Memoirs OF THE Queensland Museum Brisbane Volume 20 June, 1980 Part 1 Mem. Qd Mus. 20(1): 1-47, pis. 1^. [1980] ALGAL-BEARING DIDEMNID ASCIDIANS IN THE INDO-WEST-PACIFIC Patricia Kott Queensland Museum ABSTRACT New material from the fringing reefs of Viti Levu and from the Great Barrier Reef form the basis for a review of the taxonomy, phytogeny, habitat, distribution and larvae of 18 species of the family Didemnidae displaying symbiosis with prokaryotic algae. The species are in the genera Didemnum, Trididemnum, Lissoclinum Echinoclinum and Diplosoma, and include two new species, one of which is endemic and probably isolated in an unusual habitat on the southwestern fringing reefs of Viti Levu. A remarkable organ for the transference of algal cells from generation to generation in Diplosoma spp. is described for the first time. The name rastrum, or plant rake, is proposed for this organ. The mechanism for gene flow in species with and 30°W and 1 50°E longitude is discussed. Nine algal-bearing Ascidiacea of the family Didemnidae from the Great Barrier Reef, were formerly reported by Kott (1977). The present paper discusses these and a further 9 species, and represents a review of all the known Pacific Ocean ascidian species that display an intimate association with blue-green algae. A large quantity of material from a wide range of locations (including the larvae and the types of most of the species) has been examined. It has been jxissible to resolve much of the confusion relating to the identity and synonymy of this interesting, albeit neglected, group of didemnid ascidians. The species in which plant cells are known to occur are in 5 genera (viz. Didemnum, Trididemnum, Lissoclinum, Echinoclinum and Diplosoma). The relationship does not appear to have evolved in the two remaining genera of the family (Leptoclinides and Polysyncraton). The view that the relationship between plant cells and host is symbiotic is supported by the presence of peculiar adaptations for the transferance of plant cells to the next generation in the 15 species for which larvae are known. In the genus Didemnum the larva is known only for D. molle and its mechanism for carriage of plant cells is similar to that in Lissoclinum spp. Trididemnum spp. have different mechanisms; and in Diplosoma spp. a unique and previously a wide range between 30°N and 30°S latitude. undescribed organ, the rastrum or plant-rake is present (see D. virens, below). In some cases the similarity between these larval adaptations for carriage of plant cells tends to support the view that there are close phylogenetic relationships between groups of species in each of the genera represented. Studies of the plant cells could provide further evidence in this regard. However, the fact that the ascidian/plant cell relationship appears to have evolved separately, and sometimes more than once, in each genus further supports the view that some mutual advantage is derived from the association. The ascidian colony at least provides a habitat for the algae, but it is not known what adaptive advantage is conferred on the ascidian colony by the algae. It should not be overlooked, however, that while other ascidians (including didemnid species) invariably occupy more cryptic habitats in the rubble zone behind the reef crest or in caves and crevices in deeper waters down the slope, certain of these plant-bearing species are extremely common on the reef flat where they are exposed to great light intensity {Diplosoma spp. and Lissoclinum voeltzkowi). There is a latitudinal gradient in the occurrence of these reef flat species that may result from increasing diurnal temperature range to the south, but may also be affected by light. Olsen (1979) has shown that colonies of T. cyanophorum do not grow in the absence of light. 2 MEMOIRS OF THE QUEENSLAND MUSEUM Much of the new material that comprises the basis of this report is from the reef flats of the fringing reefs of Viti Levu. Material from the Great Barrier Reef (Lizard L and Green I. in the north; Townsville in the central section; and Heron I. in the south) supplements the Fijian collections. Of the wide ranging species in the region, only Didemnum /nolle (> D. ternatanum: Kott, 1977) and Lissoclinum patellum were not taken from the Fijian reefs. These two species are more often found on the reef slope at a wide range of locations in the Indo-west-Pacific. When deeper waters are sampled, they may be found in the Fijian waters. Only a single new species of Trididemnum appears to be endemic in Great Barrier Reef waters. A few endemic species of more restricted range occur but are a relatively rare phenomenon. A new species of Diplosoma from Fiji, two new species of Trididemnum from the Philippines, and Diplosoma handi (Eldredge) from the Caroline Is. are the only plant bearing didemnids that appear to be endemic. It is apparent that there is some strategy available to most species that will ensure gene flow over a wide geographic area. Although didemnid/algal symbiosis was not thought to occur in the Atlantic Ocean (Kott 1977) it is now known in Trididemnum cyanophorum Lafargue and Duclaux, 1979, which occurs in profusion in shallow water coralline habitats off Panama and Guadaloupe. In this large investing species the algal cells are found embedded in the superficial layer of test and not in the common cloaca. Abbreviations: The following abbreviations are used below: AM Australian Museum, Sydney; AMNH American Museum of Natural History, New York; AMPI Australian Marine Photogra- phic Index; BM British Museum (Natural History), London; QM Queensland Museum, Brisbane, Australia; USNM United States National Museum, Washington D.C.; ZMA Zoological Museum, Amsterdam; ZMH Zoologis- chen Staatsinstitut und Zoologischen Museum, Hamburg. Didemnum moHe (Herdman, 1886) (Figs 1, 2; Plate M) Diplosomoides /nolle Herdman, 1886, p. 310. Sluiter, 1909, p. 85; 1913, p. 78. Not Lissoclinum molle: Kott, 1977, p. 618 (< L. bistratum). Newcomb and Pugh, 1975, p. 533 (< L. punctatum). Didemnum ternatanum: Van Name, 1918, p. 152. Tokioka, 1955, p. 47; 1967, p. 77; 1975, p. 326. Kott, 1966, p. 287; 1977, p. 618. Vasseur, 1970, p. 213. Newcomb and Pugh, 1975, p. 533. Millar, 1975, p. 229. Not Didemnum ternatanum: Kott, 1972, p. 179. Didemnum sycon Michaelsen, 1920, p. 44. Figs. 1-2: Didemnum molle (QM G9780) — 1, zooid; 2, larva (a, from right side; b, ectodermal scales on posterior haemocoelic chamber). Fig. 3: Didemnum viride (BM 07.8.30.41) — zooid. KOTT: ALGAL-BEARING ASCIDIANS 3 Material Examined New Records: Heron I.: January 1976, LWM, QM G9953 (juveniles); March 1975, 18 m, QM G9438 (juveniles); July 1975, 5 m, QM G9439 (juveniles); October 1976, LWM QM G9794 (juveniles); October 1979, LWM QM G 12629 (juveniles); December 1976, 9 m, QM G 11900 (mature gonads, larvae). Lewellyn Reef: August 1975, 3-9 m, QM G9765 (mature c^). Lizard L: June 1976, on coral rock shallow water, LWM, QM G9914 and G9979 (juveniles), QM G9777 (mature gonads some embryos), QM G9778 (larvae); on artificial reef 3—4 m between Lizard L, Palfrey L and Solomon L, QM G9780 (mature gonads and larvae). Palau Is., 1979, LWM, QM G 12680-2. Previously Recorded: Diplosomoides molle Herdman, 1886, Holotype BM 87.2.4.446. Didemnum sycon Michaelson, 1920, Types ZMH K1088, 1089; Van Name det., USNM 7384 (Philippine Expedition). Didemnum molle > D. ternatanum: Van Name, 1918, AMNH Chordata 2138, 2139, 2140. Didemnum ternatanum. Van Name det., USNM 5982, 2988 (Albatross), 6339; Tokioka, 1967, USNM 11404; Kott, 1977, QM G9652 (juveniles). Distribution Range: Malagasy, Zanzibar (Michaelsen 1920, Vasseur 1970). Western Australia: Cockburn Sound (Kott 1977). Northern Australia (Kott 1966). Great Barrier Reef: Lizard I. to Heron I. (Kott 1977). Indonesia (Sluiter 1909). Aru I. (Herdman 1886). Palau I. (Tokioka 1955, 1967). Philippines (Van Name 1918, Tokioka 1967, Millar 1975). Okinawa (Tokioka 1975). Habitat: Specimens have been taken from intertidal to 69 m (Sluiter 1909). Van Name (1918) reported the species growing on coral, shells, eel grass and other ascidians in shallow water. However, most of the stations from which specimens were reported by that author were in waters of 16—40 m, and only a few stations may have been in shallower water. Sluiter (1909) recorded specimens from reefs (intertidal?) to 69 m. In the present collection, the most robust colonies are in fairly protected habitats (in caves, under ledges, and in lagoons) in waters of 2-10 m, usually on dead coral, or rocky substrates. Smaller more or less flat colonies present under boulders near the reef crest, appear only to be juvenile colonies, and seldom achieve the typical hemispherical or stalked facies, although some have been found in pools near the LWM at Lizard I. These soft, flaccid colonies that secrete massive amounts of mucus when disturbed, do not appear to be adapted to intertidal habitats, or habitats on other than firm surfaces. Description Colony: The colonies are characteristically dome- to flask-shaped, with a common cloacal aperture in the centre of the upper surface. Some reach a diameter of 10 cm. They are firmly attached to the substrate by fine strands of spicule-filled test. The surface is smooth with a thin layer of very small spicules (0.005 to 0.015 mm) in the surface test. The spicules, together with a variable amount of brown pigment (especially dense in the vicinity of the common cloacal aperture) affect the colour of the colonies which may be from grey to reddish brown. Spicules are absent from the surface test between zooid openings only in very immature colonies when the green colour of the symbiotic algae is clearly visible. The 6 branchial lobes are conspicuous on the surface, each covered with corresponding lobes of spicule-filled test. One of the most conspicuous characteristics of this species is the soft and easily torn test, and the excessive amounts of mucus that are secreted by the living colonies. The common cloacal cavity is extensive, occupying most of the centre of the colony, and is lined with symbiotic algal cells. Zooids are contained in the soft test strands that cross the cloacal space, connecting the surface to the basal test, or to the core of test that projects from the base up into the centre of the colony. Juvenile colonies are small, flat and predomin- antly green, the small spicules being present only around the apertures. In preservative the juveniles are an almost completely transparent reddish brown colour. ZOOIDS; Zooids are up to 1-5 mm long. The thorax is especially long with 8-10 elongate rectangular stigmata. The wide atrial opening exposes most of the branchial sac. The 6 branchial lobes are pointed and their position where they open onto the surface is clearly marked by the spicules that are contained in the test. There is a long slender retractor muscle. The gut loop posterior to the stomach is curved ventrally. The vas deferens coils 6-5 times around the rather flat d* follicle. Larvae: These incubate in the central or basal test, the embryo having moved down the test connective in which the zooid is embedded. They are never present in the transparent juvenile colonies, found near the low water mark. The occurrence of these juvenile colonies suggests that breeding occurs throughout the year. Mature 4 MEMOIRS OF THE QUEENSLAND MUSEUM colonies with ripe gonads and sometimes with embryos taken in June, August and December, support this hypothesis. The larvae are about 0-9 mm long, excluding the tail, which is long and wound three quarters of the way around the larval trunk. There is an oozooid, with ocellus and otolith, and two blastozooids. Anteriorly a more or less rectangular frontal plate, supports the short rather thick stalks of the 3 median adhesive organs. The adhesive cells are arranged in a long rather narrow cone, surrounded by a deep ectodermal cup. As the larva matures, the frontal plate extends forwards on a rather narrow stalk and four long cylindrical ectodermal ampullae are produced from each corner of the frontal plate. There are modified columnar cells on the rounded ends of the ectodermal ampullae. The posterior haemocoelic chamber (erroneously referred to by Kott, 1977, as a ‘pouch’) into which the tail is inserted, is large and slightly flattened antero-posteriorly. The vacuolated cells referred to by Millar (1975) appear to be yolk cells from the spherical mass of yolk behind the blastozooids. These cells gradually disperse and some are found in the haemocoelic chamber. When the tail is extended the haemocoelic chamber becomes almost spherical and the plant cells adhere to the larval test outside it. A few larvae have been found with the larval tail prematurely withdrawn into this chamber while they are still in the common cloacal cavity of the adult colony (Kott 1977). In the mature larva there are flattened, overlapping scales on the ectoderm of the haemocoelic cavity. Their function is not known, but they resemble ectodermal cells of the rastrum or plant rake in Diplosoma spp. and possibly they are associated with the adhesion of the plant cells that cover the test in this region. The two abdominal buds are on the right side of the larva and the thoracic buds are on the left. Remarks: The colonies, their spicules, and the copious quantities of mucus that are produced are quite distinctive and the species is readily recognised. The atrial opening that exposes most of the branchial sac, so that the peribranchial cavity is almost entirely lost, is reminiscent of the condition in Lissoclinutn patellum and Diplosoma spp. The larval blastozooids are also reminiscent of the larva of Lissoclinum bistratum, as are the modified columnar cells on the ectodermal ampullae. The deep, narrow ectodermal cups and elongate cone of the adhesive organs are unique features of the larva. Van Name (1918) erroneously identified specimens of this very characteristic species as D. ternatanum, and illustrated his description liberally. Subsequent authors accepted Van Name’s identification and so perpetuated his mistake. Examination of his specimens, and of the types of D. molle and D. sycon has established their true identity. Van Name himself was aware of his error, and corrected the labels of these specimens in the American Museum of Natural History. Colonies have been observed to move up sides of aquaria at 1-5 cm per day over a period of at least 7 days (D. Griffiths, pers. comm.). Movement may be in response to subminimal conditions by successive growth and resorption of adhesive test strands as in Diplosoma listerianum (Carlisle 1961). It may cause clustering of colonies observed at the top of Acropora debris in lagoons. Didemnum viride (Herdman, 1906) (Fig. 3; Plate 1-2) Leptoclinum viride Herdman, 1906, p. 34. Didemnum viride Vasseur, 1970, p. 216. Not Trididemnum viride: Tokioka, 1967, p, 87 (< Trididemnum spp.). Material Examined New Records: None. Previously Recorded: Leptoclinum viride Herdman, 1906. Holotype, BM 07.8.30.41. Distribution Range: Ceylon (Herdman 1906). Malagasy (Vasseur 1970). Habitat: The holotype is from 2-5 m, south of Periya Paar, Ceylon. It is investing a skeletal fragment of Campanularia juncea (Jide Herdman 1906). The specimen from Malagasy is from a blade of sea grass. KOTT: ALGAL-BEARING ASCIDIANS 5 Description Colony: The colony is an extensive encrusting sheet, about 3 mm thick. The surface is marked into rounded and slightly raised areas by slight depressions over the common cloacal canals. The zooids are arranged along either side of these canals. The branchial apertures are evident on the surface of the colony as minute prominences. In the preserved specimen there is a very thin superficial layer of bladder cells mixed with algal cells. The remainder of the test is crowded with the conspicuously stellate calcareous spicules 0 03-0 04 mm in diameter. The thoraces of the zooids are firmly embedded in the test along their ventral surface. The cloacal canal is fairly restricted and is at thoracic level. Zooids: These are very small, about 0-7 mm. The branchial aperture is 6-lobed and the atrial opening is wide. There is a large oval lateral organ in the middle of each side of the thorax. There are four rows of only five oval stigmata. The three languets of the dorsal lamina are especially long. There are the usual very fine muscle fibres on the thorax that extend from between the rows of stigmata into two bands in the pharyngeal wall either side of the dorsal lamina and into the short retractor muscle where they are joined by fibres from the outer wall of the atrial cavity. The abdomina are firmly embedded in the solid basal test. There is a narrow duodenal area. The intestine curves ventrally and to the left before curving dorsally and then anteriorly so that the part of the gut loop distal to the stomach bends ventrally at an angle to the long axis of the zooid. There is conspicuous glandular material in the loop of the gut. The zooids in the colony are not sexually mature. Remarks: Owing to the lack of gonads in both recorded specimens, the generic status of this species is unconfirmed. However the oval stigmata, the flat lateral organ and the secondary curve of the gut loop suggest that Didemnum, rather than Lissoclinum, is the appropriate genus. The small size of the zooid, conspicuously stellate spicules and flat oval lateral organ together with embedded plant cells distinguish this species from other algal containing didemnid species that are presently known. Trididemnum clinides Kott, 1977 (Figs. 4, 5; Plates 3Ta,b) Trididemnum clinides Kott, 1977, p. 617. Trididemnum viride: Tokioka, 1967, p. 87 (part, zooids with atrial siphons). Trididemnum sp. Eldredge, 1967, p. 184. Material Examined New Records: Viti Levu, Fiji, August 1979: Votualailai, under cascades, fringing reef, QM G 12620. Previously Recorded: Trididemnum clinides Kott, 1977, Holotype QM G9928 (with larvae); Paratypes QM G9931. Trididemnum viride: Tokioka, 1967, USN M 1 1 646. Distribution Range: Great Barrier Reef: Heron I. (Kott 1977). Fiji; Votualailai (new record). Philippines (Tokioka 1967). Eniwetok (Eldredge 1967). Habitat: At Heron I. the species occurs just below the low tide mark in cryptic habitats near the reef edge and in the lagoon (where it is more common). The Fijian specimens were taken where the water of the fringing reef flat drains into the river channel that bisects it. The colonies are found deep in the dense algal mat that covers the reef beneath these cascades, where a fast unidirectional current flows for about half of each tidal cycle. The Philippine specimens were also taken from shallow water but little other information is available concerning their habitat. Description Colony; The colonies are small, almost spherical or oval or slightly irregular. Large colonies are more or less flattened on the upper surface. The zooids are arranged in a single circle toward the periphery of the colony, and there is a single central common cloacal opening in the centre of the upper surface. The test is very soft. The basal test and sometimes extensions of it attach the colonies to the substrate and they are often difficult to remove. There is a thin superficial layer of bladder cells. Spicules are sometimes but not always, dense in the border and base of the colonies. They are quite sparse in the surface of the colony except where there are dense patches over the anterior end of each zooid. The branchial apertures actually open toward the outer border of each of these patches of spicules. Elsewhere the spicules are evenly distributed and mixed with the algal cells that are embedded throughout the soft test. Algal cells are most 6 MEMOIRS OF THE QUEENSLAND MUSEUM Figs. 4—5: Trididemnum clinides (QM G1 2620) — 4, colony from the upper surface showing distribution of spicules, and branchial and cloacal apertures; 5, zooid with buds and stolonic vesicles (gonads not shown). Figs. 6-8: Trididemnum miniatum (QM G 12478) — 6, colony; 7, zooid; 8, larva. Figs. 9-10: Trididemnum strigosum (USNM 1 1681) — 9, zooid; 10, larva. Fig. 11: Trididemnum nubilum (USNM 1 1641 ) — zooid. KOTT: ALGAL-BEARING ASCIDIANS 7 common in the surface test. The colonies are a pale cloudy mustard green owing to the mixture of spicules and plant cells throughout the test. The common cloacal cavity is thoracic, but rather deep, extending the whole length of the thorax. There are small accumulations of spicules where the atrial openings enter the cloacal cavity. There do not appear to be any plant cells lying free in the common cloacal cavity. Spicules are stellate, from 0 03 to 0 04 mm with a variable number of rounded or conical rays. ZOOIDS: These are small (about 1 mm). The branchial lobes are deeply incised and there is a conspicuous short, wide atrial siphon with its border produced into six shallow obtuse lobes. There are 3 rows of 5 long rectangular stigmata in each of the 3 rows. There is a rounded lateral organ in the centre of the thorax and a very short retractor muscle. The gut loop is of the usual form with the vas deferens wound 6-5 times around the d" follicle. The zooids are colourless but there is often some brown pigment in the test. Larvae: These have been taken from specimens collected at Heron I. in January (QM G9922) and in the Philippines in January. They were not present at Votualailai in July. They are 0-6 mm long and the mature larva is enveloped in a dense coat of plant cells that are absent only from areas around the anterior end of the larva to expose the 3 median adhesive papillae, and a circular area over the otolith and ocellus. Ectodermal ampullae are characteristically short and rounded, at either side of the base of the stalks of the adhesive organs. Remarks; Tokioka (1967) believed Didemnum viride (Herdman, 1906) to be synonymous with a species of the genus Trididemnum which has been found to include specimens of the present species. All the species have plant cells embedded in the superficial layer of test. However, the 4 rows of stigmata that is characteristic of the genus Didemnum has been confirmed for Herdman’s species. Trididemnum viride: Tokioka, 1967, includes specimens of T. miniatum and two new species of Trididemnum in addition to the present species. The possible habitat preferences of T. miniatum and the present species are discussed below. The atrial siphon also distinguishes the present species from T. cyclops (which has similar spicules). They are further distinguished by the distribution of spicules, which in T. cyclops appear from the surface to be in a dense concentration around each branchial siphon that is continuous with the concentration of spicules around the border of the colony. Didemnopsis globuliferum Sluiter, 1913, which Tokioka (1967) has suggested as synonymous with his specimens, appears to be a junior synonym of Trididemnum discrepans (Sluiter, 1909). Trididemnum miniatum Kott, 1977 (Figs. 6-8; Plate 3-2) Trididemnum miniatum Kott, 1977, p, 617. Trididemnum viride: Tokioka, 1967, p. 8 (part, colonies with smaller spicules). Material Examined New Records: Green L; August, 1979, on seagrass, south of Jetty, close inshore (with larvae), QM G 1 2478. Heron I.: October 1979, cryptic near reef edge below LWM (with larvae) QM G 12622. Previously Recorded; Trididemnum miniatum Kott, 1977, Syntypes QM G9927 (with larvae); Paratypes QM G9945. Trididemnum viride: Tokioka, 1967, USNM 1 1661 (part), USNM 11796 (part). Distribution Range: Great Barrier Reef: Heron I. (Kott 1977); Green I. (new record); Philippines (Tokioka 1967). Habitat; The species has been taken with T. cyclops on coral debris and on weed below the low tide mark in cryptic habitats behind the reef crest and in the lagoon at Heron I. At Green I. however, the species was taken on the sandy inner reef flat, attached to sea grass blades. Philippine specimens are also taken in very shallow water. Description Colony: These are small and rounded to elongate. Some colonies are up to 1 0 cm in their maximum dimension, but more often they are about 4 mm in diameter. They are white to lime green or emerald green depending on the concentration of spicules in the test where they are mixed with algal cells. There are some fine veins of red pigment in the surface test in fresh material. Spicules are either present or absent in the superficial layer of test where the algal cells are most dense, and are quite dense throughout the remainder of the test, providing a white background for the green cells in the surface MEMOIRS OF THE QUEENSLAND MUSEUM layers of the colony that emphasises their colour. Spicules are usually absent from the surface test in the areas just over the zooids, although they are often present in the tips of the branchial lobes. The algal cells become progressively less dense toward the base of the colony where they are absent altogether. They are also present in the common cloacal canal. The spicules are only 0-01 to 0 02 mm, spherical, with numerous blunt ended rays. There is no conspicuous superficial layer of bladder cells in this species. The common cloacal canal is shallow and thoracic. The colonies are attached only lightly to the substrate by strands of test and are easily removed. ZooiDS: These are about 0-8 mm long. They are orange in fresh material but become colourless in preservative. The 6 branchial lobes are sharply pointed. The atrial opening is wide and transverse exposing most of the middle portion of the branchial sac. There is a rounded lateral organ half way down the branchial sac. There are 7 longish oval stigmata in each of the three rows. There is a fairly long retractor muscle from the posterior end of the thorax, extending most of the length of the abdomen. The gut loop is of the usual form, and the vas deferens coils 5-5 times around the d" follicle. Larvae: Larvae are present in most of the colonies that have been collected. They are known to be present from August to January. It is possible that this small and inconspicuous species breeds throughout the year. Mature d' gonads were present in the Green I. colonies but not in those from Heron I. in either October or December and it is probable that testes mature before the ovary. The larvae are of the usual Trididemnum form with a coat of algal cells present in the mature specimens. This algal coat is interrupted over the otolith and ocellus, and at the anterior end of the larval trunk, exposing the adhesive organs. There are 3 adhesive organs in the median line and 3 pairs of elongate ectodermal ampullae. The larval trunk is 0-7 mm long. Remarks: The relatively small spherical spicules and their absence from the test immediately over the anterior end of the zooids helps distinguish this species from two others from the Philippines with which Tokioka (1968) had confused it. All three species have small zooids and algal cells embedded in the test and all lack the atrial siphon of T. clinides. The present species often shares a habitat with T. clinides. Although both have a wide geographic range that includes the Philippines and the Great Barrier Reef, T. miniatum was not taken with T. clinides at Votualailai, nor was T. clinides taken with T. miniatum at Green I. Thus, they may have slightly different ecological requirements, the softer more delicate T. clinides possibly being less tolerant of reef flat conditions than the smaller T. miniatum, which however, makes a less robust attachment to the substrate than T. clinides, and is less often found where strong water flow and turbulence is likely to affect it. Further to the south (at Heron I.) T. miniatum is found in the lagoon and near the edge of the reef in cryptic habitats, but not on the sandy inner reef flat where it is found at Green I. Trididemnum strigosum n. sp. (Figs. 9, 10; Plate 3-3) Trididemnum viride: Tokioka, 1967, p. 87 (part, spicules with fewer rays). Material Examined Previously Recorded: Trididemnum viride: Tokioka, 1967, Holotype and Paratypes USNM 11681 (part, with larvae), 11640 (part), 11641 (part), 11642, 1 1649, 1 1661 (part), 11672 (part), 11796 (part). Distribution Range: Philippine Is (Tokioka, 1967). Habitat: The species has been taken principal- ly on algae and sea grass at depths of 1-6 m. Description Colony: Colonies are very thin investing sheets up to 2 cm in length. They are closely applied to the surface of the substrate, investing weed and sometimes coral fragments, and are consequently very irregular. Sometimes the colonies are elongate and lobulating. The test is densely packed with large stellate spicules, 0-05-0-08 mm in diameter, with 7 conical pointed rays in optical section. The test is brittle and its surface is granular owing to the density of the spicules contained in it. Plant cells are found embedded in the superficial layer of test above the spicules. They are especially small (0 004-0-01 mm). The cloacal cavity is thoracic and very limited. Zooids are evenly spaced, about 1 mm apart, and are seen from the surface as dark points interrupting the spicules. KOTT: ALGAL-BEARING ASCIDIANS 9 ZCX)IDS: Zooids are extremely small, only about 0-5 mm in total length. The dark pigmentation is usually conspicuous around the branchial aperture. There is a distinct branchial siphon with 6 triangular lobes. The atrial aperture is a transverse incision. There are 3 rows of 7 elongate-oval stigmata. A stalked lateral organ projects outwards from opposite the second row of stigmata on either side of the endostyle. The retractor muscle is of moderate length. The gut loop is fairly long, and flexed ventrally with a distinct right-angle bend in the rectum where it extends anteriorly. The follicle is deep and with 5-5 coils of the vas deferens. Larvae: Larvae are present in the colonies from Basilau I. in January. They lie in the test where they become surrounded by a capsule of plant cells that become embedded in the larval test. They are about 11 mm long with a short tail extending only half way around the larval trunk. There are 6 to 8 pairs of ectodermal ampullae that arise from the frontal area around the stalks of the three adhesive organs. The variation in the number of ectodermal ampullae is caused by their subdivision. The frontal area is separated from the central area of the larval trunk where the developing oozooid expands. The posterior haemocoelic chamber tapers posteriorly around the base of the tail. The plant cells embedded deep in the larval test are absent only from the area over the sense organs and anteriorly in front of the adhesive apparatus. As the oozooid swells in the centre of the larval trunk, the test and its contained plant cells becomes thinner in this area and the developing oozooid can be seen. Remarks: The zooid of the present species resembles that of T. miniatum, although it is even smaller. The species is distinguished by its very large and dense stellate spicules, and its large larva with embedded plant cells. The thin investing colonies are also characteristic. Trididemnum nubilum n. sp. (Fig. 1 1; Plate 34) Trididemnum viride: Tokioka, 1967, p. 87 (part, specimens with numerous rays). Material Examined Previously Recorded: Trididemnum viride: Tokioka, 1967, Holotype and Paratypes USNM 11641, 11640 (part), 11659 (part), 11661 (part), 11672 (part), 11680, 11681 (part), 11796 (part). Distribution Range: Philippine Is. (Tokioka 1967). Habitat: The habitat is apparently the same as that of the previous species, and is found on weed, in shallow water, and on or in the vicinity of coral reefs. Description Colony: The colonies are small and irregular but rather solid. The spicules are evenly spaced in the solid gelatinous test which is translucent and a slightly pink-brown colour in preservative. Spicules are seldom dense in any part of the test and sometimes they are sparse in the middle layers of test. The spicules are large (0 03-0 05 mm diameter). They are almost spherical, with about 10-14 short pointed rays in optical section. There are occasional spicules with blunt rays. The plant cells are slightly larger than those of T. strigosum (about 0 015 m diameter). They are mixed with the spicules in the upper layer of test, above the common cloacal cavity. It is this mixture of spicules, plant cells and test that confer the rather fluffy, soft appearance to the colony that is also reminiscent of T. clinides. ZooiDS: Zooids are evenly spaced in the colony and are sometimes brown in the preserved specimens. They are small, about 0-6 mm in total length. The branchial lobes are minute, the atrial aperture is a deep transverse incision and there is a conspicuous flat circular mass of spicules in the lateral organ which projects outwards on either side of the endostyle. There are three rows of 5 elongate-oval stigmata. The retractor muscle is strong and is sometimes very long. The gut forms a simple horizontal loop and the single d" follicle with 5-5 coils of the vas deferens, is rather flat in comparison with T. strigosum. Testes were mature in most of the colonies but no larvae were found. Remarks: This species with its incised atrial aperture, very small zooids and embedded plant cells, is distinguished from T. miniatum and T. strigosum by its large, almost spherical spicules, strong retractor muscle, and tough gelatinous test and translucent appearance. T. clinides resembles the present species in external appearance but is readily distinguished by its atrial aperture, larger zooids and smaller stellate spicules with fewer rays. The spicules resemble those of T. paracyclops but in the latter species zooids are larger and plant cells are confined to the common cloacal cavity. 10 MEMOIRS OF THE QUEENSLAND MUSEUM Trididemnum cyclops Michaelsen, 1921 (Figs. 12—14; Plates 41 a,b) Trididemnum cyclops Michaelsen, 1921, p. 19. Hastings, 1931, p. 89 Kott, 1962, p. 581; 1966, p. 286; 1977 (part), p. 616. Eldredge, 1967, p. 183. Tokioka, 1967, p. 85 (part). Thorne, Newcomb and Osmond, 1977, p. 575. Not T. cyclops Newcomb and Pugh, 1975 p. 534 (< Didemnum sp.) Lissoclinum pulvinum: Tokioka, 1967, p. 97 (part). Material Examined New Records: Fiji (Viti Levu) fringing reefs, July 1979: Vuda Point, in pools below the low water mark, outer reef flat, QM G 12453 (some larvae); Tailevu, amongst living coral at low water mark on reef edge, QM G 12455; Makaluva reef, in pools at low water outer reef flat, QM G 12460. Fiji (Viti Levu), fringing reefs, August 1979: Votualailai, beneath cascades tangled in algal mat, QMG12619. Previously Recorded: Trididemnum cyclops Michaelsen, 1921, Syntypes, 2MH KlllO; Tokioka, 1967, USNM 1 1482, 11483; Kott, 1977, QM G9942 (with larvae), G9944. T viride: Tokioka, 1967, USNM 11661 (part). Lissoclinum pulvinum Tokioka, 1967, USNM 11480 (part), 11643 (part), 1 1684 (part), 11669 (part). Distribution Range: Malagasy (Michaelsen 1921). North- ern Australia (Kott 1966). Great Barrier Reef: Lizard I. to Heron I. (Hastings 1931, Kott 1977). Philippines (Tokioka 1967). Gilbert I. (Tokioka 1967). Fiji; Viti Levu (new record). Eniwetok (Eldredge 1967). Habitat: The species lives in cryptic habitats on weed, coral and rocks just below the low water mark in pools behind the reef crest and in the lagoon, and between coral branches and in the interstices of algal mats at the reef edge. It is a common species but is found generally where water flow is not strong, or in habitats where it is well protected, and usually where it is well shaded. It is often found with Lissoclinum bistratum and T. miniatum. With Trididemnum miniatum and Diplosoma virens it is found in deeper and less exposed locations at Heron I. than at locations further to the north. Description COLONY: These are characteristically small and oval, usually less than 1 cm in length. Dense spicules line the borders and base and there is a thick superficial layer of bladder cells that surround the colony, usually extending around onto the base. This bladder cell layer is less dense on the upper surface. The spicules are usually absent from the upper surface, which is green owing to the plant cells in the thoracic common cloacal cavity. There is usually a single system in each colony, with a central, sessile cloacal aperture. Colonies appear to divide when they exceed 0-5 cm in length. A constriction in the border eventually spreads across the surface and divides the colony from the surface toward the base (see specimens QM G 12453, Vuda Point, Viti Levu, July, 1979). The zooids are arranged around the border, their ventral surface embedded in the test. The presence of the zooids around the circumference of the colony creates their characteristic surface appearance. Each whitish zooid projects into the green algal-filled cloacal cavity from the white spicule-filled border of the colony. The dark endostylar pigment cap is visible on the anterior surface of each of the zooids. There may be a single layer of spicules, of varying concentrations over the remainder of the surface test beneath the bladder cell layer. The spicules are fairly dense in a layer beneath the shallow thoracic common cloacal cavity but become less dense toward the base of the colony. There are 0 03-0 04 mm in diameter with some at 0 06 mm. There are two types of spicules, the most numerous with about 12-14 conical pointed rays in optical cross section. There are also spicules with more numerous, almost parallel sided rays. ZOOIDS: These are almost 1-5 mm long when extended. The branchial aperture has 6 pointed, deeply divided triangular lobes. The atrial aperture is a deeply incut opening. There are 7 long oval stigmata in each of the three rows. The retractor muscle is long and broad where it arises from the neck of the zooid. The abdomina are embedded in basal test and are only about one quarter of the length of the thorax. The gut loop is flexed ventrally. The vas deferens coils 5-5 times in an anti-clockwise direction around the d" follicle. LarvaE: Although they are present in colonies from Viti Levu in July, the more prolific colonies were those from Heron I. lagoon in December 1976 (QM G9942), and from Keeper Reef off Townsville in August 1977. In some colonies mature eggs and larvae are present but S glands appear to be spent suggesting their earlier maturation. The species may breed throughout the year, with a peak in larval production in the summer months. Mature embryos are 0-5 mm long. They begin their development in the basal KOTT: ALGAL-BEARING ASCIDIANS 11 Figs. 12-14; Trididemnum cyclops (QM G9942) — 12, colony showing endostylar pigment caps and distribution of spicules beneath the superficial bladder cell layer; 13, zooid; 14, larva. Figs. 15-16; Trididemnum paracyclops (QM G12628) — 15, zooid; 16, larva. 12 MEMOIRS OF THE QUEENSLAND MUSEUM test and subsequently move into the common cloacal cavity where they acquire a coating of plant cells that is interrupted over the sense ceils and around the front of the larval body to expose the two adhesive organs. The tail is stout and is wound three quarters of the way around the larval trunk. There are two pairs of ectodermal ampullae and the characteristic two adhesive organs. Remarks: The species is distinguished by the size of the colonies, by the single systems in which zooid are restricted to the border of the colony. The solid translucent superfical layer of bladder cells and the endostylar pigment cap occur in a related species of this genus (see below) whieh is distinguished by the size of the colony, a relatively large abdomen and greater number of vas deferens coils and the size of the larva and the number and modification of its ectodermal ampullae. Trididemnum paracyclops n. sp. (Figs. 15, 16; Plates 4-2 a,b) Trididemnum spiculatum Kott, 1962, p. 281 (part, specimen from Heron I.). Trididemnum cyclops: Kott, 1977, p. 47 (part, extensive sheets). Material Examined New Records: Heron I.: October 1979, cryptic habitats in pools behind reef crest, Holotype QM G 12627, Paratypes QM G 12628 (with larvae). Previously Recorded: Trididemnum cyclops: Kott, 1977, QM G9446, G9788, G9909, G9910, G9912, G9915, G9916, G9917, G9918, G9940, G9941 (with larvae), G9943 (with larvae), G12630 (with larvae). Trididemnum spiculatum Kott, 1962 (part, specimen from HeronI.),AM Y1627. Distribution Range; Great Barrier Reef (Lizard I. to Heron I.). It has been recorded principally from Heron I. where it appears to be very common but is also present amongst the specimens previously identified as T. cyclops (Kott 1977) from Lizard I. (QM G9912), Nymph L (QM G9788), Wilson I. (QM G9940) and Northwest I. (QM G9917). Habitat: The species is found investing dead coral (Acropora) branches and rock in shaded cryptic habitats below the LWM, in deeper pools behind the reef crest. Description Colony: Extensive and irregular investing sheets, up to 5 cm in greatest dimension and 2-3 mm thick. In one colony (QM G9943) a strip of the colony has overgrown the surface and fused with another part of the colony. This is a simple example of the phenomenon which, in Trididemnum cerebriforme, creates a most complex pattern. There is a superficial layer of bladder cells which surround the colony, extending around the border and onto its lower surface. In the border of the colony the spicules are fairly dense beneath this bladder cell layer. They are of variable, though even, density in the upper surface. The green colour of the colony is a result of the green plant cells seen through the surface test and it varies with the density of the white spicules in the surface layer of test. There is a dense layer of spicules below the shallow thoracic cloacal cavity (in which the algal cells occur) and beneath this they are absent altogether except for a single layer which is sometimes present where the test is attached to the substrate. Dark pigment occurs in the basal layer of test and sometimes in the zooids and this pigment often stains the preserving fluid most conspicuously. The spicules are large, from 0 03 mm to almost 0 08 mm. They are all of the one type with about 14 conical rays in optical transverse section. Zooids are evenly and quite densely distributed in the colony. Sometimes the peripheral zooids are embedded in the test of the border of the colony, as in T. cyclops, but more often the cloacal cavity extends around between the border and the zooids, which appear, from the surface, to be surrounded by a thin layer of spicules (contained in the thoracic test sheath) and the green cells within the cloacal cavity. The zooids can be observed through the surface test, although they are not as conspicuous as in T. cyclops. Small colonies differ from T. cyclops in the large number of zooids which are crowded evenly in the colony and not just arranged around the periphery and the endostylar pigment cap is not always visible from the surface in the present species. ZOOIDS: These are about 1-5 mm long when fully extended. The branchial lobes are fairly shallow. The atrial aperture is deeply incised. There are 7 stigmata in each of the three rows. The retractor muscle is relatively short and broad and does not extend beyond the abdomen. The abdomen is almost the same size as the thorax. The gut loop is fairly straight and is not flexed ventrally. There is a very large d" follicle with the vas deferens wound anti-clockwise 9-5 times around it. Larvae: Although most of the specimens collected are fairly robust, larvae are only present in small numbers in colonies collected in October KOTT; ALGAL-BEARING ASCIDIANS 13 and December. They are not present in colonies taken in August, September, January and March at Heron Island, nor in the Nymph I. specimens in June. Mature d" follicles are present in colonies collected in October. The available data suggest a restricted breeding period in early summer (October to December). The larvae are exception- ally large, with a body 1 mm long, the tail wound only half way around. There are only two adhesive organs but there are four pairs of ectodermal ampullae and an additional unpaired dorsal one. The plant cells do not form a coat around the larva as in other species in this genus but are attached to an area at the posterior end of the body where the posterior haemocoelic chamber is extended around the base of the tail. The algal cells adhere to the larval test either side of the mid line. As the tail straightens, its proximal end, in the posterior haemocoelic chamber, is drawn down and a pocket develops just above it in the postero dorsal part of the larval test. The plant cells that adhere to this part of the larval test are accordingly enclosed in an incipient cloacal cavity that, in due course, receives the atrial openings of the zooids that surround it. The larvae are apparently free swimming for a very limited period and a larva has been observed in which the tail is in the process of resorption in the posterior haemocoelic chamber before the larva is released from the cloacal cavity. A round yolk mass persists until a late stage in the elongate frontal section of the larva. Remarks; The species was formerly thought to represent large colonies of T. cyclops. The larvae are almost twice the size of those of T. cyclops, they have more ectodermal ampullae, and the plant cells are differently distributed on them. Closer examination of both species has demon- strated differences in the colony, the spicules, the proportions of the zooids and their retractor muscles, the course of the gut, and the number of vas deferens coils. It is also possible that colonies of T. cyclops lobulate when they reach a certain size while in T. paracyclops neither the size of the colony nor the number of zooids and systems in it are restricted in this way. The species is distinguished from T. natalense and T. savignyii by the absence of an atrial siphon and the presence of only two larval adhesive organs. Lissoclinum voeltzkowi (Michaelsen, 1920) (Figs. 17-19; Plates l -3a,b and 1-4) Didemnum voeltzkowi Michaelsen, 1920, p. 54. Hastings, 1931, p. 97. Material Examined New Records; Heron L: March 1975, under rocks, rubble zone, near eastern end of south reef, QM G 12633; December 1976, under rocks, rubble zone, eastern end of south reef; QM G 12632. Green I.; August 1979, inner sandy reef flat, on brown algae, QM G1 2474 (with larvae). Lizard L: June 1976, blue lagoon, on coral in shallow water, low tide, QM G9913, G 12621 (with larvae). Fiji (Viti Levu) July 1978: Malevu, fringing reef, sandy reef flat, QM G 12479 (with larvae); Suva Barrier Reef, sandy reef flat, G 12472 (with larvae); Makaluva reef, sandy reef flat, G 12473 (with larvae). Fiji (Viti Levu) August 1979; Votualailai under cascades in interstices of algal mat QM G 126 17. Philippines: Van Name del. AMNH Chordata 2128. Previously Recorded: Didemnum voeltzkowi Michaelsen. 1920, Types ZMH K1099, Kllll; Hastings, 1931, BM 30.12.17.44 (one colony in the Australian Museum, Sydney), BM 30.12.17.43. Distribution Range: Malagasy (Michaelsen 1920). Great Barrier Reef; Low Is. (Hastings 1931); Green I. and Heron I. (new records). Fiji; Viti Levu (new records). Philippines (new record). Habitat: The habitat requirements of this species are possibly the most restricted of all the plant-bearing didemnid species (with the excep- tion only of the endemic diplosomid from Viti Levu). They are found near the low tide mark, on the outer part of the open sandy reef flat. They are not normally cryptic. They often occur on weed (including Halimeda). On the fringing reefs of the Fijian island of Viti Levu a mosaic of colonies of this species, each colony contiguous along its borders with adjacent colonies, covers vast areas of the sandy reef flat inshore from the living coral zone. They were absent only from the northern fringing reefs at Tai Levu. On the Great Barrier Reef at Green I. the species is also found in abundance inshore from the living coral zone of the reef flat although at this location it is attached to sea-weed and is not lying on the sandy sediments. At Heron I. the relatively small number of small colonies are from the rubble zone near the reef edge at the eastern end of the reef, and at one location on the southern edge of the reef. Description Colony: The living colonies are a dirty greyish brown or cream or greenish cream with dense brown-black pigment contained in spherical cells that are scattered in the surface test and also gathered into small evenly spaced patches between the zooids around the edge of the colony and 14 MEMOIRS OF THE QUEENSLAND MUSEUM between papillae. Colonies are circular, oval or polygonal flattened cushions up to 4 cm in greatest dimension and up to 5 mm thick. Some of the smaller living colonies appear almost spherical. The borders of the colony are rounded and the test around the periphery of the upper surface is produced into pointed spicule filled conical papillae just outside the peripheral zooid openings. There are occasionally similar papillae from other parts of the upper surface (especially in the colonies from Green I.). The colonies are only loosely attached to the substrate by the basal test. When removed from the substrate the upper surface is depressed and the whole colony seems to contract, the border is elevated and the conical projections around the upper surface become more conspicuous. The colonies from Heron Island are less robust than others, the spicules are relatively sparse in the surface test and it is not produced into conical papillae. There are one or two cloacal apertures on short tubular projections in the centre of the upper surface. These are made conspicuous by the dense brown pigment in the surrounding test. The branchial apertures are short slits and those around the border are aligned parallel to it. In preserved material these slits are conspicuously green owing to the plant cells that crowd around the branchial sac where it is exposed to the cloacal cavity. Spicules are stellate with 8-12 conical or blunt ended rays in optical section. They are 0 02-0 04 mm in diameter. They are crowded always in the border and in the base of the colony, and usually they are crowded in the surface test and in the test strands that enclose the zooids. In those specimens that are in shaded positions, however, the surface layer of spicules is not dense and the surface of the colony is green owing to green cells in the common cloaca which can be seen through the surface test. Spicules are absent from the area surrounding the cloacal apertures. Cloacal canals perforate the upper layer of test in which the zooids are embedded, and are continuous with voluminous spaces posterior to the zooids. ZOOIDS: These are crowded in the test. They are about 0-8 mm long. The branchial lobes are reduced to a ventral and dorsal lip and the aperture is a slit. The atrial aperture is wide exposing much of the branchial sac. There are four rows of 8 longish oval stigmata. A deep egg shaped lateral organ projects into the peribran- chial space opposite the third row of stigmata at either side of the endostyle. Its attachment to the body wall is narrow so that in section it is comma shaped (Michaelsen 1920). Some very fine muscle fibres extend from the posterior end of the zooid and into the test but these are most inconspicuous and no actual retractor muscle can be demonstrat- ed. The gut loop is short and horizontal. The d" gland is undivided and the vas deferens is hooked around it, the proximal end of the duct extending posteriorly from the left side of the follicle, postero- ventral ly, and around onto the opposite side where it runs anteriorly along the inner side of the rectum. Larvae: These are present in specimens from Fiji in July and a few were present in colonies from Lizard I. on June 1973 (G9913). They were not present in specimens from Green I. (August 1979). Testis follicles appear mature in most specimens. The body of the larva is 1-5 mm long and the short tail is wound only about one third of the distance around it. There are usually three large adhesive organs, rather close together. However, one of these appears to have arisen by subdivision since the two uppermost organs share a basal common stalk. There is a ring of 8 pairs of ampullae around the stalks of the adhesive organs, each with a cap of modified columnar cells. The body of the mature larva contains an oozooid with otolith and ocellus, and two blastozooids. The larval test contains scattered patches of brown pigment cells that are absent only from the area around the posterior end of the body at the base of the tail, where a cap of plant cells adheres in the mature larva. As the larva matures the test appears to be growing back to enclose the posterior end of the larva, together with its adherent plant cells, in a pocket around the base of the tail, that is very likely an incipient cloacal cavity. Remarks: Some of the smaller almost spherical living colonies of this species bear a superficial resemblance to Didemnum tnolle in shape and colour. However the cloacal cavity is not so extensive, they lack the copious mucous secretion of D. molle, the spicules are relatively large and stellate and the arrangement of pigment, and the conical projections around the border of the colony distinguish them. Those green colonies with less dense spicules in the upper surface are readily distinguished from L. bistratum by the form of the spicules and usually by the large spherical pigment cells. The modification of the branchial lobes is also found in Lissoclinum bistratum and L. patellum. 0.1 ram KOTT: ALGAL-BEARING ASCIDIANS 15 Figs. 17-19: LissocUnum voeltzkowi (QM G 12474) — 17, part of the border of a colony showing pigment patches (dark shading), branchial apertures, and conical papillae around the borders of the colony; 18, zpoid (a, abdomen and thorax; b-d, gut loop and testis from the right, the left, and dorsally, respectively); 1 9, larya (a, showing test extending posteriorly over area where plant cells are present; b, later larva showing cap of plant cells and pigment patches on test). ''' 16 MEMOIRS OF THE QUEENSLAND MUSEUM Hastings (1931) apparently accepted Michael- sen’s interpretation of relationship of the vas deferens in this species with the characteristic spiral duct of the genus Didemnum. However its course is characteristic of Lissoclinum and is hooked around the posterior border of the gland from the ventral side (against the gut loop) and onto the opposite or dorsal surface. In the genus Didemnum the origin of the duct is from the centre of gland on the opposite side. Both the habitat the appearance of this species over most of its range is most characteristic. Specimens from Heron I. however are few and these have been taken only from cryptic habitats near the eastern end of the reef. Heron I. is possibly at the southern end of the range for the species. The density of spicules in the surface test of these specimens is also much less than most of the specimens further north. The spicules of the surface test undoubtedly protect the colonies from the direet sunlight of the reef flat. Those colonies that occur in shade, or even those parts of a colony that may be curved around the side of a rock into a shaded part of its surface, like the cryptic Heron I. colonies, have fewer spicules in the surface test so that the green plant cells in the cloacal cavity are fully exposed to the light. Light is therefore, possibly an important environmental parameter for this species. Either this factor, or the diurnal temperature range on the reef flat, may preclude its occurrence on the reef flats at the southern end of the Great Barrier Reef. Lissoclinum bistratum (Sluiter, 1905) (Figs. 20, 21; Plates 21a,b) Didemnum bistratum Sluiter, 1905a, p. 103; 1905b, p. 18; 1909, p. 46. Hartmeyer 1909, p. 1449. Michaelsen, 1920, p. 48. Didemnum gottschaldtiTokioka., 1950, p. 118. Lissoclinum pulvinum Tokioka, 1954, p. 247; 1967, p. 97 (part). Leptoclinum mode: Kott, 1962, p. 309; 1966, p. 30. Lissoclinum mode: Tokioka, 1967, p. 95. Kott, 1977, p. 618. Didemnum patella: Millar, 1963, p. 701. Didemnum chartaceum: Kott, 1962, p. 324. Material Examined New Records: Heron I.: October 1979, below low tide amongst rubble near reef edge, QM G 1 2626 (with larvae). SE. Queensland September 1977: Mooloolabah, 18 m on reef slope QM G 10108; Mudjimbah, 17 m on reef slope, QM G10125. Green I.: August 1979, rubble zone, QM G 12481. Coral Sea (Marion Reef): August 1977, 8 m on top of reef, QM G10170 (AMPl ascidian 204). Fiji (Viti Levu) July 1979: Suva Barrier Reef, in rock pools outer reef flat, LWM, QM G 12466; Makaluva, fringing reef, in pools outer reef flat, LWN, QM G 12464; Malevu fringing reef, in pools outer reef flat, QM G 12467; Tai levu, fringing reef, in pools amongst coral near reef edge, LWM, QM G 1 2483; Deuba, fringing reef, amongst living coral near reef edge, LWM, QM G 12581. Palau Is., LWM, QM G 12676. Previously Recorded; Didemnum bistratum: Michaelsen, 1920, ZMH K1607, K1108. Lissoclinum pulvinum: Tokioka, 1967, USNM 11386, 11396, 11480 (part), 11487, 11538, 11643 (part), 11647, 11663 (part), 11684 (part), 11669 (part), 11678 (part). Lissoclinum bistratum > L. mode: Kott, 1977, QM G9908 (with larvae), G9911, G9948, G9949 (with larvae), G9950, G9951 (with larvae), G9952. Distribution Range: Gulf of Aden (Sluiter 1905a, 1905b). Red Sea (Michaelsen 1920). Northern Australia (Kott 1966). Great Barrier Reef; Lizard I. to Heron I. (Millar 1963; Kott 1962, 1977). SE. Queensland (new records). Indonesia (Sluiter 1909). Japan: Tokara Is (Tokioka 1967). Palau Is (Tokioka 1950). Fiji: Viti Levu (new records). HABITAT: The species is taken in cryptic habitats with T. cyclops from just below the low water mark in the outer reef flat and in the rubble zone behind the reef crest. Macroscopically it resembled T. cyclops, with which it is often confused. Description COLONY: The colonies are small oval cushions or irregular investing sheets, never more than 5 mm thick. They seldom exceed 2 cm, but sometimes are as much as 5 cm in length (specimens from Mooloolaba QM G 101 08, G10125). Sometimes the common cloacal aperture is sessile, but it is often raised off the surface of the colony on a cylindrical projection, which may lie over the surface, presumably in the direction of the current flow (see Tokioka 1967). They are grey green on the upper surface. The green plant cells in the common cloacal cavity are visible through the surface test. There is usually a layer of bladder cells superficially. Spicules are dense in the borders, and in the basal test. They are often less dense at the level of the abdomina. KOTT: ALGAL-BEARING ASCIDIANS 17 19a Figs, 20-21: Lissoclinum bistratum (QM 12626) — 20, zooid; 21, larva. Figs. 22-23: Lissoclinum patellum (QM G9930, zooid; AM Y1335, larva) — 22, zooid (a, from the left latero-dorsal aspect; b, abdomen from the right); 23, larva. 18 MEMOIRS OF THE QUEENSLAND MUSEUM Spicules are evenly distributed in the surface test beneath the bladder cell layer, but their concentration varies from colony to colony and in some specimens they completely occlude the bladder cell layer. There are usually spicules associated with the branchial openings and when the zooids are contracted these are drawn down into the branchial siphon in a thick plug. The spicules are 0 03-0 05 mm in diameter. They are spherical, with many crowded blunt ended rays. The test is relatively delicate in this species. Black or reddish pigment is present in the bladder cell layer and sometimes in the body walls of the zooids. Zooids are most often white with a yellow stomach. The branchial lobes are reduced to a ventral and a dorsal lobe and the apertures appear as slits on the surface of the colony as in L. voeltzkowi. The zooids are surrounded by thoracic common cloacal cavities which connect with posterior abdominal spaces, although these are often quite restricted. The basal test is relatively thick. ZooiDS: These are about 0-9 mm long. The atrial aperture is wide and open exposing most of the branchial sac. There is a deep vertical egg shaped lateral organ projecting into the peribran- chial cavity on either side of the endostyle (near the border of the atrial aperture), between the second and third rows of stigmata. There are 6 to 8 long and rectangular stigmata in each row. The gut loop is simple and vertical. The cC gland is undivided and the vas deferens is hooked around the posterior end of the gland. LarvaE: These are present in colonies from Darwin in October 1965 and from Heron I. in December (QM G9951), January 1976 (QM G9949), March 1975 (G9908 and October 1979 (QM G 12626). Colonies from Fiji had some large eggs but no embryos in July (G 12581). The species may breed all the year with a peak in early summer. The large body is about 1 mm long the tail is wound half way around it. There are 3 large adhesive organs surrounded basally by 8 pairs of ectodermal ampullae. There is an oozooid and two blastozooids. The ‘pouch’ referred to by Kott (1977) is actually the posterior extension of the posterior haemocoelic cavity, with a layer of ectoderm and test outside it. The posterior cap of algal cells adhere to the larval test in this region to form the posterior cap, around the base of the tail described by Kott (loc. cit.j. The cells at tips of the 8 pairs of conspicuous ectodermal ampullae are modified columnar cells and form a concavity on the end of each ampulla. Remarks: The species is very closely related to L. voeltzkowi, which has a similar (though more extensive) cloacal cavity, similar zooids, lateral organ, branchial sac, gut loop, and larva. In its more characteristic form the colony of L. voeltzkowi is readily distinguished from L. bistratum. However in cases where spicules are not dense in the surface test of L, voeltzkowi or when they are dense in the surface test of L. bistratum the spicules constitute a reliable means to distinguish the species. This species has been reported further to the south than has so far been recorded for any other algal-bearing didemnid. Lissoclinum patellum (Gottschaldt, 1898) (Figs. 22, 23; Plate 2-2) Didemnoides patella Gottschaldt, 1898, p. 653. Didemnoides sulcatum Gottschaldt, 1898, p. 651. Didemnoides ternatanum Gottschaldt, 1898, p, 648. Didemnum patella: Michaelsen, 1920, p. 63. Tokioka, 1950, p. 115. Millar 1975, p. 228. Leptoclinum patella: Kott, 1962, p. 310. Lissoclinum patella: Tokioka, 1967, p. 97. Kott, 1977, p. 619. Diplosomoides tropicum Sluiter, 1909, p. 88. Didemnum meandrium Sluiter, 1909, p. 64. Not Didemnum ternatanum: Kott, 1977, p. 618 and synonyms (< D. molle). Not Didemnum patella: Millar, 1963, p. 701 (< L. bistratum). Material Examined New Records; Swain Reefs; October 1974, reef slope, QM G9930. Heron I.; November 1978, reef slope, QM G1 190. Palau Is., QM G12679. Previously Recorded: Didemnoides patella Gottschaldt, 1898, Holotype ZMH K1087. Lissoclinum parW/o.