The behaviour of organisms can influence how their populations are structured. Marine organisms generally have a hierarchy of recognisable population structures. The broadest distinction in the hierarchy is the geographic range of a species, which often equates to a metapopulation. The metapopulation can be further divided into a series of mesopopulations (also known as stocks) which are largely self-contained. Local populations with strong connectivity may also be identified within a stock. Because the scales and structure of stocks are determined by the movement of animals between locations, data on the behaviour and mobility of animals in relation to local currents are crucial for understanding the population dynamics of species. Connectivity may be influenced by the combined effects of oceanographic transport and the behaviour of larvae, juveniles and adults. The medusae of Jellyfish in the class Cubozoa (Box Jellyfish) generally have strong swimming and orientation abilities, and so they may have great capacity to influence the structuring of their populations. Chironex fleckeri is the largest Cubozoan species, and their swim speed is amongst the fastest recorded for any Jellyfish (16.6 cm per second). The swim speeds of smaller species have been investigated in laboratory settings and were also found to be considerable (Chiropsella bronzie, 12 cm per second, Chiropsalmus sp., 6.7 cm per second, and Tripedalia cystophora, 4 cm per second against a 1.5 cm per second current). Furthermore, Cubozoan medusae are nearly fully formed after they metamorphose from polyps, so Cubomedusae can probably swim well and influence their dispersion soon after metamorphosis. Cubozoans also have sophisticated visual systems. They have four rhopalia and each contains two image forming eyes, similar in structure to the eyes of Vertebrates and Cephalopods. Given the sophistication of their visual system, it is unsurprising that Cubozoan species exhibit an array of visually guided behaviours including: obstacle avoidance (Chironex fleckeri, Tripedalia cystophora, Carybdae rastonii, and Chiropsella bronzie), navigating via terrestrial cues (Tripedalia cystophora), and orienting toward prey species (Copula sivickisi and Tripedalia cystophora).
In a paper published in the journal Marine Biology on 3 March 2020, Jodie Schlaefer of Marine Biology and Aquaculture at James Cook University, the ARC Centre of Excellence for Coral Reef Studies, and TropWATER at James Cook University, Eric Wolanski, also of Biology and Aquaculture and TropWATER at James Cook University, Shreya Yadav of the Marine Biology Graduate Program at the University of Hawaiʻi at Mānoa, and Michael Kingsford, again of Marine Biology and Aquaculture at James Cook University, the ARC Centre of Excellence for Coral Reef Studies, present the results of a study intended to elucidate the fine scale spatial structure of a population of the Box Jellyfish Copula sivickisi inhabiting a fringing reef, and to determine the importance of behaviour (namely diel cyclic behaviours, attachment to habitat and swimming) in maintaining this structure.
The species-specific data required to evaluate how behaviour might influence the structuring of populations are largely lacking for Cubozoan species. The Cubozoan Copula sivickisi (formerly known as Carybdea sivickisi) has a cosmopolitan distribution and is one of the better studied species in the class. A complex suite of behaviours has been observed in Copula sivickisi medusae. They have adhesive pads on the apex of their bells which they can use to attach to substrates. The species is nocturnal; SCUBA divers have observed Copula sivickisi foraging at night in the wild, and extensive plankton tows in daylight hours have previously yielded no Copula sivickisi medusae. Further, in laboratory experiments, medusae attached themselves to the sides of their tanks during the day and were actively swimming, foraging and mating in the water column at night. Copula sivickisi also seem to preferentially attach to favourable substrata. In a laboratory experiment, when given a choice between Coral, stone, Red Algae, Seagrass and the tank control, most medusae attached to the undersides of hard structures (Coral or stone) or Seagrass. SCUBA divers have also observed Copula sivickisi medusae swimming close to and adhering to the benthic Macroalgae Sargassum spp. and Colpomenia spp. in the wild. The habitat associations of Copula sivickisi and how these preferences influence patterns of distribution have rarely been addressed in the literature; data on their associations in the wild are especially lacking. Together, the behaviours exhibited by Copula sivickisi medusae have the potential to greatly influence their dispersal and population structure.
The Box Jellyfish, Copula sivickisi. Garm et al. (2015).
Copula sivickisi has a broad distribution, inhabiting nearshore waters in the Pacific and Indian Oceans at tropical to temperate latitudes. The substructure of the metapopulation(s) inhabiting this broad distribution is largely unknown. However, a previous study analysed the shapes of Copula sivickisi statoliths (calcareous particles that stimulate sensory receptors in response to gravity, so enabling balance and orientation) taken from the medusae from three different locations on the east Australian coast, separated by hundreds of kilometres. Twenty statoliths were analysed per location and their shapes differed significantly between locations, suggesting that the populations inhabiting the locations were separate stocks. Further divisions may exist within these stocks, but data have not been collected at sufficiently small spatial scales to distinguish these divisions. Collecting data to assess Cubozoan population structures can be difficult as the abundance of Cubozoans can vary greatly spatially and temporally. Many Cubozoan medusae are photopositive and night lighting, where a light is used as an attractant, has been utilised to sample Cubomedusae. Furthermore, digital cameras mounted to jetties have effectively been used to monitor the presence of relatively large photopositive Cubomedusae (Chironex fleckeri and Morbakka spp.) attracted to the field of view by strong lights. Schlaefer et al. developed a similar system in which an underwater camera and light were paired to record the abundance of smaller Cubomedusae.
