INTRODUCTION
The swimming crab Charybdis smithii MacLeay, Reference MacLeay and Smith1838 (Portunidae) and the mantis shrimp Natosquilla investigatoris (Lloyd, Reference Lloyd1907) (Squillidae) are common crustaceans of the tropical Indian Ocean. They exhibit peculiar life cycles with an extended pelagic phase for larvae, juveniles, sub-adults and even adults (Losse & Merrett, Reference Losse and Merrett1971a; Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009). They are found periodically offshore in dense aggregations, both at the surface and through the upper mixed layer.
Adult C. smithii are demersal; their reproduction is associated with shelf and slope habitats (Apel & Spiridonov, Reference Apel, Spiridonov, Krupp and Mahnert1998; Türkay & Spiridonov, Reference Türkay, Spiridonov, Krupp and Mahnert2006, Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009). This swimming crab is a short-lived species with an annual lifespan (van Couwelaar et al., Reference van Couwelaar, Angel and Madin1997). Individuals may reach 75 mm in carapace width (CW) (van Couwelaar et al., Reference van Couwelaar, Angel and Madin1997) and become mature at 45–50 mm of CW (Balasubramanian & Suseelan, Reference Balasubramanian and Suseelan1998). The biology of the mantis shrimp N. investigatoris is poorly known. Apart from infrequent extreme abundance events reported from pelagic waters (Losse & Merrett, Reference Losse and Merrett1971a; Potier et al., Reference Potier, Marsac, Cherel, Lucas, Sabatie, Maury and Ménard2007a), this species has been recorded in a few rare instances from the benthic environment of the western Indian Ocean (Lloyd, Reference Lloyd1907; Chopra, Reference Chopra1939; Venema, Reference Venema1975, Reference Venema1976). Most of the time, Natosquilla investigatoris likely persists at very low population densities or occupies areas and depths that are rarely surveyed.
Charybdis smithii and N. investigatoris play an important role in the pelagic ecosystem both as predators of small fish and cephalopods (Losse, Reference Losse1969; Losse & Merrett, Reference Losse and Merrett1971a; Balasubramanian & Suseelan, Reference Balasubramanian and Suseelan1998; Mincks et al., Reference Mincks, Bollens, Madin, Horgan, Butler, Kremer and Craddock2000) and as prey for top predators such as tunas, billfishes, lancetfish, dolphinfish, sharks and rays (Potier et al., Reference Potier, Marsac, Cherel, Lucas, Sabatie, Maury and Ménard2007a; Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009). At times, these swimming crabs and mantis shrimps may constitute the main prey item of tunas, contributing more than 50% by wet weight to the diet (Losse & Merrett, Reference Losse and Merrett1971a; Potier et al., Reference Potier, Marsac, Cherel, Lucas, Sabatie, Maury and Ménard2007a; Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009).
The western Indian Ocean is a highly dynamic monsoon-driven ecosystem exhibiting seasonally alternating oceanographic regimes that follow seasonal variability in the atmospheric circulation. In particular, the north-west part of this area is the location of one of the strongest seasonal upwellings in the world ocean (Longhurst, Reference Longhurst2007). Monsoon seasonality and upwelling-induced productivity has a profound effect on the pelagic ecology of the region, and is likely to be an important influence on both the distribution and periodic extreme abundance of the two species discussed here (Potier et al., Reference Potier, Ménard, Cherel, Lorrain, Sabatie and Marsac2007b; Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009).
Both species have occasionally been found stranded (Losse & Merrett, Reference Losse and Merrett1971a; Branch, Reference Branch1984) but such events remain poorly documented. Stranding of swimming crabs, including the event which led to the original taxonomic description of this species (MacLeay, Reference MacLeay and Smith1838), have been reported from South African coasts (Figure 1A) (Branch, Reference Branch1984; van Couwelaar et al., Reference van Couwelaar, Angel and Madin1997; Branch et al., Reference Branch, Griffiths, Branch and Beckley2002), even though mass pelagic occurrences of C. smithii are unknown from this region (Kensley, Reference Kensley1977; van Couwelaar et al., Reference van Couwelaar, Angel and Madin1997; Türkay & Spiridonov, Reference Türkay, Spiridonov, Krupp and Mahnert2006; Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009). Strandings of N. investigatoris were noted in 1967 on two remote islands of the Seychelles Archipelago during a large-scale episode of pelagic swarming in the equatorial western Indian Ocean (Losse & Merrett, Reference Losse and Merrett1971a). More recently, strandings were observed along the east African coast (Tanzania) (Kamukuru & Mgaya, Reference Kamukuru and Mgaya2004; MBREMP, 2005; Asseid et al., Reference Asseid, Drapeau, Crawford, Dyer, Hija, Mwinyi, Shinula and Upfold2006) and on islands of the Seychelles Archipelago (Anonymous, 2002; Fonteneau et al., Reference Fonteneau, Ariz, Hallier, Lucas, Pallares and Potier2004) (Figure 1B).
