Table 3.1. Sampling dates and number of b i l l loads sampled from rhinoceros auklets provisioning chicks at Triangle Island, B C : 41 Table 3.2. Calculation of total energy (in kJ) in forage fish delivered to rhinoceros auklet nestlings on Triangle Island, British Columbia 43 Table 3.3. Sizes and masses o f forage fish species sampled from Rhinoceros Auklet nestling diets at Triangle Island, British Columbia 46 Table 3.4. Effects of year and chick age on energy delivered to rhinoceros auklet nestlings at Triangle Island, British Columbia 49 Table 3.5. Growth and survival statistics for all rhinoceros auklet chicks measured at Triangle Island, British Columbia 51 Table 3.6. Growth and survival statistics for rhinoceros auklet chicks from eggs laid on or prior to the annual median lay date of rhinoceros auklets at Triangle Island, British Columbia 52 iv List of Figures Chapter Two Figure 2.1. Wet mass diet composition of rhinoceros auklet chicks at Triangle Island, British Columbia across all sampling sessions ; 18 Figure 2.2. Dry mass energy densities and relative contributions of protein and lipid constituents of forage fish obtained from rhinoceros auklets provisioning chicks at Triangle Island, British Columbia 20 Figure 2.3. Energy density (kJ/g dry mass) variation in juvenile pacific sand lance (Ammodytes hexapterus) by time of year in 2004 23 Chapter Three Figure 3.1. Composition of rhinoceros auklet nestling diet by date (days after mean hatch date) in (a) 2003 (b) 2004 48 Figure 3.2. Energy content (a) and mass (b) of rhinoceros auklet chick meals as a function of average chick age in days 50 Appendix A Figure A . l . Comparison of results from two lipid extraction methods estimating percent l ipid by wet mass for forage fish (pacific sand lance, pacific saury, and chum salmon) 68 Appendix B Figure B . l . Relationship between fresh mass and frozen mass for species of forage fish <25g in mass 70 v Acknowledgments I would like to extend sincere thanks to my supervisors, Jamie Smith and Diane Srivastava, and my committee members Mark Hipfner and Peter Arcese. I particularly want to thank Jamie for his inspiring enthusiasm for the natural world, and for the patience and encouragement he showed me during our short time working together. I would also like to thank Diane for generously inviting me to jo in her lab, and for her valuable advice on the statistical aspects of my thesis. I would very much like to thank Mark for his constant positive support, encouragement, and insightful comments. This project would not have happened without him. Special thanks to Judy Myers for looking out for me throughout my thesis and stepping in to help when I was stuck. I greatly appreciate the technical expertise of Glenn Crossin, Shayne McLel lan , Trevor Haynes and Marc Trudel, and the helpful lab assistance of L i sa Cassidy. I am also indebted to Jon Shurin, Bart Van Der Kamp, and Scott Hinch for generously lending me lab space and equipment. I would also like to extend many thanks to the Triangle field crews of 2003 and 2004, particularly Jen Greenwood, Moi ra Lemon, Laurie Savard, and Sam Franks. This project would not have been possible without funding, and for this I am grateful to N S E R C , the Canadian Wildlife Service, and Werner and Hildegaard Hesse. I am greatly indebted to four people for being incredibly supportive over the past years. To my parents, thank you so much for your unending support and encouragement and for the formative canoe trips that inspired my love of the natural world. Most sincere thanks to Jackie Shaben and Andrew Cameron for your incomparable friendship: you are my very best grad school rewards. vi Co-Authorship Statement This thesis is primarily my own work, but has benefited from the ideas, advice, and editing of both Dr. Mark Hipfner and Dr. Diane Srivastava. I identified the study questions in the third chapter, and designed and conducted all sampling protocols, field/lab work, and statistical analyses. The written work in this thesis is my own though I incorporated revisions and approaches suggested by Drs. Hipfner and Srivastava. v i i Chapter One: General Introduction Long term data sets from a variety of seabird monitoring programs have recently provided convincing evidence of strong positive links between ocean primary productivity and the annual productivity of breeding seabirds (e.g. Gjerdrum et al. 2003, Hedd et al. 2006). However, the mechanisms underlying these productivity relationships are poorly understood. This poses a problem when attempting to gain a holistic picture of the trophic interactions driving seabird population dynamics, and when attempting to discern actual mechanisms resulting in seabird reproductive failure. Proper management of seabird populations requires a solid understanding of the trophic processes involved, particularly in attempting to assess and mitigate the impacts of human activities including fisheries, climate change, and offshore oil and gas development. S T U D Y O R G A N I S M Rhinoceros auklets are piscivorous alcids that forage primarily in the epipelagic zone and target small pelagic fish like Pacific sand lance {Ammodytes hexapterus), Pacific saury (Cololabis saira), juvenile rockfish (Sebastes spp.), juvenile salmonids (Oncorhychus spp.), and Pacific herring (Clupea harengus) (Bertram and Kaiser 1993, Burger et al. 1993, Davoren and Burger 1999). The. species range extends northwards along the Pacific Coast from southern California, through British Columbia and Alaska, and into Russia and Japan (Gaston and Dechesne 1996). 1 Like many alcids, rhinoceros auklets lay a single egg clutch (Gaston and Deschesne 1996) and are constrained to foraging within approximately 80km of their breeding colony throughout incubation and chick rearing (Triangle Island Research Station, unpubl. data). Eggs are laid in burrows in early May , and chicks are reared entirely in the burrow until fledging at approximately 80% of adult mass. Rhinoceros auklets have some of the longest incubation and chick rearing times (average of 45 and 50 days, respectively, Gaston and Deschesne 1996) among the alcids and adults incur significant commuting and provisioning costs throughout the entire period. Adult breeding pairs provision nestlings with multiple whole prey items brought in a single b i l l -load, usually delivered after nightfall (Bertam et al. 1991, Davoren and Burger 1999). At Triangle, chick provisioning occurs mid June through late August and peaks in mid July (Bertram et al. 1991). Rhinoceros auklets have adapted to their fluctuating marine environment through high annual adult survival (0.8296 ± 0.095, Bertram et al. 2000) and flexible breeding effort (Bertram et al. 1996). Annual reproduction is therefore highly variable and reproductive parameters are particularly sensitive to oceanic conditions and resultant food availability (Hedd et al. 2006). T H E P R O B L E M Over the past ten years, rhinoceros auklets breeding at Triangle Island, British Columbia, have shown high interannual variation in annual productivity. The timing and magnitude of springtime primary productivity explain much of the observed reproductive 2 variation (Bertram et al. 2001, Hedd et al. 2006) but the intermediate mechanisms causing the relationship have not been identified. However, trends in rhinoceros auklet nestling diet composition suggest that prey availability and quality likely each play a large role. Rhinoceros auklets breeding at Triangle Island provision their chicks with 5 main prey types: pacific sand lance, pacific saury, juvenile rockfish, juvenile salmon, and pacific herring (Vermeer 1979, Hedd et al. 2006). The relative proportion of each species in nestling diets varies by year and time of year and both the timing and nature of diet shifts have been correlated with rhinoceros auklet reproductive success (Vermeer 1980, Bertram et al. 2001). Reproductive success is low in years when pacific saury . appear early on in nestling diets, and high in years when nestlings are raised primarily on sand lance (Vermeer 1980, Bertram and Kaiser 1993, Hedd et al. 2006). Although the mechanisms underlying this dynamic are not well described for Triangle, prey availability and the energetic quality of different prey species used both influence reproductive success in other seabird populations (Ainley and Boekelheide 1990, Litzow et al. 2002). For example, common murres at Southeast Farallon island have higher reproduction in years when juvenile rockfish are readily available (Ainley and Boekelheide 1990). Pigeon guillemots in Alaska have higher productivity when sand lance are abundant (Litzow et al. 2002). Both the abundance and l ip id content (quality) of sand eels influence the reproductive success of guillemots in the North Sea (Wanless et al. 2005). The environment around Triangle makes measuring prey availability very 3 challenging. However, information on the relative quality of different diet items may offer insight into the respective roles of prey availability and quality at Triangle Island. r 4 O B J E C T I V E S For my thesis I explored whether differences in prey quality were related to interannual differences in rhinoceros auklet reproductive success at Triangle Island, British Columbia. M y objectives were: 1. To determine the relative quality of four main prey species used by rhinoceros auklets breeding at Triangle Island (Chapter Two) 2. To determine i f diet composition affects energy delivered to rhinoceros auklet nestlings, whether differences in energy were due to prey quality, and whether or not energy differences correlated with measures of rhinoceros auklet reproductive success (Chapter Three) 5 REFERENCES Ainley, D . G . , and R.J . Boekelheide. 1990. Seabirds of the Farallon Islands: Ecology, Dynamics, and Structure of an Upwelling-System Community. Stanford, C A : Stanford University Press. Bertram, D.F . , G . W . Kaiser, and R . C . Ydenberg. 1991. Patterns in provisioning and growth of nestling rhinoceros auklets. The Auk 108: 842-852 Bertram, D . R., C . V . J . Welham,and R.C.Ydenberg. 1996. Flexible effort in breeding seabirds: adjustment of provisioning according to nestling age and mass. Canadian Journal of Zoology 74: 1876-1881 Bertram, D.F . , L L . Jones, E . Cooch, H . Knechtel, andF. Cooke. 2000. Survival rates of Cassin's and Rhinoceros Auklets at Triangle Island, British Columbia. Condor 102: 155-162 Bertram, D.F . and G .W. Kaiser. 1993. Rhinoceros auklet (Cerorhinca monocerata) nestling diet may gauge pacific sand lance (Ammodytes hexapterus) recruitment. Canadian Journal of Fisheries and Aquatic Sciences 50: 1908-1915 Burger, A . E . , R .P . Wilson, D . Gamier, and M . P . T . Wilson. 1993. Div ing depths, diet and underwater foraging of rhinoceros auklets in British Columbia. Canadian Journal of Zoology 71: 2528-2540 Davoren, G . and A . E . Burger. 1999. Differences in prey selection and behaviour during self-feeding and chick provisioning in rhinoceros auklets. Animal Behaviour 58: 853-863 Gaston, A . J., and S. B . C. Dechesne. 1996. Rhinoceros Auklet {Cerorhinca monocerata). In The Birds of North America, No. 212 (A. Poole and F. G i l l , Eds.). The Academy of Natural Sciences, Philadelphia, P A , sand The American Ornithologists' Union, Washington, D . C . Gjerdrum, C , A . M . J . Vallee, C . C . St.Clair, D .F . Bertram, J .L. Ryder, and G.S. Blackburn. 2006. Tufted puffin reproduction reveals ocean climate variability. Proceedings of the National Academy of Science 100: 9377-9382 Hedd,A., D .F . Bertram, J .L. Ryder and L L . Jones. 2006. Effects of inter-decadal climate variability on marine trophic interactions: Rhinoceros Auklets and their fish prey. Marine Ecology Progress Series 309: 263-278 Litzow, M . A . , J.F. Piatt, A . K . Prichard, and D . D . Roby. 2002. Response of pigeon" guillemots to variable abundance of high lipid and low lipid prey. Oecologia 132: 286-295 6 Vermeer, K . L . and L . Cullen. 