Nudibranchs are one of the most fascinating and intriguing life forms under the ocean’s surface yet, due partly to their small size and comparative scarcity they are frequently overlooked. Furthermore, and as noted by Todd (1981), their frequently exotic colouration and body-form has rendered them subject to frequent incidental examination, with the result that throughout the literature there are a large number of observations and somewhat incomplete data relating to their general biology.

With this is mind, it is useful to review the ecological literature concerning nudibranchs and their taxonomy, beginning with an overview of their hierachical classification, before discussing in detail their individual feeding behaviour and apparatus.

The term nudibranch means "naked gill" and refers to the nudibranch’s dorsal external gills, which take the form of branchial plumes, club- or leaf–like processes, or arborescent processes located along the sides of the body or in a lateral groove (Picton and Morrow, 1994). Nudibranchs are members of the subclass Opisthobranchiata belonging to the class gastropoda in the Phylum Mollusca. Their taxonomic classification is as members of the class Gastropoda, subclass Opisthobranchiata, order Nudibranchia.

The order Nudibranchia is the major representative of the molluscan subclass Opisthobranchia which, in the words of Morton (1967), "ranks first among the Gastropoda in evolutionary enterprise". With over 3000 species worldwide, opisthobranchs are found in habitats ranging from the Antarctic to the Indo-pacific, and on substrates including the soft sediments (where representatives may be epibenthic or interstitial), to hard substrata, and the truly pelagic oceanic environments (Todd, 1981).

Since the time of Alder and Hancock (1845-1855), two general types of nudibranchs have been recognised. One, the suborder Holohepatica (the Dorids), have the digestive gland compact and undivided. The other, the suborder Cladohepatica (the Aeolids), have the digestive gland branched, with the branches extending into dorsal outgrowths of the mantle called cerata (See Fig 1.1). Odhner (1937), proposed the division of the nudibranchia into four suborders, since it appeared to him that the evolution of the branched digestive gland had occurred more than once, and following taxonomy by Thompson (1976), the nudibranchia are now recognised as being separated into the suborders Dendronotacea, Doridacea, Arminacea, and Aeolidae. Figure 1.1.1 illustrates the main morphological features of these four suborders.

The Dendronotacea are considered to be the most primitive of the four suborders. This group is typified by an elongate body form, retractile and sheathed rhinophores, and a variously elaborated dorsum. Both types of digestive gland organisation are found in this suborder; the holohepatic condition is considered the more primitive. Extensive secondary gills along the mantle margin are displayed by, e.g., Tritonia, and Dendronotus, while in the Lomanotidae these are conservatively phylliform.

The Doridacea are characterised principally by a circle of branchial plumes (adaptive or secondary gills) about the anus, which is mid-dorsal on the posterior half of the animal. The gills are frequently contractile and, in some cases, retractile within a protective pouch. The digestive gland is compact and does not extend into dorsal projections of the mantle. The mantle is heavily impregnated with endoskeletal calcareous spicules, with the dorsum being drawn-up into many short papillae. A ‘buccal pump’ (a muscular extension of the buccal mass) is well developed in some dorids (particularly Onchidoris spp.) to facilitate the suctorial feeding method (Crampton, 1977).

The suborder Arminacea constitutes a heterogeneous assemblage of genera (Todd, 1981) including, Armina and Phyllidia (with rows of secondary gill lamellae located beneath the mantle edge) and Antiopella (an aeolid-like genus with a mid-dorsal posterior anus).

Aeolid nudibranchs typically have an elongate body shape, with two cephalic, chemosensory tentacles protruding from their anterior end. They posses a branched digestive gland which extends into finger-like projections, called cerata. The skin of the cerata is quite transparent, so that inside each ceras can often be seen the thin brown gut contents. These organisms have no branchial plume, so oxygen exchange occurs across the skin surface. An area of high energy need (the digestive system) is therefore exposed closely to the oxygen-laden seawater, and the small size of these structures greatly increases the animal’s surface area (relative to its total volume), maximising the volume of oxygen that can be absorbed (Bertsch, 1999).

The evolution of the Aeolidacea is closely related to their association with Coelenterata as prey organisms (Thompson, 1976). The cerata, as well as housing the diverticula of the digestive gland, also contain nematocysts (derived from the prey) in cnidosacs at their tips. The nematocysts or distasteful secondary metabolites can be used by the nubibranch for its own protection, as a pore connects each cnidosac to the exterior of the organism. Aeolids have been known to feed on Cnidaria including sea anemones, corals, alcyonarians, jellyfish, and even the Porteguese man-of-war (Rudman, 1999).

A study by Slattery et al., (1998) on the Aeolid nudibranch Phyllodesmium guamensis, showed that this nudibranch, a grazer of the soft corals on Guam, sequesters a diet-derived diterpene, which acts as a feeding deterrent to reef fish. In particular, the puffer fish Canthigaster solandri is deterred from feeding on P. guamensis due to the high levels of diterpene (an unpalatable secondary metabolite) in its cerata.

It is apparent that the diversity of body form- which is only possible after the total loss of the shell in the benthic stage- exhibited by nudibranchs is primarily attributable to their respective predator-prey associations. Thus, for example, those grazing encrusting sponges and bryozoans are of bulky, flattened shape, while those associated with the delicate hydroids and erect bryozoans are more slender and elongate. The Aeolids have gone one step further, as by deploying their digestive gland into cerata, they are able to attain increased slenderness, buoyancy, and crypsis (Todd, 1981).

