Medusae (Cnidaria) of Reunion Island (South West Indian Ocean): Diversity, Abundance and Distribution

Abstract

Numerous studies have been conducted on the benthic stages of Medusozoa in Reunion Island, but none on the pelagic stages. This study is the first to investigate the shallow waters of the island for the diversity, abundance, and spatio-temporal distribution of jellyfish. During a one-year survey, samples were collected with a plankton net weekly or biweekly at four sites (two reef/two non-reef) and two depths (10/50 m). Of the 267 samples, 3450 medusae were sorted and 56 species identified. The meroplanktonic Hydroidolina (Antho- and Leptomedusae) were the most diverse (38 species), while the holoplanktonic Trachylinae (13 species) were the most abundant. Hydromedusa species richness was higher at coastal stations than offshore, but similar between reef and non-reef sites. There was no significant variation in species richness or abundance between months or seasons. Including some other catches, the total number of species reached 62. Eight species are new records for the Indian Ocean (all Anthomedusae). Indian Ocean literature references are given in the species list, and some photographs are provided. This initial study, which greatly expands the local hydrozoan fauna knowledge, will serve as a reference for future research, especially regarding climate change and coastal management in Reunion Island.

1. Introduction

Medusae (jellyfish) are the free-swimming adult stages of Medusozoa cnidarian species and include three classes: Hydrozoa, Scyphozoa, and Cubozoa. Most medusae are meroplanktonic, originating from polyps that are part of sessile benthic communities, while others are holoplanktonic, having no benthic stage. Because of their benthic and planktonic nature, polyps and medusae have long been studied separately, as they required different sampling methods, resulting in two distinct nomenclatures.

Most of the knowledge about hydrozoan medusae in the Indian Ocean comes from Kramp’s monograph [1]. This work is based on an extensive collection of specimens collected during oceanographic expeditions and a thorough review of the literature. The other publications are either the result of a single cruise covering the entire Indian Ocean (except the eastern Australian and Indonesian parts) [2,3,4,5] or the result of local studies. Most of the local studies concern India and its surroundings: the Indian coasts [6,7,8,9,10,11], the Nicobar and Andaman Islands [12,13], and the Laccadive Islands [14]. Other important works include areas from Java to Perth and Bass Strait (Australia) [15], the Arabian and Red Seas [16,17], the Seychelles and Mozambique [18], and the Agulhas Current (eastern South Africa) [19]. Some jellyfish species are also reported in the Guide to the Coastal and Surface Zooplankton of the Southwestern Indian Ocean [20].

There are no published data on jellyfish in Reunion Island. However, the benthic hydrozoan fauna of coral reefs, volcanic areas, and deep waters has been studied previously by collecting benthic polyps. A total of 160 benthic species have been reported from the western coral reefs [21], a number reaching 260 when including the rocky shore and continental shelf around Reunion Island [22]. Within these species, a small percentage (23%) belongs to Anthoa- and Leptothecata (Hydroidolina), known to have a medusa in their life cycle [23]. However, based on a recent inventory of hydroids in lower mesophotic coral ecosystems (MCEs) that showed a high specific richness [24], this percentage reached 43%. These cumulative data demonstrate the great diversity of the local hydrozoan fauna. However, many species described from benthic colonies remain unidentified due to a lack of knowledge of their life cycles, the sterility of sampled specimens, or the production of newly released medusae that are too different from adults to be identified.

The collection of medusae from the plankton was initiated to provide additional information to continue this process of benthic species identification, to gain access to holoplanktonic species (Trachylinae), and to estimate the abundance of jellyfish in the shallow waters of Reunion Island. This paper presents the first overview of the diversity of medusae in the waters of the fringing coral reefs and in the vicinity of the west coast of Reunion Island. It also provides data on the abundance and seasonality of jellyfish based on a full year of plankton sampling (October 2005–October 2006).

2. Materials and Methods

2.1. Study Area

Reunion Island is a French overseas volcanic island located in the southwest of the Indian Ocean (21°07′ S, 55°32′ E) about 700 km east of Madagascar and about 300 km north of the Tropic of Capricorn (Figure 1a). It is the youngest island in the Mascarene archipelago (along with Mauritius and Rodrigues), formed 2.1 million years ago [25,26], ellipsoidal in shape, with a main diameter of 70 km. The region lies at the southern limit of the South Equatorial Current. Trade winds blow from the east-southeast all year round, with a peak in austral winter (May–October), causing heavy swells (3 m on average). Occasionally, larger swells (up to 10 m) come from the southwest due to austral depressions. In addition, during the wet and rainy summer, cyclones cause violent but rare swells (up to 4 m) from the north-northeast [27]. The tidal range is about 0.1 m during neap tides and 0.9 m during spring tides [28,29]. Sea surface temperatures vary from 22 °C (in September/October) to 28 °C (in December/January) and are increasing by about 0.1 °C per 10 years due to climate change [30]. As a young volcano, Reunion has a narrow continental shelf, with the depth dropping rapidly from the coast. The coral reefs form a narrow, discontinuous belt of fringing coral reefs (maximum width 520 m) extending along 25 km of the west leeward coast of the island and covering an area of approximately 12 km² [31]. The 9 km long Saint-Gilles/La Saline reef complex is the most developed, while those of Saint Leu and Saint Pierre, further south, are smaller (Figure 1a). The reef profile from the open sea towards the beach consists of an outer slope formed by a lower sloping platform and a steep and narrow spur and groove zone that extends from 20–15 m depth to the front, a reef flat, and a straight back reef zone or boat channel (geomorphology from [32,33]).

