Introduction

Elasmobranchs (sharks and batoids) are important top and mesopredators in marine ecosystems, being considered keystone species because of their ecological role in marine food webs (Baum and Worm 2009; Cortes 1999; Wetherbee and Cortés 2004). However, population declines of these species have been documented worldwide due to several human impacts such as fisheries or habitat degradation, having impacts on the functioning and structure of marine ecosystems (Baum 2003; Dulvy et al. 2014, 2008; Ferretti et al. 2010; Stevens 2000). For example, the decrease of large-sized sharks decreases the mortality of their main prey, contributing to variations in the abundance of other less competitive mesopredators (Ferretti et al. 2010) or affecting the numbers of primary consumers (Ruppert et al. 2013). This global situation leads to an international focus on the conservation of their populations, given the importance they have at the ecosystem level (Dulvy et al. 2014; Ferretti et al. 2010; Moore and Grubbs 2019; Myers et al. 2007; Worm et al. 2013).

A key step in the management and conservation of elasmobranch populations is to understand their trophic niche and trophic interactions (Barría et al. 2015; Navia et al. 2017; Papastamatiou et al. 2006; Vidal et al. 2023). This information is required to predict the potential consequences of natural or human-related changes in the availability of particular prey for a particular elasmobranch species (Bascompte et al. 2005; Coll et al. 2013; Dulvy et al. 2014; Simpfendorfer et al. 2001). For example, knowing the ecological interactions among ecologically similar elasmobranchs is fundamental for understanding how these species could coexist in the same ecosystem (Albo-Puigserver et al. 2015; Navia et al. 2017). Furthermore, understanding the trophic interactions amongst elasmobranch species could also allow for the pairwise comparison of species as ecologically similar or distinct (Matich et al. 2011). Based on the principle of competitive exclusion, predators occupying equivalent ecological niches are expected to be segregated (Pianka 2000). However, related species do occasionally coexist in the same ecosystems if there is no limitation on prey availability (e.g. Albo-Puigserver et al. 2015; Navia et al. 2017). Trophic niche partitioning by prey or habitat has been indicated as the central mechanism explaining the presence of sharks and batoids in the same ecosystem (Albo-Puigserver et al. 2015; Heithaus et al. 2013; Papastamatiou et al. 2006; Platell et al. 1998). Similarly, the degree of trophic specialization (i.e. trophic niche width) at the species level could affect the capability of this species to cope with changes in the main consumed prey, being the specialist potentially more vulnerable to anthropogenic pressures (Barbini et al. 2020; Coll et al. 2013; Gallagher et al. 2015; Isaac and Cowlishaw 2004).

The Gulf of Oman and the Persian Gulf, two adjacent productive areas of the northern part of the Arabian Sea, are considered global hotspots of marine elasmobranch diversity (Jabado et al. 2015a, b; Valinassab et al. 2006). The Arabian Sea hosts around 153 species (Jabado et al. 2015a, b) including some worldwide iconic species of the order Rhinopristiformes (sawfishes, guitarfishes, wedgefishes, and banjo rays) or endemic species such the oceallate eagle ray (Aetomylaeus milvus), the smoothtooth blacktip shark (Carcharhinus leiodon) or the Arabian carpetshark (Chiloscyllium arabicum) (Hsu et al. 2022; Jabado 2018; Jabado et al. 2017). Despite the high conservation value of these marine areas, many elasmobranchs inhabiting the Arabian Sea seem to be declining associated with a high fishing pressure focused on obtaining meat and fins (Hsu et al. 2022; Jabado and Spaet 2017). This intense fishing pressure in combination with other threats associated with marine pollution or habitat alterations has resulted in the population declines of most of these predators in this area (Jabado et al. 2017). From a conservation point of view, 63% of the elasmobranchs present in this area are classified as threatened species by the IUCN (International Union for the Conservation of Nature) (Hsu et al. 2022; Jabado et al. 2017). However, the research conducted on these marine predators in the Persian Gulf and the Gulf of Oman, and in general along the Arabian Sea, is very scarce (Jabado et al. 2017; Jabado et al. 2015a, b; Randall 1995; Rastgoo et al. 2018). Therefore, there is a necessity to increase the biological and ecological information of these predators inhabiting this basin, including a better understanding of their trophic habits, trophic interactions, and levels of trophic redundancy among them.