In one sense, the adaptive radiation of the opisthobranchs can be seen as an evolutionary spreading out into various feeding niches (Bertcsh, 1999). Although there are exceptions, most of the divisions of opisthobranchs are rooted in feeding specificity. The Anaspidea and Sacloglossa are herbivores, but each feed differently. The sea hares rasp algal fronds, while the Sacoglossa delicately slice open each algal cell to suck out its contents. The Nudibranchs are all carnivores. They feed on other invertebrates; sponges, soft corals, anemones, sea pens, bryozoans, ascidians, hydroids, the eggs of other nudibranchs, and some species are parasitic.

Many nudibranchs are specialised predators (monophagous), thus, the species range is limited, and the body form is adapted for predation of a specific organism. Other nudibranchs, however, are generalised browsers. Aeolids, including the genera Aeolidiopsis, make use of microscopic green algae (zooanthellae) in their own bodies to supply them with food. These algae supply the excess sugars produced during photosynthesis to the nudibranch, and in return they are provided with a safe niche in which to grow and reproduce (Picton, 1997).

For convenience the nudibranchs can be divided in four feeding classes, sponge-grazers, bryozoan-grazers, hydroid-grazers, and the ‘miscellaneous’ grouping (Todd, 1981). The ‘miscellaneous’ grouping includes representatives of all four suborders preying upon coelenterates, cirripedes, tunicates, other nudibranchs and their eggs, and teleost fish eggs. The bulk of the literature concerning dietary data for the British nudibranch species was comprehensively reviewed by Thompson (1964).

The sponge-grazing nudibranchs contains all of the largest British Doridacea, as well as representatives from around the world. In common with other groups, the sponge feeders range from the absolute specialist species-e.g. Jorunnu tormentosa on Halichondria panicae- through to the generalist predators- e.g. Archidoris pseudoargus. The nudibranchs that feed on sponges tend to have bulky, yet flattened bodies, and many of the dorids are cryptically camouflaged on the host species.

The Bryozoan-grazing nudibranch group is dominated by the Doridacea, and totally lacking in the Dendronotacea and Aeolidaecea. Most bryozoan-nudibranch associations concern the sheet-like encrusting cheilostome species, although the gelatinous ctenostome species- which are extensively grazed by the common littoral dorids- and some erect sublittoral cheilostomes, are exploited.

Polycera quadrilineata is unusal amongst bryozoan-grazers in consuming the skeletal elements of cheilostomes in adddition to the soft tissues (Todd, 1981). Interestingly, P. quadilineata only grazes on the peripheral (uncalcified) zooecia of Membranipora colonies. This behaviour has mutualistic consequences in terms of the persistence of the predator-prey association, in that the nudibranch gains nutritional advantage in not having to pass large quantities of inorganics through the digestive system, and only the actively growing regions of the bryozoan colony are cropped.

The largest trophic category is that of the hydroid-grazing nudibranchs, and it is dominated by the Aeolidacea. Calyptoblastic and gymnoblastic hydroids are exploited, and as in previous groups, the predators cover the whole range from broadly generalist (e.g. Dendronotus frondosus) to absolute specialist (e.g. Doto pinnatifida) species.

Other nudibranchs, classed in the ‘miscellaneous’ trophic group by Todd, include Tritonia hombergi, the largest British nudibranch which feeds on Alconim digitatum, and Onchidoris bilamellata an exclusive predator of acorn barnacles. Thompson and Brown (1976) suggest that the juvenile O. bilamellata graze bryozoans including Cryptosula and Umbonula.

Many nudibranchs exhibit resource partitioning between species, where different species graze different areas of the same colony. This however, is difficult to establish in the field, and as Rudmann (1979) reports, many species may appear to exhibit resource partitioning, but actually be eating the same tissue and just utilising it in different ways. This phenomenon is particularly common in the coral-feeding nudibranchs, notably species of Aeolid nudibranch.

Having discussed the feeding groups which make up the nudibranchia, it is now fitting to discuss the apparatus that allows this diversity of food preference, the radula. The radula is the characteristic buccal structure common to all gastropods and has undergone considerable diversification in size, form, and function within the nudibranchia (Behrens, 1993). Located in the mouth region within the nudibranch (See Fig 1.4), the radula is a tongue-like ‘ribbon’ on which are positioned rows of chitonous teeth. There is a great variability in the number of rows of teeth, the number of teeth per row, and the shape of the teeth themselves. Figure 1.4.1 shows the radulae of several coral feeding species of nudibranch.

The structure of the nudibranch radula is strongly associated with the substrate on which the organism feeds. Variations in food type such as hard corals and sponges, compared to the softer filamentous algae or hydroid species, have led to the evolution of variable radula morphologies in the species concerned. For example, sponge-feeding dorids and notaspideans have many rows of scythe-like teeth which rasp and scrape over the surface of the sponge like a garden rake. The teeth of aeolids, however, consist of few rows, with only one or three teeth in a row. These nudibranchs, however, have very well developed jaws which crop, or hold the cnidarian prey as the nudibranch rasps with its hooked teeth. In general, the more filamentous and articulating the nudibranch prey, the narrower the radula, but somewhat conversely, the nudibranchs with the thinnest radulae feed on the hardest organisms (Vietti and Balduzzi., 1991). However, many feeding adaptations exist, and one such extreme radular adaptation is illustrated by the sacoglossa. Here, the radular ribbon is reduced to a single row of teeth which slide into position to pierce the eggs of other nudibranchs or siphonalean green algae cells, before the nudibranch delicately sucks out the cytoplasm (Jenson, 1994).

These radular characteristics, specifically adapted to the nudibranchs particular prey preference are distinct for individual species, and therefore, they produce a ‘fingerprint’ upon which decisions concerning the classification of species can be based.

The use of radular structure as a classificatory tool, however, may not always be definite. Within a particular genus, and sometimes even at the family level, a great deal of overlap in the number of teeth and in variation in tooth morphology may exist. In a study by Bertsch (1976) regression analysis of radular counts from Discordis evelinae indicated that intraspecific variation was ontogenic. Hence, Bertsch suggests that "although radula morphology has an important use in nudibranch taxonomy, it must be treated as a biological entity subject to variation, and not as the product of a topological mould."