1979. Growth of rhinoceros auklets and tufted puffins, Triangle Island, British Columbia. Ardea 67: 22-27 Vermeer, K . , 1980. The importance of timing and type of prey to reproductive success of Rhinoceros Auklets Cerorhinca monocerata. Ibis 122: 343- 350 7 C H A P T E R T W O : Energy Density and Proximate Composition of Rhinoceros Auklet Fish Prey INTRODUCTION Although seabird nestling diets are well described in terms of species consumed (Ainley and Boekelheide 1990, Hedd et al. 2006), much less is known about the quality of individual prey species used, and the energetic implications of shifting prey availability. Prey quality information can be expensive and time consuming to obtain and many studies instead use biomass as a means of comparing the relative energy content of nestling meals. These studies treat all prey species as energetic equivalents (e.g. Harfenist 1995, Bertram et al. 1996). However, fish species of equal biomass can differ over four-fold in total amount of energy (Anthony et al. 2000), due to variability in lipid and protein content. Predators exploiting lower quality prey must subdue and consume many more individuals to satiate energetic requirements, with concomitant increases in energy expenditure (Trites and Donnelly 2003). The energetic value of a particular prey species may also vary over time, which can influence both its value to predators and the accuracy of estimating energy content from biomass. Energy densities can vary within species as a factor of size, sex, reproductive status, and nutritional status, all of which can create highly seasonal patterns in energy density, particularly for high lipid fish species (Foy and Paul 1999, Robards et al. 1999a). During peak seabird demand, young of year fish are generally allocating resources to growth rather than to energy reserves, which decreases l ipid content and energy density (Love 1970). A version of this chapter will be submitted for publication. Beaubier, J.E., J .M. Hipfner, and D.S. Srivastava. Proximate composition of rhinoceros auklet prey species. 8 In contrast, adult fish are replenishing reserves for reproduction, with proportionally less energy devoted to growth (Calow 1985). Although trends vary by species, adult individuals of high l ipid fish like sand lance and saury, can undergo 5-fold changes in energy density within a year, (Hislop et al. 1991, Kurita 2003), depending on phenological state. Further variation arises from local and broad scale oceanographic change, which influences food availability to fish (Beamish et al. 2004, Robards et al. 1999a), and the base amount of energy available for fish to allocate to fat reserves, reproduction and growth (Love 1970). This variation in food availability can change energy densities by an order of magnitude in some species and have significant implications for foraging seabirds (Diamond and Devlin 2003, Wanless et al. 2005). Seabirds operate on very slim energy budgets, particularly during the breeding season, and efficient foraging is critical (Whittow 2002). Prey quality can increase both chick provisioning rates and reproductive success (Golet et al. 2000, Diamond and Devlin 2003). Many seabirds are dependent on one or two species of high quality forage fish (e.g. capelin, sand lance, eulachon) during chick rearing and can suffer reduced reproductive success i f unable to switch to alternative high quality prey (Gjerdrum et al. 2003) or forced to exploit lower quality prey items (Wanless et al. 2005). Rhinoceros auklets breeding at Triangle Island, British Columbia, provision their chicks with a diet of 5 staple species which together comprise approximately 95% of chick diets in most years. Pacific sand lance (Ammodytes hexapterus) and pacific saury (Cololabis saira) have dominated chick diets in recent years and juvenile rockfish 9 {Sebastes spp.) and juvenile salmon (Oncorhynchus spp.) have made up much of the remainder. While pacific herring (Clupea harengus) has been important in some other years, recently it has been more incidental. Other incidental species include sablefish (Anoplopoma fimbria) and market squid (Loligo loligo) among others. The relative proportion of these species in chick diets varies by year and time of year and reproductive success is positively correlated with percent sand lance in chick diets, and negatively correlated with percent pacific saury (Vermeer and Westrheim 1984, Bertram et al. 1991, Hedd et al. 2006). To date there has only been one cursory study on local prey energy densities (Vermeer and Devito 1986) and no studies have evaluated what these diet changes imply energetically, or i f biomass may be an adequate proxy for energy content. To address this lack of knowledge about prey quality, I have approached this study with three main objectives. M y first objective was to estimate the respective energy densities of the four main prey species taken by rhinoceros auklets at Triangle Island. If similar in energy density, chick meal biomass is l ikely an adequate proxy for energy content. M y second objective was to identify factors that may compromise the use of biomass as a proxy for energy content and incorporate these covariates into develop models of total prey energy content which can be used to estimate total energy delivered to chicks. M y third goal was to consider the potential ramifications of differences in prey energy content on breeding rhinoceros auklets. 10 METHODS Study Site Triangle Island, British Columbia (50° 52' N , 129° 05' W ) is located at the fork of the downwelling Alaskan current system and the upwelling California current, approximately 45 km east of the continental shelf. Primary productivity within the system fluctuates dramatically within and between years (Beamish et al. 2004) and on decadal scales (Mantua and Hare 2002) and influence the food available to higher trophic levels (Bertram et al. 2001). The timing and magnitude of the springtime pulse of primary production seems particularly important for the annual production of seabird species breeding at Triangle Island (Bertram et al. 2001). Sample Collection and Processing I collected fish from rhinoceros auklet adults provisioning their chicks, over the course of two breeding seasons (2003, 2004). Rhinoceros auklets are epipelagically foraging alcids that breed in burrows and provision their chicks with whole prey brought in to the colony at night (Gaston and Dechesne 1996). On average, each parent makes a single provisioning trip to the colony each night, and carries multiple fish (ranging from 1 to over 30) in a single delivery (Bertram et al. 1991). A l l contents carried in a single delivery are termed a "b i l l load." 11 Sampling sessions (10 bi l l loads per session) occurred approximately every 10 days from 20 June onwards (Table 1), depending on weather conditions. A l l sessions occurred in the same vicinity on the colony, and commenced at around 22h30. Table 2.1. Sampling dates and number of b i l l loads sampled from rhinoceros auklets provisioning chicks at Triangle Island, B C . Date Number of Date Number of (Julian Date) Bill loads (Julian Date) Bill loads 2003 June 20 (171) 10 2004 June 20 (172) 7 June 30(181) 7 June 30 (182) 9 July 12 (193) 9 July 10 (192) 7 July 20 (200) 11 . July 20 (202) 8 July 27 (207) 12 July 30 (212) 10 August 9 (220) 9 Aug 9 (222) 9 August 16 (227) 1 Adults en route to provisioning their chicks were caught with small fishing nets and as much of the b i l l load retrieved as possible. B i l l loads that lost fish during capture were marked as incomplete. Fish were daubed dry of excess water, weighed on an electronic mass balance (± O.lg), and fork and standard length were measured (± 1mm) (Cailliet et al. 1996). Rockfish were classified only to genus due to difficulty in determining species. Rockfish previously identified as prey at Triangle Island include yellowtail (Sebastes flavidus) and widow (Sebastes entomelas) (Vermeer and Westrheim 1984). With the exception of some samples in 2003, which were frozen in Ziploc® bags in species-specific bundles, fish over 6g were frozen individually in Whirlpaks ®. Small fish (rockfish and juvenile sand lance) were frozen in bundles of 8g, keeping fish brought in by the same bird together where possible. Samples were stored in a propane freezer at 12 -10°C in the field and then at -20 °C once in the laboratory. Samples thawed at least partially during transport. Samples were also at least partially thawed once in the lab for the removal of sagittal otoliths and to evaluate mesenteric fat depots and reproductive maturity. A subset of samples was re-weighed to determine loss of water between capture and time of energy density processing. Differences (accounting for removed otoliths) were used to establish a water content correction factor for specimens that were not re-weighed (Equation 1, Appendix A ) . (1) Dessicated mass = 0.896(fresh mass)-0.233 (R 2 = 0 . 9 9 3 , p<0.0001) There is a risk of pseudoreplication within the bi l l loads, as in some instances I sampled more than one fish of the same species from a single b i l l load. However, the variation in proximate composition between fish of the same species within b i l l loads was similar to that observed across the entire sample of species. I therefore felt comfortable treating fish from the same bi l l load as independent samples. Ag ing Sandlance Pacific sand lance were classified as adults (1+) or young of year juveniles (0+) based on otolith measurements. Sagittal otoliths were removed, cleaned of tissue using a damp cloth, and patted dry. Each otolith was then cross-sectioned along the transverse plane using a pair of nail-clippers, and burnt to a light brown over an alcohol flame. The 13 sectioned otolith was then mounted in modeling cement and examined under mineral oil using a dissecting scope at 40x. Individuals with no annuli outside of the nucleus were recorded as first year fish (0+). Individuals with one annulus outside of the nucleus were counted as one year old fish (1+), and so on. Other Species Pacific saury were all classified as young of year juveniles (0+) based on published growth estimates (Watanabe et al. 1988, Suyama et al. 1994). Rockfish and salmon were all classified as young of year (0+) and new smolts (1+), respectively, based on morphology and size (Woodbury and Ralston 1991, Moser and Boehlert 1991). Energy Content I determined energy content of individual prey using proximate composition analysis, which constructs energy densities by measuring total l ipid, protein, water and ash (mineral content) in prey items. A l l prey above 6g were processed individually, with the exception of juvenile sand lance and juvenile rockfish, which were processed in 8g batches due to their small size. Samples were homogenized using a stainless steel mortar and pestle. A sub-sample (2g) of homogenate was placed in a dried ceramic crucible, dessicated in a drying oven at 100°C for 24 hours to determine water content, weighed, and then ashed in a 600°C muffle furnace for two hours to determine ash content by subtraction. Crude lipid content was determined using a modified Bl igh and Dyer method (described in Higgs 1979, modified by Crossin 2003) wherein 2g of wet tissue was extracted with 40ml of 1:1 chloroform methanol, and 8 ml of distilled water. This 14 extraction was performed twice for each sample and results were only retained i f the l ipid content of the two sub-samples were within 1% of one another, in which case the average of the two measures was taken. The original Bl igh and Dyer (1959) method has been found in some instances to underestimate l ipid quantities, with underestimation increasing as l ipid content increases (Iverson et al. 2001). However, the modified method used here greatly increases the tissue to solvent ratio and in doing so extracts lipids as efficiently as the high tissue:solvent Folch method (Appendix B) . Protein content was determined by subtracting l ipid, water, and ash mass from that of the total sample. This method is an accepted approach (Lawson et al. 1998, Crossin 2003, Dempson et al 2004) as fish have negligible carbohydrate values (Sidwell et al. 1974, Craig et al. 1978, Hartman and Brandt 1995). Energy densities were calculated using proximate composition results multiplied by published energy values for fish lipids (39.3 kJ/g) and protein (17.8 kJ/g) (Equation 2, Schmidt-Nielsen 1997, p. 171). The latter value is an estimate of energy derived from protein metabolism in uricotelic vertebrates, adjusted for the metabolism-specific energy return of protein (Schmidt-Nielsen 1997). This is also the value commonly used in seabird prey studies (e.g. Anthony et al. 2002, Van Pelt et al. 1997). (2) Energy density (kJ/g wet mass) = (% lipid wet mass * 39.3 + % protein wet mass * 17.8) / 100 15 Wet mass reflects the energetic return on prey items more accurately than dry mass. However, prey are subject to significant dessication during transport to the colony, which can skew interpretation of proximate composition trends within the fish population (Montevecchi and Piatt 1987). To reduce this potential bias, I used dry mass energy densities for all data analyses and present information on proximate composition in this format. Statistical analyses I used SPSS (SPSS 1999) for all analyses, with the exception of Welch's approximation for unequal variances, for which I used J M P 5.1 (SAS 2003). Between species differences Data on energy density and respective contributions of l ip id and protein to energy density were normally distributed but of unequal variance, which could not be remedied by transformation. I therefore evaluated between-species differences in these variables using Welch's approximation and Tamhane's T2 post-hoc comparisons (Tamhane 1979). Otherwise I used A N O V A . Factors accounting for within-species variation I used stepwise forward multiple regression (entry probability: 0.05, exit probability: 0.1) to identify factors accounting for variation in quality between individuals of the same species. The effects of time of capture (year, time within year) 16 and body mass on energy density were examined using forward stepwise multiple regression analyses on untransformed dry mass energy density data, with entry and exit probabilities of 0.05 and 0.1, respectively. Predicting the energy content of a single fish I calculated total fish energy content by multiplying wet mass energy density by wet mass at time of capture. To increase accuracy, I developed year-specific regression models for species captured in both 2003 and 2004. For each year I evaluated wet mass and Julian date as estimators of square-root transformed total prey energy using forward stepwise multiple regression with entry and exit probabilities of 0.05 and 0.10 respectively. Total fish energy was not square root transformed for adult sand lance in 2004 as residuals were already normally distributed. The same analysis was done using dry biomass but not including day of year as I was interested in how much variation could be accounted for solely by dry biomass. Dry biomass data met regression assumptions and did not require transformation. 