Retrospective stable isotopes of vertebrae reveal sexual ontogenetic patterns and trophic ecology in oceanic whitetip shark, Carcharhinus longimanus

Abstract There is a common phenomenon in nature whereby some animals have differences in their ontogenetic changes in dietary preferences between sexes, especially apex predators. These reflect changes in the needs of development during their lifetimes. Apex predators potentially have diverse dietary niches and a large impact on the trophic dynamics within ecosystems. However, the difference in life history between males and females often leads to increased difficulty in management and conservation. In this study, 25 oceanic whitetip sharks, Carcharhinus longimanus, were collected from the central and eastern tropical Pacific. Retrospective stable isotope analysis of vertebrae was used to evaluate the potential ontogenetic differences in feeding habits and niche width between sexes. Results showed that C. longimanus had a wide range of δ13C values (−18.1 to −12.3‰) and δ15N values (8.9–14.8‰). However, males and females had similar trophic positions with large niche overlap at similar growth stages. Both sexes had increasing δ13C values but relatively constant δ15N values along the vertebrae. These results indicated that male and female C. longimanus may share similar feeding strategies and movement patterns. The results presented in this study enhance our understanding of sexual ontogenetic patterns and ecological role of C. longimanus and highlighted the applicability of vertebrae for characterizing shark life‐history traits.


| INTRODUC TI ON
Most oceanic pelagic sharks are highly migratory predators playing complex but critical roles in marine ecosystems (Bonfil et al., 2008). The oceanic whitetip shark, Carcharhinus longimanus, is an apex marine predator potentially completing its entire life cycle in the open ocean (Backus et al., 1956;Bonfil et al., 2008;Mather & Day, 1954). Characteristics associated with high longevity, late maturity, slow growth rate, and low fecundity make this once abundant species experience severe population declines throughout its global range due to overfishing (D'Alberto et al., 2017;Myers et al., 2007;Ward & Myers, 2005). The decline of large predatory species was reported to reduce the natural mortality in a range of their preys and trigger trophic cascade changes in many marine ecosystems (Ferretti et al., 2010). Recently, the oceanic whitetip shark was listed as Critically Endangered by the International Union for the Conservation of Nature (IUCN), as well as being classified to Appendix II of the Convention on International Trade in Endangered Species (CITES) in 2013 (Rigby et al., 2019).
The feeding habits of large predatory sharks usually change through ontogeny, primarily due to morphological changes accompanying growth, age-specific habitat use, foraging tactics, and reproductive requirements (Estupiñán-Montaño et al., 2018;Heupel et al., 2007;Werner & Hall, 1976). Spatial and sexual segregations were observed in several oceanic shark species, which makes their stock assessment and conservation more complicated (Tsai et al., 2015).
Although previous observations speculated on the opportunistic feeding behavior of the C. longimanus, knowledge of its foraging ecology is still fragmentary, especially its ontogeny and sexual variation due to its highly migratory behavior and the inaccessibility of oceanic habitat (Baum & Worm, 2009;Hussey et al., 2010;Newman et al., 2012). Such information is crucial for their conservation and efficient fishery management but also to preserve pelagic ecosystem functioning since sharks are often the keystone species in marine food webs (Grubbs, 2010;Madigan et al., 2021).
Stable carbon and nitrogen isotope ratios (δ 13 C and δ 15 N, respectively) in metabolically inert tissue, such as shark vertebrae, are efficient intrinsic markers for elucidating ontogeny/sexual variations in diet and/or habitat use of many shark species, such as the white shark Carcharodon carcharias (Estrada et al., 2006;Kerr et al., 2006;Kim et al., 2012;Wolf et al., 2009), blue shark Prionace glauca (Curnick et al., 2019;Estupiñán-Montaño et al., 2019), salmon shark Lamna ditropis (Carlisle et al., 2014), and three hammerhead shark species Sphyrna mokarran, S. lewini, and S. zygaena (Loor-Andrade et al., 2015;Raoult et al., 2019). This approach is based on the fact that shark vertebrae are related to lifetime information deposited in the growth bands reflecting the entire life histories of individuals (Estrada et al., 2006).
In this study, the isotopic time series from the successive vertebral sections of C. longimanus were investigated to evaluate the  (Table 1). Vertebrae columns were taken from the dorsal-anterior part of the shark body, between the head and first dorsal fin, cleaned using ultrapure water, and stored at −40°C until being transported to the laboratory.