Schlaefer et al .specifically aimed to: (1) expand on previous laboratory experiments examining the behaviour of Copula sivickisi medusae through time, (2) experimentally determine their habitat preferences in a fringing reef environment, (3) perform depth stratified plankton tows in day and night hours to investigate how wild Copula sivickisi medusae are vertically distributed in the water column through time, (4) deploy underwater Jellyfish cameras to investigate the habitat use, and related geographic distribution, of a population of Copula sivickisi medusae on a fringing reef and, (5) quantify the swimming ability of Copula sivickisi medusae and compare it with the speeds of currents measured in their natural habitat.
Copula sivickisi medusae were collected at night from September to November of 2012 and 2016 from Geoffrey Bay and Nelly Bay on the eastern side of Magnetic Island The island lies on the central section of the Great Barrier Reef, approximately 8 km from the coast of Townsville, Queensland, Australia. Bays on the eastern side of Magnetic Island contain fringing reefs that are dominated by Sargassum sp. Algae and Coral species.
The study region. (a) Australia, the North Queensland coastline. The locations of panes (a) and (b) are indicated by the grey boxes. (b) Magnetic Island. The locations of Middle Reef (MR), Picnic Bay (PB), Nelly Bay (NB), Geoffrey Bay (GB), Alma Bay (AlB), Arthur Bay (ArB) and Florence Bay (FB) are shown. The extents of panes (c) and (d) are indicated by the dark grey and light grey boxes, respectively. The designs of the (c) 2015 and (d) 2016 Jellyfish Camera unit surveys. The white squares mark the locations of the sampled sites. The bathymetry of NB and GB is shown, the colour bar indicates the depth (m). Depths of more than 10 m are shown in black. In all panes, land is filled with a hatch pattern. Reefs are filled with solid grey in panes (a) and (b) and they are outlined in black and filled with dots in panes (c) and (d). Schlaefer et al. (2020).
The Copula sivickisi medusae were collected using light attraction. In 2012, a 1000-watt light was placed within the top 2–4 m of the water column for at least 30 min and the attracted medusae were gathered from the surface with pool scoop nets (rectangular mesh size: 1 × 2 mm). The medusae were transported back to a temperature controlled wet lab at James Cook University. The indirect sunlight in the lab gave the medusae a light/dark cycle of approximately 12:12 hours. The physical characteristics of the water column were measured during flood and ebb tides on 15 November 2012 using a conductivity, temperature and depth device. The water column was vertically well mixed. The average water temperature ranged from 26.7 °C on flood tide to 26.9 °C on ebb tide. The temperature in the lab was set to 26 °C, near the measured average. The average water salinity ranged from 36.0 parts per thousand on flood tide to 36.1 parts per thousand on ebb tide. The salinity in the lab was maintained at approximately 36 parts per thousand.
In 2016, weighted lights were submerged in 1.5–5 m of water for up to 1.5 hours; medusae aggregated around the lights and were collected by snorkelers with pool scoop nets. They were transported back to James Cook University and fed plankton that had been caught from the collection site on the night of capture. Medusae were held in an artificially lit, temperature controlled wet lab. The lights were set to turn on at 06:45 and off at 18:45, giving the medusae a light dark cycle of 12:12 hours. The temperature was again set to 26 °C and the salinity was maintained at approximately 36 parts per thousand.
In both years, medusae (size range: 1–10 mm inter pedalial distance) were kept in 100 litre holding tanks (depth 47 cm, radius 28 cm) and half water changes were performed daily. Medusae were fed Artemia nauplii (Brine Shrimp larvae) daily after the night of capture. The Artemia naulii were hatched in brackish water without an enrichment medium. While Cubomedusae can live on a diet of Artemia, it may be nutritionally deficient. Consequently, fresh medusae were caught on a weekly or bi-weekly basis for use in experiments to avoid confounding holding effects. All experiments were performed within two weeks of capture and the condition of the medusae did not deteriorate in this time.
The behaviour of Copula sivickisi medusae was monitored through time. In each trial, medusae were transferred from the holding tanks into a 25 litre (depth 35 cm, radius 14 cm) bucket at 3.00 pm. Live Artemia nauplii were then added to the bucket. The number of medusae performing predefined behaviourswas recorded after a 1 hour acclimation period. Observations were made every half an hour between 4.00 pm and 9.30 pm. The room light was shut off at 6.:45 pm, which corresponded to the normal time of dusk in the region during the Copula sivickisi season (September to November). Therefore, in each trial, six day time observations (4.00–6.30 pm) were made under a fluorescent light that recreated the tropical blue sky light spectrum, and six night time observations (7.00–9.30 pm) were made in darkness under a fluorescent light with a red filter. A previous study used red fluorescent light to film Copula sivickisi at night and concluded that the medusae performed natural behaviours in the tanks, suggesting that the red light did not disrupt their natural activity pattern. In Schlaefer et al.'s study, adults were observed in five trials, each with five males and five females, and juveniles were observed in five trials, each with ten juvenile medusae.
All statistical analyses were performed in R version 3.4.3. Schlaefer et al. tested the null hypothesis that the relative proportions of active and inactive behaviours performed by Copula sivickisi medusae were independent of the time of day. Separate tests of this hypothesis were done for the adult and juvenile trials. First, the data in each trial were pooled by time (day or night) and behaviour (active: swimming, feeding and mating; inactive: attached and bobbing) to produce a 2 × 2 contingency table for each trial. Therefore, a set of five contingency tables were generated from the adult trials and a set of five were generated from the juvenile trials. Repeated test of independence needed to be performed on the contingency table sets so.