17 R E S U L T S Sample collection and general diet description Four genera of fish were collected over the course of two seasons from rhinoceros auklets provisioning chicks at Triangle Island, British Columbia: Pacific saury, pacific sand lance, juvenile rockfish, and juvenile salmon (numbers and morphometric information in Appendix C). Rockfish and salmon, though present in small amounts in chick diets during 2003, were only sampled in 2004. The 2003 diet was dominated by pacific saury while that of 2004 was a much more even mix of sand lance, rockfish, saury and salmon (Figure 2.1). 2003 2004 • Pacific Sand lance • Pacific Sand lance (j) • Pacific Saury 0 Rock f i sh • S a l m o n Herr ing M Other Figure 2.1. Wet mass diet composition of rhinoceros auklet chicks at Triangle Island, British Columbia across all sampling sessions. 18 Differences among species Energy densities and contribution of lipid and protein Energy densities of fishes collected ranged from 20.31± 1.74 kJ/g dry mass (mean ± sd) in sand lance to 17.01± 0.45 kJ/g dry mass in pacific salmon (Figure 2.2). Salmon and pacific saury (2003) had significantly lower energy density (Welch's F 7 > 2 5 = 45.15, p<0.0001) than other species: all other species had similar energy densities. Both pacific saury in 2003 and pacific salmon in 2004 had lower lipid-derived energy (Figure 2.1, Welch's F7,24.9i = 40.66, p<0.0001) and higher protein-derived energy than other species. (Figure 2.2, Welch's F 7 ,42.20 = 28.00, p<0.0001). Between-species Variation in Proximate Constituents Between-species differences in l ipid and protein content are captured in the above analysis of their respective contributions to energy density. Proximate composition i values for each of lipid and protein may be found in Table 2. Water content (% wet biomass) was greatest in juvenile salmonids (mean ± sd = 80.22% ± 0.845) and lowest in juvenile saury captured in 2004 (mean ± sd = 73.08% ± 2.10). These were the only two species that differed from the mean water content of remaining species ( A N O V A F 7 J 2 5 =18.423, p<0.0001, Table 2.2). 19 25 20 S 15 DO ^10 E Q G I Lipid • Protein A A -I-ABC B - f a b c b e A -9E— a b Sand Lance Sand Sand Sand Saury Saury Rockfish Salmon Lance Lance j Lance j (2003) (2004) (2003) (2004) (2003) (2004) Species (year) spp spp (2004) (2004) Figure 2.2. Dry mass energy densities and relative contributions of protein and lipid constituents of forage fish obtained from rhinoceros auklets provisioning chicks at Triangle Island, British Columbia. Similar labels on the same horizontal plane indicate statistically similar groups, capital letters represent similarity in total energy density (Tamahane multiple comparisons, p<0.0001). Error bars represent 95% confidence intervals for each of lipid and protein. (j= juvenile) Mineral content, measured as ash, ranged from 10.47% ± 0.43 (mean + sd) in juvenile pacific saury in 2004 to 13.09% ± 1.10 in juvenile pacific salmon. Pacific saury from 2004 and juvenile sand lance from 2004 had lower ash content than juvenile rockfish and juvenile salmon but there was otherwise no difference between species (Welch's F 7 , 2 5 = 12.5381, p<0.0001, Table 2.2). 20 Within-species variation in proximate composition Each of sand lance, pacific saury, and juvenile rockfish showed large variation in proximate composition, within all years and ages (Figure 2.2, Table 2.2). Variation in l ipid and ash content were highest, while protein and water content varied little in all species. Wet Body Mass Wet body mass weakly predicted dry mass energy density of Pacific saury in 2003 only (Table 2.3). No other significant relationships between wet mass and energy density were observed. Year and Season Effects Pacific saury captured in 2003 had lower energy densities than those captured in 2004 (Welch's F 7,i 25=41.15, p<0.001, F ig . 2.2). There was some indication that juvenile pacific sand lance had lower energy densities in 2003 than in 2004 (Figure 2.2) but low power due to small 2003 sample size may have prevented detection of differences. Juvenile pacific sand lance in 2004 had higher energy densities later in the season due to increased lipid content (Table 2.3, Figure 2.3). As no juvenile sand lance were captured late in 2003, the seasonal effect could not be assessed for that year. N o other species varied in energy density within or between years. 21 Table 2.2. Water content (percent wet mass), proximate composition (percent dry mass), and energy density (percent dry mass) of forage fish fed to rhinoceros auklet nestlings at Triangle Island, British Columbia, (mean ± sd, coefficient o f variation in brackets). Species Year N %Water % A s h * %Lip id* %Protein* Energy Density (kJ/g)* 2003 16 74.63 ± 1.57 11.70 ± 1.68 13.04 ± 4 . 3 6 75.45 ± 3 . 6 7 18.37 ± 1.10 Pacific Saury (j) (2.1) (14.4) (33.4) (4.9) (6.0) 2004 14 73.09 ± 2 . 1 0 10.47 ± 0 . 9 8 18.32 ± 3 . 7 1 71.21 ± 3 . 6 9 19.79 ± 0 . 8 6 (2.1) (9.3) (20.2) (5.2) (4.3) 2003 35 76.12 ± 1.80 12.01 ± 3 . 1 7 19.74 ± 5 . 8 3 68.38 ± 5 . 1 7 19.81 ± 1.59 Pacific Sand lance (a) (2.4) (25.8) (29.5) (7.6) (8.0) 2004 27 74.57 ± 2 . 1 2 11.47 ± 2 . 9 9 21.48 ± 6 . 9 5 67.05 ± 6.84 20.31 ± 1 . 7 4 (2.8) (26.0) (32.4) (10.2) (8.6) 2003 76.12 ± 0 . 8 1 11.71 ± 2 . 0 0 10.62 ± 5 . 0 2 77.80 ± 3 . 3 8 17.85 ± 1.42 Pacific Sand lance (j) (1.05) (17.1) (47.2) (4.3) (8.0) 2004 16 76.18 ± 1.50 10.51 ± 0 . 4 3 20.31 ± 2 . 8 2 69.18 ± 3 . 1 0 20.22 ± 0.58 (2.0) (0.4) (13.9) (4.5) (2.9) Rockfish spp (j) 2004 16 75.57 ± 1.61 (2.1) 12.62 ± 1.06 (8.4) 19.75 ± 3 . 1 5 (15.9) 67.70 ± 3 . 3 9 (5.0) 19.71 ± 0 . 7 1 (3.6) Salmon spp. (j) 2004 12 80.22 ± 0.85 (1.0)) 13.09 ± 1.10 (8.4) 7.78 ± 1.78 (22.9) 79.13 ± 2 . 0 5 (2.6) 17.01 ± 0 . 4 5 (2.7) *Wet mass values can be calculated. Wet Mass value = Dry Mass value x (1-proportion water) Table 2.3. Factors accounting for intra-specific variation in energy density in species R 2 Total df Body Ju l i an Species N R 2 (1,2) F P mass (g) Date Pacific Saury (2003) 14 0.293 1,13 6.810 0.022 0.293 0 Pacific Sand lance (juveniles, 2004) 16 0.335 1,13 8.055 0.001 0 0.335 Prey types for which Julian date or body mass did not relate to energy density Pacific Sand lance (adults, 2003) 32 2,32 3.037 0.091 Pacific Saury (2004) 13 2,11 1.146 0.353 Pacific sand lance (adults, 2004) 26 2,24 1.292 0.670 Rockfish (juveniles, 2004) 16 2,14 2.519 0.116 Salmon (juveniles, 2004) 11 2,9 0.047 0.818 25 20 15 10 0 Lipid • Protein June 30 August 9 Figure 2.3. Energy density (kJ/g dry mass) variation in juvenile pacific sand lance (Ammodytes hexapterus) by time of year in 2004. Top error bars represent 95% confidence intervals for energy density, bottom error bars represent 95% confidence intervals for lipid-derived energy. 23 Predicting total prey energy content (kJ) within species Wet mass and in some instances Julian date accounted for most variation (87 -99%) in total prey energy content (Table 4). Dry mass alone was an even better predictor of total prey energy content, accounting for 95-99% of variation (Table 5). 24 Table 2.4. Prediction of total energy (kJ) in forage fish delivered to rhinoceros auklet nestlings on Triangle Island, British Columbia. Species Year Equation for Total Energy (kJ) R 2 Model p n Pacific Saury (j) 2003 y = (0.361 • wt + 2.05-10"2 • (ID) - 1.239)2 0.968 <0.001 15 2004 y =(0.279 • wt + 4.344) 2 0.942 <0.001 14 Pacific 2003 y =(0.308 • wt + 3.705) 2 0.993 <0.001 35 Sand lance (a) 2004 y =(0.291 • wt + 4.170) 2 0.882 <0.001 27 Pacific Sand lance (j) 2004 y =(0.879 • wt + 5.03 • 10- 3 • (ID) + 0.316) 2 0.909 <0.001 15 Rockfish spp. (j) 2004 y : =(0.666 • wt + 8.75 • IO"2 • (JD) + 0.246) 2 0.868 <0.001 17 Salmon spp. (j) 2004 y=(0.198 • wt + 4.087) 2 0.989 <0.001 12 Where: wt = wet mass in g, JD = Julian Date Table 2.5. Prediction of total energy content (in kJ) by dry mass (in g) of forage fish species taken by rhinoceros auklets breeding at Triangle Island, British Columbia. Species Year Equation for Total Energy (kJ) R2 F P n Pacific Saury (j) 2003 y=20.05x - 4.33 0.988 1.21 x 10 3 <0.001 15 14 2004 y=20.11x-0.81 0.994 2.05 x 10 3 <0.001 Pacific 2003 y=21 .78x -5 .50 0.986 2.39 x 10 3 <0.001 35 Sand lance (a) 2004 y=21.29x-2.92 0.963 6.74 x 10 2 <0.001 27 Pacific Sand lance (j) Both y=21.64x-0.42 0.968 4.30 x 10 2 <0.001 15 Rockfish spp (j) 2004 y=17.65x + 1.32 0.947 2.84 x 10 2 <0.001 17 Salmon spp. (j) 2004 y=17.50x- 1.96 0.994 1.89 x 10 3 <0.001 12 D I S C U S S I O N I investigated energy density and proximate composition of four main prey types used by rhinoceros auklets at Triangle Island, British Columbia. The goals of the study were to determine energy density of common prey species, evaluate whether biomass of chick meals can be used as a proxy for total meal energy, develop models of total prey energy content, and better understand the energetic implications of different prey types for rhinoceros auklets. The relative biomass of chick meals is likely not an accurate estimator of relative meal energy content. Different prey types'had different energy densities (Figure 2.2) and different water contents (Table 2.2), both of which wi l l skew any energetic comparisons of different diets on the basis of biomass alone. Further, differences between species did not remain consistent between years (Figure 2.2). Intraspecific variation in quality arising from year (saury and juvenile sand lance, Figure 2.2), time of year (juvenile pacific sand lance, Figure 2.3, Table 2.3) and body size (pacific saury, Table 2.3) also suggests that the relative quality of these prey species varies, and affects their value to rhinoceros auklets accordingly. The most plausible cause of interannual variation in energy density is differing food availability and resultant growth conditions between the two years. Independent marine sampling studies showed higher densities of boreal shelf copepods in 2004 than in 2003 (DFO 2003, D F O 2004). While the diets of saury and sand lance are not well known for the Triangle Island area, copepods are heavily exploited by both species in 26 other areas (Hart 1973, Watanabe et al. 2003). There is also evidence that both species decrease in growth rate and lipid deposition if feeding conditions are poor (Robards et al. 2002, Watanabe et al. 2003). Although adult sand lance did not differ in energy density or other constituents between years, interannual variation should still be considered i f these data are to be applied across years varying substantially in ocean productivity. Sand lance energy content varies with oceanic conditions in other locations (e.g. Robards et al. 2002, Wanless et al. 2005), due to both interannual and geographic variation in food availability. This is likely the case for sand lance near Triangle, and it is possible that the short window of this study did not capture adequate variation in ocean productivity. Energy densities of juvenile rockfish were substantially higher than one previously published value (15.93 kJ/g dry mass, Van Pelt et al. 1997) but on par with local values reported during the 1980s (21.77 kJ/g dry mass, Vermeer and Devito 1986). L ip id contents were also on par with those found in California rockfish during years of good food availability (Rau et al. 2001). While the physiology of settling juveniles is not very well understood (Love et al. 2002), in aquaculture environments l ipid storage in juvenile rockfish occurs when food is in excess and when diets have high ratios of l ipid to protein (Lee 2002). There is evidence that storage of lipids during food excess may also occur in natural systems (Rau et al. 2001). In 2004, juvenile rockfish numbers were high along many parts of the Pacific coast (Baltz 2004), and were also unusually abundant in Common Murre nestling diets at Triangle (Triangle Island Research Station, unpubl. 