Canadian Journal of Fisheries and Aquatic Sciences ! "# $" ( # ) % % & % ' ( * +, ! ! ) / *! / +! $. / $0 * +, / $. / *( ) *2 $ +. / ( 0 3 0# * +!# 4 . / " * +. / ( / 3 55 * %+ ) $ 0 ) " $. / % 1 / * # 6 $ " * af Dr 2 * $* # % # $ t https://mc06.manuscriptcentral.com/cjfas-pubs *( 7 Page 1 of 36 Canadian Journal of Fisheries and Aquatic Sciences 1 2 3 4 Aaron B. Carlisle1*, Steven Y. Litvin1, Daniel J. Madigan 2, Kady Lyons3, Jennifer S. Bigman4, Melissa 5 Ibarra5 & Joseph J. Bizzarro6 6 7 1 Hopkins Marine Station of Stanford University, 120 Oceanview Blvd, Pacific Grove, CA, USA 93950 8 2 Harvard University Center for the Environment, 24 Oxford Street, Cambridge, MA, USA 02138 9 3 University of Calgary, 2500 University Dr NW, Calgary, Alberta, Canada T2N 4N1 10 4 Simon Fraser University, 8888 University Drive, Burnaby, British Columbia, Canada V5A 1S6 11 5 University of California Davis, One Shields Avenue, Davis, CA, USA 95616 12 6 Moss Landing Marine Laboratories, 8272 Moss Landing Road, Moss Landing, CA 95039 13 * Corresponding author Dr 14 Aaron B. Carlisle – aaroncar@stanford.edu 16 Steven Y. Litvin – litvin@stanford.edu 17 Daniel J. Madigan – danieljmadigan@fas.harvard.edu 18 Kady Lyons – kady.lyons@sbcglobal.net 19 Jennifer S. Bigman – jbigman@sfu.ca 20 Melissa Ibarra – mibarra08@yahoo.com 21 Joseph J. Bizzarro – jbizzarro@mlml.calstate.edu t af 15 22 23 24 25 26 27 28 29 1 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 2 of 36 30 Stable isotope analysis (SIA) is becoming a commonly used tool to study the ecology of elasmobranchs. 32 However, the retention of urea by elasmobranchs for osmoregulatory purposes may bias the analysis and 33 interpretation of SIA data. We examined the effects of removing urea and lipid on the stable isotope 34 composition of fourteen species of sharks, skates, and rays from the eastern North Pacific Ocean. While 35 effects were variable across taxa, removal of urea generally increased δ15N and C:N. Urea removal had 36 less influence on δ13C, whereas extracting urea and lipid generally increased δ15N and C:N while also 37 increasing δ13C. Because C:N values of nonFextracted tissues are often used to infer lipid content and 38 adjust δ13C, shifts in C:N following urea extraction will change the inferred lipid content and bias any 39 mathematical adjustment of δ13C. These results highlight the importance of urea and lipid extraction and 40 demonstrate the confounding effects of these compounds, making it impossible to use C:N of nonFureaF 41 extracted samples as a diagnostic tool to estimate and correct for lipid content in elasmobranch tissues. 42 43 Stable isotopes, urea, lipid, carbon, nitrogen, C:N, elasmobranch, mathematical lipid correction, elasmobranch t 44 af Dr 31 45 46 47 48 49 50 51 52 53 54 55 2 https://mc06.manuscriptcentral.com/cjfas-pubs Page 3 of 36 Canadian Journal of Fisheries and Aquatic Sciences 56 57 Stable isotope analysis (SIA) uses the stable isotope composition of organismal tissue to 58 understand a diverse suite of biological and ecological processes. SIA is increasingly being used to 59 investigate the ecology of marine taxa (Peterson and Fry 1987, Michener and Kaufman 2007), including 60 sharks, skates, and rays (elasmobranchs) (Hussey et al. 2012b). Since SIA makes inferences based on the 61 chemical composition of tissues, certain compounds found in specific taxa can interfere with analysis and, 62 therefore, conclusions. Here, we investigate the effects of urea and lipid extraction on tissues from 63 fourteen elasmobranch species and report results that demonstrate the necessity to account for these 64 compounds when using SIA in elasmobranch studies. 65 The physiology and anatomy of elasmobranchs present unique challenges when applying SIA to study their ecology. In particular, elasmobranchs retain urea ((NH2)2CO) and trimethylamine oxide 67 (TMAO (C3H9NO)) in their tissues for osmoregulatory processes (Ballantyne 1997, Olson 1999, Hazon et 68 al. 2003). This retention of urea can differentially bias stable isotope results depending upon the tissue 69 type examined (Hazon et al. 2008, Kim and Koch 2011, Hussey et al. 2012b, Churchill et al. 2015). As a 70 waste product, urea is expected to have low δ15N values (Minagawa and Wada 1984, Balter et al. 2006) 71 because 14N is preferentially concentrated in urea by deaminases and transaminases (Gannes et al. 1998). 72 We were unable to find any comparable data on TMAO, but as a waste product it also would be expected 73 to be depleted in 15N. As a result, the relative concentrations of urea and TMAO in a tissue may influence 74 the δ15N value of that tissue. As urea and TMAO (hereafter referred to together as urea) both contain 75 carbon, they could potentially affect δ13C. Kim and Koch (2011) reported that the carbon in urea is 76 enriched in δ13C in some terrestrial taxa; however information on the isotopic composition of these waste 77 products, especially in aquatic taxa, remains lacking. Further complicating the effect of urea on SIA is its 78 varying concentration within organisms, which is influenced by a variety of factors including tissue type 79 (Ballantyne 1997), ambient salinity (Hazon et al. 2003, Pillans et al. 2005) and diet (Wood et al. 2010). 80 Information on how to address the effects of urea on SIA results is needed, both in terms of appropriate t af Dr 66 3 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences 81 sample treatment methodology and data interpretation (Martinez del Rio et al. 2009, Logan and 82 Lutcavage 2010, Kim and Koch 2011, Hussey et al. 2012b, Li et al. 2016). 83 Page 4 of 36 In addition to the potential effect of urea on the stable isotope composition of elasmobranchs, the 84 presence of lipids is known to influence the δ13C values of tissues (Post et al. 2007, Martinez del Rio et al. 85 2009, Hussey et al. 2012a). Because lipids are depleted in 13C relative to protein, the presence of lipid in 86 tissues can bias δ13C values and increase the tissue carbonFtoFnitrogen ratio (C:N) (Pinnegar and Polunin 87 1999, Post et al. 2007). Tissue samples with high lipid concentrations have lower δ13C values than 88 samples of the same tissue with lipids removed (Post et al. 2007). To account for variation in lipids across 89 tissue types, researchers either chemically extract or mathematically correct for lipids based on the tissue 90 C:N, which has been used as a proxy for relative lipid content in tissues (Post et al. 2007). 91 The influence of lipid content on SIA data of elasmobranch tissues has been relatively well studied (Kim and Koch 2011, Hussey et al. 2012a) compared to that of urea (Hussey et al. 2012b). Logan 93 and Lutcavage (2010) and Kim and Koch (2011) directly assess the effects of urea extraction on SIA data 94 of elasmobranchs. Logan and Lutcavage (2010) reported no effect of urea extraction on elasmobranch 95 tissues, whereas Kim and Koch (2011) reported a significant increase in δ15N in ureaFextracted tissues. 96 However, treatment methods differed between studies, with Kim and Koch (2011) using a more extensive 97 deionized water (DIW) extraction, which potentially resulted in more complete urea removal. Given that 98 lipid has a high C:N and urea has low C:N (0.5), removal of these compounds will influence tissue C:N. 99 Several studies examining the effect of lipid extraction on elasmobranch tissue noted increases in δ15N t af Dr 92 100 and C:N following lipid extraction in a manner consistent with the removal of urea, suggesting that lipid 101 extraction may effectively remove urea as well as lipid (Hussey et al. 2010, Kim and Koch 2011, Hussey 102 et al. 2012a, Churchill et al. 2015, Li et al. 2016). However Kim and Koch (2011) reported that 103 elasmobranch tissues should have both urea and lipidFextracted to obtain the most reliable results. Li et al. 104 (2016) recently conducted the most thorough study of the interactive effects of urea and lipid toFdate, 105 examining the effects of urea and lipid extraction on six species of pelagic sharks. They reported 106 significant increases in δ15N and C:N following lipid extraction, urea extraction, and lipid and urea 4 https://mc06.manuscriptcentral.com/cjfas-pubs Page 5 of 36 Canadian Journal of Fisheries and Aquatic Sciences 107 extraction, with δ13C also increasing significantly in treatments with lipid extraction. Li et al. (2016) 108 supported the conclusion of Kim and Koch (2011) that both urea and lipid should be removed when 109 analyzing elasmobranch tissues for SIA. 110 Despite the reported shifts in C:N following urea and lipid extraction and recommendations to make lipid extraction a standard practice when processing elasmobranch tissues for SIA, estimating lipid 112 content based on the C:N of unextracted bulk samples remains a common practice. Specifically, it is 113 typically assumed that tissues of aquatic organisms with C:N values < 3.3 – 3.5 have low lipid content 114 and do not warrant lipid extraction (Post et al. 2007). This is of potential concern since this does not 115 account for the influence of urea on C:N, which is then used to mathematically adjust δ13C to account for 116 inferred (based on C:N) lipid content. As a result, the δ15N and potentially δ13C values of samples 117 processed without urea and/or lipid extraction may be biased, with any resulting analyses or ecological 