Quantifying ontogenetic stable isotope variation between dermis and muscle tissue of two pelagic sharks

Comparative analysis of isotope values from different tissues can capture temporal variation in the trophic and foraging behavior of difficult to study large marine predators, revealing either uniform or variable ecological roles over time. The isotopic values (δ13C and δ15N) of dermis, and muscle tissue of silky Carcharhinus falciformis and blue sharks Prionace glauca sampled in the northeast central Pacific were analyzed to quantify ontogenetic interand intra-tissue isotopic variation. Consistent differences in δ15N values occurred between dermis and muscle tissue for both species (2.5 ± 0.4‰ and 2.1 ± 0.3‰, respectively), while tissue differences in δ13C values were more variable between species (2.3 ± 0.6‰ and 0.7 ± 0.6‰, respectively), likely a result of tissue composition. The overall δ15N and δ13C values of dermis and muscle were highly correlated for blue sharks and for silky sharks with the exception of silky shark δ13C values. This pattern indicates that dermis isotope values are able to provide a proxy for muscle tissue, similar to that previously reported for fin, accepting dermis-specific diet−tissue discrimination factors. Tissue-specific ontogenetic isotopic variation for the silky shark, and the low regression slope value between dermis and muscle δ13C values, however, may suggest that dermis and muscle tissue have different isotopic turnover rates. These data demonstrate that dermis yields valuable isotope data to examine the trophic ecology and feeding/movement behavior of sharks, but further work is required to address dermis-specific turnover rates and diet−tissue discrimination factors.


INTRODUCTION
In recent years, our understanding of the temporal and spatial variation in trophic roles and foraging dynamics of shark species has improved based on the application of carbon and nitrogen stable isotopes (Matich et al. 2010, Kim et al. 2012a). Compared to the instantaneous 'snapshot' of dietary information obtained from gut content analysis, stable isotope analysis (SIA) allows exami-ABSTRACT: Comparative analysis of isotope values from different tissues can capture temporal variation in the trophic and foraging behavior of difficult to study large marine predators, revealing either uniform or variable ecological roles over time. The isotopic values (δ 13 C and δ 15 N) of dermis, and muscle tissue of silky Carcharhinus falciformis and blue sharks Prionace glauca sampled in the northeast central Pacific were analyzed to quantify ontogenetic inter-and intra-tissue isotopic variation. Consistent differences in δ 15 N values occurred between dermis and muscle tissue for both species (2.5 ± 0.4 ‰ and 2.1 ± 0.3 ‰, respectively), while tissue differences in δ 13 C values were more variable between species (2.3 ± 0.6 ‰ and 0.7 ± 0.6 ‰, respectively), likely a result of tissue composition. The overall δ 15 N and δ 13 C values of dermis and muscle were highly correlated for blue sharks and for silky sharks with the exception of silky shark δ 13 C values. This pattern indicates that dermis isotope values are able to provide a proxy for muscle tissue, similar to that previously reported for fin, accepting dermis-specific diet−tissue discrimination factors. Tissue-specific ontogenetic isotopic variation for the silky shark, and the low regression slope value between dermis and muscle δ 13 C values, however, may suggest that dermis and muscle tissue have different isotopic turnover rates. These data demonstrate that dermis yields valuable isotope data to examine the trophic ecology and feeding/movement behavior of sharks, but further work is required to address dermis-specific turnover rates and diet−tissue discrimination factors.
Considering the threatened status of many shark species (Dulvy et al. 2014) and the requirement to limit mortalities, nondestructive or minimally invasive sampling methods are increasingly being adopted in field studies (Hammerschlag & Suli kowski 2011). Such tissue sampling from large sharks often includes blood (plasma and red blood cells), fin, and/ or a biopsy sample of skin, connective tissue, and muscle. Fin is considered an attractive tissue to sample, given the easy, quick, and minimally invasive sampling protocol (Matich et al. 2010. Skin, or more specifically the dermal collagen fiber layers that underlie the dermal denticles of shark skin (hereafter referred to as dermis), potentially provides another tissue that is easy to sample, and is likely more homogeneous in terms of structural composition compared to fin (Matich et al. 2010, Carlisle et al. 2012. To date, the isotopic composition of dermis has received limited attention (but see Carlisle et al. 2012, Jaime-Rivera et al. 2013 despite the above advantages. In this study, we analyzed the isotopic values (δ 13 C and δ 15 N) of dermis and muscle tissue sampled from silky Carcharhinus falciformis and blue sharks Priona ce glauca, 2 pelagic species that are commonly caught in commercial fisheries. Both species are known to predominantly feed on cephalopods in pelagic waters, but silky sharks are considered to forage both in nearshore and pelagic habitats in contrast to the blue shark (Rabehagasoa et al. 2012. Both species, however, undertake large-scale movement patterns between foraging locations (Bonfil 2008, Stevens et al. 2010. The objectives of this study were (1) to quantify intertissue isotopic differences and (2) to examine interand intra-tissue isotopic variation by age for both pelagic shark species. Overall, the aim was to determine if δ 13 C and δ 15 N values of dermis can provide valuable temporal insights into the tro phic ecology and feeding behavior of these impacted pelagic species.