2.2. Sampling

The material examined was obtained from daytime sampling on the west coast of Reunion Island during a one-year survey from 21 October 2005 to 17 October 2006 (63 dates). Four sites were sampled from north to south: Cap La Houssaye (CAPL), Boucan Canot (BOUC), Ermitage (ERMI), and Saint-Leu (SLEU) (Figure 1b). The first two sites consist of basalt slopes colonized by coral fauna, called the coral bank and coral platform, respectively [32]. The other two sites are part of the fringing coral reefs of Saint Gilles-La Saline and Saint Leu, respectively. At each of the four sites, two stations were sampled, one near the coast at 10 m depth (bottom at 12–15 m, named SITE-10, e.g., CAPL-10), and one “offshore” about 800–1000 m away, perpendicular to the coast, at 50 m depth (bottom at about 80 m depth, named SITE-50, e.g., CAPL-50) (Figure 1 and

Table 1. Geographical coordinates, sampling depth, and bottom geomorphology of the eight stations distributed along the west coast of Reunion Island, numbered from north to south and from the coast (SITE-10, 0–10 m depth) to the outer sea (SITE-50, 0–50 m depth).

Station Abbreviation	Site (Depth/ Localization)	Latitude (S)	Longitude (E)	Geomorphology (Bottom Nature)
CAPL-10	Cap La Houssaye (10 m/coastal)	21°01′03.0″	055°14′18.2″	Coral bank (basalt slope)
CAPL-50	Cap La Houssaye (50 m/offshore)	21°00′28.1″	055°13′59.0″	Coral bank (lower sloping platform 1)
BOUC-10	Boucan Canot (10 m/coastal)	21°01′35.9″	055°13′23.0″	Reef platform (basalt slope)
BOUC-50	Boucan Canot (50 m/offshore)	21°01′18.0″	055°12′57.6″	Reef platform (lower sloping platform)
ERMI-10	Ermitage (10 m/coastal)	21°05′12.2″	055°13′13.9″	Coral reef (spurs and grooves)
ERMI-50	Ermitage (50 m/offshore)	21°05′44.8″	055°11′57.0″	Coral reef (lower sloping platform)
SLEU-10	Saint Leu (10 m/coastal)	21°11′01.5″	055°17′03.2″	Coral reef (spurs and grooves)
SLEU-50	Saint Leu (50 m/offshore)	21°11′10.7″	055°16′19.5″	Coral reef (lower sloping platform)

¹ Lower sloping platform: area of gentle slope colonized by algal formations, sandy stretches, and scattered coral colonies.

).

Plankton samples were collected using a homemade net with a 37 cm opening diameter, 109 cm filter side length, and 56 µm mesh size. Samples were collected weekly from the northernmost sites (CAPL and BOUC) and every two weeks from the southernmost sites (ERMI and SLEU). The net was towed vertically at a speed of 1 m·s⁻¹ from a boat. Coastal stations were sampled using three successive plankton hauls (10–0 m) pooled together, while offshore stations were sampled using a single plankton haul (50–0 m). A total of 267 samples were collected from these eight stations (

Table 2. Field sampling.

Site-depth (m)	CAPL-10	CAPL-50	BOUC-10	BOUC-50	ERMI-10	ERMI-50	SLEU-10	SLEU-50	Total
Number of samples	40	40	40	39	26	27	28	27	267

) and immediately preserved in salted formalin. In addition, some qualitative samples were collected using a 168 µm mesh net (52 cm mouth diameter) when jellyfish were observed from the boat. Samples were also collected by hand using a large bottle near the surface while snorkeling. These additional samples were kept alive until they reached the laboratory.

2.3. Laboratory Work

The 267 samples were sorted, medusa specimens were separated, observed using a dissecting and a compound microscope, and counted. Specimens were identified to species level where possible and classified according to Bouillon and his colleagues [23], WoRMS [34], and the literature available to us. Live medusae from the additional samples were isolated, reared in aquaria containing “living stones”, and fed with Artemia nauplii for a few days. These additional specimens were marked with an asterisk (*) in the taxonomic species list (Appendix A) if they were not present in the 267 quantitative samples, and photographed by David Caron (Appendix B). These additional individuals were not included in the analyses. All specimens were preserved in 4% seawater formalin and stored at the University of Reunion Island. Siphonophores were not sorted or examined, but were preserved for further study.

2.4. Terminology

2.4.1. Nomenclature

According to WoRMS [34], this article uses two taxonomic new terms: “Hydroidolina” (ex-Hydroidomedusae) includes the medusae of the two orders Anthoathecata (Anthomedusae) and Leptothecata (Leptomedusae)—the order Siphonophorae has not been studied—and “Trachylinae” (ex-Automedusae) includes those of the three suborders Limnomedusae (with Liriope tetraphyla and Geryonia proboscidalis moved recently from Trachy- to Limnomedusae), Narcomedusae, and Trachymedusae. However, we maintain the term “Hydromedusae”, which encompasses all hydrozoan medusae. For the analysis, we separated the Hydroidolina, which buds from a benthic polyp, and the Trachylinae, which have direct development, based on their life cycles.

2.4.2. Occurrence, Species Richness, and Abundance

The occurrence of a species is defined as the number of samples in which the species was found (n), expressed as a percentage of the total number of samples (n × 100/267). In order to standardize the sampling effort, medusa species richness is the number of species per 100 m³ of filtered seawater, and medusa abundance is the number of individuals per 10 m³ of filtered seawater.