Geographic distribution of stranding and sedimentation records of Charybdis smithii MacLeay, Reference MacLeay and Smith1838 and stranding observations of Natosquilla investigatoris (Lloyd, Reference Lloyd1907) in the Western Indian Ocean: (A) Swimming crab Charybdis smithii. This study: dots are strandings. Published records: stars are strandings (MacLeay, Reference MacLeay and Smith1838*; Branch, Reference Branch1984*; van Couwelaar et al., Reference van Couwelaar, Angel and Madin1997*), and crosses are sedimentation from pelagic realm (Christiansen & Boetius, Reference Christiansen and Boetius2000); (B) Mantis shrimp Natosquilla investigatoris. This study: dots are strandings, and oblique crosses are reports from telephone survey in the Maldives. Published records: stars are strandings (Losse & Merrett, Reference Losse and Merrett1971a; Anonymous, 2002**; Fonteneau et al., Reference Fonteneau, Ariz, Hallier, Lucas, Pallares and Potier2004**; Kamukuru & Mgaya, Reference Kamukuru and Mgaya2004**; MBREMP, 2005*; Asseid et al., Reference Asseid, Drapeau, Crawford, Dyer, Hija, Mwinyi, Shinula and Upfold2006**). For references marked with * positions of observations were estimated from the description published in the original paper, for references marked by ** central position of island, where stranding occurs was used. On all panels the 200 m isobath (dark line) and bathymetry from 1000 to 5000 m (in 1000 m steps, light lines) are shown. Positive latitude is north from equator, negative is south from equator; longitude is east from the prime meridian. Coastline and bathymetry data are from GEBCO (GEBCO, 2010).
Mass mortalities may reflect either the consequences of a natural life cycle (semelparity, i.e. single reproductive episode before death (Green, Reference Green2008)) or a catastrophic event, as has been suggested for C. smithii stranding (Branch, Reference Branch1984). However, no explanation has yet been proposed for stranding in N. investigatoris.
In this work we report for the first time observations of mass mortalities associated with stranding of C. smithii in the equatorial Indian Ocean within the main area of its pelagic distribution, as well as several similar events affecting the mantis shrimp N. investigatoris in the same region. For N. investigatoris we document an extension of its known range to the Maldives Archipelago. We provide also a summary of all stranding events for both species recorded in the Indian Ocean. Finally for each species we discuss hypotheses of likely causes of stranding based on field observations and available biological information.
MATERIALS AND METHODS
Stranding sites are tiny oceanic islands and/or coral atolls scattered in the western Indian Ocean from east African waters to the Maldives (Figure 1). Geographic positions of strandings and observational details are presented in Supplementary Tables S1 and S2.
Charybdis smithii 2005–2011
SEYCHELLES
A stranding of crabs was recorded at North Island by the staff of Wilderness Safaris, which managed a large-scale island rehabilitation programme. Beach patrollers noticed and photographed unusual crab occurrences and reported areas of stranding, estimated numbers and associated data. A dive instructor reported in-water observation of crab behaviour, distribution and environmental conditions before stranding. The event was documented with a digital photo camera, but samples of crabs were not collected.
MALDIVES
Strandings of C. smithii were observed and documented in four islands of North Malé Atoll by co-author RCA. In 2005 seventeen crabs were sampled for identification and morphometric measurements. In 2009 and 2011 stranded crabs were photographed.