- Tokioka, 1967, USNM 11419, 11637, 11664, 1 1688; Kott, 1962, AM Y1335 (with larvae); 1977, QM G9278, G9456, G9887, Didemnoides ternatanum Gottschaldt, 1898, Holotype ZMH 595. Lissoclinum tropicum: Van Name det., AMNH Chordata 2143. Didemnum maeandrium Sluiter, 1909, Syntypes ZMA TU.459, TU.457. KOTT: ALGAL-BEARING ASCIDIANS 19 Distribution Range: Western Australia: Pt. Peron, Rottnest I. (Kott 1962). Great Barrier Reef: Heron I., Swain Reefs (Kott 1962, 1977). Indonesia (Gottschaldt 1898, Sluiter 1909, Millar 1975). Borneo (Sluiter 1909). Palau Is. (Tokioka 1950, 1967). Habitat: The species is one of the few plant bearing didemnids that does not occur intertidally. It is taken on vertical surface and under ledges from very shallow water to about 10 m. Larger specimens are taken only from deeper water. The species is large and robust and it is unlikely that adequate attachment to the substrate could be effected unless the substrate was free of sediments. Description Colony: Large colonies up to 1 cm in thickness. They have firm rounded gelatinous ridges on the surface. Cloacal and zooid openings are between these ridges. The basal test is also firm and gelatinous and as thick as the surface ridges. Small colonies consist of a single system that is surrounded by a ridge of gelatinous test around a central depressed area. Subsequently raised surface swellings develop (Tokioka 1967) and in due course assume the characteristic facies of the larger colonies (see Millar 1975, fig. 20). The thin zooid layer is conspicuous owing to the plant cells that fill the common cloacal cavity that surrounds the zooids and extends posterior to them. Spicules are also present in the zooid layer of test. They are characteristically spherical, and have a large size range, from 0 01 to 0 08 mm in diameter. ZooiDS: These are large. There are 4 rows of 14 very elongate stigmata. There are very fine muscle fibres in the transverse vessels that fan out either side of the endostyle and form circular bands around each zooid. There do not appear to be any longitudinal fibres along the dorsal or ventral borders of the pharynx and there is no retractor muscle. The long rectangular stigmata are almost completely exposed and the peribranchial cavity is almost completely lost, a narrow strip along either side of the endostyle being all that remains of the outer body wall. There is a small egg-shaped lateral organ projecting inwards between the second and third rows of stigmata. The two lobes of the branchial siphon, each with longitudinal muscle fibres that curve toward each other in the border, are marked by clumps of spicules where they are inserted into pockets of test. The oesophagus is long and the stomach and intestine form a simple loop that is flexed ventrally at an angle to the thorax. The single egg shaped d follicle lies in the usual position beneath (or dorsal to the loop of the gut). The proximal end of the vas deferens is from the right side almost at the end of the gland, and extends in a slight loop around on to the dorsal side of the gland to extend anteriorly alongside the rectum. Larvae: These are present only in colonies from Point Peron in January 1948 (AM Y1335). There are mature d" follicles and sperm-filled vas deferens in Swain Reefs specimens of October, 1974 (QM G9930). Neither mature gonads nor larvae were present in large colonies from Heron I. in either March or December. The larvae are large and almost spherical, 2-5 mm. The adhesive organs have unusually short stalks and the adhesive cone and surrounding ectodermal cup are unusually shallow. There is a single blastozooid unlike the larvae of L. voeltzkowi and L. bistratum which have two. There are plant cells around the posterior end of the larva outside the posterior haemocoelic chamber. There is also an unusually large number of ectodermal ampullae (about 15 pairs) with a rounded cap of modified columnar ectodermal cells on the tip of each. Remarks: Michaelsen (1920) interpreted the course of the vas deferens as an incipient spiral. However, unlike Didemnum spp. the ascending part of the vas deferens is on the opposite side of the gland to its point of origin. Its relationship to other species of the genus Lissoclinum rather than Didemnum is also unmistakable in respect of other characters. The cloacal system, the presence of blastozooids in the larvae, the modification of the branchial lobes, the virtual loss of the peribran- chial cavity, the absence of a retractor muscle and the long narrow stigmata all resemble the condition observed in L. voeltzkowi, L. bistratum and, with the exception only of the branchial lobes, L. punctatum. It is distinguished from all others by the great range in the size of the spicules and by the solid thick gelatinous colony that it forms. Although the larval form has similarities with the other Lissoclinum spp. above, it is also distinguished by its greater size. Didemnoides ternatanum Gottschaldt is des- cribed with a coiled vas deferens. An examination of the type in the Hamburg Zoologisches and Staats Institut Museum has shown it to be a small colony of the present species. 20 MEMOIRS OF THE QUEENSLAND MUSEUM This is the most robust of the plant-bearing didemnids, and with Didemnum molle, is found at greater depths than is usual for this group of species. Despite the greater depth at which the species is found, the plant cells are further screened from light by the high gelatinous translucent ridges on the upper surface. Lissoclinum punctatum Kott, 1977 (Figs. 24, 25; Plate 2-3) Lissoclinum punctatum Kott, 1 977, p. 620. Lissoclinum molle: Newcomb and Pugh, 1975, p. 533. Material Examined New Records; Green I.: August 1979, amongst rubble near reef edge, LWM, QM G 12456. Nelly Bay, Magnetic I.: August 1978, QM G11916. Heron I.: October 1979, cryptic, in pools below LWM near reef edge, QM G 12624. Fiji, Viti Levu, July 1979: Vuda Pt. fringing reefs in pools outer reef flat, QM G 12451 (with embryos); Malevu, fringing reefs in pools, outer reef flat, QM G 12452, (with embryos); Makaluva reef, in pools, outer reef flat, QM G 1 2462. Previously Recorded; Lissoclinum punctatum Kott, 1977, Holotype QM G9920; Paratypes QM G9426, G9926. Lissoclinum molle: Newcomb and Pugh, 1975, QM G8592. Distribution Range: Great Barrier Reef: Lizard I., Green I., Heron I. (Kott 1977; new records). Fiji: Viti Levu (new records). Habitat: The species is found just below low tide, in cryptic habitats, usually on coral skeletons. Description Colony: These are small, investing and very thin. They are bright green, with capsules of white spicules around each zooid. These are clearly seen through the test. The colonies are very soft and often disintegrate into a stream of mucus where attempts are made to scrape them from the substrate. In life, the plant cells are contained in common cloaca. However the test disintegrates so readily when the colony is disturbed that they are often found mixed with the test in the preserved material. There may be some dark spherical pigment cells in the surface test. The common cloacal cavity is large and extends posterior to the zooids but only occupies the upper half of the colony. The lower half consists of basal test. Spicules are present only in the capsules that sur'-ound the zooids, mature ova and embryos. They are spherical, 0 01-0 03 mm in diameter and are composed of radiating flat ended rods. ZCX)IDS: The zooids are small and each is completely enclosed in a capsule of spicules. The branchial aperture has 6 shallow lobes. The atrial aperture is wide exposing most of the branchial sac. There is a small, comma-shaped lateral organ from either side of the endostyle between the second and third rows of stigmata. There are about 8 long narrow stigmata in each of the 4 rows. Fine muscle fibres from the thorax extend down onto the oesophageal region but do not form a retractor muscle. The gut loop is simple and horizontal. There is a straight vas deferens that hooks around the posterior end of the single d follicle at its proximal end. Larvae: Mature eggs and some tailed embryos are present in specimens from Malevu and Vuda Point (QM G 12451 and G 12452) in July 1976 and mature ova were present at Green I. in August, 1976 (G 12456). Other colonies taken in May, August and September at various locations on the Great Barrier Reef did not contain mature gonads. However, the fact that these small colonies are present at all suggests that breeding may occur throughout the year. The tailed embryos are large (0-6 mm) and the tail is wound right around the larval trunk. They are not sufficiently well developed, however, to determine their structure. No blastozooids can be distin- guished, although the larval thorax is very small and the developing oesophageal neck is long as in L. voeltzkowi, L. bistratum and L. patella, all of which do have larval blastozooids. However, the yolk mass is anterior to the developing oozooid as in other species in which there is no larval blastozooid. (In those larvae with buds the yolk mass is always posterior to the oozooid and its buds). Remarks: The zooid of this species is similar to others in the genus Lissoclinum. It is character- ised by the small very soft colony and the capsule of small spherical spicules around the zooid. The 6 branchial lobes distinguish it from the three lissoclinid species discussed above in which the branchial lobes are modified. Echinoclinum triangulum (Sluiter 1 909) (Fig. 26; Plate 2-4) Diplosomoides triangulum Sluiter, 1909, p. 86. Lissoclinum triangulum: Kott, 1977, p. 620. Echinoclinum triangulum: Millar, 1975, p. 241. Echinoclinum philippinensis Tokioka, 1967, p. 93. Material Examined New Records: Heron L: October 1979, cryptic below LWM near reef edge, QM 12623. KOTT: ALGAL-BEARING ASCIDIANS 21 Previously Recorded: Lissoclinum triangulum: Kott, 1977, QM G9466, G9793, G9932. Echinoclinum philippinensis Tokioka, 1967, Types USNM 11791, 11790. Distribution Range: Indonesia: Saleyer, Amboina {Sluiter 1909, Millar 1977). Philippines (Tokioka 1967). Great Barrier Reef: Heron I. (Kott 1977). Habitat: The species occupies cryptic habitats in the rubble zone, and on the edge of the reef below the low water mark down to 20 m. Description Colony: The colonies are flat, soft, and yellowish green. Most records are of colonies less than 1 cm in length. One specimen from Heron Island at 20 m (G9466) is 6 cm long and 0-5 cm thick. There are spherical brown pigment cells in the test and the body walls of the zooids in some of the larger specimens. The whole test is filled with larger bladder cells. Beneath the superficial layer of test there is (in the larger colonies) a layer of spicules that become increasingly sparse towards the base. Spicules are always more common around the zooids, and in the smaller colonies the spicules are absent from the remainder of the test and only form a sparse coat around the zooids. 22 MEMOIRS OF THE QUEENSLAND MUSEUM The spicules are made up of fine pointed rays of varying length, grouped together to form a thickened central area that gradually flattens toward the periphery, where it is drawn out into 3-6 points, sometimes resembling the outline of a small asteroid echinoderm. The spicules are 0 03 to 0 08 mm in greatest diameter (from point to point). The secondary common cloacal cavity between the zooids is thoracic. However a very extensive posterior abdominal space extends to the surface around large groups of zooids that are joined to the basal test by a single test connective crossing the posterior abdominal cavity. Plant cells are present in the test and in the common cloacal cavity. ZOOIDS: The thorax and abdomen are each about 0-5 mm in length and the abdomen is bent up alongside the thorax. There are 6 shallow branchial lobes and a distinct sphincter muscle. The atrial aperture is wide and exposes most of the branchial sac when it is extended. There are 4 rows of 12 very delicate narrow elongate stigmata. There is a small narrow lateral organ that projects inwards on either side of the endostyle between the second and third rows of stigmata. There is no retractor muscle. The gut forms a straight simple loop. In freshly preserved material the stomach and the corresponding (glandular) part of the ascending limb of the intestine are clear yellow. The proximal part of the vas deferens extends anteriorly from the mid-dorsal surface of the single d" follicle. It is not hooked around the posterior border of the gland from its ventral side as is usual in Lissoclinum. Larvae: Although specimens from Heron I. taken in December have mature cf follicles no colonies have yet been taken from the Great Barrier Reef with mature 9 gonads. Colonies without mature gonads have been collected in September and October at Heron I. Millar (1975) found larvae in the base of one of the colonies he examined. They are 045-0-6 mm with a deep, almost spherical trunk and a short slender tail wound only about one third of the distance around the larval trunk. The adhesive organs have very slender stalks and shallow ectodermal cups. There are 6 ectodermal ampullae, two on each side of the base of the stalks of the adhesive organs and a dorsal and ventral ampulla. The cells on the tips of the ectodermal ampullae appear to be modified. The larval thorax is large and there are no blastozooids. There is an anterior yolk mass at the base of the ampullae. Remarks: The origin of the straight vas deferens from the mid dorsal surface, is the same as in Echinoclinum verilli. The present species and E. verilli are accordingly afforded a generic status that is distinct from the closely related Lissoclinum, A vas deferens of this type has also been described for Diplosoma handi Eldredge 1967 (see below: D. virens). The present species is also distinguished by the presence of plant cells in the test, the form of the spicules, the thoracic cloacal system, and the orientation of the abdomen that is flexed up against the thorax. The larva is also different from Lissoclinum larvae, having no blastozooids, a deep larval trunk, deep thorax, very slender adhesive organs and stalks, an anterior yolk mass and few ectodermal ampullae. Kott (1977) mistakenly recorded 2 d" follicles. The testis is undivided in this species. Diplosoma virens (Hartmeyer, 1909) (Figs. 27-30) Diplosoma viride Herdman, 1906, p. 341 . Leploclinum virens Hartmeyer, 1909, p. 1456. Tokioka 1943, p. 500; 1967, p. 68. Kott, 1966, p. 291. Diplosoma virens: Hastings, 1931, p. 102. Newcomb and Pugh, 1975, p. 533. Thinh and Griffiths, 1977, p. 673. Thinh, 1978, p. 617. Leptoclinum simile Sluher, 1909, p. 77 (part). Leploclinum varium Sluiter, 1909, p. 80. Leploclinum calificiforme Sluiter, 1909, p. 82. Van Name, 1918, p. 160 Not Leploclinum viride Herdman, 1906, p. 34 (< Didemnum viride) Not Diplosoma virens: Eldredge, 1967, p. 228. Kott, 1977, p. 620. Thorne, Newcomb and Osmond, 1977, p. 575. (< Diplosoma similis). Material Examined New Records; Heron I.: October 1979, cryptic, below LWM near edge of reef, QM G 12631. Green 1.: August 1979, inner sandy reef flat on Halimeda, QM G 12484. Fiji (Viti Levu) July 1979: Vuda Point, fringing reef, in pools, outer reef flat, QM G 12486; Malevu, fringing reef, in pools behind cascades and on inner sandy reef flat, QM G 12485 (some larvae); Makaluva reef, inner sandy reef flat, QM G 12461. Previously Recorded: Diplosoma viride herdman, 1906, Holotype BM 1907.8.30.42. Diplosoma virens: Newcomb and Pugh, 1975, QM G8593; Kott, 1977 (Heron I.), QM G9797, G9445, G9882, G9888, G9937 (with embryos); Kott, 1966 (Darwin), AM 1119 (with larvae) Leploclinum varium Sluiter, 1909, Syntypies ZMA TU.599.6, TU.599.8, ZMA TU.597. Leploclinum calciforme Sluiter, 1909, Holotypie ZMA TU.573. Leploclinum simile Sluiter, 1909, Syntype ZMA TU.591.2 (with larvae). KOTT: ALGAL-BEARING ASCIDIANS 23 Figs. 27-30: Dipiosoma virens (QM G1285, colonies; QM G 12484, zooids; QM G9937, larva) — 27, colonies (a-c, upper surface; d, side view); 28, section through part of a colony; 29, zooid (a, with extended thorax; b, contracted thorax); 30, larva. 24 MEMOIRS OF THE QUEENSLAND MUSEUM Distribution Range; Ceylon (Herdman 1906). Northern Australia (Kott 1966). Great Barrier Reef: Lizard I., Low Is., Green L, Heron I. (Hastings 1931, Kott 1977). Indonesia (Sluiter 1909). Palau Is., Gilbert Is., Philippine Is., Marshall Is (Van Name 1918, Tokioka 1967). Fiji: Viti Levu (new records). Habitat: The greatest depth recorded for this species is from off Amboina at 40 m (ZMA TU.599.6) and there are other records from similar depths. It is also found often on the open reef flat and around the sides of pools and on weed where it is just covered with water at low tide. At Heron Island the colonies are larger than is usual for the species, and are found along the reef edge and in surge channels amongst strands of living coral down to 5 m. Description Colony; The usual appearance of the living colonies is very adequately described by Herdman (1906, p. 341); ‘....numerous small rounded colonies closely placed .... The centre of the colony where the common cloaca is placed is depressed and of a paler green colour. The zone of ascidiozooids is also paler, while the outer ring of the colony, outside the ascidiozooids is the darkest and is usually of a very rich green colour.’ The living colonies are rounded egg-shaped or almost spherical. They closely resemble the sea-weed, Caulerpa, with which they are often associated. The depression is the centre of the upper surface that is referred to by Herdman is present only in contracted or preserved specimens. In larger specimens this depression causes the colonies to assume the ‘saucer’ shape described by Sluiter (1905) for D. caliciforme. The preserved specimens are yellowish mustard-green grey or brownish-yellow and opaque. The colonies are oval, up to 1 cm long and 2-5 mm thick. Colonies on the reef edge, in surf channels, at Heron I. are up to 2 cm long and about 5 mm thick. The common cloacal aperture, from the centre of the upper surface, is on a conical or occasionally a cylindrical protruberance. The latter may be bent over, to lie more or less flat over the surface of the colony, with the upper rim incised in a V extending back along the surface of the cylinder. The borders of the colonies are vertical and rounded. Zooids may be numerous, arranged fairly densely around the central area, but are often in a single circle around the periphery of the colony. The colonies are only loosely attached to the substrate by a small area of the lower surface, or by numerous slender test extensions from the under surface, or from around the border or by a stout, stalk-like extension of the test from the centre of the under surface. Many of the colonies are irregularly shaped and appear to be dividing. Such a growth pattern would explain the dense aggregations that are so characteristic of the species. The colonies are quite firm and the test is very tough and apparently fibrous. There is a surface layer of bladder cells. Beneath this, the fibrous test in which the zooids are embedded is perforated by a network of rather narrow interconnecting cloacal canals that open into a more extensive basal cavity that is posterior to the zooids and traversed by short basal test connectives that anchor large groups of zooids to the base of the colony. The colony is thus divided horizontally into four layers all of approximately equal thickness, viz,: the surface test, the zooid bearing layer perforated by a network of narrow canals, the posterior abdominal cloacal cavities, and the basal test. These layers are present in even the smallest colonies. The complicated network of thoracic canals, and the tough, rigid colonies cause the plant cells to remain trapped inside the preserved colonies and contribute to the opacity of the colonies. They are translucent only in the depressed central area where the cloacal cavity lies beneath the central aperture. The zooids can be removed from their tough test sheaths only with the greatest difficulty. Stolonic vessels extend from the abdominal regions of the zooids into the test connectives and especially into the basal test around the periphery of the colony. ZooiDS: In the preserved material the zooids are whitish with clear yellow stomach and glandular portion of the intestine. Sometimes there is some brown pigment in the abdominal body wall. Zooids are small, about 1 mm long. The thorax and abdomen are of equal length. There is active budding from the neck, with up to two sets of blastozooids present at the one time. The oesophageal neck is often very long. There are 5-6 rounded to oval or elliptical stigmata in each of the 4 rows. The stigmata are lined by rather flat epithelial cells. The atrial aperture is very wide, exposing most of the branchial sac. The branchial siphon is quite narrow with a pronounced sphincter muscle. There is a long retractor muscle that is free from about half way down the KOTT; ALGAL-BEARING ASCIDIANS 25 oesophageal neck. The thoracic musculature between the rows of stigmata and extending into the retractor muscle from the dorsal and ventral borders of the zooid, is fine but conspicuous. The abdomen forms an angle with the thorax and is flexed from the base of the oesophagus. The stomach is roomy and elongate. There are two d" follicles dorsal to the gut loop and a single straight vas deferens swollen and proximally hooked around the posterior border between the two lobes of the gland. Larvae: Larvae were present in colonies from Darwin in October 1965 (Kott 1966). A few were present in colonies from Heron I. in March 1975 and from Malevu in July 1979 and from the Celebes in January 1900. They are large, the larval trunk being slightly more than 1 mm long. The tail is wound about three quarters of the way around the trunk. There are usually three large median adhesive organs (one embryo was observed to have only two). There are two thick pairs of ectodermal ampullae, with modified columnar cells around their free ends. The oozooid has a large otolith and ocellus. There is a single blastozooid. In mature larvae there is a deep division between the oozooid and blastozooid. The latter remains associated with the frontal stalk from which the ectodermal ampullae and the adhesive organs arise. In each of the Diplosoma species with algal symbionts there is an identical organ for the transference of plant cells from the parent colony. The name rostrum is suggested for the organ. In the specimens from Darwin it has not evolved beyond the stage where it is a bilateral swelling of the posterior haemocoelic chamber above the base of the tail but in the specimens from the Celebes (ZMA TU.591.2) it reaches its full development. It first appears as bilateral outgrowths of the postero-dorsal part of the posterior haemocoelic chamber, either side of the mid line above the base of the tail. The ectoderm on these swellings is modified and produced into minute scale-like projections similar to those observed on the ectoderm covering the whole posterior haemocoelic chamber in Didemnum molle. As the embryo matures, the bilateral swellings become continuous across the mid line and the whole postero-dorsal aspect of the larva is constricted off from the larval trunk as a more or less cylindrical transverse arc with anteriorly curving horns at either end. A narrow median stalk from that part of the larval trunk over the base of the tail supports the organ. Its central cavity is continuous with the posterior haemocoelic chamber. The test over it is differentiated into a mass of very fine threads with terminal swellings that entangle the plant cells. As the tail is withdrawn into the posterior haemocoelic chamber, the rastrum is pulled into the larval test, which also overgrows it and its adherent plant cells as the cloacal cavity of the colony is formed. This parallels the process in Trididemnum paracyclops (see above). Remarks: The species is closely related to Diplosoma similis with which it has often been confused. The most conspicuous distinction is the shape, consistency, size and form of attachment of the colony, and the form of the common cloacal system. The zooids of D. virens are smaller and more narrow than those of D. similis and the stomach is shorter. D. virens larva has two pairs of ectodermal ampullae instead of three. The species occupy different habitats at low latitudes although at Heron I. they are often taken together. In his description of Leptoclinum varium Sluiter has not mentioned the presence of plant cells, however it is possible that he mistook these for what he reports as numerous blood corpuscles. Although the branchial siphon in the present specimens is much shorter than that figured by Sluiter, this character varies with the thickness of the surface layer of test. It is also affected by the degree of contraction of this rather muscular organ. One of the syntypes of L. simile (Sluiter 1909) is also identical with the present species and also contains plant cells, confirming the identification. It is possible that in these specimens plant cells were mistaken for pigment cells. The colonies of Diplosoma handi Eldredge, 1967 from the Caroline Is. resemble this species in their size, thickness, tough solid test, high cloacal apertures and the presence of vascular stolons in the thick basal test. They are distinguished by the absence of the posterior abdominal portion of the cloacal cavity. In this and other characters they resemble D. similis. The course and position of the gut loop, bent up alongside the thorax, the restricted cloacal cavity, the long branchial siphon and the vas deferens (which appears to extend anteriorly from the centre of the dorsal surface of the single d" follicle and is not hooked around from the other surface of the gland) are all reminiscent of Echinoclinum triangulum. 26 MEMOIRS OF THE QUEENSLAND MUSEUM Diplosoma similis (Sluiter, 1909) (Figs. 31-33) Leptoclinum simile Sluiter, 1909, p. 77. Diplosoma virens: Eldredge, 1967, p. 228. Millar, 1975, p. 241. Kotl, 1977, p. 620 (part). Thorne, Newcomb and Osmond, 1977, p. 575. Matfriai. Examined New Records: Green L: August 1979, outer reef flat on rubble, QM G12480. Fiji, Viti Levu, July 1979: Deuba, in pools, reef edge, LWM, QM G 1201 3 (turquoise colour), QM G 12014 (green colour), QM G 12450 (teal blue colour); Malevu, in pools behind reef crest, QM G 12454 (with larvae); Makaluva in pools behind reef crest, LWM, QM G 12465; Tailevu, in crevasses near reef edge, LWM, QM G 12006; Suva Barrier Reef, in pools near reef edge, LWM QM G12456; Serua in depression in sandy Reef flat, LWM, QM G 120 12. Line Is., Christmas L; February 1979, on reef flat, QM Gl 2010 (with larvae). Previously Recorded: Diplosoma virens: Kott 1977, QM G9924, G9789 (Lizard L); G9923 (Townsville, Magnetic L); G9795, G9880, G9883 (with larvae), G9884, G9885, G9886, G9889, G9933, G9934, G9935, G9936, G9938, G9939, G12634, G12635, G12636, G 12637 (Heron L). Leptoclinum simile Sluiter, 1 909, Lectotype ZM A TU.59 1.1. Distribution Range: Great Barrier Reef: Lizard 1. to Heron Is. (Kott 1977). Indonesia (Sluiter 1909). Eniwetok, Mashall Is., Hawaii (Eldredge 1967). Line Is.; Palmyra (Eldredge 1967), Christmas 1. (new records). Fiji: Viti Levu (new records). Philippines (Millar 1975). ? Japan: Tokara Is. (Tokioka 1954). Habitat; The species is usally found near the edge of reefs where it binds together the rubble (including coral skeletons) that collects in pools and crevices. At Heron Island it is seldom exposed directly to the light but occupies more cryptic habitats deeper down amongst the rubble. At lower latitudes, including Fiji, Low Isles and Darwin it is often found directly exposed to the light, lining the sides of pools behind the reef crest. Description COLONIES: These form thin encrusting sheets that are often very extensive. They adhere closely to the substrate by the whole of their basal surface, and their surface contours are irregular, corresponding with the contours of the substrate. They are seldom more than 2 mm thick. Their growth is not limited by the size of the substrate, for around the borders of the colony it often Figs. 31-32: Diplosoma similis (QM G9883) — 31, cross section through a colony; 32, zooid. KOTT: ALGAL-BEARING ASCIDIANS 27 28 MEMOIRS OF THE QUEENSLAND MUSEUM extends beyond the surface to which it is fixed until it attaches to an adjacent surface. Its capacity to bind rubble together results from this growth pattern that enables it to span the spaces between adjacent surfaces. The basal test is uneven and is often thickened where it fills in the irregularities in the surface to which it is attached. The colour of the living colony is very variable and may be almost any shade of dark, almost navy blue, bright green, or a whole range of teal or turquoise blue colours. The colonies at Heron I. are invariably green with patches of irridescent blue on the surface. Further to the north green colonies still occur but colonies of various shades of blue occur more often. At Low Is blue colonies of this species are the most common ascidian of the reef flat (I. Bennett, pers. comm.). In preservative the colonies become brownish green to grey as the plant cells lose their colour, and are themselves lost from the open common cloacal cavity. The test is translucent and without any other pigmentation. It is firm and gelatinous, but is not hard or tough, and is easily torn. The cloacal apertures are inconspicuous and sessile. The surface layer of test occupies about one third of the thickness of the colony, with a superficial layer of bladder cells. The anterior ends of the zooids are embedded in this surface layer. The common cloacal cavity is open, surrounding the test connectives that enclose the zooids between the surface and basal lest. Zooids are not clumped together to any extent and are often suspended in the cloacal cavity by individual strands of test. Around the borders of the colony especially the common cloaca does not extend posterior to the zooids, and the abdomens are embedded on the basal test. In the centre of larger colonies, however, the cloacal cavity is more extensive and extends posterior to the zooid around the base of the long test connectives. Normally the cloacal cavity and the basal test each occupy about one third of the thickness of the colony, except where the basal test is thickened to accommodate irregularities in the substrate (see above). In some colonies there are clouds of an opaque white or brown particulate deposit in the test. ZOOIDS: Zooids are 1 mm long. They are very evenly spaced in the test. The gut is at right angles to the thorax and there is a moderately long oesophageal region. The branchial siphon is well developed with a conspicuous sphincter muscle and 6 triangular branchial lobes. There is a moderately long prepharyngeal region. The atrial aperture is wide exposing most of the branchial sac. There are 4 rows of 6 round to oval stigmata. There is a fairly long retractor muscle that usually extends beyond the posterior end of the body. It is free from the body from a point near the branchial sac. There are stolonic vessels with the usual terminal ampullae projecting from the abdomen into the basal test. There are up to two sets of blastozooids. The body wall over the gut loop usually contains brown pigment although this is often lost in preservative. Occasionally the whole body wall of the zooid contains brown pigment. The abdomen is usually bent at an angle to the thorax from the base of the oesophagus. The stomach is especially long and roomy and gradually narrows at the pyloric end. It occupies almost the whole of the proximal limb of the gut loop. There are two d follicles and the vas deferens is straight. Larvae: These are present in the larger colonies in July, February, December, and March and there is no indication of seasonal breeding. They are large, 0-8 mm long, excluding the posteriorly projecting rastrum or plant rake. The tail is wound three quarters of the way around the larval trunk. There is an oozooid with otolith and ocellus, and a single blastozooid. There are 3 pairs of slender ectodermal ampullae, each with modified columnar cells forming a swelling at the free end. The ectodermal ampullae branch off either side of the frontal stalk that supports the three median adhesive organs. In the mature larva, there is a deep division between the oozooid and the blastozooid. The latter remains associated with the frontal stalk. There are many larvae present in the common cloacal cavity that display a fully developed rastrum (see above, D. virens). REMARKS: The species is distinguished from Diplosoma virens principally by the form and size of the colony, the simple two dimensional common cloacal system and absence of the conspicuous protruding cloacal apertures. The colony is less solid than that of D. virens and the plant cells are completely lost from the common cloacal cavity much more readily. The branchial sac is larger than in D. virens. The dark pigment in the body wall of the zooids is also characteristic of the present species — although it has often completely disappeared in the preserved material. The stomach is long and roomy and occupies almost the whole of one limb of the gut loop. The larvae of the two species are also similar and can be distinguished only by three pairs of ectodermal ampullae in the present species and a better KOTT: ALGAL-BEARING ASCIDIANS 29 developed rastrum. The synonymy of specimens previously assigned to D. virens with D. similis Sluiter, is based on the consistency, the thickness and colour of the colony and the size and form of the zooids, the gut loop and the branchial sac. The synonymy of D. midori (Tokioka 1954) was suggested by Eldredge (1967). The pyloric end of the stomach is more enlarged than in the present species. The retractor muscle of D. midori is shorter and more delicate, and the thoracic musculature is also more delicate than in the present species. Diplosoma midori (Tokioka, 1954) (Figs. 34, 35) Leptoclinum midori Tokioka, 1954, p. 1 1 . Material Examined New Record; Tonga: coral reef, LWM, coll. R. Lewin, 1 7.4.76 QM G 1 2669 Distribution Range; Japan: Tokara Is. (Tokioka 1954). Tonga (new record). Habitat: The species is taken on coral reefs. Description Colony: A single colony is available. It is an irregular sheet 4 cm in maximum extent and up to 5 mm thick. Part of the border projects as a free lamella with thick internal test and a layer of zooids on both sides and around the border. The test is gelatinous and translucent, the zooids showing as dark and whitish points. The surface layer of test is very thin indeed. The zooids are contained in test connectives that cross the cloacal cavity joining basal to surface test. The basal test is very thick. Here and there the surface test is produced into protruding common cloacal apertures. The plant cells are contained in the common cloacal cavity. Zooids; The zooids are evenly distributed. They are about 1-5 mm long. The branchial opening is almost sessile and muscles are very inconspicuous. The retractor muscle is very fine and short. There are 4 rows of 6 delicate oval stigmata. The thorax is delicate. The gut loop is relatively short and slightly flexed ventrally. The stomach is long and roomy and occupies the whole length of the proximal limb of the gut loop. There are two d" follicles. The abdominal body wall of the preserved specimens is darkly pigmented. Larvae: These are 1-6 mm long. The tail is wound only one third of the distance around the larval trunk. There is a single blastozooid and the rastrum is especially well developed in the mature embryoes. In a few specimens the proximal end of the tail is withdrawn into the haemocoelic chamber of the left rastral horn. This may be an artefact resulting from the fixation and preserva- tion of the colony. There are the usual 3 adhesive organs and 4 pairs of long slender ectodermal ampullae with modified terminal cells. Remarks: The species is distinguished from D. similis and D. virens by the very thin surface test, sessile branchial aperture, very delicate thorax and short fine retractor muscle. The zooid is distinguished from that of D. multipapillata by its delicate collapsible stigmata and the long roomy stomach. The colony of the present specimens is also distinctive in its double sided lamellae, that project free from the colony. In this it bears some resemblance to Diplosoma perspicum Sluiter, 1909 (> D. translucidum Hartmeyer, 1919). No plant cells have been located in the latter species which is further distinguished by its long, narrow stigmata. The free lamellae have not been described for the Japanese specimens, many of which were small colonies. It is most probable that there is considerable variation in this character, however, as there is in Diplosoma virens (above). The 4 pairs of ectodermal ampullae distinguish the larvae from these of all other species. Diplosoma multipapillata n. sp. (Figs. 36-39) Material Examined New Records; Malevu, Viti Levu, Fiji: fringing reef, close inshore, under cascades, coll. PK, 11.7.79. Holotype: QM G 12487, Paratypes: QM G 12488. Distribution Range; The species has been taken only from the type location, where it is prolific and conspicuous. 30 MEMOIRS OF THE QUEENSLAND MUSEUM Figs. 34—35: Diplosoma midori (QM G 12699) — 34, zooid; 35, immature larva. Figs. 36-37: Diplosoma multipapillata (QM G12487) — 34, colony, upper surface; 35, cross section through colony. KOTT: ALGAL-BEARING ASCIDIANS 31 0.1 mn Figs. 38-39: Diplosoma multipapillata (QM G 12487) — 36, zooid (a, zooid with blastozooids; b, branchial aperture and tentacles); 37, larva (a, whole larva; b, rastral hairs entangling algal cells). 32 MEMOIRS OF THE QUEENSLAND MUSEUM Habitat: It occurs close inshore, along the secondary or riverine rim of the fringing reef where it is bissected by the river channel that extends at right angles from the shore. The small colonies lie in profusion on the surface of the algal mat that covers this riverine rim and extends for a short distance down the upper gently terraced riverine slope. Here the waters of the reef flat drain into the river channel from about mid tide when the crest is exposed. The habitat is therefore swept by fast undirectional cascades of reef flat water for about half of each tidal cycle. There are similar cascades of reef flat water into river channels at other points along the south western fringing reefs of Viti Levu but the species apparently occurs only at Malevu. The habitat is a stringent one, and the species appears to be isolated in it (see below. Remarks). Description Colony: The colonies are a brilliant blue colour. In preservative this fades to a dull green, and later green-brown. Colonies are seldom more than 1 cm long, and are roughly elongate, elliptical, oval or irregular. They are about 1-5 mm in maximum thickness. The borders of the colony and parts of the basal surface are produced into branched holdfasts. The shape of some of the colonies suggests that they are lobulating. There is only a very narrow clear area around the border, the remainder of the surface being the brilliant blue of the plant cells lining the cloacal cavity. The whitish ampullae of the stolonic vessels from peripheral zooids are conspicuous in this translucent border. The surface and the basal test are especially thin and the common cloacal cavity is very extensive, and is crossed by zooids embedded in individual test sheaths. The common cloacal apertures are on fairly long narrow cylinders that are bent to lie horizontally across the upper surface. Their upper rim is often incised in a deep V that extends back along the upper surface of the tube. ZooiDS: Zooids are crowded in the colony. They are about 0*8 mm long. There is no branchial siphon and the aperture is sessile, with 6 rounded lobes. There does not appear to be a sphincter muscle. The ring of 6 large and rather swollen branchial tentacles that alternate with shorter ones is just inside the aperture. The prepharyngeal area is narrow. There are 4 rows of 5 absolutely circular stigmata with relatively small openings. The cells lining the stigmata are deep and the pore is actually a short cylinder through the gut wall. The circular diameter of the pore is firmly maintained, tending also to maintain the branchial sac in a fully expanded condition. The atrial apertures are wide exposing all of the branchial sac. There are 3 and sometimes 4 sets of oesophageal blastozooids. There is a short, slender retractor muscle with very few fibres. The oesophageal neck is very long. The gut is bent at a slight angle to the thorax. Although most colonies contained larvae, there were no mature cC follicles and it is assumed that they were spent. There appears to be a single cJ follicle. The vas deferens is straight, its proximal end hooked around the posterior border of gland. Larvae: These are large, about 1*3 mm long. The tail is wound almost all the way around the mid line of the trunk. On the right side of the anterior end of the larva there is a swelling from which 30 separate adhesive organs develop. In the mature larvae these spread out over the front. There is an oozooid with an ocellus and otolith and a single blastozooid. There are 3 pairs of long ectodermal ampullae with especially large columnar cells over the free end. The rastrum (see D. Virens, above) develops from the postero-dorsal part of the larval trunk, just above the base of the tail as in D. virens. There is a deep furrow that develops in the larval ectoderm and test, around the rastral area. In due course, this is separated from the larval trunk as an inflated horizontal arc with upturned horns that lies transversely across the posterior end of the trunk supported by a narrow median stalk. The central cavity of the rastrum is in continuity with the posterior haemocoelic chamber. The rastral hairs are thick and stronger than those of D. virens and appear to become less tangled. They clearly differentiate from the larval test in this region, and they appear to be associated with the enlarged overlapping ectodermal cells of the rastrum. The exact details of the relationship between the ectoderm, the test and the rastral hairs will be determined through histological studies that are beyond the scope of the present work. Before metamorphosis the posterior end of the larval trunk narrows, and the rastrum, with the plant cells gathered from the parental cloacal cavity, is drawn inwards, partly by the overgrowth of the larval test, and partly by KOTT: ALGAL-BEARING ASCIDIANS 33 the withdrawal of the posterior haemocoelic chamber, as the tail is drawn in, prior to its resorption. Remarks; The species is clearly most closely related to D. similis through its two dimensional cloacal cavity, well developed iarval rastrum and ectodermal ampullae. The colonies superficially resemble those of D. virens but they are much flatter. The species is most clearly distinguished from both by its extraordinary larval adhesive apparatus. Adhesive strands of test that often form a continuous fringe around the borders of the colony are also much more conspicuous than in D. virens. The cylindrical cloacal apertures lying across the upper surface in the direction of the current flow are also a feature that sometimes occurs in D. virens but is more consistent and conspicuous in the present species. The prolific vegetative blastozooid production that is associat- ed with proliferation of the colonies through lobulation, the sessile branchial aperture and large tentacles, the reduced musculature and very short, slender retractor, the small circular cylindrical stigmata and inflated thorax are also distinctive. The adaptive characters that distinguish this species from others are consistent with the requirements of its stringent habitat and suggest a possible explanation for its isolation and speciation. The adaptations are those that enable the colonies to effect and maintain rapid and firm attachment to the substrate, to maintain their population density for sexual reproduction (Kott 1974) and to accommodate the fast unidirectional current that flows over them for about half of each tidal cycle (while the salinity in this inshore extent of the river channel is at its lowest). Thus, it is possible that the river provides the barrier to dispersal of those larvae that have not effected immediate fixation following their release from the parent colony. This barrier could be effective only if larvae are not released when the tide rises above the reef crest and the freshwater runoff from the river is minimal. This requirement would be satisfied if larvae are released through the narrow cloacal aperture only when a strong current flows over the surface of colony. No data are available at present to support this hypothesis. However the release of larvae on ebbing spring tides is known for many marine species (Johannes 1978) and enhances selection for free swimming larvae to maintain gene flow in tropical waters (Kott 1974). Such a stimulus to larval release, if it is present in this species, would have resulted in its isolation in the unique conditions that prevail. CONCLUSIONS Phylogeny The characters of zooids and larvae of plant-bearing ascidians are summarised in Table 1 . It will be seen that there are certain characters common to the groups of species in each genus. Species in the genus Diplosoma show the closest affinity with one another and differences are principally found in the larvae and colonies. It is probable that the capacity to accommodate algal symbionts occurred only once in this genus, and that speciation occurred only after the algal/ ascidian relationship had been established. Distinguishing characters between species in each of the genera Trididemnum, Lissoclinum and Didemnum do not indicate such close affinities and there is no clear indication regarding the origin of the algal symbiosis. It is likely that a capacity for algal symbiosis was selected for separately in each of the species of Didemnum that demonstrate the phenomenon. Lissoclinum voeltzkowi, L. bistratum, and L. patella may represent a sister grouping having in common relatively large zooids, deeply invaginat- ed lateral organs, larval blastozooids and similar ectodermal ampullae. Lissoclinum punctatum and E. triangulum share some adult characters with other Lissoclinum spp., but their larvae are different and do not suggest a relationship either with one another or with the other three species. Thus, the algal-ascidian symbiosis may have evolved on two separate occasions in Lissoclinum and again in Echinoclinum. In Trididemnum the situation is confusing. The close relationships between adult zooids of T. paracyclops and T. cyclops are not supported by either the larvae or the respective methods of algal transfer. Further, although T. clinides differs markedly from others in the group in several respects, it does have the same mechanism of algal transfer as do T. miniatum and T. cyclops. At this stage an hypothesis that the unique method of algal transfer found in these species has been inherited from a common ancestor seems most convenient. If this hypothesis is accurate, speciation in T. clinides, T. miniatum and T. cyclops did not affect their plant transfer mechanism. In T. paracyclops however, isolation and speciation from T. cyclops, with which it shares many larval and adult characters, would have to have involved changes in the plant carrying capacity of the larvae that could be associated with its increased size. The larva of T. strigosum is the only one in which the plant cells are deeply embedded in the test. 34 MEMOIRS OF THE QUEENSLAND MUSEUM Geographic Range From Table 2, it can be seen that a range in the Indian and West Pacific Oceans is not unusual. It is also clear that records are far from complete at most locations: the majority of the widely ranging species are recorded from both the Great Barrier Reef and Fiji, It is likely that further surveys will show that a more comprehensive catalogue of species will be present over a great part of this region and that an extended range will be demonstrated for some of the species that appear at present to be relatively restricted. Habitats for some of the more common species are summarized in Table 3. Latitudinal effects on habitat are especially evident for species that occupy exposed reef flat locations at low latitudes. Lissoclinum voeltzkowi, Trididemnum cyclops, T. miniatum, Diplosoma Virens and D. similis all occur at Heron I. in cryptic habitats below low water mark near the edge of the reef. Here light values are considerably less than those they would encounter in the exposed habitats they usually occupy. It is possible that in this southern extent of their geographic range a greater diurnal temperature range on the reef flat may preclude their normal occurrence. L. voeltzkowi, however, only occurs occasionally at Heron I. and its range may be TABLE 1. Summary OP Characters OF AlCxAL-bearing Didemnidsofthe Indo-west Pacific* ZOO IDS COLONY Branchial lobes** Atria! aperture Retractor muscle Gut loop cf follicles Vas deferens spirals Size (mm) Endostylar pigment cap Lateral organ Stigmata Number of systems/ colony*** Spicules (u): spher- ical (sp). Stellate (st), flattened (fl) Plant cells:* Didemnum molle 6 wide long long i 6'A 1-5 — shallow 8 long 1 5- 1 5 sp CC viride 6 wide medium long ? ? 0-7 — shallow 5 oval >1 3(M0 St E Trididemnum ciinides 6 siphon short medium 1 m 1-0 none shallow 5 long 1 3(M0 St E miniatum 6 transverse medium medium ] 5'/: 0-8 none shallow 7 oval 1 10 20 sp E strigosum 6 transverse long medium 1 6'/2 0-5 none projects 7 oval >1 50-80 St E nubilum 6 transverse medium medium 1 5'/2 0-6 none projects 5 oval >! 30-50 sp E cyclops 6 transverse long medium 1 5'A 1-5 present shallow 7 long 1 40-60 St CC paracyclops 6 transverse medium medium 1 9>A 1-5 present shallow 7 long >1 30-80 St CC Lissoclinum voeltzkowi 2 wide none short 1 - 0’8 — deep 8 medium 1 2(M0 St CC bistratum 2 wide reduced short 1 — 0-9 — deep 8 long >1 30-50 sp CC patellum 2 wide none short 1 - 3-0 — deep 14 long >1 10-80 sp CC punctatum 6 wide none short 1 - 1-0 — deep 8 long >1 10-30 sp CC Echinoclinum Iriangulum 6 wide none medium 1 — 1-4 — deep 12 long >1 30-80 fl E Diplosoma Virens 6 wide long long 2 - 10 — none 5 oval 1 — CC similis 6 wide long long 2 - 10 — none 6 oval >1 — CC midori 6 wide short medium 2 - 1-5 ■ none 6 oval 1 — CC multipapillata 6 wide short long 1 — 08 — none 5 round 1 — CC handi 6 wide short long 1 — 10 — none 6 oval 1 — CC * Those characters that may be considered diagnostic of a species or of groups of species appear in italics. ** A branchial siphon is present, except in D. multipapillata and D. midori where the aperture is sessile. *** Colonies that form extensive sheets have >1 system; those with single systems may lobulate and divide to maintain the characteristically small sized colonies. KOTT: ALGAL-BEARING ASCIDIANS 35 restricted by lower light levels at higher latitudes. Records show that of all the reef flat species, L. voeltzkowi, L. bistratum and T. cyclops have the greatest range in both oceans. Of these, L. voeltzkowi which normally occurs in the most exposed locations on the reef flat is most restricted latitudinally. L. bistratum has the greatest latitudinal range of the reef flat species, its southern limit being off south eastern Queensland and its northern limits at the Tokara Is and the Red Sea. Diplosoma virens and D, similis have a similar recorded east-west range in both oceans. Although neither is known to extend west of Sri Lanka at present, D. similis is the only plant-bearing species known from Hawaii. The reef slope species D. molle and L. patellum also occur in both oceans. They are both common at Heron I. and are also known from Cockburn Sound. The former is also recorded from Okinawa and its latitudinal range is as extensive as that of L. bistratum. Didemnum viride (Ceylon and Madagascar) and Trididemnum paracyclops (Great Barrier ReeQ have more resctircted ranges and only Trididemnum strigosum and T. nubilum (Philippines), Diplosoma multipapillata (Fiji), and D. handi (Caroline Is.) are endemic. TABLE I (cont.) LARVAE DistributiontH Ectodermal cups Adhesive organs Ectodermal ampullae Modified ectoder- mal cells on ampullae Blastozooids Yolk massft Plant cell transfer Size (mm) Didemnum molle deep 3 2 prs none 2 posterior posterior barrel 0-9 IWP viride 7 7 7 7 7 7 7 7 WI Trididemnum clinides medium 3 3 prs none none anterior trunk encased 0-6 WP mimatum medium 3 3 prs none none anterior trunk encased 0-7 WP strigosum medium 3 6-8 prs none none anterior embedded 1-1 E(P) nubilum 7 7 7 7 7 7 7 7 E(P) cyclops medium 2 2 prs none none anterior trunk encased 05 IWP paracyclops medium 2 4'^ prs none none anterior posterior cap 10 GBR Lissoclinum voeltzkowi medium 3-2 8 prs present 2 posterior posterior cap 1-5 IWP bistratum medium 3 8 prs present 2 posterior posterior cap 1-0 IWP patellum shallow 3 15 prs present 1 posterior posterior cap 2-5 WP punctatum medium 3 (?6 prs) 7 (?none) anterior 7 0-6 WP Echinoclinum triangulum medium 3 3 prs (?present) none anterior 7 0-6 WP Diplosoma virens medium 3 2 prs present 1 posterior rastrum 1-0 IWP similis medium 3 3 prs present 1 posterior rastrum 10 WP midori medium 3 4 prs present 1 posterior rastrum 1-5 WP multipapillata shallow 30 3 prs present 1 posterior rastrum 1-3 E(F) handi 7 7 7 7 7 7 7 7 E(C) t E=embedded, cc= in cloacal cavity. 1 1 The position of the yolk mass is apparently related to the presence or absence of blastozooids. Iff IWP — Indo-west Pacific; WP ~ West Pacific; GBR - Great Barrier Reef; E (P,C,F) — Endemic (Philippines, Caroline Is, Fiji); WI — West Indian Ocean 36 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 2: Geographic Range of Plant-bearing Didemnid Ascidians * Species S .E -o r— ce: OS O 10 to J ^ = J2 .E J ,| S I a — ,E c o u. f- 5<’S 20°S 20°N 8°N 10°S iO°S 32°S 24°S 28°S 0° lO^N 30°N 26°N 10°N lO-'N 10°N 20°N 0“ 0° 20°S 23°S D. viride D. molle T. Cyclops L voeltzkowi X X X X X X X X X X XX X X X X X L bistratum D. Virens D. similis X XX XX XXX X X X X X X X X XXX XXX XX X T. clinides L punctatum D. midori XX X X X X X X L. patellum L triangulum T. miniatum X XX X X X T. paracyclops X X X X * Endemic species recorded from single geographic Locations only are not included. The mechanism whereby these ubiquitous species avoid isolation and speciation is not known. Kott (1974) suggested that pelagic larvae are the means by which gene flow is maintained, despite opposite pressures to avoid dispersal. Crisp (1977) and Johannes (1978) have also discussed the role of the pelagic larvae of marine organisms. Where habitats are transient the role of the larvae to disperse and colonise is acknowledged. Yet it is in the tropics where habitats are not transient that sessile organisms with the free swimming larvae are most common. Johannes (1978) has explored an hypothesis that fish gametes and larvae are dispersed as a result of selective pressures imposed by predators in the parental shallow water habitats, and he has argued that some progeny are likely to be recruited back into the parent population by means of gyres or other circulatory patterns. Crisp (1977) pointed out that dispersal per se is wasteful; and that adaptive advantages associated with it must therefore transcend opposite pressures to maintain populations. It is most unlikely that pressures to avoid wastage (through predators) would have caused the evolution of other strategies (such as those described by Johannes) that also promote dispersal and loss from the adult population. TABLE 3: Characteristic Habitats of Reef Flat Species at Low Latitudes. Inner sandy reef flat Outer sandy reef flat Living coral zone, outer reef flat (with pools) Rubble zone and Reef rim with surge channels, pools Reef Slope T. miniatum L. voeltzkowi L. bistratum D. similis L. triangulum D. Virens T. Cyclops L. punctatum L. patellum T. clinides D. molle KOTT: ALGAL-BEARING ASCIDIANS 37 1. Red Sea 2. Gulf of Tadjourah 3. Zanzibar 4. Malagasy 5. Ceylon 6. Indonesia 7. Darwin 8. Cockburn Sound 9. Borneo 10. Philippines 11. Tokara Is. 12. Palau Is. 13. Caroline Is. 14. Eniwetok 15. Marshall Is. 16. Hawaii 17. Gilbert Is. 18. Line Is. 19. Fiji 20. Lizard 1. 21 . Green I. 22. Heron I. Fig. 40: Locations between Latitude 30*^S and 30°N, and Longitude 30°E and 130°W where didemnids with prokaryotic symbionts have been recorded (see Table 2). Therefore it seems likely that some advantage other than population maintenance must be associated with the existence of free swimming larvae. The length of larval life of the present species is not known. The condition of the developing adult organs and larval adhesive organs before their release from the parent colony do not suggest a long larval life. Further, the larvae are generally large and it is unlikely that their relatively short tails are especially effective swimming organs. The larvae do not appear to be ideal as vehicles for gene flow between very isolated populations (see also Crisp loc. cit.). Nevertheless, they appear to be the only vehicle by which gene flow can occur. Their recruitment through the chains of islands and reefs that exist around the Indo-west Pacific coralline region would explain the lack of endemism that is characteristic of this region. If they have been selected for in order to fulfil this role, it is probable that there are associated strategies that will ensure their maximum dispersal. The probability of larval release on ebbing tides when there is maximum water flow over the parent colonies is discussed above (see Diplosoma multipapillata, Remarks). ACKNOWLEDGMENTS The opportunity to visit Fiji and collect from the fringing reefs of Viti Levu resulted from an invitation from Professor John Ryland of the School of Natural Resources and Dr Uday Raj of the Institute of Marine Resources, both in the University of the South Pacific, Suva, Fiji. The visit to Fiji was jointly sponsored by the Queensland Museum and by the University of the South Pacific. Dr S. van der Spoel of the (Zoologisches Museum, University of Amsterdam), Dr Michael Dzwillo (Zoologisches und Staats Institut, Flamburg), Dr Ernst Kirsteuer (American Museum of Natural History), Dr David Pawson (American National Museum, Smithsonian Institution), Miss Ailsa Clark (British Museum) and Dr Frank Rowe (Australian Museum) have made it possible for me to examine specimens from the collections in their care. Carlisle (1961) has observed metamorphosed colonies of Diplosoma listerianum either in mid water or attached to the surface film. Distribution of these species may therefore occur irrespective of the length of larval life. I am also grateful to Lucille Crevola-Gillespie who drew the text figures, and to Robert Raven and Alan Easton for the Scanning Electron Micrographs. Janet Byrne typed the manuscript. 38 MEMOIRS OF THE QUEENSLAND MUSEUM LITERATURE CITED Carlisle, D.B., 1961. Locomotory powers of adult ascidians. Proc. zool. Soc. 136 : 141-6. Crisp, D., 1977. In ‘Genetics, Ecology and Evolution.’ Eds B. Battaglia and J. A. Beardmore. NATO Conference Series IV Marine Sciences (Plenum Press: New York). Eldredge, L. G., 1967. A taxonomic review of Indo- Pacific didemnid ascidians and descriptions of twenty-three central Pacific species. Micronesica 2 : 161-261. Hartmeyer, R., 1909. Ascidien (continuation of work by Seeliger). In Bronn, H. G., Klassen und Ordnungen des Tier-reichs. Leipzig, 3, suppl., (89-98): 1281-772. (Abstract, repeating lists of species by Schepotieff, A., in Arch. Naturgesch., 1911,6: 3-27). Gottschaldt, R., 1898. Synascidien von Ternate. Ahh. Senckenb. Naturf. Gesellsch. 24 : 641-66. Hastings, A. B., 1931. Tunicata. Scient. Rep. Gt Barrier Reef Exped. 4 (3): 69-109. Herdman, W. a., 1886. Report on Tunicata collected during voyage of H.M.S, ‘Challenger’ during years 1873-76 Pt. II. Ascidiae compositae. Zool. Chall. Exp. 14 (38): 1^25. Michaelsen, W., 1920. Die krikobranchen ascidien des westlichen Indischen ozeans: Didemniden. Jb. hamb. Wiss. Anst. 37: 1-74. 1921. Ascidien vom westlichen Indischen Ozean aus dem Reichmuseum zu Stockholm. Ark. Zool. 13 (23): 1-25. Millar, R.H., 1963. Australian Ascidians in the British Museum (Natural History). Proc. zool. Soc. Lond. 141 (4): 689-746. 1975. Ascidians from the Indo-west-Pacific region in the Zoological Museum, Copenhagen (Tunicata, Ascidiacea). Steenstrupia 3 (20): 205-336. Newcomb, E.H. and Pljgh, T.D., 1975. Blue-green algae associated with ascidians of the Great Barrier Reef. Nature 253 : 533^. Olson, R., 1979. Ecological investigation of Trididemnum cyanophorum, Lafargue and Du- claux, 1979: a compound tunicate with symbiotic algae. Unpublished manuscript. Harvard University. Sluiter, C.P., 1905a. Tuniciers recueillis en 1904 par M. Ch. Gravier dans la Golfe de Tadjourah (Somalie francaise). Bull. Mus. Hist, nat., Paris 11 : 100-3. 1905b. Tuniciers recuellis en 1904 par M. Ch. Gravier dans le Golfe de Tadjourah (Somalie francaise). Mem. Soc. zool. Fr. 18 : 5-21. 1909. Die tunicaten der Siboga Expendition Pt. 2. Die merosomen Ascidien. Siboga Exped. 56B: 1-112. 1913. Ascidien von den Aru-Inseln. Abh. senckenb. naturforsch. Ges. 35 : 65-78. Thinh, L.V., 1978, Photosynthetic lamellae of Prochloran (Prochlorophyta) associated with the ascidian Diplosoma virens (Hartmeyer) in the vicinity of Townsville. Aust. J. Bot. 26; 617-20. 1906. Report on the Tunicata. Ceylon Pearl Oyster Fisheries, suppl. rept., 39 : 295-348. Johannes, R.E., 1978. Reproductive strategies of coastal marine fishes in the tropics. Environmental Biology of Fishes. Env. Biol. Fish. 3: 65-84. Korr, Patricia, 1962. The ascidians of Australia III Aplousobranchiata Lahille: Didemnidae Giard. Aust. J. mar. Ereshw. Res. 13 (3): 265-334. 1966. Ascidians of northern Australia. Pap. Dep. Zool. Univ. Qd 2 (15): 279-304. 1972. The ascidians of South Australia II. Eastern Sector of the Great Australian Bight and Investigator Strait. Trans. R. Soc. S. Aust. 96 (4): 165-96. 1974. The evolution and distribution of Australian tropical Ascidiacea. In ‘Proc. Second International Coral Reef Symp.’ Vol. 1; 406-23. (Great Barrier Reef Committee: Brisbane). 1977. Algal supporting didemnid ascidians of the Great Barrier Reef. In ‘Proc. Third International Coral Reef Symp.’ Vol. 1, pp. 615—621. (University of Miami: Florida). Lafargue, F. and G. Duclaux, 1979. Premier example en Atlantigue tropical d’une association symbio- tique entre une ascidie didemnidae et une cyanophycee chroococale: Trididemnum cyan- ophorum nov. sp. et Synechocystis trididemni nov. sp. Ann. Inst. Oceanogr. Paris. 1979. Prochloran (Prochlorophyta) associated with the ascidian Trididemnum cyclops Michaelsen. Phycologia 18 (1): 77-82. Thinh, L.V. and Griffiths, D.J., 1977. Studies of the relationship between the ascidian Diplosoma virens and its associated microscopic algae. Aust. J. mar. Ereshw. Res. 28 : 673-8 1 . Thorne, S.W., Newcomb, E.H., and Osmond, C.B., 1977. Identification of chlorophyll b in extracts of prokaryotic algae by fluorescence spectroscopy. Proc. Natl Acad. Sci USA 74 (2): 575-578. Tokioka, T., 1942. Ascidians found on the Mangrove Trees in Iwayama Bay, Palao. Palao trop. biol. Stn Stud. 2 (3): 497-507. 