The assessment of the trophic ecology of elasmobranchs is complex, because these marine predators include diverse ecological guilds with considerable variation in diet and trophic positions both within and between species (Barría et al. 2015; Cortes 1999; Hussey et al. 2010; Kiszka et al. 2014; Raoult et al. 2015; Rastgoo et al. 2018). Also, the high frequency of empty stomachs observed in most elasmobranchs (Ehemann et al. 2019) entails a great limitation when analyzing the diet of these species by using traditional stomach content analyses (SCA) (Hyslop 1980). Moreover, SCA requires a relatively high number of individuals to quantify the diet, which could be a logistic constraint for threatened or rare species (Cortes 1999; Hyslop 1980; Navarro et al. 2014). As alternative or complementary to SCA, the analysis of the stable isotopic of carbon (δ13C) and nitrogen (δ15N) values, and the utilization of different isotopic metrics have been used to investigate the trophic ecology of elasmobranchs (Bird et al. 2018; Shiffman et al. 2012). Both δ13C and δ15N together provide quantitative information into the trophic niche of a particular consumer, reflecting what it assimilates from its diet and the habitat in which it forages (Bearhop et al. 2004; Newsome et al. 2007). These biomarkers are useful for investigating niche partitioning and trophic relationships between sympatric species (Gaspar et al. 2022; Giménez et al. 2018), and have been used in elasmobranchs too (Albo-Puigserver et al. 2015; Barría et al. 2015; Kiszka et al. 2014; Shiffman et al. 2019). Therefore, stable isotope approaches may provide, therefore, the necessary information to quantify the trophic niche overlap/segregation between sympatric species and identify those drivers of trophic segregation within species.

In the present study, our main aims were to assess the trophic relationships, trophic niche breadth, and trophic redundancy among 21 batoids and 12 sharks inhabiting the Iranian waters of the Persian Gulf and Gulf of Oman. All these 33 elasmobranchs are poorly studied and commonly caught in artisanal fisheries operating in the Arabian Sea and based on the regional assessments of the IUCN, 9 of the sampled species are considered Endangered, 9 Vulnerable, 12 Near Threatened, and 1 Data Deficient (Table 1). We hypothesized that a certain niche partitioning must be occurring within this great number of species and within the basin in order to coexist. Our results will provide novel insights into the trophic relationships between different elasmobranch species and it will contribute to a better understanding of the ecological role of these elasmobranchs.

Table 1 Sample size (n), mean ± SD of δ15N and δ13C values, and mean of Bayesian Standard Ellipse Areas (SEAB; mean, minimum and maximums) and Bayesian trophic level (TL) for 33 elasmobranch species (grouped by Order and Family) sampled in the Persian Gulf and Gulf of Oman (north of Arabian Sea) during 2016–2017

Methodology

Fieldwork procedures

We conducted this study in the Iranian waters of the Persian Gulf and Gulf of Oman. In total, 278 specimens of 33 elasmobranch species were opportunistically collected from the bycatch generated by fisheries in 2017. All individuals were collected from fishing vessels of two fishing ports: Bandar Abbas—latitude = 27.17°, longitude = 56.26°, and Qeshm–latitude = 26.98°, longitude = 56.26°). Because both fishing ports were very close, we assumed that all the individuals were collected in a similar marine area, reducing the potential effect of spatial variability on the stable isotopic values (Kjeldgaard et al. 2021; Post 2002). After collection, each individual was identified at the species level, and the sex, body length (in the case of sharks), and disk width (in the case of batoids) were recorded (± 1 cm). From each collected individual, a sample of muscle was collected and frozen at − 20°C. The sampling included 21 batoids from 3 orders and 12 families and 12 sharks from 2 orders and 6 families (Table 1).

