As impacts on the Mediterranean Sea are expected to grow in the future, especially with climate change, habitat degradation, and displacement of native species by non-indigenous species (NIS), the investigation of significant alterations to trophic levels in this diverse marine habitat is important. Analysis of stable isotopes from targeted consumers has previously been shown to reliably reflect that of primary producers, thus enabling us to describe and highlight potential shifts in the food web of a particular ecosystem. In this study, we used δ¹³C values of essential amino acids (AA) in order to examine the dietary composition of established non-native, Lessepsian fish migrants in the Eastern Mediterranean Sea compared to that of the same fish species from their original population in the Gulf of Aqaba, Red Sea. Our data show that a clear variance in carbon isotopic signatures exists in food sources consumed by the same species between the different environments, with the exception of the classic herbivore, Siganus rivulatus (Forsskål & Niebuhr, 1775), whose very similar isotopic patterns reflect the algal source they predominantly consume in both locations. With the results of this research, we propose that Lessepsian fishes with the ability to maintain their nutritional patterns, though not necessarily that of their original food source, will acclimatize better in their new habitat. Consequences of flourishing Lessepsian fish populations include a further tropicalization of the Eastern Mediterranean Sea and the likely restructuring of local food webs.

Stable isotope, Biological Invasion, Climate Change, Sargocentron rubrum, Siganus rivulatus, Parupeneus forsskali, Pterois miles

The Mediterranean Sea is a semi-enclosed basin comprising a complex marine ecosystem. (Bethoux et al. 1999). Currently considered among the most impacted of the regional seas, the Mediterranean basin has become susceptible to changes in its biodiversity through various anthropogenic pressures such as climate change, overfishing, habitat degradation, and non-indigenous species (NIS) introductions (Galil 2007; Golani and Bogorodsky 2010). With the opening of the Suez Canal in 1869, many hundreds of marine species (including over 100 fish species) entered the Mediterranean Sea from the Red Sea (Galil 2007). These invading species, called Lessepsian migrants, arrive through ship ballast water or directly via the Suez Canal (attached to ships, swimming, larval transport) (Padilla and Williams 2004; Zenetos et al. 2005; Bianchi 2007; Galil 2007).

Concomitant with an increase in water temperature over the past decades (Bethoux and Gentili 1999), this massive migration of species from an Indo-Pacific, Indian Ocean, or Red Sea origin over the past 150 years has contributed to a “tropicalization” of the Eastern Mediterranean Sea in particular (Bianchi 2007). The successful establishment of migrating species in a new environment often demands adaptations to core behaviors, essential for the production of new generations or recruitment, leading to a substantial population size and a steady period of reproduction. Given that Lessepsian fish species often constitute over 80% of the commercial catch in the Eastern Mediterranean as of today (Edelist et al. 2013; Öztürk 2021), the ability of these NIS to adapt to different food sources, habitats, and other inorganic environmental factors show the high possibility of their success and a further tropicalization in the Mediterranean Basin as a whole (Golani and Ben-Tuvia 1985; Arndt et al. 2018).

Previous conventional approaches for tracking animal movements have largely relied on extrinsic markers and on stomach content analysis (Rindorf and Lewy 2004; McMahon and Newsome 2019). However, the use of contemporary methods such as molecular DNA profiles and the measurement of naturally occurring stable isotopes in animal tissues offer more reliable and distinct analyses for determining which adaptations have enabled migratory species to thrive in a new environment (Fanelli et al. 2015). Amino acid compound-specific stable isotope analysis (AA-CSIA) has shown to be a powerful tool for examining diet and habitat use, providing key information regarding the isotopic signature of primary producers at the base of a food web (McMahon et al. 2010, 2016).

In this study, we compare the isotopic signature of five essential amino acid (AA) δ¹³C values of two key Lessepsian species Sargocentron rubrum (Forsskål, 1775), and Siganus rivulatus (Forsskål & Niebuhr, 1775) that entered the Mediterranean a few years after the opening of Suez Canal (Haas and Steinitz 1947) and two relatively “young” immigrants Red Sea goatfish - Parupeneus forsskali (Fourmanoir & Guézé, 1976) and Devil firefish Pterois miles (Bennett, 1828) (Bariche et al. 2013; Sonin et al. 2013) to baseline isotopic compositions presented by the same species in their native habitat (the Indo-Pacific waters). P. forsskali was chosen as a species that outcompeted the native fish species from the same family. P. miles was chosen as a species that in the last two decades introduced to many new places all over the world (e.g., the Caribbean and South Atlantic Ocean as in Brazil), therefore six specimens were also collected in Dominica Caribbean. Comparing trophic position indicators of successful invasive species in their native and invaded range can provide insights into behavioral shifts, such as dietary habits, and help us understand how adaptations to environmental differences may shape food web structure and contribute to current ecosystem functioning in the Eastern Mediterranean Sea. The aim of this research was to study the dietary habits change of “old” and “new” Lessepsian migrants’ fish from the Red Sea to the Eastern Mediterranean Sea and the establishment of those species.