The rows of teeth that make up the radula are of two types- lateral and central (McDonald, 1983). Within the Doridacea the lateral teeth are always well developed, but the central teeth (if present) show less development. The breadth of the radula is thus dependent on the number and size of the laterals. Amongst the Dendronotidae, Lomanotus has a broad radula, which lacks central teeth. Dendronotus and Tritonia, on the other hand, have both well-developed centrals and large number of laterals and hence a large radula.

The sponge-feeding dorids lack a central tooth within the radula, and the radula as a whole is very broad and simple due to the replication of large numbers of plain hooks. Most of the remaining dorids posses a small simple radula-often lacking the central tooth-with the first lateral well developed and the remaining laterals rudimentary. The Arminacea characteristically have a broad radula, with the central tooth much the same in form as the laterals.

All the aeolids have a large pair of chitinous jaws-usually with a denticulated cutting edge. The radula invariably consists of a single row of central teeth, although simple or serrated laterals are sometimes present. Such radulae are characteristic of the hydroid-grazers, while the anemone-predators have radulae with very broad central teeth bearing many denticulations.

Bryozoan-grazing species can be divided into two subgroups: those feeding suctorially (including Onchidoris muricata) and those feeding in rasping manner (e.g. Polycera quadrilineata) (Todd, 1981). Studies on the feeding behaviour of O. muricata indicate that the buccal pump is the major structure used in feeding (Crampton, 1977). The oral veil is positioned close to the bryozoan colony surface, and the individual zooids appear to be removed without the direct use of the radula. Feeding in P. quadilineata, differs from the system described above as P. quadrilineata possess a pair of large jaws, and has a more complex radula used in the consumption of both the soft and skeletal parts of the bryozoan colony.

A complex series of muscles control the movement of the radula, rotating in and protruding the teeth into a position orientated where they can scrape, pierce, cut or tear the food source. Growth of the radular ribbon is in a forward direction and may occur at a rate of five to six rows per day. Growth begins at the outer edge of the newest half row, and more teeth are added in the middle of the radula as the developing tow moves forward. In this way a maximum number of teeth will then be present just before the area of the radula during feeding. The exact rate of growth depends not only on the particular species but numerous environmental factors (Behrens, 1993). Depending on the food source, and the hardness of the food surface teeth wear out and are broken off, therefore, as the rows progress through the anterior part of the radula, the number of teeth per half row significantly decrease (Bertsch, 1996). A comprehensive review of radula morphology may be found in MacFarland (1966) and McDonald (1983).

The radula is a complex morphology because it contains different kinds of information at many different scales. It is composed of many repeated rows of teeth attached to a basal membrane; the teeth that make up the rows are variable in shape and size and each tooth often has cusps that are also of distinct shapes and sizes. Traditionally, teeth have specific names based in their position in the row, and these names reflect the importance of the radula in nudibranch systematics. The inner laterals (IL) are the most medial and anterior teeth, and the outer laterals (OL) are placed more laterally and posteriorly. Similarly, many nudibranchs have a central rachidian tooth, this is found in the chromodorid nudibranch Chromodoris willani for example.

Two different types of apocrine cells, odontoblasts and membranoblasts form the teeth and membranes. Both odontoblasts and membranoblasts are located in the posterior of the radula sac, a blind pouch formed as an outpocketing of the stomodaeum (Raven, 1958). Both types are active throughout the life of the organism and continually produce new rows and membranes to supplement those lost during feeding. Odontoblast cells secrete cyclically and are organised into discrete adjacent groups isolated from one another by separate cells; each group secretes a single radula tooth. In contrast, membranoblast cells are not organised into discrete groups and secrete the radular ribbon continuously. Two other cell fields, the subradular and supraradular epithelia, add minerals onto the cusps and bases of the teeth, thus strengthening them.

Newly formed teeth have a different morphology than fully formed teeth. When the radula (both stained and unstained) is viewed under a light microscope, visual and structural differences between the teeth are immediately apparent. Young teeth are weak, thinner, and more transparent than the strong, heavier, darker older teeth. These growth differences are illustrated clearly by scanning microscopy. The erect hook of the tooth forms first, as a long, sharp, narrow point, projecting closely along the plane of the radular surface. The basal portion grows later, becoming angled to the hook as the diameter of the hook more than triples. These thickened, erect teeth can thus effectively rasp and gouge bits from the prey during the radular feeding stroke (Bertsch, 1996).

The radula of a nudibranch is readily visible with the unaided eye in large specimens, but requires dissection in the smallest specimens. The technique of removing, and preparing the nudibranch radula is described by Thompsom (1976). Thompson recommends the removal of the entire buccal bulb. The buccal bulb contains both the jaws (if present) and the radular ribbon. It may be obvious and protruding, or tightly contracted into the body cavity. The tissue mass can then be soaked in a caustic solution to soften and loosen the tissue (19% solutions of NaOH or KOH are most frequently used), and the radula teased from the tissue.

Rudman (1982) describes the methods involved in SEM of nudibranch radulae. The radula can then described using the radular formulae, for example 40.0.40*61 (+3), refers to a radula with 40 lateral teeth on each side of the midline in each row, there being 61 rows of teeth and three developing jaws (Rudman, 1982).

The organisms to be investigated in his study include four species endemic to the tropical west pacific, and five British species. . The Sulawesian opisthobranchs include Chromodoris annae, Chromodoris willani, Phyllidiella pustulosa and Phyllidiella nigra (see Fig 1.6). The British nudibranch species to be studied are Onchidoris bilamellata, Dendronotus frondosus, Facelina auriculata, Aeolidae papillosa and Archidoris pseudoargus (see Fig 1.6.1).