Sample collection
Tissue samples were obtained from silky and blue sharks caught as bycatch in the Chinese tuna longline fishery operating in the northeast central Pacific (approximately 8° to 10°N, 115° to 125°W) between June and November 2014. A skin tissue sample, in cluding both dermis and dermal denticles, was excised from the anal fin region of each individual to minimize the effects of specific variation in skin thickness, and a small portion (~10 g) of white muscle tissue was excised adjacent to the vertebral column. Tissue samples were immediately placed on ice and frozen within 30 min of sampling. For each individual, precaudal length (PCL) was measured as the straight-line distance from the tip of the snout to the beginning of the caudal fin. The age of all sharks sampled was estimated from PCL data using von Bertalanffy growth parameters for the 2 species from the same sampling region (Oshitani et al. 2003, Nakano & Stevens 2008. The length at 50% maturity was set at 151 to 156 cm PCL and 150 to 159 cm PCL for the silky and blue shark, respectively (Oshitani et al. 2003, Nakano & Stevens 2008.

Stable isotope analysis
Following the removal of dermal denticles and epidermis, isolated dermis tissue was washed and soaked overnight in de-ionized water. Dermis was then freeze-dried at −55°C for 48 h using a Christ Alpha 1-4 LD plus Freeze Dryer (Martin Christ), and each sample was homogenized using a Retsch Mixer Mill MM 400 (Retsch). Sharks maintain urea and trimethyl amine oxide in their tissues for osmotic balance, which are considered to be 15 N-depleted and may also affect δ 13 C values (Fisk et al. 2002, Li et al. 2016. Consequently, urea was removed from muscle tissue using a standard water washing technique (Kim & Koch 2012, Li et al. 2016). For δ 15 N analysis, all muscle tissue samples were urea-extracted, while for δ 13 C, both urea and lipids were removed using combined water washing and standard chloroformmethanol extraction (Li et al. 2016). Tissue samples were then freeze-dried to remove remaining solvent and divided into approximately 1 to 1.5 mg subsamples for stable isotope analysis using an IsoPrime 100 isotope ratio mass spectrometer (IsoPrime) and a vario ISOTOPE cube elemental analyzer (Elementar Analysensysteme) at Shanghai Ocean University Stable Isotope Laboratory. Reference standards for quantifying 13 C and 15 N stable isotope values, were USGS 24 (−16.1± 0.04 ‰ V-PDB) and USGS 26 (53.7 ± 0.4 ‰ V-AIR), respectively. The analytical errors of δ 13 C and δ 15 N values were ± 0.05 ‰ and ± 0.06 ‰, respectively.

Statistical analysis
To examine the relationship of isotopic values between tissues, paired t-tests and correlation analyses were used to compare individual δ 13 C and δ 15 N values between dermis and muscle for silky and blue sharks. The mean differences in δ 13 C and δ 15 N values between dermis and muscle were then calculated to examine the isotopic offset between tissues for both species. To examine tissue isotopic differences over ontogeny, least squares linear regressions were fitted to determine the relationship between estimated age and muscle and dermis δ 13 C and δ 15 N values. The absolute isotopic differences between dermis and muscle per individual were calculated to further quantify age-based inter-and intra-tissue isotopic variation and to identify whether the observed agerelated trends were comparable between species. The above analyses were conducted on all the data grouped by species because an initial examination found no regional differences in isotope values. The normality of δ 13 C and δ 15 N values for dermis and muscle tissue were examined using a Shapiro-Wilk normality test. All statistical analyses were conducted using R version 3.1.2 (R Development Core Team 2014), and the level of statistical significance was set at α = 0.05.