The searching and attachment behaviour of Copula sivickisi medusae was examined in a habitat choice experiment. The behaviour of medusae was compared over three habitats and one control as follows. Tufts of Sargassum and a fragment of dead coral rubble (Montipora sp.) were gathered from the sites of Copula sivickisi collection. The natural substrates were placed into a 9 litre rectangular tank (length 28.5 cm, width 18.5 cm, depth 17 cm) with sand so each substrate took up approximately one quarter of the bottom of the tank. The natural substrates and the sand were chosen for inclusion in this experiment as they were the predominant substrates at the sites of Copula sivickisi collection. The final quarter was left empty as a control. Black shade cloth was wrapped around the sides of the tank to obscure the observer from the medusae’s view. In each trial, an individual medusa was taken from the holding tank and placed in the centre of the and the quadrat they were in were recorded every 15 seconds for ten minutes. The number of times the medusa was recorded in each quadrat was summed. The Sargassum tended to overhang the boundaries of its quarter and the fragment of coral rubble did not fill its entire quarter.
Diagrammatic representation of the tank set up in the habitat choice experiment. Schlaefer et al. (2020).
Fifty medusae were trialled in daylight hours, between 10.00 am and 6.00 pm, when they were most likely to be inactive. Thirty nine of the 50 trialled medusae attached to a substrate within the 10-minute period. The search time was calculated as the time it took these medusae to attach to a substrate. Once attached, most medusae remained attached for the duration of the trial. However, four medusae attached to more than one substrate during the trial period. The time it took these medusae to find the substrate they spent the most time on was taken as the search time.
The swimming abilities of Copula sivickisi medusae were tested against increasing water velocities in a swimming chamber (length 45.5 cm, width 25 cm, depth 6.5 cm). The water speed within the chamber was controlled by multi-turn gate valves and calibrated by measuring how fast food dye moved through the chamber. A total of 41 medusae (23 males, 18 females), ranging in size from 4 to 11 mm, were trialled. Each medusa was acclimated in the swimming lane in a 1 cm per second current for 5 minutes before the start of each trial. The water speed was then increased to 3 cm per second for 5 minutes, and subsequently increased by 3 cm per second every 5 minutes up to a speed of 18 cm per second. This was done to simulate the strengthening of a current with a rising tide. The range of speeds used matched the current speeds measured near the site of Copula sivickisi collection. The trial was suspended once the medusa became fatigued. A medusa was considered fatigued if it was being pressed against the end of the tank by the current, or if it attached itself to the tank via the adhesive pads on its bell. The highest velocity maintained for a whole five minutes and the time elapsed at fatigue elapsed at fatigue were used to calculate the maximum swim speeds of the Copula sivickisi medusae.
The sprint swim speed was defined as the maximum speed a medusa could perform in a burst and was calculated as the highest speed a medusa swam against for at least 2.5 minutes.
The vertical distribution of Copula sivickisi in the water column was examined in the wild during the day and at night to determine if the nocturnal behaviour of medusae produced changes in their distribution over a 24 hour period. Depth stratified plankton tows were performed in Geoffrey Bay and Nelly Bay, during day (3.50–6.30 pm) and night (7.30–9.50 pm) hours. Tows were performed in all states of the tide (high, ebb, low, flood). A 320 μm mesh net with a circular mouth, 75 cm in diameter, was towed slowly (0.25–0.93 m per second) for 5–10 minutes and a flow meter attached to the net recorded the distance travelled in each tow. Water volumes between 60 and 120 m³ were sampled in the tows and the measure of Copula sivickisi medusae abundance was accordingly standardised to the number recorded per 100 m³. The neuston layer (thin surface layer with many organisms) was skimmed in the shallow tows. The bottom 1 m of the water column, corresponding to depths between 2 and 7 m, was sampled with diver-controlled deep tows. Each bay, depth and time combination was sampled twice per day over three non-consecutive days, from 8 October 2014 to 7 November 2014. A total of 48 tows were, therefore, performed, with 12 replicates per depth and time combination. No statistical analyses were performed on these data due to the absence of Copula sivickisi medusae in most tow samples.
The geographic distribution of Copula sivickisi was determined over two bays (Geoffrey Bay and Nelly Bay) on the eastern coast of Magnetic Island over two medusae seasons (2015 and 2016). Their distribution was mapped using underwater Jellyfish camera units paired with an underwater light. Medusae were attracted to the light and recorded by the video camera. The video cameras were set to record at 1080 pixel resolution, with a frame rate of 30 frames per second. The Jellyfish camera were deployed in 30-minute intervals, in organised grid patterns. The maximum number of medusae in any single frame of video during the 30 minute deployment period was used as the measure of medusae abundance; this is the convention used when managing potential repeat counts.
The availability of fringing reef habitat at each of the mapped sites was also qualitatively determined from the Jellyfish camera footage to investigate the link between medusae abundance and habitat availability. Sargassum and Coral dominated the fringing reef habitat. The habitat availability at a site was classified as high if reefal habitat took up greater than 66% of the substrate in still images from the Jellyfish camera footage recorded at the site. The habitat availability was classified as moderate if between 66 and 33% of the substrate was covered by reef and it was classified as low if less than 33% of the substrate was covered. Finally, the habitat availability was classified as absent at a site if no Sargassum or Coral were visible in any of the stills extracted from the Jellyfish camera footage.
