Records of the Museums and Art Galleries of the Northern Territory Volume 24 December 2008 The Beagle, Records of the Museums and Art Galleries of the Northern Territory (formerly ‘Records of the Northern Territory Museum of Arts and Sciences ) EDITORIAL COMMITTEE R. C. Willan Editor D. Carment C. J. Glasby Associate Editor H. K. Larson D. Megirian P. Murray P. Short The Beagle is a refereed journal published by the Museums and Art Galleries of the Northern Territory to disseminate the results of research in the following areas: Systematic and other studies of the terrestrial, marine and freshwater flora and fauna of the Northern Territory, tropical Australia, Southeast Asia and the Indo-Pacific; Australian Aboriginal, Southeast Asian and Oceanic art, material culture and archaeology; Northern Territory and Oceanic history and archaeology. The Beagle is published once or twice a year, depending upon the material available. All contributions to The Beagle are reviewed by two referees and, where possible, at least one is internationally-based. Whilst articles for The Beagle will normally be 10-50 pages in length, shorter communications, notes and review articles may also be acceptable. Longer articles, significant works or substantial revisions, which form integral studies, may be considered for separate publication as a Supplement. Prospective authors should contact the Editor, Academic Publications. The Beagle may be obtained by subscription or by exchange. The subscription rate for one year for individuals and institutions is S66.00 (includes postage and GST). All orders, subscriptions, back numbers and exchange enquiries should be addressed to the Library Manager, Museum and Art Gallery Northern Territory, GPO Box 4646, Darwin NT 0801, AUSTRALIA, or e-mail library. dam@nt.gov.au AUTHOR'S OFFPRINTS Twenty-five offprints are provided free for each published paper. Additional offprints may be ordered when returning proofs. SUBMISSION OF MANUSCRIPTS A Guide to Authors is provided on the inside back cover of each volume. Contributions should be posted or e-mailed to: The Editor, Academic Publications Museum and Art Gallery’ Northern Territory GPO Box 4646, Darwin NT0801, AUSTRALIA E-mail: richard. willan@nt.gov. an ISSN 1833-7511 ©Museums and Art Galleries of the Northern Territory, 2008. Printed by the Government Printing Office of the Northern Territory % DARWIN200 MUSEUM AND ART GALLERY NORTHERN TERRITORY v» Northern Territory Government Front cover: Pencil-blue butterflies of the Candalides absimilis species group from eastern Australia (see Braby, pages 33-54). Lower left: C. absimilis absimilis; all others: C. absimilis edwardsi. Photographs by Michael Braby. The Beagle Records of the Museums and Art Galleries of the Northern Territory Volume 24, December 2008 CONTENTS BERRA, T. M. - Charles Darwin’s paradigm shift.1 DURETTO, M. F. — A reassessment of Boronia (Rutaceae) in the Northern Territory with a key to species, the description of one new species and the reduction, in synonymy, of another species.7 PAXTON, H and SAFARIK, M. - Jaw growth and replacement in Diopatra aciculata (Annelida: Onuphidae).15 SALGADO KENT, C. P. and McGUINNESS, K. A. - Feeding selectivity of sesarmid crabs from northern Australian mangrove forests.23 BRABY, M. F. - Taxonomic review of Candalides absimilis (C. Felder, 1862) and C. margarita (Semper, 1879) (Lepidoptera: Lycaenidae), with descriptions of two new subspecies.33 BAEHR, M. - Two new species of the genus Pseudaptinus Castelnau from northern Australia (Insecta: Coleoptera: Carabidae: Zuphiinae). 55 KOTT, P. - Biogeographic implications of Ascidiacea (Tunicata) from the Wessel Islands (Arafura Sea).63 THORBURN, D. C. and ROWLAND, A. J. - Juvenile bull sharks Carcharhinus leucas (Valenciennes, 1839) in northern Australian rivers.79 M0LLER, P. R. and SCHWARZHANS, W. - Review of the Dinematichthyni (Teleostci: Bythitidae) of the Indo-west Pacific. Part IV. Dinematichthys and two new genera with descriptions of nine new species.87 LARSON, H. K. - A new species of the gudgeon Bostrychus (Teleostei: Gobioidei: Eleotridae), from peninsular Malaysia. 147 MOHD-AZLAN, J., TAHA, S. H., LAMAN, C. J. M. and ABDULLAH, M. T. - Diversity of bats at two contrasting elevations in a protected dipterocarp forest in Sarawak, Borneo.151 Short communication TENNENT, W. J. - A previously unpublished record of Liphyra brassolis Westwood, 1864 (Lepidoptera: Lycaenidae: Miletinae) from Vella Lavella, New Georgia group, Solomon Islands.157 Corrigendum HORNER, P. - Corrigendum to Horner, P. and Adams, M. (2007). A molecular systematic assessment of species boundaries in Australian Cryptoblepharus (Reptilia: Squamata: Scincidae) - a case study for the combined use of allozymes and morphology to explore cryptic biodiversity. The Beagle , Records of the Museums and Art Galleries of the Northern Territory, Supplement 3: 1-19, 159 . The Beagle, Records of the Museums and Art Galleries of the Northern Territory, 2008 24 : 1-5 Charles Darwin’s paradigm shift TIM M. BERRA Research Associate, Museum and Art Gallery Northern Territory and Department of Evolution, Ecology> and Organismal Biology, The Ohio State University, Mansfield, Ohio 44906, USA berra. l@osu.edu ABSTRACT The publication ofDarwin’s On the Origin oj Species in 1859 created a paradigm shift from creation to evolution. Darwin showed that humans are part ot nature, not above it. and that all animal life, including human, is related by descent from a common ancestor. His mechanism of evolution via natural selection is a powerful creative force that provided an explanation for the diversity of life. This dramatic change in world view from supematuralism to methodological naturalism has allowed staggering scientific advances in the past 150 years that transcend science and impact on the human psyche. Keywords: Charles Darwin, evolution, natural selection, naturalism, paradigm shift. INTRODUCTION Charles Darwin (1809-1882) was an extraordinary man by any standard. The theory of evolution by natural selection as elaborated in his book On the origin of species (1859) is considered by historians and philosophers of science to be one of the most important ideas ever had by the human mind (Dennett 1995). Before exploring this grandiose statement, a brief review ofDarwin’s life and scientific accomplish¬ ments is in order. Then I will address the implications of his very useful insight that extended beyond science and profoundly impacted on the human mind. An outline of Darwin’s Life Charles Darwin was bom into a wealthy English family on 12 February 1809. His father, Robert Waring Darwin (1766-1848), was a prominent physician as was his grand¬ father Erasmus Darwin (1731-1802). His mother was Susannah Wedgwood (1764-1817), the daughter of Josiah Wedgwood (1730-1795), the pottery manufacturer and entrepreneur, who was a close friend of Erasmus Darwin. Darwin’s father sent Charles to medical school at Edinburgh University in 1825 and removed him in 1827 when it became obvious that Charles was not interested in a medical career. Robert Darwin then decided that Charles should study to be a clergyman in the Church of England, and sent him to Cambridge University in 1828. Charles graduated 10"' in his class in 1831 and then received an invitation orchestrated by his professor, John Stevens Hens- low (1796-1861), to be an unpaid naturalist-companion to Captain Robert FitzRoy (1805-1865) on a surveying voyage around the world on I I.M.S. Beagle (1831-1836). Darwin later described this opportunity as “the first real training or education of my mind”. Upon return from the nearly five-year Beagle voyage, Darwin found that he was accepted as a serious scientist, and he had no desire to become a clergyman. He began working on the specimens collected on the voyage. He married his first cousin, Emma Wedgwood (1808-1896), and they eventually moved from London to Down House in Kent (Fig. 1). They had 10 children, seven of whom sur¬ vived to adulthood. After the voyage, he was often ill, but nevertheless, highly productive. He entered his ideas about how species form in a series of notebooks This included a Fig. 1. At rear of montage, wedding portrait of Charles Darwin (watercolour) in 1840 at age 30 by George Richmond. At centre, 1IMS Beagle drawing by American artist Samuel L. Margolies (1897-1974) from Dibncr (1960) in the Bumdy Collection at the Huntington Library, San Marino, CA. Used with permission of the Huntington Library. In foreground, one of the last photographs of Charles Darwin, by Elliot and Fiy. T. M. Berra branching, tree-like diagram that reflected the common ori¬ gin and relatedness of organisms. This first evolutionary tree showed that classification should be genealogical. However, he kept his revolutionary ideas private for 20 years except for his closest scientific colleagues: geologist Charles Lyell (1797-1875), botanist Joseph Dalton Hooker (1817-1911), zoologist Thomas Henry Huxley (1825-1895), and his American botanist correspondent at Harvard University, Asa Gray (1810-1888). In 1858 Darwin received of a let¬ ter from naturalist Alfred Russel Wallace (1823-1913), who, like Darwin, was inspired by the writings of Thomas Malthus (1766-1834). Wallace outlined ideas nearly iden¬ tical to Darwin’s. This letter and urging from Lyell and Hooker prompted him to complete and publish On the Origin of Species in 1859. Darwin continued to do experi¬ ments and publish on a variety of topics right up to the time of his death. He died of heart disease on 19 April 1882 and was laid to rest with pomp and ceremony in Westminster Abby a few feet from Isaac Newton. Further details can be pursued in three of the most comprehensive biographies of Darwin (Desmond and Moore 1991; Browne 1995, 2002), a concise biography (Berra 2009), and, of course, Darwin’s autobiography (Barlow 1958). SYNOPSIS OF DARWIN’S SCIENTIFIC ACHIEVEMENTS The educated citizen is generally aware of The origin and Darwin’s account of his voyage around the world in H.M.S. Beagle through his book now universally known as The voyage of the Beagle. These two books have never been out of print. Most people are surprised to learn that Darwin also made many other major contributions to geology, zoology, and botany through his observations, experiments and writings. His books have been chronicled (Berra 2009), so I will just briefly outline the breadth of his influence. Darwin explained how coral reefs form (1842) and contributed to geological observations on earth movements (1844) and deformation theory of metamorphic rock (1846). In a pio¬ neering four-volume work that took eight years to complete, he described all known barnacle species, fossil and living (1851-1854). Darwin explained how orchids are fertilized by insects (1862), how plants climb (1865), and catalogued the bewildering amount of variation in domestic plants and animals (1868). He explained human origins and sexual selection in ways never before articulated (1870-71), and discussed human and animal emotions in similar terms (1872). The latter work was one of the first books to use photographs to illustrate a point. Darwin showed how insectivorous plants on impov¬ erished soils utilise nitrogen-rich insects (1875), and demonstrated that the offspring of cross-fertilised plants were more numerous and vigorous than self-fertilised ones (1876, 1877). His observations of climbing plants laid the foundation for the field of plant growth hormones (1880), and his work on earthworms (1881) is a classic study in ecology. Any one of these achievements could constitute a life’s work for most scientists. DARWIN’S LEGACY Darwin was bom and educated at a time when special creation was the prevailing scientific view. That is, God cre¬ ated the universe and all species a few thousand years ago, and they were unchangeable. “Revelation”, not research, provided this view. Darwin began the Beagle voyage with this belief. During his lifetime the age of the earth was in¬ creasingly recognised as ancient as suggested by Georges Cuvier (1769-1832) and Charles Lyell (Bowler 1984; Larson 2004). Observations made during the voyage made him question the Genesis creation myth and immutability of species. He found marine fossils thousands of feet above sea level and reasoned that the land had been elevated by earth movements, not inundated in a great biblical flood. The fossil mammals he uncovered in South America resembled living mammals from the same area. He wondered why this should be if each species were specially created. Extinction was hardly recognised in those days. Why did the animals on islands off continental areas resemble those of the nearest land mass if each species were created in place? Why were there so many species in an island group that looked very similar but with slight differences from island to island? It is as if “one species had been taken and modified for different ends”, he wrote in Voyage of the Beagle. None of these things made sense from a creationist perspective. In 1844 he wrote to Hooker that “I am almost convinced (quite contrary to the opinion I started with) that species are not (it is like confessing a murder) immutable.” The elegant simplicity of Darwin’s reasoning can be distilled as follows. There is variation in nature, many more offspring are generated than can survive, therefore there is a struggle for life in which favorable variations arc preserved and unfavorable variations arc removed. This leads to evolu¬ tion which he defined as “descent with modification” and to the formation of new species. Nature is doing the selecting for the forms best adapted to a particular environment so he called the process natural selection as opposed to artificial selection that breeders impose. We now know that muta¬ tion, chromosomal rearrangements, sexual reproduction, etc. are the sources of genetic variation, but Darwin had no knowledge of such topics. Today we can speak of “descent with modification” as “a change in gene frequency”, and natural selection is simply “differential reproduction”, that is, one genetic variant leaves more offspring than another (Berra 1990). Darwin borrowed the expression “survival of the fittest” from economist Herbert Spencer (1820-1903). Evolutionary fitness means reproductive fitness. In modem terms, the fittest is the one who gets the most genes into 2 Charles Darwin’ paradigm shift Fig. 2. Cartoon of Charles Darwin as a monkey, from Fun, 16 November 1872, just after The expression of the emotions in man and animals was published. Many other similar personal attacks were published during his lifetime. the next generation, not necessarily the biggest or strongest individual. By the time of Darwin's death in 1882, most scientists of the world had accepted the concept of common descent, but some were still skeptical of natural selection as a creative mechanism (Bowler 1984). The public was less accepting (Fig. 2). The publication of On the origin of species on 24 No¬ vember 1859 precipitated one of those rare events in the history of science, a paradigm shift. Philosopher Thomas Kuhn used this term to refer to the replacement of one world view by another (Kuhn 1962). Examples ofa paradigm shift in science include the replacement of the earth-centered Ptolemaic system by the sun-centered Copemican system and the replacement of Newtonian physics by relativity and quantum physics. Darwin’s work neatly dove-tailed into the wider pattern of scientific advances that were occurring during his life. Lyell and others had provided the necessary' geological time for evolution to operate. The writings of Malthus, Spencer, Wallace, and many others help set the evolutionary stage. By 1859 evolution by natural selection was an idea whose time had come. Darwin and the publication of The origin closed the deal. Darwin changed the way humans view their place in nature. He showed that humans were not above nature, but pail of it. He supplied the explanation for the great diversity of life and showed that all life, including human, is related by descent from a common ancestor. His explanation of evolution via natural selection is the basis of all of biology and its applied subdisciplines of medicine, agriculture, and biotechnology. No other biologist in the history of our species has had an impact of this magni¬ tude. In the words of the eminent geneticist Theodosius Dobzhansky, “Nothing in biology makes sense except in the light of evolution” (Dobzhansky 1973). The paradigm shift from creation to evolution has moved intellectual endeavors from untestable belief to rational un¬ derstanding that flows from the scientific method. This, in turn, has allowed a vast array of advances in knowledge. DARWINIAN IMPLICATIONS One of the attributes of a powerful scientific theory is that it enables future research and understanding. Darwin¬ ian or evolutionary medicine as formulated by Nesse and Williams (1996) explains how some disease symptoms, such as fever, may be a response favored by natural selec¬ tion as a defense against pathogens. Some genetic diseases such as sickle cell anaemia may allow differential survival of its victims in malarial zones, a phenomenon called a balanced polymorphism (Berra 1990). Evolutionary think¬ ing explains the arms race waged by pathogens and hosts that prevents either from being completely eliminated. The development of antibiotic resistant bacteria through the flagrant overuse of antibiotics is easily explained by Darwinian reasoning. A drug kills the susceptible bacteria leaving bacteria with a pre-existing resistant mutation to build up the next generation. Then when you actually need the antibiotic for a bacterial infection, you find that the drug is ineffective. This is evolution, pure and simple. A similar process occurs in agriculture with the over application of pesticides and the formation of pesticide resistant pathogens, insects, and noxious plants. Austral¬ ians are very familiar with the myxomatosis versus rabbit “anus race” whereby the virus initially killed 99 percent of the rabbits, but given enough time the surviving rabbits returned in force as the virus evolved in the direction of less virulence and the rabbits were selected for more resistance to the virus (Berra 1998). Evolutionary psychology and evolutionary ethics, as explored by Barkow et al. (1992) and popularised by Wright (1994) help explain the origin of morality. Peacemaking among non-human primates by the calming effect of mutual grooming to diffuse aggression may be seen as the precur¬ sor of what became morality in humans (de Waal 1989). Modem religions are recent human inventions - a mere few thousand years old. The antecedents of morality, on the other hand, clearly evolved before humanity as reflected in the empathy exhibited by bonobos (Pan paniscus ) and the reciprocity of chimpanzees (P. troglodytes) (de Waal 2005). Kin selection, whereby an individual sacrifices for a close genetic relative, makes sense in an evolutionary context 3 T. M. Berra because some of the same genes of the individual making the sacrifice will be passed on by the kin who survives. This is referred to as inclusive fitness by Hamilton (1972). Realisation that humans share kinship with all animal life has helped to raise consciousness of how we treat other animals (Singer 1977). The ancestry of the AIDS virus, HIV-1 (human immu¬ nodeficiency virus-1) has been traced to SIVcpz (simian immunodeficiency virus) carried by our closest living rela¬ tive, the chimpanzees. Pan troglodytes (Bailes et al. 2003). This is not surprising from an evolutionary perspective. Somewhere in high school today there is a student who may contribute to the control of the AIDS epidemic. What chance of that would there be if creationism were taught as science in high school? Even religion is now being explained as having an evolutionary origin as a natural phenomenon once the brain evolved a critical mass and complexity (Dennett 2006). Bloch (2008) suggested that the evolution of imagination was a requisite for the emergence of religion which he con¬ sidered a logical extension of human sociality. This occurred about the time of the Upper Palaeolithic ''revolution" as manifested by an explosion of image-making and cultural transformations (White 2003). Acceptance of authority necessary for group cohesion and survival enforced by tool use and language combined with confusion of cause and effect and coincidences can result in the establishment of a belief that becomes dominate in a culture (Wolpcrt 2007). Those whose religion requires a literal interpretation of the Bible fear that the paradigm shift from supernatural ism to methodological naturalism threatens their beliefs. The 1925 Scopes trial, nicknamed "monkey trial” and “trial of the century'” in Dayton, Tennessee, has come to symbolise the struggle of religion against science in popular culture that later inspired the play and film Inherit the wind (Larson 1977). Such creationists arc particularly vocal in America which has a long standing tradition of anti-intellectual ism (Numbers 1992; Pigliucci 2002). This has resulted in a series of creationist legal challenges to evolution which have been decided in favor of evolution (Berra 1990). The most recent of these was the Intelligent Design creationist challenge of the Dover, Pennsylvania, School Board. The Intelligent Design creationist philosophy that life is too complex to have arisen by natural means and therefore had a supernatural origin has been critiqued in Pennock (2001) and exposed as a threat to science education by Forrest and Gross (2004). In the conclusion of his decision Judge John E. Jones III determined that the school board’s policy of teaching Intelligent Design violated the Establishment Clause of the First Amendment [separation of church and state] of the U.S. Constitution, and he wrote, “...in making this determination, we have addressed the seminal question whether Intelligent Design is science (Jones 2005). We have concluded that it is not, and moreover that Intelligent Design cannot uncouple itself from its creationist, and thus religious, antecedents.” He further wrote, "The breathtaking inanity of the board’s decision is evident when considered against the factual backdrop which has now been fully revealed through this trial.” Padian (2007) reviewed three books based on the Dover trial. Biotechnology, whether in the form of genetically modified crops, designer drugs, gene therapy, or the hu¬ man genome project all derive from Darwin’s profound insight. Darwin had no knowledge of genes, chromosomes, or how inheritance worked. This required additional input from the understanding of Gregor Mendel’s (1822-1884) genetic work. The modem evolutionary synthesis grew from Darwin’s explanation of natural selection and Mendel’s demonstration of inheritance augmented by the research of mathematically oriented population geneticists such as J.B.S. Haldane, Ronald Fisher, Sewall Wright, Thomas Hunt Morgan, Theodosius Dobzhansky, palaeontologist George Gaylord Simpson, botanist G. Ledyard Stebbins, Jr., biologist Julian Huxley (T.H. Huxley’s grandson), and the most important evolutionary biologist since Darwin, Ernst Mayr. This fu¬ sion of knowledge moved evolutionary science forward to the middle of the 20th century (Larson 2004). James D. Watson’s and Francis Crick’s 1953 demonstration that the molecular structure of DNA allowed for genetic coding was a huge breakthrough that ultimately made it possible to sequence the three billion chemical base pairs that com¬ pose the human genome and identify the approximately 20,000-25,000 genes in human DNA (Lander et al. 2001; Venter et al. 2001). Recent discoveries in evolutionary developmental biol¬ ogy, known as evo-devo, have shown that very similar genes are present in very dissimilar animals. These body-shaping genes are controlled by DNA switches called enhancers that turn them on or off at various times in development. Such enhancers are a major factor in the evolution of anatomy (Carroll 2005). The above examples are just a sample of the benefits to society that flow directly from the creative power of Charles Darwin’s theory of evolution by means of natural selection. The paradigm shift instigated by Darwin has made obvious the superiority of the scientific method as a means of understanding the world around us. It is ironic that the legacy of a man once destined for the church has been to replace supernaturalism with methodological naturalism. ACKNOWLEDGMENTS I am grateful to the Director of the Museums and Art Galleries of the Northern Territory, Anna Malgorezwicz, and to the Editor of the Beagle , Richard Willan, for the invitation to write this article in celebration of the bicenten¬ nial of Charles Darwin’s birth and the 150th anniversary of the publication of On the origin of species in 2009. This manuscript benefited greatly by a critique from Edward J. Larson. 4 Charles Darwin’ paradigm shift REFERENCES Bailes, E., Gao, F., Bibollet-Ruche, F., Courgnaud, V., Peeters, M., Marx, P., Han, B.H. and Sharp, P.M. 2003. Hybrid origin of SIV in chimpanzees. Science 300: 1713. Barkow, J.H., Cosmides, L. and Tooby, J. (eds). 1992. The adapted mind. Oxford University Press: Oxford. Barlow, N. (ed.) 1958. The autobiography of Charles Darwin 1809- 1882: with original omissions restored. W.W. Norton and Company : New York. Berra, T.M. 1990. Evolution and the myth of creationism. Stanford University Press: Stanford. Berra, T.M. 1998. A natural history of Australia. University of New South Wales Press/Academic Press: Sydney/San Diego. Berra, T.. 2009. Charles Darwin: the concise story of an extraordinary man. Johns Hopkins University Press: Baltimore. Bloch, M. 2008. Why religion is nothing special but is central. Philosophical Transactions of the Royal Society B 363: 2055-2061. Bowler, P.J. 1984. Evolution: the history of an idea. University of California Press: Berkeley. Browne, J. 1995. Charles Darwin voyaging, volume 1 of a biography. Princeton University Press: Princeton. Browne, J. 2002. Charles Darwin: the power of place, volume II of a biography. Knopf: New York. Carroll, S.B. 2005. Endless forms most beautiful: the new science of evo devo and the making of the animal kingdom. W.W. Norton: New York. Darwin, C. 1859. On the origin of species by means of natural selec¬ tion. or the preservation of favoured races in the struggle for life. John Murray: London. dc Waal, F. 1989. Peacemaking among primates. Harvard University Press: Cambridge. de Waal, F. 2005. Our inner ape. Riverhead Books: New York. Dennett, D.C. 1995. Darwin’s dangerous idea. Touchstone: New York. Dennett, D.C. 2006. Breaking the spell: religion as a natural phe¬ nomenon. Viking: New York. Desmond, A. and Moore, J. 1991. Darwin: the life of a tormented evolutionist. Warner Books: New York. Dibner, B. 1960. Darwin of the Beagle. Bumdy Library: Norwalk. Dobzhansky, T. 1973. Nothing in biology makes sense except in the light of evolution. American Biology’ Teacher 35: 125-129. Forrest, B. and Gross, P.R. 2004. Creationism's trojan horse: the wedge of intelligent design. Oxford University Press: Ox¬ ford. Jones, J.E. III. 2005. Memorandum Opinion. In the United States District Court for the Middle District of Pennsylvania. Tammy Kitzmiller, et al. Plaintiffs vs Dover Area School Board De¬ fendants. Case No. 04cv2688. 1-139 p. Hamilton, W.D. 1972. Altruism and related phenomena, mainly in the social insects. Annual Review of Ecology’ and Systematics 3: 193-232. Kuhn, T.S. 1962. The structure of scientific revolutions. University of Chicago Press: Chicago. Lander, E.S., and 19 coauthors 2001. Initial sequencing and analysis of the Human genome. Nature 409: 860-921. Larson, E.J. 1997. Summer for the gods: the Scopes trial and America s continuing debate over science and religion. Basic Books: New York. Larson. E.J. 2004. Evolution', the remarkable history of a scientific theory. The Modern Library: New York. Nesse, R.M. and Williams, G.C. 1996. Why we get sick. Vantage Books: New York. Numbers, R.L, 1992. The creationists: the evolution of scientific creationism. Alfred A. Knopf: New York. Padian. K. 2007. The case of creation. Nature 448: 253-4. Pennock, R.T. 2001. Intelligent design creationism and its critics: philosophical, theological, and scientific perspectives. Mas¬ sachusetts Institute of Technology Press: Cambridge. Pigliucci, M. 2002. Denying evolution: creationism, scientism, and the nature of science. Sinauer Associates: Sunderland. Singer, P. 1977. Animal liberation: towards an end to man s inhuman¬ ity to animals. Granada Publishing: London. Venter, J.C., and 273 coauthors 2001. The sequence of the human genome. Science 291: 1304-1351. White, R. 2003. Prehistoric art: the symbolic journey ofhumankind. 11.N. Abrams. Inc.: New York. Wolpert, L. 2007. Six impossible things before breakfast: The evolu¬ tionary origin of belief. W.W. Norton and Co.: New York. Wright, R. 1994. The moral animal. Pantheon Books: New York. Accepted 1 May 2008 5 ■ The Beagle, Records of the Museums and Art Galleries of the Northern Territory, 2008 24 : 7-13 A reassessment of Boronia (Rutaceae) in the Northern Territory with a key to species, the description of one new species and the reduction, in synonymy, of another species Marco F. Duretto Tasmanian Herbarium, Tasmanian Museum and Art Gallery, Private Bag 4, Hobart, Tasmania 7001, AUSTRALIA. Email: marco.duretto@tmag.tas.gov.au ABSTRACT Boronia (Rutaceae) is reassessed for the Northern Territory of Australia. Three informal names, B. aff. laxa 1 (Northern Plateau, Arnhem Land), B. aff. laxa 2 (Nabarlek, Arnhem Land) and B. aff. prolixa 1 (Red Lily Lagoon, Arnhem Land), are re-evaluated in the light of the significant number of new collections made over the previous decade. Ol the three, only one, B. aff. laxa 2 (Nabarlek, Arnhem Land), is considered to be worthy of formal taxonomic recognition and is here newly described as B. zeteticorum Duretto. The remaining two informal taxa are conspecific with B. laxa Duretto. Additional collections of B. gravicocca Duretto have shown that the taxonomic boundary between it and B. filicifolia A.Cunn. ex Benth. is not clear and thus B. gravicocca is synonymised under B. filicifolia. Boronia filicifolia was previously considered to be endemic to the Kimberley region of Western Australia. High levels of endemism of Boronia in the Northern Territory are discussed. The north-west portion of the Arnhem Land Plateau is determined to be a centre of diversity and endemism for the genus in the Territory and so is an area of high conservation value. A key to the species and subspecies of Boronia found in the Northern Territory is provided. Keywords: Rutaceae, Boronia, Boronia filicifolia, B. gravicocca. B. laxa, B. zeteticorum sp. nov. INTRODUCTION Boronia Sm. (Rutaceae) is an Australian genus of approximately 150 species found in all states and mainland territories (Wilson 1998; Duretto 1999,2003). The Northern Territory is a minor centre of diversity for the genus with 17 currently accepted species (12 endemic; one with two endemic subspecies) (Duretto 1999, 2003; Kerrigan and Albrecht 2007). In addition, three taxa, B. aff. laxa I (Northern Plateau, Arnhem Land), B. aff. laxa 2 (Nabarlek, Arnhem Land) and B. aff. prolixa l (Red Lily Lagoon, Arnhem Land), were informally recognised by Duretto and Ladiges (1997) and Duretto (1999). These taxa were not formally described at the time as there was insufficient material available and the names have not been adopted in the literature (e.g. Kerrigan and Albrecht 2007). Since 1997, a significant number of botanical surveys have occurred that have resulted in additional material for many rare species of Boronia , including B. amplectens Duretto, B. gravicocca Duretto and B. laxa Duretto. This additional material has allowed several taxonomic problems to be re-evaluated for the Northern Territory including the status of the informal taxa. Of the three informal taxa, two, B. aff. laxa 1 (Northern Plateau, Arnhem Land) and B. aff. prolixa I (Red Lily Lagoon, Arnhem Land), arc assignable to B. laxa. Boronia aff. laxa 2 (Nabarlek, Arnhem Land) is distinct enough to warrant formal taxonomic recognition, and is formally described below as B. zeteticorum Duretto, despite being known only from a single collection. Boronia zeteticorum sp. nov. is placed in B. subseries Grandisepalae and differs from all other species in that subseries (and B. subsection Grandisepalae ) by the leaves having distinctly different indumentum densities on the abaxial and adaxial surfaces. The species is similar to B. amplectens in that both have very narrow leaves; a character that could be considered an apomorphy for the two species and has proven to be robust with the addition of many collections of both B. amplectens and B. laxa over the last decade. Boronia amplectens has stellate hairs with appressed rays while the stellate hairs of B. zeteticorum are multi-angular (viz. not appressed). Boronia gravicocca Duretto was described from a small number of collections from Bradshaw Station (western Northern Territory) (Duretto 2003). It was subsequently targeted for population surveys by the staff of the Northern Territory Herbarium. Their collections have rendered B. gravicocca impossible to maintain as distinct from B. filicifolia A.Cunn. ex Benth. and so the former is here synonymised under the latter (see below). MATERIALS AND METHODS Herbarium material from mainly DNA, but also CANB, MEL and HO, were studied for this study. Herbarium abbreviations follow Holmgren et al. (1990). M. F. Duretto CLASSIFICATION AND ENDEMISM OF BORONIA IN THE NORTHERN TERRITORY Boronia is widely distributed across the ‘Top End’ of the Northern Territory, which is part of a biogeographical region that includes north-western Queensland and the Kimberley region of Western Australia. Boronia is classified into six sections (Wilson 1998; Duretto 1999, 2003, In press) of which only B. section Valvatae occurs in north-western Australia. The species of Boronia in the Northern Territory are placed in either B. subsection Valvatae ( B. lanceolata F. Muell., B. rupicola Duretto) or B. subsection Grandisepalae (remaining 16 species) (Duretto 1999, 2003) (Table 1). Boronia subsection Valvatae Within B. subsection Valvatae , B. lanceolata (Northern Territory, north-western Queensland) is most closely related to two species from central Queensland in B. series Valvatae subseries Lanceolatae (Duretto and Ladiges 1999; Duretto 1999, In press). Boronia rupicola (Northern Territory) is the sole representative of B. series Rupicolae (Duretto and Ladiges 1999; Duretto 1999). Interestingly, the only other member of B. subsection Valvatae in north-western Australia, B. hoipolloi Duretto (north-western Queensland), is placed in B. series Valvatae subseries Valvatae, and is closely related to species on Cape York Peninsula (Queensland) (Duretto and Ladiges 1999; Duretto 1999). Boronia subsection Valvatae is not found in the Kimberley region of Western Australia and the three representatives found in the Northern Territory and north-western Queensland represent three quite distinct lineages. This relationship is in contrast to what is found in B. subsection Grandisepalae. Table 1. Classification of Boronia in the Northern Territory. Boronia (all Australian states and mainland territories; c. 150 spp. in 6 sections) Boronia section Valvatae (W.A., N.T.. Qld, N.S.W., Vic.; 62 spp. in 5 subsections) subsection Valvatae (N.T., Qld. N.S.W., Vic.; 35 spp. in 4 series) series Rupicolae (NT.; t sp.) B. rupicola series Valvatae (N.T., Qld. N.S.W., Vic.; 24 spp. in 3 subseries) subseries Lanceolatae (N.T., Qld; 7 spp.) B. lanceolata (N.T., NW Qld) subsection Grandisepalae (N W. A., N.T., NW Qld; 20 spp. in 3 series) series Quadrilatae (N.T.; 2 spp.) B. cptadrilata, B. viridiflora series Grandisepalae (N.T.; 8 spp. in 2 subseries) subseries Verecundae (N.T.; 2 spp.) B. verecunda, B. xanthastrum subseries Grandisepalae (N.T.; 6 spp.) B. amplectens, B. grandisepala (2 subsp.: aciinthopsida and grandisepala), B. laxa, B. prolixa, B. suberosa, B. zeteticorum sp. nov. series Lanuginosae (N W.A., N.T., NW Qld; 10 spp. in 3 subseries) subseries Lanuginosae (N W.A., N.T.. NW Qld; 2 spp.) B. lanuginosa (NE W.A., N.T.. NW Qld), B. wilsonii (N W.A., NW N.T.) subseries Jucundae (N W.A., N.T.; 3 spp. ) B. dccumbens (N.T.), B. jucunda (N W.A., NW N.T.), II tolerans (N.T.) subseries Filicifoliae (N W.A.. NW N.T.; 5 spp.) B. filicifolia (N W.A., NW N.T.) Boronia subsection Grandisepalae Boronia subsection Grandisepalae contains 20 species, is restricted to north-western Australia (i.e. northern Western Australia, Northern Territory, north-western Queensland) and has three scries, Quadrilatae , Grandisepalae and Lanuginosae (Duretto and Ladiges 1999; Duretto 1999). The first two series are endemic to the Northern Territory. Boronia series Quadrilatae (B. quadrilata Duretto, B. viridiflora Duretto) is restricted to the north-western Arnhem Land Plateau. Boronia series Grandisepalae is more widespread and is classified into two subseries. Boronia subseries Verecundae (i.e. B. verecunda Duretto, B. xanthastrum Duretto) is found on the Arnhem Land Plateau and a few outliers to the west mainly in southern Kakadu National Park though B. xanthastrum is also found in eastern Arnhem Land. Boronia subscries Grandisepalae is more widespread, though only B. grandisepala F. Muell. is found outside the north-west portion of the Arnhem Land Plateau in eastern Arnhem Land and south of Mount Brockman to Nitmiluk National Park and also in the north-western parts of the Northern Territory. The remaining laxa (i.e. B. amplectens , B. laxa, B. prolixa Duretto, B. suberosa Duretto, B. zeteticorum sp. nov.) are narrow endemics of the north-western Arnhem Land Plateau and nearby outliers, an area that virtually matches the catchment area of the East Alligator River. Boronia series Lanuginosae is classified into three subseries that are all found in the Northern Territory and the Kimberley Region (northern Western Australia). Boronia subseries Lanuginosae contains two widespread species, B. lanuginosa Endl. (Western Australia, Northern Territory, north-western Queensland) and B. wilsonii Boronia (Rutaceae) in the Northern Territory (F. Muell. ex Benth.) Duretto (Western Australia, Northern Territory), the first species being the only representative of B. subsection Grandisepalae in north-western Queensland. Boronia subseries Jucundae contains three species. Two, B. tolerans Duretto and B. decumbens Duretto, are rare species of Nitmiluk and Kakadu National Parks. The third species, B.jucunda Duretto, is found in north-western parts of the Northern Territory and the eastern Kimberley region. Boronia subseries Filicifoliae contains five species of the Kimberley region, one of which, B. filicifolia, extends to the north-western parts of the Northern Territory. Of the 18 species of Boronia found in the Northern Territory, 13 are endemic; eight of these endemics are restricted to the north-western area of the Arnhem Land Plateau and nearby outliers roughly within the catchment area of the East Alligator River. This area measures approximately 75 x 75 km with the Mount Brockman outlier in the south-west comer. The remaining endemics are found to the south of this area on or near the western parts the Arnhem Land Plateau and outliers south to Nitmiluk National Park, though B. grandisepala and B. xanthastmm are found further afield (see above). The distributional pattern of the species of Boronia found in the Northern Territory matches that found for all endemic species of the Northern Territory as presented by Woinarski et al. (2006). That is, the greatest diversity is found on the north-west portion of the Arnhem Land Plateau and outliers with the number of taxa increasing dramatically from south to north (see also discussion in Duretto and Ladiges 1997). This area, or rather the catchment area of the Alligator River, was also identified as a minor hot-spot of endemism in an Australia-wide study on endemism by Crisp et al. (2001) and other studies (c.g. Bowman et al. 1988; Dunlop and Webb 1991). The north-west is typified by many species with very narrow distributional ranges. In some cases they do represent aggregates of closely related species. Boronia subseries Grandisepalae is an excellent example of an aggregate of many narrow endemics over a small area with live of its six species restricted to the north-west plateau area (see above). The four species with scrambling or lax habit (i.e. B. amplectens, B. laxa, B. prolixa, B. zeteticorum sp. nov.) form a monophylctic clade within this group (Duretto and Ladiges 1999: Fig. 10). The relationships of this clade of scrambling species with B. grandisepala and B. suberosa is equivocal as they form a trichotomy (see Duretto and Ladiges 1999). As B. grandisepala is also found outside the region the subseries may be represented in the area by more than one lineage. The remaining three species of Boronia endemic to the north-west Arnhem Plateau area are classified into two different series, B. series Quadrilatae (with two narrow endemics) and B. series Rupicolae (monotypic), both of which are endemic to the region and taxonomically isolated in their subsections (Duretto and Ladiges 1999; Duretto 1999; Table 1). In a cladistic analysis presented by Duretto and Ladiges (1999), these series, plus the species in B. series Quadrilatae , each have a large number of autapomorphies (see also Duretto 1999) suggesting a long period of isolation and/or a rapid accumulation of apomorphies. Also present in the north-west Arnhem Plateau region are two widespread species, B. lanceolata and B. lanuginosa, each placed in a different scries to the taxa discussed above (Table 1). Both species show considerable morphological variation over their ranges (Duretto 1997, 1999). Neobyrnesia J.A. Arrnstr., a monotypic genus closely related to Zieria Sm. (a genus of eastern Australia and New Caledonia) and Boronia , is also restricted to the north-west Arnhem Plateau region. Interestingly this genus shares a number of characters with species of Boronia section Valvatae, particularly those of B. subseries Grandisepalae, in the region - the valvatc and persistent petals and similar seed morphology (Paul G. Wilson pers. comm.; Stace and Leach 1994; Duretto and Ladiges 1997); the longitudinal rows on the seed testa are each made of a single line of tubercles (Armstrong and Powell 1980) as in B. subseries Grandisepalae and in contrast to Zieria (Duretto and Ladiges 1997). These characters may indicate a closer relationship than the current classification suggests or parallel adaptations to similar niches in the dissected sandstone landscape. Neobyrnesia suberosa J.A. Arrnstr. is a cliff specialist, as are B. suberosa (which superficially resembles N. suberosa as both are shrubs with corky stems and simple, opposite leaves), B. viridiflora, B. rupicola and species in other families found in the area (see discussion under species in Duretto (1999)). Cliff faces are a specialist niche exploited in a variety of ways by divergent taxa in the north-west Arnhem Land Plateau region. Woinarski et al. (2006) reviewed the literature and discussed several possible mechanisms behind the high level of endemism found in the north-west region ol the Arnhem Land Plateau. Within Boronia, the diversity seen in the region may be the combination of recent species radiation in conjunction with long-term isolation or refugia. A recent species radiation, where a large number of closely related species are found in close proximity, may be what is seen in B. subseries Grandisepalae. The endemic series (and Neobyrnesia ) probably also represent autochthonous taxa that have been in the region and/or isolated for a longer time. In effect, for these taxa, the region has acted as a refugium. In addition, the region is also home to widespread species that may be relatively recent additions to the local flora. When the species of Boronia found in the north-west portion of the Arnhem Land Plateau are highlighted on the cladogram of/?, section Valvatae presented by Duretto and Ladiges (1999: Figs 9 and 10) at least five distinct lineages are identified: B. lanceolata, B. lanuginosa, B. rupicola, B. scries Quadrilatae, B. subscrics Grandisepalae (see above). Only two of these lineages are represented by more than one species. The area is not only the most diverse area of the Northern Territory (and north-western Australia) in terms of the number of species of Boronia, but it is also the most diverse in terms of the number of infrageneric taxa present. 9 M. F. Duretto Thus the north-west portion of the Arnhem Land Plateau (i.e. the catchment area of the East Alligator River) is an area of extremely high conservation value. TAXONOMY Boronia filicifolia A.Cunn. ex Benth., FI. Austral. 1: 311 (1863) Type: Australia, Western Australia, Montague and York Sounds, N.W. Australia, 1820, A.Cunningham 220. third voyage of the “Mermaid” (lectotype. /zV/e Duretto Nuytsia 11: 332 (1997): K n.v. (photograph AD, cibachrome MEL); isolectotype: BM n.v. (transparency MEL, photograph PERTH). Synonymy. Boronia gravicocca Duretto (2003): 123, syn. nov. Illustrations. Wheeler et al., (1992) FI. Kimberley Region 669, Figs 206 A1 and A2; Duretto (2003) Muelleria 17: 102, figs 12 K-M, as B. gravicocca; Duretto (2005) Australian Plants 23 (183): 90 (2005). Remarks. For a full description on this species see Duretto (1997, 1999, 2003). Distribution and ecology. Boronia filicifolia occurs in the Mitchell River and Port Warrender areas (north-west Kimberley region. Western Australia) and disjunctly at Bradshaw Station (Victoria River area, Northern Territory). It is found in heath and open woodland on sandstones and quartzites. Flowering December-June; fruiting (December, Feb.) June-July. Boronia laxa Duretto, Austral. Syst. Bot. 10: 279 (1997) Type. Australia, Northern Territory, Site FF, c. 30 km SE of Jabiru, 12°55’S 132°59’E, 30 March 1981, L.A. Craven 6600 (holotype: CANB 338123 (photograph HO); isotypes: AD, DNA 20968 (transparency MEL), MEL 2041245, P, US). Synonymy. Boronia aff. laxa l (Northern Plateau, Arnhem Land) Duretto and Ladiges (1997): 282. Boronia atf. prolixa 1 (Red Lily Lagoon, Arnhem Land), Duretto and Ladiges (1997): 285. Illustration. Duretto and Ladiges (1997), Austral. Svst. Bot. 10: 280, fig. 20 a-b. Remarks. For full descriptions as well as ecological and distributional data on this species see Duretto and Ladiges (1997) and Duretto (1999). Boronia zeteticorum Duretto, sp. nov. (Fig. 1) A B. laxa Duretto foliis angustioribus (1.5-3.5, non 2.5-10 mm latis) indumenta abaxiali denso, adaxiali moderate denso dijfert. Differs from B. laxa by its narrower leaves (1.5-3.5 mm wide, not 2.5-10 mm wide) with a dense indumentum on the abaxial surface and a moderately dense indumentum on the adaxial surface. Type. Australia, Northern Territory, Nabarlek, Arnhem Land, 12°19’S 133°19’E, 23 March 1989, R. Hinz 467 (holotype: DNA 43905 (transparency MEL); isotype: CANB). Synonymy. Boronia aff. laxa 2 (Nabarlek, Arnhem Land) Duretto and Ladiges 1997: 282. Illustration. Duretto and Ladiges, Austral. Syst. Bot. 10. 270, fig. 16i-j (leaf hairs); 276, fig. 19i (seed surface) (1997), all as B. sp. aff. laxa 2. Specimen examined. Known from the type material only. Description. Semi-prostrate, much branched subshrub to 50 cm long; multi-angular stellate hairs sessile, 6-25+ rays per hair; rays unicellular, unfused, firm, straight, glossy, smooth, yellow-white, 0.1-0.2 mm long. Branches not obviously glandular, with little cork development, with a moderately dense stellate indumentum. Leaves simple; petiole 0.5-1.5 mm long; lamina 10-35 mm long, 1.5-3.5 mm wide, narrowly elliptic, tip acute, base attenuate, margins fiat, slightly discolorous, paler beneath, not obviously to slightly glandular, midrib raised abaxially, slightly impressed adaxially; adaxial surface with moderately dense stellate indumentum (Duretto and Ladiges 1997: Fig 16i); abaxial surface with dense stellate indumentum (Duretto and Ladiges 1997: Fig 16j). Inflorescence axillary, 1-flowered, with a dense stellate indumentum; peduncle 0.5 mm long; prophylls to 2 mm long and 0.5 mm wide; metaxyphylls minute to 1 mm long; pedicels c. 1.5 mm long. Sepals white, longer and wider than petals, 3.5-4 mm long, c. 2 mm wide, enlarging to 6 mm long and 3.5 mm wide with fruit, ovate to dcltate, tip acute to acuminate, valvate in bud, persistent with mature fruit; adaxial surface with dense and minute indumentum along margins, becoming glabrous towards centre and base; abaxial surface with moderately dense stellate indumentum. Petals white, 3-3.5 mm long, 1-1.5 mm wide, enlarging to 4.5-5 mm long in fruit, valvate in bud, persistent with mature fruit; adaxial surface with sparse stellate indumentum becoming glabrous towards base; abaxial surface with moderately dense to dense stellate indumentum. Stamens: filaments bearing stiff, simple hairs abaxially and on margins below glandular tip; antesepalous filaments clavate and suddenly narrowing to anther connective, 1.5 mm long, distal 0.75 mm prominently glandular; antepetalous filaments slightly glandular, c. 1 mm long; antepetalous anthers much larger than antesepalous anthers. Ovary glabrous; style glabrous. Cocci with moderately dense simple and stellate indumentum, c. 4 mm long, c. 2 mm wide. Seeds black, shining, c. 3 mm long, 1.5 mm wide, elliptical with adaxial side flattened and with prominent ridge; testa striated, at magnification tuberculate to colliculate with units fused to form longitudinal ridges 33-50 pm apart; ridge units 10 Boronia (Rutaceae) in the Northern Territory DNA 43908 "Boronitc. Co r por.