AEOLIDIIDAE

Miller, M. C., 2001, Aeolid nudibranchs (Gastropoda: Opisthobranchia) of the family Aeolidiidae from New Zealand waters, Journal of Natural History 35 (5), pp. 629-662 : 630-634

Family AEOLIDIIDAE

Diagnosis

Generally large aeolids, body usually broad and depressed except for prominent pericardial swelling; foot usually much wider than visceral part of body (except in slender forms), sides tapering gradually, then abruptly to tip of tail, tail short, front corners of foot rounded or drawn out as tapered processes, short to long; oral tentacles fairly short to very long, stout at base tapering to a rounded tip; rhinophores stout, sometimes quite short, simple, lamellate or papillate; cerata often very numerous and covering most of back, short to long, cylindrical or compressed, inserted irregularly dorsolaterally, in arches or vertical or oblique rows, sometimes in groups, sometimes on low swellings, set in single or partly double series; anus cleioproctic (amongst the anterior post-cardiac cerata), rarely notoproctic (on the back) or pleuroproctic (on the side, i.e. below the cerata); renal pore ad- or abanal, below or in front of anus, rarely behind, sometimes in space between cerata of right (anterior or pre-cardiac) and left (posterior or post-cardiac) digestive glands: oral glands usually composite and large, occasionally small or absent, one pair (rarely more); radula uniseriate, teeth broad, comb-like blade on single curve or double with cusp between two sides, denticles uniform or graded; jaws large, masticatory process with smooth edge, exceptionally, toothed; bursa copulatrix at inner end of distal part of oviduct (vagina not distinct externally), glans penis unarmed.

Genera. Aeolidia Cuvier, 1798 ; Spurilla Bergh, 1864 ; Aeolidiella Bergh, 1867 ; Cerberilla Bergh, 1873 (probably5 Fenrisia Bergh, 1888 ); Berghia Trinchese, 1877 ; Baeolidia Bergh, 1888 ; Protaeolidiella Baba, 1955 ; Aeolidiopsis Pruvot-Fol, 1956 ; Limenandra Haefel ®nger and Stamm, 1959; Anteaeolidiella gen. nov.; Burnaia gen. nov.

This family name, as the Eolididae (previously the`Enterobranches proprement’), was introduced by d’Orbigny (1839) to separate those members of his order Phlebenterata which possessed back processes Ðthus some sacoglossans, as we know them today, were grouped with aeolid nudibranchs. The root of d’Orbigny’s name comes from Eolidia Cuvier which, as Lemche (1964a) has shown, was originally spelt Aeolidia . Some years later (see Alder and Hancock, 1845±55) Allman (1845) used this family name but not J. E. Gray (see Gray, M. E., 1850) who created Glaucidae founded on the ®rst aeolid genus Glaucus . Gray’s family embraced ceratabearing nudibranchs other than aeolids. At the beginning of their monograph Alder and Hancock employed the family name Eolididae and, as well as sacoglossans, included other non-aeolid process-bearing sea slugs (classi®ed as dendronotaceans and arminaceans in the modern system); at the end they restricted the usage to the aeolids (though Embletonia Alder and Hancock, 1851 is considered now to be a dendronotacean). In his system of classi®cation Bergh (1892) used this family name in the same way except that he changed the spelling from Aeolidiidae, the correct form introduced by him in 1879, to Aeolidiadae (a spelling he had used earlier in 1859). He also included non-aeolids within the family (the last subfamilies Janidae and Heroidae ). The Aeolidiadae Propriae, ®rst of the eleven subfamilial divisions of his family Aeolidiadae , recognized for the ®rst time the distinctiveness of those aeolids with a broad rachidian tooth. These particular aeolids were recognized by VayssieÁre (1913) who named the grouping in the vernacular,`Aeolidiades’. Pelseneer (1906) presented a classi®cation of the aeolids somewhat diOEerent from that of Bergh, Eliot (1910) likewise except that he con®ned it to true aeolids (though still including Embletonia ); neither of these authors used Bergh’s subfamilies except in raising to familial rank his Fionidae and Janidae (by both authors, the Proctonotidae of Pelseneer5 Janidae), and Glaucidae (by Pelseneer only) and Heroidae (by Eliot only). Eliot (1910), however, separated the aeolid genera into three groups on diOEerences in the form of the radula (Triseriatae, Uniseriatae, and Serratae), these groups being adopted by Iredale and O’Donoghue (1923) and given the family names Calmidae (5 Serratae), Aeolidiidae (5 Uniseriatae) thereby re-introducing the correct spelling, and Flabellinidae (5 Triseriatae). It was Odhner (1939) who again recognized clearly the distinctiveness of those aeolids with a broad rachidian tooth, placing them in a separate family, the Aeolidiidae (5 Aeolidiadae Propriae of Bergh). The spelling of this family name was decided o cially in 1966 (see Lemche, 1964a; I.C.Z.N., 1966a). In his last classi®cation Odhner (in Franc, 1968) separated the three genera Spurilla , Berghia and Baeolidia as the Spurillidae on diOEerences in the form of the rhinophores and arrangement of the cerata. Recent classi®cation still recognizes these two families though the generic composition is diOEerent (see Vaught, 1989; Cattaneo-Vietti et al., 1990).