Chromodorid nudibranchs, suborder Doridacea, have been found throughout the warm temperate and tropical seas of the world (Rudman, 1973). They are usually brilliantly coloured and patterned, and the principal genus of the family, Chromodoris, is the largest in the opisthobranchia. Chromodoris annae is distributed in the tropical west pacific and Indian oceans. It is a member of the Chromodoris quadricolour colour group of species, characterised by black longitudinal lines running down the dorsum on a blue mantle ground colour, with orange borders. C. annae is distinguishabe from the other members of this group due to dark specks in the blue background areas (Rudman,1998).

Chromodoris willani has a pale blue mantle ground colour with a thin white band at the edge. The border beween the white and the blue is diffuse. A thin black band runs around the mantle, outside the rhinophores and gills, some distance from the edge. There is a similar black line down the centre of the body infront of the rhinphores to the gill pocket (Ruman, 1982c). Chromodoris willani differs from other species in its colour group by the shape of its mantle, and the colour of the gills and rhinophores. The major anatomical difference, in terms of the radular structure, is the massive central tooth in C. willani. All the chomodorids are sponge-feeders (Alder and Hancock, 1855).

Nudibranchs of the family Phyllidiidae occur throughout the Indo-Pacific region (Brunckhorst, 1993). However, phyllidiid fauna of the Indo-West Pacific has not been examined comprehensively since Bergh’s (1869) monograph. Phylidiella pustulosa, suborder Doridacea, is an elongate ovoid species with pink tubercles on a black notomand the mantle margin is edged with pale pink. Phylidiella nigra is broad and oval in shape, and convex dorsoventrally. Dorsal colouration consists of a black background with red tubercles. The tubercles are evenly distributed over the notom, whereas they are clustered in P.pustulosa. There is no continuous pale edge to the mantle margin (Brunckhorst, 1993).

The Phyllidiid nudibranchs lack a radula and are instead suctorial feeders. In situ observations of phyllidiid species feeding seem to be rare. But, sponge feeding is reported for the first time in the species P. pustulosa, feeding on the sponge, Halichondria by Brunkhorst (1993). This species has a relatively elongate pharynx and retractor muscles which suggest the foregut is everted for feeding. Field observations of Phyllidia, Phyllidiella and Phyllidiopsis confirm that eversion of the foregut does take place. The leaf-like oral glands of P. pustulosa possibly secrete other unidentified, digestive fluids in addition to weakly acidic, sulphated mucosubstances. Histological sections of the pharyngeal bulb of Phyllidiella species by Brunkhorst (1996) indicated the presence of both strongly acidic and weakly acidic mucosubstances. It is most likely that the function of these acidic secretions is for external digestion, probably including the breakdown of calcareous sponge spicules.

The British species to be studied include Onchidoris bilamellata. This is one of the most common British dorids, and also one of the largest. It is pale in ground colour, but with blotchy brown markings on the dorsum. The mantle bears abundant spiculose club-like tubercles, of various sizes, lacking in the brown pigment. There may be as many as 29 simple pinnate gills with, characteristically, a few tubercles within the branchial circlet. Each rhinophore may bear up to 16 lamellae, and the head is dilated to form an oral veil, without tentacular processes. O. bilameillata is known to feed on a variety of acorn barnacles notably Balanus balanoides (Thompson, 1988).

The species Dendronotus frondosus (Acananius, 1774) is the second British species to be studied. Dendronotus is very variable in colour and has been suggested to be a complex of several species (Picton and Morrow, 1994). Individuals may be white or mottled with yellow, red or brown pigment. There may be up to nine pairs of gills along the pallial rim. The gills, oral veil and rhinophore sheaths are extended to form branched processes, and adults may grow up to 100mm in length. D. frondosus feeds on a variety of hydroids. Juveniles usually feed on Obelia, Halecium and Sertularia whilst the adults feed on Tubularia. Dendronotus can be found all around the British Isles, the arborescent gills, oral processes and the rhinophore sheaths distinguishing it form any other nudibranch in the U.K.

Facelina auriculata (Muller, 1776) may reach up to 38mm in length when fully grown. The ground colour is translucent white and there is a rose coloured hue around the mouth. The animal’s red oesophagous is visible, slightly behind the rhinophores (Picton and Morrow, 1994). The oral tentacles are very long and the rhinophores are annulate. F. auriculata is recorded all around the British Isles and from Norway to the Mediterranean. It is known to feed on hydroid species, but has mainly been found on Obelia geniculata on kelp fronds and on Tubularia.

Aeolidia papillosa (Linnaeus, 1761) is the largest British aeolidacean, and one of the world’s largest and most widely distributed aeolid species. Aeolids reach a length of 120mm and are found around the world. The broad, depressed body bears up to 25 obliquely transverse rows of elongated, somewhat flattened cerata, which leave bare a substantial median zone behind the rhinophore bases. This bare area usually exhibits white pigment in the shape of a diamond or a crescent, often linked to similar pigment streaking the upper surfaces of the oral tentacles.

A. papillosa is one of the most common and successful shallow-water nudibranchs on the British coastline, detecting, overcoming and consuming sea-anemones when fully grown. Species of anemome which seem to be important in the diet include Actinia and Metridium.

Archidoris pseudoargus (Rapp, 1827) is known in many places around the U.K. as the "sea-lemon". The mantle bears many blunt tubercles. The mottled colouration of this nudibranch probably aids camouflage. The colours include yellow, brown, pink green and white. There is also a bright red variety known as A. pseudoargus var. flammea. Some of the large individuals may grow up to 120mm in length. A. pseudoargus is usually found on the low shore, and feeds on "bread-crumb" sponge Halichondria bowerbanki, and Suberites ficus in the sublittoral.

