Introduction

More than 70% (n = 55) of all recent mushroom coral species (Scleractinia: Fungiidae) have a free-living adult phase in their life history (Gittenberger et al. 2011; Benzoni et al. 2012), which is the so-called anthocyathus stage (Wells 1966; Hoeksema 1989). Only as juveniles in the anthocaulus stage, free-living fungiids are still attached to a solid substratum by a stalk, regardless of whether they have a single mouth or more than one (Hoeksema and Gittenberger 2010; Hoeksema and Benzoni 2013). Subsequent detachment takes place in the upper part of the stalk by dissolution of the calcareous skeleton across a distinct plane just below the polyp disc (Yamashiro and Yamazato 1996). After detachment, initially a scar remains visible on the aboral side of the polyp disc (Vizel et al. 2009; Hoeksema and Yeemin 2011).

As free-living corals, fungiids are able to colonize all kinds of substrata. These include sandy bottoms at > 20 m depth, which are usually not inhabited by many other bottom organisms and where mushroom corals can form dense aggregations (Goreau and Yonge 1968; Fisk 1983; Hoeksema 2012). They may also inhabit reef slopes where they can dominate over other benthic organisms (Goffredo and Chadwick-Furman 2000; Hoeksema and Matthews 2011; Hoeksema and Benzoni 2013). Free-living fungiids may use their mobility in densely inhabited habitats to escape from harmful interactions, or they may aggressively damage competitors (Chadwick 1988; Chadwick-Furman and Loya 1992; Hoeksema and De Voogd 2012; Hoeksema et al. 2014). The flip side of a free mode of life is that mushroom corals may accidentally get overturned by other animals or by tumbling down-slope (Jokiel and Cowdin 1976; Hoeksema 1988; Chadwick-Furman and Loya 1992). This is unfavorable because it would incapacitate their ability to capture food, for which an upright position would be most suitable (Hoeksema and Waheed 2012; Mehrotra et al. 2015).

Although it is known that mushroom corals may undergo sessile dispersal by their migration over the seafloor (Hoeksema 1988; Nishihira and Poung-In 1989; Gittenberger and Hoeksema 2013), only few reports are known concerning active mobility resulting from their own movements (Abe 1939; Horridge 1957; Hubbard 1972; Jokiel and Cowdin 1976; Yamashiro and Nishihira 1995). In the present report, we present detailed information on a flip-over experiment in which mushroom corals were overturned to document their self-righting technique with the help of time-lapse photography.

Materials and methods

Three specimens of three mushroom coral species were collected from the reef of Heron Island (Great Barrier Reef, Australia): the elongate, polystomatous Herpolitha limax (Esper, 1797), the oval-shaped, monostomatous Lobactis scutaria (Lamarck, 1801), and the disk-shaped, monostomatous Pleuractis granulosa (Klunzinger, 1879). First, they were kept in seawater basins with flowing seawater at Heron Island Research Station in normal daylight regime. For the duration of the experiment, they were brought to a laboratory and placed upside-down on a substrate consisting of coarse sand inside a self-contained aquarium. This tank is a commercially available Dymax® IQ5 Mini Acryclic Aquarium (Coburg, Australia) with a built-in filtration system and a 43 dual colour LED lamp burning continuously during the experiment. The tank contained 13 l of seawater (25 °C) and its bottom measures 27×22 cm2.

Coral movements were documented by use of time-lapse photography (Bongaerts et al. 2012). The corals had their tentacles expanded and showed rhythmic pulses of inflation and exhalation, and slightly shifted their position, but the specimens of L. scutaria and P. granulosa, were not able to turn over within an available time frame of 10 h.

One small individual of ∼7 cm in length of H. limax , with a relatively large primary mouth and three small secondary mouths aligned along its main axis, managed to overturn itself within only 6 h. The process of this flip-over is described and discussed by showing still shots extracted from a 900× time-lapse movie (see ESM). A second sequence of movements (2 h 15 min long) without overturning has been filmed to obtain additional information on the pulse frequency, which was measured as the mean time interval between two pulses, visible as shuffles and hops.

Results

The complete flip-over process consisted of four phases, which took less than 6 h to complete: (1) a stationary phase of ∼1 h in which the coral only waved its tentacles (Fig. 1a); (2) a shuffle phase (3 h 15 min long), in which the coral hopped and shifted 11 times over the sand with the oral side remaining close to the bottom and occasionally appearing to become detached from it, and with increasing force (Fig. 1b-f); (3) the flip-over phase, in which the oral side became lifted away from the sand by the coral turning over its side (Fig. 1g-i), which was concluded by a sudden flip-over at 4:51 h (Fig. 1j); and (4) the subsequent recovery phase (Fig. 1k-l). During the second movie, only part of a shuffle phase was recorded.