2.5. Statistical Analysis

Linear models (ANOVA, function “lm” in R) were performed to compare species richness and abundance of Hydromedusae, Hydroidolina, and Trachylinae between reef and non-reef sites, between coastal and offshore stations, and between months and seasons. Statistical analyses were performed using R 4.2.1 software, Vienna, Austria [35]. The reef sites (108 samples) include Ermitage (ERMI) and Saint-Leu (SLEU), while the non-reef sites (159 samples) include Cap La Houssaye (CAPL) and Boucan Canot (BOUC). The coastal stations (134 samples) include CAPL-10, BOUC-10, ERMI-10, and SLEU-10, while the offshore stations (133 samples) include CAPL-50, BOUC-50, ERMI-50, and SLEU-50 (Table 2). The number of Hydromedusa specimens identified only to genus level (Genus spp.) is retained for abundance calculations, but not for species richness calculations. The spatial distribution of medusae was analyzed from 21 October 2005 to 17 October 2006, while their temporal distribution (monthly and seasonal) was analyzed from November 2005 to October 2006. Two seasons were taken into account: the austral summer from November 2005 to April 2006, and the austral winter from May to October 2006.

3. Results

In total, 3542 jellyfish were examined in the laboratory for all samples (quantitative and qualitative), either fixed or alive. The 267 quantitative samples included three Scyphomedusae: Atolla wyvillei Haeckel, 1880, Cephea cephea (Forskal, 1775), and Thysanostoma flagellatum (Haeckel, 1880) (not T. thysanura according to Morandini et al. [36]), one Cubomedusa Alatina alata (Raynaud, 1830), and one unidentified juvenile Stauromedusa (a class of medusozoans comprising stalked jellyfish). Hydrozoan medusae (Hydromedusae) were by far the most numerous, with 3443 individuals, of which 194 could not be identified (5.6%). The total dataset, therefore, comprised 3249 Hydromedusae identified at the species or genus level (

Table 3. Number of Hydromedusae identified at the species or generic level collected at each of the eight stations during the sampling period (October 2005–October 2006).

Site-depth (m)	CAPL-10	CAPL-50	BOUC-10	BOUC-50	ERMI-10	ERMI-50	SLEU-10	SLEU-50	Total
Hydromedusae	464	636	401	549	169	396	237	397	3249

).

The maximum number of medusae found in a single coastal sample (3 × 10 m depth) reached 115 individuals, while a maximum of 58 medusae were collected in a single offshore sample (1 × 50 m depth) (Figure 2). The mean number of medusae found in coastal samples was 9.6 (±13.7) medusae per sample (median = 5), and 15.1 (±11.4) medusae in offshore samples (median = 13). During the sampling period, only 13 samples (of which 12 were coastal) contained no medusae (i.e., 4.9% of all quantitative samples), and half of the samples contained at least nine medusae (Figure 2).

By standardizing the sampling effort between coastal and offshore samples, the average abundance of Hydromedusae for the entire study and all sites together was 28 (±33) individuals per 10 m³. The Trachylinae showed the maximum abundance with 24 (±29) individuals per 10 m³, while the Hydroidolina reached 4 (±9) individuals per 10 m³. The standard deviations, which are higher than the means, highlight the high heterogeneity of the sample abundance.

3.1. Species Diversity and Community Assemblage

From the 267 samples, a total of 51 species of Hydromedusae were distinguished, distributed in 39 genera and 25 families (

Table 4. Hydromedusa community structure: number of families, genera, species, and individuals collected.

Taxa	Families	Genera	Species	Specimens
Anthomedusae	13	20	27	181
Leptomedusae	6	8	11	247
Hydroidolina indet	-	-	-	194
Total Hydroidolina	19	28	38	622
Narcomedusae	3	4	4	457
Limnomedusae	1	2	2	685
Trachymedusae	2	5	7	1679
Total Trachylinae	6	11	13	2821
Total Hydromedusae	25	39	51	3443

) (Appendix A). The meroplanktonic ones (Hydroidolina) were the most diverse with 19 families, 28 genera, and 38 species (i.e., about 75% of the total species diversity), while the holoplanktonic ones (Trachylinae) gathered 6 families, 11 genera, and 13 species (i.e., about 25% of the total). The most speciose group was the Anthomedusae (27 spp.) against the Leptomedusae (11 spp.) for the Hydroidolina, and the Trachymedusae (7 spp.) against the Narcomedusae (4 spp.) and the Limnomedusae (2 spp.) for the Trachylinae (Table 4) (several photos of living specimens are included in Appendix B). Thus, the meroplanktonic Anthomedusa species represented more than half of the Hydromedusa species diversity (53%). In contrast, the holoplanktonic Trachylinae exhibited a higher number of individuals sampled (2821 specimens), representing 82% of the total (Table 4). Among them, three species were largely dominant both in the number of specimens collected and in their occurrence: the Trachymedusa Aglaura hemistoma Péron & Lesueur, 1810, the Limnomedusa Liriope tetraphylla (Chamisso & Eysenhardt, 1821), and the Narcomedusa Solmundella bitentaculata (Quoy & Gaimard, 1833), with, respectively, 1546, 673, and 450 specimens (i.e., together 77.5% of all Hydromedusa specimens) and 83.2%, 74.2%, and 50.6% of occurrence (

Table 5. Number of specimens and occurrences of the 10 Hydroidolina taxa with more than 10 specimens and of the 4 dominant Trachylinae species collected during the study period.

Species or Taxa	Number of Specimens	Occurrences (%)
	Hydroidolina
Clytia spp.	170	24.0
Corymorpha forbesii	26	6.0
Cytaeis spp.	26	6.7
Amphinema dinema	22	5.6
Proboscidactyla ornata	22	3.3
Hydractinia spp.	21	3.3
Eucheilota tropica	16	0.4
Laodicea sp.	12	3.4
Cirrholovenia tetranema	11	1.1
Phialella quadrata	11	1.5
	Trachylinae
Aglaura hemistoma	1546	83.2
Liriope tetraphylla	673	74.2
Solmundella bitentaculata	450	50.6
Rhopalonema velatum	114	25.8

). Apart from these species, the fourth most abundant species was the Trachymedusa Rhopalonema velatum Gegenbaur, 1857, with 114 individuals (occurrence: 25.8%).