As with the other aeolid families, the constituent genera of the Aeolidiidae were clearly distinguishable early on when few species were known. However, the addition of new species (and genera) has blurred the generic diOEerences. Since Engel’s (1925) very detailed examination of the genera Spurilla , Aeolidiella and Berghia there have been attempts to either sharpen the existing de®nitions of certain of the aeolidiid genera or rede®ne them. Only two workers have reviewed all of the genera, Marcus (1961) and, more recently, Gosliner (1985). Having appraised the generic diagnoses myself, I can appreciate the problem of separating them clearlyÐthere is still much overlap or intergrading of the characters. On the present knowledge, the easiest, seemingly logical, solution would be to place all of the known aeolidiid species under the earliest generic name, Aeolidia . Characterization at the generic level would be clear and simple. Unfortunately, by doing this, the morphological and anatomical variation would be far greater than is desirable for a genus, going far beyond that shown by the large aeolid genera Flabellina Voigt, 1834 , and Cuthona Alder and Hancock, 1855 .

So, the alternative is the recognition of some or all of the existing genera with some rede®nition and the creation of new ones. A radical revision will still be required but must await information on the structure of the poorly known species, and ®ne structure, including molecular, of all of them. Gosliner’s (1985) revision of the genera is the most recent and thorough. He has maintained all of the genera except Limenandra ( Fenrisia had been synonymized already with Cerberilla ) which would seem a most satisfactory arrangement on the present knowledge. However, I would de®ne the genera somewhat diOEerently using the shape of the body and ceratal arrangement as the principal characters. My use of the second character diOEers from that of Gosliner’s. Though useful, I place less emphasis on the position of the anus and renal pore ( Marcus, 1957 and Rudman, 1982, 1990 have shown clearly that the positions can vary intraspeci®cally) and morphology of the rhinophores. The shape of the radular tooth, the principal diagnostic feature of the family, now shows little in the way of sharp distinctions which could be used to separate the genera. However, I have found that tooth shape is still useful in that the lines of variation follow the changes in ceratal arrangement.

In de®ning the genera here I have followed my previous concept ( Miller, 1971) of the basic or early aeolid having the cerata in a line along the pallial ridge, smoothedged jaws, a radula with a denticulate rachidian and several lateral teeth, and no, or at the most simple, oral glands ( Notaeolidia species , see WaÈgele, 1990). Thus the ®rst aeolidiids would have had these characters except for the subsequent loss of the lateral teeth and the (slight) widening of the rachidian tooth. Nearest to this stage is Protaeolidiella (two species according to Baba, 1992). However, this genus diOEers in having all of the groups of cerata joined by an extra-ceratal duct (a specialization) and the radular tooth lacking a distinct cusp. Protaeolidiella juliae (Burn, 1966) (as P. atra Baba, 1955 in Rudman, 1990), feeds on the athecate hydroid Solanderia fusca (Grey) and P. atra on S. secunda (Inaba) (see Baba, 1992). Hydroids are probably the ancient food of aeolid nudibranchs, rather than sea anemones which are the common food of the aeolidiids. Notaeolidia gigas , a species of what I consider to be the most primitive aeolid genus, eats a hydroid, Tubularia sp. (Barnes and Bullough, 1996). A notable feature of Protaeolidiella atra is the simple oral glands packed around the oral tube ( Rudman, 1990). Most other aeolidiids have composite oral glands. Rudman believes that the narrowish, evenly denticled rachidian tooth of P. atra (as P. juliae Burn, 1966 ) is a secondary development for eating hydroid prey, implying that it was derived by the narrowing of the typically broad aeolidiid tooth, rather than by widening of the narrow rachidian possessed by other aeolids including the primitive Notaeolidia . My view is that the radular tooth size and shape relationship with food type is a general one, broad or narrow, cusp projecting, retracted or absent, not specialized with a particular denticle and cusp size and arrangement. Thus, in the case of Protaeolidiella the basic form of the tooth, widish and slightly curved, could be primary, and the absence of a cusp secondary, which is my view and the one adopted here.