-tf'o Dim HOCOT HF r. f-A, -u.p Tasmanian herbarium ihoi NATIONAL HERBARIUM OF VICTORIA j! fine- jLrv$? '2— ■ [let? Tl-I L-A'y, ^-‘7. Svrr. ier. It?: ?n f ' C '“'V.3 Dcterminavit /^/ ^ /Jk Seen for f/ora of Australia northern territory herbarium DARWIN (DNA). AUSTRALIA rlek, Arnhca Land. 12.19S 133.19E E/DIST. t HT/DC jtonc country) rambling to protrate with white flowcrn. - , •■»!-> tOflQ rutackak Fig. 1. Holotype of Boronia zeteticorum sp. nov. (Hinz 467 , DNA 43905). Illustration 55% actual size. M. F. Duretto unicellular, smooth, anticlinal walls more or less visible, 30-43 jam across (Duretto and Ladiges 1997: Fig 19i). Distribution and ecology. Boroniazeteticorum is known only from the Nabarlek area (Northern Territory) (Duretto and Ladiges 1997: Fig. 2; or Duretto 1999: Fig 16). The only ecological information given with the single collection is that it was made in sandstone country. The material was collected in March and has both flowers and fruit. Remarks. Boronia zeteticorum differs from other species in B. subsection Grandisepalae by the surfaces of the leaves having contrasting indumentum densities, i.e. a dense indumentum on the abaxial surface but a moderately dense indumentum on the adaxial surface (Duretto and Ladiges 1997: Fig. 16ij). It is placed in B. subseries Grandisepalae and is semiprostrate with weak stems, as are B. amplectens, B. laxa and B. prolixa with which it is probably forms a monophyletic group. It differs from B. laxa and B. prolixa by its very narrow leaves: a character it shares with B. amplectens. Boronia amplectens has stellate hairs with rays apressed to the surface while the stellate hairs of B. zeteticorum are multi-angular. Conservation Status. The species is known from a single collection made in 1989. Field surveys are required urgently to ascertain the correct conservation status for this species. Etymology. The specific epithet is derived from the Greek, zetetikos (disposed to search), as coined by Craven and Jones (1991: 529), and ‘honours those persons who for their enjoyment explore natural vegetation communities to become familiar with, and collect, their constituent species’. The systematic study of Boronia , like many genera of plants, would not be where it is today without the collections made by such people in remote places. KEY TO SPECIES AND SUBSPECIES OF BORONIA FOUND IN THE NORTHERN TERRITORY For descriptions, ecology and distributional data see Duretto (1999); for B. filicifoliae see also Duretto (2003) to include information given under B. gravicocca. la. Sepals much shorter than petals (< 'A length); antepetalous anthers c. equal to antesepalous anthers.2 1 b. Sepals nearly as long (> 'A) to much longer than petals; antepetalous anthers much larger than antesepalous anthers.....3 2a. Leaves simple, midrib on the abaxial surface prominently raised; erect or very rarely pendulous shrubs; petals 2-7 mm long; staminal filaments glabrous or with 1-3 hairs. B. lanceolata 2b. Leaves 1-7-folilate, midrib of abaxial surface not prominently raised; pendulous shrubs growing on cliff faces; petals 2-3 mm long; staminal filaments hirsute. B. rupicola 3a. Leaves pinnate.4 3b. Leaves simple.9 4a. Sepals shorter and narrower than petals, 2-3.5 mm long; at least some leaves 27—41 -loliolate, with a sparse to moderately dense indumentum (epidermis clearly visible); lateral leaflets c. rhombic to circular to broadly elliptic; petiole to 2 mm long. . B. filicifolia 4b. Sepals longer and wider than petals, 3-15 mm long; leaves 3-27(-35)-foliolate (if > 30 leaflets then some petioles > 3 mm long), glabrous to densely stellate tomentose; lateral leaflets linear to narrow-elliptic, rarely elliptic; petiole to 7 mm long.5 5a. Leaves petiolate, though petiole sometimes as small as c. 0.5 mm long; leaflets linear-elliptic to elliptic, the margins plane or recurved to revolute; lamina glabrescent or with a sparse to dense, stellate indumentum. 6 5b. Leaves sessile; leaflets linear-elliptic, plane or slightly recurved along margins; lamina glabrescent or with a sparse, stellate indumentum.7 6a. Leaflets linear to narrowly elliptic, so revolute that abaxial surface not usually visible; sepals (4—)5—14 mm long, glabrous or with a sparse to dense indumentum (Top End). B. lanuginosa 6b. Leaflets elliptic to lanceolate, abaxial surface visible; sepals 5-9 mm long, with a dense indumentum (Victoria River area). B. wilsonii 7a. Branches obviously glandular; leaves 3-foliolate. . B. jucunda 7b. Branches eglandular; leaves (3-)5-9-foliolate.8 8a. Plants decumbent, with a sparse to moderately dense, simple indumentum, stellate hairs rare; leaflet-margins slightly recurved. B. decumbens 8b. Plants erect, with a sparse stellate indumentum; leaflets plane. B. tolerans 9a. Plants, other than flowers, glabrous, glaucous (especially leaves); stems distinctly quadrangular, at least on young shoots...10 9b. Plants sparsely to densely hirsute, not glaucous; stems terete to slightly quadrangular.11 10a. Erect shrub growing on ridges; leaves sessile; sepals 6-13 mm long; petals 4-5 mm long....5. quadrilata 10b. Horizontal shrub growing from cliff faces; leaves petiolate; sepals and petals 2.5-3 mm long. . B. viridiflora 12 Boronia (Rutaceae) in the Northern Territory 11a. Stellate hairs prominently stalked, rays 0.5-1 mm long; cocci glabrous.12 lib. Stellate hairs without prominent stalks, rays to 0.5 mm long; cocci hirsute.13 12a. Hairs white and flexuous, new shoots pinkish to white; leaves narrowly elliptic; adaxial surface of petal glabrous or hirsute. B. verecunda 12b. Hairs yellow and straight, new shoots yellow; leaves elliptic; adaxial surface of petal hirsute. ... B. xanthastrum 13a. Older stems with massively developed cork; indumentum of leaves usually difficult to see with the unaided eye, rays of hairs to 0.1-0.3 mm long... ... B. suberosa 13b. Older stems not corky; indumentum of leaves clearly visible to the unaided eye, rays of hairs 0.1-0.5 mm long.14 14a. Plants erect (rarely sprawling but then with a hoary, dense indumentum), with a moderately dense to dense indumentum; sepals greater than 7 mm long at anthesis (rarely less than 7 mm long but then plant with a hoary, dense indumentum).15 14b. Plants sprawling, sparsely to moderately hirsute (rarely densely hirsute on the abaxial leaf-surface only); sepals less than 8 mm long at anthesis, sometimes enlarging to 11 mm long as fruit matures .16 15a. Leaves with a very dense indumentum, epidermis not visible. B. grandisepala subsp. grandisepala 15b. Leaves with a moderately dense indumentum, leaf epidermis visible. . B. grandisepala subsp. acanthopsida 16a. Leaves narrow-elliptic, 1-4 mm wide.17 16b. Leaves elliptic, lanceolate, ovate, > 5 mm wide... 18 17a. Leaves: indumentum of both abaxial and adaxial surfaces similar, sparse; rays of hairs appressed. . B. amplectens 17b. Leaves: indumentum of abaxial and adaxial surfaces markedly different; that of the abaxial surface dense; that of the adaxial surface moderately dense; hairs multi-angular, rays not appressed. B. zeteticomm 18a. At anthesis Dowers stalks (peduncle and pedicel) 2-7 mm long, not bent at prophylls (bracts); leaves elliptic to lanceolate. B. laxa 18b. At anthesis at least some flowers stalks (peduncle and pedicel) > 10 mm long, often bent at prophylls (bracts); leaves ovate to lanceolate. B. prolixa ACKNOWLEDGMENTS I would like to thank the Director and staff of DNA for access to the collections and rapid processing of loan material as well as Neville Walsh (MEL) for the Latin diagnosis. REFERENCES Armstrong, J.A. and Powell, J.J.M. 1980. Neobyrnesia (Rutaceae), a new genus endemic to northern Australia. Telopea 1: 399-408. Bowman, D.M.J.S.. Wilson, B.A. and Dunlop, C.R. 1988. Preliminary biogeographic analysis of the Northern Territory vascular flora. Australian Journal of Botany 36: 503-517. Craven, L.A. and Jones, S.R. 1991. A taxonomic review of Homoranthus and two new species of Darwinia (both Myrtaceac, Chamelaucieae). Australian Systematic Botany 4:513-533. Crisp, M.D., Laffan, S., Linder, H.P. and Monro, A. 2001. Endemism in the Australian flora. Journal of Biogeography 28: 183— 198. Dunlop, C.R. and Webb, L.J. 1991. Flora and vegetation. Pp. 41-60 In: C.D. Haynes, C.D., Ridpath, M.G. and Williams, M.A.J. (eds) Monsoonal Australia: landscape, ecology and man in northern lowlands. A.A. Balkema: Roterdam. Duretto, M.F. 1997. Taxonomic notes on Boronia species of north-western Australia, including a revision of the Boronia lanuginosa group (Boronia section Valvatae: Rutaceae). Nuytsia 11: 301-346. Duretto, M.F. 1999. Systematics of Boronia section Valvatae sensu lato (Rutaceae). Muelleria 12: 1-131. Duretto, M.F. 2003. Notes on Boronia (Rutaceae) in eastern and northern Australia. Muelleria 17: 19-135. Duretto, M.F. In press. Description of a new subsection and two new subseries for Boronia section Valvatae (Rutaceae). Austrobaileya. Duretto. M.F. and Ladiges, P.Y. 1997. Morphological variation within the Boronia grandisepala group (Rutaceae) and the description of nine taxa endemic to the Northern Territory, Australia. Australian Systematic Botany 10: 249-302. Duretto, M.F. and Ladiges, P.Y. 1999. A cladistic analysis of Boronia section Valvatae (Rutaceae). Australian Systematic Botany 11:636-665. Holmgren, P.K., Holmgren, N.ll. and Barnett, L. 1990. Index Herbariorum. Part /. The Herbarium of the World, 8th edn. New York Botanical Gardens: New York. Kerrigan, R.A. and Albrecht, D.E. 2007. Checklist of NT vascular plant species. Northern Territory' Herbarium: Darwin. Stace, H.M. and Leach, G.J. 1994. Cytological notes in Rutaceae. 2: Neobyrnesia suberosa. Telopea 6: 167-168. Wilson, P.G. 1998. New names and new taxa in the genus Boronia (Rutaceae) from Western Australia, with notes on seed characters. Nuytsia 12: 119-154. Woinarski, J.C.Z., Hempel, C., Cowie, I.,Brennan, K., Kerrigan, R., Leach, G. and Russcll-Smilh, J. 2006. Distributional pattern of plant species endemic to the Northern Territory, Australia. Australian Journal of Botany 54: 627-640. Accepted 26 August 2008 13 The Beagle, Records of the Museums and Art Galleries of the Northern Territory, 2008 24: 15-21 Jaw growth and replacement in Diopatva aciculata (Annelida: Onuphidae) HANNELORE PAXTON 1 AND MILADA SAFARIK 2 'Department of Biological Sciences, Macquarie University, Sydney, NSW 2109, AUSTRALIA Email: hpaxton@rna.bio.mq.edit.au 2 Aquabait Pty Ltd, PO Box 5107, Dora Creek, NSW2264, AUSTRALIA ABSTRACT Jaw development was studied in 1 - to 45-day old Diopatra aciculata (fertilisation to 44-chaetigers). Both mandibles and maxillae became visible in 4- to 6-chaetiger worms. The initial protomandibles become embedded in the adult mandibles. The larval maxillae consist of a left fang-like and right small serrated clement, a pair of serrated plates, distal ridges, and a single carrier. Larval maxillae were replaced by juvenile maxillae in 14- to 15-chaetiger worms. Juvenile maxillae have adult-like carriers. Mil and MIV, but MI are proximally dentate with a distal hook; Mill and MV are absent. First adult maxillae were seen in 21 -chaetiger worms. Each subsequent moult results in larger but otherwise identical apparatuses. The moult increment (difference in size between successive moults) was established as 1.34 and used to extrapolate estimates of size classes. Based on these calculations our largest wonns had their fourth set of adult maxillae, having moulted the adult maxillae every three to four days. Keywords: Polychacta, Eunicida, jaw apparatus, mandibles, maxillae, size classes, moulting. INTRODUCTION The tubicolous onuphid polychaete Diopatra acicu¬ lata Knox and Cameron. 1971 occurs in shallow depths along the southern shores of Australia (Paxton 1993). It is collected as prized fishing bait and commercially farmed as bait and food in the conditioning of prawn broodstock (Safarik et al. 2006). The worms have a lifespan of about five years (Safarik et al. 2006) and are dioecious. Mature eggs are about 230 pm in diameter. After a free-swimming lecithotrophic larval stage of 4 to 6 days D. aciculata settles, starts to build its tube and feed. As typical for Eunicida, D. aciculata has a complex jaw apparatus consisting of ventral mandibles and dorsal maxillae. While mandibles grow during the lifetime of an individual, the maxillae arc moulted or replaced many times. Larval and juvenile jaws have been reported for several species of Onuphidae (Table 1), although most of these reports were anecdotal. The most complete study is of Kinbergonuphis simoni Santos et al., 1981, following the development of larval to adult maxillae and confirming the replacement of larv al maxillae by adult types (Hsieh and Simon 1987), previously only known for Mooreonuphis jonesi Fauchald, 1982 (Fauchald 1982). Shedding and replacement of adult maxillae in D. aciculata, comparable to arthropod moulting, has been reported (Paxton 2005). Replacement of maxillary jaws in Eunicida has been recently reviewed, suggesting that the type of moulting observed in Onuphidae may also take place in Eunicidae and Lumbrineridae (Paxton 2006). However, it is not known how often the maxillae are replaced. While those of older adult worms may moult infrequently or not at all, it is expected that those of young worms undergo frequent moults to keep pace with the rapidly growing body. The aim of the present paper was to study the de¬ velopment of jaws in very young (1 to 45 days old) D. aciculata in order to describe and illustrate the growth of the mandibles and replacement of maxillae. It was hoped that relating the size of the maxillary elements to the age and size of the worms might reveal age classes and thus a possible moult cycle. 9 MATERIALS AND METHODS Diopatra aciculata were reared in outside ponds at the Aquabait Pty. Ltd. aquaculture facility at Dora Creek, NSW, Australia. The study material ranged from newly fertilised eggs (day 1) to animals consisting of44 chaetigers (maximum of 45 days). The three stages of the maxillary apparatus are referred to as larval, juvenile and adult, even though they may not correspond to the state of maturity of the animals. Animals were mounted on a slide with diluted glycerine added to the edge of the coverslip and examined with a compound microscope. The jaws were studied in situ. The number of chaetigers was used as an expression of body size. The elements of the adult maxillary apparatus are numbered in the conventional way in Roman numerals progressing from posterior to anterior, so that for instance MIL refers to the left most posterior maxilla, following the maxillary carrier (Fig. 1A). The length of maxilla 1 was taken as an indicator of the size of the complete maxillary apparatus. The maxilla I was drawn with the aid of a camera lucida and measured on the drawing. H. Paxton and M. Safarik Fig. 1. Jaws of Diopatra aciculata: A, adult maxillae of 42-chactigcr specimen, dorsal view; B, protomandibles of 4-chaetiger specimen, dorsal view; C, protomandibles of 5-chaetiger specimen, dorsal view; D. mandibles of 42-chactiger specimen, showing incorporated protomandibles, ventral view; E, developing larval maxillae in 4-chaetiger specimen, dorsal view; F, larval maxillae with short carrier in 4-chaetiger specimen, dorsal view; G, carrier of 5-chaetiger specimen, dorsal view; H, same of 12-chaetiger specimen, dorsal view; 1, juvenile maxillae of 19-chaetiger specimen. 16 Jaw growth and replacement in Diopatra aciculata The moult increment is a term used in entomology to describe the increase in size between two successive moults. Dividing the postmoult size by the premoult size is a constant for a certain species and represents the moult increment (Gullan and Cranston 1994). RESULTS Overall growth. The overall growth of the worms as expressed in numbers of chaetigers over time is shown in Figure 2. The largest animals attained a size of 44 chacti- gers. We are aware of the shortcomings of the quantitative nature of this data resulting from some time periods when sampling was not possible. Due to the scarcity of the mate¬ rial we have combined the results from two separate cohorts. While this has bolstered the numbers, it has introduced a greater range of variation due to the different growth rates of the two cohorts. I lowever, in the absence of any previous study of this kind, we feel that the qualitative nature of the data overrides the quantitative problems. Jaw growth - mandibles. The anterior part of the cut¬ ting plate is the first part of the mandibles to become visible in 4-chaetiger worms, when they are four to six days old. The plates initially appear as lightly sclerotised structures (Fig. 1B) and become H-like as more material is deposited (Fig. 1C). With continuous growth they develop the typi¬ cal adult form which has proximal long shafts and a distal cutting plate (Fig. 1D). Jaw replacement - maxillae. The maxillary apparatus becomes visible at the same time as the mandibles. The appearance and moulting sequence of the maxillae are summarised in Table 2. The tips of the teeth or serrations are the first parts of the larval maxillae to become visible (Fig. 1E) and within a few days the whole apparatus is darkly sclerotised (Fig. 1F). It consists of a single narrow carrier, a large serrated left fang¬ like element, a very small serrated right element, a large pair of serrated plates, and a distal pair of ridges. The initial carrier is a short cup-shaped structure that articulates with the left fang-like element, has a small extension articulating with the right small plate, and a short posterior extension. This extension rapidly grows and appears to form a tube¬ like structure (Fig. 1G), that in some preparations opens posteriorly, displaying growth lines (Fig. 1 FI). The larval maxillae are shed, swallowed, and replaced by the juvenile maxillae (Fig. II) in 14- to 15-chaetiger worms. That apparatus is weakly sclerotised, particularly the distal part. The juvenile maxillae consist of carriers, paired asymmetrical MI that are proximally denticulated and have a distal hook, paired denticulated Mil and less distinct paired MIV. Maxillae III and V appearto be absent. The first adult maxillae (Fig. I A) appear in 21-chaetiger worms, consisting of the typical adult form. Subsequent moults result in larger but otherwise identical maxillae. Size classes. The length of maxilla 1 was taken as an indicator of the size of the total maxillary apparatus and was found to range from 80-105 pm for juvenile and o 5 10 15 20 25 30 35 40 45 50 Age in days Fig. 2. Growth curve of Diopatra aciculata indicating stage of maxillae of each specimen, as shown in key in lower right: (L) larval- (J) juvenile; (A) adult stage A; (B) adult stage B; (C) adult stage C; (D) adult stage D. 17 H. Paxton and M. Safarik Fig. 3. Frequency histogram of MI length of juvenile (J) and adult maxillae of stages (AD) o (Diopatra aciculata. The mean length of each stage is indicated by asterisk and value above the respective column. 120-310 pm for adult jaws. A frequency histogram of the length of maxilla I was prepared to reveal size classes (Fig. 4). It demonstrated a well defined juvenile class. The other classes appear less clear, partly due to lack of material (class A) and continuity between classes B and C. In the absence of clear size classes we endeavoured to discover the moult cycle utilising the moult increment (Table 3). The known moult increments came from two individuals with recently moulted maxillae in situ and discarded old sets recovered from the gut. A 31-chaetiger specimen (35 days old) had a moult increment of 1.32, while an 89-chaetiger specimen (82 days old) from a different study had a moult increment of 1.38. Assuming that the moult increment between the juvenile and adult maxillae equals that of consecutive adult maxillae, we calculated the value by utilising the mean size of four earliest adult and Table 1. Reports of protomandibles, larval and juvenile maxillae for the family Onuphidae. Species Protomandiblcs Size of norm Larval maxillae Size of norm Juvenile maxillae Size of norm Reference Diopatra aciculata Knox and Cameron, 1971 4- to 20-chaetiger same 14- to 28-chaetiger present study Diopatra cuprea Bose, 1802 5- to 6-chaetiger same - Allen 1959 Diopatra lilliputiana Paxton, 1993 15-chaetiger same - Paxton 1993 Diopatra cf. marocensis Paxton et at., 1995 as D. cuprea 11-chaetiger same - Monro 1924 Hyalinoecia araucana Carrasco, 1983 13-chaetiger same - Carrasco 1983 Hyalinoecia incubans Orensanz, 1990 7-chaetiger same - Orensanz 1990 Kinbergonuphis simoni (Santos et al, 1981) 5- to 14-chaetiger same 15-chaetiger Hsieh and Simon 1987 Leptoecia abyssorum Chamberlin, 1919 as Paronuphis abyssorum 9-chaetiger same - Averincev 1972 Mooreonuphis jonesi Fauchald, 1982 - 8-chaetiger 13-chaetiger Fauchald 1982 Onuphis elegans (Johnson, 1901) 5- and 7-chaetiger 5-chaetiger 14-chaetiger Blake 1975 18 Jaw growth and replacement in Diopatra aciculata Table 2. Appearance and moulting sequence of maxillary apparatus of D. aciculata. Type of maxillae Larval (Fig. IF) Juvenile (Fig. II) Adult (Fig. 1 A) Number and type of elements in posterior to anterior order Single carrier Left fang-like element; small right serrated element (?M1) Pair of large serrated plates (?M1I) Pair of ridges (?MIV) Paired carriers Paired asymmetrical MI, denticulate with distal hook Paired denticulated Mil Paired M1V Paired carriers Paired falcate MI Paired denticulate Mil Single left denticulate Mill Paired MIV Paired MV Age and size of worm Days since fertilisation Number of chaetigers 4 to 6 4 to 6 28 14 or 15 33 21 to 29 ten juvenile MI, establishing the value as 1.33. Thus, the available data indicate a mean moult increment of 1.34. To obtain an estimate of size classes we used the moult increment lor extrapolation. The smallest and largest ju¬ venile MI values were multiplied by the moult increment to obtain the range of the adult (A) size class. Three more size classes were extrapolated in the same manner to give the ranges for four size classes (Table 4). The estimated size classes have been indicated for each individual in the growth curve (Fig. 2), demonstrating that the first and second moults occurred more or less at a certain chaetiger number but that the later stages overlapped. The estimated size classes were superimposed on the frequency histogram of MI length (Fig. 3), indicating again a well defined juve¬ nile class but indistinct adult size classes. DISCUSSION It has been stated that the mandibles appear before the maxillae in the labidognath taxa (Kielan-Jaworowska 1966). This is not the case for D. aciculata and most likely other labidognath taxa as well. While larval mandibles have been illustrated in several studies (Table 1), their relationship to the adult mandibles is generally not fully appreciated. The sclerotiscd larval mandibles are here referred to as protomandibles and can be noted as a dark, X-shapcd structure through the body wall (Fig. 1C). They will be enlarged by the exterior depo¬ sition of sclerotiscd proteins and carbonates, throughout life, at the areas in contact with the cuticular epithelium, i.c. ventrally and laterally. The protomandibles arc visible in young worms in a ventral view (Fig. 1D) and sometimes still visible in adults as small dark lines. In decalcified mandibles the protomandibles are sometimes longer than the sclerotiscd plate and have been referred to as ‘Stacheln’ or thorns by von 1 latTner (1959). Larval maxillae are known for a number of species belonging to several genera of Onuphidae (Table 1). The main elements of the larval maxillae of all but one species were reported to consist of proximal narrow carriers, a pair of plates, a single left fang-like element and a distal pair of ridges. The only exception, Onuphis elegans (Johnston 1901), was reported as having paired fang-like elements (Blake 1975). The otherwise remarkable similarity of the Table 3. Calculation of moult increment of maxillary apparatus of D. aciculata. *) specimen from another study (Paxton 2006) Method Number of worms Stage of maxillae Size of worms Size of premoult Ml Size of postmoult MI Moult increment Observation 1 early adult 31-chaetiger 4^ OO T: 3 195 pm 1.32 Observation 1 adult 89-chaetiger* 840 pm 1160 pm 1.38 Estimate 14 earliest juv./ 14- to 21-chae- 80-105, x=93 pm 120-125, x=124 1.33 adults tiger Mean of 3 values = 1.34 19 H. Paxton and M. Safarik larval maxillary elements of O. elegans to those of the other species needs investigation. The larval maxillae of D. aciculata (Fig. 1F) differ from those reported for other onuphids in having an additional very small serrated right element. This element is difficult to observe and presumably has been overlooked in all but one report. Krohn and Schneider (1867) reported several annelid larvae. One of these, observed at Madeira, was followed to the 5-chaetiger stage when it possessed five rudimentary prostomial appendages and the larval jaw apparatus. The authors concluded that the jaws indicated that it belonged to the family Eunicidae and because of the number of ‘antennae’ probably to the genus Eunice. We now know that Eunice at that stage has only one median antenna, and that the larva must have been an onuphid. The amazing part of the description and small drawing of the jaw apparatus is that Krohn observed the small right maxillary element that has presumably been overlooked by all subsequent researchers. The larval maxillae are replaced by the juvenile maxillae (Fig. 11). These resemble the adult apparatus with short, broad carriers, paired forceps-like Ml and paired Mil, and were referred to as adult jaws by Hsieh and Simon (1987). The juvenile MI differs remarkably from the adult type in that it is proximally medially dentate with a distal hook. The right hook is longer than the left one, and is similar to the prionognath type MI of the Oenonidae (e.g. Arabella). The apparatus is weakly sclerotised, particularly the distal part. In our specimens MI and Mil were clearly visible, and MIV weakly visible. The single left Mill and paired MV of the adult maxillae appeared to be absent. However, Fauchald (1982) described and figured Mill but not MV for Mooreonuphis jonesi while Hsieh and Simon reported MV but not Mill for Kinbeigonupliis simoni. We think it is probable that the Mill is not present in the juvenile apparatus and makes its first appearance in the adult apparatus. The presumed homology of the larval and juvenile maxillary elements and their evolutionary significance will be discussed elsewhere (Paxton in press). Once the juvenile maxillae were replaced by the adult type, each additional moult resulted in slightly larger but otherwise identical maxillae. Figure 3 is a frequency histogram of maxilla I in an attempt to demonstrate size classes. A clear juvenile class was demonstrated while the other classes were inconclusive. This is not only a result of the scarcity of the data. A similar attempt to estimate size classes of the fossil polychaete Kettnerites ( Kettnerites ) bankvaetensis Bergman, 1987, was carried out by Bergman (1989). He measured 134 jaws (MI) ranging in size from 200 pm to 890 pm and did not obtain distinct size classes. Bergman discussed some factors affecting a polymodal size frequency distribution of Eunicida jaws and re (erred to a statement by P.J.W. Olive that with polychaetcs it is virtu¬ ally impossible to identify size classes beyond the second class (P.J.W. Olive pers. comm, to C. Bergman, 1986). The Table 4. Estimated size classes of adult maxilla 1 of D. aciculata. Size class Range of size (pm) Mean (pm) Adult A 107-141 124 Adult B 144-189 167 Adult C 192-253 223 Adult D 258-339 299 larger classes tended to merge due to individual growth rates, food supply and genetics. The difference in size between two instars in insects is called the moult increment. The increase of sclerotised parts, such as the head capsule, a regular linear progression in successive instars, is known as Dyar’s rule (Gullan and Cranston 1994). Our moult increment was established as 1.34 (Table 3) and is comparable to the usual insect value of 1.4 (Hinton and Mackerras 1970). Using this index, we have estimated size classes of adult maxillae (Table 4) based on the range of juvenile jaws, indicated these on Figure 3 and marked the points on Figure 2 according to the extrapolated size classes. During the 45 days of the experiment, the maxillae retained for the longest period were the larval maxillae. The worms consisted of 14-15 chaetigers (28-35 days old) at the moult to the juvenile jaws. This timing is identical with the corresponding moult in the brooding A-/, jonesi and K. simoni. Hsieh and Simon (1987) linked this event with leaving the maternal tube to construct tubes on their own and the commencement of feeding. Our D. aciculata, lecithotrophic only fora brief free-swimming larval stage, settles, starts to construct its lube and to feed at the age of 4-6 days (5-chaetigers) as evidenced by fecal pellets in the gut. This indicates that the timing of the first moult, i.e. larval to juvenile maxillae, is not governed by ecological but developmental clues. Based on these estimations (Figs 2,3), our largest wonns (42^14 chaetigers) had their fourth set of adult maxillae, having undergone a moulting process every 3-4 days since attaining the juvenile maxillae. As the growth rate slows down, the frequency of moulting presumably slows down. Specimens from a different study (Safarik et al. 2006) that were raised for seven months and consisted of 150-186 chaetigers, had maxillae I of 1.6-1.9 mm length, indicating that these were their tenth adult maxillae, leading to the conclusion that the rate of moulting slows down dramatically with time, unless it occurs with less or no size increase in later life. ACKNOWLEDGMENTS We are grateful to Les Safarik for allowing us to use the facilities of Aquabait Pty. Ltd. for this study. We also thank Gunter Purschke and Mats Eriksson for review of the manuscript and valuable suggestions. 20 Jaw growth and replacement in Diopatra aciculata REFERENCES Allen, M.J. 1959. Embryological development of the polychaetous annelid, Diopatra cuprea (Bose). The Biological Bulletin of the Marine Biological Laboratory Woods Hole 116: 339-361. Averincev, V.G. 1972. Benthic polychaetes Errantia from the Antarctic and Subantarctic collected by the Soviet Antarctic Expeditions. (In Russian). Explorations ofthe fauna ofthe seas XI (XIX). Biologial Results of the Soviet Antarctic Expeditions 5: 88-293. Bergman, C.F. 1989. Silurian paulinitid polychaetes from Gotland. Fossils and Strata 25: 1-128. Blake, J.A. 1975. The larval development of Polychaeta from the northern California coast. II. Nothria elegans (Family Onuphidae). Ophelia 13: 43-61. Carrasco, F.D. 1983. Description of adults and larvae of a new deep water species of Hyalinoecia (Polychaeta, Onuphidae) from the southeastern Pacific Ocean. Journal of Natural Histoiy 17: 87-93. Fauchald. K. 1982. Description of Mooreonuphisjonesi , a new species of onuphid polychaete from shallow water in Bermuda, with comments on variability and population ecology. Proceedings of the Biological Society of Washington 95: 807-825. Gullan, P.J. and Cranston, P.S. 1994. The insects: an outline of entomology. Chapman and Hall: London. Haffner, K. von 1959. Ober den Bau und den Zusammenhang der wichtigsten Organe des Kopfendes von Hyalinoecia tubicola Malmgren (Polychaeta, Eunicidae, Onuphidinae), mit Bcriicksichtigung der Gattung Eunice. Zoologische Jahrbiicher, Abteilungfiir Anatomic 77: 133-192. Hinton, 11.E. and Mackerras, M.J. 1970. Reproduction and metamorphosis. Pp 83-106. In: CSIRO (eds) The insects of Australia. Melbourne University Press: Carlton. Hsieh, H.-L. and Simon, J.L. 1987. Larval development of Kinbergonuphis simoni, with a summary of development patterns in the family Onuphidae (Polychaeta). Proceedings of the Biological Society of Washington 7: 194-210. Kielan-Jaworowska, Z. 1966. Polychaete jaw apparatuses from the Ordovician and Silurian of Poland and a comparison with modem forms. Palaeontologia Polonica 16: 1-152. Knox, G.A. and Cameron, D.B. 1971. Port Philip Surveyl957-1963, Victoria, Australia. Part 2, No. 4. Polychaeta. Memoirs of the National Museum Victoria 32: 21^12. Krohn, A. and Schneider, A. 1867. Ucbcr Annelidlarven mit porosen \MMen. Aivhivfiir Anatomic. Physiologic und wissenschaftliche Medicin 1867: 498-508. Monro. C.C.A. 1924. On the post-larval stages of Diopatra cuprea, Bose, a polychaetous annelid of the family Eunicidae. Annals and Magazine of Natural Histoiy (London) (9) 14: 193-199. Orensanz, J.M. 1990. The Eunicemorph polychaete annelids from Antarctic and Subantarctic seas. Biology of the Antarctic seas XXL Antarctic Research Series 52: 1-183. Paxton. H. 1993. Diopatra Audouin and Milne Edwards (Polychaeta: Onuphidae) from Australia, with a discussion of developmental patterns in the genus. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 10: 115-154. Paxton, H. 2005. Molting polychaete jaws - eedysozoans arc not the only molting animals. Evolution & Development 7: 337-340. Paxton, H. 2006. Replacement of adult maxillary jaws in Eunicida (Polychaeta). Scientia Marina 70S3: 331-336. Paxton, H. In press. Phylogcny of Eunicida (Annelida) based on morphology of jaws. Proceedings of the 9th International Polychaete Conference. Safarik, M., Redden, A.M. and Schreider, M.J. 2006. Density- dependent growth of the polychaete Diopatra aciculata. Scientia Marina 70S3: 337-341. Accepted 26 August 2008 21 * % The Beagle, Records of the Museums and Art Galleries of the Northern Territory, 2008 24: 23-31 Feeding selectivity of sesarmid crabs from northern Australian mangrove forests CHANDRA P. SALGADO KENT* AND KEITH A. McGUINNESS School of Science and Primary Industries, Faculty of Education, Health & Science, Charles Darwin University, Darwin, NT0909, AUSTRALIA * Corresponding author: Centre for Marine Science and Technology’, Department of Applied Physics, Curtin University of Technology, Bentley, WA 6102, AUSTRALIA c.saIgado@cmst.curtin.edu.au ABSTRACT Sesarmid crabs are possibly one of the most important components of mangrove fauna because of their influence on nutrient cycling and forest structure by feeding on litterfall. Little is known about the influence of clcctivity on the role of crabs in mangrove forests, and how this is affected by the availability of litter items. This study investigated electivity of three northern Australian sesarmid crabs ( Perisesarma semperi, Perisesanna darwinensis and Neosannatium meinerti) from leaves of various conditions and from common species, as well as the effect of the availability of propagules on feeding electivity. In almost every experiment, decayed and senescent leaves were selected over fresh leaves, and typically decayed leaves were selected over senescent. Electivity for mangrove species, however, varied among crab species and depended upon availability of litter type to select from. In experiments that included propagules as well as leaves, leaves were selected over propagules. These results suggest that the sesarmid crab species included in this study may have a greater role in nutrient cycling than in forest structuring because of their selectivity of leaves over propagules. Keywords: sesarmid, Sesarmidae, feeding behaviour, selectivity, mangrove leaves, consumption. INTRODUCTION Sesarmid crabs are one of the most common and abundant faunal groups in mangrove forests (Golley et al. 1962; Jones 1984; Smith et al. 1991). Recent research indicates that they play important roles in the ecology of these ecosystems (Lee 1998) and may occupy a keystone position in Australian mangrove forests (Smith et al. 1991). For example, crabs may affect forest structure by attacking mangrove propagules (Smith 1987; McGuinness 1997), influence nutrient cycling by feeding on litterfall (Robertson 1986), alter the properties of the soil by their burrowing activities (Smith et al. 1991) and be involved in competitive interactions with other species (Fratini et al. 2000). Studies on the feeding ecology of sesarmid crabs have contributed to our understanding of the fate of mangrove litter nutrients (Camilleri 1984, 1989). These studies have shown that sesarmid crabs have an important role in retaining nutrients within mangrove forests and reducing export to nearby coastal systems (Lee 1997, 1998). Crabs process a variety of food items, mainly dead leaves, into smaller particles, and in this way make these nutrients more readily available for other fauna to consume (e.g. gastropods, crabs and other crustaceans). If sesarmid crabs display selectivity for particular food items, such as leaves and propagules from certain species of mangroves, this is likely to affect the quantity, type and nutritional value of mangrove litter that is recycled. The feeding behaviour of some sesarmid crabs, in particular those common to mangrove forests in north¬ eastern Australia and Kenya, has been studied (Camilleri 1989; Micheli 1993; Dahdouh-Guebas et al. 1997; Kathiresan and Bingham 2001; Cannicci et al. 2007), but the range of species and locations studied, is still limited. Furthermore, previous studies did not test the effect of the availability of litter on feeding electivity. The availability of material may depend on factors such as mangrove assemblage and season. In mangrove forests in tropical Australia, for instance, most propagules drop during the wet season (Ball and Pidsley 1988) and are, therefore, only available at this time. This study investigates feeding electivity of three species of sesarmid crabs from common mangrove species occurring in tropical mangrove forests in northern Australia, and how electivity is affected by the seasonal availability of propagules. The specific aim is to examine electivity of mangrove leaves from three different species and three conditions during wet and dry season conditions (with and without the presence of propagules). Finally, other variables which might affect consumption - crab size (see Emmerson and McGwynne 1992) and sex (see Olafsson et al. 2002) - were also investigated. C. P. Salgado Kent and K. A. MeGuinness MATERIALS AND METHODS Crabs. The three most abundant sesarmid crabs in the dominant mangrove assemblages in Darwin Harbour (Salgado Kent 2004) were studied: Perisesarma semperi, P. darwinensis, and Neosarmatium meinerti. Neosarmatium meinerti is most abundant in mid to upper shore mixed woodland and hinterland assemblages (Salgado Kent 2004) dominated by Ceriops australis and Avicennia marina (naming of assemblages follows Brocklehurst and Edmeades 1996). Perisesarma darwinensis is common in tidal flat assemblages dominated by C. australis, but P. semperi is found in tidal bank assemblages dominated by A. marina and Rhizophora stylosa (Salgado Kent 2004; pers. obs.). All crabs were collected from the forest at Jones Creek, Darwin Harbour. Twenty individuals - ten of each sex - of the two Perisesarma species were collected by hand: half the crabs were larger, and half smaller, than the average size (1.3 cm in carapace width). Average size did not differ between species and was estimated from 106 crabs that were collected and measured prior to these experiments. Neosarmatium meinerti crabs were captured in tunneled pitfall traps (similar to the pitfall traps used by Warren (1987)). Results for A. meinerti did not include analyses on sex and size because only six individuals could be captured, only one of which was female. All crabs were placed in separate containers as soon as possible upon return to the laboratory, to reduce stress and injury, in particular among aggressive competing males. Leaf preparation. The leaves and propagules used in the experiments were taken from the three dominant mangrove species in the four assemblages inhabited by the crabs; A. marina, C. australis and R. stylosa. Leaf conditions included fresh (green leaves), senescent (yellow leaves) and decayed (brown leaves). Fresh and senescent leaves were collected directly from trees. Fresh leaves can frequently occur on the forest floor when storms, which are common in the wet season, knock them down. Decayed leaves were prepared by collecting senescent leaves from trees and leaving these to decompose for fifteen days, enclosed in 2 mm mesh bags tied to mangrove roots (as in Robertson (1988)). All leaves w'ere stored at 4°C for 1 to 1.5 days, until the experiments began. Circular sections of leaf, 2 cm in diameter, were used in experiments here to reduce possible influences of leaf size on electivity (as in Camilleri 1989). Propagules were collected from trees and were also cut into similar sized pieces (to each other and to the leaves). Experiment preparation. All experiments were conducted in a shaded, outdoor laboratory in which crabs experienced a regular diurnal cycle, and conditions similar to those prevailing in the field at that season. Crabs were placed individually into clear, plastic containers, 14 cm in diameter and 10 cm high. A circle 6 cm in diameter in the centre of each container’s lid was cut out and this allowed air to enter but prevented the crabs from escaping. Containers without crabs were included in experiments to control for weight changes due to leaching of dissolved organic matter (DOM) and fungal and microbial activity. Seawater was added to all containers to a level of 0.5 cm and was changed daily (to ensure that lack of moisture was not a factor affecting results, since the crabs inhabit waterlogged environments). Most habitats where these crabs were prevalent were saturated with water (with the exception of N. meinerti). Crabs were starved for 24 hours before each experiment. During the experiments, Perisesarma crabs were offered a choice of mangrove material for a period of 12 hours. In each experiment, the amount offered was one disc or piece for each type x species (and each species x condition, for leaves) to be included for testing. The amount of material offered was great enough to allow for detection of significant patterns in electivity, but small enough so that sufficient quantities of material from discs and/or pieces remained to retain information on patterns in electivity. For this reason N. meinerti crabs were offered mangrove material for a period of 18 hours, as these crabs took longer to consume a significant amount of material. The material offered to crabs was randomly placed within the containers. Each batch of crabs was used in no more than two experiments to ensure that their condition did not deteriorate significantly and affect the results of the experiments. Experiments. Two experiments were conducted (Table 1). The first included leaves from all three mangrove species and three leaf conditions, and the second included the addition of propagules from all three mangrove species. In mangrove forests of Darwin Harbour, Northern Territory, the dominance of different mangroves species differs among habitats where N. meinerti, P. semperi and P. darwinensis are most abundant. However, all three mangrove species overlap in distribution with the three crab species to various extents (Brocklehurst and Edmeades 1996). Hence, in the natural environment, the chance of encounter of leaves and propagules of these mangrove species by the three species is realistic, and in most cases relatively high. To test for seasonal differences, experiments were done under wet and dry season conditions (outdoor laboratory experiments ensured that crabs experienced seasonal changes in humidity and temperature). Dry season experiments used only leaves but wet season experiments used both propagules and leaves. A pilot study found no significant effect of propagule dimension on feeding electivity (Salgado Kent 2004). Each experiment described below simultaneously examined feeding electivity of the different species of crabs. ‘Dry Season’ experiments were done simultaneously (and in the dry season), and ‘Wet Season’ experiments were done simultaneously (in the wet season). 24 Feeding selectivity of sesarmid crabs Table 1. Material included in experiments testing electivity of sesarmid crabs for mangrove leaves and propagules. The mangrove species were Am = Avicennia marina ; Ca = Ceriops australis; Rs = Rhizophora stylosa. The sesarmid crabs were Ps = Perisesarma semperi ; Sd = Perisesarma darwinensis; Nm = Neosarmatium meinerti. Each experiment included six N. meinerti but twenty of each of the other species (half male and half female; half small and half large). See the text for further details Experiment 1: dry season Electivity on common species Experiment 2: wet season Crab species Ps Pd Nm Ps Pd Nm Am Am Am Am Am Am Ca Ca Ca Ca Ca Ca Leaf species Rs Rs Rs Rs Rs Rs Fresh Fresh Fresh Fresh Fresh Fresh Senesc Senesc. Senesc. Senesc. Senesc. Senesc. Leaf condition Decay Decay Decay Decay Decay Decay Am Am Am Ca Ca Ca Propagule species Rs Rs Rs Experiment I. Electivity of leaves from three mangrove species during dry season conditions. All crab species were offered a choice of nine types of leaves, comprising all combinations of the three species (C. australis, A. marina and R. stylosa ) and three conditions (fresh, senescent, and decayed). Experiment 2. Electivity of leaves and propagules from three mangrove species during wet season conditions. Experiment 1 was repeated but with pieces of the propagules of all three mangrove species also offered. Processing measurements and calculations. Consumption was measured in two ways, by weight change and by area removed, and the results of these two methods compared. The material offered had to be weighed wet but the material remaining could be weighed wet or dry. As dry weights were likely to be less variable, the material remaining was dried at 60°C for two days and then weighed. To calculate loss, the initial wet weights were converted to dry weights. This was done by taking wet and dry weights of twenty samples ol the leaves and propagules of each species. These samples were then dried at 60°C for two days and remeasured. Regression equations predicting dry weight from wet were then derived and used (all R2 > 0.8). After the experiments, the area processed was calculated by overlaying a clear plastic grid on top of each litter item, counting the total number of 5 x 5 mm squares that each leaf disc originally filled and the number of squares consumed. The percent of area processed was converted into dry weight processed by multiplying percent area by the converted initial leaf dry weights. The weight loss in control treatments (treatments with no crabs) was subtracted from the weight loss of leaves and propagules in treatments with crabs, to correct for weight loss of propagules and leaves due to leaching of DOM and fungal and microbial activity. Statistical analyses. As the crabs were presented with an array of choices simultaneously in each experiment, the amounts of the different items consumed may not have been independent (e.g. greater consumption of one item is likely to result in reduced consumption of others). Because of this, repeated measures analysis of variance (ANOVA) was used for analysing all data in this study. Assumptions for ANOVA were tested with Cochran’s homogeneity of variances test and data were transformed when appropriate. Mauchley’s Sphericity test was used to check the assumptions required for repeated measures ANOVA and, when these could not be met, the Greenhouse Geisser correction was applied (Winer et al. 1991). Several analyses were done: each included as many factors as possible (chosen to test hypotheses a priori) so as to limit the total number required. Data for P. semperi and P. danvinensis in Experiments I and 2 could be analysed together as the same numbers, sizes and sexes of crabs were used and all were offered similar choices. Analyses including crab size were based on absolute mean consumption rates for the two groups (and was not weight specific). Data for N. meinerti in Experiments 1 and 2 had to be analysed separately as the choices offered to the crabs differed. For all experiments, two sets of analyses were done. One set compared the seasons but used only the data for leaves (as propagules were not offered in the dry season). The second set considered only the wet season but included the data for propagules. Factors in analyses varied, depending upon the design, but included season (wet, dry), crab species ( P. semperi, P. darwinensis), crab sex (male, female), size class of crab (small, large), species of material (A. marina, R. stylosa, C. australis) and type/condition of material (fresh, senescent or decayed leaf; propagule). Some of the resulting analyses were complex (i.e. Table 3). Inspection of every higher order interaction in such analyses is likely to be tedious and potentially unrewarding. Following Mead (1988), our interpretation of such analyses focused on sources of variation which were significant and which, judged by the magnitude of the relevant mean square, accounted for a substantial proportion of the variance. 