118 interpretations being potentially based on inaccurate δ15N and δ13C values. Given the common use of δ15N 119 to estimate trophic level and δ15N and δ13C to understand habitat use and trophic relationships in 120 elasmobranchs (Fisk et al. 2002, Estrada et al. 2003, Dale et al. 2011, Kim and Koch 2011, Vaudo and 121 Heithaus 2011, Carlisle et al. 2012, Hussey et al. 2012b), these biases may have important effects on the 122 ecological interpretation of SIA data. t af 123 Dr 111 To examine how common this issue may be, we surveyed 50 recent scientific publications (2013F 124 present) that used SIA to study the ecology of elasmobranchs (Google scholar; search terms: “stable 125 isotope”, “elasmobranch”, “shark”, “ray”; selected the first 50 pertinent results, Table S1). We found 28% 126 used low C:N values (< 3.5) of tissues containing urea to support not extracting lipid from their samples, 127 another 16% used mathematic corrections to adjust δ13C based on lipid content estimates inferred from 128 C:N values of tissues containing urea, and 12% did not account for urea or lipid in any manner. Thus, 129 56% of the surveyed studies potentially had results biased due to not accounting for the combined effects 130 of urea and lipids. While it is not possible to know if the lack of urea or lipid extraction had any 131 meaningful effect on the isotopic results or their interpretation in these studies, it is clear previous 5 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences 132 recommendations to make urea and lipid extraction the standard practice when analyzing elasmobranch 133 tissues for SIA (Kim & Koch 2011, Hussey et al. 2012a, Li et al. 2016) have not been fully adopted. 134 Page 6 of 36 In this study, we expand upon previous work to better understand and account for the interactive effects of urea and lipids on SIA in elasmobranch tissues. The importance of urea and lipid extraction has 136 been demonstrated in leopard sharks (Triakis semifasciata, Kim and Koch (2011) and a suite of pelagic 137 sharks (Li et al. 2016), yet published studies commonly do not appropriately account for the potential 138 effects of urea and lipid on elasmobranch tissues. In addition, the effects of applying multiple chemical 139 treatments to remove urea and lipid, and how to determine an appropriate methodological course for 140 individual species or taxa of interest, remain unaddressed for many elasmobranch taxa across broad 141 ranges of tissue compositions and habitats. In particular, while the interactive effects of urea and lipid 142 have been explored in sharks, they have not been investigated in batoids (skates and rays), a group that 143 comprises over 50% of extant elasmobranchs (Dulvy et al. 2014). Finally, it may not always be feasible or 144 desirable to perform lipid extraction (i.e. avoid the cost, chemical waste generated, and time associated 145 with chemical extraction or to preserve information on the movement of lipids through foodwebs). While 146 Li et al. (2016) provides species specific isotopic correction models to account for urea content in lipid 147 extracted samples, there is a lack of specific guidance in the literature on the appropriate development and 148 application of mathematical correction models for δ13C based on inferred tissue lipid content (C:N) for 149 elasmobranchs that account for urea’s effects on C:N, δ13C and δ15N. t af 150 Dr 135 The goals of this study were to 1) assess the relative effects of urea and lipid extraction on the 151 stable isotope composition of muscle tissue from a variety of shark and batoid species, including pelagic, 152 demersal and benthic species, with variable lipid content, 2) address the utility of using C:N as a 153 diagnostic tool to understand and adjust for lipid content in elasmobranch tissue, particularly in the 154 context of the influence of urea on C:N, 3) develop models to mathematically adjust δ13C of ureaF 155 extracted samples to account for lipid content and 4) provide a conceptual framework to understand how 156 urea and lipid interact to influence SIA results in elasmobranchs in order to facilitate proper application of 157 the technique. 6 https://mc06.manuscriptcentral.com/cjfas-pubs Page 7 of 36 Canadian Journal of Fisheries and Aquatic Sciences 158 159 To examine the effects of urea extraction and lipid extraction on elasmobranch tissues, we collected white muscle samples from fourteen species, including six species of sharks from the families 161 Lamnidae, Carcharhinidae, Squalidae, and Triakidae, and eight species of batoids (skates, rays, and their 162 allies) from the families Arhynchobatidae, Rajidae, Myliobatidae, Gynmnuridae, Dasyatidae, and 163 Urolophidae. All samples were collected in the eastern North Pacific, primarily off California but ranging 164 as far north as the Gulf of Alaska. Samples were collected from juvenile salmon sharks (Lamna ditropis) 165 stranded on beaches in California and Oregon as described by Carlisle et al. (2015). Juvenile white sharks 166 (Carcharodon carcharias) caught as incidental bycatch in the coastal gillnet fisheries in southern 167 California were sampled as part of the Monterey Bay Aquarium juvenile white shark research program as 168 described in Mull et al. (2012). Samples from shortfin makos (Isurus oxyrinchus), blue sharks (Prionace 169 glauca), and pelagic rays (Pteroplatytrygon violacea) were collected during the annual National Oceanic 170 and Atmospheric Administration (NOAA) Juvenile Shark Longline Survey off southern California. 171 Round stingray samples (Urobatis halleri), leopard shark (Triakis semifasciata), butterfly ray (Gymnura 172 marmorata), and bat rays (Myliobatis californica) were collected in southern California as described by 173 Lyons et al. (2014). Skates (Bathyraja aleutica, Bathyraja interrupta, Beringraja binoculata, Raja rhina) 174 were collected from the western Gulf of Alaska as described in Bizzarro et al. (2014). Spotted spiny 175 dogfish (Squalus suckleyi) were collected off central California during the National Marine Fisheries 176 Service Northwest Fisheries Science Center West Coast Groundfish Bottom Trawl Survey. All muscle 177 samples were collected from the dorsal musculature and stored frozen (F20°C) until processed and 178 analyzed. t af 179 Dr 160 Individual tissue samples were homogenized and subdivided into three parts for analysis, with 180 one as the control sample (Control), one for urea extraction (U), and one for both urea and lipid extraction 181 (UL). Methods to process tissues and extract lipids and urea generally followed Kim and Koch (2011). To 182 extract urea, tissue samples were placed in scintillation vials and rinsed three times in 10 mL of DIW 183 (Kim and Koch 2011). A rinse entailed sonication for 15 minutes and then decanting the supernatant. 7 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences 184 Lipids were extracted from all tissues except skate tissues using a 2:1 chloroform:methanol solution 185 (Bligh and Dyer 1959, Logan and Lutcavage 2010) by immersing tissues in the solution for 24 hours in 186 glass scintillation vials (Bligh and Dyer 1959, Logan and Lutcavage 2010). Following both treatments, 187 tissue samples were lyophilized and homogenized using a Spex/CertiPrep 5100 mill. Page 8 of 36 Skate samples were processed slightly differently than the other species as part of another study. 189 For urea extraction of skate tissue, 10 mL of DIW were added to each homogenized sample, and then the 190 samples were mixed using a vortex mixer (Fisher Scientific). After 30 minutes, the sample was 191 centrifuged and the supernatant was decanted. For samples that were lipid and ureaFextracted, lipids were 192 extracted using Petroleum Ether (PE) following Kim and Koch (2011). Briefly, samples were immersed 193 in PE, mixed in a vortex mixer and left uncapped in a fume hood for 8 hours, centrifuged for 10 minutes, 194 and the supernatant decanted. The sample was then rinsed in DIW using the method described for urea 195 extraction of skate tissue. Following urea or urea and lipid extraction samples were dried in an oven at 196 60˚C for 24 hours. af 197 Dr 188 For each treatment, approximately 500 Qg of tissue was weighed into tin boats and analyzed at the Stable Isotope Laboratory at the University of California Santa Cruz (UCSC) using an elemental 199 analyzer coupled to an isotope ratio monitoring mass spectrometer (Delta XPFEA,ThermoF Finnagen 200 IRMS). For skate and dogfish samples, 500 Qg of tissue was weighed into tin boats and analyzed at Idaho 201 State University (ISU) using an elemental analyzer coupled to an isotope ratio monitoring mass 202 spectrometer (Elemental Combustion System (ECS) 4010 interfaced with a Delta V Advantage mass 203 spectrometer through the ConFlo IV System). Isotopic composition is expressed using standard δ 204 notation, using Vienna Pee Dee Belemnite limestone as the standard for carbon and AIR for nitrogen. For 205 runs at UCSC, analytical precision, based on an internal lab standard (Pugel), was 0.11‰ for δ15N and 206 0.07‰ for δ13C across multiple runs. For runs at ISU, analytical precision, based on internal lab standards 207 of ISU Peptone, Costech Acetanilide, and DORMF3, was 0.08, 0.04, and 0.04 ‰ for δ15N respectively and 208 0.05, 0.05, and 0.04 for δ13C, respectively. Where parametric assumptions were met (assessed with OneF 209 Sample KolmogorovFSmirnov and Levene’s Tests and visual inspection of residuals) a single factor t 198 8 https://mc06.manuscriptcentral.com/cjfas-pubs Page 9 of 36 Canadian Journal of Fisheries and Aquatic Sciences 210 ANOVA, followed by Tukey’s post hoc tests, was used to test for differences in δ13C and δ15N among 211 treatments for species with sample sizes > 3. When assumptions were not met for δ13C, δ15N, or C:N 212 differences were tested using MannFWhitney 2Fsample tests with sequential Bonferroni adjustments (Rice 213 1989). To show the magnitude and direction of the effects of treatments U and UL, differences between 214 treatment and control samples (U – Control, UL – Control, UL F U) were calculated. 