Natosquilla investigatoris 2002–2005
SOUTH-WEST INDIAN OCEAN
Strandings were observed on several islands located in the Mozambique Channel (Europa and the Glorieuses) and on Bird Island (Seychelles). Strandings were commonly preceded by dense swarms of mantis shrimps aggregated along the coastline or submerged shoals (Geyser Bank). Observations were documented by co-authors JPQ and FM and several French scientific teams with photo and video cameras. Eight fresh individuals were collected at Glorieuses Islands.
MALDIVES
Swarming associated with strandings was recorded by RCA around two islands in South-west Ari Atoll. A few days later a telephone survey was performed by SAS among administration offices of principal Maldivian atolls. Presence/absence of mantis shrimp swarms and stranding were documented throughout the archipelago.
All collected specimens of both species were measured to 0.1 mm: total length (TL) for mantis shrimps and carapace width (CW) for crabs.
RESULTS
Charybdis smithii (Figure 1A, Supplementary Table S1)
SEYCHELLES
Charybdis smithii (Supplementary Figure S1A, Supplementary Table S1) stranded in August 2009 on western beaches of the North Island (Figure 1A). The crabs washed ashore were still alive but lethargic, dying shortly afterwards. An observer estimated that half of the beach edge was covered with crabs, amounting in total to 300–500 crabs (10–15 crabs per linear metre of beach). A week before the stranding, crabs were observed in the water by divers. Crabs were reported several times in large numbers, aggregated near the surface at 1 m depth. Surface water temperature was 26–27°C. Crabs were very active in the water, swimming energetically without obvious interaction. No mating, spawning or other specific behaviours were observed and females did not carry any eggs on their abdomen. On days following the mass stranding, only one individual swimming crab was encountered though dives were performed regularly.
MALDIVES
During the three stranding events observed in the Maldives Islands from 2005 to 2011 (Figure 1A, Supplementary Table S1), large numbers (hundreds to thousands) of C. smithii (Supplementary Figure S1B) were seen swimming in shallow waters (inside lagoons). No obvious signs of parasites or other epifauna were found on any of the sampled crabs. Some individuals exhibited signs of recent moulting. Size (CW) of stranded crabs ranged within 20–25 mm, corresponding to juvenile individuals.
Natosquilla investigatoris (Figure 1B, Supplementary Table S2)
GLORIEUSES ISLANDS
During the mass stranding event of N. investigatoris observed in November 2002 (Supplementary Figure S2), the abundance of stomatopods accumulated at the upper tide mark was estimated at about 500 ind m−2. Individuals washed ashore were already dead. Before stranding, large dense swarms of N. investigatoris were observed three times from the deck of a vessel, and by divers over a wide area of sandy shoal (Supplementary Figure S3). After the stranding, no more swarms of stomatopods were observed in the vicinity of the islands.
Swarms were swimming in the water column from the surface to the bottom (~5 m depth). Swimming individuals were very active, and predation on a small flatfish (Pleuronectidae) was observed. Swarming mantis shrimps exhibited polarized swimming with a distance of around 15 cm between individuals. Over a sandy bottom mantis shrimps actively burrowed into the sand. Our records are the first documented observations of N. investigatoris burrowing behaviour in the natural environment. TL of live sampled mantis shrimps ranged from 51 to 63 mm. No egg was noticed on sampled individuals although developed cement glands and ventral telson with fused ovary (characteristic for reproductively active females) (Wortham-Neal, Reference Wortham-Neal2002a) were noted. Ovaries on advanced stage of maturation were found in one preserved sampled female dissected in January 2015.
In May 2003 and 2004 few stranded individuals (less than 100, i.e. no mass stranding) were observed (M. Le Corre, personal communication). No samples were collected, but TL was estimated between 30–40 mm.
GEYSER BANK
In November 2004 some N. investigatoris were observed swarming in the water column, and exhibited burrowing behaviour at the bottom. Sampled female individuals had developed cement glands and ventral telson with fused ovary. One year later (October 2005) a few small individuals (exact size unknown) were observed at the sea surface.
SEYCHELLES, BIRD ISLAND
A few stranded individuals of ~30–40 mm TL (less than 100, i.e. no mass stranding) were observed in June 2004 (M. Le Corre, personal communication). No samples were collected.
