Spatiotemporal variability of adult Antarctic krill (Euphausia superba) lipids in relation to sea surface temperature and Chlorophyll a

Nicole Hellessey

Lipids are key biochemicals that form both cell membranes and energy stores. Lipids are of particular importance in energy poor environments where animals require stores to survive long periods of food shortage. In the Antarctic, food availability is dominated by extreme seasonal shifts in the environment and energy rich food is scarce for a substantial period of time each year. Antarctic krill (Euphausia superba) have adapted to have large lipid stores (over a third of their dry weight) during winter for this reason. Krill are a key species in the Antarctic environment; their biomass links lower and higher trophic levels and forms the main energy conduit for the system. Krill feed on diatoms, dinoflagellates and other algal species year-round, resulting in high omega-3 polyunsaturated fatty acids which are essential for krill health, growth and reproduction. Krill-derived omega-3 containing products (particularly eicosapentacnoic acid (20:5n-3) and docosahexaenoic acid (22:6n-3)) are sold as nutraceuticals for human consumption. Krill oil tablets (sold as an omega-3 supplement) are now the fastest growing nutraceutical globally. Understanding the krill life cycle is hampered by the restricted nature of scientific sampling. Knowledge of krill diet and krill lipid dynamics is lacking for the Indian and Pacific Ocean sectors, as most studies have focused on the South Atlantic Ocean sector where the krill fishery is based. Most scientific research voyages are conducted during summer months and all scientific studies are restricted in their spatial and temporal scale. Information on krill recruitment and reproduction in the Indian and Pacific Ocean sectors is also not as developed as in the South Atlantic, where fishery-derived samples are also available. A major gap in current ecosystem models is the link between environmental drivers (such as upwelling of nutrients, sea surface temperature and height, sea ice extent and thickness and salinity) and their impact on primary production and therefore food availability during extreme seasonal shifts in Antarctica. One way of measuring these environmental drivers is through remote-sensing via satellite, which can gather data over large geographic areas and over long timeframes. Satellite-derived data for biological and ecological measures is still developing as a tool for oceanographers and other end users. However, one area of growing importance is in the use of ocean colour data which can be converted into chlorophyll a concentrations (a proxy for primary production) via a standard algorithm. By linking the GPS locations of commercial krill harvesting, and therefore krill swarms, to environmental data obtained through remote-sensing from the same date, the relationship between the environment krill live in and their biochemical composition can be examined in ways not previously explored. My study used samples collected by a member of the krill fishery, Aker BioMarine, over a continuous three-year period in the South Atlantic Ocean to look at the seasonal and interannual trends in krill total lipids and lipid classes (such as those used for energy storage and the structure and function of cells). This dataset is unprecedented in its seasonal and spatial coverage in the South Atlantic Ocean. This study has been able to establish the sinusoidal shape of the seasonal and interannual trend in krill total lipids and its associated lipid classes. No samples from the fishery were available from other sectors. These South Atlantic Ocean krill samples were contrasted to krill samples collected from scientific expeditions in the other two ocean basins surrounding Antarctica (Pacific and Indian Oceans). Krill diet was investigated at a regional scale during the crucial late-summer spawning period. Results from my study revealed that krill diet varies between ocean basins, with Indian Ocean krill showing a distinctly different diet to Pacific and Atlantic Ocean krill, during the late-summer. The fishery-derived samples were also related back to environmental data collected via satellite, for both chlorophyll a and sea surface temperature, to investigate if environmental drivers influenced krill lipid biochemistry. This study showed that both sea surface temperature and chlorophyll a concentrations (derived from ocean colour data) can be related to krill lipid and fatty acid dynamics. Krill lipid composition and content were shown to be correlated to these environmental factors through simple models. The combination of results from this study will help fill the data gaps in ecosystem models and enable better determination of krill diet, recruitment and reproduction in all ocean basins surrounding Antarctica. These advances in krill knowledge will help improve fishery management policies.

1 2 3 4 5 __________________________________________________________________________________ Spatiotemporal variability of adult Antarctic krill (Euphausia superba) lipids in relation to sea surface temperature and Chlorophyll a 6 7 __________________________________________________________________________________ 8 by 9 Nicole Hellessey, BMarSci, GradCertRes, MAntSci 10 Institute for Marine and Antarctic Studies 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy (Biological Sciences) University of Tasmania November 2019 25 1 26 Declaration of originality 27 28 29 30 31 This thesis contains no material which has been accepted for a degree or diploma by the University or any other institution, except by way of background information and duly acknowledged in the thesis, and to the best of my knowledge no material previously published or written by another person except where acknowledgement is made in the text of the thesis, nor does the thesis contain any material that infringes copyright. 32 33 29th July 2019 34 35 Nicole Hellessey Date 36 37 38 39 40 41 42 43 Statement of authority of access 44 45 46 47 The publisher of the paper comprising Chapter 2 holds the copyright for that content. Access to the material should be sought from the respective journals. The remaining non-published content of the thesis may be made available for loan and limited copying in accordance with the Copyright Act 1968. 48 49 29th July 2019 50 51 Nicole Hellessey Date 52 53 54 2 55 Statement of co-author contributions 56 57 The following collaborators and institutions contributed to the publication of the work undertaken as part of this thesis: 58 59 60 61 62 63 64 65 Nicole Hellessey, Institute for Marine & Antarctic Studies - PhD Candidate Patti Virtue, Institute for Marine & Antarctic Studies - Primary Supervisor Peter D. Nichols, Commonwealth Scientific & Industrial Research Organisation - Cosupervisor So Kawaguchi, Australian Antarctic Division - Co-supervisor Stephen Nicol, Institute for Marine & Antarctic Studies - Co-supervisor Nils Hoem, Aker BioMarine - Co-supervisor Jessica Ericson, Institute for Marine & Antarctic Studies - PhD Candidate 66 Chapter 2 67 68 69 70 Paper 1: Hellessey N, Ericson JA, Nichols PD, Kawaguchi S, Nicol S, Hoem N, Virtue P (2018). Seasonal and interannual variations in the lipid content and composition of Euphausia superba Dana, 1850 (Euphausiacea) samples derived from the Scotia Sea krill fishery. Journal of Crustacean Biology, 1-9. doi: 10.1093/jcbiol/ruy053. 71 Author Contributions: 72 73 74 All authors contributed to project design. NH and JAE performed laboratory analysis. NH completed all data analysis, wrote the manuscript and attended to reviews. PDN, SK, SN, N Hoem and PV assisted in manuscript editing. 75 Chapter 3 76 77 78 Paper 2: Hellessey N, Ericson JA, Nichols PD, Kawaguchi S, Nicol S, Hoem N, Virtue P (2019). Regional variability of Antarctic krill (Euphausia superba) diet during the late-summer as determined using lipid, fatty acid and sterol composition. Polar Biology (under review) 79 Author Contributions: 80 81 82 All authors contributed to project design. NH and JAE performed laboratory analysis. NH completed all data analysis, wrote the manuscript and attended to reviews. PDN, SK, SN, N Hoem and PV assisted in manuscript editing. 83 Chapter 4 84 85 86 87 Paper 3: Hellessey N, Johnson, R, Ericson JA, Nichols PD, Kawaguchi S, Nicol S, Hoem N, Virtue P (2019). Antarctic Krill (Euphausia superba Dana 1850) Lipid and Fatty acid Content Variability is associated to Satellite Derived Chlorophyll a and Sea Surface Temperatures in the Scotia Sea. Scientific Reports (In Press) 88 Author Contributions: 89 90 91 All authors contributed to project design. NH and JAE performed laboratory analysis. RJ sourced all environmental data and assisted with analysis. NH completed all data analysis, wrote the manuscript and attended to reviews. RJ, PDN, SK, SN, N Hoem and PV assisted in manuscript editing. 3 92 93 94 We the undersigned agree with the above stated proportion of work undertaken for each of the above published (or submitted) peer-reviewed manuscripts contributing to this thesis: 95 ……………………………….. 96 97 98 99 100 101 102 Associate Professor Patti Virtue Primary Supervisor Institute for Marine and Antarctic Studies University of Tasmania Date: 2/8/2019 Professor Craig Johnson Head of School Institute for Marine and Antarctic Studies University of Tasmania Date: 103 4 104 Acknowledgements 105 I would like to start by thanking my wonderful supervisors and contributors of the aptly named 106 “Team Krill” - Patti Virtue, Peter D. Nichols, So Kawaguchi, Steve Nicol and Nils Hoem. The 107 resources made available to me through your joint efforts were astounding and I will forever 108 cherish having such a brilliant team standing behind me. You were all endlessly supportive, 109 with advice, ideas, hugs and criticism when needed. It was truly an honour to work with all of 110 you as you are each so knowledgeable and respected in the community. Patti and Peter, in 111 particular, went above and beyond in their supervisory roles and always made me feel like I 112 was part of something bigger than just my PhD, and that with their support anything was 113 possible. Thank you so much, I will never forget it. 114 To my lab partner for life, Jessica Ericson – you were my rock, lab partner, co-conspirator, 115 lunch buddy and all-purpose friend during this whole candidature. I’m sorry for all of my 116 horrendous jokes and my terrible singing in the lab. I still don’t understand how you haven’t 117 punched me in the face yet. We laughed, cried, learned, schemed, sung and swore together. I 118 can’t imagine having done this without you there to turn to on the good days and bad. You’re 119 my pick for MVP on “Team Krill” and I can’t wait to hear about all your successes in the future. 120 To Andy Revill, Peter Mansour, Mina Brock and Ben Gaskell, thank you for all of your 121 assistance and patience during my laboratory work at CSIRO. I know having Jess and I singing 122 as we worked in the labs was trying for everyone at times. To Robert Johnson, thank you for 123 being just as crazy as me when I came up with an idea for the last chapter of my thesis and 124 helping it come to life. You went above and beyond as a friend and mentor, I can’t thank you 125 enough. To Natasha Waller, Ashley Cooper, Blair Smith and Rob King at the Australian 126 Antarctic Division Krill Aquarium – thank you for letting me into your space for a whole year 127 while I assisted Jess with her experiment. Thank you to everyone at Aker BioMarine, 128 particularly the crew of the FV Saga Sea. I had the pleasure to meet some of you during my 129 candidature and it was astounding how many people in the industry were not just interested in, 130 but also supportive of our research. 131 Thank you to the many IMAS staff and students who assisted me, particularly to everyone 132 involved with BOTES and APECS Oceania. To the S.C.A in general and particularly to S.C.A 133 Tasmania – you are the family I choose for myself, you supported me behind the scenes and 134 helped keep me sane. To the Young Tassie Scientists led by Adele Wilson – I cannot tell you 135 what learning about Science Communication has done for me. You have opened up doors for 5 136 me, taught me about myself and what I’m passionate in, as well as given me friends, contacts 137 and a renewed love of Science. To all of my friends, both near and far - I love each and every 138 one of you and I can’t thank you enough for all of the kind words of encouragement I’ve 139 received over the past 3 and a half years. 140 To Mum, Dad, Chris and Michael – thank you for your unwavering love and support. Your 141 constant reassurance, encouragement, advice and support has meant that I could do this. 142 Lastly, thank you to Jacob. I’m so sorry Mum hasn’t been there for you like we both wanted 143 me to over the last few years but you're the reason I keep going and pushing on. I love you. 144 Dedication 145 This thesis is dedicated to my Grandmother, Elizabeth Gertrude Mary Hellessey. 146 You have been a pillar of unwavering support for so many years now. You housed me for 3 147 years whilst I did my Bachelor’s degree and again for a month when I came back to Tasmania 148 to start my Master’s degree, with a 1 year old in tow. You will never know the assistance and 149 stress relief having someone like you there that I could count on meant in those early days of 150 my career. For all the meals you made me, all the sheets you washed, all the times I forgot to 151 tell you I wasn’t coming home, and you were worried about me. For all the little things you’ve 152 done for me that I never said thank you for back then. 153 Thank you, Nanna. I appreciate it more than you will ever know. 154 6 155 Glossary 156 2F – sub-adult female 157 2M – sub-adult male 158 3F – mature female 159 3F-G – gravid female 160 3F-S – spent female 161 3M – mature male 162 3x3 – 3 km x 3km pixel 163 8D – 8 day average 164 Area 48 – South Atlantic Ocean sector 165 Area 58 – South Indian Ocean sector 166 Area 88 – South Pacific Ocean sector 167 CCAMLR – Convention for the Conservation of Antarctic Marine Living Resources 168 Chl a – chlorophyll a (mg m-2) 169 DAG - diacylglycerol 170 DHA – docosahexaenoic acid (22:6n-3) 171 DM – dry mass 172 EPA – eicosapentaenoic acid (20:5n-3) 173 FAME – fatty acid methyl ester 174 FFA – free fatty acids 175 Fishery-derived samples – samples collected by/from the fishery 176 GC-FID – gas chromatography flame ionisation detector 177 GC-MS – gas chromatography – mass spectrometer 178 GPS – global positioning system 179 HC – hydrocarbons 180 LC-PUFA – long chain (≥C20) polyunsaturated fatty acids 181 MODIS - moderate resolution imaging spectroradiometer 182 MUFA – monounsaturated fatty acids 183 MSI – marine snow indicators 7 184 n-3 – omega 3 185 PCA – principal component analysis 186 PL – phospholipids 187 PUFA – polyunsaturated fatty acids 188 RRS – remote sensed reflectance wavelengths 189 SD – standard deviation 190 SDA – steariadonic acid (18:4n-3) 191 SE - steryl esters 192 SFA – saturated fatty acids 193 SG – South Georgia (Sub-Area 48.3) 194 SO – Southern Ocean 195 SOI – South Orkney Islands (Sub-Area 48.2) 196 SST – sea surface temperature 197 ST – sterols 198 TAG – triacylglycerols 199 TLC-FID – thin layer chromatography – flame ionisation detector 200 TL – total lipid (mg) 201 TL DW – total lipid dry mass (mg g-1) 202 TSE – total solvent extract 203 TSN - total non-saponifiable neutral lipids 204 WAP – West Antarctic Peninsula (Sub-Area 48.1) 205 WE – wax esters 8 206 Table of Contents 207 DECLARATION OF ORIGINALITY ....................................................................................................... 2 208 STATEMENT OF AUTHORITY OF ACCESS ......................................................................................... 2 209 STATEMENT OF CO-AUTHOR CONTRIBUTIONS .............................................................................. 3 210 ACKNOWLEDGEMENTS ........................................................................................................................ 5 211 DEDICATION ............................................................................................................................................ 6 212 GLOSSARY ............................................................................................................................................... 7 213 FIGURE AND TABLE CAPTIONS ......................................................................................................... 11 214 ABSTRACT .............................................................................................................................................. 16 215 CHAPTER 1: INTRODUCTION ............................................................................................................. 19 216 217 218 219 220 BACKGROUND KRILL INFORMATION ................................................................................................................. 19 LIPIDS AND KRILL BIOCHEMISTRY .................................................................................................................... 22 KRILL OIL AND THE KRILL FISHERY ................................................................................................................. 25 REMOTE-SENSING ENVIRONMENTAL CONDITIONS ........................................................................................... 27 AIMS AND STRUCTURE OF THESIS ..................................................................................................................... 30 221 222 223 CHAPTER 2: SEASONAL AND INTERANNUAL VARIATION IN THE LIPID CONTENT AND COMPOSITION OF EUPHAUSIA SUPERBA DANA, 1850 (EUPHAUSIACEA) SAMPLES DERIVED FROM THE SCOTIA SEA FISHERY ..................................................................................................... 32 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 ABSTRACT ..................................................................................................................................................... 32 INTRODUCTION............................................................................................................................................ 33 METHODS ...................................................................................................................................................... 36 Sample collection and storage ..................................................................................................................... 36 Lipid-extraction technique ........................................................................................................................... 38 Statistical analyses ....................................................................................................................................... 39 RESULTS ........................................................................................................................................................ 40 Krill length and mass ................................................................................................................................... 40 Total lipid content ........................................................................................................................................ 42 Lipid class composition................................................................................................................................ 45 Quantitative lipid class analysis .................................................................................................................. 49 DISCUSSION .................................................................................................................................................. 54 Effect of Sex on the lipid seasonal cycle in krill........................................................................................... 54 Effect of krill length and mass on the seasonal lipid cycle .......................................................................... 55 Triacyclglycerol and Phospholipid seasonal cycles in krill......................................................................... 56 ACKNOWLEDGEMENTS ............................................................................................................................. 58 240 241 242 CHAPTER 3: REGIONAL VARIABILITY OF ANTARCTIC KRILL (EUPHAUSIA SUPERBA) DIET DURING THE LATE-SUMMER AS DETERMINED USING LIPID, FATTY ACID AND STEROL COMPOSITION ....................................................................................................................................... 59 243 244 245 246 247 248 ABSTRACT ..................................................................................................................................................... 59 INTRODUCTION............................................................................................................................................ 60 METHODS ...................................................................................................................................................... 63 Krill sample collection ................................................................................................................................. 63 Sample preparation...................................................................................................................................... 65 Total lipid, fatty acid and lipid class extraction and analysis...................................................................... 65 9 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 Column chromatography ............................................................................................................................. 67 Saponification .............................................................................................................................................. 68 Statistical analyses ....................................................................................................................................... 68 RESULTS ........................................................................................................................................................ 69 Total lipid content ........................................................................................................................................ 69 Lipid class composition................................................................................................................................ 72 Fatty acid content and composition ............................................................................................................. 77 Sterols .......................................................................................................................................................... 83 DISCUSSION .................................................................................................................................................. 86 Total lipid content ........................................................................................................................................ 86 Lipid class composition................................................................................................................................ 87 Lipid class composition by sex ..................................................................................................................... 88 Diatoms and dinoflagellates ........................................................................................................................ 90 Marine snow and detritus ............................................................................................................................ 92 Copepod and carnivory markers.................................................................................................................. 93 ACKNOWLEDGEMENTS ............................................................................................................................. 95 265 266 267 CHAPTER 4: ANTARCTIC KRILL (EUPHAUSIA SUPERBA DANA 1850) LIPID AND FATTY ACID CONTENT VARIABILITY IS ASSOCIATED TO SATELLITE DERIVED CHLOROPHYLL A AND SEA SURFACE TEMPERATURE IN THE SCOTIA SEA ..................................................................... 96 268 269 270 271 272 273 274 275 276 277 278 279 280 281 ABSTRACT ..................................................................................................................................................... 96 INTRODUCTION............................................................................................................................................ 97 METHODS .................................................................................................................................................... 101 Krill sample collection and analysis .......................................................................................................... 101 Satellite data extraction and analysis ........................................................................................................ 102 Data and Statistical analysis ..................................................................................................................... 104 RESULTS ...................................................................................................................................................... 105 Seas Surface Temperatures ........................................................................................................................ 105 Chlorophyll a Concentrations.................................................................................................................... 107 Geographic distribution ............................................................................................................................. 109 Lipid and fatty acid general trends ............................................................................................................ 113 Using environmental parameters as predictors of fatty acid biomarkers .................................................. 118 DISCUSSION ................................................................................................................................................ 122 ACKNOWLEDGEMENTS ........................................................................................................................... 129 282 CHAPTER 5: GENERAL DISCUSSION ................................................................................................131 283 284 285 286 287 Seasonality in krill lipid content and composition ..................................................................................... 132 Regional differences in krill diet ................................................................................................................ 133 Relating SST and Chl a to krill diet ........................................................................................................... 136 Future research.......................................................................................................................................... 140 Conclusions................................................................................................................................................ 146 288 APPENDICES .........................................................................................................................................148 289 290 APPENDIX 1 ................................................................................................................................................. 148 APPENDIX 2 ................................................................................................................................................. 151 291 REFERENCES ........................................................................................................................................153 292 10 293 Figure and Table captions 294 295 Figure 1.1. The Antarctic Food Web from http://www.classroomatsea.net/JR161/about.html accessed 2/8/16 296 297 298 299 Figure 1.2. CCAMLR Statistical Areas and Sub-Areas from http://www.antarctica.gov.au/magazine/2001-2005/issue-1-autumn-001/international/ccamlrthe-first-20-years accessed 17/2/16 300 301 302 303 304 Figure 2.1: Convention on the Conservation of Antarctic Marine Living Resources (CCAMLR) Statistical Area 48 and its Sub-Areas (from http://www.fao.org/fishery/area/Area48/en accessed 12/2/16) (A). Locations of collections of Euphausia superba by FV Saga Sea (Aker BioMarine) from January 2014 to September 2016 (B). 305 306 307 308 309 310 311 Figure 2.2. Total lipid content (mg g–1 of dry mass) of Euphausia superba showing the year and location of sample collection. The boxes from left to right show the two-week period of sample collection (where period 1 is January 1 to 15, 2014). Each box is the combination of three male and three female Euphausia superba from that period (N = 6). Each box represents 1 SD, with the whiskers the second SD and the bold line the mean. WAP, West Antarctic Peninsula; SOI, South Orkney Islands; SG, South Georgia. 312 313 314 315 Figure 2.3. The seasonal relationships between phospholipids and triacylglycerol (as % of the total lipid content) from fortnightly samples of Euphausia superba collected from January 2014 to September 2016. 316 317 318 319 320 321 322 Figure 2.4. Quantitative trends in the major lipid classes during 2014 to 2016; TAG mass (mg per krill) (A). TAG of krill dry mass (mg g–1) (B). PL mass (mg per krill) (C). PL of krill dry mass (mg g–1) (D). Each box is the combination of three male and three female Euphausia superba collected from a two-week period (N = 6). The first period of sample collection was January 1–15, 2014. Each box represents 1 SD, with the whiskers the second SD and the bold line the mean. TAG, triacylglycerols; PL, phospholipids. 323 324 325 Figure 2.5. TAG mass (mg) as a proportion of the total lipid mass (mg) of Euphausia superba (A). PL mass (mg) as a proportion of the total lipid mass (mg) of Euphausia superba (B). TAG, 11 326 327 triacylglycerols; PL, phospholipids. Each box represents 1 SD, with the whiskers the second SD and the bold line the mean. 328 329 330 Figure 3.1: Euphausia superba sample collection locations coloured by their Southern Ocean basin and showing CCAMLR management areas. 331 332 333 334 335 Figure 3.2: Total lipid content (mg g-1) dry mass of Euphausia superba in different Southern Ocean sectors (Atlantic, Pacific and Indian) and tissue types (stomach, digestive gland and whole krill) per animal. Each box represents 1 SD, with the whiskers the second SD and the bold line the mean. 336 337 338 339 Figure 3.3: Lipid class composition (% of total lipids) between Southern Ocean sectors (Atlantic, Pacific and Indian) and tissue types (stomach, digestive gland and whole krill) of Euphausia superba samples. Hydrocarbons includes wax and steryl esters. 340 341 342 343 344 Figure 3.4: Principal Component Analysis of the fatty acid composition (% data) of Euphausia superba samples from different Southern Ocean sectors (Atlantic, Pacific and Indian) and sites from: (A) the total lipid of whole krill, (B) the neutral lipid fraction of krill stomachs and (C) neutral lipid fraction of krill digestive glands. 345 346 347 348 Figure 3.5: Principal Component Analysis of the total lipid fatty acid composition (% composition) compared by sex between Southern Ocean sectors (Atlantic, Pacific and Indian) of whole Euphausia superba samples. 349 350 351 352 Figure 4.1: Sea surface temperatures (°C) from January 2014 – September 2016 coloured by Euphausia superba sample location (SG: South Georgia, SOI: South Orkney Islands, WAP: West Antarctic Peninsula). The x-axis is the season and year of krill sample collection. 353 354 355 356 357 Figure 4.2: Chlorophyll a concentrations (mg m-2) from January 2014 – September 2016 coloured by Euphausia superba sample location (SG: South Georgia, SOI: South Orkney Islands, WAP: West Antarctic Peninsula). The x-axis is season and year of krill sample collection. 358 12 359 360 361 362 363 364 Figure 4.3: The geographic distribution of krill total lipid (mg g-1) dry mass (TLDM), the chlorophyll a (Chl a) concentration (mg m-2) and the sea surface temperature (°C, SST) of Euphausia superba samples collected in the West Antarctic Peninsula. Locations are points that krill were harvested by FV Saga Seas from January to May 2014 - 2016. Maps were produced using the RStudio (version 1.0.153 © 2017) package ggmaps (Kahle and Wickham, http://journal.r-project.org/archive/2013-1/kahle-wickham.pdf). Map data © 2018 Google. 365 366 367 368 369 370 371 Figure 4.4: The geographic distribution of krill total lipid (mg g-1) dry mass (TLDM), the chlorophyll a (Chl a) concentration (mg m-2) and the sea surface temperature (°C, SST) of Euphausia superba samples collected in the South Orkney Islands. Locations are points that krill were harvested by FV Saga Seas from January to May 2014 - 2016. Maps were produced using the RStudio (version 1.0.153 © 2017) package ggmaps (Kahle and Wickham, http://journal.r-project.org/archive/2013-1/kahle-wickham.pdf). Map data © 2018 Google. 372 373 374 375 376 377 378 Figure 4.5: The geographic distribution of krill total lipid (mg g-1) dry mass (TLDM), the chlorophyll a (Chl a) concentration (mg m-2) and the sea surface temperature (°C, SST) of Euphausia superba samples collected at South Georgia. Locations are points that krill were harvested by FV Saga Seas from June to September 2014 - 2016. Maps were produced using the RStudio (version 1.0.153 © 2017) package ggmaps (Kahle and Wickham, http://journal.rproject.org/archive/2013-1/kahle-wickham.pdf). Map data © 2018 Google. 379 380 381 382 383 Figure 4.6: Multi Y axis plot of sea surface temperature (°C; black), total lipid dry mass (mg g-1; blue), chlorophyll a levels (mg m-2; green) and eicosapentaenoic acid (20:5n-3; EPA) percentage (%; yellow) for dates of krill (Euphausia superba) sample collection. Lines drawn for illustrative purposes to show general trends. 384 385 386 387 388 389 390 Figure 4.7: Slopes of the models of best fit (red) and the 95% confidence interval for that model (blue) for docosahexaenoic acid (DHA; 22:6n-3) percentage (%) in Euphausia superba sampled in the different Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) sub-areas (West Antarctic Peninsula (WAP), South Orkney Islands (SOI) and South Georgia (SG)) against sea surface temperature (°C, SST) and chlorophyll a (mg m2, Chl a). 391 392 393 394 395 Table 2.1: Average total lipid (mg g–1 dry mass), length (mm) and mass (g) of Euphausia superba by sex, season and year. Seasons are defined as summer (1 December to 28 February), autumn (1 March to 31 May), winter (1 June to 31 August), and spring (1 September to 30 November). 396 13 397 398 399 400 401 Table 2.2: Lipid content and lipid class composition in Euphausia superba expressed by year and season. Seasons are defined as: summer (1 December to 28 February), autumn (1 March to 31 May), winter (1 June to 31 August) and spring (1 September to 30 November). HC, hydrocarbons (includes steryl esters and wax esters present in trace amounts); TAG, triacylglycerols; FFA, free fatty acids; ST, sterols; DAG, diacylglycerols; PL, phospholipids. 402 403 404 405 406 407 Table 2.3: Lipid composition of Euphausia superba by sex and maturity stage for 2015–2016 samples (%, mean ± SD). HC, hydrocarbons (includes steryl esters and wax esters in trace amounts); TAG, triacylglycerols; FFA, free fatty acids; ST, sterols; DAG, diacylglycerols; PL, phospholipids; 3M, mature males; 3F, mature females; 3F-G, gravid females; 3F-S, spent females; 2F, sub-adult females; 2M, sub-adult males. 408 409 410 411 412 Table 3.1: Length, weight and total lipid content (mean ± SD) of different Euphausia superba tissue sample types from the Atlantic, Indian and Pacific Southern Ocean sectors. TL: Total lipid, DM: dry mass. N = number of samples (1 krill per whole krill sample, 2 krill digestive glands per sample and 12 krill stomachs per sample). 413 414 415 416 417 418 419 Table 3.2: Mass (mg g-1, mean ± SD) of each lipid class for each Euphausia superba sample and tissue types and Southern Ocean sectors (Atlantic, Indian and Pacific). Mass for the stomach and digestive gland samples are for the whole sample (2 digestive glands per sample and 12 stomachs per sample) and are not on a per krill basis. HC - hydrocarbons (including wax and sterol esters); TAG - triacylglycerols; FFA - free fatty acids; ST - sterols; PL phospholipids. 420 421 422 423 424 425 426 Table 3.3: Euphausia superba fatty acid groups (expressed as mg g-1 sample; mean ± SD) and selected major dietary fatty acid markers in different sectors of the Southern Ocean (Atlantic, Indian and Pacific) and tissue types (whole krill, stomach and digestive glands). MUFA: monounsaturated fatty acids, PUFA: polyunsaturated fatty acids, SFA: saturated fatty acids, MSI: marine snow indicator [Ʃ C15, C17 and C19 isomers], Copepods: [Ʃ 20:1n-9c + 22:1n-9c], Diatoms: [Ʃ 16:1n-7c +16:4n-1]. 427 428 429 430 Table 3.4: Sterols of Euphausia superba digestive gland and stomach samples by Southern Ocean sector (Atlantic, Indian and Pacific) expressed as a percentage (%, mean ± SD) of the total sterol profile. 431 432 433 Table 4.1: Decision table for why we increased our temporal and spatial fields for the red, green and blue wavelengths in MODIS to generate chlorophyll data 14 434 435 436 437 438 439 440 441 442 443 Table 4.2: Euphausia superba total lipid (mg g-1) dry mass (TLDM), and lipid class composition (phospholipid (PL) and triacylglycerol (TAG) percentage) and fatty acid (20:5n3 (EPA), 22:6n-3 (DHA), and 18:4n-3 (SDA)) percentage composition (%) and mass (ug) per krill against sea surface temperature (SST), chlorophyll a (Chl a) and their interaction terms for all seasons and pooled locations across the South Atlantic sector. Chl a was measured at both an overall scale (overall) and an 8-day 3 km x 3 km (8D 3x3) pixel scale for the entire South Atlantic sector. Values given are for: P values, r2 values (italics) and χ2 values (bold) for the model of best fit. Cells that are greyed out have a P value < 0.05, an r2 of >0.5 and a χ2 value > 0.1. 444 445 446 447 448 449 450 451 452 453 Table 4.3: Euphausia superba (collected from the West Antarctic Peninsula) total lipid (mg g1 ) dry mass (TLDM), lipid class (phospholipid (PL) and triacylglycerol (TAG)) and fatty acid (20:5n-3 (EPA), 22:6n-3 (DHA) and 18:4n-3 (SDA)) percentage composition (%) and mass (ug) in relation to sea surface temperature (SST), chlorophyll a (Chl a) and their interaction terms. Chl a was measured at both a Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) region wide scale (CCAMLR region) and at an 8-day 3 km x 3 km (8D 3x3) pixel scale. Values given are for: P values, r2 values (italics) and χ2 values (bold) for the model of best fit. Cells that are greyed out have a P value < 0.05, an r2 of >0.5 and a χ2 value > 0.1. 454 455 456 457 458 459 460 461 462 463 Table 4.4: Euphausia superba (collected from the South Orkney Islands) total lipid (mg g-1) dry mass (TLDM), lipid class (phospholipid (PL) and triacylglycerol (TAG)) and fatty acid (20:5n-3 (EPA), 22:6n-3 (DHA) and 18:4n-3 (SDA)) percentage composition (%) and mass (ug) in relation to sea surface temperature (SST), chlorophyll a (Chl a) and their interaction terms. Chl a was measured at both a Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) region wide scale (CCAMLR region) and at an 8-day 3 km x 3 km (8D 3x3) pixel scale. Values given are for: P values, r2 values (italics) and χ2 values (bold) for the model of best fit. Cells that are greyed out have a P value < 0.05, an r2 of >0.5 and a χ2 value > 0.1. 