The meroplanktonic species (622 medusae collected, including 194 unidentified individuals) were generally rare and episodic, with 1 to 26 specimens (per species) sampled throughout the study, except for Clytia spp. (170 medusae, occurrence: 24%) (Table 5). A total of 181 Anthomedusae and 247 Leptomedusae were collected. The most abundant species or genera among the Anthomedusae were Amphinema dimena (Péron & Lesueur, 1810), Corymorpha forbesii (Mayer, 1894), Cytaeis spp., Hydractinia spp., and Proboscidactyla ornata (McCrady, 1859), and for the Leptomedusae, Cirrholovenia tetranema Kramp, 1959, Clytia spp., Eucheilota tropica Kramp, 1959, Laodicea sp., and Phialella quadrata (Forbes, 1848) (Table 5).

3.2. Hydromedusa Spatial Distribution (October 2005–October 2006)

3.2.1. Spatial Distribution of the Species Richness and Nominal Species

For the entire study, the Hydromedusa species richness (Genus spp. excluded, e.g., Clytia spp. excluded) found at each of the eight stations ranged from 7.15 species per 100 m³ (BOUC-50) to 17.72 species per 100 m³ (SLEU-10) (Figure 3). The species richness of the coastal stations (mean = 16.54 ± 0.92 species per 100 m³) was always higher than that of the offshore stations (mean = 9.34 ± 1.49 species per 100 m³) (ANOVA, t = −8.23, p < 0.0001), even when Hydroidolina (ANOVA, t = −4.74, p = 0.003) and Trachylinae (ANOVA, t = −2.95, p = 0.025) were analyzed separately. In contrast, no significant differences were found between sites (coastal and offshore stations combined) or between reef and non-reef zones.

Most of the Hydroidolina (28 out of 38 species) were collected from either non-reef (18 species) or reef (10 species) stations (Table S1). The 18 non-reef species were Amphinema australis, Bougainvillia bitentaculata, Cytaeis nassa, Eucheilota tropica, Euphysilla pyramidata, Halitiara formosa, Hybocodon sp., Laodicea indica, Leuckartiara sp., Obelia sp, Protiara tetranema, Pseudoclytia gardineri, Teissiera australe, Turritopsis chevalense, Vellela vellela, Zanclella diabolica, Zanclea polymorpha, and Zanclea ?sessilis, while the 10 reef species were Bougainvillia aurantiaca, Bougainvillia platygaster, Bougainvillia principis, Cirrholovenia polynema, Cirrholovenia tetranema, Clytia hemisphaerica, Euphysa sp., Halocoryne frasca, Hydractinia sp., and Staurodiscus tetrastaurus. Moreover, among the non-reef species, eight of them were caught only at the coastal stations (A. australis, E. tropica, H. formosa, Obelia sp., P. tetranema, T. australe, V. vellela, and Z. ?sessilis), while five species were collected only at the offshore stations (B. bitentaculata, C. nassa, E. pyramidata, Hybocodon sp., and Leuckartiara sp.). Among the reef species, B. principis, C. polynema, C. hemisphaerica, Euphysa sp., H. frasca, Hydractinia sp., and S. tetrastaurus (6 species) were only collected at coastal stations, while B. aurantiaca and B. platygaster were collected exclusively at offshore stations. Besides these species, ten were found in both zones, of which seven were found in both coastal and offshore stations: Amphinema dinema, Clytia mccradyi, Cnidocodon leopoldi, Corymorpha forbesii, Laodicea sp., Phialella quadrata, and Proboscydactyla ornata. The remaining three species were recorded in non-reef and reef zones: Corymorpha bigelowi and Teissiera sp. in coastal stations, and Podocorynoides minima in offshore stations.

Conversely, of the 13 species of Trachylinae identified, 11 were found in both reef and non-reef zones, of which 8 were found in both coastal (0–10 m depth) and offshore (0–50 m depth) stations (Table S1). The remaining two species, Amphogona sp. and Sminthea eurygaster, were only collected in the non-reef zone. More precisely, S. eurygaster was collected in the non-reef offshore station CAPL-50, but only once. None of the holoplanktonic species were collected only in the reef zones. Aegina citrea, Haliscera conica, and Amphogona pusilla were recorded in both non-reef and reef zones, but only in the offshore stations for the first two and only in the coastal stations for the latter. Considering only the five species with more than 10 specimens collected throughout the study (A. hemistoma, G. proboscidalis, L. tetraphylla, R. velatum, S. bitentaculata), they were all found at all stations (Table S1).

3.2.2. Spatial Distribution of Abundance

For the entire study, the lowest mean abundance of Hydromedusae was recorded at the coastal reef station ERMI-10 with 20.15 (±21.82) individuals per 10 m³, while the highest mean abundance was recorded at the coastal non-reef station CAPL-10 with 35.96 (±65.88) individuals per 10 m³ (Figure 4). These average abundances of Hydromedusae per station mirrored those of Trachylinae, which were also lowest at ERMI-10 with 16.93 (±19.39) ind·10 m⁻³ and highest at CAPL-10 with 29.30 (±56.54) ind·10 m⁻³. Furthermore, these results accurately reflected the abundance of the Trachylinae Aglaura hemistoma, which was lowest at ERMI-10 and highest at CAPL-10. However, no significant differences were found for Hydromedusae and Trachylinae between stations, between sites (coastal and offshore stations combined), between coastal and offshore stations, or between reef and non-reef zones (Figure 4).