MALDIVES
Swarms and stranding were first recorded in shallow waters around two islands (Hanghghaameedhoo and Huvahendhoo) in South-west Ari Atoll in December 2002. Several swarms comprised thousands of individuals. A sample of mantis shrimps collected at Huvahendhoo was lost in transit; the size of the individuals collected was about 80 mm TL. No other information on size was recovered. A telephone survey within the Maldives produced reports of mantis shrimps swarming from islands throughout almost the entire length of the archipelago (from 7°N to the equator) (Figure 1B). These reports suggest that the swarming and stranding episode lasted from late November to late December (Supplementary Table S2). Mantis shrimps were observed mostly as freely swimming swarms in the ocean in close proximity to atolls but in several cases were reported swarming in lagoons and stranded on reefs and beaches. The numbers of mantis shrimps occurring in the Maldives was in the millions of individuals.
DISCUSSION
Strandings of pelagic and benthic crustaceans have been observed in upwelling systems sporadically (Branch, Reference Branch1984; Stewart et al., Reference Stewart, Yochem and Schreiber1984; Gracia et al., Reference Gracia, Díaz-Garcia and Velázquez-Simental1986) or on a regular basis (Aurioles-Gamboa et al., Reference Aurioles-Gamboa, Castro-González and Pérez-Flores1994, Cockcroft et al., Reference Cockcroft, van Zyl and Hutchings2008). However the reasons usually remain unknown.
In the Indian Ocean, mass mortality and strandings have been observed in several diverse groups including coelenterates (Billett et al., Reference Billett, Bett, Jacobs, Rouse and Wigham2006; Daryanabard & Dawson, Reference Daryanabard and Dawson2008), crustaceans (Losse & Merrett, Reference Losse and Merrett1971a; this study), and fishes (Foxton, Reference Foxton1965; Panikkar & Jayaraman, Reference Panikkar and Jayaraman1966; Smith et al., Reference Smith, Hammond and Close2010). However, crustacean strandings in the Indian Ocean are rarely reported, even for such occasionally abundant species as C. smithii and N. investigatoris.
Charybdis smithii is present throughout the year in the pelagic environment of the equatorial Indian Ocean. Its biomass is highly variable, both on seasonal and annual scales, with densities exceeding 15,000 ind km−2 during abundance peaks (Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009). Nevertheless, mass strandings of C. smithii have previously only been reported on few occasions from the south-western periphery of the species’ range, along the southern coast of South Africa from East London to Cape Town: in 1834–36, 1982–83 and 1993–1996 (MacLeay, Reference MacLeay and Smith1838; Branch, Reference Branch1984; van Couwelaar et al., Reference van Couwelaar, Angel and Madin1997). Although it has been suggested that crab strandings in South Africa are annual events (B. Newman in van Couwelaar et al., Reference van Couwelaar, Angel and Madin1997), no subsequent evidence of such periodicity appears to have been reported. Prior to this study, no stranding of C. smithii had been reported from its main pelagic distribution area, namely the western equatorial and northern Indian Ocean. This supports the hypothesis of Branch (Reference Branch1984) who attributed strandings along the South African coast to environmental anomalies. He reported mass mortalities of several species that are usually rare at the southern tip of Africa in association with a warm event in 1982–1983: C. smithii, Portuguese man of war (Physalia sp.) and blue button (Porpita sp.). In the equatorial Indian Ocean C. smithii inhabits waters within the temperature range of 30°C (at the surface) to 10°C (below the thermocline) (Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009). Thus warm oceanographic anomalies may facilitate the southward expatriation this species, with subsequent mortality apparently caused by recovering normal (relatively cold) conditions. However, pelagic swarms of C. smithii have never been reported from South Africa.
No reason was found for the crab strandings in Seychelles and Maldives reported here. No abnormality was reported, beyond the existence of daytime surface aggregations with relatively high swarm density (less than 1 m between crabs) and their advection into shallow waters. In normal conditions C. smithii are extremely rare in shallow waters (Apel & Spiridonov, Reference Apel, Spiridonov, Krupp and Mahnert1998; Türkay & Spiridonov, Reference Türkay, Spiridonov, Krupp and Mahnert2006) and do not aggregate at the surface during daytime (Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009).