464 465 466 467 468 469 470 471 472 473 Table 4.5: Euphausia superba (collected from South Georgia) total lipid (mg g-1) dry mass (TLDM), lipid class (phospholipid (PL) and triacylglycerol (TAG)) and fatty acid (20:5n-3 (EPA), 22:6n-3 (DHA) and 18:4n-3 (SDA)) percentage composition (%) and mass (ug) in relation to sea surface temperature (SST), chlorophyll a (Chl a) and their interaction terms. Chl a was measured at both a Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) region wide scale (CCAMLR region) and at an 8-day 3 km x 3 km (8D 3x3) pixel scale. Values given are for: P values, r2 values (italics) and χ2 values (bold) for the model of best fit. Cells that are greyed out have a P value < 0.05, an r2 of >0.5 and a χ2 value > 0.1. 15 474 Abstract 475 Lipids are key biochemicals that form both cell membranes and energy stores. Lipids are of 476 particular importance in energy poor environments where animals require stores to survive long 477 periods of food shortage. In the Antarctic, food availability is dominated by extreme seasonal 478 shifts in the environment and energy rich food is scarce for a substantial period of time each 479 year. Antarctic krill (Euphausia superba) have adapted to have large lipid stores (over a third 480 of their dry weight) during winter for this reason. Krill are a key species in the Antarctic 481 environment; their biomass links lower and higher trophic levels and forms the main energy 482 conduit for the system. Krill feed on diatoms, dinoflagellates and other algal species year- 483 round, resulting in high omega-3 polyunsaturated fatty acids which are essential for krill health, 484 growth and reproduction. Krill-derived omega-3 containing products (particularly 485 eicosapentacnoic acid (20:5n-3) and docosahexaenoic acid (22:6n-3)) are sold as nutraceuticals 486 for human consumption. Krill oil tablets (sold as an omega-3 supplement) are now the fastest 487 growing nutraceutical globally. 488 Understanding the krill life cycle is hampered by the restricted nature of scientific sampling. 489 Knowledge of krill diet and krill lipid dynamics is lacking for the Indian and Pacific Ocean 490 sectors, as most studies have focused on the South Atlantic Ocean sector where the krill fishery 491 is based. Most scientific research voyages are conducted during summer months and all 492 scientific studies are restricted in their spatial and temporal scale. Information on krill 493 recruitment and reproduction in the Indian and Pacific Ocean sectors is also not as developed 494 as in the South Atlantic, where fishery-derived samples are also available. 495 A major gap in current ecosystem models is the link between environmental drivers (such as 496 upwelling of nutrients, sea surface temperature and height, sea ice extent and thickness and 497 salinity) and their impact on primary production and therefore food availability during extreme 16 498 seasonal shifts in Antarctica. One way of measuring these environmental drivers is through 499 remote-sensing via satellite, which can gather data over large geographic areas and over long 500 timeframes. Satellite-derived data for biological and ecological measures is still developing as 501 a tool for oceanographers and other end users. However, one area of growing importance is in 502 the use of ocean colour data which can be converted into chlorophyll a concentrations (a proxy 503 for primary production) via a standard algorithm. By linking the GPS locations of commercial 504 krill harvesting, and therefore krill swarms, to environmental data obtained through remote- 505 sensing from the same date, the relationship between the environment krill live in and their 506 biochemical composition can be examined in ways not previously explored. 507 My study used samples collected by a member of the krill fishery, Aker BioMarine, over a 508 continuous three-year period in the South Atlantic Ocean to look at the seasonal and interannual 509 trends in krill total lipids and lipid classes (such as those used for energy storage and the 510 structure and function of cells). This dataset is unprecedented in its seasonal and spatial 511 coverage in the South Atlantic Ocean. This study has been able to establish the sinusoidal shape 512 of the seasonal and interannual trend in krill total lipids and its associated lipid classes. No 513 samples from the fishery were available from other sectors. These South Atlantic Ocean krill 514 samples were contrasted to krill samples collected from scientific expeditions in the other two 515 ocean basins surrounding Antarctica (Pacific and Indian Oceans). Krill diet was investigated 516 at a regional scale during the crucial late-summer spawning period. Results from my study 517 revealed that krill diet varies between ocean basins, with Indian Ocean krill showing a distinctly 518 different diet to Pacific and Atlantic Ocean krill, during the late-summer. 519 The fishery-derived samples were also related back to environmental data collected via 520 satellite, for both chlorophyll a and sea surface temperature, to investigate if environmental 521 drivers influenced krill lipid biochemistry. This study showed that both sea surface temperature 522 and chlorophyll a concentrations (derived from ocean colour data) can be related to krill lipid 17 523 and fatty acid dynamics. Krill lipid composition and content were shown to be correlated to 524 these environmental factors through simple models. The combination of results from this study 525 will help fill the data gaps in ecosystem models and enable better determination of krill diet, 526 recruitment and reproduction in all ocean basins surrounding Antarctica. These advances in 527 krill knowledge will help improve fishery management policies. 18 528 Chapter 1: Introduction 529 Background Krill Information 530 Antarctic krill (Euphausia superba, hereon krill) are a keystone species in the Southern Ocean 531 (SO) ecosystem. Mature adult krill range from 30 to over 60 mm in length (Standard length 1; 532 Kirkwood, (1982)), with a life span of 5-6 years (Siegel, 1987) in the wild. Krill are found at 533 latitudes as far north as 50 °S (near South Georgia) to the Antarctic continental edge and can 534 survive in sea temperatures from -1.5 ℃ to +5 ℃ (Miller and Hampton, 1989, Atkinson et al., 535 2008). Krill are generally found in greater densities and are more abundant in high nutrient 536 waters that are rich in phytoplankton (Siegel, 2005). Krill can be found from the surface of the 537 ocean to depths of 4,500 m, but are more often seen in the upper 200 m of the water column 538 (Morris et al., 1984, Schmidt et al., 2014). 539 Krill reproduction and recruitment can be highly variable from year to year (Yoshida, 540 2009, Kawaguchi, 2016) and depends largely on feeding conditions. The availability of sea ice 541 algae, which is directly related to sea ice extent, is known to influence juvenile recruitment in 542 krill (Atkinson et al., 2002, Virtue et al., 2016, Schaafsma et al., 2017). During the productive 543 spring and summer seasons, krill populations grow from the new recruitment of juveniles 544 before diminishing over the harsher winter season (Kawaguchi and Satake, 1994, Siegel and 545 Loeb, 1995, Wiedenmann et al., 2009). Krill also aggregate to form swarms that can be up to 546 100 km2 (Tarling et al., 2009a). These swarms can range in densities from 100 krill m-3 to over 547 1,000 krill m-3 (Klevjer et al., 2010). The sex ratio of individuals in swarms can vary from 548 100% male to 100% female in swarms during the reproductive spawning season in late summer 549 to 50:50 male:female during winter (Atkinson et al., 2006, Tarling et al., 2016a). The patchy 550 distribution and vast geographic range of krill makes it difficult to provide robust estimates of 551 krill biomass and production (Nicol and Endo, 1999). Circumpolar estimates place krill 19 552 biomass at roughly 80-200 million metric tons (Demer, 2004, Atkinson et al., 2009). This large 553 biomass within the SO ecosystem makes krill both a keystone species and the center of the 554 wasp-waisted ecosystem, where energy is funneled from lower to higher trophic levels through 555 a single species (Figure 1.1). 556 557 558 Figure 1.1: The Antarctic Food Web http://www.classroomatsea.net/JR161/about.html accessed 2/8/16. 559 560 Most of the charismatic megafauna associated with Antarctica such as seals, penguins 561 and whales, as well as fish, squid and seabirds rely on krill as a primary food source (Hill et 562 al., 2006, Murphy et al., 2007, Ward et al., 2012a). Krill are often an indirect secondary food 563 source as well, due to it being an abundant energy-dense nutrient source (Hagen et al., 2001) 564 in the ecosystem. Krill form the key link between primary production and higher trophic levels 565 in Antarctica (Everson, 2008). Most apex predators in the SO tend to breed during the austral 20 566 summer, due to the higher abundance of available prey at this time (Bryden, 1983, Trivelpiece 567 et al., 2011, Watts and Tarling, 2012). Similarly, krill also spawn at this time as they are feeding 568 on the increased levels of primary production due to the spring/summer algal blooms (Garabotti 569 et al., 2005; Vernet et al., 2008). Krill become more energy-dense throughout summer due to 570 their grazing on the increased levels of primary production, which in turn increases their fat 571 levels. During the other times of the year, marine predators forage at dynamic oceanographic 572 features where increased levels of primary production (and hence krill) can be found year- 573 round (Charrassin et al., 2002, Heerah et al., 2013). 574 Krill diet varies greatly. Krill consume phytoplankton, predominantly diatoms and 575 dinoflagellates, which are high in chlorophyll a (a green pigment) and cause krill digestive 576 glands to be green throughout periods of high feeding (Virtue et al., 1993a; Yoshida et al., 577 2009). They also consume zooplankton, such as copepods, as well as bacteria, marine snow 578 and sea ice microbiota (Passmore et al., 2006, Kohlbach et al., 2015). Krill diet is also known 579 to shift with environmental conditions. Krill are known to be cannibalistic during harsher 580 winter conditions (Atkinson et al., 2002, Ju and Harvey, 2004, O’Brien et al., 2011, Virtue et 581 al., 2016, Meyer et al., 2017) and can be almost entirely herbivorous during good summers 582 (Virtue et al., 2011; Ericson et al., 2018a). Krill diet has been studied using various techniques 583 including DNA sequencing (Passmore et al., 2006), microscopy (Töbe et al., 2010) and 584 signature fatty acid analysis (Virtue et al., 1993a, Virtue et al., 2000, Phleger et al., 2002, 585 Schaafsma et al., 2017). Such studies are often restricted by the number of samples and the 586 season and location of sample collection. Thus, how krill diet changes in time and space and 587 especially how krill diet will be impacted by climate change is still an emerging area of research 588 (Kawaguchi et al., 2011, Ericson et al., 2018b). It has been suggested there will be a shift from 589 energy rich (e.g. krill) to energy poor species (e.g. salps) in polar regions under climate change 590 (Loeb et al., 1997, Atkinson et al., 2004). As primary production and microbial community 21 591 assemblages shift with climate change (Deppler and Davidson, 2018; Hancock et al., 2018), 592 energy transfer at lower trophic levels will be affected. This shift in krill diet will in turn have 593 major implications for the transfer of energy up to higher trophic levels (Kattner et al., 2007). 594 There has been a large focus on krill-centric studies in the Antarctic because of krill’s 595 large biomass and central role in the wasp-waisted SO ecosystem. From the Discovery voyages 596 in the early 1900s (Hardy, 1928, Deacon et al., 1939) to the more recent acoustic surveys in 597 the 2000s (Nicol et al., 2000a, Demer, 2004, Jarvis et al., 2010), krill have been the focus of 598 many large multinational expeditions (Nicol et al., 2010). However, most scientific expeditions 599 are only able to sample and undertake voyages during the summer season due to logistical, 600 financial, and time constraints. Because of this constraint on scientific voyages, there is only 601 limited information available about their winter adaptations (Huntley et al., 1994, Kolakowska 602 et al., 1994, Daly, 2004, O’Brien et al., 2011). The winter period will impact on all other aspects 603 of krill life such as their metabolism, growth, reproduction and diet. Modern facilities, such as 604 laboratories and aquariums (e.g. the Australian Antarctic Division’s Krill Aquarium; 605 Kawaguchi et al. (2010a)), have helped fill this gap by enabling the study of krill life history 606 and their seasonal diet. Laboratory studies have also improved the understanding of krill 607 genomics, krill aging and how micro-plastics impact krill. Aquarium use has also enabled 608 ocean acidification studies to be undertaken; such studies cannot be performed in situ (Virtue 609 et al., 1997, Yoshida et al., 2009, Kawaguchi et al., 2011, Kilada et al., 2017, Dawson et al., 610 2018, Ericson et al., 2018b). These facilities, however, cannot replicate the large regional and 611 interannual environmental stressors impacting krill. 612 Lipids and Krill Biochemistry 613 Lipids are biochemical building blocks that are key to the structure and function of living cells 614 and can have multiple roles in organisms. Examples of lipids include fats and oils used in 22 615 energy storage (Hagen et al., 1996, Varpe et al., 2009, Yin et al., 2016), as well as waxes, 616 certain vitamins (Burton and Ingold, 1986), hormones (Fernlund and Josefsson, 1972) and most 617 of the non-protein membrane of all cells (Caraveo-Patiño et al., 2009, Tancell et al., 2012). 618 Lipids are classified into lipid classes, such as: polar lipids (includes phospholipids, PL) and 619 neutral lipids (triacylglycerols (TAG), wax esters (WE), sterols (ST), free fatty acids (FFA), 620 hydrocarbons (HC) and diacylglycerols (DAG)). The major lipid classes are predominantly 621 made up of smaller building blocks known as fatty acids which themselves come in various 622 forms. Fatty acids can be: without any carbon-carbon double bonds, so called saturated fatty 623 acids (SFA), or can be unsaturated fatty acids, which can have either one carbon-carbon double 624 bond (monounsaturated fatty acids (MUFA)) or multiple carbon-carbon double bonds, 625 polyunsaturated fatty acids (PUFA). Long-chain (≥C20) PUFA (LC-PUFA) are essential for 626 health and survival in organisms, particularly the omega-3 (n-3) LC-PUFA which are required 627 for lipid derived cell signaling, cell membrane fluidity, reproduction and growth (Kolakowska 628 et al., 1994, Ross and Quetin, 2000, Yoshida et al., 2011, Kawaguchi, 2016). The variability of 629 lipid and fatty acid content and composition in krill has been related to sex, developmental 630 stage, nutrient conditions and nutritional requirements (Clarke, 1984, Saether et al., 1985, Pond 631 et al., 1995, Skerratt et al., 1995, Mayzaud, 1997, Virtue et al., 1997, Mayzaud et al., 1998, 632 Atkinson et al., 2002, Ju and Harvey, 2004, Alonzo et al., 2005, Yoshida et al., 2011). 633 Most marine organisms use TAG and/or WE as their primary form of energy storage 634 and may store their n-3 LC-PUFA within these stores (Wakefield et al., 2008, Mayzaud et al., 635 2011, Pond et al., 2014). Krill can also use PL as an energy store (Itonori et al., 1991, Hagen 636 et al., 1996, Yin et al., 2016), in addition to its structural role, and their n-3 LC-PUFA are 637 primarily found in this lipid class (Kolakowska et al., 1994, Mayzaud et al., 1998, Hill, 2013). 638 The fatty acids found within an organism are accumulated through dietary inputs 639 (Iverson et al., 2004), as well as being broken down and metabolized into other fatty acids 23 640 (Huntley et al., 1994). Some fatty acids cannot be broken down or converted in higher level 641 organisms and can therefore be traced and quantified as a biomarker of their dietary input 642 (Volkman and Nichols, 1991, Iverson et al., 2004) through thin-layer chromatography and 643 signature fatty acid analysis. Signature fatty acid analysis allows us to interpret the links 644 between trophic levels and to also quantify the prey items seen in the diets of predator tissues. 645 These fatty acid biomarkers can show seasonal variability due to environmental and energy 646 demands of various organisms. For example, krill have seasonal fluxes in their total lipids (up 647 to 40% dry mass at the start of winter (Falk-Petersen et al., 2000, Atkinson et al., 2002)) and 648 in their n-3 LC-PUFA, which increase with the phytoplankton blooms in the austral summer 649 (Ericson et al., 2018a). Tracing the energy flow in the Antarctic food web is particularly 650 important due to the relative simplicity and small number of trophic levels within the system 651 (Figure 1.1) compared to most other marine ecosystems. Any breakdown of energy flow within 652 the food web has far reaching effects, as the alternative pathways are often few and do not 653 contain the required nutrients to support higher trophic levels if the original pathway is not 654 restored. For example, higher trophic level predators require krill-derived n-3 LC-PUFA for 655 their health, growth and reproduction; the alternative trophic pathway, mainly through salps, 656 is lipid and therefore n-3 LC-PUFA poor (Loeb et al., 1997, Atkinson et al., 2004). 657 Most of the lipids in the SO originate from primary producers and in particular the 658 essential n-3 LC-PUFA; eicosapentaenoic acid (20:5n-3, EPA), docosahexaenoic acid (22:6n- 659 3, DHA) and stearidonic acid (18:4n-3, SDA), which are diatom and dinoflagellate markers 660 respectively (Nichols et al., 1986, Nichols et al., 1988, Schmidt et al., 2012, Kohlbach et al., 661 2017). If primary production alters on a large scale in the SO due to an environmental shift, 662 then all of the consumers up the trophic links will be influenced by that change, particularly in 663 their fatty acid biomarkers. Hence, a decrease in primary production can be linked to a decrease 664 in n-3 LC-PUFA levels in primary consumers such as zooplankton (Phleger et al., 1998, Hagen 24 665 and Kattner, 2014, Turner, 2015, Kohlbach et al., 2018). These n-3 LC-PUFA, particularly 666 EPA and DHA, are also the main oil parameters targeted by the commercial krill fishery (Butler, 667 2007, Hill, 2013, Schutt, 2016). 668 Krill Oil and the Krill Fishery 669 The Antarctic krill fishery began in the late 1960s, and is governed by the Commission for the 670 Conservation of Antarctic Marine Living Resources (CCAMLR), a world leader in sustainable 671 fishing practices (Nicol and Endo, 1999, CCAMLR, 2017). During the 2000s the fishery 672 expanded to include the production of commercially processed krill oil (Nicol et al., 2012). 673 Krill oil recently emerged internationally as an important alternative to fish oil as a source of 674 n-3 LC-PUFA, as a nutraceutical product for human consumption. Krill n-3 LC-PUFA are 675 mainly present in the phospholipid form, whereas other commercially available n-3 LC-PUFA 676 supplements, such as fish and squid oil, are generally in the conventional TAG form. Krill oil 677 also contains natural antioxidants such as astaxanthin, which krill themselves contain. With an 678 estimated value of $AUD 3.3 million/tonne (Kawaguchi, pers. comm.), several new krill oil 679 products are now in the top five omega-3 supplements sold within Australia (Schutt, 2016). 680 The trend of commercial krill fishing operators seeking to expand into this niche is growing 681 rapidly (CCAMLR, 2012, Nicol et al., 2012). The impacts of this expanding fishery on the SO 682 ecosystem and krill biomass is not well understood, particularly in sectors outside of the South 683 Atlantic Ocean (Area 48, Figure 1.2). In 2016, the Chinese started an exploratory krill fishery 684 in East Antarctica, within the Indian Ocean Sector (Area 58, Figure 1.2), but knowledge of krill 685 diet, health, growth, recruitment and reproduction in this region is mostly lacking (Nicol et al., 686 1992, Mayzaud et al., 1998, Pakhomov, 2000, Swadling et al., 2010, Virtue et al., 2010). The 687 Pacific Ocean Sector (Area 88, Figure 1.2) has rarely been fished and there is even less 688 available data on krill from this region. 25 689 690 691 692 Figure 1.2: CCAMLR Statistical Areas and Sub-Areas from http://www.antarctica.gov.au/magazine/2001-2005/issue-1-autumn-001/international/ccamlr the-first-20-years accessed 17/2/16. 693 694 CCAMLR currently sets krill fishing catch limits based on precautionary ecosystem 695 models to ensure the health of the entire Antarctic ecosystem. As lipid storage is vital for krill 696 reproduction and recruitment (Yoshida et al., 2011, Kawaguchi, 2016), providing CCAMLR 697 with data that includes krill lipid dynamics throughout all seasons and from different regions 698 will be useful to help fill gaps in these models. 26 699 Due to the ability of the fishery to cover large geographic areas and to harvest krill year- 700 round, the utilisation of samples from the krill fishery for scientific research purposes is 701 becoming more common (Kawaguchi and Nicol, 2007, Hill, 2013, Descamps et al., 2016, Nicol 702 and Foster, 2016, Tarling et al., 2016a). Gaining knowledge and samples of krill from the 703 fishery is a cheaper alternative to scientific voyages. Aker BioMarine, a Norwegian krill fishing 704 company, lands over half the total global krill catch for use in aquaculture and human 705 consumption (CCAMLR, 2017). The company harvests 24 hours a day, from December to 706 September every year, using two high technology vessels that are fitted with a mid-water 707 continuous pumping system (Siegel, 2016). Aker BioMarine harvests krill from the South 708 Atlantic Sector (Area 48), specifically the West Antarctic Peninsula (WAP, Sub-Area 48.1), 709 the South Orkney Islands (SOI, Sub-Area 48.2) and South Georgia (SG, Sub-Area 48.3). 710 Fishing companies generally have a greater geographic coverage of krill in the region than even 711 some large multinational expeditions such as the CCAMLR 2000 survey (Nicol et al., 2000a, 712 Demer, 2004). Fisheries-based research can cover a wide area and krill samples can be 713 collected for much of the year, unlike scientific voyages. 714 Remote-Sensing Environmental Conditions 715 Another alternative to traditional scientific voyages to gain insights into the Antarctic 716 environment is the use of data from remotely operated sensors, such as satellite-based 717 instrumentation. Satellites are able to measure multiple environmental variables 718 simultaneously including, but not limited to: sea surface temperature (SST), sea surface height, 719 fluorescence, ocean colour, sea ice extent and thickness, wind speed and direction, and 720 chlorophyll a concentration (a proxy for primary production) (Moore and Abbott, 2002, 721 Johnson et al., 2013, Zeng et al., 2016, Kahru et al., 2017). Unfortunately, the use of such 722 remote-sensing techniques to gain environmental data, such as ocean colour, is incredibly 723 difficult in polar regions due to cloud cover, sea ice reflectance (albedo effect) and the sun 27 724 angle in winter (IOCCG, 2015). Satellites can, however, collect data over a range of geographic 725 areas (small 1 km x 1 km pixels to large 100 km x 100 km boxes (Moore and Abbott, 2002, 726 Zeng et al., 2016)) and time periods (as short as 5 minutes and as long as 10+ years, (IOCCG, 727 2015, Kahru et al., 2017)). 728 environmental data to be used in relation to other variables, such as biological factors. Such broad data collection allows for multiple scales of 729 Relating biological data to environmental data from satellites is still in its infancy, 730 particularly in polar regions. The use of accessible remote-sensed data is still not common for 731 biologists, ecologists and physiologists (Moore and Abbott, 2002, Rayner, 2003, Mackey et 732 al., 2012, Tarling et al., 2018). One of the few often-used remotely-sensed pieces of biological 733 data is ocean colour data, which can be converted into chlorophyll a concentrations using 734 algorithms such as those reported by Johnson et al. (2013). Greater expertise in biological 735 remote-sensing data could rapidly increase knowledge and understanding of how climate 736 change impacts the biology and ecology of polar regions. This includes the effect of the 737 reduction of sea ice extent on juvenile krill diet and recruitment through the loss of sea ice 738 algae (Loeb et al., 1997, Atkinson et al., 2004). 739 In terms of climate change, how krill will deal physiologically and ecologically with 740 warming oceans and ocean acidification is a developing area of research (Kawaguchi et al., 741 2011, Kawaguchi et al., 2013, Loeb and Santora, 2015, Tarling et al., 2016b, Ericson et al., 742 2018b, Atkinson et al., 2019, Ericson et al., 2019a). Studies on the effects of climate change 743 and ocean acidification on krill have reported negative effects on embryonic development 744 (Kawaguchi et al., 2011), but negligible effects on krill lipid dynamics both in terms of ocean 745 warming (Brown, 2010) and ocean acidification (Ericson et al., 2018b and 2019a) when fed a 746 standard algal diet. Krill diet, however, may be impacted into the future; particularly if warming 747 or acidification changes the primary production community composition such as in Deppeler 748 and Davidson (2017), or by a microbial community assemblage change as in Hancock et al. 28 749 (2018). How such changes will impact on krill diet is currently unknown. Further controlled 750 experimental studies need to be conducted to investigate how increased temperatures, different 751 primary production and microbial community compositions and ocean acidification will 752 interact and affect the content and composition of krill lipids, with emphasis on the key n-3 753 LC-PUFA in all life stages of krill. 754 29 755 Aims and Structure of Thesis 756 The aims of my research are to provide a synthesis on the biochemistry and inferred diet of 757 krill over large spatial and temporal scales. Krill lipid dynamics and krill diet was related to 758 remotely-sensed environmental data. Specifically, the aims were to: 759 760 I. Investigate seasonal and interannual variations in krill by using total lipid and lipid class 761 content and composition. An investigation into the content and composition of krill lipid and 762 lipid classes using samples obtained from the krill fishery in the South Atlantic Ocean is 763 presented in Chapter 2. This study is the first detailed examination of both the total lipid and 764 lipid classes of krill across all seasons, spanning three consecutive years. 765 766 II. Examine the total lipid, fatty acid and sterol content and composition of whole krill, as well 767 as from their digestive glands and stomachs, within all three of the ocean basins surrounding 768 Antarctica (Pacific, Indian and Atlantic Ocean sectors). Chapter 3 presents results on krill diet, 769 particularly as determined via analysis of the neutral lipid fraction, and it’s variations at a 770 regional scale during the crucial late-summer spawning period. 771 772 III. Explore whether the use of remotely-sensed environmental data can be linked to krill lipid 773 dynamics. Specifically, sea surface temperature and chlorophyll a concentrations will be used, 774 as they are known to be related to krill diet and can be measured via satellite instrumentation. 775 This study combines and applies these varied methodologies within the polar region to show 776 how end users might utilise this information. Results of this research are presented in Chapter 777 4. 30 778 In Chapter 5 the results from the combined chapters are synthesized into a General Discussion 779 on the spatiotemporal variability of krill lipid dynamics and how krill diet (and therefore their 780 dietary-derived lipids) might be affected by present day environmental parameters. 781 Outcomes presented in the individual chapters and overall thesis will be important for and 782 directly relevant to krill fishing companies, CCAMLR, non-government organisations and 783 Antarctic ecologists. The results are also relevant for ongoing management of the krill fishery 784 by providing real-world examples of potential feedback management data. This includes 785 provision of a baseline for these key biochemical and environmental parameters, and the 786 possible effects of climate change to krill lipid dynamics. 787 Additional outcomes such as co-authored papers, conference presentations and scientific 788 outreach from this project are recorded in Appendix 2. 31 789 Chapter 2: Seasonal and interannual variation in the lipid content and 790 composition of Euphausia superba Dana, 1850 (Euphausiacea) samples 791 derived from the Scotia Sea fishery 792 793 This chapter has been published: 794 795 796 797 Hellessey N, Ericson JA, Nichols PD, Kawaguchi S, Nicol S, Hoem N, Virtue P (2018). Seasonal and interannual variations in the lipid content and composition of Euphausia superba Dana, 1850 (Euphausiacea) samples derived from the Scotia Sea krill fishery. Journal of Crustacean Biology, doi:10.1093/jcbiol/ruy053. 798 799 ABSTRACT 800 The Antarctic krill (Euphausia superba Dana, 1850) is an important trophic link between 801 phytoplankton and higher trophic levels. Knowledge of the lipid biochemistry of krill assists 802 in understanding their seasonal biology and predicting their responses to ecological changes. 803 We collected daily samples of krill from a commercial fishing vessel operating in the Atlantic 804 Sector of the Southern Ocean from 2014 to 2016. We analysed the total lipid content of the 805 krill and the relative distribution of lipid class levels to examine seasonal trends. Krill total 806 lipid content varied significantly within and between seasons and sexes. An annual sinusoidal 807 trend was seen in total lipid content with the highest values in autumn and the lowest in spring 808 (average 380 and 87 mg g-1 dry mass, respectively). Total lipids in krill increased during 809 summer, peaking in autumn, with the total lipids in winter individuals decreasing towards 810 spring. The relative distribution of lipid class levels varied between season and year. Levels of 811 triacylglycerol showed the same seasonal trend as total lipid content, whilst phospholipid 812 showed the inverse trend indicating the contrasting roles of these two dominant lipid classes. 813 These data provide high-resolution information on the seasonality of krill lipid content and 814 composition. This information has both ecological and commercial utility. 815 32 816 INTRODUCTION 817 Antarctic krill (Euphausia superba Dana (1850)) is a keystone species in the Antarctic 818 ecosystem as many of the marine predators in the region consume it due to its high abundance 819 and nutritional value. Krill have high lipid (oil) content (Falk-Petersen et al., 2000, Iverson et 820 al., 2004) when compared to other prey species such as myctophid fishes (Phleger et al., 1997, 821 Phleger et al., 1999). Lipids are an important energy reserve for krill (Falk-Petersen et al., 1981), 822 as well as being necessary for cell membrane structure and function. The omega-3 long-chain 823 (≥C20) polyunsaturated fatty acids (n-3 LC-PUFA), largely eicosapentaenoic acid (EPA, 20:5n- 824 3) and docosahexaenoic acid (DHA, 22:6n-3), are required for reproduction and growth in krill 825 (Clarke, 1980, Quetin et al., 1994, Mayzaud, 1997, Ross and Quetin, 2000, Tarling et al., 2009b) 826 and animals at higher trophic levels (Suhr et al., 2003, Wakefield et al., 2008, Caraveo‐Patiño 827 et al., 2009, Mayzaud et al., 2011, Trivelpiece et al., 2011). Lipid content in krill has been 828 related to sex, developmental stage, nutrient conditions, and nutritional requirements (Clarke, 829 1984, Virtue et al., 1997, Mayzaud et al., 1998, Ju and Harvey, 2004). Krill are primarily 830 herbivores feeding on phytoplankton in the summer and sea-ice algae and associated organisms 831 in the winter (O’Brien et al., 2011, Virtue et al., 2016, Schaafsma et al., 2017). They can also 832 be carnivorous and are known to consume copepods (Kohlbach et al., 2017, Schaafsma et al., 833 2017) and may even become cannibalistic in harsher winter seasons (Ju and Harvey, 2004, 834 Schaafsma et al., 2017). These changes in krill diet can be detected in their lipid content and 835 composition (Clarke, 1984, Kohlbach et al., 2015, Schaafsma et al., 2017). 836 Understanding the seasonal cycle of krill lipids and lipid storage mechanisms will assist 837 in better understanding krill life history (Quetin et al., 1994, Falk-Petersen et al., 2000, Yoshida 838 et al., 2011). The role of lipids in krill during periods of limited food supply is unclear due to 839 a lack of high-frequency year-round sampling, and in particular, adequate winter sampling 840 (Ikeda and Dixon, 1982, Marschall, 1988, Hagen et al., 1996, Atkinson et al., 2002, Kohlbach 33 841 et al., 2017, Schaafsma et al., 2017, Kohlbach et al., 2018). Many questions on the key biology 842 and ecology of krill remain unanswered because research in the Antarctic is logistically 843 difficult and costly, particularly in winter. There is conflicting evidence as to whether krill store 844 lipid for overwintering (Ikeda and Dixon, 1982, Marschall, 1988, Virtue et al., 1993b, Quetin 845 et al., 1994, Hagen et al., 1996, Atkinson et al., 2002, Schaafsma et al., 2017) or whether they 846 use their body lipid and protein and shrink in response to food shortage (Ikeda and Dixon, 1982, 847 Nicol et al., 1992, Sun et al., 1995, Alonzo and Mangel, 2001, Tarling et al., 2016a). Better 848 data from krill collected during winter will be critical in clarifying their over-wintering 849 strategies and fishery samples have already contributed to this effort (Tarling et al., 2016a, Kim, 850 2014). 851 Lipid storage is the key to reproductive success in krill (Clarke and Morris, 1983, 852 Cuzin-Roudy and Buchholz, 1999, Ross and Quetin, 2000). The cost of reproduction is also 853 one of the main parameters involved in the formulation of an energy budget (Clarke and Morris, 854 1983, Perissinotto et al., 2000, Ross and Quetin, 2000). Female krill store lipids in their eggs 855 prior to spawning and larval krill require these lipid stores to survive their initial growth and 856 development (Clarke, 1980, Atkinson et al., 2002, Tarling et al., 2009a, Yoshida et al., 2011). 857 Reproductive success and recruitment in krill populations are known to fluctuate markedly 858 from year to year (Ross and Quetin, 2000). Understanding the causes of these variations in 859 reproductive output requires samples collected at regular intervals throughout the year. Whilst 860 regular sampling is possible near research stations in Antarctica, long-term open-ocean 861 scientific sampling has proved difficult. Krill fishing vessels, however, operate year-round and 862 are a potential source of krill in all seasons (Nicol et al., 2012). Samples from the krill fishery 863 can provide a valuable insight into annual, seasonal, and short-term changes in krill biological 864 parameters (Kim, 2014, Tarling et al., 2016a). Analyses of these samples can help inform the 865 fishery on changes in krill habitat, ecology, physiology, and lipid biochemistry. 34 866 Krill has been commercially harvested for its oil, meal, and other components for the 867 last four decades (Nicol and Foster, 2016). Krill fishing is the second-largest, single-species 868 crustacean fishery in the world and the largest fishery in the Southern Ocean; 260,151 tonnes 869 in 2016 (Nicol and Foster, 2016). The krill fishery has been located in recent years in the Scotia 870 Sea, South Atlantic, namely FAO statistical Area 48 (Fig. 1A) (CCAMLR, 2017), although a 871 small catch was taken from the Indian Ocean Sector in 2017 (CCAMLR, 2017). The fishing 872 fleet migrates from the South Orkney Islands (SOI) (Sub-Area 48.2) in autumn, up to the West 873 Antarctic Peninsula (WAP) (Sub-Area 48.1) as winter progresses then up to ice-free South 874 Georgia (SG) (Sub-Area 48.3) when the WAP and SOI are either ice-covered or these fishing 875 grounds have reached their target catch limits for the season (Fig. 1A, B). 876 The krill fishery produces meal and oil (Nicol and Foster, 2016). Antarctic krill oil has 877 emerged over the past decade as an important and alternative source of omega-3 LC-PUFA for 878 use in nutraceutical and pharmaceutical products for human consumption as well as in 879 aquaculture and livestock feed (Schutt, 2016). In humans, omega-3 from krill oil may be more 880 easily absorbed than fish oil because a substantial portion of the omega-3s are attached to 881 phospholipids compared to the mainly triacylglycerol form found in fish oil (Kwantes and 882 Grundmann, 2015). These two lipid classes dominate krill oil composition, but their seasonal 883 and interannual changes are not clearly understood (Clarke, 1980, Pond et al., 1995, Mayzaud 884 et al., 1998, Falk-Petersen et al., 2000, Atkinson et al., 2002). 885 We analysed fisheries-derived samples to gain insights into specific aspects of krill lipid 886 production and storage, aiming to investigate the seasonal and interannual variation in the lipid 887 content and lipid class content and composition of krill. We compared the lipid content and 888 lipid class composition of krill caught at various fishing locations within CCAMLR 889 (Convention on the Conservation of Antarctic Marine Living Resources) Area 48 throughout 890 three years of continuous sampling. 35 891 METHODS 892 Sample collection and storage 893 One hundred individual krill of random sex, age, length, and mass were collected daily from a 894 continuous live flow system on-board the Aker BioMarine fishing vessel, the FV Saga Sea, 895 from 1 January 2014 to 11 September 2016 operating in CCAMLR Sub-Areas 48.1, 48.2 and 896 48.3 (Fig. 2.1A, B). 36 897 898 899 900 Figure 2.1. Convention on the Conservation of Antarctic Marine Living Resources (CCAMLR) Statistical Area 48 and its Sub-Areas (from http://www.fao.org/fishery/area/Area48/en accessed 12/2/16) (A). Locations of collections of Euphausia superba by FV Saga Sea (Aker BioMarine) from January 2014 to September 2016 (B). 37 901 Krill samples were snap-frozen upon harvesting and stored at −80 °C. Samples were 902 either packaged in aluminium foil in ten groups of ten krill (January 2014–September 2014), 903 or in vacuum packaging in two groups of ten krill and one bulk bag of ~80 krill (December 904 2014–September 2016). All samples were labelled with the date of their collection and bar- 905 coded, linking them to their geographical location at the time of their collection. The krill were 906 then freighted on dry ice to Hobart, Tasmania, then stored at −80 °C for later analysis. 907 The length of all krill was measured using ‘Standard 1’ from Kirkwood (1982), sexed 908 using a dissecting microscope and weighed (wet mass) prior to analysis. Dry mass was 909 calculated by multiplying the wet mass by 0.2278 (Virtue et al., 1993a). Only adult or sub- 910 adult krill (30 – 60 mm in length) were used in the analyses presented here. All juvenile krill 911 remained frozen for later analysis. The maturity stages of krill were obtained for 2015 and 2016 912 samples only, therefore krill were categorised as “male” or “female” and maturity stages were 913 not used for interannual comparisons. Once sexed, three females and three males were analysed 914 in two-week periods (N = 6), where period 1 started on 1 January 2014 and ended 15 January 915 2014. All subsequent periods followed on from this date. The variance observed within a two- 916 week period was examined by sampling the total lipid content of 10 random krill (five from 917 each sex) from within the same period and undertaking a one-way ANOVA for power analysis. 