A small-scale grid was sampled in Geoffrey Bay in 2015 to determine the habitat and depth usage of Copula sivickisi medusae in the bay at a fine spatial scale. The grid had 12 sites; the GPS locations of the sites were predetermined from satellite imagery to guide the night field sampling. The sites were separated from each other by 100 m and set out in a 3 × 4 pattern, so the grid covered a 300 × 400 m (0.12 km²) area. The three grid rows corresponded to three distinct depth strata (under 4.1 m, 4.2 to 7 m, over 7.1 m) and were set at incremental distances from a band of fringing reef habitat that was dense enough to be visible from the satellite images. The shallowest sampling row lay within the visible habitat band, the mid-depth row lay just outside of the band (range: 35–70 m from band) and the deepest row was set far from the band (range: 124–168 m from band). The placement of the sites in relation to reef habitat was later ground-truthed from the Jellyfish camera footage. The grid was sampled four times over four non-consecutive nights, from the 1st of October to the 3rd of September.
Jellyfish Camera unit design. A diagram of a deployed Jellyfish camera, with labelled components. A close up photograph of a Jellyfish camera is provided in the insert. Schlaefer et al. (2020).
In 2016, a larger grid, spanning both Geoffrey Bay and Nelly Bay, was sampled to investigate the fidelity of Copula sivickisi medusae to fringing reef habitat, and the related potential extent of the local population. The grid tracked the bands of fringing reef habitat located in the bays and was extended beyond the edges of the bands. The grid had 12 sites, with two replicate locations per site, and the GPS locations of the replicates were again predetermined from satellite imagery. The sites were separated from each other by 350 m in the longshore direction. The replicates within sites were separated by 100 m in the cross-shelf direction. The 2 × 12 grid, therefore, covered an approximate area of 100 × 3850 m (0.385 km²). In the post-hoc analysis of the 2015 Jellyfish camera results, the shallow and mid-depth rows were grouped together, and they were both grouped separately from the deep row where Copula sivickisi medusae were rare. Consequently, when a site in the 2016 grid was adjacent to dense fringing reef habitat, the GPS locations of the replicates in the site were set to mirror the placement of the shallow and mid-depth sites in 2015. The shallower replicate was placed on the habitat dense enough to be visible from the satellite images and the deeper replicate was placed just outside of the dense habitat (range: 38–109 m from habitat). Sites that lay away from the visible habitat were placed along the same longshore contour as the habitat. Again, the placement of the sites in relation to habitat was ground-truthed from the Jellyfish camera footage post sampling. The depths of the shallower row of replicates ranged from 1.3 to 7.8 m, and the depths of the deeper row ranged from 4.8 to 11.6 m. The whole grid was sampled three times over six nonconsecutive nights, from the 27th of September to the 31st of October.
The grid was then expanded to determine the extent of the horizontal distribution of medusae. Additional sampling was conducted on the 3rd of November and the 9th of November. Six sites (three in Geoffrey Bay and three in Nelly Bay) which intersected dense fringing reef habitat were sampled for a fourth time. Additional sites in Alma Bay, to the north east of Geoffrey Bay, and Middle Reef, south west of Magnetic Island, were each sampled once.
Current speeds in waters inhabited by medusae were measured using drogues. The drogues consisted of two A4 sized acetate sheets placed one on top of the other and tied together with fishing line to form a thicker, more robust A4 rectangle. The joined sheets were weighed down by small fishing sinkers. The sheets were tied to 1 m long ropes with two small buoys on the end, chosen to reduce wind drag. A small strobe light also attached to each drogue so they could be recovered at night. A total of 95 drogue deployments were done at sites on the eastern coast of Magnetic Island in September, October and November of 2015, 2016 and 2017. Drogues were dropped in Geoffrey Bay, Nelly Bay, Alma Bay, Middle Reef, Picnic Bay, Florence Bay, and Arthur Bay. Deployments were made at the surface over water depths corresponding to the rows of the 2015 Jellyfish camera deployments. Drogues were left to drift for 10 min, and the mean current speeds and directions of drift were calculated from the GPS points taken at deployment and pick up. Drogues were deployed in all states of the tide (high, ebb, low, flood) and at spring and neap tides. Drogues could only be tracked in winds of less than 15 knots (7.72 m per second).
Copula sivickisi adult and juvenile medusae were most active at night. The relative proportions of medusae performing active (swimming, feeding and mating) and inactive (attached and bobbing) behaviours changed significantly with the time of day for both adults and juveniles. There was a 30% increase in active behaviour from day to night and a 60% decrease in inactive behaviour. The change was more pronounced in the juveniles than the adults.
Analyses by hour indicated contrasting patterns of activity between day and night for adults and juveniles. Adult medusae were observed actively mating more and feeding more at night, which coincided with a reduction in the number of medusae that were swimming. Medusae were observed to mate almost exclusively at night; the greatest incidence was recorded at 8.00 pm, when an average of 20%of adult medusae were observed performing the characteristic ‘wedding dance’ mating behaviour. There was a trend for feeding in adult medusae to increase toward dusk and stay high at night, with the percentage feeding remaining around 60%. Increases in the percentage of adult medusae feeding were mirrored by reductions in the percentage swimming; an average of only 13.3% of adult medusae swam at night.