1950. Ascidians from the Palao Is. I. Pubis Seto mar. biol. Lab. 1 (3): 115-50. 1954. Contributions to Japanese Ascidian Fauna VII. Invertebrate Fauna of the Intertidal zone of the Tokara Islands. VII. Ascidians. Pubis Seto mar. biol. Lab. 3 (2): 239-64. 1955. Ascidians from the Palao Is. II. Pubis Seto mar. biol. Lab. 5 {ly. 43-51. 1967. Pacific Tunicata of the United States National Museum. Bull. U.S. natn. Mus. 251 : 1-242. 1975. Contributions to the Japanese ascidian fauna XXVII. Some ascidians from Okinawa, with notes on a small collection from Hong Kong. Pubis Seto mar. biol. Lab. 22 (5); 323-341. Van Name, W.G., 1918. Ascidians from the Philippines and the adjacent waters. Washington Smithsonian Inst. Bull. U.S. natn. Mus. 100 (1): 49-174. Vasseur, P., 1970. Contributions a I’elude des ascidie de Madagascar (region de Tulear) HI. La faune ascidiologique des herbiers de phanerogames marine. Rec. Trav. Sta. mar. Endoume Fasc. hors serie suppl. 10 : 209-221. MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 1: Didemnum molle, spherical spicules with loose flat ended rays, 0 005-0-01 5 mm (scale 0-005 mm). 2: Didemnum viride, stellate spicules with few long pointed rays, 0-03-0-04 mm (scale 0-1 mm). 3a, b: Lissoclinum voeltzkowi, spicules stellate or spherical (with blunt ended rays), 0-02-0-04 mm (scale 0-05, 0-005 mm). 4: Lissoclinum voeltzkowi, living colonies (Green I.). KOTT: ALGAL-BEARING ASCIDIANS Plate 1 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 2 la, b: Ussoclinum bistratum, spicules spherical with tight blunt ended rays, 0 03-0 05 mm {scale 0 012 mm). 2: Ussoclinum patellum, spicules spherical with tight needle-like rays 0 01-0 08 mm (scale 0-005 mm). 3: Ussoclinum punctatum. spicules spherical, with slightly irregular needle-like rays, 0-0 1-0-03 mm (scale 0-005 mm). 4: Echinoclinum triangulum, spicules flat, triangular, square or trajjezoid with loose needle-like rays of varying length, 0-03 to 0-08 mm (scale 0-005 mm). KOTT: ALGAL-BEARING ASCIDIANS Plate 2 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 3 la, b: Trididemnum clinides, stellate spicules with rounded or conical rays, 0 03-0 04 mm (scale 0 005 mm). 2: Trididemnum miniatum, spherical spicules with loose flat ended rays, 0*01-0 02 mm (scale 0 005 mm). 3: Trididemnum strigosum, stellate spicules with few conspicuously projecting rays, 0-05-0 08 mm (scale 0-005 mm). 4: Trididemnum nubilum, stellate spicules with numerous short conical rays projecting, 0-03-0-05 (scale 0-01 mm). KOTT: ALGAL-BEARING ASCIDIANS Plate 3 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 4 la, b: Trididemnum cyclops, stellate spicules with conical or flat ended rays, 0 04-0 06 mm (scales 0 05 mm, 0 005 mm). 2a, b: Trididemnum paracyclops, almost spherical spicules with projecting conical rays, 0 03-0 08 mm (scale 0 01 1 mm). KOTT: ALGAL-BEARING ASCIDIANS Plate 4 Mem. Qd Mus. 2{i{\): 49-63 [1980]. A NEW FAMILY OF ENSIFEROUS ORTHOPTERA FROM THE COASTAL SANDS OF SOUTHEAST QUEENSLAND D.C.F. Rentz CSIRO, Division of Entomology, Canberra ABSTRACT A new family, Cooloolidae, of Orthoptera is proposed, based on a single new genus and species Cooloola propator Rentz, from southeast Queensland, Australia. The cricket-like insect, referred to as the Cooloola monster, is an aberrant member of the suborder Ensifera. It is unusual in possessing very short antennae which, in turn, have a reduced number of articles, 10 in number. Nearly all species in all families of Ensifera have antennae with 30 or more articles. The structure of the mouthparts is unique in the Orthoptera in that the lacinia of the maxilla is larger and more developed than the mandibles. The lacinia is knife-shaped and not hooked. This suggests a piercing rather than the usual sorting and cleaning function for that structure. The mandible is elongate, indicative of a predaceous habit as opposed to short, stout mandibles which are possessed by plant feeders. C. propator has no close relatives but superficially resembles several species of the stenopelmatid Oryctopus. This genus is known from south India and Ceylon. Both genera share modification for a burrowing habit as illustrated by shortened antennae and legs, excavate, trowel-like modified spines, pale coloration, reduced eyes, and brachypterous males and apterous or micropterous females. The tarsal claw of both genera is sexually dimorphic. In females it is greatly reduced but normal in males. This may indicate a more sedentary role for females. Detailed examination of mouthparts, alimentary tract, and wing venation reveal that Oryctopus possesses all of the characteristic structures of the Stenopelmatidae, whereas Cooloola does not. The convergence of characters is interpreted as adaptation to a fossorial existence. Cooloola propator lives in sandy, moist soils and is not particularly associated with a single vegetation type. In rainforest and open forest, the soil is densely laced with roots. There is little accumulated litter on the surface and the soil is devoid of humus. The sand is usually bare and moist. Wandering adult males have been collected at night and their above-ground activity seems to be associated with wet weather. They appear to be active after periods of rain. Females are apparently completely subterranean. Every taxonomist is at first sceptical, and, hopefully, subsequently convinced of the reality of a new family of organisms. In an insect order such as the Orthoptera (grasshoppers, crickets, and allies), such an occurrence is extremely rare. This order is conservatively structured at the family level, despite recent escapades of several authors into the field of taxonomic escalation (see Kevan, 1977, for review). The last time a family was proposed based on an entirely new taxon was when E.M. Walker (1914) described the Grylloblattidae as a new family of Orthoptera. This paper describes a new family of cricket-like insects, the only known representative of which is known popularly as the ‘Cooloola monster’. The history of the discovery is traced, and various structures are compared with those in other Orthoptera. The new family Cooloolidae is placed in the superfamily Gryllacridoidea based on its similarities with Oryctopus of the Stenopelma- tidae. Although Cooloola Rentz gen. nov. is highly specialized in many characters, it can be assigned to the Gryllacridoidea on the basis of its tarsal formula and structure, lack of tegminal stridulatory structure, and general appearance. History of the Discovery of THE Cooloola Monster Shortly after arriving at my new post as Curator of Orthoptera, CSIRO, Canberra in 1977 I was presented with a small parcel from Mr E.C. Dahms, Curator of Insects, Queensland Museum, Brisbane with a note ‘Here’s something to introduce you to the Australian fauna’. After some amusement at the technical excellence of the 50 MEMOIRS OF THE QUEENSLAND MUSEUM Fig. 1 : Adult male of Cooloola propator. RENTZ: FAMILY OF ENSIFEROUS ORTHOPTERA 51 apparently manufactured monster, it was determined that it was a genuine complete cricket-like insect. The specimen had been taken by Dr V. Davies, Curator of Arachnids, Queensland Museum, in a pitfall trap set in mid-February 1976, in the rainforest of Cooloola National Park, near Gympie, Queensland. It was an adult male. A field trip to the area was planned and in the company of my wife, Mr E.C. Dahms, Mr l.D. Galloway and Dr G.B. Monteith, nearly two weeks were spent searching for the insect, but to no avail. All sorts of techniques, including digging, night searching, pitfall trapping and the oatmeal trail were used but no further specimens were located. Subsequently a popular article concerning the interesting discovery was prepared for the news media. The article received wide circulation and had a photograph of the insect, dubbed the ‘Cooloola monster’. The second specimen, a small nymph, was collected on nearby Fraser Island by Mr T. Bade, a ranger from the National Park, on 1 July 1978. He recognised the insect from details given in the article in ‘The Courier-Mail’, and rushed the live specimen to the Queensland Museum, from where it was forwarded to me in Canberra. Another trip was planned, at a different time of the year and including a few days on Fraser Island. The field work involved extensive digging and sieving soil, but the monumental efforts of those involved yielded no specimens. Shortly after conclusion of the trip, another specimen, a live adult male, was brought into the Queensland Museum by Mr R. Nicholas. He had been camping at Freshwater Lake camping area, Cooloola National Park, and the insect crawled into his tent at dusk after a rainstorm. He recognised it as the one featured in ‘The Courier-Mail’ some 4 months previously. The absence of females from the small samples was contrary to my experience with similar- looking Jerusalem crickets of the western United States. Adult males of these crickets are always difficult to obtain, and females and juveniles make up the bulk of collections (see Tinkham and Rentz, 1969, for discussion). The discovery of the fourth specimen, a female, possibly last instar nymph, ended conjecture. Its large, physogastric abdomen and short legs and tarsi indicate that the female is an obligate borrower and probably seldom walks on the surface of the ground. The specimen was found by Park Ranger Mr M. Johnston who dug it from a depth of 45 cm; the site was within 50 metres of where Mr R. Nicholas found his specimen. The fifth specimen, an adult male was collected by Ms K. Plowman in a pitfall trap in the vicinity of Lake Cooloola. The trap had been in the ground from mid October to mid November, 1978 when it was checked. Mr Paul Nicholas was preparing camp one evening in Aug. 1979 1 km south of Freshwater Lake in the vicinity where his father had previously found a male in Oct. 1978, and discovered a struggling cricket killed by the wheel of his automobile. It had been apparently brought to the surface by the wheels of the vehicle. The seventh specimen was found by Mr R. Perry, who was camping near Rainbow Gorge, Fraser Island in Oct. 1979. He found it while digging a shallow hole just above the beachline. Gutter traps based on 3 m lengths of roof guttering set in the ground so as to drain into buried buckets of formalin were designed by Dr G.B. Monteith and the author and constructed by CSIRO Division of Entomology. These were installed at the Freshwater Lake camping ground on February 4, 1979. They were checked in early May, 1979, without success, but when they were terminated on January 28, 1980, one was found to contain 5 adult males of the new insect. It is significant that a full year’s trapping of insects moving on the ground surface yielded neither females nor juveniles. Mr E. Zillman, of Gin Gin, Queensland, who was formerly with the National Park service on Fraser Island, submitted photographs of an adult female he had found on Fraser Island in June 1978. The specimen was uncovered at a depth of 22 cm around the roots of a brush box tree {Tristania sp.). He has seen the Cooloola monster on more than one occasion but was not aware of its importance until recently. He found one female under Banksia logs in sandy soil at a depth of 30 cm in a compost heap “some years before’. He is also convinced he dug one at Walla Lagoon, near Bundaberg, Queensland at 38 cm in 1952. Mr Zillman notes that his specimens were all found in sandy, and shaded soil, rich in surface roots. He considers the monster ‘fairly common’. Ecology Although ecological information is based on only 12 specimens, we can draw a number of conclusions regarding its habits and habitat preferences. The living specimens were observed to have a radically different gait from what one expects of a cricket-like insect. The actions of the 52 MEMOIRS OF THE QUEENSLAND MUSEUM 1.0cm 1 Fig. 2: Female of C. propator, subadult, see text. RENTZ: FAMILY OF ENSI FERGUS ORTHOPTERA 53 monsters have been described as ‘stomping’ or tractor-like. The monster’s movements are those of a fossorial creature out of its habitat. Males apparently venture forth, perhaps during rainy periods, in search of females. Otherwise they are subterranean. All specimens to date have been found within 45 cm of the surface, none deeper, although considerable digging to 3 m has been done. Mr Zillman’s notes indicate the species may be associated with buried rotting wood. Southern coastal Queensland from Coolangatta (just south of Brisbane) to Bundaberg and adjacent sandy islands, has been the subject of intensive continuous investigation by CSIRO Division of Soils and others. The area includes rainforests, sand dunes, and coastal heath habitats. It is well known locally because of the coloured ‘rainbow sands’. A list of the twelve classes of landforms in the area was presented by Thompson (1975). Watson and Arthington (1978) noted the restricted distribution of two dragonflies to acid dune lakes on Fraser Island and several other islands to the south. Ingram and Corben (1975) reported on four species of ‘acid frogs’ which breed only in water of low pH in wallum area of southeastern Queensland. These areas are within the range of the Cooloola monster. The wallum areas have apparently acted as islands during dry interpluvial periods. The discovery of Cooloola propator n. sp. in this rather limited but unique area is another indication that the habitat is peculiar and deserves protection. The first Cooloola monster was collected from near Poona Lake, a habitat which corresponds to Thompson’s Class 5, high transgressive dunes. This type of dune formation with elevations from 100-200 m is found at Cooloola, Fraser Island, and Moreton Island to the south. The dunes are formed from windblown sands from the adjacent beach. The high transgressive dunes support open forest and rainforest. The rainforest is restricted to bottoms and slopes of closed corridors, that is, dry valley bottoms. The other 1 1 specimens were found in Thompson’s Class 4, yellow-brown transgressive dunes. These range from 30 to 60 m above sea level and have been formed from sands blown off the beaches relatively recently. This zone is relatively rich in minerals and supports moderately dense vegetation. The Cooloola monster has been found in both rainforest and open forest with 11 of the 12 specimens coming from the latter. On Fraser Island it has not yet been found in rainforest. In areas where the insect has been found, the soil, to a considerable depth, is interlaced with tree roots. The sandy substrate is moist and there is little in the way of accumulated litter or humus on the surface. Large earthworms were not uncommon and their burrows extend to consider- able depths. Other organisms were seldom encountered. Family COOLOOLIDAE, new family Type Genus: Cooloola, new genus (Feminine gender) Characterisation: The family Cooloolidae is unique among the families of Orthoptera in the following combination of characters. Antennae reduced in length and number of articles. Mandibles reduced in size relative to maxillae; in structure designed for grasping or tearing, not chewing; lacinia lanceolate, without hooks, designed for piercing. Foregut extraordinarily long, extending nearly to apex of abdomen; proventriculus not hightly modified, similar to the Haglidae. Pro-, meso-, and metapleura dorso-ventrally expanded, presenting a highly ‘muscular’ appearance. Legs highly modified, femoral shape and spination of tibiae sexually dimorphic, fore femur greatly expanded dorso-ventrally, in length subequal to hind femur (more nearly equal in female); fore and hind legs (excluding tarsi) of nearly equal length; middle legs less modified, more slender. Tarsi very elongate, sexually dimorphic; in males very elongate, in females proportionately shorter; male tarsi weakly dorso-ventrally flattened, less so in female; male tarsal claws normal, in female reduced, minute, laterally compressed; all tarsi 4-segmeted. Tegmina present only on males, females apterous; tegmen highly sclerotized, convex, its caudal apex moulded around metathorax; tegmina separated dorsally, appearing to be capable of movement (perhaps used in burrowing); wing represented as minute, sclerotized bud. 54 MEMOIRS OF THE QUEENSLAND MUSEUM Abdomen not highly modified in male, exceedingly bulbous in female; male cercus not suited for grasping, paraprocts platyform, each with dorsal hook; subgenital plate bearing styli. Neither sex with any kind of stridulatory apparatus. Ovipositor short, 6-valved, not extending much beyond apex of abdomen; female subgenital plate very short, with digitiform caudal projection lying between ovipositor valves. Fig. 3: Map of area inhabited by C. propator. See text for discussion. Cooloola Rentz, new genus Type Species: Cooloola propator Rentz, new species, by monotypy. Description Head large, not seated in pronotum, cordate in frontal outline. Eye situated high on head, rather shallow, prominent (male) or considerably reduced (female). Labrum longer than broad; mandible obscured by lacinia, only its base visible anteriorly; maxillary palpus elongate, 5- segmented. Antenna short, with 10 articles. Pronotum broader than long, caudal margin not produced not hoodlike, only covering base of tegmen; cephalic and caudal margins both truncate to weakly obtuse, lateral lobes shallow, ventral margin straight; surface of disk relatively smooth, with two oblique lateral impressions, indicated mostly by colour. Prosternum armed with a pair of quadriform plates, divided mesad, then preceded by a pair of lateral sclerites, in turn, preceded by two pairs of linearly arranged setaceous unsclerotized lobiform processes, the more caudal pair more developed and more setaceous; mesosternum with a similar pair of plate-like processes, somewhat larger; metaster- num with massive plate-like processes, caudal internal angles acutely produced. Fore coxa enormously enlarged, commencing ventrad of lateral pronotal lobe; middle coxa about half size of fore coxa; hind coxa massive. Fore and middle trochanters small, elongate; hind trochanter larger, quadrate. Fore femur short, laterally flattened, dorso-ventrally expanded, and more emphasized in female. Fore tibia only slightly longer than femur, feebly laterally compressed, not dorso-ventrally flattened; ventral surface with a single spine positioned in the middle, subapically between the two internal spurs; apex with 2 spurs on anterior margin; the more dorsal of which is short, blunt, excavate, ventral spur more aciculate, positioned ventrad of metatarsus; posterior margin with a single, subapical spine and with a pair of much longer spurs, the more dorsad of which is expanded on its dorsal surface, the more ventral more aciculate, its apex directed mesad; dorsal surface unarmed except for minute, short, stout spine positioned on external margin, this spine highly variable in size even on the same individual. Middle tibia unarmed dorsally, elongate, slender in male, short, stout, dorso-ventrally flattened in female; ventral surface with a single aciculate subapical spine positioned ventral to metatarsus; RENTZ: FAMILY OF ENSIFEROUS ORTHOPTERA 55 apex bearing 6 spurs; 3 on anterior and 3 on posterior margin. Hind tibia unarmed dorsally and ventrally; apex with 6 spurs, sexually dimorphic; (male) with 3 on internal, 3 on external margin, most dorsal of internal spurs the largest, spatulate, its internal surface minutely striate, remaining spurs smaller in size, more elongate; (female) all spurs short, broad, apically excavate, surfaces striate, dorso-internal spur, in outline, ovoid, remaining spurs short, stout. Tarsi (male) tarsi elongate, claw well developed; (female) tarsi short. Prothoracic auditory structure absent; metathorax and 6 abdominal segments with well-defined spiracles. Internal surface of hind femur and adjacent portion of abdomen without stridulatory pegs. Dorsal surface of abdomen without median Carina. Male terminalia; tenth tergite with a pair of small widely separated hooks; supra-anal plate simple; cercus tubular, apex blunt, not serving a grasping function; paraprocts developed, with apical hook. Subgenital plate bearing styli; no sclerotized concealed genitalia. Female terminalia: similar to male except tenth tergite without hooks, cercus somewhat shorter, more robust, subgenital plate short, supra-anal plate unmodified. Oviposi- tor short, scarcely projecting beyond abdominal apex, 6-valved. External margins of pronotum, legs, sterum and external ventral margins of abdomen setaceous. Cooloola propator*, new species (Figs. 1-14) Material Examined FIolotype: MALE (pinned), Queensland, Cooloola National Park, Rainbow Beach, near Poona Lake, February 1976, V. Davies collector, In pitfall trap, QM T8309. Description Holotype: Internal surface of hind femur and adjacent portion of abdomen without stridulatory pegs; tenth tergite basically unmodified, median portion weakly concave, lateral portions feebly produced; supra-anal plate triangular, minute; paraprocts projecting somewhat from abdomen, not appearing highly mobile; subgenital plate quadrate, fairly short; median portion without incision. Female: Differs from male in following characters: size larger, form much more robust. Fore, middle, and hind tibiae proportionately stouter, more depressed, thickened; fore tibia rugulose dorsally on external margin at apex; middle tibia very short, 1-5 times as long as greatest width. All tibial spines more robust, shorter, more blunt, trowel-shaped. Cercus tubular, apically blunt; supra-anal plate trian- gular. Ovipositor with dorsal valves aciculate, apex highly sclerotized, extending slightly beyond abdomen; internal valves short, stout, obtuse; ventral valves short, broadly triangular, complete- ly concealing internal valves. Colouration: General colour tawny butterscotch brown; nymphs and female pale whitish, with following darker areas; eyes black except dorso-internal angles whitish; tegmen dark brown, veins somewhat darker; dorsal surface of pronotum with irregular brownish patches; outer pagina of hind femur with faint whitish herringbone pattern; apices of all spines and spurs dark brown or blackish; ventral surface of entire insect pale, thoracic portions darker. Nymphs and female (subadult?) almost white except for eyes and spines. Paratypes: IcC, Queensland, Cooloola National Park, 7 km NE of Lake Cooloola, mid-Oct. - mid-Nov. 1978, K. Plowman, in pitfall trap QM T8310. id". Freshwater Lake, 25 Oct. 1978, R. Nicholas, ANIC collection, Canberra. I 9 , Freshwater Lake camping area, 30 Oct. 1978, M.R. Johnston, QM T8311. 1 nymph, 25° lO’S 153°17’E, 19 km SSW of Indian Head, Fraser Island, National Park headquarters. Camp A July 1978, T. Bade, QM T8312. 1 9, 1 km S of Freshwater Lake Camping area, 4 Aug. 1979, P. Nicholas, QM T8313. id", I 6 km N of Rainbow Gorge, Fraser Island, 5 Oct. 1979, R. Perry, QM T8319. 5d"d", Freshwater Camp Ground, Cooloola Nat. Park SE. Qld., 4.V.1979 - 29.i.l980, G.B. Monteith, ex guttertrap, open forest, 2 in QM (T8389 & T8390) and 3 in ANIC. Relationships The peculiar combination of characters exhibited by C. propator makes it difficult to determine its relationships. The presently accepted conservative classification of the Orthoptera (see Key, 1970, for example) recognizes two suborders. This division is based primarily on the number of antennal segments, 30 or more = Ensifera, i.e. katydids and crickets; 30 or fewer = Caelifera, i.e. grasshoppers, pygmy mole crickets, cylindrache- tids. Were C. propator to have antennae with more than 30 articles, it would unquestionably be placed in the Ensifera. Its habitus is not unlike that of several of the families included therein. And the structure of the ovipositor, also of critical importance, seems within the range found in the Ensifera. I am considering the Cooloolidae as very aberrant ensiferans in the Gryllacridoidea. *Denoting first of a kind. 56 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1; Measurements (in mm). Maximum Dimensions of each Appendage holotype paratype (male) paratype (male) paratype (female) Length body 23.2* 29.0 27.0 30.0 Head: depth 8.7 9.3 8.1 11.5 width 6.5 7.0 7.0 7.8 Length fore leg: femur 8.1 7.3 8.1 7.5 tibia 7.5 7.6 7.0 6.5 tarsus 8.8 7.2 X 4.0 Length middle leg: femur 6.5 7.3 6.5 7.0 tibia 6.0 5.8 5.4 4.7 tarsus 8.7 7.4 7.5 5.1 Length hind leg: femur 11.0 11.3 10.5 10.0 tibia 9.0 8.2 8.0 7.0 tarsus 9.8 9.8 9.0 6.8 Pronotum: length 5.2 6.2 5.5 6.5 width 7.5 7.7 7.7 7.1 Length lateral lobe 4.4 4.7 4.2 6.4 Length tegmen 6.4 7.1 5.3 — •Length of body of holotype reduced because specimen is pinned, and abdomen shrunken. All others in alcohol. Table 2 presents characters of evolutionary importance in the more closely related ensiferan families (the Grylloidea are excluded since they possess 3-segmented tarsi and a 4-valved ovipositor. These primary characters are not borne by any of the examples in Table 2. Similarly, the Tettigoniidae are excluded because most members possess a tegminal file-scraper apparatus and tibial auditory apparatus, characters, in combina- tion, not shared by any of the listed families). The data in Table 2 indicate that there is little similarity between the Cooloolidae and either the Gryllacrididae or the Rhaphidophoridae. I have had the opportunity to study the male type of the type species of Oryctopus, O. bolivari Brunner (Paris Museum) and the types of the following species of the genus: O. prodigiosus Bolivar, O. bouvieri Karny, and O. lagenipes Karny all in the collection of the Paris Museum. An additional subadult female apparently taken with the type of O. prodigiosus was sent to me from the Vienna Museum. The above species were described from the Madras area of southern India. An additional species, O. sordellii Griffini, not seen by me, is known from Ceylon, Not all of these species are generically related. For example, O. lagenipes was described without locality data. It is a late instar nymph of some other genus, perhaps the American Stenopelmatus. It was reported to have single-segmented antennae, but the type clearly shows this is the result of breakage. Additionally, O. lagenipes has mandibles typical of most stenopelmatids. The species is definitely misplaced in Oryctopus. The female of O. prodigiosus was reported to be antenna-less, but my examination of the subadult female indicates that they have been broken at the bases. With the exception of O. lagenipes. the above-mentioned Oryctopus species show remark- able similarity in a number of characters to Cooloola propator. Both genera are adapted for burrowing. The legs are shortened and muscular, the spines are excavate, but in Oryctopus they are sharply pointed. Both genera are very pale in colour. Adult males are brachypterous or micropterous, females are apterous. The antennae are reduced in length and segment number, and although some of the types of Oryctopus have broken antennae, those that are entire are clearly reduced in size and segment number. The eyes of both genera are reduced, and in both this is sexually dimorphic with females showing greater reduction in the number of eye facets and in overall size of the eye. The tarsal claw of both genera is greatly reduced in females but is normal in males. On external appearance the two genera are remarkably similar (compare Bolivar’s 1899 figs. 21, 21a, 21b, and Karny 1932, figs. 25, 26 with those presented here). Detailed closer examination reveals a wide dissimilarity between the two genera. The tarsi of Oryctopus species are typically stenopelmatid. They are distinctly laterally compressed and the metatarsus is very elongate, longer than the remaining tarsal articles combined. In Cooloola the tarsi are depressed and not laterally compressed. The metatarsus is short. The mouthparts of the two genera are extremely different. As normal in the Orthoptera, the mandibles of Oryctopus are larger and more prominent than the lacinia of the maxilla. The reverse is true in Cooloola. In the latter, the mandible is not lanceolate (see Fig. 10), whereas in O. prodigiosus the mandible is lanceolate, without any dentition, very similar in shape to the lacinia. This cutting type of mandible in RENTZ: FAMILY OF ENSIFEROUS ORTHOPTERA 57 Oryctopus appears unique in the Orthoptera. At the base of the lacinia in Oryctopus, several stout bristles occur; these are absent in Cooloola. The tremendous expansion of the thoracic region reflected in Cooloola (Fig. 1) has not occurred in Oryctopus. In this respect, Oryctopus is similar to such burrowing genera as Stenopelmatus where the base of the legs is inserted just under the lower margin of the lateral lobe of the pronotum. In Cooloola (Fig. 1), the thoracic region has expanded dorso-ventrally greatly separating the base of the legs from the ventral margin of the lateral lobe of the pronotum. The sternal plates described for C. propator are wholly absent from Oryctopus. The male terminalia are quite dissimilar. In the male of O. prodigiosus before me, the subgenital plate is reduced, its apex acute and without styli. The cerci are erect and conical and apically digitiform. 2.0mm 2.0mm 2.0mm 7 Fig. 4-7; Diagnostic structures in C. propator. 4, dorsal view, male terminalia; 5, caudal view, female terminalia; 6, frontal view, head of paratypic male: note protruding lacinia; 7, internal view, apical spurs left tibia, female: note striations. 58 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 2; Characters thought TO be of Evolutionary Importance jn Several Ensiferan Families. CHARACTER GRYLLACRIDIDAE RHAPHIDOPHORIDAE STENOPELMATIDAE COOLOOLIDAE Antennal sockets widely separated nearly touching widely separated widely separated Fastigium of vertex flat or convex sulcate or strongly compressed convex or sulcate flat, not sulcate Lacinia with 2 hook-like processes with 3 hook-like processes with 2 spiniform processes 1 minute without any appendages Tarsi depressed, tobate compressed, sclero- tized, often with a minute pulvillus apically compressed, pulvillate compressed, pulvillate Metatarsi depressed, pulvillate compressed, sclerotized compressed with 2 pulvilli feebly compressed, with 2 pulvilli Tibial auditory tympanum absent absent present or absent absent Wings apterous or aiate apterous great majority of species apterous sexually dimorphic, males brachypterous Sclerotization of tegmina tegmina and wings soft, flexible — tegmina toughened tegmina toughened Femoro-abdominal stridulatory apparatus present absent present absent Inserting angle of hind femur present absent present present, but minute basicostale basisubcostale basirodioie Anal bai ? ScA or ScP ? ScA RS MA CuA AA]^ on AA2 AP, 2 mm ? Sc or C ScA AP2 d CuP Fig. 8: Right tegmen of C propator. Nomenclature after Kululova-Peck (1978), RENTZ; FAMILY OF ENSIFEROUS ORTHOPTERA 59 The apex of the abdomen and the paraprocts are not modified but there are 2 pairs of digitiform hook-like, lightly sclerotized appendages on the genital orifice. These appendages are absent from C. propat or. In summary, it appears that Oryctopus is only superficially similar to Cooloola, the similarities being apparently associated with adaptation to a wholly fossorial habit. Since Oryctopus seems stenopelmatid in all but a few characters (most notably the reduced number of antennal articles and the piercing or cutting mouthparts) I feel that it should remain in the Stenopelmatidae, with the reservation that examination of additional material might lead to a change of its placement but that this would not affect the classification of the Cooloolidae. Comparative Morphology of Selected Structures in the Cooloolidae Wing Venation: The brachyptery and modification of the tegmen render identification of the veins difficult. A drawing of the tegmen was examined by Dr J. Kukalova-Peck who stated (in litt.) that she found it difficult to trace the homologies of the veins because of the sclerotization at the base of the tegmen. As a result, the origin of several of the veins is obscure. Some of the problems are as follows: it is not possible to determine with certainty whether ?C or Sc starts from the basicostale (and is therefore a part of the costal system) or from the basisubcostale (and is therefore part of the subcostal system); similarly with ?ScA or ScP the three branches may, in fact, be all ScA. In Figure 8 the veins in doubt are labelled with a query. The venation of C. propator does not reflect relationship with any of the Haglidae, e.g. Cyphoderris monstrosa (see Sharov, 1971, fig. 25). Dr Kukalova-Peck feels that the venational characters are quite different and C. propator shows more primitive characters in the subcostal system but, conversely, reflects more advanced specialization in the anal field. The strong anal bar which is not present in the Haglidae, indicates that this area was more strongly developed in the past for some purpose, perhaps flight, burrowing, or courtship, but I am not now able to determine it. In sum, it can be stated that the venation of C. propator is quite distinct from that of any of the presently known orthopteroids and may indicate that the species is the sole survivor of a very archaic line of hitherto unknown orthopteroids. Mouthparts: Mouthparts in Orthoptera, though variable, show adaptive modification based on the food preferences of the groups involved. Such was demonstrated by Gangwere (1965) for the North American Orthoptera. The mandibles of each subfamily, and often in groups below that level, are usually distinctive. Of all the mouthparts, the labium appears to remain relatively constant throughout the Orthoptera. The maxillae and labrum show a lesser degree of modification with the maxillae showing more substantial differences than the labrum (contra Isely, 1944, who stated that only mandibles show significant adaptive modification). Because of the conservative nature of mouth- parts, they can serve as a character for taxonomic differentiation. Gangwere (1965) was able to show that the tettigoniid subfamilies Conocephalinae, Copiphorinae, and Decticinae were mqre similar in the morphology of mouthparts, than to the Phaneropterinae of Pseudophyllinae. Such is an accepted fact based on more usual morphological characters. Gangwere (1965) listed six kinds of mandibles for the Orthoptera. According to his scheme, C. propator fits the Carnivorous- Forbivorous or Flesh-Forb feeding subtype. The mandible is elongate and hook-like. The incisor and molar dentes are well defined and sharp, with the latter surrounding a minute but distinct concavity. However, his name of the subtype cannot be applied to C. propator since it appears unlikely that this insect feeds upon forbs. It is more probable that it feeds upon roots or organisms encountered underground. Other burrowing Orthoptera, such as mole crickets, have mandibles which are short and stout. They have well-defined molar regions designed for chewing rather than holding or tearing and they are classed in the Omnivorous category. The mouthparts of C. propator are unique among the Ensifera in that the maxillae are larger and more prominent than the mandibles. The galea is small and slender and, as usual in Ensifera, lies above the lacinia. It is shorter than the lacinia, not apically modified and bears two rows of setae dorsally. This is in contrast to the Stenopelmatidae, Rhaphidophoridae and Grylla- crididae. In the first the galea is usually short and broad (see previous discussion for Oryctopus as exception). In the rhaphidophorids and the gryllacridids the galea is massive and its lateral and apical margins overhang the lacinia. The galea is a robust, quadrate structure and in the gryllacridids it is apically modified into a 60 MEMOIRS OF THE QUEENSLAND MUSEUM sclerotized papillate sensory organ as is found in the Tettigoniidae. Laterally the galea possesses well-defined, although irregular, rows of setae. The peculiar knife-like structure of the lacinia of C. propator (Fig. 9) is very different from that of the three related families. The lacinia of C. propator is unique in that it is modified to form a simple, broad, knife-like structure and bears no appendages. The dorsal surface has two minute tubercles at its broadest point, on the internal margin. Lateral to the tubercles lies a pair of pits, the internal pit is large and circular, the external minute. The apex of the lacinia is a smoothly bent cutting edge without hooks or teeth. Apically it is Fig. 9-11: Mouthparts, male paratype C. propator. 9, left maxilla; note blade-like lacinia; 10, left mandible: note elongate shape for grasping or tearing; 1 1, labium. RENTZ: FAMILY OF ENSIFEROUS ORTHOPTERA 61 distinctly spatulate. In length it is half as long again as the mandible. In the stenopelmatids the lacinia is dorsally tuberculate on the internal margin and bears at least two large tubercles on its dorsal surface. There is an elongate spiniform tooth on the internal margin at the base of the apical portion. A tiny spiniform tooth is found on the internal cutting edge of the lacinia itself. In the rhaphidophorids, the lacinia bears no tubercles and has three apical teeth. It shows no modification for cutting, but appears to have a tearing, or perhaps a sorting, function. The gryllacridids have an elongate non-tuberculate lacinia bearing two apical teeth on the internal margin. Near the proximal tooth lies an inconspicuous, short, stout spine. The labrum and labium (Fig. 1 1) appear similar to those of other Orthoptera. In general, the mouthparts of the Cooloola Monster suggest adaptation for piercing and chewing rather than chewing only, as is normal for Orthoptera. The reduced nature of the molar area (Fig. 10) and extraordinary development of the lacinia relative to the mandible supports this hypothesis. The structure of the gut also supports it. The enlarged foregut (see below) may be a modification for storage of liquid food and may explain the absence of solid food from the foregut in field-collected specimens. The reduced armature of the proventriculus may reflect its lessened function as a sieving mechanism. One can only speculate that the insects feed on insect larvae or, perhaps earthworms which are abundant in the habitat, or may feed on the toughened roots of trees and shrubs which lace the habitat. The Alimentary Tract: The alimentary canal was removed from one of the fluid-preserved specimens and studied as described by Judd (1948). For the sake of comparison, the terminology of structures follows that of Judd and his code is used in my illustrations. The alimentary canal of C. propator is peculiar in that the crop is extremely long and convoluted. The proventriculus lies in the posterior part of the abdomen, thereby by necessity, severely limiting the length of the midgut and hindgut. Although this is common in gryllacridids, stenopelmatids and Cyphoderris (Haglidae), it is not normal in tettigoniids or rhaphidophorids. In those families, the proven- triculus lies in the thoracic region. Strangely, these differences were not registered by Judd. Perhaps some stenopelmatids, the cooloolids and Cyphoderris are sporadic and opportunistic feeders, feeding only occasionally when the situation warrants it, and then consuming large quantities of food. Evidence suggests that sporadic feeding may depend on local weather conditions. Many of these insects are extremely sensitive to dry conditions and all are nocturnal. They may not venture forth each night unless atmospheric conditions are suitable. Such an explanation was offered for several species of the henicine stenopelmatid Cnemotettix of California which spend periods of dry weather in their burrows (see Rentz and Weissman 1973). At the same time Rentz and Weissman noted the extraordinarily large size of the faecal pellets of Cnemotettix. Curiously, however, I have found that the faecal pellets of certain other stenopelmatids and gryllacridids are very large. The same was also mentioned by Richards (1973, p.226) in the deinacridine stenopelmatid, Hemideina spp., the giant weta of New Zealand. Comparative studies of the proventriculi of several examples each of the Tettigoniidae, Rhaphidophoridae, Stenopelmatidae, Grylla- crididae, Haglidae {Cyphoderris) and Grylloidea indicate that there is considerable overlap in the morphology of this structure from one group to the other. It appears that the taxonomic value of the proventriculus may lie at the generic level. However, a few generalisations can be drawn. Based on the characters of the proventriculus, the Cooloola monster shows no similarities with any of the Grylloidea. In this group the median tooth normally has lateral projections and two lateral lobes. This is a more complex picture than exists in the Tettigonioidea and Gryllacridoidea. In these groups the general structure is simpler, the lateral teeth and inner barbate lobes of the sclerotized appendage of the proventriculus are absent. It thus seems that, on the basis of this structure, Sharov (1971) was wrong to transfer the Haglidae from the Tettigonioidea to the Gryllidea (= Grylloidea). He did this primarily on wing venation. Fresh preparations of the proventriculi of a species of the two extant genera of the Haglidae show none of the typical grylloid characters. Judd (1948), in an extensive survey of the Orthoptera, made a point that the Rhapidophoridae were distinct in that the median tooth of the sclerotized appendage of the proventriculus always bore a tuft of hairs. He illustrated this with several genera. However, I found in the rhaphidophorid Macrobaenetes valgum (Strohecker), a more or less typical example of the family from the southwestern 62 MEMOIRS OF THE QUEENSLAND MUSEUM United States, there are no hairs at all. The lateral 14 is, indeed, astounding. The most obvious lobes, however, are lightly sclerotized as Judd differences between the proventriculus of indicated. He considered the proventriculi of the Cyphoderris and Cooloola are the following: Stenopelmatidae, Tettigoniidae, and Prophalan- proportionally smaller size of proventriculus in gopsidae (= Haglidae) indistinguishable from one Cooloola', much lesser degree of sclerotization; another at the family level. With this I concur. lack of distinct protuberances or modification of The most striking similarity of the proven- the cushions (Fig. 1 3) (C) of the sclerotized neck triculus of the Cooloola monster is with that of region in Cooloola', presence of tubular neck Cyphoderris monstrosus Uhler. Judd did not region in Cooloola (in Cyphoderris this is absent illustrate an entire section of one of the and the sclerotized portion of the neck is longitudinal folds of the proventriculus of this contiguous with the main portion); more angulate important insect, but if he had it would have projection of the caudal portion of the barbated looked much like Fig. 13. He did illustrate the lobe (BL) in more lightly sclerotized sclerotized appendage of the proventriculus (Judd loop of hairs (CT) in Cyphoderris. It can be 1948, figs. 73, 74) and the similarity with my Fig. concluded that in all aspects of the structure of the F[G. 12-14: Structures of foregut. Nomenclature after Judd (1948). 12, proventriculus, posterior end of crop and anterior end of midgut. SN = sclerotized neck, UN = tubular neck, P = proventriculus, GC = gastric caeca. 13, sclerotized portion of one longitudinal fold of proventriculus, CT = V-shaped loop of tubercules, C = cushion, BL = barbated lobe, MT = median tooth, LL = lateral lobe, CP = sclerotized partition, Oes.V. = oesophageal valve. 14, sclerotized appendage of proventriculus. LL = lateral lobe, MT = median tooth. RENTZ: FAMILY OF ENSIFEROUS ORTHOPTERA 63 proventriculus, Cooloola propator shows more similarities with Cyphoderris monstrosa than any other ensiferan. The proventriculus of C propator is 2-2 mm long, and globular, it is joined to the crop by a tubular neck which is a distinctly unsclerotized zone. Where joined to the midgut, it is surrounded by two moderately large broad gastric caecae (GC). Beyond the constricted neck there are 5 or 6 longitudinal rows {Fig. 13) of poorly defined cushions composed of hairs. Each row of cushions appears faintly divided down the middle and consists of 5 well-defined pairs preceded by 3-5 smaller more poorly defined ones. The anterior cushions bear fewer hairs and are narrower. There is no median projection from any of the cushions. Between each row of cushions there is a V-shaped loop (CT) composed of minute tubercules. Each of the 6 longitudinal folds of the main part of the proventriculus consists of 10-12 sclerotized appendages. The median 5-9 are the widest and best developed, those on either end decreasing in size. Each longitudinal fold is spearated by a weakly sclerotized partition (CP) at the end of which there is a fleshy flap of the oesophageal valve (Oes.V.) which is not clothed with hairs. Each median appendage (Fig. 14) has a median tooth (MT) with minute tubercles at the base. There are no lateral teeth. From the side of each tooth are rows of extremely fine setae. Lateral to the seta is a minute blunt lateral lobe (LL). Each lateral lobe is bounded by a larger more quadrate barbated lobe (BL) which bears a blunt tooth posteriorly. ACKNOWLEDGMENTS The author would like to first thank Mr E.C. Dahms of the Queensland Museum who directed his attention to the Cooloola monster and Dr G.B. Monteith, Queensland Museum, who subsequently has aided in the project. To all the National Park and Forestry Rangers and other interested individuals who have joined in the search, I convey my thanks. Dr A. Kallenbach, Vienna Museum and Dr M. Donskoff, Paris Museum lent the author types and critical material of Oryctopus species without which adequate comparisons could not have been made. Mr S.P. Kim is thanked for his excellent illustrations and Mr J. Balderson and Mrs B, Rentz are thanked for assistance in the field and with the manuscript. I have benefited from the comments of Drs K.H.L. Key and J.A.L. Watson on the draft of the manuscript and lastly I thank Mrs D. McShane who deciphered the handwritten draft. LITERATURE CITED Bolivar, L, 1899. Les Orthopteres de St. Joseph’s College a Trichinopoly (sud de I’lnde). Ann Soc. Entomol. Fr. 68: 761-812. Gangwere, S.K., 1965. The structural adaptation of mouthparts in Orthoptera and allies. Eos. 41: 67-85. Ingram, G.J., and Corben, C.J., 1975. The frog fauna of North Stradbroke Island, with comments on the ‘acid’ frogs of the Wallum. Proc. R. Soc. Qd 86(9): 49-54. ISELY, F.B., 1944. Correlation between mandibular morphology and food specificity in grasshoppers. Ann. Ent. Soc. Am., 37: 47-67. Judd, W.W., 1948. A comparative study of the proventriculus of orthopteroid insects with reference to its use in taxonomy. Can. J. Res. D26: 93-161. Karny, H.H., 1932. Die Gryllacrididen des Parise Museums und der Collection L, Chopard. Eos. 7: 235-309. Kevan, D.K. McE. (Ed.), 1977. The higher classifica- tion of the orthopteroid insects. XV. Int. Congr. Ent. Lyman Entomological Museum and Research Laboratory, Memoir, 4(12), pp. 1-26. Key, K.H.L., 1970. Chapter 21; Orthoptera (grasshop- pers, locusts, crickets). In The Insects of Australia. (CSIRO, Melbourne Univ. Press.) Kukulova-Peck, j., 1978. Origin and evolution of insect wings and their relation to metamorphosis, as documented by the fossil record. J. Morph. 156: 53-125. Rentz, D C., Weissman, D.B., 1973. The origins and affinities of the Orthoptera of the Channel Islands and adjacent mainland California. Part 1 . the genus Cnemotettix. Proc. Acad. nat. Sci. Philad. 125: 89-120. Richards, A.M., 1973. A comparative study of the biology of the giant wetas Deinacrida heteracantha and D. fallal (Orthoptera; Henicidae) from New Zealand. J. Zool. Lond. 169: 195-236. Sharov, A.G., 1968. Filogeniya Ortopteroidnykh Nasekomykh. Trud. paleont. Inst., Moskva, 118; 1-217, [English version, 1971, Phylogeny of the Orthoptera, Jerusalem, 251 pp.]. Thompson, C.H., 1975. Coastal areas of southern Queensland. Some land use conflicts. Proc. R. Soc. QdS6: 109-20. Tinkham, E.R. and Rentz, D.C., 1969. Notes on the bionomics and distribution of the genus Stenopelmatus in central California with the description of a new species. Pan-Pacific Entomol. 45:4-14. Walker, E.M., 1914. A new species of Orthoptera, forming a new genus and family. Can. Ent. 45: 93-9. Watson, J.A.L. and Arthington, A.H., 1978. A new species of Orthetrum Newman from dune lakes in eastern Australia. (Odonata; Libellulidae). J. Aust. ent. Soc. 17: 151-7. Mem.Qd Mus.lH{\): 65-75 pis. 1-3 [1980], PROCOELOUS CROCODILE FROM LOWER CRETACEOUS OF LIGHTNING RIDGE, N S W. R.E. Molnar Queensland Museum ABSTRACT Restudy of the type material and study of referred material of the Lower Cretaceous Crocodylus (Bottosaurus) selaslophensis indicates that the material is referable to neither Crocodylus nor Bottosaurus. Referred material includes procoelous cervical centra and suggests that the crocodile is an eusuchian. Distinguishing characters are a dentary with alveolar groove and medical shelf, procoelous cervicals, caudal transverse processes excavated posteriorly at their bases, and caudal neurocentral suture extending onto the transverse process. All known material derives from a small area of the Griman Creek Fm. at Lightning Ridge, New South Wales. Crocodilian fossils are common in the Quaternary and late Tertiary of Australia (e.g. De Vis 1885; Longman 1924, 1925; Molnar 1976; Archer and Wade 1978; Gorter and Nicoll 1978). Miocene crocodilians, still largely under study, are also common at some localities. Earlier crocodilians have only rarely been reported in Australia. Riek (1952) reported crocodilian skin impressions (QUF 10625) from the probably Eocene Redbank Plains Series. These impressions show roughly hexagonal scales with distinct growth lines and lacking keels, a combination of features not known to occur in crocodilians. The general pattern of the scales could not be matched with that of any living crocodilian from the Australasian region, and thus it is unlikely that these skin impressions derive from a crocodilian. Riek (1952) also reported an angular from the Redbank Plains Series, thus demonstrating that crocodilians were present. Etheridge (1917) described the earliest known Australian crocodile as Crocodylus (Bottosaurus) selaslophensis, from the Cretaceous opal-bearing beds at Lightning Ridge, New South Wales. While other crocodilian material from this site had already been obtained by the Australian Museum, Etheridge described only the type jaw fragment. Since 1917 further material has been discovered there, including procoelous cervical vertebrae. This material, like almost all found at the Ridge, was not in association, so it is possible that more than one taxon is involved. The duplicated left tibiae demonstrate that at least two individuals are represented. However all known material derives from individuals of approximately the same size, and among the latest discoveries I have seen teeth matching those in the type jaw indicating that elements possibly deriving from the same individual are still being found. In the absence of any indication that more than one taxon is represented, 1 shall assume that all pieces derive from a single species. All fossil bones known from Lightning Ridge are opalized, at least in part, and thus many of the specimens have been retained by their discoverers and are represented in museum collections by plaster or resin casts. Although Etheridge assigned the jaw fragment to Crocodylus (Bottosaurus) there is no indication in the literature that anyone else ever considered Bottosaurus a subgenus or a synonym of Crocodylus. Indeed Bottosaurus is usually classified (e.g. Steel 1973) as an alligatorine rather than a crocodyline. Collection Designations: AM — Aus- tralian Museum; QM — Queensland Museum; QUF — University of Queensland, Dept, of Geology. 66 MEMOIRS OF THE QUEENSLAND MUSEUM Geology of Lightning Ridge Nowhere at Lightning Ridge are there any extensive exposures of the opal-bearing beds. The geology of these beds has thus been interpreted entirely from features observed in the opal mines and the single large open cut and from the fossils collected from these mines. Earlier work on the stratigraphy of the Ridge (e.g. Whiting and Relph 1961) has been superseded by that of Byrnes (1977). He has concluded that the opal-bearing beds at the Ridge form the southern portion of the Griman Creek Formation which is more extensively exposed in southern Queensland. In New South Wales this formation is subdivided into two members, the Wallangulla Sandstone and the Coocoran Claystone. The opals (and fossils) occur in the Finch clay facies of the Wallangulla Sandstone Member. Byrnes (1977) concludes that the Wallangulla Sandstone was deposited in estuarine conditions. In addition to fossil bone, pelecypods and gastropods have been found. None of the pelecypods represent clearly marine taxa (Byrnes 1977), while some of the gastropods are members of the family Viviparidae, which are fresh water forms. The occurrence of the fossil bones is such as to suggest transport: only in one instance were any found in articulation. Most of the elements seen are such as can be easily transported, e.g. femora, phalanges, teeth, centra, with no flat bones or ribs. The largest elements identifiable were plesiosaurian. Some of the bones appear to have been worn or broken prior to fossilisation, and one plesiosaur element had been bored. Plant material, including conifer cones, is reasonably common. Byrnes (1977) reports that some cross-bedding is present, and what appear to be impressions of plant roots occur in the clay facies. All of this suggests an estuarine environment of deposition. Description Mandible: Etheridge figured and described a fragment of crocodilian dentary with six teeth in situ (AM FI 58 18) as the type of Crocodylus selaslophensis (Plate 1, E and G). The teeth are thecodont (as in all other known crocodilians) and not pleurodont as Etheridge stated: the bony partitions separating sequential alveoli are distinct. These partitions terminate below the level of the dorsal margin of the dentary so that the teeth are set in an alveolar groove. Slight anterior and posterior carinae are present on the crowns and striae are absent. A medial shelf extends along the tooth row, giving the dentary the appearance of being wider than in Crocodylus. The lateral surface of the dentary is sculptured with longitudinal ridges and grooves much as modern Crocodylus. A prominent mass of opal along the medial surface (Plate 1, E) represents either a displaced bone fragment or an opalised mass of matrix, and not a projection from the dentary. There is no indication that this piece is from the symphyseal region as reported by Etheridge. Maxilla: A small fragment (AM FI 8628) probably represents the anterior extremity of the right maxilla, with one tooth in place (Plate 1, F and H). Two other alveoli are present, one empty, the other occupied by only the root. The single crown is conical, medially flexed, and has fine striae but no carinae. The fragment as a whole is dorsoventrally compressed, suggesting a low, broad snout, and is lightly sculptured. CervicalS: Two cervical vertebrae have been seen, the more complete represented by a cast (QM F9507), and the other a centrum in the Anderson collection that is also represented by a cast (QM FI 0240). Both centra are clearly procoelous, the anterior faces deeply concave, and the posterior convexities rimmed by a flange as in modern crocodilians. The posterior convexity of the Anderson centrum (Plate 2) is incomplete, and the anterior central face is worn, but appears to be slightly inclined upwards. This centrum is 18-5 mm long as preserved. Both centra are constricted at the middle with marked ventral keels that descend anteriorly to form small, blunt hypophyses. Bases of both parapophyses and diapophyses are present on QM F9507 (Fig, 1), but absent from Anderson’s specimen. The centrum of QM F9507 is 16-4 mm long, 1 5 mm high at the anterior face and 1 3 mm wide across that face. In general form and proportions both centra resemble cervical centra of the living species of Crocodylus. Two portions of a cervical neural arch (AM F60081), although not sharing a contact, may derive from the same vertebra (Plate 2, D). Neurocentral articular facets, the base of the neural spine, the diapophyses, the left prezy- gapophysis and the base of the right are preserved. The arch is from a mid-cervical about 30 per cent larger than QM F9057. The neural spine is set anterior to the postzygapophyses, and both pre- and postzygapophyseal facets were more nearly horizontal than in Crocodylus americanus (Mook MOLNAR: PROCOELUS CROCODILE 67 1921, fig. 3). The arch resembles those of Crocodylus porosus in these features. The diapophyses are placed well above the neurocen- tral suture, as in the second cervical of C. porosus. Fig. 1: Crocodilian cervical centrum from Lightning Ridge. (Specimen lost, represented by cast QM F8507.) Bar represents 2 cm, A, anterior view; B, dorsal view; C, lateral view. Sacral: The single sacral centrum (AM FI 58 19) includes the lower portion of the neural arch with prezygapophyses, the entire left sacral rib and the base of the right (Plate 3). In general form this element greatly resembles the first sacrals of ‘Leidyosuchus’ multidentatus (Mook 1930, fig. 5) and of Diplocynodon hantoniensis (Owen 1884, there given as Crocodilus hastingsiae). The prezygapophyseal facets are more nearly horizontal in this sacral than in the first sacral of either of these taxa, being inclined at only 25® to the horizontal. The anterior face of the centrum is shallowly concave, broad (23 mm wide by 12 mm high), and slightly inclined downwards. The posterior face is incomplete, less broad, almost flat (with but a shallow central concavity) and is slightly inclined dorsally. The centrum is 24 mm long, as is the sacral rib which bears a posteriorly inclined iliac articular surface. A distinct ridge extends from the pre- zygapophysis laterally along the anterodorsal edge of the rib. The ventral surface of the centrum is faintly concave with slight longitudinal ridges along both sides. Caudal: The proximal caudal centrum (AM F60080) includes portions of both diapophyses and the right prezygapophysis (PI. 2), and is 28 mm long, 13 mm high and 12 mm wide at the anterior face. The central articular faces are inclined as in the single known sacral. The centrum is constricted at the middle, with two sharp ventral keels much like those of the caudals of Holopsisuchus brevispinus (Cope 1869, PI. IV, fig. 4). Both central faces are concave as preserved, but both are worn so it is not certain that the centrum was amphicoelous. The neurocentral suture was dorsal to the diapophyses and zygapophyses and extended laterally 12 mm onto the dorsal surface of the transverse process. A distinct pocket extends into the posterior margin of the transverse process just lateral to the postzygapophysis. Cervical Rib: The proximal portion of a cervical rib (AM F60082), from the left side, does not differ from those of Crocodylus porosus. Femur: An incomplete element possibly representing the distal one-third of a right femur is present in the collection of K. Barlow (Plate 1, A and B). The specimen is badly worn, but has less curvature in the shaft than femora of Crocodylus. 