Stable isotopic analysis

Prior to the stable isotope determination, urea content from the muscle samples was removed to prevent potential bias in the estimation of the stable isotopic values (see Kim and Koch 2012; Shiffman et al. 2012). To extract urea, all muscle samples were placed in 2 ml vials with 1.5 mL of deionized water in a sonication machine for 15 min to remove the supernatant of urea. After this process, samples were dried, powdered, and between 0.28 and 0.33 mg was packed into tin capsules. All stable isotope determinations were conducted at the Stable Isotopes Facilities (LIE) of the Estación Biológica de Doñana CSIC (http://www.ebd.csic.es/lie/index.html). Samples were combusted at 1020 °C using a continuous flow isotope ratio mass spectrometry system (Thermo Fisher Scientific, Bremen, German) by means of a Flash HT Plus elemental analyzer interfaced with a Delta V Advantage mass spectrometer. Replicate assays of standards routinely inserted within the sampling sequence indicated analytical measurement errors of + 0.1‰ and + 0.2‰ for δ13C and δ15N, respectively. The standards used were: EBD-23 (cow horn, internal standard), LIE-BB (whale baleen, internal standard) and LIE-PA (feathers of Razorbill, internal standard). These laboratory standards were previously calibrated with international standards supplied by the International Atomic Energy Agency (IAEA, Vienna). Stable isotope ratios were expressed in the standard δ-notation (‰) relative to Vienna Pee Dee Belemnite VPDB for δ13C values and atmospheric N2 (Air) for δ15N values. Based on laboratory standards, the precision (± standard deviation) was + 0.1 and + 0.2 for δ13C and δ15N, respectively. As the ratio between the percentage of carbon and the percentage of nitrogen was lower than 3.3 ‰ for all samples, no lipid correction on the δ13C values was applied (Logan et al. 2008; Post et al. 2007).

Isotopic metrics and statistical analyses

Standard Ellipses Area (SEAc, i.e. area containing the 40% of the isotopic data) and Bayesian Standard Ellipse Areas (SEAB) were computed with the SIBER package (Jackson et al. 2011) in the open-source program R v.3.5.2 (http://cran.r-project.org/) to estimate the niche width and niche overlap between species (Jackson et al. 2011). The Bayesian framework takes into account the uncertainty in the data and incorporates the error arising from the sampling process, propagating it through to the derived metric (Jackson et al. 2011). Furthermore, the estimation via Bayesian inference allows robust comparison between species with uneven sample sizes (Jackson et al. 2011). SEAB areas were computed using 10,000 posterior draws. A total of 10,000 posterior draws were generated using a sampling–importance–resampling (SIR) algorithm (Jackson et al. 2011) to calculate SEAB.

δ13C and δ15N values were compared between species and families by using ANOVA and post hoc Tukey tests, only with species and families of more than five sampled individuals. We also explored the relationships between isotopic values and body size (total length and disk width; a proxy of ontogenetic effect) by linear regressions. Due to the reduce number of individual samples for most of the species, we only evaluate the relationships between body size and stable isotopes for the species that have more than ten individuals.

Identification of the potential diet of species with no published data

We revised the published information of the diet based on SCA of all collected elasmobranchs. For the species with no published diet data, we assigned its potential diet based on the knowing diet of isotopic-similar species within the same taxonomic family. That is, we assumed that if two species of sharks or batoids showed similar stable isotope values within the same family, their diet might be similar.

Results

Stable isotopic differences between species and taxonomic orders

δ15N values differed among families (ANOVA tests, F11, 260 = 20.91, p < 0.001; Fig. 1). In particular, Mobulidae and Torpedinidae showed the lowest and highest δ15N values, respectively (Fig. 3) and post hoc tests grouped the different families in seven groups (Fig. 1). Regarding δ13C values, we did not find differences among orders (F11, 260 = 1.48, p = 0.14; Table 1, Fig. 1).

Fig. 1
figure 1

A Mean and standard error of δ15N and δ13C values of 33 elasmobranch species sampled in the Persian Gulf and Gulf of Oman during 2017. Standard ellipse area (SEAC) of each family grouped by Orders (BE)

At inter-specific level, we found significant differences in δ15N (F32, 278 = 15.41, p < 0.0001) and δ13C values (F32, 278 = 1.08, p = 0.02) (Table 1, Figs. 1 and 2). Eight groups were defined according to the similar δ15N values, with the lowest and highest values shown in M. kuhlii and T. sinuspersici, respectively (Table 1, Figs. 1 and 2). In the case of δ13C values, only two groups were defined according to post hoc tests, with the lowest and highest values showed in M. randalli and H. leoparda, respectively (Table 1, Fig. 1).

