INTRODUCTION
In many marine ecosystems, sea stars often play an important ecological role as keystone species (e.g. Pisaster ochraceus in Paine Reference Paine1966; Acanthaster planci in Paine Reference Paine1969; Heliaster helianthus in Paine et al., Reference Paine, Castillo and Cancino1985). The understanding of the ecology of such sea star populations is thus essential especially if they are to be harvested or conserved. Ecological data to document dispersal pattern, growth rate and life span of larger invertebrates are usually collected through mark–recapture studies. For animals to be tracked effectively underwater, individuals are usually tagged with acoustic transmitters.
Acoustic telemetry has provided important insights into the biology of a wide range of marine animals, including gastropods (Delgado & Glazer, Reference Delgado and Glazer2007), cuttlefish (Aitken et al., Reference Aitken, O'Dor and Jackson2005), horseshoe crabs (Moore & Perrin, Reference Moore and Perrin2007), crustaceans (MacArthur et al., Reference MacArthur, Hyndes, Babcock and Vanderklift2008; Taylor & Ko, Reference Taylor and Ko2011), sharks (Meyer et al., Reference Meyer, Clark, Papastamatiou, Whitney and Holland2009), turtles (Okuyama et al., Reference Okuyama, Shimizu, Abe, Yoseda and Arai2010) and manatees (Deutsch et al., Reference Deutsch, Reid, Bonde, Easton, Kochman and O'Shea2003). Despite the ecological and commercial importance of sea stars, there is only one previous study (Lamare et al., Reference Lamare, Channon, Cornelisen and Clarke2009) that attempted to monitor asteroids using acoustic telemetry. This lack of studies is partly due to the notoriously low rate of success in marking or attaching tags on sea stars (see e.g. Asterias rubens (Kvalvagnaes, Reference Kvalvagnaes1972); Acanthaster spp. (Barham et al., Reference Barham, Gowdy and Wolfson1973; Glynn, Reference Glynn1982); Pisaster ochraceus (Paine, Reference Paine1976); and Anasterias rupicola (Blankley, Reference Blankley1984)). Tags are usually pierced through the body of the starfish and these are secured with wires or monofilament. Acoustic tags are relatively heavy due to the size of the batteries. As a result, tags tear through the tissues easily and become detached or lost after a short period (Lamare et al., Reference Lamare, Channon, Cornelisen and Clarke2009).
Many ecologically important species of sea stars (e.g. Pisaster ochraceus and Protoreaster nodosus) have a long life span, with a maximum of 34 years (Menge, Reference Menge1975; Bos et al., Reference Bos, Gumanao, Alipoyo and Cardona2008). Lamare et al. (Reference Lamare, Channon, Cornelisen and Clarke2009) tagged five Coscinasterias muricata individuals each with an underwater acoustic tag, but only one tag remained attached after two weeks. This is a relatively short period for any meaningful ecological data to be collected. As part of a larger study to monitor the movements of a starfish population, we developed a tagging method with at least a medium term (about three months) tag retention time with a low detachment rate. We were also able to test this method on a relatively large number of individuals. The effectiveness of the method, in terms of its impact on the behaviour and well-being of the tagged individuals, was also examined.
MATERIALS AND METHODS
Study species
The knobby sea star Protoreaster nodosus (L. 1758) is a large asteroid with minor and major radii measuring up to 74 mm and 199 mm in length, respectively (Scheibling & Metaxas, Reference Scheibling and Metaxas2008; personal observation). In Singapore, P. nodosus is widely distributed in low intertidal and subtidal habitats in the Johor and Singapore Straits including Pulau Ubin, Pulau Semakau, Sentosa and Cyrene Reef (Lane & Vandenspiegel, Reference Lane and Vandenspiegel2003; personal observation). This species has a heavily calcified body, numerous tubercles on the aboral surface, and five distinct arms that narrow distally.
Tagging
ACOUSTIC TRANSMITTERS AS TAGS
In this study, continuously emitting underwater transmitters VEMCO V13-1L (Amirax Systems Inc., Halifax, Nova Scotia, Canada) were attached to individual knobby sea stars in the field. These tags ping continuously upon activation until the non-rechargeable battery is used up or deactivated by the user. The acoustic transducer is located at one end of the tag. The tags were activated within one month after delivery. The estimated life span of each tag was 118 days, when the output power and ping period were set at ‘low' level and 1500 ms, respectively. Each tag was identified by its respective operating frequency, which ranges between 60 and 84 kHz. The length and diameter of these cylindrical tags were 36 mm and 13 mm, respectively. Each tag weighed 11 g in air and 6 g in water. This tag size was chosen as it provided the best size–performance ratio.