A recent paper by Reid and Mak (1999) on Littorinids has thrown into question the use of radular morphology as an identification tool. In this study an examination of radulae from all but one of the 36 species of littorinid, genus Littoraria, found extraordinary intraspecific variation in those occuring on a range of substrates. This finding has led to the suggestion that radular morphology may show phenotypic plasticity, and this plasticity is induced by substrate or diet. Furthermore, they suggest that

"Ecotypic variations in the radula may be widespread in the littorinids, and the radular characters should therefore be used with caution in studies of taxonomy, phylogeny and adaptation"

Another study by Padilla (1998) on the radula structure in Lacuna (Gastropoda: Littorinidae) also demonstrated plasticity of radular morphology. Here, two species of snail of the genus Lacuna produced differently shaped radular teeth when fed different foods, displaying intraspecific variability as extreme as would usually be considered to define different species. Furthermore, the new tooth morphology is produced on the non-feeding end of the constantly regenerating radula, and is thus different from use-induced feeding morphologies seen in arthropods or vertebrates. Padilla concludes that tooth shape is in Lacuna is phenotypically plastic, inducible with different food conditions, and reversible within an individual’s lifetime, allowing rapid response to new environmental conditions.

Additionally, Bertsch (1976) in a series of papers on the chromodorid nudibranchs of the west coast of America, suggests that the opisthobranch radula is subject to variation within a species, and numbers of teeth could simply reflect the ontogenic stage, and not a second species (Marcus, 1967). He goes on to note that:

"The opisthobranch radula has an important use in taxonomy, but it must be treated as a biological entity subject to variation, and not as the product of a topological mould" (Bertsch, 1976)

The aim of this study therefore, is to describe the feeding behaviour of the opisthobranch species studied. To discuss and relate the functional morphology of the nudibranch radula to the substrate on which the species feeds, and to discuss whether variation, if it exists, is species or substrate dependent. If there is variation, are these differences inter- or intraspecific or ontogenic, and is there variation in the radula morphology of species from distinct locations (i.e. Britain and Indonesia) which feed on the same prey organisms? Finally, and through comparison with radula-less nudibranchs, the use of the radula as an identification tool will be discussed. Are there alternative methods for identification of species using other buccal structures? How reliable are these methods, and which morphological characteristics, if any, should be used for identification of new species of nudibranch in the future?

The hypothesis H₁:

There is significant variation in the radula morphology of the species studied. Variation in radula morphology is a function of the prey preference, and there are variations of tooth morphology within an individual radula.

The distributional data will prove that the Sulawesian nudibranchs exhibit species- specific prey preferences.

The null hypothesis H_o:

There is no significant variation in the radula morphology of the nudibranch species studied. Variation in radula morphology between species is not affected by prey preference, and there are no variations in tooth morphology within an individual radula. The distributional data will show that the Sulawesian nudibranchs are generalist predators and show no species–specific prey preferences.

To investigate the functional morphology and feeding behaviour of nudibranchs requires particular methodology and consideration. The methodology not only involves techniques for data collection in the field, but also encompasses laboratory based procedures such as dissection, dehydration, sectioning and scanning microscopy. In all cases selection of appropriate methodology was based on time frame available for work, facilities available in the field, techniques appropriate to the material and procedures which would yield the maximum possible sample size. The choice of methodology was also based on techniques described by authors such as Rudman (1973) and Picton (1994).

In order to maintain clarity the methodology is laid out in such a way as to divide up the two main areas of work, firstly, the field based data collection on behaviour and distribution, and secondly, the laboratory methodology for the radula and buccal apparatus analysis.

2.2.1: Indonesian species:

The four Sulawesian nudibranchs, C.willani, C. annae, P. pustulosa and P. nigra were collected from Hoga, one of the Tukang Besi Islands in S.E. Sulawesi, Indonesia (see Figures 2.2 and 2.2.1). The specimens were brought up to the surface in tubes containing seawater and photographed in a narrow tank using flash light photography aimed at 45^o to the water surface. The specimens were then identified and recorded, before being drawn, and measurements taken as to length, colour and patterning. Additional information such as location found, depth, inclination etc were also recorded.

Next the specimens were preserved, it is important to label the dishes or tubes in which preservation is being carried out as the nudibranchs lose their colour and form when they are fixed, making identification difficult at this stage.

The live nudibranchs were placed in a 1:1 solution of seawater and 7% MgCl₂ for 2-8 hours, until the animal no longer responded to poking. This solution narcotises the nudibranch so that it does not contract when fixed. The specimens were then left to stand in 10% Formalin for 24 hours and shaken occasionally. The nudibranch specimens were then washed in water and stored in 70% Alcohol with a paper label for easy identification.

In order to transport the nudibranch specimens to the U.K. the specimens were wrapped in a piece of material soaked in 70% alcohol and placed in an air tight tube with their identification label.

2.2.2: British species:

The British species were collected from the west coast of Scotland. The specimens were narcotised and preserved in the same way as above.

Collection of the behaviour and distribution data was carried out during three weeks of dives on Hoga. The species C. willani, C. annae, P. pustulosa and P. nigra were located during the dives and measurements recorded as to length of the specimen in mm, the depth (m) the specimen was found, time of sighting and information as to location. The location information included dive site, the surface the nudibranch was found on (i.e. sponge, coral, tunicate etc), inclination of the nudibranch (i.e. horizontal or vertical on the substrate), the exposure of the nudibranch sighted (i.e. Exposed, Semi-exposed of sheltered when found) and the activity of the specimen when sighted (i.e. whether feeding, moving, stationary etc). This information, recorded underwater on a slate for each organism found, was transferred to a database on the island, and specimens of each species collected for preservation.