| Preparation for age reading
The neural arch and connective tissue were removed from each vertebra.
Then, the vertebrae were dried in drying ovens at 60°C. To distinguish the growth bands, we polished the vertebrae with 120, 600, and 1200 μm grit in sequence to optimize the visualization of growth bands on the sagittal plane. After the age estimation by two readers, the vertebrae were sampled from the birth ring toward the outside edge to obtain vertebral collagen samples, using a micro-drill with a 0.5 mm drill bit ( Figure 2). F I G U R E 1 Sampling locations in the central and eastern tropical Pacific

| Stable isotope analysis (SIA)
Due to the high density of the edge, we sampled every two-growth band after the eighth annulus band. To remove residual inorganic carbon, powdered vertebral samples were placed in 1.5 ml of ethylenediaminetetraacetic acid (EDTA) solution at 0.5 M for a week.
EDTA was preferred over hydrochloric acid (HCl) because it was less likely to dissolve the sample. Once the process was complete, the samples were rinsed five times with deionized water and placed into a drying oven at 60°C for 24 h. Approximately 0.2-2 mg of samples were weighed into tin capsules and analyzed using an IsoPrime 100 isotope ratio mass spectrometer (IsoPrime Corporation, Cheadle, UK) and a vario IsoPrime cube elemental analyzer (Elementar Analysensysteme GmbH, Hanau, Germany).
The C/N ratio can be estimated to determine whether the treatment applied to the vertebral collagen was effective. A ratio ≤3.5 indicates that demineralization has been effective (Hussey et al., 2012). The isotope compositions of the samples are expressed in δ 13 C and δ 15 N notation and were calculated using the following equations: where ‰ is parts per thousand; 13 C/ 12 C and 15 N/ 14 N are the atomic ratios of 13 C and 15 N in the sample or standard, respectively; δ is the heavy-to-light isotope ratio in the sample. The standard reference materials for C and N were Pee Dee Belemnite carbonate and air, respectively. The reference standards USGS 24 (−16.049‰ vPDB) and USGS 26 (53.7‰ vN 2 ) were used for the quantification of the 13 C and 15 N stable isotope values, respectively. Every tenth sample was run in triplicate with a laboratory reference standard (Protein (−26.98‰ vPDB and 5.96‰ vN 2 )) to assess the within-run precision, and a blank sample was run every ten samples to remove residual gases. The analytical errors of the δ 13 C and δ 15 N values were approximately 0.13‰ and 0.06‰, respectively.

| Statistical analysis
Due to male sharks matured at 8.8 ± 1.2 years, and the females matured at 8.6 ± 1.2 years (Shen et al., 2021), the sharks were grouped into eight categories: those from 1-7 years old were divided into seven groups, and those older than 8 years were classified as adult individuals. The ontogenetic isotopic enrichment patterns were inferred from a sampling starting point located at the birth ring ( Figure 2). The relative enrichment of 13 C and 15 N was calculated using the following equations by Estrada et al. (2006): where Y is the element of interest ( 13 C or 15 N), z is the atomic mass of the element, and n is the number of growth bands. The relative trophic position (TP) was estimated using the following equation: where TP base is the trophic position of baseline species, δ 15 N lim is the saturating isotope value, k represents the rate at which δ 15 N TP approaches δ 15 N lim , and δ 15 N TP is the δ 15 N value of the shark.
The signatures of the zooplankton (δ 13 C = −20.1 ± 0.7‰ SD and δ 15 N = 5.3 ± 0.8‰ SD; Estupiñán-Montaño et al., 2019) sampled in the study area were used as an isotopic baseline with TP base = 2. The δ 15 N lim and k values of 21.93 and 0.14, respectively, were derived from a meta-analysis of experimental isotope data .
The niche width and isotopic overlap between individuals, sexes, and growth stages were estimated using the Stable Isotope Bayesian Ellipses method in R, with analysis using ellipses, calculated δ 13 C( ‱ ) = 13 C∕ 12 C sample 13 C∕ 12 C standard − 1 × 1000

| RE SULTS
A total of 181 samples of vertebral collagen were obtained from 25 individuals (14 males, 101 vertebral sections; 11 females, 80 vertebral sections). After removing residual inorganic carbon, the range of the C/N ratio was 2.7 ~ 3.5, indicating that the demineralization was sufficient (Table 2).
There was no difference in δ 15 N values among growth stages (for males and females combined: ANOVA, p = .063, for male and female separately: Tukey's test, P male = 0.696 and P female = 0.832, between sexes: paired t-test, p = .695).
The mean trophic position of C. longimanus was estimated to be 3.7 ± 0.1 with a range of 2.8-4.7 ( Table 2). The TP estimated by age suggested similar TP (p > .05) throughout ontogeny. And the estimated TP for males and females was also similar at each growth stage. The estimated TPs for the growth stages of males (3.7 ± 0.1) and females (3.5 ± 0.1) showed no significant differences (p > .05).