215 We considered four previously used lipid correction models (Post et al. 2007, Logan et al. 2008, Reum 2011) to examine the utility of using C:N as a diagnostic tool to understand and adjust for lipid 217 content in ureaFextracted elasmobranch tissue. Lipid correction models were used to characterize U13C, 218 the difference between lipid and urea (UL) and ureaFextracted (U) δ13C values (U13C = δ13CU F δ13CUL) as 219 a function of the C:N of ureaFextracted tissue (C:NU). The first (model 1) is a threeFparameter model 220 derived by Logan et al. (2008) from McConnaughey and McRoy (1979): U13C = (aC:NU + b)( C:NU + c)F1 221 , where a, the yFasymptote, corresponds to proteinFlipid δ13C discrimination and –baF1, the xFintercept, is 222 the urea and lipid free C:N value (C:NUL), and bcF1, the y'intercept, is the value of U13C at C:NU = 0. The 223 second (model 2) is a two parameter model (Fry 2002): U13C = P F PF(C:NU)F1, where P represents 224 proteinFlipid δ13C discrimination and F is C:NUL. The third and fourth are linear models: (model 3, Logan 225 et al. 2008) U13C = β0 + β1Ln(C:NU) and (model 4, Post et al. 2007) U13C = b + aC:NU, where t 227 af 226 Dr 216 ( ) and –baF1 are estimates of C:NUL, respectfully. We modeled the relationship between C:NU and U13C for five groups: all species, batoids, all 228 sharks other than S. suckleyi, S. suckleyi, and C. carcharias. We modeled S. suckleyi independently since 229 its lipid content was higher than all other taxa and its ureaFextracted samples had the widest range of C:N 230 values (Results). For C. carcharias, we wanted to attempt to develop a speciesFspecific relationship and 231 this species had the largest sample size. To compare the performance of potential lipid correction models, 232 the corrected Akaike Information Criterion, AICc (Burnham and Anderson 2002), was calculated for each 233 model. The model with the lowest AICc is considered the best fit, but any model(s) with AICc values 234 within two units of the lowest value have strong support as well (Burnham and Anderson 2002). In 9 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences 235 addition for those models that preformed best based on AICc, we calculated the mean and standard 236 deviation of the absolute values of the residuals errors, to further evaluate model fit, and compared 237 estimates of proteinFlipid δ13C discrimination and C:NUL (Logan et al. 2008, Reum 2011). All models 238 were fitted with leastFsquares procedures using R and the libraries nlme and AICcmodavg (www.rF 239 project.org). Page 10 of 36 240 241 The removal of urea and lipid from elasmobranch tissue influenced δ15N, δ13C, and C:N in most species, although the direction and magnitude of the effects varied by species and with the lipid content of 243 the sample (Table 1, Figures 1F2, Figure S1). The C:N of control samples was consistently low for all 244 species, with 13 of the 14 taxa having values < 3.5 (mean ± SD of all species 3.0 ± 0.4). The only 245 exception to this was S. suckleyi, which had a high C:N of 4.5 due to higher lipid content of its muscle. 246 Urea extraction (treatment U) generally increased δ15N and C:N, but generally did not Dr 242 significantly change δ13C, results that are consistent with the removal of isotopically light nitrogen present 248 in urea (Table 1, Figures 2a and 3a). In seven of the ten species (4 of 5 sharks, 3 of 5 batoids) that were 249 statistically tested, δ15N increased significantly following urea extraction (mean 0.8‰ ± 0.2). This result 250 was very similar to the overall trend across all taxa, which showed an average increase of 0.7‰ ± 0.2. 251 Three taxa had increases in δ15N greater than 1‰ (B. aleutica 1.1‰, L. ditropis 1.1‰ and I. oxyrinchus 252 1.0 ‰). δ13C only changed significantly in B. binoculata (F0.9‰) and U. halleri (F0.5‰). Overall, there 253 was a consistent, though generally nonFsignificant, decrease in δ13C across the batoids that was not 254 evident in sharks (mean F0.4‰ for batoids, 0.0‰ for sharks, and F0.2‰ ± 0.3 for all taxa). C:N increased 255 significantly in nine of the ten taxa statistically tested (4 of 5 sharks, 5 of 5 batoids), with the exception 256 being S. suckleyi, which had a high initial C:N that increased from 4.5 to 5.6 (+ 1.1‰ ) following urea 257 extraction. Overall, C:N increased by an average of 0.7 (± 0.2) across all taxa following urea extraction. 258 C:N values increased to values above 3 (mean C:N of 3.6) in all species, and the C:N of C. carcharias, L. 259 ditropis, G. marmorata, and U. halleri increased to values above 3.5. t af 247 10 https://mc06.manuscriptcentral.com/cjfas-pubs Page 11 of 36 260 Canadian Journal of Fisheries and Aquatic Sciences Extracting urea and lipid (treatment UL) from samples consistently increased δ15N and C:N in a fashion similar to what was observed with urea extraction only, while also generally increasing δ13C 262 (Table 1, Figures 2b and 3b). Seven out of the ten taxa tested had significantly higher δ15N values 263 following urea and lipid extraction, although all taxa showed some increase (mean 0.8‰ ± 0.2). L. 264 ditropis showed the largest increase in δ15N following lipid extraction (1.1‰). Four of the five 265 statistically tested sharks had significantly higher δ13C following urea and lipid extraction, with dogfish 266 (mean 2.0‰), salmon sharks (1.5‰) and white sharks (1.1‰) having the largest increases. Three batoids 267 showed a decrease in δ13C following urea and lipid extraction, with B. binoculata exhibiting a significant 268 decrease (mean F0.9‰). Except for S. suckleyi, all taxa exhibited an increase in C:N (mean increase 0.3 ± 269 0.3, mean of species C:N 3.3), with nine of ten taxa tested statistically having significant changes. C:N of 270 S. suckleyi decreased, though nonFsignificantly, following urea and lipid extraction (control C:N 4:5, UL 271 C:N 3.9). 272 Dr 261 The differences between the U and UL treatments were more obvious in sharks than in the af batoids, in which the differences were relatively small (Table 1). All taxa showed an increase in δ13C in 274 the UL treatment relative to the U treatment, although only three of the five sharks (C. carcharias, L. 275 ditropis and S. suckleyi), and one of five batoids (U. halleri), had significant increases. There were no 276 significant differences in δ15N between U and UL treatments in any species examined (p > 0.05), 277 indicating that lipid extraction did not affect δ15N. C:N was generally lower in the UL treatment relative 278 to the U treatment, especially in the sharks. In the five shark species tested, all had significant decreases 279 in C:N in UL treatments relative to U treatments, whereas only two of the five tested batoids had 280 significant decreases. 281 t 273 Lipid correction models were created to adjust ureaFextracted tissue δ13C to account for lipid 282 content. Model performance varied across elasmobranch groups (Figure 3, Figure S2), with no single 283 model amongst those exhibiting the lowest AICc values across all groups (see supplementary Table S2 for 284 AICc Values, r2 (linear models only) and model parameters). For all species pooled, models 1 and 2 (nonF 285 linear models) had the lowest AICc, with identical mean ± SD of the absolute values of the residual errors 11 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 12 of 36 (MRE, 0.29 ± 0.24) and similar estimates of proteinFlipid δ13C discrimination (5.39 and 6.12) and C:NUL 287 (3.20 and 3.18). All four models for sharks (excluding S. suckleyi) had similarly low AICc values (i.e. 288 within 2 units), MREs (0.20 – 0.25 ± 0.25 – 0.29) and estimates of C:NUL (3.10 – 3.13). However, 289 estimates of proteinFlipid δ13C discrimination for sharks varied widely between models 1 (1.94) and 2 290 (7.86, Table S2). The two linear models (3 and 4) performed equally for batoids (Table S2), with identical 291 MRE (0.21 ± 0.16) and estimates of C:NUL (3.27). Model performance varied between the two single 292 species groups. For S. suckleyi, which had both the widest range and highest C:N values, model 2 293 provided the singular best fit (MRE 0.26 ± 0.24) with estimates of proteinFlipid δ13C discrimination and 294 C:NUL of 6.31 and 3.35, respectively (Table S2) . In contrast, for C. carcharias, which had only two ureaF 295 extracted C:N values > 4, models 2, 3 and 4 had similarly low AICc, MREs (0.12 – 0.13 ± 0.12 – 0.13) 296 and estimates of C:NUL (3.05 – 3.14). The estimate of urea extracted proteinFlipid δ13C discrimination for 297 C. carcharias was the highest in the study (8.30, Table S2). 