Rhopilema nomadica in the Mediterranean: Molecular Evidence for Migration and Insights into Its Proliferation

Abstract

Since it was first observed in Israel in the 1970s, and due to its subsequent negative impact on human activities, the nomad jellyfish Rhopilema nomadica has earned itself a spot on the list of the 100 Worst Invasive Alien Species in the Mediterranean. It was assumed to originate in the Red Sea, or in the Indo-Pacific region, but in the absence of additional reports of live specimens outside the Mediterranean, its origins have remained a mystery. Here, via molecular analysis, we present the first verified results of the existence of R. nomadica in the Western Indian Ocean. Moreover, using additional evidence from Cassiopea andromeda and R. nomadica, we propose that the construction of the Aswan High Dam may have led to the proliferation of R. nomadica in the Levantine Basin.

1. Introduction

The scyphomedusa Rhopilema nomadica was first reported in coastal waters of the southeastern Mediterranean Sea in the 1970s [1]. By the mid-1980s, large aggregations of this species appeared annually along the Israeli coast, mainly during the summer [2,3,4]. Thereafter, R. nomadica expanded north and westward into the central Mediterranean, appearing in Turkey [5], Greece [6], Malta [7], Tunisia [8], and more recently, Sardinia and Sicily [9]. Dense swarms of R. nomadica, often exceeding 10 individuals m⁻² (Figure 1) and encompassing many kilometers, cause environmental, economic, and social damage, reducing biodiversity [10], clogging water intake pipes of coastal infrastructure [11,12], causing damage to the fishing industry [13,14], and disrupting coastal recreation and tourism [15]. As a result of its expansion and the effects of the swarms, R. nomadica is considered one of the “100 worst invasive species” in the Mediterranean [10,16] and one of the most “impacting species” in European seas [17].

Rhopilema nomadica is listed among many other species that have been introduced to the Mediterranean; since opening in 1869, the Suez Canal has served as a corridor for introducing >400 multicellular species [18]. However, there are gaps in our understanding of its ecology, origin, and geographical expansion. Until recently, as there were no reports of R. nomadica in the Mediterranean Sea prior to the 1970s [1], it was assumed to have arrived around this time. However, a unique photograph taken in 1938 (Rakotsh, Figure 2), presented at a public exhibition in the city of Haifa, Israel, in 2002, shows a dense aggregation of R. nomadica in Haifa Port. Furthermore, despite it being considered a Lessepsian migrant, based on a specimen collected off Kamaran, Yemen, southern Red Sea, in 1939 (holotype, RMNH 7038) [1], there have been no additional reports of R. nomadica from the Red Sea since then. Only Berggren [19] and Tahera & Kazmi [20] have reported R. nomadica outside the Mediterranean, and the report by Tahera & Kazmi [20] was found to be a misidentification [21]. Divers and researchers in eastern Africa have recently reported the presence of “nomad medusae” with similar morphology to R. nomadica near Dar es Salaam, Tanzania (Figure 3) (e.g., https://seaunseen.com) [22], suggesting that the Western Indian Ocean may be the source of the Mediterranean population. However, this conjecture has yet to be tested.

Here, we provide evidence that the western Indian Ocean “nomad medusa” is R. nomadica. We then explore why it has only relatively recently established persistent annual swarms in the Mediterranean. We propose that R. nomadica medusae were introduced into the Mediterranean over the last century, possibly soon after the Suez Canal was constructed. Moreover, we suggest that its apparent appearance in the late 1970s was due to changing environmental conditions, favoring the success of its reproductive benthic stage [Box 1].

Box 1. Summary of the metagenic life cycle of Rhopilema nomadica.

Rhopilema nomadica has a metagenic life cycle composed of the free-swimming, planktonic medusa, and the benthic sessile polyp (Figure 4). Sexual reproduction occurs in the medusa stage and asexual reproduction occurs in the polyp stage. The medusa is gonochoric, releasing sperm or oocytes. Fertilization occurs in the water column. Fertilized eggs develop into free-swimming ciliated planula larvae that seek a suitable surface to settle on. After settlement, the planula metamorphoses into a sessile polyp called ‘scyphistoma’. The scyphistoma then begins to reproduce asexually through podocyst formation, with the potential to create dense populations of scyphistomae within weeks. Given suitable conditions (e.g., temperature, food), the scyphistoma starts to produce medusae through a process called ‘strobilation’. The success of the benthic/scyphistoma stage is undoubtedly responsible for the formation of the medusa swarms [23,24]. Nevertheless, for most species this stage is cryptic, and scyphistoma populations have rarely been observed in the wild.