| Isotopic niche
The overall estimated SEA B was 6.01‰ 2 , suggesting that C. longimanus has a broad isotopic niche. A broad isotopic niche was F I G U R E 2 Photograph of a vertebral section of a female Carcharhinus longimanus, estimated to be 16 years old, at 235 cm total length estimated for both sexes (for the male, SEA B was 5.89‰ 2 , and for the female, SEA B was 5.62‰ 2 ). In every growth stage, the niche width of males was higher than that of females, except for 1-and 4-year-olds (  Figure 5).

| DISCUSS ION
This study was the first attempt to use the stable isotope ratios of vertebrae to reveal the long-term trophic patterns of C. longimanus and improved our understanding of its dietary ontogeny, trophic positions, and isotopic niche width.

| Habitat use and trophic position
Variance in isotopic compositions from vertebrae reflects the integration of metabolism, growth, protein composition, and feeding  intraspecific variation in dietary habits and sources. Generally, 12 C of aqueous CO 2 is preferentially uptake by phytoplankton during photosynthesis, resulting in the enrichment of 13 C. Higher δ 13 C values, therefore, can be observed in waters with high phytoplankton productivity or warmer water closer to the equator (Graham et al., 2010). Major marine and marginal marine habitat types had distinct δ 13 C values and could provide different basic carbon sources. Such variation can be transferred along the food chain and reflected in the tissues of top predators (Hobson et al., 1994). It was reported that the C. longimanus typically lives offshore, along the edges of continental shelves or around oceanic islands in water deeper than 184 m, and from the surface to at least 152 m depth (Backus et al., 1956;Bonfil et al., 2008). And C. longimanus was considered to be an oceanic species and was able to undertake long-distance migration and vertical movement throughout its entire lifetime (Madigan et al., 2015). It was also possible that this species regularly migrated to breeding grounds as a foraging strategy. Although the species spent most of its time in less than 200-m depth, it could conduct deep dives into the mesopelagic zone for a short period, appearing to tolerate colder waters (Howey et al., 2016;Tolotti et al., 2017). Meanwhile, small C. longimanus have been observed to inhabit deep reef areas along the continental shelf where upwelling processes lead to enhanced productivity (i.e., high δ 13 C values) (Seki et al., 1998).
In addition, Howey-Jordan et al. (2013) reported that C. longimanus depart from habitats associated with reproduction or take roundtrip between favorable expansion of warm water and more northerly latitude areas. However, there was currently no definitive evidence of parturition or mating occurred (e.g., the presence of neonates or juveniles, or observations of sharks with physical F I G U R E 3 δ 13 C and δ 15 N values obtained from vertebrae of Carcharhinus longimanus sampled (n = 25). (a) Individual patterns, (b) Mean ± SE signs of mating) in the Pacific (Madigan et al., 2015). Such complex movement patterns might contribute to the wide range of δ 13 C values. Differences observed in δ 13 C values between males and females at each growth stage may indicate the sexual dimorphism and site segregation (Canani & Oddone, 2020). However, isotopic discrimination at higher trophic positions also could be driven by physiological processes (Madigan et al., 2021). This might be also the reason for the difference between sexes in δ 13 C, but further investigation is needed.
In this study, we found a broad range of δ 15 N values in C. longimanus, possibly driven by maternal transfer or ontogenetic changes in diets (Kiszka et al., 2014). The consumption of prey at different trophic positions could be reflected in the vertebrae. In addition, the wide range of δ 15 N values could also be attributed to environmental factors (Lin, 2013 (Palacios, 2002).
Typically, opportunistic predators with long-distance feeding migration behaviors have wide ranges of δ 15 N values (Howey et al. 2016;Madigan et al., 2015;Young & Carlson, 2020). Moreover, individual physiological and biochemical factors might also contribute to the isotopic variation (Shipley & Matich, 2020). Boggs et al. (2016) have confirmed nutritional condition and reproduction could affect isotopic fractionation. Both the males and females showed similar δ 15 N values at each growth stage. This pattern indicated similar feeding strategies in different habitats, supported by the differences in the δ 13 C values between sexes at each growth stage.
The estimated TP of C. longimanus is 3.7 ± 0.1 across all growth stages. It was slightly lower than the previous studies using stomach content analysis (TP mean = 4.2; Cortés, 1999), but was similar with the bulk stable isotope analysis using white muscles (TP mean = 3.92 ± 0.25; Li et al., 2014). It was worth noting that unidentified turnover rates and the enrichment mechanism of stable isotopes in vertebrae could affect the result of isotopic values and further investigation is required. Although there was a wide range across the studied period, the males and females in this study had similar TP among all growth stages.