918 This process indicated that three individuals of each sex produced a representative sample from 919 within that period, giving the statistical power required for analysis. 920 Lipid-extraction technique 921 Whole krill samples were quantitatively extracted overnight using a modified Bligh and Dyer 922 (1959) method consisting of a methanol:dichloromethane:water (MeOH:CH2Cl2:H2O) solvent 923 mixture (20:10:7 ml). Phase separation was achieved the next day by adding 10 ml CH2Cl2 and 924 10 ml Milli-Q H2O (saline), giving a final methanol:dichloromethane:water solvent ratio of 925 1:1:0.85. The lower layer was drained and the total lipid (TL) was concentrated using rotary 38 926 evaporation. The TL was transferred into a pre-weighed 2 ml vial and the solvent was blown 927 down under nitrogen gas to quantify total lipid content mass (mg). Solvent was added until 928 further procedures were carried out to avoid oxidation. 929 The lipid class composition of samples was determined of the krill total lipid content 930 using an Iatroscan MK-5 TLC/FID analyser (Iatron Laboratories, Tokyo, Japan). A standard 931 solution sourced from Sigma; with known quantities of wax esters (WE), triacylglycerols 932 (TAG), free fatty acids (FFA), sterols (ST), and phospholipids (PL), was used to calibrate the 933 flame ionisation detector, with hydrocarbon (HC) (squalene) also used in a separate solution. 934 Hydrocarbons (HC) and WE and steryl esters (SE) were combined for all statistical analyses 935 (HC refers to this combination hereon), as these were only minor peaks and not all samples 936 contained all of these lipid classes. 937 Aliquots (1µl) of the total lipid extract were spotted on chromarods and developed in 938 accordance to Sutton (2015), with slight modifications for krill lipid classes such as a solvent 939 system of hexane:diethyl-ether:acetic acid (90:10:0.1 ml) and a drying time of 5 min at 50 °C. 940 Total lipid content per gram of krill mass is expressed as mg g−1 and abbreviated as TL DM 941 (milligrams of total lipid extract per gram of krill dry mass). 942 Statistical analyses 943 The R statistical package (R Core Team, 2017) used for analysis was R version 3.4.2 (2017 - 944 09-28). The platform x86_64-w64-mingw32/x64 (64 bit) was utilised for all analyses. The R 945 packages used to undertake the statistics and create the figures were reshape2 (Wickham, 2007), 946 ggplot2 (Wickham, 2009), gapminder (Bryan, 2015), effects (Fox, 2003), gplots (Warnes, 947 2016), ggpmisc (Aphalo, 2016), devtools (Wickham, 2017), knitr (Xie, 2018), maps (Becker, 948 2017), pwr (Champely, 2018), and lattice (Sarkar, 2008). 949 Total lipid, TL DM, and lipid class quantitative and percentage data for each season 950 were analysed in the RStudio statistics package (version 0.99.893) using a multifactorial 39 951 ANOVA, with sexual maturity and year as factors. Type 3 SS analyses were used to check 952 statistical outputs for data levels that were unbalanced, but did not significantly alter output 953 results, so Type 1 SS were used. To identify significant differences between factor levels, 954 Tukey post-hoc comparisons were used. Log or square root transformations were utilised when 955 data did not meet assumptions of normality. Location was not included as a factor as the fishing 956 locations differed between seasons and years. Data in tables are expressed as mean ± standard 957 deviation. For all analyses, α was set at 0.05. 958 RESULTS 959 Krill length and mass 960 The mean body length (Standard length 1; Kirkwood (1982)) of krill was 46.0 mm (± 4.8) and 961 the mean dry mass was 0.16 g (± 0.05). Krill were longest in autumn (47.43 mm ± 3.76 mm) 962 and were generally lighter and shorter during winter (0.264 g ± 0.09 g and 44.73 mm ± 4.26 963 mm) and heavier in summer (0.362 g ± 0.12 g). Krill sampled in the winter of 2015 were 964 heavier and longer than krill in the previous or subsequent winters (p: 0.03; Table 2.1). Krill 965 length and mass followed a seasonal trend (p < 0.001; Table 2.1). Krill mass and length 966 significantly correlated with krill total lipid content (r2 = 0.87 and 0.93, respectively, p < 0.001; 967 Table 2.1). 40 968 969 970 971 Table 2.1: Average total lipid (mg g–1 dry mass, mean ± SD), length (mm) and mass (g) of Euphausia superba by sex, season and year. Seasons are defined as summer (1 December to 28 February), autumn (1 March to 31 May), winter (1 June to 31 August), and spring (1 September to 30 November). Total lipid content Average Average (mg g–1 dry mass) length dry mass Females (N = 201) (mm) (g) Males (N = 190) Summer 2014 148.9 ± 116.2 148.0 ± 58.7 48.19 0.21 Autumn 2014 329.8 ± 68.3 268.6 ± 101.3 47.23 0.17 Winter 2014 210.6 ± 86.2 203.8 ± 66.4 42.04 0.11 Spring 2014 59.8 ± 12.7 114.8 ± 30.0 45.66 0.14 Summer 2015 168.1 ± 130.1 166.7 ± 102.1 46.49 0.17 Autumn 2015 309.2 ± 60.8 303.8 ± 74.2 45.63 0.14 Winter 2015 234.1 ± 60.7 233.5 ± 64.3 48.42 0.17 Spring 2015 138.9 ± 17.2 131.3 ± 16.5 48.95 0.20 Summer 2016 271.8 ± 128.3 217.2 ± 94.3 46.77 0.19 Autumn 2016 399.7 ± 51.8 361.3 ± 96.6 48.51 0.21 Winter 2016 209.0 ± 46.1 208.0 ± 57.2 44.01 0.13 Spring 2016 90.7 ± 26.6 120.1 ± 9.3 42.97 0.13 972 41 973 Total lipid content 974 Total lipid content varied significantly between seasons (p < 0.001; Fig. 2), years (p: 0.006; 975 Fig. 2), and between sexes (p: 0.039; Table 2.1), but these variations were not evident in the 976 sex and season interaction (p: 0.167; Table 2.1). A clear seasonal trend in lipid content of krill 977 was seen across all years (Fig. 2.2), with the highest lipid content measured during autumn 978 (508 mg g–1 TL DM) and the lowest during summer (41 mg g–1 TL DM). Total krill lipid 979 content followed the same trend across all three years, increasing from summer to an autumn 980 high, before declining throughout winter into spring lows. Total lipid in samples from each 981 season were distinctly different from each other, except for summer and winter as total lipid 982 levels in these two seasons were similar though they were increasing in summer and decreasing 983 in winter (p: 0.113). The absolute levels in total lipid for each season differed between years 984 (Table 2.2). The season and year interaction had a significant effect on the total lipid content 985 (p < 0.001; Fig. 2.2), but this was driven mostly by the differences between 2014 and 2016 (p: 986 0.004). The differences between 2014 and 2016 were driven by their summer and autumn TL 987 DM values being so dissimilar (p < 0.0001 for both); similarly, total lipid in autumn 2016 was 988 significantly higher than in autumn 2015 (p: 0.018; Table 2.1). 42 989 990 991 992 Table 2.2: Lipid content and lipid class composition in Euphausia superba expressed by year and season (mean). Seasons are defined as: summer (1 December to 28 February), autumn (1 March to 31 May), winter (1 June to 31 August) and spring (1 September to 30 November). HC, hydrocarbons (includes steryl esters and wax esters present in trace amounts); TAG, triacylglycerols; FFA, free fatty acids; ST, sterols; DAG, diacylglycerols; PL, phospholipids. Lipid class composition (%) Total lipid Total lipid N mass (mg) (mg g–1 dry mass) HC TAG FFA ST DAG PL Unknown Summer 2014 33 33.6 148.4 1.4 33.2 9.7 4.7 0.9 48.9 0.1 Autumn 2014 45 54.1 300.6 0.7 45.9 2.8 2.4 1.2 47.0 0.1 Winter 2014 44 23.6 201.9 1.3 35.6 1.4 1.8 0.6 59.2 0.0 Spring 2014 6 12.1 87.3 0.8 28.1 1.9 2.4 0.4 66.3 0.0 Summer 2015 30 30.9 183.9 0.9 31.2 1.4 2.2 1.0 62.6 0.7 Autumn 2015 42 44.9 315.1 1.6 42.1 0.1 2.7 2.8 50.5 0.0 Winter 2015 38 42.9 233.8 1.1 40.9 0.2 2.3 0.7 54.6 0.0 Spring 2015 12 27.4 135.1 1.1 34.9 0.8 2.5 0.7 59.9 0.0 Summer 2016 44 45.5 243.6 0.6 39.1 2.5 2.2 1.5 53.9 0.3 Autumn 2016 37 82.6 379.9 0.4 46.6 0.5 2.7 2.9 46.7 0.2 Winter 2016 38 27.4 208.5 0.5 41.9 0.7 1.5 0.9 54.4 0.1 Spring 2016 12 14.0 105.4 0.5 27.6 1.2 1.6 0.5 68.5 0.0 43 993 994 995 996 997 Figure 2.2. Total lipid content (mg g–1 of dry mass) of Euphausia superba showing the year and location of sample collection. The boxes from left to right show the two-week period of sample collection (where period 1 is January 1 to 15, 2014). Each box is the combination of three male and three female Euphausia superba from that period (N = 6). Each box represents 1 SD, with the whiskers the second SD and the bold line the mean. WAP, West Antarctic Peninsula; SOI, South Orkney Islands; SG, South Georgia. 44 998 The differences seen between sexes in their TL DM were variable over time (Table 2.1). 999 All krill sampled in summer showed similar values for TL DM, except 2016, where males had 1000 higher TL DM than females (Table 2.1). Males exhibited higher average TL DM than females 1001 in autumn 2014 and 2016; however, no difference between sexes was seen in autumn 2015 (p: 1002 0.155; Table 2.1). Winter krill showed no difference between the sexes (p: 0.999). Spring krill 1003 exhibited more variability, with females in 2014 and 2016 showing higher average TL DM 1004 than males, although no difference between the sexes was seen in spring 2015 (Table 2.1). This 1005 pattern of difference between the sexes was significant (p: 0.004), but the Sex*Season 1006 interaction (p: 0.167), Sex*Year interaction (p: 0.966), and Sex*Season*Year interactions (p: 1007 0.704) were not. The analyses of female and male krill were therefore combined for all 1008 subsequent analyses on total lipid content as they had exhibited the same interannual and 1009 seasonal patterns in their total lipid content. 1010 Lipid class composition 1011 Lipid class percentages (expressed as % of TL) varied between seasons and years. There were 1012 significant differences in TAG levels between years (p: 0.028), seasons (28% in spring 2016 1013 to 47% in autumn 2016; p < 0.001) and sex (p: 0.01). A significant Year*Season interaction (p: 1014 0.002; Table 2.2) as well as a strong Sex*Year effect (p < 0.001) and Sex*Season*Year 1015 interaction (P = 0.005; Table 2.1) could be seen. Spent female krill (having spawned their eggs) 1016 had different levels of TAG compared to other female maturity stages and males (p: 0.02; Table 1017 2.3). 45 1018 1019 1020 Table 2.3: Lipid composition of Euphausia superba by sex and maturity stage for 2015–2016 samples (%, mean ± SD). HC, hydrocarbons (includes steryl esters and wax esters in trace amounts); TAG, triacylglycerols; FFA, free fatty acids; ST, sterols; DAG, diacylglycerols; PL, phospholipids; 3M, mature males; 3F, mature females; 3F-G, gravid females; 3F-S, spent females; 2F, sub-adult females; 2M, sub-adult males. Total lipid (mg Lipid class composition (%) Sex N g–1 dry mass) HC TAG FFA PL Unknown 3M 62 239.9 ± 128.1 0.5 ± 0.3 37.9 ± 11.8 1.4 ± 2.9 1.4 ± 1.7 1.7 ± 4.2 56.1 ± 11.5 0.2 ± 0.4 3F 37 216.3 ± 100.5 0.5 ± 0.4 37.2 ± 9.0 1.7 ± 3.1 1.8 ± 0.9 0.9 ± 0.7 57.5 ± 9.1 0.1 ± 0.2 3F-G 5 113.3 ± 39.4 0.3 ± 0.1 25.1 ± 18.8 1.8 ± 1.5 2.6 ± 0.9 0.8 ± 0.8 68.1 ± 16.4 1.1 ± 1.0 3F-S 4 96.7 ± 43.0 1.0 ± 0.6 19.7 ± 18.6 2.6 ± 2.3 2.9 ± 1.4 0.7 ± 0.5 70.3 ± 13.4 2.1 ± 2.3 2F 94 254.5 ± 100.1 0.6 ± 0.4 40.6 ± 7.3 0.8 ± 1.2 1.9 ± 2.0 1.1 ± 1.0 54.6 ± 7.1 0.2 ± 0.5 2M 51 257.4 ± 106.0 0.5 ± 0.4 38.9 ± 9.5 1.3 ± 1.5 1.8 ± 1.3 1.1 ± 1.2 55.9 ± 8.2 0.3 ± 1.1 1021 46 ST DAG 1022 The dominant lipid class for all krill samples was PL (58 ± 7.4%), which had the widest 1023 range of values between samples (23 to 88%; Table 2.2). Levels of PL in krill varied 1024 significantly with sex (p: 0.018; Table 2.3) and season (p: 0.041) and showed strong Year*Sex 1025 effects (p: 0.025; Table 2.3). Levels of PL in the summer and spring of 2015 were significantly 1026 different to those in 2014 and 2016 (p < 0.001; Table 2.2), whereas PL levels in all other 1027 seasons and years were not (p > 0.05 for all). PL levels in gravid females and spent females 1028 were not significantly different, but PL levels in both gravid and spent females differed from 1029 sub-adults and other mature krill of both sexes (Table 2.3). 1030 When all seasons were combined, there was a relationship between PL and TAG (y = 1031 −0.835x + 86.955, where y is the phospholipid percentage and x is the triacylglycerol 1032 percentage). This relationship is much stronger and has a better fit in the winter (r2 = 0.956) 1033 and spring (r2 = 0.94) than in the summer (r2 = 0.618) and autumn (r2 = 0.651) (Fig. 2.3). The 1034 greater variation in this relationship in the summer was driven by 2014 samples, which had a 1035 much greater spread of values. 47 1036 1037 1038 1039 Figure 2.3. The seasonal relationships between phospholipids and triacylglycerol (as % of the total lipid content) from fortnightly samples of Euphausia superba collected from January 2014 to September 2016. 48 1040 FFA levels also did not significantly differ between years and seasons. Male krill had 1041 a significantly lower FFA level than all female stages except spent females (p < 0.001). FFA 1042 had large Year*Sex and Season*Sex interactions (p < 0.001 for both sexes; Table 2.3). FFA 1043 levels also exhibited a significant Year*Sex*Season interaction (p: 0.013; Tables 2.2 and 2.3), 1044 with differences between sexes and all maturity stages particularly in 2014. Sterol (ST) levels 1045 also showed a significant sex effect (p < 0.001). Whilst levels of diacylglycerol (DAG) in krill 1046 were fairly consistent (Table 2.2), there were a few outliers early in 2014 and one two-week 1047 period of significantly higher DAG levels in 2016 which were driven by higher levels in mature 1048 male krill. 1049 The levels of hydrocarbons (HC) were significantly different between years, seasons 1050 and sexes (p < 0.001 for all) when tested independently by ANOVA. When analysed using a 1051 multifactorial ANOVA and passed through Type III SS and F tests, however, there were no 1052 significant differences or interactions (p > 0.1 for all; Tables 2.2 and 2.3). No larger pattern is 1053 evident even if a Tukey test showed some significant differences between individual seasons 1054 or years or sexes. 1055 Quantitative lipid class analysis 1056 The absolute mass of TAG (mg) in individual krill varied seasonally and interannually (Fig. 1057 2.4A), with TAG content being more variable in the summer and autumn. The variations align 1058 more closely to the seasonal and interannual trend seen in the lipid class composition data when 1059 TAG was corrected for by krill mass (and expressed as mg g–1 DM) (Fig. 2.4B). FFA content 1060 increased slightly in samples collected at the beginning of 2016. ST content was highly stable 1061 across all three years and between seasons, whether looking at absolute ST mass, or the ST 1062 mass scaled to krill mass or total lipid content. PL showed the same seasonal and interannual 1063 trend as in the percentage data when looking at the absolute mass of PL (Fig. 2.4C). The 49 1064 anomalous highs in autumn 2016 are captured in this trend, but not when PL are standardised 1065 against krill dry mass (Fig. 2.4D). 50 1066 1067 1068 1069 1070 Figure 2.4. Quantitative trends in the major lipid classes during 2014 to 2016; TAG mass (mg per krill) (A). TAG of krill dry mass (mg g–1) (B). PL mass (mg per krill) (C). PL of krill dry mass (mg g–1) (D). Each box is the combination of three male and three female Euphausia superba collected from a two-week period (N = 6). The first period of sample collection was January 1–15, 2014. Each box represents 1 SD, with the whiskers the second SD and the bold line the mean. TAG, triacylglycerols; PL, phospholipids. 51 1071 The seasonal trend seen in TAG content (Fig. 2.4A) was muted significantly when TAG 1072 content was scaled to total lipid content (Fig. 2.5A). Such a result shows that the amount of 1073 total lipid content in the krill affects the quantity of TAG content more than krill size (g) does. 1074 Although the same seasonal trend is seen in PL when data are scaled to krill mass, the trend 1075 seen in PL all but disappears once the amount of total lipid content is used for scaling (Figs. 1076 2.4D and 2.5B). No discernible trend in the PL content as a proportion of total lipid content 1077 could be detected, although this may be due to the high levels of variability throughout 2014 1078 and early 2015 (Fig. 2.5B). 52 1079 1080 1081 1082 Figure 2.5. Triacylglycerol (TAG) mass (mg) as a proportion of the total lipid mass (mg) of Euphausia superba (A). Phospholipids (PL) mass (mg) as a proportion of the total lipid mass (mg) of Euphausia superba (B). Each box represents 1 SD, with the whiskers the second SD and the bold line the mean. 53 1083 DISCUSSION 1084 Our study has resolved the sinusoidal shape of the seasonal trend in krill lipid content and 1085 composition by utilising high-resolution fisheries-derived samples. Krill lipids exhibit a clear 1086 seasonal trend in their total content and in their component lipid classes. The seasonal 1087 differences observed in the major lipid classes reflect their differing roles in key physiological 1088 and biochemical processes: growth, storage, and reproduction. These seasonal cycles have 1089 become clearer through the use of high-resolution sampling. 1090 Effect of Sex on the lipid seasonal cycle in krill 1091 The lipid content and the relative proportions of lipid classes in krill differed 1092 significantly between seasons. These seasonal trends have similarly been seen at a coarser 1093 resolution in Antarctic krill (Clarke, 1984, Fricke et al., 1984, Quetin et al., 1994, Phleger et 1094 al., 2002, Hagen and Kattner, 2014). Pond et al. (1995) and Ju and Harvey (2004) reported that 1095 lipid accumulation is tightly linked to seasonal factors such as the timing of reproduction in 1096 krill. Such differences were also shown in the northern krill Meganyctiphanes norvegica (Sars, 1097 1857) by Falk-Petersen (1981) and Cuzin-roudy et al. (1999), who reported that the timing of 1098 lipid accumulation in relation to reproduction was vital. 1099 The sex of krill had little effect on the amount of total lipid regardless of year or season. 1100 This result differs from the results of previous studies which reported that gravid female krill 1101 have higher total lipid content due to the development of the ovary prior to spawning in the late 1102 summer or early autumn (Virtue et al., 1996, Cuzin-roudy et al., 1999). Males at this time of 1103 year also had elevated total lipids (Table 2.1), indicating that extra energy stores may be 1104 required during the mating process as Virtue et al. (1996) suggested. The lower total lipid 1105 values in females compared to males during the spawning season might be due to females 1106 losing large quantities of lipid when they spawn (Table 2.3). Females, however, have higher 1107 lipid stores in spring leading into the spawning season (Table 2.1). 54 1108 The lipid content and lipid class composition of reproductive krill is well studied 1109 (Clarke, 1984, Pond et al., 1995, Virtue et al., 1996, Mayzaud et al., 1998, Atkinson et al., 2002) 1110 but the differences between the sexes and how differing lipid content and composition impacts 1111 on their reproduction is not well understood. Mayzaud et al. (1998) reported that sub-adults 1112 and females had a similar wet weight to TAG content relationship, whereas males showed a 1113 particularly different relationship. Age-dependent use of lipid stores has been reported at the 1114 onset of winter, with adult krill having higher lipid stores than sub-adults but similar lipid class 1115 compositions (Atkinson et al., 2002). Female krill from South Georgia had higher lipid levels 1116 on average and were longer and heavier than male and sub-adult krill in summer (Pond et al., 1117 1995). Tarling et al. (2016a) reported that growth and shrinkage in krill across the South 1118 Atlantic was sex-dependent. This suggests that mature males and females in the spawning 1119 seasons should have sex-dependent growth, although this may not necessarily match their sex- 1120 dependent lipid content and composition. 1121 Effect of krill length and mass on the seasonal lipid cycle 1122 Total lipid content varied with krill mass and length, with bigger krill having higher 1123 overall lipid content than smaller ones (as reported by Falk-Petersen et al. (2000) and Gigliotti 1124 et al. (2011)). We found that the longest krill were not always the heaviest. Krill in summer 1125 were heavier than the longest krill in autumn except in 2016. This may be due to many krill in 1126 summer being gravid females heavy with eggs. Post-spawn females were larger but lighter than 1127 gravid females due to the loss of egg mass. If lipid contents in krill are proportional to size, as 1128 well as maturity stage in females, this could be significant for recruitment and energy flow in 1129 the krill life cycle. 1130 Krill were significantly smaller and leaner in the summer of 2014; these krill also 1131 generally showed higher PL and lower TAG levels (Table 2.2). These low TAG levels could 1132 be due to the summer of 2014 being exceptionally warm (0.57 °C warmer than average for the 55 1133 global oceans; NOAA (2013) and Wiedermann et al. (2016)), causing food supplies to be lower 1134 and diatoms to be less abundant (Ericson et al., 2018a). A change from a predominantly diatom 1135 diet to a dinoflagellate diet in M. norvegica halted ovary development due to the reabsorption 1136 of lipids from ovaries, which was mostly TAG (Cuzin-Roudy et al., 2004). Significantly lower 1137 levels of TAG could be a sign of poor krill health and therefore recruitment in the summer of 1138 2014 as TAG, along with PL, is one of the major energy sources in krill. 1139 Triacyclglycerol and Phospholipid seasonal cycles in krill 1140 The seasonal levels of TAG followed the sinusoidal trend seen in total lipid content 1141 with a peak in autumn. PL also had a seasonal response but with the peak occurring in spring 1142 (Table 2.2). The relationship between PL and TAG was highly variable in all summer samples. 1143 This variability may be due to differences in the timing and spatial extent of algal blooms after 1144 the spring melt (Skerratt et al., 1995, Janout et al., 2016). Such effects have been suggested for 1145 northern krill (Falk-Petersen, 1981). Gravid krill have much higher levels of TAG than spent 1146 females because krill eggs are high in TAG. Observed TAG levels in female krill would thus 1147 be dependent on the timing of krill spawning. 1148 TAG, however, is a storage lipid and would be expected to fluctuate according to the 1149 energetic needs of the krill and the availability of food throughout the year (Hagen, 1996, 1150 Atkinson et al., 2002). Krill increase their TAG levels throughout late spring and summer when 1151 food is abundant, causing TAG levels to rise and be at their highest in autumn (Hellessey et al., 1152 2018). Krill use these stored lipids over winter and hence TAG levels are at their lowest at the 1153 end of spring when food starts to become more abundant again (Korb et al., 2005, Vernet et 1154 al., 2008, Schmidt et al., 2012, Kohlbach et al., 2018). Food supplies increase throughout spring 1155 and summer, and krill have a higher energy demand to fuel reproduction and growth during 1156 this time of year (Kawaguchi, 2016), so storing lipids at this same time is unwise. TAG levels 1157 have the same sinusoidal pattern seen in total lipid content and underlie the changes in total 56 1158 lipid content at the seasonal scale. PL levels, however, showed the reverse seasonal relationship 1159 to TAG. A decrease in PL levels in autumn is accompanied by a rise in TAG from summer 1160 levels. When krill are at their smallest (shortest and lightest) during winter, their PL levels are 1161 low, whilst their total lipid content and TAG levels are still high. Similarly, krill are large in 1162 summer, and their PL levels are high, whilst their TAG levels and total lipid contents are low. 1163 Phospholipids are vital for healthy cell membranes and as krill grow and accumulate 1164 more cells, their PL levels would increase proportionally. Because of the ability of krill to 1165 shrink and re-grow to full size (Ikeda and Dixon, 1982, Tarling et al., 2016a), there are likely 1166 to be fluctuations in their PL levels accompanying their growth and shrinkage. The length and 1167 mass of krill were significantly correlated with PL levels and these inversely followed the peaks 1168 and troughs of the seasonal cycle of their lipid content (Tables 2.1 and 2.2). PL may be 1169 conserved preferentially as it is used in the structure of krill cell membranes. Krill in this study 1170 had inconsistent PL content throughout the year and it was not conserved preferentially. PL 1171 has been suggested as a storage fat (Hagen et al., 1996, Daly, 2004, Ju and Harvey, 2004) and 1172 the inconsistent content of PL seems to suggest the same. PL levels are also important in the 1173 production of krill oil rich in PL-containing omega-3 fatty acids. 1174 There was one exception to the proportional seasonal relationship, between the content 1175 of PL and TAG. Krill had disproportionally low TAG levels in summer. Greater increases in 1176 TAG levels from the spring could be driving this pattern, but further sampling throughout the 1177 spring would be required to clarify this. 1178 Sterol content showed little seasonal trends and was highly consistent whether in terms 1179 of its total mass, or when scaled to krill mass or total lipid content (Table 2.2). This observation 1180 suggests that sterols may not be used as much as previously thought for the mobilisation of 1181 krill lipids, which is why these levels are consistent throughout the krill life cycle and across 1182 years and seasons. 57 1183 Our results also show a clear seasonal trend and some interannual variation in adult krill 1184 lipid profiles. TAG content followed the same seasonal pattern as TLDM and could be shown 1185 to behave as a storage lipid alongside PL in winter in adult krill. The seasonal trends of lipid 1186 content and composition in juvenile krill are not well understood with limited available data 1187 regarding interannual trends (Atkinson et al., 2002, O’Brien et al., 2011, Virtue et al., 2016, 1188 Schaafsma et al., 2017). Future analyses of fishery-derived samples may assist with 1189 understanding lipid trends in juvenile krill. 1190 Our study also indicates the utility of samples collected by the krill fishery. Other 1191 studies have utilised data from such samples, but these have not been at a high level of temporal 1192 resolution (Kim, 2014, Tarling et al., 2016b). Fisheries samples are not ideally collected for 1193 ecological studies and because our samples were collected across the Scotia Sea, and in 1194 different seasons, our results may be confounded to a degree. Separating the spatial and 1195 temporal elements from the lipid data will require more samples from multiple locations at the 1196 same time of year, or year-round sampling from a single location (e.g. South Georgia, which 1197 is accessible year-round). Alternatively, a combination of sampling by multiple fishing vessels 1198 and by scientific research vessels may ensure complete seasonal or regional coverage. 1199 ACKNOWLEDGEMENTS 1200 This project is performed under the funding and approval of the ARC Linkage Project 1201 LP140100412, in partnership with Aker BioMarine, Commonwealth Scientific and Industrial 1202 Research Organisation (CSIRO), Australian Antarctic Division, University of Tasmania, and 1203 Institute for Marine and Antarctic Studies. We thank Peter Mansour, Mina Brock, Andy Revill, 1204 and Ben Gaskell who helped us immeasurably in the CSIRO lipid laboratory. We gratefully 1205 acknowledge the captain and crew of FV Saga Sea for their time and great care in collecting, 1206 packaging, and storing krill samples. We would also like to thank Stig Falk-Petersen and two 1207 anonymous reviewers for their useful comments and suggestions to the manuscript. 58 1208 Chapter 3: Regional variability of Antarctic krill (Euphausia superba) diet 1209 during the late-summer as determined using lipid, fatty acid and sterol 1210 composition 1211 1212 This Chapter is currently under review at Polar Biology. 1213 1214 ABSTRACT 1215 Antarctic krill (Euphausia superba) are a circumpolar species with an omnivorous diet. 1216 Knowledge of krill diet in different oceanic regions will help predict how regional-scale 1217 environmental change may impact local krill populations. Krill from the Atlantic, Indian and 1218 Pacific sectors of the Southern Ocean were compared. The total lipid, lipid class, and neutral 1219 fraction fatty acid and sterol content and composition of whole krill, their digestive glands and 1220 stomachs during the late-summer were examined. Krill from the Indian sector had a distinctly 1221 different diet to the Atlantic and Pacific sectors based on their fatty acid profiles (p: < 0.001). 1222 Indian sector whole krill had higher phospholipids (55.0 ± 8.9%, as % total lipids) compared 1223 to Pacific (45.9 ± 3.6%) and Atlantic sector krill (43.7 ± 8.2%). Indian sector krill digestive 1224 glands showed lower phospholipid levels (Indian: 29.4 ± 8.5%, Pacific: 52.5 ± 5.7%, Atlantic: 1225 52.5 ± 5.9%). Indian sector krill had a more carnivorous and diatomaceous diet (higher levels 1226 of 16:1n-7c, 14:0 and 20:1 and 22:1 isomers), with less flagellate input (lower 18:4n-3, 21:5n- 1227 3 and 18:3n-6) than other regions. One site in the Indian sector had particularly high 22:6n-3. 1228 Indian Ocean sector krill had lower cholesterol levels in their stomachs (52.5 ± 14.1%, as % 1229 total sterols) than Pacific and Atlantic sector krill stomachs (62.8 ± 1.9 % and 60.9 ± 4.9%, 1230 respectively). This study is, to our knowledge, the first to detail the regional differences in late- 59 1231 summer krill diet by assessing the lipid, neutral fraction fatty acid and sterol content and 1232 composition of different tissue types. 1233 INTRODUCTION 1234 Antarctic krill (Euphausia superba, hereon krill) are a keystone species in the Antarctic 1235 ecosystem (Murphy et al., 2007, Barnes and Tarling, 2017). Krill are extremely lipid rich (up 1236 to 40% of dry mass in winter (Hagen et al., 2001, Atkinson et al., 2002)), and are therefore 1237 very energy dense, making them an ideal food source for other animals in the harsh Antarctic 1238 environment. Krill have a circumpolar distribution and live in a variety of environments 1239 (Atkinson et al., 2008, Atkinson et al., 2009, Jarvis et al., 2010, Kawaguchi et al., 2010b, 1240 Leonori et al., 2017). Krill lipid profiles have been used to detect their health, condition and 1241 diet (Schaafsma et al., 2017, Ericson et al., 2018a, Hellessey et al., 2018). It is unclear how 1242 lipid profiles vary with shorter term dietary and environmental changes. 1243 Krill diet varies seasonally and regionally (Ericson et al., 2018a) and with maturity 1244 stage (Virtue et al., 1997), although these observations generally have been based on analyses 1245 of the lipid profiles from whole krill. Such observations may not reflect the shorter term diet 1246 of the krill which can be better observed in measurements from individual body parts such as 1247 the digestive gland and stomach (Mayzaud et al., 1998, Schmidt and Atkinson, 2016, 1248 Schaafsma et al., 2017). 1249 The fatty acid profile in the neutral lipid fraction, containing mostly the storage lipid 1250 triacylglycerol, reflects the diet to a greater degree than fatty acids in phosphoplipids or total 1251 lipids. As triacylglycerol is accumulated during feeding, the use of neutral lipid fatty acid 1252 profiles of digestive and stomach tissue, rather than whole animal tissue, allows for even 1253 greater understanding of dietary components. To date only a few studies have been conducted 1254 using fatty acid markers from the neutral lipid fraction, however, these studies used whole krill 1255 tissue (Pond et al., 1995, Ju and Harvey, 2004, Schaafsma et al., 2017). Only one study by 60 1256 Cabrol et al. (2019) has looked at the fatty acids in neutral lipid fractions of different tissue 1257 types in Northern Atlantic krill to determine diet. No studies to date, however, have 1258 investigated the neutral lipid fraction in different krill tissue types for Antarctic krill. 1259 Krill accumulate lipids in the austral summer during the spawning season and these 1260 reach their highest concentrations in autumn (Hellessey et al., 2018). Krill are known to have 1261 divergent lipid class stores and fatty acid profiles depending on their sex (Mayzaud et al., 1998) 1262 and particularly if they are gravid or are post-spawn females (Mayzaud et al., 1998, Hellessey 1263 et al., 2018). Krill lipid profiles are also a function of their diet, however, differences in krill 1264 diet at a circumpolar scale at this time of year are generally unknown. The lipids in krill are 1265 utilised by both the megafauna that prey on krill (Mori and Butterworth, 2004, Trivelpiece et 1266 al., 2011) and the commercial krill fishery (Nicol et al., 2012, Kwantes and Grundmann, 2015). 1267 Although predation by vertebrates is highest in the summer months, the krill fishery is most 1268 active in autumn and winter (Nicol et al., 2012). 1269 Currently the Antarctic krill fishery is largely based in the Atlantic sector of the 1270 Southern Ocean despite being almost circumpolar in nature when it first came into operation 1271 in the 1970’s (Hill et al., 2016, CCAMLR, 2017). There has been little recent krill fishing in 1272 the Pacific and Indian sectors of the Southern Ocean (Nicol and Foster, 2016, CCAMLR, 2017). 1273 The South Atlantic, especially the West Antarctic Peninsula, is an area of extremely 1274 rapid environmental change and is the fastest warming region in the world (Ducklow et al., 1275 2007), having experienced a midwinter surface air temperature warming of 5 – 6 °C in the last 1276 50 years (Rayner, 2003, Ducklow et al., 2007). An associated 40% reduction in sea ice duration, 1277 extent and concentration has been recorded over the last 26 years (Ducklow et al. (2007)). This 1278 has led to a community level shift in the lower trophic levels (Deppeler and Davidson, 2017, 1279 Hancock et al., 2018), particularly in the primary producers which krill feed upon. This 61 1280 reduction of sea ice and its associated communities are likely to affect krill diet and population 1281 dynamics. Sea ice variability has been associated with krill recruitment and population stability 1282 (Brierley et al., 2002, Wiedenmann et al., 2009, Schmidt et al., 2014), because juvenile krill 1283 are thought to rely on sea ice algae during their first winter (Ross and Quetin, 2000). The 1284 combination of reduced sea ice and warming seas in this region may affect both the short- and 1285 long-term diet of krill and thus their health and ability to reproduce (Kawaguchi, 2016). Lipid 1286 analysis can be a useful tool to investigate any shift in krill diet and condition (Virtue et al., 1287 2016, Schaafsma et al., 2017, Ericson et al., 2018a). 1288 The Indian and Pacific sectors of the Southern Ocean have lower temperatures and more 1289 stable sea ice coverage than the South Atlantic sector (Nicol et al., 2000c, Ducklow et al., 2007, 1290 Barnes and Tarling, 2017). Both the Indian and Pacific sectors harbour large populations of 1291 krill, but these populations have not been as well studied as those in the southwest Atlantic. 1292 Circumpolar krill studies use data from disparate time points (different months and 1293 seasons) or from large datasets using varying methodologies (e.g. KRILLBASE (Atkinson et 1294 al., 2017)) and these circumpolar datasets rarely have information on krill biochemistry or diet. 1295 Understanding regional differences in krill diet will assist in identifying how krill cope 1296 with varying regional environmental conditions. Our aims were: 1) to investigate krill diet 1297 using neutral fraction fatty acid and sterol content and composition of whole krill, 2) to use 1298 both krill stomach and digestive gland samples as comparisons for short term diet vs long term 1299 diet in whole krill samples and 3) to investigate differences in krill diet from samples collected 1300 in all three ocean basins surrounding Antarctica during the late-summer season. We 1301 hypothesise that krill from the three different oceanic basins around Antarctica will have vastly 1302 different diets at the crucial late-summer spawning period. 62 1303 METHODS 1304 Krill sample collection 1305 We utilised krill from three sources. Krill were collected by Aker BioMarine on-board FV Saga 1306 Sea throughout February 2016 from the Atlantic sector of the Southern Ocean (CCAMLR Area 1307 48 (Figure 3.1)) using a harvesting technology that continuously pumps krill from the cod-end 1308 of a submerged net (Hellessey et al., 2018). After capture, krill were individually stored on- 1309 board at −80 °C. Krill were collected on-board the RSV Aurora Australis throughout February 1310 2016 as part of the K-Axis voyage from the Indian Ocean sector of the Southern Ocean 1311 (CCAMLR Area 58 (Figure 3.1)) using a Rectangular Midwater Trawl and stored individually 1312 on-board at −80 °C. Krill were collected on the RSV Akademik Treshnikov throughout 1313 February 2017 as part of the Antarctic Circumpolar Expedition in the Pacific Ocean sector of 1314 the Southern Ocean (CCAMLR Area 88 (Figure 3.1)) using a Bongo net. Krill were stored on- 1315 board at −20 °C, but showed no signs of lipid or fatty acid degradation (total free fatty acids: < 1316 5% in the whole krill) due to this temperature storage difference. All krill samples were 1317 transported on dry ice (approx. −80 °C) to Hobart, Australia. 63 1318 1319 Figure 3.1: Euphausia superba sample collection locations coloured by their Southern Ocean basin and showing CCAMLR management areas. 64 1320 Sample preparation 1321 Individual krill were sized (Standard Length 1; Kirkwood (1982)), weighed and their maturity 1322 stage and sex were determined prior to analysis. Wet mass was converted to dry mass by 1323 multiplying by 0.2278 to account for the 77.2% water content in krill (Virtue et al., 1993a). 1324 Only adult or sub-adult krill were used in this study. 1325 Lipids in whole krill reflect the long-term feeding history (Schmidt and Atkinson, 1326 2016)); the digestive gland reflects shorter-term feeding history and the stomach lipids are 1327 reflective of what has been immediately eaten (Virtue et al., 1993a, Yoshida et al., 2009). Three 1328 gravid females, three spent females, three sub-adult females and three male krill were used for 1329 whole krill extractions (these whole krill samples included the digestive gland and stomach). 1330 The digestive glands were dissected from additional krill samples. Two digestive glands were 1331 used per sample to ensure a large enough lipid fraction at the end of the extraction process. 1332 Similarly, the stomachs were dissected out, weighed and combined (12 stomachs per sample) 1333 for extraction. All dissections took place in a glass petri dish sitting in an ice bath. Krill were 1334 consistently sexed, weighed and dissected in under 5 minutes to keep thawing to a minimum 1335 whilst samples were on ice. 1336 Total lipid, fatty acid and lipid class extraction and analysis 1337 Samples were quantitatively extracted overnight using a modified Bligh and Dyer 1338 (1959) method as described in Hellessey et al. (2018) to produce the total solvent extract (TSE). 