Figure 4. Rhopilema nomadica life cycle (clockwise). Scale for each of the life stages: mature medusa, X = 10 cm; planula, X = 90 µm; young polyp, X = 100 µm; mature polyp, X = 700 µm; podocysts, X = 600 µm; strobila, X = 630 µm; ephyra, X = 800 µm; young medusa, X = 1.5 cm (sketch by Rahel Wachs).

Figure 4. Rhopilema nomadica life cycle (clockwise). Scale for each of the life stages: mature medusa, X = 10 cm; planula, X = 90 µm; young polyp, X = 100 µm; mature polyp, X = 700 µm; podocysts, X = 600 µm; strobila, X = 630 µm; ephyra, X = 800 µm; young medusa, X = 1.5 cm (sketch by Rahel Wachs).

2. Materials and Methods

DNA Sampling, Amplification, and Phylogenetic Analyses

In order to genetically confirm the identity and understand the phylogenetic position of R. nomadica from Tanzania (three medusae stranded on the beach in Dar es Salaam, Tanzania, 6°44′28.76″ S, 39°16′33.39″ E, sampled on 18 September 2016 by A. Sutton, a co-author based in Tanzania), a genus-level phylogeny was reconstructed. For the molecular analyses, 217 individuals of Rhopilema were included, of which three were the newly collected specimens from Tanzania (labeled RN1P, RN2P, and RN3P; GenBank accession numbers PX760500–2, respectively), and 214 available specimens were retrieved from GenBank (48 of R. esculentum, 50 of R. hispidum, 113 of R. nomadica, and three of R. verrilli). It is noteworthy that all the available sequences of R. nomadica were solely from the Mediterranean Sea, i.e., the newly generated Tanzanian sequences were the only ones from outside the Mediterranean region. Additionally, four sequences of the genus Rhizostoma (also from the family Rhizostomatidae) were retrieved from GenBank to root the phylogenetic trees (two sequences each for R. luteum and R. pulmo, from the northeast Atlantic Ocean and from Tunisia, respectively). Table S1 provides details on the identification, collection localities, and GenBank accession numbers of all sequences included in this study.

DNA was extracted using a CTAB-phenol/chloroform protocol [25]. The samples were PCR-amplified and bi-directionally sequenced for the mitochondrial protein-coding gene fragment of cytochrome oxidase subunit 1 (COI). The mitochondrial COI was amplified for a fragment of 316 bp using the species-specific primer set RNF (5′-AGACATACCAGGGGCTCTCA-3′) and RNR (5′-CCGTTTTAATCGGAGGGTTT-3′) [26]. PCR cycling conditions included initial denaturation at 94 °C for 3 min; 35 cycles at 94 °C for 30 s, 58 °C for 30 s, and 72 °C for 1 min; and a final extension at 72 °C for 5 min. Both strands of the PCR products were sequenced, and chromatographs were checked, assembled, and edited using Geneious v.7.1.9 (Biomatter Ltd., Auckland, New Zealand). The Mafft plugin of Geneious was used with default settings to align all sequences. The sequences were translated to amino acids, and no stop codons were detected.

For the phylogenetic analyses, the COI dataset of 316 bp was analysed under Maximum Likelihood (ML) and Bayesian Inference (BI) methods. The ML analyses were performed using IQ-TREE v. 1.6.12 [27] through the web interface [28]. The best-fit substitution model was selected automatically by ModelFinder [29], as implemented in IQ-TREE, determined as TPM2u+F+I+G4 according to the BIC model selection criteria. Branch support was assessed with the Shimodaira-Hasegawa-like approximate likelihood ratio test (SH-aLRT; [30]) and the ultrafast bootstrap (UFBoot; [31,32]), both with 1000 replicates. The BI analyses were carried out using MrBayes v.3.2.7 [33]. The best substitution model was determined using JModelTest v.2.1.7 [34,35] which resulted in the HKY+G model. Nucleotide substitution model parameters were unlinked across partitions, and the different partitions were allowed to evolve at different rates. Two simultaneous parallel runs were performed with four chains per run (three heated, one cold) for 10 million generations with sampling frequency of every 1000 generations. Stationarity was determined by the standard deviations of the split frequencies being lower than 0.01. The standard deviation of the split frequencies between the two runs and the Potential Scale Reduction Factor (PSRF) diagnostic were examined. The first 25% of trees were conservatively discarded as burn-in. Nodes with SH-aLRT ≥ 80, UFBoot ≥ 95, and a Bayesian posterior probability ≥ 0.95 were considered strongly supported. We calculated inter- and intraspecific uncorrected p-distance between Rhopilema and Rhizostoma species for the COI mitochondrial fragment, with pairwise deletion, in MEGA11 [36].