2 7 data), suggesting an exceptionally good year for young of year production. It is possible that these good conditions resulted in higher energy densities than might be observed under poorer conditions and may have contributed to their higher than average use by rhinoceros auklets in 2004. Seasonal Trends The energy density of juvenile sand lance improved over the course of the season and corroborates patterns reported for Alaskan pacific sand lance (Robards et al. 1999a), and in the genus Ammodytes in general (Hislop et al. 1991). The increase in energy density with Julian date is likely due to a transition from growth to reserve accumulation in preparation for poor food availability during the winter months (Robards et al. 1999b). The energetic value of juvenile sand lance to rhinoceros auklets w i l l thus vary depending on when during the year sand lance are exploited. No seasonal trends in energy density were observed in adult pacific sand lance, though they have been reported in other areas (Robards et al. 1999a, Hislop et al. 1991). In those other areas, energy and lipid content peaked in mid-June and remained high through early August, at which time both energy and l ipid content started to decline. The chick-rearing season of rhinoceros auklets at Triangle may overlap this stable peak in sand lance energy density and explain why no seasonal trends were observed in this study. 28 Body mass Pacific saury showed a weak positive relationship between energy density and size, though only for 2003. Similar relationships have been reported for other fast growing fish (Anthony et al. 2000). While fast growth is generally favoured amongst younger fish, there is a point at which reserves become important for preventing starvation during periods of low food availability. However, this relationship should be evident in all years and does not account for the lack of a similar observed trend in 2004. More importantly, pacific saury had markedly lower energy densities than those published elsewhere (Kurita 2003, Suh et al. 2000), primarily due to substantially lower l ipid content. However, these published accounts were of adult fish (1-2+) only and highlight the importance of considering both size and age in this species when applying the results of other studies. Est imating total prey energy content Some factors that were not significant in estimating dry mass energy density appeared to influence estimation of total energy density using wet mass. For instance, Julian date was a significant factor in predicting total energy content of adult sand lance (2003) and juvenile rockfish, suggesting that wet mass energy densities changed within both of these seasons. Because of the possible influence of dehydration during transport to chicks (Montevecchi and Piatt 1987), particularly for small prey, it is difficult to assess 29 whether this wet biomass result reflects different transport distances, or a true dynamic within sand lance and rockfish populations. While both dry and wet masses were good predictors of total prey energy density within species, the interannual variation within species suggests that models should be re-calibrated each year for applications requiring accurate estimates of energy content. Juvenile salmon might be the exception to the need to recalibrate as they showed relatively low variation in energy density within years (Table 2.2) and low interannual variation in other studies (Trudel et al. 2005). Implications for rhinoceros auklets Based on the above results of inter and intraspecific variation in prey quality, it seems reasonable that differences in prey quality could help account for interannual differences in rhinoceros auklet reproductive success at Triangle Island. For instance, to achieve equal energy delivery, rhinoceros auklets provisioning with salmon would have to deliver 1.5 times more wet mass than those provisioning with pacific sand lance (calculated from water content and dry biomass energy densities of salmon and sand lance). The effects of this energy density difference on rhinoceros auklet chicks and parents could be significant i f salmon were exploited for any significant amount of time. This difference in quality may in part explain why salmon are used to a lesser extent than all other examined species. The 2003 coincidence of lower quality pacific saury yet higher prevalence in diets also suggests that rhinoceros auklets may be most dependent on saury when they are less energetically rewarding. This difference in quality could 30 reasonably have a negative effect on energy delivery rates and hence rhinoceros auklet reproduction. Future studies measuring both prey quality and reproductive success would offer further insight into the extent to which prey quality influences the reproductive success of rhinoceros auklets at Triangle Island. 31 R E F E R E N C E S Ainley, D . G . , and R.J . Boekelheide. 1990. Seabirds of the Farallon Islands: Ecology, Dynamics, and Structure of an Upwelling-System Community. Stanford, C A : Stanford University Press. Anthony, J .A. , D . D . Roby and K . R . Turco. 2000. L i p i d content and energy density of forage fishes from the northern Gul f of Alaska. Journal of Experimental Marine Biology and Ecology 248:53-78 Baltz, K . 2004. Cruise report DS-04-03, David Starr Jordan, M a y 4 - June 11, 2004: rockfish recruitment assessment. National Marine Fisheries Service, Santa Cruz, California. 24 pp. Beamish, R.J . , A . J . Benson, R . M . Sweeting and S . M . Nevil le 2004. Regimes and the history of the major fisheries off Canada's west coast. Progress in Oceanography 60: 355-385 Bertram, D .F . , D . L . Mackas, and S . M McKinne l l . 2001. The seasonal cycle revisited: interannual variation and ecosystem consequences. Progress in Oceanography 49: 283-307 Bertram, D . R., C . V . J . Welham, and R.C.Ydenberg. 1996. Flexible effort in breeding seabirds: adjustment of provisioning according to nestling age and mass Canadian Journal of Zoology 74: 1876-1881 Bertram, D.F . , G . W . Kaiser, and R . C . Ydenberg. 1991. Patterns in provisioning and growth of nestling rhinoceros auklets. The Auk 108: 842-852 Bl igh, E . G . and W.J . Dyer. 1959. A rapid method for total l ipid extraction and purification.1 Canadian.Journal of Biochemical Physiology. 37: 911-917 Cailliet, G . M . , M . S . Love, and A . W . Ebling. 1996. Fishes: a field and laboratory manual on their structure, identification and natural history. Wadsworth Inc. 194pp. Calow, P. 1985. Adaptive aspects of energy allocation. Pp. 13-31 in Fish Energetics: New Perspectives. (P. Tytler and P. Calow Eds.) John Hopkins Press. 349pp. Craig. J.F., M.J.Kenley, and J.F. Tailing. 1978. Comparative estimations of the energy content of fish tissue from bomb calorimetry, wet oxidation and proximate analysis. Freshwater Biology 8: 585-590 Crossin, G.T. 2003. Effects of ocean climate and upriver migratory constraints on the bioenergetics, fecundity, and morphology of wild, Fraser River salmon. M.Sc . thesis, University of British Columbia, Vancouver. 32 Dempson I. B . , C.J . Schwarz, M . Shears and G . Furey. 2004. Comparative proximate body composition of Atlantic salmon with emphasis on parr from fluvial and lacustrine habitats. Journal of Fish Biology 64: 1257-1271 D F O , 2004. 2003 Pacific Region State of the Ocean. D F O Science Ocean Status Report 2003 D F O , 2005. 2004 Pacific Region State of the Ocean. D F O Science Ocean Status Report 2004 Diamond, A . W. , and C M . Devlin. 2003. Seabirds as indicators of changes in marine . ecosystems: ecological monitoring on Machias Seal Island. Environmental Monitoring and Assessment 88: 153-175 Foy, R .J . and A . J . Paul. 1999. Winter feeding and changes in somatic energy content of age-0 pacific herring in Prince Wi l l i am Sound, Alaska. Transactions of the American Fisheries Society. 128: 1193-1200 Gaston, A . J., and S. B . C . Dechesne. 1996. Rhinoceros Auklet (Cerorhinca monocerata). In The Birds of North America, No. 212 (A. Poole and F. G i l l , Eds.). The Academy of Natural Sciences, Philadelphia, P A , and The American Ornithologists 'Union, Washington, D C . Golet, G . H . , K . J . Kuletz, D . D . Roby, and D . B . Irons. 2000. Adult prey choice affects chick growth and reproductive success in Pigeon Guillemots. A u k 117: 82-91 Gjerdrum, C , A . M . J . Vallee, C . C . St.Clair, D .F . Bertram, J .L. Ryder, and G.S. Blackburn. 2003. Tufted puffin reproduction reveals ocean climate variability. Proceedings of the National Academy of Science 100: 9377-9382 Harfenist, A . 1995. Effects of growth-rate variation on fledging of rhinoceros auklets (Cerorhinca monocerata). Auk 112: 60-66 Hart, J .L . 1973. Pacific fishes of Canada. Bulletin of the Fisheries Resource Board of Canada. 180: 740pp. Hartman, K . , and S. B . Brandt. 1995. Estimating energy density of fish. Transactions of the American Fisheries Society 124: 347-355 Hedd,A., D . F . Bertram, J .L. Ryder and I.L. Jones. 2006. Effects of inter-decadal climate variability on marine trophic interactions: Rhinoceros Auklets and their fish prey. Marine Ecology Progress Series 309: 263-278 Higgs, D . A . , J.R. Markert, D . W . MacQuarrie, J.R. McBride , B .S . Dosanjh, C . Nichols, and G . Hoskins. 1979. Development of practical dry diets for coho salmon, 33 Oncorhynchus kisutch, using poultry-by-product meal, feather meal, soybean meal and rapeseed meal as major protein sources. In Finfish nutrition and fishfeed technology. V o l . 2. (J.E. Halver and K . Tiews Eds.). Heenemann Verlagsgesellschaft M b H , Berlin, Germany. Hislop, J .R.G. , M . P . Harris, and J . G . M . Smith. 1991. Variation in the calorific value and total energy content of the lesser sandeel (Ammodytes marinus) and other fish preyed on by seabirds. Journal of Zoology 224: 51-517 Iverson, S.J., S.C. Lang, and M . H . Cooper. 2001. Comparison of the Bl igh and Dyer x and Folch methods for total l ipid determination in a broad range of marine tissue. Lipids 36: 1283-1287 Kurita, Y . 2003. Energetics of reproduction and spawning migration for Pacific saury (Cololabis saira). Fish Physiology and Biochemistry 28: 271-272 Lawson, J.W., A . M . Magalhaes, E . H . Mi l le r . 1998. Important prey species of marine vertebrate predators in the northwest Atlantic: proximate composition and energy density. Marine Ecology Progress Series 164: 13-20 Lee, S . M . , L G . Jeon, and J .Y . Lee. 2002. Effects of digestible protein and lipid levels in practical diets on growth, protein utilization and body composition of juvenile rockfish (Sebastes schlegeli). Aquaculture 211: 227-239 Love, M . , M . Yklavich, and L . Thorsteinson. 2002. The Rockfishes of the Northeast Pacific. University of California Press. 405 pp. Love, R. M . (1970). The Chemical Biology of Fishes. Academic Press, London. Mantua, N . J . and S.R. Hare 2002. The Pacific decadal oscillation. Journal of Oceanography 58: 35-44 Montevecchi, W . A . and J.F. Piatt. 1987. Dehydration of seabird prey during transport to the colony: effects on wet mass energy densities. Canadian Journal of Zoology. 65:2822-2824 Moser, H . G . , and G . W . Boehlert. 1991. Ecology of pelagic larvae and juveniles of the genus Sebastes. Environmental Biology of Fishes 30: 203-224 Rau, G . H . , S. Ralston, J. Southon, and F.P.Chavez. 2001. Upwell ing and the condition and diet of juvenile rockfish: A study using 1 4 C , 1 3 C , and 1 5 N natural abundances. Limnology and Oceanography 46: 1565-1569 Robards, M . D . , J .A. Anthony, G . A . Rose and J.F. Piatt. 1999a. Changes in proximate composition and somatic energy content for Pacific sand lance (Ammodytes hexapterus) from Kachemak Bay, Alaska relative to maturity and season. Journal of Experimental Marine Biology and Ecology 242: 245-258 34 Robards, M . D . , J.F. Piatt, and G . A . Rose. 1999b. Maturation, fecundity, and intertidal spawning of Pacific sand lance in the northern Gul f of Alaska. Journal of Fish Biology 54: 1050-1068 Robards, M . D . , G . A . Rose, and J.F. Piatt. 2002. Growth and abundance of Pacific sand lance, Ammodytes hexapterus, under differing oceanographic regimes. Environmental Biology of Fishes 64: 429-441 S A S Institute. 2003. JMP-In Version 5.1 Schmidt- Nielsen, K . 1997. Animal physiology: adaptation and environment, 5 t h Ed . Cambridge University Press, New York. Sidwell, V . D . , P.R. Foncannon, N .S . Moore, and J .C. Bonnet. 1974. Composition of the edible portion of raw (Fresh or frozen) crustaceans, finfish, and mollusks. 