Concerning the Hydroidolina, the lowest mean abundance was recorded at the offshore reef station ERMI-50 with 2.07 (±2.33) ind·10 m⁻³, while the highest mean abundance was recorded at the coastal non-reef station CAPL-10 with 6.66 (±15.55) ind·10 m⁻³ (Figure 4). The mean abundance of Hydroidolina was higher in coastal stations (5.67 ± 11.31 ind·10 m⁻³) than in offshore stations (2.56 ± 4.92 ind·10 m⁻³) (ANOVA, t = −2.91, p = 0.004), but there were no significant differences between sites, and between reef and non-reef zones.

A few swarms, defined as any sample in which the abundance of Hydromedusae was at least one standard deviation greater than the average abundance of Hydromedusae (see [37]), were detected at each station. The highest number of samples containing such swarms was found at station CAPL-10 (7 samples out of 40). These swarms were primarily made up of the dominant species of Trachylinae, and on occasion, Clytia spp.

3.3. Hydromedusa Temporal Distribution (November 2005–October 2006)

3.3.1. Temporal Distribution of the Species Richness and Nominal Species

The Hydromedusa species richness was lowest in January 2006 (6.8 species per 100 m³) and highest in October 2006 (17.4 species per 100 m³) (Figure 5). A similar trend was observed for the Trachylinae species richness, which reached its lowest level in January 2006 (3.9 species per 100 m³) and its highest level in October 2006 (11.6 species per 100 m³). In contrast, the Hydroidolina species richness was lowest in September 2006 (1.9 species per 100 m³) and highest in March 2006 (9.0 species per 100 m³). However, statistical tests revealed no significant differences between months or seasons.

Among the Hydroidolina, with the exception of the Clytia spp., which were present all year round, of the species Amphinema dinema, Corymorpha forbesii, and Zanclea spp., which were present during both the hot and the cool seasons, and of Cytaeis spp., which were not recorded when the SST was at its maximum (January–March), it is difficult, if not impossible, to define a clear seasonality for the many species rarely observed in this study (Table S2). However, if we take into account species that were collected in at least two samples (at two stations or on two dates in the same station), a few taxa appeared to show seasonality. Four species were recorded during two successive months, either in summer or winter: Proboscidactyla ornata (December 2005 and January 2006), Teissiera sp. (February–March 2006), Zanclella diabolica (April–May 2006), and Zanclea polymorpha (May–June 2006). On another hand, Cirrholovenia tetranema, Corymorpha bigelowi, Laodicea indica, Protaria tetranema, Pseudoclytia gardineri, and Turritopsis chevalense were only sampled for a few days or at a few stations during a single month (Table S2). In contrast, the dominant Trachylinae (including Geryonia proboscidalis) were collected almost all year round. However, Aegina citrea was collected only in December 2005, while Amphogona pusilla and Amphogona sp. were collected from May to July 2006, Haliscera conina in September and October 2006, and Cunina sp. and Solmaris sp. in September 2006 (the number of specimens of each species and the date on which they were collected during the study are given in the taxonomic list in Appendix A).

3.3.2. Temporal Distribution of Abundance

During the study year, August 2006 (at the height of the austral winter) was the month with the lowest abundance of Hydromedusae (15 ± 18 ind·10 m⁻³), while December 2005 (at the height of the austral summer) showed the highest abundance (39 ± 33 ind·10 m⁻³) (Figure S1). However, the heterogeneity of the sample abundances is such that no significant difference was found between months, even when a seasonal analysis (over 3 or 6 months) was performed. In fact, these monthly mean abundances mirrored those of the dominant species Aglaura hemistoma (linear correlation: R² = 0.844) and, to a lesser extent, those of Solmundella bitentaculata (R² = 0.206) (Figure 6). A few swarms (sensus [37]) were observed every month, except January and April 2006. The largest number of swarms was recorded in December 2005 (7 samples out of 24). Once again, these swarms corresponded to high densities of the dominant Trachylinae and the Hydroidolina Clytia spp.

4. Discussion

This study is the first to be conducted on medusae in Reunion Island. It significantly expands knowledge of the island’s hydrozoan fauna, which was previously known only through its benthic stages. To date, although general knowledge of both stages has been acquired for several large regions of the world, no site in the Indian Ocean has been so extensively studied based on polyps and medusae, with the exception of the east coast of South Africa (polyps by Millard [38] and medusae by Buecher et al. [19], and India (polyps by Mammen [39,40,41] and medusae from numerous publications, see Section 1. Introduction).