Conversely, in the juvenile trials, the increase in activity at night was driven by an increase in the incidence of swimming, and not feeding. Feeding peaked in the hour before lights were out and remained high at the first night time observation. Feeding levels then nearly returned to where they had been before the peak. Comparatively, the percentage of juvenile medusae swimming nearly doubled from an average of 27.7% in day light hours to an average of 54.7% at night.
Night-time reductions in the observed frequency of inactive behaviours occurred in both the adult and the juvenile trials. The average percentage of medusae attached dropped by approximately 10% from day (27.7%) to night (16.7%) in the adult trials, and it dropped by 13% (from 19.3% to 6.3%) in the juvenile trials. The bobbing behaviour was not observed after 7.45 pm in any of the adult trials and it was not recorded in any of the night time observations in the juvenile trials.
Copula sivickisi medusae showed a preference for Sargassum sp. Algae over the other substrates in the habitat choice experiment. Eight of the ten (80%) medusae that attached to the Sargassum had corrected search times less than or equal to 1 minute 38 seconds. Medusae generally took much longer to attach to the other substrata; the majority of the medusae that attached to the tank control (10 of 14; 71%) and to the sand (10 of 14; 71%) had corrected search times longer than 1 minute 38 seconds. On average, medusae attached to the Sargassum 1.6 times faster than they attached to both the tank control (2 minutes 49 seconds) and the sand substrates (2 minutes 40 seconds). However, this difference was not statistically significant. Only one medusa attached to the dead Montipora sp. coral rubble, and it had a relatively long corrected search time of 6 minutes 55 seconds.
The Copula sivickisi medusae that encountered the Sargassum and remained in contact with it for a short time tended to attach to the Algae quickly. The medusae that attached to the Algae, therefore, spent little time in the other substrate quadrants before attaching. In contrast, the medusae that attached to the tank and to the sand spent more time searching before they chose to attach; they spent considerable time in both the tank and sand quadrants, and generally had little contact with the Algae. The medusa that attached to the Montipora Coral rubble was only counted in the sand quadrant and the coral quadrant in the period before it attached. Furthermore, 11 of the 50 medusae trialled in the habitat choice experiment 2 did not attach to any substrate. These medusae had nearly 5 times as much exposure to the tank control and sand quadrants compared to the Algal quadrant. They were also counted in the Coral quadrant more than twice as many times as they were counted in the Algal quadrant.
The abundance of Copula sivickisi medusae in the small-scale grid in Geoffrey Bay differed significantly with depth/habitat availability. Copula sivickisi medusae were most abundant at sites in the shallowest depth stratum, which all had high Sargassum density and/or Coral cover. There was a trend for medusae to be less abundant at the mid-depths where habitat availability was moderate or high, but the difference between shallow and mid-depths was not significant. Medusae were rare in the deepest stratum where habitat availability was either low or absent. The abundance of Copula sivickisi medusae at the deep sites was 11 and 7 times less than the abundances at the shallow and mid-depth sites, respectively; these differences were large and statistically significant. Depth accounted for 38.6% of the variability in medusae abundance.
The population of Copula sivickisi medusae inhabiting Nelly Bay and Geoffrey Bay was largely restricted to reef habitat. Copula sivickisi medusae were almost always present, and often highly abundant, at sites in the 2016 sampling grid where the habitat availability was high in at least one replicate. Medusae were absent at two of the three sites where there was no reefal habitat and they were rare at the third. A Jellyfish camera recorded a single medusa attaching to a leaf of Sargassum, further indicating that medusae interact with Algae.
Sampling at two other locations indicated that the presence of reef habitat did not guarantee the presence of medusae. All sampled sites within Alma Bay and Middle Reef had either high or moderate reef availability. Copula sivickisi medusae were present in Alma Bay in moderate abundance and were absent from Middle Reef. Alma Bay is relatively sheltered while Middle Reef lies in open water.
The Copula sivickisi medusae swam at speeds comparable to the surface currents measured in the field. The greatest burst swim speed sustained by a medusa for more than half of a trial interval was 12 m per second. However, only 15% of medusae tested reached this speed. The average maximum swim speed of the trialled medusae was 4.9 m per second. There was high variation in maximum swim speed over the full size range of medusae. Seventeen of the 41 medusae were observed attaching to a side of the tank during their trial to avoid being washed backwards; 13 attached in slow currents (less than 3 cm per second) and four attached in faster currents (above 6 cm per second).
The surface current speeds increased with total depth of the water column. he weakest currents were generally measured by the drogues deployed above depths equivalent to the shallow sites in the 2015 Jellyfish camera survey. The greatest burst swim speed of the trialled Copula sivickisi medusae was faster than all but one of the shallow water current measurements. The maximum swim speed of the medusae exceeded most of the shallow water surface current speeds and was approximately double the median. The greatest burst swim speed of the medusae exceeded most of the surface current speeds measured by the drogues that were deployed in waters with total depths equivalent to the mid and deep sites. The maximum swim speed of the Copula sivickisi medusae was slower than a majority of the mid and deep surface currents; although, maximum swim speed was approximately 85% of the median speed measured at mid-depths and 90% of the deep-water median.