Fig. 1
figure 1

Still shots from time-lapse video (6 h long) showing cycles of tissue inflation and deflation in a Herpolitha limax specimen (length ∼7 cm) after being overturned: a stationary phase, b-f shuffling and hopping, g-i turning phase, j flip-over, k-l recovery. Time moments are indicated in hours and minutes, h:min). Red square: reference point

Full size image

The mean time interval between two jumps as registered in both film shots together was 22.5 min (s.d. 7.9 min, n  = 17). After every inflation the aboral side became lifted upward and sank down again, resembling hopping, with increasing amplitude but without a clear change in frequency. The body mass swelled up and curled somewhat around the exposed aboral side while the tentacles expanded and retracted. During the turning, one side gradually rose more than the other and one flip-over attempt failed to be completed (Fig. 1i). During the second try, the coral succeeded (Fig. 1j). At the start of the recovery the mouth was open (Fig. 1k), suggesting that water had been ejected from it. Eventually, the coral polyp became deflated and shrank (Fig. 1l).

Discussion

This is the first time that self-righting in free-living corals is documented, analysed and published with the help of time-lapse photography. Abe (1939) used a kimograph and direct observations to study the righting in Heliofungia actiniformis (Quoy and Gaimard, 1833), a mushroom coral species with a large, fleshy polyp and long tentacles. According to him the mouth of the coral opened just after the flip-over while it was shut during the move itself. Hubbard (1972) studied self-righting in a small mushroom coral species, Cycloseris distorta Michelin, 1842, and observed that it is able to use jet propulsion before the final flip-over. The mushroom coral of the video movie had its primary mouth open just after landing (Fig. 1k). This suggests that it was used for jet propulsion during the flip-over, which could not be seen because the oral side was directed downward and away from the camera.

Previously, time-lapse photography in laboratory conditions was applied to study sediment shedding in a mushroom coral (Bongaerts et al. 2012). This method was also used to study locomotion in Cycloseris distorta, which showed that this mushroom coral uses peristaltic movements to crawl toward light (Yamashiro and Nishihira 1995). Besides mushroom corals, righting and migration have also been observed in other free-living corals, such as in the Caribbean Manicina areolata (Linnaeus, 1758) (see Fabricius 1964; Hubman et al. 2002; Uhrin et al. 2005) and the Indo-Pacific Heteropsammia cochlea (Spengler, 1781) (see Goreau and Yonge 1968). The latter studies were based on field studies.

Righting in free-living corals may take much time and can, therefore, be difficult to observe. It has to be stressed that the present observations concern a single small, coral individual. All observed corals, regardless of their species identity, showed similar behavior (see also Hubbard 1972). Duration of the hopping intervals and the entire flip-over process is expected to vary among individual corals because mushroom corals of some species are more frequently found in upside-down position than those of others, which may depend on their weight, size, and shape, as well as depth, substrate type, and steepness of the reef slope (Hoeksema 1988; Gittenberger and Hoeksema 2013; Mondal and Raghunathan 2015). The largest and heaviest free-living corals are probably the least acrobatic.

An elongated corallum shape and a low coral weight as in the present specimen may have helped to complete this observation. Both corals that did not manage to right themselves within the available filming time, may have been too heavy (in the case of Lobactis scutaria) or not slender enough for a sideways flipover (in the case of Pleuractis granulosa). Many mushroom corals are round and may indeed become large and heavy (Hoeksema 1991). Such corals can probably become displaced during storms (Hoeksema 1988), after which they can still free themselves from sediment by the use of pulsed body inflation (Bongaerts et al. 2012).

Earlier observations (Yamashiro and Nishihira 1995; Bond 2013) and the present results show that lime-lapse photography can be of much help in observing mobility of corals and other invertebrates. In future studies, all kinds of free-living coral species can be studied in this way, such as the West Atlantic species Manicina areolota (Linnaeus, 1758) and Meandrina danae (Milne Edwards and Haime, 1848) (Johnson 1988; Meesters et al. 2012), and the Indo-Pacific Heteropsammia spp. and Heterocyathus spp. (Goreau and Yonge 1968; Hoeksema and Best 1991; Hoeksema and Matthews 2015). The latter species are famous for their symbiosis with a sipunculan worm, which slowly pulls the corals over the seafloor. Time-lapse photography may be able to show what kind of behavior is performed by such free-living corals when they are placed in various densities and on different kinds of substrates.