4.1. Species Diversity, Abundance, and Medusae Assemblage

Of the 267 quantitative samples collected, 56 species were identified: 51 Hydromedusae (Table 4), three Scyphomedusae (Atolla wyvillei, Cephea cephea, and Thysanostoma flagellatum), one Cubomedusa (Alatina alata), and one Stauromedusa (unidentified medusozoan juvenile stalked medusa). Among the 51 Hydromedusae species collected, the Hydroidolina were the most diverse (38 species), especially the Anthomedusae (27 species), but presented few individuals. In contrast, the Trachylinae were the least diverse (13 species), but they were present in large numbers. From additional qualitative samples, six more Hydromedusae species were found, including three Anthomedusae (Dicnida rigida (Appendix B: Plate 3F), Porpita porpita, and Zanclea medusopolypata (Appendix B: Plate 3B), two Leptomedusae (Aequorea sp. (Appendix B: Plate 4A–B), and Laodicea ?undulata), and one Narcomedusa (Pegantha sp. (Appendix B: Plate 5D)). Thus, the total medusae species richness reached 62 species (belonging to 48 genera and 33 families, see Appendix A), of which more than 78% were meroplanktonic. Eight species are new Indian Ocean records: Bougainvillia aurantiaca, Bougainvillia principis, Halocoryne frasca, Protiara tetranema, Zanclea medusopolypata, Zanclea polymorpha, Zanclea ?sessilis, and Zanclella diabolica. They are all Hydroidolina budded by tiny hydroids from benthic stolonal colonies (except for Bougainvillia, which can have erected ones) that are very difficult to see when diving and are often neglected [42]. Moreover, they belong to the order Anthoathecata (Anthomedusae), whose fragile polyps are not covered by a perisarc. In addition, polyps of the genus Halocoryne, Zanclea, and Zanclella often live in association with other invertebrates such as sponges, bryozoans, and scleractinians, and thus must be sampled from living substrates in order to be found [43,44,45]. Furthermore, the polyp of the genus Protiara is unknown [23]. As a result, such benthic hydroids that give rise to these medusae have not yet been identified in the Indian Ocean, and particularly in the waters around Reunion Island, despite four decades of inventory (by scuba diving) and taxonomic work carried out by Gravier-Bonnet. These species have warm affinities and have been recorded in the western Pacific Ocean, except Zanclea ?sessilis (Appendix A). However, Reunion Island’s jellyfish diversity is here obviously underestimated, as the samples in this study were taken over the course of one year on a short section of the west coast and only during the day. Numerous medusae migrate vertically and are more abundant at night [46]. Moreover, our study’s method was effective for catching Hydromedusae but not Scyphomedusae and Cubomedusae. Furthermore, much of the coastline was not studied because it is exposed to strong waves, far from harbors, and thus difficult to reach by the small boats available on the island.

In this study, three Trachylinae species were largely dominant in terms of abundance: Aglaura hemistoma, Liriope tetraphylla, and Solmundella bitentaculata (Appendix B: Plates 5 and 6). The fourth most common species was Rhopalonema velatum (Plate 6D). These four species accounted for 98.7% of the Trachylinae collected and 81% of all Hydromedusae collected. Moreover, A. hemistoma was present throughout the year, with the highest occurrence value of 83.2%. These results are consistent with previous data published on the hydrozoan fauna of the Indian Ocean, from the west coast of India to the east coast of South Africa, where Hydroidolina (ex-Hydroidomedusae) are the most diverse but occur at low densities, while Trachylinae (ex-Automedusae) are the most abundant in number of individuals but comprise fewer species [5,10,19]. A compilation of these previous studies shows that the Anthomedusae and Leptomedusae (Hydroidolina) were the two most speciose orders, each accounting for 75–90% of the diversity, and that the Limnomedusae, Trachymedusae, and Narcomedusae (Trachylinae) accounted for 70–90% of the total catch. Furthermore, the four most abundant species were A. hemistoma, L. tetraphylla, S. bitentaculata, and R. velatum, in decreasing order, as found in this study for Reunion Island. Thus, the present results confirm that these four species of Trachylinae are the «characteristic quatuor» of the tropical and subtropical regions of the Indian Ocean [47]. These holoplanktonic species are also globally dominant in all three oceans [47] and even in the Mediterranean Sea [48,49], where they are associated with the Scyphomedusa Pelagia noctiluca, depending on years, depth, and seasons [50,51]. Although it is difficult to compare abundances between the different studies because of the different depths of sampling, plankton nets used, and sampling methodology (daylight or night sampling, oblique or vertical hauls), the abundances found in Reunion Island were of the same order of magnitude as in other works in the Indian Ocean [8,9,10,52].

Among the Hydroidolina, the species of the genus Clytia were the most abundant throughout the year and had the highest occurrence among the meroplanktonic species (approximately 25%, including C. hemisphaerica, C. mccradyi, and Clytia spp.). The second place went to species of the genus Cytaeis, as well as the two species Amphinema dinema and Corymorpha forbesii, although these three taxa had lower occurrences (approximately 6% each, Table 5). The diversity of the Hydroidolina around the world is such that it is difficult to identify a characteristic assemblage of tropical species. However, according to Navas-Perreira & Vannucci [5], the ones collected on Reunion Island include several species inhabiting the “Indian Ocean central water system” and/or the “Indian Ocean equatorial system” (such as Bougainvillia platygaster, Corymorpha bigelowi—ex Euphysora bigelowi-, Corymorpha forbesii—ex Vannuccia forbesii -, Euphysilla pyramidata, and Proboscidactyla ornata). In addition, some of the rarest species of their large study (45,000 specimens) were also present in our collection: Staurodiscus tetrastaurus, Amphogona apsteini, Clytia mccradyi, Cirrholovenia tetranema, Eucheilota tropica, Phialella quadrata, and even a few species that, at that time, had only been observed in the Bay of Bengal, such as Laodicea indica. The rarity observed by Navas-Perreira & Vannucci [5] could be explained by the location of their samples, which were mainly collected at deep offshore stations, whereas these meroplanktonic species are probably issued from colonies living on continental shelves and, therefore, likely poorly represented in such offshore areas. In contrast, the Hydroidolina species Bougainvillia fulva, Crossota alba, and Cytaeis tetrastyla, common on the east and west coasts of India [10], were not recorded in our study. However, we found numerous unidentified individuals of Cytaeis which could include Cytaeis tetrastyla (Appendix B: Plate 1A–B), reported as the fourth most numerous meroplanktonic hydrozoan species in the Indian Ocean [16] and recorded in the Agulhas Current along the east coast of South Africa, i.e., the study area closest to Reunion Island in the southwestern Indian Ocean [19]. Bougainvillia fulva (but not Crossota alba) has also been reported along the east coast of South Africa [19]. It is noteworthy that only seven species of Hydroidolina from Reunion Island are common to the east coast of South Africa: Amphinema australis, Cnidocodon leopoldi, Corymorpha bigelowi, Corymorpha forbesii, Euphysilla pyramidata, Laodicea ?undulata, and Proboscidactyla ornata. Conversely, Amphinema dinema, Cirrholovenia tetranema, Eucheilota tropica, and Phialella quadrata (Reunion’s species with more than 10 specimens collected) were not listed [19]. The invasive Blackfordia virginica reported in different oceans and seas all over the world (see [53] for a review), including the east coast of South Africa [19], was not found, probably because our study was carried out in an area facing coral reefs, whereas this species is known to inhabit estuaries. However, once again, the eastern South African samples came from oceanic areas, making these results difficult to compare with ours. Furthermore, although the South African study was conducted over a 12-year period, many of the species collected consisted of only one or two specimens, as in our one-year study.