3. Results and Discussion

The results of the phylogenetic analyses, based on the COI dataset and using BI and ML analyses, produced similar trees, differing mostly in the less supported nodes at the intraspecific level (Figure 5). Rhopilema was recovered and found to be monophyletic (SH-aLRT = 86.4/UFBoot = 92/Bayesian posterior probability = 1.0; support values are given in the same order hereafter). Monophyly of all species of the genus was strongly supported (R. esculentum: 100/100/1.0; R. verrilli: 93.2/100/1.0; R. hispidum: 99.5/100/1.0; R. nomadica: 99.4/100/1.0). The species R. esculentum was recovered as the sister taxon to all the other Rhopilema species included in the analysis. A sister-taxon relationship between R. hispidum and R. verrilli was recovered, though with weak support (44.1/66/0.79), as was their inferred relationship with R. nomadica (63.6/71/0.91). The species R. nomadica showed low genetic differentiation, with all three Tanzanian specimens being phylogenetically grouped within the same clade as those from the Mediterranean Sea. Among these three Tanzanian specimens, two (RN1P and RN3P) had identical sequences, whereas the third (RN2P) differed from them by a single nucleotide at position 307.

The uncorrected genetic distance in the COI fragment was in the range of 16.5–18.9% among Rhopilema species, and 15.8–19.9% between Rhopilema and Rhizostoma. The lowest distance among Rhopilema species was 16.5% between R. verrilli and R. hispidum, and the highest was 18.9% between R. verrilli and R. esculentum. The level of genetic variability within Rhopilema species was relatively low and ranged between 0 and 3% (R. esculentum, 3%; R. nomadica, 1%; R. hispidum, 1%; R. verrilli, 0%).

The phylogenetic analyses confirm that the medusae collected from Dar es Salaam, Tanzania, were R. nomadica (Figure 5). These analyses reveal strong genetic affinities within R. nomadica between the Indo-Pacific and Mediterranean regions. This is the first genetic indication that the R. nomadica in the Mediterranean may have originated in the Indo-Pacific region, possibly via ballast water in ships from east African waters [37,38].

Galil et al. [1] reported that R. nomadica was found in the Red Sea and that it entered the Mediterranean via the Suez Canal (i.e., it is a Lessepsian migrant), yet there have been no published reports of its presence elsewhere in the Red Sea, nor anywhere else, aside from the Mediterranean. It was only in 1994, in the waters of Mozambique, 2500 km south of Tanzania, that a medusa, serving as a host to a pelagic shrimp, was identified as R. nomadica [19]; however, this identification was based on photographs (unpublished), taken while snorkeling (Berggren, personal communication). Hence, our results are the first to provide sound molecular evidence for the presence of R. nomadica in the waters off East Africa. It appears that this medusa, now identified as R. nomadica, is seasonally abundant offshore of Dar es Salaam, Tanzania (A. Sutton, personal communication), generally between October and December.

In common with other ephemeral jellyfish that have a metagenic life cycle, we propose that R. nomadica may exist and persist as cryptic populations (albeit in small numbers) in a variety of habitats and may remain unnoticed if they do not swarm or if no one is looking for them. In 2000, a swarm of the previously unreported rhizostome Phyllorhiza punctata occurred in the northern Gulf of Mexico [39]. This scyphomedusa likely entered the Atlantic Ocean from the Pacific through the Panama Canal some 50 years earlier [39], but the cryptic population was undetected until it bloomed. Likewise, in Brazil, the first report of the genus Cassiopea was published in 2002 [40]; however, based on molecular findings, within that genus, the species C. andromeda may have existed unnoticed in Brazilian waters for as long as 500 years before this medusa was first described in Brazil [41].

It seems that the date of first observation of a migrant species does not necessarily correspond to the date of introduction, as it may take some time for this species to be detected [42]. This applies even more strongly to scyphozoans, which have a metagenic life cycle. As Walsh et al. [42] suggested, a long time lag between introduction and detection, as observed for R. nomadica, is not uncommon, but the real challenge is identifying the reasons for the proliferation of a species that had previously maintained a low population density.