Crab ‘swarms’ observed at the surface during twilight and at night in the open ocean are always ‘loose’ aggregations of a density 0.1–0.2 ind m−2 (Della Croce & Holthuis, Reference Della Croce and Holthuis1965; van Couwelaar et al., Reference van Couwelaar, Angel and Madin1997; Christiansen & Boetius, Reference Christiansen and Boetius2000). Maximum density reported earlier in the upper surface layer (0–150 m) reached 0.016 ind m−3 (van Couwelaar et al., Reference van Couwelaar, Angel and Madin1997) but much lower densities, about 0.0001 ind m−3 are usually found in the same stratum (Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009). The noticeably higher densities (about 1 ind m−3) reported here before stranding suggest apparently unusual crab behaviour before they were washed ashore. Environmental stress or disease might alter crab behaviour. However, no major anomaly in sea temperature or in overall weather conditions was observed before or during our stranding observations. The recorded temperatures during stranding in Seychelles (26–27°C) do not exceed the range commonly observed in the crab's pelagic habitat (Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009). Observed temperature variations are much lower than the temperature gradient regularly crossed by the crabs during their diurnal vertical migrations in the open ocean, which can exceed 15°C (Romanov et al., Reference Romanov, Potier, Zamorov and Ménard2009). These crabs apparently also tolerate low oxygen concentrations (Karuppasamy et al., Reference Karuppasamy, Balu, George, Persis, Sabu and Menon2007; Longhurst, Reference Longhurst2007). The active swimming behaviour observed during stranding did not correspond to animals possibly incapacitated by stress. No external visible parasites were noted on stranded crabs.
Mass mortalities do not always result in stranding. In September–October 1995 mass sedimentation of immature (CW = 34–44 mm) dead individuals of C. smithii was observed at three deepwater stations (3190–4030 m depth) in the Arabian Sea (Christiansen & Boetius, Reference Christiansen and Boetius2000). Within a period of 15 days, fresh dead individuals of swimming crab were found at the stations separated by 250–370 nmi (station positions: 16°10′N 60°30′E; 14°30′N 64°30′E; 20°00′N 65°35′E. 1 nautical mile = 1.852 km.). The oceanic area inside the triangle formed by the stations where crab mortality was observed exceeded 40,000 nmi2 (1 nmi2 = 3.43 km2) i.e. over 130,000 km2. Christiansen & Boetius (Reference Christiansen and Boetius2000) avoided the assumption of a single large-scale crab mortality episode, treating each observation as a non-connected event. Similar observations of large-scale mass mortality of jellyfish in the Arabian Sea were reported in 2002 (Billett et al., Reference Billett, Bett, Jacobs, Rouse and Wigham2006; Daryanabard & Dawson, Reference Daryanabard and Dawson2008).
While no obvious natural factors responsible for the mortality of these pelagic organisms were identified (Christiansen & Boetius, Reference Christiansen and Boetius2000; Billett et al., Reference Billett, Bett, Jacobs, Rouse and Wigham2006, Daryanabard & Dawson, Reference Daryanabard and Dawson2008) the scale and irregularity of such events, involving unrelated taxonomic groups, suggests the possibility of ‘catastrophic’ causes. Several factors may trigger mass mortalities of marine organisms, e.g. algal blooms, very high temperatures and/or oxygen depletion. The Arabian Sea is indeed known for anoxic conditions and the presence of hydrogen sulphide in subsurface waters (Longhurst, Reference Longhurst2007).