TAG HOLDER
A holder was designed for the attachment of an acoustic tag to a sea star. Each holder was made from a rectangular strip of neoprene with dimensions 80 mm long × 23 mm wide × 2.8 mm thick. The neoprene strip was folded into half and the middle was used to wrap the tag tightly by stretching the material. Both sides of the strip were then stitched together using a monofilament nylon fishing line of 0.3 mm diameter (Figure 1). To protect the stitches from abrasion, another running stitch was added parallel to it, at approximately 3 mm away. Both ends of the neoprene strip were stitched together, resulting in an approximately 23 mm × 12 mm ‘blank area'. A monofilament nylon fishing line of 0.5 mm diameter and 50 cm length was then sewn across the this area with a running stitch.
An acoustic transmitter (VEMCO) in a neoprene holder designed in this study. (A) Front; (B) back.
TAG ATTACHMENT
Sea stars (N = 36) were tagged in the field and returned to their original location immediately after tagging. Among these individuals, 9 were retagged on a different arm after their first tags have detached, usually after a period of one year. Thus a total of 45 tagging observations were made. The minor radii of tagged individuals averaged 52.9 ± 4.2 mm (range = 42–61 mm). To attach a tag to a sea star, two small perforations were made along the mid-ridge of one arm for the monofilament line (length = 50 cm; diameter = 0.5 mm) to be threaded through (Figure 2). A steel needle with a straight shaft (length = 43 mm; diameter = 1.3 mm) was used to make these slots. The two slots were approximately 2 cm apart from each other and were located mid-way along the length of one arm. The location was chosen to prevent injuries to internal organs close to the disc and also to minimize the probability of tag detachment as the arm is more likely to be damaged at the distal end. The arm was pierced from the aboral side, avoiding the tubercles, and the needle was made to emerge from the ambulacral groove. To minimize damage to the arm, piercing was done by gently rotating the needle. The tag holder was attached to the arm with the help of another steel teasing needle with a straight shaft (length = 32 mm; diameter = 1.3 mm). The tapered tip of this needle was modified into a hook with an inner diameter of approximately 1 mm. Both ends of the 0.5 mm diameter nylon line sewn into the tag holder were guided through the holes of the arm using this curved needle. This was achieved by inserting the needle into the hole from the aboral side, inserting one end of the nylon line into the loop of the needle's curved tip, and then pulling the line from the ambulacral groove and out to the aboral side. The ends of the line were attached to the tag holder with multiple running stitches and then tied together with a dead knot. The entire process of tag attachment on one sea star individual took about 15 minutes to complete in the field.
Method of attachment of an acoustic transmitter to the arm of a knobby sea star Protoreaster nodosus. (A) Aboral side of the animal; (B–G) lateral side of the arm; (H) oral side of the arm.
Prior to this finalized technique, several variations of it were tried out in a preliminary study. Variations include making one instead of two perforations for tag attachment. Perforations were tried at several locations along one arm of the sea star, including at the arm lateral mid-way between the dorsal and the ventral. The line was made to emerge from the opposite lateral side of the arm. Perforations were located at the distal third quarter of the arm length. Monofilament lines with diameter ranging 0.2–0.8 mm were also tried. Individuals were tagged with different permutations of the variations and most of the tags detached after less than a week.
Tracking of tagged and untagged sea stars
Knobby sea stars occurring on an intertidal reef in the East Johor Strait, Singapore were tracked. The approximate position of each sea star was obtained (± 5 m accuracy) using a portable Global Positioning System unit (Garmin eTrex). Body measurements (i.e. minor radii and major radii, to the nearest mm) were also obtained from each individual using a tailor's measuring tape.
UNDERWATER ACOUSTIC TELEMETRY
The positions of tagged sea stars were noted between December 2007 and December 2010 every two weeks during spring low tides at a tidal height of up to 0.8 m. During the initial phase of the study, tracking was carried out at various time intervals (i.e. hourly, daily, weekly, fortnightly and monthly) and tidal regimes (i.e. spring low tides, neap low tides, neap high tides and spring high tides). Tracking was carried out by either wading on the reef flat or rowing a small boat. An ultrasonic telemetry and tracking receiver (VEMCO VR100) and a directional hydrophone (VEMCO VH110) were used to detect ultrasonic signals transmitted by the tags. A total of 36 individuals were tracked in this study.