298 af Dr 286 Our results provide further evidence of the substantial, direct affect that urea can have on δ15N. 300 We also show the important, and often unconsidered, indirect role urea plays in influencing δ13C values 301 by lowering the C:N, effectively masking lipid content and leading to inaccurate assessments of lipid 302 content and inappropriate mathematical corrections (Figure 4). Our results indicate that urea must be 303 removed to obtain reliable δ15N and C:N values, and that only with ureaFextracted tissues can C:N be used 304 as a diagnostic tool for understanding and mathematically adjusting for lipid content. 305 t 299 Urea extraction resulted in an increase in δ15N across all taxa, and a significant increase in 7 of 10 306 statistically tested taxa, ranging from ~0.5 to 1.1‰ (mean 0.7 or 0.8‰ for U and UL respectively), with 307 urea and lipid extraction producing similar changes. When using δ15N to infer trophic level, this shift is 308 equivalent to an inferred trophic level difference of ~22 – 50% or ~15 – 30% assuming a trophic 309 discrimination factor for nitrogen of 2.3‰ (Hussey et al. 2010) or 3.7‰ (Kim et al. 2012), respectively. 310 This shift is similar to that reported by Hussey et al. (2012a) for elasmobranch tissues following lipid 311 extraction as well as by Li et al. (2016) following both urea extraction and urea and lipid extraction. 12 https://mc06.manuscriptcentral.com/cjfas-pubs Page 13 of 36 Canadian Journal of Fisheries and Aquatic Sciences 312 Clearly the presence of urea in analyzed tissues will directly bias the use of δ15N as a tracer in ecological 313 studies, whether it is being used to assess trophic level, reconstruct diet, habitat or migration patterns, or 314 even for simple qualitative comparisons. 315 The urea effects we describe here also have important implications for the use of C:N as a metric of lipid content of elasmobranch tissue. The C:N of nonFextracted elasmobranch tissues are consistently 317 very low (< 3) across studies (Logan and Lutcavage 2010, Matich et al. 2010, Dale et al. 2011, Kim and 318 Koch 2011, Vaudo and Heithaus 2011, Hussey et al. 2012a, Hussey et al. 2012b), and are often much 319 lower than would be expected of pure protein. Frequently, when tissues with C:N values < 3.5 are 320 assumed to have little lipid content (Post et al. 2007), the low C:N values are used to infer that lipid 321 extraction is not warranted. However, results from this study and previous work (Kim and Koch 2011, 322 Hussey et al. 2012a, Li et al. 2016) demonstrate that urea extraction generally increases the C:N as 323 nitrogen is removed (Figures 1 and 2). In this study, extracting urea through DIW rinses caused 324 significant increases in C:N in every species tested, increasing it by as much as 1.2‰ (mean 0.7‰). af 325 Dr 316 Removal of the nitrogen contributed by urea will increase the C:N value of a sample, thereby changing the estimated lipid content that are based on C:N (Figure 4). Following removal of urea, C:N 327 can increase from very low values to values above threshold levels that are used to indicate low lipid 328 content (e.g. 3.5). In effect, the presence of urea and its lowering of C:N has the potential to mask lipid 329 content. In four of the species examined in this study (C. carcharias, L. ditropis, G. marmorata and U. 330 halleri) the C:N value shifted from values < 3 to values > 3.5, and other species had C:N values of ~3.3 331 following DIW rinses, which is similar to pure protein values (Post et al. 2007). The interpretation of 332 these ureaFextracted samples would then be that lipid extraction is warranted, and in two of these species 333 (C. carcharias, L. ditropis), there was a significant increase in δ13C following lipid extraction. 334 Importantly, despite having C:N values ~3.3 following urea extraction, most of the species exhibited an 335 increase in δ13C following lipid extraction. These findings suggest that lipid extraction can significantly 336 affect δ13C even when C:N is < 3.5. Hence, lipid extraction may be required even in a tissue that is 337 relatively lean, a result that is concordant with the findings of Li et al. (2016). Our results indicate that t 326 13 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 14 of 36 338 failure to lipid extract elasmobranch tissues based on low C:N values of untreated tissue, where urea has 339 not been removed, will result in biased δ15N values, due to the inclusion of the isotopically light nitrogen 340 of urea, and potentially biased δ13C values as well, due to inclusion of lipid content that was masked by 341 low C:N values. 342 An additional important, and generally unrecognized implication of the effect of urea on C:N is that it will bias C:N based arithmetic corrections that are used to adjust δ13C in lieu of lipid extracting 344 tissues. Any adjustment to δ13C values that is based on C:N values from nonFureaFextracted tissue will be 345 biased, although the magnitude of the effect will vary based on the urea and lipid content of the tissue. 346 Thus, failing to extract urea will not only bias δ15N, but by affecting C:N it will lead to incorrect estimates 347 of inferred lipid content upon which mathematical correction models rely. As our results indicate, 348 however, it is possible to develop models to adjust ureaFextracted tissues to account for lipid content 349 when lipid extraction is not feasible. For example, the model we developed for C. carcharias 350 demonstrates that even with a relatively small sample size (n = 19), we were able to generate a robust 351 speciesFspecific simple linear model based on ureaFextracted tissue C:N (model 4, r2 = 0.92, with no 352 systematic prediction biases based on visual inspection of residuals). Deriving taxaFspecific relationships 353 is always desirable, but for sharks (excluding dogfish) all four models examined seem to provide 354 potentially suitable lipid correction models. However, examination of model parameters reveals that for 355 model 1 the estimation of proteinFlipid δ13C discrimination is unrealistically low, ~3 times, or more, lower 356 than other estimates from this study and the generally reported range of 5–8‰ (Fry 2002, Post et al. 2007, 357 Logan et al. 2008, Reum 2011), demonstrating the need to consider other factors beyond AICc and fit (r2 358 and residual distribution) when determining the suitability of a correction model. Sharks (excluding 359 dogfish) have C:N of ureaFextracted tissues below ~4.5 and there appears to be a linear relationship 360 between C:N and U13C. This suggests that linear correction models might be most appropriate for sharks 361 with relatively low C:N values (~4.5), which is concordant with the findings of Post et al. (2007) for 362 aquatic organisms over a similar C:N range. This relationship is likely nonFlinear when tissues span a 363 wide range of C:N values, such as with S. suckleyi (Logan et al. 2008, Reum 2011). It is therefore t af Dr 343 14 https://mc06.manuscriptcentral.com/cjfas-pubs Page 15 of 36 Canadian Journal of Fisheries and Aquatic Sciences 364 important to consider the range in lipid content in species and tissues of interest when developing and 365 applying lipid correction models. Batoids had a less clear C:N relationship relative to other taxa in the 366 study. While exhibiting a significant linear relationship between C:Nu and U13C, the models explained a 367 relatively low proportion of the variability in U13C (models 3 and 4, r2 = 0.56 and 0.58) and may not 368 provide the same relative correction across all batoids (Figure 3). This emphasizes that the interaction 369 between urea and lipid content may change across disparate elasmobranch taxonomic groups. 370 Although our results indicate that urea directly influences δ15N values and potentially indirectly influences the δ13C of elasmobranch tissues by affecting the C:N and inferred lipid content, the effects are 372 variable and speciesFdependent (Figures 1, 2). For species that have low lipid content in their muscle, 373 such as batoids (e.g. ~0.2% in U. halleri, (Lyons unpublished data), ~1F2% in B. binoculata and R. rhina 374 (Farrugia et al. 2015)), the effect on δ13C will be minimal, but δ15N may change substantially. In species 375 with higher lipid content, such as L. ditropis, which can have lipid content as high as 6.5 to 14.6% (mean 376 9%) in their muscle (data source: https://dec.alaska.gov/eh/vet/fish.htm), the effect on δ13C and δ15N will 377 be significant. The relatively low lipid content of batoid muscle compared to shark muscle may underlie 378 the observed general decrease in δ13C of batoids following urea extraction while urea and lipid extraction 379 showed less overall change in δ13C relative to the control samples. As described by Kim and Koch (2011), 380 the carbon in urea is enriched in 13C relative to the diet in humans (enriched 3 – 5‰) and cattle (0 – 381 3.5‰), suggesting that its removal may reduce the δ13C (Ivlev et al. 1996, Knobbe et al. 2006). In batoids 382 with low lipid content, removal of the 13C enriched urea would reduce δ13C, while lipid extraction, by 383 removing a small amount of 13C depleted lipids, would offset the removal of urea and result in little net 384 change in δ13C. This effect would vary based on both the concentration of urea as well as lipid content of 385 the tissue and again highlights the importance