Introduction

The Eurythoe complanata belongs to the class amphinomidae, of the superclass Polychaeta. Commonly known as the fireworm or bristleworm, this cryptic animal is characterized by an impressive array of calcareous bristles, containing a mild but uncomfortable neurotoxin, extending from the parapodia (Barroso et al, 2010). The class polychaeta is exceptionally diverse in both morphological characteristics and habituation of marine environments (Paiva and Barroso, 2007). This diversity has driven a correspondence in diversity in the feeding mechanisms and preferences within this class (Amaral & Nonato, 1996). Often associated with tropical, intertidal waters the cryptic amphinomidae polychaetes are generally found beneath coral heads or boulders. The amphinomidae family is broadly classed as carnivorous, preying on a range of sponges, molluscs, carrion or even other polychaetes (Pardo and Amaral, 2005) . Further introspection on a species level will investigate the feeding preference of E. complanata to develop a greater understanding of this species ecological role.

With a worldwide distribution, this animal exhibits remarkable similarity in its morphological components, despite genetic studies defining multiple subspecies (Glasby, 2007; Barroso et al, 2010). Its simple morphology, with the absence of jaws, teeth or papillae alludes to the hypothesis that the fireworm will prefer decomposing prey as it seemingly lacks the necessary jaw structure to hunt and subdue its quarry. Instead, they possess a strong muscular pharynx that may extend the lip 4 times its natural size, functioning as a suction pump. The exposed inner pharynx harbors a series of small grooves, driving the food into the gut (Pardo and Amaral, 2005). To compensate this pharyngeal simplicity the fireworm has adapted a well-developed chemosensory system it utilises to locate and acquire food. The head bears a nuchal (chemosensory) organ projecting from the posterior end of the prostomium (Pardo and Amaral, 2005). Therefore it is hypothesized that the fireworm will exhibit a preference and a swifter feeding response to a food that excretes high levels of oils and smell, such as a decaying pilchard. This study aims to address the E.complanata’s feeding preferences that may potentially provide an insight to its ecological role.

Material and Methods

Study site:

This study took place at Heron Island, off the tropical Queensland coast, Australia. Heron Island is a large coral cay surrounded by a reef crest, enclosing an expansive lagoon. The specimens were collected from beneath intertidal boulders within close proximity to the land on the Southeastern margin of the island.

Organisms:

Fifteen Eurythoe complanata specimens were collected and separated into 3 separate holding jars adjoined to a recirculating water system, nondescript of size or sex. All animals were in normal condition when collected over a two day period, ie. None exhibited recent damage, and/or regeneration. Animals ranged in size from 7cm to 15cm, but were observed to extend to double that length whilst moving. All fireworms were housed in holding jars for a minimum 48 hours and were not fed.

Feeding preference tests:

Time taken to approach food source:

Two 30 x 50cm trays were filled with water and assigned as the testing tank and acclimatization holding tank. All 15 animals were placed in the holding acclimatization tank in regulated conditions (ie. no distinct shadows, shaded cool temperature) for 2 hours to familiarize them with the identical testing environment. The three food sources were categorized as one alive source (limaria bivalve), one freshly killed local source (white bivalve) and one decaying fish source (pilchard).

The food source was placed in one corner of the tray and the individual organism was placed in the distal corner of the tank. Each animal was tested once, and were circulated through the testing tank using the same food source. In between each individual trial the tank was ‘3 rinsed’ and rotated randomly, in conjunction with random corner placement of the food source. Animal was removed once contact or feeding behavior was observed with the food source to ensure continued starving.  A maximum of 5 minutes was delegated to each trial to ensure opportunity for the animal to sense the food source, ‘non-feeding’ behaviors were recorded. 