According to Bouillon & Boero [54], the diversity and abundance of the different orders of Hydromedusae are linked to their life cycle. More recently, a global study using integrative metabarcoding and environmental data (Tara Oceans expedition) suggests that the loss of the benthic stage (considered an evolutionary stage of hydrozoans) leading to holoplanktonic cycles results in a decrease in diversity, while facilitating colonization of the open ocean [55]. Indeed, the four dominant holoplanktonic species appear to be highly tolerant of different ecological conditions, as they have a cosmopolitan distribution. According to Santhakumari & Nair [10], these cosmopolitan holoplanktonic species (independent of the bottom to complete their life cycle) are warm water species, euryhaline, eurythermal, and fairly tolerant of variations in dissolved oxygen levels. However, cosmopolitanism (or circumtropicalism) for hydrozoan species is sometimes questioned for both medusae (e.g., refs. [56,57]) and polyps (e.g., refs. [58,59,60,61]) due to the existence of cryptic species revealed by genetic analysis.

All of these results confirm the trend of a general inverse relationship between biomass (i.e., abundance in our study) and species diversity already pointed out in India [52]. Since then, numerous studies conducted in different regions of the world have reported the same trend, not only for cnidarian species (e.g., refs. [62,63,64]), but also for various gelatinous zooplankton [65].

4.2. Hydromedusae Spatial Distribution

4.2.1. Coastal Versus Offshore Stations

The species richness of Hydromedusae, Hydroidolina, and Trachylinae, and the abundance of Hydroidolina were consistently higher at coastal stations than at offshore stations (Figure 3 and Figure 4). This result is not surprising concerning the meroplanktonic Hydroidolina, as it could be correlated with the decreasing species richness of benthic hydroids from the fringing reef’s outer slope towards the open sea (pers. obs.). The same pattern should apply to benthic hydroids on the reef bank and reef platform of CAPL and BOUC, respectively. Indeed, hydroids are generally the first organisms to colonize available space, settling on various hard substrates, either inert or living [66,67]. These hard substrates suitable for hydroids are rare in our offshore stations, due to more sandy bottoms. Moreover, many individuals were juveniles; therefore, they were probably sampled not far from their mother colonies. A few species seem to prefer deeper habitats: Bougainvillia auriantiaca, B. bitentaculata, B. platygaster, Euphysilla pyramidata, Hybocodon sp., Leuckartia sp., and Podocorynoides minima were only recorded in offshore stations. More surprisingly, the species richness of holoplanktonic Trachylinae was also higher in coastal stations than offshore. However, among them, three species (Aegina citrea, Haliscera conica, and Sminthea eurygaster) were only collected in offshore stations. Conversely to the Hydroidolina, the abundances of Trachylinae and Hydromedusae were similar between coastal and offshore stations and are greatly correlated with the abundances of the main Trachylinae species: Aglaura hemistoma.

4.2.2. Between Sites, and Reef Versus Non-Reef Zones

The species richness and abundance of Hydromedusae, Hydroidolina, and Trachylinae were similar between sites (coastal and offshore stations combined) and between reef and non-reef zones. The most numerous meroplanktonic species, such as Amphinema dinema, Corymorpha forbesii, Laodicea sp., Phialella quadrata, and Proboscidactyla ornata, were present in both reef and non-reef zones. However, several Hydroidolina seem to prefer coral reef habitat (Bougainvillia principis, Cirrholovenia polynema, C. tetranema, Clytia hemisphaerica, Euphysa sp., Halocoryne frasca, Hydractinia sp., and Staurodiscus tetrastaurus), while others seem to prefer non-reef zones (Amphinema australis, Eucheilota tropica, Halitiara formosa, Obelia sp., Protiara tetranema, Teisseria australe, Vellela vellela, and Zanclea ?sessilis). However, these trends should be viewed with caution because many Hydroidolina were only collected once throughout the entire study (Table S1), and above all, 194 individuals among them were not identified as they were juveniles or damaged.

Regarding Trachylinae, the similar species richness and abundance between sites, as well as between reef and non-reef zones, was not surprising. Since Trachylinae are independent of the bottom, ocean currents transport them to Reunion Island. They come from the center of the Indian Ocean (via the South Equatorial Current) and occasionally from the south due to strong swells that sometimes occur during the winter months (July to September). The island’s oceanic location, combined with its small size, meant that no differences were observed between sites, which are located only a few kilometers apart (Figure 1b).

4.3. Temporal Distribution

The monthly species richness and abundance of Hydromedusae, Hydroidolina, and Trachylinae were similar during the entire study, from November 2005 to October 2006. The majority of Trachylinae species were present all year, as well as the Hydroidolina Amphinema dinema, Clytia spp., Corymorpha forbesii, Cytaeis spp., and Zanclea spp. Some species appear to show seasonality, but given the small number of individuals collected, it is difficult to draw any clear conclusions on this point.