25 C. P. Salgado Kent and K. A. McGuinness 0.15 r E2 0 13 Propagule r 1 1 Fresh 0.04 - zz/> Senescent ■1 Decayed 0.02 0.00 A. marina stylosa C. australis Fig. 1. Consumption of material (g dry weight per 12 hours) from three mangrove species -A. marina, R. stylosa and C. australis - by N. meinerti in electivity experiments (mean + SE). Note change in y-axis scale on graph for Experiment 2 (E2). For further details see Table 1. RESULTS Comparison of analyses based on leaf weight processed and leaf area processed. Overall, results from analyses based on weight and area were usually very similar, although analyses of area consumed gave more significant effects in the statistical analyses. This was probably due to greater variability in estimates based on weight (than on area). With measurements based on weight, the initial observations, as they were of wet weights, were likely to be more variable and the conversion of these to dry weight, using the regression equations, probably introduced additional errors. As there were relatively few such differences, and they did not affect the overall interpretation of the experiments, results in the rest of the study are presented here from weights estimated from area consumed. Neosarmatium meinerti. Processing of leaves, in general, depended upon the season, species and type/condition of material when crabs were offered (see Table 2 for specific factors affecting processing in each experiment; Figure 1) . The amount of decayed leaves processed was nearly always greater than that of senescent or fresh leaves, with the latter usually least preferred. Crabs processed more R. stylosa than A. marina or C. australis, particularly in the wet season when processing of some items increased markedly (Figure 1: El v E2). When propagules were offered with leaves (Experiments 2) , processing still depended upon the species and type/ condition of material (Table 2). Of the propagules, only A. marina were processed and only in moderate amounts; senescent and decayed leaves were processed in greater quantities. Perisesarma semperi. Patterns in processing of leaves by P. semperi were complex, with numerous interactions between the different factors in the analysis (Table 3). Judged by the magnitude of the mean squares, the major effects were of the species and type/condition of material, and their interaction, and the interaction between crab size and season. The overall pattern of results was very similar to that seen with N. meinerti - greater processing of decayed material than the other types, and of R. stylosa than of the other species-although there was no increased processing in the wet season (Fig. 2: El v E2). An interaction between season and crab size (Table 3) occurred because large crabs processed more leaf material in the dry season than small crabs, but there was no difference between sizes in the wet (means = 0.015 g/12 hours for large crabs in the dry season, and 0.007 g/12 hours for small; 0.019 g/12 hours for large crabs in the wet season, and 0.020 g/12 hours for small). Interactions between size, season and other factors (Table 3), however, indicate that the strength of this pattern depends on the crab and mangrove species, and the condition of the material. Interactions with crab sex were also significant, although of less importance (Table 3). When propagules were offered, the results were again similar to those for N. meinerti : a moderate amount of A. marina propagules was processed while other species Table 2. Results of repeated measures analyses of variance of amount of material processed (weights based on areas) by Neosarmatium meinerti in all experiments using all common species (El, E2). The table gives the df, MS and significance (* = P < 0.05; *** = P < 0.001) for two analyses: the first used data for leaves only but included two seasons (wet and dry); the second used data for leaves and propagules but only for the wet season. Only the main effects and significant interactions are shown. Effects with fractional df have the Geisser-Greenhousc correction applied. See Table 1 for further information on the design of the experiments. Electivity for common species Leaves only: Leaves, El, E2 propagules: E2 Factor df MS df MS Season 1,9 0.032 Species (Sp) of material 1.25, 18 0.076 2, 10 0.730* Type/Condition of material 2, 18 0.075* 3, 15 0.901*** Season x Sp 2, 18 0.013 Season x Type 2, 18 0.102* Sp x Type 4, 36 0.025 6, 30 0.230 Season x Sp x Type 4, 36 0.066* 26 Feeding selectivity of sesarmid crabs 0.08 El 0.08 E2 A. marina R. stylosa C. australis Fig. 2. Consumption of material (g dry weight per 12 hours) from three P. semperi in electivity experiments (mean + SE). 0.06 0.04 0.02 0.00 gggg Propagule l l Fresh A. marina R. stylosa C. australis mangrove species - A. marina, R. stylosa and C. australis - by were ignored (Fig. 2: E2). In Experiment 2, when materials including propagules were offered, there was a substantial effect of crab size (Table 3), but the interactions indicated that the pattern was only observed for P. danvinensis (see below). Perisesarma danvinensis. Results for the consumption of commonly available species by P. danvinensis were broadly similar to those of the other two crab species (Table 3; Fig. 3) - the major effects were of the species and type/ condition of material and their interaction - but there were differences in detail. While P. danvinensis still tended to consume greater amounts of decayed leaves, greater amounts of senescent and even fresh, leaves were taken by this species (Fig. 3). In fact, in the dry season, roughly equal amounts of all three types of C. australis leaves were taken; and there was similar consumption of R. stydosa decayed, and A. marina senescent and decayed, leaves (Fig. 3: El). As with N. meinerti, there was a markedly increased consumption of R. stylosa material in the wet season (Fig. 3: E2). In contrast to the dry season, fresh and senescent C. australis leaves were not taken at this time. In Experiment 2, as noted above, when commonly available materials, including propagules, were offered there was a substantial effect of crab size (Table 3), Table 3. Results of repeated measures analyses of variance of amount of material processed (weights based on areas) by Perisesarma semperi and P. danvinensis in Experiments 1 and 2. Sec Table 1 for the design of the experiments and Table 2 for the format of the table. Source Leaves only: El, E2 Leaves, propagules: E2 df MS df MS Season 1, 60 0.018 Crab species 1,60 0.055 * 1,28 0.198 * Sex of crab 1,60 0.031 1,28 0.036 Size class of crab 1,60 0.009 1, 28 0.578 ** Species (Sp) of material 1.31, 120 0.337 *** 1.6, 56 1.416 *** Type/Condition of material 1.46, 120 0.826 *** 2.25, 84 1.871 ** Season x Size 1, 60 0 152 *** Sex x Sp 1.31, 120 0.032 * 1.6, 56 0.055 Sex x Type 1.46, 120 0.049 ** 2.25, 84 0.049 Sp x Type 4, 240 0.140** 6, 168 1.004 *** Season x Size x Sp 1.31, 120 0.054 ** Crab x Sex x Type 1.46, 120 0.036 * 2.25, 84 0.148 * Season x Size x Type 1.46, 120 0.072 ** Crab x Size x Type 1.46, 120 0.005 2.25, 84 0.172 * Sex x Size x Type 1.46, 120 0.024 2.25, 84 0.192* Crab x Sp x Type 4, 240 0.045 *** 6, 168 0.071 Sex x Sp x Type 4, 240 0.023 * 6, 168 0.025 Season x Crab x Size x Type 1.46, 120 0.031 * Season x Sex x Size x Type 1.46, 120 0.021 * Crab x Sex x Sp x Type 4, 240 0.047 *** 6, 168 0.063 Season x Size x Sp x Type 4, 240 0.053 *** Sex x Size x Sp x Type 4, 240 0.002 6, 168 0.166 ** Season x Crab x Sex x Size x Sp 1.31, 120 0.046 ** Season x Crab x Sex x Size x Type 1.46, 120 0.038 * Season x Sex x Size x Sp x Type 4, 240 0.035 ** Season x Crab x Sex x Size x Sp x Type 4, 240 0.043 ** 27 C. P. Salgado Kent and K. A. McGuinness 0.08 r 3 0.06 - 0.04 0.02 0.00 El T I Jt lyJ A. marina R. stylosa C. australis 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 E2 Propagule Fresh ZZa Senescent Decayed ik i A. marina R. stylosa C. australis Fig. 3. Consumption of material (g dry weight per 12 hours) from three mangrove species - A. marina , R. stylosa and C. australis - by P. darwinensis in electivity experiments (mean + SE). Note change in y-axis scale on graph for Experiment 2 (E2). For further details see Table 1. with large crabs processing more material than smaller crabs (means = 0.013 g/12 hours, and 0.002 g/12 hours, respectively). DISCUSSION A common trend in the present study was selectivity for older material. When offered common species, N.meinerti and P. semperi consumed more decayed R. stylosa leaves than any of the other options (Table 4). Perisesarma darwinensis displayed a similar pattern in the wet season, although not in the dry. The second and third most consumed materials were also usually decayed or senescent leaves, although P. darwinensis electivity again differed in the dry season. Other studies have documented similar electivity for older material. In northeastern Australia, Camilleri (1989) found Sesanna erythodactyla preferred aged leaves to freshly fallen leaves and Micheli (1993) found Sesanna messa preferred decayed leaves over senescent leaves. In Kenya, Neosannatiwn smithii (= Sesanna smithii) preferred old or decaying leaves over young leaves (Micheli 1991; Table 5). Ashton (2002), however, found two Malaysian species, Sesanna eumolpe and Sesanna onychophorum, preferred fresh to senescent Avicennia officinalis leaves. Studies have shown that decaying leaves have a lower concentration of tannins than fresh leaves, so may be more Table4. Summary ofelcctivity exhibited by the three speciesofsesarmids. Am = A. marina ; Ca = C. australis; Rs = R. stylosa; f = fresh leaf; s = senescent leaf; d = decayed leaf; p = propagule. 1st 2nd 3rd Species choice choice choice Dry Season N. meinerti d-Rs d-Am d-Ca P. semperi d-Rs d-Am d-Ca, s-Ca P. darwinensis d-Am f-Ca several Wet Season N. meinerti d-Rs s-Rs d-Ca, p-Am P. semperi d-Rs s-Rs, p-Am P. darwinensis d-Rs s-Rs d-Ca easily digested and preferred for this reason (Giddins et al. 1986; Neilson et al. 1986). A study by Micheli (1993), testing for the effects of tannin, found no significant correlation on the feeding preferences of N. smithii and S. messa. These tests were, however, done with senescent, rather than decaying, leaves (including R. stylosa, C. australis and A. marina) and Robertson (1988) found that the tannin content of these leaves decreased rapidly over the first 14 days of decomposition. Thus, the range of tannin concentrations in senescent leaves might not be large enough to affect electivity. Studies of sesarmid electivity for material from different mangrove species, in contrast to material of different ages, have given more variable results. Micheli (1993) found that N. smithii preferred R. stylosa to A. marina , Bruguiera exaristata and C. australis; results similar to P. semperi here. Camilleri (1989) found S. erythrodactyla selected R. stylosa least, after A. marina and B. exaristata. Ashton (2002), with S. eumolpe and S. onychophorum , found a preference for A. officinalis but only in fresh leaves. With S. messa , however, Micheli (1993) found no significant preference among the species tested and Dahdouh-Guebas et al. (1997), studying N. meinerti in Kenya, also found no preference, although only fresh material was offered. Olafsson et al. (2002), however, also tested N. meinerti from Kenya and obtained results similar to Camilleri (1989) for 5. erythrodactyla. Greater electivity of A. marina , as also exhibited on occasion by P. darwinensis in the present study, can be explained by the particularly low tannin and high nitrogen levels characteristic of this species (Robertson 1988; Camilleri 1989; Michelli 1993). Leaf nitrogen, in particular, is usually a reliable predictor of herbivore preference in both laboratory and field situations (Perez-Harguindeguy et al. 2003). Mature leaves from A. marina in Darwin Harbour have been shown to have significantly higher nitrogen concentrations than R. stylosa and C. australis (Coupland 2002). Rhizophora stylosa leaves, in contrast, have lower nitrogen concentrations and higher percentage of tannin, than A. marina (Robertson 1988; Coupland 2002) and electivity for this species is more difficult to explain. 28 Feeding selectivity of sesarmid crabs Table 5. Summary of elcctivity/preferences for leaves of different species and conditions exhibited by sesarmid crabs in other studies. Am = A. marina ; Ca = C. australis ; Ct = C. tagal; Bg Bruguiera gymnorltiza', Be = B. exaristata ; Rm = R. mucronata', Rs = R. stylosa ; Sa = Sonneratia alba', f = fresh leaf; s = senescent leaf; d = decayed leaf; Nm = N. meinerti', Ns = N. smithi; Sm = 5. messa ; Se = S. erythrodactyla; Cc = C. carnifex. Note: all species/conditions that were used in the trials are included below. *No pattern in electivity/preference. Study Field/lab Species 1st choice 2nd choice 3rd choice 4th choice 5th choice Steinke et al. 1993 Field Nm s-Bg s-Am f-Bg f-Am Giddins et al. 1986 Lab Ns d-Ct s-Ct f-Ct Micheli 1993 Lab Sm s-Am, s-Rs, s-Ct, s-Be* Lab Nm s-Rs s-Am, s-Ct, s-Be Lab Sm d-Am, d-Rs, d-Ct, d-Be s-Am, s-Rs, s-Ct, s-Be Field Nm/ Sm s-Ct s-Rs, s-Re, s-Am Camilleri 1989 Lab Se d-Am d-Bg d-Rs s-Am s-Bg; s-Rs Lab Se f-Am, d-Am s-Am Lab Se d-Rs s-Rs Micheli etal. 1991 Lab Cc s-Bg s-Sa s-Rm s-Ct s-Am Lab Nm s-Bg. s-Sa, s-Rm, s-Ct, s-Am* Camilleri (1989) found that crabs preferred thicker R. stylosa leaves, so other leaf attributes may be influential (such as moisture, fibre content and other chemical constituents). Kennish and Williams (1997), in a study of the tropical rocky shore crab Grapsus albolineatus, concluded that algal morphology, through effects on feeding efficiency, was more important than nutritional value or digestibility. And Chavanich and Harris (2002) suggested that several factors, including morphology and nutritional value, probably influenced the feeding preferences of the subtidal gastropod Lacuna vincta. Previous experience by the crabs may also be important (Perez-Harguindeguy el al. 2003) and explain some of the varying results for A. meinerti, which has shown no particular pattern in electivity (Micheli et al. 1991; Dahdouh-Guebas et at. 1997) and electivity for (Olafsson et al. 2002) and against (Steinke et al. 1993; this study) A. marina. It does not, however, appear to hold true for crabs in the present study, since they did not usually, when offered a range of common species, select species from the assemblage in which they were most abundant (in contrast to the results of Ashton (2002)). Crab electivity within a restricted range of material, representative of probable encounter rates, may differ from electivity when all mangrove leaves and conditions are offered in equal proportions. In a recent study on the gypsy moth ( Lymatria dispar) for example, Raffa et al. (2002) found that results could be affected by the combination and arrangement of choices, and also by total consumption. Perez-Harguindeguy et al. (2003) concluded that laboratory preference experiments can predict relationships in the field, but ecological factors such as variations in accessibility and specialised plant-herbivore relationships can cause differences. However, according to the model they developed, predictions are likely to be reliable for generalist herbivores and plants of high accessibility, both conditions which apply here to various extents. Together, these points suggest that strong, general trends, such as the electivity for older material, are likely to apply in the field. In contrast, more precise distinctions between species and ages of material may be situation-specific. Further study of weight-specific consumption rates, and experiments offering different proportions of materials would shed light on the different patterns observed here. Given the importance of sesarmid crabs as propagule predators (Smith 1987; McGuinness 1997; Lee 1998), the limited consumption of propagules in the present study, particularly by N. meinerti , is surprising. The Perisesarma species may be too small to deal effectively with propagules but N. meinerti is known to consume them in the field (McGuinness 1997; Dahdouh-Guebas et al. 1997). On the basis of these results, however, leaf material is preferred, when it is available. Further, studies have found from 75% (Steinke etal. 1993) to 90% (Dahdouh-Guebas et al. 1997) of the material in N. meinerti stomachs to be leaf material, although Skov and Hartnoll (2002) reported only 10% of crabs feeding on leaves compared to 76% feeding on mud. In part, the difference may result from the differing digestibility of material: this would be consistent with the results of Bouillon et al. (2002), whose stable isotope studies indicated that sesarmid crabs fed on a wider range of material than just mangrove leaves. Effects of size and sex of crab were inconsistent, although in accordance with previous results. Emmerson and McGwynne (1992) reported a correlation between size and consumption for N. meinerti in southern Africa. Furthermore, Olafsson et al. (2002) found greater consumption by female crabs, and they suggested this could be due to either the difference in size between the sexes, with females being smaller and having a higher potential for energy loss, or to reproductive demands. The latter explanation is perhaps more likely as size was controlled in experiments here. Further, Micheli (1993) observed more ovigerous crabs during the late dry season and this might partially explain greater consumption at this time. In conclusion, this study confirms the general pattern in electivity of sesarmids for older material observed in 29 C. P. Salgado Kent and K. A. McGuinness most other studies but also demonstrates that the range, and types, of material offered can affect electivity. The amount consumed depends upon the size of the crab, and is likely to depend upon its sex and reproductive state. Furthermore, the results suggest that sesarntid crabs included in this study have a greater role in nutrient cycling than in forest structuring because of their selectivity of leaves over propagules. Hence, sesarmids in northern Australia appear to have a distinct overall ecological role than in other regions such as shown in some studies in new world mangrove forests (Smith et al. 1991). Future studies should attempt to confirm the findings here by; investigating weight specific consumption rates, determining whether ‘feeding preferences’ differ from ‘feeding electivity’ (Underwood et al. 2004), and ultimately determining the reasons underlying the high variability in sesarmid feeding selectivity for leaves of different mangrove species, observed in the majority of studies conducted thus far. ACKNOWLEDGMENTS This study was supported by a Large ARC grant (to K. A. McGuinness). We thank other local mangrove researchers for support and assistance. J. Warren provided helpful advice on the design and use of pitfall traps. We offer particular thanks to P. Davie for his valuable assistance in the identification of Northern Territory sesarmid crabs. Finally, the quality of the manuscript presented here has been markedly improved by the input and insight of the many reviewers, including J. Martin and K. Metcalfe. The experiments presented in this study comply with the current laws of the country in which the experiments were performed. REFERENCES Ashton, E.C. 2002. Mangrove sesarmid crab feeding experiments in Peninsular Malaysia. Journal of Experimental Marine Biology and Ecology 273: 97-119. Ball, M. and Pidsley, S. 1988. Establishment of mangrove seedlings in relation to salinity. Pp. 123-134. In: Larson, H„ Michie, M., and Hanley, J. (eds) Proceedings of a Workshop on Research and Management held in Darwin. Australian National University Press: Canberra. Brocklehurst, P. and Edmcadcs, B. 1996. The mangrove communities of Darwin Harbour. Resource capability assessment branch technical memorandum No 96/9. Department of Lands, Planning and Environment: Darwin. Bouillon, S., Dahdouh-Guebas, F., Rao, A., Koedam, N. and Dehairs, F. 2003. Sources of organic carbon in mangrove sediments: variability and possible ecological implications. Hydrobiologia 495: 33-39. Bouillon, S., Koedam, N., Raman, A.V. and Dehairs, F. 2002. Primary producers sustaining macro-invertebrate communities in intertidal mangrove forests. Oecologia 130: 441-448. Camilleri, J. 1984. Litter decomposition and resource partitioning in a mangrove forest: a case study in systems ecology and modelling. Unpublished PhD Thesis. Griffith University: Queensland, Australia. Camilleri, J. 1989. Leaf choice by crustaceans in a mangrove forest in Queensland. Marine Biology 102: 453-459. Cannicci, S., Gomei. M., Dahdouh-Guebas, F.. Rorandelli, R. and Terlizzi, A. 2007. Influence of seasonal food abundance and quality on the feeding habits of an opportunistic feeder, the intertidal crab Pachygrapsus marmoratus. Marine Biology’ 151: 1331-1342. Chavanich, S.. and Harris, L.G. 2002. The influence of macroalgae on seasonal abundance and feeding preference of a subtidal snail. Lacuna vincta (Montagu) (Littorinidae) in the Gulf of Maine. Journal of Molluscan Studies 68: 73-78. Clough. B. 1992. Primary productivity and growth of mangrove forests. Pp 225-251. In: Robertson, A.I. and Alongi, D.M. (eds) Tropical mangrove ecosystems. American Geophysical Union: Washington. Conde, J.E. and Diaz, H. 1989. The mangrove tree crab A rants pisonii in a tropical estuarine coastal lagoon. Estuarine and Coastal Shelf Science 28: 639-650. Coupland, G. 2002. The ecological interaction between insects and mangroves in Darwin Harbour, Australia. PhD Thesis. Northern Territory University: Darwin. Dahdouh-Guebas, F., Vcmcirt, M.,Tack, J.F., Van Speybroeck, D. and Koedam. N. 1997. Food preferences in Neosarmatiwn meinerti de Man (Decapoda: Sesarminae) and its possible effect on the regeneration of mangroves. Hydmbiologia 347: 83-89. Emmerson, W. and McGwynne, L. 1992. Feeding and assimilation of mangrove leaves by the crab Sesarma meinerti de Man in relation to leaf-litter production in Mgazana, a warm temperate southern African mangrove swamp. Journal of Experimental Marine Biology’ and Ecology 157: 41-53. Fratini, S., Cannicci, S. and Vannini, M. 2000. Competition and interaction between Neosarmatium smithi (Crustacea : Grapsidae) and Terebralia palustris (Mollusca : Gastropoda) in a Kenyan mangrove. Marine Biology 137(2): 309-316. Giddins. R.I., Lucas, M.J., Neilson, M.J. and Richards, G.N. 1986. Feeding ecology of the mangrove crab Neosarmatium smithii (Crustacea: Decapoda: Sesarmidae). Marine Ecology Progress Series 33: 147-155. Golley, F., Odum, H.T. and Wilson, R.F. 1962. The structure and metabolism of a Puerto Rican red mangrove forest in May. Ecology’ 43: 9-19. Jones, D.A. 1984. Crabs of the mangal ecosystem. In: Por, F.D. and Dor. I. (eds) Hydrobiology> of the mangal. Dr W. Junk: The Hague, Netherlands. Kathiresan K. and Bingham, B. L. 2001. Biology of mangroves and mangrove ecosystems. Advances in Marine Biology 40: 84-193. Kcnnish, R. and Williams, G.A. 1997. Feeding preferences of the herbivorous crab Grapstts albolineatus: The differential influence of algal nutrient content and morphology. Marine Ecology Pmgress Series 147: 87-95. Lee, S.Y. 1997. Potential trophic importance of the faecal material of the mangrove crab Sesarma messa. Marine Ecology Progress Series 158: 275-284. Lee, S.Y. 1998. Ecological role of grapsid crabs in mangrove ecosystems: a review. Marine and Freshwater Research 49: 335-343. McGuinness, K.A. 1997. Dispersal, establishment and survival of Ceriops australis propagules in a north Australian mangrove forest. Oecologia 109: 80-87. Mead, R. 1988. The design of experiments: statistical principles for practical applications. Cambridge University Press: Cambridge. 30 Feeding selectivity of sesarmid crabs fvjjcheli, F., Gherardi, F. and Vannini, M. 1991. Feeding and burrowing ecology of two East African mangrove crabs. Marine Biology 111:247-254. M jcheli, F. 1993. Feeding ecology of mangrove crabs in North Eastern Australia: mangrove litter consumption by Sesarma messa and Sesanna smitliii. Journal of.Experimental Marine Biology and Ecology’ 171: 164-186. Ngilson, M.J., Painter, T.J., and Richards, G.N. 1986. Flavoglycan: a novel glyconjugate from leaves of mangrove ( Rhizophora stvlosa Griff.). Carbohydrate Research 147: 315 324. Oiafsson, E.. Buchmaycr, S. and Skov, M. 2002. The East African decapod crab Neosarmatium meinerti (de Man) sweeps mangrove floors clean of leaf litter. Ambio 31: 569-573. p e rez-Harguindeguy, N., Diaz, S., Vendramini, F., Cornclissen, J.H.C., Gurvich, D.E. and Cabido, M. 2003. Leaf traits and herbivore selection in the field and in cafeteria experiments. Austral Ecology’ 28: 642-650. R a ffa, K.F., Havill, N.P. and Nordheim, E.V. 2002. How many choices can your test animal compare effectively? Evaluating a critical assumption of behavioural preference tests. Oecologia 133: 422-429. Robcrston, A.I. 1986. Leaf-burying crabs: their influence on energy flow and export from mixed mangrove forests (Rhizophora spp.) in northeastern Australia. Journal of Experimental Marine Biology and Ecology 102: 237-248. Robertson, A.I. 1988. Decomposition of mangrove leaf litter in tropical Australia. Journal of Experimental Marine Biology' and Ecology 116: 235-257. Salgado Kent, C.P. 2004. The significance of plant-animal interactions: how crabs affect structure and function. Unpublished PhD Thesis. Charles Darwin University: Darwin. Skov, M„ and Hartnoll, R. 2002. Paradoxical selective feeding on a low-nutrient diet: why do mangrove crabs eat leaves. Oecologia 131: 1-7. Smith, A.N. 1998. Benthic macroinvertebrates of tropical and subtropical Australian mangroves: relationships with litterfall and latitude. Unpublished MAppSc Thesis. Southern Cross University: Lismore, New South Wales. Smith, T.J. III. 1987. Seed predation in relation to tree dominance in distribution in mangrove forests. Ecology 68: 266-273. Smith, T.J. Ill, Boto, K.G., Frusher. S.D. and Giddins, R.L. 1991. Keystone species and mangrove forest dynamics: the influence of burrowing by crabs on soil nutrient status and forest productivity. Estuarine and Coastal Shelf Science 33: 419-432. Sousa, W. and Mitchell, B. 1999. The effect of seed predators on plant distributions: is there a general pattern in mangroves. Oikos 86(1): 55-66. Steinkc, T., Rajh, A. and Holland, A. 1993. The feeding behaviour of the red mangrove crab Sesarma meinerti dc Man, 1887 (Crustacea: Decapoda: Grapsidae) and its effect on the degradation of mangrove leaf litter. South African Journal of Marine Science 13: 151-160. Underwood. A.J., Chapman, M.G. and Crowe, T. 2004. Identifying and understanding ecological preferences for habitat or prey. Journal of Experimental Marine Biology’ and Ecology 300: 161-187. Warren, J. 11.1987. Behavioural ecology of crabs in temperate mangrove swamps. Unpublished Ph.D. Thesis. University of Sydney: Sydney. Winer, B., Brown, D., and Michels, K. 1991. Statistical principles in experimental design, 3rd edition. McGraw-Hill Inc.: New York. WoodrolTe, C„ Bardslcy, K.. Ward. P. and I lanley, J. 1988. Production of mangrove litter in a macrotidal embaryment, Darwin I larbour. N.T., Australia. Estuarine. Coastal and Shelf Science 26: 581-598. Accepted 8 October 2008 31 . *•- . The Beagle. Records of the Museums and Art Galleries of the Northern Territory, 2008 24: 33-54 Taxonomic review of Candalides absimilis (C. Felder, 1862) and C. margarita (Semper, 1879) (Lepidoptera: Lycaenidae), with descriptions of two new subspecies MICHAEL F. BRABY Biodiversity Conservation Division. Department of Natural Resources, Environment, the Arts and Sport, PO Box 496, Palmerston, NT0831, AUSTRALIA School of Botany and Zoology, The Australian National University, Canberra, ACT 0200, A USTRALIA michael.braby@nt.gov.au ABSTRACT The taxonomic status of two members of the Candalides absimilis (C. Felder, 1862) species group from Australia is reviewed. Two new subspecies, C. absimilis eastwoodi ssp. nov. from north-eastern Queensland and C. absimilis edwardsi ssp. nov. from south-eastern Australia, are described, diagnosed and compared with related taxa. C. gilberti Waterhouse, 1903 from north-western and central northern Australia is shown to comprise a geographical race of C. margarita (Semper, 1879) based on comparative evidence of the genitalia, adult and immature stage morphology, life history and adult behaviour, and accordingly is treated as a subspecies of that taxon. Information on the distribution, habitat, conservation status and biology are summarised for each of the three taxa, with additional notes provided on the larval food plant, behaviour and life cycle of C. absimilis edwardsi ssp. nov, which specialises as larvae on the flush growth of Bracfiychitan populneus (Schott and Endl.) R.Br. (Sterculiaceae) growing in eucalypt woodland. Females of both C. absimilis edwardsi ssp. nov. and C. margarita gilberti are remarkable in that they comprise the only taxa within the C. absimilis species group from Australia in which the white patches on the fore and hind wing are replaced with blue. A striking association between habitat (broad vegetation type) and adult female phenotype in the C. absimilis species group is noted, with highly contrasting ‘black and white forms’ occurring in closed forest and less striking ‘blue forms’ in open forest/woodland, and this relationship is found to extend more generally within the Australian Lycaenidae. A general hypothesis, that ambient light properties among divergent habitats or vegetation types (i.e. different light environments) is a potent selective force shaping sex-limited phenotype, is proposed for these butterflies. Keywords: ambient light environment, Australian monsoon tropics, butterfly, Candalidini, Candalides, Cape York Peninsula, Iron Range, East Gippsland, sexual dimorphism. INTRODUCTION The tribe Candalidini of lycaenid butterflies includes about 33 species referred to two genera, Candalides Hiibner and Nesolycaena Waterhouse and Turner, both of which are endemic to the Australian Zoogeographic Region (Tite 1963; Edwards and Kerr 1978; Braby 1996, 2000; Braby and Douglas 2004). Braby (2000) and Braby and Douglas (2004), building on the foundational work of Tite (1963) and Edwards and Kerr (1978), recognised three species groups within Candalides, viz : the C. absimilis (C. Felder, 1862) species group, the C. erinus (Fabricius, 1775) species group, and the C. xanthospilos (Hiibner, [1817]) species group. Phylogenetic relationships among the three species groups of Candalides and Nesolycaena are unknown at present, but each taxonomic unit is assumed to comprise a monophyletic group. By far the largest species group is the C. absimilis species group, containing 21 described species (Parsons 1998; Braby 2000; Tennent 2006). The group is characterised by having a comparatively long labial palpus, especially in the female; the androconia! scales in the male fore wing being concentrated into a diffuse trident¬ shaped patch along the basal half of veins M v CuA, and CuA„ as well being distributed on the adjacent veins; and the underside pattern of the wings possessing a series of small dark markings and wavy lines but usually without terminal spots, although some species have one or more black tomal spots on the hind wing. In males, the termen of the hind wing is often produced towards the tomus, and the upperside ground colour is lustrous blue; while in females, the upperside ground colour of both wings is frequently black with a prominent white central patch. Members of this species group are distributed predominantly in rainforest habitats of mainland New Guinea and its adjacent islands and the coastal areas of eastern Australia. One species also extends across the monsoon tropics of northern Australia, while another, C. cvprotus (Olliff, 1886), occurs in heathland habitats in the coastal, semi-arid and arid areas of southern, central and south-western Australia (Braby 2000; Grand 2006). In the past few decades, six species in the Candalides absimilis species group have been recognised from the M. F. Braby Australian mainland (Common and Waterhouse 1972, 1981; Edwards 1996; Braby 2000; Edwards et al. 2001). One of these, C. absimilis, shows complex geographical variation, with divergent populations at the extreme northern and southern ends of the range, and three distinct forms were recognised by Braby (2000). Another taxon, C. margarita gilberti Waterhouse, 1903 stat. rev. from the Kimberley and Top End of north-western and central northern Australia, has hitherto been treated as a distinct species since its description, but Braby (2000) postulated that it may be conspecific with C. margarita (Semper. 1879), comprising a geographical subspecies closely related to the nominate subspecies from eastern Australia. The purpose of this paper is to clarify and revise the taxonomic status of the geographical forms of C. absimilis and the taxonomic status of C. margarita gilberti. Information is also summarised on the distribution, habitat, conservation status, and biology of each species. Attention is drawn to a striking association between adult female wing colour pattern and light characteristics among divergent habitats (broad vegetation types) in Candalides , and this relationship is found to extend more broadly within the Australian Lycaenidae. METHODS The male and female genitalia of Candalides absimilis were dissected and examined from six populations spread across the species’ latitudinal range, viz: Cape York Peninsula (Iron Range, Mt White Coen); Wet Tropics (Shiptons Flat, Townsville); central Queensland (Mackay, Eungella); south-eastern Queensland (Eudlo); north-eastern New South Wales (Broken Head); and eastern Victoria (Mitchell River). The male and female genitalia of C. margarita were examined from four populations distributed across the geographical range of the species within Australia, viz: Top End of the Northern Territory (Darwin, Adelaide River); Gulf Country of western Queensland (Musselbrook, near Lawn Hill National Park); Wet Tropics (Paluma Range west of Townsville); and north-eastern New South Wales (Grafton). The size of the white central patch in female Candalides absimilis was analysed quantitatively across the species geographical (latitudinal) range. Specimens (N= 173) from seven populations spread along the eastern coast, viz: (1) Cape York Peninsula (Iron Range); (2) Wet Tropics (Caims- Townsville); (3) central Queensland (Mackay-Eungclla); (4) south-eastern Queensland (Brisbane); (5) central New South Wales (Sydney); (6) south-eastern coastal New South Wales (Central Tilba-Bega); and (7) eastern Victoria (East Gippsland), were measured to assess the extent of geographical variation. Thirty specimens were sampled from each population, except for Cape York Peninsula and central Queensland for which only eight and 15 specimens were available respectively. The shape of the white patch on both wings is either an ellipse or approximately circular; the area of the patch was therefore calculated using the formula Tt^r,, where r, = radius 1, and r,=radius 2. For the fore wing, r, was measured along the cubitus and vein M 3 , while r was measured perpendicular to r p from the mid costa to tornus; for the hind wing, r, was measured along vein M„ while r, was measured perpendicular to r,, from the apex to tornus. Radial measurements were made with vernier callipers to a precision of 0.1 mm. Fore wing length, from the base to the apex of the wing, was measured for each specimen as a proxy for body size. The following abbreviations refer to repositories where material has been examined: AMS: Australian Museum, Sydney ANIC: Australian National Insect Collection, Canberra BMNH; Natural History Museum, London (formerly British Museum of Natural History) MCZ: Museum of Comparative Zoology, Harvard University, USA MFBC: private collection of Michael Braby MTQ: Museum of Tropical Queensland, Townsville NMV: Museum Victoria, Melbourne (formerly National Museum of Victoria) NTM: Museum and Art Gallery of the Northern Territory, Darwin (formerly Northern Territory Museum) SAM: South Australian Museum, Adelaide Label data of all material are given as depicted on the specimen, with quotation marks used to designate each separate label for specimens with two or more labels. The symbol “|” is used to denote text on the reverse side of a label. TAXONOMY Candalides absimilis eastwoodi ssp. nov. (Figs 7-10, 26) Type material. HOLOTYPE - c? labelled “MT. WHITE, COEN, 7 JAN 1994, R. EASTWOOD” (ANIC). PARATYPES- 14cfc?, 399. QUEENSLAND; 19 labelled “CLAUDIE R., 29 12 13”, “LEP-9841” (NMV); 19 labelled similarly but with date “18 2 1914”, registration number"LEP-9837” (NMV); 19 labelled similarly but with date “5 4 14”, registration number “LEP-9840” (NMV); 2c? labelled “Claudie. R. Nth. Q., 5/16-V-61, J. Macquecn” (ANIC); 1 cf labelled “Mount White, Coen, Qld, I3°58’S 143° 11 ’E, 29 Apr 1989, G. and A. Daniels”, “G. Daniels Collection” (AMS); 4c? labelled “Mt White, Coen, Qld, 13°58’S 143°] l’E, 7 Jan 1994 447m, G. and A. Daniels”, “G. Daniels Collection” (AMS); 3cf labelled “MT. WHITE, COEN, 7 JAN 1994, R. EASTWOOD” (ANIC); 4c? labelled similarly but with date “12 JAN 1994” (ANIC). 34 Taxonomic review of Candalides Figs 1-25. Type material of Candalides absimilis and C. margarita: 1-3, Holochila absimilis C. Felder, 1862 lectotype male from Ash Island, NSW (BMNH), showing upperside, underside and specimen labels; 4-6, C. persimilis Waterhouse, 1942 lectotype male from Mackay, Qld (AMS), showing upperside, underside and specimen labels; 7-8, C. absimilis eastwoodi ssp. nov. holotype male from Mount White, Coen, Qld (ANIC), showing upperside and underside; 9-10, C. absimilis eastwoodi ssp. nov. paratype female from Iron Range. Qld (NMV), showing upperside and underside; 11-12. C. absimilis edwardsi ssp. nov. paratype male from Mitchell River National Park, Vic. (NMV), showing upperside and underside; 13-14, C. absimilis edwardsi ssp. nov. holotype female from Suggan Buggan. Alpine National Park, Vic. (ANIC), showing upperside and underside; 15-16, 23, Holochila margarita Semper, 1879 lectotype male from Bowen, Qld (BMNH), showing upperside, underside and specimen labels; 17-18, C. margarita margarita female from Bluewater State Forest west of Townsville, Qld (MFBC), showing upperside and underside; 19-20, 24, C. gilberti Waterhouse, 1903 lectotype male from Darwin, NT (AMS), showing upperside, underside and specimen labels; 21-22,25, C. gilberti paralectotype female from Darwin, NT (AMS), showing upperside, underside and specimen labels. Scale bar = 20 mm. 35 M. F. Braby Other material examined. 23cfcT, 699- QUEENSLAND: 1U labelled “1 mile NE Mt. Lamond, Iron Range. Qld., 21 Dec. 1971”, “D.K. McAlpine, G.A. Holloway, D.P. Sands” (AMS); 19 labelled “Claudie R, 2 miles S Mt. Lamond, Iron Range, Qld., 13 Jan. 1972”, “D.K. McAlpine & G.A. Holloway”, “Australian Museum K231664” (AMS); 19 labelled “ Candalides consimilis toza? 9, Claudie Rv area. N.Q.. 4 January 1974, Shane F McEvey.”, “Genitalia Slide M345” (ANIC); lcf labelled “Iron Range, Cape York. Qld, 13.xi. 1991, S.J. Johnson” (MTQ); lef labelled “IRON RANGE, N. Qld, 25 Nov. 1995, D J FERGUSON” (ANIC); 2cf, 1 9 labelled “Iron Range, Cape York Pen. Qld., 5-11 .xii. 1995, S.J. Johnson” (MTQ); 19 labelled “Wcnlock River crossing, Coen-Iron Range road, Cape York Pen., N. Qld., 10 Sep. 1974, M.S. & B.J. Moulds”, “M.S. MOULDS COLLECTION”, “Australian Museum K231665” (AMS); 19 labelled “Rocky River (nth of Silver Plains), Qld, 13°48.415’S, 143 o 28.20UE, 27 Jul 2005, R.P. Field”, “LEP-480S8” (NMV); lcMabelled “Mt White, Coen, N. Qld, 5 Nov. 1979, M.S. & B.J. Moulds”, “M.S. MOULDS COLLECTION” (AMS); 1 cf labelled “Mount White, Coen, Qld, 13°58’S 143°1 HE, 29 Apr 1989, G. and A. Daniels”, “G. Daniels Collection” (AMS); 1 cf labelled “Mt. White, Coen. Qld., 19,i. 1993, S.J. Johnson” (MTQ); 1 cf labeled “Mt White, Coen, Qld, 10 Jul 1993, R.P. Field”, “LEP-44854” (NMV); 1 cf labelled similarly but with date “11 Jul 1993” and registration number “LEP- 44855” (NMV); 1 cf, 19 labelled “Mt White, Coen, Qld, 13°58’S 143°1 l’E, 7 Jan 1994 447m, G. and A. Daniels”, “G. Daniels Collection” (AMS); 1 cf labelled similarly but with date and collector “12 Jan 1994, G. Daniels R. Eastwood” (AMS); 2cT labelled “MT. WHITE, COEN, 7 JAN 1994, R. EASTWOOD” (MCZ); 3cf labelled “Mt. White, Coen. QLD., 27.iv. 1997, S.J. JOHNSON” (MTQ); 3cf labelled similarly but with date “9.v. 1998” (MTQ); 1 cf labelled “Mt White, Coen, Qld, I3°58.035’S, 143°11.538’E, 4 Jul 2005, R.P. Field”, “LEP-48086” (NMV); 1 cf labelled similarly but with registration number “LEP-48087” (NMV); 1 cf labelled similarly but with date “5 Jul 2005” and registration number “LEP-48085” (NMV). Adult description. Male. Fore wing length 16.4-19.2 mm. Upperside of wings lilac-blue, with costa, termen and outer half of dorsum narrowly edged black, radial, median, cubital and anal veins black towards wing margin; fore wing with a broad distinct median patch of androconial scales, patch more prominent at anterior end of cell and base of veins M : and M, and their origin with the discocellulars, at base of veins M,, CuA, and CuA, and their origin with the cubital vein (Cu), and along vein IA+2A; hind wing with a conspicuous black tomal spot between ends of veins CuA, and 1A+2A, and a smaller indistinct black terminal spot between ends of veins CuA, and CuA,. Underside of wings silvery white, with a series of small dark brown markings and a narrow dark brown termen; hind wing with a conspicuous dark brown tomal spot between ends of veins CuA, and 1A+2A, and sometimes a series of four smaller indistinct dark brown terminal spots between ends of veins M, and CuA,. Female. Fore wing length 17.2-19.8 mm. Upperside of wings black enclosing a large white central patch, base suffused with purplish blue scales; fore wing with purplish blue scales extending along posterior margin of white central patch above vein 1A+2A and sometimes along anterior margin of patch in cell; hind wing with scattered purplish blue scales bordering posterior and outer margins of white central patch. Underside of wings with ground colour and markings similar to male, but markings usually more distinct. Variation. Males show minor variation in the extent of the tomal spot on the hind wing, and in the shape of the hind wing in which the termen and tomus are sometimes rounded. In females, the extent of the white central patch on the upperside of the hind wing varies. Comparison. Candalides absimilis eastwoodi ssp. nov. males are distinct and can be readily distinguished from males of C. absimilis absimilis (Figs 1 -6) and C. absimilis edwardsi ssp. nov. (Figs 11-12) by differences in the shape of the fore wing in which the termen is straighter and less convex, and the apex more sharply pointed; the shape of the hind wing in which the tornus is generally produced to a point; the extent of the underside markings, which are less conspicuous and only faintly visible, especially the spots and wavy lines on the hind wing; the upperside ground colour, which is paler purplish blue; the median patch of sex scales on the fore wing, which is more clearly visible; and the extent of the black/dark brown tomal spot on the hind wing, which is generally more prominent on the upperside and always present on the underside. Candalides absimilis absimilis males from the northern end of their range (i.e. Wet Tropics - between Cooktown and Townsville) are similar to C. absimilis easbvoodi ssp. nov. in that the tornus of the hind wing is often produced and the tomal spot is usually present (though usually not as well developed), but otherwise more closely resemble the lectotype male of C. absimilis absimilis from Ash Island, New South Wales (Figs 1, 2), in that the termen of the fore wing is more rounded, the underside markings are more distinct, and the upperside ground colour is brighter/deeper blue with the patch of sex scales less clearly visible. Candalides absimilis eastwoodi ssp. nov. females differ from C. absimilis edwardsi ssp. nov. females by possessing a large, conspicuous white central patch on the upperside of each wing. The white patch on the fore wing is similar in size to that of C. absimilis absimilis populations from south of the Wet Tropics (i.e. Mackay-Sydney), but apparently smaller than those from the Wet Tropics (i.e. Cooktown-Townsville). The white patch on the hind wing of C. absimilis eastwoodi ssp. nov., however, is substantially larger than populations of C. absimilis absimilis from south of the Wet Tropics, but significantly smaller than those from the Wet Tropics (see quantitative analysis below under Remarks). 