Figure 1. Time taken to approach food source experiment

Food Preference:

A specifically designed ‘half-channeled’ testing tank was used to test food preference of E. complanata. Half of the 30cm x 25cm tank was separated, using Perspex and plastacene, into 5 watertight channels, with the remaining half remaining open to provide the fireworm withroom to move (figure 2). This tank was designed to provide an equal opportunity for the fire worm to identify and choose its preferred food source, which was placed at the end of each channel. The fire worm was place in the centre of the ‘open half’ of the tank and its preference of food was recorded. The food sources were Pilchard, Limaria Bivalve (alive), freshly killed white bivalve, porifera (broken apart) and some small sedentary polychates (killed). A five minute maximum was allocated to each individual replicate.


Figure 2. Food Preference experiment set-up.

Statistical analysis:  

Statistical analysis methods will be completed using the Excel statistical program. The time taken to identify and approach food sources will be tested for significance using the ANOVA statistical method, in conjunction with standard error bars.

Results:

There was a no significant difference in the time taken to identify and approach the three food sources (figure 3, p= 0.17,df=2). The decaying pilchard food source exhibited the swiftest mean approachtime at 170.7 ± 17.7s. This appeared to be a significantly faster than the response time exhibited with the freshly killed bivalve food source with a mean of 213.9 ± 13.8s (p=0.047). The live Limaria bivalve food source exhibited no significantly different time taken with a mean of 187.8 ± 16.4s.

The feeding preference test indicated a clear preference for the decaying pilchard food source with 7 of the 15 specimens selecting that source (figure 4).

Figure 3. Mean time taken (s) to approach and make contact with three different food sources. The error bars represent standard error for each food source trialed. The mean time taken and standard error for the freshly killed bivalve source was 213.9 and 13.83 respectively.The mean time taken and standard error for the decaying pilchard source was 170.7 and 17.7 respectively. The mean time taken and standard error for the live lima bivalve food source was 187.8 and 16.4 respectively.

Figure 4. Number of times each of the food sources was selected by organisms (n=15) using the ‘half channeled’ testing tank. The decaying pilchard source was selected as the preferred food source on 7 occasions. The freshly killed bivalve, lima bivalve, dead polychaete and porifera were selected 4, 3, 2 and 2 times respectively.

Discussion:

The hypothesis was dismissed with the timed response test yielding no significant difference in response time between the three selected food sources; decaying pilchard, live Limaria bivalve and a freshly killed, local, bivalve (figure 3, p = 0.17, df=2). The significance of the data was confounded by the small sample size (n=15), with high variance, restricted by the small window of opportunity for testing whilst on Heron Island. Additionally, exposure to unnatural environmental conditions such as handling stress, excessive light and high animal densities was regulated to the best of abilities but was identified as the likely cause of the overall insignificance of the data (figure 1). The fireworms exhibited a tendency for crypsis and maintaining close grouping behavior during the acclimatization period. This would indicate a species more comfortable in a sheltered environment, with an emphasis on nocturnal activity, in deference to the daytime experimentation (Pardo and Amaral, 2005). The 48 hour purging periodmay not have been adequate to successfully starve the fireworms.    

As hypothesized, there was a significantly swifter response to the decaying pilchard food in comparison to the freshly killed bivalve source (p=0.047, df=2). This difference may be related to its well-developed chemosensory system of the E. complanata and the strength and dispersal of the smell associated with the food source. Previous research has indicated the importance of the chemosenseory system in the ‘searching’ feeding behaviors exhibited by the fireworm (Pardo and Amaral, 2005). The nuchal organ can identify a potential food source from dispersed, suspended particles (Barosso et al, 2010). The softer and oilier flesh of the decaying pilchard had a greater dispersal potential than the freshly killed local bivalve. The sensitivity of this nuchal organ allowed swifter identification and approach of a potential food source due to the properties of the food source (Pardo and Amaral, 2005). This was supported by the food preference test with the pilchard being the selected food source, among the four other known food sources, by nearly half the specimens (figure 2 and 4). Its developed nuchal organs, simple jaw structure and decomposing food preference indicated the E. complanata is a scavenger, an important ecological role, in the intertidal coral reef.

Future research utilizing a greater sample population would be ideal to confirm the validity of the food preferences of E. complanata. This study system also provides a framework for further feeding preference tests in a variety of marine invertebrates.