In contrast to the swimming crab, strandings of the mantis shrimp N. investigatoris are always associated with recurrent ‘bloom’ episodes. These were reported in 1944, 1965–1967, 1971–1976 and 2000–2005, when dense swarms of thousands or even millions of individuals were observed swimming at the surface or stranded on beaches (Barnard, Reference Barnard1950; Losse & Merrett, Reference Losse and Merrett1971a, Reference Losse and Merrettb; AzCherNIRO, 1972; Venema, Reference Venema1975, Reference Venema1976; Fonteneau et al., Reference Fonteneau, Ariz, Hallier, Lucas, Pallares and Potier2004; Kamukuru & Mgaya, Reference Kamukuru and Mgaya2004; MBREMP, 2005; Asseid et al., Reference Asseid, Drapeau, Crawford, Dyer, Hija, Mwinyi, Shinula and Upfold2006). Strandings reported here coincide with the most recent mantis shrimps ‘bloom’, probably the strongest one ever observed, covering six contiguous years (2000–2005) (authors' unpublished data). During this period stomatopod swarms and strandings were spread from East African waters (Kamukuru & Mgaya, Reference Kamukuru and Mgaya2004; Asseid et al., Reference Asseid, Drapeau, Crawford, Dyer, Hija, Mwinyi, Shinula and Upfold2006) including the North Mozambique Channel (this paper), through the open ocean and islands of Seychelles archipelago (Anonymous, 2002; Potier et al., Reference Potier, Marsac, Lucas, Sabatie, Hallier and Ménard2004, Reference Potier, Marsac, Cherel, Lucas, Sabatie, Maury and Ménard2007a, Reference Potier, Ménard, Cherel, Lorrain, Sabatie and Marsacb; Malone et al., Reference Malone, Buck, Moreno and Sancho2011) to the Maldives (this paper) (Figure 1B). Swarms of N. investigatoris had never been reported from the Maldives before. The present observations considerably expand the known range of this species: previously the eastern-most record was collected from the continental slope off the eastern tip of the Arabian Peninsula in ~60°E during the R/V ‘Mabahiss’ cruise (Chopra, Reference Chopra1939). During 2000–2005 event, active burrowing behaviour was observed on sandy shoals, and the presence of developed cement glands and ovaries on sampled individuals were both suggestive of active reproductive status of females in these abundant swarms (Hamano & Matsuura, Reference Hamano and Matsuura1984; Wortham-Neal, Reference Wortham-Neal2002a).
Mantis shrimps are bottom-dwellers inhabiting a wide range of habitats: from shallow waters to bathyal depths down to 1500 m, and from clear sandy shoals and coral reefs to turbid muddy estuaries (Reaka, Reference Reaka1980; Reaka & Manning, Reference Reaka and Manning1981; Kodama et al., Reference Kodama, Shimizu, Yamakawa and Aoki2004). They also exhibit a diversity of lifestyles, behaviours and adaptations from solitary dispersed individuals (Ahyong, Reference Ahyong and Poore2004) to abundant species supporting target fisheries (Kodama et al., Reference Kodama, Shimizu, Yamakawa and Aoki2004; Maynou et al., Reference Maynou, Abelló and Sartor2004). The Stomatopoda are divided into two broad functional and morphological groups relating to the manner of prey capture: ‘smashers’ and ‘spearers’ (Caldwell & Dingle, Reference Caldwell and Dingle1976). Most ‘spearers’ occur in dense aggregations inhabiting soft substrates (sand or mud) on the shelf and slope, living in self-excavated burrows where they lay their eggs (Caldwell & Dingle, Reference Caldwell and Dingle1976; Wortham-Neal, Reference Wortham-Neal2002a, Reference Wortham-Nealb). Morphologically N. investigatoris is a typical ‘spearer’: burrowing in soft sediments may represent a natural attempt of reproductively active females to construct burrows and to reproduce.
While semelparity is rare among crustaceans and not previously reported for stomatopods (Thiel & Duffy, Reference Thiel, Duffy, Thiel and Duffy2007), mass mortality observed immediately after swarming might indicate such a mode of reproduction in N. investigatoris.
These swimming crabs and mantis shrimps are common and abundant species in the western Indian Ocean, but their biology and life cycles are still poorly known. Mortality occurring in response to environmental variability or as a natural part of the life cycle, and sometimes resulting in strandings, is an important aspect of their natural history. Further studies, including meticulous registration of such events in combination with biological sampling and environmental observations should provide further information on the biology of both species, and further our understanding of their roles and interactions in the pelagic ecosystem of the tropical western Indian Ocean.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit http://dx.doi.org/S002531541500096X
ACKNOWLEDGEMENTS
We would like to thank Jenny Huggett (Oceans and Coasts, Department of Environmental Affairs, South Africa) who first attracted our attention to crab stranding at North Island and to Linda Vanherck (Environment Officer, North Island, Seychelles) for her willingness to share this information with us. We are thankful to Sheena Talma (Environment Assistant of North Island) who reported onshore observations and provided us with photos of stranded crabs and to Steve Shipside (Dive Instructor of North Island) for information on crab behaviour in the water. In the Maldives, our thanks are due to Angus Nicholl and the many island officials who reported strandings, as well as to Ahmed Hafiz who assisted with field work. Critical comments of Dr Vassily Spiridonov and an anonymous referee improved the manuscript.