Two of the species brought back from Indonesia Phyllidiella pustulosa and Phyllidiella nigra did not possess a radula as these species are suctoral feeders. Thus, in order to study their buccal apparatus wax sections of their mouth parts were produced. In this way, it could be noted whether there were any differences between the two species of the same genus, and hence decisions drawn as to whether these differences could be used in classification of Phyllidiid nudibranchs.

2.4.1: Dissection

First the nudibranchs had to be dissected as they were too large to section the entire organism (being approx. 33mm in length), and only the head region was going to be studied. The head section of the nudibranch, from behind the rhinophores, was cut off using a razor blade so that the head could be sectioned longitudinally.

2.4.2: Dehydration and Embedding:

The dissected head was then dehydrated. This involved the tissue being taken through the alcohol dehydration series:

1. 30% alcohol 1 hour (with 4 changes of alcohol)
2. 50% alcohol 1 hour (with 4 changes of alcohol)
3. 70% alcohol 1 hour (with 4 changes of alcohol)
4. 95% alcohol 1 hour (with 4 changes of alcohol)
5. 100% alcohol 24 hours (Alcohol changed as often as possible)

The alcohol concentrations were made up in glass vials and placed in a rotator to speed up the dehydration process, and to ensure that the alcohol and distilled water were properly mixed.

The tissue was then further dehydrated in a 50:50 solution of Absolute Alcohol and Histoclear. This solution was changed every half an hour over 4 hours. The specimens were then placed in 100% Histoclear for 24 hours, the histoclear being changed regularly.

Embedding:

The dehydrated specimens were embedded in molten wax, the head orientated so it could be sectioned longitudinally on a student microtome. The molten wax embedded with the specimens was then cooled in a water bath until the wax had set. The wax block was then cut, using a razor blade, to remove the excess wax, and the wax melted onto a wooden block. The wax was adhered to the block by heating up a metal spatula in a bunsen flame and the bottom of the wax melted so that it would stick to the wood block when cooled.

2.4.3: Sectioning, Staining and Mounting

The wax blocks were sectioned longitudinally on a student microtome. Serial sections of 10m thickness were taken and placed shiny side down on a slide with a drop of water. Approximately 3 rows, with 8 sections per row, were placed per slide. In total seventy-four slides were produced from the P. nigra block, and forty slides form the P. pustulosa block. The slides were labelled and placed on a hot plate to dry before staining.

Th slides were stained with Erlich’s Haematoxylin and eosin, the procedure for which is as follows:

1. Place slide in histoclear for 6 minutes to remove wax from the section
2. Absolute Alcohol for 2 minutes
3. 95% alcohol for 2 minutes
4. 70% alcohol 2 minutes
5. Distilled water 2 minutes
6. Ehrlich’s haematoxylin 5 minutes
7. Rinse in distilled water
8. Alkaline alcohol 15 seconds (to blue the nuclei)
9. Rinse in distilled water
10. 70% alcohol 1 minute
11. 95% alcohol 1 minute
12. Eosin 30 seconds
13. Wash quickly in 95% alcohol
14. Absolute alcohol 25 seconds
15. Histoclear 4 minutes
16. Place a drop of D.P.X. mountant on stained section and gently lower a cover slip onto the slide.

2.4.4: Photography and Procesing:

The slides were photographed using a digital camera linked to a light microscope. The digital pictures were then converted to Adobe photoshop files so that work could be carried out in processing the pictures using Asobe Photoshop Version 5.0.

In order to create a scale bar for the images a picture of the stage micrometer was taken as the photographs of the sections were taken. In this way the length of the scale bar could be calculated and added to the image in Photoshop.

2.5.1: Dissection and Dehydration:

The species investigated using S.E.M were Onchidoris bilamellata, Dendronotus frondosus, Aeolidia papillosa, Facelina auriculata, Archidoris pseudoargus, Chromodoris willani and Chromodoris annae.

Before the Scanning Electron Microscopy could be carried out the specimens needed to be dissected in order to find the buccal mass, and thus the radula within the buccal mass. Dissections of the species Onchidoris bilamellata were carried out first, and the procedure was followed right through to the S.E.M. in order to gain experience using the technique and to see if there were any changes that needed to be made. However, in order not to have to repeat the procedure for each species it will only be described once, as the same techniques were used throughout.

The specimens were dissected using a dissection microscope. The instruments to complete the dissection were fine tweezers, and dissecting scissors. First, the buccal buls was identified within the organism. In order to achieve this a cut was made along the dorsal surface of the organism from the gill circlet to the tip of the head. This allows the body wall to be pulled back on either side using the fine tweezers and exposes the mantle cavity. The mantle cavity exposed the buccal bulb can be seen in the head of the organism inside the mouth. The buccal bulb contains both the jaws, if present, and the radula. The dissected buccal bulbs were then preserved in a 70% alcohol solution before further dissection to the radula level (see Fig 2.5)

Dissection to the radula level was achieved through gently teasing apart the buccal bulb tissue and muscle (using fine tweezers) to expose the radula. The radula is present within a radula sac, it may appear in ribbon form as in the species Onchidoris bilamellata, Facelina auriculata, Dendronotu frondosus, Archidoris pseudoargus and Aeolidia papillosa, or in a plate-like form, as in the species Chromodoris annae and C. willani. The radula sac tissue and supporting muscle needs to be removed in order to clearly see the radula using the S.E.M. The procedure for removing this tissue is as follows:

1. Dissect out radula and radula sac from buccal bulb.
2. Place in solution of 0.1% bleach (made up with distilled water) and place in rotator, change solution every 15 minutes.
3. Monitor digestion over half an hour, if digestion is too slow (i.e. there is no softening and removal of tissue) increase to 0.5% bleach solution, changing solution regularly.
4. If radular sac tissue is still not digested increase to 0.75% bleach solution but check regularly. Do not leave the radulae in any bleach solutions over night. Increase concentration of bleach solution or leave for longer in current lower concentration solution until radula tissue digested.
5. Once radula sac tissue digested, wash radula in distilled water. Wash the radula by pipetting out the bleach solution and gradually increasing the volume of water in the vial. Keep washing radula until no bleach solution remains.
6. Radulae may be stored in distilled water until dehydration procedure.