of understanding how these relationships change across 386 taxa. It is possible that other differences in the composition of batoid tissue may play a role as well, such 387 as differences in urea concentration or the presence of ceratotrichia, but this remains unclear. t 389 af 388 Dr 371 Although lipid extraction by itself may remove lipids and much of the urea present in tissues (Hussey et al. 2012a, Churchill et al. 2015), we reiterate the recommendations of Kim and Koch (2011) 15 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 16 of 36 and Li et al. (2016) to make lipid and urea extraction the standard practice when analyzing elasmobranch 391 tissues for SIA. Since extracting urea is simple and inexpensive, there is no practical reason not to remove 392 it. In addition, our results indicate that δ13C may change significantly following lipid extraction even in 393 apparently lean tissues with relatively low C:N, suggesting that lipid extraction may be warranted in all 394 situations as suggested by Li et al. (2016). However, in some taxa that are very lean, such as the batoids in 395 this study, lipid extraction may not be required. However, urea would still need to be extracted to evaluate 396 the need for lipid extraction or correction. In instances where it is not feasible or desirable to lipid extract 397 every sample, we demonstrate that it is possible to develop speciesF or groupFspecific correction curves to 398 adjust for lipid content in ureaFextracted tissues. Though lipid extraction did not affect δ15N in our study, 399 a potential benefit of using mathematical correction models is the ability to account for the effect of lipid 400 content on δ13C while avoiding potential effects of chemical extraction on δ15N, which have been reported 401 previously in other taxa (Post et al. 2007). However, our results show that the confounding effects of urea 402 and lipids make it impossible to use C:N of nonFureaFextracted samples as a diagnostic tool to determine 403 the proper method of tissue treatment, something that occurs regularly in the literature. af The changes in the stable isotope composition of elasmobranch tissue resulting from urea and t 404 Dr 390 405 lipid extraction will be mediated by the relative concentration of those compounds (Figure 4), which vary 406 across taxa, and their differential effects on δ13C and δ15N (Figures 1 & 2). We conclude that at a 407 minimum, urea should be removed to evaluate lipid content and a species or group specific lipid 408 correction relationship created to account for lipid content. The most robust approach to most confidently 409 eliminate bias and to facilitate comparisons across studies will be to apply urea and lipid removal 410 techniques in SIAFbased ecological studies of elasmobranchs. 411 412 We would like to thank Owyn Snodgrass, James Wraith, and Heidi Dewar (NOAA) for assistance 413 in obtaining samples from the NOAA Juvenile Shark Longline Survey, and John O’Sullivan, Christopher 414 Lowe, Salvador Jorgensen and the Monterey Bay Aquarium for help collecting juvenile white shark 415 samples. Funding was provided by PADI Foundation and the Dr. Earl Myers and Ethel Myers 16 https://mc06.manuscriptcentral.com/cjfas-pubs Page 17 of 36 Canadian Journal of Fisheries and Aquatic Sciences 416 Oceanographic and Marine Biology Trust. For funding for materials and processing dogfish samples at 417 ISU, we would like to thank Bruce Finney. Funding and materials for skate processing were provided by 418 the Pacific Shark Research Center at Moss Landing Marine Laboratories and the North Pacific Research 419 Board. Support for M. Ibarra was provided by Mount Holyoke College. 420 Ballantyne, J.S. 1997. Jaws: the inside story. The metabolism of elasmobranch fishes. Comp Biochem Physiol B Biochem Mol Biol !(4): 703F742. Balter, V., Simon, L., Fouillet, H., and Lécuyer, C. 2006. BoxFmodeling of 15N/14N in mammals. Oecologia "#(2): 212F222. Bizzarro, J.J., Broms, K.M., Logsdon, M.G., Ebert, D.A., Yoklavich, M.M., Kuhnz, L.A., and Summers, A.P. 2014. Spatial segregation in eastern north Pacific skate assemblages. Bligh, E.G., and Dyer, W.J. 1959. A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology $#: 911F917. Burnham, K.P., and Anderson, D.R. 2002. Model selection and multimodel inference: a practical informationFtheoretic approach. Springer Science & Business Media. Carlisle, A.B., Kim, S.L., Semmens, B.X., Madigan, D.J., Jorgensen, S.J., Perle, C.R., Anderson, S.D., Chapple, T.K., Kanive, P.E., and Block, B.A. 2012. Using stable isotope analysis to understand the migration and trophic ecology of northeastern Pacific white sharks (Carcharodon carcharias). PLoS ONE #(2): e30492. Carlisle, A.B., Litvin, S.Y., Hazen, E.L., Madigan, D.J., Goldman, K.J., Lea, R.N., and Block, B.A. 2015. Reconstructing habitat use by juvenile salmon sharks links upwelling to strandings in the California Current. Mar Ecol Prog Ser %&%: 217F228. Churchill, D.A., Heithaus, M.R., and Grubbs, R.D. 2015. Effects of lipid and urea extraction on δ15N values of deepFsea sharks and hagfish: Can mathematical correction factors be generated? Deep Sea Research Part II: Topical Studies in Oceanography %: 103F108. Dale, J.J., Wallsgrove, N.J., Popp, B.N., and Holland, K.N. 2011. Nursery habitat use and foraging ecology of the brown stingray Dasyatis lata determined from stomach contents, bulk and amino acid stable isotopes. Mar Ecol Prog Ser "$$: 221F236. Dulvy, N.K., Fowler, S.L., Musick, J.A., Cavanagh, R.D., Kyne, P.M., Harrison, L.R., Carlson, J.K., Davidson, L.N.K., Fordham, S.V., Francis, M.P., Pollock, C.M., Simpfendorfer, C.A., Burgess, G.H., Carpenter, K.E., Compagno, L.J.V., Ebert, D.A., Gibson, C., Heupel, M.R., Livingstone, S.R., Sanciangco, J.C., Stevens, J.D., Valenti, S., and White, W.T. 2014. Extinction risk and conservation of the world’s sharks and rays. eLIFE $: e00590. Estrada, J.A., Rice, A.N., Lutcavage, M.E., and Skomall, G.B. 2003. Predicting trophic position in sharks of the northFwest Atlantic Ocean using stable isotope analysis. J Mar Biol Assoc U K !$(6): 1347F1350. Farrugia, T.J., Oliveira, A.C., Knue, J.F., and Seitz, A.C. 2015. Nutritional content, mercury, and trace element analyses of two skate (Rajidae) species in the Gulf of Alaska. Journal of Food Composition and Analysis "&: 152F163. Fisk, A.T., Tittlemier, S.A., Pranschke, J.L., and Norstrom, R.J. 2002. Using anthropogenic contaminants and stable isotopes to assess the feeding ecology of Greenland sharks. Ecology !$(8): 2162F2172. Fry, B. 2002. Stable isotopic indicators of habitat use by Mississippi River fish. J North Am Benthol Soc & (4): 676F685. Gannes, L.Z., Martinez Del Rio, C., and Koch, P. 1998. Natural abundance variations in stable isotopes and their potential uses in animal physiological ecology. Comparative Biochemistry and Physiology ' : 725F737. t af Dr 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 17 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Hazon, N., Wells, A., Anderson, G., Good, J., Pillans, R., and Franklin, C. 2008. Osmoregulation in elasmobranch fishFA review. Comparative Biochemistry and Physiology AFMolecular, pp. S87F S87. Hazon, N., Wells, A., Pillans, R.D., Good, J.P., Anderson, W.G., and Franklin, C.E. 2003. Urea based osmoregulation and endocrine control in elasmobranch fish with special reference to euryhalinity. Comparative Biochemistry and Physiology BFBiochemistry & Molecular Biology $((4): 685F 700. Hussey, H.E., Olin, J.A., Kinney, M.J., McMeans, B.C., and Fisk, A.T. 2012a. Lipid extraction effects on stable isotope values (δ13C and δ15N) of elasmobranch muscle tissue. J Exp Mar Biol Ecol "$") "$%: 7F15. Hussey, N.E., Brush, J., McCarthy, I.D., and Fisk, A.T. 2010. δ15N and δ13C dietFtissue discrimination factors for large sharks under semiFcontrolled conditions. Comp. Biochem. Physiol. AFMol. Integr. Physiol. %%(4): 445F453. Hussey, N.E., MacNeil, M.A., Olin, J.A., McMeans, B.C., Kinney, M.J., Chapman, D.D., and Fisk, A.T. 2012b. Stable isotopes and elasmobranchs: tissue types, methods, applications and assumptions. J Fish Biol: 1449F1484. Ivlev, A., Knyazev, Y.A., and Logachev, M. 1996. Daily average carbon isotope composition of CO2 of expired air and urine urea in norm and some endocrine pathologies in man. Biofizika " (2): 508F 516. Kim, S.L., Casper, D.R., GalvánFMagaña, F., OchoaFDíaz, R., HernándezFAguilar, S.B., and Koch, P.L. 2012. Carbon and nitrogen discrimination factors for elasmobranch soft tissues based on a longF term controlled feeding study. Environ Biol Fishes '%(1): 37F52. Kim, S.L., and Koch, P.L. 2011. Methods to collect, preserve, and prepare elasmobranch tissues for stable isotope analysis. Environ Biol Fishes '%(1): 53F63. Knobbe, N., Vogl, J., Pritzkow, W., Panne, U., Fry, H., Lochotzke, H., and PreissFWeigert, A. 2006. C and N stable isotope variation in urine and milk of cattle depending on the diet. Analytical and bioanalytical chemistry $!