1. Protein, fat, moisture, ash, carbohydrate, energy values, and cholesterol. Marine Fisheries Review 36: 21-35 SPSS 1999. SPSS for Windows Standard Release 10.0.5 (27 Nov 1999) Suh, Junghyuck, D . Choi , E . Lee, and M . Hong. 2000. Surveillance program on dioxin-like compounds in fatty food in Korea. Organohalogen Compounds 66: 2023-2026 Suyama, S., Sakurai, Y . and K . Shimazaki. 1994. Age and growth of Pacific saury Cololabis saira (Brevoort) in the western North Pacific Ocean estimated from daily otolith growth increments. Fisheries Science 62: 1-7 Tamhane, A . C . 1979. A comparison of procedures for multiple comparisons. Journal of the American Statistical Association 74: 471-480 Trites, A . W . and C P . Donnelly. 2003. The decline of Steller sea lions Eumetopias jubatus in Alaska: a review of the nutritional stress hypothesis. Marine Mammal Review 33: 3-28 Trudel M , S. Tucker, J.F.T. Morris, D . A . Higgs, and D . W . .Welch. 2005. Indicators of energetic status in juvenile coho salmon and Chinook salmon. North American Journal of Fisheries Management 25: 374-390 Van Pelt, T .L , J.F. Piatt, B . K . Lance and D . D . Roby. 1997. Proximate composition and energy density of some North Pacific forage fishes. Comparative Biochemistry and Physiology A . 118: 1393-1398 Vermeer, K . , and K . Devito. 1986. Size, caloric content, and association of prey fishes in meals of nestling rhinoceros auklets. Murrelet 67: 1-9 35 Vermeer, K . and S. J. Westrheim. 1984. Fish changes in diets of nestling rhinoceros auklets and their implications, pp. 96-105 in Marine Birds: Their feeding ecology and commercial fisheries interactions. (D. N . Nettleship, G . A . Sanger, and P. F. Springer, Eds.). Special Publication, Canadian Wildlife Service, Ottawa. Wanless S., M . P . Harris, P. Redman, and J.R. Speakman. 2005. L o w energy values of fish as a probable cause of a major seabird breeding failure in the North Sea Marine Ecology Progress Series 294: 1-8 Watanabe, Y . , J .L: Butler, and T. M o r i . 1988. Growth of Pacific saury, Cololabis saira, in the northeastern and northwestern Pacific Ocean. Fisheries Bulletin. U .S . 86: 489^198 Watanabe, Y . , Y . Kurita, M . Notto, Y . Oozeki, andD. Kitagawa. 2003. Growth and survival processes of Pacific saury (Cologalis siara) in the huroshio-Oyashio transitional waters. Journal of Oceanography 59: 403-414 Whittow, G . C . 2002. Seabird Reproductive Physiology and Energetics pp. 409-438 in The Biology of Marine Birds. (E.A.Shreiber and J . Burger, Eds.) C R C Press, Boca Raton. Woodbury, D . and S. Ralston. 1991. Interannual variation in growth rates and back-calculated birthdate distributions of pelagic juvenile rockfishes (Sebastes spp.) off the central California coast. Fisheries Bulletin. U .S . 89: 523-533 36 C H A P T E R T H R E E : Correlations of Prey Quality and Reproductive Success in Rhinoceros Auklets I N T R O D U C T I O N Temperate piscivorous seabirds operate on marginal energy budgets, particularly during the breeding season when parents must meet both their own needs and those of their offspring (Whittow 2002, Ni izuma and Yamamura 2001). Further, they must do so within commuting-distance of the breeding colony, and within the context of fluctuating prey communities that vary over wide spatiotemporal scales (Litzow et al. 2000, Robards etal. 1999a). The annual reproductive success of piscivorous seabirds is often correlated with the breeding season availability and condition of one or two high quality prey species, usually small-bodied schooling fish (e.g Rindorf et al. 2000). Changes in oceanic production can substantially alter the distribution (Abookire and Piatt 2005), abundance (Cury et al. 2000), and l ipid content (Robards et al. 1999b) of these prey species, which can have significant consequences for predators (e.g. Wanless et al. 2005). The abundance and condition of high lipid fish species like Japanese anchovy (Engraulis japonicus), pacific sand lance (Ammodytes hexapterus) and lesser sand eels (Ammodytes marinus), are linked to population-level reproductive success of rhinoceros auklets (Takahashi et al. 2001), pigeon guillemots (Litzow et al. 2002), and common guillemots (Rindorf et al. 2000), respectively. A version of this chapter will be submitted for publication. Beaubier, J.E., J.M. Hipfner, and D.S. Srivastava. Correlations of prey quality and productivity of rhinoceros auklets. 37 In optimal condition, each of these prey species provides high amounts of energy per unit biomass (Anthony et al. 2000, VanPelt et al. 1997) which, all else being equal, significantly increases both provisioning rates and return on the energy expended during their capture (Golet et al. 2000). Each of these factors has the potential to affect individual predator fitness through increased survival of both parents and offspring. Chicks fed higher quality prey generally receive more energy (Golet et al. 2000, Diamond and Devlin 2003) and digest it more efficiently (Dahdul and Horn 2003, Niizuma and Yamamura 2004). As a result, these chicks generally grow more quickly (Oyan and Anker-Nilssen 1996, Kitaysky 1999, Visser et al. 2002), fledge at a younger age (Harfenist 1995), and with higher l ipid reserves (Romano et al. 2000, Dahdul and Horn 2003, Takenaka et al 2005) than chicks fed lower quality prey. Faster growth rates and younger age at fledging can decrease the length of time adults must provision both self and offspring and be limited to foraging areas within commuting distance to the colony. Younger age at fledging also decreases the length of time adults must spend in the terrestrial colony environment, which may reduce corresponding risk of predation and injury (Ydenberg 1989). Chicks fledging at a higher mass may also have higher post-fledging survival, as chicks with larger fat reserves may be better able to buffer against food shortages while learning how to forage on their own (Ricklefs 1983, Phillips and Hamer 1999, Reid et al. 2000). Short term ( 1 - 2 weeks) nutritional stress can also cause impaired cognitive ability in seabird fledglings, regardless of fledging mass, and bear negative implications for post-fledging survival (Kitaysky et al. 2006). 38 Rhinoceros auklets breeding at Triangle Island provision their chicks with 5 main prey types: pacific sand lance (Ammodytes hexapterus), pacific saury (Cololabis saira), juvenile rockfish (Sebastes spp.), juvenile salmon (Oncorhynchus spp.) and, though less frequent in recent years, pacific herring (Clupea harengus) (Vermeer 1979, Hedd et al. 2006). Changes in nestling growth rates and fledging masses at Triangle have been correlated with both the timing and proportion of different species in chick diets (Vermeer 1980, Bertram et al. 2001). Growth rates are lower in years when pacific saury appear early on in chick diets, and are higher in years when nestlings are fed a high proportion of sand lance (Vermeer 1980, Bertram and Kaiser 1993, Hedd et al. 2006). New information on local prey quality (Chapter Two of this thesis) can now be used to help determine i f the above correlations between diet and reproductive success are due to differences in prey quality, or i f availability in the vicinity of the colony is a more likely factor. If chicks receive lower masses of food, yet prey quality between years remains constant, prey availability should be considered the more likely cause of differences in reproductive success. Here I reconstructed the energy delivered to chicks over the course of two breeding seasons during which chicks were fed distinctly different diets. Through this reconstruction I determined i f either prey energy or biomass delivered to rhinoceros auklet chicks varied between years of different diets. I also compared the different diets for average quality (energy density). To determine i f the different diets correlated with changes in rhinoceros auklet reproductive success, I compared growth rates and fledging 39 characteristics of nestlings between years. I also tested if parental quality played a role in mediating population level responses to changing prey quality and availability. 4 0 METHODS Study Species Rhinoceros auklets are epipelagically foraging alcids that breed in burrows and provision their chicks with whole prey brought in to the colony at night (Gaston and Dechesne 1996). On average, each parent makes a single provisioning trip to the colony each night, and carries multiple fish in a single delivery (Bertram et al. 1991). A l l contents carried in a single delivery are termed a "b i l l load." Diet sampling I collected fish from rhinoceros auklet adults provisioning their chicks at Triangle Island, British Columbia. Sampling sessions (10 bi l l loads per session) occurred approximately every 10 days, starting when chicks were approximately 10 days old (Table 3.1). A l l sessions occurred in the same vicinity on the colony, and commenced at around 22h30. Table 3.1. Sampling dates and number of bi l l loads sampled from rhinoceros auklets provisioning chicks at Triangle Island, B C . Date Number of Date Number of (Julian Date) Full Bill (Julian Date) Full Bill loads loads 2003 June 30 (181) 7 2004 June 20 (172) 7 July 12 (193) 9 June 30(182) 9 July 20 (200) 10 July 10 (192) 7 July 31 (221) 10 July 20 (202) 8 August 9 (220) 9 July 30 ( 212) 9 August 16 (227) 8 A u g 9 (222) 9 T O T A L 53 T O T A L 49 41 Adults en route to provisioning their chicks were caught with small fishing nets and all contents of their bi l l load retrieved. Only bi l l loads for which all fish were retrieved were used in diet analyses. Fish were daubed dry of excess water, weighed on an electronic mass balance (± O.lg), and measured for fork and standard length ( ± l m m ) . Rockfish were classified to genus due to difficulty in determining species. Rockfish previously identified as prey at Triangle Island include yellowtail (Sebastes flavidus) and widow (Sebastes entomelas) (Vermeer and Westrheim 1984). Ch ick meal energy content Energy content of fish in chick meals was calculated from species-specific regression equations (Chapter 2 of this thesis), or from published energy densities when local information was not available (Table 3.2). Biomass and energetic values of all fish in a chick meal (bill load) were then summed to find total meal biomass and total meal energy content. Analyses To test for the influence of chick age and year on chick meal biomass and energy content, I fit a quadratic model of chick age and year using a backwards regression in R statistical package (R Development Core Team 2006). Energy was modeled with Poisson errors as they followed a Poisson distribution, while mass was modeled with normal errors as they followed a normal distribution. I used a Mann-Whitney U-test (Zar 1998) to determine i f diet quality (kJ/g) differed between years as data could not be transformed to meet the assumptions of parametric tests. 42 Table 3.2. Calculation of total energy (in kJ) in forage fish delivered to rhinoceros auklet nestlings on Triangle Island, British Columbia. Species Year Equation for Total Energy (kJ) Pacific Saury (j) 2003 2004 y y = (0.361 • wt + 2.05-10~2 • (JD) - 1.239)2 = (0.279 • wt + 4.344) 2 Pacific Sand lance (a) 2003 y= = (0.308 • wt + 3.705)2 2004 y : = (0.291-wt + 4.170)2 Pacific Sand lance (j) 2004 y =(0.879 • wt + 5.03 • IO"3 • (JD) + 0.316) Rockfish (j) 2004 y= =(0.666 • wt + 8.75 • IO"2 • (JD) + 0.246) 2 Salmon spp. (j) 2003 2004 y= =(0.198 • wt + 4.087) 2 Sablefish (j) 2003 y= =2.64- wt (Van Pelt et al. 1997) Herring (a) 2004 y= =5.80-wt (Anthony et al. 2000) Herring (j) 2004 y= =3.69- wt (Anthony et al. 2000) Where: wt = wet mass in g, JD = Julian Date Productivi ty and C h i c k Growth I monitored all rhinoceros auklet burrows located within an established (approx. 