68 MEMOIRS OF THE QUEENSLAND MUSEUM Tibia: Two left tibiae are preserved, the distal end missing from both. The smaller (AM FI 8630) is less than 10 per cent smaller than the larger (AM FI 5821) (Plate 1, C and D). In both the cnemial crest is badly worn. These two tibiae are identical and closely resemble those of C. porosus, from which they differ only in that the lateral surface of the shaft is Oat with a sharper anterolateral border and a less marked poster- olateral border. Discussion Etheridge’s type specimen (AM FI 58 18), a dentary, is characterised by: 1) an alveolar groove and 2) a medial shelf. Assuming that the other crocodilian material is referable to the same taxon, it is further characterised by: 3) a low, probably broad snout; 4) procoelous cervicals, with a small hypophysis; 5) keeled caudal centra; 6) excavated caudal diapophyses; and 7) caudal neurocentral suture extended onto diapophysis. Postcranial elements resemble those of crocodylids in the forms of the cervical and sacral vertebrae, and of the tibia. This general resemblance, together with the procoelous character of the cervical centra, suggests that this crocodilian was eusuchian. The type jaw fragment of Crocodylus selaslophensis was compared by Etheridge with a jaw fragment referred to Bottosaurus harlani (most recently figured by Mook, 1925, fig. 8). The latter specimen is attributed to an immature individual. There is no clear evidence of immaturity of the Lightning Ridge crocodile, although the caudal neural arch has separated from the centrum at the neurocentral suture, as has the cervical neural arch (but not that of the sacral). Mature teeth of B. harlani are considerably more bulbous and blunt than in the Lightning Ridge form, and neither the type jaw of B. harlani nor the referred immature specimen exhibit an alveolar groove. Thus the Lightning Ridge crocodilian is not referable to Bottosaurus. The alveolar groove and the excavated caudal transverse processes suggest that this form is also not referable to Crocodylus. The genus Crocodylus is not otherwise known to range into the Lower Cretaceous (cf. Steel 1973), and the three reported Lower Cretaceous English species (C. cantabrigiensis, C. icenicus, and C. saulii) have not been reviewed since the nineteenth century and are of uncertain significance. C saulii may be congeneric with Bernissartia fagesii (Buffetaut, 1975). Other than 'Crocodylus’ selaslophensis. only two crocodilians exhibit an alveolar groove: Edentosuchus tienshanensis (Young, 1973) and Macelognathus vagans (March 1884, Ostrom 1971). Both are currently referred to the Mesosuchia, and both differ from the Lightning Ridge crocodile (and from each other) in several other characters; no close relationship is warranted. Procoelous vertebrae, considered characteristic of eusuchians, also occur in a mesosuchian (Joffe 1967), and reportedly in a sebecosuchian (Arid and Vizotto 1965). The mesosuchian is Theriosuchus, an atoposaurid. Atoposaurids have been suggested as possible ancestors of eusuchians (Joffe 1967; see also Langston 1973). However the known atoposaurids are all considerably smaller than the individuals represented at Lightning Ridge and all date from the Upper Jurassic. None are reported in the literature to exhibit excavated caudal diapophyses, an alveolar groove, or several of the other characters of the Lightning Ridge crocodile. In the absence of further evidence it may be assumed that the Lightning Ridge crocodilian is not a large atoposaurid. The sebecosuchian Baurusuchus also reportedly had procoelous vertebrae (Arid and Vizotto 1965). The type jaw and maxillary fragment from the Ridge show none of the dental specializations of sebecosuchians, so that reference to this group can also be ruled out. Unfortunately the vertebrae of Baurusuchus were not illustrated so that comparison of the vertebrae cannot presently be made. The occurrence of a procoelous crocodilian in the Lower Cretaceous (Aptian or Albian) of Australia is unexpected. Hylaeochampsa vectiana, usually recognised as the earliest known eusuchian (Romer 1966, Steel 1973), comes from the Lower Cretaceous of England. The Lower Cretaceous Bernissartia fagesii, often considered eusuchian (e.g. Charig 1967) has recently been demonstrated to be a mesosuchian (Buffetaut, 1975). Heterosuchus valdensis, also from the Lower Cretaceous of western Europe, is often considered congeneric with Hylaeochampsa (Romer 1966, Steel 1973). While Hylaeochampsa is known only from cranial material, Heterosuchus is known from postcranial material, including procoelous vertebrae. The three Lower Cretaceous species attributed to Crocodylus, as mentioned previously, need restudy. The described vertebrae of C. cantabrigiensis and C. icenicus are procoelous, and generally resemble those from Lightning MOLNAR: PROCOELUS CROCODILE 69 Ridge (Seeley 1874, 1976). All these forms are from western Europe. The earliest extra-European eusuchians {Aegyptosuchus, Stomatosuchus, and Stromerosuchus) appear in North Africa around the Cenomanian, although a procoelous vertebra has been found in the Albian of Algeria (Buffetaut pers. comm., 1979) and by the latest Cretaceous eusuchians were widespread and diverse. Thus the indication of a possible eusuchian in the Aptian or Albian of New South Wales suggests a considerably wider range of procoelous crocodilians during the Lower Cretaceous than has been generally recognised. The holotype of Etheridge’s Crocodylus (Bottosaurus) selaslophensis is too incomplete for confident comparison with other specimens. However it is sufficiently unique (see the character states listed on p. 1 36) that should more complete material be found, the taxon would be both recognisable and diagnosible. Further, the Griman Creek Formation has been so little explored for fossils that it is premature to relegate Etheridge’s species to the status of nomen vanum until it is clear that more complete topotype material is not forthcoming. Summary Procoelous cervical vertebrae from the Griman Creek Fm. (probably Albian) of Lightning Ridge, New South Wales, demonstrate the existence of a procoelous crocodilian in Australia during the Lower Cretaceous. Assuming that all the remains pertain to a single taxon, that taxon is characterised by procoelous cervicals, excavated caudal transverse processes and caudal neurocen- tral suture extending onto the transverse processes. This material probably pertains to Crocodylus selaslophensis Etheridge (1917). The type dentary fragment of that species has an alveolar groove and a medial shelf establishing that there is no reason to refer this species to the genus Crocodylus. The species is regarded as indeterminate pending the discovery of further material from the Griman Creek Fm. ACKNOWLEDGMENTS Dr Alex Ritchie provided access to the specimens in the Australian Museum, and Dr Mary Wade drew my attention to the material in her care at the (Queensland Museum. Their encouragement at the beginning of this study is greatly appreciated. Dr D. Baird, of Princeton University, kindly provided information on adult Bottosaurus. Mr G. Anderson and Mr K. Barlow kindly loaned specimens from their respective private collections for study, Mr R. Jones of the Australian Museum and Miss K. Brown of the School of Zoology, University of New South Wales, also provided material assistance for this study. The photographic illustrations were prepared with the assistance of the Department of Biomedical Illustration, Prince of Wales Hospital, and the photographic department of the Queensland Museum; the drawings are by Ms L. Crevola-Gillespie. LITERATURE CITED Archer, M. and Wade, M., 1976. Results of the Ray E. Lemley expeditions, part 1. The Aliingham Formation and a new Pliocene vertebrate fauna from northern Queensland. Mem. Qd Mus. 17 : 379-97. Arid, F.M., and Vizotto, L.D., 1965. Crocodilideos fosseis nas proximidades de Santa Adelia (SP). Cienc. Cultura 17 : 138-9. Buffetaut, E., 1975. Sur I’anatomie el la position systematique de Bernissartia fagesii Dollo, L., 1883, crocodilien du Wealdien de Bernissart, Belgique. Bull. Inst. r. Sci. nat. Belg., Sci. Terre. 51 : 1-20. Byrnes, J.G., 1977. Notes on the Rolling Downs Group in the Milparinka, White Cliffs and Angledool 1:250,000 sheet areas. Geol. Surv. New South Wales. Rept. GS1977/005: 1-17. Charig, A.J., 1967. Reptilia. In Appleby. R.M., et al. ‘The Fossil Record, a Symposium with Documenta- tion’. 827 pp. (Geological Society of London: London). Cope, E.D., 1869. Synopsis of the extinct Batrachia, Reptilia and Aves of North America. Trans. Amer. Phil.Soc. 14 : 34-61. De Vis, C.W., 1885. On remains of an extinct saurian. Proc.Roy.Soc. 2: 181-91. Etheridge, R., Jr, 1917. Reptilian notes: Megalania prisea, Owen, and Notiosaurus dentatus, Owen; lacertilian dermal armour; opalized remains from Lightning Ridge. Proc. Roy. Soc. Vic. (ns) 29; 127-33. Gorter, J.D., and Nicoll, R.S., 1978. Reptilian fossils from Windjana Gorge, Western Australia. J. Roy. Soc. W. Australia 60: 97-104. JOFFE, J., 1967. The ‘dwarf crocodiles of the Purbeck Formation, Dorset; a reappraisal. Palaeontology 10 : 629-39. Langston, W., Jr., 1973. The crocodilian skull in historical perspective. In, Gans, C., and Parsons, T.S., eds. ‘Biology of the Reptilia’, vol. 4 Morphology D. 539 pp. (Academic Press: London). Longman, H.A., 1924. Some Queensland fossil vertebrates. Mem. Qd Mus. 8: 16-28. 1925. A crocodilian fossil from Lansdowne Station. Mem. Qd Mus. 8: 103-8. 70 MEMOIRS OF THE QUEENSLAND MUSEUM Marsh, O.C., 1884. A new order of extinct Jurassic reptiles (Macelognatha). Amer. J. Sci. (3) 27: 341. Molnar, R.E., 1976. Crocodile with laterally com- pressed snout: first find in Australia. Science 197; 62 ^. Moor, C.C., 1921. Notes on the postcranial skeleton in the Crocodilia. Bull. Amer. Mus. Nat. Hist. 44; 67-100. 1925. A revision of the Mesozoic Crocodilia of North America. Bull. Amer. Mus. Nat. Hist. 51; 319-432. 1930. A new species of crocodilian from the Torrejon beds. Amer. Mus. Nov. 447: 1-11. OsTROM, J.H., 1971. On the systematic position of Macelognathus vagans. Postilla 153: 1-10. Owen, R., 1884. ‘A History of British Fossil Reptiles.’ 856 pp. (Cassell & Co.: London). Riek. E.F., 1952. The fossil insects of the Tertiary Redbank Plains Series. Part I: an outline of the fossil assemblage with descriptions of the fossil insects of the orders Mecoptera and Neuroptera. Univ. Qd Papers, Dept. Geol. 4: 3-14. Romer. A.S., 1966. ‘Vertebrate Paleontology’, third edition. 468 pp. (Univ. of Chicago Press: Chicago). Seeley, H.G., 1874. On cervical and dorsal vertebrae of Crocodilus cantabrigiensis (Seeley) from the Cambridge Upper Greensand. Quart. J. Geol. Soc. London 30: 693-5. 1876. On Crocodilus icenicus (Seeley), a second and larger species of crocodile from the Cambridge Upper Greensand, contained in the Woodwardian Museum of the University of Cambridge. Quart. J. Geol. Soc. London 32: 437-9. Skwarko, S.K., 1973. A correlation chart for the Cretaceous system in Australia. Bur. Miner. Resources, Bull. 126; 55-63. Steel, R., 1973. Crocodylia. In, Kuhn, O., ed. ‘Encyclopedia of Paleoherpetology’, vol. 16. 116 pp. (Gustav Fischer Verlag: Stuttgart). Young C., 1973. A new fossil crocodile from Wuerho. Mem. Inst. Vert. Paleo. Paleoanthro., Acad. Sin. 11:37^4. (In Chinese). Whiting, J.W., and Relph, R.E., 1961. The occurrence of opal at Lightning Ridge and Grawin, with geological notes on County Finch. Tech. Rept., N.S. W. Dept. Mines 6: 7-2 1 . Plate 1 A: Distal end of ?femur from Lightning Ridge (Barlow collection) posterior aspect. B: The same, medial aspect. Bar represents 1 cm. C: Crocodilian tibia from Lightning Ridge (AM FI 5821) medial aspect. D: The same, anterior aspect. Bar represents 2 cm. E: The mandibular fragment (holotype) of Crocodylus (Bottosaurus) selaslophensis (AM FI 58 18), dorsal aspect. F: Crocodilian maxillary fragment from Lightning Ridge (AM FI 8628), lateral aspect. G: The mandibular fragment, lateral aspect. Bar represents 1 cm. Unbroken surface is present around the specimen beneath the third and fourth teeth, thus the full depth of the mandible is represented at that point. H: The maxillary fragment, ventral aspect. Bar represents 1 cm. MOLNAR: PROCOELUS CROCODILE Plate 1 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 2 A and B: Crocodilian cervical centrum from Lightning Ridge, A, dorsal and B, lateral aspect of centrum in the Anderson collection. Bar represents 2 cm. C: Crocodilian caudal vertebra from Lightning Ridge (AM F60080) lateral aspect. Bar represents 2 cm. D: Incomplete cervical neural arch of crocodilian from Lightning Ridge, anterior aspect (AM F60081). Diagonal hatching indicates glue connecting the two portions. Bar represents 2 cm. E: The caudal vertebra (AM F60080), ventral aspect. F: The same, dorsal aspect. MOLNAR: PROCOELUS CROCODILE Plate 2 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 3 A: Crocodilian sacral vertebra from Lightning Ridge (AM FI 57 19) ventral aspect. B: The same, posterior aspect. MOLNAR; PROCOELUS CROCOOrLE Plate 3 Mem. Qd Mus. 20(1): 77-87, pi. 1 [1980]. AN ANKYLOSAUR (ORNITHISCHIA: REPTILIA) FROM THE LOWER CRETACEOUS OF SOUTHERN QUEENSLAND R.E. Molnar Queensland Museum ABSTRACT A small ankylosaur, described as a new genus and species (Minmi paraverlebra), has been found in the Lower Cretaceous Bungil Fm. near Roma, southeast Queensland. The specimen is unique in the possession of a ventral armour of small ossicles, and possession of bony elements alongside of the neural spines of the dorsals, herein called paravertebrae. This is the first ankylosaur discovered in Australia. The ankylosaurs include the prominent armoured dinosaurs of Cretaceous times. Well known specimens have been found in Asia, North America and Europe, but little is known of this group in the southern hemisphere (e.g. von Huene 1929, Nopcsa 1929). Two reports of armoured dinosaurs from Australia (Etheridge 1904, Barlholomai 1966a) have proved incorrect. The earlier was due to a misidentification of bivalve steinkerns, and the latter to a misidentification of sauropod limb material. The specimen described herein thus represents the first ankylosaur known from Australia, as well as the most complete material yet described from the southern hemisphere (although admittedly incomplete compared with much northern hemisphere material). In addition to its geographic significance, this specimen is the first to show extensive regions of the ventral armour. The presence of a series of previously undescribed ossifications alongside of the neural spines of some posterior dorsal vertebrae is also clearly demonstrated. The specimen consists of eleven dorsal vertebrae and associated elements, an incomplete right pes, and much of the ventral armour, and was collected by Alan Bartholomai in 1964 probably from a gulley 1 km south of Mack Gulley, north of Roma, Queensland. The material was preserved in calcareous concretions, and was articulated when discovered. Stratigraphy The site lies on the mapped Minmi Member of the Bungil Formation (Day 1964, Exon and Vine 1970). This member consists of medium to fine sands, silts and muds, with calcareous concretions (Day 1964, Exon 1976), and was deposited under freshwater (Day 1964) or brackish marine conditions (Exon 1976). No other vertebrate material is known from this unit, which is considered of Lower Cretacous age, probably Aptian (Exon and Vine 1970), although dates as early as Neocomian have been suggested (Exon 1976). Occurrence At least the trunk and pes of the creature were articulated when it was buried, and the ventral armour was still in position. Some wear, breakage, and displacement of elements did however occur, with one instance of considerable re-orientation. The proximal portion of the left(?) rib associated with the sixth vertebra from the front preserved, was displaced, rotated about 180 degrees and came to rest across the posterior central articular face. The other instances of breakage did not result in such re-orientation, usually with no movement of the detached portion. All of the left ribs, except that fused to the vertebra, have been ventral ly displaced as a unit, and some of the vertebrae have also slipped downwards. 78 MEMOIRS OF THE QUEENSLAND MUSEUM In addition to yielding the osseous remains, the matrix of the concretions showed some other structure. The texture of the matrix varied across the pavement of ventral scutes, and was uniformly argillaceous external to the pavement, but with large regions of coarser grain size internal to that layer. These regions formed no discernible pattern within the concretion. Large, lobate masses of clay also occurred between the level of the vertebral column (and ribs) and that of the ventral ossicles. Unfortunately because of the fragmented condition of the concretions no overall pattern of these clay masses was discernible. There appears to have been a sheet of clay below the vertebrae, with other possible isolated masses lying just above the ventral pavement. The sheet is considerably thinned beneath the vertebral column and deflected upwards on either side of the row of centra, as if, after formation of the sheet, the vertebral column had subsided into it. While it is tempting to identify these clay masses with some visceral structure there is no good evidence for this. The obvious structures that might be filled with clay would be the lungs or the gastrointestin- al tract. The continuity of the sheet from one side to the other beneath the vertebral column suggests that it does not represent a fill of the lungs, and the great breadth of the sheet, together with the absence of any internal structures that could possibly be attributed to ingested food, suggests that a fill of the gut tract is likewise not represented. Five ossicles from the ventral pavement, together with one small, water-worn quartz pebble, were found between the ventral pavement and the level of the ribs and vertebrae. This relocation of the ossicles suggests that they fell to these positions from the ventral pavement while the carcase lay upside-down on the substrate. Identification Although ankylosaurs have been recovered from marine deposits (e.g. Mehl 1936), the occurrence of an incomplete specimen on a continent not previously known to yield ankylosaurs requires Justifying this identification. Eight groups of Mesozoic reptiles are known to have developed extensive dermal armour: chelonians, placodonts, aetosaurs, phytosaurs, crocodilians, sauropods, stegosaurs, and ankylosaurs. Three of these, the placodonts, aetosaurs, and phytosaurs, are not known to have survived the Triassic, and also differed in the form of the ventral armour and vertebrae from the Minmi saurian. Chelonians are characterised by the development of a ventral plastron rather than a pavement of ossicles, and by fusion of the ribs to the carapace, a feature absent in the Minmi saurian. Eusuchian crocodiles have procoelous dorsals, which the Minmi saurian does not, while mesosuchians have more nearly horizontal transverse processes and zygapophyses than the Minmi saurian. When armoured, mesosuchians have broad ventral plates rather than a pavement of small ossicles. Most sauropods apparently did not possess dermal armour, but recently discovered material from Argentina (Bonaparte, Salfity, Bossi and Powell 1977) does exhibit an armour of small ossicles, resembling those of the Minmi saurian. The first sacral of this form is opisthocoelous (Bonaparte and Powell 1978) and this suggests opisthocoelous dorsals, while those of the Minmi saurian are amphipla- tyan. Stegosaurs have elongate, elevated dorsal neural arches quite unlike those of the Minmi saurian, and the spines or arches are posteriorly inclined, while those of the Minmi saurian are vertical. Thus the Minmi saurian may be excluded from seven of these eight groups, the ankylosaurs alone remaining. The Minmi saurian shows several ankylosaurian features, such as a presacral rod, fusion of ribs to vertebra, centra with expanded amphiplatyan articular faces, and transverse processes inclined dorsally (Coombs 1978), and therefore the Minmi saurian is considered an ankylosaur. In spite of the incompleteness of the specimen, which does not render it ideal type material, the unique features represented (ventral armour and paravertebrae) make it desirable to provide a name for this taxon. Order ORNITHISCHIA Seeley Suborder ANKYLOSAURIA Osborn Family incertae sedis Genus Minmi nov. Type Species Minmi paravertebra nov. Diagnosis: Small ankylosaur with paraverte- bral elements present; ventral armour formed by a pavement of small ossicles; dorsal vertebrae amphiplatyan, without notocordal knobs; trans- verse processes slender and triangular, not T-shaped, in section; neural canal broad; posterior intervertebral notch shallow. MOLNAR: ANKYLOSAUR (ORNITHISCHIA: REPTILIA) 79 Etymology: Minmi from the Minmi Crossing, near the site of discovery. ‘Minmi’ seems to be of aboriginal origin, but uncertain meaning, either being a corruption of Min Min (a kind of will-o-the-wisp light), or referring to a large lily (Reed 1967). Justification of the erection of a new genus for this specimen presupposes knowledge of the detailed description. Thus this justification is postponed to the discussion. Minmi paravertebra nov. Type Specimen: Queensland Museum F10329. Eleven dorsal vertebrae, with associated bases of eleven ribs and three paravertebrae; five incomplete ribs; one incomplete pes; ventral armour; two unidentified elements. Locality: Probably a gulley about 1 km south of Mack Gulley, on the Injune Road, north of Roma, southeast Queensland. Horizon: Minmi Member of the Bungil Formation. Age: Lower Cretaceous, probably Aptian. Etymology; The specific name refers to the unique paravertebral elements. Diagnosis As there is only a single species attributed to the genus, the specific diagnosis cannot be separated from the generic diagnosis. Description VERTEBRAE: Eleven vertebrae are preserved. They were all articulated, but are not demonstra- bly consecutive. They were collected in two groups of three, and one of five. Most retain both centra and arches, while four retain the transverse processes (of at least one side) and the spines as well. The two most posterior centra of the group of five are fused together with no sign of exostoses or other pathological features, and thus this is identified as the anterior end of the presacral rod. The centra are uniformly amphiplatyan, without notocordal knobs. They are laterally and ventrally constricted, with expanded articular faces. These faces are subcircular without obvious compression. Slight lateral depressions are found in the centra Just below the pedicles of the arches. Only one centrum has a keel, and it is faint. (Plate l,fig.B). The neural arches are but slightly elevated (Fig. 1). The neural canals are high but broad, the ratio of breadth to height ranging from 0-76 to 102. The transverse processes are slender, dorsally inclined, and triangular (not T-shaped) in cross-section. The diapophyseal facets face ventrolaterally, and are situated just below the ends of the processes, giving the tips an unswept look when viewed from the front or back. These facets are shallowly concave, and just dorsal to them, at the tips appear smooth, shallowly convex surfaces like articular surfaces. The parapophy- seal facets are cuplike, and situated on the arches at the level of the top of the neural canal. No variation in this level is apparent, however on the posterior vertebrae a rugose band joins the parapophyseal to the diapophyseal facet, running along the ventral margin of the transverse process. The parapophyseal facets are each backed by a buttress supporting the transverse process and running ventrally onto the centrum. A second buttress runs posteriorly along the top of the postzygapophyseal facet. The neural spines are uniformly thin, rectan- gular and anteroposteriorly long. They overhang the posterior ends of the centra. The prezy- gapophyses are borne on short processes that, however, increase in length posteriorly. The facets are inclined at about 65 degrees to the horizontal and are united along their ventral edges to form a V-shaped trough. The postzygapophyses are borne directly on the overhanging posteroventral portion of the neural spine. Paravertebrae: On the left side of the first three of the group of five vertebrae preserved in articulation, are a set of three posteroventrally inclined, elongate bony rods. These rods superficially resemble those identified as ossified tendons, but one may be seen to attach to the Fig. 1: Dorsal vertebra of Minmi paravertebra in left lateral (left) and posterior (right) aspects. Broken line indicates reconstructed portion. Bar represents 2 cm. 80 MEMOIRS OF THE QUEENSLAND MUSEUM posterior margin of one of three elements located between the neural spines and the tips of the right transverse processes. Such elements have not been previously reported, and are here called paravertebrae (Plate 1, figs. C and D). These paravertebrae are flattened and lie in the horizontal plane (Fig. 2). The rod-Hke processes, all incomplete posteriorly, extend back at least the length of one vertebra, and attach to the posteromedial corner of the paravertebra. The anterior two paravertebrae are triangular in form when viewed from above, with the posteromedial angle a right angle, and the hypotenuse shallowly emarginate around the tip of the transverse process. This margin is serrate. The posleriormost paravertebra is elongate, extending well lateral to the tip of the transverse process, and appears to have been rectangular in form when viewed from above.' The medial margins of both this and the middle paravertebra are deflected ventrally to lie along the lateral faces of the neural spines. The dorsal face of the posteriormost paraverte- bra is preserved (the others are eroded dorsally) and shows no sign of sculpture as is found in dermal armour. Mantell (1851) reported that Hylaeosaurus had a row of small ossicles on either side of the neural spines. In Euoplocephalus tutus Fig. 2 : Posterior dorsal vertebrae, ribs and paravertebrae in posterior aspect: posteriormost vertebra preserved on right, preceding vertebra on left. A, paravertebra; B, dorsal process of rib; C, presumed ossified tendons. (the specimen originally described as Scolosaurus cutleri) the dorsal tips of the neural spines extend up to the plane of the dorsal armour, as may be seen in Plate 7 of Nopcsa (1928). Thus the paravertebrae may have been situated in the plane of the dorsal armour, since they are at the level of the tops of the neural spines, but there is no evidence that they either formed part of the dorsal armour or supported elements of the dorsal armour. Maryanska (1977) describes a set of plate-like processes attached to the ribs in Saichania chulsanensis. These processes, which attach to the posterior margins of the distal portions of the ribs, are obviously not paravertebrae which are found proximal to the tubercles of the ribs and are not known to attach to the ribs at all. Nonetheless these processes, which Maryanska also believes may have been present in Ankylosaurus and Panoplosaurus (= Edmontonia}, as well as the paravertebrae, exemplify a tendency among the ankylosaurs toward the development of (presuma- bly) dermal plates in the body wall. The existenee of the elongate posterior processes suggests some attachment of the dorsal epaxial muscles to the paravertebral elements, and hence mobility of the paravertebrae relative to the vertebrae. If the paravertebrae were associated with armour this could imply mobile dorsal dermal armour. The adaptive advantage of this is obscure. The posteriormost paravertebra extends sufficiently far laterally to reaeh the rib, and eould thus have contacted it. The other paravertebrae preserved, however, seem not to have reached their respective ribs, so that any arrangement resulting in rib motion would seem to have affected only one pair of ribs. In conclusion while the paravertebrae may well have exhibited some limited motion relative to the vertebrae, the function of this motion cannot be presently stated. Two short rod-like segments inclining poster- odorsally lie against the neural spines medial to the elongate processes: these seem to represent actual ossified tendons. Ribs: The bases of eleven ribs are preserved along the left side of the dorsals, but only of three on the right, the others having been lost due to erosion. Portions of the more distal regions of five successive ribs are also present. The bases of the ribs are strongly compressed anterolaterally and the capitulum is bluntly convex and dorsoventrally elongate. The tuber- culum is a low, convex facet on the medial side of the low tubercle. On at least one rib these two MOLNAR: ANKYLOSAUR (ORNITHISCHIA: REPTILIA) 81 facets are connected by a thin, shallow trough, suggesting a continuous articulation with the vertebra as is found in some other ankylosaurs (Coombs, pers. comm.). One fragment of rib shows the edge of a marked, medially recurved dorsal process lateral to the tubercular process (Plate 1, fig. E); unfortunately the rib is broken at this point so that the extent of the process is not known. The other proximal portions of ribs are all broken just lateral to the tubercular process. More distally the ribs take on the characteristic T-shaped cross-section of ankylosaur ribs (also found in some sauropods and some hadrosaurs, Gilmore 1914). Pes: Three metatarsals, four phalanges (including an ungual) and a single tarsal remain of the right pes. (Plate 1, fig. A). The disposition of the metatarsals and phalanges indicates that at least four digits were present. The ventral portions of all of the pedal elements are badly worn. The tarsal could not be identified, and is irregularly quadrangular in form. The metatarsals are robust, expanded proximally, and constricted at mid- shaft. The ungual is blunt and about 1-5 times as long as wide. Ventral Armour: The preserved ventral surface is covered by a pavement of small dermal ossicles. This surface is apparently from the belly just anterior to the pelvis, as the portions of the pavement were preserved on the same blocks as the dorsal vertebrae. The ossicles are uniform in size and shape, small, 6 to 8 mm in diameter and 4 to 5 mm thick. They are cushion-shaped, flattened, roughly hexagonal to roughly square in form from below. The dorsal (internal) surface exhibits the common surface texture of trabeculae running in two mutually perpendicular directions, described for Nodosaurus by Lull (1921). The ventral (external) surface is mildly sculptured in a radiate pattern; there is no indication of any keel. The ossicles occur in a single layer and there is no obvious variation in size or form. A flat, roughly triangular, badly weathered element, 6-5 by 5 cm as preserved was found 5 mm above a fragment of rib. This piece may indicate a dorsal armour composed of scutes larger than those found on the venter. Discussion Comparison: Comparison of Minmi with the known taxa of ankylosaurs must be based on vertebral form, since paravertebrae are absent in the other taxa, the ventral armour is either missing or poorly preserved, the ribs seem not to be diagnostic and the pes is incomplete in the only known specimen of Minmi. A survey of the available relevant literature (Bohlin 1953, Brown 1908, Bunzel 1871, Coombs 1978, Eaton 1960, Gilmore 1930, Hulke 1881, Huxley 1867, Lull 1921, Maleyev 1952, Maryanska 1977, Moodie 1910, Nopsca 1928, 1929, Ostrom 1970, Owen 1884, Seeley 1879, 1881) reveals that while ankylosaur dorsal vertebrae are similar in form, there are character states that vary from genus to genus; these states are presented in Table 1 . One problem of the comparison of ankylosaur- ian dorsals is that the complete series of dorsals has been studied in only one form, Ankylosaurus magniventris (Coombs pers. comm.). There appears to be almost no progressive variation in the dorsals of that species, with the exception that the posterior intervertebral notch becomes increasingly shallow as one proceeds posteriorly in the series. For the type specimen of Minmi paravertehra the group of five vertebrae including the anterior end of the presacral rod must be the most posterior set of dorsals preserved, thus comparison of this group with the other six dorsals can suggest progressive variation. Such compari- son suggests that the prezygapophyses become increasingly long posteriorly, and that the centra become increasingly broad. Thus for other taxa for which only dorsals of unknown position are described and figured these characters were not used. In vertebral form Minmi agrees most closely with Nodosaurus {sensu Coombs 1978), from which it differs in that the diapophyses originate below the zygapophyses in Nodosaurus textilis (Lull 1921), but dorsal to them in Minmi. Minmi differs from Hylaeosaurus (sensu Coombs 1978) in having amphiplatyan dorsal centra, those of Hylaeosaurus being planoconcave (Hulke 1881), and in having relatively shorter prezygapophyseal processes. These processes of the dorsals of Hylaeosaurus extend anteriorly to the middle of the preceding centrum (Owen 1884). From Sauropelta, Minmi differs in lacking notocordal knobs, and having vertically orientated spines. From Crataeomus (see note to Table 1) Minmi differs in having amphiplatyan centra, in lacking the sulcus between the postzygapophyses des- cribed in that form (Seeley 1881), and in lacking the fossa anterodorsal to the parapophysis (Seeley 1881). Differences from other nodosaurids are given in Table 1. Ankylosaurids have neural canals less broad than in Minmi, and with the exception of Talarurus, parapophyses placed level with the top of the centrum, rather with the top of the neural canal. 82 MEMOIRS OF THE QUEENSLAND MUSEUM Forms for which too little information is available for inclusion in Table 1 may also be distinguished from Minmi. Acanthopholis has bioconcave centra (Seeley 1879), and a neural canal much higher relative to the central height (Huxley 1867). Stegosaurides dorsals are represented by a centrum with a flat ventral surface (Bohlin 1953), quite unlike other known ankylosaur dorsal centra. That attributed to Princonodon (Lull 1911) tapers in dorsal view from anterior to posterior, also unlike other known ankylosaur dorsal centra. Dorsals of Hoplitosaurus are biconcave and described as having flattened sides (Gilmore 1914). None of these features are found in Minmi. Saichania (Maryanska 1977) and Pinacosaurus (Young 1935) have dorsal neural canals that are strongly compressed from side to side. This seems to be common in the dorsals of Asian Late Cretaceous ankylosaurs, as it is also true of Talarurus (Maleyev 1952, Fig. 7) and Heishansaurus (Bohlin 1953, Fig. 30b). Thus Minmi can be distinguished from other ankylosaurs by its dorsal vertebral morphology, as well as by its possession of paravertebrae. Table 1 suggests that Minmi agrees more closely with some nodosaurids (e.g. Nodosaurus) than with any ankylosaurid. Coombs (1978) gives no firm criterion applicable to the dorsal vertebrae for distinguishing nodosaurids from ankylosaurids. Minmi is thus classified as family incertae sedis, but may be a nodosaurid. Geographic Considerations: Minmi is the first ankylosaur known from Australia, and one of the few reported from the Gondwana continents. The other reported forms are: Loricosaurus scutatus (von Huene 1929), known from armour and referred caudal vertebral centra (von Huene 1931) from north-central Argentina; Lametasaurus indicus, known from armour (Matley 1924), Brachypodosaurus gravis, known from a humerus (Chakravarti 1934), and undescribed material (Coombs 1978) from India; and undescribed nodosaurid material from Madagascar (Russell, Taquet and Thomas 1976). To this latter are possibly referable teeth described as Stegosaurus madagascariensis (Piveteau 1926) and recognised as ankylosaurian by Hoffstetter (1961). None of the described TABLE !: Dorsal Vertebral Character States of A//a a// and other Ankylosaurs. Genus Central form Notocordal knob Keel Neural spine Neural canal Transverse process Parapophyses location Fused ribs Crataeomus* proc. absent absent vert. broad slender top of centrum 9 Hylaeosaurus plan. absent present ? narrow ? top of canal ? Nodosaurus amph. absent absent vert. narrow 9 top of canal present Panoplosaurus amph. present present** post. broad broad varies! varies! Sauropelta amph. present absent post. broad 9 top of canal 9 Silvisaurus amph. absent absent post. broad broad top of canal? absent Struthiosaurus plan. absent present post. narrow slender top of canal 9 Minmi amph. absent absent! t vert. broad slender top of canal present Ankylosaurusi amph. absent absent vert. narrow broad top of centrum present Heishansaurus amph. absent present 9 narrow 9 ? 9 Talarurus amph. present present 9 narrow ? top of canal present * The genera here used follow Coombs 1978, and Maryanska 1977, with the exception of Crataeomus. This genus is treated as valid because of the different position of the pseudacromial process in Crataeomus pawlowitschii from that in C. lepidophorus {=Struthiosaurus austriacus) and Struthiosaurus transylvanicus. This difference is comparable to, but greater than, that between the scapulocoracoids of Ankylosaurus and Euoplocephalus. ** Keel present anteriorly on the centra only (Gilmore 1930). f Variation between species not within a species. 1 1 Faint keel present on one of eleven centra. I Euoplocephalus does not differ from Ankylosaurus in these states (Coombs, pers. comm.), and so this entry represents Euoplocephalus as well. Abbreviations: proc. = procoelous amph. = amphiplatyan post. = posteriorly inclined, plan. - planoconcave vert. = vertical MOLNAR: ANKYLOSAUR (ORNITHISCHIA; REPTILIA) 83 forms are represented by material comparable to that of Minmi, so taxonomic relationships between them cannot presently be established. The existenee of an ankylosaur in Australia, in the Lower Cretaeeous, underscores the widespread distribution of this group soon after its first appearance in the fossil record in the Neocomian (Coombs 1978). It also suggests the existence of a route into Australia during the early Cretaceous, or (accepting the existence of ankylosaurs during the Jurassic as postulated by Coombs 1978) the late Jurassic. Currently accepted dates for the separation of Australia from Antarctica are early Tertiary (Weissel and Hayes 1972, Deighton, Falvey and Taylor 1976), while the breakup of Gondwanaland is usually dated as Jurassic (e.g. Dietz and Holden 1970, Seyfert and Sirkin 1973); Rich (1975) in surveying the possible tetrapod dispersal routes across Antarctica into Australia concluded that any routes from South America would have been archipelagic after the mid- Mesozoic. The discovery of ceratodontid tooth- plates resembling those of Neoceratodus forsteri, the living Australian lung-fish, in the Upper Cretaceous Coli-Toro Formation of Argentina (Pascual and Bondesio 1976) suggests that such a route may have existed into the Upper Cretaceous. Molnar (1980) has presented evidence that such a route may have involved a filter, consonant with Rich’s conclusion that the route would have been archipelagic. Routes involving Africa or India are also possible, as these land masses were in connection with Antarctica into the Lower Cretaceous (Dietz and Holden 1970). Ankylosaurs are known from India and it seems that routes from Africa, at least, into the northern continents were open in the Lower Cretaceous (Molnar 1980, Sues and Taquet 1979). The presence of the freshwater plesiosaur Leptocleidus in the Lower Cretaceous of Queensland (Bartholomai 1966b), as well as in South Africa and western Europe, suggests that such a route could well have been taken by the ancestors of Minmi. SUMMARY A small Lower Cretaceous ankylosaur has been discovered in southeast Queensland. This is the first known ankylosaur from Australia. The ankylosaur, Minmi paravertebra, derives from the Minmi Mem. of the Bungil Fm. The existence of a presacral rod and fusion of one pair of ribs to the vertebra indicates that it is an ankylosaur. A set of peculiar elements of unknown function lie parallel to the dorsal neural spines, medial to the ribs. These elements, herein named paravertebrae, are preserved with the posterior three dorsals preserved. A fragment of one of the ribs shows a dorsal process just lateral to the tubercular process. Minmi paravertebra had a tetradactyl pes, and a ventral armour of small ossicles. Its dorsal vertebrae and ribs closely resemble those of other ankylosaurs. The sequence of dorsal vertebrae of Minmi suggest that ankylosaur dorsals change slightly in form from anterior to posterior; the posterior vertebrae having longer prezygapophyses and broader centra than the anterior. This specimen, although consisting of little more than the posterior portion of the trunk, minus pelvis, is still the most complete described from the southern hemisphere. It demonstrates that ankylosaurs were quite widespread soon after their first appearance (in the Neocomian) in the fossil record. They probably entered Australia via South America. ACKNOWLEDGMENTS Dr Alan Bartholomai brought this specimen to my attention, and kindly permitted the prepara- tion and study of it. I am also grateful to Dr W.P. Coombs Jr, Mr G. Czechura, Dr R.W. Day and Dr M. Wade for their gracious assistance. LITERATURE CITED Bartholomai, A., 1966a. Fossil footprints in Queens- land. Aust. Nat. Hist. 15: 147-50. 1966b. The discovery of plesiosaurian remains in freshwater sediments in Queensland. Aust. J. Sci. 28 : 437. Bohlin. B., 1953. Fossil reptiles from Mongolia and Kansu. Sino-Swedish Expeditions, Publ. 37 (6): 11-67. Bonaparte, J.F., and Powell, J.E., 1978. A continental assemblage of tetrapods from the Upper Cretaceous of northwestern Argentina, (Sauropoda — Coeluro- sauria — Carnosauria — Aves). Abs. lere Table ronde Ecosystemes continentaux Mesozoique, Paris, 1978. Bonaparte, J.F., Salfity, J.A., Bossi, G., and Powell, J. E., 1977. Hallazgo de dinosaurios y aves Cretacicas en la formacion Lecho de El Brete (Salta), proximo al limite con Tucuman. Acta geol. lilloana 14 : 5-17. Broom, R., 1910. Observations on some specimens of South African fossil reptiles preserved in the British Museum. Trans. R. Soc. S. Afr. 2 : 19-25. Brown, B., 1908. The Ankylosauridae, a new family of armored dinosaurs from the Upper Cretaceous. Bull. Amer. Mus. nat. Hist. 24 : 187-201. Bunzel, E., 1871. Die Reptilfauna de Gosau- Formation in der Neuen Welt bei Wiener-Neustadt. Abh. K. K. Geol. Reichs-Anstalt, Wein 5(1): 1-20. 84 MEMOIRS OF THE QUEENSLAND MUSEUM Chakravarti, D.K., 1934. On a stegosaurian humerus from the Lameta beds of Jubbulpore. J. Geol. Miner. Metall. Soc. India 6 (3): 75-9. Coombs, W.P., Jr.. 1978. The families of the ornithischian dinosaur order Ankylosauria. Palaeontology 21: 83-127. Day, R.W., 1964. Stratigraphy of the Roma- Wallumbilla area. Geol. Surv. Qd, Publ. 318: 1-23. Deighton, [., Falvey, D.A., and Taylor. D.J., 1976. Depositional environments and geotectonic framework: southern Australian continental mar- gin. APEA Jounral 16: 25-36. Dietz. R.S., and Holden. J.C., 1970. Reconstruction of Pangaea: break-up and dispersion of continents, Permian to Recent. J. geophys. Res. 75: 4939-56. Eaton, T., Jr.. 1960. A new armored dinosaur from the Cretaceous of Kansas. Paleont, Contrib. Univ. Kans. Vert. 8: 1-24. Exon, N.F., 1976. Geology of the Surat Basin in Queensland. Bull. Bur. Miner. Resour. Geol. Geophys. Aust. 166: 1-160. Exon, N.F., and Vine, R.R., 1970. Revised Nomencla- ture of the ‘Blythsdale’ sequence. Qd Govt Min. J. 71:48-52. Gilmore, C.W., 1914. Osteology of the armored Dinosauria in the United States National Museum, with special reference to the genus Stegosaurus. Bull. U.S. natn. Mus. 89: 1-136. 1930. On dinosaurian reptiles from the Two Medicine Formation of Montana. Proc. U.S. natn. Mus. 77(16): 1-39. Hoffstetter, R., 1961. Nouveaux restes d’un serpent boide'^ {Madtsoia madagascariensis nov. sp.) dans le Crelace''superieur de Madagascar. Bull. Mus. Hist, not.. Pam (2)33: 152-60. V. Huene, F., 1929. Los Saurisquois y Ornitisquios del Cretaceo Argentine. An. Mus. La Plata 3 2: 1-196. 1931. Verschiedene mesozoische Wirbeltierreste aus Siidamerika. Neu. Jb. Miner. Geol. Palaont. (B)66: 181-98. Hlilke. J.W., 1881. Polacanthus foxii, a large undescribed dinosaur from the Wealden Formation in the Isle of Wight. Phil. Trans. R. Soc. 172(3): 653-62. Huxley, T.H., 1867, On Acanthopholis horridus, a new reptile from the Chalk-Marl. Geol. Mag. 4(4): 65-7. Lull, R.S., 1911. Systematic paleontology of the Lower Cretaceous deposits of Maryland. Vertebra ta. Md. geol. Surv. gen. Ser., Lower Cretaceous: 183-21 1. 1921. The Cretaceous armored dinosaur, Nodosaurus textilis. Am. J. Sci. (5)1; 97-126. Maleyev, E.A., 1952. Novii ankilosavr iz verkhnego Mela Mongolii. Dok. Akad. Sauk SSSR 87: 273-6. Mantell, G.A., 1851. ‘Petrifactions and Their Teachings’. (Henry G. Bohn: London). Maryanska, T., 1977. Ankylosauridae (Dinosauria) from Mongolia. Palaeont. Pol. 37: 85-151. Matley, C.A., 1924. Note on an armoured dinosaur from the Lameta beds of Jubbulpore. Rec. geol. Surv. India 55: 105-9. Mehl, M., 1936. Hierosaurus coleii: a new aquatic dinosaur from the Niobrara Cretaceous of Kansas. J. scient. Labs Denison Univ. 31(4): 1-20. Molnar, R.E., 1980. Australian late Mesozoic terrestrial tetrapods: survey and implications. Mem. Soc. geol. Fr.. In press. Moodie, R., 1910. An armored dinosaur from the Cretaceous of Wyoming. Kans. Univ. Sci. Bull. 5: 257-73. Nopcsa, F., 1928. Palaeontological notes on reptiles. Geologica Hung. 1(1): 51-74. 1929, Dinosaurierreste aus Siebenbiirgen. V. Geologica Hung. 4: 1-72. OSTROM, J.H., 1970. Stratigraphy and paleontology of the Cloverly Formation (Lower Cretaceous) of the Bighorn Basin area, Wyoming and Montana. Bull. Peabody Mus. nat. Hist. 35: 1-234. Owen, R., 1884. ‘A History of British Fossil Reptiles’. (Cassell and Company: London). Pascual, R., and Bondesio, P., 1976. Notas sobre vertebrados de la frontera Cretacica-Terciaria. Ill: Ceratodontidae (peces osteichthyes. Dipnoi) de la Formacion Coli-Toro y de otras unidades del Cretacico tardio de Patagonia y sur de Mendoza. Sus implicancias paleobiogeograficas. Actas de Sexto Congr. Geol. Arg. 1: 565-77. PiVETEAU, J., 1926. Contribution a I’Etude des Formations Langunaires du Nord-Ouest de Madagascar. Bull. Soc. geol. Fr. (4)26: 33-8. Reed, A.W., 1967. ‘Aboriginal Place Names and Their Meanings’. (Reed and Reed: Sydney). Rich, P.V., 1975. Antarctic dispersal routes, wandering continents, and the origin of Australia’s non- passeriform avifauna. Mem. Nat. Mus. Vic. 36: 63-126. Russell, D.D., Taquet, P., and Thomas, H., 1976. Nouvelles recoltes de Verlebres dans les terrains continentaux du Cretace'^ superieur de la region de Majunga (Madagascar). Bull. Soc. geol. Fr. Suppl. 5: 204-8. Seeley, H.G., 1879. On the Dinosauria of the Cambridge greensand. Q. Jl. geol. Soc. Lond. 35: 591-636. 1881. The reptile fauna of the Gosau Formation preserved in the Geological Museum of the University of Vienna. Q. Jl. geol. Soc. Lond. 42: 620-77. Seyfert, C.K., AND SiRKiN, L.A., 1973. ‘Earth History and Plate Tectonics’. (Harper and Row: New York). SuF.S, H.D., AND Taquet, P., 1979. A pachycephalosaur- id dinosaur from Madagascar and a Laurasia- Gondwanaland connection in the Cretaceous. Nature 279: 633-5. Weissel, J.K., AND Hayes, D.E., 1972. Magnetic anomalies in the southeast Indian Ocean. In D.E. Hayes (Ed) ‘Antarctic Oceanology II. The Australian-New Zealand sector’. Amer. Geophys. Union, Antarctic Res. Ser. Mon. 19: 165-96. Young C.-C., 1935. On a new nodosaurid from Ninghsia. Palaeont. sin. (C) 11: 5-28. MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 . Fig. A: The pes of Minmi paravertebra as preserved, in dorsal view. Bar represents 2 cm. Fig. B; The last five dorsal vertebrae preserved, in left lateral aspect. The proximal portions of the ribs are still in place. Note the beginning of the presacral rod at right. Bar represents 5 cm. Fig. C: The same, in dorsal aspect. Paravertebrae are visible on the right. Bar represents 5 cm. Fig. D: Block of matrix with three transverse processes and two paravertebrae adherent. Anterior is to the left. Bar represents 2 cm. Fig. E: Proximal portion of last left rib present, showing dorsal process just lateral to tubercular process. Bar represents 2 cm. Abbreviations: B, tuberculum; C, capitulum; D, dorsal process; N, neural spine; P, paravertebra; R, rib; T, transverse process. MOLNAR: ANKYLOSAUR (ORNITHISCHIA: REPTILIA) Plate 1 Mem.Qd Mus.ia{\): 89-93, pi. 1 [1980] A REVIEW OF THE DE VIS FOSSIL PIGEONS OF AUSTRALIA G.F. VAN TETS Division of Wildlife Research, CSIRO and P.V. Rich Earth Sciences Department, Monash University, and National Museum of Victoria ABSTRACT A review of Australian avian fossils Nyroca effodiata De Vis, Leucosarcia proevisa De Vis and Lithophaps uinaris De Vis has shown that they cannot be distinguished from living species of pigeons, Phaps histrionica and P. chalcoptera. The type specimens of N. effodiate and L. proevisa are shown to be two parts of a single humerus. From a large collection of fossil birds in the Queensland Museum, Charles W. De Vis described only three as extinct new species of pigeons: Progura gallinacea (1888); Lithophaps uinaris (1891); and Leucosarcia proevisa (1905). Progura gallinacea was shown to be a megapode by van Tets (1974). While checking the De Vis material we found that the type of Nyroca effodiata (1905), described and named as a duck by De Vis is the distal part of the same humerus designated as the type of the pigeon Leucosarcia proevisa (Rich 1976, p. 59, fig. 4). We agree that this material is columbid but is not separable from modern species. We have compared the type material of Lithophaps uinaris, Leucosarcia proevisa, and Nyroca effodiata with every genus and almost all species of living Australian pigeons in the collections of the National Museum of Victoria (NMV), Melbourne; the Australian National Wildlife Collections, Division of Wildlife Research, CSIRO (ANWC), Canberra; and the Queensland Museum (QM). We have used the classification of Condon (1975) with amendments {Emu 76: 216-7; 78: 8(U7). SYSTEMATICS ‘Leucosarcia proevisa’ De Vis, 1905 and ‘Nyroca effodiata’ De Vis, 1905 Type Specimens: Leucosarcia proevisa, QM F5511, proximal right humerus; Nyroca effodiata, QM F5544, distal right humerus; these two fragments fit together to form one nearly complete element lacking only part of the midshaft on the internal side (Plate 1). The combined specimen is now registered as QM F55 1 1 . Type Locality and Age: Wurdulumankula, eastern Lake Eyre Basin, South Australia, Quaternary (Fig. 1). Diagnosis Differs from all other Australian pigeons except those in the genus Phaps by combining the following character states; head is highly domed, not flattened proximally as in Petrophassa; distinct ridge runs from external end of head across anconal surface of shaft diagonally to capital groove, whieh is absent in several pigeons including Macropygia and absent or only slightly indicated in Ptilinopus; long axis of ligamental furrow forms small acute angle with long axis of shaft, not large one as in Ptilinopus, Petrophassa, Macropygia, Ducula, Leucosarcia, Geopelia, Chalcophaps and Ocy phaps; proximal projection of long axis of shaft passes through external segment of head instead of lateral to it as it does in Petrophassa: internal surface of shaft just distal to pneumatic fossa deep, markedly planar, and delimited anconally by a sharp ridge that continues distally on the shaft crossing onto the anconal surface (extremely reminiscent of Phaps histronica) differing from most other Australian pigeons (including some Phaps) in which this area is more curved; internal margin of shaft not highly curved unlike that of Petrophassa; tubercle along 90 MEMOIRS OF THE QUEENSLAND MUSEUM external border of shaft elevated further proximaliy from distal end (25% of total shaft length) than in other pigeon genera except Phaps, Geopelia, and Columba; distal end does not flare broadly unlike that in Leucosarcia, Petrophassa, Columba and Ocyphaps; ectepicondylar promin- ence is subdued, not protruding far laterally, unlike in Petrophassa; external condyle is deep relative to width i.e. distal end is not markedly palmoanconally compressed unlike in Petrophassa. some Phaps (not P. histrionica), Ptilinopus, Ducula, and Geopelia; overall proportions of moderate length and indicate a gracile bird, not robust as in Leucosarcia or most Columba, and not as elongate as in Ducula; about the same size as some Phaps, Columba, and Leucosarcia (see Table 1), but smaller than Lopholaimus. Only the distal end of the humerous of Lopholaimus was available. It demonstrated no diagnostic qualitative differences, only differences in size. Comment When compared to every genus and most species of Australian pigeons, 'Leucosarcia proevisa’ clearly shares the greatest similarity with Phaps, the Bronzewings, and in fact should be included in that genus. The fossil exhibits no qualitative differences from any of the species of Phaps. P. elegans is distinctly smaller than the fossil, however. Both P. histrionca and P. chalcoptera are so similar in size and proportions to the fossil that precise assignment to either of these species seems impractical (see Table 1 and Fig. 2). Thus, ‘Leucosarcia proevisa’ should be re-assigned to Phaps within the P. histrionica/ chalcoptera assemblage. *Lithophaps ulnaris’ De Vis, 1891 Type Specimen: Complete right ulna, QM FI 119 (Plate 1). Type Locality and Age: Darling Downs, South- eastern Queensland; Quaternary (Fig. 1). Fig. 1: Localities of fossil pigeons in Australia and ranges of Phaps chalcoptera and P. histrionica. VAN TETS AND RICH: FOSSIL PIGEONS 91 TABLE 1 : Measurements (in mm) of Humeri of recent and fossil Pigeons of Australia Taxa Total length of bird Museum numbers, where applicable N A Total length Measurements B C Maximum Width Width of proximal end distal end D Depth of external condyle Ptilinopus cinctus 330 ANWC BS233 1 41-0 12-5 8-9 5-5 Ptilinopus superbus Ptilinopus 230 ANWC BS244 ! 30-8 11-1 7-7 4-6 magnificus Lopholaimus 430-560 ANWC BS255 1 46-0 15-9 11-5 6-7 antarcticus 430 NMV B9229 1 — — 13-6 8-4 Columba livia Streptopelia 330 NMV - 1 41-8 17-3 10-3 c.6-9 chinensis Macropygia 280 NMV W6447 1 34-8 11-6 8-1 5-0 amboinensis 400 NMV R6470 1 39-5 13-0 9-2 5-6 Geopelia cuneata 200 NMV W6185 I 19-4 6-8 4-9 2-8 Geopelia humeralis 290 NMV W4834 1 31-6 l!-7 7-9 4-8 Chalcophaps indica 240 NMV W5915 1 32-3 111 8-8 4-8 Phaps chalcoptera 305 — 21 42-2^7-6 14-7-16-8 10-6-1 1-9 5-8-7-5 Phaps elegans 280 — 9 38-6-41-1 12-3-13-7 9-5-10-4 5-8-6-3 Phaps histrionica 280 — 4 44-2^5-3 14-9-16-8 9-9-10-7 6-7-7-3 Ocyphaps lophotes 320 NMV B8873 1 38-8 14-4 9-5 5-7 Petrophassa smithii Petrophassa 250 ANWC BS234 2 31 -6-34-! 12-3 8-4- 8-9 5-1-5-6 albipennis Petrophassa 280 ANWC BS226 1 32-4 11-2 8-5 4-9 plumifera Leucosarcia 200-230 NMV B8536 1 27-0 9-3 6-9 4-3 melanoleuca ‘Leucosarcia proevisa 430 ANWC BS232 1 47-7 17-7 12-5 8-1 and Nyroca effodiata’ QM F5511/ F5544 1 45-5 >16-0 10-5 6-6 Diagnosis Tubercles on convex rear edge of fossil are of similar size and evenly spaced as in Geopelia, Phaps, Petrophassa, Ocyphaps and Leucosprcia and not with fourth turbercle from distal end being smaller than three more distal tubercles as in Ptilinopus, Ducula, Lopholaimus, Chalco- phaps, Columba and Streptopelia; or with several median tubercles smaller than proximal and distal ones as in Macropygia; ulnae of Geopelia and Petrophassa are much smaller (see Table 2) and of Leucosarcia much larger and stouter; compared with species of Phaps, fossil decidedly longer than P. elegans, but comparable in size to P. chalcoptera and P. histrionica, neither of which exhibit significant differences in size and shape. Comment In overall morphology we can find no differences between ‘Lithophaps’ and Phaps. Xithophaps ulnaris’ falls within the size range of both Phaps chalcoptera and P. histrionica (see Table 2 and Fig. 3), and we cannot distinguish between these two species. Thus, ‘Lithophaps’ should be included within Phaps and ‘L. ulnaris’ referred to the P. chalcoptera/histrionica assemblage. TABLE 2: Measurements (in mm) of Ulnae of living and fossil Phaps Pigeons in Australia Taxa n Total length Proximal width Minimum shaft width Distal width Phaps chalcoptera 19 46-7 - 52-7 6-7 - 8-3 2-9 - 3-6 6-1 - 6-8 Phaps elegans 16 41-3 - 44-3 5-9 - 6-9 2-6 - 3-0 5-2 - 5-6 Phaps histrionica 3 50-4 - 52-5 5-8 - 8-0 3-0 - 3-2 6-1 - 6-5 ‘Lithophaps ulnaris’ 1 >47-3 6-7 2-9 5-1 92 MEMOIRS OF THE QUEENSLAND MUSEUM i?" £ It- £ I I0-- 9 -- P ctialcoptera . \ P elegans B-| 1 1 1 1 ( 1 » 1 ( 1~— I- 37 33 39 40 41 42 43 44 46 46 47 40 Total Length imm I Fig. 2: Plot of humeral measurements of extant species of Australian Phaps and F, the fossil ‘Leucosarcia proevisa and Nyroca effodiata', QM F55 1 1 /F5544. 8 5 . 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 Total Length (mm) they fall within the range of variability exhibited by both the common Bronzewing {P. chalcoptera) and the Flock Pigeon {P. histrionica), and we can see no decisive method for distinguishing these species on the basis of the humerus or the ulna. We can thus assign the fossils only to the chalcoptera/histrionica assemblage. Although somewhat south or southeast of the modern geographic range of P, histrionica, the fossil occurrences in northern South Australia and southeastern Queensland are within the range of the currently widespread P. chalcoptera. ACKNOWLEDGMENTS Sincere thanks are due to A. McEvey (National Museum of Victoria, Melbourne) and W. Bock and W. Lanyon (American Museum of Natural History, New York), for providing comparative material; and to A. Bartholomai and M. Wade (Queensland Museum, Brisbane) for kindly allowing us to review the De Vis material. Coral Armstrong and Wendy Guy typed drafts of the manuscript, Frank Coffa photographed the fossil and recent material, and Frank Knight drew the map. LITERATURE CITED Fig. 3; Plot of ulnar measurements of extant species of Australian Phaps and F, the fossil ‘Lithophaps ulnaris\QM FI 119. CONCLUSIONS A review of the fossil pigeon material described by C.W. De Vis has demonstrated that ‘Leucosarcia proevisa’, 'Nyroca effodiata’ and ‘Lithophaps unlnaris’ should be included within the living genus Phaps, the Bronzewings. Although the fossils represent birds distinctly larger than the Brush Bronzewing (P, elegans). Condon, H.T., 1975. Checklist of the birds of Australia, Pt. 1. Non-passerines. (RAOU: Melbourne). De Vis, C.W., 1888. A glimpse of the post-Tertiary avifauna of Queensland. Proc. Linn. Soc. N.S.W. (2)3(3^): 1277-92. 1891. On the trail of an extinct bird. Proc. Linn. Soc. N.S.W. 2(4): 117-22. 1905. A contribution to the knowledge of the extinct avifauna of Australia. Ann. QdMus. 6: 1-25. Rich, P.V., 1976. The history of birds on the island continent Australia. Proc. 16th Int. Ornith. Cong., Canberra. Aust. Acad. Sci.: 53-65. VAN Tets, G.F., 1974. A revision of the fossil Megapodidae, including a description of a new species of Progura De Vis. Trans. Roy. Soc. S. Aust. 98(4): 213-24. Plate 1 Humeri and ulnae of fossil and recent pigeons. Right humeri, anconal views, stereo pairs: a, a’ Phaps elegans (NMV B12395); b, b’ ‘Leucosarcia proevisa and Nyroca effodiata’, QM F551 1 /F5544); c, c’ Phaps chalcoptera (NMV B12400); d, d’ Phaps histrionica (NMV B8588). Right humeri, palmar views, stereo pairs: e, e’ see a; f, F Lithophaps Ulnaris. Ulnaris QM-1119 g, g’ see c; h, h’ see d. Right ulnae lateral views: i, see a; j, see b; k, see c; and I see d. VAN TETS AND RICH: FOSSIL PIGEONS Plate 1 5cm Mem. Qd Mus. 20(1); 95-101, pi. 1 [1980]. TWO NEW SPECIES OF SKINKS FROM MID-EASTERN QUEENSLAND RAINFOREST J. COVACEVICH Queensland Museum and K.R. McDonald National Parks and Wildlife Service, Townsville ABSTRACT Two species of skinks (Sphenomorphus luteilateralis, Sphenomorphus amplus spp. nov.) are described from restricted rainforest habitats in mid-eastern Queensland, The former is confined to high altitude, mixed notophyll vine forest in the Eungella National Park, near Mackay. Sphenomorphus amplus occurs at lower altitudes, usually associated with rocks along creeks, in simple and complex notophyll vine forest in the Eungella National Park, and in complex notophyll vine forest in the Conway State Forest, near Proserpine. Until recently the reptiles of mid-eastern Queensland had not been surveyed methodically. Surveys by the Queensland National Parks and Wildlife Service commenced in the area in 1975 and, in that year, the Australian and Queensland Museums, supported financially by the Australian Biological Resources Study, undertook fauna surveys in the area. Twelve sites representative of the different rainforest structural types and floristic provinces of mid-eastern Queensland were examined. Eight sites were surveyed from Mackay to Proserpine, an area that was virtually unknown herpetologically. Preliminary results of surveys by the Australian and Queensland Museums have been presented (Broadbent and Clark 1976, Queensland Museum 1976, Covacevich 1977). As expected, many undescribed species of frogs and reptiles were collected, particularly in the Eungella National Park sites near Mackay. Amongst the undescribed material are two large, lygosomine skinks. Genera of the sub-family Lygosominae, to which Sphenomorphus Fitzinger 1843 belongs, have been defined and their relationships discussed by several authorities: Gray (1845), Boulenger (1887), Smith (1937), Mittleman (1952), and Greer (1970). Greer (1970, p.l71) has noted . , . ‘the greatest single taxonomic problem with the lygosominae is the delineation of genera’ . . . Within this subfamily the genus Sphenomorphus presents special problems of delineation. Skinks currently assigned broadly to Sphenomorphus occur widely in tropical Africa, southern Asia, Australia, New Zealand (Storr, 1967), and New Guinea (Scott, Parker, and Menzies 1977). Storr (1967) and Cogger (1975) have recently defined this genus for Australia to include about 20 species, all of which lack supranasal scales. A New Guinea genus, Otosaurus, was erected by Gray (1845, p.93) for specimens which differed from Sphenomorphus' essentially in having ‘super nasal 2, contiguous’. This genus was maintained by both Smith (1937, p.218) and Mittleman (1952, pp.6-7), whose work maintaining Otosaurus for species with supra- nasals (e.g. O. cumingi, from New Guinea and Southeast Asia) is the most recent ‘total’ discussion of all lygosomine genera and their broad relationships. The species with supranasal scales have, however, been referred to Sphenomorphus by Loveridge (1948), Greer (1970), and Scott, Parker, and Menzies (1977) without discussion. The first species (S. luteilateralis sp. nov.) is easily referred to Sphenomorphus as it is defined for Australian species by Storr (1967) and Cogger (1975). It is apparently related to members of the S. murrayi group within Sphenomorphus — S. murrayi (Boulenger), 5. tenuis (Gray), S. tigrinum (De Vis). All share the following features in addition to those described for Sphenomorphus — large size, robust form, and ground-dwelling habits. 