VISUAL SURVEY
As part of a separate monitoring programme, untagged sea stars were tracked between March 2002 and December 2010 once a month during spring low tides at a tidal height of up to 0.3 m. About 1800 individuals were tracked passively using visual survey while wading on the reef flat. A non-invasive method based on morphological characters (Chim et al., submitted) was used for individual recognition. Data collected from untagged sea stars were compared with those from tagged sea stars. Visual surveys also allowed us to locate sea stars that have lost their tags or those still with tags but with low or no battery power.
Duration of tag retention
The duration of tag retention was calculated as the number of days between the day of tag attachment and the day when the tag was last found attached to the sea star. If a tag was detached before the expiry of its battery life, it was reused either on the same individual or a different one. Sea stars that were not recaptured after tagging were not included in the data analysis.
Effects of tagging on sea stars
The effects of tagging on sea stars were assessed based on the extent of injury, feeding frequency and movement. An injury severity index (ISI) was developed to quantify the extent of visible damage caused by tagging (Figure 3). Injury assessment was made after the tag has detached or after its battery life has expired, whichever came first. An animal with no visible injury at the point of attachment was given an ISI score of 0. Animals that were scarred from tagging had an ISI score of 1. If an arm had a slight indentation of less than half its arm width, the animal was given an ISI score of 2. For indentations that affected half of arm width, the ISI score was 3. Shortened arms were given an ISI score of 4. If a tagged individual was not encountered again throughout the study period, it was assumed to have died and was given an ISI score of 5.
Injury severity index. (A) 0: no visible injury; (B) 1: scar; (C–D) 2: indentation at less than half of arm width; (E, F) 3: indentation at half of arm width; (G, H) 4: shortened arm. Black arrow indicates point of tag attachment.
The frequencies of tagged and untagged individuals observed feeding were tested for statistical difference. A sea star was considered to be feeding when its stomach was everted at the time of observation. The GPS coordinates of tagged and untagged sea stars were plotted to detect any difference in their dispersal patterns. The monthly displacement of the two groups of sea stars was tested for statistical difference to determine if tagging has any effect on movement. Monthly displacement is the straight-line distance calculated from an individual's GPS coordinates obtained in consecutive months.
RESULTS
Duration of tag retention
Tags were retained for an average of 60 ± 30 days (range = 14–131 days, N = 45). Eleven of these tags had retention duration of at least 12 weeks (Figure 4).
Protoreaster nodosus. Duration of acoustic tag retention on arms of individuals (N = 45).
Effects of tagging on sea stars
INJURY AND MORTALITY
Out of 36 individuals that were tagged, the majority of them (N = 20) suffered only scarring (ISI = 1) (Figure 5). There was no obvious injury (ISI = 0) in ten individuals. Three individuals had slight indentations (ISI = 2) in their arms. The arms of two individuals were badly damaged (ISI = 3) while the arm of one individual was shortened (ISI = 4). No mortality (ISI = 5) was observed in any of the tagged sea stars. In summary, only three out of 36 individuals, or 8.3% of the individuals, incurred severe injury (ISI ≥ 3) through tagging.
Protoreaster nodosus. Injury severity in tagged individuals (N = 36). See Figure 3 for explanation of index.
FEEDING FREQUENCY AND MOVEMENT
Based on data collected from 596 observations, there was no significant difference in feeding frequency between tagged and untagged sea stars (χ2 = 0.015, df = 1, P = 0.903). The percentages of tagged and untagged sea stars observed feeding were 46% (119 of 259) and 45% (152 of 337), respectively. There was also no significant difference in monthly displacement between tagged and untagged sea stars (Table 1) over three out of four years of monitoring their movements.
Two-sample t-tests on monthly displacements (mean ± SE) of tagged and untagged Protoreaster nodosus between 2007 and 2010 on an intertidal reef in Johor Strait, Singapore.