Fig. 2
figure 2

Standard ellipse areas (SEAc) of each species by elasmobranch family sampled in the Persian Gulf and Gulf of Oman during 2016–2017. Only species with more than two individuals sampled are represented

Regarding the relationship between isotopic values and body length, we found that only the sharks C. arabicum and I. omanensis showed significant relationship between δ15N and their total body length (p < 0.05; Fig. 3). For the other species with more than ten sampled individuals we did not find significant relationships between δ15N and body length (all cases p > 0.05).

Fig. 3
figure 3

Relationships between the body length (anal length, cm) and δ15N of Iago omanensis and Chiloscyllium arabicum sampled in the Persian Gulf and Gulf of Oman

Standard ellipse areas

Regarding the isotopic niche, we found that C. macrostoma (mean [5% Credible Interval-95% Credible Interval], 0.87 [0.45–1.02]) presents the smallest isotopic niche, while S. lewini (14.79 [7.60–17.07]) is the largest within the shark species. For rays, the smallest isotopic niche is found in P. fai (0.30 [0.15–0.35]) and the largest in M. gerrardi (14.39 [8.58–17.04]) (Tables 1 and S1). By family, we found differences in the isotopic niche, with Narkidae (0.57 [0.43–0.79]), Hemigaleidae (0.68 [0.47–1.04]) or Mobulidae (1.49 [1.14–1.99]) with the lower niches, and Gymnuridae (10.49 [8.81–12.61]), Dasyatidae (10.64 [9.71–11.70]) or Sphyrnidae (11.05 [7.56–17.07]) with the higher niches (Fig. 2; Table 1).

Regarding the isotopic niche overlap, we found that the families of each order overlapped to a certain degree between them, in special for the families of the order Myliobatiformes (Table 2). The higher overlap was found between Myliobatidae and Aetobatidae (0.29 [9.71–11.70]). The lowest overlap between families was found within the orders Rhinopristiformes and Carcharhiniformes (Table 2). Nevertheless, when analyzed each species individually, some of them were segregated from the rest of the species of the same family, such as H. leoparda and M. gerrardi for Dasyatidae, or A. milvus and A. ocellatus for Aetobatidae (Fig. 2; Table S1). Others, such R. oligolinx and R. acutus for Carcharhinidae showed high overlap (Fig. 2; Table S1).

Table 2 Median, 5% and 95% Credible Interval of the overlap between the Bayesian Standard Ellipse Areas (SEAB) of each elasmobranch family within each order from the Persian Gulf and Gulf of Oman (north of Arabian Sea) during 2016–2017

Niche breath was dispersed in both axes (δ13C and δ15N) such as in M. gerrardi or G. poecilura but some exceptions were present in this elasmobranch community where the dispersion is more pronounced in one of the isotopic axes. For example specially for δ13C, M. randalli, P. sephen, C. dussumieri, C. macloti and C. arabicum present a larger dispersion indicating greater variability in the assimilation of prey living in dissimilar habitats. On the other hand, for δ15N, A. milvus, A. nichofii, H. leoparda, R. laevis, G. granulatus and C. macrostoma present a larger dispersion suggesting the assimilation of prey of diverse trophic levels (see Fig. 2).

Assignation of the potential diet based on SIA similarity

We found 11 species without any published data on their diet (Table 1). After comparing the stable isotope values of these 11 species with other sharks and batoids with published diet data, we assigned a potential diet for these species. In particular, we assigned a diet composed mainly by crustaceans for A. milvus and M. gerrardi, by fish for C. macloti, and by fish and crustaceans for H. leoparda (Table 1) (Fig. 2).