| Dietary ontogeny
Several studies had reported ontogenetic changes in the habitat or diet of pelagic sharks (Kim et al., 2012;Li et al., 2016;Young et al., 2010). Such patterns also occurred in our study with a high degree of individual specialization within the population.  (Howey-Jordan et al., 2013;Young & Carlson, 2020).
In addition, maternal transfer might also affect δ 13 C values in early life. As aplacental viviparous sharks, female C. longimanus trans- Similar trends of δ 13 C values in this species were also found in Atlantic (Madigan et al., 2015).
Though variable shifts in δ 15 N values along the vertebrae for both sexes were observed, no discernible trend was detected.
Essentially, the δ 15 N values of 1-year-old were similar to the values of adult. This might reflect the maternal isotopic signature acquired through maternal transfer (McMeans et al., 2009). Moreover, higher δ 15 N values were found in 7-or 8-year-old, which might be explained by a diet toward large size fish and squid (Cortés, 1999). This hypothesis was supported by our isotopic enrichment analysis in which the enrichment in 15 N was observed in adults, especially in the females.
The result of isotopic enrichment analysis suggested that compared with females, male sharks have lower energy requirement change with growth (Papastamatiou et al., 2018). In addition, the enrichment of δ 15 N values also could be affected by environmental factors (Lin, 2013). Oceanic sharks are believed highly migratory, undertake ontogenetic migration among different habitats within 1 year (Howey-Jordan et al., 2013). Alternatively, in the oligotrophic open ocean, sharks exhibit opportunistic feeding in their life span. Thus, due to these dilution effects, the dietary ontogeny of C. longimanus may be even more pronounced than the results suggest (Madigan et al., 2015).
The isotopic enrichment of 25 individuals found in this study suggested that the maternal δ 13 C signatures were "erased" by the offspring at 5 years old, and the maternal δ 15 N values were "erased" at 5 years old (males) or 6 years old (females). These findings indi- It should be noted that variance in isotopic compositions from bulk tissue (e.g., vertebra) may reflect mixed effects, such as primary producers, trophic position, and metabolic routing (Magozzi et al., 2021). Compound-specific stable isotope analysis of amino acids (CSIA-AA) has been increasingly employed in trophic ecology research (Magozzi et al., 2021;McMahon et al., 2015), since it can reduce uncertainty in estimates of change in location, diet, and nutrient source. Future studies are needed to confirm the results of this study using CSIA-AA.

| Niche width and overlap
The trophic niche width is associated with diversity and availability of food resources which were considered to be the most important factors in shaping trophic niche (Bearhop et al., 2004;Páez-Rosas et al., 2020). The broad SEA B values suggested diversity of food sources and habitat utilization. The difference exhibited by males and females suggested their complex foraging and migratory behavior. The expansion on isotopic niche of large pelagic sharks due to the opportunistic feeding was reported by Vaudo and Heithaus (2011). Larger niche width of females across growth stages indicated their presence in higher productive areas with abundant resource than the males and possibly move horizontally and vertically within the greater range during growth of the same overall ecosystem. The isotopic niche overlap can reflect the similarity in resource utilization and potential competition among individuals. And the high degree of isotopic overlap between sexes at each growth stages in this study suggested both sexes similar prey species.

F I G U R E 5
The niche overlap between sexes and among growth stages of Carcharhinus longimanus in the central and eastern tropical Pacific. The ellipses represent the estimated standard ellipse area determined by the SIBER analysis

| CON CLUS ION
The SIA of oceanic whitetip shark (C. longimanus) vertebrae provided lifetime records of diet and revealed feeding patterns. In this study, similar trends in δ 13 C and δ 15 N values of both sexes were observed, and a high degree of isotopic overlap was found among its entire life histories. Stable isotopic data suggested that C. longimanus at different stages share similar feeding strategies and both sexes have similar trophic positions in the study area. Moreover, we found that C. longimanus might occupy different habitats between sexes due to their migratory behaviors. However, it remains unknown whether this sexual dimorphism is prevalent across its geographical range or what environmental conditions trigger sexual segregation in habitat. Further research is needed to understand differences in physiological response and habitat use between males and females.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest. Writing -review & editing (equal).

DATA AVA I L A B I L I T Y S TAT E M E N T
Data used in this manuscript were submitted to Dryad and are preliminarily available at https://doi.org/10.5061/dryad.pvmcv dnm6.