36 Taxonomic review of Candalides Candalides absimilis eastwoodi ssp. nov. has previously been confused with C. consimilis Waterhouse, 1942 (Dunn et al. 1994; Dunn 1995) to which it closely resembles. On Cape York Peninsula, C. consimilis is known only from Iron Range and has been separated as the northern subspecies C. consimilis toza Kerr, 1967. Dunn et al. (1994) initially placed material from Mount White near Coen with C. consimilis toza, which is known only from a few male specimens (Kerr 1967), but later considered them to belong to an undescribed subspecies of C. consimilis closely allied to C. consimilis toza (Dunn 1995). However, the valva of the male genitalia (Fig. 26) and sinus vaginalis of the female genitalia (not illustrated) clearly indicate that the population from Mount White belongs with C. absimilis and not with C. consimilis. Also, males of C. absimilis eastwoodi ssp. nov. are distinguished from those of C. consimilis toza by the absence of a conspicuous series of black terminal spots on the uppersidc of the fore and hind wings, particularly between the ends of veins M, and CuA,. Remarks. Waterhouse (1942) described Candalides persimilis from Mackay, Qld, and placed material from the Wet Tropics (Cairns, Kuranda) and coastal north-eastern New South Wales (Byron Bay) with that species. The taxon was differentiated on the basis of the paler lilac uppersidc ground colour of the male with a more conspicuous patch of sex-scales, a large white central patch on the fore wing of the female, the presence of a distinct black tornal spot on the underside of the hind wing, and minor differences in the male genitalia in which the apices of the valvae are smaller and truncate. Titc (1963) and McCubbin (1971) questioned the validity of C. persimilis Waterhouse, 1942, but Tindale (1965) recognised the taxon and indicated a wide sympatric distribution with C. absimilis along the eastern coast, from Cairns, Queensland, to Narara, New South Wales. Tindale crudely illustrated and compared the male genitalia, although his Figure 4 does not match that of Waterhouse’s (1942) Figure 1 d for the same structural character. Common and Waterhouse (1972, p. 415) did not recognise the species, noting that “A close study of C. persimilis Waterhouse suggests that it is simply a form of C. absimilis." The species group name persimilis was subsequently treated as a junior synonym of C. absimilis by Edwards (1996) and Edwards el al. (2001). Examination of the lectotype male (Figs 4-6) and paralectotypes of C. persimilis in the AMS, including the male genitalia, indicate that this synonymy is justified. The minor differences in the genitalia alluded to by Waterhouse (1942) and Tindale (1965) are part of the overall geographical variation observed in C. absimilis. The male genitalia from three widely dispersed populations (Cape York Peninsula, south-eastern Queensland, eastern Victoria) are illustrated in Figures 26-28. The genitalia show minor geographical variation, particularly in their overall size and in the shape of the valvae. In the northernmost population (C. absimilis eastwoodi ssp. nov.), the genitalia are consistently smaller, and the bifurcated apices of the valvae are smaller, less pronounced and rotated inwardly so that they lie perpendicular to the length of the valva (Fig. 26). The genitalia from the Wet Tropics (C. absimilis absimilis) appear transitional with those from Cape York Peninsula (C. absimilis eastwoodi ssp. nov.) and Mackay- Eungella in central Queensland (C. absimilis absimilis) in having relatively small apices that are partially rotated. Quantitative analysis of the size of the white central patch of Candalides absimilis females (,V= 140) revealed that patch size, when present, was positively correlated with body size, and this relationship was more tightly correlated in the fore wing (r = 0.755, d.f 138, P < 0.001) than in the hind wing (r= 0.560, d.f 138, P< 0.001) (Fig. 33). After controlling for the effects of body size, patch size was found to exhibit clinal variation, being negatively correlated with latitude (Fig 34). Analysis of variance of mean patch size, measured for both fore and hind wing from the seven populations sampled across the species’ geographical range, revealed that the relationship was highly significant for each (fore wing: F= 7.348, d.f. 5, P = 0.0422; hind wing: F= 11.990, df 5 ,P = 0.0180). Closer inspection of the relationship, however, suggests a steep cline in this character for the hind wing but not the fore wing between Caims-Townsville and Mackay-Eungella, Qld, negligible change between Mackay, Queensland, and Sydney, New South Wales, followed by another steep cline approaching a step for both wings south of Sydney, New South Wales, to East Gippsland, Victoria. (Fig. 34). Patch size reached its maximum extent in the Wet Tropics, and then appeared to diminish in size further north on Cape York Peninsula, although the sample size for the latter population (n = 8 for C. absimilis eastwoodi ssp. nov.) was too small for comparison with the Wet Tropics. Although C. absimilis absimilis females from the Wet Tropics have the largest white patches on both wings, this population is better treated as a local form rather than as a distinct subspecies until more C. absimilis eastwoodi ssp. nov. females become available for comparison. C. absimilis absimilis females from Mackay more closely resembled populations from Brisbane and Sydney than those from Caims-Townsville with respect to patch size (see also comments by Tite 1963). Etymology. The taxon is named in honour of Dr Rodney Eastwood who has made a substantial contribution to the scientific study of Australian butterflies, especially the natural history and ecology of ant-lycaenid associations and the phylogcography of Lycaenidae. Distribution. Candalides absimilis eastwoodi ssp. nov. is restricted to Cape York Peninsula, from Iron Range (McEvey 1977) south along the Mcllwraith Range to the Rocky River north of Silver Plains (R.P. Field) and Mt White near Coen, Queensland (Braby 2000) (Fig. 35). It is known from six sites (i.c. point localities > 1 km apart) representing four discrete locations (i.e. areas > 10 km apart). The subspecies is probably endemic to the mid¬ peninsula rainforest block (Iron Range-Mcllwraith Range) 37 M. F. Braby 1 mm Figs. 26-32. Male genitalia of Candalides absimilis and C. Margarita , showing ventral view, with aedeagus removed: 26. C. absimilis eastwoodi ssp. nov. from Iron Range, Old; 27, C. absimilis absimilis from Eudlo, Old: 28. C. absimilis edwardsi ssp. nov. from Mitchell River, Vic.; 29, C. margarita gilberti from Darwin, NT; 30, C. margarita gilberti from Musselbrook, Gulf of Carpentaria, Qld; 31, C. margarita margarita from the Paluma Range west of Townsville, Qld; 32, C. margarita margarita from Grafton, NSW. Scale bar = 1 mm. 38 Taxonomic review of Candalides Body size (mm) Body size (mm) Fig. 33. Relationship between body size (measured as fore wing length) and size of white patch for each wing in Candalides absimilis females: A, fore wing; B, hind wing. Correlations are significant for both wings (P < 0.001). of Cape York Peninsula, which is geographically separated from the Wet Tropics where the nominate subspecies C. absimilis absimilis reaches its northernmost limit at Cooktown, Queensland. Despite reports to the contrary, Candalides absimilis eastwoodi ssp. nov. has not been recorded further north than the tip of Cape York Peninsula and the Torres Strait islands. These reports were found, on closer inspection, to be erroneous. A male labelled “Prince of Wales Is., 9 June 1908, H. Elgner.”, “LEP-9890”, “Seems to belong with Candalides absimilis Feld., det K.L. Dunn 1989” (NMV) is considered to have been mislabelled and excluded from the material examined. Dunn and Dunn (1991) considered the record to be reliable and included Prince of Wales Island from southern Torres Strait islands in their review of the distribution of the species. A detailed chronology of Elgner’s geographical movements in the Torres Strait islands and Cape York, based on insect specimens that he collected (Moulds 1977; Dunn 2007), indicate that for the period 28 April-19 July 1908, Elgner was stationed on Prince of Wales Island, so he would have had access to the island in June 1908. However, Elgner’s specimen of C. absimilis does not agree with the characteristics of C. absimilis eastwoodi ssp. nov. It lacks the straight termen and pointed apex of the fore wing, paler upperside ground colour, and indistinct markings on the underside; it more closely resembles C. absimilis absimilis from south¬ eastern Queensland and New South Wales. Moreover, examination of the genitalia indicates that it conforms with typical C. absimilis absimilis (Fig. 27) and not to C. absimilis eastwoodi ssp. nov. (Fig. 26). Dunn (2007) noted a high level of accuracy in Elgner’s label data; therefore, it is possible the specimen was mislabelled by subsequent curators. Two male specimens of C. margarita margarita in AMS labelled “Horn Is. Q„ 25.6.43” and “Cape Yk. Q, 25.6.43” were found to have been misidentified as C. absimilis in the drawers of the main reference collection. Biology. The early stages and larval food plants of Candalides absimilis have not been recorded. Adults have been collected in September, from November to February and in April, May and July, with most specimens collected during the ‘build-up’ and early wet season. Available data on adult phenology suggests the subspecies is on the wing throughout the year. Near Coen, males have been collected hilltopping at the summit of Mount White (R.G. Eastwood, personal communication; D.A. Lane, personal communication), a behaviour that appears to be largely absent in the two other subspecies (except for the population of C. absimilis absimilis in the Wet Tropics, which has been recorded hilltopping at Mt Stuart Townsville, M.F. Braby unpublished data). Conservation status. Candalides absimilis eastwoodi ssp. nov. has a narrow geographical range, with an estimated extent of occurrence of4,500 km 2 . Available data suggests the subspecies is restricted to tropical forest of the mid¬ peninsula rainforest block of Cape York Peninsula (i.e. Iron Range-Mcllwraith Range), most of which is now protected in the Iron Range and Mcllwraith National Parks. There is no information on population trends, fragmentation of populations, or immediate threatening processes impacting on the known sites, and the ecology of the subspecies is very poorly known. Although the range of the taxon is suspected to be relatively circumscribed, there is inadequate information to make a direct assessment of the risk of threat. Its conservation status should therefore be regarded as Data Deficient (DD) according to 1UCN criteria. 39 M. F. Braby Fig. 34. Relationship between size of white patch divided by body size and latitude for both fore and hind wing in Candalides ahsimilis females sampled from seven populations across the species’ geographical range. Points are mean values with error bars representing one standard deviation. Sample sizes are shown above graph. Regression equations for the significant relationship between mean patch size and latitude are as follows: y = 1.984 - 0.0388.V for fore wing (F 7.348, d.f 5 ,P = 0.0422), and y 1.733 - 0.0409.V for hind wing (F 11.990, d.f 5, P = 0.0180). Accompanying map of eastern Australia shows distribution of C. absimilis (modified from Braby 2000), with examples from each of the seven geographical areas, as follows: A, Cape York Peninsula, Qld (Iron Range-Mcllwraith Range); B, Wet Tropics, Qld (Cairns- Townsville); C, central Queensland (Mackay-Eungclla): I), south-eastern Queensland (Brisbane): E, central New South Wales (Sydney); F, south-eastern coastal New South Wales (Central Tilba-Bcga); and G, eastern Victoria (East Gippsland). Candalides ahsimilis edwardsi ssp. nov. (Figs 11-14, 28) Type material. HOLOTYPE - 9 labelled “36°56’S, 148°21’E; 600m, 7 km N Suggan Buggan (by rd), Alpine NP, VIC., emg. 6 JAN. 2003, M.F. Braby & L.J. Aitchison”, “Reared from larva on Brachychiton populneus, coll. 8-9 DEC. 2002, pupated 25 DEC. 2002“ (AN1C). PARATYPES - 15CTCT, 1499. NEW SOUTH WALES: 19 labelled “36°45’S, 148°25’E; 300m, 18 km N Willis (by road), Kosciuszko NP, NSW; emg. 13 JAN. 2003, M.F. Braby & LJ Aitchison”, “Reared from larva on Brachychiton populneus , coll. 10 DEC. 2002, pupated 2 JAN. 2003” (AMS). VICTORIA: 19 labelled “36°56’S, 148°21 ’E; 600m, 7 km N Suggan Buggan (by rd), Alpine NP, VIC., emg. 4 JAN. 2003, M.F. Braby & L.J. Aitchison”, “Reared from larva on Brachychiton populneus, coll. 8-9 DEC. 2002, pupated 23 DEC. 2002” (NMV); lef labelled similarly but with dates “emg. 5 JAN. 2003”, “pupated 22 DEC. 2002” (AMS); 19 labelled similarly but with dates “emg. 12 JAN. 2003”, “pupated 31 DEC. 2002” (AMS); 1 Cf labelled similarly but with dates “emg. 15 FEB. 2003”, “pupated 29 DEC. 2002” (NMV); lef labelled similarly but with dates “emg. 17 OCT. 2003”, “pupated 30 DEC. 2002” (ANIC); IcT labelled “37°05'S, 148°24’E; 200m, McKillops Bridge, Snowy River NP, VIC., 3 JAN. 2004, M.F. Braby” (ANIC); 1 cf labelled “37°05’S, 148°24’E, McKillops Bridge, Snowy RiverNP, VIC., emg. 23 JAN. 2004, M.F. Braby”, “Reared from larva on Brachychiton populneus, coll. 3 JAN. 2004” (NMV); I cf labelled similarly but with date “emg. 27 JAN. 2004” (AMS); 1 cf labelled similarly but with date “emg. 28 JAN. 2004” (AMS); 19 labelled similarly in NMV; 1 cf. I 9 labelled but with date “emg. 29 JAN. 2004” (ANIC); 1 cf labelled similarly in NMV; 19 labelled similarly in AMS; IcT labelled “37°29’S, 148°09’E, Buchan Caves Reserve, VIC; 100m, emg. 10 JAN. 2004, M.F. Braby”, “Reared from larva on Brachychiton populneus, coll. 19-21 DEC. 2003” (ANIC); 1 cf labelled similarly but with date “emg. 12 JAN. 2004" (AMS); 19 labelled similarly in NMV; 1 cf labelled similarly but with date “emg. 15 JAN. 2004” (NMV); 19 labelled similarly but with date “emg. 16 JAN. 2004” (ANIC); 19 labelled “37°42’S, I47°21’E, Den of Nargun, Mitchell River NP, VIC.., emg. 7 JAN. 2003, M.F. Braby & L.J. Aitchison”, “Reared from egg on Brachychiton populneus, coll. 6-7 DEC. 2002, pupated 24 DEC. 2002” (ANIC); 1 cf labelled similarly but with dates “emg. 9 JAN. 2003”, “pupated 28 DEC. 2002” (NMV); 40 Taxonomic review of Candalides distribution of Candalides absimilis eastwoodi ssp. nov. (■). 1 cT labelled similarly but with date “emg. 10 JAN. 2003” (ANIC); 19 labelled similarly but with dates “emg. 12 JAN. 2003”, “pupated 30 DEC. 2002” (AMS); 19 labelled similarly but with date “pupated 31 DEC. 2002” (NMV); 19 labelled “37°42’S, 147 0 21’E, DenofNargun, Mitchell River NP, VIC., 23 DEC. 2003, M.F. Braby & L.J. Aitchison” (ANIC); 19 labelled similarly in AMS; 19 labelled “37°42’S, 147°21’E, Den of Nargun, Mitchell River NP, VIC., emg. 15 JAN. 2004, M.F. Braby”, “Reared from larva on Brachychiton populneus , coll. 21-23 DEC. 2003” (NMV); lef labelled similarly but with date “emg. 19 JAN. 2004” (AMS). Other material examined. 61 cfcf, 329 9- NEW SOUTH WALES: lef labelled“36°45’S, 148°25’E; 300m, 18kmN Willis (by road), Kosciuszko NP, NSW; 10 DEC. 2002, M.F .Braby & L.J. Aitchison” (ANIC); 1 cf labelled similarly in NMV; lef labelled similarly in AMS; lef labelled “36°45’S, 148°25’E; 300m, 18 km N. Willis, Kosciuszko NP, NSW; emg. 3 MAR. 2003, M.F. Braby”, “Reared from larva on Brachychiton populneus, coll. 10 DEC. 2002, pupated 2 JAN. 2003” (MFBC). VICTORIA: lef labelled “36°56’S, 148°21’E; 600m, 7 km N Suggan Buggan (by rd), Alpine NP, VIC., emg. 4 JAN. 2003, M.F. Braby & L.J. Aitchison”, “Reared from larva on Brachychiton populneus, coll. 8-9 DEC. 2002, pupated 22 DEC. 2002” (MFBC); 19 labelled similarly but with date “pupated 23 DEC. 2002” (MFBC); 19 labelled “37°05’S, 148°24’E, McKillops Bridge, Snowy River NP, VIC., emg. 29 JAN. 2004, M.F. Braby”, “Reared from larva on Brachychiton populneus, coll. 3 JAN. 2004,” (MFBC); lef labelled similarly but with date “emg. 30 JAN. 2004” (MFBC); 19 labelled similarly but with dates “emg. 30 AUG. 2004, pupated late JAN. 2004” (MFBC); 19 labelled similarly but with date “emg. 10 SEP. 2004” (NMV); lef labelled “Buchan Caves, Buchan, Vic, 37°30.16’S, 148°9.51’E, 11 Jan 2003, R.P. Field”, “LEP- 46133” (NMV); 1 cf labelled similarly but with registration number “LEP-46134” (NMV); lef labelled similarly but with registration number “LEP-46135” (NMV); 19 labelled “Buchan Caves, Buchan, Vic, 37°29.63’S, 148°10.08’E, 13 Dec 2003, R.P. Field”, “LEP-46197” (NMV); 19 labelled similarly but with date “17 Jan 2004” and registration number “LEP-46201” (NMV); 19 labelled similarly but with date “ern 20 Feb 2004” and registration number “LEP- 46202” (NMV); 19 labelled similarly but with date “cm 27 Feb 2004” and registration number “LEP-46203” (NMV); lef labelled similarly but with registration number “LEP- 46204” (NMV); 1 cf labelled similarly but with date “em 22 Sept 2004” and registration number “LEP-46789” (NMV); 1 cf labelled similarly but with registration number “LEP- 46790” (NMV); 1 cf labelled similarly but with date “em 29 Sept 2004” and registration number “LEP-46785” (NMV); 1 cf labelled similarly but with registration number “LEP- 46786” (NMV); 1 cf labelled similarly but with date “em 5 Jan 2005” and registration number “LEP-46799” (NMV): 1 cf labelled similarly but with registration number “LEP- 46800” (NMV); 1 cf labelled similarly but with date “em 25 Jan 2005” and registration number “LEP-46903” (NMV); lef labelled similarly but with registration number “LEP- 46904” (NMV); 1 cf labelled similarly but with registration number “LEP-46905” (NMV); lef labelled similarly but with registration number“LEP-46906” (NMV); 19 labelled similarly but with date “em 26 Jan 2005” and registration number “LEP-46901” (NMV); 1 9 labelled similarly but with registration number“LEP-46902” (NMV); 1 cf labelled similarly but with date “em 27 Jan 2005” and registration number “LEP-46897” (NMV); lef labelled similarly but with registration number“LEP-46898” (NMV); 1 cf labelled similarly but with registration number “LEP-46899” (NMV); 19 labelled similarly but with registration number “LEP-46900” (NMV); 19 labelled similarly but with date “em 28 Jan 2005” and registration number “LEP-47218” (NMV); 19 labelled similarly but with registration number “LEP-47219” (NMV); 19 labelled similarly but with date “em 29 Jan 2005” and registration number “LEP-47220” (NMV); 1 9 labelled similarly but with date “em 30 Jan 2005” and registration number “LEP-47221” (NMV); 19 labelled similarly but with registration number “LEP-47222” (NMV); 19 labelled "37°29’S, 148°09’E, Buchan Caves Reserve, VIC; 100m, emg. 11 JAN. 2004, M.F. Braby”, “Reared from larva on Brachychiton populneus, coll. 19- 21 DEC. 2003” (MFBC); 19 labelled similarly but with 41 M. F. Braby date “emg. 12 JAN. 2004” (MFBC); lcT labelled similarly but with date “emg. 17 JAN. 2004” (MFBC); lcf labelled similarly but with dates “emg. 17 FEB. 2004. pupated DEC. 2003” (MFBC); 1 cf labelled similarly but with dates “emg. 10 JUL. 2004, pupated JAN. 2004” (MFBC); Id labelled similarly but with dates “emg. 12 SEP. 2004, pupated late DEC. 2003” (MFBC); Id labelled “VIC Den ofNargun, 40 km NE Stratford, 37°42’S 147°21'E, 7 December 1977, Shane F. McEvey”, “S.F. McEvey B1 OOOd” (AMS); Id labelled similarly but with registration number “S.F. McEvey B100Id” (AMS); Id labelled similarly but with registration number “S.F. McEvey B1002d” (AMS); Id labelled “VIC The Amphitheatre, Billy Goat Bend, Mitchell River, 37°40’S 147°22’E, 8.xii. 1977 S.F. McEvey”, “S.F. McEvey B1008d” (AMS); Id labelled similarly but with registration number “S.F. McEvey B1009d” (AMS); Id labelled similarly but with registration number “S.F. McEvey BlOlOd” (AMS); Id labelled similarly but with registration numbers “S.F. McEvey B1011 d”, “Australian Museum K233023” (AMS); 19 labelled similarly but with registration numbers “S.F. McEvey B10069”, “Australian Museum K233021” (AMS); 19 labelled “VIC 1km upriver from The Amphitheatre on Billy Goat Bend, Mitchell River, 37°39’S 147 0 21’E, 9.xii.l977 S.F. McEvey” “S.F. McEvey B10209”, “Australian Museum K233024” (AMS); 1 d labelled similarly but with registration number “S.F. McEvey B102Id” (AMS); Id, 29 labelled “DEN OF NARGUN, Mitchell River, Vic., 17 Dec. 1989, K.L. Dunn, J.M. Dunn” (ANIC), d with additional label “AN1C genitalia slide No. 3183 d” and genitalia preparation “LYC 3183, Candalides absimilis (Feld.), AUST. ENT. INS. COLL. Slide E.D. Edwards 1990”, 9 with additional label “Holotype, Candalides pseudogoodingi”; Id, 19 labelled “VIC., Den ofNargun, Mitchell River N.P., 17 Dec. 1989, K.L. & J.M. Dunn”, “Candalides absimilis, det KL Dunn 1989”, “Reference: DUNN 1990, Victorian Entomologist 20: 49-53” (ANIC); 19 labelled similarly in SAM; Id labelled similarly but with registration number “LEP- 9732” (NMV); Id labelled similarly but with registration number “LEP-9733” (NMV); 1 9 labelled similarly but with registration number“LEP-9754” (NMV); 1 9 labelled similarly but with registration number “LEP-9734” (NMV); Id labelled “DEN OF NARGUN. V., MITCHELL R. N.P.,21 FEB. 1991, D.F. CROSBY”(ANIC); Id labelled “37°42’S, 147°21’E, Den ofNargun, Mitchell River NP, VIC., emg. 1 JAN. 2003, M.F. Braby & L.J. Aitchison", “Reared from larva on Brachychiton populneus, coll. 6-7 DEC. 2002, pupated 19 DEC. 2002” (MFBC); 19 labelled similarly but with dates “emg. 5 JAN. 2003, pupated 22 DEC. 2002” (MFBC); 1 d labelled similarly but with date “pupated 23 DEC. 2002” (MFBC); Id labelled similarly but with dates “emg. 7 JAN. 2003, pupated 24 DEC. 2002” (AMS); Id labelled similarly but with dates “emg. 8 JAN. 2003, pupated 25 DEC. 2002” (MFBC); 19 labelled similarly but with dates “emg. 10 JAN. 2003, pupated 30 DEC.2002”(MFBC); Id, 19 labelled“37°42’S, 147°2UE, Den ofNargun, Mitchell River NP, VIC., 2 JAN. 2004, M.F. Braby” (ANIC); 2d, 19 labelled similarly in NMV; 4d labelled similarly in AMS; 3d labelled similarly in MFBC; Id labelled “37°42’S, I47°21’E, Den ofNargun, Mitchell River NP, VIC., emg. 12 JAN. 2004, M.F. Braby”, “Reared from larva on Brachychiton populneus , coll. 21 -23 DEC. 2003” (NMV); Id, 19 labelled similarly but with date “emg. 13 JAN. 2004” (MFBC); 2d labelled similarly but with date “emg. 14 JAN. 2004” (MFBC); Id labelled similarly but with date “emg. 19 JAN. 2004 (ANIC); Id labelled similarly but with date “emg. 23 JAN. 2004” (MFBC); 19 labelled similarly but with dates “emg. 18 SEP. 2004. pupated JAN. 2004” (MFBC); Id labelled similarly but with dates “emg. 23 SEP. 2004, pupated 11 JAN. 2004” (MFBC). Adult description. Male. Fore wing length 15.5 - 16.6 mm. Upperside of wings purplish blue, with costa and termen narrowly edged black, radial, median, cubital and anal veins black towards wing margin; fore wing with a broad indistinct median patch of androconial scales, patch more prominent at anterior end of cell and base of veins M, and M, and their origin with the discocellulars, at base of veins M,, CuA, and CuA, and their origin with the cubital vein (Cu), and along vein 1A+2A; hind wing with a small black tornal spot between ends of veins CuA, and 1A+2A. Underside of wings silvery white, with a series of dark brown markings and a narrow dark brown termen; hind wing usually with an obscure dark brown tornal spot between ends of veins CuA, and 1A+2A, and a series of smaller indistinct dark brown terminal spots between ends of veins Rs and CuA,. Female. Fore wing length 15.3- 16.9 mm. Upperside of wings with costa, termen and outer half of dorsum broadly black enclosing a large purplish blue central patch which extends from base and inner half of dorsum to postmedian area; fore wing usually with a small postmedian patch or suffusion of pale whitish blue scales distal to discocellulars between veins M, and CuA! or CuA,. Underside of wings with ground colour and markings similar to male. Variation. The brightness of the upperside purplish blue ground colour varies in males. In females, the purplish blue central patch on the upperside occasionally extends narrowly as a suffusion of purplish blue scales to the costa between veins R, and R, and to the subterminal area between veins CuA, and 1A+2A on the fore wing, and more broadly to the termen between veins M, and 1A+2A on the hind wing. The upperside of the hind wing very occasionally possesses a small subapical patch or suffusion of pale whitish blue scales between veins Rs and M, or M r Comparison. Candalides absimilis edwardsi ssp. nov. females are distinguished from females of C. absimilis absimilis and C. absimilis eastwoodi ssp. nov. by pronounced differences in the extent of the white central patch on the upperside. In C. absimilis edwardsi ssp. nov., the white patch on the fore wing is absent, being replaced with a small postmedian patch of pale whitish blue scales 42 Taxonomic review of Candahdes between veins M, and CuA,; occasionally this patch is restricted to a narrow postmedian streak between veins M, and M, distal to the discal cell. The white patch is also absent on the hind wing, but occasionally there is a narrow subapical whitish blue streak between veins Rs and M r Some C. absimilis absimilis females from south-eastern coastal New South Wales (Central Tilba-Bega) approach C. absimilis edwardsi ssp. nov. in having much reduced or no white central or subapical patch on the upperside of the hind wing and an extensive basal suffusion of purplish blue scales, but these specimens have a more conspicuous white central patch on the fore wing that is approximately circular in shape and which extends from veins M, to CuAj or CuA,. Candalides absimilis edwardsi ssp. nov. males do not differ from those of nominate C. absimilis absimilis in external morphology. The male genitalia (Fig. 28), however, are slightly larger, with the apical projections of the valvae more pronounced. Candalides absimilis edwardsi ssp. nov. females are similar to those of C. eonsimilis goodingi (Tindale, 1965) from eastern Victoria, but the latter are readily distinguished by the more pointed apex of the fore wing, less rounded tomus of the hind wing, presence of a small but conspicuous white central patch on the upperside of the fore wing which extends below vein M r and the well-defined basal spots on the underside of the hind wing. In addition, the central iridescent area in C. eonsimilis goodingi is a brighter, richer blue, whereas in C. absimilis edwardsi ssp. nov. it is purplish blue. The two species also differ significantly in their genitalia (see Waterhouse 1942; Tite 1963). Remarks. Candalides absimilis edwardsi ssp. nov. was previously regarded as a distinct local form that was believed to be restricted to the Mitchell River National Park of East Ciippsland (Dunn 1990; Dunn and Dunn 1991; Dunn et al. 1994). Dunn and Dunn (1991, p. 391) concluded that in relation to the white central patch of females “This Victorian population represents the opposite extreme of the continuum ... at the subspecific level this population would be considered a poorly differentiated taxon”. However, quantitative analysis of this character over the broad geographical range of the species (Fig. 34) indicates negligible variation between Mackay-Eungella, Queensland, and Sydney, New South Wales, but a steep cline approaching a step between Sydney, New South Wales, and East Gippsland, Victoria. C. absimilis edwardsi ssp. nov. lies at the southernmost limit of the species’ geographical range and represents the extreme end of variation in the extent of the white patch. Its disjunct distribution, together with a broader geographical range than hitherto believed, and the striking female phenotypic divergence in which the reduced white patch comprises a consistent and diagnosable character state that lies at the end of a step or steep latitudinal cline, indicate that the known populations warrant formal recognition at the subspecific level. Examination of label data of museum material of Candalides absimilis edwardsi ssp. nov. revealed several inconsistencies. The original Dunn material comprised 10 specimens (4cf, 69) collected from Den of Nargun on 17 December 1989. Most specimens have an additional label "Candalides absimilis, det KL Dunn 1989”; however, one female in the ANIC is labelled in Dunn’s original handwriting “Holotype” and a manuscript name of Dunn’s is associated with it. Because this name was never formally published, neither the specimen nor the name carry any nomenclatural validity. A male sent to the ANIC for taxonomic identification (see Dunn 1990) has its accompanying genitalia slide labelled “LYC 3183, Candalides absimilis (Feld.), Den of Nargun, Mitchell River, Vic., 17 Dec. 1989, K.L. Dunn & J.M. Dunn, AUST. ENT. INS. COLL. Slide E.D. Edwards 1990”, indicating that correct taxonomic diagnosis of the species was not made until 1990. Etymology. The taxon is named in honour of Mr Ted Edwards in recognition of his outstanding contribution to the study ofAustralian Lepidoptera, curation and development of the ANIC, and continuous support, generous help and mentoring to others, including the author, over many decades. Fig. 36. Map of south-eastern Australia showing geographical distribution of Candalides absimilis edwardsi ssp. nov. (■) and C. absimilis absimilis (A). Stars (★) indicate occurrence of the natural larval food plant, Brachychiton populneus, of both subspecies in the region (data based on records registered in Australia’s Virtual Herbarium http://www.cpbr.gov.au/cgi-bin/avh.cgi). 43 M. F. Braby Distribution. Candalides absimilis edwardsi ssp. nov. is known from a limited region in south-eastern Australia (Fig. 36). It has been recorded from three disjunct areas in south-eastern New South Wales inland of the Great Escarpment and East Gippsland of eastern Victoria at altitudes between 100-600 m. These three areas include: (1) the Upper Snowy River area on the New South Wales- Victoria border (M.F. Braby); (2) Buchan Caves Reserve (New et cd. 2007); and (3) Mitchell River National Park (Dunn 1990). In the Upper Snowy River area, it has been recorded from three locations, viz: the southern sections of Kosciuszko National Park, New South Wales, close to the Victorian border; at Suggan Buggan and Willis in Alpine National Park, Victoria; and at McKillops Bridge in Snowy River National Park, Victoria. At Buchan Caves Reserve, it is known only from one site. In the Mitchell River National Park, it is known from three sites along the Mitchell River: 1 km upstream from The Amphitheatre on Billy Goat Bend, The Amphitheatre on Billy Goat Bend, and Den ofNargun. Thus, C. absimilis edwardsi ssp. nov. is known from a total of eight sites representing five locations. Candalides absimilis edwardsi ssp. nov. is geographically separated from C. absimilis absimilis , which reaches its southernmost limit in the coastal/subcoastal areas of south¬ eastern New South Wales (Fig. 36). C. absimilis absimilis was previously knowrn as far south as Tilba Tilba, New South Wales (Braby 1998), but surveys by the author have detected the species further south near Cobargo and several sites at Bcga. The two subspecies are separated by a minimum distance of 120 km in south-eastern New South Wales, but are divided by the Great Escarpment and montane plateau of the Monaro Plains. C. absimilis absimilis occurs east of the escarpment in moist coastal eucalypt open-forest, whereas C. absimilis edwardsi ssp. nov. occurs w'est of the escarpment in the dry box woodlands in the rain-shadow area of the Upper Snowy River south of the Great Dividing Range. The natural larval food plant of both subspecies throughout this region is Brachychiton populneus (Schott and Endl.) R.Br. (Braby 1998, and unpublished data), although at 13 km west of Cobargo, G. Guy (personal communication 2007) has also recorded C. absimilis absimilis breeding on several planted (non- indigenous) rainforest trees, viz: Macadamia integrifolia Maiden and Betche, Harpullia pendula Planchon ex F. Muell., and Stenocarpus sinuatas Endl. The latter two species have not previously been recorded as larval food plants for C. absimilis , although near Brisbane females were observed ovipositing on new shoots of Harpullia pendula (Braby 2000). Comparison of the spatial distribution of Brachychiton populneus in south-eastern Australia with that of C. absimilis shows a close geographical relationship (Fig. 36). The larval food plant is apparently absent from the Monaro Plains (between Bredbo and Bombala, New South Wales) west of the Great Escarpment, suggesting that the two butterfly subspecies are indeed isolated. Searches for C. absimilis edwardsi ssp. nov. by the author in the relatively dry woodland areas inland of the Great Dividing Range at Mount Majura and the Murrumbidgee River corridor, Australian Capital Territory, and the Wee Jasper district. New South Wales, where Brachychiton populneus grows abundantly failed to detect the butterfly. Presumably C. absimilis edwardsi ssp. nov. is also absent from the upper reaches of the Murray River inland of the Divide where the food plant likewise is present. Dunn and Dunn (1991) recorded C. absimilis absimilis from the Australian Capital Territory, based on a single female (lodged in the ANIC) collected from the Brindabella Range, but the phenotype of this specimen does not match specimens from south-eastern coastal New South Wales, suggesting the specimen is mislabelled. C. absimilis absimilis in south-eastern coastal New South Wales is otherwise known only from the coastal areas east of the Great Escarpment. Four specimens of Candalides absimilis recorded from three localities in the outer Melbourne region (Healesville, Dandcnong Ranges, Mordialloc) registered in the AMS and NMVare considered to be erroneous and accordingly their data excluded from the material examined. These specimens are labelled as follows: 1 cf “Dandcnong Ranges, Victoria | 3765”, "KL16576”, “G.A. Waterhouse Collection” (AMS); 19 “Dandcnong Ranges, Victoria | C. French, 3766”, “KL 16576”, “KL22003”, “G.A. Waterhouse Collection” (AMS); 19 “HEALESVILLE, J.A.K.” (NMV); 19 “MORDIALLOC, J.A.K.” (NMV). In addition, there is a female specimen in NMV reputedly from Victoria labelled “J.A. Kershaw, Collection, PURCH JUNE 1941”. In all cases, the females resemble typical C. absimilis absimilis in that they have a prominent white central patch on the upperside of each wing and do agree with populations of C. absimilis edwardsi ssp. nov. from eastern Victoria in which the white patches are absent. Waterhouse (1942), Tindale ( 1 965) and Dunn (1990) referred to this material in part and regarded the specimens as authentic, but Dunn and Dunn (1991) seriously questioned their validity and regarded the label data as unreliable. Dunn (1990) concluded that the specimens resulted from multiple accidental introductions to Melbourne but which failed to establish a resident breeding population, but he did not identify the mechanism or source of the introductions. However, the inadequate data on the labels and limited number of specimens suggest the material is probably mislabelled. Crosby (1951) recorded ''Candalides absimilis' from eastern Victoria based on observations at Gypsy Point, Victoria; however, these records refer to C. consimilis goodingi, the species of which was not widely recognised at that time and which had previously been confused with C. absimilis (Waterhouse 1942). Crosby (1951) also indicated that material of‘C. absimilis' had been collected from the northern reaches of the Macalister River north of Heyfield (by V. Smith), but this material was subsequently described by Tindale (1965) as Holochila goodingi, with Macalister River as the type locality. McEvey (1979) 44 Taxonomic review of Ccindalides Figs 37-57. Habitat and life histories of Candalides absimilis and C. margarita: 37, eucalypt woodland habitat of C. absimilis edwardsi ssp. nov. on the slopes ofthc Mitchell River, Vic., showing larval food plant Brachychiton populneits in deciduous form; 38, eucalypt woodland habitat of C. absimilis edwardsi ssp. nov. at McKillops Bridge, Snowy River, Vic., showing larval food plant Brachychitonpopulnens in foreground with new flush growth; 39. eucalypt woodland habitat of C. absimilis edwardsi ssp. nov. at Suggan Buggan. Vic; arrow indicates larval food plant Brachychiton populneits with new flush growth; 40, open woodland habitat of C. absimilis edwardsi ssp. nov. at Buchan Caves, Vic., showing larval food plant Brachychiton populneus with mature foliage; 41. Brachychiton populneits new leaf growth and flowers; 42. egg of C. absimilis edwardsi ssp, nov.; 43, egg of C. margarita gilberti ; 44, egg of C. margarita margarita ; 45, early instar larva of C. absimilis edwardsi ssp. nov.; 46-47, final instar larva of C. absimilis edwardsi ssp. nov. showing dorsal and lateral views and colour forms; 48-49. pupa of C. absimilis edwardsi ssp. nov. showing dorsal and lateral views; 50-51, final instar larva of C. margarita gilberti showing dorsal and lateral views; 52-53, pupa of C. margarita gilberti showing dorsal and lateral views; 54-55, final instar larva of C. margarita maigarita showing dorsal and lateral views; 56-57, pupa of C. maigarita margarita showing dorsal and lateral views. 45 M. F. Braby recorded C. consimilis goodingi from the Mitchell River during surveys in 1977-78; however, careful examination of this material (13 specimens in AMS) indicated a mixed series comprising two species - three specimens (399) of C. consimilis goodingi and 10 specimens (8cfcF, 299) of C. absimilis edwardsi ssp. nov. McEvcy surveyed three sites (Den of Nargun, The Amphitheatre on Billy Goat Bend, 1 km upstream from The Amphitheatre on Billy Goat Bend) over a three day period (7-9 December 1977). C. absimilis edwardsi ssp. nov. was recorded from each of these sites, but C. consimilis goodingi was collected only at The Amphitheatre, indicating that the two taxa occur together but presumably with limited overlap. The Mitchell River is the only known location where these two species are sympatric in Victoria. McEvey’s material from the Mitchell River thus appears to represent the first genuine collection of C. absimilis from Victoria. Habitat. In the Upper Snowy River area (Figs 38, 39), the breeding population occurs in dry eucalypt woodland, chiefly dominated by Eucalyptus albens Benth. (White Box) with some Callitris glaucophylla Joy Thomps. and L.A.S.Johnson (White Cypress Pine) and Eucalyptus melliodora A.Cunn. ex Schauer (Yellow Box) in the valley of the Snowy River, including banks of the river itself as well as gentle slopes several hundred metres above the river. In this habitat, the larva! food plant is sparsely distributed over a relatively wide area, often occurring as single isolated trees, compared with those growing on the slopes of the Mitchell River (Fig. 37). The area lies in a pronounced rainfall shadow in which the mean annual rainfall (c. 800 mm) is much lower than the coastal areas of New South Wales of similar latitude further east. At Buchan Caves Reserve (Fig. 40), the breeding population occurs in open woodland on limestone outcrops on exposed slopes with a north or west facing aspect, or along the valley floor of Fairy Creek, where the larval food plant grows. The larval food plants vary greatly in size and age, from very old large trees to young saplings only a few metres high and, within the Reserve itself, occur in relatively low density (a total of only 16 plants were located). A few scattered trees (mostly ornamental garden and street trees) also occur in residential areas of Buchan close to the Reserve where the butterfly also breeds. At Mitchell River National Park (Fig. 37), the breeding population occurs in woodland dominated by Brachychiton populneus and Eucalyptus melliodora, with a sparse understorey of shrubs and small trees of Acacia implexa Benth. (Lightwood) and A. mearnsii De Wild. (Black Wattle), and grasses. The habitat occurs as a narrow strip on both sides of the Mitchell River in a limited zone where the larval food plant grows abundantly on steep rocky slopes between the top of escarpment and the bottom of the gorge/ river valley (i.e. Mitchell River and its tributaries, such as Woolshed Creek). This habitat type is patchy in distribution because the food plants generally grow on rocky slopes with a predominantly a north to west facing aspect where conditions are drier and more exposed. As a result, the larval food plants are more concentrated and less widely dispersed than those occurring in the Upper Snowy River area. Warm temperate rainforest dominated by Tristaniopsis laurina (Sm.) Peter G.Wilson and J.T.Waterh., Acmena smithii (Poir.) Merr. and L.M.Perry, Elaeocarpus reticulatus Sm., Pittosporum undulation Vent, and Myrsinehowittiana (Mcz) Jackcs prevails along the gorges; however, the larval food plant does not grow, and the butterfly does not breed, in this habitat. Biology. Larval food plant. Brachychiton populneus (Schott and Endl.) R.Br. (Sterculiaceae) (Kurrajong). The larval food plant typically grows as a tree up to 10 m in height, with a dense, dark green crown of foliage. Long¬ term observations made on the phenology of Brachychiton populneus in south-eastern Australia by the author indicated that the trees flowered and regenerated their foliage over a short interval during summer, mainly in December and January, although some trees produced new flushes of growth as late as February. Although foliage regeneration was very seasonal and occurred rapidly among individual trees, not all trees regenerated synchronously, and there was a succession of new annual leaf growth during the summer period. A proportion of trees were also deciduous (Fig. 37) - some plants dropped all of their leaves during late spring and early summer before regeneration, especially in dry years or those growing in the more exposed situations. Other trees produced new growth before shedding most of their older mature leaves, and thus remained vegetated. Behaviour. Eggs (Fig. 42) were laid singly or sometimes in pairs, usually on the stem immediately below a terminal branchlet of new developing leaf buds, or on the new soft shoots and stipules of the young developing leaves. Many eggs were laid on individual trees, but females oviposited only on trees producing flushes of new growth or, for those trees that had shed all of their older leaves, in the early phases of regeneration; they generally did not oviposit on trees in which the leaves had reached an advanced phase of regeneration. Females were observed to oviposit on the same tree over a minimum period of 10 days. Individual trees probably remain suitable foroviposition for about two to three weeks due to the rapid growth of new foliage. Early instar larvae (Fig. 45) fed singly on the new soft leaves (Fig. 41) at night, resting by day on the underside of the young leaves where they avoided direct sunlight. Usually only one larva occurred on a single terminal cluster of foliage, but occasionally up to four larvae were found. The later instar larvae, when not feeding, typically rested lower down on the petiole of a leaf where they remained well concealed, often in the fork between the petiole and the branchlet. Between instars, the larvae settled on the leaf petiole or further down on the thicker stems of the branchlet to moult. In the late instars (Figs 46, 47), larvae varied considerably in colour, from uniformly green to almost entirely reddish with some green at the posterior end; others were intermediate between these two phenotypic 46 Taxonomic review of Candalides extremes, being green with a red mid-dorsal line and a red lateral line and a slight pinkish tinge at the anterior end. Variation in larval colour was associated with the colour of juvenile foliage, which was either bright red or bright green (Fig. 41). Larvae were not observed attended by ants, despite the abundance of Crematogaster sp. ants on many trees. Larvae, when reared in captivity, showed a preference to pupate in well-concealed, dark situations, such as inside rolled bark. The pupa was attached, to a silken platform spun over the substrate, by anal hooks and a silken central girdle. Pupae (Figs 48, 49) were noted to stridulate when stimulated (e.g. when gently stroked or sprayed with water). In the field, pupae were not found on the foliage, flowers or trunk and branches of the larval food plant, and it is likely that the larvae leave the food plant to pupate amongst dry leaf litter (dead leaves, bark etc) or under rocks, stones or inside rock crevices at the base of the tree. In captivity, adults always emerged during the morning, well after sunrise, but usually before midday. After emergence, adults quickly expanded their wings and were ready for flight within 30-60 mins. In the field, males were noted to fly rapidly throughout much of the day in the mid- to upper canopy, usually between 2-8 m above the ground, around the outer foliage of larger trees of the larval food plant. They typically patrolled around a tree and then quickly flew to the next; they were also observed to fly around the outer foliage of Pittospomm undulatum, the leaves of which arc superficially similar to Brachychitonpopulneus. Males rarely alighted to settle, especially when many individuals were present, but once settled, they usually perched for a few seconds, or sometimes for up to about 1 min, high up on horizontal leaves of the outer foliage of the larval food plant. When fewer individuals were present, males perched for longer periods on the upper foliage, especially during the late afternoon, to establish territories. In these situations, males would leave their perch briefly to patrol around the crown of the tree before settling again, usually in a different place on the tree. When settled, the wings remained closed or half opened at an angle of about 90° towards the sun. The patrolling and perching behaviour of males suggest that newly emerged females seek the mid- to upper canopy of the larval food plant for mate location. Females were usually observed on or near the larval food plant. Their flight was considerably slower than that of males but, like males, they sometimes settled on the outer foliage of the food plant, with wings opened up to 90° towards the sun. They were observed ovipositing from late morning to early afternoon (1000-1415 hrs EST) during warm, sunny weather. Multiple females were occasionally observed to oviposit simultaneously on the same tree. Life cycle. Adults have been collected from December to February. At several locations, early instar larvae were found in early December which, given the developmental time of the immature stages (see below), indicates that females were ovipositing in November. Thus, the putative flight period of Candalides absimilis edwardsi ssp. nov. is from late November to late February, which coincides with the period of seasonal flushes of new growth of the larval food plant. In captivity, eggs hatched about a week after being laid, and the larvae completed development in about 3.5 weeks. Pupal developmental time was more variable: from a sample of 52 pupae, most (81%) reared from larvae collected in December and January developed directly, with the pupal duration lasting 11-14 days, but the others (19%) remained dormant for two months or for 6-10 months, indicating presence of a facultative diapause. The life cycle, from egg to adult for directly developing individuals, was completed in approximately six weeks. These observations suggest that most of the population of the subspecies is bivoltine, with at least two overlapping generations during the flight season. Conservation status. Candalides absimilis edwardsi ssp. nov. has a narrow geographical range, with an estimated extent of occurrence of 3,200 km 2 . It is known from eight sites (> 1 km apart) representing five locations (>10 km apart) in three disjunct areas. Although the taxon is known from a limited area, there is no evidence of decline, recent fragmentation of populations, or threatening processes impacting on the known sites or habitat, and all known populations occur in conservation reserves. Its conservation status should therefore be regarded as Least Concern (LC) according to IUCN criteria. At Buchan Caves, the low number of trees of Brachychiton populneus suggests the butterfly population may be small, and further planting of B. populneus is recommended to augment the existing population. Dunn et al. ( 1994, pp. 236-237) provisionally regarded Candalides absimilis edwardsi ssp. nov. as secure on the basis that the known site was protected within a single national park, but cautioned that the population “...may become threatened because of their perching behaviour on bushes along the escaqrment where adults could be taken in large numbers by hand or with a short handled net.” and that “Regular wildfires through the main breeding colony could be a threat and females appear to be associated with remnant warm temperate rainforest patches which have escaped burning.” Both of these statements in relation to threatening processes must be seriously challenged. First, the breeding area at Mitchell River, based on the extent and abundance of the larval food plant, is concentrated along the steep rocky fire-prone escarpment between the rainforest gully and woodland plateau. Moreover, not all adults perch along the escarpment, and there is little documented evidence that butterfly collecting per se is a threatening process, except perhaps in those situations where populations have already been substantially reduced and fragmented by habitat change (New 1991). Second, the bivoltine life cycle and behaviour of larvae, which undoubtedly leave the food plant to pupate in concealed situations near or below the ground, suggests the species would be well protected from fire. 47 M. F. Braby Candalides margaritagilberti Waterhouse, 1903 stat. rev. (Figs 19-22, 24, 25,29,30) Candalides gilberti Waterhouse, 1903a, p. 181; Waterhouse 1903b, p. 23; Waterhouse and Lyell 1914, pp. 78-79, pi. 15; Waterhouse 1932, p. 129, pi. 19; Common 1964, p. 122; Peters, 1971, p. 30; Common and Waterhouse 1972, p. 413, pi. 40; Common and Waterhouse 1981, pp. 527-528, pi. 47; Common and Waterhouse 1982, pp. 301- 302, pi. 26; Puccetti 1991, p. 144, 146; Dunn and Dunn 1991, pp. 389-390; Samson and Wilson 1995, pp. 71-73; Edwards 1996, p. 252; Daniels and Edwards 1998, p. 90; Braby 2000, pp. 757-758, pi. 54; Edwards et al. 2001, p. 139; Braby 2004, p. 258; Meyer et al. 2006, p. 13; Franklin 2007, pp. 12-13. Holochila gilberti (Waterhouse). - Tite 1963, p. 205; Tindale 1965, p. 173; McCubbin 1971, p.71; D’Abrera 1971, pp. 366-367. Comments on synonymy. Holochila margarita was described by Semper (1879) based on an unspecified number of specimens, but he did not designate a holotype. Gabriel (1932) gave the type locality as Bowen, Queensland, and he appears to be the first author to have referred to a type. Edwards et al. (2001) interpreted Gabriel’s (1932) reference to a ‘holotype’ as a lectotype designation. The lectotype male is illustrated in Figures 15-16 and 23; a typical example of the female of Candalides margarita margarita (Figs 17, 18) is also illustrated for comparison. C. margarita gilberti was originally described as a distinct species by Waterhouse (1903a) based on a pair of specimens (Figs 19-22) collected from Darwin (given as Port Darwin), Northern Territory, by Gilbert Turner. Although both syntypes in AMS possess Waterhouse’s labels indicating they are type specimens (Figs 24,25), Waterhouse (1903a) did not designate a holotype. Both specimens were subsequently illustrated by Waterhouse and Lyell (1914, p. 79) who noted that “The figures arc drawn from the types, both of which are in poor condition.” Tindale (1965, p. 173) remarked that “The type is from Port Darwin...”, but did not refer to a holotype. Peters (1971) referred to the syntype male as the ‘holotype’ and the syntype female as the ‘allotype’. Edwards et al. (2001) interpreted Tindale’s (1965) reference to a type as a lectotype designation, but this interpretation must be rejected since it is not clear which of the two syntypes Tindale was referring to. In this work, Peters (1971) incorrect reference to a holotype is regarded as a lectotype designation according to Article 74 of the 1CZN (International Commission of Zoological Nomenclature 1999). In his extensive revision of the genus, Tite (1963) regarded C. maria Bethune-Baker, 1908 from Waigeo, Misool, Aru and mainland New Guinea to be the Papuan subspecies of C. margarita based on similarities in the male genitalia, and this arrangement was followed by Parsons (1998). Waterhouse (1903a) drew attention to the distinct underside markings, especially the postmedian series, and the reduced white areas on the upperside of the female by which Candalides margarita gilberti was distinguished. He considered C. margarita gilberti to be most closely allied to C. absimilis, but did not compare the taxon with C. margarita maigarita. Tite (1963) considered C. margarita gilberti to be specifically distinct based on differences in the male upperside ground colour, shape of the fore wing apex, and extent of black spots on the underside of the hind wing, with C. maigarita gilberti being pale lavender-blue, having the fore wing apex noticeably produced, and with sharp black spots on the underside. He also drew attention to the distinct upperside colouration of the female, which is pale lavender-blue with the fore wing costal margin broadly black, but without darkened veins. Tite (1963, p. 205) regarded C. margarita gilberti to be closely related to C. margarita, C. tringa (Grose-Smith, 1894) and C. biaka (Tite, 1963) on account of similarities of the male genitalia, but remarked that “Surprisingly.. .the male genitalia arc identical with those of maigarita.'” Parsons (1986), however, considered C. maigarita sensu stricto to be closely allied to C. afretta Parsons, 1986 from the lowland savanna belt of the Western Province of Papua New Guinea, but he did not refer to C. maigarita gilberti. Braby (2000, pp. 757-756) stated that “C. gilberti is very closely related to, and possibly conspccific with, C. margarita ” and recommended that “Further study is needed to determine if C. gilberti is merely a subspecies of C. maigarita.” Type material. LECTOTYPE - cf labelled “Port Darwin, 24 Nov. 1902, G. Turner | L2008”, “C. gilberti cf. Type specimen | KL21869”, “G.A. Waterhouse Collection”, “FIG. 266 Upperside IN ‘THE BUTTERFLIES OF AUSTRALIA', by WATERHOUSE & LYELL, was taken from this specimen | KL21869”, “K191348” (AMS). PARALECTOTYPE - 9 labelled “Port Darwin, 25 Nov. 1902, G. Turner | L2009”, “C. gilberti 9 , Type specimen | KL2I870”, “G.A. Waterhouse Collection”, “FIG. 267 Upperside IN ‘THE BUTTERFLIES OF AUSTRALIA’, by WATERHOUSE & LYELL, was taken from this specimen | KL21870”, “K191349” (AMS). Other material examined. 