((1): 104F108. Li, Y., Zhang, Y., Hussey, N.E., and Dai, X. 2016. Urea and lipid extraction treatment effects on δ15N and δ13C values in pelagic sharks. Rapid Communications in Mass Spectrometry $*(1): 1F8. Logan, J.M., Jardine, T.D., Miller, T.J., Bunn, S.E., Cunjak, R.A., and Lutcavage, M.E. 2008. Lipid corrections in carbon and nitrogen stable isotope analyses: comparison of chemical extraction and modelling methods. J Anim Ecol ##(4): 838F846. Logan, J.M., and Lutcavage, M. 2010. Stable isotope dynamics in elasmobranch fishes. Hydrobiologia ("": 231F244. Lyons, K., Lavado, R., Schlenk, D., and Lowe, C. 2014. Bioaccumulation of organochlorine contaminants and EROD activity in southern California round stingrays (Urobatis halleri) exposed to planar aromatic compounds. Environ Toxicol Chem $$(6): 1380F1390. Martinez del Rio, C., Wolf, N., Carleton, S.A., and Gannes, L.Z. 2009. Isotopic ecology ten years after a call for more laboratory experiments. Biol Rev !": 91F111. Matich, P., Heithaus, M.R., and Layman, C.A. 2010. SizeFbased variation in intertissue comparisons of stable carbon and nitrogen isotopic signatures of bull sharks (Carcharhinus leucas) and tiger sharks (Galeocerdo cuvier). Can J Fish Aquat Sci (#: 877F885. McConnaughey, T., and McRoy, C. 1979. FoodFweb structure and the fractionation of carbon isotopes in the Bering Sea. Mar Biol %$(3): 257F262. Michener, R.H., and Kaufman, L. 2007. Stable isotope ratios as tracers in marine food webs: an update. In Stable isotopes in ecology and environmental science. Edited by R. Michener and K. Lajtha. Blackwell, Boston. pp. 238 F 282. Minagawa, M., and Wada, E. 1984. Stepwise enrichment of 15 N along food chains: further evidence and the relation between δ 15 N and animal age. Geochim Cosmochim Acta "!(5): 1135F1140. Mull, C.G., Blasius, M.E., O’Sullivan, J.B., and Lowe, C.G. 2012. Heavy metals, trace elements, and organochlorine contaminants in muscle and liver tissue of juvenile White Sharks, Carcharodon t af Dr 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 Page 18 of 36 18 https://mc06.manuscriptcentral.com/cjfas-pubs Page 19 of 36 carcharias, from the Southern California Bight. Global perspectives on the biology and life history of the White Shark. CRC Press, Boca Raton, Florida: 59F75. Peterson, B.J., and Fry, B. 1987. Stable isotopes in ecosystem studies. Annu Rev Ecol Syst !: 293F320. Pillans, R.D., Good, J.P., Anderson, W.G., Hazon, N., and Franklin, C.E. 2005. Freshwater to seawater acclimation of juvenile bull sharks (Carcharhinus leucas): plasma osmolytes and Na+/K+F ATPase activity in gill, rectal gland, kidney and intestine. Journal of Comparative Physiology B #%(1): 37F44. Pinnegar, P.K., and Polunin, N.V.C. 1999. Differential fractionation of δ13C and δ15N among fish tissues: implications for the study of trophic interactions. Funct Ecol $: 225F231. Post, D.M., Layman, C.A., Arrington, D.A., Takimoto, G., Quattrochi, J., and Montana, C.G. 2007. Getting to the fat of the matter: models, methods and assumptions for dealing with lipids in stable isotope analyses. Oecologia %&(1): 179F189. Reum, J.C.P. 2011. Lipid correction model of carbon stable isotopes for a cosmopolitan predator, spiny dogfish Squalus acanthias. J Fish Biol #'(7): 2060F2066. Rice, W.R. 1989. Analyzing tables of statistical tests. Evolution "$: 223F225. Vaudo, J.J., and Heithaus, M.R. 2011. Dietary niche overlap in a nearshore elasmobranch mesopredator community. Mar Ecol Prog Ser "&%: 247F260. Wood, C.M., Walsh, P.J., Kajimura, M., McClelland, G.B., and Chew, S.F. 2010. The influence of feeding and fasting on plasma metabolites in the dogfish shark (Squalus acanthias). Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology %%(4): 435F444. t af Dr 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 Canadian Journal of Fisheries and Aquatic Sciences 19 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences 560 561 + ,- , Species Carcharodon carcharias ID CC Isurus oxyrinchus IO Lamna ditropis LD Prionace glauca PG Squalus suckleyi SS Triakis semifasciata TS . +/ Treatment C U UL C U UL C U UL C U UL C U UL C U UL δ13C mean F16.8 F16.6 F15.6 F18.0 F17.8 F17.6 F19.4 F19.0 F17.9 F18.0 F18.1 F17.8 F19.4 F19.7 F17.4 F16.0 F16.1 F15.7 sd 0.9 0.8 0.5 0.2 0.7 0.2 0.4 0.4 0.6 0.2 0.3 0.2 1.1 1.0 0.2 0.3 0.5 0.4 Treatment C U UL C U UL C U UL C U UL C U UL C U UL C U UL C U UL δ13C mean F16.7 F16.9 F16.7 F16.3 F16.6 F16.5 F17.0 F17.2 F16.1 F16.4 F16.5 F16.2 F18.6 F18.8 F18.5 F15.2 F16.2 F16.1 F16.2 F16.8 F16.7 F15.0 F15.4 F14.9 sd 0.5 0.4 0.4 0.4 0.3 0.4 1.1 1.5 0.9 1.2 1.2 1.2 0.5 0.5 0.4 0.5 0.4 0.4 1.0 0.6 0.6 0.3 0.3 0.3 C test F ns * F ns ns F ns * F ns * F ns * U test F F * F F ns F F * F F ns F F * δ15N mean 16.8 17.3 17.3 16.4 17.4 17.4 14.4 15.4 15.5 16.1 16.9 17.0 14.0 14.5 14.9 15.3 16.0 16.0 sd 0.6 0.7 0.7 0.2 0.3 0.3 0.3 0.4 0.4 0.6 0.5 0.6 0.7 0.6 0.6 0.2 0.2 0.2 δ15N mean 15.0 16.1 15.9 15.7 16.2 16.1 16.6 17.0 17.2 16.2 16.9 17.0 13.5 13.9 14.1 15.0 15.5 15.5 15.7 16.5 16.4 15.7 16.5 16.5 sd 0.5 0.5 0.5 0.4 0.5 0.2 0.9 0.9 0.7 0.4 0.1 0.0 0.8 1.2 0.9 0.1 0.3 0.3 0.7 0.8 0.7 0.3 0.3 0.4 C test F * * F * * F * * F * * F ns * U test F F ns F F ns F F ns F F ns F F ns C:N mean 2.8 3.6 3.2 2.8 3.3 3.2 2.9 3.7 3.2 2.6 3.2 3.1 4.5 5.6 3.9 2.9 3.3 3.1 sd 0.2 0.3 0.0 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.0 0.0 1.1 1.2 0.1 0.3 0.0 0.0 C:N mean 2.7 3.4 3.4 2.9 3.4 3.3 3.1 3.7 3.2 3.0 3.3 3.1 2.6 3.4 3.2 2.9 3.4 3.4 2.9 3.4 3.4 2.7 3.8 3.4 sd 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.4 0.0 0.4 0.0 0.0 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 C test F * * F * * F * * F * * F ns ns U test F F * F F * F F * F F * F F * n 19 TL (cm) mean (SD) 161 (21) 8 147 (22) 10 106 (9) 10 116 (26) 9 73 (15) 2 U n 7 TL (cm) mean (SD) 128 (20) 3 72 (14) 3 U 2 U 7 55 (9) 5 132 (24) 6 104 (11) 11 15 (4) , Bathyraja interrupta BI Gymnura marmorata GM Myliobatis californica MC Pteroplatytrygon violacea PV Beringraja binoculata BB Raja rhina RR Urobatis halleri UH C test F ns ns U test F F ns C test F * * U test F F ns t ID BA af Species Bathyraja aleutica Dr 562 563 564 565 566 567 568 569 570 571 572 Page 20 of 36 F ns ns F * * F ns ns F * ns F F ns F F ns F F ns F F * F ns ns F * * F ns ns F * * F F ns F F ns F F ns F F ns C test F * * F * * F * * F * * F * * U test F F ns F F * F F ns F F ns F F * Table 1: Effects of urea and lipid extraction on stable isotope composition of various sharks (top panel) and batoids (bottom panel). Treatments are C (control), U (ureaFextracted), UL (urea & lipidFextracted). C test shows results of statistical comparisons between the control (C) and U and UL treatments, whereas U test shows comparison between U and UL (ns = not significant, * = significant, p <0.05). Note that for taxa with low sample sizes (<=3) we did not test for statistical differences. 20 https://mc06.manuscriptcentral.com/cjfas-pubs Page 21 of 36 0 1 Figure 1: Differences in δ13C, δ15N, and C:N between ureaFextracted and control samples (A) and ureaF and lipidFextracted and control samples (B) in sharks. Statistically significant differences between control and U and UL treatments are indicated (* p <= 0.05). Species are Prionace glauca (PG), Squalus suckleyi (SS), Triakis semifasciata (TS), Isurus oxyrinchus (IO), Lamna ditropis (LD), and Carcharodon carcharias (CC). We did not test for differences in TS due to low sample size. Figure 2: Differences in δ13C, δ15N, and C:N between ureaFextracted and control samples (A) and ureaF and lipidF extracted and control samples (B) in batoids. Statistically significant differences between control and U and UL treatments are indicated (* p <= 0.05). Species are Bathyraja aleutica (BA), Bathyraja interrupta (BI), Beringraja binoculata (BB), Raja rhina (RR), Myliobatis californica (IO), Gymnura marmorata (TS), Pteroplatytrygon violacea (PV), and Urobatis halleri (UH). We did not test for differences in BI, MC and GM due to low sample sizes. Figure 3: Relationships between the C:N of ureaFextracted tissue (treatment U, C:Nu) and changes in δ13C between the U and UL treatments (Uδ13C = δ13CU F δ13CUL). Lines show selected best fit modeled relationship between C:Nu and Uδ13C using best fit models based on AICc, which for groups with relatively low lipid content and C:Nu (batoids and all sharks except for dogfish) was a linear model, but for groups with a higher lipid content and C:Nu was based on the two parameter model from Fry (2002). Figure 4: Conceptual diagram showing relative effects of urea and lipid extraction on elasmobranch muscle tissue. Axes show relative change in δ13C and δ15N following different treatments, and color bar shows C:N. The secondary axes show the relative effect of urea and lipid removal on δ13C and δ15N, with the size of arrows indicating relative magnitude of effect. Urea extraction will generally increase δ15N and C:N, and potentially also affect δ13C as 13C enriched urea is removed. Lipid extraction does not influence δ15N, but increases δ13C and reduces C:N. The degree to which the different treatments affect δ13C, δ15N (depicted by the magnitude and direction of the “Urea extraction” and “Urea & lipid extraction” arrows) and C:N (depicted by the shading gradient within each treatment arrow) will vary based on the urea and lipid (dashed arrow) content of the tissue. t af Dr 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 Canadian Journal of Fisheries and Aquatic Sciences Figure S1: Changes in δ13C and δ15N of ureaFextracted (grey) and ureaF & lipidFextracted samples (black) relative to control samples in sharks (A) and batoids (B). Dotted lines show relative shift in values between ureaFextracted and urea & lipidFextracted samples. Shark species are Prionace glauca (PG), Squalus suckleyi (SS), Triakis semifasciata (TS), Isurus oxyrinchus (IO), Lamna ditropis (LD), and Carcharodon carcharias (CC). Batoid species are Bathyraja aleutica (BA), Bathyraja interrupta (BI), Beringraja binoculata (BB), Raja rhina (RR), Myliobatis californica (IO), Gymnura marmorata (TS), Pteroplatytrygon violacea (PV), and Urobatis halleri (UH). Figure S2: Relationship between the C:N of urea extracted tissue (U, C:Nu) and changes in δ13C between the U and UL treatments (Uδ13C = δ13CU F δ13CUL) for individual species. Inset: magnified view of data with low C:N (< 4.4). 