Discussion

In this study, we present new ecological data regarding the trophic interactions and isotopic niches of 33 species of uncommon and threatened elasmobranchs inhabiting a data-poor region of the Persian Gulf and Gulf of Oman. In comparison with the accurate information provided with stomach content or DNA barcoding methodologies (Davis and Pineda-Munoz 2016), stable isotope values only allow to estimate the isotopic niche (as a proxy of the trophic niche) and trophic relationships among species (Layman et al. 2012). Nonetheless, as stable isotope values in muscle reflect the diet of elasmobranchs during the 9–10 months before sampling (Shiffman et al. 2012), this approach informs us about the trophic habits in the long-term (Navarro et al. 2014).

Stable isotope results revealed differences in the trophic niche among the elasmobranch sampled at two levels: at family level and at the species level within each family. Regarding the differences between families, we found differences in their isotopic position (indicated by mean δ15N values) and their isotopic niche width. For example, the batoids of the family Torpedinidae and the sharks of the order Carcharhinidae and Triakidae occupied higher trophic position and the sharks of the family Narkidae showed the lower trophic niche width, whereas the batoids of the family Mobulidae showed lower trophic position but a high trophic niche width. These differences reflected the inter-specific differences in the stable isotope values and the trophic habits described by the species of each family.

In marine environments, the consumers that include prey occupying high trophic levels in the ecosystem were expected to have high δ15N values (Bird et al. 2018; Shiffman et al. 2012). This is the case of the elasmobranchs of the family Carcharhinidae or the Torpedinidae, with a diet composed mainly by fish, with some of them preying also on other shark species (Jabado et al. 2015a, b; Raeisi et al. 2015; Randall 1995; Rastgoo et al. 2019, 2018). In contrast, other species that prey on lower trophic levels, such the Myliobatiformes M. kulhii or A. ocellatus with a diet composed by gastropods and bivalves (Randall 1995). In the case of M. kulhii, although its diet has not been well-studied (Barbato et al. 2019), dietary research on similar mobulids has described the group as mostly planktivorous, feeding on various zooplankton and small fish (Couturier et al. 2012).

In the case of the trophic diversity at family level, we found that two families of the Order Myliobatiformes, exhibit a great isotopic diversity, indicating that the different species included in this group, exploit a wider range of resources. For example, some of these batoids such A. flagellum, A. milvus, or A, nichofii mainly prey on crustaceans, others such U. asperrimus and R. javanica prey on polychaetes and bivalves and others such G. poecilura, H. leoparda or H. uarnak prey of fish (Carassou et al. 2017; Randall 1995; Rastgoo et al. 2018). In contrast, the different species included in the families order Carcharhinidae and Hemiscyliidae showed diet mainly composed by fish and bivalves, respectively (Randall 1995).

Unexpectedly, only two sharks, C. arabicum and I. omanensis, presented positive and significant relationships between δ15N and their total body length. This probably was due to the small range of body lengths sampled for the rest of the species due to the opportunistic nature of the sampling. For this reason, a reassessment of these relationships with the other elasmobranch species should be carried out if smaller and larger individuals will be obtained. In any case, these relationships could be reflecting an ontogenetic change in the trophic level of the prey consumed by these two sharks. In particular, the increase of δ15N values in larger individuals could be indicating the consumption of large-size prey and/or an increase in the quantity of fish or other prey such sea snakes or cephalopods (Liu et al. 2020; Rastgoo et al. 2019; Waller and Baranes 1994).

We have provided some insights into the ecology of poorly known species. This is especially important for those that are classified endangered by the IUCN at global or regional scale. The comparison of their stable isotope values to other more studied species provides us with some but limited information, as isotopic similarity not always means trophic similarity as different prey can have similar isotopic values. On the other hand, when presenting dissimilar isotopic values, we are more certain that there are feeding on different prey. Nevertheless, higher taxonomical precision is needed for a proper conservation strategy for those endangered elasmobranch species. We encourage performing stomach content analysis or DNA metabarcoding using bycatch individuals to gain more precise information on their diet.

In conclusion, in the present study, we provide information relating to the trophic preferences, trophic interactions, and levels of trophic redundancy among batoids and sharks inhabiting the Persian Gulf and Gulf of Oman, highly useful to future predictions on the consequences of changes in their abundance due to natural or human-related factors.