58cfcf, 3999. WESTERN AUSTRALIA: 1 cf labelled “16.31S 125.16E, Synnot Ck W.A., 17-20 Jun. 1988, T.A. Weir” (ANIC); 19 labelled “14.25S 126.38E 12 km S of Kalumburu Mission W.A., 7-11 June 1988, T.A. Weir” (ANIC). NORTHERN TERRITORY: 19 labelled “48 mi. SW. of Daly River, NT. 14.11S, 130.08E, 3 Sept. 1968, M. Mendum” (ANIC); 4eftf labelled “60 km S of Daly River N.T., 12 June 1981, D.P. Sands” (ANIC); 1 9 labelled “DALY R N.T., 30 JUNE 1969, JC LE SOUEF” (ANIC); I cf labelled “16°07’39”S, 130° 19’00”E, Gregory NP, NT, 31 JUL. 2006, M.F. Braby”, “NT Museum 1004226” (NTM); 1 cf labelled “N. Territory, BunellsTrig., I3°30’S, 131°02’E, 244m, 23 July 1983, D.P. Sands” (ANIC); 1 cf labelled “Mt Burrill [sic] NT, 18 Jan 1992, R.N. STOODLEY”, “L.R. Ring Collection” (ANIC); 1 cf labelled “13°29’46”S, 131 °02’08”E, Burrells Trig, NT 250m. 6 MAY 2006, M.F. Braby & DA Young”, “MFB 48 Taxonomic review of Candalides Collection 00175” (MFBC); 1 cf labelled similarly but with registration number “MFB Collection 00083” (MFBC); 3cfcf, 299 labelled “P. Darwin, F.P. Dodd | 4918-4922”, “KL21874”, “G.A. Waterhouse Collection” (AMS); 19 labelled “P. Darwin, 13 Sep. 08, F.P. Dodd”, "Passed through C.W. Wyatt Theft-Coil. 1946-1947”, “LEP-9746” (NMV); 19 labelled “P. Darwin, Nov. 08, F.P. Dodd”, “Passed through C.W. Wyatt Theft-Coil. 1946-1947”, “LEP-9745” (NMV); 19 labelled “P. Darwin, Bd. Nov 08, F.P. Dodd | 4349”, “KL21871”, “G.A. Waterhouse Collection”, pupal exuvia mounted separately adjacent to specimen and labelled similarly (AMS); 19 labelled “P. Darwin, Bred Jan 09, F.P. Dodd”, “Passed through C.W. Wyatt Theft-Coil. 1946-1947”, pupal exuvia mounted separately adjacent to specimen and labelled similarly (AMS); 1 cf labelled“P. Darwin, Jan 09, F.P. Dodd”, “From F.P. Dodd, N.Q. Land, 12.12.10”,“LEP-9739”(NMV); 1 cf labelled “P. Darwin, Feb. 09, F.P. Dodd”, “LEP-9740” (NMV); 19 labelled similarly but with registration number “LEP-9743” (NMV); 19 labelled similarly but with registration number “LEP9-747” (NMV); 2cfcf, 19 labelled similarly but with additional labels “KL21872”, “G.A. Waterhouse Collection” (AMS); lcf labelled similarly (ANIC); 19 labelled similarly but with additional label “Figured in “AUSTRALIAN BUTTERFLIES” Jacaranda Press, 1964, I.F.B. Common” (ANIC); 1 cf labelled “P. Darwin, Mch 09, F.P. Dodd”, “Passed through C.W. Wyatt Theft-Coil. 1946-1947” (AMS); 1 cf labelled “P. Darwin, Apr. 09, F.P. Dodd”, "Passed through C.W. Wyatt Theft- Coil. 1946-1947”, “LEP-9741” (NMV); 19 labelled “P. Darwin, Apl 09, F.P. Dodd | 4464”, “KL21873”, “G.A. Waterhouse Collection” (AMS); 1 cf labelled “P. Darwin, May 09, F.P. Dodd”, “Passed through C.W. Wyatt Theft- Coil. 1946-1947”, “LEP-9742” (NMV); 1 cf labelled “P. Darwin, May 09, F.P. Dodd”, “LEP-9737” (NMV); lcf labelled similarly but with registration number "LEP-9738” (NMV); 19 labelled similarly but with registration number “LEP-9744” (NMV); 19 labelled similarly but with registration number “LEP-36622”, “Collection A.N. Bums” (NMV); 19 labelled “Darwin, Nov., Purcell", “Label data very doubtful, 22.10.68 J.V.P.”, "Passed through C.W. Wyatt Theft-Coil. 1946-1947” (AMS); 19 labelled “Darwin NT, Jan 1949”, “F.E. Parsons Collection, Donated A.N.l.C. 1967” (ANIC); 1 cf labelled “AUSTRALIA, NT DARWIN, Stuart Pk, T. FENNER ex D. Wilson”, “larva on Decaisnina signata. cm. 31 May 91” (ANIC); 1 cf labelled “Stuart Park N.T., 18 June 1991, D.N. WILSON”, “NORTHERN TERRITORY MUSEUM, 1003598 LYCAEN1DAE, Candalidesgilberti, Waterhouse, 1903, Det; D.N. Wilson" (NTM); 19 labelled “Stuart Park N.T., Ex pupa 24 June 1991, D.N. WILSON”, “NORTHERN TERRITORY MUSEUM, 1003599 LYCAENIDAE, Candalides gilberti, Waterhouse, 1903, Det: D.N. Wilson” (NTM); 19 labelled “Stuart Park N.T., Ex pupa 24 June 1991, D.N. Wilson”, “G. Daniels Collection” (AMS); 299 labelled “Stuart Park N.T., Ex pupa 25 June 1991, D.N. Wilson” (ANIC); 19 labelled similarly but with additional label “Figured in Butterflies of Australia (1999) [sic], CSIRO Publishing, M. F. Braby” (ANIC); 19 labelled “Egg collected 13 May 1991, Egg hatched 15 May 1991, Larvae pupated 31 May 1991, Adult emerged 10 July 1991, Nudle Street Stuart Park N. T., D.N. WILSON”, “Figured in Butterflies of Australia (1999) [sic], CSIRO Publishing, M.F. Braby” (ANIC); lcf labelled “Stuart Park NT, 07 Apr 1992, D.N. Wilson”, “G. Daniels Collection” (AMS); 19 labelled "NIGHTCL1FF, NT, 23 AUG 92, C. M EYER XP”, “G. Daniels Collection” (AMS); lcf labelled “NIGHTCLIFF, NT, 3 SEP 92, C. MEYER XP”, “G. Daniels Collection”, pupal exuvia pinned beneath specimen (AMS); lcf labelled “Darwin, N.T., Coconut Grove, 2 OCT. 1994 B., C. Meyer". “MFB Collection 00079”, “genitalia No. 014” (MFBC); lcf labelled “Fannie Bay, Darwin, N.T., R.P. WeirXP, 3.8.98”, “NORTHERN TERRITORY MUSEUM, 1003481 LYCAENIDAE, Candalides gilberti, Waterhouse, 1903, Det: D.F. Trembath 16 Mar 2006” (NTM); lcf labelled “Darwin, NT. emg. 28 FEB. 2005, M.F. Braby & R. Weir, The Gardens”, “reared from larva on flowers of Decaisnina signata, coll. 2 FEB. 2005”, “MFB Collection 00174”, “genitalia No. 023” (MFBC); 1 9 labelled “Darwin, NT, emg. 28 FEB. 2005, M.F. Braby & R. Weir, The Gardens”, “reared from larva on flowers of Decaisnina signata, coll. 2 FEB. 2005”, “MFB Collection 00176” (MFBC); 19 labelled similarly but with date “emg. 8 MAR. 2005” and registration number “MFB Collection 00080”, pupal exuvia pinned beneath specimen (MFBC); 19 labelled “Stuart Park. Darwin, NT, emg. 26 MAY 2006, D.A. Young, reared from larva on Decaisnina signata", “MFB Collection 00178” (MFBC); 1 cf labelled similarly but with date “emg. 20 MAY 2006” and registration number “MFB Collection 00081” (MFBC); lcf labelled “12°22’S, 130°53’E, Leanyer, Darwin, NT, emg. 26 JUL. 2006, M.F. Braby”, “reared from larva on Decaisnina signata, coll. JUL. 2006”, “MFB Collection 00171” (MFBC); 19 labelled similarly but with date “emg. 31 JUL. 2006” and registration number “MFB Collection 00177” (MFBC); 1 cf labelled similarly but with registration number “MFB Collection 00082” (MFBC); 19 labelled similarly but with date “emg. 2 NOV. 2007”, “reared from larva on foliage of Decaisnina signata, coll. OCT. 2007” and registration number “MFB Collection 00179” (MFBC); 19 labelled similarly but with date “emg. 4 NOV. 2007” and registration number “MFB Collection 00180” (MFBC); 1 cf labelled similarly but with registration number “MFB Collection 00172” (MFBC); lcf labelled similarly but with date “emg. 24 APR. 2008”, “Reared from larva on foliage of Decaisnina signata, coll. APR. 2008” and registration number “MFB Collection 00187” (MFBC); lcf labelled“12.37308°S, 130.88657°E, Wanguri, Darwin, NT, emg. 10 FEB. 2007, M.F. Braby”, “reared from egg on Decaisnina signata, coll. 12 JAN. 2007”, “MFB Collection 00173”, pupal exuvia pinned beneath specimen (MFBC); 19 labelled “21 August, 1970., J. V. Peters, Howard Springs, N.T., 16 mis, South Darwin.” (AMS); lcf labelled “Tidy 49 M. F. Braby Hill 233’ N.T., 12°45’S. 130*54’E., 25 August 1970., J.V. Peters” (AMS); lcf labelled "13.21S 131.08E, Hill above Robin Falls N.T., 8 June 1993, E.D. Edwards” (AN1C); 19 labelled “Marrakai Rd, N.T., R.P. Weir XP, 26.10.97”, “NORTHERN TERRITORY MUSEUM, 1003480 LYCAENIDAE, Candalides gilberti, Waterhouse, 1903, Det: D.F. Trembath 16 Mar 2006” (NTM); 1 cf labelled “20 KM STH ADELAIDE R. NT., 23 FEB. 1992, RN STOODLEY”, “L.R. Ring Collection” (ANIC); lcf labelled “30 KM STH PINE CK NT, 14 APRIL 1992, D.N. WILSON”, “Figured in Butterflies of Australia (1999) [sic], CSIRO Publishing, M.F. Braby” (ANIC); 19 labelled “12.35S 131.17E, Fogg Dam N.T., 15 June 1981, D.P. Sands” (ANIC); lcf labelled similarly but with date “17 June 1981” (ANIC); 2cfcf labelled “Mt. Bundey - Mary R. District. N.T., 14 May, 1972. J. Kerr” (ANIC); 1 9 labelled “11.09S 132.09E, Black Point Cobourg Pen., NT, 26 Jan. 1977, E.D. Edwards” (ANIC); 19 labelled “3 mi. S. of Oenpelli, NT., 11 Dec. 1970, W. Omer Cooper” (AMS); lcf labelled “South Alligator RiverN.T., 15 May 1981, D.P. Sands” (ANIC); 1 cf labelled “Little Menagerie Rock, N.T., 16 May 1981, D.P. Sands” (ANIC); lcf labelled “12.48S 132.42E, Nourlangie Creek, 8 km N of Mt. Cahill, N.T., 19 Nov. 1972, E.D. Edwards & M.S. Upton” (ANIC); lcf labelled “12.50S 132.5 IE, 15km NE of Mt. Cahill, N.T., 23 May 1973, E.D. Edwards & M.S. Upton" (ANIC); lcf labelled “12.86488°S, 132.70468°E, Mt Cahill (Mirrai lookout) Kakadu NP, NT, 3 MAY 2008, MF Braby & LJ Aitchison”, registration number “MFB Collection 00186” (MFBC); lcf labelled 12°39’19”S, 135°51’31”E, Dbamiyaka outstation, NE Arnhem Land, NT, 14 AUG. 2007, M.F. Braby & I. Morris”, “NT MUSEUM 1004303” (NTM); I cf labelled “Hideway Motel, Gove, 17.iv.l976,A.J. DartncU”, “NORTHERN TERRITORY MUSEUM, 1003479 LYCAENIDAE, Candalides gilberti , Waterhouse, 1903, Det: DF Trembath 16 Mar 2006” (NTM); 19 labelled “12°10’41”S, 136°47’01 ”E, Nhulunbuy, NT, emg. 12 FEB. 2008, MF Braby & LJ Aitchison”, “Reared from larva on Decaisninasignata, coll. 27 JAN. 2008”; “MFB Collection 00181”, pupal exuvia pinned beneath specimen (MFBC). QUEENSLAND: 3cfcf, 299 labelled “18.35S 138.03E, Murray’s Spring 8 km W by N of Musselbrook Camp Q, 9 May 1995 GPS, E.D. Edwards” (ANIC); 3cfcf labelled similarly but with date “11 May 1995” (ANIC); lcf labelled “18.38S 138.08E, Gorge 2 km S of Musselbrook Camp Q, 20 May 1995 GPS, E.D. Edwards” (ANIC); lcf labelled “18.42S 138.29E, Lawn Hill Ck, Q, 17 May 1995 GPS, E.D. Edwards” (ANIC). Comparison. Adults of Candalides margarita gilberti are very similar to those of C. maigarita margarita in wing shape and pattern, although in C. margarita gilberti the underside markings are more distinct, and the fore wing has a straighter termen with the apex more pointed. The upperside ground colour of males of the two subspecies is similar, but in C. maigarita gilberti the colouration is somewhat paler and the black margin is narrower than in C. maigarita maigarita. C. maigarita gilberti females have narrower black margins with the large white central patches on the upperside replaced with blue. A characteristic feature of C. maigarita gilberti noted by Waterhouse (1903a), that the postmedian spot between veins Rs and M, displaced proximally from the remainder of the postmedian series on the underside of the hind wing, is also shared with C. margarita margarita but not by other members of the C. absimilis species group from Australia. Remarks. The structure of the male genitalia of Candalides maigarita , which is substantially different and more complex than those of C. absimilis , is remarkably uniform across the species range (Figs 29-32). Although the genitalia are slightly smaller from the Top End and western Gulf Country (C. maigarita gilberti) compared with those of from the Wet Tropics and north-eastern New South Wales (C. maigarita maigarita), there are no significant differences in the form of the valvae or brachia. The female genitalia (not illustrated) show a similar geographical pattern to that of the males. Larvae of both Candalides margarita gilberti and C. maigarita margarita specialise on loranthaceous food plants (Smales and Ledward 1943; Samson and Wilson 1995), and comparison of the immature stages of the two taxa indicate that the morphology of the egg (Figs 43,44), final instar larva (Figs 50, 51, 54, 55) and pupa (Figs 52, 53,56,57) are identical. Comparison of the first instar larva of the two taxa also revealed no differences. The first instar larva is characterised by a pronounced dorsal ridge along the length of the body, which bears a series of long paired dorsal primary setae on the metathorax and abdominal segments I -7; on each segment lies a pair of smaller setae posterior to the longer paired setae. The mesothorax also possesses a pair of long setae, but they are located dorsolaterally so that they are not as close together as the setae on the other segments. Final instar larvae of the two subspecies vary in pattern and colouration depending on the colour of the food consumed (red flowers or green leaves), although in C. maigarita gilberti the darker patches on the body appear to be more pronounced. In summary, evidence from the genitalia, adult and immature stage morphology, together with similarities in the life history and adult behaviour (see below), indicate that gilberti is conspecific with Candalides margarita ; gilberti should be regarded as a geographical subspecies of that taxon on account of its disjunct distribution and morphological divergence in the adult stage. Thus, three subspecies are now recognised within C. margarita : C. maigarita maria from mainland New Guinea and the nearby islands of eastern Indonesia; C. maigarita gilberti from north-western and central northern Australia; and C. maigarita maigarita from eastern Australia (Thursday Island, Queensland, to Tuncurry, New South Wales). Presumably, C. margarita gilberti is sister to C. margarita margarita + C. margarita maria based on the greater phenotypic divergence of C. maigarita gilberti and the close 50 Taxonomic review of Candalides similarity among the females of C. margarita margarita and C. margarita maria. Distribution. Candalides margarita gilberti occurs in the monsoon tropics of north-western and central northern Australia, from the Kimberley, Western Australia, across the Top End, Northern Territory, to the Gulf Country of north-western Queensland (Braby 2000; Franklin 2007). It is allopatric from C. margarita margarita, the known ranges of the two subspecies in Queensland being separated by a minimum distance of about 700 km. C. margarita gilberti reaches its northernmost limit at Black Point on Cobourg Peninsula (Garig Gunak Barlu National Park), Northern Territory, about 550 km south-south-west of Aru Island in the Arafura Sea which supports the nearest population of C. margarita maria. Habitat. Candalides margarita gilberti, in contrast to C. margarita margarita, occurs in savanna woodland and patches of monsoon forest and vine-thicket where the larval food plant grows. In Darwin, it is widespread and the early stages may be relatively common in suburban areas, parklands and areas with native vegetation where the subspecies breeds throughout the year. Biology. The larval food plant, description of the early stages, general biology and habitat of Candalides margarita gilberti have been well documented (Samson and Wilson 1995; Braby 2000). The recorded larval food plant is the aerial-stem hemiparasite Decaisnina signata (F.Muell. ex Benth.) Tiegh. (Loranthaceae) (Samson and Wilson 1995), but eggs have also been found on Decaisnina petiolata (Barlow) Barlow at Kakadu National Park (M.F. Braby unpublished data). Although the life history was documented only relatively recently, Waterhouse and Lyell (1914, p. 79) noted earlier that “The species has since been bred by Mr. F.P. Dodd...” Examination of the Dodd material collected from Darwin during 1908-09 in AMS,ANICandNMV(13cfcf, 139 9) revealed two reared female specimens each with their pupal exuviae mounted separately and labelled “P. Darwin, Bd. Nov 08, F.P. Dodd” (AMS) and “P. Darwin, Bred Jan 09, F.P. Dodd” (AMS); however, no information on the identity of the larval food plant was provided on the labels of these specimens. Little has been documented on adult behaviour of Candalides margarita gilberti, but the following observations suggest it is similar, if not identical, to that of C. margarita margarita (M.F. Braby unpublished data). In the Northern Territory, C. margarita gilberti males regularly congregate on hilltops where they perch on foliage of trees to defend territories during the morning (from 7.45 am onwards) and the afternoon. When settled, they usually perch, with wings closed, 3-5 m above ground level. Males also establish territories along edges of gallery forest where they typically fly and perch, with wings closed and oriented head downwards towards the sunlight, on outer foliage of rainforest trees about 5-8 m above ground level during early to late afternoon. Females have been observed flying in similar places, but when settled frequently perch with wings opened at about 90° towards the late afternoon sun. They have been observed ovipositing on the larval food plant during the early afternoon. Conservation status. The conservation status of Candalides margarita gilberti should be regarded as Least Concern (LC) on account of its broad distribution and lack of threatening processes. No populations are known to have been extirpated and there is no evidence that the subspecies has declined or is in decline. However, its relative abundance in savanna woodland across the monsoon tropics of the Kimberley and Top End may have declined over the past century with changed fire regime because mistletoes, the larval food plant of the butterfly, are highly susceptible to fire (Kelly et al. 1997). In these fire-prone habitats, there has been a general increase in the extent, frequency and intensity of fires, in contrast to the general suppression of fire since European settlement in agricultural areas of temperate southern Australia where some mistletoe species appear to have increased in abundance (Kelly et al. 1997). DISCUSSION Geographical variation in Candalides absimilis males parallels that of C. consimilis males (Tindale 1965; Kerr 1967), both species of which are broadly sympatric along the eastern coast of Australia. Not only do these species diverge at the extreme ends of their geographical ranges, but they also exhibit similar patterns of phenotypic divergence, particularly on Cape York Peninsula where they both have narrow range endemic subspecies restricted to the mid¬ peninsula rainforest block (Iron Range-Mcllwraith Range). Males of both C. absimilis eastwoodi and C. consimilis toza, which are superficially similar, are characterised by possessing a paler upperside purple ground colour, a straightcr termen of the fore wing, a more prominent black tomal spot on the hind wing, and less distinct markings on the underside compared with their respective nominate subspecies. This suggests that the northernmost populations of the two species of Candalides, C. absimilis eastwoodi and C. consimilis toza, have diverged under similar selective pressures in both space and time. The mid-peninsula block of rainforest on Cape York Peninsula is well known for its high level of local endemism and insularity (Kikkawa etal. 1981), and the biogeographical patterns observed in the two species of Candalides suggest these butterflies have been subjected to the same underlying historical processes that have led to isolation and differentiation. Interestingly, differences observed between males of the two Australian subspecies of Candalides margarita parallels the pattern of geographical variation observed in fore wing shape in C. absimilis and C. consimilis males in which the subspecies from the Kimberley and Top End, C. margarita gilberti, has the termen straighter and apex sharply pointed, similar to that observed in C. absimilis 5 M. F. Braby eastwoodi and C. consimilis toza. C. margarita gilberti also has more prominent black tomal spots on the hind wing underside than in C. margarita margarita, but the underside basal spots and wavy lines are more distinct, whereas in C. absimilis eastwoodi and C. consimilis toza the underside markings are less distinct than their respective nominate subspecies. In C. margarita, the geographical variation is partitioned across the ‘Carpentarian Gap’, a barrier comprising the Gulf of Carpentaria and arid plains of the Gulf Country between the Top End and Cape York Peninsula, whereas in C. absimilis and C. consimilis geographical variation among the northernmost populations is partitioned across a putative barrier comprising the dry lowland plains south of Princess Charlotte Bay of Cape York Peninsula (i.e. between Iron Range-Mcllwraith Range and the Wet Tropics). Although the blue colouration differs between females of Caudal ides absimilis edwardsi and C. margarita gilberti, the upperside pattern of these two subspecies is broadly similar in that the white central patches are absent and the basal and central areas of both wings are broadly blue. Indeed, these two subspecies are remarkable in that they comprise the only taxa in the C. absimilis species group in Australia in which the white patches are replaced with blue. The only other species in the C. absimilis species group that possess this trait are C. riuensis (Tite, 1963) from Tagula (Sudest) Island, C. lamia (Grose-Smith, 1897) from the D’Entrecasteaux Islands, and C. coeruleus (Rober, 1886) from Aru Island and the lowland areas (< 200 m) of mainland New Guinea (Parsons 1998). These taxa, together with C. cyprotus , are exceptional as females of most members of the C. absimilis species group have the upperside ground colour black with contrasting patches of white in the central areas of the wings, and their morphological pattern poses interesting questions as to how this character may have evolved, its functional significance and whether the ‘blue form’ represents a plesiomorphic (ancestral) or apomorphic (derived) state. Answers to such questions may come from comparative analysis of female phenotype in relation to the species’ ecology and phylogenetic history of the genus. For instance, among the Australian taxa, both C. absimilis edwardsi and C. margarita gilberti are associated mainly with habitats comprising eucalypt woodland, open woodland and savanna woodland (i.e. vegetation types in which the projective foliage cover < 30%), whereas C. absimilis absimilis, C. absimilis eastwoodi, C. maigarita margarita, C. helenita (Semper, [1879]), C. consimilis consimilis and C. consimilis toza are associated primarily with closed forest habitats, particularly rainforest (i.e. vegetation types in which the projective foliage cover > 70%) (Braby 2000). Although C. margarita gilberti also occurs in dry rainforest (monsoon forest), this vegetation type in the Kimberley and Top End, compared with the evergreen tropical forests of eastern Australia, occurs as a vast network of smaller patches (mostly < 5 ha in extent) that is naturally fragmented throughout the extensive savanna landscape so that these habitats have greater edge effects (i.e. larger perimeter/area ratios) where the canopy is more open and, depending on water availability, the dominant canopy trees are frequently deciduous or semi-deciduous during the dry season. Thus, the ‘blue forms’ in female Candalides are primarily associated with open habitats, whereas the striking and highly contrasting ‘black and white forms’ are associated with closed habitats. In this context, it is noteworthy that C. absimilis absimilis females from the extreme southern end of its range in south-eastern coastal New South Wales (Central Tilba-Bega), where it breeds primarily in eucalypt open-forest and tall open-forest (projective foliage cover 30%-70%), have substantially reduced white patches on the wings (Fig. 34F) compared with populations further north that arc mainly associated with closed forest. A similar trend also occurs in C. consimilis goodingi females from south¬ eastern Australia, which occurs primarily in eucalypt tall open-forest, in which the white patch on the fore wing is reduced in extent compared with the nominate subspecies C. consimilis consimilis, which occurs mainly in subtropical and warm temperate rainforest. The association between sex-limited phenotype and habitat (i.e. broad vegetation type, measured in terms of projective foliage cover) in Candalides also extends to other species in the family from Australia (Table 1). Of the 49 species of Lycaenidae that occur predominantly in closed forest vegetation types, 30(61 %) have the upperside ground colour black with prominent white patches in the female sex and the sexes are frequently strongly sexually dimorphic, whereas of the 90 species that live primarily in open forest/ woodland vegetation types the proportion of ‘black and white forms’ among females drops to 3%. The three species in the latter category include two [ Hypolycaena phorbas (Fabricius, 1793), Erysichton palmyra (C. Felder, 1860)] which also occur in rainforest and one [Leptotes plinius (Fabricius, 1793)] in which the white patches are very obscure. Differences in the frequencies of adult female forms among the two broad vegetation types tabulated in Table 1 are significant (%= 55.8, d.f 1, P<0.001), supporting Tabic 1 . Associations between female phenotype (upperside wing pattern) and habitat (broad vegetation type) among Australian Lycaenidae (data compiled from Braby 2000). Differences in frequency of female forms among the two vegetation types are highly significant (P < 0.001). Habitat Phenotype Total Ground colour black, with white patches present* Ground colour variable, but white patches absentf Closed forest 30 19 49 Open forest/ woodland 3 87 90 Total 33 106 139 * white patches may have some traces of blue pigmentation, t ground colour frequently brown or blue. 52 Taxonomic review of Candalides the hypothesis that habitat characteristics affects female wing phenotype in Candalides, at least in terms of the visible spectrum. Further analysis is needed to establish if other factors, such as phylogenetic history and mate location or ovipositing behaviour, are correlated with the observed pattern, and if the relationship also holds in the non-visible spectrum such as ultraviolet light, which butterflies frequently use for mate recognition. Female wing pattern does not appear to be explained by shared ancestry because the trend for habitat associated sex-limited phenotype among the Australian Lycaenidae also occurs in Jamides and Nacaduba, two genera which occur in both open and closed forest. The evolutionary forces that may have selected for female wing pattern among Candalides, and the Australian Lycaenidae in general, have not been established, but since the two habitats (broad vegetation types) analysed differ greatly in projective foliage cover and therefore their light environment, differences in ambient light properties may be a critical factor. Closed forest environments are optically complex, varying greatly in intensity (brightness) and spectral composition (colour) under sunny conditions (Endler 1992, 1993). Therefore, selection is expected to favour phenotypes that are more striking to maximise brightness contrast and conspicuousness among conspecifics or potential mates. A recent study (Douglas et al. 2007) has demonstrated that nymphalid butterflies inhabiting understorey microhabitats of tropical forest, in which light intensity is very low, are more likely to exploit polarised light as a signal for communication than related species occupying open habitats. In contrast, in open forest/woodland habitats or seasonally deciduous monsoon forests, Endler (1992) predicted that selection should favour exploitation of the shorter wavelengths (ultraviolet, blue, green) for visual signals in these ‘woodland shade’ light environments. These findings suggest that a more general hypothesis may be proposed, namely that ambient light characteristics among divergent habitats or vegetation types (i.e. different light environments) is a potent selective force in shaping sex-limited phenotype among Australian lycaenid butterflies. ACKNOWLEDGMENTS I am grateful to D.R. Britton and S.F. McEvey (AMS), E.D. Edwards and V. Rangsi (ANIC), B. Done (MTQ), R Lillywhite and C. McPhee (NMV) and J. Forrest (SAM) for access to and/or loan of specimen material held in their care. R.G. Eastwood, S.F. McEvey, R.P. Field and C.E. Meyer also provided material and field notes from their collections for detailed examination. B. Huertas (BMNH) kindly provided digital images of type material lodged in the Natural History Museum, London, for inclusion in this work, D. Schmidt provided live material of the early stages of C. margarita margarita from Brisbane for comparative morphological work, D.J. Kemp greatly assisted with literature and ideas on animal communication in relation to light environments, and R.P. Weir introduced me to the life history of C. margarita gilberti in Darwin. I thank L.J. Aitchison for field assistance in south-eastern Australia, and R.P. Field, D.A. Lane and G. Guy for biological observations on C. absimilis. D. Schmidt and C.J. Muller provided constructive editorial suggestions and thought- provoking comments on the manuscript. Specimens of C. absimilis edwardsi were collected under Research Permit Number 10002433 issued by the Victorian Department of Sustainability and Environment. REFERENCES Braby, M.F. 1996. A new species of Nesolycaena Waterhouse and Turner (Lepidoptera: Lycaenidae) from northeastern Australia. Australian Journal of Entomology 35: 9-17. Braby, M.F. 1998. Notes on the biology of some Hesperiidae and Lycaenidae in south-eastern Australia. The Victorian Naturalist 115: 4-8. Braby, M.F. 2000. Butterflies of Australia, their identification, biolog)’ and distribution. CSIRO Publishing: Melbourne. Braby, M.F. and Douglas, F. 2004. The taxonomy, ecology and conservation status of the Golden-rayed Blue, a threatened butterfly endemic to western Victoria, Australia. Biological Journal of the Linnean Society 81: 275-299. Common, I.F.B. 1964. Australian Butterflies. Jacaranda Press: Brisbane. Common, I.F.B. and Waterhouse, D.F. 1972. Butterflies of Australia. Angus and Robertson: Sydney. Common. I.F.B. and Waterhouse, D.F. 1981. Butterflies of Australia. revised edition. Angus and Robertson: Sydney. Common, I.F.B. and Waterhouse, D.F. 1982. Butterflies of Australia. abridged field guide edition. Angus and Robertson: Sydney. Crosby, D.F. 1951. Notes on some eastern Victorian butterflies with a new Victorian record. The Victorian Naturalist 68: 97-101. D’Abrera, B. 1971. Butterflies of the Australian Region. Lansdownc Press: Melbourne. Daniels, G. and Edwards, E.D. 1998. Butterflies from Lawn Hill Nationa. Park and Musselbrook Reserve, Queensland. Pp. 89-91 In: Transactions of the Royal Geographical Society of Queensland (ed.) Musselbrook Resene Scientific Study Report, Geography Monograph Series No. 4. The Royal Geographical Society of Queensland Inc.: Brisbane. Douglas, J.M., Cronin, T.W., Chiou, T.-H. and Doniiny, N.J. 2007. Light habitats and the role of polarized iridescence in the sensory ecology of neotropical nymphalid butterflies (Lepidoptera: Nymphalidae). The Journal of Experimental Biology 210: 788-799. Dunn, K.L. 1990. Candalides absimilis (Felder) in Victoria. Victorian Entomologist 20: 49-53. Dunn, K.L. 1995. New distribution records for some eastern Australian butterflies with comments on subspecies taxonomy. Victorian Entomologist 25: 26-30. Dunn, K.L. 2007. The geographical movements ofllermann Elgner (d. 1913): a lepidopterist based in Torres Strait, Australia. Calodema 10: 19-34. Dunn, K.L. and Dunn. L.E. 1991. Review of Australian Butterflies: distribution, life history and taxonomy. Parts 1-4. Privately published by the authors: Melbourne. Dunn, K.L., Kitching, R.L. and Dexter, E.M. 1994. The Conservation Status of Australian Butterflies. Unpublished report to Australian National Parks and Wildlife Service: Canberra. 53 M. F. Braby Edwards, E.D. 1996. Lycaenidae. In: Nielsen, E.S., Edwards, E.D. and Rangsi,T.V. (eds) Checklist of the Lepidoptera ofAustralia, monographs on Australian Lepidoptera, volume 4. Pp 249-254, 360-363. CSIRO Publishing: Melbourne. Edwards, E.D. and Kerr. J.F.R. 1978. A new species of Candalides from eastern Australia and notes on Candalides hyacinthinus (Semper) (Lepidoptera: Lycaenidae). Australian Entomological Magazine 4: 81-90. Edwards, E.D., Newland, J. and Regan, L. 2001. Lepidoptera: Hesperioidea, Papilionoidea. Zoological Catalogue of Australia. Vol. 31.6. CSIRO Publishing: Melbourne. Endler, J. 1992. Signals, signal conditions, and the direction of evolution. The American Naturalist 139: S125—S153. Endler, J. 1993. The colour of light in forests and its implications. Ecological Monographs 63: 1 -27. Franklin, D.C. 2007. Dry season observations of butterflies in the “Gulf Country" of the Northern Territory and far north-west Queensland. Northern Territory Naturalist 19: 9-14. Gabriel, A.G. 1932. Catalogue of the type specimens of the Lepidoptera Rhopalocera in the Hill Museum. John Bale, Sons & Danielsson: London. Grund, R. 2006. Some new butterfly observations for Central Australia. Victorian Entomologist 35: 111-114. International Commission of Zoological Nomenclature 1999. International Code of Zoological Nomenclaturefourth edition. International Trust for Zoological Nomenclature: London. Kelly, P, Reid, N. and Davies, 1. 1997. Effects of experimental bunting, defoliation, and pruning on survival and vegetative resprouting in mistletoes (Amyema miquelii and Amyema pendula). International Journal of Plant Sciences 158: 856-861. Kerr, J.F.R. 1967. New records of Lycaenidae (Lepidoptera) in Australia and a description of a new subspecies. Journal of the Australian Entomological Society 6: 49-51. Kikkawa, J., Monteith, G.B. and Ingram, G. 1981. Cape York Peninsula, major region of faunal interchange. In: Kcast, A. (ed.) Ecological Biogeographv of Australia. Pp 1695-1742. Dr. W. Junk Publishers: The Hague. McCubbin, C. 1971 .Australian Butterflies. Nelson: Melbourne. McEvey, S.F. 1977. Notes from a collecting trip to Caims and Iron Range including a new record for the Australian mainland. Victorian Entomologist 7: 59-62. McEvey, S.F. 1979. Butterflies of some Gippsland rain forests. Victorian Entomologist 9: 12-13. Meyer, C.E., Weir, R.P. and Wilson, D.N. 2006. Butterfly (Lepidoptera) records from the Darwin region. Northern Territory. Australian Entomologist 33: 9-22. Moulds, M.S. 1977. Bibliography of the Australian butterflies (Lepidoptera: Hesperioidea and Papilionoidea) 1773-1973 Australian Entomological Press: Greenwich. New, T.R. 1991. Butterfly Conservation. Oxford University Press Australia: Oxford. New, T.R., Field, R.P. and Sands, D.P.A. 2007. Victoria’s butterflies in a national conservation context. The Victorian Naturalist 124: 243-249. Parsons, M.J. 1986. A new genus and twenty-six new species of butterflies (Lepidoptera: Hesperiidae, Lycaenidae, Nymphalidac) from Papua New Guinea and Irian Jaya. Tyo To Gail: 103-177. Parsons, M.J. 1998. The Butterflies of Papua New Guinea. Their systematics and biology. Academic Press: London. Peters. J. V. 1971.4 catalogue of the type specimens of the Hesperioidea and Papilionoidea (Lepidoptera) in the Australian Museum. Australian Entomological Press: New South Wales. Puccctti. M. 1991. Butterflies of Doomadgec - northwestern Queensland. Victorian Entomologist 2 1: 142-147. Samson, P.R. and Wilson. D.N. 1995. The life history of Candalides gilberti Waterhouse (Lepidoptera: Lycaenidae). Australian Entomologist 22: 71-73. Semper, G. 1879. Bcitrag zur Rhopalocerenfauna von Australien. Journal des Museum Godeffroy 14: 138-194. Smales, M. and Ledvvard, C.P. 1943. Notes on the life histories of some lycacnid butterflies - Part II. (with notes on some skippers). Queensland Naturalist 12: 47-52. Tennent. W.J. 2006. A checklist of the butterflies of Melanesia, Micronesia, Polynesia and some adjacent areas. Zootaxa 1178: 1-209. Tindalc, N.B. 1965. A new species of Holochila (Rhopalocera, family Lycaenidae) from Victoria and southern New South Wales. Records of the South Australian Museum 15: 165-174. Tite, G.E. 1963. A revision of the genus Candalides and allied genera (Lepidoptera: Lycaenidae). Bulletin of the British Museum (Natural History) Entomology 14: 197-259. Waterhouse, G.A. 1903a. Notes on Australian Rhopalocera: Lycaenidae. Part III Revisional. Proceedings of the Linnean Society of New South Wales! 8: 132-275. Waterhouse, G.A. 1903b. A catalogue of the Rhopalocera of Australia. Memoirs of the New South Wales Naturalists' Club No. 1: 1-49, i-ii. Waterhouse, G.A. 1932. What Butterfly is That? Angus and Robertson: Sydney. Waterhouse, G.A. 1942. Notes on Australian butterflies in the Australian Museum. Records of the Australian Museum 21: 122-125. Waterhouse, G.A. and Lyell, G. 1914. The Butterflies of Australia. A monograph of the Australian Rhopalocera. Angus and Robertson: Sydney. Accepted 11 July 2008. 54 The Beagle, Records of the Museums and Art Galleries of the Northern Territory, 2008 24: 55-61 Two new species of the genus Pseudaptinus Castelnau from northern Australia (Insecta: Coleoptera: Carabidae: Zuphiinae) MARTIN BAEHR Zoologische Staatssammlung Miinchen, Miinchhausenstr. 21, D-81247 Miinchen, GERMANY martin.baehr@zsm.mwn.de ABSTRACT Two new species of the zuphiine genus Pseudaptinus Castelnau are described from the Northern Territory of Australia: P. parallelus sp. nov. and P. gregoryensis sp. nov. According to the size of eye and the structure of surface of the elytra, the first species is probably most closely related to P punctatostriatus Bachr and P. cyclophthalmus Baehr, but il is smaller than either of them; the second species is probably nearest to P. iridescens Baehr, but it is distinguished by its smaller eyes and shorter antennae. Additional records are provided for P. hirsutulus Baehr and P. iridescens Baehr. A new key to all the Australian species of this genus is included. Keywords: Coleoptera, Carabidae, Zuphiinae, new species, Pseudaptinus parallelus, P. gregoryensis, taxonomy, Australia, Northern Territory. INTRODUCTION During a recent collecting trip by the author through the northern parts of the Northern Territory and Western Australia, two new species of the zuphiine genus Pseudaptinus Castelnau were collected. These are described here as P. parallelus sp. nov. and P. gregoryensis, sp. nov., and this is the fifth supplement to my revision of the Australian Zuphiinae (Baehr 1984, 1985a, 1985b, 1986a, 1985b, 1988, 1991, 1992, 1995, 2001, 2008). Additional specimens of one of the new species, P. parallelus, were also received from the Australian National Insect Collection, Canberra, and the South Australian Museum, Adelaide. Additional records of two other species of this genus occurring in northern Australia, P. hirsutulus Baehr and P. iridescens Baehr, are also provided. The key to the 11 Australian species of this genus is revised to include all species described since the revision of the genus by Baehr (1985). With about 40,000 described species and subspecies worldwide, carabid beetles are one of the largest families of beetles. They belong to the suborder Adephaga which also includes a few families of aquatic beetles and is mainly characterised by a rather plesiomorphic structure of their leg insertions and by their primitive campodeiform larvae which still possess well articulated legs. From Australia presently about 2,600 species and subspecies are recorded, but the number of actual existing species certainly is much greater. Zuphiinae (or Zuphiini, according to the opinion of authors) is a highly evolved subfamily (or tribe) of carabid beetles, characterised by the completely pilose antennae and in particular by the elongate 1st antennomere, the scapus. Opinions about the status (subfamily or tribe) of the main subgroups of the family Carabidae differ markedly between authors, but the delimitation of the subgroups is not in doubt, hence the question whether zuphiines should be regarded a subfamily or rather a tribe, is of minor importance. Zuphiinae presently contain eight genera in Australia, of which the genus Acrogenys Macleay and two genera of cavernicolous species ( Speothalpius Moore and Speozuphium Moore) are endemic to Australia. With respect to their phylogenetic status, the Australian zuphiine fauna is particularly diverse and includes as well the probably most plesiotypic zuphiine genus in the world ( Acrogenys), as well as the highly evolved genera Speothalpius and Speozuphium. At present, 45 zuphiine species and subspecies are recorded from Australia (Baehr 1984,1985a, 1985b, 1986a, 1986b, 1988, 1991, 1992,1995,2001,2008; Moore 1995), but specimens of most species still are rare to very rarely encountered. This is probably due to the presumably rather secretive way of life of most species which, at least during daytime, may be hidden under debris or in cracks in the earth. As a consequence, most recorded specimens were collected using a light at night. The genus Pseudaptinus was founded by Castelnau (1835) on an American species. Later Castelnau (1867) described a single Australian species which was included in the subgenus Thalpius Leconte, 1851. Baehr (1985) revised the genus, transferred Acrogenys australis Blackburn, 1890 to Pseudaptinus, and described six new species. Later Baehr (1995) described another additional species. Hence, at present nine species are recorded from Australia, all are included in the subgenus Thalpius, and most of them occur in the tropical northern parts of M. Baehr Australia. The genus Pseudaptimis is characterised by the quite compact build of all species with fairly convex dorsal surface, their posteriorly rounded, not square head, the only moderately elongated 1 st antennomere, and the erect and hirsute pilosity. The subgenus Thalpius is rather weakly characterised, distinguished from the nominate subgenus mainly by the more depressed dorsal surface and the wider, more cordiform pronotum. The genus Pseudaptimis also occurs in South America and in the southern parts of North America, where specimens of both subgenera are found. Hence this is one of the few carabid genera with an Australian-Neotropical range, which is very interesting in the light of their biogeographical history. METHODS For the taxonomic treatment standard methods were used. The male genitalia were removed from specimens soaked overnight in a jar under a wet atmosphere, then cleaned for a short while in hot potassium hydroxide. For examination of the fine, though taxonomically important, surface microsculpture (i.e., punctation and microreticulation) a high resolution stereomicroscope with up to 64x magnification was used, supported by a lamp of higli intensity giving natural light that could be focused. For exact definition of the microsculpture such light is preferable, because fibre-optics lights substantially change perception of the surface structures. The habitus photographs were taken with a digital camera using SPOT Advanced for Windows 3.5 and subsequently edited with Corel Photo Paint 11. Measurements were made using a stereomicroscope with an ocular micrometer. Length has been measured from the apex of labrum to the apex of the elytra. Length of eye includes a small ring of dark coloured rhabdomes that in some instances is present behind the light area. Length of orbit was taken in a straight line parallel to midline of body, from the posterior margin of the eye to the ‘neck' incision. Length of pronotum was measured from the most advanced part of base to the most advanced part of apex. Length of elytra was taken from the most advanced part of humerus to the most advanced apex of elytra. The length/width ratio of the 6th antennomere was chosen for better comparison with the measurements in the revision. It should be noted, however, that the 7th to 9th antennomcres usually are slightly shorter than the 6th antennomere. ABBREVIATIONS ANIC Australian National Insect Collection, Canberra CBM Working Collection M. Baehr in Zoologische Staatssammlung, Miinchen NTM Museum and Art Gallery of the Northern Territory, Darwin (formerly Northern Territory Museum) SAMA South Australian Museum, Adelaide TAXONOMY Pseudaptimisparallelus sp. nov. (Figs 1-3, 5, 7) Material examined. HOLOTYPE-male, Australia07, NT22, Roper River Crossing, 3 km E. Roper Bar, 14°42.83'S, 134 o 30.51’E, 10 m, 6-7 November 2007, M. Baehr (NTM 1005298). PARATYPES - 2 females, same dataasholotype(CBM); 1 female, Australia07, NT9, Mary River National Park, Maiy River Billabong, 5 km NNW. Mary River Crossing, I2°53.49’S, 131°38.33’E, 28 m, 1-2 November 2007, coll. M. Baehr (CBM); 1 male, 1 female, 12°40’S 142°40’E, Queensland, Batavia Downs Homestead. 