96 MEMOIRS OF THE QUEENSLAND MUSEUM The generic placement of the second species (S. amplus sp. nov.) is more difficult to decide upon, because it has distinct supranasal scales, a feature which would exclude it from Sphenomorphus as it has been defined by Storr (1967) and Cogger (1975) for Australian species, and from Sphenomorphus as it is defined fully by Mittleman (1952). In this case S. amplus sp. nov. would most appropriately be assigned to Otosaurus. However, considering the current lack of understanding of lygosomine generic boundar- ies and relationships generally, and the status of Sphenomorphus species ‘groups’ in particular, and noting that recent authorities have ‘lumped’ Otosaurus in Sphenomorphus, it seems reasonable to place amplus in Sphenomorphus while recognising its affinity with species of the Otosaurus group within Sphenomorphus from New Guinea. Sphenomorphus iuteilateralis sp. nov. (Plate l,a) Material Examined Holotype: QM J31685 Eungella National Park 35’E, 2P 03’S coll. K.R. McDonald, D. Vlasak,9.i.l976, closed forest. Paratypes: J31674— J3I684, J31686, same locality and data as holotype; J3 1687-8, 15.iii,1975, Eungella National Park, as for holotype, col. K.R. McDonald, J. James; J31689, 14.iii.l975, same data as for J31687; J31690, data as for holytype; J3 169 1-7, new born young of three females, same data as holotype; AM R47765-70, 14.iv.l975, Eungella National Park, Dalrymple Heights, coll. P. Webber; R47763^, 14.iv.l975, Eungella National Park, coll. H. Posamen- tier; R47855-6, 3.iv.I975. Eungella National Park near Vlasak property, coll. P. Webber; R47941-3, 6.V.1975, Mt William, Eungella National Park, coll. P. Webber; R47497, 21-26. iv. 1975, Mt William, Eungella National Park, coll. P. Webber. Diagnosis S. luteilateralis most closely resembles the southeastern Queensland/northeastern New South Wales skink S. murrayi (Boulenger 1887). It may be distinguished from 5. murrayi as follows: prominent orange coloured sides with white ocellations, distinct dark patch above fore limb; 37-41 mid-body scales vs 29-33 in S. murrayi. Description of Holotype Snout-vent length 87 mm; tail regenerated. Habit stout. Tip of snout-forelimb/axilla-groin = 33/46 mm. Snout short, obtuse. Rostral broad, forming a suture with the nasals and frontonasal. Frontonasal slightly broader than long, bordered posteriorly by two large prefrontals which just touch. Frontal about 2-5 times as long as broad, and narrow posteriorly; about as long as the frontoparietals and interparietal together. Frontal in contact with first and second supraocular. Supraoculars four, the second largest and the fourth smallest. Supraciliaries 7. Frontoparietals and interparietal distinct, the frontoparietals slightly larger. Supralabials 6, the fifth contacting the eye and immediately below the eye centre. Ear opening large, oval (about the same size as eye) with a deeply set tympanum. No auricular lobules. Mid-body scales 37. Mid-dorsal scales slightly larger than lateral and mid-ventral scales. Two enlarged preanals. Limbs and digits moderately long. Twenty-two lamellae under fourth toe. Colour: In preservative, basically brown dorsally with a few scales with darker markings. A conspicuous large black patch occurs above the forelimb with a small dark area on the dorsal edge of ear and another dark mark exists dorsolaterally between ear and forelimb patches. Ventrally white. Labials and chin shields white. Laterally, between head and forelimb, scales off-white; between forelimb and hindlimb faded orange with distinct white occellations. Limbs are basically similar in colour to the dorsal surface but they have more dark scales. In life the dorsal surface is darker brown and the faded orange is a bright burnt-orange. Variation in Paratypes SVL 52-92 mm. Tail 77-1 12 (only 5 specimens have original tails and most of these are juveniles). The following variation from the holotype description has been observed in the paratypes. Frontonasal scale: J31682 has a fragmented scale. Colour: Differences are minor. Juveniles are similar to adults except that the orange lateral sides are not as intensely coloured. In some specimens the dark patches near the forelimb are not as intensely coloured as in the type. Mid-body scales: Counts in 38 paratypes vary from 36—41 as follows — 36 (2 specimens, 5-5%), 37 (6, 15-5%), 38 (17, 44-5%), 39 (9, 23-5%), 40 (2, 4-5%), 41 (2, 5-5%). No mid-body count could be taken from one damaged juvenile specimen. COVACEVICH AND McDONALD: NEW RAINFOREST SKINKS 97 Lamellae: Counts under the fourth toe vary from 17-22 as follows in 39 paratypes 17 (1, 2-5%), 18 (5, 13%), 19 (9, 23-5%), 20 (11, 28%), 21 (7, 18-5%), 22 (6, 15-5%). Supraciliaries: These vary from 6-8 as follows 6 (2, 5%), 7 (34, 87%), 8 (3, 8%). Supraoculars: These vary only slightly with one having 3 and two having 4 supraoculars on one side and three on the other. Prefrontals: These meet (10 specimens) or are separated (29 specimens). Distribution and Habitat Sphenomorphus luteilateralis is known only from the Eungella National Park on the slopes of Mount William. All specimens were collected in mixed notophyll vine forest above 900 metres. S. luteilateralis is a common species where it occurs. Eighteen specimens were seen in 1 Vj hours in about 2 ha of rainforest. Notes Specimens of S. luteilateralis were observed ‘sunning’ on palm trunks and fronds on the forest floor as well as on dead logs and were also found in rotting logs and palm frond bases. They forage in leaf litter, are always found on the ground, and show no tendency to climb when disturbed. Females bear five individuals viviparously. SVL and tail lengths of 15 individuals at birth varied between 29-0-32-7 (mean 30-64, SD 1-31) and 38-6-45-3 (mean 41-60, SD 3-53). S. murrayi and S. luteilateralis are similar in external morphology as well as habitat preferences and habits. Both species usually bear 5 live young, are found in rainforest at higher elevations, are crepuscular in habit, and are found in similar niches. Both are terrestrial and live in decaying logs and forest floor vegetation. Further, both species are strongly territorial. Single pairs (cC, 9) of both species live harmoniously in captivity but, where more specimens are kept in the one cage, fighting occurs constantly. Warnings are given to intruders of territory by vigorous waving of the tail in the leaf litter. The defendant of the territory invariably faces the intruder when making these aggressive warnings. Fighting sometimes even- tuates, usually in the form of biting around the head until one specimen retreats. Juveniles are not attacked and are tolerated in the adult territory. This pattern has also been observed in the field. S. murrayi and S. luteilateralis are, however, widely separated geographically. S. murrayi is known from above 150 m from northeastern New South Wales and southeastern Queensland (as far north as the Conondale Range) while S luteilateralis is confined to the high altitude slopes (above 900 metres), of Mt William, mid-eastern Queensland. Sphenomorphus amplus sp. nov. (Plate l,b) Material Examined Holotype: QM J26054, Finch Hatton Creek, Eungella National Park, ME.Q., coll. K.R. McDonald, 7.11.1975, Paratypes; QM J31647, 10.i.l976, Eungella Hotel Grounds, coll. K.R. McDonald, P. Wilson; J31648, 14.1.1976, Broken River, Eungella National Park, coll. K.R. McDonald, J. Miller; J31649, 7.i.l976, Wishing Pool, Eungella National Park, coll. V.R.J. Hansen, K.R. McDonald; J31650 - 2, 15.i.l976, Finch Hatton Gorge, Eungella National Park, coll. V.R.J. Hansen, K.R. McDonald; J31653, 2.ix.l975, Finch Hatton Gorge, Eungella National Park, coll. V.R.J. Hansen; J31654, 13.ix.l975, Finch Hatton Gorge, Eungella National Park, coll. V.R.J. Hansen; J31655, 15.i.l976, Finch Hatton Gorge, Eungella National Park, coll. K.R. McDonald, V.R.J. Hansen; J31656, 17.iii.l975, Finch Hatton Gorge, Eungella National Park, coll. K.R. McDonald, J. James; J31657, 25.iii.1975, Finch Hatton Gorge, Eungella National Park, coll. V.R.J. Hansen; J31658, 17. ii. 1975, Finch Hatton Gorge, Eungella National Park. coll. K.R. McDonald, V.R.J. Hansen; J31659, 15.i.l976, Finch Hatton Gorge, Eungella National Park, coll. K.R. McDonald, V.R.J. Hansen; J31660 21.vii.l975, Finch Hatton Gorge, Eungella National Park, coll. V.R.J. Hansen; J31661 - 73, 5.1.1976, Finch Hatton Gorge, Eungella National Park, coll. K.R. McDonald, V.R.J. Hansen; J32784-5, 19-22.iv.l975, Brandy Creek, Conway State Forest, coll. J. Covacevich, R. Monroe, P. Filewood; J33980, J33983, 7-14.iv.l975, Finch Hatton Gorge, Eungella National Park, coll. R. Monroe, J. Covacevich, P. Filewood; J33984, 8.iv.l975, Finch Hatton Gorge, Eungella National Park, coll. V.R.J. Hansen; J33985, J34022 -3, J34026-9, J34034-5, J34061, J34084, 7-1 4.iv. 1975, Finch Hatton Gorge, Eungella National Park, coll. J. Covacevich, R. Monroe, P. Filewood; AM R61483, R61482 (formerly J33981-2), 7-14.iv.I975, Finch Hatton Gorge, Eungella National Park, coll. J. Covacevich, R. Monroe, P. Filewood. Diagnosis S. amplus most closely resembles the Irian Jaya (Western New Guinea) species Otosaurus mimikanus (Boulenger 1914). It may be distinguished from O. mimikanus as follows — supranasal scales (usually 2 pairs* vs one pair); mid-body scale count (41-52 40); lamellae under fourth toe (22-26 vs 15); colour and pattern * See discussion of variation in the paratypes 98 MEMOIRS OF THE QUEENSLAND MUSEUM (see type description Boulenger, 1914). These features also readily distinguish S. amplus from all Australian species of Sphenomorphus. No other Australian Sphenomorphus species have supranasals and only one species group has more than 40 mid-body scales. Species of the S. quoyi group have 24-44 mid-body scales (Cogger, 1975) but in addition to lacking supranasals, these species are strikingly coloured olive-brown dorsally with prominent yellow-cream dorsolateral longitudinal stripes, a pattern which is never present in S. amplus. Description of Holotype SVL 100 mm; tail (original) 143; head width 15-2. Habit stout. Tip of snout-forelimb/ axilla-groin = 40/48 mm. Snout short, obtuse. Rostral broad, forming a suture with the single nasal and the anterior supranasal. Two supra- nasals; on each side the anterior ones form a median suture. Frontonasal approximately equally broad as long, bordered posteriorly by the large prefrontals which form a median suture. Frontal about three times as long as broad and narrow posteriorly; about as long as the frontoparietals and the interparietal together. Frontal in contact with the first and second supraocular. Supra- oculars 5, the anterior one largest and posterior one smallest. Supraciliaries 9. Frontoparietals and parietal distinct and about equal in size. Supralabials 7; fourth, fifth, and sixth under the eye. Ear opening large, oval about the same size as the eye, and with deeply set tympanum; no auricular lobules. Mid-body scales 46. Mid-dorsal scales about twice as large as mid-ventral and lateral scales. Two enlarged preanals. Limbs and digits moderately long; lamellae under fourth toe 23. Colour: In preservative, grey-brown dorsally and cream ventrally, with a conspicuous black patch above the forelimb. Labials and chin shields blue-grey, edged with narrow dark brown-black lines which are broader on the labials, giving the appearance of irregular blotching. Laterally, between the head and the forelimb, scales are also basically blue-grey. These have small dark spots which form a series of regular narrow lines. Dorsally between the neck and tail there is a series of irregular transverse light brown bands. Limbs also irregularly banded with light brown. On the tail the light bands are more numerous and regular. Colour pattern in life was basically similar but some colours were stronger e.g. olive-brown vs grey-brown dorsally; lemon vs cream. Variation in Paratypes SVL 40-1 15 mm. Tail 78-126 mm (only 12 specimens have original tails and most of these are juveniles). The following variation from the holotype description has been observed. Supranasal scales: In three specimens (J31666, J31664, J34023) there are two supranasals on each side. The anterior pair do not form a medium suture. One specimen (J31660) has two supra- nasals on one side and a single, large supranasal on the other. They form a median suture. In J34027 the supranasals are as in J31660 but they do not form a median suture. J31652 has only one large supranasal on both sides (presumably representing fused smaller scales) which form a median suture. In three juvenile specimens (J34026, J34061, J33984) the posterior suprana- sals fuse with the loreals. The anterior supranasals in all three specimens form median sutures. Frontonasal scale: J32785 (a juvenile) has fragmented frontonasal and right prefrontal scales. Colour: Differences are minor. Shading of basic colouring is more intense in some specimens and the head scales in juveniles are sharply lined in black. Mid-body scales: Counts in the 45 paratypes vary from 41-52 as follows. 41 (4 specimens, 9%); 42 (6, 13-5%); 43 (5, 11%); 44 (10, 22%); 45 (6, 13-5%); 46 (8, 18%); 48 (3, 7%); 51(1, 2%); 52 (2, 4%). Toe lamellae: Counts under the fourth toe vary from 22-26 as follows — 22 (7, 16%); 23 (21, 47%); 24 (11, 24%); 25 (5, 1 1%); 26 (1, 2%). Supraciliaries: These vary from 8-10 as follows — 8 (11, 24%); 9 (23, 52%); 10(11,24%). Supraoculars: There is only slight variation — 5 (42 specimens) or 6 (3). Distribution and Habitat S. amplus is known from only two areas in mid-eastern Queensland — Eungella National Park, west of Mackay and Conway State Forest, east of Proserpine. It is confined to closed forest. In the Eungella National Park it occurs between 300 and 900 m usually amongst and near granite outcrops near streams. It is a common element of the reptile fauna of this reserve. S. amplus is apparently less common in the Conway State Forest closed forests. Here it frequents buttresses and roots of large trees, especially Ficus sp. and rocks near creeks. It has not been found with S. luteilateralis which occurs in the same area but which is apparently confined to higher altitudes. COVACEVICH AND McDONALD; NEW RAINFOREST SKINKS 99 Notes Sphenomorphus amplus is an unusual species. In the Eungella National Park specimens were frequently observed at night sleeping on open rock faces, apparently for warmth. When handled or fighting they emit a sharp, loud ‘squeaking’ sound. The former feature has not been reported for any other Australian skinks and the latter is apparently an unusual feature amongst Australian skinks. It has been reported in some Tiliqua and in two other Australian species (Greer 1976) — Sphenomorphus murrayi, Tropidophorus queenslandiae. S. amplus and S. tenuis are sympatric in the Eungella National Park and are often found very close together, but apparently occupy different niches. S. amplus is invariably found on or very near rocks in moist areas, usually near creeks. S. tenuis, by comparison, favours rotting logs and the bases of targe trees, especially those with hollows and crevices. S. amplus is viviparous. A gravid female (J31652) carried five well developed young along with an undeveloped egg in the oviducts. The external morphological affinities of S. amplus lie with Southeast Asian and New Guinea species referred to Otosaurus (incorporating Parotosaurus) as listed by De Rooij (1915) and Smith (1937), but referred to Sphenomorphus by Loveridge (1948), Greer (1970), and Scott, Parker, and Menzies (1977). O. mimikanus, is described and well illustrated by Boulenger (1914). It shares the following features with S. amplus — large, robust form; high mid-body scale count; supranasals; colour pattern (see diagnosis for differences). Whether these similarities indicate a real relationship of this species (which is confined to mid-eastern Queensland) to Asian and New Guinea species, or whether they are purely superficial, remains to be determined. This question highlights current unresolved problems in the relationships between lygosomine skinks of the Australian region, particularly those currently referred broadly to Sphenomorphus. ACKNOWLEDGMENTS The Queensland National Parks and Wildlife Service gave permission to collect specimens in the Eungella National Park and the Australian Biological Resources Study financed field trips on which many of the specimens described here were collected. Helpful advice has been received from Dr A.E. Greer, Dr G.M. Storr, and Mr C.J. Limpus. The manuscript has been improved by Mr B. Campbell. LITERATURE CITED Boulenger, G.A., 1887. ‘Catalogue of the Lizards in the British Museum (Natural History)’, Voi. Ill; 1-575, 2nd edition, (British Museum Trustees: London). 1914. An annotated list of the batrachians and reptiles collected by the British Ornithologists Union Expedition and the Wollaston Expedition in Dutch New Guinea. Trans. Zool. Soc. Land. 20(5): 247-72, plates 27-30. Broadbent, J. and Clark, S., 1976. A fauna survey of the East Australian rainforests. Interim report. 132 pp. (Australian Museum: Sydney). Cogger, H.G., 1975, ‘Reptiles and Amphibians of Australia’. (Reed: Sydney). Covacevich, J., 1977. ‘Fauna of Eastern Australian Rainforests II: Preliminary report on sites surveyed by the Queensland Museum in southeastern and far northeastern Queensland, with additional results from sites surveyed previously in northeastern Queensland’, pp. 1-87, pis., 1-17, (Queensland Museum: Brisbane). De Rooij, N., 1915. ‘The reptiles of the Indo-Australian Archipelago. I. Lacertilia, Chelonia, Emydosauria. 1-334. (E.J. Brill: Leiden). Gray, J.E., 1845. ‘Catalogue of Lizards in the British Museum (Natural History)’. (British Museum: London). Greer, A.E., 1970. A sub-family classification of Scincid lizards. Bull. Mus. comp. Zool. Harv. 139: 151-83. 1976. A most successful invasion. The diversity of Australia’s skinks. Aust. nat. Hist. 18(12): 428-33. Loveridge, A., 1948. New Guinea reptiles and amphibians in the Museum of Comparative Zoology and United States National Museum. Bull. Mus. comp. Zool. Harv. 101: 305-430. Mittleman, M.B., 1952. A generic synopsis of the lizards of the sub-family Lygosominae. Smthson. Misc. collect. 117: 1-35. Queensland Museum, 1976. ‘Fauna of Eastern Australian Rainforest: Preliminary report on sites surveyed by the Queensland Museum in mid- eastern and northeastern Queensland’, pp. 1-78. (Queensland Museum; Brisbane). Scott, F., Parker, F., and Menzies, J.L, 1977. A checklist of the Amphibians and Reptiles of Papua New Guinea. Wildlife Publication 11 fh'. 1-18. (Department of Natural Resources: Papua New Guinea). Smith, M.A., 1937. A review of the genus Lygosoma (Scincidae: Reptilia) and its allies. Rec. Indian Mus. (Calcutta) 39: 213-34. Storr, G.M., 1967. The genus Sphenomorphus (Lacertilia: Scincidae) in Western Australia and the Northern Territory. J. Proc. R. Soc. West. Aust. 50: 10 - 20 . MEMOIRS OF THE QUEENSLAND MUSEUM PLATE 1 a. Sphenomorphus luteilateralis sp. nov., Mt William, Eungella National Park, via Mackay. b. Sphenomorphus amplus sp. nov.. Finch Hatton Gorge, Eungella National Park, via Mackay. COVACEVICH AND McDONALD: NEW RAINFOREST SKINKS Plate 1 Mem. Qd Mus. 20(1): 103-9, pi. 1 [1980], THE EMERALD MONITOR VARANVS PRASINUS (SCHLEGEL): AN ADDITION TO THE AUSTRALIAN MAINLAND HERPETOFAUNA Gregory V. Czechura Queensland Museum ABSTRACT The occurrence of the Emerald Monitor Varanus prasinus (Schlegel) on the Australian mainland, is confirmed by the collection of three specimens from the Mcllwraith Range area. Cape York Peninsula. These specimens are compared with the four currently recognized subspecies of this monitor and assigned tentatively to the nominate race. Difficulties encountered in the definition of these subspecies are briefly outlined. The Emerald Monitor {Varanus prasinus) is widely distributed throughout mainland Papua New Guinea, West Irian and adjacent island groups. This monitor may be distinguished from other Varanus species by the combination of the following features; round nostril, slightly com- pressed tail which lacks a median double keel, and the presence of transverse supraoculars (Cogger 1975; De Rooij 1915; Mertens 1942, 1959). Mertens (1959), in the most recent treatment of Indo- Australian Varanus species, recognises four subspecies of V. prasinus, all from New Guinea and adjacent islands: 1. Varanus prasinus prasinus (Schlegel): mainland New Guinea, Misol, Salawatti, Goodenough, ‘islands of Torres Strait’ and the d’Entrecasteaux Archipelago. 2. Varanus prasinus beccarii (Doria): Aru Islands, and possibly from Fak Fak (West Irian). 3. Varanus prasinus bogerti Mertens: the d’Entrecasteaux Archipelago. 4. Varanus prasinus kordensis (A.B. Meyer): Biak, northwest New Guinea. Australian records of V. p. prasinus (from the Torres Strait only) are based on Gunther (1877, 1879) and Boulenger (1885) who record Varanus prasinus as occurring on islands of Torres Strait, Cornwallis (=Dauan), and Murray Islands. Cogger (1975) notes unconfirmed reports of this monitor from northern Cape York Peninsula but. until now, there have been no specimen records of Varanus prasinus from mainland Australia. In August 1978, an unusual dark specimen of Varanus was collected by the author at Buthen Buthen, east of Coen, far northern Queensland. Subsequently, two additional similar specimens were collected in the same area by officers of the National Parks and Wildlife Service. Examination of these specimens confirms the presence of V. prasinus on the Australian mainland. Materials and Methods The three Cape York Peninsula specimens were compared with specimens of V. prasinus in the Queensland Museum (QM) and Australian Museum (AM) collections. In addition, the type of Varanus prasinus bogerti in the American Museum of Natural History (AMNH 41639), was also examined and compared with these specimens. The following measurements were recorded in millimeters; snout-vent length (SVL), tail length (TL), head length (HL), head depth (HD), head width (HW), eye-naris distance (EN), internarial distance (IN), and snout to anterior edge of naris (SN). The values are expressed by the mean value followed by the range in parentheses. The term ‘New Guinea’ used below refers to the whole island exclusive of political subdivisions. 104 MEMOIRS OF THE QUEENSLAND MUSEUM Description of Australian Mainland Secimensof Varanus prasinus SVL 237-6 (214-250); TL 479-6 (425-524); HL 42-0 (36-0-49-7); HD 20-6 (16-2-23-9); HW 23-7 (19-2-26-7); EN 13-5 (12-6-14-2); IN 8-3 (7 4-8-8); SN 10-2(9-5-11-2). Head long and narrow, about twice as long as broad and 1-7 times as long as high. Snout rounded in profile. Canthus rostralis indistinct, rounded and slightly elevated above the nostril which is closer to tip of snout than eye. Nostril rounded, inclined posteriorly in lateral aspect and visable from above (see Plate 1). Ear opening large, oval and more or less vertically orientated. Dorsal head scales variable in size and shape, largest between snout and interorbital region (Plate 1). Individual scales well demarcated by deep suture lines, not forming a continuous layer. Individual scales may be pitted, weakly tuber- culate or smooth, with structures demarcated by black pigmentation, A series of about five transversely enlarged supraoculars, some or most of which may be divided. Occipital scales similar, but smaller in size and less smooth. 35-36 scales around head from rictus to rictus. Temporal scales, smaller than the dorsal head scales; individual temporal scales subequal, and elliptical in shape, forming oblique rows (Plate 1). Nuchal scales transversely enlarged, rounded in profile. Neck scales becoming longer than wide posterior- ly. Low keels present on last third of neck. All neck scales rounded in profile. Anterior chin shields little differentiated except for enlarged pair immediately posterior to mental. Mental groove weak. Dorsal body scales longitudinally keeled, rectangular or slightly rounded in profile; in 90-99 scale rows at midbody. Lateral scales subequal. Ventral scales with apical pits, in 86-89 rows from gular fold to hindlimb insertion. Limbs well developed with strong claws. Palms and soles black. Approximate- ly 35-38 smooth lamellae under 4th toe, (except for 6-10 distal lamellae) divided and gradually merging with the similarly shaped scales of sole. Caudal scales keeled, subequal, and in regular annuli to tail tip. Proximal caudal scales with apical pits, each being replaced distally by a short spine. Plate 1 shows QM J31566 in dorsal view. Colouration: All three specimens are predominantly black dorsally and laterally on both body and limbs. The area from the tip of snout to the interorbital area is light bluish green, with black suture lines (except on snout itselQ. A short yellowish bar is present above the ear opening and may extend forward towards the posterior corner of the eye. individual body scales may have tiny yellowish spots, which when viewed from a distance, combine to form indistinct chevrons across the body. These indistinct chevrons are separated by narrow areas of black. These spots are more noticeable after preservation than in life. Yellowish colour is more noticeable on the tail where thin bands are present in middle of each annuli. Ventral surface of body and limbs pale yellowish-green. Comparison with Recognized Subspecies Using Mertens (1942, 1952) data, these specimens may be distinguished from V. p. beccarii (Doria) and V. p. kordensis (A.B. Meyer) by respective colouration differences (dark dorsum/light venter vs dark dorsum and venter vs green and black network pattern on dorsum/light venter). The Cape York specimens also lack the prominent neek keels and narrow, pointed snouts of these two subspecies. Comparison with the type of V. p. bogerti Mertens (AMNH 41369) shows that the three Australian specimens differ from this subspecies on the basis of more rounded snouts, larger dorsal head scales, weakly keeled neck scales and possession of a regular, oblique series of temporal scales. Separation from the nominate race, V. p. prasinus (Sehlegel), is difficult. Only two consistent differenees were apparent when comparing New Guinea V. p. prasinus with the Cape York specimens. These differences being the black (as opposed to green) dorsal colouration and the more prominent body keels of the Australian specimens. Ventral seale counts overlap (New Guinea; 71-86, N=9: Cape York; 86-89, N = 3), while midbody scale counts for Cape York Peninsula specimens (90-99: N = 3) fall within the range of variation of New Guinea material (74-105: N=9). Minor differences were detected with respect to the degree of ‘pitting’ on head scales, depth of dorsal head scale sutures and the appearance of nuchal and neck scales. However, as New Guinea V. p. prasinus showed variation among themselves in regard to these features, their diagnostic value was considered questionable. Field Observations The Buthen Buthen specimen (QM J31566) was collected in semi-deciduous mesophyll vine forest (altitude 60-100 m). It was discovered climbing through the crowns of secondary story vegetation, which formed a more or less continuous layer three to four meters above ground level. The Leo Creek Road specimen (QM J35450) was collected from the outer surface of CZECHURA: EMERALD MONITOR 105 the tree canopy in deciduous vine thicket (altitude 280-300 m). It was estimated to be some fifteen meters above ground level (J. Winter, pers. comm.). The specimen from Lankelly Creek (QM J35451), was collected in Eucalyptus-dovcnndiiQd open forest within 50 meters of deciduous vine thicket (altitude 520—540 m). It was first seen on the ground, then collected from a standing eucalypt (J. Winter, pers. comm.). Mertens (1942, 1950) notes that the tail tip in both V. p. prasinus and V. p. bogerti is ‘somewhat prehensile’. The prehensile nature of the tail tip of QM J31566 was readily apparent before collection and prior to preservation. In addition to the prehensile tail-tip, climbing in these monitors appears to be facilitated by the structure of the surface tissue on the soles of fore and hind feet. These surfaces are covered by soft black tissue, which feels sticky on contact and appears to give additional support to the climbing animal. An investigation of the structure of this tissue is in progress. The localities from where the species were collected are shown in Fig. 1. Fig. 1: Distribution of Varanus prasinus in Queensland. Discussion Although Cape York Peninsula specimens are distinct from V. p. prasinus in colour and body keels, there are no apparent consistent differences in the structure and shape of head scales and the condition of the nuchal and neck scales. In the absence of convincing evidence to the contrary, it is felt that the Cape York population should be tentatively assigned to the nominate race of this monitor (Varanus prasinus prasinus). Two considerations have largely influenced this decision. The first involves the rather confused taxonomy of prasinus subspecies. Mertens (1959), while recognizing the four subspecies, drew attention to the existence of a number of problematical populations of this monitor; kordensis-Wkc individuals from the mainland and small island groups; beccarii-\\k& individuals from the south-west mainland of New Guinea; and a bogerti ‘relative’ from the d’Entrecasteau group. A specimen (QM J1190) apparently collected from Rossel Island (Louisiade Archipelago) presents a further complication. V. prasinus has not previously been reported from this island group, and this specimen was not examined by Mertens. It is not possible to satisfactorily assign it to any existing subspecies using available literature or specimens. The second consideration concerns the apparent trend of V. prasinus to become increasingly melanistic in outlying populations. The greatest degree of melanism is found in the insular forms V. p. bogerti and V. p. beccarii, a lesser degree in the ‘kordensis’ pattern and the least in predominantly green V. p. prasinus. Cape York specimens appear intermediate between the fully melanistic and ‘kordensis’ forms. Until a suitable series of this monitor from southern New Guinea, Torres Strait and northern Cape York Peninsula is available for investigation, the possibility of clinical variation along this route precludes the description of a Queensland subspecies. Specimens Examined Varanus prasinus bogerti: AMNH 41693, Fergusson Island, d’Entrecasteax Archipelago. Varanus prasinus prasinus: New Guinea — QM J2218, St. Joseph’s River; AM RI205, no data; R6234, ‘New Guinea’; R11492, R12533, Bulolo Valley; R16760, mouth of Oriomo River near Daru; R17963, ‘Papua’; R24343, Fly River; R64790, Bristow Island near Daru; R65300, Daru Island, Western Province. Cape York — QM J31566, Buthen Buthen, Nesbit River; J35450, Leo Creek Road, 17 km NE. Mt Croll; J35451, Lankelly Creek, 10 km NE. Coen. Varanus prasinus (subsp.?): QM J1190, Rossel or Russell Island New Guinea. ACKNOWLEDGMENTS Two of the specimens on which this work is based were collected by Mr R. Atherton and Mr K. McDonald of the National Parks and Wildlife Service while they were conducting fauna surveys 106 MEMOIRS OF THE QUEENSLAND MUSEUM sponsored by the Australian National Parks and Wildlife Service and the Department of Environment, Housing and Community Develop- ment. Dr J. Winter co-ordinator of the surveys donated these two specimens to the Queensland Museum for Study. Dr R. Zweifel (American Museum of Natural History) and Dr A. Greer (Australian Museum) made specimens in their care available for this work. Assistance in translating Mertens’ work has been received from Ms S. Maywald, Mr B. Hoffer, and Mr P. Davie. Helpful suggestions in preparing the manuscript were made by Mr B. Campbell and Ms J. Covacevich (Queensland Museum), Dr A. Greer (Australian Museum), Mr G. Ingram (Queens- land Museum), Dr G. Storr (Western Australian Museum), and Dr J. Winter (National Parks and Wildlife Service). I wish to thank Ms D. Fitton for typing the manuscript. LITERATURE CITED Boulenger, G.S., 1855. ‘Catalogue of the Lizards in the British Museum (Natural History)’. Vol. 2. (British Museum Natural History: London). Cogger, H.G., 1975. ‘Reptiles and Amphibians of Australia’. (Reed: Sydney). De Rooij, N., 1915. ‘The Reptiles of the Indo-Australian Archipelago’, (Brill: Leiden). Gunther, A., 1877. Description of three new species of lizards from islands of Torres Straits. Ann. Mag. nat. Hist. (4) 19: 413-5. 1879. Notice of a collection of reptiles from islands of Torres Straits. Ann. Mag. nat. Hist. (5) 3: 84-7. Mertens, R., 1942. Die Familie der Warane (Varanidae). Abh. senckenberg. naturf. Ges. 466: 235-391. 1950. Notes on some Indo-Australian monitors (Sauria, Varanidae). Amer. Mus. Novit. 1456: 1-7. 1959. Liste der Warane Asiens und der Indo- australischen Inselwelt mit systematischen Bemer- kungen. Senck. biol. 40: 221-40. MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 Varanus prasinus (QM J31566) Buthen Buthen, Nesbit River, Cape York Peninsula (scales in centimetres). Top left: Lateral aspect of head. Top right: Dorsal aspect of head. Bottom: Dorsal view of body. CZECHURA: EMERALD MONITOR Plate 1 Mem. Qd Mus. 20(1): 111-19, pi. 1 [1980]. A NEW FROG OF THE GENUS TAUDACTYLUS {MYOBATRACHIDAE) FROM MID-EASTERN QUEENSLAND WITH NOTES ON THE OTHER SPECIES OF THE GENUS Glen Ingram Queensland Museum ABSTRACT Taudactylus liemi sp. nov. is described from montane rainforest mid-eastern Queensland. New data on the distribution and natural history of the other species of the genus {T. acutirostris, T. diurnis, T. eungellensis and T. rheophilus) are presented. A phylogeny has been devised for Taudactylus and aspects of the biogeography of the genus are discussed. In 1975, the Australian Biological Resources Survey provided funds to enable the Australian and Queensland Museums to survey frogs and reptiles in rainforest sites of mid-eastern Queensland, as part of an overall survey of the rainforests of Queensland. The preliminary results of these surveys have been presented in Broadbent and Clark (1976), Queensland Museum (1976) and Covacevich (1977). A large series of specimens of a new species was collected from two sites (Mt William and Crediton). These frogs were clearly referable to the genus Taudactylus because of the presence of T-shaped terminal phalanges (Fig. ID), but were distinctly different from T. eungellensis described from the same area by Liem and Hosmer (1973). The new species is here described as T. liemi, in recognition of the contribution to herpetology made by Dr David Liem. Measurements are in millimetres and ratios are expressed as percentages. Specimens are housed in the Queensland Museum (QM), and the Australian Museum (AM). Abbreviations follow Liem and Ingram (1977). Taudactylus liemi sp. nov. (Figs 1, Plate IB) Material Examined Holotype, Adult female, QM J32625, Crediton, ME.Q. (2P 12’S, 148° 33’E), 15-22 April 1975, collected by J. Covacevich, P. Filewood and R. Monroe. Paratypes. am R47499-505, Mt William, 21-26 April, 1975, P. Webber; AM R47831, Eungella, 1975, P. Webber; QM J34420 18 km N. of Dalrymple Heights, December 1978, G.J. Ingram; QM J31 5 1 5-8, Dalrymple Heights, 3 July, 1974, G. Czechura; QM J326 18-24 (J32618 cleared specimen), J32626-33, 332660^8, J32694, same data as holotype. Diagnosis: Differs from T. diurnis and T. eungellensis by very small discs on fingers and toes; from T. acutirostris by the lack of dorsolateral skinfolds and in snout shape (in profile, rounded vy wedge shaped and curved upwards); and from T. rheophilus by the lack of extensive brown mottling ventrally, and by dorsal markings (presence of a dark triangle between the eyes and a dark lyre on the back). T. rheophilus is also more robust and has a larger HW/SWL. Liem and Hosmer (1973) give 37-42 for the latter whereas the range for T. liemi is 26-36. Description of Holotype: SVL 28- 1, TL 11-7, TL/SVL 42, HW 8 4, HW/SVL 30, ED 3-1, ED/HW 37, EN 2-5, IN 3-4, EN/IN 74. Dorsal aspect of snout blunt, acuminate, rounded in profile. Loreal region sloping. Canthus rostralis distinct, curving in from eye and then out to nose, then converging anteriorly to form an acuminate snout. Pupil horizontal and oval shaped. Tympanum concealed. Tongue hinged in front, widest posteriorly where it is rounded; narrow and stright anteriorly. Vomerine teeth absent. Fingers unwebbed, slightly expanded distally. Length of Fig. 1; Taudactylus liemi. A, Dorsal view; B, Ventral aspect of foot; C, Ventral aspect of hand; D, T-shaped terminal phalange of 4th toe; E, Side of head. MEMOIRS OF THE QUEENSLAND MUSEUM INGRAM: A NEW TAUDACTYLUS 113 fingers shortest to longest 1 -2-4-3. Large rounded tubercles at base of fingers; rounded outer palmer tubercle twice the size of oval inner tuberele. Toes not webbed but with distinct toe fringes, slightly expanded distally. Length of toes from shortest to longest 1-2-5-3-4. Low rounded tubercles at base of toes with smaller tubercles on joint above on 3rd, 4th and 5th toe; a small oval inner metatarsal tubercle. Skin smooth. Cloacal opening directed posteriorly at mid-level of thighs. Dorsal ground colour grey-brown; a black triangle between eyes, with the hypotenuse stretching from eye to eye and the apex directed backwards; a black lyre marking on neck and back; ‘V’ marking in front of hind legs with apex directed forwards; two cross bars on forearm, 8 on hind legs; fingers and toes barred. Laterally, ground colour grey-brown; a dark line from snout through eye and curving down into loreal region; side of face dark brown, lips barred; an indistinct dark upper lateral mark from above forelegs and fading at midbody; a dark blotch at beginning of foreleg and on knees. Posterior surface of thighs brown finely speckled with dirty yellow; black markings on either side of cloaca; heels dark. Ventrally, cream with fine brown speckling on legs. Description of Paratypes: SVL 20-9-29-3 (N = 36, mean 26-1). HW 7 2-9-5 (N = 36, mean 8- 4). HW/SVL 26-36 (N = 36, mean 32-0). TL 9- 6-12-5 (N = 36, mean 11-5). TL/SVL 36-48 (N = 36, mean 44-2). ED 2-2-3-5 (N = 36, mean 2-9). ED/HW 27-44 (N = 36, mean 35-1). EN 2-0-2-5 (N = 36, mean 2-2). IN 2-8-4 0 (N = 36, mean 3-4). EN/IN 55-77 (N = 36, mean 65-5). Vocal sacs present in males. Dorsal and lateral colouring varies from light brown to dark brown when ground colour very dark the markings can be difficult to discern, and when light, they can be faint. Lyre marking on back can extend forward joining up with the triangle between the eye and it can extend backwards breaking up into blotches in front of the ‘V’ marking in front of the cloaca. Apex of this ‘V’ is often missing. The hidden tympanum may be defined by a light patch. The upper lateral stripe in some specimens extends backwards and down to the inguinal region. Ventrally there may be more intense brown speckling. Nasal bones narrow, widely separated and not touching sphenethmoid or the maxillary; frontoparietal fontanelle moderately large; zygomatic rami of squamosal as long as otic rami; omosternum present; sternum rounded; terminal phalanges of fingers and toes T-shaped (Fig. 2B). Habitat: Rocky streams and their environs in montane rainforest. Distribution: Eungella area west of Mackay, ME.Q., from Mt. William in the north to Crediton in the south. Field notes: T. liemi is a secretive frog. Males call on land under rocks along the side of rocky streams. In December 1978, they called all day but more commonly at night. G. Czechura (pers. comm.) noted in July 1974 that they were calling in large numbers. There appeared to be no difference between the numbers calling day or night. He found them mostly under rocks during the day and under roots and in the mouth of crayfish burrows during the night. Amplexus was noted as inguinal. In August 1976, he heard no vocalization. At Crediton, in April 1975, individuals were caught inside the rolled up ends of palm fronds {Archontophoenix sp.) in an area where no running streams were observed (J. Covacevich, pers. comm.). A number of these individuals were gravid females. The trunk ends of these palm fronds have moist micro-environments inside and are a catchment for water during rain. Gravid females carry 34-51 large eggs measuring 1 -7-2-5 mm. Egg masses and tadpoles have not been identified. Species synchronosympatric with T. liemi were Adelotus brevis, Taudactylus eungellensis, Mixophyes fasciolatus, Litoria chloris and L. lesueurii. Call: To the ear the call is a short series of ‘tinks’. A sound spectograph of this call is given in Plate 1(B). This is based on a recording by Chris Corben at Dalrymple Heights, Eungella, ME.Q., in January 1976. The energy of the call is concentrated between 3000-4000 HZ, and has a duration of 41 1 milliseconds. It is composed of 4 pulses, with an individual duration of 8 1 milliseconds. The number of pulses however, may vary between 1 and 6 but 3 predominates. An individual calls on the average 27 times per minute. The call is easily distinguished from the multipulsed, sharper ‘ting’ call of T. rheophilus, the long multipulsed ‘eek, eek, eek . . .’ call of T. acutirostris. and the soft, short ‘eek eek eek’ call of T. diurnis. NOTES ON THE OTHER SPECIES OF TAUDACTYLUS Frogs of the genus Taudactylus, Straughan and Lee (1966), occur only in isolated montane 114 MEMOIRS OF THE QUEENSLAND MUSEUM rainforest of eastern Queensland, south of Mt Hartley, some 20 km south of Cooktown, They are a conspicuous element of the rainforest fauna where ever they occur, and are almost invariably associated with cool, clear, fast-flowing rocky mountain streams. Three species are sun-loving and diurnal, an unusual habit amongst frogs. Two species have high pitched, very distinct, bell-like calls. The most recent work on this genus was undertaken by Liem and Hosmer (1973). Since that study, new data on the natural history and distribution of the four previously known species have been compiled. These are summarized below. T. acutirostris (Andersson 1916) Distribution: Great Dividing Range, north- eastern Queensland from Ml Hartley, Bloomfield area in the north, to Tully Falls in the south. Call; To the ear, the call is a series of ‘eek-eek-eek . . sometimes ending or beginning with sharp metallic notes. Plate 1(c) is a sound spectograph of part of a call based on a recording by the author at Mt Lewis on 7 Nov., 1975. The call has a dominant frequency of 3000 HZ and the pulses have an individual duration 63 milliseconds. The call is usually from 4-6 seconds long with 17-25 pulses per call. An individual calls on the average 7 times per minute. Males form a chorus. Field Notes: T. acutirostris is mainly active during the day especially on sunny days when they will often ‘bask’ in the sun. Males call exposed on rocks and will interrupt calling when clouds move in front of the sun. Typically, after a period of exposure on a rock in the sun, individuals move off to forage along the sides of creeks and on the rainforest floor nearby. When distrubed they show no hesitation in jumping into water, be it a still pool or a waterfall. In still pools they lie exposed on the bottom amongst the leaf litter or rocks for several minutes before resurfacing, COMMENTS: There is much confusion in the literature regarding the call of this species. Clyne (1969) calls it the Tinker Frog and gives the call as a series of metallic ‘links’. Liem and Hosmer (1973) describe the call as a series of sharp tapping sounds repeated 3-4 times in quick succession. Neither have been verified by further field observations. Males typically call in a chorus sometimes giving sharp metallic notes at the beginning and end of calls but especially during male-male aggression. The typical call of T. acutirostris has been recorded at the upper reaches of Mulgrave River and Charmillan Creek, Ravenshoe (G. Czechura pers. comm.). T. eungellensis Liem and Hosmer, 1973 Distribution; In the ranges west of Mackay, ME.Q., from Clark Range in the north (A. Greer pers. comm.) to Finch Hatton Gorge and Credition in the south. Call: Liem and Hosmer (1973) record the call of this species as a high pitched metallic tinkering noise, like a little hammer tapping on metal repeated 4-5 times in quick succession. This has not been verified either by my own field observations or by C. Corben and G. Czechura (pers. comms.). Indeed we have failed so far to discover if it calls at all. T. liemi is synchronosympatric with T. eungellensis and was at that time undiscovered. It appears Liem and Hosmer have confused the two, and their description may apply to the T. liemi call. It would be surprising if a frog without vocal sacs could produce a high pitched, sharp tinkering call. Field Notes: T. eungellensis is similar in behaviour to T. acutirostris except in the case of calling males. T. rheophilus Liem and Hosmer, 1973 Distribution; Great Dividing Range, NE. Q., from Thornton Peak, Daintree area, in the north to Mt Lewis in the south. Call: To the ear, the call is a series of sharp metallic ‘link-tink-tink Plate 1(A) is a sound spectograph of the call based on a recording by the author at Mt Lewis on 6 December, 1975. It shows a dominant frequency of 5500 HZ and a lower harmonic containing nearly as much energy around 2750 HZ. The pulses have an individual duration of 31 milliseconds. The call is usually between 4-5 seconds long with 11-15 pulses per call. On the average an individual will call 5 times per minute. Field Notes: This species is a very secretive frog. It calls day and night but mainly during the day. Male calling sites are usually under rocks or roots and individuals may be partly in water. Calls form a chorus. INGRAM: A NEW TAUDACTYLUS 115 T. diurnis Straughan and Lee, 1966 Distribution: Conondale Ranges (Czechura 1975) and Blackall Ranges in the north to Mt Nebo, D’Aguilar Range in the south. Call: As Liem and Hosmer (1973) have noted, males do not have vocal sacs but can vocalize. This is especially evident in captivity. To the ear, the call is soft ‘eek-eek’ sometimes with sharper notes, and is reminiscent of T. acutirostris. In the field, the call is heard especially during male-male aggressive encounters. No breeding choruses have been noted. Field Notes: T. diurnis and T. eungellensis are very similar in behaviour. Individuals forage into late evening. PHYLOGENY The following characters were used in the preparation of Table 1, and the construction of the cladogram (Fig. 2). These show the relationships between the five species of Taudactylus. The data for species other than T. liemi are taken from Liem and Hosmer (1973). (1) represents the apomorphic character-state and (0) the plesiomor- phic character-state. TABLE 1: Character-states of the Species of Taudactylus Species I 2 Characters 3 4 5 6 7 T. diurnis 0 0 1 1 1 1 I T. eungellensis 0 0 I 1 I 1 1 T. acutirostris 1 1 0 0 0 0 1 T. rheophilus I 1 0 0 0 0 1 T. liemi 1 1 0 0 0 0 1 0 = plesiomorphic, 1 = apomorphic. 1. Exposure of Frontoparietal Fonta- NELLE: Following the reasoning of Lynch (1978), the lack of an exposed fontanelle, or the presence of a very small fontanelle (1), is considered to be apomorphic and an extensive fontanelle (0), plesiomorphic. 2. Nasals Contacting the Maxillary: Large nasals contacting the maxillary (0) are considered to be plesiomorphic, and small nasals not contacting (1), apomorphic. 3. Omosternum: Liem and Hosmer (1973) are followed in considering the lack of an omosternum (1) apomorphic, and presence (0), plesiomorphic. 4. Vocal SacS: Because most male frogs possess vocal sacs, the absence of sacs (1) is considered apomorphic, and presence (0), plesiomorphic. 5. Digital Discs: Liem and Hosmer (1973) are followed in considering broad digital discs (1) as apomorphic, and narrow discs (0) as plesiomorphic. 6. SUBARTICULAR TUBERCLES OF FINGERS: Liem and Hosmer (1973) regarded absence of these tubercles (1) as apomorphic, and their presence (0) as plesiomorphic. 7. T-shaped Terminal Phalanges: The T-shaped terminal phalanges is unique to Taudactylus and is thereby considered auto- apomorphic (1) for the genus. From the cladogram in Figure 2, it can be seen that there are two sister groups in Taudactylus — the T. diurnis complex (T. diurnis and T. eungellensis) and the T. acutirostris complex (T. acutirostris, T. rheophilus, and T. Uemi). The latter group is presented as a trichotomy. Dichotomies are to be preferred; however, unless they are supported by characters, they are invalid (Platnick and Shadab 1978). The T. diurnis complex is regarded as apomorphic because it contains more apomorphies than its plesiomorphic sister group, the T. acutirostris (Table 1). Fig. 2: Cladogram of the sf>ecies of Taudactylus. Td = T. diurnis, Te = T. eungellensis, Ta = T. acutirostris, Tr = T. rheophilus. Tl = T. liemi. 116 MEMOIRS OF THE QUEENSLAND MUSEUM BIOGEOGRAPHY Hennig (1966) has stated that sister groups should be allopalric and that the most primitive sister groups remain near the centre of origin of the overall group. The T. acutirostris complex is the pleisiomorphic sister group, and is centred in northeast Queensland. Thus, it would seem reasonable to assume that this area was the centre of origin of the genus Taudactylus. Probably, the T. diurnis complex evolved from a T. acutirostris complex-like ancestor that invaded the rainforests of southern Queensland and became isolated there during rainforest contractions. Later, as the rainforest once again extended, the T. diurnis- eungellensis ancestral group may have reinvaded northern Queensland, and again become isolated by subsequent contractions, such that a northern group (T. eungellensis), and a southern group {T. diurnis) were formed. The ancestor of T. liemi may have been similarly isolated by rainforest contractions in mid-east Queensland. It is difficult to reconstruct the evolution T. acutirostris and T. rheophilus with an allopatric model. The latter is synchronosympatric with the former, although T. acutirostris is distributed a little further north and south, and occurs at lower altitudes. T. rheophilus and T. liemi are restricted to areas that acted as rainforest refugia (Mt Lewis, Thornton Peak, Eungella — Webb and Tracey, in press) during dry periods of the Pleistocene. The other species, although having important refugia within their range, have colonized rainforest areas nearby. No species of Taudactylus occur in New Guinea or in rainforests north of the Daintree River ‘block’. This is not surprising as it is unlikely that the high altitude, clear cool, fast-flowing rainforest streams necessary for Taudactylus existed during the dry Pleistocene periods when land connections were extant (Kikkawa, Monteith, and Ingram, in press; Covacevich and Ingram, in press). ACKNOWLEDGMENTS I thank J. Covacevich, C. Corben, G. Czechura, P. Davie, and R. Raven for their assistance with this paper. A. Greer loaned specimens in his care at the Australian Museum. S. Sands and T. Low prepared the drawings. LITERATURE CITED Andersson, L.G., 1916. Results of Dr E. Mjoberg’s Swedish scientific expeditions to Australia 1910-1913. IX Batrachians from Queensland. K. Svenska VetenskAkad. Handl. 52(9): 1-20. Broadbent, J. and Clark, S., 1976. ‘A fauna survey of the East Australian rainforests. Interim report.’ (Australian Museum: Sydney). Covacevich, J., (Ed), 1977. ‘Fauna of Eastern Australian Rainforests II: Preliminary report on sites surveyed by the Queensland Museum in southeastern and far northeastern Queensland, with additional results from sites surveyed previously in northeastern Queensland’. (Queensland Museum: Brisbane). Clyne, D., 1969. ‘Australian frogs’. (Lansdowne Press: Melbourne). Covacevich, J. and Ingram, G., (in press). The endemic frogs and reptiles of Cape York Peninsula. Proc. R. Soc. Qd. Czechura, G., 1975. Notes on the frog fauna of Conondale Range, southeast Queensland. Herpetofauna 7(2); 1-A. Hennig, W., 1968. ‘Phylogenetic syslematics’. (Univer- sity of Illinois Press: Illinois). Kikkawa, J., Monteith, G. and Ingram, G., (in press). Cape York Peninsula: the major region of faunal interchange. In Keast, A. (Ed.) ‘Ecological Biogeography in Australia’ (Junk: The Hague). Liem, D.S. and Hosmer, W., 1973. Frogs of the genus Taudactylus with descriptions of two new species (Anura: Leptodaclylidae) Mem. Qd Mus. 16 : 435-57. Liem, D.S. and Ingram, G.J., 1977. Two new species of frogs (Anura: Myobatrachidae, Pelodryadidae) from Queensland and New South Wales. Vic. Nat. 94(6): 255-62. Lynch, J.D., 1978. A re-assessment of the telematobiine leptodactylid frogs from Patagonia. Occas. Pap. Mus. Nat. Hist. Univ. Kansas 72: 1-57. Platnick, N.I. and Shadab, M.U., 1978. A review of the spider genus Anapis (Araneae, Anapidae), with a dual cladistic analysis Amer. Mus. Novit. 2663: 1-23. Queensland Museum, 1976, ‘Fauna of Eastern Australian Rainforest: Preliminary report on sites surveyed by the Queensland Museum in mid- eastern and northeastern Queensland’. (Queensland Museum: Brisbane). Straughan, I. and Lee, A.K., 1966. A new genus of leptodactylid frog from Queensland. Proc. R. Soc. Qd 77: 63-6. Webb, L. and Tracey, G., (in press). The rainforests. In A. Keast (Ed.) ‘Ecological Biogeography in Australia’ (Junk: The Hague). MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 Sonographs of the calls of Taudactylus. A: T. rheophilus B: T. liemi C: T. acutirostris KHz INGRAM; A NEW TAUDACTYLUS Plate 1 Mem. Qd Mus. 20(1): 121-3, pi. I [1980] RADIOGRAPHIC PATHOLOGY IN MANDIBLES OF ANTECHINOMYS LANIGER FROM HORSESHOE CAVE, WESTERN AUSTRALIA W.A. Miller Department of Oral Biology, State University of New York at Buffalo ABSTRACT Jaw fragments of Antechinomys laniger (Gould) were radiographed. It was found that the cave population contained mature animals and a number of older individuals with dentitions at the end of their functional period due to excessive wear. A number of mandibles of Antechinomys laniger (see Archer 1977 for revision of the genus) collected by Dr M. Archer from Horseshoe Cave (N59), on the Western Australian Nullarbor, in February 1970, have been radiographed. One is Queensland Museum number J23103; the others have no acession numbers. Jaw fragments were radiographed from both sides using fine grain photographic film and a modified dental radiographic apparatus with the aluminium filter removed (Miller and Radnor 1970). Results Judging by the presence or absence of teeth, and the wear on the occlusal surface, individuals were either mature adults or old adults. Mature Animals: The jaw fragments mostly contained the post-canine dentition: P ]_3 and M ,_4 were present. When viewed bucco-lingually all teeth appeared to be two rooted; the premolars slightly shorter than the molars. The premolar cusp pattern was also simpler. With increasing occlusal tooth wear, presum- ably associated with increase in age, it may be observed (Plate 1) that the cheek teeth have migrated occlusally so that the distance between root-apex and lower border of mandible has increased. There is also an increase in the distance between the bifurcation of the root and the tip of the alveolar crest in some teeth. The roots, however, appear to be otherwise correctly seated in their bony sockets so that one may assume this to be biological rather than artefactual, but probably pathological. The most likely cause is that it is due to the loss of bone, associated with periodontal inflammation, during life. The horizontal distance between the distal surface of M 4 and the ascending ramus of the mandible was larger in animals with the most wear. This implies that some type of physiological mesial drift has taken place with age. This is a process which is more marked in macropodids, namely molar migration. Old Adult Animals: Partially edentulous jaws were of two kinds. One type had lost teeth almost certainly post mortem. The sockets were clearly visible on the radiographs. The remaining teeth were incompletely embedded in bone, and this would facilitate post mortem loss of teeth. The sockets of all teeth were at a comparable level however, implying similarities in the teeth and their supporting structures ante mortem. One jaw (J23103) had considerable ante mortem tooth loss (Plate lA). It so happened that C] and a fragment of I 3 were also present. Only three molars were still in the jaw. M 4 was severely worn and extruded to a certain degree. Bone loss around the distal root and bifurcation was considerable, with cratering behind the tooth and an inflammatory sclerosis in the bone behind M 4 as indicated by increased radiodensity. M 3 and M 2 were severely worn but their bony support within normal limits. M, was lost but its socket was still visible although abnormal in outline. The mesial part was much wider than normal and rounded in outline and the distal slightly so. A dark ova! area can be seen on the radiograph related to the mesial root which on the bone was an open buccal sinus 122 MEMOIRS OF THE QUEENSLAND MUSEUM about 2 mm in diameter. This should be termed a sinus and not a cloaca, the latter term is usually associated with the much more severe inflamma- tory changes and bone necrosis of osteomyelitis when the marrow cavity is also severely effected by the inflammatory process and the periosteum has been stripped from the bone, due to the presence of migrating pus, to subsequently give rise to a new bone known as an involucrum. The remaining cheek teeth P ,_3 were missing but odd irregularities in bone density indicate that they had been present but were lost sometime ante mortem. The most distal missing tooth is identified as M] because no space is found behind the terminal molars to accommodate another tooth, and the socket depth indicates a root length similar to the other molar teeth. When the tooth row is complete (P 1 -M 4 ), the premolar root length is distinctly shorter than that of the molars. This may be observed on other specimens on the radiograph. Discussion Although inflammatory jaw bone diseases are relatively common in animals in captivity they are rare in wild animals (Colyer 1936, Potkay 1977). Such is the importance of the dentition that a defect such as these usually results in the death of the animal. This cave population (Horseshoe Cave N59) contained a number of older individuals with dentitions at the end of their functional period due to excessive wear and one with considerable tooth loss, one quadrant being almost nonfunctional. It is most unusual to find populations with this degree of nonfunction. One might ask whether these were the remains of old individuals killed by some predator and subsequently dumped or maybe stored in the cave. Associated with this it is likely that the oldest specimens were the longer lived females at the end of the breeding season. From their dentitions alone they would be incapable of much extension of their life span. This is relatively recent material (pers. comm. M. Archer) so one might expect the post cranial skeleton to be present if natural death had occurred. The finding of mostly dentary fragments also argues for the effects of a predator on exhausted animals at the end of the short life span. ACKNOWLEDGMENTS I must thank Dr M. Archer for making this material available for study. Travel support in part by a Grant in Aid from Melbourne University and Departmental funds, Dental Surgery and Medicine, School of Dental Science, University of Melbourne where the author was on sabbatical leave. LITERATURE CITED Archer, M., 1977. Revision of the dasyurid marsupial genus Antechinomys Krefft. Mem. Qd Mus. 18 : 17-29. Colyer. J.F., 1936. ‘Variations and diseases of the teeth of animals’. (John Bale, Sons and Danielson: London). Miller, W.A. and Radnor, C.J.P., 1970. Tooth replacement patterns in young Caiman sclerops. J. Morph. 130 : 501-10. Potkay, S., 1977. Diseases of marsupials. In Hunsaker, D. (ed.) ‘The biology of marsupials’. (Academic Press: London). MILLER: RADIOGRAPHIC PATHOLOGY Plate 1 Plate 1 Radiographs of seven dentaries of Antechinomys laniger collected by Dr M. Archer at Horseshoe Cave. All of right side except C. Compare D with F to appreciate movement of teeth away from lower border of the bone with increased wear on the cheek teeth. Shorter tooth roots of premolar teeth are evident. A is J23103. Mem. Qd Mus. 20(1); 125-33, pi. 1 [1980], TWO LARGE AUSTRALIAN ORB-WEAVING SPIDERS, ERIOPHORA TRANSMARINA (KEYSERLING 1865) AND ERIOPHORA BIAPICATA (L. KOCH 1871) Valerie Todd Davies Queensland Museum ABSTRACT Eriophora transmarina and E. biapicata have been re-described. A neotype has been established for E. biapicata. E. transmarina is found along the eastern coast of Australia from New South Wales to New Guinea. E. biapicata occurs in the western areas of the eastern states, in Central, South and Western Australia. There has been continual confusion about the identity of the large orb-weaving spiders in Australia. The controversy has mainly centred on Eperia transmarina Keyserling 1865 and Epeira producta Koch 1867. Koch (1871) redescribed and figured Kumbarilla, SEQ, QM S369; 9, College View, via Gatton, SEQ, QM S370; 9, Cunnamulla, SWQ, QM W1477; 9, Longreach, CQ QM S371; 3 9, Canberra, A.C.T., QM S372; d, Eastwood, N.S.W., AM KS3144; 9, Broken Hill, N.S.W., AM KS3145; 9, Ardlethan, N.S.W., AM KS3146; 9, Menindee, N.S.W., AM KS3147; 9, Bingara, N.S.W., AM K61643; 9, Strathfield, Sydney, N.S.W., AM K56167; 9, Canterbury, Sydney, N.S.W., AM KS3148; 9, Mandurama, N.S.W., AM KS3149, 9, Earlwood, N.S.W., AM K66688; 9, Woy Woy, N.S.W., AM KS3150; 9, Campsie, Sydney, N.S.W., AM K61645; 9, Katoomba, N.S.W., AM K56200; Putney, Sydney, N.S.W., AM K57451; 9* Bexley, Sydney, N.S.W., AM K55478; 9, Willoughby, Sydney, N.S.W., AM K57602; 9, Lithgow, N.S.W., AM K53530; 9, Wahroonga, Sydney, N.S.W., AM K61649; 9, Dundas, N.S.W., AM KS3151; 9, Naremburn, N.S.W., AM K61120; 9, Mudgee, N.S.W., AM KS3152; 9, Annangrove, N.S.W., AM KS3153; 9, Northam, W.A., AM KS3154. Description Male Neotype: QM S361: Cephalothorax length lO'O mm; abdomen length 12-8 mm; total length 19 0 mm. There are 2 long setae on the anterior margin of the cephalothorax between lateral eyes and median projection; 2 long setae between the anterior and posterior median eyes. The latter are smaller than the anterior eyes 5:6; the eyes are an equal distance apart. The abdomen has 2 latero-dorsal tubercles, a posterior one and another posterior tubercle below this (Fig. 9). There is a distal ventral hook on coxa I and a distal spur on coxa II (Fig. 10) as noted by Dondale (1966). There are 5 spines on femur III and 9 retro-ventral spines on femur IV. Tibia II is swollen prolaterally. It has 3 dorsal spines, 7 ventral spines and 12 prolateral spines — 2 apical, 4 upper and 6 lower prolaterals (Fig. 11). d" palp (Figs. 12, 13); see also Dondale (1966, fig. 2G). This is very similar to E. transmarina however the terminal apophysis is relatively larger. VARIATION: Males examined varied in length from 11 to 19 mm. The abdomen is either uniform in colour or has a dark median stripe. Femur III with 4-7 spines. Tibia II with 10-15 prolateral spines and 6-10 ventral spines. Female: Length 14-27 mm. The pattern is variable. Anterior median eyes are larger than posterior median with an equal space between eyes. The clypeus is wide, twice the diameter of an anterior median eye. The seape length is 3 0-410 (average 3-6 mm). There are 5-7 spines on proventral femur II. Epigynum (Figs. 14, 15); see also Dondale (1966, fig. 2E). Apart from the presence of the second posterior tubercle on the abdomen in E. biapicata, the females of both species are very similar though the scape of E, biapicata tends to be shorter, femur II has more proventral spines, and the clypeus is DAVIES: ERIOPHORA 129 terminal apophysis Figs. 9-13: cf Eriophora biapicata (QM S361). 9, abdomen, lateral; 10, cephalothorax, ventral; 11, 1. tibia II, prolatero-ventral; 12, 1. palp, dorsoprolateral; 13, palp, mesal. Figs. 14-15: 9 Eriophora biapicata (QM S364). 14, epigyne, posterior; 15, epigyne, lateral. 130 MEMOIRS OF THE QUEENSLAND MUSEUM wider than in E. transmarina. The males of E. biapicata are separated on the presence of the spur on coxa II and the absence of the 4 heavy thorn-like spines on the ventral surface of femur III. Distribution E. biapicata is found in Victoria, South Australia, Central Australia, West Australia and in areas west of the Great Dividing Range in New South Wales and Queensland. Occasionally a spider, probably transported by vehicles, has been found in more eastern areas of Queensland {Fig. 16). In Sydney, N.S.W., female E. biapicata were found in several localities. Neither E. biapicata nor E. transmarina has been found in Tasmania (Hickman, pers. comm.). Araneus brouni from New Zealand has a short epigyne and is distinct from these species. Fig. 16: Map showing distribution of E. transmarina (a) and E. biapicata (•). DAVIES: ERIOPHORA 131 ACKNOWLEDGMENTS I should like to thank the following who kindly gave me access to their collections or sent specimens on loan from their collections: Dr Gisela Rack (ZMH), Mr F. Wanless (BM), Dr M. Grasshoff (SMF), Dr C.D. Dondale for (CNC), Dr R.R. Forster (OM), Dr A. Neboiss (NM), Mr M. Gray (AM), Mr D.C. Lee (SAM), Dr L. Koch (WAM), Mr M. Upton (ANIC). I am indebted to Professor H.W. Levi who, on a 24 hour visit to Brisbane told me that these large spiders were Eriophora. Thanks are also due to the Trustees of the C.S.LR.O. Endowment Fund, Canberra for financial support to travel to Europe and to the Interim Council of the Australian Biological Resources Study which funded field trips to the northeastern coast of Australia where some of these spiders were collected. LITERATURE CITED Archer. A.F., 1951. Studies in the Orbweaving Spiders (Argiopidae) 1. Am. Mus. Novit. 1487: 1-52. Bonnet, P., 1945-61. ‘Bibliographia Araneorum.’ (Toulouse.) Chrysanthus, O.F.M., 1960. Spiders from south New Guinea III. Aova Guinea (ns) 10(3): 23-42. 1970. Further notes on the spiders of New Guinea (Argyopidae). Zool. Verh., Leiden 113: 3-52. Dondale, C.D., 1966. The spider fauna (Araneida) of deciduous orchards in the Australian Capital Territory. Aust. J. Zool. 14: 1157-92. Hogg, H.R., 1900. A contribution to our knowledge of the Spiders of Victoria; including some new species and genera. Proc. R. Soc. Viet. 13: 68-1 23. Keyserling, E., 1865. Beitrage zur Kenntnis der Orbitelae. Latrl. Verh. zool. -hot. Ges. Wien 15: 799 856. 