Copula sivickisi medusae inhabited shallow/mid-depth reefs where Sargassum sp. Algae and Coral was abundant and current speeds were low. Where structured habitats were lacking, Copula sivickisi medusae were rare or absent. Complex daily behaviour cycles, where they were most active at night, combined with active swimming and a near substratum distribution suggested that dispersal from local populations would be low under normal conditions.
The diurnal activity pattern of Copula sivickisi medusae recorded in the laboratory and in the wild (plankton tows) affirms the species is nocturnal. In Schlaefer et al.'s study, medusae were detected in the night time plankton tows and were practically absent from the day time tows. A previous study also conducted day and night plankton tows, although their nets were drawn up vertically through the water column (approximately 5 m deep), starting at different depths. That study similarly found that Copula sivickisi medusae were present in the night samples and absent from the day samples. Furthermore, Schlaefer et al. observed major daily cycle behavioural shifts in the laboratory, where the level of activity increased from day to night over the scale of hours. In contrast the previous study recorded behavioural shifts over the scale of minutes. For example, in an experiment where the previous researchers turned a light on at night, all of the observed medusae stopped swimming and attached themselves to the tank within 30 min of the introduction of the light. The differences in day time and night time activity levels reported in Schlaefer et al.'s study were also less pronounced than those reported in the earlier study. Nearly 100% of the medusae observed by the earlier researchers were attached in day light hours and an average activity level of greater than 70% was maintained in all night time hours. In the present study, greater than 60% of adult medusae were still active during the day on average, and the percentage increased to above 80% at night.
It is possible that there were some tank artefacts that influenced behaviour. Medusae were observed swimming and feeding in the tank during the day. In the wild, Copula sivickisi medusae may hunt planktonic Crustaceans at night by swimming to areas of high prey density, guided by the flashes given off when bioluminescent Dinoflagellates (e.g. Pyrocystis noctiluca) contact zooplankton. Bioluminescent flashes were observed during the night sampling in Schlaefer et al.'s study. The Copula sivickisi medusae Schlaefer et al. kept in the laboratory were fed Artemia nauplii, which are not bioluminescent. Consequently, the medusae that hunted in the tank at night would not have been guided by the flashes they associate with their normal prey. During the day, the orange colour of the Artemia would have contrasted against the white walls of the tank. This contrast may have triggered a feeding response in the Copula sivickisi medusae. Contrast has been found to guide the behaviour of the Cubozoan Tripedalia cystophora.
Copula sivickisi medusae displayed a preference for Sargassum and avoided Coral in the laboratory. SCUBA divers have previously observed Copula sivickisi medusae swimming near and attaching to macroalgae (Sargassum spp. and Colpomenia spp.) in the wild. Similarly, medusae were captured swimming close to and attaching to Sargassum in the Jellyfish camera footage. Contrastingly, the underside of Coral was the preferred habitat of Copula sivickisi medusae in the habitat choice experiment conducted previously. Few medusae attached to the Calcified Red Alga Gracilaria sp. in the same experiment. The association between Copula sivickisi medusae and fringing reef habitat revealed in the Jellyfish camera surveys was most likely driven by a preference for Sargassum and not Coral.
Copula sivickisi medusae are capable swimmers. The average maximum swim speed of the medusae (4.9 cm per second) exceeded most of the surface current speeds measured in the shallow waters they inhabit. Furthermore, they could swim much faster in bursts (up to 12 cm per second). They were also collected within one meter of the bottom in the stratified plankton tows, where the currents would be expected to be weaker than the surface currents due to current shear.
The swim speeds of other cubomedusae in the size range of Copula sivickisi medusae (under 1 cm interpedalial distance; the distance between the bases of pedalia on one side of bell) have been measured and are comparable to the speeds measured in Schlaefer et al.'s study. Chiropsella bronzie medusae with interpedalial distances less than 1 cm were recorded swimming at maximum speeds of approximately 3 cm per second. Another study measured Tripedalia cystophora medusae ranging in size from 0.8 to 1.2 cm bell diameter swimming at maximum speeds of 3–4 cm per second against a 1 to 1.5 cm per second current in a flow chamber. Adding the highest speed and current measurements, the Tripedalia cystophora medusae could have been swimming at maximum speeds of up to 5.5 cm per second, similar to the maximum swim speed of Copula sivickisi medusae recorded by Schlaefer et al.. Larger Cubomedusae have been recorded swimming at faster speeds. The study of Chiropsella bronzie recorded a specimen with an interpedalial distance of 5.6 cm swimming at a maximum speed of approximately 12 cm per second, while another study reported a members of the same species with 3–5 cm bell diameters swimming at maximum speeds of 9.5 cm per second. Chironex fleckeri medusae ranging in size from 4 to 12 cm interpedalial distance have been measured swimming at speeds of up to 16.6 cm per second in bursts in the wild. Increases in cubomedusae swimming performance with size have been reported previously. One study found the opposite relationship but this was likely due to the observed wild medusae not swimming at their greatest capacity.
The formula used to determine the maximum swim speed of Copula sivickisi medusae in Schlaefer et al.'s study was developed for Fish larvae. To Schlaefer et al.'sknowledge, no such formula has been defined that directly relates to Jellyfish. Jellyfish medusae have been found to travel more energetically efficiently than other swimming organisms, expending less energy per meter travelled. Further, compared to medusae that swim by rowing, the jetting propulsion of Cubomedusae is significantly greater at utilising the passive energy recapture mechanism that partially enables the energetically inexpensive transport of Jellyfish. Despite the efficiency of Cubomedusae swimming, the maximum speeds of Copula sivickisi medusae determined from the Fish larvae method matched the reported maximums of similarly sized Cubomedusae derived from alternate techniques. This suggests that the Fish larvae method provided a true representation of the swimming capabilities of Copula sivickisi medusae.