22 November 1992, light trap, coll. A. Caldcr and P. Zborowski (ANIC, CBM); 1 male, Old Greenvale, 70 km SW. at light, 14-23 February 96, coll. A. J. Watts (SAMA). Diagnosis. Species of subgenus Thalpius Leconte, characterised by small size, elongate and remarkably parallel-sided elytra, and short antenna. Description. Measurements. Length: 4.75-4.95 mm; width: 1.5-1.6 mm. Ratios: Length of orbit/lcngth of eye: 0.45-0.5: width/length of pronotum: 0.94-0.98; widest / narrowest diameter of pronotum: 1.31-1.35; length/width of elytra: 1.77-1.79; length/width of 6th antennomere: 1.75-1.8. Colour. Reddish-black to blackish, according to age of the specimen. In dark specimens, suture of elytra indistinctly paler. Mouthparts, antennae, and legs dirty yellow. Head (Fig. 3). Slightly narrower than pronotum. Eyes large and laterally well protruded, about twice as long as orbits (when measured in a straight line). Orbits markedly Fig. I. Pseudaptimis parallelus sp. nov. Male genitalia: aedeagus, lateral and ventral view, left and right parameres. Scale: 0.25 mm. 56 New species of Pseudaptimis from northern Australia Fig. 2. Pseudaptimis parallelus sp. nov. Female gonocoxites 1 and 2. Scale: 0.1 mm. convex, even obtusely ungulate, with sharp angle to neck. Suture between occiput and summit deep, neck dorsally and laterally deeply separated. Orbits separated from summit by sulcus. Posterior supraorbital seta located at level with posterior margin of eye. Labrum anteriorly straight. Mandibles short, incurved at apex. Mental tooth triangular, short and wide. Glossa corneous, apically square, paraglossae membraneous, short, fused to glossa, laterally little surpassing the glossa. Galea distinctly pilose. Both palpi densely pilose, labial palpus small, terminal palpomcre cylindrical. Maxillary palpus large, though rather short (in group). Antenna short and stout, just surpassing base of elytra, median antennomeres distinctly less than twice as long as wide. 1 st antennomere barely longer than 2nd and 3rd antennomeres together. Surface glossy, not microreticulate, coarsely punctate, diameter of punctures slightly larger than distance between punctures, pilosity fairly dense, hirsute, slightly inclined anteriad. Prothorax. Slightly longer than wide, depressed, widest in apical fifth. Apical angles broadly rounded, apex faintly concave, lateral margins moderately convex on anterior half, rather deeply sinuate in front of the acute, Figs 3,4. Left side of head, showing ratio eye/orbit; 3. Pseudaptimis parallelus sp. nov.; 4. P. gregoryensis sp. nov. about rectangular.markedly projecting basal angles. Base distinctly projecting medially, laterally deeply excised, hence basal angles removed from base. Median line deep, anterior transverse sulcus indistinct. Central part of pronotum slightly raised, pronotum near apex with an elongate, rather deep impression on either side. Base and apex both unmargined, lateral borders margined, marginal channel very narrow. Anterior marginal seta situated at apical fifth, posterior seta at basal angle. Anterior seta considerably longer than the posterior seta. Surface without microreticulation, glossy, densely and rather coarsely punctate, diameter of punctures wider than distance between them. Pilosity rather elongate, hirsute, irregularly inclined posteriorly. Elytra (Figs 5, 7). Narrow and elongate, remarkably parallel-sided, not perceptibly widened in apical part, surface depressed. Humeri projected, rounded. Apex laterally rounded, in middle transversal, barely incurved towards suture. All striae well impressed, coarsely punctate. Intervals depressed, without microreticulation, glossy, with coarse, irregularly uni- to biseriate punctation. Pilosity hirsute but rather depressed, inclined posteriorly, hairs about as long as one interval wide. Third interval with three punctures and erect setae which are well recognisable when seen laterally. Ventral surface. Thorax with coarse and moderately dense punctures and pilosity. Abdominal sterna with finer and denser punctures and with dense adpressed pilosity. Metepisternum elongate, greater than twice as long as wide at anterior margin. Apical margin of sternum VII in male bisetose, in female quadrisetose, in both sexes with an additional seta on either side removed from apex. Legs. Of normal shape and length, densely punctate and pilose. Three basal tarsomeres of male protarsus slightly widened and asymmetrically biseriate'y squamose. Male genitalia (Fig. 1). Aedeagus small, fairly elongate, apex rather elongate (in group), with remarkably acute tip. Lower surface very gently concave. Orificium very elongate, with a median cleft at apex. Internal sac with two large, rather symmetrical, slightly sclerotised and somewhat folded plates on either side. Parameres very dissimilar, both rounded at apex, right paramere remarkably short. Female genitalia (Fig. 2). Gonocoxite 2 wide, trapezoidal, with wide, oblique apex, apical part less sclerotised than basal-central part and densely setose latero-apically. No ensiform setae present. Apex of gonocoxite 1 asetose, though apical margin of lateral plate with a few elongate setae. Variation. Apart from differences in colour which are probably due to the age of the specimens, little variation was noted, even between specimens from Northern Territory and those from northern Queensland. Collecting circumstances. The specimens from the Northern Territory were sampled with a black light near the sandy bed of the Roper River and at the margin of a billabong with dense marginal vegetation of trees and undergrowth. 57 M. Baehr Figs 5,6. Habitus. Body lengths in brackets; 5. Pseudaptinus parallelus sp. nov. (4.85 mm); 6. P. gregoiyensis sp. nov. (5.35 mm). Distribution. North-eastern part of Northern Territory and tropical northern Queensland, from Cape York Peninsula southwards to about Townsville. Etymology. The name refers to the remarkably parallel¬ sided elytra. It is an adjective. Relationships. Although similar to P. hirsutulus Baehr in body shape, the new species is certainly closer to P. punctatostriatus Baehr and P cyclophthalnuis Baehr, according to the shape of the pilosity and the structure or the aedeagus. However, it differs from both these species by its smaller size, shorter antenna, and the conspicuously parallel-sided elytra. Pseudaptinus gregoiyensis sp. nov. (Figs 4, 6, 8) Material examined. HOLOTYPE-female,Australia07, NT37, Gregory National Park, Old Victoria River Crossing, Figs 7, 8. Centre of elytra showing elytra! microsculpture. 7. Pseudaptinusparallelus sp. nov.; 8. P. gregoiyensis sp. nov. 5 km W. Victoria River Roadhouse, Victoria Highway, 15°34.87’S, 131°06.24'E, 35 m, 11-12 November 2007 coll. M. Baehr (NTM 1005299). Diagnosis. Characterised by moderately large eyes (in group), rather dense punctures on the elytra! intervals and fine punctures on the striae. Distinguished from most similar P. iridescens Baehr by smaller size, slightly smaller eyes, distinctly shorter antenna, and moreover by far less distinct iridescent lustre of the elytra. Description. Measurements. Length: 5.35 mm; width: 1.75 mm. Ratios: Length of orbit/length of eye: 0.75; width/ length of pronotum: 0.91; widest/narrowest diameter of pronotum: 1.31; length/width of elytra: 1.69; length/width of 6th antennomere: 2.05. Colour. Dark blackish brown to almost black, only pronotum very slightly paler, and suture of elytra narrowly dark reddish. Mouth parts, antennae, and legs dirty yellow. Head (Fig. 4). Slightly narrower than pronotum. Eyes of moderate size, little protruded laterally, about 1.5 x as long as orbits (when measured in a straight line). Orbits evenly convex, forming sharp angle with neck. Suture between neck and summit of head deep, neck dorsally and laterally deeply separated. Orbits separated from summit by sulcus. Posterior supraorbital seta located slightly behind posterior margin of eye. Labrum anteriorly straight. Mandibles short, incurved at apex. Mental tooth triangular, short and wide. Glossa corneous, apically square, paraglossae membraneous, short, fused to glossa, laterally little surpassing the glossa. Galea distinctly pilose. Both palpi densely pilose, labial palpus small, terminal palpomere cylindrical. Maxillary palpus large, though rather short (in group). Antenna rather short and stout, sutpassing base of elytra by about 2.5 antennomeres, median antennomeres distinctly only twice as long as wide, 1st antennomere slightly longer than 2nd and 3rd antennomeres together. Surface glossy, with extremely superficial microreticulation which is not visible when seen from above, rather densely punctate, diameter of punctures about as large as distance between punctures, but laterally punctation less dense, pilosity fairly dense, hirsute, slightly inclined anteriad. Prothorax. Considerably longer than wide, depressed, widest in apical fifth. Apical angles broadly rounded, apex faintly concave, lateral margins in anterior half moderately convex, with elongate sinuation in front of the acute, about rectangular, markedly projecting basal angles. Base in middle distinctly projecting, laterally deeply excised, hence basal angles removed from base. Median line deep, anterior transverse sulcus indistinct. Central part of pronotum slightly raised, near apex with an elongate, rather deep impression on either side. Base and apex both not margined, marginal channel veiy narrow. Anterior marginal seta situated at apical fifth of length, posterior seta at basal angle. Anterior seta considerably longer than posterior seta. Surface with extremely superficial microreticulation, visible only in oblique light, glossy, densely and moderately 58 New species of Pseudaptinus from northern Australia coarsely punctate, diameter of punctures about as wide or even slightly wider than distance between them. Pilosity rather elongate, hirsute, irregularly inclined posteriorly. Elytra (Figs 6, 8). Rather narrow and elongate, fairly parallel-sided, slightly widened in apical part, surface in middle depressed. Humeri projected, rounded. Apex laterally rounded, in middle slightly oblique, barely incurved towards suture. All striae well impressed, moderately coarsely punctate. Intervals depressed, with superficial traces of slightly transverse microreticulation, rather glossy, with fairly coarse, irregularly triseriate punctation. Pilosity hirsute but rather depressed, inclined posteriad, hairs about as long as one interval wide. Third interval with three punctures and erect setae which are well recognisable when seen laterally. Lower surface. Thorax with coarse and moderately dense punctures and pilosity. Abdominal sterna with finer and denser punctures and with dense adpressed pilosity. Metepistemum elongate, about twice as long as wide at anterior margin. Apical margin of sternum Vll in female quadrisetosc and with an additional seta on either side removed from apex. Legs. Of normal shape and length, densely punctate and pilose. Squamosity of male protarsus unknown. Male genitalia. Unknown. Female genitalia. Similar to those of P. parallelus sp. nov. Variation. Unknown. Collecting circumstances. Collected at a black light near the bed of the Victoria River. Here the substrate consists of pebbles and more or less coarse sand. Distribution. North-western Northern Territory. Known only from type locality. Etymology. The name refers to the collecting area, Gregory National Park. It is an adjective. Relationships. The species is probably most closely related to P. iridescens Baehr which is widely distributed in tropical northern Australia, but it is distinguished from P. iridescens mainly by the shorter antenna, smaller eyes, coarser punctures of the elytral striae, and barely iridescent elytra. Pseudaptinus hirsutulus Baehr Pseudaptinus hirsutulus Baehr, 1985a: 51. - Baehr 1986b: 162; Baehr2001: 85. Distribution. Northern parts of the Northern Territory from Darwin eastwards at least to the western boundary of Arnhem Land. New records. Australia07, NT9, Mary River National Park, Mary River Billabong 5 km NW. Mary River Crossing, 12°53.49’S, I31°38.33’E, 28 m, 1-2 November 2007, coll. M. Baehr (CBM, NTM). Collecting circumstances. All specimens collected at a black light at the margin of a billabong with dense vegetation of trees and undergrowth. At the Maty River Billabong this species was collected together with P. parallelus sp. nov. (see above). Pseudaptinus iridescens Baehr Pseudaptinus iridescens Baehr, 1985a: 43. - Baehr 1986b: 162; Baehr2001: 84. Distribution. North-western Victoria, New South Wales and Queensland west of the Great Dividing Range, northern parts of Northern Territory and of Western Australia. New records. NT22, Roper River Crossing, 3 km E. Roper Bar, 14°42.83’S, 134°30.51’E, 10 m, 6-7 November 2007, coll. M. Baehr; NT29, 70 km N. Top Springs on Buntine Highway, 16°00.94’S, 131°56.33’E, 253 m, 9-10 November 2007, coll. M. Baehr (CBM, NTM). Collecting circumstances. All the specimens were collected at a black light, those near Roper Bar close to the bank of the Roper River; the one specimen from the ‘Buntine Highway’ was taken in open woodland, apparently far away from any body of water. At the Roper River this species was collected together with P. parallelus sp. nov. (see above). KEY TO GENUS PSEUDAPTINUS Because the three additional species must now be included in the key for the genus that was previously published (Baehr 1985), a completely new key to all the species of Pseudaptinus is provided. For the benefit of the users, figures from previous papers arc included as BA85 and BA95. 1 a. Eyes not longer than orbits (BA85 figs 3a-c, g).... 2 lb. Eyes considerably longer than orbits (BA85 figs 3d-f, h).6 2a. Antenna very elongate, median aniennomeres >3x longer than wide (BA85 figs 4c, f).3 2b. Antenna shorter, median antennomeres <2.5 x longer than wide (BA85 figs 4a, b).4 3a. Eyes perceptibly shorter than orbits (BA85 fig. 3g); pronotum elongate, narrow, basal angles very acute (BA85 fig. 12); pronotum and elytra rather convex; striae very coarsely punctate, punctures of intervals in two rows (BA85 fig. 5g); pilosity unequal, rather erect; aedeagus short and thick, apex abruptly sloping, left paramere very large, broadly rounded at apex, right paramere short (BA85 fig. 19).... P. monteithi Baehr 3b. Eyes as long as orbits (BA85 fig. 3c); pronotum wider, basal angles far less acute (BA85 fig. 8); pronotum and elytra fairly depressed; striae rather finely punctate, punctures of intervals in 3-4 rows (BA85 fig. 5c); pilosity equal, rather depressed; aedeagus long, narrow, and rather low, apex gently sloping, left paramere narrow, square at apex, right paramere rather long and narrow (BA85 fig. 15). P. brittoni Baehr 59 M. Baehr 4a. Larger species (body length >5.5 mm); antenna longer, median antennomeres well >2 x longer than wide (BA85 fig. 4a); longer and narrower species, fore body considerably shorter than elytra (BA85 fig. 6); aedeagus small and depressed, thickened before apex, then gently sloping, both parameres very long and narrow (BA85 fig. 14). P. fulvus (Castelnau) 4b. Smaller species (body length <5 mm); antenna shorter, median antennomeres <2 x as long as wide (BA85 fig. 4b; BA95 fig. 1); shorter and wider species, fore body almost as long as elytra (BA85 fig. 7; BA95 fig. 1) ; aedeagus far less depressed, almost equally wide, or unknown, parameres shorter and wider (BA95 fig- 2).5 5a. Eyes smaller, orbits 1/3 x longer than eyes (BA85 fig. 3b); lateral margins of pronotum near base less sinuate, basal angles less projecting (BA85 fig. 7); elytra not markedly depressed on disk, 5th interval not raised; aedeagus unknown. South Australia. ... P australis (Blackburn) 5b. Eyes larger, orbits little longer than eyes (B A95 fig. 1); lateral margins of pronotum near base deeply sinuate, basal angles well projecting (BA95 fig. 1); elytra markedly depressed on disk, 5th interval distinctly raised; aedeagus short and rather compact (BA95 fig. 2) . North Queensland. P. depressipennis Baehr 6a. Eyes at most slightly longer than orbits (BA85 fig. 3c); pronotum and elytra rather depressed; antenna very elongate, median antennomeres >3 x longer than wide (BA85 fig. 4c); aedeagus elongate, depressed, apex gently sloping, apex of left paramerc square (BA85 fig. 15). P. brittoni Baehr 6b. Eyes much longer than orbits (Figs 3, 4; BA85 figs 3d-f, h); body more convex; antenna shorter, median antennomeres <3 x longer than wide (BA85 figs 4d-f, h); aedeagus shorter and stouter, apex more abruptly sloping or apical lamellae with pilose appendices (Fig. 1; BA85 figs 16-18,20).7 7a. Smaller species (body length < 5.1 mm); antenna short, median antennomeres distinctly <2 x longer than wide (BA85 fig. 4h).8 7b. Larger species (body length >5.35 mm); antenna longer, median antennomeres >2 x longer than wide (BA85 figs 4d-f).9 8a. Colour brownish to black, surface glossy; head short, neck very distinctive; elytra shorter, <1.67 x longer than wide and perceptibly widened towards apex; striae very coarsely punctate, punctures of intervals fine (BA85 fig. 5h); pilosity elongate, erect, but quite sparse; hairs c. 2 x as long as width of an interval; apex of aedeagus with pilose lateral appendices (BA85 fig. 20). p hirsutulus Baehr 8b. Colour reddish to light brown, surface less glossy; head longer, neck less distinctive; elytra longer, >1.77 x longer than wide and markedly parallel-sided (Fie. 5); punctures of striae and intervals less dissimilar; pilosity shorter and denser; apex of aedeagus without such appendices (Fig. 1). P. parallelus sp. nov. 9a. Antenna rather elongate, median antennomeres > 2.5 x longer than wide (BA85 fig. 4d); striae finely punctate, punctures of intervals dense, in 3-4 rows, punctures not much smaller than those of striae (BA85 fig. 5d); usually elytra distinctly iridescent; aedeagus very thick and broad, both parameres large, apex of right paramere very short and wide (BA85 fig. 16).. . P. iridescens Baehr 9b. Antenna shorter, median antennomeres c. 2 x longer than wide, or little longer (BA85 figs 4e, fj; striae coarsely punctate, punctures of intervals considerably finer than those of striae (BA85 figs 5e, 0; elytra not, or but faintly iridescent; aedeagus not so large, right paramere longer and more acute (BA85 figs 17, 18), or unknown.10 10a. Eyes very large, semicircular and remarkably protruded (BA85 fig. 4f); punctures of intervals in 2-3 rows (BA85 fig. 5f); pilosity rather elongate, unevenly erect and depressed; aedeagus fairly small and narrow, lower surface concave, internal sac small, right paramere very narrow and elongate (BA85 fig. 18)... .P. cyclophthalmus Baehr 10b. Eyes less semicircular and protruded (Fig. 3; BA85 fig. 4e); pilosity shorter, more regular and depressed; either punctures of striae less coarse and punctures of intervals at least in 3 rows, or punctures of striae very coarse, almost as wide as the intervals, and punctures of intervals in 2 rows; aedeagus short and thick, widened apicad, internal sac very large, right paramere much shorter (BA85 fig. 17), or unknown . 11 1 la. Punctures of striae very coarse, punctures almost as wide as one interval; punctures of intervals sparse, in 2 rows (BA85 fig. 5e); pilosity rather sparse; aedeagus short and thick, widened apicad, internal sac very large, right paramere much shorter (BA85 fig. 17).. . P. punctatostriatus Baehr lib. Punctures of striae less coarse, punctures at most half as wide as one interval; punctures of intervals denser, in 3 or even 4 rows; pilosity denser; aedeagus unknown. P. gregoryensis sp. nov. 60 New species of Pseudaptinus from northern Australia REMARKS The Australian species of the genus Pseudaptinus are remarkably similar in many aspects of body shape and structure, except for the unique P. hirsutulus, in which the external morphology as well as that of the aedeagus differs considerably from all the other Australian species. The question, why so many morphologically similar species arc able to coexist in tropical northern Australia, commonly at the same locality, is not solvable given the lack of information about the ecological preferences and habits of any species of this genus. Almost all specimens with label data on collecting circumstances, and almost all specimens that I have collected during several trips in northern Australia, were sampled at light. Although it may be supposed that all Pseudaptinus species are more or less hygrophilous and during the daytime they probably hide under leaf litter, or debris, or in cracks in the soil, nothing has been ever recorded about their nutrition, courtship, or reproduction, and the larvae are also unknown. It would be interesting to know whether species of Pseudaptinus can be sampled cither by systematic pitfall trapping or by Berlese Funnel extraction of leaf litter and debris. Both these collecting methods would probably yield more exact information about the habitat preferences of the species. ACKNOWLEDGMENTS I am grateful to the Deutsche Forschungsgemeinschaft who supported my collecting trip and visits to certain Australian museums under grant No. Ba 856/10-1. The authorities of the Parks and Wildlife Commission of the Northern Territory kindly granted me a permit to collect in certain Northern Territory national parks. Thanks are also due to Jan Forrest and Eric Matthews, Adelaide, and Tom Weir, Canberra, for the loan of specimens. REFERENCES Baehr, M. 1984. Revision der Australischen Zuphiinae. 1. Gattung Acrogenys Maclcay (Insecta, Coleoptera, Carabidae). Spixiana 7: 115-134. Baehr. M. 1985a. Revision of the Australian Zuphiinae. 3. The genus Pseudaptinus Castelnau (Insecta, Coleoptera, Carabidae). Spixiana 8: 33-57. Baehr, M. 1985b. Revision of the Australian Zuphiinae. 4. The genus Parazupliium Jcannel (Insecta, Coleoptera, Carabidae). Spixiana 8: 295-321. Baehr, M. 1986a. Revision of the Australian Zuphiinae. 5. The genus Zuphiunt Latreille (Insecta, Coleoptera, Carabidae). Spixiana 9: 1-23. Baehr, M. 1986b. Revision of the Australian Zuphiinae. 6. The genus Planetes Macleay. Supplement to the other genera (Insecta, Coleoptera, Carabidae). Spixiana 9: 151-168. Baehr, M. 1988. Revision of the Australian Zuphiinae 2. Colasidia monteitlii sp. nov. from North Queensland, first record of the tribe Leleupidiini in Australia (Insecta: Coleoptera: Carabidae). Memoirs of the Queensland Museum 25: 135-140 (1987). Baehr, M. 1991. On new and rare Leleupidiini from the Oriental and Australian Regions (Coleoptera, Carabidae, Zuphiinae). Mitteilungen der Munchner Entomologischen GesellschaJ) 81: 193-202. Baehr. M. 1992. A new Acrogenys Macleay from Central Australia. Supplement to the revision of the Australian Zuphiinae (Insecta, Coleoptera, Carabidae). Spixiana 15: 75-80. Baehr, M. 1995. A new species of Pseudaptinus Castelnau from Australia. 2nd supplement to the ‘Revision of the Australian Zuphiinae’ (Insecta, Coleoptera, Carabidae, Zuphiinae). Koleoptemlogische Rundschau 65: 15-18. Baehr, M. 2001. New species and new records of Zuphiinae from Australia (Coleoptera: Carabidae). - Sukunahikona. Special Publication of the Japan Coleopterological Society 1: 83-101. Baehr, M. 2008. New species of the zuphiine genus Acrogenys Macleay from Australia (Carabidae: Zuphiinae). Coleoptera 11 (for 2007): 113-124. Moore, B. R 1995. Two remarkable new genera and species of troglobiotic Carabidae (Coleoptera) from Nullarbor Caves. Journal of the Australian Entomological Society 34: 159-161. Accepted 9 September 2008 61 . The Beagle, Records of the Museums and Art Galleries of the Northern Territory, 2008 24: 63-77 Biogeographic implications of Ascidiacea (Tunicata) from the Wessel Islands (Arafura Sea) PATRICIA KOTT Honorary Associate, Queensland Museum, PO Box 3300, South Brisbane, QLD 4101, AUSTRALIA patricia.mather@qm.qld.gov.au ABSTRACT A collection of ascidians from the Wessel Islands (off north-eastern Arnhem Land) confirms the Indo-West Pacific affinities of the Australian tropical ascidian fauna. The collection, dominated by species of the family Didemnidae previously observed to be diverse in tropical waters, includes several of the species with obligate Prochlonm symbioses. The collection also contains three large solitary species (Polycarpa aurita, P. papillata and Microcosmus helleri) known to have a pan-tropical range. The geographic range and life history strategies of the majority of species in this collection support the view that gene flow in the Indo-West Pacific and through the straits to the north of the Australian continent that connect the two oceans is not constrained by either a short free-swimming larval life, or internal fertilisation in the fixed, colonial organisms that dominate the tropical fauna. In colonial species at least, population maintenance may be a more significant selective advantage than gene flow through larval dispersal. Keywords: Indo-West Pacific, temperate, tropical, pan-tropical, Didemnidae, gene flow. INTRODUCTION The 35 ascidian specimens (colonies and solitary individuals) reported on below were taken by SCUBA from the Wessel Islands (in the Arafura Sea off north¬ eastern Arnhem Land; see Table 1). Colonial, especially didemnid species, dominate the species list, only five of the 25 species represented being solitary (Table 2). The majority of the species in this collection have been recorded previously from northern Australia and are confirmed as common components of the benthic fauna occurring in tropical Australian waters. The collection contains species with a range in tropical and occasionally temperate waters of the western Pacific from Fiji to the eastern Australian coast (including the Great Barrier Reef), the northwestern Australian coast and sometimes further, occasionally to the West Indian Ocean. Only one species (Trididemnum marmorattun) known from the western Pacific is not yet recorded further west than the northern coast of Australia. At present the possibly indigenous Australian species known only from tropical seas are few (five being present in this collection, see Table 2). Four of the species present (including three that are solitary, viz. Polycarpa aurita, P. papillata and Microcosmus helleri ) have a pan-tropical range. The Ascidiacea arc sessile (fixed) organisms with larvae free-swimming for relatively short periods and it is possible that they could be vulnerable to isolation and spcciation. In fact, collections (including the present one, see Table 2 below) made in shallow tropical waters, where species diversity is high, are dominated by colonial species, which are invariably internally fertilised, brooding their embryos internally and with larvae free-swimming for particularly short periods. Further, these characteristics of colonial ascidians are particularly conspicuous in the Didemnidae, the most speciose family in the tropical ascidian fauna, in which larvae arc free-swimming for periods of ten minutes or less. On the other hand, solitary species (with the exception of some species of Polycarpa and Molgida) are externally fertilised, embryos develop into tailed larvae in a planktonic phase, and be more likely agents of gene flow than the short-lived larvae of colonial species. However, although the characteristics of the life histories of colonial ascidian species (in particular, the Didemnidae), tend to support an hypothesis of population isolation resulting in their spcciation, the geographic range of the majority of shallow water tropical ascidian species (confirmed in the present collection) suggest that this hypothesis of isolation is incorrect; and that gene flow around the tropical Indian and West Pacific Oceans and through the straits that connect these oceans, separating the Australian continent from the islands to the north, is not impeded by the sessile habit of the adults, nor by internal fertilisation of eggs and brooding of embryos within the parental organism and a short free-swimming larval life. Probably there are other selective advantages of a colonial habit not directly related to gene flow, viz. patterns and rates of growth to rapidly occupy substrata and exclude other benthic organisms and a tendency to population P. Kott Table 1. Collection data for specimens reported on below (NTM, Museum and Art Gallery of the Northern Territory registration number; NCI, National Cancer Institute voucher specimen). All locations are the Northern Territory (Australia). NTM reg. no. Date collected Depth of Depth NCI No. Latitude Longitude Site Description collection (m) range (m) Habitat Substrate E500 0M9H2629-Y 11°47.89"S 136°29.04'E Cotton Island, 100 m off Northern Bay, English Company Islands. 30 Mar 04 14 13-16 Coral heads Fine sand E501 0M9H2633-F 11°47.89"S 136°29.04‘E Cotton Island, 100 m off Northern Bay, English Company Islands. 30 Mar 04 14 13-16 Coral heads Fine sand E502 0M9H2657-G 11°38.60'S 136°17.83'E Raragala Island, large bay on SW side of island, Wessel Islands. 30 Mar 04 18 17-20 Rocky Coarse sand, rubble E503 0M9H2668-R ll°33.9rs 136°22.34'E Raragala and Guluwuru Islands (Gugari Rip). Wessel Islands. 30 Mar 04 23 19-27 Channel Rock, sand E504 0M9H2669-S 11°38.60’S 136°I7.83'E Raragala Island, large bay on SW side of island, Wessel Islands. 30 Mar 04 18 17-20 Rocky Coarse sand, rubble E505 0M9H2670-T 11°32.85'S 136°21.28‘E Raragala Island, 700 m NE of tip, Wessel Islands. 31 Mar 04 14 13-16 Muddy bottom Mud, rock E506 0M9H2674-X 11°32.85'S 136°21.28'E Raragala Island, 700 m NE of tip, Wessel Islands. 31 Mar 04 14 13-16 Muddy bottom Mud, rock E507 0M9H2676-Z 11°06.69'S 136°41.26'E Trafalgar Bay Headland, Marchinbar Island. Wessel Islands. 31 Mar 04 10 9-12 Rocky reef Rock, sand E508 0M9H2677-A 11°06.69'S 136°41.26'E Trafalgar Bay Headland, Marchinbar Island. Wessel Islands. 31 Mar 04 10 9-12 Rocky reef Rock, sand E509 0M9H2678-C 1l°06.69'S 136°41.26'E Trafalgar Bay 1 leadland, Marchinbar Island, Wessel Islands. 31 Mar 04 10 9-12 Rocky reef Rock, sand E510 0M9H2687-N 1 r01.29'S 136°43.55 E Marchinbar Island, 900 m NW of Emu Islet. Wessel Islands. 31 Mar 04 18 11-17 Rocky reef Rock, sand E511 0M9H2718-V iri0.95'S 136°39.24 E Shark Point, 1 km W of Jensen Bay, Marchinbar Island, Wessel Islands. 02 Apr 04 18 17-20 Rocky reef Rock, silt, sand Hopeful Bay. 340 m NW of Breakwater 10-19 Rocky reef Mud, rock E512 0M9H2727-H 11°24.69'S 136°28.85'E Point. Marchinbar Island, Wessel Islands. 02 Apr 04 18 E513 OM9H2740-U 11°28.30'S 136°25.43'E Guluwuru Island, 580 m NE of tip, Wessel Islands. 03 Apr 04 18 14-15 Rocky Rock, silt E514 0M9H2741-V H°41.17'S 136°02.03’E Drysdale Island, 600 m offshore of SE side of island, Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E515 0M9H2742-W 11°41.17S 136°02.03'E Drysdale Island, 600 m offshore of SE side of island. Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E516 0M9H2743-X 11°41.17'S 136°02.03'E Drysdale Island, 600 m offshore of SE side of island, Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E517 0M9H2744-Y 11°41.17'S 136°02.03'E Dry sdale Island. 600 m offshore of SE side of island, Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E518 0M9H2745-Z 11°41.17'S 136°02.03’E Drysdale Island, 600 m offshore of SE side of island. Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E519 0M9H2746-A 11°41.17'S 136°02.03 E Drysdale Island, 600 m offshore of SE side of island, Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E520 0M9H2747-C 11°41.17'S 136°02.03'E Drysdale Island. 600 m offshore of SE side of island, Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E521 0M9H2748-F 11°41.17’S 136°02.03'E Drysdale Island, 600 m offshore of SE side of island, Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E522 0M9H2749-G 11°41.17'S 136°02.03 E Drysdale Island, 600 m offshore of SE side of island, Wessel Islands, 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E523 0M9H2750-H 11°41.17'S 136°02.03 E Drysdale Island. 600 m offshore of SE side of island. Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E524 0M9H2751-I 11°41.I7’S 136°02.03'E Drysdale Island, 600 m offshore of SE side of island. Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E525 0M9H2753-K 11°41.19'S I36°02.13'E Drysdale Island. 600 m offshore of SE side of island, Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral bommies, sand E526 0M9H2754-L H°41.19’S 136°02.13E Drysdale Island, 600 m offshore of SE side of island, Wessel Islands. 04 Apr 04 18 17-20 Coral heads Coral Bommies, sand E527 0M9H2768-Z 11°38.56'S 136°17.85'E Raragala Island, 240 m NE of bay on SW side, Wessel Islands. 05 Apr 04 18 11-20 Rocky reef Coral, sand, rubble E528 0M9H2782-Q 11°38.27'S 136°17.52E Raragala Island, 600 m NE of bay on SW side, Wessel Islands. 05 Apr 04 14 13-14 Sandy bottom Sand, mud E529 OM9H2783-R 11°38.27'S I36°17.52'E Raragala Island, 600 m NE of bay on SW side. Wessel Islands. 05 Apr 04 14 13-14 Sandy bottom Sand, mud E530 0M9H2802-N 11°23.26'S 136°29.36'E Marchinbar Island, 3 km N of Breakwater Point, Wessel Islands. 02 Apr 04 18 15-17 Rocky Mud, rock E531 0M9H2803-O 11°06.69”S 136°41.26'E Trafalgar Bay Headland, Marchinbar Island, Wessel Islands. 31 Mar 04 10 9-12 Rocky reef Rock, sand 64 Ascidians from the Wessel Islands Table 2. Species of ascidians taken from the Wessel Islands and English Company Islands, showing their previously known geographic range and NTM registration number(s) of relevant specimens. * IWP. Indo-West Pacific; NE, northeastern Australia; NW, northwestern Australia; WP, West Pacific. Taxon name Recent reference Previously known geographic range* NTM rcg. no. No. specs. Suborder Aplousobranchia Family Polycitoridae Polycitor circes Michaelsen, 1930 Kott 2004 WP, NE, NW and northern Australia E518 i Eudistoma ehoreum Kott, 1990a Kott 2008 NW, NE and northern Australia E502 i E513 i E524 i Eudistoma ovatuin (Herdman, 1886) Kott 2004 Northern Australia E508 3 E520 1 Eudistoma pyriforme (Herdman, 1886) Kott 1990a Torres Strait, NW, NE Australia E522 1 Family Didemnidac Leptoclinides braiuli Kott, 2001 Kott 2004 NE, northern Australia E514 1 Leptoclinidcs cf. rigidus Kott, 2001 Kott 2005a WP, NE, NW and northern Australia E521 1 Polysyncraton oceanium Kott, 2001 Kott 2004 WP and northern Australia E519 1 Didemnuin clavum Kott, 2001a Kott 2004 WP NW Australia E515 1 Didemnum molle (Herdman, 1886) Kott 2001 IWP tropieal/temperate Australia E527 1 + Didemnuinpsammaiode (Sluiter, 1895) Kott 2002 IWP tropical/temperate Australia E510 1 E517 1 Didemnum roberli Michaelsen, 1930 Kott 2004 NW, northern Australia E501 | Didemnum viride (Herdman, 1906) Kott 2002 IWP, NW, NE Australia E525 1 Trididemnum marmoratum (Sluiter. 1909) Kott 2004 WP, northern Australia E516 1 E509 1 Trididemnum savignyi (Herdman, 1886) Kott 2001 Pan-tropical, NW, NE Australia E526 1 Trididemnum sibogae (Hartmeyer 1910) Kott 2007 IWP tropical / temperate Australia E529 1 Lissoclinum badium F. and C. Monnot, 1996 Kott 2004 WP, NW, NE Australia E523 1 Lissoclinum multijidum (Sluiter, 1909) Kott 2004 IWP, northern and southern Australia E503 1 Suborder Phlebobranchia Family Ascidiidae Pliallusia arabica Savigny, 1816 Kott 1985 IWP, NW, NE Australia E530 1 Family Perophoridae Perophora modificata Kott, 1985 Kott 2004 WP, NE Australia, Ashmore Reef E507 1 Suborder Stolidobranchia Family Stvelidae (Subfamily Styelinae) Polycarpa aurita (Sluiter 1890) Kott 2008 Pan-tropical, NW, NE Australia E594 1 Polycarpa longiformis Tokioka, 1952 Kott 1985 NW, NE Australia, Japan E500 1 Polycarpa papillata (Sluiter, 1885) Kott 2006 Pan-tropical, tropical/temperate E505 1 Australia E512 1 Family Styelidae (Subfamily Polyzoinae) Stolonica australis Michaelsen, 1927 Kott 1985 Temperate Australia E528 1 Symplegma brakenhielmi (Michaelsen, 1904) Kott 2004 IWP and temperate Australia E531 1 Family Pyuridae Microcosmus helleri Herdman. 1882 Kott 1985 Pan-tropical, tropical/temperate E511 1 Australia E506 1 maintenance by concentration of recruits (resulting from reduction in the larval exposure to dispersal, albeit retaining larvae that are free-swimming for minimum periods for site selection). The maintenance of stable populations may enhance opportunities for fertilisation (both internal and external) that transcend the advantages of increasing gene flow through a longer larval life; and form links in chains of recruitment around the margins of the continents and between the islands and reefs of the western Pacific and Indian Oceans. Unfortunately, at this stage there is little data available on the patterns of life histories, population dynamics and selective advantages reflected in the dominant groups of the diverse ascidian fauna in the tropical Indo-West Pacific (see Kott 1985, 1990,2001; glossary entries for fertilisation and gene flow). 65 P. Kott BIOGEOGRAPHIC NOTES Generally the species of ascidians represented in this collection are those that would be expected in the north of Australia. They can be divided into the following groups: 1. Species with a wide pan-tropical range. Microcosmus helleri and Polycarpa papillata are known to extend also around Australia into temperate waters, but Trididemnum savignyi and Polycarpa aurita do not extend further south than the tropics. 2. Species known from the tropical West Pacific (including north-eastern Australia) and the Indian Ocean. For some species in this group, the only Indian Ocean record is offnorth-westem Australia ( Polycitor circes, Leptoclinides cf. rigidus, Didemnum clavum, Lissoclinum badium and Pewphora modificata), but others ( Didemnum molle, Didemnum viride, Phallusia arabica, Didemnum psammatode, Trididemnum sibogae and Symplegma brackenhielmi) extend to the West Indian Ocean and the last two species listed also extend from the tropics into temperate waters around the southern coast of Australia. Lissoclinum multifidum is recorded from the western Pacific, the Indian Ocean and northern Australia as well as southern Australia, but has not been recorded from north-eastern or north¬ western Australia and it is possible that more than one species is involved. 3. Only one species is known only from the western Pacific including northern eastern Australia (Trididemnum marmoratum ). 4. Species, possibly indigenous, known from the north of Australia, including the north-western and / or north-eastern coasts ( Eudistoma eboreum, E. ovatum, E. pyriforme, Leptoclinides brandi, Polysyncraton oceanium, Didemnum roberti, Polycarpa longiformis) have not been recorded outside Australian waters. 5. Stolonica australis is the only species not previously reported outside temperate Australian waters and the present new record from the Wessel Islands is an extension of its known geographic range. TAXONOMIC NOTES The specimens in this collection were photographed in situ by their respective collectors. Unfortunately, the photographed specimen (Figs 1-3) is not necessarily the same specimen as the one examined and reported on here. Polycitor circes Michaelsen, 1930 (Fig.l A) Polycitor circes Michaelsen, 1930: 495; Millar 1975: 205, part, specimens from Marongas (20.iii.14); Monniot 1988:207; F. and C. Monniot 1996: 184; F. and C. Monniot 2001: 249; Kott 1990a: 169 and synonymy; Kott 2002: 26. Not Millar 1975: 205 part, specimen from Marongas (19.iii. 14) and those from other locations (see Remarks, below). Distribution. Previously recorded (see Kott 1990a): Western Australia (north-western Australia, Shark Bay, Cockbum Sound); Queensland (Martha Ridgeway Reel); Northern Territory (Darwin), Papua New Guinea, New Caledonia, Indonesia, Philippines. New record: Northern Territory (Northern Territory (Wessel Islands, NTM E518). Description. The newly recorded colony is a firm gelatinous translucent vertical column, expanded slightly at the top where the test is slightly softer and more translucent than the firm test that forms the stalk. The surface of the colony is smooth and even. The zooids extend from the upper surface of the head where each zooid opens to the exterior by two separate 6-lobed apertures. They extend parallel to one another toward the base of the stalk. Zooids arc very long, with a long oesophageal neck. The abdomen continues past the pole of the gut loop and terminates in a vascular stolon. Longitudinal muscles extend along the length of the zooid and terminate in a short horn-like protrusion on each side of the anterior part of the posterior abdominal stolon (posterior to the pole of the gut loop). About 25 rows of stigmata are in the large thorax. The stomach, at the posterior end of the abdomen, has about 20 longitudinal folds. The gonads are in the gut loop and consist of a few large eggs and many small testis follicles scattered through and around the pole of the gut loop. Large embryos arc present moving up the oviduct as it extends up the oesophageal neck toward the atrial cavity. Larval adhesive organs are triradially arranged at the anterior end of the trunk. Remarks. The present colony resembles many assigned to Diazona futtgia (see F. and C. Monniot 2001, Fig. 123E). However it is readily distinguished by the absence of internal longitudinal vessels in the branchial sac, the conspicuous folds in the stomach (rather than the internal striations of Diazona) and by the embryos moving up through the abdomen (fertilisation having taken place in the base of the oviduct). Colonies from Deo Roa (Philippines) with small flat- topped lobes branching off basal stolons assigned to this species by Millar (1975) are probably specimens of Clavelitut arafurensis Tokioka, 1952. They arc quite different from the massive robust vertical lobes of the type specimens of the present species (Michaelsen 1930) and others described by Kott (1990a) from Western Australia. Another specimen Millar (1975) described from Deo Roa with smooth margins around the apertures appears to be a Pycnoclavella sp. Of the other specimens Millar (1975) assigned to this species, one from Marongas (20.iii.14, Millar 1975) appears to be properly assigned, although the other specimen from that location (19.iii.l3) without stomach folds may be a specimen of Polycitor translucidus. 66 Ascidians from the Wessel Islands Fig 1. A, Polycitor circes ; B, Eudistoma eboreum ; C, Eudistoma ovatunv, D, Leptoclinides cf. rigidus; E, Didemnum clavum', F, Didemnum psammatode. Photos: A, P. Colin; B, D-F, D. de Maria; C, M. Browne. 67 P. Kott The specimens assigned to Polycitor giganteus by F. and C. Monniot (2001) and Kott (2005b) from New Caledonia and the Solomon Islands, respectively, lacking stomach folds and having larval adhesive organs in a median vertical line, probably are also specimens of Polycitor translucidus. Polycitor giganteus Herdman, 1899 has zooids of similar proportions to the present species but is known mainly from temperate Australian waters and has only four stomach folds (sec Kott 1990a). Although not described, specimens from Papua New Guinea appear to have been accurately assigned to Polycitor circes by F. and C. Monniot (2001). Specimens from New Caledonia and Indonesia (F. Monniot 1988; F. and C. Monniot 1996) also appear to be members of the present species. Eudistoma eboreum Kott, 1990 (Fig. IB) Eudistoma eboreum Kott, 1990a: 205; Kott 2002: 27; Kott 2008: 1118. Distribution. Previously recorded (see Kott 2008): Western Australia (Kalbarri); Northern Territory (Darwin); Queensland (Lizard Island). New records'. Northern Territory (Wessel Islands, NTM E502, E513, E524). Description. The newly recorded colonies are purple- brown, translucent, firm gelatinous slabs with evenly distributed groups of about four to six zooids arranged in a circle with their atrial apertures opening to the surface in the centre of the circle and their branchial openings around the circumference. In the living colonies there is some variation in the extent to which these systems are depressed into the surface, although in preserved specimens the surface is more or less even. Also photographs of two of the newly recorded specimens show the colonies to have been a whitish colour and pale, but the other colony (NTM E513) is distinctly grey-blue. Collector’s notes describe one of these colonies (NTM E524) to have been brown-black internally. However, in preservative all three specimens are grey with diffuse brown pigment in the test of the upper half of the colony. Zooids have the characteristic layer of transverse muscles in the thoracic body wall that overlies strong longitudinal bands extending the length of the long zooids. Atrial apertures are on a long siphon. Three rows of stigmata are in the branchial sac, which is separated from the smooth-walled stomach at the posterior end of the zooid by a long oesophageal neck. A small vascular stolon is at the posterior end of the body. Remarks. Although the colour of living specimens of this species is variable, their colour in preservative is more or less the same. The robust zooids in circular rudimentary systems in the firm grey translucent slabs of test are characteristic of this species. It is known from relatively few specimens across the northern tropical coast of Australia. Eudistoma ovation Herdman, 1886 (Fig. 1C) Eudistoma ovation Herdman, 1886: 246; Kott 2004: 42 and synonymy. Not Kott 1990a: 222 and synonymy (see E. pyriforme below). Distribution. Previously recorded (see Kott 2004): Northern Territory (Torres Strait, Gulf of Carpentaria, Bynoe Harbour). New records : Northern Territory (Wessel Islands, NTM E508: 3 specimens, E520: 1 specimen). Description. Colonies are vertical to conical or rounded lobes to 2 cm diameter and 3 cm high, sessile or with a thick short stumpy basal stalk. Muddy-looking plant cells, faeces pellets and sand arc embedded in the soft test around the long robust zooids that extend from the upper surface toward the base of the colony. The zooids are arranged in circles with the long atrial siphons opening in the centre of the circle and the branchial openings in an outer circle. Zooids are muscular with an outer layer of transverse muscles and inner longitudinal bands along the length of the zooids, which extend from the upper surface toward the base of the colony. As is usual in this genus, the oesophageal neck is very long. In this species, the gut forms a spiral in the posterior end of the loop. Remarks. The spiral in the gut loop and the inclusions of various particles in the test are characteristic of this species. The former character distinguishes the species from the otherwise similar Eudistoma amplum (Sluiter, 1909). Polycitor multiperforatus Sluiter, 1909 also is a similar species, although it appears to have more stigmata per row (30) and the circular arrangement of zooids was not detected in the syntypes (see Kott 1990a). Eudistoma pyriforme (Herdman, 1886) Psammaplidium pyriforme Herdman, 1886: 419. Eudistoma pyriforme. — Kott 1990a: 226 and synonymy. Eudistoma ovation. -— Kott 1990a: 222. Distribution. Previously recorded (see Kott 1990a): Western Australia (Onslow); Queensland (Bundaberg); Torres Strait. New record'. Northern Territory (Wessel Islands, NTM E522). Description. The colony is a flat slab with the margins raised around the flattened upper surface. Fine sand is on the upper surface, interrupted where the zooids open separately to the exterior. Sand also is present throughout the colony. Some excrescences or root-like projections are around the sides of the colony enmeshed with sand and other particles. Zooids are contracted and the abdomen is contracted to about the same length as the thorax. Three rows of stigmata were detected in the branchial sac but other details of the zooids were not observed. Two small larvae arc in the thorax of the newly recorded specimen. Remarks. The present colony is not unlike those previously reported for this species. Although some sand and possibly symbionts are embedded in the test, it lacks the upright lobes, the conspicuous circular systems and Ascidians from the Wessel Islands the spiral at the distal end of the gut loop of E. ovatum, a species that often has been confused with it. Leptoclinides brandi Kott, 2001 Leptoclinides brandi Kott, 2001a: 40. — Kott 2004: 2468. Distribution. Previously recorded (see Kott 2004): Queensland (Great Barrier Reef); Northern Territory (Darwin and Bynoe Harbour). New record : Northern Territory (Wessel Islands, NTM E514). Description. The tough, fleshy-looking colony has a smooth surface raised into mounds and ridges. A layer of large, stellate spicules (to 0.08 mm diameter) with 9-13 moderately long and pointed conical rays in optical transverse section are in the surface of the colony but the remainder of the translucent test has only sparse spicules. In life, the colony is greyish white. The arrangement of the common cloacal systems is obscured by compression of the colony. Zooids are very contracted, although the branchial siphon is long with a bulbous expansion halfway along it. The atrial siphon is long and oriented posteriorly. Stigmata are relatively short in these contracted zooids but they appear to be pointed at each end. About six male follicles are in a circle and the vas deferens coils five times. Remarks. The newly recorded colony resembles those previously described, although its condition obscures some of its characters, especially the fusiform stigmata. The stigmata and their distribution arc characteristic of the species and these together with the testis follicles and coils of the vas deferens distinguish the species from those in the dubius group, which resemble the present colony in some respects. Leptoclinides cf. rigidus Kott, 2001 (Fig. ID) Leptoclinides rigidus Kott, 2001 a: 77; Kott 2005a: 2425 and synonymy. Distribution. Previously recorded (see Kott 2005a): Western Australia (Ashmore Reef, Montebello Islands); Queensland (Great Barrier Reef); Northern Territory (Darwin, Wessel Islands); Papua New Guinea. New record: Northern Territory (Wessel Islands, NTM E521). Description. Colonies arc gelatinous sheets overgrowing the substrate. The small zooids are arranged along each side of canals that surround zooid-free circular areas about 3 mm diameter, each containing a patch of spicules beneath a superficial bladder cell layer. Spicules are sparse in the remainder of the test. Large sessile common cloacal apertures are at the junctions of some of the common cloacal canals, which all contain some faecal pellets. The spicules are to 0.05 mm diameter with 7-9 conical rays in optical transverse section. Some of the ray tips are relatively blunt. In life, the colonies arc a reddish brown. Little of the zooid structure can be determined. As in all species of this genus a posteriorly orientated atrial siphon is present and a retractor muscle was not detected Remarks. Superficially the specimen resembles the temperate Leptoclinides rigidus , which has been recorded previously from this location. However specimens previously assigned to the species are reported to have been blue in life. The present specimen resembles Didemnum viride in life, but it could be an undescribed species of the genus Leptoclinides. Leptoclinides volvus Kott, 1975 has similar common cloacal systems, but it is a stalked colony and is also a temperate species. Polysyncraton oceanium Kott, 2001 Polysyncraton oceanium Kott, 2001a: 115. Distribution. Previously recorded (see Kott 2001a): Queensland (Great Barrier Reef); Fiji. New record: Northern Territory (Wessel Islands, NTM E519). Description. The colony is an investing sheet with firm translucent test. A thin layer of evenly spaced but not crowded spicules is in the surface test beneath a superficial layer of bladder cells. A sparse layer of spicules also is on the base of the colony but there are no spicules in the test between the surface and basal layers. Spicules are stellate, to 0.035 mm diameter, with 11-15 short pointed rays that break up readily. A common cloacal cavity was not detected in this specimen, although the zooids sometimes were seen to be arranged in double rows, presumably along each side of canals. Five coils of the vas deferens were detected, but the gonads were not seen, although some large eggs are present in the basal test. Remarks. The form and distribution of the spicules resemble those of Didemnum caesium Sluiter, 1909, Polysyncraton rica Kott, 2001. P otuetue C. & F. Monniot, 1987 and P. scobinum Kott, 2001, but they are smaller. The specimen generally conforms to those previously assigned to Polysyncraton oceanium. Didemnum davitm Kott, 2001 (Fig. IE) Didemnum clavum Kott, 2001a: 163; Kott 2004: 53. Distribution. Previously recorded (see Kott 2004): Western Australia (NW Australia to Port Hedland); Northern Territory (Darwin); Indonesia. New record: Northern Territory (Wessel Islands, NTM E515). Description. The colony is the characteristic branched structure with red pigment particles in the surface and spicules crowded throughout. The spicules are small, stellate, with occasional giant spicules, with 4-6 long pointed rays, scattered amongst them Remarks. Although other tropical Didemnum species are a similar red-pink colour, the present species is readily identified by the form of the colonies and the occasional giant spicules with relatively few rays scattered amongst the smaller stellate spicules. 