Of the 4 specimens of O. bilamellata specimens 1 and 2 underwent tissue digestion, specimens 3 and 4 had no digestion. In this way it was established that digestion of the radular sac tissue was needed for adequate viewing of the radula structure in the species.

It was decided that digestion of the tissue was either not needed of inappropriate (due to possible damage occuring to the radula) in the species F. auriculata, D. frondosus, A. papillosa and A. pseudoargus.

The dehydration of the radulae is similar to that undertaken for the dehydration before wax sectioning of the Phyllidiids, but the time needed in each concentration is reduced due to the small size of the radulae. The dehydration series is, therefore, as follows:

1. 30% alcohol 50 mins (with 4 changes of alcohol)
2. 50% alcohol 50 mins (with 4 changes of alcohol)
3. 70% alcohol 50 mins (with 4 changes of alcohol)
4. 95% alcohol 1 hour (with 5 changes of alcohol)
5. 100% alcohol 4 hours (Alcohol changed as often as possible)
6. Dry Acetone 24 hours (Dry Acetone changed as often as possible)

The radulae are then left in Dry Acetone until Critical Point Drying

2.5.2: Critical Point Drying and Mounting:

The procedure for Critical Point Drying is as follows:

1. Wash, fix and dehydrate radulae
2. Fill the aluminium boat with Acetone, put the specimens into gauze baskets and cover, taking care to do allow any air drying. (The radulae are wrapped individually in pieces of lens tissue soaked in acetone, this prevents them floating out of the baskets during drying due to their small size)
3. Close all the valves to start. If the temperature of the chamber is above 20^oC run cold water through to cool it before starting.
4. Open the valve on the CO₂ cylinder, load specimen boat, close door (do not over tighten). Very quickly open the inlet valve and the quantity of CO₂ and acetone mixture will rise inside the cylinder. To release any back pressure open the inlet valve.
5. FLUSHING: Leave the inlet valve fully open (the vent valve slightly open to maintain the liquid level) and open the drain valve to remove the substitution liquid. This flushing action should be kept up for 1-5 mins depending on the size of the specimen.
6. After flushing the bulk of the substitution liquid, fill the chamber to boat level, close all valves and leave for one hour to allow impregnation.
7. Close the inlet valve and allow the level of the liquid to fall to the top level of the boat
8. HEATING TO CRITICAL POINT: Close all the valves. Using a combination of hot and cold taps raise the temperature(1^o/20secs) over about 5-10 mins so that the final pressure of ~12000psi is reached. At 11000psi and 31^oC the critical point is reached and the liquid level eventually disappears. The CO₂ is now gasified without any change in volume. Continue heating to above 12000psi and a temperature of 36-38^oC, then turn off water supply.
9. Carefully vent the CO₂ gas by slowly opening the vent valve
10. Remove the radulae from the baskets (under the dissection microscope) and mount the radulae on the S.E.M. stubs using tweezers.

The radulae, after being mounted on stubs, were then photographed using Scanning electron microscopy.

2.5.3: Scanning Electron Microscopy:

First the specimens were coated with Gold Palladium, then they are put into the specimen chamber of the S.E.M, the chamber is pumped, and the specimens moved into a vacuum. In this vacuum the S.E.M. creates a beam of electrons which are scanned across the surface of the radulae. The radulae, being coated in gold palladium, emit secondary electrons that are detected by a scintillation material which produces flashes of light from them. It is by correlating the scanning position and the consequent light signal that an image can be produced.

The S.E.M. images were saved to Zip disc and the magnification of the images recorded. The scale bar for each image was calculated using the following formulae:

M = K.L_bim.M_oi

Where:

M = Total Linear Magnification

L_bim = Width of printout

M_oi = "Object Image" (magnification on the screen)

K = The Constant 8.33´ 1 0 - 3 1/mm

Then by using the formula:

Scale (m) = Magnification ´ Chosen length of Scale Bar (m)

1000

The Scale bar was then added using Adobe Photoshop Version 5.0

The behavioural and distribution data was analysed using standard statistical tests, such as ANOVA (one way) and chi-squared tests. The analysis of the S.E.M. images was carried out by creating a 200m² grid across the images to be analysed, and looking at the teeth within three 400m² areas, one at the posterior end of the radula, one in the mid-region of the radula and one at the anterior end of the radula. Within each of the 400m² areas measurements were taken including:

• The number of teeth,
• The length of teeth,
• The angle of tooth curvature,
• The length of distance between the cusp of tooth and first denticle,
• The distance between consecutive rows.

This data was then analysed using a 2 way ANOVA without replications.

In addition to quantitative information the S.E.M images were also analysed visually. The information noted included:

• Level of wear of the teeth,
• The shape of teeth across the radula,
• The overall shape of the radula,
• The level of fusion between the teeth making up a row.

The pictures of the sections of the buccal apparatus of P. pustulosa and P. nigra are analysed visually, with comparisons and differences between the two species discussed. Statistical analysis of the sections is not possible due to the small number of samples.

Alder, J. and Hancock, A. (1845-1855) A monograph of the British Nudibranchiate Mollusca. The Ray Soc., London.