Some medusa swarms (sensus [37]) were detected during the survey, but they were always associated with those of Clytia spp. or of the dominant Trachylinae. However, the Scyphomedusa Thysanostoma flagellatum and the Cubomedusa Alatina alata beached periodically throughout the years, depending on the meteorological conditions in the open ocean (pers. obs.). Moreover, Pruski & Miglietta [68] showed that it is only through highly frequent sampling (i.e., several times per week) that days of high abundance can be detected due to the brief presence of the medusae. Therefore, all swarms of Hydromedusae may not have been detected by our weekly or biweekly samplings.

4.4. Hydroidolina vs. the Known Local Hydroid Fauna

Regarding the meroplanktonic Hydroidolina species, when we compare the present results with what we know about the benthic hydroid fauna of Reunion Island, some conclusions become evident. First, half of the genera (12 out of 24) are inventoried for the first time, thus greatly increasing the local knowledge of the hydrozoan fauna. Second, the Anthoathecata exhibit a much higher species richness than the Leptothecata in this study. This result contrasts with the Leptothecata’s common prevalence in the benthic hydroid communities in shallow waters of islands from the Mozambique Channel (southwest Indian Ocean) [69,70,71], Reunion Island, and of the central Indian Ocean [72]. It is explained by the important presence of the Macrocolonia families (e.g., Sertulariidae and Aglaopheniidae), which lack medusae in their life cycle.

Leptomedusae hydroid polyps were checked locally for the following genera, but matching them with the medusae at the specific level is often not possible. For example, the several colonies of Aequorea sp. sampled over time were all sterile, and only juvenile medusae were obtained for the numerous colonies of Clytia spp. (gracilis, hummelincki, linearis, etc…) and Obelia spp. (dichotoma, geniculata) observed. It is also more difficult for hydroids previously attributed to the Campanulinidae, and usually sampled as tiny sterile colonies, to be attributed to a family (Laodiceidae, Lovenellidae, Phialellidae) as they are based on medusa morphology and have similar hydroid polyps. Therefore, local hydroid species of the genera Cuspidella and Campanulina, for example, cannot be linked to the medusae sampled during this study. The case of Cirrholovenia tetranema Kramp, 1959, is an exception. Indeed, after the discovery of its life cycle [73], we learned that it is the medusa of Egmundella amirantensis Millard and Bouillon, 1973. This species has been reported several times in Reunion Island in its benthic stage, and it is documented for the first time as a medusa in this study. The medusa Staurodiscus tetrastaurus can also be linked to Hebella scandens [74], whose hydroid polyps are frequently found in our waters, associated with Aglaophenids.

Among the Anthoathecata, several medusa genera were reported to have an unknown hydroid stage, including Cnidocodon, Dicnida, Euphysilla, Podocorynoides, and Protiara [23]. For others, conversely to the Leptothecata, the benthic stage that provides medusae has never been found in Reunion Island until now, including Amphinema, Corymorpha, Euphysa, Euphysilla, Halitiara, Halocoryne, Leukartiara, Podocorynoides, Proboscidactyla, Protiara, Teissiera, and Zanclella. About the Aplanulata (Appendix A), the two specimens of Hybocodon sp. in this collection are probably matching with hydroids sampled recently in Reunion Island [24]. Inside the Filifera, small colonies of hydroids of the family Bougainvilliidae were found but rarely and always with few stolonal sterile polyps, which therefore could not be attributed to an appropriated genus; the single Hydractinia species known is without medusa; two Turritopsis hydroid species are recorded in Reunion, which differ in the size of their nematocysts, and one of them, which had produced juvenile medusae, has been provisionally assigned to T. chevalense, thus matching with the present results; at last, the species Cytaeis nassa was reported previously only from the benthic stage, then it is here the first sample of the medusa. For the Capitata, it is interesting to notice that Halocoryne orientalis was the single species of this genus found in the Indian Ocean literature. The present findings indicate that this genus is present in the local benthos, though it has not been reported until now, and that a second species, H. frasca, is now recorded in the Indian Ocean (Table S1, Appendix A). Colonies of several Zanclea species have already been checked in Reunion, but it is not possible one more time to match them with the present medusa records, as the colonies were sterile or had produced juvenile medusae that were not reared to maturity. Notable are the captures of the two rare species whose polyps were not reported in the area: Zanclea medusopolypata (Appendix B: Plate 3B), for which this is only the third record of this Indo-Pacific species, and Zanclella diabolica (Appendix B: Plate 3C–E), of which one more medusa was recently checked with benthic material, suggesting the presence of hydroid polyps in the lower mesophotic area (Gravier-Bonnet, pers. obs.). Polyps of Euphysilla were recently described for the first time in the Maldives as being similar to those of Sphaerocoryne based on genetic analyses [75]. Therefore, it is possible that some of the Sphaerocoryne spp. specimens sampled in Reunion Island correspond to the medusa Euphysilla pyramidata discovered during this study. Unfortunately, the medusae were not preserved with alcohol at the time of sampling (2005–2006); only formalin was used. Therefore, it was not possible to match the DNA of the polyps with that of the medusae.

5. Conclusions

This study, carried out in the shallow waters off the west coast of Reunion Island, provided the opportunity to get a small view of jellyfish, an important group of the marine fauna that had never been studied in our area until now. This study allowed us to identify 62 species of medusae, including 57 species of Hydromedusae. This small oceanic island is home to a few holoplanktonic species characteristic of tropical and subtropical zones, as well as numerous meroplanktonic species with warm water affinities. The abundance of medusae is mainly due to four holoplanktonic species, which are equally distributed between reef and non-reef zones, as well as between coastal and offshore stations, throughout the study year. Due to its geographical position and limited continental shelf, Reunion Island appears to be protected from jellyfish blooms. However, this initial study will serve as a baseline for future research, particularly in estimating the evolution of jellyfish composition and abundance in the context of climate change and new coastal management in Reunion Island.