21 https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences t af Dr Figure 1: Differences in δ13C, δ15N, and C:N between urea extracted and control samples (A) and urea and lipid extracted and control samples (B) in sharks. Statistically significant differences between control and U and UL treatments are indicated (* p <= 0.05). Species are (PG), (SS), (TS), (IO), (LD), and (CC).We did not test for differences in TS due to low sample size. 205x256mm (300 x 300 DPI) https://mc06.manuscriptcentral.com/cjfas-pubs Page 22 of 36 Page 23 of 36 Canadian Journal of Fisheries and Aquatic Sciences t af Dr Figure 2: Differences in δ13C, δ15N, and C:N between urea extracted and control samples (A) and urea and lipid extracted and control samples (B) in batoids. Statistically significant differences between control and U and UL treatments are indicated (* p <= 0.05). Species are (BA), (BI), (BB), (RR), (IO), (TS), (PV), and (UH). We did not test for differences in BI, MC and GM due to low sample sizes. 207x261mm (300 x 300 DPI) https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences af Dr t Figure 3: Relationships between the C:N of urea extracted tissue (treatment U, C:Nu) and changes in δ13C between the U and UL treatments (∆δ13C = δ13CU δ13CUL). Lines show selected best fit modeled relationship between C:Nu and ∆δ13C using best fit models based on AICc, which for groups with relatively low lipid content and C:Nu (batoids and all sharks except for dogfish) was a linear model, but for groups with a higher lipid content and C:Nu was based on the two parameter model from Fry (2002). 122x98mm (300 x 300 DPI) https://mc06.manuscriptcentral.com/cjfas-pubs Page 24 of 36 Page 25 of 36 Canadian Journal of Fisheries and Aquatic Sciences Dr t af Figure 4: Conceptual diagram showing relative effects of urea and lipid extraction on elasmobranch muscle tissue. Axes show relative change in δ13C and δ15N following different treatments, and color bar shows C:N. The secondary axes show the relative effect of urea and lipid removal on δ13C and δ15N, with the size of arrows indicating relative magnitude of effect. Urea extraction will generally increase δ15N and C:N, and potentially also affect δ13C as 13C enriched urea is removed. Lipid extraction does not influence δ15N, but increases δ13C and reduces C:N. The degree to which the different treatments affect δ13C, δ15N (depicted by the magnitude and direction of the “Urea extraction” and “Urea & lipid extraction” arrows) and C:N (depicted by the shading gradient within each treatment arrow) will vary based on the urea and lipid (dashed arrow) content of the tissue. 246x171mm (300 x 300 DPI) https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences t af Dr Figure S1: Changes in δ13C and δ15N of urea extracted (grey) and urea & lipid extracted samples (black) relative to control samples in sharks (A) and batoids (B). Dotted lines show relative shift in values between urea extracted and urea & lipid extracted samples. Shark species are (PG), (SS), (TS), s (IO), (LD), and (CC). Batoid species are (BA), (BI), (BB), (RR), ! (IO), " (TS), # (PV), and $ (UH). 209x273mm (300 x 300 DPI) https://mc06.manuscriptcentral.com/cjfas-pubs Page 26 of 36 Page 27 of 36 Canadian Journal of Fisheries and Aquatic Sciences Dr Figure S2: Relationship between the C:N of urea extracted tissue (U, C:Nu) and changes in δ13C between the U and UL treatments (∆δ13C = δ13CU ' δ13CUL) for individual species. Inset: magnified view of data with low C:N (< 4.4). af 123x78mm (300 x 300 DPI) t https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Study # C:N metric Adjustment Nothing Lipid ext. 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 Urea ext. 0 0 1 0 1 0 0 1 0 1 0 0 1 1 1 0 1 1 1 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 1 1 1 1 0 t 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 1 1 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 1 af 0 1 0 1 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 1 Dr 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 https://mc06.manuscriptcentral.com/cjfas-pubs Page 28 of 36 Page 29 of 36 Canadian Journal of Fisheries and Aquatic Sciences 48 49 50 Number 0 1 0 14 0 0 0 8 0 0 1 6 1 0 0 19 0 0 0 7 Table S2: Survey of 50 recent (2013-present) studies using stable isotope analysis to study the ecology of elasmobranchs. We c “stable isotope”, “elasmobranch”, “shark”, “ray”; and selected the first 50 pertinent results. We assessed each study to determin indicates the study used the C:N of bulk, non-extracted samples (i.e. contain urea and lipid) to assess the lipid content of their sa whether lipid extraction was required or not. “Adjustment” indicates the study used the C:N of non-extracted samples to mathem t af Dr https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 30 of 36 t af Dr Study Couturier et al. 2013. Stable Isotope and Signature Fatty Acid Analyses Suggest Reef Manta Rays Feed on Demersal Zooplankt Matich & Heithaus 2014. Multi ‐tissue stable isotope analysis and acoustic telemetry reveal seasonal variability in the trophic i Hussey et al. 2014. Rescaling the trophic structure of marine food webs. Ecology Letters, 17(2), pp.239-250. Daly et al. 2013. Comparative feeding ecology of bull sharks (Carcharhinus leucas) in the coastal waters of the Southwest India McMeans et al. 2013. The role of Greenland sharks (Somniosus microcephalus) in an Arctic ecosystem: assessed via stable isot Meneses et al. 2016. Trophic overlap between blue sharks (Prionace glauca) and shortfin makos (Isurus oxyrinchus): Trophic li Heithaus et al. 2013. Apparent resource partitioning and trophic structure of large-bodied marine predators in a relatively pristin Shiffman et al. 2014. Feeding ecology of the sandbar shark in south carolina estuaries revealed through δ13C and δ15N stable i De Lecea et al. 2013. Processes controlling the benthic food web of a mesotrophic bight (KwaZulu-Natal, South Africa) reveale Caut et al. 2013. Diet-and tissue-specific incorporation of isotopes in the shark Scyliorhinus stellaris, a North Sea mesopredator Tilley et al. 2013. Diet reconstruction and resource partitioning of a Caribbean marine mesopredator using stable isotope Bayes Navarroet al. 2014. Short-and long-term importance of small sharks in the diet of the rare deep-sea shark Dalatias licha. Marine Polo-Silva et al. 2013. Trophic shift in the diet of the pelagic thresher shark based on stomach contents and stable isotope analy Olin et al. 2013. Seasonal variability in stable isotopes of estuarine consumers under different freshwater flow regimes. Marine Munroe et al. 2015. Regional movement patterns of a small ‐bodied shark revealed by stable ‐isotope analysis. Journal of fis Li et al. 2014. Trophic ecology of sharks in the mid-east Pacific ocean inferred from stable isotopes. Journal of Ocean Universit Churchill et al. 2015. Trophic interactions of common elasmobranchs in deep-sea communities of the Gulf of Mexico revealed Kiszka et al. 2015. Plasticity of trophic interactions among sharks from the oceanic south-western Indian Ocean revealed by sta Espinoza et al. 2015. Feeding ecology of common demersal elasmobranch species in the Pacific coast of Costa Rica inferred fro McMean et al. 2015. Impacts of food web structure and feeding behavior on mercury exposure in Greenland Sharks (Somniosu Malpica‐Cruz et al. 2013. Tissue ‐specific stable isotope ratios of shortfin mako (Isurus oxyrinchus) and white (Carcharodon ca Torres et al. 2014. Trophic ecology and bioindicator potential of the North Atlantic tope shark. Science of The Total Environme McCauley et al. 2014. Reliance of mobile species on sensitive habitats: a case study of manta rays (Manta alfredi) and lagoons. Papastamatiou et al. 2015. Movements and foraging of predators associated with mesophotic coral reefs and their potential for l de Moura et al. 2015. Assessment of trace elements, POPs, 210 Po and stable isotopes (15 N and 13 C) in a rare filter-feeding sh Madigan et al. 2015. Diet shift and site-fidelity of oceanic whitetip sharks Carcharhinus longimanus along the Great Bahama B Speed et al. 2012. Trophic ecology of reef sharks determined using stable isotopes and telemetry. Coral Reefs, 31(2), pp.357-36 Drymon et al. 2012. Trophic plasticity in the Atlantic sharpnose shark (Rhizoprionodon terraenovae) from the north central Gul Fanelli et al. 2013. Trophic webs of deep-sea megafauna on mainland and insular slopes of the NW Mediterranean: a comparis Courtney & Foy, R., 2012. Pacific sleeper shark Somniosus pacificus trophic ecology in the eastern North Pacific Ocean inferre McCauley et al. 2012. Assessing the effects of large mobile predators on ecosystem connectivity. Ecological Applications, 22(6 Lyons et al. 2013. Effects of trophic ecology and