1339 The total lipid content (TL, expressed as mg) of each sample was weighed gravimetrically in a 1340 pre-weighed 2 ml glass vial. To account for differences in krill size, the TL was divided by 1341 krill dry mass (g) and is expressed as mg of total lipid content per gram of krill dry mass (mg 1342 g-1, TL DM). The TL DM was standardised by the number of krill per sample (whole krill = 1, 1343 digestive gland = 2, stomach = 12) to give the TL DM (mg g-1) per krill. Lipid class composition 1344 was determined by analysis of the TSE on an Iatroscan TLC-FID analyser following Hellessey 65 1345 et al. (2018). Briefly, an aliquot of the TSE was drawn up a 0.1 µl capillary tube and spotted 1346 onto glass chromarods and placed into a solvent bath (hexane (C6H14):diethyl ether (Et2O): 1347 acetic acid (CH3COOH), 90:10:0.1, v:v:v) for 25 minutes before being dried in an oven at 50 °C 1348 for 10 minutes. The chromarods were then run through an Iatroscan TLC-FID analyser (Parrish 1349 and Ackman, 1985). The identification and quantification of lipid classes (expressed as % total 1350 lipid content) was conducted in comparison to a known laboratory standard (sourced from 1351 Sigma) of wax esters (WE), triacylglycerols (TAG), free fatty acids (FFA), sterols (ST), and 1352 phospholipids (PL) and was used to calibrate the flame ionisation detector, with hydrocarbon 1353 (HC) (squalene) also used in a separate solution. Aliquots of sample TSEs were methylated to 1354 extract the fatty acid methyl esters (FAME) of the sample (Ericson et al. (2018a). In brief, an 1355 aliquot of the TSE was transferred to a glass test tube and methylated with 3 ml of solution 1356 (methanol (MeOH): dichloromethane (CH2Cl2): hydrochloric acid (HCl), 10:1:1, v:v:v). 1357 Samples were then heated to 90 – 100 °C for 75 minutes and cooled for 5 minutes before 1 ml 1358 H2O and 1.8 ml (C6H14:CH2Cl2, 4:1, v:v) solution were added to extract FAME. Samples were 1359 centrifuged for 5 minutes and the upper layer transferred to a vial before an additional 1.8 ml 1360 of C6H14:CH2Cl2 was added. This process was carried out 3 times to ensure all FAME were 1361 extracted, and the vial was kept under nitrogen (N2) gas between transfers. Sample vials were 1362 blown down to remove all solvents and were made up with 1 ml internal injection standard 1363 (23:0 FAME) (Iverson et al., 2004). Samples were analysed via gas chromatography using an 1364 Agilent Technologies 7890A GC-FID System (Palo Alto, CA, USA) equipped with a non-polar 1365 Equity.-1 fused-silica capillary column (15 m length x 0.1 mm internal diameter, 0.1 μm film 1366 thickness). Samples (0.2 μl) were injected in splitless mode at an oven temperature of 120 °C 1367 with helium as the carrier gas. The oven temperature was raised to 270 °C at a rate of 10 °C 1368 per minute, then to 310 °C at 5 °C per minute. Peaks were quantified using Agilent 1369 Technologies ChemStation software (Palo Alto, CA, USA) with initial identification based on 66 1370 comparison of retention times with known (Nu Chek Prep mix; http://www.nu-chekprep.com) 1371 and fully characterised laboratory (tuna oil) standards. Fatty acid peaks were expressed as a 1372 percentage of the total fatty acid area. 1373 Confirmation of component identification was performed by GC-MS of selected 1374 samples and was carried out on a Thermo Scientific (Waltham, MA, USA) 1310 GC coupled 1375 with a TSQ triple quadruple. Samples were injected using a Tripleplus RSH (Waltham, MA, 1376 USA) auto sampler using a non-polar HP-5 Ultra 2 bonded-phase column (50 m length x 0.32 1377 mm internal diameter x 0.17 μm film thickness). The HP-5 column was of similar polarity to 1378 the column used for GC analyses. The initial oven temperature of 45 °C was held for 1 min, 1379 followed by temperature programming at 30 °C per minute to 140 °C, then at 3 °C per minute 1380 to 310 °C, where it was held for 12 min. Helium was used as the carrier gas. Mass-spectrometer 1381 operating conditions were as follows: electron impact energy 70 eV; emission current 250 1382 μamp, transfer line 310 °C; source temperature 240 °C; scan rate 0.8 scan/sec and mass range 1383 40–650 Da. Mass spectra were acquired and processed with Thermo Scientific XcaliburTM 1384 software (Waltham, MA, USA). Identification and quantification of peaks was conducted using 1385 the same standards as GC-FID analysis. 1386 1387 Column chromatography 1388 Lipid class fractions of krill sample TSE were separated via column chromatography. The 1389 columns were packed with 1 g of activated silica and enough chloroform (CHCl3) to just cover 1390 the silica. An aliquot of the lipid extract (approximately 10 mg) was syringed onto the top of 1391 the column. In some instances, due to samples having less than 10 mg of total lipid, the entire 1392 lipid extract was used. Lipid fractions were collected by eluting 20 ml of CHCl3, then acetone 1393 (C3H6O) and finally MeOH, which were then each rotary evaporated to concentrate the 1394 individual lipid fractions. An Iatroscan TLC-FID analyser was used to confirm lipid class 67 1395 separation. All of the neutral lipid fractions from the digestive glands and stomachs were used 1396 to prepare FAME of these samples using the same method as above. Neutral lipid FAME were 1397 analysed by GC-FID and GC-MS to identify and quantify fatty acid markers as described above. 1398 Saponification 1399 An aliquot of the neutral lipid fraction was treated with 2 ml saponifying solution (5% 1400 potassium hydroxide (KOH) in MeOH:MilliQ H2O 80:20, v:v) and heated to 60 °C for 3 hours 1401 before being cooled and 1 ml MilliQ H2O and 1.8 ml C6H14:CH2Cl2 added to extract the total 1402 non-saponifiable neutral (TSN) lipid. The TSN lipid fractions were then silyated with the 1403 addition of 50 µl N, O-bis (trimethylsilyl) trifluroacetamide and heated at 60 °C overnight. 1404 Prior to GC-FID analysis, samples were blown down under N2 gas and 1 ml of internal injection 1405 standard (23:0 FAME) added. Silyated TSN samples were analysed by GC-FID and GC-MS 1406 similarly to the FAME samples to identify and quantify sterol markers. 1407 Statistical analyses 1408 All statistical analyses were conducted in RStudio (version 1.0.153 Copyright (C) 2016) using 1409 the packages: nlme (Pinheiro et al., 2017), maps (Becker, 2017), effects (Fox, 2003), ggplot2 1410 (Wickham, 2009), multcomp (Hothorn et al., 2008), multcompView (Graves et al., 2015), 1411 lsmeans (Lenth, 2016). Due to the low number of males in the sample collections, sub-adult 1412 and mature males were pooled so that each site had three males for comparison to the three 1413 sub-adult, three gravid and three spent females from each site. Linear mixed effect models 1414 including a random factor, least squares comparisons and multifactorial ANOVAs were used 1415 to see which variables of Sector (Atlantic, Indian, Pacific), Site (2 in Atlantic, 3 in Indian, 1 in 1416 Pacific), Type (whole, stomach or digestive gland) and Sex (gravid female, spent female, sub- 1417 adult female and male krill) and their interaction terms were significant to the TL DM per krill 1418 content, as well as the individual lipid classes, and the neutral lipid fraction fatty acids and 1419 sterol levels and their associated masses. Nested ANOVAs were undertaken to look at 68 1420 differences in the sites from the same sector and the sexes from the same site and sector for all 1421 other variables. Year was not included as a factor despite the Indian and Atlantic sectors being 1422 sampled in February 2016, whilst the Pacific was sampled in February 2017. Some 1423 confounding of year/site is to therefore be expected. Data were log or square root transformed 1424 when the assumptions of normality and homogeneity of variances weren’t met. 1425 Principal component analyses (PCA) were performed on the major fatty acids (> 0.5% 1426 of the total fatty acid profile) in Primer6 (version 6.1.13 Copyright (C) 2009) using Bray-Curtis 1427 similarity matrices. All data were pre-treated with a log transformation (logx+1) prior to PCA 1428 analyses. 1429 RESULTS 1430 Total lipid content 1431 Krill from the Indian sector were generally shorter than krill from the other sectors, however, 1432 they were not significantly different in weight in any of their sample types (Table 3.1). Atlantic 1433 sector samples had higher TL DM (mg g-1) per krill in all tissue sample types compared to 1434 Indian sector samples (p < 0.006). No significant differences in TL DM (mg g-1) per krill were 1435 observed between samples from the Indian and Pacific sectors (p: 0.904) and there were only 1436 small differences between the Pacific and Atlantic sectors (p: 0.092; Figure 3.2). There was a 1437 strong tissue type effect and sector effect, although no significant interaction (Type*Sector) 1438 effect (p < 0.001, 0.04 and 0.80 respectively; Figure 3.2). 69 1439 1440 1441 1442 Figure 3.2: Total lipid content (mg g-1) dry mass of Euphausia superba in different Southern Ocean sectors (Atlantic, Pacific and Indian) and tissue types (stomach, digestive gland and whole krill) per animal. Each box represents 1 SD, with the whiskers the second SD and the bold line the mean. 70 1443 1444 1445 1446 Table 3.1: Length (mm), wet weight (g) and total lipid content (mean ± SD) of different Euphausia superba tissue sample types from the Atlantic, Indian and Pacific Southern Ocean sectors. TL: Total lipid, DM: dry mass. N = number of samples (1 krill per whole krill sample, 2 krill digestive glands per sample and 12 krill stomachs per sample). Sample Type Whole Sector N Atlantic 18 Length (mm) Wet Weight (g) TL (mg) TL (mg g-1) DM 50.1 ± 3.2 0.94 ± 0.18 66.3 ± 28.5 306 ± 89.2 Indian 24 46.4 ± 5.5 0.83 ± 0.32 44.3 ± 17.0 257 ± 97.8 Pacific 12 45.1 ± 4.3 0.80 ± 0.29 46.3 ± 14.3 263 ± 56.7 Stomach Atlantic 4 51.8 ± 6.9 0.01 ± 0.00 5.51 ± 2.94 254 ± 97.0 Indian 6 46.7 ± 5.2 0.01 ± 0.01 7.36 ± 7.24 232 ± 108 Pacific 2 52.5 ±7.8 0.01 ± 0.00 11.1 ± 4.83 300 ± 63.1 Digestive Atlantic 12 50.8 ± 2.4 0.08 ± 0.02 14.9 ± 7.76 414 ± 48.2 Krill gland Indian 18 44.5 ± 6.8 0.07 ± 0.03 8.24 ± 4.11 279 ± 73.0 Pacific 6 53.6 ± 5.5 0.15 ± 0.06 20.8 ± 11.2 304 ± 67.6 1447 71 1448 Sex had a small but insignificant overall effect on the TL DM per krill of samples (p: 1449 0.053), although there were no significant differences between individual sex classes and the 1450 small differences seen were mainly due to males having higher TL DM per krill (304.4 ± 89.3 1451 mg g-1) than spent females (236.1 ± 98.2 mg g-1) (p: 0.07). Tukey tests revealed that Sector 1452 had a more significant impact on TL DM per krill content than the sex of the sample did (p > 1453 0.1 for all sex classes). 1454 Lipid class composition 1455 Digestive gland, stomach and whole krill lipid class compositions were not significantly 1456 different from each other (p: 0.25, Figure 3.3). Most differences in lipid class composition 1457 were driven by regional differences, with the Indian sector being significantly different to the 1458 Atlantic and Pacific sector samples (p < 0.001 for both, Figure 3.3). Atlantic and Pacific sector 1459 samples were not significantly different to each other (p: 0.22). There was, however, a strong 1460 Type*Sector interaction for overall lipid class composition (p: 0.002; Figure 3.3). 72 1461 1462 1463 Figure 3.3: Lipid class composition (% of total lipids) between Southern Ocean sectors (Atlantic, Pacific and Indian) and tissue types (stomach, digestive gland and whole krill) of Euphausia superba samples. Hydrocarbons includes wax and steryl esters. 73 1464 Both the mass (mg) and percentage (%) levels of FFA (0.0 ± 0.1 mg and 2.9 ± 5.6%), 1465 ST (ST: 0.0 ± 0.0 mg and 0.8 ± 3.6%), DAG (0.0 ± 0.0%) and PL (0.7 ± 0.6 mg and 29.4 ± 1466 8.5%) were lower in the Indian sector digestive gland samples than the Atlantic and Pacific 1467 sector samples (Table 3.2 and Supplementary Table 3.1 (Appendix 1)). Pacific sector digestive 1468 glands had the highest FFA and PL mass and percentages, followed by Atlantic sector samples, 1469 with Indian sector digestive glands having the lowest mass and percentages of FFA and PL 1470 (Tables 3.2 and Supp. Table 3.1). Pacific digestive gland samples had the lowest HC (includes 1471 wax and steryl esters) mass and percentages (0.01 ± 0.01 mg and 0.2 ± 0.3%) and the lowest 1472 TAG percentages (33.7 ± 3.9%) of all the digestive gland samples (Atlantic: 36.3 ± 6.6%, 1473 Indian: 57.9 ± 19.1%). 74 1474 1475 1476 1477 1478 1479 Table 3.2: Mass (mg g-1, mean ± SD) of each lipid class for each Euphausia superba sample, tissue types (digestive gland, stomach and whole krill) and Southern Ocean sectors (Atlantic, Indian and Pacific). Mass for the stomach and digestive gland samples are for the whole sample (2 digestive glands per sample and 12 stomachs per sample) and are not on a per krill basis. HC - hydrocarbons (including wax and sterol esters); TAG - triacylglycerols; FFA - free fatty acids; ST - sterols; PL - phospholipids. Sample Type Sector HC TAG FFA ST PL Whole Krill Atlantic 0.4 ± 0.6 18.7 ± 11.2 1.7 ± 0.7 1.3 ± 1.9 17.0 ± 13.0 Indian 0.1 ± 0.2 12.0 ± 6.5 0.1 ± 0.0 0.5 ± 0.6 13.8 ± 5.7 Pacific 0.1 ± 0.0 16.7 ± 5.6 0.6 ± 1.2 0.5 ± 0.3 13.7 ± 4.7 Atlantic 0.1 ± 0.0 1.0 ± 0.7 0.1 ± 0.0 0.0 ± 0.0 1.8 ± 1.1 Indian 0.0 ± 0.0 0.8 ± 0.7 0.5 ± 1.1 0.4 ± 0.8 1.5 ± 1.7 Pacific 0.1 ± 0.0 0.6 ± 0.5 4.2 ± 3.3 0.0 ± 0.0 2.3 ± 0.3 Digestive Atlantic 0.0 ± 0.0 3.3 ± 2.5 0.7 ± 0.3 0.1 ± 0.1 3.7 ± 2.2 gland Indian 0.2 ± 0.3 1.4 ± 0.7 0.0 ± 0.0 0.0 ± 0.1 0.7 ± 0.6 Pacific 0.0 ± 0.0 3.8 ± 2.8 1.5 ± 1.1 0.0 ± 0.0 5.0 ± 3.7 Stomach 1480 75 1481 Pacific sector stomach samples had the lowest PL mass (2.3 ± 0.3 mg) and percentages 1482 (43.0 ± 16.4%), as well as the lowest TAG mass (0.6 ± 0.5 mg) and DAG percentages (0.2 ± 1483 0.3%) (Tables 3.2 and Supp. Table 3.1). Indian sector krill stomachs had low HC and FFA 1484 mass (0.00 ± 0.01 mg and 0.5 ± 1.1 mg, respectively) and HC percentages (0.00 ± 0.02%) 1485 compared to other stomach samples, whilst Atlantic sector stomachs had low ST mass and FFA 1486 percentages (Tables 3.2 and Supp. Table 3.1). 1487 Indian sector whole krill samples followed the same trend as their digestive glands with 1488 low HC and TAG mass (0.1 ± 0.2 mg and 12.0 ± 6.5 mg, respectively) and low TAG and FFA 1489 percentages (37.6 ± 10.7% and 0.1 ± 0.1%, respectively). Atlantic sector whole krill samples 1490 had the highest mass and percentages for most lipid classes (Tables 3.2 and Supp. Table 3.1). 1491 There were no significant differences in lipid class levels between sexes, except for PL 1492 (p: 0.036). This difference was driven by spent females in the Indian sector which were 1493 significantly higher in their percentage of PL (66.5 ± 13.8%) than spent females in the other 1494 regions (Atlantic: 41.6 ± 9.5%, Pacific: 46.6 ± 5.7%). Indian sector krill were generally higher 1495 in PL levels than observed for the other regions except in gravid females which were consistent 1496 in their PL levels regardless of the region. 1497 Quantitatively, TAG mass varied slightly but insignificantly with sex class (p: 0.053), 1498 due to differences between males and sub-adult females (18.3 ± 10.4 mg and 9.9 ± 4.2 mg, 1499 respectively) (p: 0.06). There was also a large difference in PL mass depending on sex class (p: 1500 0.01), with sub-adult females and males (5.5 ± 1.9 mg and 18.7 ± 12.3 mg, respectively) (p: 1501 0.042) having the largest difference in PL mass, but sexes within the same region didn’t show 1502 a significant difference (p: 0.68). 76 1503 Fatty acid content and composition 1504 The sum of fatty acid groups (monounsaturated (MUFA), polyunsaturated (PUFA) and 1505 saturated fatty acids (SFA)) did not vary significantly between tissue types or regions (Table 1506 3.3). Some fatty acids were combined or used as ratios to look at particular dietary indicators 1507 such as: carnivory (18:1n-9c/18:1n-7c), copepod consumption (Ʃ 20:1n-9c+22:1n-9c, Hagen 1508 et al., 1995), marine snow and bacterial indicators (Ʃ C15, C17 and C19 isomers, hereon termed 1509 MSI), diatom consumption (Ʃ 16:4n-1+16:1n-7c), and the copepod to diatom ratio (Ʃ 20:1n- 1510 9c+22:1n-9c/16:4n-1+16:1n-7c) for an estimate of their herbivory: omnivory: carnivory ratio. 1511 77 1512 1513 1514 1515 Table 3.3: Euphausia superba fatty acid groups (expressed as mg g-1 sample; mean ± SD) and selected major dietary fatty acid markers in different sectors of the Southern Ocean (Atlantic, Indian and Pacific) and tissue types (whole krill, stomach and digestive glands). MUFA: monounsaturated fatty acids, PUFA: polyunsaturated fatty acids, SFA: saturated fatty acids, MSI: marine snow indicator [Ʃ C15, C17 and C19 isomers], Copepods: [Ʃ 20:1n-9c + 22:1n-9c], Diatoms: [Ʃ 16:1n-7c +16:4n-1]. Sample Sector MUFA PUFA SFA MSI Phytol Copepods Diatoms Whole Atlantic 13.2 ± 5.6 16.3 ± 4.9 16.1 ± 6.6 0.3 ± 0.1 1.3 ± 0.9 1.0 ± 0.6 4.2 ± 1.7 Krill Indian 6.5 ± 5.3 6.8 ± 4.1 8.2 ± 8.1 0.1 ± 0.5 0.8 ± 0.4 0.3 ± 0.2 2.2 ± 2.1 Pacific 11.1 ± 2.5 15.1 ± 3.5 14.6 ± 4.2 0.4 ± 0.1 0.8 ± 0.3 0.6 ± 0.2 2.7 ± 0.8 Atlantic 17.3 ± 19.5 16.9 ± 16.4 16.8 ± 14.8 0.3 ± 0.2 0.5 ± 0.7 1.0 ± 1.0 4.2 ± 3.8 Indian 10.4 ± 7.0 11.5 ± 7.8 11.1 ± 5.8 0.2 ± 0.2 0.0 ± 0.0 0.5 ± 0.3 2.9 ± 1.4 Pacific 12.2 ± 4.8 17.7 ± 6.8 13.9 ± 5.6 0.4 ± 0.2 0.4 ± 0.6 0.8 ± 0.2 2.6 ± 0.9 Digestive Atlantic 15.8 ± 5.1 19.9 ± 4.8 20.3 ± 6.9 0.4 ± 0.1 1.1 ± 0.5 1.3 ± 0.4 5.0 ± 1.7 gland Indian 9.2 ± 3.5 10.2 ± 3.0 13.0 ± 5.8 0.2 ± 0.1 0.0 ± 0.0 0.5 ± 0.2 3.7 ± 1.9 Pacific 12.9 ± 2.8 17.6 ± 3.6 16.1 ± 4.1 0.4 ± 0.0 0.8 ± 0.2 0.8 ± 0.2 3.1 ± 1.0 Type Stomach 1516 78 1517 The biggest differences in the fatty acids of the total lipid between tissue types were 1518 seen in the major fatty acids 20:5n-3 (EPA), 22:6n-3 (DHA), 18:4n-3 (SDA), MSI, copepod, 1519 carnivory and diatom markers, phytol (side chain of chlorophyll, used as a primary production 1520 marker) and the copepod to diatom ratio (Tables 3.3 and Supplementary Table 3.2 (Appendix 1521 1)). Indian sector samples had the lowest fatty acid mass (MSI, copepod and diatom markers) 1522 and relative levels (MSI and carnivory markers) in their total lipids overall, although not in 1523 their 16:0, 16:1n-7c, copepod biomarker percentages and copepod to diatom ratio levels 1524 (Tables 3.3 and Supp. Table 3.2). 16:4n-1 and EPA and copepod marker levels were 1525 consistently highest in Atlantic sector digestive glands and whole krill, followed by Pacific 1526 sector samples, with the lowest relative levels in Indian sector samples (Tables 3.3 and Supp. 1527 Table 3.2). Similarly, DHA, SDA, MSI and carnivory marker levels were highest in Pacific 1528 sector samples, followed by the Atlantic sector and were lowest in Indian sector samples, 1529 regardless of sample type (Tables 3.3 and Supp. Table 3.2). 1530 The differences seen between whole krill from the different regions are clearly 1531 established in a PCA plot of the total lipid fatty acids (Figure 3.4A), which shows the distinctly 1532 different Indian sector krill. This difference was mostly driven by higher levels of 16:1n-7c, 1533 14:0 and 20:1n-9c and lower levels of 18:4n-3 and DHA in Indian sector krill compared with 1534 krill from the other regions (cumulative percentage variation: 86.8%). The fatty acids having 1535 the biggest influence were SDA, DHA and 16:1n-7c (Figure 3.4A, see Supplementary Table 1536 3.3 (Appendix 1) for PCA eigenvalues). PC1 was driven by 16:1n-7c and SDA, PC2 was driven 1537 by DHA and 14:0 and PC3 was driven by 14:0 and 20:1n-9c (Figure 3.4A and Supp. Table 1538 3.3). Other fatty acids made up smaller parts of each component, but were not the driving 1539 factors of the differences seen (e.g. EPA, 18:1n-7c, 20:4n-3 and 16:4n-1) (Figure 3.4A and 1540 Supp. Table 3.3). 79 1541 1542 1543 1544 1545 Figure 3.4: Principal Component Analysis of the fatty acid composition (% data) of Euphausia superba samples from different Southern Ocean sectors (Atlantic, Pacific and Indian) and sites from: (A) the total lipid of whole krill, (B) the neutral lipid fraction of krill stomachs and (C) neutral lipid fraction of krill digestive glands. 80 1546 The differences in regional dietary influences are more clearly seen using stomach and 1547 digestive gland neutral lipid fraction fatty acids (Figures 3.4B and 3.4C, respectively). 1548 Digestive glands showed a smaller difference between regions, with more overlap observed 1549 between their neutral lipid fraction fatty acid profiles (Figure 3.4C) than the distinctly different 1550 regional neutral lipid fraction fatty acid profiles observed in the stomach samples (Figure 3.4B). 1551 The regional differences seen in the stomach neutral lipid fraction fatty acid profiles are driven 1552 by 14:0, 18:0, DHA and 16:1n-7c (cumulative percentage variation: 95.3%, see Supplementary 1553 Table 3.3 (Appendix 1) for full PCA values). The differences between regions in digestive 1554 gland profiles are driven by 14:0, EPA, 18:1n-9c and SDA (cumulative percentage variation 1555 92.4%, see Supp. Table 3.3 for full PCA values). SDA made up a significant but smaller portion 1556 of the variation in the stomach PCA (Figure 3.4B and Supp. Table 3.3). 1557 The different fatty acid groups (MUFA, PUFA and SFA) in the total lipid of whole krill 1558 varied slightly, but not significantly between sex classes. Again, the Indian sector krill were 1559 distinctly different in their fatty acid composition, and this was driven mostly by sub-adult 1560 females which were high in 16:1n-7c (Figure 3.5). The principle components that caused this 1561 separation are identical to those seen in Figure 3.4A (Supp. Table 3.1). However, a significant 1562 difference between the same sexes at different sites within the Indian sector could also be seen 1563 (Figures 3.4A and 3.5). Krill from the site closest to the continental shelf (Indian 6) were more 1564 similar to the krill of the Atlantic and Pacific sectors (higher DHA and SDA, lower 16:1n-7c), 1565 whereas krill from the more oceanic sites (Indian 4 and 5) had distinctly different dietary 1566 markers. This continental shelf site had no sub-adult females present, which was the driving 1567 sex class in the other Indian sector sites. 81 1568 1569 1570 Figure 3.5: Principal Component Analysis of the total lipid fatty acid composition (% composition) compared by sex between Southern Ocean sectors (Atlantic, Pacific and Indian) of whole Euphausia superba samples. 82 1571 Fatty acid masses were quantitatively lower in all sexes of Indian sector krill. Atlantic 1572 sector males had more 16:0 (12.7 ± 5.5 mg, mean ± SD), 16:1n-7c (4.8 ± 2.2 mg), 16:4n-1 (0.3 1573 ± 0.1 mg), EPA (9.8 ± 3.7 mg), and copepod marker (1.3 ± 0.8 mg) masses than Pacific sector 1574 males (9.5 ± 3.9 mg, 2.5 ± 1.2 mg, 0.2 ± 0.1 mg, 7.5 ± 2.3 mg and 0.7 ± 0.3 mg, respectively). 1575 Pacific sector sub-adult females had higher fatty acid masses than Indian sector sub-adult 1576 females, regardless of fatty acid type or ratio, except in their 16:1n-7c mass which was lower 1577 (Pacific: 3.4 ± 0.5 mg, Indian: 3.7 ± 3.9 mg). Spent females in the Atlantic sector were higher 1578 in all fatty acid masses except for their MSI and carnivory markers (0.4 ± 0.1 mg and 1.9 ± 0.2 1579 mg, respectively), which were higher in spent females in the Pacific sector (0.5 ± 0.1 mg and 1580 2.1 ± 0.2 mg, respectively). Gravid females in the Pacific sector were higher in DHA and SDA 1581 (4.2 ± 0.5 mg and 1.6 ± 0.2 mg), and in their MSI and carnivory marker (0.6 ± 0.1 mg and 2.2 1582 ± 0.1 mg) masses than gravid females in the Atlantic (DHA: 4.0 ± 1.6 mg, SDA: 1.5 ± 0.5 mg, 1583 MSI: 0.5 ± 0.1 mg, Carnivory: 1.8 ± 0.2 mg) and Indian sectors (DHA: 1.7 ± 0.9 mg, SDA: 0.3 1584 ± 0.3 mg, MSI: 0.2 ± 0.1 mg, Carnivory: 1.7 ± 0.1 mg). Atlantic sector gravid females were 1585 higher in 16:0, 16:4n-1, EPA and copepod marker masses. 1586 Sterols 1587 Cholesterol levels did not differ between regions or sample types (p: 0.55 and 0.69) and there 1588 was no Type*Sector interaction (p: 0.79; Table 3.4). Pacific sector samples were generally 1589 higher in the major sterol percentages (cholesterol: 62.8 ± 1.89% (stomach) and 47.8 ± 11.24% 1590 (digestive gland), desmosterol: 26.5 ± 2.87% (stomach) and 28.0 ± 9.48% (digestive gland) 1591 and brassicasterol: 3.8 ± 0.39% (stomach) and 6.27 ± 1.79% (digestive gland)) on a mass basis 1592 than samples from the other regions, except for 24-methylenecholesterol in their digestive 1593 gland (2.0 ± 0.54%) and 24-ethylcholesterol in their stomach (0.1 ± 0.45%) samples being low 1594 (Table 3.4). Indian sector samples were lower in selected sterols except 24-ethylcholesterol 1595 levels which were higher in digestive gland (1.0 ± 0.58%) and stomach samples (0.8 ± 0.56%) 83 1596 than the other regions (Atlantic sector stomachs: 0.3 ± 0.11% and digestive glands: 0.4 ± 0.18%, 1597 Pacific sector stomachs: 0.1 ± 0.45% and digestive glands: 0.3 ± 0.06%. 1598 84 1599 1600 Table 3.4: Sterols of Euphausia superba digestive gland and stomach samples by Southern Ocean sector (Atlantic, Indian and Pacific) expressed as a percentage (%, mean ± SD) of the total sterol profile. Ʃ Unknown 24-Ethyl-cholest-5-ene-3ß–ol 24-(Ethylcholesterol) 24-Methyl-cholesta- 5,24(28)E-diene-3ß–ol 24-(Methylenecholesterol) 24-Methyl-cholesta-5,22E- diene-3ß–ol (Brassicasterol) 24-Methyl-cholesta-5,24- diene-3ß–ol (Desmosterol) Cholest-5-ene-3 ß –ol (Cholesterol) Cholesta-5,22Z- dien-3ß–ol gland (Trans-22- Digestive dehydrocholesterol) Stomach 24-Norcholesta-5,22E-dien- Type 3ß-ol (24-Norcholesterol) Sample Sector N Atlantic 4 1.01 ± 0.22 2.48 ± 0.40 60.90 ± 4.93 28.65 ± 6.61 2.30 ± 0.87 2.72 ± 0.87 0.31 ± 0.10 1.56 ± 0.17 Indian 6 0.95 ± 0.73 7.38 ± 5.55 52.50 ± 14.10 18.31 ± 9.56 3.52 ± 2.18 1.74 ± 0.93 0.81 ± 0.56 14.81 ± 6.43 Pacific 2 0.91 ± 0.19 4.40 ± 0.53 62.81 ± 1.89 26.51 ± 2.87 3.80 ± 0.69 1.08 ± 0.45 0.14 ± 0.45 0.37 ± 0.02 Atlantic 12 1.41 ± 0.59 3.63 ± 2.03 54.02 ± 22.91 27.09 ± 12.02 2.37 ± 1.14 6.34 ± 3.61 0.36 ± 0.18 4.76 ± 1.98 Indian 18 1.56 ± 0.91 8.64 ± 4.13 54.90 ± 14.03 23.20 ± 6.57 4.99 ± 2.78 3.82 ± 1.17 0.96 ± 0.58 1.91 ± 0.54 Pacific 6 12.50 ± 5.12 47.83 ± 11.22 27.99 ± 9.48 6.27 ± 1.79 2.03 ± 0.54 0.32 ± 0.06 1.02 ± 0.08 1.99 ± 0.52 1601 85 1602 DISCUSSION 1603 Krill diet varied by region during the late-summer season as seen in their lipid, fatty acid and 1604 sterol biochemistry. Large differences in the shorter-term diet of krill can be seen in their 1605 stomach (days) and digestive gland (days-weeks) as assessed by neutral lipid fraction fatty acid 1606 profiles. Indian sector krill had a more diatomaceous and carnivorous based diet than krill in 1607 the Pacific and Atlantic sectors. Within each region, krill of different sexes showed little 1608 variation. The Indian sector was the exception, where sub-adult females and spent females had 1609 significantly different dietary signals. Differences in the diet of sex classes was predominantly 1610 due to a regional effect. Males and gravid females at Indian site 6 had vastly different diets 1611 with females generally having higher DHA and males generally higher SDA. 1612 Total lipid content 1613 Atlantic sector digestive glands and whole krill samples showed much higher TL DM (mg g-1) 1614 per krill than observed for the other regions. There were no significant differences between TL 1615 DM (mg g-1) per krill between spent females and males. Earlier studies have reported large 1616 differences between krill sexes and TL content (Fricke et al., 1984, Mayzaud et al., 1998). 1617 However, sub-adult females were high in TL DM (mg g-1) per krill, although lipid content 1618 decreased in gravid females and decreased further in spawned females in the Indian and Pacific 1619 sectors. This trend is not seen in females in the Atlantic sector. This difference may be due to 1620 primary production in the Atlantic sector being higher than in the Indian and Pacific sectors 1621 (El-Sayed and Weber, 1982, Vernet et al., 2008, Westwood et al., 2010). 1622 The TL DM (mg g-1) per krill in stomach samples was higher in the Pacific sector than 1623 in the Indian sector, despite Pacific sector krill digestive glands and whole krill sharing similar 1624 TL DM (mg g-1) per krill as Indian sector samples. Indian sector krill may be utilising their 1625 digestive gland to store lipids (Dall et al., 1992, Virtue et al., 1993a). Krill stomachs digest and 86 1626 break down lipids but they lack any short-term storage capability (Mayzaud et al., 1998, 1627 Schaafsma et al., 2017). 1628 Changes in whole krill lipid variability has been shown to be related to variability in 1629 their digestive glands (Alonzo et al., 2005). Additionally, the fatty acid dietary signal observed 1630 in krill is due mostly to the dietary markers within the digestive gland (Alonzo et al., 2005). 1631 Stomach samples had lower TAG levels than the other tissue types, suggesting that this storage 1632 lipid is produced further down the digestive system (Mayzaud et al., 1998, Hagen et al., 2001). 1633 Krill stomachs are rarely used for lipid or fatty acid analysis due to their lack of TAG and 1634 neutral lipid fatty acids (Bottino, 1974, Ju and Harvey, 2004). These analyses are more 1635 commonly conducted on the digestive gland (for short term dietary signals) (Virtue et al., 1993a, 1636 Yoshida et al., 2009) which accumulates TAG during feeding. For longer-term dietary signals, 1637 whole krill are generally the preferred sample type used (Schaafsma et al., 2017, Ericson et al., 1638 2018a, Hellessey et al., 2018). The use of neutral lipid fatty acid profiles of digestive and 1639 stomach tissue, rather than whole animal tissue, allows for even greater understanding of 1640 dietary components as it reflects the diet to a larger degree than fatty acids from phospholipids 1641 or total lipids. 1642 Lipid class composition 1643 Krill from the Indian sector had low PL levels in their digestive glands, but high PL levels in 1644 their whole body, showing a need for krill in this region to both metabolise and synthesise this 1645 lipid class. Pacific sector krill had high PL levels in their stomach and digestive glands and 1646 lower PL levels in their whole body, compared to krill from the other regions. Krill conserve 1647 PL and TAG based on their available dietary choices, the season, and their health status (e.g. 1648 PL is conserved until it is at required levels for adequate health, then TAG is stored) (Hagen et 1649 al., 1996, Hagen et al., 2001). 87 1650 Stomachs and digestive glands had higher FFA as the organs where dietary items (such 1651 as lipids) are broken down (Henderson et al., 1981, Pond et al., 1995). Our results show higher 1652 FFA in the stomach than the digestive gland in two regions (Indian and Pacific sectors), which 1653 may be a function of the higher residence time of food in stomachs as a result of reduced 1654 feeding activity (Virtue et al., 1993a). The FFAs were extremely high in Pacific sector krill stomach 1655 and digestive gland samples and this may be due to these samples being stored at -20 °C whereas krill 1656 from the Indian and Atlantic sectors were stored at -80 °C. This may mean that some enzymatic activity 1657 was still present in the Pacific sector krill samples as they were stored differently. However, whole krill 1658 samples from the Atlantic sector also had high FFAs present, despite the FFAs being low in their 1659 stomach and digestive gland samples. This potentially shows two different outcomes: 1) that the 1660 enzymatic activity even at -80 °C is still present in some tissues, or 2) that high FFAs were present in 1661 samples regardless of the difference in storage temperatures. 1662 Lipid class composition varied between the regions for krill digestive gland and 1663 stomach samples, but whole krill samples had a consistent lipid class composition within all 1664 regions. Whole krill lipids would be consistent for lipid classes required for reproduction, and 1665 growth, which would be similar in all regions (Pond et al., 1995, Ju and Harvey, 2004, 1666 Kohlbach et al., 2015), whilst stomach and digestive gland lipids would fluctuate greatly based 1667 on dietary input. 1668 Lipid class composition by sex 1669 Sub-adult females were generally higher in TAG percentage levels than gravid and spent 1670 females within the same region. Sub-adult females may be storing more fats going into their 1671 second winter (Atkinson et al., 2002). Older females that are spawning have used and are still 1672 using these stored lipids to reproduce (Mayzaud et al., 1998, Atkinson et al., 2002). Hence, 1673 older females have less TAG available and this was particularly evident in Indian sector krill 1674 samples. A large decrease in TAG could be due to 2016 being a less productive year (Bestley 88 1675 et al., 2018, Schallenberg et al., 2018), so stored fats were being more readily used for 1676 reproduction in both mature females and males (Virtue et al., 1993a). Sub-adult females may 1677 also be outcompeting mature females and males for the food that is available (Atkinson et al., 1678 2002). 1679 Gravid female krill from the Indian and Atlantic sectors had higher sterol and lower PL 1680 percentage levels than spent females in the same region. Higher sterol levels in gravid females 1681 would help facilitate females moving lipids from their internal lipid stores into their eggs before 1682 spawning (Tarling et al., 2009b). Krill eggs are high in lipids, particularly sterols (Pond et al., 1683 1995, Mayzaud et al., 1998), which accounts for the high sterol levels in gravid females. 1684 Atlantic sector male krill had much higher sterol levels than male krill in other regions due to 1685 energetic differences (Virtue et al., 1996). 1686 Male and sub-adult female krill had significantly higher percentages of TAG than spent 1687 and gravid females and this may be why their PUFA levels are lower. Gravid female krill in 1688 particular do not store non-reproductive lipid (Mantel, 1983) and spent females are low in lipid 1689 (Mayzaud et al., 1998). Gravid females had large differences in their PUFA levels depending 1690 on the diet and productivity level in different regions, whilst sub-adult females and males show 1691 large differences in their PUFA levels in areas of lower production (Pacific and Indian sectors). 1692 MUFA levels were high in gravid females as well as sub-adult females and males in the Indian 1693 sector compared to other regions, further showing there is a major difference in their diet in 1694 this region. 1695 These differences in diet can be seen more clearly once fatty acid markers and sterols 1696 are used to tease apart specific prey item groups within the digestive glands and stomachs. It is 1697 becoming more common for neutral lipid fractions to be utilised in analysis to see clearer 1698 dietary signals (Cabrol et al., 2019). Using only the fatty acids from the neutral lipid fractions 89 1699 ensures that any signal from the structural cells of the digestive gland and stomach are not 1700 included into the dietary marker profile (Virtue et al., 1993b, Yoshida, 2009, Cabrol et al., 1701 2019). 1702 Diatoms and dinoflagellates 1703 The fatty acid and sterol profiles revealed fine scale differences between regions. Digestive 1704 gland fatty acid profiles did not show as clear of a distinction between regions as the stomach 1705 sample profiles. This may be due to digestive glands being used for lipid storage within the 1706 body, so some markers could be stored here long after the shorter-term diet of the krill in that 1707 region has changed (Virtue et al., 1993a, Mayzaud et al., 1998, Yoshida et al., 2009). Fatty acid 1708 profiles from Atlantic and Pacific sector whole krill suggest they have a similar long-term diet 1709 despite their stomach profile being more unique. The storage of particular fatty acids in the 1710 digestive gland could be driving this similarity between the two regions. Stomachs showed the 1711 clearest distinction between regions, with krill from the Indian sector having a more 1712 carnivorous and diatomaceous diet and krill from the Atlantic and Pacific being higher in 1713 herbivory and flagellate markers. 1714 Sterol levels (expressed as % of total sterols) in the stomach showed that diatoms 1715 (proposed source of 24-methylenecholesterol (Phleger et al., 2002)) were nearly double the 1716 dietary input in the Atlantic sector compared to the Pacific sector and 1.5 times that of the 1717 Indian sector, consistent with the fatty acid markers. Digestive gland sterol levels showed that 1718 this trend was amplified, as Atlantic sector krill showed nearly threefold higher levels of the 1719 diatomaceous sterols compared to Pacific sector krill and double that of the Indian sector krill. 1720 Fricke et al. (1984) showed that krill sterols were dominated by cholesterol and that most of 1721 the other major sterol markers (such as desmosterol) came from diatoms and other 1722 phytoplankton inputs. Phleger et al. (2002) reported that these algal sterol levels fluctuated 90 1723 interannually and seasonally just like algal fatty acid markers as shown in Fricke et al. (1984) 1724 and Ericson et al. (2018a). 1725 The essential and health-benefitting omega-3 (EPA (a diatom marker), DHA (a 1726 dinoflagellate marker) and SDA (a flagellate marker)) (Virtue et al., 1996, Ross and Quetin, 1727 2000, O’Brien et al., 2011) were present at similar relative levels in all regions as they are used 1728 for krill growth and reproduction. Interestingly, Pacific sector krill had very high EPA, but was 1729 consistently lower in other diatom markers (16:1n-7c, 14:0 and 16:4n-1). These krill may be 1730 preferentially storing EPA, and using other, more readily metabolized fatty acids, for 1731 reproduction (Gigliotti et al., 2011) which is also seen in North Atlantic krill (Saether et al., 1732 1986). 1733 Omega-3 fatty acid levels in whole krill normally reflect the environmental omega-3 1734 levels in that region (Hagen et al., 2001, Schaafsma et al., 2017). Higher environmental omega- 1735 3 levels might be driving krill in the Atlantic to conserve higher body levels, as this region has 1736 both diatoms and dinoflagellates in abundance in the late summer period (Korb et al., 2005). 1737 Atlantic sector krill also had higher PL levels in their stomachs, which may be due to an 1738 abundance of dietary items which are high in PL in this region (e.g. diatoms and 1739 dinoflagellates). 1740 Indian sector krill had lower levels of omega-3 fatty acids in all tissue types at sites 4 1741 and 5 and only DHA was elevated at site 6. The elevated levels of DHA at this site may be due 1742 to these krill being mostly spent females, whereas sub-adult females were analysed from sites 1743 4 and 5. Males and gravid females from Indian site 6 had elevated DHA and SDA levels, 1744 consistent with a more dinoflagellate based diet at this more inshore site (El-Sayed and Weber, 1745 1982, Korb et al., 2005, Ducklow et al., 2007). Male krill from Indian site 6 had higher DHA, 1746 SDA and 16:4n-1 showing a more herbivorous diet than males from the other Indian sector 91 1747 sites. Pacific sector males were low in 16:0 (palmitic acid) and phytol (chlorophyll side-chain) 1748 levels, suggesting a more complex phytoplankton diet than occurs for the Indian and Atlantic 1749 sector male diets. 1750 Atlantic and Pacific sector sub-adult females had mostly flagellate fatty acid markers 1751 (very high SDA), whilst Indian sector sub-adult females had a predominantly diatomaceous 1752 diet. Spent females were especially high in diatom and flagellate marker levels in the Indian 1753 sector (EPA and DHA). Their copepod to diatom ratio suggests an almost entirely 1754 phytoplankton-based diet. This may be due to the vital omega-3 fatty acids used for egg 1755 production (Mayzaud et al., 1998, Ross and Quetin, 2000, Tarling et al., 2009b) being found 1756 in diatoms and dinoflagellates and spent females needing to replenish these losses as quickly 1757 as possible once the eggs are spent. However, spent females in the Pacific sector had a diet that 1758 was lower in phytoplankton and spent females in the Atlantic sector showed a mixed diet with 1759 some phytoplankton markers (DHA, SDA, 16:4n-1 and palmitic acid). These varied strategies 1760 show that spent females are opportunistic and may be feeding on anything available to them as 1761 they must quickly build up fat reserves throughout the autumn to prepare for winter when krill 1762 use up most of their storage fats (Hellessey et al., 2018). The differences between gravid 1763 females could be due to the Pacific sector having lower levels of primary production and 1764 therefore to compensate, krill are needing to eat more detritus and marine snow (Schmidt et al., 1765 2011). 