50m x 30m) quadrat containing on Triangle Island, British Columbia during 2003 and 2004. Approximately 145 burrows were monitored in each year, though only 46 were actually used by rhinoceros auklets in 2003 and 55 were used in 2004. From late Apr i l through mid-July, burrows were searched every 5 days for incubating or prospecting adults. Once an incubating adult was found, the burrow was marked as occupied and left undisturbed for 40 days, at which point we returned every 5 days to check for hatching. I marked hatch date as the Julian date (numerical day of year) on which chicks hatched. Newly hatched chicks were aged by flattened wing cord length and hatch dates back-calculated based on an established index of known age wing cord (Triangle Island Research Station, unpubl. data). Chick mass, flattened wing cord and tarsal length were all measured on the day of first finding the chick, at 10, 40, and 45 days of age, and every 43 second day following until fledging. Chick mass was measured using Pesola © weigh-scales (± 2g), wing cord with a wing cord ruler (± 0.1mm), and tarsus with calipers (± 0.1 mm). I recorded as dead all chicks that disappeared prior to 45 days of age, or that were found dead within the burrow. I used the age at which the chick was last observed in the burrow as the age at fledging and measurements taken on that day as those at fledging. Rhinoceros auklet chicks show linear growth rates between approximately 10 and 40 days of age (Harfenist 1995). Linear daily growth rates were therefore calculated by dividing by 30 the total mass gained between 10 and 40 days of age (Harfenist 1995). Analyses I used t-tests for 2 independent samples to test for differences between years for all continuous variables except hatch date, using SPSS (SPSS 1999). Data on mass at fledging and wing cord at fledging were both squared to achieve normality. Age at fledging was log-transformed for the same reason. Hatch date data were analyzed with non-parametric Mann-Whitney U-test (Zar 1998) as data could not be transformed to meet t-test assumptions. A l l other data met t-test assumptions without transformation. I tested for interannual differences in hatching and fledging rates using %2 goodness of fit as all %2 assumptions were met. Controlling for differences in parental quality 44 Interannual differences in hatching success can confound the interpretation of interannual differences in population level measures of chick growth and survival (Williams and Croxall 1990, Pyle et al. 2001), and this may have occurred in this study. In 2004, rhinoceros auklets had the highest hatching success on record at Triangle Island, whereas 2003 was among the lowest on record (Triangle Island Research Station, unpubl. data). Thus, chick measurements in 2003 and 2004 are likely drawn from populations that are quite different. 2003 chick measurements represent only top quality parents as all but the best failed early on. In contrast, chick measurements from 2004 wi l l be affected by a much broader range of parental quality, as all but the most incompetent of breeders were able to hatch chicks. The chicks of high quality parents generally receive more food (DeForest and Gaston 1996) and grow faster (Hipfner 1997, Hipfner and Gaston 2002-2003) than those of lower quality parents. The inclusion of low quality parents thus has the potential to downwardly bias estimates of both food delivery and chick fledging parameters. In an attempt to control for variation in parental quality between years, I isolated measurements taken from chicks hatching from eggs laid prior to the median laying date in 2004 (JD =119) and compared them to those of parents laying prior to the median lay date in 2003 (JD =124). Positive relationships between parental quality and early timing of reproduction have been reported in many avian species (DeForest and Gaston 1996, Hipfner 1997, Arnold et al. 2004) and thus prior-to-median chicks should belong to higher quality parents. I used 2-sample t-tests to test for differences as data met all test assumptions. 45 RESULTS Diet Composition Eight types of prey were collected over the course of two seasons (Table 3.3). In 2003, diets were dominated by pacific sand lance and juvenile salmon during the early season, and pacific saury mid-season onwards (Figure 3.1a). In contrast,' early season 2004 diets were composed primarily of juvenile rockfish and adult sand lance, while juvenile sand lance and salmon dominated the latter part of the season (Figure 3.1b). Table 3.3. Sizes and masses of forage fish species sampled from Rhinoceros Auklet nestling diets at Triangle Island, British Columbia, (mean ± sd) Species Age Year N Fork Length (mm) Mean ± S.D. Range Mass (g) Mean + S.D. Range Pacific saury Juvenile 2003 2004 96 19 133.83 ±25.53 144.63 ± 39.60 59-189 79-206 8.43 ± 4.52 12.21 ±8.68 0.4-20.3 1.6-29.5 Pacific sandlance Adult Juvenile 2003 2004 2003 2004 42 39 95 115 145.83 ± 18.99 146.26 ±20.14 71.30 ±9.48 77.16 ±6.46 113- 199 114- 190 51-90 55-92 10.81 ±4.77 11.19 ± 4.85 1.05 ±0.48 1.49 ±0.43 4.7-26.9 5.6-26.1 0.2-2.2 0.6 - 2.7 Rockfish spp. Juvenile 2003 2004 39 81 48.90 ± 6.42 62.33 ±6.51 36-58 30-70 1.09 ±0.38 2.57 ± 0.67 0.30-1.7 0.6-3.7 Salmon spp. Juvenile 2003 2004 15 16 114.27±17.27 124.71 ±28.17 69-140 72-154 14.26 ±6.16 19.19 ± 10.15 3.2-25.3 2.7-38.0 Pacific herring Pacific herring Adult Juvenile 2004 2004 2 15 154* 68.50 ±8.59 154* 55 -79 29.40 ± 9.05 1.88 ± 1.01 23.0-35.8 0.4-4.0 Sablefish Juvenile 2003 5 98.40 ± 11.80 90-118 6.58 ±2.65 4.6-11.0 Market squid ? 2003 1 n/a n/a 5.0 n/a Slender Juvenile 2003 5 122 ± 19.04 102 - 145 . 1.64 ±0.83 0.8-2.6 Barracudina 2004 1 97 Na 0.6 Na :only one fish could be measured for length Nestling meal mass and energy content Energy content varied non-linearly with date and, once we controlled for this covariate, energy content was higher in 2004 than in 2003, albeit marginally (Table 3.4, 46 Figure 3.2a). Total meal biomass was also higher in 2004 than in 2003 and varied non-linearly with date (Table 3.4, Figure 3.2b). Chick diet quality (kJ/g) did not vary between years (2003: 4.76 ± 0.16, 2004: 4.74 ±0.30, z =-0.03, p=0.976). Productivity Rhinoceros auklet productivity was substantially lower in 2003 than in 2004 (Table 3.5). This difference was driven by low hatching success and moderate hatchling survival in 2003. Fledglings in 2003 had shorter tarsi and hatched an average of 8 days later than those in 2004 but did otherwise not differ from 2004 fledglings. Chicks of early laying parents in 2004 only differed in tarsus length when compared to those of early laying parents in 2003 (Table 3.6). 47 a) c IE 5 = o C L 100% 8 0 % 6 0 % 4 0 % 2 0 % 0 % b) OS ed o o 5^ c u GJ I I 1 1 30 Jun (10) 12 July (22) 20 July 31 July 9 Aug 16 Aug (30) (41) (50) (57) • Pacific Sand lance 0 Rockfish m Other • Pacific Sand lance (j) • Salmon • Pacific Saury • Herring 100% 80% 60% -40% 20% 0% Mm nn 20 Jun 30 Jun 10 July 20 July 30 July 9 Aug (10) (20) (30) (40) (50) (60) Figure 3.1. Composition of rhinoceros auklet nestling diet by date (days after mean hatch date) in (a) 2003 (b) 2004 48 Table 3.4. Effects of year and chick age on energy delivered to rhinoceros auklet nestlings at Triangle Island, British Columbia. Chick meal energy content Chick meal mass F(dfl, df2) P F(dfl, df2) P Chick age 8.17 (1,99) 0.005 6.89 (1,99) 0.010 Chick age 2 8.57 (1,99) 0.004 7.86 (1,99) 0.006 Year 3.86 (1,99) 0.052 4.00(1,99) 0.048 Age-year interactions 0.991 (2,98) 0.375 0.151 (2,98) 0.860 Model R z 0.120 0.100 2003 2004 2003 2004 M e a n ± 9 5 % C I 124.69 ± 11.19 139.06 ±14.66 26.54 ± 2.34 29.53 ± 2.66 n= 63 n= 49 n= 63 n= 49 49 a) 250 -i c o S3 o u o o £ u ss > < 200 150 H 100 50 0 • 2003 B 2004 10/10 22/20 30/30 41/40 50/50 57/60 Days after mean nestling hatch date (2003/2004) b) ca W U < 45 40 35 30 25 20 15 10 5 0 • 2003 E3 2004 10/10 22/20 30/30 41/40 50/50 57/60 Days after mean nestling hatch date (2003/2004) Figure 3.2. Energy content (a) and mass (b) of rhinoceros auklet chick meals as a function of average chick age in days. Error bars represent 95% confidence intervals. Both energy and mass delivered to chicks varied within and between years. 50 Table 3.5. Growth and survival statistics for all rhinoceros auklet chicks measured at Triangle Island, British Columbia. 2003 2004 n Estimate ± 95% CI (min-max) n Estimate ±95% CI (min-max) P Power Production (g of chick/pair) 99.2 198.25 Number eggs laid 46 55 % eggs hatching * 19 0.41 40 0.73 <0.001 1.00 % chicks fledging 16 0.84 38 0.95 0.06 0.196 % eggs laid fledging * 16 0.35 38 0.69 <0.001 1.00 Hatch Date (JD)* 19 172.21 ± 4 . 1 39 164.36 ± 2 . 5 3 0.010 0.926 Daily growth rate (mass, linear phase) 13 4.46 ± 0 . 5 1 31 4.83 ± 0.42 0.294 0.219 Mass at fledging (g) 14 284.50 ± 19.51 33 287.29 ± 17.13 0.842 0.052 Wing cord at fledging (mm) 14 154.14 ± 2 . 5 33 154.88 ± 2 . 4 1 0.711 0.096 Age at fledging 14 52.86 ± 2 . 0 5 33 54.97 ± 1.35 0.080 0.525 Tarsus length at fledging (mm)* 14 29.44 ± 0.42 34 30.72 ± 0.42 0.01 0.985 Table 3.6. Growth and survival statistics for rhinoceros auklet chicks from eggs laid on or prior to the annual median lay date of rhinoceros auklets at Triangle Island, British Columbia Variables marked with an asterisk were significantly different (a = 0.05). 2003 2004 n Estimate ±95% CI (min-max) n Estimate ±95% CI (min-max) P Power Production (g of chick/egg laid) 122.92 258.07 Number eggs laid 24 30 % eggs hatching * 11 0.45 28 0.93 <0.001 1.00 % chicks fledging 10 0.91 26 0.93 0.8092 0.89 % eggs laid fledging * 10 0.42 26 0.87 <0.001 1.00 Hatch Date (JD) 11 166.36 ± 2.02 28 160.48 ± 1.59 Daily growth rate (mass, linear phase) 9 4.60 ± 0.57 21 5.23 ± 0.47 0.102 0.54 Mass at fledging (g) 10 292.60 ± 24.62 24 296.63 ± 17.42 0.786 0.09 Wing cord at fledging (mm) 10 154.70 ± 3.04 24 156.58 ± 2 . 5 8 0.383 0.25 Age at fledging 10 53.60 ± 2 . 6 2 24 53.25 ± 1 . 6 1 0.806 0.08 Tarsus length at fledging (mm)* 10 29.59 ± 0.34 23 30.98 ± 0.42 <0.001 1.00 D I S C U S S I O N I examined whether chick diets of significantly different composition led to differences in prey energy or biomass delivered to rhinoceros auklet nestlings at Triangle Island, British Columbia. Nestlings were fed different diets between years (Figure 3.1) and both the energy and biomass of meals delivered to nestlings were lower in 2003 than in 2004 (Table 3.4). However, as the energy density of nestling meals did not differ between years, differences in meal biomass likely accounted for the bulk of interannual differences in chick meal energy. This difference suggests chicks were fed less in 2003, possibly indicating that prey availability was lower in 2003 than in 2004. Concurrent studies on stress hormones in chick rearing rhinoceros auklets at Triangle showed that adult rhinoceros auklets had higher peak corticosterone responses, indicative of long-term food shortages, in 2003 than in 2004 (B. Addison unpubl. data). Parents were working harder to provision chicks and themselves in 2003, yet still unable to provision chicks at the same levels observed in 2004. While energy differed between years, seasonal trends in energy and mass delivery were consistent between years. Total energy and total mass of b i l l loads each showed a parabolic relationship with sampling time, in both 2003 and 2004. The mid-season peak during both 2003 and 2004 (Figure 3.2) may reflect changing chick energetic demands (Koelink 1972, Kitaysky 1999), a seasonal decline in prey availability (Birt et al. 1997) and/or a decline in parental investment over the course of the season (Deguchi et al. 2004). Overall, the difference in energy delivery between years was due to consistently lower energy delivery over the course of 2003. 5 3 Compared to 2004, 2003 nestlings received an average 10% less energy over the course of the breeding season. However, the amount of food and energy being delivered to chicks did not differ enough to elicit statistically-significant population-level changes in chick growth, nor fledging age, mass, or wing cord length (Table 3.5). For most morphometric variables, the biological significance of differences presented in Table 3.5 is likely small, regardless of statistical differences, as average values between years were close. Growth rates and mass at fledging were particularly similar, given the broad range of observed values for rhinoceros auklet chicks at Triangle Island (average linear growth rate: 3.01 - 7.00 g/day, average annual fledging mass: 224-341g, Hedd et al. 2006). Differences in parental quality did not seem to explain a lack of difference in chick growth characteristics between the two years (Table 3.6). However, it is possible that the analysis used simply failed to isolate birds of higher quality, as it is unlikely that no differences in parental quality arose from such a large difference in hatching success (Table 3.5). Although high quality birds tend to lay earlier in the season (DeForest and Gaston 1996, Hipfner 1997), there is variation around this trend and it is possible that the sample size was not large enough to capture the trend within the variation observed here. This idea is supported by the low power of the analyses (Table 3.6) and the presence of a low outlier in 2004, the removal of which created a significant difference in chick growth rates between 2003 and 2004. While the removal of the outlier is not biologically justified, it does suggest that interannual comparisons in chick growth should be explored with larger sample sizes. Tracking both the diet and growth of chicks whose parents are 54 of known quality would clarify the potential role of parental quality in mediating population-level responses to changing prey conditions. Tarsus length was the only structural measure that did differ between birds in 2003 and those in 2004. Chick tarsus length, which in some species reflects overall body size and in waterfowl is correlated with post-fledging survival (Cooch et al. 1991, Van der Jeugd and Larsson 1998) was shorter in 2003 fledglings. Positive relationships between energy intake and tarsus length have been observed in rhinoceros auklets (Takenaka et al. 2005) and Atlantic, Tufted, and Horned puffins (Ovan and Anker-Nilssen 1996, Kitaysky 1996), though only under differences in diet that exceed those observed in this study. Relationships between tarsus length and general body size or survival are not well documented for rhinoceros auklets and it is therefore difficult to estimate what a 0.5 mm difference in tarsus length would imply for chick survival. In Japan, tarsus length differed between rhinoceros auklet chicks fed high and low energy diets but overall structural size did not (Takenaka et al. 2005), suggesting that tarsus length and body size are not necessarily related in rhinoceros auklets fledglings. In adult seabirds, the relationship between tarsus length and survival varies. Tarsus length did not seem to affect survival in crested auklets (Jones et al. 2004) while stabilizing selection on tarsus length was observed in adult male snow petrels (Barbraud 2000). It is difficult to tell whether the growth responses to variation in energy delivery observed here are normal, as most experimental studies manipulating energy delivery did so to a much great extent than the 10% reported here (e.g. Harfenist 1995, Takekana et al. 55 2005). Under large manipulations (-25-50% differences in energy delivered), rhinoceros auklet chicks fed higher energy diets generally grew at faster rates and reached higher mass at fledging than those fed lower energy diets (Harfenist 1995, Takenaka et al. 2005). This growth effect has also been observed in tufted puffins, a close relative of rhinoceros auklets (Romano et al. 2000), though also at large manipulations of energy consumption. Atlantic puffins also showed reductions in chick mass and mass-based growth rates with only a 5% reduction in meal energy content throughout the nestling period (Oyan and Anker-Nilssen 1996). The lack of difference in growth rates between years may also have been partly driven by factors other than energy intake. While energy delivery to seabird nestlings is the most commonly reported factor influencing chick growth, chick growth is also affected by weather-related thermoregulation costs (Weathers 1996), parasite loads (Duffy 1983, Morbey 1996), and micronutrient availability (Murphy 1996). It is unlikely that weather altered growth rates differently in 2003 than in 2004, as air temperature at nearby Cape Scott was similar between years (Environment Canada 2004), and burrows are well- insulated by the surrounding soil (Triangle Island Research Station unpubl. data). However, differing rates of parasitism is a more likely confounding factor. Rhinoceros auklet chicks at Triangle are sometimes parasitized by the common avian tick Ixodes uria. This species of tick has led to decreased nestling growth rates in other seabird species (Morbey 1996, Ramos et al. 2001). A s the intensity of tick infestations can vary between years (Duffy 1983) and by location within the colony 56 (Morbey 1996), it is possible that an increase in parasite load during 2004 could have led to a decline in chick growth rates. Chick growth rates may also have been affected by differing amounts of micronutrients available in the different diets observed in 2003 and 2004. It is difficult to evaluate the likelihood of this effect as the micronutrient profiles of most forage fish species are unknown (Murphy 1996), as are the micronutrient requirements of seabird nestlings. While chick growth rates and fledging characteristics did not seem to be very affected by variation in prey availability or quality, total productivity (measured as grams of chick per egg laid) was affected (Table 3.5). The large difference in hatching success was the primary factor driving interannual differences in rhinoceros auklet productivity, which suggests that the effects of prey availability were strongest prior to chick rearing. The majority of seabird prey studies focus on response variables like fledging success and chick growth rather than on factors occurring prior to chick rearing. The results of this study indicate that chick growth may be of relatively little influence in the context of all stages contributing to rhinoceros auklet productivity. In conclusion, interspecific differences in prey quality (energy density) were not substantial enough to exert an influence on population-measures of chick growth or fledging success in the two years of this study. Instead, interannual differences in energy delivery to nestlings were driven primarily by biomass, which suggests fluctuating prey 57 availability as an underlying mechanism. The large difference in hatching success between the two years also suggests that prey availability may have been stronger in influencing hatching success than chick growth. Although prey quality did not have a strong effect in this study, it should not be discounted as an important mechanism at Triangle Island. The availability of only two years of prey quality data, and measuring variables at the population level makes it difficult to discern the true role of prey quality in rhinoceros auklet reproductive success. Using repeated and concurrent measures of both diet and growth for individuals would overcome many of the power issues caused by the large variation associated with using population level diet information. Obtaining prey quality information over a broader range of oceanic conditions, and testing effects of diet over more than two years would also provide greater understanding of how diet influences reproductive success in rhinoceros auklets at Triangle Island. 58 R E F E R E N C E S Abookire, A . A . , and J. F . Piatt. 2005. Oceanographic conditions structure forage fish into lipid-rich and lipid-poor communities in lower Cook Inlet, Alaska. Marine Ecology Progress Series 287: 229-240 Anthony, J .A. , D . D . Roby and K . R . Turco. 2000. L ip id content and energy density of forage fishes from the northern Gul f of Alaska. Journal of Experimental Marine Biology and Ecology 248: 53-78 Arnold, J . M . , J.J. Hatch, and I.C.T. Nisbet. 2004. Seasonal declines in reproductive success of the common tern Sterna hirundo: timing or parental quality? Journal of Avian Biology 35: 33-45 Barbraud, C. 2000. Natural selection on body size traits in a long-lived bird, the snow petrel Pagodroma nivea. Journal of Evolutionary Biology 13: 81-88 Bertram, D.F . , Kaiser, G .W. , and R . C . Ydenberg. 1991. Patterns in the provisioning and growth of nestling rhinoceros auklets. A u k l 0 8 : 842-852 Bertram, D.F . , D . L . Mackas, and S . M McKinne l l . 2001. The seasonal cycle revisited: interannual variation and ecosystem consequences. Progress in Oceanography 49: 283-307 Bertram, D . F . and G . W . Kaiser. 1993. Rhinoceros auklet (Cerorhinca monocerata) nestling diet may gauge Pacific sand lance (Ammodytes hexapterus) recruitment. Canadian Journal of Fisheries and Aquatic Science. 50: 1908- 1915 Birt, V . L . , T. P. Birt, D . Goulet, D . K . Cairns, and W . A . Montevecchi. 1987. Ashmole's halo: direct evidence for prey depletion by a seabird. Marine Ecology Progress Series 40: 205-208 Cooch, E . G . , D . B . Lank, R.F. Rockwell , and F. Cooke. 1991. Long-term decline in body size in a snow goose population: evidence of environmental degradation. Journal of Animal Ecology 60: 482-496 Cury, P., A . Bakun, R . J . M . Crawford, A . Jarre, R . A . Quinones, L . J . Shannon, and H.M.Verheye. 2000. Small pelagics in upwelling systems: patterns of interaction and structural changes in "wasp-waist" ecosystems. ICES Journal of Marine Science 57: 603-618 Dahdul, W . M . and M . H . Horn. 2003. Energy allocation and postnatal growthin captive elegant tern (Sterna elegans) chicks: responses to high- versus low-energy diets. Auk 120: 1069-1081 59 DeForest, L . N . and A . J . Gaston. 1996. The effect of age on timing of breeding and reproductive success in the thick-billed Murre. Ecology 77: 1501-1511 Deguchi, T., Takahashi, A . , and Y . Watanuki. 2004. Proximate factors determining age and mass at fledging in Rhinoceros Auklets (Cerrohinca moncerata): intra-and interyear variations. Auk 121: 452-462 Diamond, A . W. , and C M . Devlin. 2003. Seabirds as indicators of changes in marine ecosystems: ecological monitoring on Machias Seal Island. Environmental Monitoring and Assessment 88: 153-175 Duffy, D . C 1983. The ecology of tick parasitism on densely nesting Peruvian seabirds. Ecology. 64: 110-119 Environment Canada. 2004. Monthly data report for 2003, 2004: Cape Scott. http://www.climate.weatheroffice.ec.gc.ca/climateData/monthlydata_e.html Gaston, A . J., and S. B . C. Dechesne. 1996. Rhinoceros Auklet (Cerorhinca monocerata). In The Birds of North America, No. 212 (A. Poole and F. G i l l , Eds.). The Academy of Natural Sciences, Philadelphia, P A , and The American Ornithologists' Union, Washington, D . C . Golet, G . , K . J . Kuletz, D . D . Roby, and D . B . Irons. 2000. Adult prey choice affects chick growth and reproductive success in Pigeon Guillemots. Auk 117: 8 2 - 91 Harfenist, A . 1995. Effects of growth-rate variation on fledging of rhinoceros auklets, Cerorhinca monocerata. Auk 112: 60-66 Hedd, A . , D .F . Bertram, J .L. Ryder, and L L Jones. 2006. Effects of inter-decadal climate variability on marine trophic interactions: rhinoceros auklets and their fish prey at Triangle Island, British Columbia, Canada, 1976-2001. Marine Ecology Progress Series.309: 263-278 Hipfner, J . M . 1997. The effectos of parental quality and timing of breeding on the growth or nestling thick-billed murres. Condor 99: 353-360 Hipfner, J . M . , and A . J . Gaston. 2002-2003. Growth of Thick-Bi l led Murre (Uria lomvia) chicks in relations to parental experience and hatching date. Auk 119: 827- 832 and 120: 235 Jones, I.L., F . M . Hunter, G.J.Robertson, and G . Fraser. 2004. Natural variation in the sexually selected feather ornaments of crested auklets (Aethia cristatella) does not predict future survival. Behavioural Ecology 15: 332-337 60 Kitaysky, A . S . , 1999. Metabolic and developmental responses of alcid chicks to experimental variation in food intake. Physiological and Biochemical Zoology. 72: 462^173 Kitaysky, A . S . , E . V . Kitaiskaia, J.F. Piatt, and J.C. Wingfield. 2006. A mechanistic link between chick diets and decline in seabirds? Proceedings of the Royal Society of London: Biological Sciences 273: 445-450 Litzow, M . A . , J.F. Piatt, A . A . Abookire, A . K . Prichard, and M . D . Robards. 2000. Monitoring temporal and spatial variability in sandeel (Ammodytes hexapterus) abundance with pigeon guillemot (Cepphus columba) diets. ICES Journal of Marine Science 57: 976-986 Litzow, M . A . , J.F. Piatt, A . K . Prichard, D . D . Roby. 2002. Response of pigeon guillemots to variable abundance of high l ipid and low l ipid prey. Oecologia 132: 286-295 Murphy, M . E . 1996. Nutrition and metabolism, pp. 31-60 in Avian Energetics and Nutritional Ecology (C. Carey, Ed.). Chapman and Hal l , New York. Niizuma, Y . , and Yamamura, O. 2004. Assimilation efficiency of rhinoceros auklet, Cerorhinca monocerata, chicks fed Japanese anchovy, Engraulis japonicus, and Japanese sand lance, Ammodytespersonatus. Comparative Biochemistry and Physiology A . 139: 97-101 Oyan, H.S. , and T. Anker-Nilssen. 1996. Allocation of growth in food-stressed Atlantic puffin chicks. Auk 113: 803-841 Phillips, R. A . and K . C . Hamer. 1999. L i p i d reserves, fasting capability and the evolution of nestling obesity in procellariiform seabirds. Proceedings of the Royal Society of London: Biological Sciences 266: 1329-1334 Pyle, P., W.J . Sydeman, and M.Hester. 2001. Effects of age, breeding experience, mate fidelity and iste fidelity on breeding performance in a declining population of Cassin's auklets. Journal of Animal Ecology 10: 1088-1097 R Development Core Team. 2006. R: A language and Environment for Statistical Computing. R Foundation for Statistical Computing. Vienna. Ramos, J .A. , J .Bowler, L . Davis, S. Venis, J. Quinn, and C. Middleton. 2001. Activity patterns and effect of ticks on growth and survival of tropical Roseate Tern nestlings. Auk 118: 709-716 Reid, K . , P . A . Prince, and J.P. Croxall . 2000. F ly or die: the role of fat stores in the growth and development of grey-headed albatross Diomedea chrysostoma chicks. Ibis 142: 188-198 61 Ricklefs, R. E . 1983. Avian postnatal development, pp. 1-183 in Avian Biology. (D. S. Farner, J. R. K ing , and K . C. Parkes, Eds.). New York: Academic Press. Rindorf, A . , S. Wanless, and M . P . Harris. 2000. Effects of sandeel availability on the reproductive output of seabirds. Marine Ecology Progress Series 202: 241-252 Robards, M . D . , J.F. Piatt, A . B . Kettle, and A . A . Abookire. 