Isoleucine, methionine, and phenylalanine Amino acid δ¹³C isotopic ratio analysis was performed on four representative fish species, P. forsskali, S. rubrum, P. miles, S. rivulatus and on green algae (Ulva sp.) from two locations – the Israeli coast of the Mediterranean Sea and the Gulf of Aqaba, in the Red Sea. Statistics show a clear and significant separation exists between the Red and Mediterranean Sea sampling groups, S. rivulatus and green algae (Ulva sp.) showed no difference between the Red and the Mediterranean Sea (see also Suppl. material 1: tables S1–S4).

The highest difference in stable isotope signatures was found between P. forsskali individuals caught in the Eastern Mediterranean Sea (Med. n = 9, Red. n = 4) basin and those originating in the Red Sea (R = 0.979, p = 0.001), as seen in the clear separation of groups in Figure 1. Similarly, S. rubrum individuals (Med. n = 7, Red. n = 7) also showed a significant difference (R = 0.848, p = 0.001) between the two locations (Figure 2).

Figure 1. 

PCA of the δ13C of five essential AAs (valine, leucine, isoleucine, methionine, and phenylalanine) of Parupeneus forsskali from the Red Sea and the Mediterranean Sea indicates a significant difference in the isotopic signature (ONE-WAY ANOSIM R = 0.979 p = 0.001). Eigenules (%) variation PC1 = 73.4 PC2 = 19.7

Figure 2. 

PCA of the δ13C of five essential AAs (valine, leucine, isoleucine, methionine, and phenylalanine) of Sargocentron rubrum from the Red Sea and the Mediterranean Sea indicates a significant difference in the isotopic signature (ONE WAY ANOSIM R = 0.848 p = 0.001). Eigenvalues (%) variation PC1 = 72.5 PC2 = 24.9.

Representing one of the more recent NIS arrivals to the Mediterranean Sea, PCA of samples collected from P. miles (Figure 3) show differences in isotopic makeup between the locations (Med. n = 6, Red. n = 6, Carib. n = 6), though with some overlap (R = 0.487, p = 0.001).

Figure 3. 

PCA of the δ13C of five essential AAs (valine, leucine, isoleucine, methionine, and phenylalanine) of Pterois miles from the Red Sea and the Mediterranean Sea indicates a significant difference in the isotopic signature (ONE WAY ANOSIM R = 0.487 p = 0.001). Pterois miles from the Mediterranean Sea and the Caribbean (R = 0.378 p = 0.01). Pterois miles (Bennett, 1828) from the Red Sea and the Caribbean (R = 0.904 p = 0.002). Eigenvalues (%) variation PC1 = 77.6 PC2 = 11.2.

Lastly, very little isotopic change (R = 0.053, p = 0.144) was observed in δ13C of S. rivulatus populations from the Red Sea compared with those sampled from the Eastern Mediterranean sites (Med. n = 15, Red. n = 13) (Figure 4), suggesting close similarity in carbon source. Furthermore, within the PCA of the two Ulva sp. groups, we see a dominant representation of AA isotope signatures. (R = 0.168, p = 0.073) (Figure 5), also collected from both locations.

Figure 4. 

PCA of the δ13C of five essential AAs (valine, leucine, isoleucine, methionine, and phenylalanine) of Siganus rivulatus from the Red Sea and the Mediterranean Sea indicates a significant difference in the isotopic signature (ONE WAY ANOSIM R = 0.053 p = 0.144). Eigenvalues (%) variation PC1 = 72.4 PC2 = 17.4.

Figure 5. 

PCA of the δ13C of five essential AAs (valine, leucine, isoleucine, methionine, and phenylalanine) of green algae Ulva sp. from the Red Sea and the Mediterranean Sea indicates a significant difference in the isotopic signature (ONE WAY ANOSIM R = 0.168 p = 0.073). Eigenvalues (%) variation PC1 = 86.3 PC2 = 8.2.

The Eastern Mediterranean Sea and the northernmost tropical Red Sea host a variety of differences between their environments, which include varying temperatures, salinity, habitat structure, as well as bio-community abundance and diversity. Despite these differences, which are themselves on a course of rapid change, a growing number of non-native fish species have exploited local resources and niches to establish successful populations in the Levant basin (Coll et al. 2010) and see also primary results. In the Suppl. material 1: table S1. In assessing any change in trophic level position driven by NIS, analysis of δ¹³C values of essential AAs from targeted consumers (here, sampled fishes) reliably reflect that of primary producers, thus enabling us to describe and highlight potential shifts in the food web of a particular ecosystem.

When foragers from the same species exploit resources across numerous locations, a different isotopic signature is expected. Isotopic analysis in our study indeed revealed significantly different food sources among the following Lessepsian migrants and their original Red Sea population - the squirrelfish, S. rubrum; the common Red Sea goatfish, P. forsskali; and the lionfish P. miles.