1884-89. ‘Die Arachniden Australiens.’ (Niirnberg). Koch, L., 1867. Beschreibung neuer Arachniden und Myriopoden. Verh. zool. -hot. Ges. Wien 17: 173-250. 1871-83. ‘Die Arachniden Australiens’. (Nurnberg). Levi, H.W., 1970. The ravilla group of the orbweaver genus Eriophora in North America (Araneae: Araneidae) Psyche, Camb. 77: 280-302. Main, B.Y., 1976. ‘Spiders’. (Collins: Sydney). Musgrave, a., 1933. ‘The Garden Orb-Weaving Spider’. Australian Museum Leaflet 5, one page. Rainbow, W.J., 1909. Notes on the architecture, nesting habits and life histories of Australian Araneidae. Rec. Aust. Mus. 1: 212-34. 1911. A census of Australian Araneidae. Rec. Aust. Mus. 9: 107-319. Roewer, C.Fr., 1942-54. ‘Catalogue der Araneae.’ (Bremen). Simon, E., 1892-1903. ‘Histoire naturelle des Araig- nees.’ 2nd Ed. (Paris). Strand, E., 1913. Uber einige australische Spinnen des Senckenbergischen Museums. Zool. Jh. Syst. 35: 599-624. Thorell, T., 1881, Studi sui Ragni Malesi e Papuan! III. Annali Mus. civ. Stor. not. Giacomo Doria 17: vii-xxvii, 1-720 MEMOIRS OF THE QUEENSLAND MUSEUM o a , .2 *n u o 00 t a < m DAVIES: ERIOPHORA Plate 1 Mem. Qd Mus. 20(1): 135^1, pi. I [1980], MEGADOLOMEDES NOV. GEN. (ARANEAE: PISAURIDAE) WITH A DESCRIPTION OF THE MALE OF THE TYPE-SPECIES, DOLOMEDES AUSTRALIANUS KOCH 1865 Valerie Todd Davies and Robert J. Raven Queensland Museum ABSTRACT Males of this east Australian water spider are described for the first time. They show differences from Dolomedes which warrant the establishment of a new genus, Megadolomedes. Megadolomedes has very long legs with long, flexible tarsi. The male is very much smaller than the female and the palpal organ differs essentially from that of Dolomedes in having a well developed club-shaped apophysis at the base of the fulcrum. The epigyne and variations of pattern in the female are illustrated and some notes on the biology given. Dolomedes australianus Koch 1865, D. cervinus Koch 1876, and D. trux Lamb 1911 are considered synonyms of Megadolomedes australianus. Dolomedes australianus was described by Koch (1865) from an immature female from New South Wales. Later (1876) he described the mature female. The spider is found along the waterways of the east coast of Australia from Tasmania to Cape York and has been collected on Prince Edward Island in Torres Strait. The females are large spiders with body lengths varying from 20 to 30 mm and with leg spans up to 150 mm or more. The male is very much smaller with a leg span of less than 60 mm. Because of its long legs and flexible tarsi the spider bears a striking superficial resemblance to the Neotropical genus, Trechalea. Measurements were made with an ocular micrometer and converted to millimetres. Abbreviations: AM, Australian Museum, Sydney; HZMI, Zoologisches Staatinstitut und Zoologisches Museum, Hamburg; QM, Queens- land Museum, Brisbane. AME, ALE, PME, PLE, anterior median, anterior lateral, posterior median, posterior lateral eyes; MOQ, median ocular quadrangle. Megadolomedes nov. gen. Large spiders with long legs and flexible tarsi only a little shorter than metatarsi. Eyes in two recurved rows, anterior row much narrower than posterior, AME slightly larger than ALE. Retromargin of cheliceral furrow with 4 large, subequal teeth; promargin with 3 teeth, distal minute, middle largest, proximal intermediate in size. 7 spines on dorsal femur of pedipalp. Legs 4(12) 3, Superior lorum of pedicel composed of 4 sclerites, 2 median and 2 lateral. Male much smaller than female, with short palpal tibia bearing medial retrolateral apophysis. Cymbium long and slender with non-excavated distal region. Embolus very long; fulcrum well developed with a flat club-shaped apophysis basally. Megadolomedes can be distinguished from Dolomedes by the characters listed in Table 1. Type-SPECIES: Dolomedes australianus Koch 1 865, the only species known for this genus. TABLE 1: DIFFERENCES BETWEEN MEGADOLOMEDES AND Dolomedes. Megadolomedes Dolomedes tarsi flexible; more than not flexible; barely % length of metatarsi Vi length metatarsi male much smaller than 9 about the same size, slightly smaller cymbium long, slender; excavated excavated for most proximally, of length; conical flat distally distally lateral flat and club- absent sub-terminal apophysis shaped 136 MEMOIRS OF THE QUEENSLAND MUSEUM Megadolomedes australianus (Koch) Dolomedes australianus Koch, 1865, p. 863; 1876, p. 873. Dolomedes cervinus Koch, 1 876, p. 872. Dolomedes trux Lamb, 1911, p. 173. Rainbow, 1912, p. 210. Material Examined Cedar Creek Falls, Mt Nebo, near Brisbane, SE.Q., R. Raven, 16.xi.l973, I d", QM S230; Home Rule, near Helenvale, NE.Q., V. Davies, D. Joffe, 19.xi.l974, 2 d", QM S231; Dolomedes australianus, id, L. Koch, 1876, Wollongong, N.S.W., 1 9 , H2MI, Mus. Godeffroy 302; Dolomedes trux, holotype, Ithaca Creek, Brisbane, SE.Q., J. Lamb, 1 penultimate 9, QM G55; Dolomedes trux id. W.J. Rainbow 1912, Blackall Ranges, SE.Q., C.J. Wild, 1 juvenile, QM W2146; Pennant Hills, N.S.W., 10.viii.l953, 1 penultimate 9 , AM; Richmond Range, N.S.W., R. Raven, 14. ix. 1974, 1 penultimate 9> QM S232, Home Rule, NE.Q., V. Davies, D. Joffe, 18.xi.l974, 1 9 , QM S233, ibid., 19.xi.l974, 2 9, QM S234; Davies Creek, Mareeba, NE.Q., N.C. Coleman, 30.iii.l975, 1 9, QM S235; Mount Molloy, N.E.Q., F. Little, ix.l969, 1 9 . QM S236; Booloomba, Kenilworth State Forest, SE.Q., R. Raven, ii.1973, 2 9. QM S237; Kondalilla National Park, SE.Q., R. Raven, 10.xii.l973, 2 9 t QM S238; Goodna reach, Brisbane River, SE.Q., R. Monroe, 13.V.75, 1 9> QM S239; Beerburrum, SE.Q., G.J. Male, 1 9, QM S240; Little Yabba Creek, Kenilworth State Forest, SE.Q., G. May, 1 9i QM S241 ; East branch Kilcoy Creek, SE.Q., K.R. McDonald, 1 9^ QM S242; Scrubby Creek, near Eden, N.S.W., 13.xii.l971, 1 9 , AM. Description Male (QM S230); Carapace length 4-48, width 3-68. Abdomen length 4-20, width 2-28. Legs 4(12)3 (Table 2). Colour (in alcohol) of carapace, legs and chelicerae yellow brown. Dorsal abdomen light brown with anterior median brown rectangular marking to about half the length of the abdomen, irregular darker areas behind. Carapace lightly covered with short brown simple hairs. Fovea length 0-25 x length of carapace, lateral striae present. Clypeus twice the longest diameter of an anterior median eye (Fig. 1). Ratio of eyes AME;ALE:PME:PLE is 8:7:14:15. MOQ width at front: width at back: length is 21:35:33. Maxillae longer than wide, 35:14 and broadest distally. Labium about as long as wide 30:27 constricted at base and rounded distally. Sternum shield-shaped, as long as wide, pointed posteriorly. Tarsi very long, curved and flexible with transverse striations, no scopula (Plate 1, C). Superior claws similar with 7-8 long teeth (Plate 1, D); inferior claw smooth and short. TABLE 2: Leg Measurements (mm) of d" Megadolomedes australianus I II III IV Palp Femur 613 6-44 5-44 7-00 2-63 Patella 1-88 2-19 1-75 1-94 0-88 Tibia 5-94 5-94 4-38 5-94 0-88 Metatarsus 6-25 5-88 4-75 6-88 — Tarsus 5-13 4-83 3-50 5-13 3-75 Total 25-33 25-28 19-82 26-89 8-14 Notation of spines. First leg: Femur, p.LLl.l.l.d.l.O.l.O.l.r. 1.1.1. 1.1. Patella, p-L d.l.r.L Tibia, p.O.l.O.l.d.O.l.O.l.r.O.l.O.l.v.2.2. 2.2. Metatarsus, p. 1.1.1. r. 1.1.1. v. 2. 2. 3. Tarsus, 0. Second leg: Femur, p.l.l.l.l.l.d. 1.0. 1.0.1. r.Ll.1.1.1. Patella, p.l.d.l.r.l. Tibia, p.0.1.0. 1. d.l .0.0.1 .r.0.0.1 .l.v.2.2.2.2. Metatarsus, p.1.1. l.r.l.l.l.v.2.2.3. Tarsus, 0. Third leg: Femur, p.l.l.l.l.l.d.l.O.l.O.l.r.1.1.1.1.1. Patella, p.l. d.l.r.L Tibia, p.O.l.Ld.l.O.l.r.O.l.l.v.2.2.2. Metatarsus, p.l . 1 . 1 .r. 1 . 1. 1 .v.2.2.3. Tarsus, 0. Fourth leg: Femur, p.l .1 .1 .1 .1 .d. 1 .0. 1 .0.1 .r. 0. 0. 1.1.1. Patella, p.l.d.l.r.l. Tibia, p.0.1.1. d.l .0.1 .r.O.l.l.v.2.2.2. Metatarsus, p.l .l.l.r.l .1. 1. v.1.1 .1 .1.1 .3. Tarsus, 0. Palp: Femur, p.0.0.1. 0.d.l.l.l.2.r.0.0.1.0. Patella, p.l.d.l.r.l. Trichobothria are present in an irregular line of 1 1-13 running the full length of prolateral femora i-iv; some on proximal femora i-iv; some on the prolateral, dorsal and retrolateral surfaces of tibiae i-iv (most numerous proximally); 6-7 on dorsal metatarsi; and 5-6 in a single row and an additional 3^ distally on tarsi i-iv. Palp (Figs. 2, 3, Plate 1, B) has a slender,, elongate cymbium, not excavated distally. Embolus long and whip-like, supported by the fulcrum distally. The membraneous conductor tapers to a tongue which supports the tips of the fulcrum and embolus. The median apophysis arises from the tegulum and is rounded and membraneous distally. From the base of the fulcrum a flat club-shaped apophysis, Comstock’s (1948) ‘lateral sub-terminal apophysis’, extends posteriorly and lies against the tegulum; near its point of attachment it has a needle-like projection. A similar sharp projection arises near the base of the embolus. The tibial apophysis is a stout, curved, retrolateral spur arising mid-way along the short tibia. DAVIES AND RAVEN: MEGADOLOMEDES 137 Variation: The body lengths of 2 other males from Brisbane; on examination this is found to be were 7-3 and 7-8 respectively. There is little a penultimate female M. australianus. An variation in pattern and spination. immature specimen QM W2146 from the Blackall Female (Plate 1, A): In 1876 Koch described Range described by Rainbow (1912) as D. trux is Dolomedes cerinus which, from the description, is also M. australianus. There is great variability in certainly an immature female Megadolomedes size and pattern in females as noted by Koch; the australianus. Lamb (1911) described D. trux spider is figured in Koch ( 1 876, LXXV, 2) though the curved flexible nature of the tarsi is not clearly shown. The superior lorum of the pedicel is Fig. 1^: d" M australianus. 1, cephalothorax, lateral; 2, palp, retrolateral; 3, palp, prolateral; 4, pedicel with sclerites. 138 MEMOIRS OF THE QUEENSLAND MUSEUM composed of 4 sclerites (Fig. 4). A scopula is present on metatarsi and tarsi of all legs. Epigynum (Figs. 5, 6) has well defined lateral ridges. The long emboli are often found blocking the epigynal ducts of the female (Fig. 7); this gives the epigynum a broader outline than usual. The atrium leads into a long twice-coiled bursa copulatrix to the ovoid spermatheca. From here a thicker, coiled fertilization tube goes to the oviduct. Variation; Females vary in length from 20 to 30 mm. Five colour patterns may be recognised: 1) The ‘type pattern as illustrated by Koch — light yellowish lateral bands on the carapace and primarily a dark foliate pattern on the dorsal abdomen. 2) The ‘uniform’ pattern as illustrated by Koch for D. cervinus — lateral cephalic bands poorly defined, if present; irregular pigmentation on the abdomen. 3) The ‘spotted’ pattern — carapace and abdomen brown-black marked with yellow spots. 4) The ‘striped’ pattern — lateral cephalic bands well defined and dorsal abdomen with a median broad dark band enclosed laterally by yellowish stripes. 5) As in Plate 1, D with more Fig. 5-7: 99 ^ australianus. 5, external epigynum S233; 6, internal epigynum S233; 7, internal epigynum with emboli in ducts S234. DAVIES AND RAVEN: MEGADOLOMEDES 139 extensive light areas than in 1). (Figs. 8-12). The presence of two or more patterns occurring together suggests that the patterns are of no ecological significance. Biology Megadolomedes australianus has a round seamless egg-sac which it carries in the chelicerae, with the aid of the palps. Just prior to the emergence of the young the egg-sac is suspended in a loosely constructed nursery in a shrub near the water (Hickman 1967). In the laboratory the young emerged from this 5-6 days later. Females are found throughout the late spring, summer and autumn in areas of still to moderately fast moving water. Males have been found only in early summer (November, December) although penul- timate males were found as early as August. Before maturity they are often found high up in vegetation overhanging water. Females show a preference for the underside of logs and branches and flatten themselves if disturbed or they may dive into the water and remain submerged for nearly half an hour (Hickman 1967). In swimming the spider sculls with the second and third pairs of legs while the first and fourth pairs are held outstretched with tips together. In pursuit of prey, females strike the water with all the legs at once, thus bouncing across the water surface. The long flexible tarsi appear to be an adaptation to support the large body on water. ACKNOWLEDGEMENTS We should like to thank Dr Gisela Rack who gave the senior author generous access to the collections in the Zoologisches Staatinstitut und Zoologisches Museum during a visit in 1977. Thanks are also due to the Trustees of C.S.I.R.O. Endowment Fund, Canberra for financial support to travel to Europe and to the Interim Council of Australian Biological Resources Study which funded the collecting trip to northeastern Queensland during which specimens of M. australianus were collected. LITERATURE CITED Comstock, J.H., 1948. The Spider Book,’ revised and edited by W.J. Gertsch. (Comstock; New York). Hickman, V.V., 1967. ‘Some Common Tasmanian Spiders’. (Tasmanian Museum and Art Gallery: Hobart). Koch, L,, 1865. Beschreibungen neuer Arachniden und Myriapoden. Verh. zool. -hot. Ges. Wien 15 : 857-92. 1876. ‘Die Arachniden Australiens’. (Niirnberg). Lamb, J., 1911. Descriptions of some new Queensland Araneida. Ann. Qd Mus. 10; 169-74. Rainbow, W.J., 1912. Araneidae from the Blackall Ranges. Mem. Qd Mus. 1 : 190-202. 1912. Note on Dolomedes trux Lamb. Mem. Qd Mus. 1 : 210 . Fig. 8-12: 99 australianus patterns. MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 A; 9 M- australianus. B-D: cS" australianus. B, palp, ventral; C, tarsus showing striations; D, tarsal claws. (Marker = 110m) DAVIES AND RAVEN: MEGADOLOMEDES Plate I Mem. Qd Mus. 20(1): 143-61, [1980]. THE ESTUARINE MACROBENTHOS OF THE CALLIOPE RIVER AND AUCKLAND CREEK, QUEENSLAND P. Saenger* W. STEPHENSONt and J. MOVERLEY* ABSTRACT Quantitative benthic sampling in the Calliope River and Auckland Creek has been conducted at approximately 3 monthly intervals since November 1974 at 1 1 transects, each with 5 stations (55 sites). In September 1976 power generation commenced, using the estuary for thermal discharge. The data to August 1976 — the ‘pre-thermal’ situation — are intended to serve as the baseline for comparison with the ‘post-thermal’ situation. The ‘pre-thermal’ data are analysed here. A total of 263 taxa were found comprising 15 species of seagrasses and algae, 75 species of polychaetes, 5! species of molluscs, 91 species of crustaceans and 30 species belonging to other phyla. Of these species, those represented by less than 5 individuals, and those occurring in less than 2 sites or times, were not considered further. The remaining 72 species (representing 88.7% of all individuals collected) were classified into site-, time-, and species-groups using the ‘Bray-Curtis’ dissimilarly measure, followed by group-average sorting. Nine site-groups were accepted; these showed good topographic coherence and they could be characterized by their depths, substrate types, and position in the river. Classification into species-groups showed that many species were scattered throughout various sites, and on the basis of the pseudo- F test, only 39 species conformed to the site groups. Species classification using species-occurrences within the already established site-groups, gave 14 species-groups which characterize by their absence or presence, the various site-groups. Classification into time-groups showed that intertime dissimilarities are high, indicating that either sampling times were too distantly spaced, or alternately that seasonal differences overlie a non-seasonal trend. Cyclical analysis of summated species data identified a non-cyclic and a cyclic component with a period of approximately 12 months. A recolonization model has been proposed to describe these data; it is postulated that the intense flooding of December/ January 1974 led to the removal of the benthos and that species reappearance was more or less linear over a period of 29 months, while the number of individuals of all species increased logarithmically. When comparisons with the ‘post-thermal’ fauna are made, it will be essential to allow for this recovery when assessing the thermal effects. The utilization of estuaries as heat-sinks for electricity generating plants has stimulated considerable overseas interest in the effects of elevated temperatures on estuarine biotas (Naylor 1965, Thorhaug et al. 1973, Young and Frame 1976). In Queensland, several estuaries are affected by thermal discharges (e.g. Brisbane R., Burrum R., Ross R.) but there are no studies on the effects of elevated temperatures on estuarine biotas. As part of a general investigation into the benthic organisms of Port Curtis, the Calliope River and Auckland Creek are under study. The investigation of the macrobenthos of Port Curtis has now been terminated and an account of these data has been given (Stephenson et al. 1979). Quantitative benthic sampling in the Calliope River and Auckland Creek has been conducted at approximately 3 monthly intervals since November 1974 at 11 transects, each with 5 stations (55 sites). With power generation commencing in September 1976, the results obtained from November 1974 to August 1976 — the ‘pre-thermal’ situation — are intended to constitute the baseline for comparison with the ‘post-thermal’ situation, and an analysis of these ‘pre-thermal’ results is given below. * Scientific Services Branch, Queensland Electricity Generating Board, Brisbane, t Zoology Department, University of Queensland. 144 MEMOIRS OF THE QUEENSLAND MUSEUM THE STUDY AREA The study area (Fig. 1) comprises the Calliope River and Auckland Creek, both of which flow into Port Curtis, a semi-enclosed bay whose outer barrier consists of Facing and Curtis Islands. Climatically, the area is sub-tropical with a mean annual rainfall of 944 mm, of which approximately 50% falls from December to February. Rainfall data, together with other meteorological and hydrological data are summar- ized in Table 1. Temperature, salinity and dissolved oxygen data for the Calliope River opposite the power station has been compiled from TABLE 1 : Mean Monthly Meteorological and Hydrological Data for Gladstone and Calliope River Month Mean Daily Temperature* (°C) Min/Max Rainfall/ Raindays* (mm) Evaporation* (mm) Humidity* (%) 9 a.m./3 p.m. Radiation** (M J/ m^/day) Mean River Discharge*** (megalitres) January 3M/22-2 190/20 206 70/63 24-7 21,695 February 30-9/22 0 153/15 172 71/63 23-5 68,342 March 301/2M 90/10 177 71/61 20-9 48,479 April 28-5/19-3 35/ 7 150 69/59 19-2 6,480 May 25-5/16-4 50/ 5 114 70/57 15-8 6,964 June 23-2/14-2 40/ 6 96 69/55 14-4 4,666 July 22-6/12-9 37/ 7 102 66/51 15-4 5,537 August 24-1/14-1 33/ 7 113 67/53 17-5 2,009 September 26-4/16-1 21/ 5 149 62/55 21-8 726 October 28-7/18-6 63/ 6 183 61/58 23-4 2,009 November 30-2/20-4 82/11 189 63/61 25-5 778 December 30-9/21-5 150/ 9 202 66/62 24-5 18,213 Annual 27-7/18-2 944/108 1855 67/58 20-5 185,718 * Supplied by Bureau of Meteorology ** from Paltridge and Proctor (1976) *** Supplied by Water Resources Council of Queensland from ‘Castlehope’, 33 km upstream of mouth. SAENGER ET AL: MACROBENTHOS OF THE CALLIOPE RIVER 145 various sources for the duration of the study period, and these are presented in Fig. 2. These data show cyclic changes in water temperature with an optimal wavelength of 12 months (Saenger et al. 1979), and somewhat less regular changes in salinity and oxygen saturations. A correlation between sharp changes in salinity and dissolved oxygen are also apparent, and the i nflow of water rich in detritus is probably involved. 160- 140- 120- 100- ■■ 80- 60- 40- 0, Saturation 1%) (dl 40- ■ 1 f 30- 20- 10- Salinity (%l Ibl 30- ’ J. ' 20- 10- Temperature (C‘) (cl N oIiIj fmamj J ason oIiU f m a M j J 1974 ' 1975 A S 0 1976 The bottom sediments of the Calliope River include up to 1-3 m of mud, which covers rock, or overlies sand. The mud generally consists of soft, dark'brown or blue-grey, usually organic, silty clay (Hofmann 1971). The sand is brown to grey, silty, fine to medium grained although occasional coarse deposits are found. Gravel and shell debris occur locally, generally in the deeper channels of the river. The present day stream sediment load of the Calliope River, which drains a hinterland of argillaceous rock (Jardine 1925), is predominantly mud. Quaternary intertidal mud deposits (predomin- antly quartzite and albite) overlie most of the Palaeozoic rocks to the east, north and west of the power station site (Hofmann 1971). The dark-grey to black organic mud includes layers of silt and sand. The bottom sediments of Auckland Creek include sandy clay, silty sand, and gravels in the deeper channels. Auckland Creek does not have an appreciable catchment area or stream sediment load (Hofmann 1971). Conaghan (1966), who studied the sedimentary processes in Port Curtis, concluded that tidal circulation dominates the hydrographic features of the area. The scouring effects of tidal currents (the tidal range is up to 4-2 m), particularly at ebb tide, maintains a relatively stable balance between deposition and erosion in all tidal channels, including the smaller creeks. 30 20 10 BOTTOM SALMTY |%.l ■ June 1974 . i November 1973 I "Means (March 1970- July 1975 I 5 1— " — I ' ' ' ' f" 10 15 20 DISTANCE UPSTREAM Ikml Fig. 2: Physico-chemical parameters of Calliope River; (a) oxygen saturation opposite power station site; (b) salinity opposite power station site; (e) bottom water temperatures from throughout the Calliope River; (d) decreasing salinity in the Calliope River with increasing distance upstream. (Based on data collected (i) by spot readings in the river, (ii) continuous salinity/temperature recorders, (iii) Dr J. Greenwood and (iv) Water Quality Council quarterly surveys). METHODS Survey Design Eleven transects were selected on the river and creeks (Fig. 1); each transect extends across the river from about 1 m below L.W.M. at each bank. Two transects were on Auckland Creek, one in Black Harry Creek, two on the anabranch of the Calliope River and 6 were distributed along the length of the main course of the Calliope River, Selection of transects was based on their spread along the rivers and the ease with which they could be accurately relocated. Relocation errors are about ± 2 m except at transects 1 and 1 1 , where they are approximately ± 4 m. At each transect, five equally-spaced stations were sampled using an 0*05 m^ Van Veen grab sampler; the duplicate grabs were combined to 146 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 2; Depth, Upstream Distance and Substrate Type of Individual Sampling Sites, 1-55. Transect Station No. No. Site No. Depth at Distance low water upstream (m) (km) Substrate type 1 1 1 1-6 1-3 Silty sand 1 2 2 1-9 1-3 Silty sand, some clayplug 1 3 3 1-8 1-3 Silty sand, some clayplug 1 4 4 6-1 1-3 Silty sand, some clayplug 1 5 5 5-6 1-3 Silty sand 2 1 6 1-3 2-8 Sandy silt 2 2 7 2-0 2-8 Silt 2 3 8 3-7 2-8 Silt 2 4 9 2-4 2-8 Silt 2 5 10 1-0 2-8 Yellow clay 3 1 11 0-8 5-6 Coarse sand, some gravel 3 2 12 3-1 5-6 Coarse sand 3 3 13 2-8 5-6 Coarse sand 3 4 14 2-3 5-6 Coarse sand 3 5 15 1-3 5-6 Coarse gravel 4 1 16 2-3 3-8 Coarse gravel 4 2 17 5-2 3-8 Coarse gravel 4 3 18 6 6 3-8 Coarse gravel 4 4 19 7-2 3-8 Fine sand 4 5 20 4-1 3-8 Silt 5 1 21 1-3 3-6 Soft mud 5 2 22 8-3 3-6 Sandy mud 5 3 23 8-0 3-6 Muddy sand 5 4 24 8-7 3-6 Fine sand 5 5 25 9-0 3-6 Silt 6 1 26 2-1 15-1 Solt mud 6 2 27 4 1 15-1 Mud 6 3 28 3-9 15-1 Coarse gravel 6 4 29 2-1 15-1 Coarse gravel 6 5 30 1-8 15-1 Coarse gravel 7 1 31 1-5 11-1 Coarse gravel 7 2 32 2-8 11-1 Coarse gravel 7 3 33 3-7 11-1 Mud, some gravel 7 4 34 4-8 IM Coarse gravel 7 5 35 2-2 111 Soft mud 8 1 36 3-4 8-3 Silty sand 8 2 37 4-6 8-3 Silty sand 8 3 38 4-4 8-3 Silty sand 8 4 39 4-6 8-3 Silty sand 8 5 40 2-0 8-3 Soft mud 9 1 41 0-6 7-1 Silty sand 9 2 42 2-7 7-1 Silty sand 9 3 43 5-8 7-1 Silty sand 9 4 44 4-1 7-1 Mud 9 5 45 1-2 7-1 Soft mud, some detritus 10 1 46 0-7 4-6 Soft mud 10 2 47 1-3 4-6 Mud 10 3 48 2-7 4-6 Mud 10 4 49 2-3 4-6 Mud Transect Station No. No. Site No. Depth at Distance low water upstream (m) (km) Substrate type 10 5 50 0-9 4-6 Mud 11 1 51 4-4 4-6 Coarse shell material 1 1 2 52 3-9 4-6 Coarse sand 1 1 3 53 3-6 4-6 Coarse sand 11 4 54 3-1 4-6 Coarse sand 11 5 55 2-2 4-6 Coarse sand give a total of 0-1 sampled at each of 55 sites in all. For each transect and station, data on depth, sediment type and distance upriver are given in Table 2. Sample Treatment Sampling was carried out at approximately three-monthly intervals as follows: November 1974; March 1975; June 1975; October 1975; February 1976; May 1976 and August 1976. These times are referred to below as T ]_7 respectively. On all occasions all sites were sampled although at some site-times, no organisms were found (Table 3). All samples were subsequently sieved through an 8 mm (to break up the sediments) and a 1 mm TABLE 3: Site-times at which no Organisms were FOUND in the Samples Times Sites 1 24 28 29 30 32 34 50 55 2 11 35 46 48 54 3 9 15 21 23 24 29 49 4 49 5 45 6 — 7 sieve in the field and that fraction retained by the fine sieve was fixed in 5% formalin/seawater and subsequently transferred to glycerol: water: ethanol (5:25:70 v/v). Seagrasses and algae were blotted dry and weighed. In order to make these weights approximately equivalent to faunal numbers, the weights were multiplied by 10. Numerous taxonomic problems were encoun- tered while working up the collections and the assistance of many specialists was enlisted (see acknowledgements). Once species had been identified, subsequent identifications were gener- ally made using validated reference material, except in a few difficult groups such as some polychaetes, echinoderms and ascidians. SAENGER ET AL: MACROBENTHOS OF THE CALLIOPE RIVER 147 ANALYSES OF DATA Initially the data were treated as three- dimensional i.e. time (1-7) x transect (1-11) x station (1-5), or as two-dimensional i.e. time (1-7) X sites (1-55). In addition there was a species dimension of 263 if all species are included. Five different sample parameters ( 5 * number of species; n number of individuals of all species; G Gleason diversity; H' standardized Shannon diversity to log base 2; J’ Shannon equitability) were derived from summations involving all species, and five matrices of times x transects x stations were considered. Classification For classificatory purposes, the less abundant species — those with less than 5 individuals in the 385 samples — were eliminated. This reduced the species number to 1 20, and gave a basic matrix for classification of 120 species x 55 sites x 7 times. Five of these species however were subsequently eliminated because they occurred in less than 2 sites. Classification into species-, time-, and site- groups were thus initially based on 1 1 5 species. For further time-group classification, an addition- al 6 species were eliminated as they occurred only once. Site classification based on 115 species suggested that many species were scattered throughout the different site-groups, and while these species may be important ones, for classificatory purposes they are in the nature of ‘background noise’. These species were ignored during the final site classification, leaving a residuum of 72 species for the determination of site-groups; these 72 species represent 88.7% of all individuals collected during this study. Data transformations were used as follows: Site classification involved use of the logio («+l) transformation followed by the ‘Bray-Curtis’ dissimilarity measure with subsequent group- average sorting. For species, the transformed values were standardized by totals followed by ‘Bray-Curtis’ dissimilarity and group-average sorting as before. Data for times were averaged over the 55 sites and transformed by logjo («-F 1), and then classified using ‘Bray-Curtis’ dissimila- rity and group-average sorting. Periodicity Analyses of periodicity followed the iterative technique described in Saenger et al. (1979) and Stephenson and Burgess (in press). It consists of running the generalized equation, Y = ViA^ + Acos (^) + Bsin (2?n) T T where Y is the recording of a species at time t, T is the wavelength of the curve, ViA° is the estimated midpoint of the curve, A is the contribution of the cosine component, B is the contribution of the sine component, through a series of wavelengths, and determining the proportion of the total sum of squares accounted for by the regression {R^). If the selected wavelength is the predominant one in the system, then will be maximal. TABLE 4: Faunistic Composition of the Calliope River and Auckland Creek Macrobenthos. Phylum and Total No. No./Station Total No. class individuals of species Thallophyta 2-71* 49** 14 Anthophyta 7-33* 133** 1 Sub-totals 1004* 182** 15 Coelenterata Anthozoa 11 (01%) 0*2 2 (0-8%) Pennatulida 2 (01%) 01 2 (0-8%) Alcyonaria 3 (0-1%) 0-1 1 (0-4%) Nemertea 29 (0-4%) 0*5 1 (0-4%) Echiura Platyhelminthes 9 (0-1%) 0-2 1 (0-4%) Turbellaria 11 (0-1%) 0-2 2 (0-8%) Bryozoa Annelida 4 (01%) 01 1 (0-4%) Polychaeta 1455 (19-0%) 26-5 76 (30-6%) Oligochaeta Mollusca I (0*1%) 0-1 1 (0-4%) Pelecyjx)da 1927 (25-2%) 35-03 33 (13-3%) Gastropoda 521 (6-8%) 9-47 17 (6-9%) Scaphopoda Arthropoda 12 (0-2%) 0-2 1 (0-4%) Crustacea 2150 (28-1%) 39-09 91 (36-7%) Pycnogonidae Echinodermata 7 (0-1%) 0-1 3 (1-2%) Ophiuroidea 399 (5-2%) 7-25 7 (2-8%) Holothuroidea 1023 (13-4%) 18-6 1 (0-4%) Phoronida Chordata 1 (0-1%) 0 1 1 (0-4%) Ascidiaceae 16 (0-2%) 0-3 3 (1-2%) Pisces 59 (0-8%) 1-07 4 (1-6%) Sub-totals 7640 138-9 248 Total 263 * as grams ** as mg/station 148 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 5: Numerically Most Abundant Benthic Species. Code Species Total No. of % of Site- No. Individuals times at which species found POLYCHAETA % of all polychaetes in surveys 76 Glycera americana 186 12-8 27-3 86 Leitoscoloplos sp. nov. 54 3-7 3-6 95 Nephtys mesobranchia 70 4-8 9-6 98 Ophelina sp. 54 3-7 8-8 104 Prionospio sp. 58 40 6-2 111 Scoloplos implex 111 7-6 3-9 124 Magelona dakini 1 10 7-6 7-8 CRUSTACEA % of all crustaceans in surveys BRACHYURA 194 Paracleistostoma mcneilli AMPHIPODA 76 3-5 5-7 238 Dryopoides sp. 228 10-6 8-3 239 Ericthonius sp. 180 8-4 3-6 242 Grandidierella sp. 1 51 2-4 4-4 243 Grandidierella sp. II TANAIDACEA 143 6-7 11-4 276 Apseudes estuarius MYSIDAE 649 30-2 4-7 298 Unident. sp. IV ISOPODA 62 2-9 1-0 302 Sphaeroma sp. 58 2-7 3-4 MOLLUSCA % of all molluscs in surveys GASTROPODA 331 Nassarius burchardi BIVALVIA 335 13-6 37-4 356 Circe cf. australis 68 2-8 9-1 364 Gouldia sp. I 109 4-4 15-1 368 cf. Laternula sp. 621 25-2 20-3 371 Modiolus auriculatus 109 4-4 7-3 372 Notospisula parva 329 13-4 19-2 377 Solecurtis sp. 269 10-9 11-7 381 Tellina sp. 236 9-6 21-3 Code Species Total No. of % of Site- No. Individuals times at which species found ECHINODERMS % of all echinoderms in survey OPHIUROIDEA 405 Amphiura micro 365 25-7 80 HOLOTHUROIDEA 408 Protankyra verrilli 1,023 71-9 23-4 FISH % of all fish in survey PISCES 452 Brachyamblyopus cf. urolepis 53 89-8 11-2 RESULTS Faunistic Composition A total of 263 taxa were recorded in the survey although some are probably oversplit (e.g. Magelona and MAGELONIDAE sp, I) whilst others require subdivision (e.g. Squilla, Alpheus). All species are listed in the Appendix together with the code numbers used throughout. Of the 263 taxa, 1 5 were seagrasses or algae (Table 4). Amongst the faunal taxa, 36.7% were crustaceans, 30.6% polychaetes, 20.6% molluscs and 12.1% belonged to other groups (Table 4). In terms of the 7640 individuals recorded, 28.1% were crustaceans, 32.2% molluscs, 19.0% poly- chaetes and 20.7% belonged to other groups (Table 4). The most numerous species within these categories are listed in Table 5. Several new species have been found in this study, including the polychaete Nephtys meso- branchia Rainer and Hutchings (1977) and the pycnogonid Propallene saengeri Staples (1979). In addition, several new Australian records were established including significant extensions to the previously known ranges; these species are indicated in the Appendix. By summating the station data for each transect in the main channel of the Calliope River, it is possible to obtain an overall view of the upstream distribution of organisms. Two trends are SAENGER ET AL: MACROBENTHOS OF THE CALLIOPE RIVER 149 apparent in this data; Fig 3a shows that the total number of species recorded at each transect decreased with increasing distance upstream. Presumably this reflects the gradual loss of marine species with increasing freshwater influence. The second trend (Fig. 3b) is in the number of organisms per m^, and shows an initial decrease — minimum approximately 8 km upstream of the mouth — followed by a rapid increase upstream of this point. This upstream increase is largely due to a few species of amphipods, Apseudes estuarius, and the mussel Modiolus auriculatus, suggesting that these species reach large numbers in the presence of a marked freshwater influence. Mean Sample Parameters The meaned data from various groupings of samples are given in Table 6. Overall means were: 5 6 08; n 20 08; G 1 -86; //' 1 -76 and J’ 0-7 1 . From each row, the mean sum of squares (mss) was DISTANCE UPSTREAM (km)(X) Fig 3: Upstream distribution of species and individuals of all species in the Calliope River. TABLE 6: Means of the 5 Sample Paramaters (5, n, G, H‘ AND J' (IN THE 7 TIMES, 11 TRANSECTS AND 5 Stations. (Also Mean Sum of Squares Variation in EACH Row.) 1 2 3 TIMES 4 5 6 7 MSS s 3-36 216 4-04 5-31 4-87 7-40 !5-44 16-905 n 7-56 3-78 8-25 12-04 12-47 20-05 76-38 550-662 G 1-25 0-78 1-48 1-86 1-66 2-27 3-71 0-759 H’ 1-33 0-80 1-5! 1-80 1-75 2-23 2-88 0-378 J’ 0-69 0-51 0-68 0-77 0-78 0-80 0-77 0-009 obtained (Table 6) and these indicate that there is much greater heterogeneity between times than between transects, and between transects than between stations. It is also evident that there is a hierarchy of heterogeneity between the par- ameters with n>s>G>H’>J’. The main cause of times heterogeneity is the high value of all parameters (except J’) at time T-, (August 1976), contrasted with the low values in T 2 (March 1975). The likely importance of this is discussed later. 1 2 STATIONS 3 4 5 MSS s 6-10 6-32 6-65 5-47 5-87 0-160 n 20-43 17-69 21-74 19-58 20-95 1-919 G 1-84 2-05 1-99 1-65 1-76 0-021 H’ 1-79 1-97 1-79 1-53 1-72 0-020 J’ 0-73 0-79 0-72 0-64 0-68 0-003 Site Classification Classifying the 55 sites using 72 species, the sites dendrogram (Fig. 4) was resolved into 9 site-groups; their topographic coherence is shown in Fig. 5 while their depths, substrate types and their position in the river are given in Table 7. Two site-groups (III, V) consist of a single site respectively; while site-group V is not easy to 1 2 3 4 TRANSECTS 5 6 7 8 9 10 II MSS s 7-49 6-49 6-00 7-00 7-20 5-60 5-54 4-71 6-97 3-89 6-03 1-114 n 23-66 21-23 21-54 12-74 16 94 43-20 28-86 9-77 15-57 7-83 19-51 88-059 G 2-20 1-97 1-88 2-27 2-34 1-32 1-52 1-56 2-19 1-43 1-78 0-120 H’ 2-06 1-83 1-67 2-09 2-12 1-29 1-53 1-61 2-12 1-37 1-67 0-084 J’ 0-78 0-75 0-73 0-81 0-77 0-51 0-64 0-69 0-81 0-66 0-68 0-007 150 MEMOIRS OF THE QUEENSLAND MUSEUM reallocate, site-group III could be easily fused with site-group IV but it has been retained separately because of the presence of characteris- tically high frequencies of the isopod Sphaeroma and the bivalve Circe cf. australis. Species Classification Species classification using 115 species gave 10 species-groups which could be characterized broadly as ‘widespread’, ‘upstream’ or ‘down- stream’. Perusal of a two-way coincidence table of species-groups in sites showed however that many species were scattered throughout various sites TABLE 7: Abiotic Characterization or Site-groups. Depth Upstream Downstream Sand/gravel Mud Coarse Sand Fine Sand Sandy Mud Sand/Mud/ Gravel Shallow <2 m IX Intermediate 2-4 m II IV V VI VII Deep >4 m III 1 VIII Fig. 4; Dendrogram of site-groupings using 72 species. (X) indicates site-groups accepted. SAENGER ET AL: MACROBENTHOS OF THE CALLIOPE RIVER 151 Fig. 5: Schematic representation of the distribution of site-groups I-IX together with the individual sites. 152 MEMOIRS OF THE QUEENSLAND MUSEUM and only 39 species conformed at an equivalent to the OiO probability level, when using the pseudo-F test (Stephenson and Campbell 1977) with log transformed data. Reclassification using only 72 species resulted in 18 species-groups but again, little conceptual insight was gained. Since the site classification made good topographic and sedimentary sense, the 1 1 5 species were classified by using the species occurrences after standardization by the total number of occurrences (log transformed) within the already established site-groups. Classification of these data gave 14 species-groups (Fig. 6) which were accepted as forming more or less natural units (Table 8). Fig. 6: Dendrogram of species-grouping using 1 1 5 species. Crosses indicate species-groups accepted. The individual species comprising these species-groups are given in Table 8. SAENGER ET AL: MACROBENTHOS OE THE CALLIOPE RIVER 153 TABLE 8: Individual Species (By code number) in Species-Groups A-N as Accepted in Dendrogram (Fig. 6). Species Group Species Composition A 427 B 197 334 C 51 88 188 327 330 D 79 86 177 187 194 229 298 E 26 69 75 90 303 311 F 67 113 114 115 136 142 156 157 277 G 102 109 122 236 295 312 332 401 H 78 100 130 134 336 1 66 87 95 124 133 140 153 238 242 243 331 368 369 372 381 403 452 J 32 46 76 82 89 91 98 104 105 151 159 206 301 333 339 364 377 408 K 77 96 119 123 160 185 244 251 281 286 287 294 352 356 365 366 380 405 L 3 234 237 378 426 M 74 302 371 N 56 85 101 111 180 183 239 241 247 276 362 Species-groups in Site-groups From the species and site classifications, a coincidence table was prepared using the means of species-groups proportions within site-groups (Table 9); heavy type has been used for values greater than 12%, From this table it is readily recognizable which species-groups characterize TABLE 9: Mean Percentages of Species - groups in Site -GROUPS Species Group Site - group I 11 III IV V VI VII Vlii IX A 0 0 0 0 0 0 0 100 0 B 0 0 0 0 12-5 0 0 250 62-5 C 0 0 0 1-8 0 10-8 20-0 0 67-4 D 0 26-0 6-7 2-9 0 1-3 9-4 8-0 45-7 E 4-5 1-7 0 12-7 4-2 40-7 50 0 31 5 F 3-9 5-1 2-8 0-7 7-9 104 42-6 16 25-4 G 6-6 3-8 3-3 3-8 0 310 12 0 39-6 0-8 H 0 10-4 0 70 2-4 1 1-6 47-6 19-8 1-0 I 26 22-6 2-6 3-5 0-4 6-9 17-5 28-8 15-1 J 5 6 5-9 08 1-3 2-4 14-4 26-0 19-1 24-7 K 60 4-1 1-4 0-5 0-6 10-7 29-4 39-1 8-4 L 25» 5-0 0 0 0 12-8 0 38-6 17-8 M 5 0 41-7 60 6-0 0 7-7 0 22-3 1 1-3 N 0 73-3 0 0 0 8-2 7-1 8-3 3-2 which site-groups, and which site-groups are the focus of a species-group. For example the species-groups characterizing the most upstream site-group (II) are readily apparent (species-group M and N), while Table 8 allows the individual species in these species-groups to be identified by code number. TABLE 10; Characterization OF SiTE-GROUPS I - IX Site Groups Mean Distance Upstream (km) Predominant Substrate Mean Depth (m) Total No. Species Flora (mg/m^) Fauna (N/m2) H' G J’ Characterizing Species-groups I 2-6 ± 1-8 Sand 6-7 ±0-8 59 157 92-1 2-01 2-09 0-88 (-f)L (-)A,B,C,D,H,N II 12-7 ±2-5 Sand + Gravel 3-1 ± 1-2 99 5 332-7 1-37 1-36 0-57 ( + )D,I,M,N (-)A,B,C III 3-8 Mud 4 1 32 71 122-9 1-74 1-81 0-74 ( + )nil. (-)A,B,C,E,H,L,N IV 4-6 Mud 2-1 ±0-7 42 152 46-2 1-15 1-24 0-59 ( + )E (-)A,B,L,N V 4-6 Gravel 3-1 27 0 118-6 1-25 1-38 0-45 ( + )B (-)A,C,D,G,L,M,N VI 3-4 ±2-1 Sand + Mud 3-7 ±3-5 113 955 143-1 2-06 2-28 0-79 ( + )E,G,J,K,L (-)A,B VII 4-1 + 1-3 Mud + Sand 3-5 ±2-1 122 3 225-1 1-83 2-02 0-78 ( + )C,F,G,H,I,J,K (-)A,B,L,M VIII 5-8 ±2-4 Sand + Mud -f Gravel 5-0 ± 1-8 162 51 170-0 2-13 2-30 0-79 ( + )A,B,G,H,I,J,K L,M (-)C,E IX 4-8 ±2 0 Mud 1-4 ±0-6 128 713 171-1 1-71 1-71 0-70 ( + )B,C,D,E,F,I,J,L (-)A 154 MEMOIRS OF THE QUEENSLAND MUSEUM Not only positive but also negative character- izations can be seen (Table 9) and these have been used to give an overall characterization of the site-groups (Table 10). Times Classification Results of the times classification using 109 species are shown in Fig. 7. The times dendrogram can be accepted at either a 2-group level or a 4-group level; the former splits the early times (7’|_5) from later times (T’ 5 _ 7 ) while the latter gives Species classification based on times gave 17 species-groups; most of these were sequential groups. Two species-groups emerged which were present throughout the period with the exception of the two later summer times — March 1975 and Ts — February 1976): Species-group A contained only one species, the bivalve Solen correctus; species-group B contained the following 10 species — 98, 113, 157, 187, 188, 242, 244, 330, 366 and 401. Pseudo-F tests of conformity of species to the two levels of time-groupings were performed using a conformity level equivalent to 010. Sixty-two species failed to conform to either grouping, and it appears that the time-groupings are partly sequential and partly seasonal. Intertime dissimilarities are high; the lowest is at approximately 21% while the highest fusion is at approximately 67% dissimilarity. This suggests that either times were too distantly spaced for clear trends to show in the data, or that seasonal differences overlie a non-seasonal trend. On the basis of at least two seasonal species-groups and of cyclical analyses, the latter alternative appears to be the correct one, and no further time- classifications were undertaken. Fig. 7: Dendrogram of time-groupings using 107 species. Circles and crosses indicate the alternate 2-group or 4-group level of acceptance. SAENGER ET AL: MACROBENTHOS OF THE CALLIOPE RIVER 155 Seasonality Because of the low and scattered recordings of many individual species, no attempt was made to subject individual species to cyclic analyses after non-cyclic components had been identified. However two-stage regressions were determined for the summated data. One stage regressions were performed between times (in months) and the summated values of s, n, H\ G and J’. Linear, power curve fit, exponential logarithmic regressions were calculat- ed and significant regressions are given below, and superimposed on Fig. 8. s, linear, 5 = 5-25 t + 28-8 = 0-80 n, exponential, « = 187 e® r'^ = 0-76 G, linear, G = 0-554 t + 6-70 = 0-87 It will be noted that the underlying linear trend in the number of species becomes zero at -5-5 months, or approximately April 1974, while G becomes zero at -12.1 months, or approximately October 1973. TIME FROM OCTOBER 1974 (MONTHS) Fig. 8: Changes in parameters, s, G and n summated for all sites, for Tj-Tj Having identified the non-cyclic components in the data, the residuals were examined for near-annual eycles, using cycles of 9-5 to 13 months in increments of 0-5 months. The values of (proportion of variance take up) in each case were maximal for cycles of 10-5 months. Considering the total regressions (i.e. linear/ exponential plus 10*5 month cycle), for signifieance at 0 05 is 0-903 and at 0*01 is 0-967. The total values were: for 5 0-974 — highly significant; for n 0-849 — not significant; for G ~\ -00 — highly significant. Although cycles of 10-5 months gave the best fit to the present s values, subsequent sampling since September 1976 indicate that maxima occur annually in September and a 12-monthly period has been used to calculate the constants in the generalized equation, s = V 2 A° + y4cos (^t) -f- fisin (277-t) T T where ViA^ = 3*20 A = 19*28 B = -17-27 T= 12-0 Combining the linear and cyclic components gives, 5=19-28 cos (27Tt)- 17-27 sin (2zn) -F 5-25 / -F 32-0 12 12 This curve, superimposed on the data in Fig. 8a shows good agreement, and while reducing the R} value from 0-974 to 0-968, it still remains highly significant to the 0-01 level. DISCUSSION Faunistic Composition The total number of species (263) is less than the total species reported from Moreton Bay (Southern Region — 394 by Stephenson et al. 1974; central region — 182 by Raphael 1974; Middle Banks — 463 by Stephenson et al. 1978) and similar to the number from Port Curtis (251 — Stephenson et al. 1979), but greatly exceeds the number of species so far recorded from southeastern Queensland estuaries (98 — Serpentine Creek, Stephenson and Campbell 1977; 64 — eight estuaries sampled once after the 1974 floods, Campbell et al. 1974). In other Australian estuarine studies, the following numbers of macrobenthic species have been recorded: 55 — Blackwood River Estuary, W.A. (Hodgkin 1978); 74 — Wallis Lake, N.S.W. (Hutchings et al. 1978); 158 — Careel Bay, N.S.W. (Hutchings and Recher 1974); 246 156 MEMOIRS OF THE QUEENSLAND MUSEUM — Port Phillip Bay, V. (Poore and Kudenov 1978) and 571 — Westernport Bay, V. (Coleman et al. 1978). In terms of the numbers of organisms per square metre, the benthos of the Calliope River and Auckland Creek is compared with other southeastern Queensland estuaries (Table 11). The large numbers of Apseudes estuarius as found in Ninghi, Cabbage Tree and Serpentine Creeks (Campbell et al. 1974; Stephenson and Campbell 1977) have not been observed. Classification of species-groups in site-groups indicated that an ‘upstream’ community exists in the Calliope River, of which A. estuarius is a predominant component; it is thus possible that sampling further upstream than at present may have resulted in higher numbers of individuals per square metre. Site/Species Classification The classification of sites using their biotic composition gave site-groups which were easily characterized abiotically (Table 7), and it indicates that sites can be grouped according to their upstream/downstream position, the sediment type and the depth below low water. Classification of species using their standardized occurrences within site-groups, enabled site-groups to be characterized by the presence or absence of certain species-groups (Table 10). TABLE 11: Numbers of Benthic Organisms per M^ Recorded from Queensland Estuaries. System n/m^ Calliope River* 0 - 3,540 Auckland Creek* 0 - 1,950 Noosa River 80 - 6,080 Maroochy River 40 - 6,240 Mooloolah River** 463 - 6,48) Ninghi Creek 100 - 15,840 Caboolture River 200 - 11,680 Pine River 340 - 8,660 Cabbage Tree Creek 520 - 33,140 Serpentine Creek*** 132 - 14,128 Logan River 20 - 3,060 Coomera River 140 - 1,360 Nerang River 0 - 360 * Present survey Saenger unpubl. data •** Stephenson and Campbell, 1977 Other data based on Campbell et al, 1974 A point that merits emphasis is that with the exception of the upstream areas, the shallow-water and deeper-water biotas at a given distance upstream were quite different; a similar pattern was observed in Serpentine Creek (Stephenson and Campbell 1977). One obvious implication of this finding in relation to baseline benthic surveys is that to adequately sample an area, across-river transects should be used. Times Classification/Periodicities Times classification showed high times dis- similarities; initially this suggested that sampling times were too infrequent; however two times showed low values of s and n {Ti and T^), both were in late summer and both lacked two times-based species-groups; this suggests a seasonal component. Cyclic analyses of summated parameters confirmed this but indicated that an underlying non-cyclic trend was also present. The underlying linear trend was zeroed at -8-8 months i.e. January 1974, using the average estimate from s and G. December 1973 and January 1974 were particularly wet months with unusually heavy rainfall and river flow rates (Table 12), and extensive flooding occurred during this period. In fact on the 20 December 1973, a flow rate of 172,000 megalitres was recorded for the preceding 24-hour period; this 24-hour flow rate is comparable to the mean annual flow rate of normal years (Table 1). In view of this flooding, it became apparent that the benthic data could be best interpreted as a recolonization situation following a ‘catastrophic’ disturbance, and a model is proposed. TABLE 12: Gladstone Rainfall and Calliope River Discharges During October 1973 to May 1974. Month Monthly Rainfall (mm)* Monthly River Discharge (megalitres)** October 43 2,065 November 127 4,751 December 433 401,863 January 640 254,718 February 43 31,652 March 38 8,590 April 29 5,051 May 27 3,514 * Supplied by Department of Meteorology ** Supplied by Water Resources Council of Queens- land, from Castlehope Measuring Station. SAENGER ET AL: MACROBENTHOS OF THE CALLIOPE RIVER 157 Recolonization Model It is postulated that the intense flooding of December-January 1974 led to a more or less complete removal of the benthic fauna as a result of the combination of reduced salinity, physical scouring of the estuary and reduced oxygen levels. These factors have been identified as reducing the benthos elsewhere (Stephenson et al. 1977; Stephenson et al. 1979; Rosenberg 1977; Boesch et al. 1976). On returning to normal conditions, recoloniza- tion would commence. Evidence of recolonizalion comes from a superficial survey carried out in November 1973 (before flooding) near the power station site, which revealed 44 benthic species (A. Maluish unpubl. data). Another survey, during which 3 sites were sampled near the power station in April 1974 (after flooding), revealed ‘no organisms where these would have been expected’ (B. Campbell pers. comm.). The present data suggest that the re-appearance of species is more or less linear over a period of 29 months (Fig. 8a) while the numbers of individuals of all species increase logarithmically (Fig. 8c) during this period. Wolff et al. (1977) followed the colonization of a newly-formed seawater lake for 4-5 years and found that the numbers of individuals increased logarithmically over the entire 4-5 years. A linear increase in the number of species was found over the initial 27 months, but after this time, the number of species did not increase further. It could be expected that after a massive disturbance, opportunistis benthic species would be the initial recolonizers (Grassle and Grassle 1974; Boesch et al. 1976; Rhoads et al. 1978) with other species reappearing later. The cumulative number of species recorded at various times (Fig. 9) during the study indicates that new species arrived throughout the study, even though the cumulative number at each transect approaches a maximum (Fig. 9). This in turn suggests that genuine changes (as distinct from sampling inadequacies) in the species composition occurred during the 20 months of the study. In terms of the time scale involved, recoloniza- tion after ‘catastrophic’ and ‘seasonal’ events must be distinguished. Recovery after seasonal disturbance (e.g. freshwater inflow, high tempera- tures etc.) seems to occur in approximately 3 months in the present study area (Fig. 8a). In Port Curtis, seasonal recovery occurred after 2-3 months (Stephenson et al. 1979); in a west African estuary a period of 1-2 months was observed (Sandison and Flill 1959) while in a Dutch estuary, seasonal recovery occurred at 2-3 months (Wolff et al. 1977). Recovery from non-seasonal ‘catastrophic’ events takes longer and depends on the severity of the disturbance: MacGinitie (1939) reports recolonization after flooding was completed in 9 months; Kaplan et al. (1975) found that the benthos of a dredged channel had not recovered after 11 months; Stephenson et al. (1977) found general recovery 13 months after flooding in Bramble Bay, Queensland; Rhoads et al. (1978) found that colonization of spoil dumps required 1-5-2 years; Watson (1973) estimated that recovery of the benthos after pipeline dredging in Port Phillip Bay, Victoria, would be complete within 2 years; Boesch et al. (1976) found that the benthos had not recovered from the effects of a tropical cyclone after 2-5 years; Wolff et al. (1977) found that species numbers approached a maximum in a new seawater lake after 2*5 years. In the present study, the number of species has not approached a maximum within 29 months after severe flooding (Fig. 8a and 9), although preliminary analyses of subsequent survey data (Tg — November 1976 to Tjg — July 1979) suggest that recovery was more or less complete after 5 years. Fig 9: Cumulative number of species recorded for all site-groups and for site-groups VII. Species-area curves for catches at site-group VII are also shown. CUMULATIVE N2 0F SPECIES RECORDED FROM SITE-GROUP M 158 MEMOIRS OF THE QUEENSLAND MUSEUM While the present model, like all models, has tentative aspects, the data summarized in Figs. 8 and 9 show that the ‘pre-thermal’ situation is not a stable one, and that progressive changes are occurring. Since these can be expected to continue to occur, two elements are likely to interact during the ‘post-thermal’ times i.e. recovery from flooding and thermal effects. It will be essential to allow for the former in assessing the latter. ACKNOWLEDGMENTS We thank the various Specialists who have grappled with the taxonomic problems; these include Dr P. Hutchings (Polychaetes), Dr W. Ponder (Molluscs), Mr R. Burns (Opistho- branchs), Mr R. Monroe (Amphipods), Mr P. Davie (Brachyurans), Dr P. Mather (Ascidians), Dr D. Hoese (Fish), Dr P. Alderslade (Octocoral- lia), Mr D. Staples (Pycnogonids), Dr S. Edmonds (Echiura, Sipunculida), Dr L. Cannon (Platyhel- minthes). Dr F. Rowe (Echinoderms), Dr A.J. Bruce (Carids), Mr M. Johnston (Cumaceans) and Dr J. Ross (Bryozoans). We also thank the Department of Meteorology for the meteorological data and the Water Resources Council of Queensland for the Calliope River flow data. Finally we thank our colleagues in the Q.E.G.B. for their generous and varied assistance, particularly Mr J. Ewald for his constructive advice and comments; and the members of the Environmental Study Advisory Committee for their continual constructive review. LITERATURE CITED Boesch. D.F., Diaz, R.J. and Virnstein, R.W., 1976. Effects of tropical storm Agnes on soft-botton macrobenthic communities of the Jones and York estuaries and the lower Chesapeake Bay. Chesapeake Sci. 17: 24CL59. Campbell, B., Wallace, C., King, H. and Mather, P., 1974. ‘The estuarine sublittoral macrobenthos of southeastern Queensland: a preliminary survey.’ Pp. 41 + 4 + 25; 4 pis. (Qld. Museum: Brisbane). Coleman. F., 1972. Frequencies, tracks and intensities of tropical cyclones in the Australian region 1909-1969. Commonw. Aust., Bur. Meteorol., Meteorol. Summ. Pp. 42. Coleman, N., Cuff, W., Drummond, M. and Kudenov, J.D., 1978. A quantitative survey of the macrobenthos of Westernport, Victoria. Aust. J. mar. Freshw. Res. 29: 445-66. Conaghan, P.J., 1966. Sediments and sedimentary processes in Gladstone Harbour, Queensland, llniv. Qld. Dept. Geol. 6: 1-52. Grassle, J.F. and Grassle, J.P., 1974. Opportunistic life histories and genetic systems in marine benthic polychaetes. J. Mar. Res. 32: 253-84. Hodgkin, E.P., 1978. An environmental study of the Blackwood River Estuary, Western Australia, 1974—1975. W.A. Dept. Conservation and Environ- ment, Report No. 1. Pp. 78. Hofmann, G.W., 1971. Geological reconnaissance of a proposed power station site at Gladstone, Central Queensland. Geol. Surv. Qld. Dept, of Mines, Rep. No. 72: 1-10. Hutchings, P.A. and Recher, H.F., 1974. The fauna of Careel Bay with comments on the ecology of mangrove and sea-grass communities. Aust. Zool. 18: 99-128. Hutchings, P., Nicol, P.I. and O’Gower, A.K., 1978. The marine macrobenthic communities of Wallis and Smiths Lakes, New South Wales. Aust. J. Ecol. 3: 79-90. Jardine. E., 1925. The physiography of the Port Curtis district. Rep. Grt. Barr. Reef Comm. 1: 35-51. Kaplan, E., Welker, J.R., Draus, M.G. and McCourd, S., 1975. Some factors affecting the colonisation of a dredged channel. Mar. Biol. 32: 193-204. Macginitie, G.E., 1939. Some effects of fresh water on the fauna of a marine harbour. Amer. Midi. Nat. 21:681-86. Naylor, E., 1965. Effects of heated effluents upon marine and estuarine organisms. Adv. Mar. Biol. 3: 63-104. Paltridge, G.W. and Proctor, D., 1976. Monthly mean solar radiation statistics for Australia. Solar Energy 18: 235^3. Poore, G.C.B. and Kudenov, J.D., 1978. Benthos of the Port of Melbourne, the Yarra River and Hobson’s Bay, Victoria. Aust. J. mar. Freshw. Res. 29: 141-56. Rainer. S. and Hutchings, P., 1977. Nephtyidae (Polychaeta: Errantia) from Australia. Rec. Aust. Mus. 31 (8): 307-47. Raphael, Y.L, 1974. The macrobenthic fauna of Bramble Bay, Moreton Bay, Queensland. Unpubl. Thesis, Zoology Dept., University of Queensland. Rhoads, D.C., McCall, P.L. and Yingst, J.Y., 1978. Disturbance and Production on the estuarine seafloor. Amer. Scientist 66: 577-86. Rosenberg, R., 1977. Benthic macrofaunal dynamics, production and dispersion in an oxygen-deficient estuary of West Sweden. J. exp. mar. Biol. Ecol. 26: 107-33. Saenger, P., Stephenson, W. and Moverley, J., 1979. The subtidal fouling organisms of the Calliope River and Auckland Creek, Central Queensland. Mem. QdMus. 19: 399^12. Sandison, E.E. and Hill, M.B., 1959, The annual repopuiation of Lagos Harbour by sedentary marine animals. In SEARS, M., ‘Reprints of the 1st International Oceanographic Congress, 31st Aug.-12th Sept,, 1959’ Pp. 584-85. (Amer. Assoc. Advancement Science: Washington). SAENGER ET AL: MACROBENTHOS OF THE CALLIOPE RIVER 159 Staples, D.A., 1979. Three new species of Propallene (Pycnogonida: Callipallenidae) from Australian waters. Trans. R. Soc. S. Aust. 103; 85-93, Stephenson. W. and Campbell, B.M., 1977. The macrobenthos of Serpentine Creek. Mem. Qd Mus. 18; 75-93. Stephenson, W., Cook, S.D. and Raphael. Y.I., 1977. The effect of a major flood on the macrobenthos of Bramble Bay, Queensland. Mem. Qd Mus. 18; 95-119. Stephenson, W., Cook, S.D. and Newlands, S.J., 1978. The macrobenthos of the Middle banks area of Moreton Bay. Mem. Qd Mus, 18: 185-212. Stephenson, W., Saenger, P. and Cook, S.D., 1979. Problems in establishing a benthic baseline during a macrobenthos survey near Gladstone, Queensland. Proc. R.Soc. Qd.90: 21-32. Stephenson, W., Williams, W.T. and Cook, S.D., 1974. The benthic fauna of soft bottoms, southern Moreton Bay. Mem. Qd Mus. 17; 73-1 23. Thorhaug. a., Segar, D. and Roessler, M.A., 1973. Impact of a power plant on a sub-tropical estuarine environment. Mar. Poll. Bull. 4: 166-69. Watson, J.E., 1973. The impact of the ethane pipeline on the marine ecosystem of Port Phillip Bay, Victoria. Viet. Nat. 90: 60-65. Wolff. N.J., Sandee, A.J.J. and de Wolf, L., 1977. The development of a benthic ecosystem. Hydrobiologia 52 (1): 107-15. Young, J.S. and Frame, A.B., 1976. Some effects of a power plant effluent on estuarine epibenthic organisms. Jnt. Revue ges. Hydrobiol. Hydrogr. 61: 37-61. APPENDIX River Benthos Species List with Code Numbers (* indicates new species or major extension of range) POLYCHAETA 66 Amaeana trilobata 67 Ampharete sp. 68 Armandia intermedia 69 Branchiomma sp. 70 Cirratulus sp. 71 Cirriformia sp. 72 Capitella sp. ALGAE 1 Champia parvula 2 Enteromorpha clathrata 3 Heterosiphonia multiceps 4 Laurencia obtusa 5 Lophocladia sp. 6 Polysiphonia flaccidissima 1 Crouania sp. nov.* 8 Sporochnus comosus 9 Spyridia Jilamentosa 1 0 Ulothrix sp. 1 1 Callithamnion sp. 1 2 Soliera robusta 1 3 Griffithsia sp. 14 Ceramium cliftonianum SEAGRASSES 26 Halophila decipiens COELENTERATA 31 ANTHOZOAsp. I 32 ANTHOZOAsp. II 33 Virgularia gustaviana 34 Dendronephtya (Morchellana) sp.* 35 Virgularia gracillima NEMERTEA 46 Unident sp. I ECHIURA 51 ECHIURIDAEsp. I PLATYHELMINTHES 56 Stylochid sp. A 57 Planocerid sp. A BRYOZOA 61 Bugula c{. uniserialis 73 Coppingeria longisetosa 74 Ceratocephala sp. 75 Diopatra sp. 76 Glycera americana 77 Glycera sp. 78 HESIONIDAEsp. I 79 Heteromastus sp. 80 HETERONEREIIDAE sp. 1 8 1 Harmothoe sp. 82 Lysilla apheles 83 Lysilla pacifica 84 Laonice sp. 85 Leitoscoloplos normalis 86 Leitoscoloplos sp. r\o\.* 87 Lumbrinereis sp. 88 Lepidonotus sp. 89 MAGELONIDAEsp. I 90 MALDANIDAE sp. I 91 M ediomastus sp. 92 Marphysa sanguine a 93 Maldane sp. 94 NEREIlDAEsp. 1 95 Nephtys mesobranchia* 96 Notomastus sp, 97 OPHELIIDAEsp. I 98 Ophelina sp. 99 Owenia fusiformis 100 Poecilochaetus serpens 101 Poecilochaetus sp. 102 Pista pectinata 160 MEMOIRS OF THE QUEENSLAND MUSEUM 103 sp. nov.* 1 04 Prionospio queenslandica 1 05 Polydora sp. 106 POLYONIDAEsp. 1 107 Phyllodoce 108 Phy Uodoce sp.W 1 09 Pect inaria sp. 110 Scyphoproctus sp. 1 1 1 Scoloplos implex 1 1 2 Scoloplos johnstonei 1 1 3 Sthenolepis sp. 1 1 4 Sternaspis scutata 1 1 5 Stenelais sp. 1 1 6 Streblosoma amboinense 117 Streblosoma sp. 1 18 TEREBELLIDAE (s.f. Thelepinae) sp. I 1 1 9 Terebellides stroemi 1 20 Thelepus setosus 1 2 1 Pseudopolydora kempi 1 22 Onuphis sp. 123 Isolda pulchella 1 24 Magelona dakini 1 25 Lao no me sp. 126 Australonereis ehlersi 127 SERPULIDsp. I 1 28 Ancistrosyllis sp. 129 OLIGOCHAETE sp. I 1 30 Cossura sp. 1 3 1 Aonides sp. 1 32 Eunice c.f. australis 133 Ceratonereis erythraeensis 1 34 Euclymene sp. 135 Mediomastus calif orniensis 1 36 Scolelepis sp. 137 Syllis sp.\ 138 CHAETOPTERIDAE sp. I 139 SABELLIDAEsp. 1 40 Spio pacifica 141 Ophiodromus sp. 142 Syllis sp.U 1 50 Phoronis sp. NATANTIA 151 Alpheus sp. 152 Metapenaeopsis sp. 1 53 Parapenaeopsis sp. 1 54 Metapenaeus bennettae 1 56 Lysmata sp. 157 Macrobrachium intermedium 1 58 Latreutes mucronatus 1 59 Ogyrides delli 160 HIPPOLYTIDAEsp.il 162 Lucifer sp. BRACHYURA 176 Acheus lacertosus 177 Australoplax tridentata 1 78 Dorippa astuta 179 Dorippa sp. 1 80 Elamenopsis lineata 181 Eucrate sp. 182 Halicarcinus australis 1 83 Halicarcinus sp. A 1 84 Halicarcinus bedfordi 185 Heteropanope sp. 1 86 Hyastenus sp. 187 Ilyoplax dentatus 1 88 Ilyoplax orientalis 1 89 Macrophthalmus punctulatus 190 Macrophthalmus latreillei 191 Matuta sp. 192 Neorhynchoplax sp. 1 93 Nursia abbreviata 1 94 Paracleistostoma mcneilli 1 95 Petrolisthes teres 1 96 Petrolisthes cf. scabriculus 197 Petrolisthes sp. noy. 1 98 Phalangipus sp. 199 Pilumnus sp. \ 200 Pilumnus sp. B 201 Pilumnus c(. hirsutus 202 Portunus pelagicus 203 Raphidopus cilatus 204 Sesarma semperi longicristatum 205 Vca sp. 206 Xenophthalmus pinnotheroides AMPHIPODA 226 AMPELISCIDAEsp. 227 AMPHIPODA sp. I 228 AMPHIPODAsp.il 229 AMPHIPODA sp. Ill 230 AMPHIPODAsp.lv 231 AMPHIPODA sp. V 232 AMPHIPODA sp. VI 233 AMPHIPODA sp. VII 234 CAPRELLIDAEsp. I 235 CAPRELLIDAE sp. II 236 Ceradocus sp. 237 Corophium sp. 238 Dryopoides sp. 239 Ericthonius sp. 240 Eriopisa sp. I 241 Eriopisa sp.W 242 Grandidierella sp. I 243 Grandidierella sp. II 244 Grandidierella sp. Ill 245 Maera sp. I 246 Maera sp. II 247 Melita sp. 248 Oedicerateda sp. I 249 Oedicerateda sp. II 250 Paracorophium sp. 25 1 Paraphoxus sp, 252 Quadrivisio sp. 253 AMPHIPODA VIII SAENGER ET AL: MACROBENTHOS OF THE CALLIOPE RIVER 161 OTHER CRUSTACEANS 276 Apseudes estuarius 277 Apseudes W 278 Apseudes \W 279 Tanais sp. 281 OSTRACODA sp. I 282 OSTRACODAsp.il 285 Potnacuma australiae 286 Dimorphostylis australis 287 Eocuma agrion 288 Cyclaspis ? usitata 289 Campylaspis latidactyla 294 PAGURIDAEsp. I 295 MYSIDAEsp. I 296 MYSIDAEsp. II 297 MYSIDAEsp. Ill 298 MYSIDAEsp. IV ISOPODA 301 Mesanthura sp. 302 Sphaeroma sp. 303 Synidotea sp. 304 Paracereis sp. 305 Eurydice sp. STOMATOPODA 3 1 1 Squilla sp. 312 Callianassa sp. 313 Thalassina anomala PYCNOGONIDAE 3 1 6 Propal lene sae tiger i * 3 1 7 Hemichela sp. nov.