1766 Marine snow and detritus 1767 Atlantic sector krill exhibited higher MSI levels in their digestive glands than their stomachs, 1768 potentially showing temporal effects of feeding on detrital type material. Marine snow marker 1769 percentages were higher in Pacific sector krill stomachs than Indian sector krill stomachs, 1770 reflecting marine snow levels in the Indian sector during summer (Turner, 2002, Stübing et al., 1771 2003, Pasquer et al., 2010, Turner, 2015). Atlantic sector males were lower in marine snow 92 1772 markers and spent females in the Pacific sector had a diet that was higher in marine snow and 1773 copepod marker levels. 1774 Copepod and carnivory markers 1775 Copepod marker levels were high in all Atlantic sector krill tissue types, indicative of a 1776 consistent level of feeding and metabolism of these fatty acid markers; e.g. no storage in the 1777 digestive gland is occurring for these markers. Copepods may be a more consistent food source 1778 for krill in the Atlantic sector (Ward et al., 2012b) and hence krill in this sector can maintain 1779 higher levels of these markers. Krill in the Indian and Pacific sectors, however, appear to be 1780 eating more copepods at this time of year; as indicated by higher copepod marker masses and 1781 levels in their stomachs and digestive glands than Atlantic sector krill (Chiba et al., 2001, 1782 Belcher et al., 2017, Schaafsma et al., 2017). In this study Indian sector spent females were 1783 low in copepod marker levels and their copepod to diatom ratio suggests an almost entirely 1784 phytoplankton-based diet. Carnivory markers, however, were found throughout all body tissues 1785 in all regions, with slightly higher levels in krill sampled from the Pacific sector. This may be 1786 due to cod-end feeding from the method of collection in the Pacific sector, a Bongo net (Morris 1787 et al., 1984), unlike the Atlantic sector collection method (a continuous pump). Carnivory has 1788 always been seen in low levels within krill diets regardless of where, when and how the studies 1789 were conducted (Ju and Harvey, 2004, Schmidt and Atkinson, 2016, Ericson et al., 2018a). 1790 Indian sector spent females were low in carnivory marker levels. 1791 This study has revealed, that lipid class, neutral lipid fraction fatty acids and sterol 1792 dynamics in krill are variable depending on the region, sex and tissue type of the krill examined. 1793 Differences found in the total lipid, lipid class and neutral lipid fraction fatty acid and sterol 1794 composition and content, could be due to the varied trophic environments within each region. 1795 Future work could focus on 6 key fatty acids for dietary analysis between regions: EPA, DHA, 1796 SDA, 16:1n7c, 16:4n1 and 14:0. These fatty acids would allow greater comparisons between 93 1797 diatomaceous and flagellate based diets in each region. Whilst MSI and copepod markers did 1798 influence krill diet these had a smaller overall impact than the 6 major fatty acids listed above. 1799 Similarly, there was no clear trend between krill sex classes, however sub-adult females were 1800 more carnivorous than mature krill, so future work looking into sexual maturity and krill diet 1801 could focus on the 18:1n9c to 18:1n7c ratio and copepod markers. Clarifying regional trends 1802 in krill diet will require further, more extensive, studies in all ocean basins – preferably 1803 throughout the year. Samples from the krill fishery offer the best opportunity for examining 1804 detailed seasonal and regional differences (Ericson et al., 2018a, Hellessey et al., 2018) 1805 although the fishery is currently regionally constrained. The standardised collection of krill 1806 from all ocean basins may be possible in the future due to an increase of research effort in the 1807 Ross Sea and the expanding fishery in the Indian sector of the Southern Ocean, allowing for 1808 greater dietary comparisons between these regions. 94 1809 ACKNOWLEDGEMENTS 1810 This project was funded by the ARC Linkage Project LP140100412, in partnership with Aker 1811 BioMarine, CSIRO, AAD, UTAS and Griffith University. We thank Andy Revill, Mina 1812 Brock and Ben Gaskell who helped us in the CSIRO lipid laboratory, Aker BioMarine and 1813 the crew of the FV Saga Seas for the Atlantic Ocean krill samples, Rob King from the 1814 Australian Antarctic Division for providing the Indian Ocean krill samples, and Giuseppe 1815 Suaria from the Antarctic Circumpolar Expedition (Swiss Polar Institute and Frederik 1816 Paulsen) for providing the Pacific Ocean krill samples. 95 1817 Chapter 4: Antarctic Krill (Euphausia superba Dana 1850) Lipid and Fatty 1818 acid Content Variability is associated to Satellite Derived Chlorophyll a 1819 and Sea Surface Temperature in the Scotia Sea 1820 1821 This Chapter has been accepted pending minor revisions at Nature Scientific Reports. 1822 1823 ABSTRACT 1824 Antarctic krill (Euphausia superba) is a key component of the Antarctic food web with 1825 considerable lipid reserves that are vital for both their own and higher predator survival. Krill 1826 lipids are primarily derived from their diet of plankton, in particular diatoms and flagellates, 1827 but few attempts have been made to link the spatial and temporal variations in krill lipids to 1828 those in their food supply. Remotely-sensed environmental parameters provide large-scale 1829 information on the potential availability of krill food, although relating this to physiological 1830 and biochemical differences has only been performed on small scales and with limited samples. 1831 Our study utilised remotely-sensed data (Chlorophyll a and sea surface temperature) coupled 1832 with krill lipid data obtained from 3 continuous years of fishery-derived samples. 1833 examined within and between year variation of trends in both the environment and krill 1834 biochemistry. 1835 Chlorophyll a levels were positively related to krill lipid levels, particularly triacylglycerol 1836 which is a storage lipid. Plankton fatty acid biomarkers analysed in krill (such as n-3 1837 polyunsaturated fatty acids) increased with decreasing sea surface temperature and increasing 1838 chlorophyll a levels. 1839 Our study demonstrates the utility of combining remote-sensing and fisheries data in examining 1840 biological and physiological relationships between Antarctic krill and the Southern Ocean 1841 environment. 96 We 1842 INTRODUCTION 1843 Antarctic krill (Euphausia superba, hereon krill) are at the centre of the wasp-waisted Southern 1844 Ocean ecosystem (Hill et al., 2006, Murphy et al., 2007). Krill, due to their high lipid (oil) 1845 content (up to 40% dry mass (Clarke, 1984, Atkinson et al., 2002)), are vital food for predators 1846 in the region (Stübing et al., 2003, Saunders et al., 2015). Krill have a naturally varied diet 1847 ranging from copepods and phytoplankton such as diatoms and flagellates, to marine snow and 1848 even cannibalism in harsh winter conditions (Hagen et al., 1996, Atkinson et al., 2002, Ju and 1849 Harvey, 2004, Ericson et al., 2018a). Krill are predominantly herbivorous during the summer 1850 and are more omnivorous from autumn to spring (Ericson et al, 2018a). Krill diet has been 1851 assessed through several different means such as microscopy (Schmidt and Atkinson, 2016), 1852 DNA extraction (Passmore et al., 2006, Töbe et al., 2010) and the use of signature fatty acid 1853 biomarkers (Ju and Harvey, 2004, Ericson et al., 2018a). 1854 Biomarkers, such as fatty acids, have been used to examine krill health and diet 1855 previously (Bottino, 1974, Fricke et al., 1984, Virtue et al., 1993a, Mayzaud, 1997, Ju and 1856 Harvey, 2004, Kohlbach et al., 2015, Schaafsma et al., 2017, Ericson et al., 2018a), as they are 1857 a reliable way of looking at the long-term diet of krill (Passmore et al., 2006, Töbe et al., 2010, 1858 Schmidt and Atkinson, 2016). Fatty acid biomarkers are useful as they not only broadly classify 1859 what krill are eating, but their relative and absolute amounts allow insights into how much of 1860 these prey items and types krill have consumed over a more extended period of time (Ericson 1861 et al., 2018a, Ericson et al., 2018b, Hellessey et al., in review). Omega-3 (n-3) long-chain (≥C20) 1862 polyunsaturated fatty acids (n-3 LC-PUFA) are mainly derived from phytoplankton (Nichols 1863 et al., 1986, Butler, 2007), and are needed for krill health, growth and reproduction. They also 1864 serve as useful biomarkers (Virtue et al., 2010, O’Brien et al., 2011, Ericson et al., 2018a) in 1865 food-chain research. In particular, eicosapentaenoic acid (EPA; 20:5n-3) and docosahexaenoic 1866 acid (DHA; 22:6n-3), which are known to be associated with the intake of diatoms and 97 1867 dinoflagellates respectively (Bottino, 1974, Ericson et al., 2018a), are needed for production of 1868 krill eggs before spawning, and for the development of the larval krill (Yoshida, 2009). EPA 1869 and DHA are consistently abundant in krill and make up a large part of the krill fatty acid 1870 profile (Kolakowska et al., 1994, Ericson et al., 2018a). Other sources for EPA and DHA may 1871 also exist. The specific source(s) of other n-3 LC-PUFA such as eicosatetraenoic acid (ETA, 1872 20:4n-3) and docosapentaenoic acid (DPA, 22:5n-3) are not as well defined. EPA and DHA 1873 are consistently abundant in krill and make up a large part of the krill fatty acid profile 1874 (Kolakowska et al., 1994, Ericson et al., 2018a), particularly in summer and autumn. EPA and 1875 DHA are also the n-3 LC-PUFA targeted by the krill fishing industry for application into 1876 nutraceutical products (Nicol et al., 2012, Hill, 2013, Schutt, 2016). The n-3 PUFA, stearidonic 1877 acid (SDA, 18:4n-3), also plays a vital role in krill diet, although its precise function is not well 1878 understood (Ericson et al., 2018a). 1879 Phytoplankton, such as diatoms and dinoflagellates, all naturally produce chlorophyll which 1880 can be remotely detected via satellites (Moore and Abbott, 2002, Johnson et al., 2013, Zeng et 1881 al., 2016) using ocean colour data. Recent studies have shown that the colour of the ocean, due 1882 to shifts in the assemblage of these phytoplankton blooms (Deppeler and Davidson, 2017), is 1883 changing with climate change (Dutkiewicz et al., 2019). As these phytoplankton assemblages 1884 change, krill diet may also be altered as the climate changes (Deppeler and Davidson, 2017, 1885 Hancock et al., 2018). Phytoplankton biomarker levels will have more pronounced changes 1886 within krill diet as lower trophic level populations will shift more rapidly with climate change 1887 (Deppeler and Davidson, 2017) than higher trophic levels (Fraser and Hofmann, 2003), such 1888 as grazers like copepods and krill (Schofield et al., 2010). Therefore, the biochemical 1889 composition of krill will shift year round from that of a seasonal winter diet which includes 1890 copepods (omnivorous) to a summer diet (herbivorous) (Ericson et al., 2018a) for a larger part 1891 of the year as sea ice is lost and waters warm (increase in sea surface temperature, SST) and 98 1892 become more acidic, allowing for greater phytoplankton blooms to occur (Montes-Hugo et al., 1893 2009, Behrenfeld et al., 2017, Deppeler and Davidson 2017). It is difficult, however, to link 1894 changes in the marine environment to changes in krill biochemistry in situ. Controlled 1895 aquarium experiments are difficult to conduct over longer time scales and cannot emulate 1896 conditions over large geographic areas. Understanding these large-scale relationships requires 1897 data collected over wide areas and long timeframes. Hence, using remotely-sensed satellite 1898 chlorophyll a (Chl a) data as a proxy for primary production is optimal for data collection that 1899 can happen simultaneously over large geographic areas and over span long timeframes (Zeng 1900 et al., 2016, Behrenfeld et al., 2017). Similarly, SST is a major environmental driver of the 1901 phytoplankton blooms during spring and summer in the Southern Ocean (El-Sayed and Weber, 1902 1982, Helbling et al., 1995, Moline et al., 1997, Garibotti et al., 2005); increases in SST are a 1903 major cause of ecosystem level shifts (e.g. phytoplankton assemblage change) with climate 1904 change (Montes-Hugo et al., 2009, Behrenfeld et al., 2017, Deppeler and Davidson, 2017). 1905 Remotely-sensed SST data from satellites can also be collected over large geographic areas 1906 and long timeframes. Unfortunately, as of yet, a proxy for zooplankton, marine snow or 1907 bacterial assemblages that can be detected via satellite are not available. However, biological 1908 and biochemical data from krill samples are able to be collected and analysed over a large 1909 region of the Southern Ocean and over a long timeframe by the krill fishery, such as in Tarling, 1910 et al. (2016a). 1911 The krill fishery has been operating since the early 1970’s (Nicol and Foster, 2016), 1912 and in the mid-1990s the industry began to produce and sell krill oil as a nutraceutical due to 1913 its high levels of omega-3 containing oils (Nicol et al., 2012). The krill fishery operates year- 1914 round in the South Atlantic Ocean (Commission for the Conservation of Antarctic Marine 1915 Living Resources (CCAMLR) Area 48) and in particular near the West Antarctic Peninsula 1916 (WAP, CCAMLR Sub-Area 48.1), South Orkney Islands (SOI, CCAMLR Sub-Area 48.2) and 99 1917 South Georgia (SG, CCAMLR Sub-Area 48.3). Over the last few years the total catch has been 1918 approximately 300,000 tonnes of krill a year, and from a scientific research perspective the 1919 commercial harvest is a useful source of biological samples (Tarling et al., 2016a, Ericson et 1920 al., 2018a, Hellessey et al., 2018). Satellites collect environmental data such as SST, ocean 1921 colour, sea surface height, fluorescence, wind direction and wind speed. While satellite data 1922 must be calibrated and validated, it can fill in gaps of data collected in situ and provides wide- 1923 area coverage. Linking remotely-sensed environmental data with analysis of samples collected 1924 by the krill fishery is a cost-effective approach to explore biological responses to environmental 1925 changes. 1926 Our study uses three consecutive years of krill lipid data from fishery-derived samples 1927 collected throughout the South Atlantic Ocean. We simultaneously accessed satellite data to 1928 link krill signature biochemical data to environmental conditions seen at the location of krill 1929 collection. We hypothesise that changes in krill biochemistry, specifically their fatty acid 1930 dietary biomarkers, will track broad scale satellite-derived environmental data. Linking 1931 environmental drivers to krill diet will assist in ecosystem energy budget and food web models. 1932 By basing such models around environmental parameters, future environmental scenarios can 1933 be modelled, and krill diet and ecosystem responses will be more reliably predicted. 1934 We aim to show if a link between the environment and krill diet exists by examining 1935 whether: 1) chlorophyll a levels (Chl a, mg m-2), total lipid (mg g-1 krill dry mass) and 1936 phytoplankton fatty acid biomarkers (percentage (%) and mass (ug)) increase as SST (°C) 1937 decreases throughout summer and autumn, post the annual spring phytoplankton blooms; 2) 1938 larger shifts in SST and Chl a closer to the pole (currently an area in a state of flux) will drive 1939 larger shifts in krill biochemistry and 3) SST and Chl a is correlated to total lipid or 1940 phytoplankton fatty acid biomarkers in a meaningful manner. 100 1941 METHODS 1942 Krill sample collection and analysis 1943 Krill lipid and fatty acid data used for this analysis are published in Hellessey et al. (2018) and 1944 Ericson et al. (2018a). Briefly, samples were collected by the FV Saga seas using a continuous 1945 underwater pumping system in the South Atlantic Ocean (Area 48). Samples were collected 1946 from January 2014 – September 2016 in the WAP (Sub-Area 48.1), the SOI (Sub-Area 48.2), 1947 and SG (Sub-Area 48.3). Krill were stored at -80°C and transported on dry ice to Hobart, 1948 Tasmania for lipid and fatty acid analysis. 1949 3 adult male and 3 adult female krill were analysed each fortnight, and these lipid and fatty 1950 acid profiles were pooled (N=6 per fortnight, total N = 391). Samples were quantitatively 1951 extracted overnight using a modified Bligh and Dyer (1959) method as described in Hellessey 1952 et al. (2018) to produce the total solvent extract (TSE). The total lipid content (TL, expressed 1953 as mg) of each sample was weighed gravimetrically in a pre-weighed 2 ml glass vial. To 1954 account for differences in krill size, the TL was divided by krill dry mass (g) and is expressed 1955 as mg of total lipid content per gram of krill dry mass (mg g-1, TL DM). Lipid class composition 1956 was determined by analysis of the TSE on an Iatroscan TLC-FID analyser following Hellessey 1957 et al. (2018). Aliquots of sample TSEs were methylated to extract the fatty acid methyl esters 1958 (FAME) of the sample (Ericson et al. 2018a). Samples were made up with 1 ml internal 1959 injection standard (23:0 FAME) and analysed by gas chromatography (GC-FID) (Iverson et al., 1960 2004). Samples were injected (0.2 µl) and the identification and quantification of fatty acids 1961 (expressed as % total fatty acid area) was conducted in comparison to a commercial standard 1962 mix (sourced from Sigma) and a known laboratory standard (tuna oil). Fatty acid identifications 1963 were further confirmed through gas chromatography-mass spectrometry (GC-MS) analyses 1964 (Iverson et al., 2004). 1965 101 1966 Satellite data extraction and analysis 1967 This study used ocean colour data from the NASA Moderate Resolution Imaging 1968 Spectroradiometer 1969 (https://oceancolor.gsfc.nasa.gov/data/aqua/) and sea surface temperature data from the 1970 GHRSST L4 gridded products (https://data.nodc.noaa.gov/ghrsst/L4/). The sea surface 1971 temperature data and the 3 different ocean colour Remote Sensed Reflectance wavelengths 1972 (RRS; red, green and blue: 443, 488 and 555 nm, respectively) were extracted for each date 1973 krill were collected in an area in the south Atlantic Ocean (bounds of 55-80 °S and 30-80 °W) 1974 in a 1 km x 1 km grid of pixels. The exact GPS location of krill collection on that date was then 1975 used to extract the pixel value (28 successful matches = 4.18% matched, Table 4.1), but due to 1976 the low match-up rate this was expanded both temporally and spatially to an 8-day (8D) average 1977 and a 3 km x 3 km (3x3) pixel area (145 matches = 21.64% matched, Table 4.1). Whenever 1978 data was patchy, it was smoothed linearly to the nearest pixel within a 4 km area. Aqua (MODIS-Aqua) 102 L3 mapped data products 1979 Table 4.1: Decision table for why we increased our temporal and spatial fields for the red, 1980 green and blue wavelengths in Moderate Resolution Imaging Spectroradiometer (MODIS) to 1981 generate chlorophyll a data. The Commission for Conservation of Antarctic Marine Living 1982 Resources (CCAMLR) regions were defined as the West Antarctic Peninsula (WAP, Area 1983 48.1), the South Orkney Islands (SOI, Area 48.2) and South Georgia (SG, Area 48.3) 1984 (www.ccamlr.org). Raw value for percent data match in brackets. Total number of days with 1985 lipid data to match against = 670. Case Temporal averaging Pixel averaging Percentage (%) match 1 Daily 1 km x 1 km 4.18 (28) 2 Daily 3 km x 3 km 7.01 (47) 3 8 Day 1 km x 1 km 21.64 (145) 4 8 Day 3 km x 3 km 27.91 (187) 5 8 Day Custom WAP - 46.26 (310) CCAMLR SOI - 51.34(344) Regions SG - 66.86 (448) 103 1986 The ocean colour RRS data were converted into Chl a concentrations using the MODIS 1987 Southern Ocean chlorophyll algorithm in Johnson et al. (2013). Once converted into Chl a 1988 concentrations, this environmental data (SST and Chl a) was merged into the same data frame 1989 as the lipid data by matching the date and GPS location of 1 day 1x1 pixel locations to the date 1990 and GPS location of krill harvest. To examine the seasonal trend for Chl a, each CCAMLR 1991 fishing sub-area within the South Atlantic (Area 48) also had its Chl a calculated for 8-day 1992 averages. These wider geographic areas of the WAP, the SOI and SG generated much higher 1993 recovery rates for Chl a (46.26%, 51.34% and 66.86% respectively, Table 4.1) and is hereon 1994 called Chl a (CCAMLR). This data was also merged into the same data frame by matching 1995 dates and GPS locations of krill sampling, as done previously. 1996 However, data for some dates did not match krill harvesting location (e.g. Chl a data 1997 from SG on a day when krill were collected from WAP), so the rates of Chl a recovery 1998 decreased once matched to krill lipid data. Therefore, to achieve the best Chl a matches to lipid 1999 data, the Chl a concentration was kept in a hierarchy from Case 1 to 5 (Table 4.1). For example, 2000 if a daily pixel match was available this was kept in preference over an 8-daya 3x3 or 8-day 2001 regional Chl a concentration. This method increased the overall match rate to 226 matches = 2002 60.59% matched. This merged Chl a data ( hereon called Chl a (overall)) was used to examine 2003 the larger scale trends in Chl a across the entire south Atlantic Ocean and krill lipids. 2004 Data and Statistical analysis 2005 Statistical analysis was done in RStudio (version 1.0.153 © 2017) using packages: nlme 2006 (Pinheiro et al., 2017), ggplot2 (Wickham, 2009), ggmap (Kahle and Wickham, 2013), maps 2007 (Becker, 2017), anytime (Eddelbuettel, 2018), and reshape2 (Wickham, 2007). Multifactorial 2008 ANOVAs were performed using SST, Chl a, as well as their interaction terms, as factors for 2009 the variables of total lipid content (mg g-1 dry mass), PL and TAG percentage, individual fatty 2010 acid percentage and mass data for the fatty acids most associated with primary production 104 2011 (mostly diatom and flagellate markers), and the ratios of 16:1/16:0 to look at diatom levels 2012 (Ericson et al., 2018a) and EPA:DHA to look at diatom or flagellate dominance in the diet 2013 (Ericson et al., 2018a). Data were log or square root transformed when the assumptions of 2014 normality and homogeneity of variances were not met - Chl a is typically log distributed. Linear 2015 models were similarly produced using the same factors as for the multifactorial ANOVAs, 2016 including interaction terms, but sub-divided into summer/autumn and winter/spring models due 2017 to the large seasonal shift in SST arising from the change of harvesting location by the FV Saga 2018 Seas. Models were tested for fit using a standard regression table, where the adjusted r2 value 2019 showed the fit of points to the confidence interval of the model. Models were additionally run 2020 through drop testing and Tuckey post-hoc tests to ensure no compounding of results was 2021 occurring. Models of best fit had adjusted r2 values > 0.5, a P value of < 0.05 from the 2022 multifactorial ANOVAs and a χ2 value above 0.1. These models are shown within all tables 2023 throughout the Results as greyed out. 2024 Maps were produced within R using the maps and ggmap packages to see the 2025 geographic distribution of krill lipid content (mg g-1 dry mass) as well as SST and Chl a. 2026 RESULTS 2027 Seas Surface Temperatures 2028 During this study (Jan 2014 – Sep 2016) satellite-derived sea surface temperatures had the 2029 greatest variability in the South Orkney Islands (SOI), ranging from -1.28 to 1.93 °C (average 2030 0.33 ± 0.86 °C), followed by South Georgia (SG) where temperatures ranged from -1.10 to 2031 1.49 °C (average 0.49 ± 0.54 °C) and the West Antarctic Peninsula (WAP) which ranged from 2032 -1.28 to 0.94 °C (average -0.25 ± 0.78 °C). Temperatures decreased from summer into autumn 2033 in the WAP and SOI, and from the start of winter to spring in SG. However, in 2016, SST at 2034 WAP increased from summer into autumn and then decreased from early autumn onwards 2035 (Figure 4.1). 105 2036 2037 2038 Figure 4.1: Sea surface temperatures (°C) from January 2014 – September 2016 coloured by Euphausia superba sample location (SG: South Georgia, SOI: South Orkney Islands, WAP: West Antarctic Peninsula). The x-axis is the season and year of krill sample collection. 106 2039 Chlorophyll a Concentrations 2040 The levels of Chl a varied depending on the location and the season (Figure 4.2). Values of 2041 more than 5.78 mg m-2 of Chl a were observed in the SOI, and as low as 0.19 mg m-2 Chl a in 2042 SG. Chl a increased in summer and decreased rapidly during the autumn of 2014 around the 2043 SOI. Chl a concentrations in other years were more consistent, although very few satellite 2044 measurements were recorded at SG due to the time of year, cloud cover, ice cover and 2045 suboptimal sun angle. 2046 Chl a varied the most around the SOI, ranging from 0.24 to 5.78 mg m-2 (average 0.88 2047 ± 0.85 mg m-2), followed by the WAP which ranged from 0.20 to 1.34 mg m-2 (average 0.62 ± 2048 0.30 mg m-2), and SG which ranged from 0.19 to 0.62 mg m-2 (average 0.39 ± 0.11 mg m-2). 2049 SG had the smallest range of Chl a concentrations and the lowest average Chl a concentration. 2050 SOI had the biggest range of Chl a concentrations and the highest average Chl a concentration. 2051 107 2052 2053 2054 Figure 4.2: Chlorophyll a concentrations (mg m-2) from January 2014 – September 2016 coloured by Euphausia superba sample location (SG: South Georgia, SOI: South Orkney Islands, WAP: West Antarctic Peninsula). The x-axis is season and year of krill sample collection. 108 2055 Geographic distribution 2056 The geographic distribution of SST, Chl a and krill total lipid (mg g-1) dry mass (TL DM), for 2057 CCAMLR Sub-Areas 48.1, 48.2 and 48.3 can be seen in Figures 4.3, 4.4 and 4.5. . Higher SST 2058 and Chl a in the Bransfield Strait (WAP) reflect bathymetric and current features (Fig. 4.3) as 2059 well as the northern advection of water from the deep canyon to the south of the SOI (Korb et 2060 al., 2005) (Fig. 4.4). These oceanographic features are known locations of higher SST and Chl 2061 a within the SG and SOI areas. Krill TL DM is also higher in the Bransfield Strait (Fig. 4.3). 2062 To the northwest of SG, higher SST and lower Chl a values can be seen due to the faster flowing 2063 current coming off of the slope of SG ( Fig. 4.5). 109 2064 2065 Figure 4.3: The geographic distribution of krill total lipid (mg g-1) dry mass (TLDM), the chlorophyll a (Chl a) concentration (mg m-2) and the 2066 sea surface temperature (°C, SST) of Euphausia superba samples collected in the West Antarctic Peninsula. Locations are points that krill were 2067 harvested by FV Saga Seas from January to May 2014 - 2016. Maps were produced using the RStudio (version 1.0.153 © 2017) package 2068 ggmaps (Kahle and Wickham, http://journal.r-project.org/archive/2013-1/kahle-wickham.pdf). Map data © 2018 Google. 110 2069 2070 Figure 4.4: The geographic distribution of krill total lipid (mg g-1) dry mass (TLDM), the chlorophyll a (Chl a) concentration (mg m-2) and the 2071 sea surface temperature (°C, SST) of Euphausia superba samples collected in the South Orkney Islands. Locations are points that krill were 2072 harvested by FV Saga Seas from January to May 2014 - 2016. Maps were produced using the RStudio (version 1.0.153 © 2017) package 2073 ggmaps (Kahle and Wickham, http://journal.r-project.org/archive/2013-1/kahle-wickham.pdf). Map data © 2018 Google. 2074 111 2075 2076 Figure 4.5: The geographic distribution of krill total lipid (mg g-1) dry mass (TLDM), the chlorophyll a (Chl a) concentration (mg m-2) and the 2077 sea surface temperature (°C, SST) of Euphausia superba samples collected at South Georgia. Locations are points that krill were harvested by 2078 FV Saga Seas from June to September 2014 - 2016. Maps were produced using the RStudio (version 1.0.153 © 2017) package ggmaps (Kahle 2079 and Wickham, http://journal.r-project.org/archive/2013-1/kahle-wickham.pdf). Map data © 2018 Google. 2080 112 2081 Lipid and fatty acid general trends 2082 Results describing the trends seen in krill TLDM, lipid classes and their fatty acids can be 2083 found in Ericson, et al. 7 and Hellessey, et al. 46. Briefly, krill TLDM and triacylglycerol (TAG) 2084 percentage (%) increased throughout summer to reach autumn highs in the WAP and SOI, 2085 whereas krill at SG had declining TLDM and TAG % throughout winter and spring. EPA and 2086 DHA (mg g-1 dry weight) followed the same seasonal trend as TLDM and TAG %. 16:1n-7c 2087 and SDA had variable quantities across all seasons, years and fishing locations. In summer, 2088 krill had high levels (% total fatty acids) of EPA, DHA and PUFA, but low 18:1n-9c/18:1n-7c 2089 ratios, indicating a more herbivorous diet. 2090 Tracking fatty acid biomarkers using satellite derived environmental data 2091 Clear seasonal trends can be seen in fatty acid biomarkers in krill throughout the fishing seasons 2092 both in percentage composition and quantitative amounts (mass, ug). Most diatom-based 2093 markers in krill (such as EPA, 16:4n-1 and 16:1n-7c) increased in percentage throughout 2094 summer to reach autumn highs in the WAP and SOI, as did 16:0 percentages, whereas krill at 2095 SG had declining diatom-based marker and 16:0 percentages throughout winter and spring. 2096 Dinoflagellate markers in krill such as DHA and SDA showed similar trends to diatom markers 2097 in their percentages. Both diatom and flagellate markers in krill showed the opposite trend in 2098 their masses (low in summer and autumn, higher in winter and spring) but this could be due to 2099 location of sampling (WAP/SOI in summer/autumn and SG in winter/spring). Table 4.2 2100 provides the p-values for 1-way ANOVAs comparing models as well as the adjusted r2 value 2101 and χ2 value for the associated model of best fit between the krill’s biochemical data (lipid and 2102 fatty acid content (mass) and composition (percentage)) and the environmental data (SST and 2103 Chl a) and their interaction terms from the South Atlantic region. Tables 4.3-4.5 show these 2104 same relationships broken down into the smaller CCAMLR management sub-areas (WAP, SOI 2105 and SG, respectively). 113 2106 2107 2108 2109 2110 2111 2112 Table 4.2: Euphausia superba total lipid (mg g-1) dry mass (TLDM), and lipid class composition (phospholipid (PL) and triacylglycerol (TAG) percentage) and fatty acid (20:5n-3 (EPA), 22:6n-3 (DHA), and 18:4n-3 (SDA)) percentage composition (%) and mass (ug) per krill against sea surface temperature (SST), chlorophyll a (Chl a) and their interaction terms for all seasons and pooled locations across the South Atlantic sector. Chl a was measured at both an overall scale (overall) and an 8-day 3 km x 3 km (8D 3x3) pixel scale for the entire South Atlantic sector. Values given are for: P values, r2 values (italics) and χ2 values (bold) for the model of best fit. Cells that are greyed out have a P value < 0.05, an r2 of >0.5 and a χ2 value > 0.1. SST Chl a (overall) Chl a (8D 3x3) SST*Chl a (overall) SST*Chl a (8D 3x3) TLDM (mg g ) PL % TAG % < 0.001 (0.144) 0.086 0.296 (0.000) 0.003 0.001 (0.026) 0.001 0.948 (-0.004) 0.018 < 0.0001 (0.061) 0.006 0.344 (-0.000) <0.001 0.434 (-0.005) 0.238 < 0.0001 (0.255) 0.250 0.035 (0.043) 0.087 0.808 (0.045) 0.082 0.176 (0.069) 0.029 0.480 (-0.006) 0.007 0.007 (0.134) 0.238 0.074 (0.376) 0.249 0.015 (0.193) 0.087 EPA % EPA (ug) 0.001 (0.025) 0.080 < 0.001 (0.167) 0.239 0.426 (-0.002) 0.018 0.664 (-0.004) 0.239 0.006 (0.078) 0.241 0.667 (-0.010) 0.238 < 0.0001 (0.085) 0.241 0.385 (0.051) 0.239 0.372 (0.068) 0.241 0.058 (0.074) 0.238 DHA % DHA (ug) < 0.001 (0.065) 0.082 < 0.001 (0.158) 0.239 0.109 (0.007) 0.081 0.534 (-0.003) 0.239 0.194 (0.009) 0.240 0.102 (0.021) 0.238 0.203 (0.059) 0.242 0.851 (0.067) 0.239 0.918 (-0.016) 0.240 0.048 (0.117) 0.238 SDA % SDA (ug) 0.078 (0.006) 0.003 < 0.001 (0.141) 0.239 0.626 (-0.003) 0.003 0.603 (-0.003) 0.239 0.446 (-0.005) 0.240 0.921 (-0.012) 0.238 0.013 (0.024) 0.082 0.397 (0.052) 0.239 0.612 (0.189) 0.240 0.422 (0.096) 0.238 16:0 % 16:0 (ug) 0.001 (0.037) 0.083 < 0.001 (0.182) 0.239 0.588 (-0.003) 0.082 0.920 (-0.004) 0.239 0.943 (-0.012) 0.244 0.555 (-0.008) 0.238 0.034 (0.034) 0.243 0.759 (0.068) 0.239 0.417 (-0.018) 0.244 0.091 (0.042) 0.239 16:4n-1 % 16:4n-1 (ug) 0.925 (-0.003) < 0.001 < 0.001 (0.103) 0.239 0.451 (-0.002) <0.001 0.407 (-0.001) 0.239 0.334 (-0.000) 0.019 0.337 (-0.000) 0.238 0.939 (-0.010) 0.003 0.787 (0.008) 0.239 0.174 (0.126) 0.019 0.899 (0.125) 0.238 6:1n-7c % 16:1n-7c (ug) 0.115 (0.004) 0.082 < 0.001 (0.143) 0.239 0.982 (-0.004) 0.080 0.972 (-0.004) 0.239 0.583 (-0.009) 0.241 0.456 (-0.005) 0.238 0.061 (0.015) 0.241 0.539 (0.059) 0.239 0.989 (-0.034) 0.241 0.069 (0.043) 0.238 16:1/16:0 ratio (ug) EPA/DHA ratio (ug) 0.588 (-0.002) 0.239 0.621 (-0.002) 0.239 0.781 (-0.004) 0.239 0.944 (-0.004) 0.239 0.543 (-0.008) 0.238 0.008 (0.073) 0.238 0.136 (0.006) 0.239 0.730 (-0.012) 0.239 0.950 (-0.034) 0.238 0.625 (0.060) 0.238 Phytanic acid % Phytanic acid (ug) 0.013 (0.014) <0.001 0.009 (0.016) <0.001 0.047 (0.013) <0.001 0.052 (0.012) <0.001 0.149 (0.014) <0.001 0.149 (0.014) <0.001 0.309 (0.012) <0.001 0.313 (0.009) <0.001 0.362 (0.098) <0.001 0.250 (0.127) <0.001 -1 114 2113 2114 2115 2116 2117 2118 Table 4.3: Euphausia superba (collected from the West Antarctic Peninsula) total lipid (mg g-1) dry mass (TLDM), lipid class (phospholipid (PL) and triacylglycerol (TAG)) and fatty acid (20:5n-3 (EPA), 22:6n-3 (DHA) and 18:4n-3 (SDA)) percentage composition (%) and mass (ug) in relation to sea surface temperature (SST), chlorophyll a (Chl a) and their interaction terms. Chl a was measured at both a Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) region wide scale (CCAMLR region) and at an 8-day 3 km x 3 km (8D 3x3) pixel scale. Values given are for: P values, r2 values (italics) and χ2 values (bold) for the model of best fit. Cells that are greyed out have a P value < 0.05, an r2 of >0.5 and a χ2 value > 0.1. SST < 0.001 (0.304) 0.240 Chl a (CCAMLR) 0.014 (0.148) 0.238 Chl a (8D 3x3) 0.185 (0.059) 0.235 SST*Chl a (CCAMLR) 0.001 (0.363) 0.238 SST*Chl a (8D 3x3) 0.002 (0.610) 0.235 0.006 (0.077) 0.028 0.007 (0.073) 0.102 0.004 (0.212) 0.089 0.001 (0.277) 0.023 0.949 (-0.071) 0.235 0.782 (-0.065) 0.235 < 0.001 (0.564) 0.089 < 0.001 (0.559) 0.089 0.001 (0.655) 0.235 0.004 (0.568) 0.235 EPA % EPA (ug) 0.001 (0.137) 0.240 < 0.001 (0.217) 0.239 0.012 (0.157) 0.238 0.079 (0.065) 0.238 0.388 (-0.014) 0.235 0.059 (0.175) 0.235 < 0.001 (0.559) 0.238 0.001 (0.311) 0.238 < 0.001 (0.689) 0.235 0.002 (0.615) 0.235 DHA % DHA (ug) 0.001 (0.122) 0.239 < 0.001 (0.285) 0.239 0.254 (0.010) 0.238 0.357 (-0.004) 0.238 0.001 (0.498) 0.235 0.624 (-0.053) 0.235 0.028 (0.148) 0.238 < 0.001 (0.430) 0.238 0.014 (0.467) 0.235 < 0.001 (0.892) 0.235 SDA % SDA (ug) 0.795 (-0.011) 0.083 < 0.001 (0.311) 0.239 0.187 (0.024) 0.086 0.020 (0.131) 0.238 0.009 (0.351) 0.235 0.372 (-0.010) 0.235 0.541 (0.098) 0.087 0.001 (0.390) 0.238 0.068 (0.295) 0.235 0.004 (0.560) 0.235 16:0 % 16:0 (ug) 0.001 (0.122) 0.240 < 0.001 (0.254) 0.239 0.048 (0.089) 0.238 0.059 (0.078) 0.238 0.012 (0.325) 0.235 0.097 (0.126) 0.235 0.014 (0.214) 0.238 0.001 (0.347) 0.238 0.014 (0.464) 0.235 0.002 (0.621) 0.235 16:4n-1 % 16:4n-1 (ug) 0.527 (-0.007) 0.021 < 0.001 (0.197) 0.239 0.945 (-0.031) 0.245 0.059 (0.079) 0.238 0.004 (0.412) 0.235 0.041 (0.213) 0.235 0.982 (0.189) 0.245 0.008 (0.280) 0.238 0.003 (0.589) 0.235 0.093 (0.253) 0.235 16:1n-7c % 16:1n-7c (ug) 0.130 (0.015) 0.240 < 0.001 (0.185) 0.239 0.625 (-0.023) 0.240 0.083 (0.063) 0.238 0.003 (0.446) 0.235 0.080 (0.145) 0.235 0.104 (0.043) 0.241 0.001 (0.302) 0.238 0.026 (0.406) 0.235 0.014 (0.468) 0.235 16:1/16:0 ratio (ug) EPA/DHA ratio (ug) 0.445 (-0.005) 0.239 0.031 (0.043) 0.239 0.897 (-0.031) 0.238 0.181 (0.026) 0.238 0.001 (0.489) 0.235 0.003 (0.436) 0.235 0.192 (0.012) 0.238 0.188 (0.162) 0.238 0.017 (0.449) 0.235 0.040 (0.358) 0.235 Phytol % Phytol (ug) 0.426 (-0.004) <0.001 0.816 (-0.011) <0.001 0.013 (0.156) <0.001 0.014 (0.149) <0.001 0.859 (-0.074) <0.001 0.799 (-0.066) <0.001 0.039 (0.246) <0.001 0.145 (0.185) <0.001 0.072 (0.308) <0.001 0.046 (0.342) <0.001 -1 TLDM (mg g ) PL % TAG % 2119 115 2120 2121 2122 2123 2124 2125 Table 4.4: Euphausia superba (collected from South Orkney Islands) total lipid (mg g-1) dry mass (TLDM), lipid class (phospholipid (PL) and triacylglycerol (TAG)) and fatty acid (20:5n-3 (EPA), 22:6n-3 (DHA) and 18:4n-3 (SDA)) percentage composition (%) and mass (ug) in relation to sea surface temperature (SST), chlorophyll a (Chl a) and their interaction terms. Chl a was measured at both a Commission for the Conservation for Antarctic Marine Living Resources (CCAMLR) region wide scale (CCAMLR region) and at an 8-day 3 km x 3 km (8D 3x3) pixel scale. Values given are for: P values, r2 values (italics) and χ2 values (bold) for the model of best fit. Cells that are greyed out have a P value < 0.05, an r2 of >0.5 and a χ2 value > 0.1. SST TLDM (mg g ) < 0.001 (0.168) 0.081 Chl a (CCAMLR) 0.003 (0.059) 0.018 Chl a (8D 3x3) 0.124 (0.034) 0.238 SST*Chl a (CCAMLR) < 0.001 (0.217) 0.081 SST*Chl a (8D 3x3) 0.672 (0.084) 0.238 PL % TAG % 0.652 (-0.006) 0.028 0.004 (0.054) 0.006 0.001 (0.102) 0.005 < 0.001 (0.121) <0.001 0.001 (0.245) 0.251 0.731 (-0.021) 0.085 0.098 (0.118) 0.027 0.049 (0.164) 0.007 0.006 (0.209) 0.251 0.587 (-0.041) 0.085 EPA % EPA (ug) 0.001 (0.082) 0.079 < 0.001 (0.122) 0.239 0.001 (0.078) 0.017 0.229 (0.004) 0.239 0.002 (0.188) 0.241 0.007 (0.146) 0.238 0.296 (0.112) 0.080 0.082 (0.184) 0.239 0.021 (0.157) 0.241 0.029 (0.142) 0.238 DHA % DHA (ug) 0.001 (0.092) 0.081 < 0.001 (0.123) 0.239 0.016 (0.037) 0.018 0.070 (0.018) 0.239 0.198 (0.017) 0.241 0.529 (-0.014) 0.238 0.426 (0.102) 0.081 0.345 (0.186) 0.239 0.364 (-0.010) 0.241 0.594 (0.026) 0.238 SDA % SDA (ug) 0.121 (0.010) 0.081 0.001 (0.089) 0.239 0.152 (0.008) 0.018 0.057 (-0.003) 0.239 0.122 (0.034) 0.241 0.417 (0.057) 0.238 0.263 (0.065) 0.081 0.910 (0.139) 0.239 0.042 (0.126) 0.241 0.415 (0.034) 0.238 16:0 % 16:0 (ug) 0.007 (0.045) 0.079 < 0.001 (0.143) 0.239 0.063 (0.019) 0.017 0.095 (0.014) 0.239 0.292 (0.003) 0.241 0.382 (-0.005) 0.238 0.997 (0.032) 0.080 0.011 (0.209) 0.239 0.501 (-0.035) 0.241 0.635 (-0.017) 0.238 16:4n-1 % 16:4n-1 (ug) 0.305 (0.000) 0.082 0.051 (0.021) 0.239 0.819 (-0.007) 0.019 0.541 (-0.005) 0.239 0.045 (0.073) 0.241 0.003 (0.171) 0.238 0.606 (-0.021) 0.083 0.263 (0.052) 0.239 < 0.001 (0.367) 0.241 0.001 (0.280) 0.238 16:1n-7c % 16:1n-7c (ug) 0.111 (0.011) 0.079 < 0.001 (0.115) 0.239 0.027 (0.030) 0.017 0.090 (0.015) 0.239 0.496 (-0.013) 0.241 0.573 (-0.016) 0.238 0.415 (0.024) 0.080 0.002 (0.192) 0.239 0.078 (0.028) 0.241 0.266 (-0.016) 0.238 16:1/16:0 ratio (ug) EPA/DHA ratio (ug) 0.289 (0.001) 0.239 0.686 (-0.006) 0.239 0.072 (0.017) 0.239 0.482 (-0.004) 0.239 0.751 (-0.022) 0.238 0.001 (0.271) 0.238 0.320 (0.010) 0.239 0.522 (-0.016) 0.239 0.077 (0.024) 0.238 0.001 (0.285) 0.238 Phytol % Phytol (ug) 0.046 (0.022) <0.001 0.199 (0.005) <0.001 0.353 (-0.001) <0.001 0.246 (0.003) <0.001 0.507 (-0.013) 0.001 0.507 (-0.013) 0.001 0.549 (0.019) <0.001 0.706 (0.002) <0.001 0.457 (-0.001) 0.001 0.457 (-0.001) 0.001 -1 2126 116 2127 2128 2129 2130 2131 2132 Table 4.5: Euphausia superba (collected from South Georgia) total lipid (mg g-1) dry mass (TLDM), lipid class (phospholipid (PL) and triacylglycerol (TAG)) and fatty acid (20:5n-3 (EPA), 22:6n-3 (DHA) and 18:4n-3 (SDA)) percentage composition (%) and mass (ug) in relation to sea surface temperature (SST), chlorophyll (Chl a) and their interaction terms. Chl a was measured at both a Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) region wide scale (CCAMLR region) and at an 8-day 3 km x 3 km (8D 3x3) pixel scale. Values given are for: P values, r2 values (italics) and χ2 values (bold) for the model of best fit. Cells that are greyed out have a P value < 0.05, an r2 of >0.5 and a χ2 value > 0.1. SST TLDM (mg g ) 0.003 (0.054) 0.246 Chl a (CCAMLR) 0.219 (0.008) 0.240 Chl a (8D 3x3) 0.084 (0.095) 0.236 SST*Chl a (CCAMLR) < 0.001 (0.299) 0.240 SST*Chl a (8D 3x3) 0.336 (0.337) 0.236 PL % TAG % 0.103 (0.115) 0.264 0.206 (0.004) 0.102 0.013 (0.083) 0.249 0.010 (0.088) 0.249 0.019 (0.195) 0.236 0.083 (0.095) 0.241 0.001 (0.297) 0.249 0.001 (0.292) 0.249 0.001 (0.301) 0.236 0.163 (0.194) 0.241 EPA % EPA (ug) 0.095 (0.012) 0.241 0.035 (0.023) 0.239 0.083 (0.033) 0.078 0.105 (0.027) 0.239 0.001 (0.419) 0.236 0.178 (0.041) 0.236 0.002 (0.235) 0.240 0.001 (0.179) 0.239 0.003 (0.437) 0.236 0.645 (0.032) 0.236 DHA % DHA (ug) 0.863 (-0.007) 0.241 0.026 (0.029) 0.239 0.003 (0.120) 0.240 0.602 (-0.012) 0.239 0.001 (0.377) 0.236 0.052 (0.128) 0.236 0.001 (0.280) 0.240 0.001 (0.149) 0.239 <0.001 (0.611) 0.236 0.817 (0.072) 0.236 SDA % SDA (ug) 0.005 (0.046) 0.087 0.007 (0.042) 0.239 0.076 (0.035) 0.019 0.083 (-0.010) 0.239 0.594 (-0.033) 0.236 0.551 (0.095) 0.236 0.583 (0.121) 0.085 0.409 (0.162) 0.238 0.991 (0.062) 0.236 0.029 (0.269) 0.236 16:0 % 16:0 (ug) 0.173 (0.006) 0.241 0.277 (0.001) 0.239 0. 