habitat use on maternal transfer of contaminants in four species of young of th Matich et al. 2015. Short-term shifts of stable isotope (δ 13 C, δ 15 N) values in juvenile sharks within nursery areas suggest rap Albo-Puigserver et al. 2015. Feeding ecology and trophic position of three sympatric demersal chondrichthyans in the northwes Rojas et al. 2014. Feeding grounds of juvenile scalloped hammerhead sharks (Sphyrna lewini) in the south-eastern Gulf of Calif Reum & Marshall 2013. Evaluating δ15N–body size relationships across taxonomic levels using hierarchical models. Oecologia Hussey et al. 2015. Expanded trophic complexity among large sharks. Food Webs, 4, pp.1-7. Valls et al. 2014. Structure and dynamics of food webs in the water column on shelf and slope grounds of the western Mediterra Teffer et al. 2014. Trophic influences on mercury accumulation in top pelagic predators from offshore New England waters of t Kiszka et al. 2014. Trophic ecology of common elasmobranchs exploited by artisanal shark fisheries off south-western Madaga Cresson et al. 2014. Mercury in organisms from the Northwestern Mediterranean slope: Importance of food sources. Science of Barría et al. 2015. Unravelling the ecological role and trophic relationships of uncommon and threatened elasmobranchs in the Lopez et al. 2013. Trophic ecology of the dusky catshark Bythaelurus canescens (Günther, 1878)(Chondrychthyes: Scyliorhinid Freedman et al. 2015. Connectivity and movements of juvenile predatory fishes between discrete restored estuaries in southern Shaw et al. 2016. Trophic Ecology of a Predatory Community in a Shallow-Water, High-Salinity Estuary Assessed by Stable Is Jaime-Rivera et al. 2014. Feeding and migration habits of white shark Carcharodon carcharias (Lamniformes: Lamnidae) from Torres-Rojas et al. 2015. Diet and trophic level of scalloped hammerhead shark (Sphyrna lewini) from the Gulf of California an https://mc06.manuscriptcentral.com/cjfas-pubs Page 31 of 36 Canadian Journal of Fisheries and Aquatic Sciences Hernández-Aguilar et al. 2015. Trophic ecology of the blue shark (Prionace glauca) based on stable isotopes ( δ 13 C and δ Raoult et al. 2015. Not all sawsharks are equal: species of co-existing sawsharks show plasticity in trophic consumption both w Frisch et al. 2016. Reassessing the trophic role of reef sharks as apex predators on coral reefs. Coral Reefs, pp.1-14. e analysis to study the ecology of elasmobranchs. We conducted the search on Google scholar using the search terms: 50 pertinent results. We assessed each study to determine how they treated samples for urea and lipid. “C:N metric” ntain urea and lipid) to assess the lipid content of their samples and use that inferred lipid content to inform them 13 e study used the C:N of non-extracted samples to mathematically adjust δ C to account for inferred lipid content. t af Dr https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 32 of 36 on Demersal Zooplankton. PLoS ONE 8(10): e77152 ariability in the trophic interactions of juvenile bull sharks in a coastal estuary. Journal of Animal Ecology, 83(1), pp.199-213. t af Dr of the Southwest Indian Ocean inferred from stable isotope analysis. PloS one, 8(10), p.e78229. assessed via stable isotopes and fatty acids. Marine biology, 160(5), pp.1223-1238. oxyrinchus): Trophic linkages between two shark species In the Eastern Pacific Ocean food web. Food Webs. ors in a relatively pristine seagrass ecosystem. Mar Ecol Prog Ser, 481, pp.225-237. δ13C and δ15N stable isotope analysis. Marine and Coastal Fisheries, 6(1), pp.156-169. l, South Africa) revealed by stable isotope analysis. Marine Ecology Progress Series, 484, pp.97-114. North Sea mesopredator. Marine Ecology. Progress Series, 492. ng stable isotope Bayesian modelling. PloS one, 8(11), p.e79560. k Dalatias licha. Marine biology, 161(7), pp.1697-1707. and stable isotope analyses. Marine Biology Research, 9(10), pp.958-971. r flow regimes. Marine Ecology Progress Series, 487, pp.55-69. analysis. Journal of fish biology, 86(5), pp.1567-1586. rnal of Ocean University of China, 13(2), pp.278-282. ulf of Mexico revealed through stable isotope and stomach content analysis. Deep Sea Research Part II: Topical Studies in Oceanography, 1 n Ocean revealed by stable isotope and mercury analyses. Deep Sea Research Part I: Oceanographic Research Papers, 96, pp.49-58. Costa Rica inferred from stable isotope and stomach content analyses. Journal of Experimental Marine Biology and Ecology, 470, pp.12-25 nland Sharks (Somniosus microcephalus). Science of The Total Environment, 509, pp.216-225. d white (Carcharodon carcharias) sharks as indicators of size ‐based differences in foraging habitat and trophic level. Fisheries Oceanog of The Total Environment, 481, pp.574-581. ta alfredi) and lagoons. Marine biology, 161(9), pp.1987-1998. and their potential for linking ecological habitats. Marine Ecology Progress Series. n a rare filter-feeding shark: The megamouth. Marine pollution bulletin, 95(1), pp.402-406. ng the Great Bahama Bank. Marine Ecology Progress Series, 529, p.185. Reefs, 31(2), pp.357-367. m the north central Gulf of Mexico. Environmental biology of fishes, 95(1), pp.21-35. diterranean: a comparison by stable isotope analysis. Marine ecology. Progress series, 490, pp.199-221. th Pacific Ocean inferred from nitrogen and carbon stable ‐isotope ratios and diet. Journal of fish biology, 80(5), pp.1508-1545. gical Applications, 22(6), pp.1711-1717. r species of young of the year lamniform sharks. Marine environmental research, 90, pp.27-38. ursery areas suggest rapid shifts in energy pathways. Journal of Experimental Marine Biology and Ecology, 465, pp.83-91. hthyans in the northwestern Mediterranean. Marine Ecology Progress Series, 524, p.255. uth-eastern Gulf of California. Hydrobiologia, 726(1), pp.81-94. hical models. Oecologia, 173(4), pp.1159-1168. of the western Mediterranean. Journal of Marine Systems, 138, pp.171-181. New England waters of the northwest Atlantic Ocean. Marine environmental research, 101, pp.124-134. south-western Madagascar inferred from stable isotopes. Aquatic Biology, 23, pp.29-38. ood sources. Science of The Total Environment, 497, pp.229-238. d elasmobranchs in the western Mediterranean Sea. Marine Ecology Progress Series, 539, p.225. drychthyes: Scyliorhinidae) in the southeast Pacific Ocean. Journal of Applied Ichthyology, 29(4), pp.751-756. d estuaries in southern California. Marine Ecology Progress Series, 520, p.191. y Assessed by Stable Isotope Analysis. Marine and Coastal Fisheries, 8(1), pp.46-61. rmes: Lamnidae) from Isla Guadalupe inferred by analysis of stable isotopes δ15N and δ13C. Revista de Biolog he Gulf of California and Gulf of Tehuantepec, Mexico. Iranian Journal of Fisheries Sciences, 14(3), pp.767-785. https://mc06.manuscriptcentral.com/cjfas-pubs ía Tropical, 62(2), p Page 33 of 36 Canadian Journal of Fisheries and Aquatic Sciences δ 13 C and δ 15 N) and stomach content. Journal of the Marine Biological Association of the United Kingdom, pp.1-8. hic consumption both within and between species. Canadian Journal of Fisheries and Aquatic Sciences, 72(11), pp.1769-1775. fs, pp.1-14. t af Dr https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences pp.199-213. Studies in Oceanography, 115, pp.92-102. apers, 96, pp.49-58. and Ecology, 470, pp.12-25. t af Dr hic level. Fisheries Oceanography, 22(6), pp.429-445. , 80(5), pp.1508-1545. pp.83-91. ía Tropical, 62(2), pp.637-647. https://mc06.manuscriptcentral.com/cjfas-pubs Page 34 of 36 Page 35 of 36 Canadian Journal of Fisheries and Aquatic Sciences Kingdom, pp.1-8. p.1769-1775. t af Dr https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences data All species All sharks (no Dogfish) All batoids Dogfish S. suckleyi Great White C. carcharias Δ13C = (a C:NU + b )( C:NU + c )-1 AICc: 94.02 a= b= 5.39 ± 0.85 -17.25 ± 2.62 c= -0.68 ± 0.75 AICc: a= b= 38.58 1.94 ± 0.23 -6.27 ± 0.75 c= -2.98 ± 0.06 a= b= nf - c= - AICc: c= -1.90 ± 1.03 a= b= -0.46 14.18 ± 16.56 -43.96 ± 50.48 c= 3.22 ± 9.12 Δ13C = β 0 + β 1 Ln(C:NU) AICc: 102.33 β0 = β1 = 6.12 ± 0.30 3.18 ± 0.03 AICc: P= F= 38.97 7.86 ± 0.70 3.15 ± 0.03 AICc: β0 = β1 = a= b= 0.88 ± 0.06 -2.57 ± 0.22 0.76 r2 : 0.70 39.54 -7.69 ± 0.82 6.74 ± 0.66 2 r: AICc: P= F= 15.41 5.63 ± 0.77 3.28 ± 0.04 Dr AICc: af β0 = β1 = 17.30 6.31 ± 0.62 3.35 ± 0.20 0.69 13.87 -6.15 ± 0.91 5.19 ± 0.73 r2 : t AICc: β0 = β1 = 0.56 20.93 -4.08 ± 1.27 3.79 ± 0.74 r2 : AICc: P= F= -5.15 8.30 ± 0.52 3.14 ± 0.03 AICc: β0 = β1 = Δ13C = b + a C:NU AICc: 121.93 -4.89 ± 0.33 4.34 ± 0.25 r2 : P= F= a= b= AICc: P= F= AICc: 22.46 4.76 ± 1.03 -16.96 ± 3.44 AICc: Δ13C = P - PF (C:NU)-1 AICc: 92.53 Page 36 of 36 0.76 -3.88 -7.80 ± 0.61 6.90 ± 0.48 2 r: 0.92 AICc: a= b= 40.50 1.81 ± 0.18 -5.61 ± 0.63 2 r: 0.68 a= b= 12.26 1.45 ± 0.19 -4.74 ± 0.69 r2 : 0.58 AICc: a= b= 24.39 0.62 ± 0.16 -1.18 ± 0.91 r2 : 0.65 AICc: AICc: a= b= -3.27 1.80 ± 0.13 -5.49 ± 0.46 2 r: 2 Table S1. Parameter estimates (± SE), r (linear models only) and corrected Akaike Information Criteria valeus (AICc) for models of Δ13C, the difference between lipid and urea (UL) and urea (U) extracted δ13C values (Δ13C = δ13CU - δ13CUL), fit to all species, batoids, all sharks other than S. suckleyi (dogfish), S. suckleyi , and C. carcharias . nf = model failed to converg. https://mc06.manuscriptcentral.com/cjfas-pubs 0.92