Here we propose that a cryptic population of R. nomadica occurred in the eastern Mediterranean years before it was described, and its presence became apparent only following changes in environmental conditions. This idea (see also Thibault et al. [43]) took root following an observation in 2014, when a sudden bloom of C. andromeda occurred in our flowing seawater aquaria. Although it was designated as one of the first Lessepsian migrants [44], reports of the medusa stage of C. andromeda in the Mediterranean coastal waters of Israel are relatively rare. A short search in our seawater aquaria revealed a rock, covered with C. andromeda polyps, that was arbitrarily collected at sea for a different purpose, and left in one of the tanks. Most likely, the protected waters in the aquarium, the mode of asexual reproduction via planuloid buds, and possibly abundant food from aquarium feedings resulted in an increased number of polyps, leading to enhanced strobilation and to the outbreak of this species. Following this introduction, C. andromeda became a dominant ‘fouling’ organism in the flowing seawater system, where polyps continue to proliferate and strobilate, and ephyrae are released seasonally.

Realizing that polyps of C. andromeda are probably common in the Israeli coastal waters of the Mediterranean, and following its proliferation in laboratory conditions, we propose a similar scenario for R. nomadica in the wild. We know what conditions C. andromeda needs to bloom, but what conditions does R. nomadica need? Perhaps therein lies the answer to the apparent absence of R. nomadica medusae until the 1970s. The findings of Edelist et al. [45] concur with our assumption that the origin of R. nomadica swarms in the eastern Mediterranean is in the Nile Delta, suggesting that this is the region where many of the polyps reside and release ephyrae. Success of the polyp stage is often inhibited by sub-optimal temperatures, insufficient food quality and quantity, or a combination of these [23,46,47,48].

Therefore, we searched for environmental changes on a regional scale that could have affected polyp reproduction and the subsequent medusa populations.

The opening of the Aswan High Dam in 1965 caused a change in the nutrient regime in eastern Mediterranean waters [49], which may have led to changes in R. nomadica abundances. Prior to 1965, nutrient enrichment from the Nile, most notably the August–October flood, led to Mediterranean phytoplankton blooms [50]. Unlike other parts of the oligotrophic eastern Mediterranean [51], these plankton-rich Egyptian coastal waters, dominated by diatoms (>90%) [52,53,54], were the source of rich planktivorous fisheries. Shortly after the Aswan High Dam became operational (early 1970s), the phytoplankton composition changed to a nearly balanced mix of diatoms and small flagellates [54], followed by the collapse of the sardine population [55]. More than a decade later, 60–80% of chlorophyll-a biomass in the south-eastern Mediterranean consisted of small (nano- and pico-) phytoplankton [54], such as flagellates, in contrast with the large diatoms of pre-Aswan High Dam construction. These changes in the lower trophic level may have led to an increase in ciliates [56] that feed on small flagellates [57] and are in turn grazed upon by zooplankton [58]. Ciliates may also sustain benthic scyphozoan polyps and ephyrae [59,60,61], enabling a substantial increase in their abundances and subsequently to adult jellyfish blooms, as observed in the late 1970s and early 1980s [1]. Since microzooplankton was also found to comprise at least 50% of adult R. nomadica’s diet [62], a shift towards smaller prey could also support a larger number of medusae.

Another factor to be considered is the ongoing warming trend of the Mediterranean Sea [63]. Recently, Dror and Angel [48] showed that the rising sea surface temperatures play a major role in polyp life history, positively affecting both polyp survival and asexual reproduction rates of R. nomadica. This is also supported by the findings of Schiariti et al. [64], who demonstrated that food sufficiency and increased temperatures enhance podocyst production in some podocyst-forming rhizostomes (e.g., R. esculentum, Rhizostoma pulmo). Mono-mode podocyst-forming species (i.e., species that reproduce exclusively via podocysts) are considered to have lower reproductive rates, and their podocysts serve as a means to maintain polyp populations under harsh environmental conditions [64,65]. However, Dror and Angel [48] found that in the ‘mono-mode’ R. nomadica podocysts serve as a mechanism for accelerated population amplification in the current season; excystment of podocysts occur within just two weeks after podocyst production. Strobilation of the proliferated polyps inevitably leads to an increase in ephyrae, the seed of the massive annual swarms that now characterize the eastern Mediterranean Sea. However, a cryptic reservoir of dormant podocysts may also have contributed to the dramatic increase in R. nomadica swarms observed since the 1980s, as environmental conditions—particularly temperature and food availability—became increasingly favorable.

While temperature and food availability are recognized as key drivers of scyphozoan polyp development [23,48], the presence and availability of suitable benthic substrate constitute an additional, and potentially limiting, factor that warrants consideration [66].