018 (0.072) 0.242 0.008 (0.095) 0.239 0.036 (0.155) 0.246 0.001 (0.381) 0.236 0.022 (0.163) 0.242 0.001 (0.281) 0.239 0.642 (0.216) 0.246 0.001 (0.497) 0.236 16:4n-1 % 16:4n-1 (ug) < 0.001 (0.104) 0.004 0.001 (0.077) 0.239 0.236 (0.007) 0.004 0.841 (-0.016) 0.239 0.046 (0.137) 0.023 0.481 (-0.022) 0.236 0.903 (0.136) 0.004 0.030 (0.119) 0.239 0.314 (0.122) 0.023 0.472 (-0.050) 0.236 16:1n-7c % 16:1n-7c (ug) 0.756 (-0.006) 0.241 0.330 (-0.000) 0.239 0.001 (0.103) 0.239 0.003 (0.125) 0.239 0.007 (0.265) 0.236 0.001 (0.443) 0.236 0.001 (0.279) 0.239 < 0.001 (0.336) 0.239 0.956 (0.247) 0.236 <0.001 (0.551) 0.236 16:1/16:0 ratio (ug) EPA/DHA ratio (ug) 0.985 (-0.007) 0.239 0.245 (0.002) 0.239 0.015 (0.078) 0.239 0.004 (0.114) 0.239 0.020 (0.195) 0.236 0.949 (-0.047) 0.236 0.004 (0.225) 0.239 0.033 (0.155) 0.239 0.978 (0.127) 0.236 0.398 (0.029) 0.236 Phytol % Phytol (ug) 0.252 (0.002) <0.001 0.059 (0.018) <0.001 0.306 (0.001) <0.001 0.783 (-0.015) <0.001 0.859 (-0.046) <0.001 0.482 (-0.023) <0.001 0.507 (0.015) <0.001 0.603 (-0.008) <0.001 0.691 (-0-070) <0.001 0.387 (-0.013) <0.001 -1 2133 117 2134 Using environmental parameters as predictors of fatty acid biomarkers 2135 The WAP had the most models that fit environmental factors with a significant relationship to 2136 the biomarkers in the krill found in that region (Table 4.3). The SG region had the next 2137 highest number of models that fit the biomarker/environment interaction relationship (Table 2138 4.5). Both the overall South Atlantic area and the SOI had models that had less significant 2139 effects and correlations between the krill dietary biomarkers and the environment in those 2140 regions (Tables 4.2 and 4.4). Based on these simple models, the best areas (due to their 2141 consistency and predictability in environmental factors) for use to examine krill dietary 2142 biomarkers are the WAP and SG, and the best environmental predictors for krill diet are SST, 2143 Chl a (8D 3x3) concentrations and the interaction of SST and Chl a (CCAMLR) 2144 concentrations (Tables 4.2-4.5). 2145 The inter-relationships between environmental factors showed that TLDM increased as SST 2146 decreased in summer and autumn, but TLDM decreased as did SST in winter and spring. This 2147 was independent of Chl a levels, which were highest in the SOI in 2014, but the highest 2148 TLDM levels were in the WAP in 2016 (Figure 4.6). The percentage of EPA increased after a 2149 decrease in Chl a levels (grazing effect), but generally followed the same yearly trends as Chl 2150 a. EPA was high then decreased in 2014, plateaued in 2015 and again started high and 2151 decreased throughout the summer and autumn of 2016 (Figure 4.6). 2152 In terms of the fit of models, no models had an adjusted r2 value greater than 0.5 for the 2153 pooled data of the entire South Atlantic region (Table 4.2). The CCAMLR region data had 2154 multiple models with good fit with many models exceeding an adjusted r2 of 0.5. DHA 2155 percentage and 16:1n-7c mass fitted well with the interaction between SST and Chl a (8D 2156 3x3) (r2 = 0.611 and 0.551, respectively) in SG (Table 4.5). Whilst no fatty acids had an r2 2157 above 0.5 at the SOI (Table 4.4). DHA percentage was close to fitting with Chl a (8D 3x3) 2158 with an r2 value of 0.498 at the WAP (Table 4.3). SST and Chl a (CCAMLR) fitted to PL and 118 2159 TAG percentages in the WAP (r2: 0.564 and 0.559, respectively). EPA percentage fitted well 2160 with an adjusted r2 of 0.559 for its WAP model of SST and Chl a (CCAMLR). 2161 Better fits were found, however, by using the SST and Chl a (8D 3x3) interaction in the 2162 model at the WAP (Table 4.3). TLDM, PL and TAG percentages fitted this interaction with 2163 adjusted r2 values of 0.610, 0.655 and 0.568, respectively. EPA percentage (r2: 0.689) and 2164 EPA, DHA and 16:1n7c masses (r2: 0.615, 0.892, 0.621) all fitted the SST and Chl a (8D 2165 3x3) interaction at the WAP (Table 4.3). Additionally, SDA mass correlated with SST and 2166 Chl a (8D 3x3) at the WAP (r2: 0.560; Table 4.3). 2167 The best model fit for all of those tested was DHA mass at the WAP using the SST and Chl a 2168 (8D 3x3) interaction with an adjusted r2 value of 0.892. This can be seen in Figure 4.7 which 2169 compares the slopes and fit of models from the different CCAMLR sub-areas to DHA mass 2170 against SST and Chl a (CCAMLR) and the 95% confidence interval around the model. 2171 χ2 values are shown for all models whether the pooled South Atlantic models (Table 4.2) or 2172 for the CCAMLR specific sub-areas (Tables 4.3-4.5). 119 2173 2174 2175 2176 Figure 4.6: Multi Y axis plot of sea surface temperature (°C; black), total lipid (mg g-1) dry mass (blue), chlorophyll a levels (mg m-2; green) and eicosapentaenoic acid (20:5n-3) percentage (%; yellow) for dates of krill (Euphausia superba) sample collection. Lines drawn for illustrative purposes to show general trends. 120 2177 2178 Figure 4.7: Slopes of the models of best fit (red) and the 95% confidence interval for that 2179 model (blue) for docosahexaenoic acid (DHA; 22:6n-3) percentage (%) in Euphausia 2180 superba sampled in the different Commission for the Conservation of Antarctic Marine 2181 Living Resources (CCAMLR) sub-areas (West Antarctic Peninsula (WAP), South Orkney 2182 Islands (SOI) and South Georgia (SG)) against sea surface temperature (°C, SST) and 2183 chlorophyll a (mg m2, Chl a). 121 2184 DISCUSSION 2185 Krill lipid content and composition, specifically their fatty acid dietary biomarkers, correlate 2186 with changes in broad scale environmental data (SST and Chl a levels) derived from 2187 satellites. 2188 Krill eat, metabolise, and store lipids and fatty acids derived from their prey throughout the 2189 summer and early autumn when waters are warmer (higher SST) with more available food 2190 (diatoms and flagellates; higher Chl a) (Garibotti et al., 2005, Ericson et al., 2018a, Hellessey 2191 et al., 2018). In turn, they use these fatty acid and lipid stores during winter and early spring 2192 (lower SST and lower Chl a), resulting in a decrease in lipid, fatty acid and therefore n-3 LC- 2193 PUFA amounts (Ericson et al., 2018a, Hellessey et al., 2018). This change of fatty acid 2194 composition causes an increase in their other fatty acid composition percentages, although the 2195 fatty acid masses may not change. During summer and autumn, decreases in specific fatty 2196 acid percentages (EPA, DHA, 16:0 and phytanic acid (derived from phytol, a side chain of 2197 chlorophyll)), follow the decrease of SST. This is predominantly due to the overall increase 2198 in krill TLDM shifting the fatty acid composition in these seasons to lipids used more for 2199 reproduction and over winter survival (Hagen et al., 1996, Ju and Harvey, 2004, Yoshida, 2200 2009, Schmidt et al., 2014, Kawaguchi, 2016). Similarly, masses of EPA, DHA, 16:0, 16:4n- 2201 1, 16:1n-7c and phytanic acid increased through summer and autumn as krill laid down lipid 2202 stores for eggs, mostly n-3 LC-PUFA, and increased their TAG percentages prior to winter. 2203 This increase in fatty acid mass had an inverse relationship to SST in summer and autumn 2204 and may be due to a grazing effect and/or the lag effect of lipids being metabolised after the 2205 spring/summer algal bloom (Schmidt et al., 2012, Behrenfeld et al., 2017). 2206 The large increase in Chl a seen during the summer of 2014 in the SOI was strongly 2207 correlated to DHA mass and percentage for that season and year. This suggests that krill were 2208 predominantly eating flagellates in the summer of 2014 near the SOI and that this flagellate 122 2209 bloom was what was detected as extremely elevated green ocean colour and hence Chl a 2210 levels at the time. Being able to detect the bloom on the same day and at the same location of 2211 krill harvest was purely coincidental. Environmental conditions such as sun angle, cloud 2212 cover and sea ice did not interfere with ocean colour data capture for that location over that 2213 period of time, allowing for one of the best coincidental match ups of environmental data and 2214 krill fatty acids throughout the sampling period. 2215 The decrease of SST at SG during winter and spring, however, had an inversely proportional 2216 relationship with the increase of these same fatty acid percentages (EPA, DHA, 16:0 and 2217 phytanic acid). The fatty acid masses decreased in proportion to the decrease in SST at SG. It 2218 is thought that this may be related to krill using their lipid stores over winter and spring, a 2219 time when SST is lower and Chl a in the open ocean is lower (Schaafsma et al., 2017, 2220 Kohlbach et al., 2018), and most algae is bound in sea ice (Schmidt et al., 2014, Kohlbach et 2221 al., 2017, Meyer et al., 2017, Schaafsma et al., 2017). However, sea ice is not as prevalent at 2222 SG during winter as it is at the WAP and SOI, so herbivorous fatty acid sources would be 2223 lacking. The 18:1n-9c/18:1n-7c ratio has been previously shown to move from a more 2224 herbivorous diet in the summer/autumn to a more omnivorous diet in winter and spring 2225 (Mayzaud et al., 1998, Schaafsma et al., 2017, Ericson et al., 2018a). This dietary shift would 2226 also be seen as a decrease of fatty acid masses from herbivorous sources (e.g. diatoms and 2227 flagellates), however, n-3 LC-PUFA are preferentially conserved in krill as they serve as a 2228 major fatty acid functional group for krill health and growth (Ju and Harvey, 2004, Alonzo et 2229 al., 2005, O’Brien et al., 2011, Virtue et al., 2016). Therefore, decreases in 16:0, 16:4n-1, 2230 16:1n-7c and phytanic acid masses would be much larger than decreases in EPA and DHA 2231 masses, causing their total fatty acid composition percentages to increase proportionally and 2232 inversely to SST at this time of year. 123 2233 Chl a levels, indicated from extremely green ocean colour data, could be derived from 2234 diatom blooms (Moore and Abbott, 2002, Johnson et al., 2013, Zeng et al., 2016). These 2235 blooms may also be from flagellates which are seen as high green ocean colour values too 2236 (Moore and Abbott et al., 2002, Zeng et al., 2016), such as those seen at the SOI in the 2237 summer of 2014. Chl a (8D 3x3) levels correlated positively with EPA percentage, and Chl a 2238 (overall) levels positively related to phytanic acid percentage and mass. However, no 2239 flagellate lipid biomarkers in krill were significantly related to Chl a data overall, so it is 2240 more likely to come from diatom sources, which would then be seen in diatom markers such 2241 as EPA, phytanic acid and a high 16:1/16:0 ratio. Diatoms are likely the major source of 2242 phytol-derived phytanic acid in this case, as EPA is a dominant FA in diatoms and high 2243 16:1/16:0 and EPA:DHA ratios were observed, reflecting greater diatom abundance (Morris 2244 et al., 1984, Mayzaud et al., 1998, Schaafsma et al., 2017, Ericson et al., 2018a). Other 2245 diatom markers weren’t as high, possibly due to the differing rates of fatty acid metabolism 2246 (Huntley et al., 1994, Mayzaud et al., 2000, Hagen et al., 2001). Biomarkers with faster 2247 uptake rates, seen in larger quantities both in the krill’s diet and in lipid storage, would be 2248 faster to track and could fluctuate more closely to what is seen in the local environment (e.g. 2249 small scale blooms and an increase in Chl a). EPA and phytanic acid are both readily 2250 absorbed and metabolised by krill and can therefore track Chl a levels in the environment 2251 more closely immediately after a bloom event. EPA can be readily absorbed and stored in 2252 the PL of krill. However, some EPA is present (at low levels) within the TAG of krill as well 2253 and this may be metabolised even faster than the EPA stored in PL, as it does not need to be 2254 converted to PL from the primary dietary source (Ericson et al., 2019b). 2255 Fatty acid biomarker percentages associate better with the Chl a (overall) data, however, fatty 2256 acid masses are more highly associated with the more specific and localised Chl a (8D 3x3) 2257 data. Krill maintain their percentages of fatty acids between years, seasons and locations 124 2258 (Ericson et al., 2018a), so large scale Chl a data might not show smaller fluctuations. Krill 2259 fatty acid masses can change dramatically (up to a 10 fold increase between summer and 2260 early winter, Ericson et al., 2018a), and hence Chl a (8D 3x3) data fluctuations are more 2261 apparent in localised areas, but not in the overall Chl a data. The interaction between SST and 2262 Chl a (overall) showed a significant relationship to EPA, SDA and 16:0 percentages, while 2263 EPA, DHA, and 16:1n-7c masses were more significantly related to the SST and Chl a (8D 2264 3x3) interaction. Therefore, using different scales of Chl a in the SST and Chl a interaction 2265 within the models can provide a better prediction for either krill fatty acid percentages or 2266 masses, depending on the scale of Chl a pixel used. Reasons for inconsistencies may be due 2267 to differences between regions as these vary with the local environment in that area. These 2268 environmental differences will influence the primary production and hence diet of krill in 2269 these regions. Therefore, the biomarkers will vary between regions naturally, but may still 2270 correlate with the environmental data from that region (e.g. SST and Chl a from WAP will 2271 correlate with biomarkers from krill within the WAP but not from SOI). Many of the major 2272 essential krill fatty acid biomarkers were correlated to SST, Chl a and their interaction terms 2273 at varying scales. 2274 Tukey tests revealed the interaction with TLDM was mainly driven by SST, and not Chl a, 2275 and that SST had a close relationship to TLDM in all locations. Because this relationship 2276 holds without the Chl a interaction term, it can be assumed that SST drives TLDM levels 2277 more than Chl a levels do. TLDM relates well to SST and Chl a (8D 3x3) interactions as it is 2278 scaled to the krill’s weight, which may be affected by both temperature (Atkinson et al., 2279 2006) and its stomach and digestive gland weighing more from being full from chlorophyll 2280 rich items (Morris et al., 1984, Virtue et al., 1993a, Alonso et al., 2005). As TLDM naturally 2281 increases throughout summer and autumn and decreases throughout winter and spring 2282 (Hellessey et al., 2018), this would also coincidentally inversely follow the decrease of SST 125 2283 in summer and autumn, and decrease proportionally to SST in winter and spring. Therefore, 2284 the relationship between SST and TLDM may be coincidental due to the seasonal shift in 2285 how krill use their lipids aligning with seasonal shifts in SST. At the regional scale, TLDM 2286 was strongly related to Chl a (CCAMLR) levels in the SOI and WAP, although not in SG. 2287 This may be due to krill having a more herbivorous diet during summer and autumn when 2288 krill are harvested from the SOI and WAP and a more omnivorous diet during winter and 2289 spring whilst they are harvested at SG. 2290 The interaction between SST and Chl a (CCAMLR) was significant for SG, however, and 2291 also the WAP, but not for the SOI. This may be due to larger SST shifts at the more extreme 2292 ends of the latitudinal scale at WAP and SG (Rayner, 2003, Murphy et al., 2007). These 2293 locations would have the greatest extremes in environment, particularly for SST (Morris et 2294 al., 1984, Rayner, 2003), and as such any variations may explain why these locations show a 2295 strong relationship between TLDM and the environment. Similarly, at SG, Chl a (CCAMLR) 2296 was consistently low for all of winter and spring (when able to be recorded), whereas TLDM 2297 decreased dramatically during this period, so this may be giving false model fits for this area 2298 at this time of year. If readings of Chl a levels were possible throughout the winter season, a 2299 more closely related trend might be seen. 2300 As TAG is a storage lipid in krill (Hagen et al., 1996, Hellessey et al., 2018), it decreases 2301 from late summer though to the following late spring, when krill build stores for both 2302 reproduction (Varpe et al., 2007, Kawaguchi, 2016) and survival over winter (Hagen et al., 2303 1996, Ju and Harvey, 2004, Schmidt et al., 2014, Kohlbach et al., 2018). This peak in TAG 2304 also follows the peak in summer algal blooms (and thus peak Chl a) and is seen as a lag 2305 effect. This lag is due to both the time it takes for krill to metabolise TAG and the rate that 2306 krill convert fatty acids in algal TAG to their own PL stores. The PL percentages in krill 2307 tracked with Chl a (overall and 8D 3x3). PL has also been reported as a storage lipid in krill 126 2308 (Hagen et al., 1996), and is known to be vital for the storage of essential n-3 LC-PUFA for 2309 reproduction (Yoshida, 2009, Ross and Quetin, 2000, Varpe et al., 2007, Schmidt et al., 2012, 2310 and krill health (Mayzaud et al., 1997, Yoshida et al., 2011). Interestingly, PL levels were 2311 not related to changes in SST, but tracked well with changes in Chl a, possibly due to the way 2312 that polar phytoplankton blooms can occur via a boom-bust cycle (Behrenfeld et al., 2017). 2313 Many types of algae in polar latitudes are high in PL (Nichols et al., 1986, Nichols et al., 2314 1988, Skerratt et al., 1995, Kohlbach et al., 2015), and are also very green in colour, which 2315 could be one reason why these are so closely related to Chl a levels. PL can be incorporated 2316 very quickly into krill tissue as krill predominantly store their lipids as PL (Hagen et al., 2317 1996). Therefore, there would be little lag between ingestion, metabolism and incorporation. 2318 Our data shows there is little to no lag seen between Chl a data and PL percentages. 2319 The EPA:DHA ratio in krill was significantly related to Chl a (8D 3x3). A higher EPA:DHA 2320 ratio suggests that krill are consuming more diatoms than flagellates in their diet (Ericson et 2321 al., 2018a and 2018b). This relationship could be due to diatom and flagellate blooms 2322 affecting the ocean colour readings (higher green values) from satellite more so than other 2323 factors (Moore and Abbott, 2002, Johnson et al., 2013, Kahru et al., 2017). The green colour 2324 can be detected relatively easily in the ocean (IOCCG, 2015) and ocean colour is changing 2325 far faster with the climate than was predicted by models (Dutkiewicz et al., 2019). These 2326 changes in plankton community assemblages could also be changing faster than expected 2327 (Deppeler and Davidson, 2017, Hancock et al., 2018), which would be reflected in a 2328 changing EPA:DHA ratio. Krill potentially prefer a more diatom-based diet when blooms 2329 occur, even if flagellates are available in the water column (Behrenfeld et al., 2017, Deppeler 2330 and Davidson, 2017, Kohlbach et al., 2018). Such a dietary preference might skew the 2331 EPA:DHA ratio of krill, and could be related to the amount of Chl a being detected via 2332 remote sensing (Behrenfeld et al., 2017), whether at the 8D 3x3 or overall scale. The 127 2333 16:1/16:0 ratio shows differences in plankton types being consumed by krill (Mayzaud et al., 2334 1998, Ericson et al., 2018a), and was not associated with any Chl a or SST data over the 2335 whole South Atlantic but did show differences at smaller CCAMLR region scales. This may 2336 be due to krill diet shifting with seasons at the same time as the krill fishery also shifts it’s 2337 fishing location at the end of autumn (from SOI or WAP) to the start of winter (SG). 2338 The krill fishery operates at SG during winter and spring when algal populations are naturally 2339 lower in the water column (Schaafsmas et al., 2017, Kohlbach et al., 2018), and most algae is 2340 bound in sea ice (Schmidt et al., 2014, Kohlbach et al., 2017, Meyer et al., 2017, Schaafsma 2341 et al., 2017). Remote sensing during winter in polar regions is particularly hard for multiple 2342 reasons e.g. cloud cover, sea ice, and sun angle (IOCCG, 2015). These difficulties create gaps 2343 in the Chl a data available at SG, which may be falsely lowering these levels. Ground 2344 truthing the Chl a concentrations in SG throughout the winter season is one way of 2345 confirming this remote sampling data bias. Large-scale and long-term studies such as the 2346 Palmer Long-Term Ecosystem Research (LTER) program (Holm-Hansen et al., 1994, 2347 Helbling et al., 1995, Smith et al., 1995, Moline et al., 1997) and the U.S. Antarctic Marine 2348 Living Resources (AMLR) program (Helbling et al., 1995, Phleger et al., 2002) can ground 2349 truth their Chl a recordings by being present year round in a location to take water samples 2350 for Chl a analysis. Future technological advancements may also assist with such ground- 2351 truthing, including the use of deployed moorings that take water samples and record algal 2352 fluorescence year-round. Improving satellite algorithms for ocean colour data to be converted 2353 into Chl a concentrations in polar regions would significantly increase the number of data 2354 points available throughout winter. In turn, such enhanced data could then be better related 2355 to other factors, such as krill diet, with the remote sensed Chl a data being closer to the true 2356 Chl a levels present at that time of year. 128 2357 Additionally, due to krill swarming the spatial and temporal scales used may not be the 2358 tightest to fit krill dynamics as more than a single krill aggregation may be present in a 3 km 2359 x 3 km grid, and krill diet may vary greatly over an 8-day period. However, total lipid content 2360 varies slowly, particularly in the whole animal, as does their fatty acid profile. So, large 2361 dietary differences in krill aren’t expected to be seen at this level over the period of 8 days or 2362 at a scale of 3 km x 3 km. If, however, this analysis was using the lipids and fatty acids of 2363 krill stomachs or digestive glands, then this may have a more significant impact as these 2364 would change greatly over the period of 8 days and would show variation within a 3 km x 3 2365 km grid. 2366 This study used satellite-derived ocean colour and SST data in conjunction with fatty acid 2367 content and composition in krill diet at different times of year and in different locations. 2368 Cross-disciplinary work such as in this study is promising as it enables remote sensing and 2369 satellite oceanography specialists to better link with biological, physiological and ecological 2370 specialists. This collaboration may enable issues such as winter sampling of Chl a through 2371 satellites to be more well understood and solutions to issues such as sea ice and cloud cover 2372 to be resolved at a scale that is meaningful to the biology associated to that Chl a reading, 2373 whether primary producers or krill. 2374 The relationship between SST, Chl a and krill lipid biochemistry presented here could be 2375 expanded to examine similar relationships in krill diet and krill lipid content and composition 2376 in other regions around Antarctica. The approach could also be used for other marine-based 2377 species both in the Antarctic and other polar areas where sampling is restricted. 2378 ACKNOWLEDGEMENTS 2379 This research was funded by an Australian Research Council Linkage Grant LP140100412 2380 between the Australian Antarctic Division, Commonwealth Scientific and Industrial Research 129 2381 Organisation, Institute for Marine and Antarctic Studies (University of Tasmania), Aker 2382 BioMarine and Griffith University. Thank you to the Australian Bureau of Meteorology for 2383 allowing Dr Robert Johnson to assist and participate in this research; his help has been truly 2384 invaluable. We would also like to thank the Editors and two anonymous reviewers for their 2385 useful comments and suggestions to the manuscript. 2386 130 2387 Chapter 5: General Discussion 2388 During this project I studied the within and between seasonal variation in the lipid and fatty 2389 acid biochemistry of krill. Such data was used to investigate temporal variation in the diet of 2390 adult krill. Krill diet was inferred from the analysis of a unique set of fishery-derived samples. 2391 The samples were collected over large spatial and temporal scales that I then related to 2392 remotely-sensed environmental data. These approaches have provided new insights into how 2393 krill diet and lipid content vary throughout and between seasons, years, geographical areas and 2394 with the environment. 2395 The three main outcomes of this research were: 2396 • A detailed interannual and seasonal examination of the cycle in krill total lipid content 2397 and individual lipid class composition and content. These cycles were observed over a 2398 near-continuous three-year time period in the South Atlantic. This aspect of the study 2399 resulted from the availability of high-resolution fishery-derived samples; 2400 • Strong regional differences in krill diet were observed during the late-summer spawning 2401 period. These differences were detected from analysis of the fatty acids and sterols from 2402 the neutral lipid fraction in both sexes of krill. Results were further supported by similar 2403 analyses performed for their main digestive organs (digestive gland and stomach); 2404 • Krill lipid dynamics, including their fatty acid biomarkers, were related to the sea 2405 surface temperature (SST) and chlorophyll a (Chl a) concentration derived from 2406 remotely-sensed satellite data. 2407 Results from my research could be of use in developing and better parameterising ecosystem 2408 and energy budget models. Utilising the links between the environment (SST and Chl a) and 2409 krill fatty acids will directly improve ecosystem models relating krill physiology and diet to 2410 their environment. Similarly, understanding krill lipid dynamics at a seasonal, regional, and 131 2411 interannual scale will assist in improving krill energy budgets and life history models, as lipids 2412 are the key energy molecule in krill used to predict their condition and recruitment potential. 2413 These results will improve the understanding of seasonality of krill biology, and also assist in 2414 predicting krill recruitment and dietary changes under changing climate scenarios. 2415 Seasonality in krill lipid content and composition 2416 The seasonal and interannual trend in krill lipid content and composition was detailed 2417 by utilising high-resolution fishery-derived samples. Krill total lipid content showed a 2418 temporally sinusoidal pattern, previously surmised, but not quantified or detailed at such a high 2419 resolution as presented here. This clear seasonal and interannual trend in krill total lipid content 2420 was also seen in their component lipid classes. The fluctuations in krill lipid classes, 2421 specifically TAG, had long been assumed to be used for lipid storage, although this had not 2422 been unequivocally demonstrated due to the lack of high-resolution winter sampling of krill 2423 (Marschall, 1988, Huntley et al., 1994, Hagen et al., 1996, Atkinson et al., 2002, Ju and Harvey, 2424 2004, O’Brien et al., 2011, Schaafsma et al., 2017). By utilising fishery-derived samples, this 2425 issue of winter sampling for TAG levels has now been further resolved. The significance of 2426 TAG as a storage lipid, alongside PL as Hagen et al. (1996) has shown, will require further 2427 investigation as to how and why krill metabolise TAG through the seasons and if this varies at 2428 different geographical scales. 2429 The seasonal differences observed in the major lipid classes reflected their differing 2430 roles in key physiological and biochemical processes: growth, storage, and reproduction. The 2431 seasonal relative levels of TAG followed the temporal sinusoidal trend seen in total lipid 2432 content, with a peak in autumn and lows in spring. This corresponds to TAG being used as a 2433 storage lipid for energy reserves over the winter months, with little other use by krill unlike for 2434 PL, which can be both a structural and storage lipid class in krill. PL relative levels in krill also 2435 showed a seasonal response, but the seasonal peaks were out of phase with highs occurring in 132 2436 spring and lows in autumn. The high PL levels seen in these seasons could be due to its primary 2437 uses; for growth during the spring and summer when conditions are optimal, as well as for 2438 energy and lipid transfer during the reproductive season, particularly in gravid females. 2439 However, gravid female krill contain much higher levels of TAG than spent females because 2440 krill eggs are high in TAG (Marschall and Hirche, 1984, Tarling et al., 2009b). Gravid females 2441 have almost no available space for storing polar lipids due to the available carapace space being 2442 used for egg storage and production instead. The late-summer spawning period is of particular 2443 interest, as the ability of krill to reproduce is also linked to their ability to store enough lipids 2444 for growth and survival throughout winter (Mayzaud et al., 1998, Atkinson et al., 2006, 2445 Schmidt et al., 2012). Observed TAG levels in female krill would thus be dependent on the 2446 timing of krill spawning and the spring/summer algal bloom. 2447 This variability in the relationship between PL and TAG stores in krill may be due to 2448 differences in the timing and spatial extent of algal blooms that occur after the spring sea ice 2449 melt (Skerratt et al., 1995, Janout et al., 2016). Increases in both PL and TAG content and 2450 relative (%) composition after algal blooms have been observed for northern krill species; 2451 Meganyctiphanes norvegica (M. Sars), Thysanoessa inermis (Krøyer) and Thysanoessa raschii 2452 (M. Sars) (Falk-Petersen, 1981, Falk-Petersen et al., 1981). Pond et al. (1995) and Ju and 2453 Harvey (2004) have previously reported that lipid accumulation, particularly TAG, is tightly 2454 linked to seasonal factors such as the timing of reproduction in Antarctic krill species and the 2455 timing of the spring/summer algal bloom, which allows for krill to lay down lipid stores. To 2456 further explore this crucial spawning period, the differences in krill diet at a regional scale were 2457 investigated during this season. 2458 Regional differences in krill diet 2459 The TL, fatty acid and sterol content and composition of whole krill, as well as their 2460 digestive glands and stomachs, were examined for samples obtained from all three of the ocean 133 2461 basins surrounding Antarctica during the spawning period. By comparing fishery-based 2462 samples from the South Atlantic Ocean to scientific samples collected from the Southern Indian 2463 and Pacific Oceans (where fishery samples are unavailable), regional differences in krill diet 2464 could be seen for the first time. 2465 Early studies reported large differences in total lipid (TL) content between krill sexes 2466 (Fricke et al., 1984, Mayzaud et al., 1998). The TL (mg g-1) content in Indian and Pacific Ocean 2467 sector female krill decreased from sub-adult females to gravid females, and decreased further 2468 to spawned females. This trend was not seen in females in the Atlantic Ocean sector, which all 2469 had high TL content (mg g-1). This difference may be due to levels of primary production in 2470 the Atlantic Ocean sector being higher and also more consistent year-round than in the Indian 2471 and Pacific Ocean sectors (El-Sayed and Weber, 1982, Vernet et al., 2008, Westwood et al., 2472 2010). There is a known abundance and consistency of primary production in the Atlantic 2473 Ocean sector (Bodungen et al., 1986, El-Sayed and Weber, 1982, Helbling et al., 1995, Korb 2474 et al., 2005, Vernet et al., 2008). The relationship between primary production and krill lipid 2475 dynamics is currently unknown, however, primary production is known to fluctuate temporally 2476 (e.g. the spring/summer bloom). The timing and type of algal blooms available throughout the 2477 spawning period will depend on local factors such as SST, light availability, nutrient levels and 2478 salinity. Regional differences in the timing of this bloom would therefore also affect krill diet 2479 and spawning in each locality. Few studies have looked at differences in the fatty acid or sterol 2480 content and composition between the sexes of krill (Mayzaud et al., 1998, Atkinson et al., 2006, 2481 Schmidt et al., 2012), but never at this large of a geographical scale. 2482 In my study, the analysis of the neutral lipid fraction of krill digestive glands and 2483 stomachs was undertaken for samples collected in all three oceans basins surrounding 2484 Antarctica for the first time. Very few previous studies have used neutral lipid-derived fatty 2485 acids in dietary analysis, as the procedure is more complex than total lipid-derived fatty acid 134 2486 analysis (Cabrol et al., 2019). The analysis of the neutral lipid-derived fatty acids has allowed 2487 an in-depth detection of both shorter- and longer-term dietary differences between sexes of 2488 krill at a regional scale. My work details the shorter-term diet of krill from neutral lipid fatty 2489 acid and sterol profiles in samples from the stomach (days) and digestive gland (days-weeks). 2490 Using only the fatty acids from the neutral lipid fractions ensures that any signals from the 2491 structural components of the cells of the digestive gland and stomach are not included into the 2492 dietary marker profiles (Virtue et al., 1993a, Yoshida et al., 2009, Cabrol et al., 2019). The use 2493 of neutral lipid-derived fatty acids therefore enabled a better understanding of krill diet, which 2494 was found to vary by region in all of the different tissue samples during the late-summer season. 2495 My results showed that during the reproductive season digestive gland and stomach samples 2496 from Indian Ocean sector krill had a more diatomaceous and carnivorous based diet than 2497 digestive gland and stomach samples from krill in the Pacific and Atlantic Ocean sectors. 2498 Similarly, differences seen in the diet of the various sex classes of krill was predominantly due 2499 to a regional effect. 2500 Within each region, krill of different sexes showed little variation in their fatty acid 2501 profiles, which may be expected as they would have the same dietary material available to 2502 them. Krill from the Indian Ocean sector were the exception, where sub-adult females and 2503 spent females had significantly different dietary signals. The cause of this dietary difference is 2504 unknown, but it is hypothesised that spent females may need to regain lipid stores quickly in 2505 preparation for the winter months post-spawn, unlike sub-adult females. Another theory is that 2506 sub-adult and spent females are feeding in different locales (e.g. continental shelf, deeper water, 2507 warmer waters) within the wider region of the Indian Ocean sector as Nicol et al. (2000b) and 2508 results for my own samples have shown. Feeding at different locations therefore may be 2509 impacting the dietary signatures for sub-adult and spent female krill. 135 2510 These analyses have allowed for a comparison of krill diet at different temporal scales 2511 (days, weeks and long-term) at a regional level. The differences found between these regions 2512 suggested that possible environmental factors could be influencing krill diet at these smaller 2513 regional scales. Understanding how the environment at these different regional locations 2514 impacts krill diet would assist in linking key environmental factors, such as SST and Chl a 2515 concentrations (a proxy for primary production), to krill lipid dynamics. 2516 The inclusion of environmental data into this regional data set, for example, may assist 2517 in addressing why males and gravid females at Indian Ocean sector site 6 had vastly different 2518 diets to krill at other sites within the Indian Ocean sector. However, such a study would need 2519 more extensive sampling to be undertaken in this region. The collection of krill from scientific 2520 expeditions in both the Pacific and Indian Ocean sectors for this work was opportunistic. The 2521 collection of krill at a larger scale in these regions would overcome some of the limitations of 2522 having smaller sample sizes; these include the limited comparisons, variables and experiments 2523 which are currently able to be done. The trends and generalisations that can be drawn from 2524 studies using larger sample sets, particularly when investigating factors such as differences in 2525 diet between krill sexes, would be greatly improved. Details of krill diet at a seasonal and 2526 interannual scale are also still required at a regional scale. 2527 Relating SST and Chl a to krill diet 2528 To examine these dietary differences further, I used remotely-sensed SST data and Chl 2529 a concentrations to investigate whether these environmental variables could be related to the 2530 long-term krill lipid data obtained from the fishery-derived samples collected in the South 2531 Atlantic Ocean. Understanding the dynamics of the seasonal cycles of krill lipid content and 2532 composition in relation to the seasonal fluxes in primary production is a well-known, but 2533 unquantified link in the Antarctic food web. Through the use of high-resolution krill fishery 136 2534 samples in conjunction with remotely-sensed satellite environmental data, particularly SST and 2535 Chl a, these patterns have been established and measured for the first time. 2536 Broad scale environmental data from satellites was collected on simultaneous dates and 2537 from the same geographic positions as the krill fishery-derived samples. This allowed for the 2538 exploration of the links between the environment at that location and the biochemistry of the 2539 krill harvested there. The results show that krill lipid biochemistry, specifically their fatty acid 2540 dietary biomarkers, track changes in broad scale environmental data such as SST and Chl a 2541 concentrations derived from satellites. 2542 Interactions between Chl a and SST correlated well with krill fatty acid masses and 2543 percentages, although it is noteworthy that TL (mg g-1) dry mass (DM) related well to SST, but 2544 not to Chl a. This observation may be due to large scale environmental changes occurring 2545 between seasons. Seasonal changes such as sea ice expansion and atmospheric temperature 2546 drops during autumn and winter drive these large scale community composition changes. 