* 318 Anoplodactylus t ubi ferns * GASTROPODA 326 Aglaja sp. nov.* 327 ^/>'5'sp. 328 Bedeva hartley i 329 Cerithideopsilla cinguiata 330 Haminoea sp. 33 1 Nassarius burchardi 332 Nassarius dorsatus 333 Nassarius sp. 334 Polynices sordida 335 Pseudoraphitoma sp. 336 Ringicuta sp. 337 Zafra sp. 338 Zeacumantus sp. 339 Philine sp. nov. 340 Tornatina sp. 341 Trinchesia sp. 342 Facelinella sp. PELECYPODA 351 Anomia descripta 352 Amygdalum sp. 353 ARCIDAEsp. I 354 Barnea sp.\ 355 Barnea sp.W 356 Circe cf. australis 357 C/Vcfsp. II 358 Corbula sp. 359 Didimacar repeata 360 Ennucula sp. 361 Ensiculus cultellus 362 Fluviolanatus amarus 363 Gari lessoni '364 Gouldia sp.\ 365 Gouldia sp.W 366 Laternula sp. I 367 Laternula sp.W* 368 cf. Laternula sp. 369 Limaria noverea 370 Mactra sp. 371 Modiolus auriculatus 372 Notospisula parva producta 373 Nuculana sp. 374 Pharella wardi 375 Placamen sp. 376 Saccostrea cf. cucullata 111 Solecurtus sp. 378 Solen correctus 379 Tapes hiantina 380 Tellina semen 381 Tellina sp. 382 Trisidos tortuosa 383 VENERIDAE sp. I 384 Volachlamys singaporinus 400 Dentalium sp. ECHINODERMATA 401 Amphioplus personatus 402 Amphioplus hastatus 403 Amphiura bidentata* 404 Amphiura leptotata* 405 Amphiura micra* 406 Amphiura phrixa* 407 Amphiura tenuis* 408 Protankyra verrilli 409 Amphiura octacantha* 4 1 0 Amphipholis squamata TUNICATA 426 Ascidia sydneyensis 427 Molgula mollis 428 Cnemidocarpa etheridgii 429 Microcosmus australis PISCES 451 APOGONlDAEsp. 1 452 Brachyamblyopus cf. urolepis 453 Drombus palackyi 454 Prionobutis microps 455 Favonigobius sp. Mem. Qd M us. 20(1): 163-70, [1980] THE ISOPOD GENUS DYNAMENE FROM AUSTRALIAN WATERS, WITH DESCRIPTION OF A NEW SPECIES FROM CORAL REEFS D.M. Holdich and K. Harrison Department of Zoology, The University, Nottingham, England. ABSTRACT A new species of the isopod genus Dynamene, D. curalii, is described from the coral habitat of Heron Island and Lizard Island in Queensland. The only other species of this genus known from the Southern Hemisphere, D. ramuscula Baker from South Australia, is redescribed. It is noted that the two Australian species differ markedly in one of the diagnostic characters when compared with species from Europe and N.W. Africa. Seven species of the sphaeromatid isopod Dynamene Leach, 1814 are known to date (Holdich 1968a, 1970). Of these species D. bidentata (Adams) is known only from the Atlantic coasts and islands of N.W. Africa and Western Europe; D. magnitorata Holdich from the coasts of continental Western Europe (Holdich 1976), the Azores, and Mediterranean; D. edwardsi Lucas has been recorded from Atlantic coasts of Europe and N.W. Africa but mainly occurs in the Mediterranean, Aegean and Red Sea; D. bifida Torelli and D. tubicauda Holdich appear to be restricted to the Mediterran- ean; and D. torelliae Holdich occurs in the Mediterranean, Adriatic, Aegean, and also extends into the Black Sea. These species therefore appear to be restricted in their distribution to Western Europe, N.W. Africa and the Mediterranean and adjacent seas. There have been frequent misidentifications in the past (Holdich 1968a for review) but the only valid record for a species outside that area is D. ramuscula Baker from South Australia. This species appears, from the literature, to closely resemble D. bifida. The only low latitude record for a Dynamene species is for D. edwardsi from southern Spanish Sahara (latitude 22°N). In this paper a new species is described from coral reefs in (Queensland, Australia which appears to have a tropical and sub-tropical distribution. When comparing the new species with the type specimens of D. ramuscula it was discovered that the latter species in fact differs notably from D. bifida and a redescription was considered necessary. SYSTEMATICS Order ISOPODA Family SPHAEROMATIDAE Sub-family EUBRANCHIATINAE Dynamene Leach, 1814 Generic Diagnosis Eubranchiate sphaeromatid with body approx- imately elliptical. Anteriorly, cephalosome separ- ating the bases of the antennules. Eyes set slightly into pereonal tergite 1 . Coxal plates of pereonites 1-7 separated from tergites by sutures. Pereonal tergite 7 reduced, not reaching lateral margins of body. Each side of the pleonal tergite bearing two short, postero-lateral incisions. Pleotelson domed, bearing an obvious terminal notch which may be enclosed forming a tube. Antennular peduncle articles 1 and 2 dilated and juxtaposed to ventral margin of cephalosome. All pereopods ambula- tory. Both rami of pleopods 1-3 bearing margin of plumose seate. Endopods of uropods juxtaposed to pleotelsonic margin; exopods posteriorly directed. Sexual dimorphism pronounced. Adult male with pereonal tergite 6 longer than those preceding, posterior margin with an elongate, posteriorly 164 MEMOIRS OF THE QUEENSLAND MUSEUM directed process either side of mid-line. Pleotelson bearing transverse, bilaterally symmetrical ornamentation. Penes small, separate. Endopod of pleopod 2 lacking appendix masculina. Ovigerous female with pereonal tergite 6 similar to those preceding. Pleotelson smooth. Young incubated in a marsupium, formed from four pairs of lamellae which arise from pereonites 1-4. Mouthparts strongly metamorphosed. Dynamene curalii sp. nov. Material Examined Holotype: Queensland Museum QM W6319, male, from Lithothamnion sp. behind reef crest, intertidal, Heron Island, SEQ, D.M. Holdich, 15 April 1976. Allotype; QM W6320, ovigerous female, same data as Holotype. Paratypes: QM W6321, juvenile, from Halimeda sp. behind reef crest, intertidal. Heron Island, SEQ, D.M. Holdich, 15 April 1976. QM W6322, sub-adult male, behind reef crest opposite research station, intertidal, Heron Is., SEQ, N. Bruce, 6 June 1978. QM W6323, non-ovigerous female, same data as W6322. Other Material Heron Island, Queensland, from Lithothamnion sp. behind reef crest. Intertidal, 4 adult males, 2 ovigerous females, 1 juvenile, Coll. D.M. Holdich, 15 April 1976. North Point, Lizard Island, Queensland, On Halimeda sp. associated with coral at 9 0 m, 1 sub-adult male. Coll. B. Barnett, 8 June 1976. Heron Island, Queensland, Behind reef crest opposite research station. Intertidal, 2 adult males, 1 sub-adult male, 4 ovigerous females, 2 juveniles. Coll. N. BTuce, 6 June 1978. Description Adult Male: (Figs, la - c, f - h; 2d - v). Body deeply vaulted, not depressed. Dorsal surface and lateral margins bearing long setae. Frontal region of cephalosome, between antennules, sub-equal in width to labrum. Pereonal tergite 1 sub-equal in length to cephalosome and twice as long as tergite 2. Tergites 2-5 sub-equal. All coxal plates directed ventrally, not expanded. Pereonal tergite 6 three times as long as tergite 5 and rugose in its posterior half. Posterior projections rugose, simple (i.e. not bifid), widely separate, just more than half length of tergite. Posterior tergal margin, lateral to the projections, straight, bearing no secondary projections. Postero-lateral margin of pereonal tergite 7 as a simple lobe, not bifid. Pleonal tergite rugose laterally. Pleotelson bearing a short, simple or multispined tubercle either side of the mid-line. Pleotelsonic foramen with sub-triangular, serrate lip; entirely enclosed ventrally, forming a tube. Antennule with peduncle articles 1 and 2 setose; 1 slightly longer than 2, bearing a distal inferior spine; 2 bearing a proximal, inferior, curved spine and a distal infero-dorsal spine. Distal region of article 2 acute. Antennular flagellum of six articles, 2-5 each bearing one long aesthetasc. Antennal peduncle not expanded; flagellum of seven articles, last very reduced. Mouthparts: labrum slightly wider than long with setose distal margin; proximal region set into the broad epistome which has a straight anterior margin. Mandibles with incisor processes, molar processes and palps well developed. Maxillule with tip of inner lobe bearing four pectinate setae and outer lobe bearing approximately ten stout, curved spines. Maxilla with two outer lobes bearing long, curved, robust, simple setae, and inner lobe bearing eight long, robust, pectinate setae. Maxillipedal endite four times as long as broad with stout, plumose, apical setae. Inner margin bearing one coupling hook. Palp of five articles; 1 reduced, 2-4 bearing pronounced, narrow lobes, 2-5 with apical, plumose setae. Pereopods all ambulatory, first slightly more robust than remainder. Dactylus of each with pronounced secondary unguis. Merus, carpus and propus bearing stout setae on superior and inferior surfaces. Penes small, separate. Pleopods: pleopod 1 with internal margin of basis extended medially beyond rami, bearing three coupling hooks. Rami sub-elliptical, each bearing apical, plumose setae equal in length to ramus, lateral simple setae and several long, non-marginal setae. Endopod two-thirds length of exopod. Basis of pleopod 2 bearing three coupling hooks. Endopod sub-triangular; exopod ovate with several proximal, non-marginal setae. Both rami bearing long, plumose, terminal setae. Basis of pleopod 3 rectangular with three coupling hooks. Endopod sub-triangular, twice as long as elliptical exopod. Both rami with terminal, and exopod also with external, plumose setae, those on exopod twice as long as those on endopod. Both rami of pleopod 4 sub-ovate, bearing respiratory folds along entire length. Apex of exopod thickened, toothed; external margin with proximal, toothed process. Apex of endopod acute, curved medially. Both rami of pleopod 5 extensively pleated. Endopod sub-elliptical with apex broadly rounded, bearing fine setae. Exopod sub-ovate with three very pronounced, toothed apical processes. Both rami of all pleopods lacking articulations. Uropods HOLDICH AND HARRISON; AUSTRALIAN DYNAMENE 165 Fig. 1: Dynamene curalii sp. nov. Adult male (a) dorsal, (b) lateral, (c) ventral, (f) right antennule, ventral view, (g) right antennular peduncle, antero-dorsal view, (h) left antenna. Sub-adult male posterior region (d) dorsal, (e) lateral. Ovigerous female (i) dorsal, (j) lateral, (k) ventral. Scales represent 1 mm. 166 MEMOIRS OF THE QUEENSLAND MUSEUM with wide, rectangular endopods, each internal corner bearing a pronounced spine. Margins straight or markedly spined, setose. Exopod setose, rugose, sub-cylindrical, tapering to an acute, medio-dorsally directed apex bearing a short, external, accessory spine. Body length: 2-44 mm. Sub-adult Male: (Fig. Id, e). Pereonal tergite 6 twice as long as 5, posterior margin bearing a short, blunt projection either side of the mid-line. Projections on pereonal tergite not contiguous. Pleotelson domed, setose, lacking obvious rugosity and ornamentation. Pleotelsonic foramen sub- circular, smooth-lipped. Penes present as minute separate papillae. Uropod with rectangular endopod bearing a spine on the medio-posterior corner. Exopod sub-cylindrical, rugose, sub-equal in length to endopod, with a curved, acute tip. Body length: 212 mm. OviGEROUS Female: (Figs, li-k; 2w-z). Pleotelsonic foramen sub-circular, directed dorsally, completely enclosed ventrally forming a tube. Pereopods as in male, no obvious increase in setation. Four pairs of brood lamellae well developed, overlapping considerably in mid-line. Approximately twelve eggs in brood pouch. Uropod with rectangular endopod, either lacking spines or with a single medio-posterior spine. Exopod rugose, sub-cylindrical, sub-equal in length to endopod, lacking curved, acute tip. Body length: 2-3 mm. Non-ovigerous Female: (Fig. 2a, b). Pleotel- son more domed than in ovigerous female, with foramen sub-circular, smooth-lipped, directed postero-dorsally. Antennule bearing two aesthe- tascs. Exopod of uropod with acute terminal spine. Body length: 2-48 mm. There is a tendency for some Dynamene species to decrease slightly in apparent length at the ovigerous moult (Holdich 1968b) hence the difference in the lengths of the non-ovigerous and ovigerous females recorded here. Juvenile: (Fig. 2c). Body slightly depressed, coxal plates directed laterally. Pleotelsonic foramen sub-circular, directed posteriorly. Exopod of uropod acute, sub-equal to endopod. Remarks Dynamene curalii can be separated from all known species of Dynamene, except D. tubicauda. by the lack of a central pleotelsonic boss in the adult male. Unlike D. tubicauda, however, the adult male of D. curalii is not depressed, the anterior region of the cephalosome is not extended as a shelf and the exopods of the uropods bear incurved, acute apices. (See also remarks following D. ramuscula). ETYMOLOGY: Dynamene + L. curalii i.e. of coral. Ecology and Life-cycle On Heron Island this species occurs intertidally on and behind the reef crest. As with other Dynamene species (Holdich 1968b, 1970, 1976) the adults take up a crytozoic existence for the terminal reproductive phase of their life-cycle, occupying small cavities amongst dead coral and lithified algae. In some cases the habitats are exposed to strong wave action. As the ovigerous females undergo marked degeneration at the ovigerous moult (as in other species of Dynamene) they probably die after release of the brood. Juveniles, after release from the brood pouch, settle and feed on various species of algae on the reef. It is not known how long the growth phase of the life-cycle is, or whether males spend one or two breeding seasons in the reproductive habitat. On Lizard Island two sub-adult males were collected (one subsequently escaped) from sub-littoral algae thus indicating that D. curalii is not restricted to the intertidal zone. Other species of isopod which have a biphasic life-cycle occur in similar reproductive habitats to Dynamene and probably compete with it for space. Notable amongst these is the genus Gnathia Leach which has recently been found to be a common element in cryptic habitats in Queens- land (Holdich and Harrison, 1980). Dynamene ramuscula Baker, 1908 Dynamene ramuscula Baker, 1908, pp, 45-6, 161, pi. 5; Hale, 1929, pp. 293^; Monod, 1932, p. 64; Holdich, 1968a, pp. 412-3; 1970, p. 422. Dynamene ramusculus Nierstrasz, 1931, p. 211. (Unjustified emendation). Material Examined South Australian Museum reg. no. C.355. 2 males, 2 ovigerous females, St. Vincent Gulf, South Australia. On sponges. Description Adult Male (Fig. 3a-k, n-p). Body deeply vaulted, not depressed. Dorsal surface, uropods HOLDICH AND HARRISON; AUSTRALIAN DYNAMENE 167 Fig. 2: Dynamene curalii sp. nov. Non-ovigerous female (a) dorsal, (b) lateral, p>osterior region. Juvenile (c) dorsal. Adult male (d-j) pereopods 1-7 respectively, (k) penes, (1-n) pleopods 1-3 respectively, (o) rami of pleopod 4, (p) rami of pleopod 5, (q) mandibles, posterior view, (r) labrum and epistome, (s) maxillule, (t) maxilla, (u) maxillipede, (v) paragnaths. Ovigerous female (w) maxillipede, (x) mandible, (y) maxillule, (z) maxilla. Scale represents 1 mm. 168 MEMOIRS OF THE QUEENSLAND MUSEUM and ventral surface of pleotelson bearing long setae. Frontal region of cephalosome, between antennules, narrow, one quarter width of labrum. Pereonal tergite 1 sub-equal in length to cephalosome and twice as long as tergite 2. Tergites 2-5 sub-equal, each rugose in posterior half. All coxal plates directed ventrally, not expanded. Pereonal tergite 6 more than twice as long as tergite 5, rugose. Posterior projections sinuous, contiguous in mid-line, diverging distally, sub-parallel terminally, each bearing a ventrally directed, sub-terminal, accessory projection. Posterior tergal margin bearing a slight, rugose tubercle mid-way between projection and coxal plate on each side. Postero-lateral margin of pereonal tergite 7 as a simple lobe, not bifid. Pleonal tergite rugose, obscure in dorsal view. Pleotelson rugose, having a quincunx of acute, conical tubercles, the central being the largest, with a smaller, conical tubercle anterior to the basis of each uropod. Pleotelsonic foramen sub-triangular, closed ventrally forming a tube. Antennular peduncle with article 1 having a short, blunt, distal, inferior spine; 2 with a larger, wide, proximal, inferior spine and a blunt apex. Flagellum of eight articles, 3-8 each bearing a single aesthetasc. Antennal peduncle not expand- ed, flagellum of nine articles, distal article reduced. Mouthparts: labrum sub-circular, semi-enclosed by V-shaped epistome. Mandible with well formed incisor process, molar process and palp. Palp article 2 with five distal, inferior setae. Left mandible with tridentate incisor process, triden- tate lacinia mobilis and setal row of six pectinate setae. Maxillule and maxilla of usual sphaeroma- tid form. Maxillipedal endite four times as long as broad. Palp articles 2—4 with pronounced superior lobes, relatively longer than those of D. curalii sp. nov. Pereopods all ambulatory, first slightly more robust than remainder, with reduced carpus and inferior margins of merus, carpus and propus bearing long, stout, simple spines. All pereopods with pronounced secondary unguis, 2-7 with variable setation on inferior margins. Penes short, widely separate at base. Pleopods: pleopod 1 with internal margin of basis not extending medially beyond rami. Rami sub-elliptical with long, terminal and external, marginal, plumose setae. Endopod three quarters length of exopod. Pleopod 2 with endopod sub-triangular, exopod sub-ovate. Rami sub-equal in length, each bearing long, terminal and external setae. Pleopod 3 with endopod twice length of exopod. Endopod sub-triangular with short. terminal, plumose setae. Exopod sub-ovate with terminal, plumose setae extending beyond tip of endopod. Pleopod 4 of specimen damaged, but external margin of exopod having proximal, toothed process. Both rami of pleopod 5 with respiratory folds. Endopod sub-elliptical with apical border of short, fine setae. Exopod sub-elliptical with truncate base. Apex as a short, wide, toothed region with a prominent sub- terminal toothed process and a toothed region on the distal, internal margin. All observed pleopodal rami lacking articulations. Uropods with basis bearing conical, dorsal tubercle. Endopod rugose, with a blunt, distal, posterior spine; antero-medial margin smoothly rounded; posterior margin uneven. Exopod one and a half times length of endopod, sub-cylindrical, rugose, acute tip lacking accessory spine. Body length; 4-23 mm. OviGEROUS Female (Fig. 3 1, m) Body sub-elliptical, setose; lateral pereonal margins strongly convex. Pleotelson with low, median tubercle. Sub-circular foramen directed postero-dorsally, enclosed ventrally forming a tube. Uropod with rami sub-equal, smoothly rounded apically. Body length: 4-23 mm. Remarks The similarity between D. ramuscula. as figured by Baker (1908) and D. bifida from the Mediterranean has been mentioned in the literature (Monod 1932, Holdich 1968a). In fact these two species can be easily separated by the very different nature of the pleotelsonic ornamentation of the males (that of D. bifida being restricted to a single, central, low, bifid boss) and by the form of the posterior pereonal projections in the male (those of D. bifida being sub-parallel, separate at the base and not markedly divergent). ETYMOLOGY; Dynamene -I- L. ramuscula i.e. branching, like a twig. Discussion The occurrence in Australia of a second species of Dynamene is very noteworthy. Previously, only D. ramuscula from South Australia was known. This species has only been recorded once. Examination of specimens of D, bifida and D. ramuscula have shown conclusively that, on morphological grounds, they are distinct species. HOLDICH AND HARRISON: AUSTRALIAN DYNAMENE 169 Fig. 3: Dynamene ramuscula Baker. Adult male (a) dorsal, (b) lateral, (c) ventral, (d) pereoptod 1, (e) pereopod 3, (0 maxillipede, (h) antennule, (i) antenna, (g, j, k) pleopods 1-3 respectively, (n) proximal external margin of damaged exopod of pleopod 4, (o) exopod of pleopod 5, (p) endopod of pleopod 5. Ovigerous female (1) dorsal, (m) lateral. Scales represent 1 mm. 170 MEMOIRS OF THE QUEENSLAND MUSEUM D. curallii sp. nov. is the first species of the genus Dynamene to be recorded for sub-tropical and tropical habitats in the Southern Hemisphere. In the Northern Hemisphere only D. edwardsi has been recorded from within the tropics. During the present study it was noted that in the two Australian species the coxal plates of pereonite 7 of the adult males have smoothly rounded posterior margins. This contrasts markedly with the strongly bidentate margins found in all the Afro-European species and clearly separates the two geographical groups. Males of D. ramuscula can easily be separated from those of D. curallii by the structure of the pleotelsonic ornamentation and pereonal spines, and by the absence of a tubercle on the postero-lateral tergal margin of pereonite 6 of D. curallii. Other factors separating the two species are: the shape of the epistome (that of D. ramuscula being acute and that of the new species being transverse); the relative width of that region of the cephalosome lying between the antennular bases (that of D. ramuscula being proportionally much narrower than that of the new species); and the relative lengths of the postero-ventral pleotelsonic shelf (that of D. ramuscula being proportionally shorter than that of the new species). Dynamene is a common component of the epifauna of seaweeds and the infauna of rock crevices, empty barnacle tests and sponges in the Mediterranean and adjacent seas and on the Atlantic coasts of Western Europe and NW, Africa. In its reproductive habitat at least it has few isopod competitors. However, in other geographical locations its intertidal niche is often filled by Dynamenella spp. This is also the case on the Queensland mainland coast and in the intertidal zone of continental islands, such as Lizard Island. Three species of Dynamenella were found (Holdich, unpubl. obs.) associated with the rocky intertidal zone on Lizard Island but D. curalii was only found associated with the coral reef. On Heron Island, however, where the bulk of the intertidal zone is composed of coral, Dynamenella was restricted to the upper eulittoral beach rock. The apparent discontinuous geographical distribution of Dynamene may not in fact be so marked now that it is known that some members of this genus occupy coral habitats. More intensive collecting is needed in the Indo-Pacific coral habitat to show whether this is the case or not. ACKNOWLEDGMENTS Thanks are due to the Nuffield Foundation for a Fellowship in Tropical Marine Biology to D.M. Holdich, to the Natural Environment Research Council for a research grant, and the Directors of the Heron Island and Lizard Islands* Research Stations for provision of research facilities. Thanks are also due to Dr W. Zeidler (South Australian Museum) for the loan of specimens of D. ramuscula, and to N. Bruce and B. Barnett for the gift of specimens of the new species described here. LITERATURE CITED Baker, W.H., 1908. Notes on some species of the isopod family Sphaeromidae, from the South Australian coast. Trans. Roy. Soc. S. Australia 32: 1 38-62. Hale. H.M., 1929. The crustaceans of South Australia Pt. 2: 269-310. ‘Handbooks of the flora and fauna of South Australia.’ (British Science Guild, S. Australia Branch: Adelaide). Holdich, D M., 1968a. A systematic revision of the genus Dynamene (Crustacea: Isopoda), with descriptions of three new species. Publ. Staz. zool. Napoli 36: 401-26. 1968b. Reproduction, growth and bionomics of Dynamene bidentata (Crustacea, Isopoda). Proc. zool. Soc. Lond. 156 : 1 37-53. 1970. The distribution and habitat preferences of the Afro-European spwcies of Dynamene (Crustacea: Isopoda). J. nat. Hist. 4 : 419-38. 1976. A comparison of the ecology and life cycles of two species of littoral isopod. J. exp. mar. Biol. Ecol. 24 : 133^9. Holdich, D.M. and Harrison, K., 1980. The isopod genus Gnathia Leach from Queensland waters with descriptions of nine new species Aust. J. Mar. Freshwat. Res. (in press). Monod, T., 1932. Tanaidaces et Isopodes aquatiques de I’Afrique occidentale et septentrionale. Pt. 3. Sphaeromatidae. Mem. Soc. Sci. Nat. Maroc. 29: 1-91. Nierstrasz, H.F., 1931. Die Isopoden der Siboga- Expedition. 3. Isopoda Genuina. 2. Flabellifera. In Weber, M,, Siboga-Expedite. No. 114. (Monogr. 32c): 123-233. Mem. Qd M us. 20(1): 171-9. [1980] A NEW TAENIACANTHID COPEPOD FROM THE ESOPHAGUS OF A SEA URCHIN IN QUEENSLAND Arthur G. Humes Boston University Marine Program, Marine Biological Laboratory, Woods Hole, Massachusetts, U.S.A. ABSTRACT The cyclopoid copepod Clavisodalis salmacidis sp. nov., the fifth taeniacanthid to be reported from sea urchins, occurs in the esophagus of the temnopleurid Saimacis belli Doderlein in Queensland. The new species may be distinguished from its only eongener by the female genital segment being about as long as wide, the caudal ramus separated from the anal segment, and the claw of the female maxilliped unguiform and recurved with two rows of spines on its attenuated tip, the claw of the male maxilliped terminating in a spinulose area. Four taeniacanthid copepods have been described from Echinoidea. Humes and Cressey (1961) described three species from Diadema setosum (Leske) in Madagascar — Echinosocius pectinatus, Echinosocius dentatus, and Echinirus laxatus. Humes (1970) described Clavisodalis heterocentroti from Heterocentrotus trigonarius (Lamarck) at Eniwetok Atoll. In addition, Gooding (1965) found two more species of Echinosocius and another species of Echinirus living with Diadema at Singapore, but did not describe them. Other taeniacanthids occur on fishes. The dissection of three Aristotle’s lanterns of Saimacis belli, preserved in alcohol, revealed the copepods which are the subject of this paper. The copepods were found clinging by means of the second antennae to the lining of the esophagus. Materials and Methods The copepods were cleared in lactic acid. Dissections were prepared using the method described by Humes and Gooding (1964). The figures were drawn with the aid of a camera lucida. The letter after the explanation of each figure refers to the scale at which it was drawn. The abbreviations used are: A, = first antenna, MD = mandible, P = paragnath, MX] = first maxilla, and P ]_4 = legs 1-4. Family TAENIACANTHIDAE Wilson, 1911 Genus Clavisodalis Humes, 1970 Clavisodalis salmacidis sp. nov. (Figs, 1-42) Material Examined Holotype; QM W7977, 9» bom oesophagus of Saimacis belli Doderlein, trawled in Moreton Bay, SE Queensland, by C. Boel, 25.V.1977. Paratypes: QM W7978, cf, allotype, same data as holotype. QM W7979, 3 cT, 5 9, QM W7980, 9 copepodids, same data as holotype. One c^, 2 9i >n the collection of the author (dissected). Description Female: Body (Figs. 1, 2) elongate and moderately flattened, dorsoventral thickness of prosome about 0-47 mm. Length (not including setae on caudal rami) 2 07 mm (1 •98-2 24 mm) and greatest width 0*68 mm (0-62-0-73 mm), based on four specimens in lactic acid. Segment of leg 2 almost completely fused with segment of leg 1; slight separation seen only in lateral view (Fig. 2). Anterior part of prosome guitar-shaped in dorsal view and wider than succeeding segments. Ratio of length to width of prosome 2-25:1. Ratio of length of prosome to that of urosome 2-34: 1 . 172 MEMOIRS OF THE QUEENSLAND MUSEUM Segment of leg 5 (Fig. 3) 1 30 x 234 f^m. Genital segment 260 ^m long, in dorsal view expanded in its anterior half (width 263 ^m) and with nearly parallel sides in its posterior half (width 164 /^m). Genital areas situated laterally on expanded part. Each area (Fig. 4) with three smooth setae 75, 37, and 10 fdm long. Three postgenital segments from anterior to posterior 101 x 148, 55 x 140, and 75 x 166 /vm. Anal segment with lateroventral row of spinules on both sides (Fig. 6 ). Caudal ramus (Figs. 5, 6 ) moderately elongate, 1 27 X 68 f/m, ratio of length to width 1-87:1. Outer lateral seta 51 ;/m and dorsal seta 40 both smooth. Outermost terminal seta 78 /vm, innermost terminal seta 75 /^m, and two median terminal setae 180 ^m (outer) and 300 /^m (inner), all with small lateral barbules. Ventral row of spinules near insertion of outermost terminal seta. Left caudal ramus of one female lacking dorsal seta and having extra row of spinules near outermost terminal seta (Fig. 7). Body surface smooth with very few hairs (sensilla). Egg sac (Figs. 1, 8 , 9, 10) with 3-14 eggs, each about 170 itm in average diameter. Largest egg sac seen (Fig. 10) 0-74 x 0-40 mm, with 1 4 eggs. Rostral area (Fig. 11) broad and projecting slightly anteriorly. First antenna (Fig. 12) 323 /^m long and 5-segmented. Lengths of segments (measured along their posterior nonsetiferous margins): 73 (143 //m along anterior margin), 75, 44, 35, and 26 ym respectively. Formula for armature: 19, 8 , 4, 2 + 1 aesthete, and 7 + 1 aesthete. Many setae on first segment with short barbules; setae on segments 2-5 smooth. Second antenna (Figs. 13, 14) 3-segmented, with formula 1, 1, 7 (terminally with three clawiike spines and three setae and subterminally with a very small seta). Seta on second segment with truncate, very finely denticulate tip. Third segment ornamented on its inner surface with spinules. Distal outer corner of segment produced to subacuminate point and bearing row of spinules and small setiform element. One of three terminal spines strongly recurved (Fig. 15). Maxillary hook (Fig. 16) elongate and slender, 118 f^m along shorter side, 156 ^m along longer side. Labrum (Fig. 17) with broadly rounded posteroventral margin bearing row of spines. Mandible (Fig. 18) terminally with stout unilaterally pectinate spine and adjacent small spiniform process. Paragnath (Fig. 17) a small smooth lobe. First maxilla (Fig. 19) with three setae. Second maxilla (Fig. 20) 2-segmented, second segment having terminally two stout spines with strong lateral spinules and one small naked seta. Maxilliped (Figs. 21, 22) 3-segmented, First segment elongate and unarmed. Second segment small and wedge-shaped. Third segment forming a recurved claw 125 ^m long with prominent subterminal spinules and bearing proximally an antero-outer seta and two very small postero-inner setules. Arrangement of mouthparts as in Figure 23. Ventral area between maxillipeds and first pair of legs without special sclerotization. Leg 1 (Fig. 24) with both rami 2-segmented. Legs 2-4 (Figs. 26, 29, 30) with 3-segmented rami. Formula for armature (Roman numerals representing spines, Arabic numerals indicating setae) as follows: P,: coxa 0-0, basis 1-1, exp. I-0;8, enp 0-1 ;7. P 2 : coxa 0-0, basis 1-0, exp I-0;I-1;II,I,3, enp 0-l;0-l;l,I,3. P 3 : coxa 0-0, basis 1-0, exp I-0;I-1,I,1,I,2,1, enp 0 -l; 0 -l,I, 2 . P 4 : coxa 0-0, basis 1-0, exp I-0;I-1 ;1, 1,3,1, enp 0-l;0-l;l,3. Legs 1 and 2 with intercoxal plates bilaterally armed with spines on free margin. Leg 1 (Fig. 24) with basis having strong spines along inner and inter-ramal margins; these inter-ramal spines absent in legs 2-4. Inner lobe of basis with barbed seta; this seta absent in legs 2-4. Second segment of exopod with eight setae as in Figure 25, but rarely with seven setae as in Figure 24. Leg 2 (Fig. 26) with outer spine on first segment of exopod bearing spines on posterior surface (Fig. 27). One female with both rami of leg 2 abnormal as shown in Figure 28. Leg 3 (Fig. 29) with outer side of third segment of exopod bearing I, 1,1. Leg 4 (Fig. 30) with hairs on outer margin of first and second segments of endopod. Leg 5 (Fig. 31) 2-segmented. First segment approximately 52 x 47 ;^m, with seta 45 f^m, and bearing ventrodistally a row of fine spinules. Second segment elongate, 114 x 44 f/m, bearing four distal setae 39, 39, 127, and 88 //m, all minutely barbed. Row of very small spinules along distal end of segment. Leg 6 probably represented by three setae on genital area (Fig. 4). Color unknown. HUMES: NEW TAENIACANTHID COPEPOD 173 Male: Body (Fig. 32) resembling in general aspeets that of female. Length (excluding setae on caudal rami) 1-50 mm (140-1-57 mm) and greatest width 0-58 mm (0-55-0-61 mm), based on seven specimens in lactic acid. Segment of leg 2 separated from segment of leg 1 by dorsal transverse suture. Ratio of length to width of prosome 1-58:1. Ratio of length of prosome to that of urosome 1-52:1. Segment of leg 5 (Fig. 33) 86 x 21 3 ^m. Genital segment 135 x 221 ^m, subrectangular and wider than long. Genital areas located laterally near posterior margin of segment (Fig. 42). Three postgenital segments from anterior to posterior 1 14 X 164, 99 X 151, and 107 X 122 ;/m. Caudal ramus (Fig. 33) similar to that of female, but smaller, 104 x49 One male with abnormal caudal rami (Fig. 34). Body surface smooth, with very few sensilla as in female. Rostral area, first antenna, second antenna, maxillary hook, labrum, mandible, paragnath, first maxilla, and second maxilla as in female. Maxilliped (Fig. 35) massive, 4-segmented. First segment short with one smooth seta. Large second segment swollen proximally, bearing on inner less swollen side two small setae and five prominent spines and other smaller spines as indicated. Small third segment unornamented. Fourth segment forming a stout, slightly recurved claw bearing four proximal setae, longest of them minutely barbed, and having on anterior surface a row of delicate spinules leading toward finely spinulose tip (Fig. 36). Ventral area between maxillipeds and first pair of legs as in female. Legs 1-4 segmented and armed as in female, with same spine and setal formula. Leg 1 (Fig. 37) with certain setae having several long proximal hairs, a few of these setae plumose rather than with short barbs. Leg 2 resembling female but terminal spine on exopod with blunt tip and naked throughout except for minute spinules near tip (Fig. 38), and endopod (Fig. 39) with outer hairs on second segment. Leg 3 similar to that of female, but terminal spine on exopod with blunt tip as in leg 2. Left endopod in one male with abnormal armature (Fig. 40), right endopod in this male normal. Leg 4 like that of female, but one male with abnormal endopods (Fig. 41). Leg 5 (Fig. 42) resembling that of female. First segment 39 x 34 ^m and second segment 91 x 36 /vm. No setae visible on genital area and leg 6 apparently absent. Extruded spermatophore not seen. Color unknown. ETYMOLOGY: The specific name salmacidis is the genitive form of the generic name of the host, from -aAfu\K(s',a nymph who in the fountain of Caria embraced a youth named Hermaphrodi- tus, and both grew together. Discussion Clavisodalis salmacidis differs from Clavisodalis heterocentroti in several features easily seen without dissection in animals cleared in lactic acid. These selected characters are shown in Table 1 . In both C. heterocentroti and C. salmacidis the armature of legs 1-4 is subject to variation. Most individuals, however, conform to the formula as indicated for the two species respectively. Accurate identification obviously requires obser- vation of more than one copepod. TABLE 1: Characters useful for the Differentiation of the Two Species of Clavisodalis. Character C. heterocentroti C. salmacidis Female genital segment much wider than long about as long as wide caudal ramus fused with anal segment separated from anal segment claw of maxilliped clavate, not recurved, with cluster of spines on blunt tip unguiform, recurved, with two rows of spines on atten- uated tip number of setae on second segment of P| Exp and Enp 7,6 8,7 armature of segments 1 and 2 of ?3 and P4 Enp 0 0; 0 0 0-1; 0-1 armature of inner side of second seg- ment of P4 Exp none one seta Male claw of maxilliped terminating in fine filament tip with spinulose area 174 MEMOIRS OF THE QUEENSLAND MUSEUM Figs. 1-9: Clavisodalis salmacidis sp. nov., female. 1, dorsal (A); 2, lateral (A); 3, urosome, dorsal (B); 4, genital area, lateral (C); 5, caudal ramus, dorsal (D); 6, caudal ramus, ventral (D); 7, abnormal caudal ramus, ventral (D); 8, egg sac, ventral (B); 9, egg sac, ventral (B). HUMES: NEW TAENIACANTHID COPEPOD 175 Figs. 10--17: Clavisodalis salmacidis sp. nov,, female. 10, egg, sac, ventral (B); 11, rostrum, ventral (C); 12, first antenna, ventral (D); 13, second antenna, outer (C); 14, second antenna, inner (C); 15, terminal spine on second antenna, posterior (E); 16, maxillary hook, ventral (D); 17, labrum and paragnaths, ventral (C). 0.05 MM 176 MEMOIRS OF THE QUEENSLAND MUSEUM Figs. 18-25: Clavisodalis salmacidis sp. nov,, female. 18, mandible, ventral (C); 19, first maxilla, postero-inner (C); 20, second maxilla, postero-inner (C); 21, maxilliped, antero-outer (C); 22, maxilliped, postero-inner (C); 23, cephalosome, ventral (B); 24, leg 1 and intercoxal plate, anterior (D); 25, exopod of leg 1, anterior (D). HUMES: NEW TAENIACANTHID COPEPOD 177 Figs 26-31: Clavisodalis salmacidis sp. nov., female. 26, leg 2 and intercoxal plate, anterior (F); 27, outer spine on first segment of exopod of leg 2, posterior (E); 28, abnormal rami of leg 2, anterior (F); 29, leg 3, anterior (F); 30, leg 4, anterior (F); 31, leg 5, dorsal (C). 178 MEMOIRS OF THE QUEENSLAND MUSEUM Figs. 32-39: Clavisodaiis salmacidis sp. nov., male. 32, dorsal (A); 33, urosome, ventral (B); 34, abnormal anal segment and caudal rami, ventral (D); 35, maxilliped, postero-inner (C); 36, claw of maxilliped, anterior (C); 37, leg 1 , anterior (D); 38, terminal spine on exopod of leg 2, anterior (C); 39, endopod of leg 2, anterior (D). HUMES: NEW TAENIACANTHID COPEPOD 179 The site of taeniacanthid copepods in sea urchins appears to be the esophagus. This is confirmed by the recovery of C. salmacidis from the esophagus and by information supplied by Gooding (personal communication to the author) that such copepods are commonly to be found at this site in tropical Pacific Echinoidea. When taeniacanthids were first discovered in 1955 in Diadema in Madagascar, they were obtained by washing the entire urchins in sea water with 5 per cent ethyl alcohol. This undoubtedly stimulated some of the copepods to crawl from the esophagus out of the mouth. They were subsequently recovered from the sediment resulting from the washing. Many more copepods could perhaps have been found if the esophagus of each urchin had been opened. Unfortunately, at that time it was not known that the real habitat of these copepods is the esophagus. The range of the host urchin, Salmacis belli, extends throughout the East Indies and North Australia (Clark and Rowe 1971). At present Clavisodalis salmacidis is known only from southeastern Queensland. Whether the copepod follows the range of the host is not known. ACKNOWLEDGMENTS The study of the copepods was aided by a grant, DEB 77 11879, from the National Science Foundation of the United States. I am greatly indebted to Dr L.R.G. Cannon, Queensland Museum, for sending me the Aristotle’s lanterns and allowing me to remove the copepods for study. LITERATURE CITED Clark. A.M. and Rowe, F.W.E., 1971. Monograph of shallow-water Indo-West Pacific echinoderms. Brit. Mus. (Nat. Hist.), publ 690, pp. 1-238. Gooding. R.U. 1965. Taeniacanthid copepods associat- ed with the black sea-urchin Diadema setosum in Madagascar and Singapore. Med. J. Malaya 20: 176. Humes, A.G, 1970. Clavisodalis heterocentroti gen. et sp. n., a cyclopoid copepod parasitic on an echinoid at Eniwetok Atoll. J. Parasitol. 56: 575-83. Humes, A.G. and Cressey, R.F., 1961. Copepodes taeniacanthides parasites d’un oursin a Madagas- car. Mem. Inst. Scient. Madagascar 1959 (F) 3: 1 - 12 . Humes, A.G. and Gooding, R.U., 1964. A method for studying the external anatomy of copepods. Crust aceana 6: 238-40. Figs. 40—42: Clavisodalis salmacidis sp. nov., male. 40, abnormal endopod of leg 3, anterior (D); 41, abnormal endopods of leg 4, anterior (D); 42, segment of leg 5, genital segment, and first postgenital segment, lateral (D). Mem. QdMus. 20(1) 181-235, pis. 1-16 [1980] A REVISION OF THE CORAL GENUS LEPTOSERIS (SCLERACTINIA: FUNGIINA: AGARICIIDAE) Zena D. Dinesen James Cook University of North Queensland ABSTRACT Seventeen of the 28 nominal recent species of Leptoseris are placed in synonymy. The remaining seven Indo-Pacific species and two species from the Caribbean region are redescribed, and two new species found in Great Barrier Reef waters, Leptoseris foliosa and Leptoseris glabra, are described. History, ecology, and zoogeography of the genus are discussed. The scleractinian genus Leptoseris has seldom been treated in any depth in taxonomic works. Although widely distributed geographically, it is a comparatively uncommon coral; it never occurs on reef flats, usually prefering deep water or shaded conditions. Hence it is poorly represented in museum collections and most previous workers have had at their disposal only small suites of specimens. Consequently, they have not appreciat- ed the range of variation exhibited by some species. The aim of this study is to provide a revision covering all recent species of Leptoseris. To this end many hundreds of specimens were collected, by SCUBA diving, from Great Barrier Reef waters, principally from the central region of the Great Barrier Reef, from the Lizard Island region, and the Palm Islands. Information about the living colonies, and their habitat, was recorded. I have also examined numerous specimens from a variety of Indo-Pacific and West Indian localities, including nearly all available type specimens, and other museum specimens referred to in the literature. The large series available for most species demonstrated considerable variability in some species, and seventeen of the 28 available names are shown to be synonyms. I recognise 1 1 recent species, seven of which occur in the Great Barrier Reef region. Taxonomic History Initially species were assigned to existing genera; subsequently several genera were created to accommodate the various species. This confusion piersisted because of the variety of growth forms found within the genus. Leptoseris papyracea (Dana), for example, was originally placed in several different genera. Indo-Pacific Species: The first described species of Leptoseris is apparently Agaricia crispa Ehrenberg, 1834. Dana (1846) moved Ehren- berg’s species to Pavonia, and added P. papyracea, although the descriptions suggest that these two species are probably synonymous. The genus Leptoseris was established in 1849 by Milne-Edwards and Haime who did not attribute to the genus the small branching species now recognised as Leptoseris; their 1851 description (p. 133) is as follows; ‘Polypier compose et adhelrent. Plateau commun, nu, delicatement strie, s’dlevant de fa^on ^ constituer un disque subcrateriforme, dont le centre est occupe'' par un individu parent et entoure par de plus petits, Calices trfes imparfaitement circonscrits, mais bien radies; rayons cloisonnaires tres longs. Columelle tuberculeuse. La Leptoseris est tres voisine des Mycedies, mais elle nous paratt devoir en ^re distinguee a cause de la diffluence des calices qui sont toujours fort rares, et de la grande etendue des rayons septo-costaux.’. Milne-Edwards and Haime (1849, 1851) also described the genus Haloseris in which they placed Ehrenberg’s crispa, but not Dana’s species papyracea. Their rather brief description of Leptoseris led Rousseau (1854) to include in it Leptoseris edwardsi, which is in fact a Lithophyllon (see Wells 1966). Duncan (1884) listed Leptoseris; however, he relegated Haloseris to a subgenus of Lophoseris 182 MEMOIRS OF THE QUEENSLAND MUSEUM {=Pavona). Quelch (1886) described two new genera — Cylloseris with its numerous hydnophoroid projections (it is not surprising that he did not recognise in it the Leptoseris of Milne-Edwards and Haime) and Domoseris which he admitted (p. 125) ‘somewhat resembles Leptoseris at the extreme marginal parts’. Bassett-Smith (1890) considered Dana’s papyracea referrable to Pavonia rather than to Leptoseris and described two further species, P. pretiosa and P. ramosa, both synonyms of L. papyracea. However Rehberg (1892) established the genus Folioseris in which he placed both papyracea and the crispa of Ehrenberg (but not H. crispa of Milne-Edwards and Haime because of its lack of columella). But his own specimens of F. papyracea are not the same as Dana’s species and are synonymous with L. gardineri van der Horst! Gardiner (1905) doubted whether Domoseris was distinct from Leptoseris, but clearly considered Cylloseris {C. incrustans) separate, for he described L. incrustans as a new species. Vaughan (1907) similarly questioned the validity of Domoseris, and placed Folioseris in synonymy with Leptoseris. Like Vaughan, van der Horst (1921) felt that Folioseris and Dana’s papyracea belonged in Leptoseris, and he recognised that Quelch’s Cylloseris is also a Leptoseris. Although Vaughan (1918) and Faustina (1927) placed both papyracea and its synonym L. digitata Vaughan in Leptoseris, they both retained Haloseris as a separate genus. Finally, Yabe, Sugiyama, and Eguchi (1936) referred several specimens of a Leptoseris species typified by the presence of collines and (often) lacking ‘normal’ Leptoseris corallites to Agaricia (?) minikoiensis, mistakenly believing them to be the same as Gardiner’s (1905) Agaricia ponderosa var. minikoiensis. Ma (1937) instead placed these specimens in a new subgenus, Agaricia {Agariciella)., but gave no reasons in the text. Wells ( 1 954) described this species as Leptoseris? mycetoseroides and included in synonymy the Japanese specimens. Atlantic Species: Madrepora cucullata was described by Ellis and Solander in 1786. Milne-Edwards and Haime (1849) made this the type of a new genus, Helioseris, but later (1851) submerged this in Mycedium. Wells (1956) considered Helioseris a synonym of Leptoseris but subsequently (1973) re-established the genus because of its lack of a columella. However, since some specimens have a distinct columellar structure, there seems to be no good reason for retaining Helioseris as a separate genus, and the species cucullata is now included in Leptoseris. A second Caribbean species, Mycedium cailleti Duchassaing and Michelotti, 1864, was retained in Mycedium by Pourtales (1874), but because of the species’ obvious agariciid characters, sub- sequent authors (Vaughan 1901, Porter 1972, Wells 1973) referred the species to Agaricia. This species is here included in Leptoseris, as it seems much more closely allied to other species of that genus than to any Agaricia. Thus in accordance with Vaughan and Wells (1943), and Wells (1956) I include Haloseris, Helioseris, Cylloseris, Domoseris, and Folioseris in Leptoseris adding Ma’s Agariciella as did Wells (1956). Ecology and Zoogeography Several species of Leptoseris may be found in soft bottom areas (coarse sand to mud) living apparently free, or attached to a piece of substrate such as coral rubble. The genus also occurs unshaded on solid substrate, and in crevices, caves, and overhangs. It may grow in conditions of strong, current but very rarely in high energy situations with waves or surf. It can occur as deep as 470 m (Vaughan 1907) and plays a significant role in the deeper hermatypic communities. Wells (1954), in his description of the seaward slope of Bikini Atoll, defines the Leptoseris zone as 91-146 m, below the Echinophyllia zone but above the ahermatypic Sclerhelia-Dendrophyllia zone. At Bikini these zones may be more clearly defined than usual and also deeper than usual because of extreme water clarity. But there is no doubt that Leptoseris is important in coral communities occurring below the normal hermatypic range, and subsequent records confirm this view. (See, for example, Barnes et al. 1971, Pichon 1973, Dinesen 1977.) Leptoseris is a coral of fairly restricted habitat, and its contribution to reef structure is limited. However, in soft bottom areas with large stands of L. gardineri, and an abundance of L. glabra and L. scabra, it may be a dominant coral in a strictly local sense. Furthermore, in some Outer Barrier situations (especially in channels and some back reef patches) which are favourable for growth of Leptoseris, large colonies of several species are not uncommon. Their contribution to the total coral cover is of some significance. Five Australian species {L. hawaiiensis, L. glabra, L. mycetoseroides, L. papyracea, and L. scabra) have been tested for the presence of DfNESEN: REVISION OF CORAL GENUS LEPTOSERIS 183 zooxanthellae, and all tests have proved positive. It is likely that all Leptoseris species are hermatypic, though quite possibly some deep-living specimens may lack zooxanthellae. The preference of many Leptoseris species for deep water or shady conditions may be due to an inability to tolerate too much light, or to competitive exclusion, or both. Specimens of Leptoseris have been found in shallow water in gullies, exposed to strong sunlight but only for limited periods. This observation suggests that if Leptoseris is sensitive to over-illumination, it is not the light intensity itself but rather the time of exposure to strong light that has an adverse effect. The distributions of the various species are given following each description. The records indicate that Leptoseris is widely distributed in the Indo- Pacific but less well represented in the Caribbean region. Definite records for some species may be widely spread geographically, but it seems reasonable to assume that many of these species do occur throughout much of the Indo-Pacific, and could be found if sought in the right biotopes. From available data, two species (L. incrustans and L. solidaj are confined to the eastern or eastern and central Pacific, and L. foliosa to the western Pacific. Few collections have been made of L. cailleti, all from localities broadly scattered in the Caribbean area. There has been little mention of L. cucullata in the literature, but Wells (1973) reported that it is not uncommon throughout the West Indies in depths below about 8 m. Terminology The terminology followed is basically that given by Moore, Hill, and Wells (1956). The specific use of the following terms should be noted. ‘Septocostae’ is used as defined by Duncan (1884, p. 201); ‘The costae may be directly continuous with the septa at the calicular margin, and may pass from calice to calice. Where there is continuity, the costae are termed septocostae.’ Although some workers may refer to the continuation of septa between centres as ‘costae’, it seems traditional in the Leptoseris literature to call these ‘septocostae’. In a genus where a distinct mural structure is generally lacking, the point at which a septum becomes a septocosta may be somewhat arbitrary. Here ‘septum’ denotes that part of the structure which lies within what is judged to be the thecal rim. The term ‘costae’ is reserved for the striations on the non-calicinal surface. ‘Septal profile’ denotes the shape of the septal margin as septa drop into the fossa. The ‘upper margin profile’ of the septocostae is the shape of the septocostal margin in section, e.g. rounded or acute. ‘Surface ornamentation’ refers to the granulations visible under the lens on the surface of the septa and septocostae. This includes ‘lateral ridges’, which may be set in rows on the sides of septa and septocostae, parallel to the upper margin. The ‘fossa’ (‘calicular axial fossa’, Duncan 1884, p. 201) is the axial depression in a calice, usually partly filled by the columella, here considered to be demarcated by the inner ends of the septa. Some species typically have corallites which are inclined towards the corallum margin. In such cases the inner or proximal side of a corallite will be raised into a cushion-like structure, here termed ‘proximal cushion’. Methods Field Methods; Material was collected by SCUBA diving. Details of depth and general habitat were always recorded; colour, growth form of entire living colonies, and microhabitat (e.g. position in a cave) were usually recorded. Some specimens encountered in caves (where indentifi- able in the field) were not collected. Laboratory Methods; Specimens were cleaned in a solution of sodium hypochlorite. Material was examined with a Nikon SMZ-2 binocular microscope, and measurements (apart from corallum size) were taken using an eyepiece with a square grid graticule. Where only a small suite was available for a species, measurements were taken from all material. With larger series, a proportion of representative specimens (20-40 per cent) was measured and examined in detail. Remaining material was checked over and all major features noted. In general, at least four measurements were made of each character on a specimen, and rather more measurements were made of septal number, size of corallites and fossae, and thickness of septocostae. Emphasis has been placed on those characters most useful in distinguishing the species. Measurement of Characters: For most measurements, the range is given first, followed by the mean or usual figure in parenthesis — e.g. corallites 10-5 0 mm (2 0-3 0 mm) by 10-4 0 mm (2-0-2-5 mm). 184 MEMOIRS OF THE QUEENSLAND MUSEUM Corallum; In general, only the dimensions of the largest coralla examined are given. These are listed as length in the larger diameter followed by the maximum width at right angles to this. Height is sometimes indicated in the case of erect colonies. Corallites: In the absence of a well defined thecal rim (e.g. in L. foliosa), it is difficult to give precise measurements of corallite size. In such cases, measurements should be treated as a guide only. Since corallites are often elliptical, the length of the greater axis is normally given first, followed by that of the shorter axis. The distance between corallites or rows of corallites was measured between fossae. Only a radial measurement is given, i.e. the distance apart parallel to the corallum margin or branch tips. Septal counts: If corallites are poorly defined, the number of septa per centre may be a somewhat subjective estimate, especially if many new septocostae are inserted in the area surrounding a corallite. For example, in his description of the holotype of L.? mycetoseroides. Wells (1954) gives the total septa as 25-35. On the same specimen I was not able to count as many as 35 septa in a calice. Presumably the figure reached depends upon one’s estimate of the corallite border. Some Leptoseris species have many more septa per centre than others, so that septal number is still a very useful character. Septocostae: The thickness of septocostae was measured using the grid graticule where each grid square = 0-5 mm. Each square could be approximately divided by eye into quarters, i.e. about 0-125 mm. Measurements are given to two decimal places; where these are quoted as increments of <0-5 mm, they should be considered merely a good estimate. Septal thickness has not been measured because the septa always taper and become thinner towards the centre of the corallite. The maximum septal thickness is never greater than the maximum thickness of the septocostae, and is usually less. Since septocostae tend to be thinner at the corallum margin, measurements were made both at the margin and elsewhere on the corallum. The number of septocostae in a given distance (5 mm) was most conveniently measured by counting with the grid at right angles to rows of parallel septocostae. For very narrow forms (e.g. L, papyracea) septocostae were counted within 2-5 mm, but for consistency these figures have been corrected for a 5 mm distance. Dentation of septocostal margin: For large dentations, the size of these is indicated. Where dentations are minute, figures are restricted to the number of dentations in a given distance of septocostal margin. Fossae: Usually the diameter of the longer axis is followed by that of the shorter axis. Depth of fossae is merely estimated (e.g. deep, very shallow). Costae: For the number of costae in 5 mm, separate counts for the corallum margin were not made. Costal spines, where set in a row, were measured within 1 mm. Collines, proximal cushions, nodules: The range or maximum size of these structures is indicated, if appropriate. Identification of Leptoseris. Because Leptoseris often occurs in a relatively restricted range of biotopes, there may be less intraspecific variation than in corals which are very widely distributed on reefs (e.g. acroporids and faviids). However, some Leptoseris species may vary considerably, both in the general morphology of the corallum and in finer skeletal structure. Much of the variation within the species may be found within a single colony, especially as regards corallite size, septal number and columellar development. Many of the species must be differentiated by taking into account a combina- tion of characters, when material comes close to the species’ limits in some characters. Which characters will be most useful in identifying a specimen may depend both on the species in question and on the material itself For this reason, it is not possible to construct a simple dichotomous key to the species of Leptoseris. The inclusion of a more complex key would probably lead to unnecessary confusion. For identification, reference should first be given to the photographic plate, then to the appropriate species descriptions bearing in mind generic characters are not always repeated in individual descriptions. Certain species of some corals may be more easily distinguished in the field than in the laboratory. (See, for example, Wallace 1978 on Acropora.) Although typical specimens of all species of Leptoseris can be recognised under- water by the experienced observer, in my experience, problems with the identification of Leptoseris are best resolved by examining specimens under the microscope. The most important characters for distinguish- ing species are as follows: corallum shape, size and arrangement of corallites, number of septa per centre, and the structure of the septocostae. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS 185 SYNONYMY: No synonymies given by other authors have been accepted without investigation. With the exception of L. cucuUata, the type of which has not been examined, synonymy lists are restricted to material examined by the author or by M. Pichon (of James Cook University), and to specimens which have been clearly illustrated in the literature. For example, Ma (1959) includes a number of plates of Leptoseris specimens collected by earlier workers. Only where his figures provide unequivocal illustrations of a species, have they been included in the synonymy. The holotype of the type species of Leptoseris, L. fragilis Milne- Edwards and Haime, appears to have been lost (see note following generic synonymy). The holotype of Leptoseris tenuis van der Horst has also apparently been mislaid. This problem is considered in the discussion section for L.foliosa. Bathymetric Records: The minimum and maximum records are given for specimens collected in Great Barrier Reef waters, but for other material only the maximum record is quoted (except if the species does not occur on the Great Barrier Reef). Distributions: Information does not include all records in the literature, but only records for material listed in the synonymy and for specimens examined by the author. Records for unregistered specimens or material received after the species descriptions were written have been included where they increase the geographical range of a species. Material Examined: Because a large amount of material is involved, collection details have been summarised. For material from the Great Barrier Reef region, relatively precise locality details are included. All such specimens were collected between 1976 and 1978, mostly by the author, but collector and actual collection dates are not listed. For other material, only the registration number and general locality are given. Collecting stations from the Lizard Island group have been divided into four areas, as follows. Area 1: Lagoon (protected). Area 2: N. Watson’s B. to W. side of N. Pt (fairly protected). Area 3: E, side of N. Pt to N. of Crystal Beach (fairly exposed). Area 4: Crystal Beach to South I. (fairly exposed). All material listed is in registered museum or university collections. These institutions are listed below, with the abbreviations used in the text. AHF: Allan Hancock Foundation (University of Southern California) AM: Zodlogisch Museum, Amsterdam BM: British Museum (Natural History) CM: University Museum of Zoology, Cambridge DB: Discovery Bay, Jamaica EN: Collection from Enewetak (specimens currently at James Cook University (JCU) but to be moved to Bernice P. Bishop Museum, Hawaii) GF: Collection of G. Faure (Centre Universi- taire. Reunion) MP: Collection of M. Pichon from Indonesia — Philippines region and Great Barrier Reef (specimens at JCU) MPIO: Collection of M. Pichon from southwest Indian Ocean (specimens at JCU) PM: Museum Nationale d’Histoire Naturelle, Paris QM: Queensland Museum TWCMS: Tyne and Wear County Council Museum, Sunderland, England UT: University of Tohoku UTA: University of Tel Aviv USNM: United States National Museum WAM: Western Australian Museum ZD: Collection of Z. Dinesen (specimens at JCU) SYSTEMATICS Family AGARICIIDAE Gray, 1847 Genus Leptoseris Milne-Edwards and Haime, 1849 Leptoseris Milne-Edwards and Haime, 1849, p. 72. (Type species Leptoseris fragilis Milne-Edwards and Haime, 1849, by monotypy. Published simultaneously with Haloseris Milne-Edwards and Haime, 1849, p. 72, and Helioseris Milne-Edwards and Haime, 1849, p. 72, and selected as valid name for the genus by Vaughan and Wells, 1943, p. 125.) Haloseris Milne-Edwards and Haime, 1849, p. 72. (Type species Haloseris lactuca Milne-Edwards and Haime, 1849, by monotypy. Placed in synonymy with Haloseris crispa (Ehrenberg, 1834) by Milne-Edwards and Haime in their 1851 description.) Helioseris Milne-Edwards and Haime, 1849, p. 72. (Type species Madrepora cucullata Ellis and Solander, 1786, by monotypy.) Cylloseris Quelch, 1886, p. 124. (Type species Cylloseris incrustans Quelch, 1 886, by monotypy.) 