Since the early 1970s, the Nile Delta coastline, which is predominantly sandy, aside from the natural rocky formations of Alexandria, has experienced extensive “hard” coastal engineering, including the construction and subsequent expansion of jetties, groins, seawalls, detached breakwaters, and revetments at multiple sites. Many of these interventions were implemented or substantially reinforced after 1981, transforming historically sandy and morphodynamically active shorelines into environments increasingly dominated by armorstone, concrete elements, and rubble-mound structures, as well as stabilizing several inlets and navigation channels [67]. However, the initial emergence of R. nomadica swarms in the Levant during the 1980s predates much of the large-scale expansion of artificial coastal structures in this region [68], suggesting that coastal engineering was unlikely to have been a primary driver of the early population outbreaks. In this context, we further suggest that the natural rocky formations present along the Alexandria coastline may have supported cryptic populations of R. nomadica polyps prior to the onset of widespread jellyfish blooms. Such populations could have persisted at low, undetected densities on naturally available hard substrates [69], without giving rise to recurrent or large-scale medusa outbreaks. Rather, broad-scale environmental changes, including rising sea temperatures and altered trophic conditions, appear to have played a more fundamental role in facilitating the establishment of the species. Nevertheless, the subsequent increase in artificial hard substrates may have contributed locally to the persistence and spatial expansion of polyp populations by providing additional firm settlement surfaces [70] and modified nearshore hydrodynamic conditions [71]. Thus, coastal structures are best viewed as a secondary or reinforcing factor that may have enhanced population maintenance in later decades, rather than as a principal cause of the initial proliferation of R. nomadica in the Levantine Basin.

Interestingly, considering the image from 1938 (Figure 2), it seems that the establishment and colonization dynamics of R. nomadica, are similar to those described for the Australian spotted jellyfish, P. punctata, and the bottom-dwelling C. andromeda in the Mediterranean. These dynamics include a slow initial establishment phase, followed by rapid dispersal, and probably a period of several decades from first observation in the Levant to a pan-Mediterranean distribution [72,73]. Nevertheless, neither P. punctata nor C. andromeda establish dense aggregations like those frequently seen for R. nomadica in the eastern Mediterranean. This is particularly intriguing with respect to P. punctata, a bloom-forming species that forms large aggregations in warm coastal waters [39,74]. The absence of large populations of P. punctata in the Mediterranean may be related to the relatively low number of semi-enclosed, sheltered coastal lagoons, required for the success of its early life stages [72]. Other traits, such as different asexual reproduction strategies, may also play a role in shaping its population dynamics. Unlike R. nomadica, which reproduces exclusively via podocysts and releases multiple ephyrae via polydisc strobilation, P. punctata and C. andromeda polyps reproduce through free-swimming planuloid buds, followed by the release of a single ephyra via monodisc strobilation. This suggests that, in addition to the different modes of reproduction, there may be other physiological and ecological traits that could explain why one species forms swarms and other species do not, as clearly observed in the Levant.

The new data presented here are consistent with an Indo-Pacific origin for this species, most likely from the Indian Ocean. An alternative scenario, whereby R. nomadica migrated from the Mediterranean to the Indian Ocean (e.g., an anti-Lessepsian migration), cannot be entirely excluded. However, this scenario appears unlikely, given that the majority of rhizostome jellyfish are of tropical and subtropical origin [43], sharing physiological traits such as branched oral arms, multiple mouthlets, and nematocyst-bearing surfaces, presumably the result of their evolutionary history in tropical oligotrophic waters [75,76]. Considering R. nomadica’s absence from the Red Sea, based on the lack of reports, we can rule out the possibility that any of its pelagic life forms swam from the Indian Ocean, through the Suez Canal, to the Mediterranean; thus, it was probably introduced via ballast water or as hull fouling. Nevertheless, the dynamics of its proliferation need to be further investigated via polyp feeding trials, prey preference analyses, growth trials, and studies of population genetics to better understand the relationship between eastern African R. nomadica populations and those of the Mediterranean.

Improved understanding of these dynamics could provide essential information to help track the dispersal of the nomad jellyfish and predict its future distributions. The fact that R. nomadica became a dominant species in the eastern Mediterranean only in the late 1970s, long after it was first documented in 1938, raises interesting questions concerning the ecological dynamics involved. Nevertheless, we may never know if it was the opening of the Aswan High Dam that triggered a series of ecological changes that resulted in R. nomadica becoming the dominant species in the Levant.