Our data support observations reported by Golani et al. (1983), describing S. rubrum as a “no preference” bottom feeder, with stomach contents mainly (90%) consisting of Eastern Mediterranean crustacean species, despite abundant access to Lessepsian invertebrates of Red Sea origin. This species has also been observed to exploit under-used niches in deeper water, thus opportunistically capitalizing on new habitats and feeding grounds (Golani and Ben-Tuvia 1985; Farrag et al. 2018). Similarly, though as a relatively new immigrant to the Eastern Mediterranean basin, reports on P. forsskali describe this species as a strong competitor for resources and space and shown to exhibit habitat flexibility in their newly invaded region (Arndt et al. 2018; Evagelopoulos et al. 2020).

Lastly, the swift and successful invasions of P. miles in diverse areas of the globe have been largely attributed to their generalist diet, along with rapid growth rates and high fecundity. Male and female P. miles are capable of reproducing year-round and the ovaries of mature females may contain up to 40,000 oocytes, with larger females containing even more oocytes (Gardner et al. 2015). Their growing presence in the Mediterranean indicates that lionfish are thriving and could potentially become a serious threat in the region, particularly in relation to the trophic disruption of local herbivores (Layman et al. 2014; Kleitou et al. 2021).

With the exception of the classic herbivore, S. rivulatus, whose very similar isotopic patterns can be attributed to the green algae they predominantly consume in both locations, it appears that successful Lessepsian species that manage to adopt a wider trophic niche breadth, through more “generalist” feeding habits (e.g., feed on a wide variety of food types), become better long-term colonizers (Golani 1998).

Diet selectivity of rabbitfish S. rivulatus has been well documented in the Red Sea, due to resource competition among the large population of herbivorous fishes (Lundberg and Golani 1995; Bariche 2006). With their immigration to the Mediterranean Sea close to a century ago, the rabbitfish enjoyed little niche competition from local herbivores, as well as increased access to their preferred food source, the green algae Ulva sp., albeit seasonally (Golani 1998; Pickholtz et al. 2018). Our findings show that these classic herbivores do not need to (drastically) change their nutritional habits, thus preserving the same trophic levels between their original habitat and invaded habitat. This possibly alleviates challenges in their ability to establish and maintain a thriving population as compared to many other species which do require a change in their feeding habits in order to survive and thrive.

Whether opportunistic foraging characteristics or relatively similar food sources, in the case of S. rivulatus, have contributed to the successful adaptation of certain Lessepsian migrants to the Eastern Mediterranean environment, we maintain that species who can effectively adapt their diet to a new habitat will survive and thus be able to establish a large and stable population there. Although declines in native species abundance and some extirpations have already been reported throughout the Eastern Mediterranean Sea. Native genotypes may be lost, food webs may be altered, and significant economic losses may occur (Kalogirou et al. 2012; Katsanevakis et al. 2014), many established Lessepsian fishes (among them, three are included in our study) have often replaced local species of the same trophic position. In addition, Mediterranean fisheries have also adapted to exploit a number of commercially valuable Lessepsian species, possibly off-setting direct impacts of competitive exclusion in these isotopic niches (Svanbäck and Bolnick 2007).

Nevertheless, the successful establishment of Lessepsian species, particularly in terms of competition and niche partitioning with native and other non-indigenous fish, will likely contribute to further tropicalization and homogenization of the Eastern Mediterranean Sea, and eventually the general Mediterranean basin. While the breadth of these environmental and economic impacts remains unclear, further research should be devoted to understanding the complex and dynamic invasion ecology of the Mediterranean basin in order to monitor these changes. We do not have the ability to prevent the entry of more species through the Suez Canal. Optimal management of fishing may help to mitigate and reduce the ecological and economic impact caused by those migrating species.

This research was funded by the Morris Kahn Marine Research Station and the Department of Marine Biology of the University of Haifa as part of Rami Tsadok’s Ph. D dissertation.

The authors’ contributions to the manuscript: Rami Tsadok – research conceptualization, sample design, methodology, investigation, and data collection, data analysis and interpretation, writing - original draft, writing - review & editing. Eli Shemesh - research conceptualization, methodology, and data collection. Stephane Martinez - data analysis, writing review. Debra Ramon - writing original draft, review & editing. Anat Tsemel - data analysis and interpretation. Ziv Zemah-Shamir - writing original draft, review & editing. Itai Kolski - review & editing. Dan Tcheronov - research conceptualization, methodology, writing - original draft, writing - review & editing.

We thank all the staff and students at the Morris Kahn Marine Research Station from the Department of Marine Biology, University of Haifa, Israel, for their assistance, especially Dr. Shai Einbinder, Dr. Dany Morick, Dr. Dalit Meron, Dr. Leigh Livne, Mr. Eran Rosen Mr. Hagai Nativ and Dr. Aviad Scheinin. We also thank Prof. Tali Mass from the Department of Marine Biology, University of Haifa, Israel, for her support on this paper. Two anonymous reviewers considerably helped improve our submission. We thank them and the Academic Editor: Jean Vitule, for their patient aid.