The Jellyfish cameras used to map the distribution of Copula sivickisi medusae proved to be highly effective at detecting the small medusae in enough detail to positively identify the species. Copula sivickisi medusae have been observed by SCUBA divers in their natural environment but such observations are rare and efforts to observe Copula sivickisi in the wild have previously yielded no medusae. Cameras have already been used to detect larger jellyfish, but from above the water. The Jellyfish camera system presented by Schlaefer et al. is novel because it used lights underwater. The technique allowed for the simultaneous sampling of medusae over relatively large areas. It was also cost effective; each unit was relatively inexpensive. However, the presence of the light likely affected the behaviour of the medusae, so their natural behaviours could not be inferred from the Jellyfish cameras footage. To observe the natural behaviours of medusae, the camera field of view could be illuminated by a red light imperceptible to the medusae. The medusae could be attracted by additional lights simulating the intermittent bioluminescent flashes from Dinoflagellates that purportedly guide hunting in Copula sivickisi medusae.
The quality of the Jellyfish camera footage varied based on several factors including the placement of the Jellyfish camera on the substratum, the position of the camera in relation to the light and the presence of dense plankton and Algae in the field of view. Some of the footage was also unusable due to instrument failure (i.e. the camera or light malfunctioning or shutting off before the end of the 30-minute deployment); though loss of samples from Jellyfish camera failure was less than 9%. Therefore, Schlaefer et al. recommend that the mapping of the distribution of Cubomedusae with Jellyfish camera or similar technology is done with high enough replication to account for the loss of replicates.
Mapping the distribution of Copula sivickisi medusae revealed that they were generally only found at sites where the availability of fringing reef habitat, rich in Sargassum, was moderate to high. Moreover, areas of reef that had lower habitat availability generally had less medusae. The behaviour of Copula sivickisi medusae is likely critical for maintaining their observed distribution on fringing reef habitat. Certainly, many of the behaviours performed by Copula sivickisi medusae have the potential to limit their dispersion including limiting their activity, attaching to hard substrates and staying near the bottom of the water column.
Copula sivickisi medusae are only active for a portion of the day due to their diurnal activity pattern, limiting their exposure to dispersive currents. Even when medusae are active at night, the swim trial results suggest it is unlikely that they would swim against dispersive currents for extended periods. In the trials, some medusae attached to the tank when exposed to currents speeds of 6 cm per second or greater for less than a trial interval (5 minutes). Wild medusae could rationally attach to reefal habitat to avoid being flushed away in strong currents and then remain attached until the currents weakened. Copula sivickisi medusae have been observed attaching to reef structures such as Sargassum, Red Algae, Hard Coral and stone. Medusae were also found to maintain positions near the bottom of the water column where they would not have to swim far to find a surface to attach to. Furthermore, medusae near the bottom would likely experience currents weakened from current shear and they could counteract dispersion by swimming into the weakened currents. Numerous Copula sivickisi medusae exhibited rheotaxis (swimming against a current to remain stationary) in the swim trials, swimming directly into the current for minutes at a time. The Cubozoans Chiropsella bronzie and Tripedalia cystophora have previously exhibited rheotaxis in a laboratory setting and the scyphozoan Rhizostoma octopus has been documented swimming counter to the mean current direction in the wild. While some insight was gained from the swim trials, more research is required to determine the maximum swimming duration of Copula sivickisi medusae and the strategies they utilise to prevent expatriation. They are small, have transparent bodies, and are inactive during the day. Further laboratory experiments would likely be required to elucidate the behaviour of Copula sivickisi medusae in currents because of the extreme difficulty in observing them in their natural environment.
Other cubozoan species have been found to perform behaviours that enable the maintenance of distribution patterns. Tripedalia cystophora medusae are most commonly found at the fringes of Mangroves; they maintain positions in this preferred habitat by using their specialised eyes to peer up through the water surface to detect and navigate by the Mangrove canopy. Chironex fleckeri medusae inhabit estuarine and nearshore coastal waters. They are highly mobile and often swim longshore, and probably toward the shore, to maintain their nearshore distribution and local populations. Like the reefal habitat inhabited by Copula sivickisi, the Mangroves, estuaries and coasts inhabited by other Cubozoans are characterised by ‘sticky water’, where complex structures reduce flow and facilitate retention.
Maintaining positions near fringing reef habitat may have numerous advantages for Copula sivickisi medusae, in addition to providing a refuge from dispersive currents. The reefal structures (e.g. Algal tufts and Hard Coral) probably provide some protection against predators, as has been shown for other reef species (e.g. Fish). There may also be important reproductive benefits. The convergence of medusae to a thin band of reef would put them near potential mates and provide an abundance of structures and niches to attach their sticky embryo sacs to.