2547 Therefore, krill lipid dynamics (and hence TL DM) are affected more by these large scale 2548 environmental changes than the small changes in algal community composition that occur 2549 throughout these same seasons. This community composition change would be seen more in 2550 the fatty acid composition of the krill, although it may not be seen in the TL DM of the krill. 2551 This may be due to krill storing lipids throughout summer and autumn, and increasing their TL 2552 DM, but not their fatty acids. These increases in krill TL DM are influenced more by the 2553 seasonal temperature shift than being due to prey community composition and availability. 2554 This study was limited by the fact that some winter and early spring Chl a 2555 concentrations were not available because of a lack of ocean colour data. Hence, any 2556 relationship between SST, Chl a and krill lipids at this time cannot therefore be accurately 2557 predicted. However, the interaction between SST and Chl a, when examined at a CCAMLR 137 2558 region scale, did give an initial indication that this relationship with krill lipids might be seen 2559 year-round. The interaction between SST and Chl a (CCAMLR region) matched well with the 2560 TL DM of krill at the more extreme ends of the latitudinal scale (South Georgia and the West 2561 Antarctic Peninsula). These locations would have the greatest extremes in environmental 2562 conditions, particularly for SST (Morris et al., 1984, Rayner, 2003). Variations in SST and Chl 2563 a at these locales may assist in explaining why they show a stronger relationship between TL 2564 DM and their corresponding environment. 2565 Krill lipid classes also showed relationships that varied with the remotely-sensed data. 2566 TAG percentage was related to Chl a (8D 3x3) and increased with decreasing SST during 2567 summer, but not in autumn. As TAG is the main storage lipid in krill (Hagen et al., 1996, Hagen 2568 et al., 2001), it would be related to a decrease in SST as the year progresses towards a time 2569 when krill need more stores for both reproduction (Varpe et al., 2007, Kawaguchi, 2016) and 2570 survival over winter (Hagen et al., 1996, Ju and Harvey, 2004, Schmidt et al., 2014, Kohlbach 2571 et al., 2018). The TAG percentage in krill follows the same sinusoidal trend seen in TL DM 2572 throughout the seasons, and as TL DM is highly related to SST, it follows that TAG percentage 2573 would have the same relationship to SST as TL DM. 2574 Percentages of PL did not correlate well with SST, as the peaks and troughs occurring 2575 for PL levels in krill are at the opposite times of year to TAG. PL percentages in krill did, 2576 however, correlate well to Chl a concentrations at both the overall and 8D 3x3 scale. This may 2577 be due to krill utilising and/or absorbing PL faster from phytoplankton that are already naturally 2578 high in PL content. It could also be due to krill metabolising TAG from phytoplankton and 2579 converting it to PL, causing it to increase, after an algal bloom occurs. Both scenarios would 2580 have an increase in Chl a concentration in the water due to the occurrence of an algal bloom, 2581 and both would result in higher PL percentages being observed in the krill. Further 2582 investigation into the mechanisms (e.g. absorption, utilisation and metabolisim) of how PL 138 2583 levels increase in krill, such as through lipidomics, would help to clarify this relationship with 2584 Chl a. 2585 Whilst this methodology was novel, it could also be expanded to be used with the 2586 regional dietary samples of krill. This expansion would facilitate further insights to the impacts 2587 of regional environments on the condition and diet of krill around Antarctica. This would also 2588 allow for meaningful regional scale changes to be modelled in different climate scenarios. 2589 At a regional scale, ecosystems will be driven by their local environments more so than 2590 the wider environment. Having both environmental and dietary food web data links at a 2591 regional scale is vital for understanding changes occurring within the different regions. As an 2592 example, my research showed that Chl a, when examined at an 8-day (8D) 3 km x 3 km (3x3) 2593 pixel scale, correlated positively with EPA percentages (as % of total fatty acids) in krill from 2594 the Atlantic Ocean sector in summer and autumn. This relationship between the environment 2595 (e.g. Chl a) and krill fatty acids is uncertain with regards to krill samples from the Indian Ocean 2596 sector, which had a more carnivorous and diatomaceous based diet. Carnivory inputs would 2597 not be able to be detected through remote-sensing. Insights into the interaction between the 2598 environment and krill diet at a regional scale will assist in understanding how krill predate 2599 differently around Antarctica. 2600 This study is the first to link remotely-sensed environmental data to krill fatty acid data 2601 and the methodologies used here potentially can be further improved, strengthened, and 2602 expanded to other areas, and applied to different krill species or other marine organisms. There 2603 is a need for further research and experimentation into the linking of remotely-sensed Chl a 2604 concentrations at different scales to fatty acid content and composition. Examining different 2605 phytoplankton functional types through remote-sensing techniques would be another key area 2606 for future research. This can be performed by further analysing the ocean colour data and 139 2607 separating out the chlorophyll values into specific phytoplankton pigment groups, such as red 2608 algae, green algae and brown algae (e.g. coccolithophores, diatoms, dinoflagellates). Ideally, 2609 these remote-sensing analyses would be ground-truthed with in situ sampling of phytoplankton 2610 by high throughput methods such as automated shape recognition assisted flow cytometery. In 2611 particular, it would be valuable to relate the variation in the relative abundance of different 2612 phytoplankton functional groups throughout the year to the seasonal and interannual trends 2613 seen in krill lipid dynamics. 2614 Future research 2615 2616 The results of this study indicate five broad areas for future research that would improve our knowledge of krill diet and lipid dynamics. 2617 Firstly, the high-density sampling that was possible through the availability of krill 2618 fishery-derived samples has revealed new insights into the seasonal and temporal cycles in krill 2619 lipids. If such sample collection could be broadened to include all three Antarctic ocean basins, 2620 this would improve our understanding of krill lipid biochemistry as well as krill biology more 2621 generally. One potential way of expanding the number of samples would be to utilise the recent 2622 expansion of the krill fishery into the Indian Ocean sector of the Southern Ocean (CCAMLR, 2623 2017). Krill distribution and density are known to differ greatly between basins, and improved 2624 sampling might help to clarify the drivers of many krill biochemical and biological parameters. 2625 Although many multinational research voyages have attempted to cover vast areas (such as the 2626 CCAMLR 2000 and 2019 surveys, BROKE, BROKE-West, K-Axis, and the recent survey of 2627 Sub-Area 58.4.1 by the Japanese), these voyages rarely cover more than a single basin, month 2628 or season. These larger research voyages also normally target specific locations with little 2629 overlap between voyage locations, which in turn does not allow for repetition or interannual 2630 effects to be accounted for. 140 2631 The Chinese krill fishery opened an exploratory fishery in the Indian Ocean sector in 2632 2016 in accordance to the CCAMLR conservation management guidelines and policies 2633 (CCAMLR, 2017). The continued presence of a Chinese based krill fishery in the Indian Ocean 2634 sector would facilitate the sampling of krill in this region in different seasons and would have 2635 a much wider geographical spread than current voyages in this sector that are undertaken by 2636 national scientific voyages. If samples were to be collected through a similar methodology as 2637 those in the Atlantic Ocean sector, the comparison of samples between regions could also 2638 become more consistent. 2639 Currently there are no large-scale biological voyages planned for the Pacific Ocean 2640 sector, whether by multinational efforts of the scientific community or by industry. Few large- 2641 scale expeditions have concentrated on the biology of the Pacific Ocean sector (CCAMLR 2642 Area 88) (Bottino, 1974, Mackey et al., 2012, Leonori et al., 2017), although the physical 2643 oceanography of the Ross Sea has been studied numerous times (Jacobs et al., 1970, Orsi and 2644 Wiederwohl, 2009, Smith Jr et al., 2014). Since the Discovery voyages in the 1820’s, only the 2645 Antarctic Circumpolar Expedition (ACE) voyage in 2017 has collected biological samples at a 2646 large scale in this area, so this region in particular warrants further investigation. 2647 Another way of accessing krill samples from these areas would be to utilise ships of 2648 opportunity, such as other fishing vessels, tourist vessels, resupply vessels, and other national 2649 scientific voyages. This would increase the number of samples of krill within these regions 2650 with minimal effort. Adoption of an agreed set of collection and analysis protocols would 2651 ensure that comparable sets of samples were obtained, making additional sample contributions 2652 invaluable. 2653 Secondly, this study focussed only on the diet of adult krill as they are numerous within 2654 the fishery samples and found year-round at all spatiotemporal scales. Future studies should 141 2655 aim to examine the diet of larval and juvenile krill, which is known to be substantially different 2656 (Atkinson et al., 2002, Virtue et al., 2016, Schaafsma et al., 2017). Further investigations at 2657 varying spatial and temporal scales would assist in understanding dietary differences between 2658 larval and juvenile krill (and adults), and how their diet may impact development, reproduction 2659 and survival. 2660 The climate of the Southern Ocean is changing with waters warming, becoming fresher, 2661 and increasing in acidity, the consequences of which on adult krill are still mostly unknown 2662 (Kawaguchi et al., 2011, Flores et al., 2012, Bijma et al., 2013, Hill et al., 2013, Kawaguchi et 2663 al., 2013, Constable et al., 2014, Barnes and Tarling, 2017, Ericson et al., 2018b, Atkinson et 2664 al., 2019). The flow on effects of such changes on larval and juvenile krill diet quality, due to 2665 a decrease in sea ice extent and thickness with climate change, warrants further study. A 2666 reduction in sea ice will likely impact future krill recruitment and population levels because 2667 these life stages require sea ice for food, shelter and protection (Wiedenmann et al., 2009, 2668 Massom and Stammerjohn, 2010, Schmidt et al., 2014, Kohlbach et al., 2017, Schaafsma et al., 2669 2017, Kohlbach et al., 2018). Research on the impacts of ocean acidification on adult krill by 2670 Ericson et al. (2018b) has included the development and use of an aquarium-based 2671 methodology that would be ideal to look into larval and juvenile krill diet, health, development 2672 and mortality in relation to ocean acidification also. The link between increased sea ice extent 2673 and thickness with increased growth, health and recruitment of juvenile krill is already well 2674 understood (Wiedenmann et al., 2009, Schmidt et al., 2014, Schaafsma et al., 2017, Mori et al., 2675 2019). The use of remotely-sensed data for determining sea ice expansion and contraction rates, 2676 thickness and extent throughout the seasons could also be useful for enhancing the 2677 understanding of how juvenile krill diet and health will change at a larger scale in real time. 2678 If a robust relationship between remotely-sensed sea ice data and larval and juvenile 2679 krill lipid dynamics can be determined, then a model of sea ice expansion and contraction with 142 2680 krill recruitment levels could be developed. Such a model could operate using near real time 2681 data which would allow for annual predictions and detection of interannual and long-term 2682 trends of krill recruitment. As sea ice contracts southward and thins with climate change, 2683 having faster predictions of potential negative effects on the ecosystem by human influences 2684 (such as the fishery) will be required to stop additional stress being added to the system. A 2685 predictable relationship between remotely-sensed sea ice data and krill recruitment would 2686 allow CCAMLR to implement within season fishing regulations as envisaged in the various 2687 proposals for feedback management (Constable et al., 2000, Constable and Nicol, 2002, Hill 2688 and Cannon, 2013, Hill et al., 2016). Similarly, Chl a concentrations and SST could be used 2689 for providing adult krill condition and reproduction estimates leading into the spawning season 2690 allowing for tighter within season fishery management. 2691 In addition, investigations at varying spatial and temporal scales would assist in 2692 understanding the influence of diet on larval and juvenile krill development and maturity. If 2693 larval and juvenile krill diet shifts with climate change, due to the assemblages of microbes 2694 and primary producers changing (Deppeler and Davidson, 2017, Hancock et al., 2018), this 2695 may have direct impacts on larval and juvenile krill growing to maturity and their ability to 2696 reproduce. Knowing how larval and juvenile krill diet varies currently at different scales is of 2697 benefit to understanding and predicting krill recruitment and survival both now and into the 2698 future. Studies on larval and juvenile krill lipid dynamics are highly recommended, particularly 2699 if they are able to be coupled with large spatial, temporal or environmental data studies such 2700 as presented here, or via the use of aquarium-based studies looking at future ocean scenarios 2701 as in Ericson et al. (2018b). 2702 Thirdly, this data adds to and expands upon the work of the Southern Ocean Diet and 2703 Energetics Database (SO-Diet; Raymond et al., 2011). This database is collecting dietary 2704 analyses from all Southern Ocean species, both historical data and more recently analysed, 143 2705 and from a variety of different dietary Methodologies including stable isotopes, gut content 2706 analysis, lipid and fatty acid analysis and DNA analysis. 2707 The inclusion of a large spatial and temporal range of adult krill diets to this database, 2708 particularly in the form of lipid and fatty acid data, will inform future studies linking species, 2709 or looking at demographics within krill diet to be enhanced. By combining this data with 2710 other krill diet and energetic studies a more complete and robust energy budget for adult krill 2711 will also be possible. This data can also be used as a baseline, especially within the Atlantic 2712 Sector where the data is richest, for future climate studies in relation to adult krill diet or the 2713 impact of the fishery on krill diet. Looking for longer term trends in adult krill diet is nearly 2714 impossible without combining data from multiple studies, regions, expeditions and nations, 2715 therefore the addition of this large and diverse dataset will be extremely useful of larger, 2716 longer term trend studies also. 2717 Forth, my research has demonstrated the utility of using remote-sensing for addressing 2718 potential ecological problems linked to krill. A new generation of multinational controlled 2719 polar orbiting satellites has recently been launched. The first new satellite, Joint Polar Satellite 2720 System-1 (JPSS-1) was launched in November 2017. JPSS-1 includes instrumentation such as 2721 the Advanced Very High Resolution Radiometer (AVHRR) and the Advanced TIROS 2722 Operational Vertical Sounder (ATOVS). The joint AVHRR/ATOVS on the JPSS-1 will 2723 provide data in the visible, infrared and microwave ranges, allowing for a multitude of 2724 applications. The JPSS fleet will eventually include five satellites and, if they are calibrated 2725 and utilised correctly, such a system could be extremely useful for gaining biologically relevant 2726 environmental data, specifically from this hard to access area of the world. Additional ocean 2727 colour and biological data from the JPSS would improve future Chl a algorithms for the 2728 Southern Ocean, provide more winter and early spring ocean colour samples, and provide a 2729 platform for other future biological measurements to be undertaken within this polar region. 144 2730 Finally, there is a need for future biochemical studies into the lipidomics and pigment 2731 signatures of krill. Lipidomics is the large-scale study of pathways and networks of 2732 cellular lipids in biological systems (Wenk, 2005). Research into krill lipidomics would allow 2733 for more detailed answers on how lipids are converted and metabolised within the krill’s 2734 biological system. Most krill lipidomic studies are focused on the metabolic fate, including a 2735 number of biologically active intermediates, of n-3 LC-PUFA when ingested as krill oil in 2736 humans (Wenk, 2005, Backes et al., 2014, Méndez et al., 2017, Sung et al., 2019). Currently, 2737 studies on krill lipidomics within the krill themselves are rare (Chen, 2012). Knowledge of the 2738 molecular conversion of one lipid into another, through a chain of complex reactions, within 2739 the cells of krill could potentially be harnessed for commercial purposes. Lipidomic studies 2740 could ascertain where and why krill store their lipids based on external influences, such as their 2741 environment; or they could find the most efficient way of converting krill dietary inputs into 2742 commercially viable n-3 LC-PUFA. Lipidomics is an expanding area of study and the 2743 implications for application of this approach at a commercial scale are not yet fully understood, 2744 but there is great potential for its use into the future. 2745 Astaxanthin, a red pigment found in krill, already has its own unique market due to its 2746 antioxidant qualities and ability to colour the flesh of higher predators (particularly salmon). 2747 How and why astaxanthin is metabolised and utilised by krill is still largely unknown (Takaichi 2748 et al., 2003). Knowledge of how lipids and their related pigments are converted, utilised and 2749 metabolised in krill is an obvious next step for gaining further understanding of krill lipid 2750 dynamics. Knowledge of how pigmentation in krill is related to their diet and lipids would be 2751 of huge benefit to the krill fishery in their production of krill oil for nutraceutical products, 2752 which contain this natural antioxidant (Tou et al., 2007, Barros et al., 2014). Studies and 2753 research such as this would be able to provide valuable insights to the fishery on many potential 2754 krill based products such as (but not limited to) astaxanthin, chitin and krill enzymes. Studies 145 2755 into other commercially relevant biochemicals in krill would benefit from also having spatial, 2756 temporal and environmental data linked to the wider krill fishery and it’s sustainability. 2757 Conclusions 2758 By utilising the spatial and temporal aspects of krill lipid data, fishery decisions can be 2759 made based on where and when n-3 LC-PUFA or TL content is highest. Knowing when and 2760 where krill lipid content is optimal will facilitate maximising lipid yield whilst minimising krill 2761 harvest. This will assist the fishery as the industry, in theory, will not require as much krill to 2762 be caught if the lipid content within each krill is greater on average. Similarly, using the 2763 environmental links as shown in my research, the fishery will be better able to predict krill lipid 2764 and n-3 LC-PUFA content in years with lower/higher Chl a, or SST, or both. This will also 2765 assist in sustainable fishery management into the future as the environment becomes more 2766 unpredictable. 2767 The research undertaken in this thesis has increased the understanding of krill lipid 2768 dynamics, which has implications for enhancing knowledge on krill diet and the factors 2769 influencing krill diet at different spatial, temporal and environmental scales. The interannual 2770 and seasonal trends in krill lipid content and lipid class composition were further resolved with 2771 the availability of fishery-derived samples. Regional differences in krill diet were found by 2772 examination of the neutral lipid fraction fatty acids of krill digestive glands and stomachs. Krill 2773 diet varied with sex at a regional scale during the late-summer spawning period. Krill lipid 2774 dynamics, including the composition of key fatty acid biomarkers, were able to be related to 2775 satellite-derived Chl a concentrations and SST data. The availability of high resolution data 2776 on krill lipid dynamics under current environmental conditions will facilitate an improved 2777 understanding on krill growth, health, reproduction and predictions of recruitment. Such 2778 information will improve both ecosystem and energy budget models, which in turn may lead 146 2779 to better krill fishery management models allowing for maximum lipid yields whilst potentially 2780 minimising krill catch. 147 2781 Appendices 2782 APPENDIX 1 2783 2784 2785 2786 Supplementary Table 3.1: Composition (as % mean ± SD of total lipids) of lipid classes for each Euphausia superba sample type and Southern Ocean sector (Atlantic, Indian and Pacific). Mass for the stomach and digestive gland samples are for the whole sample, not on a per krill basis. (HC - hydrocarbons (including wax and sterol esters); TAG - triacylglycerols; FFA - free fatty acids; ST - sterols; DAG - diacylglycerols; PL phospholipids; Unknown). Sample Type Sector HC TAG FFA ST DAG PL Unknown Whole Krill Atlantic 0.5 ± 0.4 42.6 ± 5.3 3.2 ± 1.7 2.5 ± 3.6 5.8 ± 6.9 43.7 ± 8.2 1.6 ± 2.8 Indian 0.4 ± 0.7 37.6 ± 10.7 0.1 ± 0.1 2.3 ± 1.8 3.4 ± 10.5 55.0 ± 8.9 1.1 ± 4.4 Pacific 0.2 ± 0.1 49.1 ± 4.2 1.7 ± 3.5 1.8 ± 1.1 1.2 ± 1.1 45.9 ± 3.6 0.0 ± 0.0 Atlantic 1.0 ± 1.4 28.3 ± 5.1 2.2 ± 0.7 1.4 ± 1.3 0.4 ± 0.4 66.6 ± 4.9 0.0 ± 0.0 Indian 0.0 ± 0.0 33.5 ± 29.7 17.4 ± 15.5 1.3 ± 2.3 1.8 ± 3.5 46.0 ± 22.2 0.1 ± 0.4 Pacific 0.2 ± 0.3 9.1 ± 10.3 46.5 ± 27.3 0.8 ± 0.1 0.2 ± 0.3 43.0 ± 16.4 0.2 ± 0.3 Digestive Atlantic 1.2 ± 0.9 36.3 ± 6.6 6.8 ± 4.3 0.9 ± 1.4 2.3 ± 3.9 52.5 ± 5.9 0.0 ± 0.0 gland Indian 8.7 ± 15.0 57.9 ± 19.1 2.9 ± 5.6 0.8 ± 3.6 0.0 ± 0.0 29.4 ± 8.5 0.0 ± 0.0 Pacific 0.2 ± 0.3 33.7 ± 3.9 12.5 ± 7.6 0.7 ± 0.3 0.3 ± 0.2 52.5 ± 5.7 0.2± 0.3 Stomach 2787 148 2788 2789 2790 2791 Supplementary Table 3.2: Euphausia superba fatty acid groups (as % of the total fatty acids; mean ± SD) and selected major dietary fatty acid markers in different Southern Ocean sectors and tissue types (MUFA: monounsaturated fatty acids, PUFA: polyunsaturated fatty acids, SFA: saturated fatty acids, MSI: marine snow indicator [Ʃ C15, C17 and C19 isomers], Copepods: [Ʃ 20:1n-9c + 22:1n-9c], Carnivory ratio: [18:1n9c/18:1n-7c], Diatoms: [Ʃ 16:1n-7c +16:4n-1], CvD: Copepods vs Diatoms ratio [Ʃ 20:1n-9c + 22:1n-9c/ 16:4n-1]). 2792 Sample Type Sector Whole Krill Atlantic Copepods Carnivory Diatoms CvD Phytol 33.9 ± 2.3 33.9 ± 3.4 32.1 ± 1.6 0.7 ± 0.2 1.9 ± 0.5 1.9 ± 0.2 8.4 ± 0.9 2.4 ± 0.8 2.6 ± 1.6 Indian 40.0 ± 4.1 26.5 ± 6.6 33.8 ± 3.2 0.6 ± 0.3 1.5 ± 0.4 1.7 ± 0.2 8.9 ± 2.1 3.3 ± 1.6 2.8 ± 1.2 Pacific 31.8 ± 1.5 35.0 ± 2.3 33.2 ± 1.8 0.9 ± 0.1 1.5 ± 0.3 2.3 ± 0.3 6.3 ± 0.6 2.6 ± 0.8 1.9 ± 0.7 Atlantic 37.4 ± 4.4 30.3 ± 2.3 32.3 ± 3.4 0.4 ± 0.3 1.4 ± 0.9 1.9 ± 0.0 7.7 ± 0.9 2.5 ± 0.0 1.3 ± 1.5 Indian 36.9 ± 4.4 30.7 ± 3.8 32.4 ± 6.1 0.5 ± 0.3 1.4 ± 0.1 1.8 ± 0.2 8.7 ± 1.8 3.5 ± 0.8 0.0 ± 0.0 Pacific 31.6 ± 0.1 38.5 ± 0.3 29.9 ± 0.2 0.9 ± 0.1 1.8 ± 0.3 2.0 ± 0.1 5.6 ± 0.1 3.0 ± 0.1 0.7 ± 1.0 Digestive Atlantic 33.5 ± 1.5 33.5 ± 3.3 33.0 ± 2.1 0.7 ± 0.1 2.2 ± 0.3 1.8 ± 0.2 8.2 ± 0.8 2.9 ± 0.7 1.9 ± 0.5 Gland Indian 34.9 ± 2.4 30.1 ± 5.8 34.9 ± 3.7 0.6 ± 0.2 1.6 ± 0.4 1.6 ± 0.2 6.9 ± 2.1 3.8 ± 1.8 0.0 ± 0.0 Pacific 31.9 ± 1.3 35.8 ± 2.2 32.2 ± 1.6 0.9 ± 0.1 1.7 ± 0.2 2.0 ± 0.2 6.1 ± 0.7 2.9 ± 0.6 1.7 ± 0.5 Stomach MUFA PUFA SFA MSI 149 2793 2794 2795 2796 Supplementary Table 3.3: Eigenvalues, variation (%) cumulative variation (%) and loadings for principal component analysis of the fatty acid percentage composition data of Euphausia superba. Whole krill fatty acid profiles were taken from the total lipid, and stomach and digestive gland fatty acid profiles were taken from the neutral lipid fractions. Largest loadings for positive and negative components of each principal component are highlighted in bold. Sample Type Principal component Whole Krill PC1 PC2 PC3 Eigenvalues 0.344 % variation Cumulative % variation Loadings 65.9 65.9 14:0 0.173 Stomach PC1 PC2 PC3 6.4 e-2 4.4 e-2 12.4 78.3 Digestive Gland PC1 PC2 PC3 2 0.861 0.193 1.06 0.378 9.5e-2 62.4 62.4 26.9 89.3 6.0 95.3 64.0 64.0 22.7 86.7 5.7 92.4 14:0 -0.760 0.542 0.119 0.520 -0.726 0.139 8.4 86.8 Loadings -0.530 0.521 16:4n-1 -0.176 -0.198 -0.374 16:4n-1 -0.144 0.035 -0.012 0.033 -0.206 -0.122 16:1n-7c 0.329 -0.210 -0.296 16:1n-7c 0.032 0.140 -0.273 0.294 0.027 -0.289 16:0 0.105 -0.155 -0.073 16:0 0.040 0.107 0.081 0.101 -0.064 -0.072 18:3n-6 -0.109 0.008 0.175 18:3n-3 0.095 -0.151 0.072 -0.016 0.143 0.307 18:4n-3 -0.724 -0.386 0.010 18:2n-6 -0.114 -0.218 0.270 -0.117 -0.197 0.435 18:1n-9c 0.101 -0.103 -0.025 18:4n-3 -0.257 -0.112 -0.028 -0.426 -0.224 0.535 18:1n-7c 0.139 0.286 -0.128 18:1n-9c 0.133 -0.009 -0.265 -0.030 0.206 0.162 -0.021 -0.166 0.089 18:1n-7c 0.144 0.001 -0.206 -0.042 0.180 0.016 20:5n-3 -0.178 0.207 -0.223 18:0 0.261 0.128 0.715 -0.033 0.141 0.041 20:4n-3 0.106 0.312 0.083 20:5n-3 -0.269 -0.329 -0.270 -0.450 -0.282 -0.482 20:1n-9c 0.007 -0.131 -0.386 20:1n-9c 0.018 -0.080 21:5n-3 -0.151 -0.013 -0.244 21:5n-3 -0.054 -0.086 0.052 -0.081 -0.094 -0.076 22:6n-3 22:1n-9c -0.433 0.374 0.178 0.021 -0.211 -0.378 22:6n-3 22:1n-9c -0.355 -0.666 -0.056 -0.126 0.179 0.242 -0.436 -0.348 -0.126 -0.076 -0.154 -0.107 18:0 2797 150 -0.021 -0.187 -0.133 0.080 2798 APPENDIX 2 2799 Additional Outcomes from this Research 2800 Co-Authored papers: 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 • ERICSON, J. A., HELLESSEY, N., NICHOLS, P. D., KAWAGUCHI, S., NICOL, S., HOEM, N. & VIRTUE, P. 2018a. Seasonal and Interannual Variation in the Fatty Acid Content and Composition of Euphausia superba samples derived from the Scotia Sea fishery Journal of Crustacean Biology, 38, 662-672. • ERICSON, J. A., HELLESSEY, N., KAWAGUCHI, S., NICOL, S., NICHOLS, P. D., HOEM, N. & VIRTUE, P. 2018b. Adult Antarctic krill proves resilient in a simulated high CO 2 ocean. Communications Biology, 1, 190. • ERICSON, J. A., HELLESSEY, N., KAWAGUCHI, S., NICHOLS, P. D., NICOL, S., HOEM, N. & VIRTUE, P. 2019a. Near-future ocean acidification does not alter the lipid content and fatty acid composition of adult Antarctic krill. Scientific reports, 9, 1-10. • ERICSON, J. A., HELLESSEY, N., NICHOLS, P. D., NICOL, S., KAWAGUCHI, S., HOEM, N. & VIRTUE, P. 2019b. New insights into the seasonal diet of Antarctic krill using triacylglycerol and phospholipid fatty acids, and sterol composition. Polar Biology, 1-12. Presentations at: • Antarctic Climate and Ecosystems Co-operative Research Centre (ACE CRC) Symposium 2016 , Hobart, Australia (poster) • University of Tasmania (UTAS) Graduate Research Conference 2016, Hobart, Australia (poster) • Science of Omega-3: Balancing the Scales, Omega-3 Centre Conference 2016, Sydney, Australia (poster) • Homeward Bound Symposium at Sea 2016, Ushuaia, Argentina (oral, chosen as 1 of 78 voyagers out of over 3,000 applications globally) • Association of Polar Early Career Scientists (APECS) Online Conference "Outside the Box" 2017, global webinar (oral) • 3rd International Krill Symposium 2017, St Andrew’s, Scotland (oral) • Scientific Committee for Antarctic Research (SCAR) Biology 2017, Leuven, Belgium (oral and poster) • Australasian section of the American Oil Chemists Society (AAOCS) Conference 2017, Tanunda, Australia (oral) 151 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 • University of Tasmania (UTAS) Graduate Research Conference 2017 (oral – 3 Minute Thesis Finalist, top 10 from over 500 applicants) • Marine Ecosystem Assessment for the Southern Ocean (MEASO) conference 2018, Hobart Australia (oral and poster – local organising committee) • Scientific Committee for Antarctic Research (SCAR) and the International Arctic Science Committee (IASC) POLAR2018 conference, 2018, Davos, Switzerland Visiting Researcher stay at: • British Antarctic Survey, June 2017, with Dr Geraint Tarling Science Communication and Outreach through UTAS, IMAS and CSIRO: 2857 • TAstroFest (2016) 2858 • Antarctic Festival (2016-2018) 2859 • ABC radio interviews (2016-2019) 2860 • Mercury, Examiner and the Advocate newspaper articles (2016-2019) 2861 • Online guest blog posts and interviews (2016-2019) 2862 • Young Tassie Scientists (2016-2019) 2863 • STEM Professionals in Schools (2017, with Margate Primary and St Helen’s Regional School) 2864 2865 • TV interviews (Four Corners - 2017) 2866 • Science Investigation Awards Judge (2017-2018) 2867 • Festival of Bright Ideas – IMAS (2017), WhySci and Young Tassie Scientists (2018) 2868 • Tasmanian Youth Science Forum (2017-2019) 2869 • Conoco Phillips Science Experience (2017-2019) 2870 • DataTas Keynote Presentation (2018) 2871 • RoboCup Junior Judge (2018) 2872 • “Now that’s what I call Science” radio show co-hosting at Edge Radio (May 2019) 152 2873 References 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 ALONZO, S. H. & MANGEL, M. 2001. Survival strategies and growth of krill: avoiding predators in space and time. Marine Ecology Progress Series, 209, 203-217. ALONZO, F., VIRTUE, P., NICOL, S. & NICHOLS, P. D. 2005. Lipids as trophic markers in Antarctic krill. II. Lipid composition of the body and digestive gland of Euphausia superba in controlled conditions. Marine Ecology Progress Series, 296, 65-79. APHALO, P. J. 2016. Learn R...as you learnt your native tongue. LeanHub, Helsinki, Finland. ARDELAN, M., HOLM-HANSEN, O., HEWES, C., REISS, C. S., SILVA, N., DULAIOVA, H., STEINNES, E. & SAKSHAUG, E. 2010. Natural iron enrichment around the Antarctic Peninsula in the Southern Ocean. Biogeosciences, 7, 11-25. ATKINSON, A., MEYER, B., STÜBING, D., HAGEN, W., SCHMIDT, K. & BATHMANN, U. 2002. Feeding and energy budgets of Antarctic krill Euphausia superba at the onset of winter—II. Juveniles and adults. Limnology and Oceanography, 47, 953-966. ATKINSON, A., SIEGEL, V., PAKHOMOV, E. & ROTHERY, P. 2004. Long-term decline in krill stock and increase in salps within the Southern Ocean. Nature, 432, 100-103. ATKINSON, A., SHREEVE, R. S., HIRST, A. G., ROTHERY, P., TARLING, G. A., POND, D. W., KORB, R. E., MURPHY, E. J. & WATKINS, J. L. 2006. Natural growth rates in Antarctic krill (Euphausia superba): II. Predictive models based on food, temperature, body length, sex, and maturity stage. Limnology and Oceanography, 51, 973-987. ATKINSON, A., SIEGEL, V., PAKHOMOV, E. A., ROTHERY, P., LOEB, V., ROSS, R. M., QUETIN, L. B., SCHMIDT, K., FRETWELL, P., MURPHY, E. J., TARLING, G. A. & FLEMING, A. H. 2008. Oceanic circumpolar habitats of Antarctic krill. Marine Ecology Progress Series, 362, 1-23. ATKINSON, A., SIEGEL, V., PAKHOMOV, E., JESSOPP, M. & LOEB, V. 2009. A reappraisal of the total biomass and annual production of Antarctic krill. Deep Sea Research Part I: Oceanographic Research Papers, 56, 727-740. ATKINSON, A., HILL, S. L., PAKHOMOV, E. A., SIEGEL, V., ANADON, R., CHIBA, S., DALY, K. L., DOWNIE, R., FIELDING, S., FRETWELL, P., GERRISH, L., HOSIE, G. W., JESSOPP, M. J., KAWAGUCHI, S., KRAFFT, B. A., LOEB, V., NISHIKAWA, J., PEAT, H. J., REISS, C. S., ROSS, R. M., QUETIN, L. B., SCHMIDT, K., STEINBERG, D. K., SUBRAMANIAM, R. C., TARLING, G. A. & WARD, P. 2017. KRILLBASE: a circumpolar database of Antarctic krill and salp numerical densities, 1926–2016. Earth System Science Data, 9, 193-210. ATKINSON, A., HILL, S. L., PAKHOMOV, E. A., SIEGEL, V., REISS, C. S., LOEB, V. J., STEINBERG, D. K., SCHMIDT, K., TARLING, G. A., GERRISH, L. & SAILLEY, S. F. 2019. Krill (Euphausia superba) distribution contracts southward during rapid regional warming. Nature Climate Change. 9, 142. BACKES, J., RUISINGER, J., HARRIS, K., GIBSON, C., HARRIS, W. & MORIARTY, P. 2014. Evaluating the Effects of Prescription Fish Oil, Supplemental Fish Oil and a Krill Oil Blend on Serum Lipids/Lipoproteins and the Omega-3 Index: A Pilot Study. Journal of Glycomics & Lipidomics, 4, 1. BARNES, D. K. & TARLING, G. A. 2017. Polar oceans in a changing climate. Current Biology, 27, R454-R460. BARROS, M. P., POPPE, S. C. & BONDAN, E. F. 2014. Neuroprotective Properties of the Marine Carotenoid Astaxanthin and Omega-3 Fatty Acids, and Perspectives for the Natural Combination of Both in Krill Oil. Nutrients, 6, 1293-1317. 153 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 BECKER, R. A. W., A.R 2017. maps: Draw Geographical Maps. R package version 3.2.0, [https://CRAN.R-project.org/package=maps]. BEHRENFELD, M. J., HU, Y., O’MALLEY, R. T., BOSS, E. S., HOSTETLER, C. A., SIEGEL, D. A., SARMIENTO, J. L., SCHULIEN, J., HAIR, J. W. & LU, X. 2017. Annual boom–bust cycles of polar phytoplankton biomass revealed by space-based lidar. Nature Geoscience, 10, 118. BELCHER, A., MANNO, C., WARD, P., HENSON, S. A., SANDERS, R. & TARLING, G. A. 2017. Copepod faecal pellet transfer through the meso- and bathypelagic layers in the Southern Ocean in spring. Biogeosciences, 14, 1511-1525. BESTLEY, S., VAN WIJK, E., ROSENBERG, M., ERIKSEN, R., CORNEY, S., TATTERSALL, K. & RINTOUL, S. 2018. Ocean circulation and frontal structure near the southern Kerguelen Plateau: the physical context for the Kerguelen Axis ecosystem study. Deep Sea Research Part II: Topical Studies in Oceanography. BIJMA, J., PÖRTNER, H. O., YESSON, C. & ROGERS, A. D. 2013. Climate change and the oceans – What does the future hold? Marine Pollution Bulletin, 74, 495-505. BLIGH, E. G. & DYER, W. J. 1959. A rapid method of total lipid extraction and purification. Canadian journal of biochemistry and physiology, 37, 911-917. BODUNGEN, B. V., SMETACEK, V. S., TILZER, M. M. & ZEITZSCHEL, B. 1986. Primary production and sedimentation during spring in the Antarctic Peninsula region. Deep Sea Research Part A. Oceanographic Research Papers, 33, 177-194. BOTTINO, N. 1974. The fatty acids of Antarctic phytoplankton and euphausiids. Fatty acid exchange among trophic levels of the Ross Sea. Marine Biology, 27, 197-204. BRIERLEY, A. S., FERNANDES, P. G., BRANDON, M. A., ARMSTRONG, F., MILLARD, N. W., MCPHAIL, S. D., STEVENSON, P., PEBODY, M., PERRETT, J. & SQUIRES, M. 2002. Antarctic krill under sea ice: elevated abundance in a narrow band just south of ice edge. Science, 295, 1890-1892. BROWN, M. E. C. 2010. Environmental effects on the growth, maturation and physiology in antarctic krill (Euphausia superba) over an annual cycle : an experimental approach. PhD Thesis, University of Tasmania. BRYAN, J. 2015. gapminder: Data from Gapminder. R package 0.2.0 [https://CRAN.Rproject.org/package=gapminder]. BUTLER, W. E. 2007. Omega-3 Fatty Acids—Science, Efficacy, and Clinical Use in Cardiology. 22-26 pp. CABROL, J., TROMBETTA, T., AMAUDRUT, S., AULANIER, F., SAGE, R., TREMBLAY, R., NOZAIS, C., STARR, M., PLOURDE, S. & WINKLER, G. 2019. Trophic niche partitioning of dominant North‐Atlantic krill species, Meganyctiphanes norvegica, Thysanoessa inermis, and T. raschii. Limnology and Oceanography, 64, 165-181. CARAVEO‐PATIÑO, J., WANG, Y., SOTO, L. A., GHEBREMESKEL, K., LEHANE, C. & CRAWFORD, M. A. 2009. Eco‐physiological repercussions of dietary arachidonic acid in cell membranes of active tissues of the Gray whale. Marine Ecology, 30, 437447. CCAMLR, S. C. 2012. Report of the 31st Meeting of the Scientific Committee, Hobart, Australia, 22-26 October, 2012. SC-CCAMLR-XXXI. Scientific Committee for the Conservation of Antarctic Marine Living Resources, Hobart, Tasmania, Australia 400 pp. CCAMLR, S. C. 2017. Report of the 36th Meeting of the Scientific Committee, Hobart, Australia, 21-28 October, 2017. SC-CCAMLR-XXXI. Scientific Committee for the Conservation of Antarctic Marine Living Resources, Hobart, Tasmania, Australia 154 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 CHAMPELY, S. 2018. pwr: Basic Power Functions for Power Analysis. R package version 1.2.2., [https://CRAN.R-project.org/package=pwr]. CHARRASSIN, J. B., MAHO, Y. L. & BOST, C. A. 2002. Seasonal changes in the diving parameters of king penguins (Aptenodytes patagonicus). Marine Biology, 141, 581589. CHEN, S. 2012. Lipidomics of marine biomass: fish, fish liver by-product and krill. In: CHEN, S. (ed.) Lipidomics: Sea food, marine based dietary supplement, fruit and seed. Kerala, India: Transworld Research Network. CHIBA, S., ISHIMARU, T., HOSIE, G. W. & FUKUCHI, M. 2001. Spatio-temporal variability of zooplankton community structure off east Antarctica (90 to 160°E). Marine Ecology Progress Series, 216, 95-108. CLARKE, A. 1980. The biochemical composition of krill, Euphausia superba Dana, from South Georgia. Journal of Experimental Marine Biology and Ecology, 43, 221-236. CLARKE, A. & MORRIS, D. 1983. Towards an energy budget for krill: the physiology and biochemistry of Euphausia superba Dana. Polar Biology, 2, 69-86. CLARKE, A. 1984. Lipid content and composition of Antarctic krill, Euphausia superba Dana. Journal of Crustacean Biology, 4, 285-294. CONSTABLE, A. J., DE LA MARE, W. K., AGNEW, D. J., EVERSON, I. & MILLER, D. 2000. Managing fisheries to conserve the Antarctic marine ecosystem: practical implementation of the Convention on the Conservation of Antarctic Marine Living Resources (CCAMLR). ICES Journal of marine Science, 57, 778-791. CONSTABLE, A. & NICOL, S. 2002. Defining smaller-scale management units to further develop the ecosystem approach in managing large-scale pelagic krill fisheries in Antarctica. Ccamlr Science, 9, 117-131. CONSTABLE, A. J., MELBOURNE‐THOMAS, J., CORNEY, S. P., ARRIGO, K. R., BARBRAUD, C., BARNES, D. K., BINDOFF, N. L., BOYD, P. W., BRANDT, A. & COSTA, D. P. 2014. Climate change and Southern Ocean ecosystems I: how changes in physical habitats directly affect marine biota. Global Change Biology, 20, 3004-3025. CUZIN-ROUDY, J., ALBESSARD, E., VIRTUE, P. & MAYZAUD, P. 1999. The scheduling of spawning with the moult cycle in Northern krill (Crustacea: Euphausiacea): a strategy for allocating lipids to reproduction. Invertebrate reproduction & development, 36, 163-170. CUZIN-ROUDY, J. & BUCHHOLZ, F. 1999. Ovarian development and spawning in relation to the moult cycle in Northern krill, Meganyctiphanes norvegica (Crustacea: Euphausiacea), along a climatic gradient. Marine Biology, 133, 267-281. CUZIN-ROUDY, J., TARLING, G. & STROMBERG, J. 2004. Life cycle strategies of Northern krill (Meganyctiphanes norvegica) for regulating growth, moult, and reproductive activity in various environments: the case of fjordic populations. ICES Journal of Marine Science, 61, 721-737. DALL, W., CHANDUMPAI, A. & SMITH, D. 1992. Fatty acid composition of organs and tissues of the tiger prawn Penaeus esculentus during the moulting cycle and during starvation. Marine Biology, 113, 45-55. DALY, K. L. 2004. Overwintering growth and development of larval Euphausia superba: an interannual comparison under varying environmental conditions west of the Antarctic Peninsula. Deep Sea Research Part II: Topical Studies in Oceanography, 51, 21392168. DANA, J. D. 1850. Synopsis generum crustaceorum ordinis "Schizopeda". American Journal of Science and Arts ser. 2., 9, 129-133. 