DISCUSSION
A method was developed in this study to attach acoustic transmitters to the knobby sea star Protoreaster nodosus. Using this technique, we tagged 36 sea stars with little or no detrimental effects on the health of individuals. Among the tagged sea stars, 91.3% of them had little more than a slight cut by the nylon line at the location of tag attachment. Only a few individuals were badly damaged, developing shortened or deformed arms after the tags were attached, but none of them died during the study period. Tagged sea stars were feeding as frequently as the untagged ones, suggesting that tagging did not affect feeding. Throughout the three-year study period (2007–2010), tagged and untagged sea stars appeared to be moving across the study area to the same locations and at about the same time. Furthermore, both groups of sea stars moved similar distances every year, with the exception of 2008. This suggested that it took about a year for tagging to affect the movement of an individual, and the effect subsided after a year. Tags were retained for an average of 60 days and a maximum duration of 131 days. This tag retention time is considerably longer than that in other documented studies. In Lamare et al. (Reference Lamare, Channon, Cornelisen and Clarke2009), one of three tags was found still attached to a sea star (Coscinasterias muricata) after two weeks and another two tags were not found after deployment. The extended tag retention duration achieved through this technique allows sea stars to be tagged reliably for up to three months with little or no observable effects on their movement or health.
The material used for tag attachment, and the position of the tag on the body of the sea star, appear to be important factors in determining the duration of tag retention on the animal. In this study, the tag was attached to the sea star by threading two loops using a nylon line with a diameter of 0.5 mm through the mid-ridge of the dorsal and ambulacral groove of the arm. The line selected was neither too stiff, to the extent of cutting into the tissues easily nor too flexible such that the loop will twist upon itself. In addition, the loop was neither too small such that the line is in contact with the arm, nor too large such that it will entangle easily with objects. By passing the line through the ambulacral groove ventrally, this technique further minimized contact between the line and the tissues of the sea star. As tags were normally lost due to the nylon line cutting through the arm in a downward direction, tags were retained on the animal for an extended duration by attaching them at the tallest point (i.e. mid-ridge) of the arm. The tag was attached to the arm such that it would have been resting on the substratum, thus minimizing the pulling force of the line on the arm. In Lamare et al. (Reference Lamare, Channon, Cornelisen and Clarke2009), a steel wire with a diameter of 0.7 mm was used for tag attachment. Abrasion by the thick and stiff wire could have contributed to the short duration of tag retention. The large size and highly calcified body of P. nodosus could have resulted in the relatively slow rate of the tag cutting across its arm in this study. Furthermore, the tag is relatively light, weighing 6 g in water. Although the tags used by Lamare et al. (Reference Lamare, Channon, Cornelisen and Clarke2009) were of a similar size (i.e. 5 g in water), sea stars tagged in that study (total ray diameter = 0.20–0.23 m) were smaller than those tagged in this study (total ray diameter = 0.25–0.33 m). Our technique may therefore be less useful for smaller, more delicate sea stars. The habitat of P. nodosus may also explain the extended tag retention duration. The study area is composed primarily of laterite gravel, sand and silt. As the habitat is not rocky, there was little likelihood of the tags being lodged in crevices from which they might have been difficult to extract. Although there were seaweeds and hydroids, they did not occur in such high abundance for the tags to be entangled.
In this study, the duration of tag retention and the battery life of tags used were similar, i.e. three months. In two instances, the tag holder was still attached to the sea star even after the battery life of the tag has expired. The old tag can be easily removed from the tag holder and replaced with a new tag. During the entire course of the study, none of the tags slipped off from their respective holders. By using the tag holder, it is unnecessary to retag an individual, unless the tag holder has detached, thus minimizing further manipulations of the sea star. In the event that the battery life of a tag has to be shortened intentionally, such as increasing the output power and decreasing the ping period, the tag holder can be used instead of tagging an individual repeatedly. The technique we have outlined here is likely to work equally well with data-storage tags, including depth recorders and accelerometers, which are widely deployed on both marine vertebrates and invertebrates (Gleiss et al., Reference Gleiss, Jorgensen, Liebsch, Sala, Norman, Hays, Quintana, Grundy, Campagna, Trites, Block and Wilson2011; Hays et al., Reference Hays, Bastian, Doyle, Fossette, Gleiss, Gravenor, Hobson, Humphries, Lilley, Pade and Sims2011).
ACKNOWLEDGEMENTS
We thank Justin Sih, Jeff Chouw, Eric Au Yong, Lim Swee Cheng, Lai Chien-Houng, Ken Lee, Hoe Lishan, Nicholas Yap, Ong Xueyuan and many others who have provided assistance in the field.