The absence of medusae from sites off the fringing reef habitat band that were devoid of Sargassum and Coral suggests that the Copula sivickisi medusae inhabiting Nelly Bay and Geoffrey Bay represent a local population. In addition to their swimming/attachment behaviour, the reproductive behaviour of Copula sivickisi medusae may assist in maintaining this restricted distribution. Copula sivickisi gametes are not dispersed via broadcast spawning because medusae have internal fertilisation, and the medusae may selectively attach their embryo sacs to reefal habitat. The early life history stages of Copula sivickisi medusae are also seemingly adapted to limit dispersion. The planula larvae that arise from the embryos can attach to hard surfaces. Once settled, they develop into sessile polyps. The medusae that bud from Cubozoan polyps are nearly fully formed, and could probably swim competently to counteract dispersion soon after metamorphosis. The swimming capabilities of newly metamorphosed medusae were not assessed by Schlaefer et al., and this would be an interesting avenue for future research.
The spatial extent of the stock that the identified local population belongs to is still unknown. A previous study collected statoliths from Copula sivickisi medusae inhabiting locations on the east Australian coast that were separated by hundreds of kilometres. That study found that the shapes of the statoliths differed significantly by location, suggesting that the Copula sivickisi stocks had a maximum extent of hundreds of kilometres. The potential for stock differentiation to occur at much smaller spatial scales is great given the non-dispersive reproductive behaviour of Copula sivickisi medusae, the stickiness of Copula sivickisi at all life stages, and the demonstrated strong swimming abilities of medusae. Copula sivickisi medusae were absent from Middle Reef which had suitable habitat but is relatively exposed, given its location between Magnetic Island and mainland Australia. This suggests that the island population could be isolated from any mainland Copula sivickisi populations. A separation of Magnetic Island and mainland stocks has been reported for the larger cubozoan Chironex fleckeri. The statolith elemental chemistry of Chironex fleckeri medusae caught from the island and the adjacent mainland differed significantly indicating that the medusae had experienced different environmental conditions and were, therefore, from separate stocks.
The seemingly poor dispersal potential of Copula sivickisi contradicts the expansive extent of the metapopulation. The Copula sivickisi phenotype has been recognised throughout the Pacific and in the Indian Ocean. Such a cosmopolitan distribution is rare in Cubozoan species and, given their poor dispersal abilities, it suggests a long geological history where range expansion occurred with the movements of tectonic plates (i.e. vicariance theory). It is highly likely that medusae from many locations within the metapopulation are genetically distinct enough to be recognised as incipient species (i.e. genetically distinct but morphologically similar with reproductive compatibility). Populations of Scyphozoans inhabiting environmentally similar systems have diverged into incipient species through genetic isolation (Mastigias sp., Catostylus mosaicus). Furthermore, a previous study found that different environmental conditions drove divergent selection and speciation in the cosmopolitan Scyphozoan Jellyfish genus Aurelia (Moon Jellyfish). There is great capacity for similar divergent selection across the tropical and temperate latitudes inhabited by Copula sivickisi. However, speciation is probably comparatively slower in Cubozoans given the low number of described species. Populations of the Cubozoan Carybdea marsupialis from coastal locations across the Mediterranean, covering hundreds of kilometres, were found to represent populations of the same species as they were genetically and morphologically similar. The population genetics of species in the more pelagic Cubozoan genus Alatina were also recently analysed; the species in the genus likely represent a single species with a pantropical distribution. Similar analyses are currently being conducted for the Copula genus. Genetic analyses have been combined with biophysical modelling (hydrodynamic models coupled with models of plant/animal behaviour) to effectively identify population structures in other Jellyfish. Similar analyses could help to elucidate the hierarchical structure of Copula sivickisi populations, given the complex behaviour of medusae.
Storms and the attachment of polyps to drifting Sargassum could provide mechanisms for increasing the scales of connectivity between Copula sivickisi populations or for facilitating speciation. Medusae may not be able to maintain their positions on reefal habitat in extreme weather events, and adrift medusae could be transported by the currents. Furthermore, the Sargassum species that Copula sivickisi are associated with are annuals that shed the sporophyte seasonally. The loss of Algal cover peaks in spring, coinciding with the presence of Copula sivickisi medusae at Magnetic Island. While medusae are likely to detach from drifting Algae, as has been found for other invertebrate species, the embryo sacs, planula larvae and polyps of Copula sivickisi medusae could remain attached. Drifting macroalgae can travel for kilometres transporting any attached fauna with it. To survive, the Copula sivickisi would probably need to metamorphose into medusae before the drifting Algae is washed onto the shore or into the open sea. Expatriated Copula sivickisi could establish new populations or connect existing ones. The founder effect may affect a new population if the number of founding members is low, providing a method for speciation.
Schlaefer et al. have provided strong evidence that the behaviour of Copula sivickisi medusae is likely to minimise dispersal from local populations. Medusae were most active at night, were found close to the substratum and could swim at speeds comparable to the current speeds measured at the depths they inhabit. The new Jellyfish camera technology was highly effective in mapping the distribution of Copula sivickisi medusae. The medusae were most abundant in shallow/mid-depth waters with fringing reef habitat. The swimming ability of medusae and their restriction to bands of habitat suggests that the population of Copula sivickisi medusae inhabiting Geoffrey Bay and Nelly Bay represents a local population. While the spatial extent of Copula sivickisi stocks is unknown, the island population may be relatively isolated from anymainland populations. Despite the apparently limited dispersion potential of Copula sivickisi at all life stages, Copula sivickisi is a cosmopolitan species. This discrepancy suggests a long geologic history and Schlaefer et al. predict a level of incipient speciation within the species’ cosmopolitan distribution.
Follow Sciency Thoughts on Facebook.