69 P. Kott Didenutum /nolle (Hcrdman, 1886) Diplosomoides molle Herdman, 1886: 310. Didemnum molle. — Kott 2001: 208 and synonymy. Distribution. Previously recorded (see Kott 2001): NW, NE and northern Australia to Cockburn Sound and one record from Esperance. Western Pacific, the Indian Ocean (to Mauritius, Madagascar), Vietnam. New record: Northern Territory (Wessel Islands, NTM E527). The newly recorded location is well within the vast geographic range of this shallow reefal species, one of the most frequently recorded of the didemnid-Proc/i/oron symbioses. Description. The species is readily identified, forming vase-shaped colonies with a central test core separated from the outer zooid-bearing layer of the colony by a vast common cloacal cavity lined by green symbiotic cells, which can be seen through the large terminal common cloacal aperture. The species characteristically secretes vast quantities of mucus in which the green symbiotic cells are liberated from the colony when it is disturbed. Remarks. The species is discussed in Kott (2001). Didemnum psammatode (Sluiter, 1895) (Fig. IF) Leptoclinum psamathodes Sluiter, 1895: 171. Didemnum psammatode. — Kott 2002: 38. Distribution. Previously reported (see Kott 2002): a wide recorded range in the Indian and Pacific tropical/ temperate oceans between Sri Lanka, the Red Sea to Fiji and extending to southern Australia and the South China Sea. New records'. Northern Territory (Wessel Islands, NTM E510, E517) Description. The newly recorded specimens are as previously described, with mud-coloured faecal pellets crowded in the test and minute spicules around each zooid opening. Didemnum roberti Michaelsen, 1930 (Fig. 2A) Didemnum roberti Michaelsen 1930: 516; Kott 2004: 58 and synonymy. Distribution. Previously recorded (see Kott 2004): Western Australia (north west to Cockburn Sound); Northern Territory (Gulf of Carpentaria, Darwin, Torres Strait). New record : Northern Territory (English Company Islands, NTM E501). The species is commonly recorded from the northwestern coast of the continent and from Darwin, the Gull of Carpentaria and Torres Strait, but it is not yet reported from northeastern Australia or from locations in the Western Pacific. At this stage, its records suggest that it is one of the few indigenous Australian tropical species. Description. The colony is robust with terminal large circular common cloacal apertures on the rounded surface swellings, branches and ridges. The surface is always smooth and even. Spicules are in a relatively sparse layer at the surface and are more sparsely distributed and often patchy internally. They also are in a sparse but even layer lining the extensive common cloacal cavity that separates the central test core from the surface zooid-bearing layer of the colony. Zooids arc in clumps, each attached by a short basal connective that traverses the common cloacal cavity from the centre of the under surface of each clump. Each zooid has a fine retractor muscle. The atrial aperture is large, open and sessile and the vas deferens coils eight times around the undivided testis follicle. Larvae, with four pairs of lateral ampullae along each side of the antero-median adhesive organs, are embedded in the basal or central test. Remarks. This species has robust but variable colonies with a similarly variable colour pattern. It resembles many species of Leptoclinides in its smooth outer surface and well-developed common cloacal systems with a large terminal common cloacal aperture and an extensive posterior abdominal cavity separating a central test core from an outer zooid-bearing layer. Assignation of the species to the genus Didemnum is confirmed by characteristics of zooids, such as the large open sessile atrial apertures; the fine tapering retractor muscle; and the undivided testis surrounded by the coiled vas deferens. Didemnum viride (Herdman, 1906) (Fig. 2B) Leptoclinum viride Herdman, 1906: 340. Didemnum viride. — Kott 2001a: 9. Distribution. Previously reported (see Kott 2001a): Western Australia (Montebello Islands); Queensland (Great Barrier Reef); Papua New Guinea, New Caledonia, French Polynesia, Indian Ocean (Sri Lanka). New record: Northern Territory (Wessel Islands, NTM E525). Description. In preservative, the newly reported colony is a thin cream-coloured sheet with the surface marked by a mosaic of small elevated areas separated by narrow depressions over the common cloacal canals that are lined on each side by zooids. In life, the colony is brownish red in the depressed areas over the common cloacal canals. Spicules, to 0.05 mm diameter, are crowded in the basal half of the colony but are less crowded in the zooid layer. They have 7- 11 conical rays with more or less rounded tips in optical transverse section. Remarks. The specimen generally resembles Didemnum poecilomorpha F. & C. Monniot, 1996 although it lacks the two different types of spicules of the latter species. In the present collection the colony resembles that of Leptoclinides cf. rigidus (see above), but the oval elevations that form the mosaic on the surface of the colony are smaller in the present species. 70 Ascidians from the Wessel Islands A C E Fig 2. A, Didemnum roberti; B, Didemnum viride; C, Trididemnum marmoratum ; D, Trididemnum sibogae; E, Lissoclinum badiunv F, Lissoclinum multifidum. Photos: A,B,D,E, D. de Maria; C,F, P. Colin. 71 P. Kott Trididemttum marmoratum (Sluiter, 1909) (Fig. 2C) Leptoclinum marmoratum Sluiter, 1909: 84. Trididemnum marmoratum. — Kott 2002: 38; Kott 2004: 61. Distribution. Previously reported (see Kott 2004): Northern Territory (Gulf of Carpentaria, Darwin); Indonesia. New records'. Northern Territory (Wessel Islands, NTM E509, E516). Description. One of the newly reported colonies (E5 1 6) has relatively uniform rounded elevations about 1.0 cm diameter at the base and up to 0.5 cm high, each with a terminal common cloacal aperture. The other specimen is an extensive encrusting sheet. Stellate spicules, to 0.09 mm diameter, with 9-13 conspicuously pointed or chisel¬ shaped rays in optical transverse section are in a thin layer beneath the superficial bladder cell layer but are absent from the remainder of the colony. Primary posterior abdominal common cloacal spaces beneath the surface layer of zooid- bearing test are traversed by clumps of zooids and surround a central test core in each of the conical elevations. Zooids are robust, with a posteriorly orientated atrial siphon opening into the posterior abdominal common cloacal cavity. Sometimes they have black squamous epithelium on the thorax and abdomen, although this was not always detected. Occasionally there is also a black endostylar pigment cap. A short retractor muscle is present from the posterior end of the thorax. A large spherical male follicle is surrounded by at least eight coils of the vas deferens. Larvae with a narrow waist behind a corona of eight club- shaped ectodermal ampullae with bifid tips on each side of the three antero-median adhesive organs are in the central test core. Remarks. The specimens resemble closely those previously assigned to this species by Kott (2002). Trididemnum savignyi (Herdman, 1886) Didemnum savignyi Herdman, 1886: 281. Trididemnum savignyi. — Kott 2001a: 64. Distribution. Previously reported (see Kott 2001a): Western Australia (Nares Rock); Queensland (Great Barrier Reef); Northern Territory (Darwin); tropical Atlantic. New record: Northern Territory (Wessel Islands, NTM E526). Description. The newly recorded colony is an extensive thin sheet growing over a spiral worm tube, although the in situ photograph of this specimen is difficult to reconcile with its appearance in preservative. A sparse layer of large spicules, to 0.09 mm diameter, with 9-11 conical pointed rays in optical transverse section is beneath a superficial bladder cell layer. Spicules also line the oesophageal common cloacal canals but are absent from the remainder ol the colony. Faecal pellets arc embedded in the basal layer of test. Zooids have large funnel-shaped atrial siphons from about halfway down the dorsal surface of the thorax. Sometimes abdomina have dark squamous epithelium. Remarks. The newly recorded specimen resembles previously reported specimens of this species, although the dark pigment previously thought to be characteristic is not present. This is presumably associated with intraspecific variability in this widely distributed species. Trididemnum sibogae (Hartmeyer, 1910) (Fig. 2D) Didemnum sibogae Hartmeyer, 1910: 261. Trididemnum sibogae. — Kott 2001a: 283 and synonymy; Kott 2007: 1203 and synonymy. Distribution. Previously recorded (see Kott 2007): Western Australia (Cape Jaubert); South Australia (Cape Jaffa); Victoria (Western Port); Tasmania (Port Davey); New South Wales (Port Hacking, Port Jackson, Arrawarra); Queensland (Great Barrier Reef and mainland locations between Fraser Island and Princess Charlotte Bay); Northern Territory (Darwin, Gulf of Carpentaria). Indonesia, New Caledonia, Indian Ocean (Gulf of Manaar). New record'. Northern Territory (Wessel Islands, NTM E529). The records indicate this species to be common in temperate as well as tropical waters around Australia. Description. The newly recorded colony is a fist-sized three-dimensional reticulum (sec Kott 2001a, Fig. I32A) with thin branches overgrowing and fusing with one another to form the sponge-like consistency of the colony. An in situ photograph shows the colony to be creamish-ycllow. The colony has a superficial bladder cell layer and spicules are scattered relatively sparsely through the colony. An extensive posterior abdominal common cloacal cavity separates the zooid-bcaring surface layer of test to the central test in each of the branches. Spicules are relatively small, to 0.06 mm diameter, stellate, with 7-11 conical, sharply pointed rays in optical transverse section. Zooids are small, with a short posteriorly oriented atrial siphon. Remarks. The large, 0.016 mm diameter, spicules reported to occur occasionally in this species were not detected in the present specimen, although otherwise its characters conform to those previously reported. Lissoclinunt badium F. and C. Monniot, 1996 (Fig. 2E) Lissoclinum badium F. and C. Monniot, 1996: 170; Kott 2004: 69. Distribution. Previously recorded (see Kott 2004): Western Australia (Bonaparte Archipelago); Queensland (Great Barrier Reef); Northern Territory (Darwin); Coral Sea, Timor Sea, Palau Is. New record : Northern Territory (Wessel Islands, NTM E523). Description. The characteristic soft flexible specimen, dark brown internally, has large common cloacal apertures on surface ridges that are prominent on the living inflated colony. Specimens of this species arc readily identified. Brown pigment cells are mixed with the especially small spicules in the surface of the colony and brown coloured zooids can be seen crossing the extensive thoracic common 72 Ascidians from the Wessel Islands cloacal cavity. The basal half of the colony is opaque with crowded spicules but the abdomina of the zooids are embedded in the floor of the common cloacal cavity where the spicules arc less crowded. Lissodinum miiltifidum (Sluiter, 1909) (Fig. 2F) Leptodinum multifidum Sluiter, 1909: 311. Lissoclinum multifidum. — Kott 2004: 65 and synonymy. Distribution. Previously recorded (see Kott 2004): South Australia (Flinders Island, etc); Tasmania (Forestier Peninsula); Northern Territory (Darwin); Indonesia, Gulf of Thailand, Mauritius. New record: Northern Territory (Wessel Islands, NTM E503). Description. T he colony is said to have been brown in life although in the in situ photograph it looks red. It encrusts rubble, some of which is mixed with the basal or central test mass. This is surrounded by an extensive common cloacal cavity crossed by test commissures that attach it to the zooid-bearing layer of test. Large common cloacal apertures are on surface elevations. Small sparse spicules, to 0.025 mm diameter, with a range of spicule rays from conical to rounded or rod-like are scattered through the test together with morula cells and possibly some plant cells as described previously for this species. Remarks. Although Kott (2001a) had thought this species to be aspicular, it docs in fact have minute spicules scattered through the test. Phallusia arabica Savigny, 1816 (Fig. 3A) Phallusia arabica Savigny, 1816: 164; Kott 1985: 61; Kott 2001b: 61 and synonymy. Distribution. Previously recorded (see Kott 1985): Queensland (from the Capricorn Group, southern Great Barrier Reef, to Trinity Bay), Arafura Sea, Philippines, Sri Lanka, Red Sea, Gulf of Suez. New record: Northern Territory (Wessel Islands, NTM E530). Description. The newly repotted specimen has a robust, slightly curved body attached to the substrate about halfway down the convex ventral surface with a short conical atrial protuberance from the concavity about halfway down the dorsum. Anteriorly the body narrows to the terminal branchial aperture. The test is smooth, translucent and relatively firm and thick, and both atrial and branchial siphons appear to be linn and inflexible. However, internally the siphons are long and muscular, and the body lies loosely in the test and does not adhere to it. A mesh of longitudinal and transverse muscles is over the right side of the body and on the left a similar mesh is present but only anterior to the gut loop. Branchial tentacles are short and darkly pigmented in the preserved specimen. The long dorsal ganglion is at the base of the atrial siphon. A small U-shaped slit is on the dorsal tubercle at the anterior end of the dorsal lamina but a peritubercular-V is shallow. Very small secondary openings of the neural duct open into the peribranchial cavity along the left side of the dorsal lamina, which has conspicuous ribs along the left side of the membrane. The gut, on the left side of the posterior half of the body, is filled with mud. The rim of the anal opening is divided into about 20 short rounded scallops. Remarks. This robust, widely ranging species is readily distinguished from Phallusia obesa (Herdman, 1880), P. julinea Sluiter, 1919 and P millari Kott, 1985, all commonly occurring species of this genus in the Indo-West Pacific, by the smooth surface of the present species, the lack of chromatophores in the test, the presence of the dorsal ganglion at the base of the atrial aperture rather than halfway between the atrial opening and the dorsal tubercle and the lack of the sandy holdfast characteristic of P. millari. Perophora modificata Kott, 1985 (Fig. 3B) Perophora modificata Kott, 1985: 104; Kott 2004: 40 and synonymy. Distribution. Previously reported (see Kott 2004): Western Australia (Ashmore Reef); Queensland (Great Barrier Reef), Coral Sea Plateau; Palau Islands, Philippines. New record: Northern Territory (Wessel Islands, NTM E507). Description. The newly recorded specimen is a small clump of rounded yellow zooids each on a long stalk attached to a basal mat of stolons. The branchial sac has four rows of stigmata. The zooids in this and all other respects conforms to previous accounts of this readily identified species. Polycarpa aurita (Sluiter, 1890) (Fig. 3C) Styela aurita Sluiter, 1890: 338. Polycarpa aurita. — Kott 1985: 152 and synonymy; Kott 2008: 1199. Distribution. Previously reported (see Kott 1985): Western Australia (Cape Jaubert, Shark Bay to Cockbum Sound); New South Wales (Port Jackson); Queensland (Moreton Bay to Lizard Island); Northern Territory (Gulf of Carpentaria); Indonesia, New Caledonia, Philippines; Atlantic Ocean (Venezuela, Gulf of Mexico, Caribbean). New record: Northern Territory (Wessel Islands, NTM E504). The species has not yet been reported from the eastern Pacific, the Indian or the eastern Atlantic Oceans, but otherwise it is pan-tropical. Description. The newly recorded specimen is robust with a hard leathery test. It is large and appears to be senescent, the gonads being degenerate, only traces being detected embedded deeply in the body wall. The species is identified principally by its large gut, large elliptical stomach, the strong ligament joining the two limbs of the primary loop, the long crowded vertical endocarps between the limbs of the gut loop and, on the body wall, the small 73 P. Kott rounded scallops lining the rim of the anus and the deep conspicuous slit on the dorsal tubercle. Remarks. The specimen is cryptic, the leathery wrinkled test being covered with epibionts and in the field it can be confused with other stolidobranch species in this collection, such as Microcosmus helleri, although it does not appear to form aggregates as many of these species do. Polycarpa longiformis Tokioka, 1952 (Fig. 3D) Polycarpa longiformis Tokioka, 1952: 119; Kott 1985: 170 and synonymy. Distribution. Previously reported (see Kott 1985): Western Australia (Port Hedland); Queensland (Martha Ridgeway Reef, Lizard Island, Cape Weymouth, Raine Island, Murray Island); Arafura Sea, Japan. New record : Northern Territory (English Company Islands, NTM E500). Description. The newly reported specimen is large (about 4 cm long), almost cylindrical, grey and rubbery- looking with a smooth surface. The test, though thin, is firm and the muscular body wall adheres closely to it. The apertures are both anterior on conical protuberances diverging from one another. The lobes of the aperture are completely obscured by the firm gelatinous test. Posteriorly the body narrows to a holdfast attached to rocks and other nibble. The branchial folds are narrow and widely separated with 5-7 internal longitudinal vessels in the interspace and the meshes are long. The gut forms a short loop in the posterior end of the body and the rectum continues anteriorly to the base of the atrial siphon. The anal border has long, rounded lobes. Loosely attached long polycarps in an uneven row down each side of the body are obscured by the long vertical teardrop-shaped cndocarps crowded on the body wall. Remarks. In the field, the gelatinous smooth conical protuberances with their terminal apertures projecting from crevices can be mistaken for the very similar structures in Phallusia species (P. julinea, P. arabica and P. obesa) and in specimens of Polycarpa papillata. The close relationship with P. papillata is discussed by Kott (1985). Polycarpa papillata Sluiter, 1885 (Fig. 3E) Styela (Polycarpa) papillata Sluiter, 1885: 192. Polycarpa papillata. — Kott 1985: 184 and synonymy; Kott 2006: 218. Distribution. Previously reported (see Kott 1985): Western Ausralia (Dampier Archipelago to Cockburn Sound); South Australia (St Vincent Gulf); Victoria (Portland); New South Wales (Port Jackson); Queensland (Morcton Bay to Bathurst Island); Northern Territory (Gulf of Carpentaria, Darwin); Arafura Sea, Indonesia, Palau Islands, Philippines, Marianas Islands, Sri Lanka, Madagascar. New records : Northern Territory (Wessel Islands, NTM E505, E512). Description. Long, tough specimens, pink in preservative with faint pink longitudinal stripes. The outer test is smooth and even without ridges or creases. The thin but muscular body wall is closely adherent to the test. The body narrows to the anterior terminal branchial opening which is turned to the right and slightly downwards. The atrial aperture is on a conical protrusion projecting out at an angle to the body about one-third of the distance down the dorsal surface. Branchial folds are long and narrow and widely separated from one another. The body wall has crowded upright leaf¬ like cndocarps, flattened and sometimes indented around their rounded margins and constricted at their bases where they join the body wall. They are especially crowded around the gut loop in the posterior part of the body. The anus has 16 long and slightly frilly anal lobes around the margin. Long polycarps in about three irregular rows are embedded in the body wall or attached to it by a long ligament at their proximal ends, but the dorsal end of each polycarp with its terminal gonoducts is free of the body wall and directed toward the atrial opening. Remarks. The species is common in the tropical Indo- West Pacific and although it resembles P longiformis, it is readily identified by its red colour and longitudinal red stripes. It often has ejected its gut when collected although in the newly recorded specimens the body organs are all present. Stolonica australis Michaelscn, 1927 Stolonica australis Michaelscn, 1927: 202; Kott 2003: 1642 and synonymy. Distribution. Previously recorded (sec Kott 2003): Western Australia (Albany); South Australia (Great Australian Bight, St Vincent Gulf, Investigator Strait, Kangaroo Island, Yorke Peninsula); Tasmania (Bruny Island. d’Entrecasteaux Channel); Victoria (Bass Strait, Anglesea, Portland, Western Port); New South Wales (Port Jackson). New record: Northern Territory (Wessel Islands, NTM E528). Description. Zooids are vertical with a terminal branchial aperture and the atrial aperture antero-dorsal, each on a slight conical protuberance. Zooids are joined to one another and to basal stolons by short lateral commissures to form tight aggregations that are brownish pink in life. The test is thin but rigid. Two branchial folds and a possible third ventral rudimentary fold are on each side of the body. A typical branchial formula is 9DL4(12) 5(9) 9E. A total of about 40 internal longitudinal vessels are on each side. The stigmata are in about 20 rows. The dorsal tubercle is a longitudinal slit. About 25 longitudinal folds are in the stomach wall, of which six terminate on each side of the suture line. Only a short caecum projects from the pyloric end of the suture line into the gut loop. Small cndocarps are scattered on the body wall and in the gut loop. Small male follicles are scattered in an arc around the postero-ventral curve of the right side of the body. 74 Ascidians from the Wessel Islands Fig 3. A, Phallusia arabica; B, Perophora modificata ; C, Polycarpa aurita ; D, Polycarpa longiformis : E, Polycarpapapillata: F, Microcosmus helleri. Photos: A-C, D. dc Maria, D,E, P. Colin; F, B. Alvarez de Glasby. 75 P. Kott Remarks. The present species has formerly been regarded as having a range limited to temperate waters, but the newly recorded material suggests that the range may be more extensive and include tropical waters. It does have some similarities with Stolonica alula Kott, 1985 (recorded from the south-western part of Western Australia), which has a similar zooid with a long vertical slit on the dorsal tubercle and similar distribution of gonads around the postero-ventral curve of the body, but three folds on each side of the body, a longer gastric caecum and a gastric spur that extends the stomach into the gut loop. Symplegma brakenhielmi (Michaelsen, 1904) Diandrocarpa brakenhielmi Michaelsen, 1904: 50. Symplegma brakenhielmi. — Kott 2004: 71. Distribution. Previously recorded (sec Kott 2004): Western Australia (Cape Preston to Cockburn Sound); Queensland (Moreton Bay to Martha Ridgeway Reef). Indonesia, Noumea, Palau Islands, Fiji, Thailand, Hong Kong, China, Indian Ocean including Sri Lanka. New record'. Northern Territory (Wessel Islands, NTM E531). Description. The colony forms a large robust sheet over the substrate. Zooids are dorso-ventrally flattened and completely embedded in the thin test forming a mosaic like pattern seen from the upper surface. Apertures are sessile on the upper (dorsal) surface of each zooid. Characteristic of the genus are the four longitudinal branchial vessels on each side of the branchial sac. The gut forms a tight double loop on the left side of the body and the stomach is large, but relatively short and wide with a conspicuous curved caecum in the pole of the primary gut loop. The two testis follicles, one anterior and one posterior to the small ovarian sac in the middle of the body wall, are lobed. Larvae are in the peribranchial cavity of the newly recorded specimen. The species and its relationships are discussed in Kott (2004). Microcosmus Itelleri Herdman, 1882 (Fig. 3F) Microcosmus helleri Herdman, 1882: 131; Kott 1985: 349. Distribution. Previously recorded (see Kott 1985): Western Australia (Port Hcdland, Cockburn Sound); South Australia (St Vincent Gulf); Indonesia. Java Sea, Singapore, Sri Lanka, West Indian Ocean, West Indies. New records'. Northern Territory (Wessel Islands, NTM E506, E511). The species is pan-tropical but also extends into temperate waters around the southern coast of Australia. Description. The newly recorded specimens, up to about 9 cm long, are tough, hard and leathery, with wrinkled test and some scattered epibionts. The branchial aperture is terminal and turned slightly ventrally. Internally the atrial siphon extends from near the base of the branchial siphon and turns posteriorly, lying at almost 180 degrees to it, opening in the sessile external atrial aperture about halfway down the dorsal surface of the body. Hard, pointed tubercles of the test arc around both external openings. Four hard, rounded and slightly spoon-shaped valves are around the base of the branchial siphon and each is folded back against the lining of the siphon from its base just anterior to the branchial tentacles. Siphonal spines were not detected in these specimens. The dorsal tubercle has the double spiral slit characteristic of several species in this genus. Six wide branchial folds are on each side of the body. The gut forms a narrow loop around the ventral margin of the left side and the gonads, divided into three large blocks embedded in the body wall pass from inside the pole of the gut loop, cross its descending limb and continue alongside anterior to the gut loop to the atrial aperture. Remarks. The course of the left gonad, crossing from inside the pole of the loop and extending close to the rectum, occurs also in M. exasperatus Heller, 1878 and M. squamiger Michaelson, 1927. However, this large, tough pan-tropical species, which extends into temperate waters around the southern Australian coast, is distinguished by the lack of siphonal spines, the presence of four rather cartilaginous spoon-like valves in the base of the branchial siphon and the six branchial folds on each side of the body. ACKNOWLEDGMENTS The specimens were collected by Dr Belinda Alvarez de Glasby of the Museum and Art Gallery of the Northern Territory and her party (M. Browne, H. Nguyen, D. de Maria and R Colin) under the ‘Collection and Taxonomy of Shallow Water Marine Organisms’ program for the US National Cancer Institute (Contract N02-CM-27003). Access to the area was kindly granted by Mr Terry Yumbulul, who also provided the team with advice on collection sites. I am grateful to Dr Alvarez de Glasby for making the collection and the in situ photographs of the specimens available to me. I am grateful also to the Trustees and the CEO of the Queensland Museum (Dr Ian Galloway) for providing the accommodation and other infrastructure that make it possible for me to pursue this work. REFERENCES Hartmeyer, R. 1909-1911. Ascidien (continuation of work by Sccligcr). Pp. 1281-1772. In: Bronn, H.G. (ed.) Klassen und ordnungen des Tierreichs vol. 3, suppl., part 89-98. C.F. Winter: Leipzig. [Abstract, repeating lists of species by Schepotieff, A. 1911, published in Archives de Zoologie Experimental et Generate Naturgeschicht 6: 3-27], Herdman, W.A. 1880. Preliminary report on the Tunicata of the Challenger Expedition. Ascidiidac. Proceedings of the Royal Society Edinburgh 10( 1): 458^172. Herdman, W.A. 1882. Report on the Tunicata collected during the voyage of H.M.S. Challenger during the years 1873-76. Pt. I, Ascidiae simpliees. Report on the Scientific Results of the Exploring voyage of H.M.S. Challenger duhng the years 1873- 1876 Zoology 6: 296 pp., 37 pis 6: 296 pp, 37 pis. 76 Ascidians from the Wessel Islands Herdman. W.A. 1886. Report on the Tunicata collected during the voyage of H.M.S. Challenger during the years 1873-76. Pt. II, Ascidiae compositae. Report on the Scientific Results of the Exploring voyage of H.M.S. Challenger during the years 1873-1876 Zoology 14(38): 1-425. Herdman. W.A. 1899. Descriptive catalogue of the Tunicata in the Australian Museum. Australian Museum, Sydney, Catalogue 17: 1-139. Herdman, W.A. 1906. Report on the Tunicata. Ceylon Pearl Oyster Fisheries Supplementary Report 39: 295-348. Kott, P. 1985. The Australian Ascidiacea Pt 1, Phlebobranchia and Stolidobranchia. Memoirs of the Queensland Museum 23: 1-440. Kott, P. 1990a. The Australian Ascidiacea Pt 2, Aplousobranchia (1). Memoirs of the Queensland Museum 29( 1): 1-266. Kott, P. 1990b. The Australian Ascidiacea, Phlebobranchia and Stolidobranchia, supplement. Memoirs of the Queensland Museum 29(1): 267-298. Kott, P. 1992. The Australian Ascidiacea, supplement 2. Memoirs of the Queensland Museum 32(2): 621-655. Kott, P. 2001a. The Australian Ascidiacea Pt4, Didemnidae. Memoirs of the Queensland Museum 47(1): 1-410. Kott, P. 2001b. The Australian Ascidiacea, Corrigenda. Memoirs of the Queensland Museum 47(1): 411-413. Kott, P. 2002. Ascidiacea (Tunicata) from Darwin, Northern Territory, Australia. The Beagle, Records of the Museums and Art Galleries of the Northern Territory’ 18:19-55 Kott, P. 2003. New syntheses and new species in the Australian Ascidiacea. Journal of Natural History 37: 1611-1653. Kott, P. 2004. Ascidiacea (Tunicata) in Australian waters of the Timor and Arafura Seas. The Beagle, Records of the Museums and Art Galleries of the Northern Territory 20: 37-81. Kott, P. 2005a. New and little known species of Didemnidae (Ascidiacea, Tunicata) from Australia (part 3). Journal of Natural History 39(26): 2409-2479. Kott, P. 2005b. Ascidians from the Solomon Islands. The Beagle, Records of the Museums and Art Galleries of the Northern Territory 2 1: 55-68. Kott, P. 2006. Observations on non-didemnid ascidians from Australian waters (1) Journal of Natural History 40(3-4): 169-234. Kott, P. 2007. New and little known species of Didemnidae (Ascidiacea, Tunicata) from Australia (Part 4). Journal of Natural History 41 (17-20): 1163-1211. Kott, P. 2008, Ascidiacea (Tunicata) from deep waters of the continental shelf of Western Australia. Journal of Natural History 42(15-16): 1103-1217. Michaclsen, W. 1904. Revision der compositcn Styeliden oder Polyzoinen. Jahrbuch der Hamburgischen wissenschaftlichen Anstalten 21(2): 1-124. Michaelsen, W. 1927. Einigc ncue vvestaustralischc ptychobranchiate Ascidien. Zoologisches Anzeiger 1 1: 193-203. Michaelsen, W. 1930. Ascidae Krikobranchiae. Fauna Sudwest- Australiens 5(7): 463-558. Millar. R.H. 1975. Ascidians from the Indo-West Pacific region in the Zoological Museum, Copenhagen (Tunicata, Ascidiacea). Steenstrupia 3(20): 205-336. Monniot, F. 1988. Ascidies de Nouvcllc Caledonie V. Polycitoridae du lagon. Bulletin du Museum National d 'Histoire Naturelle. Paris scrie 4 10A(2): 197-235. Monniot, F. and Monniot, C. 1996. New collections of ascidians from the western Pacific and southeastern Asia. Micronesica 29(2): 133-279. Monniot, F. and Monniot, C. 2001. Ascidians from the tropical western Pacific. Zoosystema 23(2): 201-383. Savigny, J.C. 1816. Memoires stir les animaux sans vertebres. Part 2. Paris, pp. 1-239. Sluiter, C.P. 1885. Ueber einige einfachen Ascidien von der Inscl Billiton. Natuurkundig Tijdschrift voor Nederlandsch Indie 45: 160-232. Sluiter, C.P. 1890. Die Evertebraten aus der Sammlung des Koniglichen Naturwissenschaftlichen Vercins in Niederlandisch Indien in Batavia. Natuurkundig Tijdschrift voor Nederlandsch Indie 50: 329-48. Sluiter, C.P. 1909. Die Tunicaten der Siboga Expedition Pt. 2. Die merosomen Ascidien. Siboga Expeditic. Siboga Expedition 56B: 1-112, Tokioka, T. 1952. Ascidians collected by Messrs Renzi Wada and Seizi Wada from the Pearl Oyster bed in the Arafura Sea in 1940. Publications of the Seto Marine Biological Laboratory 2(2): 91-142. Accepted 10 November 2008 77 * The Beagle, Records of the Museums and Art Galleries of the Northern Territory, 2008 24: 79-86 Juvenile bull sharks Carcharhinus leucas (Valenciennes, 1839) in northern Australian rivers DEAN C. THORBURN AND ANDREW J. ROWLAND Centre for Fish and Fisheries Research, Murdoch University, South St, Murdoch, Western Australia 6150, AUSTRALIA Corresponding author: dthorhurn@aapt.net.au ABSTRACT Juvenile bull sharks, Carcharhinus leucas , ranging in total length from 687 mm to 1365 mm, were captured in riverine waters throughout the Northern Territory and Kimberley region. Western Australia. Carcharhinus leucas was captured in highest abundance in low salinity riverine waters (<5 ppt), however, none were captured in offshore marine waters. The rivers of northern Australia act as nurseries for juvenile C. leucas within which individuals remain for approximately four years. Assuming annuli present on vertebrae were laid down annually and a birth size of approximately 700 mm TL, growth estimates suggest that C. leucas attains 863 mm TL in its first year, 993 mm TL in its second year, 1118 mm TL in its third year and 1239 mm TL in its fourth year. Stomach content analysis indicated a broad diet with the consumption of large amounts of teleost fishes within those systems, as well as numerous other prey types including terrestrial mammals, aquatic insects and other elasmobranchs. Keywords: bull shark, nursery area, Carcharhinus , river, Kimberley, Northern Territory. INTRODUCTION The occurrence of the bull shark Carcharhinus leucas in Australian rivers is well documented as it is the only carcharhinid known to penetrate well into freshwater systems for extended periods of time (Whitley 1943; Chubb et al. 1979; Merrick and Schmida 1984; Last and Stevens 1994; Allen et al. 2002). However, records of the occurrence of C. leucas in offshore waters of Australia are rare, a disparity that may be attributed to the confusion of this species with other whalers and in particular with the very similar pigeye shark ( Carcharhinus amboinensis) (Cliff and Dudley 1991; Last and Stevens 1994). Although biological data for Australian populations of C. leucas are scarce, various aspects of the life history have been studied elsewhere including Lake Nicaragua (Thorson et al. 1966; Thorson and Lacy 1982), the Florida Lagoons (Snclson et at. 1984), Gulf of Mexico (Branstetter and Stiles 1987) and South Africa (Cliff and Dudley 1991; Wintner et al. 2002). These studies indicate that C. leucas can remain in freshwater systems for extended periods and often spends a significant part of its juvenile life in upper estuarine and freshwater reaches (Thorson 1972; Snelson et al. 1984; Cliff and Dudley 1991). Breeding is thought to occur in offshore marine waters, with females entering estuarine and inshore waters to give birth (Montoya and Thorson 1982; Bishop et al. 2001). Length at birth estimates range from 55 to 90 cm total length (Thorson and Lacy 1982; Snelson 1984; Branstetter and Stiles 1987; Cliff and Dudley 1991; Wintner et al. 2002). Despite the marine affinities of C. leucas, this species' utilisation of upper riverine reaches has often led to it being included in studies of freshwater elasmobranchs. This was the case during two previous ichthyological surveys that aimed to capture elasmobranchs in fresh waters of northern Australia (Taniuchi et al. 1991; Ishihara et al. 1991). As a result of these studies and other surveys of riverine waters C. leucas has been shown to occupy numerous systems in tropical and warm temperate regions of Australia extending south to Sydney (New South Wales) and Perth (Western Australia) (Chubb et al. 1979; Last and Stevens 1994; Bishop et al. 2001; Allen et al. 2002; Thorbum et al. 2003). In light of the paucity of knowledge on elasmobranchs occurring in estuarine and freshwaters of northern Australia, ichthyological surveys were conducted in the Northern Territory' and Kimberley region of Western Australia in 2002 (Thorbum et al. 2003) and additionally in the Fitzroy River (Western Australia) throughout 2003 and 2004 (Thorbum et al. 2004). During these surveys, C. leucas was captured in higher abundance than any other elasmobranch species, and provided an opportunity to: determine the distribution and any broad-scale habitat associations of C. leucas in the rivers of the Northern Territory and Kimberley region, Western Australia; and describe the biology, age composition and diet of C. leucas occurring within these rivers. Furthermore, this study aimed to test the hypothesis that juveniles of the species utilise rivers of northern Australia as nursery grounds, by satisfying the criteria defined by Heupel et al. (2007) that, “(1) sharks are more commonly encountered D. C. Thorburn and A. J. Rowland in the area than other areas; (2) sharks have a tendency to remain or return for extended periods; and (3) the area or habitat is repeatedly used across years”. MATERIALS AND METHODS Study sites, environmental variables and habitat. One hundred and eighty-six marine, estuarine and freshwater sites were sampled between the Robinson River (NT) and the Fitzroy River (WA) from June 2002 to July 2004 (Fig. 1). The salinity (ppt), temperature (°C), and an estimate of water clarity using a secchi disc (cm) were recorded at each sample site. The depth (m) and tidal movement were also recorded. Notes on the immediate habitat, including predominant sediment type, density of aquatic vegetation types and detritus, riparian vegetation and snag density, were also made. Sample collection, measurements and field dissection. Sampling was primarily conducted with sinking monofilament gill nets (including 20 m panels of 5, 7.5, 10, 15 and 20 cm stretched mesh) that were set perpendicular to the river bank. All gill net mesh sizes were utilised during daylight hours, however, only the larger meshes were used at night to minimise by-catch. Baited long-lines and handlines were also used. Long-lines were a maximum of 40 m long and contained up to 20 5/0 tuna circle hooks. Each hook was connected to the 500 lb nylon mainline via 0.5 m of 200 lb wire trace. Locally available fish was used as bait. Long-lines were set during the day and night. Fig. 1. The sites at which Carcharhinus leucas was captured in the Northern Territory and in the Kimberley region of Western Australia. Included are records from the Museum and Art Gallery of the Northern Territory (NTM) and the Western Australian Museum (WAM). Dissections were conducted in the field as soon as practicable after capture to minimise tissue breakdown and further digestion of the gut contents. The total length (TL, mm), sex and weight (W, g) were recorded for each specimen captured. For comparison with other studies which reported in precaudal length (PCL) only, the conversion TL (cm) = (PCL + 9.16) / 0.81 (Branstetter and Stiles 1987) was used. The stomach and a minimum of six vertebrae from below the first dorsal fin were removed from 51 female and 49 male C. leucas. Stomach fullness and contributions were either determined soon after capture or the stomachs were preserved whole in 100% ethanol for later examination. Vertebral samples were stored in 100% ethanol or kept on ice until they could be frozen. Length-weight, length-frequency and age. A likelihood ratio test (Cerrato 1990) was used to determine if the length-weight relationship was significantly different between the sexes. As no difference appeared to exist the sexes were pooled and the SPSS statistical package used to determine the line of best fit for the relationship. A length-frequency histogram was generated to aid in the identification of size specific cohorts. Samples of vertebrae were defrosted and placed in 5% sodium hypochlorite solution until free of tissue. The centra were then rinsed thoroughly in water and allowed to dry for several hours. A minimum of two centra from each sample was embedded in resin and a 0.3 mm longitudinal section cut with an Isomet low speed rotary saw. Sections were then mounted on a slide with DePex mounting medium and observed under a dissection microscope with reflected 80 Juvenile bull sharks ( Carcharhinus leucas ) in northern Australian rivers light. Counts of the number of growth rings or annuli commencing after the birth mark (identified by a change of angle on the outer edge of the corpus calcerium) on the facia of whole and sectioned vertebrae were then made for each individual. The number of annuli was compared to TL and the mean size of individuals with respect to the number of annuli calculated. The lack of successive annual samples of vertebrae precluded validation that annuli were laid down annually during the current study. Marginal increment analysis of vertebrae from specimens from the Gulf of Mexico (Branstetter and Stiles 1987) and ‘mark- recapture’ analysis of captive South African specimens (Wintner et at. 2002) confirmed the annual formation of annuli. These findings, in conjunction with fact that length at age and growth rate data collected during the current study (see Results) were comparable to those collected by Branstetter and Stiles (1987) and Wintner et at. (2002), suggest that annuli may similarly be laid down annually in Australian specimens. A von BertalanfTy growth curve was fitted to the length of C. leucas at their estimated age of capture, however, this method required the provision of the individual’s birth dates. This estimation was subsequently based on the stage of healing of the umbilical scars present on a number of the individuals captured. Healing rates were assumed to be comparable to that of Carcharhinus cautus where umbilical scars were observed to be completely healed within three months of birth (Dr William White, CSIRO, Hobart, pers. comm). The birth date of an individual possessing an open umbilical scar was estimated to have been at the beginning of the month of capture. An individual possessing a closed umbilical scar was assumed to have been born at the beginning of the month prior to the date of capture, and an individual possessing a near faded umbilical scar was assumed to have been born two months prior to capture. Other individuals collected from the same river as those possessing umbilical scars, or from other rivers in close proximity, were subsequently assigned the same birth months. Based on the individuals and times captured, those from (and near to) the Roper River (NT) were assumed to be born I June, those from the Daly River (NT) on 1 August, those from the Ord River (WA) on 1 November, and those from the Fitzroy River (WA) on 1 February. A line of regression was fitted with the use of SigmaPlot 8.0 and the von Bertalanffy growth parameters generated by the equation L = L r (1 - e' A ’ ( ''" l) ), where L t is the length at age t, L : is the mean asymptotic length, K is the growth coefficient and t (j is the hypothetical age at which the estimated length is zero. Maturation. Maturation in male C. leucas was determined on the basis of clasper calcification. Individuals were considered immature when claspers were small and uncalcified, maturing if claspers were extending and becoming semi-calcified and mature when claspers were fully calcified and sperm was present. Maturity in females was assessed on observations of the ovaries and uteri. Individuals possessing undeveloped ovaries and thin, flaccid uteri were considered immature, maturing when the uterus began to enlarge and ovaries contained differentiated ova, and mature when the ovaries contain yolked ova and an enlarged uterus (Conrath 2004). Diet. Upon removal of the stomach, an estimate of fullness on a scale of 0 to 10 (zero representing an empty gut and 10 being fully distended) was made. Some smaller prey items required identification under a dissection microscope. An estimation of the percentage contribution of each food item was made for each individual, and the frequency of occurrence (%F) and mean percentage volumetric contribution (%V) calculated (Hynes 1950; Hyslop 1980). The graphical method of Costello (1990) was also used to display the broad functional prey categories found to constitute the diet of C. leucas ; %F is plotted against relative quantity (in this case %V). This method takes into account the percentage occurrence, and subsequently provides insight into population wide food habits (Cortes 1997). Unidentified items were excluded from analysis and the values of all other dietary 1 categories of that individual adjusted upward to sum to 100% (Pusey et al. 2000). Similar prey categories were combined to reflect a functional prey group. RESULTS Capture locations, sex ratio, length ranges and habitat. A total of 111 C. leucas were captured from 41 of the 186 sites sampled (Fig. 1). Fifty seven were female and 54 were male (sex ratio of 1.06 female : 1 male) which ranged in length from 691 to 1382 mmTLand 687 to 1365 mm TL, respectively (Fig. 2). Carcharhinus leucas was only captured within rivers (i.e. upstream of the mouth). The sites of capture ranged in salinity from 0 to 41.1 ppt (Fig. 3), however, 86% of individuals (all of which were Fig. 2. Length-frequency histogram for female and male Carcharhinus leucas collected during the study. 