Further evolution of the aeolidiids has been principally the hypertrophy of the rachidian tooth (except in Aeolidiopsis ransoni Pruvot-Fol, 1956 ) and the re-arrangement of the cerata (and digestive caeca) and, naturally, the lateral ducts of the digestive glands, though the two are not always exactly the same, i.e., the positions of the cerata do not mark precisely the pattern of the terminal branches. The widening of the radular tooth has been accompanied by a change in shape, being either very broad and low, bow-shaped with uniformly-sized denticles or broad and high, bilobed with a cusp in the median notch with the denticles grading in size. In most cases the oral glands have become composite and large, and in a few cases the jaws have become toothed. It seems to me that changes in these features in certain combinations have taken several paths.

What follows is a delineation of the genera recognized here using the principal de®ning characters in various combinations, and an ordering of them which represents the paths of change (possibly representing evolutionary lines) outlined above. It must be remembered that none of those genera (and any of their constituent species) considered more primitive represents exactly an earlier evolutionary stage.

1. Radular teeth narrow, slightly curved, denticles uniform, median cusp minute or

absent; ceratal rows horizontal or slightly sloping; foot corners rounded; no composite oral glands.

(i) Ceratal rows horizontal on a weakly developed notal ridge; body narrowish, long and high; anus pleuroproctic (on the side of the body).... Protaeolidiella

(ii) Ceratal rows straight, sloping very gradually downwards and rearwards; body low; anus notoproctic (on the back)........... Aeolidiopsis

2. Radular teeth a moderately wide simple arch, denticles graded; ceratal rows straight, sloping gradually downwards and forwards; body low; anus pleuroproctic; foot corners rounded; no composite oral glands; a new genus is proposed but not named here for Aeolidiopsis harrietae Rudman, 1982 .

3. Radular teeth moderately wide to wide, strongly bilobed, cusp small or large,

denticles graded; cerata in rows, clusters of rows or arches; anus cleioproctic (surrounded by cerata); foot corners angular; composite oral glands large.

(i) Cerata in clusters of rows for the most part sloping downwards and rearwards; rhinophores simple; foot corners triangular; oral glands spindle-shaped with two opposed rows of large nodules (5 giant secretory cells).. Anteaeolidiella gen. nov.

(ii) Cerata in completely or partially double rows in arches on low swellings.

(a) Cerata cylindrical, sometimes a little ¯attened and sinuous; rhinophores lamellate; oral tentacles fairly short; foot corners triangular or weakly tentaculiform. Spurilla

(b) Cerata fusiform or claviform; rhinophores papillate; oral tentacles long; foot corners tentaculiform, longish............. Berghia

(iii) Cerata in many long, crowded yet evenly spaced rows, single or double at the lower end, curving downwards and forwards.

(a) Radular teeth with denticles graded gradually, cusp large; oral tentacles and rhinophores same size, moderately long; foot corners tentaculiform, longish..................... Aeolidiella

(b) Radular teeth with denticles steeply graded or a mix of large and small, cusp small or absent; rhinophores very short, oral tentacles very long; foot corners tentaculiform, long................... Cerberilla

4. Radular teeth a wide to extremely wide simple low arch, denticles uniform except

at ends of arch; anus cleioproctic; composite oral glands small or absent.

(i) Cerata in many long, crowded yet evenly spaced rows curving downwards and forwards; body very wide; foot corners tentaculiform, short.... Aeolidia

(ii) Cerata in arch-shaped rows, ends curving downwards and forwards, anterior limb of arches longer than posterior; body moderately wide; foot corners tentaculiform, moderately long................ Baeolidia

(iii) Cerata in horseshoe-shaped rows; body long and narrow; foot corners rounded................... Burnaia gen. nov.

(iv) Cerata in near vertical, widely spaced straight rows; body long and narrow; foot corners tentaculiform, longish........... Limenandra

The reorganization presented above means that some species will have to be transferred between the genera. Taking Gosliner’s (1985) review of four of the constituent genera (includes the three established genera considered in detail here) as the most recent and discerning statement on them, the required generic changes are:

Aeolidiella faustina Bergh, 1900 to Spurilla

A. indica Bergh, 1888 to Anteaeolidiella gen. nov.

Berghia australis ( Rudman, 1982) to Baeolidia

B. major ( Eliot, 1903) to Baeolidia , returned its original genus

Baeolidia benteva Marcus, 1958 to Anteaeolidiella gen. nov.

B. cryoporos Bouchet, 1977 to Aeolidia

B. fusiformis Baba, 1949 to Limenandra

B. harrietae ( Rudman, 1982) to proposed, unnamed, gen. nov.

B. nodosa (Haefel®nger and Stamm, 1959) to Limenandra , returned to its original genus.

B. palythoae Gosliner, 1985 to proposed, unnamed, gen. nov.

Reasons for these generic transfers are given later in the sections dealing with the genera. Otherwise the composition of the genera remains as listed by Gosliner (1985). However, there are several species, e.g., Berghia dela Marcus and Marcus, 1959 , which cannot be placed in a genus with any certainty because their structure is poorly known.