1999a. Temporal and geographic variation in fish communities of lower Cook Inlet, Alaska. Fishery Bulletin 97: 962-977 Robards, M . D . , J .A. Anthony, G . A . Rose and J.F. Piatt. 1999b. Changes in proximate composition and somatic energy content for Pacific sand lance (Ammodytes hexapterus) from Kachemak Bay, Alaska relative to maturity and season. Journal of Experimental Marine Biology and Ecology 242: 245-258 Romano, M . D . , D . D . Roby, J.F. Piatt, and A . Kitaysky. 2000. Effects of diet on growth and development of nestling seabirds. Exxon Valdez O i l Spi l l Restoration Project Final Report (Restoration Project 98163N), U S G S - Oregon Cooperative Fish and Wildlife Research Unit, Oregon State University, Corvallis, Oregon SPSS 1999. SPSS for Windows Standard Release 10.0.5 (27 Nov 1999) Takahashi, A . , M . Kuroki , Y . Niizuma, A . Kato, S. Saitoh,, and Y . Watanuki. 2001. Importance of the Japanese anchovy (Engraulis japonicus) to breeding rhinoceros auklets {Cerorhinca monocerata) on Teuri Island, Sea of Japan. Marine Biology 139:361-371 Takenaka, M , Y . Niizuma, and Y . Watanuki. 2005. Resource allocation in fledglings of the rhinoceros auklet under different feeding conditions: an experiment manipulating meal size and frequency. Canadian Journal of Zoology 83: 1476-1485 Van der Jeugd, H.P. and K . Larsson. 1998. Pre-breeding survival of barnacle geese Branta leucopsis in relation to fledgling characteristics. Journal of Animal Ecology 67: 953-966 Van Pelt, T .L , J.F. Piatt, B . K . Lance and D . D . Roby. 1997. Proximate composition and energy density of some North Pacific forage fishes. Comparative Biochemistry and Physiology A . 118: 1393-1398 Vermeer, K . , 1979. Nesting requirements, food and breeding distribution of Rhinoceros Auklets, Cerorhinca monocerata, and Tufted Puffins, Lunda cirrhata. Ardea 67: 101-110 Vermeer, K . , 1980. The importance of timing and type of prey to reproductive success of Rhinoceros Auklets Cerorhinca monocerata. Ibis 122: 343- 350 62 Vermeer, K . and S. J. Westrheim. 1984. Fish changes in diets of nestling rhinoceros auklets and their implications, pp. 96-105 in Marine Birds: Their feeding ecology and commercial fisheries interactions. (D. N . Nettleship, G . A . Sanger, and P. F. Springer, Eds.). Special Publication, Canadian Wildl i fe Service. Ottawa. Visser, G . H . , 2002. Chick growth and development in seabirds. Pp. 439-465 in The Biology of Marine Birds. (E .A. Schreiber and J. Burger, Eds.). C R C Press, Boca Raton. Wanless S, M . P . Harris, P. Redman and J.R. Speakman. 2005. L o w energy values of fish as a probable cause of a major seabird breeding failure in the North Sea Marine Ecology Progress Series 294: 1-8 Weathers, W . W . 1996. Energetics of postnatal growth. Pp 461-496 in Avian energetics and nutritional ecology. (C. Carey, Ed.) Chapman and Hal l , New York. Whittow, G . C . 2002. Seabird Reproductive Physiology and Energetics pp. 409- 438 in The Biology of Marine Birds. (E.A.Schreiber and J. Burger, Eds.) C R C Press, Boca Raton. Ydenberg, R . C . 1989. Growth-mortality trade-offs and the evolution of juvenile life histories in the Alcidae. Ecology 70: 1494-1506 Zar, J .H. 1998. Biostatistical Analysis 4 t h Ed . Prentice Hal l . 63 Chapter 4: Conclusions In this thesis I have explored whether or not differences in prey quality may underlie the observed relationships between diet composition and rhinoceros auklet reproductive success at Triangle Island, British Columbia. Within this I examined how prey quality might change between years, touched on possible causes of quality changes within species, and evaluated whether or not meal biomass is a suitable proxy for total meal energy density. I then used models of prey energy content to reconstruct the energy delivered to chicks in two years of contrasting diets and productivity to evaluate whether or not differences in prey quality correlated with differences in chick growth and productivity. Prey energy content varied both within and between prey types, and caused the relative quality of different prey types to change over time. The energy density values reported in chapter one may therefore not be applicable to other years of study. It also suggests that relative meal biomass is not an accurate proxy for relative meal energy content. However, the species and year specific regression equations based on simple measures of dry biomass offer an alternative means of estimation. Simply dessicating each prey item and taking note of what species it is wi l l offer a much better idea of relative energy content. That said it is important that the regression coefficients be recalibrated in years when oceanographic conditions vary considerably from those observed in the years of this study. 64 Although different prey types did vary in quality, the effects of inter and intraspecific variation did not result in any interannual differences in total diet quality between the two years of this study. The two diets observed each contained species of varying quality, but rhinoceros auklets used the different species in proportions that resulted in the same average diet quality between years. A s a result, it was possible to discern that the lower amount of energy delivered to chicks in 2003 was a result of differences in the average mass of chick meals rather than differences in prey quality. . This observation suggests that absolute prey abundance may have differed between years and may have driven the interannual differences in reproductive success observed in 2003 and 2004. This is not to say that prey quality never accounts for differences in reproductive success at Triangle Island, the two years of this study likely only capture a very small window of this relationship. Longer term study is clearly required before a conclusive argument can be made. However, some preliminary conclusions can be drawn from these results: 1) Biomass alone is not likely an accurate estimator of relative meal energy content at Triangle island as prey species differ in quality 2) Species-specific dry biomass has the potential to be a good indicator of prey energy content 3) Interannual variation in prey quality may have little influence on rhinoceros auklet reproductive success in comparison to prey availability 65 Appendix A. Lipid extraction method testing. To estimate l ipid content in forage fish, I used a modified Bl igh and Dyer (1959) l ipid extraction procedure developed by Higgs et al. (1979) and modified by Crossin (2003). The original Bl igh and Dyer method underestimates l ipid content when at concentrations higher than 2% l ipid (Iverson et al. 2001), l ikely due to a small solvent to tissue sample ratio (2:1). The modified technique I used increased the ratio to 20:1. To ensure that this modification solved the problem of underestimation associated with the Bl igh and Dyer technique, I compared the modified Bl igh and Dyer technique to that of Folch et al. (1957). This latter technique was highlighted by Iverson et al. (2001) as reliable. Methods I selected 13 age (2+) sand lance captured at Clayoquot Sound, British Columbia during July of 2004 with high visible fat stores. I also tested 2 pacific saury, 2 sand lance, and 1 chum salmon collected from rhinoceros auklets at Triangle Island. This approach was to provide a selection of l ipid levels across which to compare the two methods. Folch: l g of tissue was homogenized with 13.2ml of chloroform and 6.8ml of methanol for 2min30 seconds. The sample was then washed with 4.5ml of xmol salt solution for 30 seconds. The contents were vacuum filtered through a buchner funnel, the filtrate poured into a 50ml graduated cylinder, and left to stand for 1 hour while the solvent front formed. The top (methanol) layer was then suctioned off and a 5 ml aliquot of chloroform removed. The aliquot was placed in a dry, pre-weighed aluminum weigh 66 boat, heated to remove chloroform, and then dried for 1 hour at 100°C. L i p i d content was then determined gravimetrically. Bligh and Dyer, l g of tissue was homogenized with 5ml chloroform and 10ml methanol for 2 minutes, at which time an additional 5ml chloroform was added and homogenized for 30 seconds. The mixture was then washed with 4ml distilled water for 30 seconds. The same isolation and drying procedure as above was used to determine lipid content gravimetrically. Differences between paired samples were normally distributed, thus I analyzed the results using a paired-samples t-test in SPSS (SPSS Inc 1999). Results There was no difference between the results obtained by the two l ipid extraction procedures (mean difference ± s.d. = 0.117% ± 0.96%, ti8,i7=0.517, p>0.05, Figure A . l ) . Discussion Although there was some variation in the results between samples, it was not beyond that which was observed when a fish was sampled with duplicates of the same procedure. Most samples were within 1 percent of one another, the range of acceptable error for duplicates used in the main part of my study. The bulk of variation likely arose from incomplete tissue homogenization. 67 14.0 12.0' 10.0-8.0' lch 6.0' o 4.0-2.0' 0.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 Bligh and Dyer % Lipid Figure A . l . Comparison of results from two l ipid extraction methods estimating percent l ipid by wet mass for forage fish (pacific sand lance, pacific saury, and chum salmon). Line represents a 1:1 ratio. References Bl igh ,E .G. and W.J . Dyer. 1959. A rapid method for total l ipid extraction and purification. Canadian.Journal of Biochemical.Physiology. 37:911-917 Crossin, G.T. 2003. Effects of ocean climate and upriver migratory constraints on the bioenergetics, fecundity, and morphology of wild, Fraser River salmon. M . S c . thesis, University of British Columbia, Vancouver. Folch, J., M.Lees, and G . H . Sloane-Stanley. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. of Biological Chemistry226: 497-509 Higgs, D . A . , J.R. Markert, D . W . MacQuarrie, J.R. McBride , B .S . Dosanjh, C . Nichols and G.Hoskins. 1979. Development of practical dry diets for coho salmon, Oncorhynchus kisutch, using poultry-by-product meal, feather meal, soybean meal and rapeseed meal as major protein sources. In Finfish nutrition and fishfeed technology. V o l . 2. (J.E. Halver and K . Tiews Eds.). Heenemann Verlagsgesellschaft M b H , Berl in, Germany Iverson, S.J., S.C. Lang, and M . H . Cooper. 2001. Comparison of the B l igh and Dyer and Folch methods for total l ipid determination in a broad range of marine tissue. Lipids 36: 1283-1287 SPSS 1999. SPSS for Windows Standard Release 10.0.5 (27 Nov 1999) 6 8 Appendix B: Dehydration Correction Specimens were stored for approximately 1 year post-capture, thawed at least once during transport from the field, and handled to remove otoliths and visually assess lipd reserves prior to processing for energy density measurements. To test for water loss, specimens from 2004 were weighed at capture and then again just prior to analysis, adjusting for the mass of removed otoliths. I used standard linear regression to construct a dehydration correction factor for specimens that had not been re-weighed. As all specimens over 25 grams were re-weighed, the correction factor was constructed using only specimens weighing less than 25g. Results There was a significant linear relationship between fresh mass and frozen mass (R 2 = 0.993, p<0.0001m Figure B . l ) . 1) Frozen mass = 0.896 (fresh mass) - 0.233 69 30.0 ® 20.0 £ 10.0 0.0 Total Population Fresh Mass (g) Figure B.l. Relationship between fresh mass and frozen mass for species of forage fish <25g in mass. Regression equation is y=0.896x-0.233, r 2 = 0.993). Species explained 0.1% of variation (p<0.001) 70 Appendix C. Sizes and masses of forage fish species sampled from Rhinoceros Auklet diets at Triangle Island, British Columbia Species Age Year N Fork Length (mm) Mass (g) Mean ±S.D. Range Mean ± S.D. Range Pacific saury Juvenile 2003 16 147.5 ± 22.62 1 1 2 - 1 8 3 11.81 ± 5 . 3 2 4.4 - 22.62 Pacific saury Juvenile 2004 14 150.38 ± 2 9 . 8 5 104 - 201 13.94 ± 7 . 9 5 5 .0 -29 .3 Pacific sandlance Adult 2003 35 142.3 ± 22.2 105 - 207 13.98 ± 6 . 3 0 6 .6 -31 .3 Pacific sandlance Adult 2004 25 143.4 ± 2 2 . 9 107 - 190 11.87 ± 5 . 4 0 5 .6 -26 .0 Pacific sandlance Juvenile 2003 4 1 71.2 ± 9 . 6 51-90 1.04 ± 0 . 4 7 0.2 -2.2 Pacific sandlance Juvenile 2004 16 2 77.1 ± 6 . 4 55-92 1.47 ± 0 . 4 2 0 .6-2 .5 Rockfish spp. Juvenile 2004 16 2 62.4 ± 6 . 6 30-75 2.69 ± 0.45 0 . 6 - 4 . 4 Salmon spp. Juvenile 2004 12 131.22 ± 18.91 103-152 21.58 ± 8 . 6 2 10 .4-38 .0 1. Juveniles were pooled into groups of -15 2. Juveniles were pooled into groups of ~8g (4-6 fish) Juvenile means and ranges are those of all individual fish in pooled groups 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092678/manifest

Comment

Related Items