Becerro, M. A., Paul, V. J., Starmer, J. (1998) Intracolonial variation in chemical defenses of the sponge Cacospongia sp. and its consequences on the generalist fish predators and the specialist nudibranch predator Glossodoris pallida. Mar. Ecol. Progress series.168, 187-196.

Behrens, D. N. (1991) Pacific Coast Nudibranchs. A Guide to the Opisthobranchs, Alaska to Baja California. Sea Challengers, Monterey, 2^nd Ed.

Behrens, D. N. (1991) Pacific Coast Nudibranchs. Suppl 1.

Bertsch, H (1977) Intraspecific and Ontogenic Radular variation in Opisthobranch Systematics. Sys. Zoo. 25, 117-122.

Brunckhorst, D. J (1990) Description of a new species of Phyllidiopsis, Bergh (Nudibranchia: Doridoidse: Phyllidiidae) from the Tropical Western Pacific, with comments on Atlantic species. J. Moll. Stud.56.577-584.

Brunckhorst, D. J (1993) The Systematics and Phylogeny of the Phyllidiid Nudibranchs (Doridoidea).Records of the Australian Museum.

Evans, C. G., Villim, F. S., Harish, O., Kupfermann, I., Weiss, K. R. (1999) Modulation of the Radular Opener Muscles in Aplysia. Jou. Neurophys. 82,1339-1351

Gosliner, T. M., and Draheim, R. (1996) Indo-Pacific gastropod biogeography: How do we know what we don’t know?Amer.Mal.Bull.12, 37-43.

Gosliner, T. M., and Johnson, R. (1999) Phylogeny of Hypselodoris (Nudibranchia: Chromodoridae) with a review of the monophyletic clade of Indo-Pacific species, including descriptions of twelve new species. Zoo. Jou. Linn. Soc.125, 1-114.

Haaften, D. F and Verway, S. (1978) Chemosensory prey perception in Aeolidia papillosa (Gastropoda: Nudibranchia).Veliger.13, 124-134.

Marcus, E. R (1967) Opisthobrnach molluscs from California.Veliger. 3(suppl.), 1-85.

Macfarland, F. M., (1966) Studies of Opisthobranchiate mollusks of Pacific Coast of North America.Calif. Acad. of Scien.6, 1-54b

McDonald, G. R. (1983) A Review of the nudibranchs of the California Coast.Malacologia,24 (1-2), 114-246.

Odhner, N. H. (1937) Opisthobranchiate Molluscs from the western coasts and the northern coasts of Norway. Kgl. Norske Vidensk. Selsk. Skrift. 1, 1-92.

Padilla, D. K. (1998) Inducible phenotypic plasticity of the radula in Lacuna (Gastropoda: Littorinidae). Veliger. 41. 201-204.

Picton, B. E., Morrow, C. C. (1994) A field Guide to the Nudibranchs of the British Isles. Immel, London.

Reid, D, G., Mak, Y, M. (1999) Indirect evidence for ecophenotypic plasticity in radular dentition of Littoraria species (Gastropoda: Littorinidae).Jou. Mollusc. Stud. 65, 355-370.

Rudman, W. B. (1982) The Chromodorididae (Opisthobranchia: Mollusca) of the Indo-West Pacific: Chromodoris quadricolor, C. lineolata and Hypselodoris nigrolineata colour groups. Zoo. Jou. Linn. Soc. 76, 183-241.

Rudman, W. B. (1984) The Chromodorididae (Opisthobranchia: Mollusca) of the Indo-West Pacific: a review of genera. Zoo. Jou. Linn.Soc.130-137

Slattery, M., Avila, C., Starmer, J., Paul, V. (1998) A sequestered soft coral diterpene in the aeolid nudibranch Phyllodesmium guamensis Avila, Ballestesteros, Slattery, Starmer and Paul. Jou. Exp. Mar. Biol. Ecol. 226, 33-49

Thollesson, M. (1999) Phylogenetic analysis of dorid nudibranchs (Gastropoda: Doridacea) using the mitochondrial 16s rRNA gene.Jou. Moll. Stud. 65, 335-353.

Thompson, T. E., and Brown, G. (1976) British Opisthobranch Molluscs. Synopsis of the British Fauna No. 8. The Linnean Soc, Acad Press, London.

Thompson, T. E. (1976) Biology of Opisthobranch Molluscs. The Ray Society, London. Vols 1 & 2.

Thompson, T. E. (1988) Molluscs: Benthic Opisthobranchs (Mollusca Gastropoda) Synopsis of the British Fauna No. 8. The Linnean Soc, Acad Press, London. 2^nd Ed

Todd, C (1981) The Ecology of Nudibranch Molluscs. Oceanogr. Mar. Biol. Ann. Rev.,19, 141-234.

Valdes, A., and Gosliner, T. M (1999) Phylogeny of the radula-less dorids (Mollusca, Nudibranchia), with the description of a new genus and a new family.Zoologica Scripta.28, 315-360.

Wollscheid, E., and Wagele, H. (1999) Initial results on the molecular phylogeny of the nudibranchia (Gastropoda, Opisthobranchia) based on the 18s rDNA data. Mol. Phyl. Evol.13, 215-226.

Websites:

Nudibranchs: Marine Slugs with Verve

http://www.siolibrary.ucsd.edu/slugsite/nudi_han.htm

The Australian Museum Online, Rudman 1999:

http://www.austmus.gov.au.seaslugs

The Sea Slug Site:

http://slugsite.tierranet.com/phil.htm

Nudibranchs: Marine Slugs with Verve, Bertsch 1999:

http://siolibrary.ucsd.edu/slusite/nudi_han.htm

Bernard Picton’s Slug Site, Picton 1999:

http://www.picto.freeserve.co.uk/nudibranchs/anatomy.html