155 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 DAWSON, A. L., KAWAGUCHI, S., KING, C. K., TOWNSEND, K. A., KING, R., HUSTON, W. M. & BENGTSON NASH, S. M. 2018. Turning microplastics into nanoplastics through digestive fragmentation by Antarctic krill. Nature Communications, 9, 1001. DEACON, G. E. R., MARR, J. W. S. & RAYNER, G. W. 1939. The Antarctic Voyages of R. R. S. Discovery II and R. R. S. William Scoresby, 1935-37. The Geographical Journal, 93, 185-204. DEMER, D. A. 2004. An estimate of error for the CCAMLR 2000 survey estimate of krill biomass. Deep Sea Research Part II: Topical Studies in Oceanography, 51, 12371251. DEPPELER, S. L. & DAVIDSON, A. T. 2017. Southern Ocean phytoplankton in a changing climate. Frontiers in Marine Science, 4, 40. DESCAMPS, S., TARROUX, A., CHEREL, Y., DELORD, K., GODØ, O. R., KATO, A., KRAFFT, B. A., LORENTSEN, S. H., ROPERT-COUDERT, Y. & SKARET, G. 2016. At-sea distribution and prey selection of Antarctic petrels and commercial krill fisheries. PloS one, 11, e0156968. DUCKLOW, H. W., BAKER, K., MARTINSON, D. G., QUETIN, L. B., ROSS, R. M., SMITH, R. C., STAMMERJOHN, S. E., VERNET, M. & FRASER, W. 2007. Marine pelagic ecosystems: the West Antarctic Peninsula. Philosophical Transactions of the Royal Society B: Biological Sciences, 362, 67-94. DUTKIEWICZ, S., HICKMAN, A. E., JAHN, O., HENSON, S., BEAULIEU, C. & MONIER, E. 2019. Ocean colour signature of climate change. Nature communications, 10, 578. EDDELBUETTEL, D. 2018. anytime: Anything to 'POSTIXct' or 'Date' Converter. R package version 0.3.3, https://CRAN.R-project.org/package=anytime. EL-SAYED, S. & WEBER, L. 1982. Spatial and temporal variations in phytoplankton biomass and primary productivity in the Southwest Atlantic and the Scotia Sea. Polar Biology, 1, 83-90. ERICSON, J. A., HELLESSEY, N., NICHOLS, P. D., KAWAGUCHI, S., NICOL, S., HOEM, N. & VIRTUE, P. 2018a. Seasonal and Interannual Variation in the Fatty Acid Content and Composition of Euphausia superba samples derived from the Scotia Sea fishery Journal of Crustacean Biology, 38, 662-672. ERICSON, J. A., HELLESSEY, N., KAWAGUCHI, S., NICOL, S., NICHOLS, P. D., HOEM, N. & VIRTUE, P. 2018b. Adult Antarctic krill proves resilient in a simulated high CO2 ocean. Communications Biology, 1, 190. ERICSON, J. A., HELLESSEY, N., KAWAGUCHI, S., NICHOLS, P. D., NICOL, S., HOEM, N. & VIRTUE, P. 2019a. Near-future ocean acidification does not alter the lipid content and fatty acid composition of adult Antarctic krill. Scientific reports, 9, 1-10. ERICSON, J. A., HELLESSEY, N., KAWAGUCHI, S., NICOL, S., NICHOLS, P. D., HOEM, N. & VIRTUE, P. 2019b. New insights into the seasonal diet of Antarctic krill using triacylglycerol and phospholipid fatty acids, and sterol composition. Polar Biology, 1-12. EVERSON, I. 2008. Krill: Biology, Ecology and Fisheries, Wiley. FALK-PETERSEN, S. 1981. Ecological investigations on the zooplankton community of Balsfjorden, northern Norway: Seasonal changes in body weight and the main biochemical composition of Thysanoessa inermis (Krøyer), T. Raschii (M. Sars), and Meganyctiphanes norvegica (M. Sars) in relation to environmental factors. Journal of Experimental Marine Biology and Ecology, 49, 103-120. 156 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 FALK-PETERSEN, S., GATTEN, R. R., SARGENT, J. R. & HOPKINS, C. C. E. 1981. Ecological investigations on the zooplankton community in Balsfjorden, Northern Norway: seasonal changes in the lipid class composition of Meganyctiphanes norvegica (M. Sars), Thysanoessa raschii (M. Sars), and T. inermis (Krøyer). Journal of Experimental Marine Biology and Ecology, 54, 209-224. FALK-PETERSEN, S., HAGEN, W., KATTNER, G., CLARKE, A. & SARGENT, J. 2000. Lipids, trophic relationships, and biodiversity in Arctic and Antarctic krill. Canadian Journal of Fisheries and Aquatic Sciences, 57, 178-191. FLORES, H., ATKINSON, A., KAWAGUCHI, S., KRAFFT, B. A., MILINEVSKY, G., NICOL, S., REISS, C., TARLING, G. A., WERNER, R., BRAVO-REBOLLEDO, E., CIRELLI, V., CUZIN-ROUDY, J., FIELDING, S., VAN FRANEKER, J. A., GROENEVELD, J. J., HARALDSSON, M., LOMBANA, A., MARSCHOFF, E., MEYER, B., PAKHOMOV, E. A., VAN DE PUTTE, A. P., ROMBOLÁ, E., SCHMIDT, K., SIEGEL, V., TESCHKE, M., TONKES, H., TOULLEC, J. Y., TRATHAN, P. N., TREMBLAY, N. & WERNER, T. 2012. Impact of climate change on Antarctic krill. Marine Ecology Progress Series, 458, 1-19. FRASER, W. R. & HOFMANN, E. E. 2003. A predator¹s perspective on causal links between climate change, physical forcing and ecosystem response. Marine Ecology Progress Series, 265, 1-15. FOX, J. 2003. Effect displays in R for generalised linear models. Journal of Statistical Software, 8, 1-27. FRICKE, H., GERCKEN, G., SCHREIBER, W. & OEHLENSCHLÄGER, J. 1984. Lipid, sterol and fatty acid composition of Antarctic krill (Euphausia superba Dana). Lipids, 19, 821-827. GARIBOTTI, I. A., VERNET, M. & FERRARIO, M. E. 2005. Annually recurrent phytoplanktonic assemblages during summer in the seasonal ice zone west of the Antarctic Peninsula (Southern Ocean). Deep Sea Research Part I: Oceanographic Research Papers, 52, 1823-1841. GIGLIOTTI, J. C., DAVENPORT, M. P., BEAMER, S. K., TOU, J. C. & JACZYNSKI, J. 2011. Extraction and characterisation of lipids from Antarctic krill (Euphausia superba). Food Chemistry, 125, 1028-1036. GRAVES, S., PIEPHO, H.-P., SELZER, L. & ASSISTED BY DORAI-RAJ, S. 2015. multcompView: Visualizations of Paired Comparisons. R package version 0.1-7, https://CRAN.R-project.org/package=multcompView. HAGEN, W., VAN VLEET, E. & KATTNER, G. 1996. Seasonal lipid storage as overwintering strategy of Antarctic krill. Marine ecology progress series. Oldendorf, 134, 85-89. HAGEN, W., KATTNER, G., TERBRÜGGEN, A. & VAN VLEET, E. S. 2001. Lipid metabolism of the Antarctic krill Euphausia superba and its ecological implications. Marine Biology, 139, 95-104. HAGEN, W. & KATTNER, G. 2014. Lipid dynamics and biosynthetic pathways of dominant krill species-a global view of Antarctic via tropical to Arctic euphausiids. SCAR Open Science Conference: Global messages from Antarctica. HANCOCK, A. M., DAVIDSON, A. T., MCKINLAY, J., MCMINN, A., SCHULZ, K. G. & ENDEN, R. L. 2018. Ocean acidification changes the structure of an Antarctic coastal protistan community. Biogeosciences, 15, 2393-2410. HARDY, A. C. 1928. The work of the royal research ship" Discovery" in the dependencies of the Falkland Islands. The Geographical Journal, 72, 209-227. HEERAH, K., ANDREWS-GOFF, V., WILLIAMS, G., SULTAN, E., HINDELL, M., PATTERSON, T. & CHARRASSIN, J.-B. 2013. Ecology of Weddell seals during 157 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 winter: Influence of environmental parameters on their foraging behaviour. Deep Sea Research Part II: Topical Studies in Oceanography, 88–89, 23-33. HELBLING, E. W., VILLAFAÑE, V. E. & HOLM-HANSEN, O. 1995. Variability of phytoplankton distribution and primary production around Elephant Island, Antarctica, during 1990–1993. Polar Biology, 15, 233-246. HELLESSEY, N., ERICSON, J. A., NICHOLS, P. D., KAWAGUCHI, S., NICOL, S., HOEM, N. & VIRTUE, P. 2018. Seasonal and Interannual variation in the lipid content and composition of Euphausia superba Dana, 1850 (Euphausiacea) samples derived from the Scotia Sea fishery. Journal of Crustacean Biology, 38, 673-681. HELLESSEY, N., ERICSON, J. A., NICHOLS, P. D., KAWAGUCHI, S., NICOL, S., HOEM, N. & VIRTUE, P. (in review). Regional variability of Antarctic krill (Euphausia superba) diet as determined using lipid, fatty acid and sterol composition. Polar Biology HENDERSON, R., SARGENT, J. & FALK-PETERSEN, S. 1981. Lipogenesis in the Arctic euphausiid Thysanoessa inermis. Marine Biology, 63, 235-240. HILL, S., MURPHY, E., REID, K., TRATHAN, P. & CONSTABLE, A. 2006. Modelling Southern Ocean ecosystems: krill, the food-web, and the impacts of harvesting. Biological Reviews, 81, 581-608. HILL, S. L., 2013. Prospects for a sustainable increase in the availability of long chain omega 3s: lessons from the Antarctic krill fishery. in Omega-6/3 Fatty Acids, 267-296. Springer. HILL, S. & CANNON, M. 2013. A potential feedback approach to ecosystem-based management: model predictive control of the Antarctic krill fishery. CCAMLR Science, 20, 119-137. HILL, S. L., PHILLIPS, T. & ATKINSON, A. 2013. Potential climate change effects on the habitat of Antarctic krill in the Weddell quadrant of the Southern Ocean. PloS one, 8, e72246. HILL, S., ATKINSON, A., DARBY, C., FIELDING, S., KRAFFT, B. A., GODØ, O., SKARET, G., TRATHAN, P. N. & WATKINS, J. L. 2016. Is current management of the Antarctic krill fishery in the Atlantic sector of the Southern Ocean precautionary? CCAMLR Science, 23, 31-51. HOLM-HANSEN, O., AMOS, A. F., VILLAFAÑE, V. & HELBLING, E. W. 1994. In situ evidence for a nutrient limitation of phytoplankton growth in pelagic Antarctic waters. Antarctic Science, 6, 315-324. HOTHORN, T., BRETZ, F. & WESTFALL, P. 2008. Simultaneous Inference in General Parametric Models. Biometrical Journal, 50, 346-363. HUNTLEY, M. E., NORDHAUSEN, W. & LOPEZ, M. D. 1994. Elemental composition, metabolic activity and growth of Antarctic krill Euphausia superba during winter. Marine Ecology Progress Series, 23-40. IKEDA, T. & DIXON, P. 1982. Body shrinkage as a possible over-wintering mechanism of the Antarctic krill, Euphausia superba Dana. Journal of Experimental Marine Biology and Ecology, 62, 143-151. IOCCG, 2015. Ocean Colour Remote Sensing in Polar Seas. Babin, M., Arrigo, K., Bélanger, S. and Forget, M-H. (eds.). IOCCG Report Series, No. 16. ITONORI, S., KAMEMURA, K., NARUSHIMA, K., SONKU, N., ITASAKA, O., HORI, T. & SUGITA, M. 1991. Characterization of a new phosphonocerebroside, N-methyl-2aminoethylphosphonylglucosylceramide, from the Antarctic krill, Euphausia superba. Biochimica et Biophysica Acta (BBA) - Lipids and Lipid Metabolism, 1081, 321-327. 158 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 IVERSON, S. J., FIELD, C., DON BOWEN, W. & BLANCHARD, W. 2004. Quantitative fatty acid signature analysis: a new method of estimating predator diets. Ecological Monographs, 74, 211-235. JACOBS, S. S., AMOS, A. F. & BRUCHHAUSEN, P. M. Ross Sea oceanography and Antarctic bottom water formation. Deep Sea Research and Oceanographic Abstracts, 1970. Elsevier, 935-962. JANOUT, M. A., HÖLEMANN, J., WAITE, A. M., KRUMPEN, T., APPEN, W. J. & MARTYNOV, F. 2016. Sea‐ice retreat controls timing of summer plankton blooms in the Eastern Arctic Ocean. Geophysical Research Letters, 43, 12-493. JARVIS, T., KELLY, N., KAWAGUCHI, S., VAN WIJK, E. & NICOL, S. 2010. Acoustic characterisation of the broad-scale distribution and abundance of Antarctic krill (Euphausia superba) off East Antarctica (30-80 E) in January-March 2006. Deep Sea Research Part II: Topical Studies in Oceanography, 57, 916-933. JOHNSON, R., STRUTTON, P. G., WRIGHT, S. W., MCMINN, A. & MEINERS, K. M. 2013. Three improved satellite chlorophyll algorithms for the Southern Ocean. Journal of Geophysical Research: Oceans, 118, 3694-3703. JU, S. J. & HARVEY, H. R. 2004. Lipids as markers of nutritional condition and diet in the Antarctic krill Euphausia superba and Euphausia crystallorophias during austral winter. Deep Sea Research Part II: Topical Studies in Oceanography, 51, 2199-2214. KAHLE, D. & WICKHAM, H. 2013. ggmap: Spatial Visualisation with ggplot2. The R journal, 5, 144-161. KAHRU, M., LEE, Z. & MITCHELL, B. G. 2017. Contemporaneous disequilibrium of biooptical properties in the Southern Ocean. Geophysical Research Letters, 44, 28352842. KATTNER, G., HAGEN, W., LEE, R. F., CAMPBELL, R., DEIBEL, D., FALKPETERSEN, S., GRAEVE, M., HANSEN, B. W., HIRCHE, H. J. & JÓNASDÓTTIR, S. H. 2007. Perspectives on marine zooplankton lipids. Canadian Journal of Fisheries and Aquatic Sciences, 64, 1628-1639. KAWAGUCHI, S. & SATAKE, M. 1994. Relationship between Recruitment of the Antarctic Krill and the Degree of Ice Cover near the South Shetland Islands. Fisheries science, 60, 123-124. KAWAGUCHI, S. & NICOL, S. 2007. Learning about Antarctic krill from the fishery. Antarctic Science, 19, 219-230. KAWAGUCHI, S., KING, R., MEIJERS, R., OSBORN, J. E., SWADLING, K. M., RITZ, D. A. & NICOL, S. 2010a. An experimental aquarium for observing the schooling behaviour of Antarctic krill (Euphausia superba). Deep Sea Research Part II: Topical Studies in Oceanography, 57, 683-692. KAWAGUCHI, S., NICOL, S., VIRTUE, P., DAVENPORT, S. R., CASPER, R., SWADLING, K. M. & HOSIE, G. W. 2010b. Krill demography and large-scale distribution in the Western Indian Ocean sector of the Southern Ocean (CCAMLR Division 58.4. 2) in Austral summer of 2006. Deep Sea Research Part II: Topical Studies in Oceanography, 57, 934-947. KAWAGUCHI, S., KURIHARA, H., KING, R., HALE, L., BERLI, T., ROBINSON, J. P., ISHIDA, A., WAKITA, M., VIRTUE, P. & NICOL, S. 2011. Will krill fare well under Southern Ocean acidification? Biology Letters, 7, 288-291. KAWAGUCHI, S., ISHIDA, A., KING, R., RAYMOND, B., WALLER, N., CONSTABLE, A., NICOL, S., WAKITA, M. & ISHIMATSU, A. 2013. Risk maps for Antarctic krill under projected Southern Ocean acidification. Nature Climate Change, 3, 843-847. KAWAGUCHI, S. 2016. Reproduction and larval development in Antarctic krill (Euphausia superba). Biology and Ecology of Antarctic Krill, Springer, 225-246. 159 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 KAWAGUCHI, S. pers. comm. Cnr Channel Hwy and Algona Rd, Kingston, Tasmania, 7050. Australian Antarctic Division, 2017. KILADA, R., REISS, C. S., KAWAGUCHI, S., KING, R. A., MATSUDA, T. & ICHII, T. 2017. Validation of band counts in eyestalks for the determination of age of Antarctic krill, Euphausia superba. PLOS ONE, 12, e0171773. KIM, M. A., JUNG, H.R., LEE, Y.B., CHUN, B.S. & KIM, S.B. 2014. Monthly variations in the nutritional composition of Antarctic krill Euphausia superba. Fisheries and Aquatic Sciences, 17, 409-419. KIRKWOOD, J. M. 1982. A guide to the Euphausiacea of the Southern Ocean, Information Services Section, Antarctic Division, Dept. of Science and Technology. KLEVJER, T. A., TARLING, G. A. & FIELDING, S. 2010. Swarm characteristics of Antarctic krill Euphausia superba relative to the proximity of land during summer in the Scotia Sea. Marine Ecology Progress Series, 409, 157-170. KOHLBACH, D., GRAEVE, M., DAVID, C., LANGE, B. & FLORES, H. 2015. Dynamics in diet sources of Antarctic krill species revealed by lipid and stable isotope analyses. KOHLBACH, D., LANGE, B. A., SCHAAFSMA, F. L., DAVID, C., VORTKAMP, M., GRAEVE, M., VAN FRANEKER, J. A., KRUMPEN, T. & FLORES, H. 2017. Ice Algae-Produced Carbon Is Critical for Overwintering of Antarctic Krill Euphausia superba. Frontiers in Marine Science, 4. KOHLBACH, D., GRAEVE, M., LANGE, B. A., DAVID, C., SCHAAFSMA, F. L., VAN FRANEKER, J. A., VORTKAMP, M., BRANDT, A. & FLORES, H. 2018. Dependency of Antarctic zooplankton species on ice algae‐produced carbon suggests a sea ice‐driven pelagic ecosystem during winter. Global change biology, 24, 46674681. KOLAKOWSKA, A., KOLAKOWSKI, E. & SZCZYGIELSKI, M. 1994. Winter season krill (Euphausia superba D.) as a source of n‐3 polyunsaturated fatty acids. Food/Nahrung, 38, 128-134. KORB, R. E., WHITEHOUSE, M. J., THORPE, S. E. & GORDON, M. 2005. Primary production across the Scotia Sea in relation to the physico-chemical environment. Journal of Marine Systems, 57, 231-249. KWANTES, J. M. & GRUNDMANN, O. 2015. A brief review of krill oil history, research, and the commercial market. Journal of dietary supplements, 12, 23-35. LENTH, R. V. 2016. Least-Squares Means: The R Package lsmeans. Journal of Statistical Software, 69, 1-33. LEONORI, I., DE FELICE, A., CANDUCI, G., COSTANTINI, I., BIAGIOTTI, I., GIULIANI, G. & BUDILLON, G. 2017. Krill distribution in relation to environmental parameters in mesoscale structures in the Ross Sea. Journal of Marine Systems, 166, 159-171. LOEB, V., SIEGEL, V., HOLM-HANSEN, O., HEWITT, R., FRASER, W., TRIVELPLECE, W. & TRIVELPLECE, S. 1997. Effects of sea-ice extent and krill or salp dominance on the Antarctic food web. Nature, 387, 897-900. LOEB, V. J. & SANTORA, J. A. 2015. Climate variability and spatiotemporal dynamics of five Southern Ocean krill species. Progress in Oceanography, 134, 93-122. MACKEY, A. P., ATKINSON, A., HILL, S. L., WARD, P., CUNNINGHAM, N. J., JOHNSTON, N. M. & MURPHY, E. J. 2012. Antarctic macrozooplankton of the southwest Atlantic sector and Bellingshausen Sea: Baseline historical distributions (Discovery Investigations, 1928–1935) related to temperature and food, with projections for subsequent ocean warming. Deep Sea Research Part II: Topical Studies in Oceanography, 59-60, 130-146. MANTEL, L. 1983. The Biology of Crustacea. Vol. 5. Academic Press. 160 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 MARSCHALL, H.P. & HIRCHE, H.J. 1984. Development of eggs and nauplii of Euphausia superba. Polar biology, 2, 245-250. MARSCHALL, H.P. 1988. The overwintering strategy of Antarctic krill under the pack-ice of the Weddell Sea. Polar Biology, 9, 129-135. MASSOM, R.A. & STAMMERJOHN, S.E. 2010. Antarctic sea ice change and variability – Physical and ecological implications. Polar Science, 4, 149-186. MAYZAUD, P. 1997. Spatial and life-cycle changes in lipid and fatty acid structure of the Antarctic euphausiid Euphausia superba. in Antarctic Communities: Species, Structure and Survival, Cambridge University Press. MAYZAUD, P., ALBESSARD, E. & CUZIN-ROUDY, J. 1998. Changes in lipid composition of the Antarctic krill Euphausia superba in the Indian sector of the Antarctic Ocean: influence of geographical location, sexual maturity stage and distribution among organs. Marine Ecology Progress Series, 173, 149-162. MAYZAUD, P., ALBESSARD, E., VIRTUE, P. & BOUTOUTE, M. 2000. Environmental constraints on the lipid composition and metabolism of euphausiids: the case of Euphausia superba and Meganyctiphanes norvegica. Canadian Journal of Fisheries and Aquatic Sciences, 57, 91-103. MAYZAUD, P., CHEVALLIER, J., TAVERNIER, E., MOTEKI, M. & KOUBBI, P. 2011. Lipid composition of the Antarctic fish Pleuragramma antarcticum. Influence of age class. Polar Science, 5, 264-271. MÉNDEZ, L., DASILVA, G., TALTAVULL, N., ROMEU, M. & MEDINA, I. 2017. Marine lipids on cardiovascular diseases and other chronic diseases induced by diet: An insight provided by proteomics and lipidomics. Marine drugs, 15, 258. MEYER, B., FREIER, U., GRIMM, V., GROENEVELD, J., HUNT, B. P., KERWATH, S., KING, R., KLAAS, C., PAKHOMOV, E. & MEINERS, K. M. 2017. The winter pack-ice zone provides a sheltered but food-poor habitat for larval Antarctic krill. Nature ecology & evolution, 1, 1853. MILLER, D. & HAMPTON, I. 1989. Biology and ecology of the Antarctic krill (Euphausia superba Dana): a review, Scientific Committee on Antarctic Research. MOLINE, M. A., PREZELIN, B. B. & SCHOFIELD, O. 1997. Palmer LTER: Stable interannual successional patterns of phytoplankton communities in the coastal waters off Palmer Station, Antarctica. Antarctic Journal of the United States, 32, 151. MONTES-HUGO, M., DONEY, S. C., DUCKLOW, H. W., FRASER, W., MARTINSON, D., STAMMERJOHN, S. E. & SCHOFIELD, O. 2009. Recent changes in phytoplankton communities associated with rapid regional climate change along the western Antarctic Peninsula. Science, 323, 1470-1473. MOORE, J. K. & ABBOTT, M. R. 2002. Surface chlorophyll concentrations in relation to the Antarctic Polar Front: seasonal and spatial patterns from satellite observations. Journal of Marine Systems, 37, 69-86. MORI, M. & BUTTERWORTH, D. 2004. Consideration of multispecies interactions in the Antarctic: a preliminary model of the minke whale–blue whale–krill interaction. African Journal of Marine Science, 26, 245-259. MORI, M., CORNEY, S. P., MELBOURNE-THOMAS, J., KLOCKER, A., KAWAGUCHI, S., CONSTABLE, A. & SUMNER, M. 2019. Modelling dispersal of juvenile krill released from the Antarctic ice edge: Ecosystem implications of ocean movement. Journal of Marine Systems, 189, 50-61. MORRIS, D. J., EVERSON, I., RICKETTS, C. & WARD, P. 1984. Feeding of krill around South Georgia. II. Relations between feeding activity, environment and vertical distribution. Marine Ecology Progress Series, 20, 203-206. 161 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 MURPHY, E. J., WATKINS, J. L., TRATHAN, P. N., REID, K., MEREDITH, M. P., THORPE, S. E., JOHNSTON, N. M., CLARKE, A., TARLING, G. A., COLLINS, M. A., FORCADA, J., SHREEVE, R. S., ATKINSON, A., KORB, R., WHITEHOUSE, M. J., WARD, P., RODHOUSE, P. G., ENDERLEIN, P., HIRST, A. G., MARTIN, A. R., HILL, S. L., STANILAND, I. J., POND, D. W., BRIGGS, D. R., CUNNINGHAM, N. J. & FLEMING, A. H. 2007. Spatial and temporal operation of the Scotia Sea ecosystem: a review of large-scale links in a krill centred food web. Philosophical Transactions of the Royal Society B: Biological Sciences, 362, 113148. NICHOLS, P. D., PALMISANO, A. C., SMITH, G. A. & WHITE, D. C. 1986. Lipids of the Antarctic sea ice diatom Nitzschia cylindrus. Phytochemistry, 25, 1649-1653. NICHOLS, P. D., PALMISANO, A. C., VOLKMAN, J. K., SMITH, G. A. & WHITE, D. C. 1988. Occurence of an isoprenoid C25 diunsaturated alkene and high neutral lipid content in Antarctic sea-ice diatom communities. Journal of Phycology, 24, 90-96. NICOL, S., STOLP, M., COCHRAN, T., GEIJSEL, P. & MARSHALL, J. 1992. Growth and shrinkage of Antarctic krill Euphausia superba from the Indian Ocean sector of the Southern Ocean during summer. Marine Ecology Progress Series, 175-181. NICOL, S., WILLIAM, K. & STOLP, M. 1995. The energetic cost of egg production in Antarctic krill (Euphausia superba Dana). Antarctic Science, 7, 25-30. NICOL, S. & ENDO, Y. 1999. Krill fisheries: development, management and ecosystem implications. Aquatic Living Resources, 12, 105-120. NICOL, S., CONSTABLE, A. & PAULY, T. 2000a. Estimates of circumpolar abundance of Antarctic krill based on recent acoustic density measurements. CCAMLR Science, 7, 87-99. NICOL, S., KITCHENER, J., KING, R., HOSIE, G. & DE LA MARE, W. K. 2000b. Population structure and condition of Antarctic krill (Euphausia superba) off East Antarctica (80–150°E) during the Austral summer of 1995/1996. Deep Sea Research Part II: Topical Studies in Oceanography, 47, 2489-2517. NICOL, S., PAULY, T., BINDOFF, N. L., WRIGHT, S., THIELE, D., HOSIE, G. W., STRUTTON, P. G. & WOEHLER, E. 2000c. Ocean circulation off east Antarctica affects ecosystem structure and sea-ice extent. Nature, 406, 504-507. NICOL, S., MEINERS, K. & RAYMOND, B. 2010. BROKE-West, a large ecosystem survey of the South West Indian Ocean sector of the Southern Ocean, 30 E–80 E (CCAMLR Division 58.4. 2). Deep-Sea Research Part II, 9, 693-700. NICOL, S., FOSTER, J. & KAWAGUCHI, S. 2012. The fishery for Antarctic krill–recent developments. Fish and Fisheries, 13, 30-40. NICOL, S. & FOSTER, J. 2016. The Fishery for Antarctic Krill: Its Current Status and Management Regime. Biology and Ecology of Antarctic Krill, 387-421. NOAA 2013. https://www.ncdc.noaa.gov/sotc/global/201313. Accessed 19th October, 2017. O’BRIEN, C., VIRTUE, P., KAWAGUCHI, S. & NICHOLS, P. 2011. Aspects of krill growth and condition during late winter-early spring off East Antarctica (110–130 E). Deep Sea Research Part II: Topical Studies in Oceanography, 58, 1211-1221. ORSI, A. H. & WIEDERWOHL, C. L. 2009. A recount of Ross Sea waters. Deep Sea Research Part II: Topical Studies in Oceanography, 56, 778-795. PAKHOMOV, E. 2000. Demography and life cycle of Antarctic krill, Euphausia superba, in the Indian sector of the Southern Ocean: long-term comparison between coastal and open-ocean regions. Canadian Journal of Fisheries and Aquatic Sciences, 57, 68-90. PARRISH, C. C. & ACKMAN, R. G. 1985. Calibration of the Iatroscan‐Chromarod system for marine lipid class analyses. Lipids, 20, 521-530. 162 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 PASQUER, B., MONGIN, M., JOHNSTON, N. & WRIGHT, S. 2010. Distribution of particulate organic matter (POM) in the Southern Ocean during BROKE-West (30° E80° E). Deep Sea Research Part II: Topical Studies in Oceanography, 57, 779-793. PASSMORE, A., JARMAN, S., SWADLING, K., KAWAGUCHI, S., MCMINN, A. & NICOL, S. 2006. DNA as a dietary biomarker in Antarctic krill, Euphausia superba. Marine Biotechnology, 8, 686-696. PERISSINOTTO, R., GURNEY, L. & PAKHOMOV, E. 2000. Contribution of heterotrophic material to diet and energy budget of Antarctic krill, Euphausia superba. Marine Biology, 136, 129-135. PHLEGER, C. F., NICHOLS, P. D. & VIRTUE, P. 1997. The lipid, fatty acid and fatty alcohol composition of the myctophid fish Electrona antarctica: high level of wax esters and food-chain implications. Antarctic Science, 9, 258-265. PHLEGER, C. F., NICHOLS, P. D. & VIRTUE, P. 1998. Lipids and trophodynamics of Antarctic zooplankton. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, 120, 311-323. PHLEGER, C. F., NELSON, M. M., MOONEY, B. D. & NICHOLS, P. D. 1999. Wax esters versus triacylglycerols in myctophid fishes from the Southern Ocean. Antarctic Science, 11, 436-444. PHLEGER, C. F., NELSON, M. M., MOONEY, B. D. & NICHOLS, P. D. 2002. Interannual and between species comparison of the lipids, fatty acids and sterols of Antarctic krill from the US AMLR Elephant Island survey area. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, 131, 733-747. PINHEIRO, J., BATES, D., DEBROY, S., SARKAR, D. & R CORE TEAM 2017. nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1-131, https://CRAN.R-project.org/package=nlme. POND, D., WATKINS, J., PRIDDLE, J. & SARGENT, J. 1995. Variation in the lipid content and composition of Antarctic krill Euphausia superba at South Georgia. Marine ecology progress series. Oldendorf, 117, 49-57. POND, D. W., TARLING, G. A. & MAYOR, D. J. 2014. Hydrostatic pressure and temperature effects on the membranes of a seasonally migrating marine copepod. PLoS One, 9, e111043. QUETIN, L., ROSS, R. & CLARKE, A. 1994. Krill energetics: seasonal and environmental aspects of the physiology of Euphausia superba. Southern Ocean ecology: the BIOMASS perspective. Cambridge University Press. R CORE TEAM 2017. R: A language and Environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. RAYNER, N. A. 2003. Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. Journal of Geophysical Research, 108 (D14). ROSS, R. & QUETIN, L. 2000. Reproduction in euphausiacea. in Krill: Biology, ecology and fisheries, Blackwell Science, 150-181. SAETHER, O., ELLINGSEN, T. E. & MOHR, V. 1985. The distribution of lipid in the tissues of Antarctic krill, Euphausia superba. Comparative Biochemistry and Physiology Part B: Comparative Biochemistry, 81, 609-614. SAETHER, O., ELLINGSEN, T. E. & MOHR, V. 1986. Lipids of North Atlantic krill. Journal of lipid research, 27, 274-285. SARKAR, D. 2008. Lattice: Multivariate data visualzation with R. Springer, New York. SARS, M. 1857. Om 3 nye norske Krebsdyr. Forhandlinger verd de Skandinaviske Naturforres Mote 1856, 7, 160-175. 163 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 SAUNDERS, R. A., COLLINS, M. A., WARD, P., STOWASSER, G., HILL, S. L., SHREEVE, R. & TARLING, G. A. 2015. Predatory impact of the myctophid fish community on zooplankton in the Scotia Sea (Southern Ocean). Marine Ecology Progress Series, 541, 45-64. SCHAAFSMA, F. L., KOHLBACH, D., DAVID, C., LANGE, B. A., GRAEVE, M., FLORES, H. & FRANEKER, J. A. V. 2017. Spatio-temporal variability in the winter diet of larval and juvenile Antarctic krill, Euphausia superba, in ice-covered waters. Marine Ecology Progress Series, 580, 101-115. SCHALLENBERG, C., BESTLEY, S., KLOCKER, A., TRULL, T. W., DAVIES, D. M., GAULT‐RINGOLD, M., ERIKSEN, R., RODEN, N. P., SANDER, S. G. & SUMNER, M. 2018. Sustained upwelling of subsurface iron supplies seasonally persistent phytoplankton blooms around the southern Kerguelen plateau, Southern Ocean. Journal of Geophysical Research: Oceans, 123, 5986-6003. SCHMIDT, K., ATKINSON, A., STEIGENBERGER, S., FIELDING, S., LINDSAY, M. C. M., POND, D. W., TARLING, G. A., KLEVJER, T. A., ALLEN, C. S., NICOL, S. & ACHTERBERG, E. P. 2011. Seabed foraging by Antarctic krill: Implications for stock assessment, bentho-pelagic coupling, and the vertical transfer of iron. Limnology and Oceanography, 56, 1411-1428. SCHMIDT, K., ATKINSON, A., VENABLES, H. J. & POND, D. W. 2012. Early spawning of Antarctic krill in the Scotia Sea is fuelled by “superfluous” feeding on non-ice associated phytoplankton blooms. Deep Sea Research Part II: Topical Studies in Oceanography, 59-60, 159-172. SCHMIDT, K., ATKINSON, A., POND, D. W. & IRELAND, L. C. 2014. Feeding and overwintering of Antarctic krill across its major habitats: The role of sea ice cover, water depth, and phytoplankton abundance. Limnology and Oceanography, 59, 17-36. SCHMIDT, K. & ATKINSON, A. 2016. Feeding and food processing in Antarctic krill (Euphausia superba Dana). Biology and Ecology of Antarctic Krill, 175-224. SCHOFIELD, O., DUCKLOW, H. W., MARTINSON, D. G., MEREDITH, M. P., MOLINE, M. A. & FRASER, W. R. 2010. How do polar marine ecosystems respond to rapid climate change? Science, 328, 1520-1523. SCHUTT, E. 2016. The Global Omega-3 Market: A Look at the Numbers and the Consumer. Science of Omega-3: Balancing the Scales, Omega-3 Centre Conference, Sydney, Australia. SIEGEL, V. 1987. Age and growth of Antarctic Euphausiacea (Crustacea) under natural conditions. Marine Biology, 96, 483-495. SIEGEL, V. & LOEB, V. 1995. Recruitment of Antarctic krill Euphausia superba and possible causes for its variability. Marine Ecology Progress Series, 123, 45-56. SIEGEL, V. 2005. Distribution and population dynamics of Euphausia superba: summary of recent findings. Polar Biology, 29, 1-22. SIEGEL, V. 2016. Biology and ecology of Antarctic Krill, Springer. SKERRATT, J., NICHOLS, P., MCMEEKIN, T. & BURTON, H. 1995. Seasonal and interannual changes in planktonic biomass and community structure in eastern Antarctica using signature lipids. Marine Chemistry, 51, 93-113. SMITH, R. C., BAKER, K. S., FRASER, W. R., HOFMANN, E. E., KARL, D. M., KLINCK, J. M., QUETIN, L. B., PREZELIN, B. B., ROSS, R. M. & TRIVELPIECE, W. Z. 1995. The Palmer LTER: A long-term ecological research program at Palmer Station, Antarctica. Oceanography, 8. SMITH JR, W. O., AINLEY, D. G., ARRIGO, K. R. & DINNIMAN, M. S. 2014. The oceanography and ecology of the Ross Sea. Annual review of marine science, 6, 469487. 164 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 STÜBING, D., HAGEN, W. & SCHMIDT, K. 2003. On the use of lipid biomarkers in marine food web analyses: an experimental case study on the Antarctic krill, Euphausia superba. Limnology and Oceanography, 48, 1685-1700. SUHR, S. B., POND, D. W., GOODAY, A. J. & SMITH, C. R. 2003. Selective feeding by benthic foraminifera on phytodetritus on the western Antarctic Peninsula shelf: evidence from fatty acid biomarker analysis. Marine Ecology Progress Series, 262, 153-162. SUN, S., DE LA MARE, W. & NICOL, S. 1995. The compound eye as an indicator of age and shrinkage in Antarctic krill. Antarctic Science, 7, 387-392. SUNG, H. H., SINCLAIR, A. J., HUYNH, K., SMITH, A. T., MELLETT, N. A., MEIKLE, P. J. & SU, X. Q. 2019. Differential plasma postprandial lipidomic responses to krill oil and fish oil supplementations in women: a randomised crossover study. Nutrition. SUTTON, A. 2015. Krill in the Leeuwin Current system: Influence of oceanography and contribution to Indian Ocean zoogeography. Ph.D. thesis, Murdoch University. SWADLING, K. M., KAWAGUCHI, S. & HOSIE, G. W. 2010. Antarctic mesozooplankton community structure during BROKE-West (30 E–80 E), January–February 2006. Deep Sea Research Part II: Topical Studies in Oceanography, 57, 887-904. TAKAICHI, S., MATSUI, K., NAKAMURA, M., MURAMATSU, M. & HANADA, S. 2003. Fatty acids of astaxanthin esters in krill determined by mild mass spectrometry. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, 136, 317-322. TANCELL, C., PHILLIPS, R. A., XAVIER, J. C., TARLING, G. A. & SUTHERLAND, W. J. 2012. Comparison of methods for determining key marine areas from tracking data. Marine Biology, 160, 15-26. TARLING, G. A., KLEVJER, T., FIELDING, S., WATKINS, J., ATKINSON, A., MURPHY, E., KORB, R., WHITEHOUSE, M. & LEAPER, R. 2009a. Variability and predictability of Antarctic krill swarm structure. Deep Sea Research Part I: Oceanographic Research Papers, 56, 1994-2012. TARLING, G. A., CUZIN-ROUDY, J., WOOTTON, K. & JOHNSON, M. L. 2009b. Eggrelease behaviour in Antarctic krill. Polar Biology, 32, 1187-1194. TARLING, G. A., HILL, S., PEAT, H., FIELDING, S., REISS, C. & ATKINSON, A. 2016a. Growth and shrinkage in Antarctic krill Euphausia superba is sex-dependent. Marine Ecology Progress Series, 547, 61-78. TARLING, G. A., PECK, V., WARD, P., ENSOR, N. S., ACHTERBERG, E., TYNAN, E., POULTON, A. J., MITCHELL, E. & ZUBKOV, M. V. 2016b. Effects of acute ocean acidification on spatially-diverse polar pelagic foodwebs: Insights from on-deck microcosms. Deep Sea Research Part II: Topical Studies in Oceanography, 127, 7592. TARLING, G. A., WARD, P. & THORPE, S. E. 2018. Spatial distributions of Southern Ocean mesozooplankton communities have been resilient to long‐term surface warming. Global change biology, 24, 132-142. TÖBE, K., MEYER, B. & FUENTES, V. 2010. Detection of zooplankton items in the stomach and gut content of larval krill, Euphausia superba, using a molecular approach. Polar Biology, 33, 407-414. TOU, J. C., JACZYNSKI, J. & CHEN, Y.-C. 2007. Krill for Human Consumption: Nutritional Value and Potential Health Benefits. Nutrition Reviews, 65, 63-77. TRIVELPIECE, W. Z., HINKE, J. T., MILLER, A. K., REISS, C. S., TRIVELPIECE, S. G. & WATTERS, G. M. 2011. Variability in krill biomass links harvesting and climate warming to penguin population changes in Antarctica. Proceedings of the National Academy of Sciences, 108, 7625-7628. 165 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 TURNER, J. T. 2002. Zooplankton fecal pellets, marine snow and sinking phytoplankton blooms. Aquatic microbial ecology, 27, 57-102. TURNER, J. T. 2015. Zooplankton fecal pellets, marine snow, phytodetritus and the ocean’s biological pump. Progress in Oceanography, 130, 205-248. VARPE, Ø., JØRGENSEN, C., TARLING, G. A. & FIKSEN, Ø. 2007. Early is better: seasonal egg fitness and timing of reproduction in a zooplankton life-history model. Oikos, 116, 1331-1342. VARPE, Ø., JØRGENSEN, C., TARLING, G. A. & FIKSEN, Ø. 2009. The adaptive value of energy storage and capital breeding in seasonal environments. Oikos, 118, 363-370. VERNET, M., MARTINSON, D., IANNUZZI, R., STAMMERJOHN, S., KOZLOWSKI, W., SINES, K., SMITH, R. & GARIBOTTI, I. 2008. Primary production within the sea-ice zone west of the Antarctic Peninsula: I—Sea ice, summer mixed layer, and irradiance. Deep Sea Research Part II: Topical Studies in Oceanography, 55, 20682085. VIRTUE, P., NICOL, S. & NICHOLS, P. 1993a. Changes in the digestive gland of Euphausia superba during short-term starvation: lipid class, fatty acid and sterol content and composition. Marine Biology, 117, 441-448. VIRTUE, P., NICHOLS, P., NICOL, S., MCMINN, A. & SIKES, E. 1993b. The lipid composition of Euphausia superba Dana in relation to the nutritional value of Phaeocystis pouchetii (Hariot) Lagerheim. Antarctic Science, 5, 169-177. VIRTUE, P., NICHOLS, P., NICOL, S. & HOSIE, G. 1996. Reproductive trade-off in male Antarctic krill, Euphausia superba. Marine Biology, 126, 521-527. VIRTUE, P., NICHOLS, P. & NICOL, S. 1997. Dietary-related mechanisms of survival in Euphausia superba: biochemical changes during long-term starvation and bacteria as a possible source of nutrition. Antarctic communities: species, structure, and survival, 193. VIRTUE, P., MAYZAUD, P., ALBESSARD, E. & NICHOLS, P. 2000. Use of fatty acids as dietary indicators in northern krill, Meganyctiphanes norvegica, from northeastern Atlantic, Kattegat, and Mediterranean waters. Canadian Journal of Fisheries and Aquatic Sciences, 57, 104-114. VIRTUE, P., KAWAGUCHI, S., MCIVOR, J., NICOL, S., WOTHERSPOON, S., BROWN, M., CASPER, R., DAVENPORT, S., FINLEY, L. & FOSTER, J. 2010. Krill growth and condition in Western Indian Ocean sector of the Southern Ocean 30-80° E in austral summer 2006. Deep Sea Research Part II: Topical Studies in Oceanography, 57, 948-955. VIRTUE, P., MEYER, B., FREIER, U., NICHOLS, P. D., JIA, Z., KING, R., VIRTUE, J., SWADLING, K. M., MEINERS, K. M. & KAWAGUCHI, S. 2016. Condition of larval (furcilia VI) and one year old juvenile Euphausia superba during the winter– spring transition in East Antarctica. Deep Sea Research Part II: Topical Studies in Oceanography, 131, 182-188. VOLKMAN, J. K. & NICHOLS, P. D. 1991. Applications of thin layer chromatographyflame ionization detection to the analysis of lipids and pollutants in marine and environmental samples. J. Planar Chromatogr, 4, 19-26. WAKEFIELD, S. L., LANE, M., SCHULZ, S. J., HEBART, M. L., THOMPSON, J. G. & MITCHELL, M. 2008. Maternal supply of omega-3 polyunsaturated fatty acids alter mechanisms involved in oocyte and early embryo development in the mouse. American Journal of Physiology-Endocrinology and Metabolism, 294, E425-E434. WARD, P., ATKINSON, A., VENABLES, H. J., TARLING, G. A., WHITEHOUSE, M. J., FIELDING, S., COLLINS, M. A., KORB, R., BLACK, A., STOWASSER, G., SCHMIDT, K., THORPE, S. E. & ENDERLEIN, P. 2012a. Food web structure and 166 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 bioregions in the Scotia Sea: A seasonal synthesis. Deep Sea Research Part II: Topical Studies in Oceanography, 59-60, 253-266. WARD, P., ATKINSON, A. & TARLING, G. 2012b. Mesozooplankton community structure and variability in the Scotia Sea: A seasonal comparison. Deep Sea Research Part II: Topical Studies in Oceanography, 59-60, 78-92. WARNES, G. R., BOLKER, B., BONEBAKKER, L., GENTLEMAN, R., HUBER, W., LIAW, A., LUMLEY, T., MAECHLER, M., MAGNUSSON, A., MOELLER, S., SCHWARTZ, M., VENABLES B. 2016. gplots: various R programming tools for plotting data. R package version 3.0.1., [https://CRAN.R-project.org/package=gplots]. WENK, M. R. 2005. The emerging field of lipidomics. Nature Reviews Drug Discovery, 4, 594-610. WESTWOOD, K. J., GRIFFITHS, F. B., MEINERS, K. M. & WILLIAMS, G. D. 2010. Primary productivity off the Antarctic coast from 30–80 E; BROKE-West survey, 2006. Deep Sea Research Part II: Topical Studies in Oceanography, 57, 794-814. WICKHAM, H. 2007. Reshaping data with the reshape package. HJournal of Statistical Software, 21, 1-20. WICKHAM, H. 2009. ggplot2: Elegant Graphics for Data Analysis. Springer, New York. WICKHAM, H., AND CHANG, W. 2017. devtools: Tools to make developing R packages easier. R package version 1.13.3., [https://CRAN.R-project.org/package=devtools]. WIEDENMANN, J., CRESSWELL, K. A. & MANGEL, M. 2009. Connecting recruitment of Antarctic krill and sea ice. Limnology and Oceanography, 54, 799-811. WIEDERMANN, M., RADEBACH, A., DONGES, J. F., KURTHS, J. & DONNER, R. V. 2016. A climate network-based index to discriminate different types of El Niño and La Niña. Geophysical Research Letters, 43, 7176-7185. XIE, Y. 2018. knitr: A general-purpose package for dynamic report generationin R. R package version 1.20., [https://CRAN.R-project.org/web/packages/knitr/knitr.pdf]. YIN, F., ZHOU, D., XI, M., ZHAO, Q., LIU, Z., LI, D., DONG, X. & ZHU, B. 2016. Influence of Storage Conditions on the Stability of Phospholipids‐Rich Krill (Euphausia superba) Oil. Journal of Food Processing and Preservation, 40, 12471255. YOSHIDA, T. 2009. Environmental influences on reproduction in Antarctic krill, Euphausia superba. Ph.D. Thesis, University of Tasmania. YOSHIDA, T., KAWAGUCHI, S., MEYER, B., VIRTUE, P., PENSCHOW, J. & NASH, G. 2009. Structural changes in the digestive glands of larval Antarctic krill (Euphausia superba) during starvation. Polar biology, 32, 503-507. YOSHIDA, T., VIRTUE, P., KAWAGUCHI, S. & NICHOLS, P. D. 2011. Factors determining the hatching success of Antarctic krill Euphausia superba embryo: lipid and fatty acid composition. Marine biology, 158, 2313-2325. ZENG, C., XU, H. & FISCHER, A. M. 2016. Chlorophyll a estimation around the Antarctica peninsula using satellite algorithms: hints from field water leaving reflectance. Sensors, 16, 2075. 167

RELATED PAPERS

RELATED TOPICS

Log In

Spatiotemporal variability of adult Antarctic krill (Euphausia superba) lipids in relation to sea surface temperature and Chlorophyll a

Spatiotemporal variability of adult Antarctic krill (Euphausia superba) lipids in relation to sea surface temperature and Chlorophyll a

Related Papers

RELATED PAPERS

RELATED TOPICS