186 MEMOIRS OF THE QUEENSLAND MUSEUM Domoseris Quelch, 1886, p. 125. (Type species Domoseris solida Quelch, 1886, so designated in this revision.) Folioseris Rehberg, 1892, p. 26. (Type species Agaricia crispa Ehrenberg, 1834, so designated in this revision.) {Agariciella) Ma, 1937, p. 149. (Type species Leptoseris mycetoseroides Wells, 1954 [misiden* tified by Ma as Agaricia {Agariciella) minikoiensis Gardiner 1905] by monotypy.) The holotype of Leptoseris fragilis Milne-Edwards and Haime, 1849, appears to have been lost. The specimen in the Paris Museum which is supposed to be the holotype is a Coscinaraea, and does not fit the description given by Milne-Edwards and Haime (in 1851). Chevalier (1961) redescribed L. fragilis from a specimen which he stated was the ‘genotype’. Since Chevalier’s description is consistent with that of Milne-Edwards and Haime, I assume that Chevalier then had before him the true type specimen. There being no evidence that the true holotype is not a Leptoseris (according to the present understanding of the genus) it seems preferable to retain the name Leptoseris, than to abandon it in favour of a little-known synonym. In their description of L. fragilis, Milne-Edwards and Haime describe small specimens with distinct central coraliites, which might be young L. hawaiiensis, L. scabra or L. glabra. Subsequent authors have referred juvenile examples of more than one species to L. fragilis. Thus although L. fragilis is given as the type species of the genus, it has not been possible to give this species further consideration in this revision. I have not been able to obtain the holotype of Agaricia crispa Ehrenberg, 1834, from the Berlin Museum. A photograph supplied by the museum is inscribed “Haloseris crispa Ehrenberg Philippinen? Kat. Nr. 820”. However, the photographed specimen, while clearly a Leptoseris, appears larger than that described by Ehrenberg, and rather resembles Leptoseris cailleti Duchassaing and Michelotti, a Caribbean species. I am therefore in doubt as to whether the photograph is of the type specimen. It seems very likely that Ehrenberg’s crispa is a Leptoseris (and judging from the literature, probably synonymous with L. papyracea (Dana, 1846)), Thus while reference is given to this species in the generic synonymy, it could not be further considered for species synonymies. Description CORALLUM: Colonial, colony formation by circumoral then marginal budding. Growth form encrusting, explanate, bowl-shaped, tubuliferous, fan-shaped or branching. Corallum unifacial, but the margin may be reflected over onto the non-calicinal surface. Size from a few centimetres to about 0-5 m; stands of L. gardineri may be several metres across. Colour; beige, brown, green, dark red, mauve, grey, either uniform, or patchy with more than one colour. Sometimes a white margin; occasionally centres and septocostae of contrasting colour. Polyps apparently very small, with few, short, pointed tentacles. CORALLITES: Round or elliptical, from 10 mm across, up to 1 2 0 mm in the longer axis of elliptical coraliites. Longer axis usually parallel to corallum margin or branch tips. Coraliites flush, slightly sunken, or better delineated, with raised rims and directed upwards or towards corallum margin. Scattered or in roughly concentric rows parallel to corallum margin, with 2->30 centres per row. Parent corallite sometimes distinguish- able; coraliites or rows adjacent or up to several centimetres apart. Proximal cushions, where present, up to 6 mm high, and raised nodular areas up to 8 mm above general corallum surface. Hydnophoroid projections, where present, up to 4 mm high and 7 mm long at base. Collines, if developed, as concentric ridges or intersecting so as to enclose 1-12 centres. Collines 1-5 mm thick at base, up to 8 mm high. Septa: Equal to strongly alternating, with first order thicker and more exsert. Rarely divisible into cycles. Profile rounded or acute, and variable according to orientation of coraliites. Septa taper to become thinner at centre of coraliites. Margin smooth or slightly uneven, but not dentate. Usually imperforate. Number per centre: 6-50. Septocostae: Equal to strongly alternating, with the first order thicker and more exsert. Straight, slightly flexuous, or very contorted. In coralla with branches which become wider towards the tip, septocostae may form a diverging, fan-like pattern as septocostae become more numerous higher up the branches. Upper margin profile rounded, acute, or flattened with a narrow ridge running down centre. Margin smooth; with long, wave-like undulations, 0-5—2-5 mm long; with minute dentations, 3-10 in 1 mm; or with large, often peg-like dentations, 6-18 in 5 mm. Septocostae perforate or imperforate. Thickness: <0- 13-10 mm. Number in 5 mm: 1 1-36. Surface Ornamentation: Surface of septa and septocostae nearly smooth, or covered with minute, irregularly shaped granulations, sparse to profuse. Granulations may coalesce to form minute, sharp, wavy, irregular ridges, visible only under the lens. Lateral ridges sometimes present; very small granulations may be set in a row along each ridge. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS 187 Fossa: Round or oval, measuring <0-25-6 0 mm by <0-25-2 0 mm. Shallow to a few millimetres deep, depending on shape and orientation of corallites. Columella: Weakly to well developed, sometimes absent; development variable within one corallum. Structure styliform; papillae in a row; one or a few twisted particles or plates, more or less fused; or a solid boss, often with sculptured, uneven surface. A few principal septa, to all septa, may reach columella. Non-calicinal Surface: Almost smooth, or marked with fine, equal or subequal costae. Costae straight, slightly flexuous, or, in branching coralla, forming a fan-like pattern as do the septocostae. Number in 5 mm: 12-34. Costae smooth or with minute spines, often visible only under the lens, irregularly spaced, scattered over and between costae or set in a row along each costa. Number in 1 mm: 3-12. Underside of colonies frequently covered with encrusting organisms, such as bryozoa. Distribution Indo-Pacific and Caribbean region. Bathymetric Range: From 3 m (GBR) to 470 m (at Hawaii, Vaughan 1907). Discussion Of the six recent agariciid genera, Leptoseris seems most elosely allied to Gardineroseris Scheer and Pillai, and espeeially to Pavona Lamarck. Indeed, the distinction between Leptoseris and some unifacial Pavona species is still somewhat uncertain. For instance, Veron and Pichon (1979) have moved Pavona yabei Pillai and Scheer to Leptoseris, though the present author considers this species a Pavona. Previous descriptions of these two genera become unsatisfactory in the case of a few species whose generic position is uncertain. To labour over unequivocal genus diagnoses is not very useful when the distinction between these genera may to some extent be artificial. Since the validity of the species themselves remains unquestioned, the best course seems to be to outline the differences at species level. Pavona yabei has close affinities with L. mycetoseroides. However, coralla of P. yabei are usually distinctive, being often much larger and forming vertically-growing vases or fluted tubes. The collines are more consistently well-developed in P. yabei, and the radial collines (i.e. those running from the centre to the corallum margin) are always much more prominent than in L. mycetoseroides. The septa and septocostae tend to be somewhat thinner and more alternating than in L. mycetoseroides. Pavona cf. explanulata (Lamarck) sensu Veron and Pichon (1979) may resemble several Leptoseris species, notably L. hawaiiensis, L. glabra, and L. mycetoseroides at marginal parts which lack collines. But on the same colony, P. cf. explanulata may have thickly encrusting or massive areas as well as laminar plates, and the corallites tend to be more numerous than in Leptoseris. Septa and septocostae are strongly alternating, which is not the case in L. hawaiiensis and L. mycetoseroides. According to Veron and Pichon (1979), P. cf. explanulata has first order septocostae thinner than those of the second order, and septocostae which frequently branch to three or more corallites. Both these features distinguish this species from any Leptoseris. L. mycetoseroides also shows some affinities with Gardineroseris planulata (Dana), with which it has previously been confused. Larger colonies of G. planulata may become very massive, and the species always lacks oval, outwardly inclined corallites which may occur near the margin in L. mycetoseroides. Collines are always well developed, and usually high and very acute. They frequently enclose single centres, which is much less often the case in L. mycetoseroides. Finally, the septa tend to be more numerous than in the Leptoseris species. Leptoseris cucullata (Ellis and Solander, 1786) (Plate 1, figs. 1-3) Madrepora cucullata Ellis and Solander, 1786, p. 157, pi. 42. Helioseris cucullata: Milne-Edwards and Haime, 1849, p. 72. Wells, 1973, p. 23, fig. 14. Leptoseris nobilis: Ma, 1959, p. 17, pi. 35. Material Examined Belize, Carrie Bow Cay, USNM 47319-22; Jamaica, DB 598, 2082, 2432; 13°11-3’N, 61°06-5’W, USNM P-878 (included with specimens of L. cailleti). Description Corallum and Corallites: Corallum a thin, irregular plate, centrally attached, or encrusting with free margin. Largest colony examined measures 17-5 cm by 13 cm, several millimetres 188 MEMOIRS OF THE QUEENSLAND MUSEUM thick, but <1 mm thick at the margin. Colour: green and brown (Colin 1978). Corallites round or more often elliptical, measuring 10-6 0 mm (2 0-4 0 mm) by 1-0-6 0 mm (2 0-3 0 mm). Usually strongly inclined towards corallum margin, but sometimes sunken in. Corallites arranged singly or in rows, with up to 12 centres per row. Usually fairly crowded, but corallites or rows may be up to 1 -5 cm apart. Parent sometimes distinguishable. Where corallites are in rows, proximal cushions may form a short low ridge, 1-5 mm (2-4 mm) above the general surface, and 1-3 mm (1-2 mm) thick at the base. Ridges usually slightly inclined towards corallum margin; one specimen (USNM 47322) has short ridges and rather acute proximal cushions; these are directed upwards and give the corallum a slightly hydnophoroid appearance. Occasional raised nodular areas on some specimens, up to 8 mm high, may have corallites on them. Septa: Usually strongly alternating, rarely subequal; three orders occasionally visible. Profile rounded or acute. Septal margin may be uneven but not dentate. Septa imperforate. Number per centre: 12-44 (21). SeptoCOSTAE: Usually strongly alternating, with first order thicker and much more exsert. Usually rather straight; rarely crenellated on raised areas. Upper margin profile acute except where septocostae well thickened. Margin non-dentate. Septocostae imperforate. Thickness: Margin (first order septocostae): <0T 3-0-25 mm (<0-25 mm). Margin: (second order septocostae): to 013 mm (<013 mm). Elsewhere (first order septocostae): 013-0-75mm (0-25 mm to nearly 0-5 mm). Elsewhere (second order septocostae): <013-0-5 mm (<0-25 mm). Number in 5 mm: Margin: 12-24 (17). Elsewhere: 11-20(15). Surface Ornamentation: Septa and sep- tocostae with small, fairly profuse granulations, though second order generally smoother. Rarely all septocostae scarcely granulated. Lateral ridges not developed. Fossa: Usually elliptical, measuring 0-5-2-5 mm (0-75-2 0 mm) by 0-5-1 -5 mm (. 443, pi. 154, figs. 1, 2. Veron and Pichon, 1979, p. 38, figs. 65, 66. Leptoseris digitata Vaughan, 1907, p. 140, pi. 42, figs. 1, 2. Matthai, 1924, p. 52, pi. 11, figs. 5, 7. Durham and Barnard, 1952, p. 31, pi. 2, figs. 10a, 10b. Nemenzo, 1976, p. 239, pi. 5, fig. 1. Leptoseris panamensis Durham and Barnard, 1952, p. 36, pi. 2, figs. 11a, 11b. Leptoseris zamboi Nemenzo, 1971, p. 158, pi. 5, figs. 4, 5. DINESEN; REVISION OF CORAL GENUS LEPTOSERIS 189 Material Examined Non-GBR Material: Amiranie Is., BM E.16; Macclesfield Bank, BM 1892.10.17.130 {Pavonia ramosa holotype); Tizard Bank, BM 16B {^Pavonia pretiosa holotype); Philippines, PM 119 (mentioned Milne-Edwards and Haime, 1851, as Haloseris crispa); Hawaii, USNM 20892 {Leptoseris digitata cotype); Panama, AHF 1 {Leptoseris panamensis holotype); Colombia, AHF 10-4 (mentioned Durham and Barnard, 1952, as Leptoseris digitata)-. Coco I., AHF 10-3 (mentioned Durham and Barnard, 1952, as Leptoseris digitata)-, Ecuador, AHF 11-2 {Leptoseris panamensis paratype). GBR Material: Townsville Region: Between 18®50’S-19°0rS and 146^58’E-147°07’E, BM 1979.4.1.1, USNM 54210, QM G 12248-9, ZD 20292, 20294-5. Description CORALLUM AND CORALLITES: Corallum small and delicate, of twisted branches which tend to widen towards the tip. Largest colony collected measures 5 cm by 4-5 cm, up to 4 mm thick at the base but <1 mm at the margin. Nemenzo (1971) described a larger colony. Colour: pale brown. Corallites round or elliptical, measuring approximately l -0-3-5mm (2O-3-0 mm) across. There is usually space for one corallite only within the branch width, but occasionally there are two abreast, Corallites superficial, slightly sunken or sometimes slightly inclined towards branch tips. Corallites usually a few millimetres apart along branches, but up to > 1 cm. Septa: Equal or subequal, sometimes a few markedly thicker and more exsert than the rest. Profile usually rounded. Margin sometimes uneven, but not dentate; septa imperforate. Number per centre: 8-22 (14). Septocostae: Equal or subequal, tending to diverge as branches widen and new septocostae are inserted. A few septocostae may be especially thickened and exsert. Upper margin profile rounded, acute, or slightly flattened. Margin may be ragged but rarely dentate. Dentations present on a few specimens (e.g. the types of P. ramosa, BM 1892.10.17.130, and I. digitata. USNM 20892), these minute, irregularly spaced, 5-10 in 1 mm. Septocostate usually imperforate. Thick- ness: Margin: <013-0-25mm (<0-25 mm). Elsewhere: <0-13-0-75 mm (0-25 mm). Number in 5 mm: (very variable according to pattern of insertion of septocostae) Margin: 16-28 (20). Elsewhere: 12-20(16). Surface Ornamentation: Septa and sep- tocostae rather well granulated; granulations may occur as sharp, wavy ridges. Lateral ridges sometimes present, usually not well developed. Fossa: Round or elliptical, measuring 0-25-2 0 mm (<10 mm) across. Rather shallow. Columella: Variably developed, sometimes absent, especially in young corallites. A small pillar; a few twisted particles or plates; or a solid boss. Few to the majority of septa reach the columella. Non-calicinal Surface: Marked with fine costae. Number in 5 mm: 16-30 (22). Costal spines in rows or scattered. Number in 1 mm: 4—12 (7). Costae sometimes reduced to rows of spines, and direction of costae may then be unclear. FIabitat and Variation Material has been collected from soft-bottom biotopes of coarse sand and mud, at stations over 25 km offshore, in deeper water. Unlike Z.. gardineri, it has not been found in the close vicinity of fringing reefs, and apparently occupies a different type of biotope. Although there is some variation in branch shape, septal number, presence of minute dentations on the septocostae, and columellar development, the series of specimens (from a wide geographical range) is very consistent. Distribution Indo-Pacific: Madagascar; Amirante Is.; Andaman Is.; South China Sea; Indonesia; Sulu Sea; Philippines; Great Barrier Reef; Marshall Is.; Hawaii; Panama; Colombia; Coco L; Ecuador. Bathymetric Range: 26-35 m (GBR); 108-176 m (at Bikini, Wells 1954). Affinities This species is easily distinguished from L. gardineri by its much smaller, thinner branches and superficial corallites. In growth form it resembles its West Indian counterpart, L. cailleti, but the latter has much better circumscribed corallites which are strongly inclined towards branch tips. 190 MEMOIRS OF THE QUEENSLAND MUSEUM Leptoseris cailleti (Duchassaing and Michelotti, 1864) (Plate 2, figs. 3-4) Mycedium cailleti Duchassaing and Michelotti, 1864, p. 93 (reprint). Pourtales, 1874, p. 44, pi. 9, figs. 1, 2 . Agaricia cailleti: Vaughan, 1901, p. 31 1, pi. 20. Material Examined 20W4’N, 7i°40-7’W, USNM P-1148; 12°15-4’N, 69°57-5’W, USNM P-760; Off Mayagiiez, Puerto Rico, USNM 54215, 54216; Off St Thomas, USNM 6079; I5°33-7’N, 6P14-7’W, USNM P-932; I3°11-3’N, 6U06-5’W, USNM P-878. Description CORALLUM AND CORALLITES: Corallum of delicate, twisted branches which widen from the base; branches may fuse, or occasionally curl to form tubes. Branches measure <5 mm thick at base, often <1 mm thick at margin; and 0-5-3 0 cm (<10-2 0 cm) across. Sometimes dead at the base. Largest colony measures 8 cm high and 12 cm wide. Colour: light brown. Corallites round or more often elliptical, measuring 10-4 0 mm (1 -5-2-5 mm) by 10-4-0 mm (1-0-<2 0 mm). Sometimes flush or slightly sunken, but usually strongly inclined towards branch tips. Usually scattered, occasion- ally in rows with up to six centres per row. Often close; rarely up to 8 mm apart along branches. Proximal cushions low, 0-5-1 0 mm above the general surface. Septa: Equal, sometimes subequal, profile normally rounded. Septa non-dentate, imperfor- ate. Number per centre: 6-26 (16). SeptocostaE: Equal, occasionally subequal with first order slightly more exsert. Straight or slightly flexuous, with upper margin profile usually acute. Septocostae non-dentate, imperfor- ate. Thickness: Margin: <0-1 3-0-25 mm (<0-25 mm). Elsewhere: 0-13-0-38 mm (0-25 mm). Number in 5 mm: Margin: 18-30 (21). Elsewhere: 16-26 (19). Surface Ornamentation: Sides of septa and septocostae with small, fairly profuse granula- tions. Lateral ridges not developed. FOSSA: Usually elliptical, measuring <0-5-2 0 mm (l-0mm) along the septocostal margin. This gives septocostae an irregular appearance; however, margin is never divided into the distinct, peg-like teeth typical of L. scabra. Septocostae sometimes perforate. Thickness: Margin (both orders): <013-0-38 mm (0-25 mm or less). Elsewhere (first order septocostae): <0-13-1-0 mm (0-25-0-5 mm). Elsewhere (second order septocostae): <0-1 3-0-5 mm (<0-25-0-38 mm). Number in 5 mm: Margin; 15-23 (19). Elsewhere: 14-20(16). Surface Ornamentation: Septa and sep- tocostae, especially those of the first order, well granulated; granulations rough, rather profuse, sometimes coalescing to form sharp, wavy ridges. Lateral ridges sometimes developed, mostly on second order septocostae. Fossa: Round or elliptical, measuring <0-25-2-0 mm (0-5-10 mm) by <0-25-0-5 mm. Usually deep. Columella: Weakly to well developed. Usually of twisted particles, more or less fused; occasionally a boss. Few to many septa, but usually at least half, reach the columella. Non-calicinal Surface: Costae equal or subequal. Number in 5 mm; 14-23 (19). Smooth, or covered with small spines, scattered or in a row along each costa. Number in 1 mm: 6-10 (7). In his description of L. paschalensis, Wells (1972) reports juvenile, mono- to tricentric offsets on the non-calicinal surface. Habitat and Variation There is little information regarding the habitat of this species. The series available is small, but shows some variation in growth form, arran- gement of corallites and thickness of septa and septocostae. In particular. Wells’ holotype of L. paschalensis (USNM 53156) has thinner septocostae with a more even margin. Distribution Indo-Pacific: Marshall Is.; Tahiti; Easter I. Bathymetric Range: 40-100 m (at Easter I., Wells 1972) to 106-165 m (at Bikini, Wells 1954). Affinities This species has previously been confused with L. scabra. However, the septocostae in L. solida may have an undulating margin, but this is never divided into the distinct teeth found in L. scabra. Furthermore, L. solida has smaller corallites with fewer septa and smaller fossae, and corallites are usually more strongly inclined towards the corallum margin. L. solida has some affinity with L. glabra in that the septocostae are fairly thick, strongly alternating and basically non-dentate. However, the latter species is easily distinguished by its larger corallites with more septa, better developed columella, and much straighter septocostae. DINESEN; REVISION OF CORAL GENUS LEPTOSERIS 193 Leptoseris hawaiiensis Vaughan, 1907 (Plate 4, fig. 4; Plates 5-7) Leptoseris striatus Saville-Kent, 1871, p. 284. {Nomen oblitum) Leptoseris hawaiiensis Vaughan, 1907, p. 137, pi. 39, pi. 40. Matthai, 1948, p. 192, pi. 4, figs. 9, 10. Wells, 1954, p. 444, pi. 154, figs. 3, 4. Veron and Pichon, 1979, p. 52, figs. 92-8, 740. Leptoseris tubulifera Vaughan, 1907, p. 141, pi. 42, fig. 3, pi. 43. Leptoseris gravieri van der Horst, 1 922, p. 422. Leptoseris incrustans Gardiner, 1905, p. 948, pi. 92, fig. 25. Leptoseris incrustans (Quelch): Wells, 1954, p. 444, pi. 1 54, figs. 5-7. not Cylloseris incrustans Quelch, 1886, p. 124, pi. 6, figs. 4-4b. not Leptoseris tubulifera: van der Horst, 1921, p. 30, pi. 5, fig. 8. (= juvenile ? L. scabra, ? £. glabra). Material Examined Non-GBR Material: Amirante Is., BM 1937.1 1.17,123 {Leptoseris gravieri holotype); Reunion, GF REU.254, REU.812, REU.831-2, REU.899; Chagos Is., TWCMS C9055, C9058-9, C9061, C9063, C9066, C9074, C9080, Cl 2403, Cl 2408; Maldives, CM Registered but not numbered. {Leptoseris incrustans Gardiner holotype); Borneo, BM 1851.1.20-12 {Leptoseris striatus holotype); Paternoster Is., AM Coel 695 (incorrectly labelled type of Leptoseris tenuis)-, Philippines, MP 1811-75, 2656-75; Bikini Atoll, USNM 44796, 44798 (mentioned Wells, 1954, as Leptoseris incrustans); Eniewetak, EN 2869; Hawaii, USNM 20843-5, 20875-6 {Leptoseris hawaiiensis cotypes); USNM, 20846 {Leptoseris tubulifera cotype); Tahiti, ZD 20325. GBR Material: Lizard I. region: Jewell Reef — Outer Slope, QM G 12254, G 12269, G 12274, G 12282, ZD 9193, 9205, 9232, 9256, 20284. Carter Reef — Outer Slope, ZD 9840; Knoll, back reef, BM 1979. 4.2.3, ZD 9600; Channel to N. of Carter Reef; BM 1979.4.2.8, 1979.4.2.12, QM G 12284, ZD 9768-9, 20302. Yonge Reef — Outer slope, QM G 12255, G12261, G12268, MP 2636, 2638, ZD 9789, 9792, 9798, 10571-4, 10872, 20169; Back reef, BM 1979.4.2.5, USNM 54203, QM G12250, G12257, G12262-3, G12278, ZD 9570, 9573, 9779-81, 9783, 10579, 10656, 10680; Channel to N. of Yonge Reef, BM 1979.4.2.4, ZD 9761, 10920, 20303. Unnamed reef S. of Yonge Reef — Channel to N. of reef, BM 1979.4.2.2. S. Ribbon Reef — Back reef, BM 1979.4.2.10-11, USNM 54202, QM G 12252-3, G 12265, G 12267, G 12272, G 12275, G 12283, ZD 20260; Plug Reef, S. end of reef, QM G12271, G12273. Lizard I. — Area 2, ZD 20067, 20086; Area 3, BM 1979.4.2.9, USNM 54204, QM G 12251, G 12258, G 12258^0, G 12264, G 12266, G 1 2276, G 1 2279, G 1 228 1 , ZD 20043; Area 4, ZD 98 1 4 Palm Is. Group: Pelorus Is., W. side, BM 1979.4.2.1, 1979.4.2.7. Eclipse Is., QM G1 2256. Townsville region: Keeper Reef, SE. side, ZD 20246. Heron I. Region: Heron Reef, N. side, BM 1979.4.2.6, QM G 1 2270, G 1 2277, G 1 2280, ZD 1 0699. Description CORALLUM AND CORALLITES: Corallum corti- monly encrusting with free margin; sometimes plate or bowl-shaped, or entirely encrusting; rarely tubuliferous or foliose. Largest colony collected measures 39 cm by 34 cm, >1 cm thick but <1 mm at the margin. Some plate-shaped and encrusting coralla may be extremely thin. Colour: Dark to light brown or beige, rarely green or dark red. Sometimes patchy, e.g. brown with green, red, or beige patches. Occasionally a white margin. Corallites round or oval, measuring 10-7-0 mm (2 0-5 0 mm) by 10-5 0 mm (2 0-4 0 mm). Often flush or slightly sunken; sometimes with raised rims and directed upwards; or inclined towards corallum margin. Corallites sparse or crowded; scatered or in short rows, with up to 12 centres per row. Corallites or rows up to >3 cm apart. Parent corallite sometimes distinguishable. Corallum oecasionally with nodules, up to 5 mm high. Tubes, when present, up to 8 mm wide, usually less. Proximal cushions up to 5 mm high. Septa: Equal or subequal, profile rounded or acute. Margin non-dentate, though sometimes uneven. Septa rarely perforate. Number per centre: 8-48 (21). Septocostae: Equal or subequal; usually straight or slightly flexuous, but sometimes very ''contorted over raised areas, especially on encrusting coralla. Upper margin profile often very acute and pointed rather than tapering. Margin non-dentate, or with minute dentations, 3-10 (7) in 1 mm. On some specimens margin is slightly pitted and discontinuous; this condition is often associated with perforate septocostae. However, septocostae most often imperforate. Thickness: Margin: <0-13-0-38mm (<0-25 mm). Elsewhere: 013-0-38 mm (0-25-0-38 mm). Number in 5 mm; Margin: 20-30 (23). Elsewhere: 15-26 (19). Surface Ornamentation: Septa and sep- tocostae lightly and minutely granulated. Lateral ridges often well developed. Fossa: Usually elliptical, measuring 0-25-5 0 mm (0-5-10 mm) by <0-25-0-75 mm (0-5 mm). Shallow to rather deep. Columella: Weakly to fairly well developed. Generally rather a loose structure of twisted particles; sometimes a solid boss. The number of septa reaching the columella varies from few to all septa, though often shorter septa do not reach columella. 194 MEMOIRS OF THE QUEENSLAND MUSEUM Non-calicinal Surface: Costae equal or subequal. Number in 5 mm; 16-27 (22). Costal spines scattered or in rows on costae. Number in 1 mm: 4-9 (6). Habitat and Variation This species is most frequently found on Outer Barrier reefs, where it is not uncommon in overhangs and caves (especially on the ceilings and back walls). It has a definite preference for a shaded environment; and rarely occurs in very turbid water conditions. There is considerable variation in growth form and corallite size and arrangement. For example, some specimens (including Vaughan’s cotype USNM 20843) are very thin with small, sparse corallites directed upwards. Heavier coralla (thick plates and thick encrusting colonies) tend to have larger, more numerous corallites, slightly sunken, or directed towards corallum margin, and sometimes contorted septocostae. A few specimens (e.g. ZD 9570) exhibit a variety of corallite types within one corallum. These differences are not readily related to differences in micro-habitat. However, Australian specimens resembling Vaughan’s cotype (USNM 20843) have been collected in very poorly illuminated conditions. Thinly encrusting colonies with numerous, rather superficial corallites have been found most often on the ceilings of caves. Distribution Indo-Pacific: Red Sea; Amirante Is.; Reunion Chagos Archipelago; Maidive Is.; Thailand Borneo; Seringapatam Atoll (NW. Australia) Indonesia; Philippines; Great Barrier Reef Marshall Is.; Hawaii; Tahiti. Bathymetric Range: 3-46 m (GBR); 470 m (at Hawaii, Vaughan 1907). Affinities This species is mostly closely allied to L. mycetoseroides. In particular, specimens with close but poorly delineated corallites may resemble L. mycetoseroides in which collines are absent or very poorly developed. In such cases, L. hawaiiensis is best distinguished by its septocos- tae, which are often fewer and less granulated, with a more pointed upper margin profile, and better developed lateral ridges. Leptoseris scabra Vaughan, 1907 (Plates 8-9) Domoseris regularis Quelch, 1886, p. 127, pi. 5, figs. 6-6 b. (Nomen obi i turn). Leptoseris scabra Vaughan, 1907, p. 139, pi. 41, figs. 1, la, 2. Wells, 1954, p. 444, pi. 155, figs. 1, 2. Veron and Pichon, 1979, p. 48, figs. 83-91, 739. Leptoseris columna Yabe and Sugiyama, 1941, p. 75, pi. 63, figs. 2-2d. Leptoseris explanata Yabe and Sugiyama, 1941, p. 75, pi. 63, figs. 3-3e. Material Examined Non-GBR Material: Reunion, GF REU.835, REU.849, REU.853, REU.893, REU.897, REU.901; Mauritius, GF MAU-849; Chagos Is., TWCMS C9070-3, C9076-8, C9086-8, C9090-1, C9097, C9099-100; Palau, UT 64333 {Leptoseris columna holotype), 64334 {Leptoseris explanata holotype); Bikini Atoll, USNM 44801 (mentioned Wells, 1954); Hawaii, USNM 20886 {Leptoseris scabra cotype); Tahiti, BM 1886.12.9.295 {Domoseris regularis cotypc). GBR Material: Lizard 1. region; Jewell Reef — Outer slope, BM 1979.4.3.1, 1979.4.3.4, QM G 12290, G12330, ZD 9192-3, 20312-13. Carter Reef— Knoll, back reef, USNM 54212, QM G 12324, ZD 9613; Channel to N. of Carter Reef, BM 1979.4.3.10, USNM 54211, ZD 9765. Yonge Reef— Outer slope, MP 2637, 3492, 3759, ZD 9748, 9791, 9819; Back reef, ZD 10578; Channel to N. of Yonge Reef, BM 1979.4.3.11, QM G 12288, ZD 9744, 9746, 9750-1. Unnamed reef S. of Yonge Reef — Channel to N. of reef, QM G 12295, ZD 10348. S. Ribbon Reef — Back reef, QM G 12321. Lizard L— Area 1, QM G12319, G 12327, G 12331, ZD 9940, 10032-3; Area 2, BM 1979.4.3.14, USNM 54214, QM G 12289, G 12291, G 12294, G 12303, G12305, G 12307, G12320, G12326, G12332-3, ZD 9037, 9125, 9273, 9577, 9579, 9581-2, 9586, 9631, 9776, 9910, 10127, 10215, 10974; Area 3, BM 1979.4.3.15, QM G12297, G12299, G12302, G12323, G12325, MP 2737, ZD 9270, 9273, 9275, 9335, 9338, 9340, 9342, 9802, 10236, 10241, 10301, 10342-3, 10349, 10354-5, 10364, 10597, 20315, 20317; Area 4, BM 1979.4.3.8, QM G12292, G 12301, G12308, G 12322, G 12328, MP 3244, ZD 9371, 9382-4, 9833, 10175, 10178, 10622, 20128, 20314, 20318-21. Nymph I., QM G12298, G12334, ZD 9285, 9292. MacGillivray Reef, QM G 123 16. North Direction L, ZD 10552. Palm Is. Group: Pelorus I„ W. side, BM 1979.4.3.5, 1979.4.3.7, QM G 12287, G 12296, G 12304, G 12306, G 12309, G 123 13, ZD 9524. Orpheus I., Pioneer B„ BM 1979.4.3.13, QM G12285, G12314, ZD 9520, 9528, 9878, 9909. S. Calliope Channel; BM 1979.4.3.2; QM G 12293, ZD 9526. Curacoa L, SW. side; BM 1979.4.3.6, 1979.4.3.12, QM G12317, ZD 20322. Great Palm L, NW. side, BM 1979.4.3.3, 1979.4.3.9, QM G 12300, G1 23 11-12, ZD 9838. Eclipse I., USNM 54213, QM G12286, G12315, G12318, G12329. Brisk L, QMG12310, ZD 20323. DINESEN; REVISION OF CORAL GENUS LEPTOSERIS 195 Description CORALLUM AND CORALLITES; Corallum usual- ly encrusting with free margin; sometimes plate- or bowl-shaped, or erect with a foliose margin which may curl around to form low tubes. The largest colony collected measures 33 cm by 30 cm, and up to >1 cm thick, but less at the margin. Colour: Usually mid-brown, sometimes green, browny-green, or beige; rarely mauve or grey. Occasionally a combination of colours, e.g. pale nodules or darker septocostae. Sometimes a white margin. Corallites usually elliptical, measuring 2 0-10 0 mm (3 0-6 0 mm) by 2 0-7 0 mm (2 0-4 0 mm). Corallites sometimes flush or slightly sunken, but usually with raised rims and directed upwards or towards corallum margin. Sparse or crowded; scattered, or arranged in roughly concentric rows (especially in flatter colonies or near the margin). Up to 20 centres per row, but often <10. Corallites or rows up to >3 cm apart; parent corallite sometimes distinguishable. Proximal cushions up to 5 mm high. Nodules, not necessarily adjacent to corallites, up to 8 mm high. Septa: Generally strongly alternating, with first order much thicker and more exsert; sometimes one or two septa particularly thickened. Septa rarely equal. Profiile usually rounded. Margin may be uneven but not dentate; septa usually imperforate. Number per centre: 10-48 (24). Septocostae: Rarely equal, usually strongly alternating even at the margin. Sometimes fairly straight between centres, but often very labyrinthine over proximal cushions and nodules. Here second order septocostae may become as thickened as those of first order. Upper margin profile of first order septocostae rounded or subacute; second order septocostae usually with flatter margin with a narrow ridge down the centre. First order septocostae dentate, bearing irregular, often blunt, peg-like dentations. Height from crest to base is 01 3-0-75 mm (0-25-0-5 mm). Spacing 6-18 (10) in 5 mm. Second order septocostae sometimes with similar though smaller dentations, but frequently non-dentate. Dentations often best developed on proximal sides of corallites. Occasionally, if septocostae equal, all septocostae may have smaller, more regular dentations along entire length of margin between centres. Septocostae perforate or imperforate. Thick- ness: Margin (first order septocostae): 0-13-0-5 mm (0-25-0-38 mm). Margin (second order septocostae): <0-13-0-38 mm (0-25 mm or less). Elsewhere (first order septocostae): 0-25-0-75 mm (0-38-0-5 mm). Elsewhere (second order septocostae): 0-13-0-5 mm (0-38 mm). Number in 5 mm: Margin: 15-28 (21). Elsewhere: 11-24(17). Surface Ornamentation: Sides of septa and septocostae, especially those of first order, well granulated; granulations rough, sometimes coalescing to form sharp, wavy ridges. Lateral ridges conspicuous on thinner, second order septocostae. Fossa: Usually elliptical, measuring 0-5-4 0 mm (10-2-0 mm) by 0-25-1 -0 mm (0-25-0-75 mm). Often deep. Columella: Weakly to well developed. Structure a style; a few twisted particles, more or less fused; a solid boss; or papillae in a row in very elliptical fossae. Number of septa reaching columella varies from a few septa, to most septa if columella well developed. In general, most first order septa reach columella, most second order septa do not. Non-calicinal Surface: Costae equal or subequal. Number in 5 mm: 16-26 (21). Spines, where present, often scattered over costate. Number in 1 mm: 4—12 (8). Habitat and Variation L. scabra is one of the commonest species of Leptoseris. It occurs unshaded on soft bottoms or firm substrates, also in crevices, overhangs and caves (most frequently on the back or side walls). Colonies from areas with high sedimentation rates tend to be more erect, with curled or foliose margins. Larger, plate-shaped colonies have been collected mostly in clearer water, from exposed stations at Lizard I. and at the Outer Barrier. More encrusting colonies are common in a variety of biotopes. Most specimens are typically very scabrous. However, some specimens (representing a variety of growth forms) have areas of the corallum in which the septocostae are much smoother and less alternating; a few specimens are scarcely dentate throughout. This variation does not seem to be related to habitat differences. Specimens with equal and often more regularly dentate septocos- tae were collected at Outer Barrier stations. 196 MEMOIRS OF THE QUEENSLAND MUSEUM alongside specimens with more typically alternat- ing and dentate septocostae, and this difference may well be genetic. Distribution Indo-Pacific: Reunion; Mauritius; Chagos Archipelago; Houtman Abrolhos Is.; Palau; Great Barrier Reef; Solomon Is.; Marshall Is.; Hawaii; Tahiti. Bathymetric Range: 5-55 m (GBR); 143-406 m (at Hawaii, Vaughan 1907). Affinities: Unusual examples of this species with less alternating, less scabrous septocostae may come close to specimens of L. glabra with less alternating septocostae. L. scabra is also closely related to L. solida, as discussed previously. Leptoseris gardineri van der Horst, 1921 (Plate 10) Leptoseris gardineri van der Horst, 1921, p. 30. Hoffmeister, 1925, p. 41, pi. 4, fig. 2. Yabe and Sugiyama, 1941, p. 73, pi. 63, figs. 1-lc. Wells, 1954, p. 444. Veron and Pichon, 1979, p. 40, figs. 67-70. Folioseris papyracea: Rehberg, 1892, p. 26, pi. 2, fig. 8, pi. 4, fig. 2. Leptoseris papyracea: Gardiner, 1905, p. 947, pi. 92, fig. 23. Material Examined Non-GBR Material: Indonesia, Lombok, AM Siboga-Expeditie, Stat. 19. GBR Material: Palm Is. Group: Pelorus L, W. side, BM 1979.4.4.1., 1979.4.4.4, ZD 9538. Orpheus I., Pioneer B., BM 1979.4.4.2, QM G 12335-7, G 12339, ZD 9532, 9534, 9540, 9891-5, 9897-8, 9902^. Great Palm L, Butler B„ QM G 12338. Eclipse I., BM 1979.4.4.3, USNM 54201 . Esk I., ZD 9536. Description CORALLUM AND CORALLITES: Corallum erect and branching, branches often >20 cm high, 1-3 cm wide; branches may fuse. Branches fairly flat, or somewhat curled with calicinal surface innermost. Margin may be reflected onto non-calicinal surface for >1 cm, and corallites may develop here so that at the margin, corallum appears bifacial. Bases of branches, buried in sediment, frequently dead. Colonies may consist of a few branches, or stands several metres across. Colour: Brown or beige, sometimes with a white margin. Corallites round or elliptical, measuring 2 0-8 0 mm (4 0-6 0 mm) by 2 0-6 0 mm (3 0-4 0 mm). Often with raised rims, directed upwards or towards branch ends; sometimes flush. Scattered or up to three abreast on a branch; close or >3 cm apart along branches. Septa: Subequal to alternating, with first order septocostae thicker and more exsert. Profile usually rounded, sometimes acute. Septa non- dentate, imperforate. Number per centre: 10-36 (24). Septocostae: Alternating, with first order septocostae especially thicker; less alternating at margin. As branches widen, second order septocostae become first order ones. Upper margin profile of first order septocostae rounded; second order septocostae with rounded profile, or margin may be flat with a narrow ridge down the centre. Septocostae non-dentate, imperforate. Thickness: Margin: 0-1 3-0-5 mm (0-25 mm). Elsewhere: 0-1 3-0-75 mm (0-25-0-38 mm). Number in 5 mm; Margin: 14-24 (18). Elsewhere: 12-18 (15). Surface Ornamentation: Surface of septa and septocostae, especially those of first order, well granulated, sometimes with sharp, wavy ridges. Lateral ridges more conspicuous on thinner, second order septocostae. Fossa: Usually elliptical, measuring 0-5-2-5mm (1 -0-2-0 mm) by <0-5-2 0 mm (0-5-1 -0 mm). Shallow to fairly deep. Columella: Poorly to well developed. A few twisted particles, or a more solid boss. All septa may reach columella, but usually only those of the first order. Non-Calicinal Surface: Often distinctly costate, with costae tending to diverge and become more numerous towards branch tips. Number in 5 mm: 12-22 (18). Costal spines minute, scattered or in rows along costae. Number in 1 mm: 4—8 (6). Habitat and Variation This species occurs on soft bottoms, often in conditions of high sedimentation and turbid water, and specimens were collected in relatively shallow water in the vicinity of fringing reefs. There may be considerable variation in branch shape and size, even within one colony. However, most skeletal characters are rather consistent, despite some variation in coralHte size, and the extent to which septa and septocostae are alternating. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS 197 Distribution Indo- Pacific: Maldives; Indonesia; Palau; Great Barrier Reef; Marshall Is.; Samoa. Bathymetric Range: 10-18 m (GBR); to 55 m (at Bikini, Wells 1954). Affinities This species is most closely allied to L. glabra. However, coralla of L. glabra, if branching, have wider branches; the corallites are larger with more septa per centre, and septa and septocostae are generally more alternating than in L. gardineri. Leptoseris mycetoseroides Wells, 1954 (Plates 1 1-13) Leptoseris? mycetoseroides Wells, 1954, p. 445, pi. 153, figs. 4-6. Leptoseris incrustans: van der Horst, 1922, p. 422, pi. 32, figs. 3, 4. Ma, 1959 (pars), p. 17, pi. 30, pi. 32, figs. 3, 4. Agaricia (?) minikoiensis: Yabe, Sugiyama, and Eguchi, 1936, p. 55, pi. 42, figs. 5-7. Agaricia (Agariciella) minikoiensis: Ma, 1937, p. 149, pi. 45, figs. 2, 3. Leptoseris mycetoseroides: Veron and Pichon, 1979, p. 57, figs. 99-103, 741. not Agaricia ponderosa var. minikoiensis Gardiner, 1905, p. 937, pi, 90, fig. 7. Material Examined Non-GBR Material: Madagascar, MPIO 127-67, 142-67, 534-67, 539-67, 1131-69; Reunion, GF REU.96, REU.217, REU.320, REU.386, REU.518, REU.872; Mauritius, GF MAU.355. MPIO MAU-142, MAU-145; Chagos Is., TWCMS C9036-54, C9056, C9062, C9064-5, C9067, C9082, C9083-5, C9092^, C9096, Cl 2402, Cl 2405-7, Cl 24 10; Houtman Albrol- hos Is., WAM 389-77, 498-77; Philippines, MP 2504-p, 2567-75; Celebes, MP 1810-76, 1812-76; Kyusyu, UT 44907 (mentioned Yabe, Sugiyama and Eguchi, 1936, as Agaricia (?) ponderosa var. minikoiensis, and Ma, 1937, as Agaricia (Agariciella) minikoiensis); Solomon Is., WAM 119-77; Bikini Atoll, USNM 44805 (Leptoseris? mycetoseroides holotype); Enewetak, EN 1501, 1642, 2207, 2325, 2447, 2918, 2929. GBR Material: Lizard I. region: Jewell Reef — Outer slope, BM 1979.4.5.2, 1979.4.5.4, 1979.4.5.7, 1979.4.5.14, USNM 54208, QM G12346, G12353, G 12358, G 12360, G 12365-6, G 12370-2, G 12375, G 12380, G 12383^, ZD 9191, 9196, 9200, 9203^, 9206, 9209-11, 9224, 9230, 9236-8, 9242-5, 9248, 9250, 20304—5. Carter Reef — Outer slope, ZD 9770; Back reef, BM 1979.4.5.11, QM G 12363, ZD 9599, 9602, 9604, 9610-11, 9615; Channel to N. of Carter Reef, BM 1979.4.5.8, ZD 10541. Yonge Reef — Outer slope, BM 1979.4.5.12, QM G12343, G12388, G12390, MP 2647, 3317, 3769, ZD 9369, 9824, 20 1 70; Back reef, QM G12373, ZD 9784-6, 10583, 10676; Channel to N. of Yonge Reef, USNM 54205, QM G 12381, ZD 9755, 9758, 9762. Unnamed reef S. of Yonge Reef — Channel to N. of reef, QM G21356. S. Ribbon Reef — Back reef, BM 1979.4.5.5, QM G12350, ZD 9311, 9320, 9328, 20268, 20272; Plug Reef, S. end of S. Ribbon Reef, USNM 54206, QM G 12378, G12389, ZD 9298, 9301-2, 9304-5, 9307, 9315, 20256. Lizard I. — Area 1, USNM 54207, ZD 9122, 9623; Area 2, BM 1979.4.5.1, QM G 12362, G 12377, ZD 9591-3, 9992; Area 3, BM 1979.4.5.6, 1979.4.5.10, 1979.4.5.13, 1979.4.5.15, USNM 54209, QM G 12341-2, G 12349, G 12351, G12357, G12361, G12367, G12369, G12382, G12386, MP 3140. ZD: 9339, 9344, 9349-51, 93^3, 9806-8, 9810, 9932, 9934, 9936-9, 9982, 9993, 10031, 10078, 10083-4, 10104, 10170, 10227, 10284, 10310, 10346, 10357, 10394, 10402; Area 4, QM G 12352, ZD 9389-90, 9621, 9831, 10216-17, 20126, 20306. Nymph L, ZD 9283. Eyrie Reef, QM G12359. MacGillivray Reef, QM G 12385. ZD 9280. N. Direction L, QM G12376. Palm Is. Group: Pelorus L, W. Side, ZD 9505-7, 20307. Curacoa I., SW. side, ZD 20308. Great Palm L, BM 1979.4.5.9, QM G12374, ZD 9541, 20309. Brisk I., QM G 1 2387, ZD 203 1 0. Dido Rock, ZD 20311. Townsville region: John Brewer Reef, ZD 9554, 9558, 9560, 9562-3. Keeper Reef, ZD 20239. Heron I. region: Heron Reef — NW. side, BM 1979.4.5.3, QM G12345, ZD 10815; — S. side, QM G12344, G12348, G12354. Wistari Reef, N. side, QM G 12347, ZD 10772. Description CORALLUM AND CORALLITES: Corallum most commonly encrusting, sometimes with a free or laminar edge; rarely a plate. Largest specimen collected measures 65 cm by 50 cm, > 1 cm thick, but thinner at the margin. Colonies often only a few millimetres thick. Colour: Usually mid-brown, occasionally light or dark brown, or beige. Sometimes green (which may be very bright); or basically brown with patches of another colour, e.g. white, red, beige, or green. Rarely mauve or greyish-brown. Rarely with bright green centres, with or without bright green septocostae. Corallites usually elliptical, but sometimes round or slightly polygonal, measuring 10-7 0 mm (10-5 0 mm) by 10-4 0 mm (10-3 0 mm). Arrangement may depend on development of collines (see below). Corallites flush, somewhat sunken, or near corallum margin sometimes inclined towards margin. Numerous or sparse; where in rows, up to >30 centres per row, but usually <20. Corallites up to >2 cm apart; parent corallite rarely distinguishable. Collines are a distinctive feature of this species. They may be poorly developed as short, roughly 198 MEMOIRS OF THE QUEENSLAND MUSEUM concentric ridges, or more prominent, bordering long rows of corallites. Concentric collines (those parallel to the corallum margin) may be intersected by other collines, so as to enclose 1-12 (<6) centres, scattered or in short rows. Concentric collines better developed than intersecting ridges; collines often weaker or absent near corallum margin. Here corallites may be inclined towards margin, with proximal cushions as in some other Leptoseris species. Collines 1-5 mm (1-3 mm) thick at base, and up to 8 mm above general surface. Septa: Equal or subequal; with profile rounded or acute. Septa non-dentate, imperforate. Number per centre: 6-32 (17). SeptocostaE: Equal or subequal; straight, flexuous, or rarely contorted. Upper margin profile usually rounded, sometimes acute. Margin not dentate, but a few specimens have occasional peaks, 0-5-0-75 mm high, extending from the septocostal margin (see Plate 13, fig. 3). These occur only sporadically (not more than once per septocosta on each colline) and have only been found on specimens with large corallites and well developed collines. Septocostal margin occasionally slightly pitted and discontinuous, especially if septocostae are perforate. However, septocostae normally im- perforate. Thickness; Margin: <01 3-0-25 mm (<0-25 mm). Elsewhere; <0-1 3-0-5 mm (0-25 mm or slightly less). Number in 5 mm: Margin: 22-36 (28). Elsewhere: 14-32 (24). Surface Ornamentation: Sides of septa and septocostae usually with rather rough and fairly profuse granulations. Lateral ridges may be present, but generally more weakly developed than in L. hawaiiensis. Fossa; Usually elliptical, measuring <0-5-2 0 mm (1 cm thick, but <1 mm at the margin. Colour: Usually brown or beige, sometimes green, or brown with beige. Often a white margin; sometimes paler septa and septocostae. Corallites usually elliptical and large, measur- ing 2 0-12 0 mm (5 0-8 0 mm) by 2 0-6 0 mm (3 0-5 0 mm). Sometimes almost flush, but usually inclined towards corallum margin. Parent DINESEN: REVISION OF CORAL GENUS LEPTOSERIS 201 corallite sometimes distinguishable. Corallites numerous to sparse, scattered or in roughly concentric rows, with up to 25 centres per row. Corallites or rows adjacent to over 6 cm apart. Proximal cushions up to 6 mm high. SEPTA: Usually strongly alternating, with those of the first order thicker and more exsert. Profile usually rounded. Margin non-dentate; septa imperforate. Number per centre: 10-50 (30). Septocostae: As septa, usually strongly alternating, even at corallum margin. Straight or slightly flexuous, rarely contorted over raised areas of corallum. Upper margin profile of first order septocostae normally rounded; margin of second order septocostae flatter, often with a narrow ridge running down the centre. Margin non-dentate; septocostae nearly always imperfor- ate. A few specimens have a slightly pitted and discontinuous septocostal margin, and some perforations in the septocostae. Occasionally three orders of septocostae can be distinguished. Thickness: Margin (first order septocostae): 013-0-5mm (0-38 mm). Margin (second order septocostae): <013-0-38mm (0-25 mm). Else- where (first order septocostae): 01 3-0-75 mm (0-5 mm). Elsewhere (second order septocostae): <0-13-0-5 mm (0-38 mm). Number in 5 mm: Margin: 12-20 (16). Elsewhere: 11-18 (14). Surface Ornamentation: Surface of septa and septocostae, especially those of first order, well granulated; granulations often coalesce to form sharp, wavy ridges, these being more conspicuous than in other species. Lateral ridges prominent on second order septocostae. Fossa: Usually elliptical, measuring 0-75-6-0 mm (1-5-4 0 mm) by <0-5-1 -0 mm (0-5 mm). Shallow to over 3-0 mm deep in prominent corallites. Columella: Usually well developed, elliptical to fit shape of fossa. Often a solid boss with uneven, sculptured surface; sometimes of twisted particles, more or less fused. Few to most septa may reach columella; in general, most first order septa reach columella, most second order septa do not. Non-calicinal Surface: Costae, when present, usually equal. Number in 5 mm: 10-22 (15). Costal spines may be scattered or in a row along each costa. Number in 1 mm: 5-9 (7). Habitat and Variation Specimens have been collected principally from two very different biotopes. Those from protected areas, with high sedimentation rates and turbid water, may be found on muddy bottoms. They are usually erect, somewhat foHose, sometimes tubuliferous. This growth form may help to reduce the effects of sedimentation. Specimens from Outer Barrier localities, found on firm substrate, are usually plate-shaped or partly encrusting. They have been found unshaded, but more often in gullies and caves (on the ceiling or floor). Apart from the growth form, there is also some variation in corallite size and septal number. In general, the structure of the septa and septocostae, and the columella, are rather consistent. A few specimens have been found with rather more equal septa and septocostae, and this condition may sometimes be associated with poor illumination. Distribution Indo-Pacific: Red Sea; Reunion; Houtman Abrolhos Is.; Great Barrier Reef: Solomon Is.; Palau. Bathymetric Range: 5-44 m (GBR); 110 m (at Palau, Yabe and Sugiyama 1941). Affinities This species is most closely related to L. gardineri. However, coralla of L. gardineri have narrow branches, with generally smaller corallites and fewer septa. The septa and septocostae, though of similar structure, are less alternating in L. gardineri. As previously discussed, L. glabra also has some affinities with L. scabra and L. solida. ACKNOWLEDGMENTS I should like to thank the many institutions and individuals who have kindly lent material for examination. In particular, I appreciate the assistance of Dr P. Cornelius of the British Museum (Natural History), and Dr F. Bayer and Dr S. Cairns of the United States National Museum. Thanks are also due to Dr M. Pichon, Dr T. Done and Mr R. Birtles for various Great Barrier Reef specimens. Finally, I should like to thank Dr M. Pichon for reading the manuscript; Mr L. Brady and Mr W. Nash for advice on photography; and Miss J. Nichols for typing the manuscript. 202 MEMOIRS OF THE QUEENSLAND MUSEUM LITERATURE CITED Barnes. J, Bellamy, D.J., Jones. D J., Whitton, B.A., Drew, E.A., Kenyon, L., Lythgoe, J.N., and Rosen, B.R., 1971. Morphology and ecology of the reef front of Aldabra. Symp. zool. Soc. Land. 28: 87-114. Bassett-Smith, P.W., 1890. Report on the corals from the Tizard and Macclesfield Banks, China Sea. Ann. Mag. Nat. Hist. (6) 6: 353-74, 443-58, pis. 12-14. Chevalier, J.P., 1961. Recherches sur les madrepor- aires et les formations recifales miocenes de la Mediterran6e occidentale. DSc Thesis. 562 pp. (Societe g^ologique de France: Paris). Colin, P.L., 1978. Caribbean reef invertebrates and plants, pp. 512. (T.F.H. Publications: Neptune City, N.J.). Dana, J.D., 1846. Zoophytes. U.S. Exploring Exped. 1: 1-740, pis. 1-61. Dinesen. Z.D., 1977. The coral fauna of the Chagos Archipelago, pp. 155-61 In ‘Proceedings of the Third International Coral Reef Symposium, vol. 1.’ (University of Miami: Florida). Duchassaing, P., and Michelotti. J., 1864. Sup- plement au memoire sur les coralliaires des Antilles. Mem. Reale Accad. Sci. Torino 23(2): 97-206 (reprint paged 1-112), pis. 1-11. Duncan, P.M., 1884. A revision of the families and genera of the sclerodermic Zoantharia, Ed. & H., or Madreporaria (M. Rugosa excepted). J. Linn. Soc. 18 : 1-204. Durham, J.W., and Barnard, L.J., 1952. Stony corals of the Eastern Pacific collected by the Valero HI and the Valero IV. Allan Hancock Pacif Exped. 16 : 1-110, pis. 1-16. Ehrenberg, C.G., 1834. Beitrage zur physiologischen Kenntnis der Corallenthiere im Allgemeinen und besonders des Roten Meeres. Abh. K. Akad. Wiss. 1832 : 250-380. Ellis, J,, and Solander, D., 1786. ‘The natural history of many curious and uncommon zoophytes.’ 208 pp., 63 pis. (London: White and Son). Faustino, L.A., 1927. Recent Madreporaria of the Philippine Islands. Monogr. 22, Bur. Sci. Manila: 1-310, 100 pis. Gardiner, J.S., 1905. Madreporaria. III. Fungida. IV. Turbinolidae. pp. 933-57, pis. 89-93 In ‘Fauna and Geography of the Maldives and Laccadives Archipelagoes,’ vol. 2. (Cambridge: Cambridge University Press). Hoffmeister, J.E., 1925. Some corals from American Samoa and the Fiji Islands. Pap. Dep. mar. Biol. Carnegie Instn Wash. 22 : 1-90, pis. 1-23. Horst, C.J. van der, 1921. The Madreporaria of the Siboga Expedition. II. Madreporaria Fungida. Siboga Exped. XVIb: 53-98 pis. 1-6. 1922. Percy Sladen Trust Expedition to the Indian Ocean in 1905. IX. Madreporaria Agariciidae. Trans. Linn. Soc. Lond. (2) 18 : 417-29, pis. 30-2. Ma, T.Y.H., 1937. On the growth rate of reef corals and its relation to sea water temperature. Palaeont. sin. (B.) 16 ( 1 ): 1-226, pis. I-IOO. 1959. Effect of water temperature on growth rate of reef corals. Oceanographia sin. spec. vol. 1: 1-116, pis. 1-320. Matthai, G., 1924. Report on the madreporarian corals in the collection of the Indian Museum, Calcutta. Mem. Indian Mus. 8: 1-59 pis. 1-11. 1948. On the mode of growth of the skeleton in fungid corals. Phil. Trans. R. Soc. (B) 233(597): 177-96, pis. 3-14. Milne-Edwards, H., and Haime, J., 1849. Me^noire sur les polypiers appartenant a la famille des Oculinides, au groupe interm^iaire des Pseudas- treides et a la famille des Fongides. C. r. hebd. Seanc. Acad. Sci. Paris 29: 67-73. 1851. Recherches sur les polypiers. Mem. 6. Monographic des Fongides. Annls Sci. Nat. Zool. (3) 15: 73-144. Moore, R.C., Hill, D., and Wells, J.W., 1956. Glossary of morphological terms applied to corals, pp. F245-51 In MoORE, R.C. (ed.) ‘Treatise on Invertebrate Paleontology’, part F. Coelenterata. (Geol. Soc. Am.). Nemenzo, F., 1971. Systematic studies on Philippine shallow-water scleractinians. VII. Additional forms. Nat. appl. Sci. Bull. Univ. Philipp. 23(3): 142-85, pis. 1-12. 1976. Some new Philippine scleractinian reef corals. Nat. appl. Sci Bull. Univ. Philipp. 28: 229-76, pis. I-IO. PiCHON, M.M., 1973. Recherches sur les peuplements a dominance d’Anthozoaires dans les recifs de Tulear (Madagascar). Arch. Originales C.N.R.S. 9255: 1-440. Porter, J.W., 1972. Ecology and species diversity of coral reefs on opposite sides of the Isthmus of Panama. Bull. Biol. Soc. Wash. 2 : 89-116. PourtalEs, L.F. de, 1874. Zoological results of the Hassler expedition. Crinoids and corals. III. Cat. Mus. comp. Zool. Harv. 8: 27-50, pis. 5-9. Quelch, J.J., 1886. Report on the reef-corals. Rep. Sci. Results Voyage H.M.S. Challenger Zool. 16 : 1-203, pis. 1-12. Rehberg, H., 1892. Neue und wenig bekannte Korallen. Abh. Naturwiss. Ver. Hamburg \2'. 1-50, pis. 1-4. Rousseau, L., 1854. Zoophytes, pp. 119-24, pis. 27-9, In Dumont d’Urville, J.S.C. ‘Voyage au Pole sud et dans I’Oceanie sur les corvettes V Astrolabe et La Zelee.' (Paris: Gide). Saville-Kent, W., 1871. On some new and little-known species of Madrepores, or stony corals in the British Museum collection. Proc. zool. Soc. Lond. 1871 ( 2 ): 275-86, pis. 13-15. Vaughan, T.W., 1901. The stony corals of Porto Rican waters. Bull. U.S. Fish Commn 20 : 289-320, pis. 1-38. 1907. Recent Madreporaria of the Hawaiian Islands and Laysan. Bull. U.S. natn. Mus. 59(9): 1-427, pis. 1-96. DINESEN; REVISION OF CORAL GENUS LEPTOSERIS 203 1918. Some shoal-water corals from Murray Islands, Cocos-Keeling Islands and Fanning Island. Pap. Dep. mar. Biol. Carnegie Instn Wash. 9: 51-234, pis. 20-93. Vaughan, T.W., and Wells, J.W., 1943. Revision of the suborders, families and genera of the Scleractinia. Spec. Pap. geol. Soc. Am. 44 : 1-363, pis. 1-51. Veron, J.E.N., AND PiCHON, M.M. 1979. Scleractinia of Eastern Australia. III. Families Agariciidae, Siderastreidae, Fungiidae, Oculinidae, Merulinidae, Mussidae, Pectiniidae, Caryophylliidae, Den- drophylliidae. Aust. Inst. Mar. Sci. Monogr. Ser. 4 : 1-443, figs. 1-857. Verrill, A.E., 1864. List of the polyps and corals sent by the Museum of Comparative Zoology to other institutions in exchange, with annotations. Bull. Mas. comp. Zool. Harv. 3: 29-60. Wallace, C.C., 1978. The coral genus Acropora (Scleractinia: Astrocoeniina: Acroporidae) in the central and southern Great Barrier Reef Province. Mem. QdMus. 18(2): 273-319, pis. 43-103. Wells, J.W., 1954. Recent corals of the Marshall Islands. Prof. Pap. U.S. geol. Surv. 260-1: 385^86, pis. 94-185. 1956. Scleractinia. pp. F328^78 In Moore, R.C., (ed.) ‘Treatise on Invertebrate Paleontology’, part F. Coelenterata. (Geol. Soc. Am.). 1966. Evolutionary development in the scleractinian family Fungiidae. Symp. zool. Soc. bond. 16 : 223-46, pi. 1. 1972. Notes on Indo-Pacific scleractinian corals. VIII. Scleractinian corals from Easter Island. Pacif. Sci. 26(2): 183-90, figs. 1-3. 1973. New and old scleractinian corals from Jamaica. Bull. Mar. Sci. 23(1): 16-58, figs. 1-36. Yabe, H., and Sugiyama, T., 1941. Recent reef- building corals from Japan and the south sea islands under the Japanese mandate. II. Sci. Rep. Tohoku Univ. Geol. (2) spec. vol. 2: 67-91, pis. 60-104. Yabe, H., Sugiyama, T., and Eguchi, M., 1 936. Recent reef-building corals from Japan and the south sea islands under the Japanese mandate. 1. Sci. Rep. Tohoku Univ. Geol. (2) spec. vol. 1 : 1-66, pis. 1-59. MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 Fig. 1: Leptoseris cucullata (Ellis and Solander, 1786). x 0-9. Belize, Carrie Bow Cay, USNM 47321. Figs. 2, 3; L. cucullata. x 0-8; x 3-4. Jamaica, DB 2432. Fig. 4: Leptoseris papyracea (Dana, 1846). x 4-4. Macclesfield Bank, BM 1892.10.17,130 (piece of holotype of Pavonia rantosa). DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 1 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 2 Fig. 1: Leptoseris papyracea (Dana, 1846). x 3. Tizard Bank, BM 16B {Pavonia pretiosa holotype). Fig. 2: L. papyracea. x 1-5. 19°01’S, 146°58’E, (off Townsville), GBR, 25 m, ZD 20292. Figs. 3, 4: Leptoseris cailleti (Duchassaing and Michelotti, 1864). x 2-8; X 0-6. Off Mayagiiez, Puerto Rico, USNM 54215. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 2 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 3 Figs. I, 2: Leptoseris incrustans (Quelch, 1886). x 0-7; x 2-9. Tahiti, BM 1886.12.9.172 {Cylloseris incrustans cotype). — Fig. 3: Leptoseris solida (Quelch, 1886). x 0-4. Tahiti, BM 1 886. 1 2.9. 1 77 {Domoseris solida holotype). DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 3 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 4 Fig. 1: Leptoseris solida (Quelch, 1886). x 3. Tahiti, BM 1886.12.9.177. {Domoseris solida holotype). Fig. 2: L. solida. x 1. Tahiti, BM 1886.12.9.296 (Domoseris porosa holotype). Fig 3: L. solida. x 3. Easter I, USNM 53156 (piece of holotype of L. paschalensis). Fig. 4; Leptoseris hawaiiensis Vaughan, 1907. x 1-5. Hawaii, USNM 20875 {L. hawaiiensis cotype). DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 4 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 5 Figs. 1, 2; Leptoseris hawaiiensis Vaughan, 1907. x 0-6; x 3-4. USNM 20843 (L. hawaiiensis co\.y^Q). Fig. 3: L. hawaiiensis. x 1-6. Borneo, BM 1851.1.20-12 (L. striatus holotype). DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 5 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 6 Fig. 1: Leptoseris hawaiiensis Vaughan, 1907. x 1-3. Hawaii, USNM 20846 {L. tubulifera cotype). Fig. 2: L. hawaiiensis. x 3-3. Maldives, CM Registered but not numbered (L. incrustans Gardiner holotype). Fig. 3: L. hawaiiensis. x 3. Amirante Is, BM 1937.11.17.123 {L. gravieri holotype). Fig. 4: L. hawaiiensis. x 3. Channel to N. of Carter Reef, GBR, 5 m, ZD 9769. DINESEN: REVISION OF CORAL GENUS LEPTOSERJS Plate 6 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 7 Fig. 1: Leptoseris hawaiiensis Vaughan, 1907. x 3. Outer slope. Carter Reef, GBR, 8 m, ZD 9840. Fig 2: L. hawaiiensis. x 3. Back reef, Yonge Reef, GBR, 6 m, ZD 9570. Fig. 3: L. hawaiiensis. x 0-6. Channel to N. of Carter Reef, GBR, 5 m, ZD 9769. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 7 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 8 Fig. 1: Leptoseris scabra Vaughan, 1907. x 2-9. Hawaii. USNM 20886 (L. scabra cotype). Fig. 2; L. scabra. x 1. Tahiti, BM 1886.12.9.295 {Domoseris regularis cotype). Fig. 3: L. scabra. x 1-3. Lizard I., Area 2, GBR, 12 m, ZD 9273. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 8 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 9 Fig. 1: Leptoseris scabra Vaughan, 1907. x 3. Lizard I., Area 2, GBR, 12 m, ZD 10215. Figs. 2, 3: L. scabra. x 3; x 1. Back reef. Carter Reef, GBR, 10 m, ZD 9613. DINESEN: REVISION OF CORAL GENUS LEPTOSER/S Plate 9 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 10 Fig. 1; Leptoseris gardineri van der Horst, 1921. x 2-9. Orpheus I., Pioneer B., Palm Is. Group, GBR, 15 m, ZD 9893. Figs. 2, 3: L. gardineri. x 3; x 0-6. Pelorus I., W. side. Palm Is. Group, GBR, 18 m, ZD 9538. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 10 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 1 Figs. 1, 2: Leptoseris mycetoseroides Wells, 1954. x 0-5; x 3. Bikini Atoll, USNM 44805 {L.? mycetoseroides holotype). Fig. 3: L. mycetoseroides. xO-7. Kyusyu, UT 44907. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 11 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 12 Fig. 1; Leptoseris mycetoseroides Wells, 1954. x 3. Outer slope, Jewell Reef, GBR, 15 m, ZD 9224. Figs. 2, 3: L. mycetoseroides. x 3-3; x 0-5. Back reef. Carter Reef, GBR, 8 m, ZD 9615. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 12 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 3 Figs. ), 2: Leptoseris mycetoseroides Wells, 1954. x 0-6; x 3. Outer slope, Yonge Reef, GBR, 40-56 m, MP 3679. Fig. 3: L. mycetoseroides. x 3-8. John Brewer Reef, GBR, 9 m, ZD 9554. DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 13 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 4 Figs. 1, 2: Leptoseris foliosa n. sp. x 0-7; x 3 9. Lizard L, Area 1, GBR, 8 m, BM 1979.4.6.1, (holotype). Fig. 3: L. foliosa. x 0-8. Lizard L, Area 1, GBR, 5-10 m. BM 1979.4.6.3. (paratype). DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 14 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 1 5 Figs. 1, 2: Leptoseris glabra, n. sp. x 0-7; x 3. Pelorus L, W. side, Palm Is. Group, GBR, 17 m, BM 1979. 4.7.1 (holotype). Fig. 3: L. glabra, x 3-2. Orpheus 1., Pioneer B., Palm Is, Group, GBR, 14 m, ZD 9496 (paratype). DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 15 MEMOIRS OF THE QUEENSLAND MUSEUM Plate 16 Figs. 1, 2: Leptoseris glabra n, sp. x 0-6; x 1-7. Orpheus I., Pioneer B., Palm Is. Group, GBR, 14 m, ZD 9496 (paratype). Fig. 3: L. glabra, x 3. Channel to N. of Carter Reef, GBR, 5 m, BM 1979.4.7.5 (paratype). DINESEN: REVISION OF CORAL GENUS LEPTOSERIS Plate 16 CONTENTS Kott, Patricia Algal-bearing didemnid ascidians in the Indo-wcst-pacific ; i Rentz, D.C.F. A new family of ensiferous Orthoptera from the coastal sands of southeast Queensland 49 Molnar, R.E. Procoelous crocodile from Lower Cretaceous of Lightning Ridge, N.S.W 65 Molnar, R.E. An Ankylosaur (Ornithischia: Reptilia) from the Lower Cretaceous of southern Queensland 77 Van Tets,G.F. and Rich, P.V. A review of the De Vis fossil pigeons of Australia 89 CovACEViCH. J. and McDonald, K.R. Two new species of skinks from mid-eastern Queensland rainforest 95 Czechura.G.V. The emerald oionitor Varanus prasinus (Schlegel): an addition to the Australian mainland herpetofauna . 103 Ingram, G. A new frog of the genus Taudactylus (Myobatrachidae) from mid-eastern Queensland with notes on the other species of the genus 1 1 1 Miller, W.A. Radiographic pathol<^y in mandibles of Antechinomys laniger from Horseshoe Cave, Western Australia . 121 Davies, Valerie Todd Two large Australian orb-weaving spiders, Eriophora transmarina (Keyserling 1865) and Eriophora biapicata (L. Koch 1 87 1 1 25 Davies, Valerie Todd and Raven, R.J. Megadolomedes nov. gen. (Arancae: Pisauridae) with a description of the male of the type-species, Doiomedes australiensis Koch 1865 1 35 Saenger, P., Stephenson, W., and Moverley. J. The estuarine macrobenthos of the Calliope River and Auckland Creek, Queensland 1 43 Holdich, D.M. and Harrison, K. The isopod genus Dynamene from Australian waters, with description of a new species from coral reefs 1 63 Humes, A.G. A new taeniacanthid copepod from the oesophagus of a sea urchin in Queensland 1 7 1 Dinesen, Zen a A revision of the coral genus Leptoseris (Scleractinia: Fungiina: Agariciidae) 1 8 1