The δ 13 C and δ 15 N values for dermis and muscle were highly correlated for both pelagic sharks. For silky sharks, the relationship between dermis and muscle δ 13 C values was significant but relatively weak compared to the dermis-muscle δ 15 N relationship (Fig. 2a,c). In contrast, the above δ 13 C and δ 15 N relationships for blue sharks were both strong with R 2 > 0.84 and regression slopes that were not statistically different from 1 (Fig. 2b,d).
Overall, the δ 13 C and δ 15 N values of each tissue exhibited similar trends with increasing age of shark. For silky sharks, the δ 13 C values of both tissues showed no relationship with age (dermis: R 2 = 0.01, p = 0.5 and muscle: R 2 = 0.01, p = 0.1; Fig. 3a). Conversely, there was a significant negative relationship between the δ 15 N values of silky shark dermis and age, and a similar but non-significant trend was observed between muscle δ 15 N values and age (Fig. 3c). For blue sharks, δ 13 C values of both tissues increased significantly with increasing age (dermis: R 2 = 0.2, p < 0.05 and muscle: R 2 = 0.2, p < 0.05; Fig. 3b). In contrast, there was no relationship be tween δ 15 N values and age of blue sharks for either dermis or muscle (dermis: R 2 = 0.01, p = 0.5 and muscle: R 2 = 0.01, p = 0.1; Fig. 3d).
The absolute difference between dermis and muscle δ 13 C values for both shark species were variable and showed no relationship with age (silky: R 2 = 0.1, p = 0.06; blue: R 2 = 0.1, p = 0.2, Fig. 4). For the silky shark, the absolute difference in dermis-muscle δ 15 N values increased significantly with age (R 2 = 0.6, p < 0.001), while the absolute differences in δ 15 N values for the blue shark were more variable, and no age relationship was evident (R 2 = 0.003, p = 0.8).

DISCUSSION
Examination of the carbon and nitrogen isotopic relationships between different tissues sampled from pelagic sharks provides insights into the amenability of these tissues to address ecological questions over the species' trophic roles, foraging behavior, and habitat use. Certain tissues may provide reliable surrogate isotope values for the most commonly ana-  lyzed muscle tissue but are less invasive to sample, while multiple tissue sampling allows opportunities to examine temporal variation in diet and feeding interactions. The data present a comparative analysis of the isotope values of dermis and identify that it provides valuable information to further explore the trophic dynamics of these commercially exploited species. The fact that δ 13 C and δ 15 N values of dermis and muscle tissues for both pelagic shark species were significantly correlated and that standard deviations of each tissue were low for both isotopes indicates that dermis isotope values may provide a potential proxy for muscle tissue for specific life stages. These findings are in agreement with those previously reported for fin-muscle tissue isotopic relationships for sharks , Matich et al. 2011). There was clear evidence, however, to suggest tissuespecific isotope values. Specifically, the weaker correlations between tissue isotope values for silky sharks compared to blue sharks support this latter point, identifying tissue-specific turnover rates and isotopic integration periods. For the silky shark, this pattern may be a result of the different life-history stages sampled relative to the blue shark, given agespecific feeding reported among juvenile, sub-adult, and adult life-stages (Cabrera-Chávez-Costa et al. 2010, Malpica-Cruz et al. 2013. Marked ontogenetic dietary shifts have been previously reported for silky sharks, with juveniles feeding predominantly on jumbo squid Dosidicus gigas, whereas sub-adults consumed teleost fish Scomber japonicus (Cabrera-Chávez-Costa et al. 2010). Variable turnover rates between dermis and muscle tissue coupled with a marked diet shift would result in the observed greater inter-tissue isotopic variation. In contrast, most blue sharks sampled were sub-adult and mature animals (Hsu et al. 2011) that consume epipelagic and mesopelagic cephalopods (Hernández-Aguilar et al. 2015), similar to the diet of juveniles (Nakano & Stevens 2008). This consumption would result in more consistent isotope values between tissues even with different turnover rates, but small sample sizes over ontogeny limit the conclusions that can be made. Shark placoid scales or dermal denticles form in the dermis and increase in size with increasing body surface or body growth during the lifetime of an animal (reviewed by Meyer & Seegers 2012). The growth of these placoid scales is continuous, with substitution taking place as scales shed, resulting in different growth stages of scales occurring between fully grown dermal denticles (Meyer & Seegers 2012). Consequently, the isotopic incorporation rate of dermis is likely relatively fast, providing insight into short-term diet and feeding behavior when compared with muscle (Carlisle et al. 2012). The low regression slopes between muscle and dermis tissue isotope values of the silky shark may support this point. Previous studies have identified relatively fast isotopic turnover rates in skin tissue of marine mammals; for example, half lives of δ 13 C and δ 15 N were < 2 mo for manatees Trichechus manatus latirostris and bottlenose dolphins Tursiops truncatus, respectively (Alves-Stanley & Worthy 2009, Browning et al. 2014. Although the dermis of sharks is structurally different to that of marine mammals, the fact that skin shedding/ replacement occurs across a broad range of marine species (Alves-Stanley & Worthy 2009, Meyer & Seegers 2012, Browning et al. 2014 would suggest a faster isotopic turnover rate. For both shark species, dermis δ 13 C values were consistently higher than those of muscle tissue. Collagen and elastic fibers, the primary constituents of dermis (Meyer & Seegers 2012), typically drive enriched 13 C values relative to diet (Kim & Koch 2012). This trend was previously reported for white shark dermal δ 13 C values relative to muscle (Carlisle et al. 2012, Jaime-Rivera et al. 2013. It is likely that 13 C enrichment of dermis is a reflection of differences in biochemical composition between dermis and muscle, because each amino acid comprising the tissue proteins can vary in δ 13 C and δ 15 N values (Car lisle et al. 2012, Vander Zanden et al. 2014. Alternatively, tissue-specific diet discrimination factors may provide an explanation for the observed isotopic differences between tissues (Matich et al. 2010, Cherel et al. 2014. It is still likely, however, that variable tissue-turnover rates explain some of the observed variation in δ 13 C and δ 15 N values between dermis and muscle   It is important to note that the thickness of dermis can vary among species and over ontogeny, and females at different stages of reproductive maturity typically have variable dermal thickness related to mating (Motta 1977, Meyer & Seegers 2012. In this study, samples were taken from the anal fin area to minimize potential variation, but given that sex was not recorded, dermal thickness may also account for some of the observed isotopic variation. Following birth, juvenile silky sharks transition from coastal waters to offshore and pelagic habitat (Dagorn et al. 2007, Bonfil 2008) but continue to use shallow waters near the edge of the continental shelf and over deep-water reefs (Bonfil 1997). The δ 13 C−age relationships for both tissues of silky shark were non-significant, while the δ 15 N values in dermis decreased and the offset between dermis and muscle tissue δ 15 N values increased significantly with age, i.e. there was a large difference in values between tissues of older animals. These δ 15 N and δ 13 C relationships indicate that older and larger animals show a marked temporal shift in diet but that animals of all ages consistently occur in the same habitat (Malpica-Cruz et al. 2013).
For blue sharks, there were no obvious tissuespecific isotopic trends with age, but large variation among individuals was observed. This variation may suggest variable diets, but more likely reflects basinscale movements across spatial gradients in baseline δ 15 N values (Cherel et al. 2007, Kiszka et al. 2014. In contrast to the silky shark, the δ 13 C values of both tissues of blue sharks increased with age, possibly reflecting an ontogenetic shift in foraging location or increased consumption of 13 Cenriched prey. Such size-related isotopic patterns are consistent with those reported for blue sharks in the Indian Ocean (Rabehagasoa et al. 2012, Kiszka et al. 2015. Pelagic sharks are top predators that couple marine ecosystems over large spatio-temporal scales; consequently, their removal is predicted to destabilize food web interactions (Heithaus et al. 2008). These data demonstrate that dermis can yield valuable isotope data to better understand the role of pelagic sharks, but observed differences between dermis and muscle tissue δ 15 N and δ 13 C values were species-specific and life stage-dependent, which may relate to variable isotopic routing, tissue-specific amino acid composition, and/or different turnover rates between tissues. Our data when compared to published fin and dermis isotope data for sharks suggest a combination of these influencing factors , Carlisle et al. 2012), but further work will be required to quantify dermis turnover rates and discrimination factors in sharks to improve confidence when interpreting multi-tissue stable isotope data.