81 D. C. Thorbum and A. J. Rowland immature) were captured in waters of less than 5 ppt (32% were captured in 0 ppt). Carcharhinus leucas was caught in waters ranging in temperature from 20.8 to 32.5 °C that were between 0.7 and 21 nr deep and which varied greatly in clarity (5-400 cm secchi depth). The species was generally captured in open waters over sandy substrates that supported low macrophyte and algal growth. Length-weight relationship, age and growth, maturity and diet. Weights of C. leucas ranged from 2360 to 20 500 g in females and from 1740 to 18 000 g in males (Fig. 4). The relationship between TL and W can be described by the formula W = 6165.14 - (21.741TL) + (0.0228TL 2 ). Observations of the annuli present on the vertebrae of 100 individuals suggested the presence of five year classes (0+ to 4+) (Fig. 5). The average TL of individuals possessing zero annuli on vertebrae was 787 mm, individuals possessing one annuli on vertebrae averaged 859 mm, individuals possessing two annulus on vertebrae o — Salinity (ppt) Fig. 3. The number of Carcharhinus leucas captured versus sample site salinity (the number of sites sampled are given in parenthesis). W = 6165.14 - (21.741TL) + (0.0228TL 2 ) Total length (mm) Fig. 4. The relationship between total length (TL) (mm) and weight (W) (g) of Carcharhinus leucas captured in northern Australia. 1600 1400 | 1200 JZ Ul g S 1000 £ 800 600 0(31) 1(30) 2(13) 3 (22) 4(4) Number of annuli Fig. 5. Number of annuli observed on vertebrae versus the total length (TL) of Carcharhinus leucas. The number of individuals possessing 0 to 4 annuli are given in parentheses. averaged 1010 mm, individuals possessing three annuli on vertebrae averaged 1 1 68 mm and individuals possessing four annuli on vertebrae averaged 1339 mm. The von Bcrtalanffy growth equation generated parameter estimates of L a = 4662 mm, K = 0.0347 and t„ = -4.88 years (Fig. 6). While the L r is above the known maximum size of-3400 mm TL (Last and Stevens 1994), both the / 0 and L m are comparable lo those published for this species elsewhere. From this equation it is estimated that C. leucas attains 863 mm TL in its first year, 993 mm TL in its second year, 1118 mm TL in its third year and 1239 mm TL in its fourth year. All 49 male C. leucas captured possessed small non- calcified claspers indicating immaturity. Similarly all 51 females dissected were found to be immature by the possession of undeveloped ovaries and thin, flaccid uteri. Seventy-six of the 100 stomachs inspected contained prey. Teleost fishes were the major constituents of the diet of C. leucas (Table 1, Fig. 7). Where identification of teleost prey to species was possible, the lesser salmon catfish Ariopsisgraeffei, bony bream Nematalosa erebi and scaly croaker Nibea squamosa were the major components Age (years) Fig. 6. von Bcrtalanffy growth curve for Carcharhinus leucas collected during the study. 82 Juvenile bull sharks ( Carcharhinus leucas) in northern Australian rivers 2. Crocodylus johnstoni Ariid catfish# 3. Terrestrial mammal 4. Insect 5. Macrobrachium rosenbergii - Other teleost • ■ • Nematalosa erehi 2 3 1* • #Plqnt matter i 0 10 20 30 40 50 60 Occurrence (%F) Fig. 7. The important broad functional prey categories of Carcharhinus leucas. Tabic 1. Percentage volumetric contribution (%V) and percentage occurrence (%F) of the different food items found in the stomachs of Carcharhinus leucas. Prey Item % V %F Vegetation Terrestrial vegetation 1.04 13.16 Aquatic macrophyte 0.22 3.95 Macrocrustacea Macrobrachium rosenbergii 1.40 7.89 Insect Orthoptera 0.03 1.32 Coleoptera 0.63 5.26 Teleost Ariopsis graeffei 26.71 32.89 Ariopsis midgleyi 2.41 1.32 Hemiarius diodes 3.56 2.63 Plicofollis argypleuron 1.27 1.32 Other ariid sp. 10.11 17.11 Harpadon translucens 1.22 1.32 Hemiramphidae (garfish) 0.20 1.32 Lates calcarifer 3.21 5.26 Liza tade 2.08 2.63 Nematalosa erebi 15.14 19.74 Nibea squamosa 9.44 7.89 Polydactylus macrochir 2.86 1.32 Rhinomugil nasutus 0.95 2.63 Sclerophages jardinii 1.26 1.32 Unidentified teleost parts 6.77 21.05 Elasmobranchii Pristis microdon 2.54 1.32 Mammal Pig 1.59 2.63 Other 0.98 2.63 Reptile Chelonidae 0.54 1.32 Crocodylus johnstoni 2.96 2.63 Unidentified 0.89 3.95 of the diet. The removal of the ‘unidentified’ prey category and consolidation into broad (functional) prey categories further reflected the dominance of teleost species in the diet of C. leucas. DISCUSSION Distribution and habitat. Carcharhinus leucas was frequently captured in fresh waters several hundred kilometres from the coast. The notion that northern Australian rivers act as nurseries for juveniles of the species is strongly supported, and satisfies the definition criteria defined by Heupel et al. (2007) in that: high abundances were encountered upstream of the rivers mouths while none were captured from marine waters; all specimens were small and immature and appeared to remain in the river until approximately three or four years of age (assuming annuli were laid down annually); and juveniles of less than one year were captured from the Fitzroy River in each of the three consecutive years of sampling. Saline estuarine waters have generally been considered to provide important nursery habitats of C. leucas more so than freshwaters (Simpfendorfer 2005). During the current study however, the highest abundances of individuals were captured in low salinity waters near the tidal limit and in non-tidal freshwaters, suggesting that in Australia low salinity waters are more important as nursery habitats than higher salinity waters associated with the lower estuary. The use of upstream waters by juvenile C. leucas is likely attributed to the high abundance of food resources available and lack of predators including large sharks and estuarine crocodiles Crocodylus porosus (Simpfendorfer and Milward 1993; Morgan et al. 2004). Upstream migration of the species was demonstrated on several occasions where the highest abundances of individuals were encountered immediately below a barrier such as an exposed rock bar, road crossing or barrage. In a number of cases, these barriers also represent the upper limit of tidal influence and may partially explain the high abundance of this species in those reaches. Size and age structure of C. leucas in northern Australian rivers. The structure of riverine populations in northern Australia appears similar to those observed elsewhere including North America where C. leucas resides in inland waters of the Florida Lagoons for a similar period of time (Snelson et al. 1984). Studies by Wintncr et al. (2002) and Cliff and Dudley (1991) suggested that C. leucas in South Africa are born slightly larger than elsewhere, including the coastal lagoons of Florida, Nicaragua and the Gulf of Mexico, i.e. 800 to 900 mm TL cf. 600 and 750 mm TL, respectively (Snelson et al. 1984; Thorson and Lacy 1966; Branstettcr and Stiles 1987). During the current study, open umbilical scars were observed on individuals between 687 and 749 mm TL. The size at 83 D. C. Thorbum and A. J. Rowland birth of Australian C. leucas was therefore found to be comparable to individuals of the latter locations. Umbilical scars were observed in individuals captured from the Roper River (NT) in June, Daly River (NT) in August, Ord River in November (WA) and from the Fitzroy River (WA) in February. Bishop el al. (2001) suggests that C. leucas in Australia are born during summer which coincides with the early portion of the tropical wet season. This apparent westward progression of birthing across northern Australia may therefore be a response to the westward onset of the wet season (considered to be between October and April in the NT and between January and March in WA) (Bureau of Meteorology, Commonwealth of Australia 2004). High river discharges associated with the wet season subsequently provide the opportunity for juvenile C. leucas to penetrate far upstream before the contraction of water levels during the tropical the dry season. Various studies have indicated that the annual growth rate of C. leucas is greatest in the first few years of life (Caillouet 1969; Thorson and Lacy 1982; Branstctter and Stiles 1987; Wintner et al. 2002). Assuming that annuli were laid down annually, length at age estimates generated during this study of863 mm TL at one year of age and 1239 mm TL at four years of age are consistent with those observed from the Gulf of Mexico (Branstetterand Stiles 1987). However, length at age estimates of northern Australian specimens appear lower than those from South Africa where a one year old C. leucas was estimated to be 1040 mm TL and a four year old estimated to be 1406 mm TL (Wintner et al. 2002). The current study also estimated that annual growth in the first four years may be approximately 1 35 mm yr 1 . This annual growth estimate is marginally greater than that observed in South Africa (an average of 110 mm yr 1 in the first three years) (Wintner et al. 2002), however, it is less than the growth rate observed in Nicaragua (an average of 160-180 mm yr' in the first two years) (Thorson and Lacy 1982) and the Gulf of Mexico (an average of 150 - 200 mm yr 1 for the first 5 years) (Branstetter and Stiles 1987). Occurrences of sharks greater than two metres in length were reported from near Fitzroy Crossing on the Fitzroy River (Western Australia) over 300 km upstream of the mouth. Furthermore, the senior author observed a C. leucas of approximately two metres TL swimming immediately below Camballin Barrage approximately 150 km from the mouth of the Fitzroy River. Based on the von Bertalanffy growth equation and length weight relationship lor younger C. leucas generated during this study, these specimens would be 11 years old and weigh 54 kg. Size of maturity data collected from elsewhere has indicated that C. leucas attains sexual maturity at between 1600 and 2200 mm TL in males and between 1800 and 2250 mm TL in females (Thorson et al. 1966; Bass 1977; Branstetter and Stiles rat ' -ml m*L Fig. 8. Stomach contents and potential prey of C. leucas, including (a) freshwater sawfish Pristis microdon, (b) other C. leucas, (c) freshwater crocodile Crocodylus johnstoni and (d) feral pig. 84 Juvenile bull sharks ( Carcharhinus leucas) in northern Australian rivers 1987; Cliff and Dudley 1991). Based on the aforementioned observations and the length at maturity estimates, the presence of mature C. leucas in northern Australian rivers can not be dismissed. Feeding behaviour and predation. Inspection of the stomach contents of C. leucas indicated the consumption of a wide variety of prey types and an opportunistic often indiscriminate hunting pattern. While a vast majority of the diet consisted of bony fishes (in particular ariid catfishes), unusual prey including mammals (pig), other elasmobranchs (.Pristis microdon) and portions of freshwater crocodiles (Crocodylus johnstoni) were observed (Fig. 8). Vegetation (aquatic and terrestrial) was also encountered in a high proportion of the stomachs inspected, further reflecting the random nature of prey selection. The authors were also witness to cannibalism by the species whereby C. leucas captured in gill nets were attacked by others. Several C. leucas were also observed leaping approximately 2.5 m clear of the water to take birds perched on overhanging branches in the Fitzroy River, Wester Australia. Anecdotal accounts of aggressive feeding described by traditional owners of the Fitzroy River may also suggest the presence of cooperative hunting behaviours. Of particular interest was an account of an attack on a feral pig that had entered the waters edge to drink. As explained, several small C. leucas attacked the legs of the pig and proceeded to pull it into deeper water. The pig eventually drowned and was consumed. An additional report of an attack by C. leucas on a freshwater crocodile was also collected. A single C. leucas was observed to rapidly disable the crocodile by biting off one of its front legs. Numerous other C. leucas subsequently appeared and proceeded to bite the remaining limbs before attacking the body. ACKNOWLEDGMENTS Sincere thanks must go to Dr David Morgan, Dr Howard Gill, Dr Peter Last, Dr John Stevens, Dr Helen Larson, Dr Barry Russell, Steven Gregg and family, Stirling Peverell, Dr William White, Michael Taylor and Matt Pember, the Natural Heritage Trust, Murdoch University, Department of Fisheries Government of Western Australia, Northern Territory Fisheries, Museum and Art Gallery of the Northern Territory, Kimberley Land Council and Department of Conservation and Land Management. REFERENCES Allen, G. R., Midgley, S. H. and Allen, M. 2002. Field Guide to the freshwater fishes of Australia. CSIRO/Western Australian Museum: Perth. Bass, A. J. 1977. Long-term recoveries of tagged sharks. Copeia 3: 574-575. Bishop, K. A., Allen, S. A., Pollard, D. A. and Cook, M. G. 2001. Ecological Studies on the Freshwater Fishes of the Alligator Rivers Region, Northern Territory: Autecology. Supervising Scientist Report 145, Supervising Scientist: Darwin. Branstctter, S. and Stiles, R. 1987. Age and growth of the bull shark, Carcharhinus leucas, from the northern Gulf of Mexico. Environmental Biology of Fishes 20: 169-181. Caillouet, C. W. 1969. Weight, length and sex ratio of immature bull sharks, Carcharhinus leucas , from Vermillion Bay, Louisiana. Copeia 1: 196-197. Cerrato, R. M. 1990. Interpretable statistical tests for growth comparisons using parameters in the von Bertalanffy equation. Canadian Journal of Fisheries and Aquatic Sciences 47: 1416-1426. Chubb, C. F., Hutchins, J. B., Lcnanton, R. C. J. and Potter I. C. 1979. An annotated checklist of the fishes of the Swan-Avon river system. Western Australia. Records of the Western Australian Museum 8: 1-55. Cliff, G. and Dudley, S. F. J. 1991. Sharks caught in the protective gill nets off Natal, South Africa. 4. The bull shark Carcharhinus leucas Valenciennes. South African Journal of Marine Science 10:253-270. Conrath, C.L. 2004. Reproductive biology. In: Musick, J. A. and Benlll, R. (Eds) Elasmobranch Fisheries Management Techniques, APEC Fisheries Working Group. Pp. 133-164. Cortes, E. 1997. A critical review of methods of studying fish feeding based on stomach contents: application to elasmobranch fishes. Canadian Journal of Fisheries and Aquatic Science 54: 726-738. Costello, M. J. 1990. Predator feeding strategy and prey importance: a new graphical analysis. Journal of Fish Biology 36: 261-263. Ebert. D. A. 2003. California Natural History Guides. Sharks. Rays and Chimaeras of California. University of California Press, Berkeley. Heupel, M. R., Carlson, J. K. and Simpfcndorfer, C. A. 2007. Shark nursery areas: concepts, definition, characterization and assumptions. Marine Ecology Progress Series 337: 287-297. Hynes, 11. B. N. 1950. The food of sticklebacks with a review of the methods used in studies of food in fishes. Journal of Animal Ecology? 19: 36-58. Hyslop, E. J. 1980. Stomach content analysis - A review of methods and their application. Journal of Fish Biology 17: 411-429. Ishihara, H., Taniuchi, T„ Sano, M. and Last, P. R. 1991. Record of Pristis elavata Gannan from the Pentecost River, Western Australia, with brief notes on its osmoregulation, and comments on the syslematics of the Pristidae. University Museum, University of Tokyo, Nature and Culture 3: 43-53. Last, P. R and Stevens, J. D. 1994. Sharks and Rays of Australia. CSIRO Division of Fisheries, CSIRO, Australia. Merrick, J. R. and Schmida, G. E. 1984. Australian Freshwater Fishes. Biology and Management. Griffin Press: Nctley, South Australia. Montoya, R. V. and Thorson, T. B. 1982. The bull shark (Carcharhinus leucas) and largetooth sawfish (Pristis perotteti) in Lake Bayano, a tropical man-made impoundment in Panama. Environmental Biology of Fishes 7: 341-347. Morgan, D. L., Allen, M. G., Bedford, P. and Horstman, M. 2004. Fish fauna of the Fitzroy River in the Kimberley region of Western Australia - including the Bunuba, Gooniyandi, Ngarinyin, Nyikina and Walmajarri Aboriginal names. Records of the Western Australian Museum 22: 147-161. Pusey, B. J., Read, M. G. and Arthington, A. H. 2000. The dry-season diet of freshwater fishes in monsoonal tropical rivers of Cape York Peninsula, Australia. Ecology of Freshwater Fishes 9: 177-190. 85 D. C. Thorbum and A. J. Rowland Sintpfendorfer, C. A., Freitas, G. G„ Wiley, T. R. and Heupel, M. R. 2005. Distribution and habitat partitioning of immature bull sharks (Carcharhinus leucas) in a southwest Florida estuary. Estuaries 28: 78-85. Simpfendorfcr, C. A. and Mil ward. N. E. 1993. Utilisation of a tropical nursery area by sharks of the families Carcharhinidac and Sphymidae. Environmental Biology of Fishes 37: 337-345. Snelson, F. F., Mulligan, T. J. and Williams, S. E. 1984. Food habitats, occurrence, and population structure of the bull shark, Carcharhinus lcucas, in Florida coastal lagoons. Bulletin of Marine Science 34: 71-80. Taniuchi, T., Shimizu, M., Sano, M., Baba, O. and Last, R R. 1991. Description of freshwater elasmobranchs collected from three rivers in northern Australia. University Museum, University of Tokyo, Nature and Culture 3: 11-26. Thorbum, D. C., Peverell, S., Stevens, J. D., Last, R R. and Rowland, A. J. 2003. Status of Freshwater and Estuarine Elasmobranchs in Northern Australia. Report to the Natural Heritage Trust. Thorbum, D. C„ Morgan, D. L., Rowland, A. J. and Gill, H. S. 2004. Elasmobranchs in the Fitzroy River, Western Australia. Report to the Natural Heritage Trust. Thorson, T, B. 1972. The status of the bull shark, Carcharhinus leucas, in the Amazon River. Copeia 3: 601-605. Thorson, T. B. and Lacy, E. J. 1982. Age, growth and longevity of Carcharhinus leucas estimated from tagging and vertebral rings. Copeia 1: 110-116. Thorson. T. B., Watson, D. E. and Cowan, C. M. 1966. The status of the freshwater shark of Lake Nicaragua. Copeia 3: 385-402. Whitley, G. R 1943. Ichthyological descriptions and notes. The Proceedings of the Linnean Societ V of New South Wales 68: 114-144. Wintner, S. P.. Dudley, S. F. .1, Kistnasamy, N. and Everett, B. 2002. Age and growth estimates for the Zambezi shark, Carcharhinus leucas, from the east coast of South Africa. Marine and Freshwater Research 53: 557-566. Accepted 15 April 2008 86 The Beagle, Records of the Museums and Art Galleries of the Northern Territory, 2008 24: 87-146 Review of the Dinematichthyini (Teleostei: Bythitidae) of the Indo-west Pacific. Part IV. Dinematichthys and two new genera with descriptions of nine new species PETER RASK M0LLER 1 AND WERNER SCHWARZHANS 2 1 Natural History Museum of Denmark, Zoological Museum, University of Copenhagen, Universitetsparken 15, DK-2100 Copenhagen 0, DENMARK pclrmoller@snm.ku.dk 2 Ahrensburger Weg 103 D, 22359 Hamburg, GERMANY wwsch warz@aol. com ABSTRACT A revision of the dinematichthyine fishes (Ophidiiformes: Bythitidae: Brosmophycinae) of the Indo-West Pacific based on more than 6500 specimens is published in several parts. Part IV is the last part and includes 4719 identified specimens in the genera Alionematichthys (new genus with three described and eight new species), Dinematichthys Bleeker, 1855 (with one described and one new species) and Porocephalichthys (new genus with one described species). A neotype of Dinematichthys iluocoeteoides Bleeker, 1855 is here designated. The genera reviewed here have in common a high anterior nostril and are considered to be related to each other. When previously described, they were all included in the genus Dinematichthys. The separating characters of the species are the pseudoclasper morphology, morphometric characters, vertebrae and fin ray counts, otolith morphology, head squamation, presence or absence of the upper preopercular pore and development of cirri on the snout. Keywords: Viviparous brotulas, Indo-west Pacific, Dinematichthys, new species, new genera. CONTENTS INTRODUCTION.88 MATERIAL AND METHODS.88 COMPARATIVE MATERIAL.89 SYSTEM ATICS.89 Tribe Dinematichthyini Cohen and Nielsen, 1978 .90 Key to the genera and the species reviewed herein.89 Alionematichthys gen. nov.90 Alionematichthys ceylonensis sp. nov.92 Alionematichthys crassiceps sp. nov.95 Alionematichthys minyomma (Sedor and Cohen, 1987).99 Alionematichthys phuketensis sp. nov.100 Alionematichthys piger (Alcock, 1890).102 Alionematichthysplicatosurculus sp. nov.109 Alionematichthys riukiuensis (Aoyagi, 1954).113 Alionematichthys samoaensis sp. nov. 117 Alionematichthys shinoharai sp. nov.120 Alionematichthys suluensis sp. nov.122 Alionematichthys winterbottomi sp. nov.125 Alionematichthys sp. 1 .126 Alionematichthys sp. 2.127 Dinematichthys Bleeker, 1855.128 Dinematichthys iluocoeteoides Bleeker, 1855.129 Dinematichthys trilobatus sp. nov.136 Porocephalichthys gen. nov.138 Porocephalichthys dasyrhynchus (Cohen and Hutchins, 1982).138 DISCUSSION.141 REGIONAL CHECK LIST.142 KEY TO THE GENERA OF THE DINEMATICHTHYINI .144 ACKNOWLEDGMENTS.145 REFERENCES.145 W. Schwarzhans and P. R. Moller INTRODUCTION The global review of the dinematichthyine fishes, a tribe within the subfamily Brosmophycinae of the viviparous family Bythitidae, is concluded with this sixth part. The first two parts dealt with the American Dinematichthyini (Moller et al. 2004a, 2005) and were followed by four publications revising the Dinematichthyini from the Indo-west Pacific (Schwarzhans et al. 2005; Moller and Schwarzhans 2006; Schwarzhans and Moller 2007; this paper). The latest comprehensive account of ophidiiform fishes (Nielsen et al. 1999) contained 12 genera now regarded as valid in the Dinematichthyini, and 25 valid species. With the completion of the world-wide review of the Dinematichthyini, the current status of recognised taxa within the tribe has risen to 26 genera with 110 species. The fishes described in this paper were previously referred to the genus Dinematichthys. Following Cohen and Nielsen (1978) and redefined by Cohen and Hutchins (1982) and Sedor and Cohen (1987), the position of the anterior nostril high above the upper lip, and about midway between the lip and the posterior nostril, was regarded as the main diagnostic character for the genus, distinguishing it from all other than known dinematichthyine fishes. In this review we conclude that the formal genus Dinematichthys “sensu Iato” in fact contains three well defined groups for which generic rank applies: two more closely related ones - Dinematichthys with two species and A lionematichthys gen. nov. with 11 species - and one more distant monospecific genus containing Pomcephalichthys gen. nov. dasyrhynchus (Cohen and Hutchins). The combining character of the high anterior nostril, reformulated to “positioned 1/3 or less of the distance from upper lip to anterior margin of eye” in Schwarzhans and Moller (2007), however, is not unique to these three genera and this distinguishing character has therefore become weakened. Similarly, high positioned anterior nostrils between 1/3 and 1/3.5 the distance from upper lip to anterior margin of eye have been observed in a few species of the genus Diancistnis Ogilby (for instance D. alleni, D. atollonm and D. katrineae), in Eusurculus pistillum Schwarzhans and Moller and in Lapitaiclithys frickei Schwarzhans and Moller. Fishes of the three genera Dinematichthys, Alionematichthys gen. nov. and Pomcephalichthys gen. nov. can be easily distinguished from Diancistnis by the otolith sulcus showing distinctly separated colliculi (vs fused, undivided), from Lapitaiclithys in the presence of two pairs of pseudoclaspers (vs single pair of outer pseudoclaspers), and from Eusurculus pistillum in the flap-like inner pseudoclasper (vs stalked with sucker-disk tip) (see also key below). The diagnostic, distinguishing characters in the species of the three genera covered here are similar to those of the previous parts of our review. Again, pseudoclasper morphology represents the most useful character. However, pseudoclaspers in the following three genera are smaller than in most other Dinematichthyini, with few exceptions, and arc also less diverse in most instances. Other useful characters for species differentiation are found to be head squamation, vertebrae and fin ray counts, morphometric measurements, otolith morphology, presence or absence of the upper preopercular pore and, new in this group, development of cirri on the snout. Of these, meristic counts show a lower degree of variation than in most other Dinematichthyini, and although individual species show statistically significant differences (average values), they are rarely separated in absolute values to be of use in diagnostic keys. Eight species have been described previously, as placed in the genus Dinematichthys, of which five are here regarded as valid, and represent the three genera dealt with in this part of the review. Three previously described species are placed in synonymy. Dinematichthys Bleeker is the earliest named genus of its tribe and hence, its definition is essential to achieving nomenclatural stability. The genus was established by monotypy (D. iluocoeteoides) based on a single type specimen, which according to extensive research by Cohen and Nielsen (1978) must be regarded as no longer extant. For the purpose of redefinition of the genus and species, another original Bleeker specimen mentioned by Gunther (1862) (BMNH 1868.2.28.65), probably the oldest extant, is here selected as neotype. Although shrivelled, it represents a male specimen with all the important diagnostic characters, and is in good accordance with Blocker’s original description except for a slight variation in eye size, which is within the range of variability observed in the species. MATERIAL AND METHODS The examination of ca. 6500 specimens of Indo-west Pacific Dinematichthyini yielded 4719 specimens which were identified in the genera treated herein. Also included are additional specimens identified in the collections of AMS and USNM but not borrowed for detailed investigations. These are listed as additional specimens and are not referred to as type specimens for any of the new species. The material described herein belongs to the following institutions: AMS (Australian Museum, Sydney), ANSP (Academy of Natural Sciences, Philadelphia), BMNH (Natural History Museum, London), BPBM (Bishop Museum, Honolulu), CAS (California Academy of Sciences, San Francisco), KAUM (The Kagoshima University Museum), KSHS (Kochi Prefectural Kochi Nishi Senior High School, Kamobe, Kochi, Japan), MNHN (Museum National d’Histoire Naturellc, Paris), NSMT (National Science Museum, Tokyo), NTM (Museum and Art Gallery of the Northern Territory (formerly Northern Territory Museum), Darwin), ROM (Royal Ontario Museum, Toronto), SAIAB (South African Institute for Aquatic Biodiversity, formerly RUSI (JLB Smith Institute of Ichthyology), Grahamstown), SMF (Senckenberg Forschungsinstutut und Museum, Frankfurt/Main), SMNS Dinematichthyine fishes of the Indo-west Pacific, Part IV (Staatliches Museum fur Naturkunde, Stuttgart), TAU (Tel Aviv University, Tel Aviv, Israel); USNM (National Museum of Natural History, Smithsonian Institution, Washington D.C.), WAM (Western Australian Museum, Perth), YCM (Yokosuka City Museum), and ZMUC (Zoological Museum, University of Copenhagen). For methodology used in analysing dinematichthyine fishes, reference is made to Moller et al. (2004a) and Schwarzhans et al. (2005). Abbreviations used in meristic counts are: D/V = anterior dorsal fin ray above vertebra number; D/A = anterior anal fin ray below dorsal fin ray number; V/A = anterior anal fin ray below vertebrae number; D-A = number of dorsal fin rays minus number of anal fin rays; V in D = number of dorsal fin rays per ray¬ bearing vertebra. The abbreviations ‘i.p.’ and ‘o.p.’ used in the figures relating to male copulatory organ structures are explained in Schwarzhans et al. (2005). The ecology of most of the species is poorly known. From available station data we have gathered some information about habitat and depth range, but we have very little data about behaviour, live colouration and feeding. A number of females were examined for reproductive data, e.g. number and size of embryos. The distribution maps were created using Microsoft Encarta 2001 digital world atlas. COMPARATIVE MATERIAL Indo-west Pacific Dincmatichthyini: see Schwarzhans et al. (2005), Moller and Schwarzhans (2006) and Schwarzhans and Moller (2007). American Dinematichthyini: sec Moller et al. (2004a) and Moller et al. (2005). Brosmophycinae and Bythitinae: see Moller et al. (2004b). SYSTEMATICS Family Bythitidac Gill, 1861 Subfamily Brosmophycinae Gill, 1862 Tribe Dinematichthyini Cohen and Nielsen, 1978 Diagnosis. Male copulatory organ with penis and 1-2 (rarely 3) pairs of pseudoclaspers in cavity of ventral body wall covered by fleshy hood. First anal fin pterygiophore slightly to strongly elongate. Head pore system generally not reduced, 6 mandibular, 2-4 preopercular, 5-7 infraorbital and 3-4 supraorbital pores, including supraorbital pore above opercular spine. Posteriormost supraorbital head- pore tubular. Table 1 summarises meristic characters used in distinguishing the species of the genera described here. Key to the genera and the species reviewed herein The following key to the genera Dinematichthys, Alionematichthys gen. nov. and Porocephalichthys gen. nov. is based on the key to the genera of the Dinematichthyini, and on the species reviewed by Schwarzhans and Moller (2007). The first dichotomy in that key refers to the anterior nostril position and immediately separates the three genera dealt with herein from all other dinematichthyine genera. Since, however, certain other dinematichthyine genera exist with almost similarly high anterior nostrils we begin the following key with a separation from those few genera. la. Anterior nostril positioned high (equal or less than 1/3 the distance from upper lip to anterior margin of eye).2 1 b. Anterior nostril positioned low (1 /4 to 1/6 the distance from upper lip to anterior margin of eye) . all other Dinematichthyini (see Schwarzhans and Moller (2007)) 2a. Single pair of (outer) pseudoclaspers. . Lapitaichthys 2b. Pair of inner and outer pseudoclaspers.3 3a. Sulcus of otolith with undivided colliculi. . Diancistrus [note: only a few species of the genus exhibit an elevated anterior nostril] 3b. Sulcus of otolith with separated colliculi (not known for Alionematichthys samoensis) .4 4a. Inner pseudoclasper stalked with sucker-disk shaped tip. Eusurculus pistillum [note: the only species in Eusurculus with an elevated anterior nostril] 4b. Inner pseudoclasper variable, flap or stick-shaped or folded, but not stalked and without sucker-disk shaped tip.5 5a. Head with continuous squamation on cheeks, opercle and occiput.6 [note: juveniles of Dinematichthys may have a scaleless gap between cheeks and opercle] 5b. Head with scales on cheeks and in some species with scale patch on opercle above opercular spine, rarely also below opercular spine {Alionematichthys) .8 6a. Upper preopercular pore present, a pair of additional mandibular pores, three additional pairs of pores on occiput, one behind eye, two above preopercular pore; sulcus of otolith with nearly equally long ostium and cauda; precaudal vertebrae 14, total vertebrae 47^49, dorsal fin rays 96-104, D/A 37-43; canine teeth absent. Porocephalichthys dasyrhynchus 6b. Upper preopercular pore absent (see Schwarzhans et al. 2005, fig. 1 A), no additional mandibular pores, no pores on occiput; sulcus of otolith with ostium at least two times as long as cauda; precaudal vertebrae 11-12, total vertebrae 41^45, dorsal fin rays 75-92, D/A< 30; canine teeth present {Dinematichthys)...1 89 W. Schwarzhans and P. R. Moller 7a. Inner pseudoclasper with 2 lobes; head squamation not extending beyond tip of maxilla. . Dinematichthys iluocoeteoides 7b. Inner pseudoclasper with 3 lobes; head squamation extending beyond tip of maxilla forward of 3rd posterior mandibular pore. . Dinematichthys trilohatus sp. nov. 8a. Upper preopercular pore present on both head sides; cirri on snout present or absent.9 8b. Upper preopercular pore absent (rarely present on one head side only); cirri on snout absent.15 9a. Cirri on snout absent, eye small (usually < 2% SL). .10 9b. Cirri on snout present (except absent in Alionematichthys suluensis sp. nov.), eye relatively large (usually > 2% SL).11 10a. Scales above opercular spine 3-10; inner pseudoclasper with anterior thorn as long as posterior lobe; otolith length to otolith height 1.8-2.0, sulcus without ventral indentation at junction of ostium and cauda . Alionematichthys crassiceps sp. nov. 10b. Scales above opercular spine absent; inner pseudoclasper with anterior thorn shorter than posterior lobe; otolith length to otolith height 2.2-2.3, sulcus with ventral indentation at junction of ostium and cauda. Alionematichthys minyomma 11a. Scales above opercular spine 5-17; pscudoclaspers not extruding beyond hood in resting position.12 [note: scales 2-6 in few tentatively assigned specimens -Alionematichthys aff. riukiuensis ] lib. Scales above opercular spine 0-2; pseudoclaspers usually extruding beyond hood in resting position .13 12a. Inner pseudoclasper thick, large, anterior thorn entirely fused to anterior rim of outer pseudoclasper; scale patch on opercle below opercular spine in large specimens. Alionematichthys riukiuensis 12b. Inner pseudoclasper thin, small, bifurcate, anterior thorn free from outer pseudoclasper; no scale patch on opercle below opercular spine . Alionematichthys winterbottomi sp. nov. 13a. Cirri on snout absent; anterior thorn of inner pseudoclasper very long, twice as long as posterior lobe. Alionematichthys suluensis sp. nov. 13b. Cirri on snout present; anterior thorn of inner pseudoclasper about as long as posterior lobe.14 14a. Outer pseudoclasper with thick knob on inner face distally; posterior nostril funnel-shaped. . Alionematichthys samoaensis sp. nov. 14b. Outer pseudoclasper thin, without knob; posterior nostril with anterior Hap, not funnel-shaped . Alionematichthys piger 15a. Scales above opercular spine 2; sulcus of otolith with extremely small cauda (ostium length to cauda length 5.5-7, ostium height to cauda height 2.2-2.7) . Alionematichthys shinoharai sp. nov. 15b. Scales above opercular spine absent; sulcus of otolith with moderately small cauda (ostium length to cauda length 3-4, ostium height to cauda height 1.5-2.0). .16 16a. Inner pseudoclasper thick, with posterior lobe folded over anterior thorn; outer pseudoclasper with thickened tip... . Alionematichthysplicatosurculus sp. nov. 16b. Innerpseudoclasperthin, unfolded; outer pseudoclasper with thin or with knob on inner face.17 17a. Inner pseudoclasper with a symmetrically developed anterior thorn and posterior lobe resembling in shape a socket fitting to a small knob on the inner face of the outer pseudoclasper; snout rounded; otolith length to height 2.3-24, with shallow dorsal rim. . Alionematichthys ceylonensis sp. nov. 17b. Inner pseudoclasper a simple flap without significant indentation; snout pointed; otolith length to height 2.1-2.3, with distinct postdorsal projection. . Alionematichthys phuketensis sp. nov. Alionematichthys gen. nov. (Tables 1-12) Type species: Dinematichthys riukiuensis Aoyagi, 1954 (type locality: Ishigaki Island, Ryukyu Islands, Japan). Gender masculine. Dinematichthys (non Bleeker, 1855) in part. — Alcock 1890:432; Aoyagi 1954: 237; Cohen and Nielsen 1978: 57; Sedorand Cohen 1987: 6; Machida 1994: 451; Nielsen et al. 1999: 129; Moller el al. 2004a: 148. Diagnosis. A genus of Dinematichthyini with the following combination of characters: anterior nostril placed high on snout at about one-third or less the distance from lip to anterior rim of eye; head with a scale patch on cheeks, in some species also above opercular spine, rarely below opercular spine; male copulatory organ with two pairs of mostly small pseudoclaspers, the outer up to twice as large as inner; inner pseudoclasper with anteriorly oriented pointed lobe containing supporter and fleshy appendix posteriorly; large size, up to 150 mm SL; precaudal vertebrae 10-12, total vertebrae 39-45, dorsal fin rays 71-92, anal fin rays 55-72, V in D 2.0-24; upper preopercular pore present or, less commonly, absent; sulcus of otolith with ostium at least twice as long as cauda; maxilla expanded posteroventrally. 90 Dinematichthyine fishes of the Indo-west Pacific, Part IV Table 1. Selected meristic characters of the species of Alionematichthys, Dinematichthys and Porocephalichthys: A, frequency distribution of precaudal vertebrae, total vertebrae and anal fin ray counts; B, frequency distribution of dorsal fin ray and pectoral fin rays counts. A precaudal vertebrae total vertebrae anal fin rays 10 11 12 13 14 39 40 41 42 43 44 45 46 47 48 49 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 Alionematichthys revlonensis 2 64 8 51 7 1 4 12 16 22 5 4 - 2 Alionematichthys crassiceps 1 41 4 30 8 1 7 2 5 10 9 6 1 Alionematichthys minvomma 4 15 3 16 1 1 6 6 3 1 1 Alionematichthys phuketensis 9 2 7 1 1 2 1 2 2 Alionematichthys nicer 309 25 5 42 206 73 9 2 2 - 5 10 23 41 57 51 46 38 19 8 2 - 1 1 Alionematichthys phcalosurculus 1 27 2 21 4 i 1 8 6 3 3 6 - 1 Alionematichthys riukiuensis 80 41 1 39 62 18 2 3 5 14 16 6 14 12 10 13 4 4 2 2 Alionematichthys samoaensis 25 8 1 2 29 1 3 8 10 3 5 2 1 1 Alionematichthys shinolutrai 2 1 1 2 Alionematichthys suluensis 15 14 i 24 4 1 2 3 4 3 7 6 2 1 Alionematichthys winterhottomi 25 1 1 3 17 5 1 3 6 1 13 Alionematichthys sp. 1 2 1 1 I 1 Alionematichthys sp. 2 1 1 1 Dinematichthys iluocoeteoides 2 147 5 12 73 46 14 8 1 3 6 9 14 21 23 23 14 10 9 5 i Dinematichthys trilobatus 15 5 10 1 2 1 1 2 3 4 1 Porocephalichthys dasvrlivnchus 9 2 4 3 1 1 3 3 1 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 1V_ 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 19 20 21 22 23 24 25 26 27 28 Alionematichthys cevlonensis 4 - 10 15 15 9 8 3 2 1 4 5 4 Alionematichthys crassiceps 1 4 7 5 8 5 5 1 4 1 2 1 4 2 Alionematichthys minvomma 1 - 1 9 2 5 - - 1 5 7 1 Alionematichthys nliuketensis 2 2 3 - - 1 - - 1 2 4 1 Alionematichthys nicer I - 2 9 21 28 34 50 42 42 41 18 7 6 1 1 1 2 1 5 1 Alionematichthys nlicatosurculus 1 2 1 4 4 8 5 2 1 1 7 3 Alionematichthys riukiuensis i 3 7 4 12 8 13 16 10 12 4 7 7 2 - 1 5 - 3 2 1 Alionematichthys samoaensis 2 3 5 6 7 3 4 3 I 3 8 Alionematichthys shinolutrai 1 - 1 2 Alionematichthys suluensis 1 - 1 2 3 8 5 2 5 1 1 3 5 3 i Alionematichthys winterhottomi 2 7 4 2 6 2 1 2 4 5 Alionematichthys sp. 1 1 - - 1 1 - 1 Alionematichthys SEii i 1 Dinematichthys iluocoeteoides i - 1 5 5 10 15 23 14 24 12 13 6 5 - 1 1 1 2 3 8 2 Dinematichthys trilobatus 5 1 2 1 3 1 - 2 i 7 5 1 Porocephalichthys dasyrhynchus 2 - 2 1 1 1 - 1 1 3 4 3 3 91 W. Schwarzhans and P. R. Moller Comparison. Alionematichthys is clearly related to Dinematichthys, from which it mainly differs by lacking a uniform and continuous head squamation, including the cheeks, opercle and occiput (vs scale patches on the cheeks and, in some species, on the opercle). The head squamation of Dinematichthys (and Porocephalichthys gen. nov.) is unique within the Dinematichthyini and it supports the stability of generic diagnoses in the tribe. Dinematichthys cannot easily be separated from Alionematichthys without that character. Species. Alionematichthys comprises II species; 10 from the Indo-west Pacific - A. ceylonensis sp. nov., A. crassiceps sp. nov., A. phuketensis sp. nov., A. piger (Alcock, 1890) (formerly placed in Dinematichthys), A. plicatosurcuhis sp. nov., A. riukiuensis (Aoyagi, 1954) (formerly placed in Dinematichthys with Dinematichthys megasoma Machida, 1994, as a junior synonym), A. samoaensis sp. nov., A. sliinoharai sp. nov., A. suluensis sp. nov., A. winterbottomi n. sp., and one species from the tropical West Atlantic - A. minyomma (Sedor and Cohen, 1987) (formerly placed in Dinematichthys). The species of the genus Alionematichthys can be arranged in three species groups according to the presence or absence of the upper preopercular pore, the presence or absence of scales on the opercle, differences in the eye size, and the presence or absence of cirri on the snout as follows: 1) Alionematichthys plicatosurcuhis sp. nov. species group: upper preopercular pore absent, no cirri on snout, large eyes and no scales on the opercle, except for A. shinoharai sp. nov. (two scales above opercular spine). This group contains four species (A. ceylonensis sp. nov., A. phuketensis sp. nov., A. plicatosurcuhis sp. nov. and A. shinoharai sp. nov.). However, strong differences in the pseudoclasper and otolith morphology of some species indicate that these species may not be closely related. 2) Alionematichthys riukiuensis species group: upper preopercular pore present, numerous cirri on snout (except cirri lacking in A. suluensis sp. nov.) and large eyes. Presence or absence of scales on the opercle could be used to separate two subgroups, one with five or more scales above the opercular spine (A. riukiuensis and A. winterbottomi sp. nov.), and the other usually without scales on the opercle (rarely up to two scales above the opercular spine) (A. piger, A. samoaensis sp. nov. and A. suluensis sp. nov.). 3) Alionematichthys minyomma species group: upper preopercular pore present, no cirri on snout, small eyes and scales on opercle present or absent (A. crassiceps sp. nov. and A. minyomma). The small eyes and the big head readily distinguish this small group, which combines a western Pacific species and the only Atlantic species of the genus. Remarks. Some of the more widespread species, namely A. crassiceps sp. nov., A. piger and A. riukiuensis, show a remarkable regionalisation usually based on variations of a singie character. Alionematichthys crassiceps sp. nov. and A. riukiuensis contain few specimens, which can only be tentatively assigned to the respective species. In the case of A. pigei- the pseudoclaspers exhibit a wider than usual variation which, however, does not seem to follow regional trends in this particular case. The distinction of these (regional) variations is not sufficient at this stage, but it is possible that future genetic investigations may result in distinguishing more species than we currently recognise. Etymology. Combined from alius (Latin = the other, different) and nematichthys, the stem of the genus name; Dinematichthys, to which Alionematichthys is most similar. The gender is masculine. Alionematichthys ceylonensis sp. nov. (Figs 1-3; Tables 1,2) Material examined. (84 specimens, 19-72 mm SL), HOLOTYPE-USNM 263707, male, 59 mm SL, Sri Lanka, Weligama, 500 yards south of rest house, 05°57’25”N, SO^fi’Q-’E^-bm.C. Koenig, 15 Feb. 1970. PARATYPES - USNM 263731,4 males, 42-65 mm SL, 6 females, 47-63 mm SL, Sri Lanka, Weligama; USNM 366532,4 specimens, 37-72 mm SL, 08 o 33’24'’N, 81°14’42”E, Trincomalee, Sri Lanka, 2-5 m, T. Iwamoto, 25 Sept. 2001; USNM 394975, 25 males, 37-69 mm SL and 40 females, 37-60 mm SL, same data as holotype. Additional material. USNM 366531, 7 specimens, 25-65 mm SL, Sri Lanka, Triconmalee; USNM 366533, 4 specimens, 40-59 mm SL, Sri Lanka, Battocalao; USNM 366534, 6 specimens, 34-60 mm SL. 07°56’N, 81°34'E, Sri Lanka. Diagnosis. Vertebrae 11-12+29-31=41-43, dorsal fin rays 75-83 (average 79), anal fin rays 57-65 (average 60); eyes moderately large (2.0-2,8% SL); snout rounded, without cirri; scales only on cheeks; no upper preopercular pore; outer pseudoclasper broad-based, triangular, with distal knob on inner side; inner pseudoclasper with symmetrically developed anterior pointed lobe and posterior lobe resembling in shape a socket fitting to small knob on inner face of outer pseudoclasper; otolith length to height 2.3-2.4, with shallow dorsal rim, otolith length to sulcus length 1.5-1.6, ostium length to cauda length 3.3^L0. Description (Figs 2-3).The principal meristic and morphometric characters are shown in Tabic 2. Mature at about 45 mm SL. Body slender, with rounded head profile. No cirri on snout. Eye size 2.0-2.8 (2.6)% SL. Head with scale patch on cheek containing 6-8 (8) vertical rows of scales on upper part and 3 vertical rows on lower part. Horizontal diameter of scales on body about 1.5% SL, in 29 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded, with small knob. Anterior nostril positioned high. 1/2.5 distance from upper lip to anterior margin of eye. Posterior nostril small, about one- fifth the size of eye. Opercular spine with free tip, thin and 92 Dinematichthyine fishes of the Indo-west Pacific, Part IV Fig. 1. Sample sites of O Alionematichthys ceylonensis sp. nov., © A. phuketensis sp. nov., © A. shinoharai sp. nov.. O A. plicatosurculus sp. nov., © A. crassiceps sp. nov., and © A. aff. crassiceps. One symbol may represent several samples. Fig. 2. Alionematichthys ceylonensis sp. nov., USNM 263707, holotype, male, 59 mm SL. sharply pointed. Anterior gill arch with 14-18 (17) rakers, 3 elongate. A single short raker is placed between the two lower elongate rakers in holotype and several paratypes. Pseudobranchial filaments 2. Head sensory pores (Fig. 3A, B). Supraorbital pores 2 to 3: first pore in front of second anterior infraorbital pore, second pore above and behind eye usually absent, third pore tubular, at upper termination of gill opening above opercular spine. Infraorbital pores 6 (3 anterior and 3 posterior): first anterior pore behind anterior nostril, second and third anterior pores covered by dermal flap of upper lip, three posterior pores on rear part of upper lip, about 1/3 the size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior): first anterior pore large and tubular, second anterior pore positioned in lateral skin fold, small, third anterior pore at anterior termination of jugular isthmus, three posterior pores on rear part of lower jaw. Preopercular pores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view; third non¬ tubular; no upper preopercular pore. [This description of the position of head sensory pores is used as reference for all subsequent descriptions.] Dentition (of holotype). Premaxilla with 6 outer rows of granular teeth and 1 inner row of larger teeth anteriorly. Anteriormost teeth in inner row up to 1/4 diameter of pupil. Vomer horseshoe-shaped, with 3 outer rows of small teeth and I inner row of large teeth up to 1/5 diameter of pupil. Palatine with 3 outer rows of small teeth and 1 inner row of long teeth up to 1/4 diameter of pupil. Dentary with 5 93 W. Schwarzhans and P. R. Moller Fig. 3. Alionematichthys ceylonensis sp. nov. A, lateral view of head, holotype; B, ventral view of head, holotype; C, ventral view of male copulatory organ, holotype; D, view of left pscudoclaspers from inside, holotype; E, inclined lateral view of male copulatory organ, holotype; F, view of left pscudoclaspers from inside, USNM 394975, 68 mm SL; G, view of left pseudoclaspers from inside, USNM 394975,48 mm SL; H, median view of right otolith, USNM 394975, male, 58 mm SL. outer rows of granular teeth and I inner row of larger teeth anteriorly, up to about 1/3 diameter of pupil. Otolith (Fig. 3H). Elongate in shape, length to height 2.3-2.4 (40-70 mm SL) and thin (otolith height to otolith thickness about 2.5). Anterior and posterior tips moderately pointed, the latter expanded and irregularly ornamented. Dorsal rim shallow; ventral rim likewise shallow, deepest anterior to middle. Inner face moderately convex, outer face slightly concave, both smooth. Otolith length to sulcus length 1.5-1.6. Sulcus slightly supramedialy positioned, not inclined, with separated colliculi and marked notch at ventral rim of sulcus at joint of ostium with cauda. Ostium length to cauda length 3.3-4.0. Ventral furrow distinct, very close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyses present from vertebrae 6 to 12(11). Pieural ribs on vertebrae 2 to 11 (10). First anal fin pterygiophore not or only slightly longer than subsequent pterygiophore. Male copulatory organ (Fig. 3C-G). Two pairs of moderately large pseudoclaspers, outer pair about 50% larger than inner; outer pseudoclasper with broad base, triangular with pointed tip and distinct knob at inner face near tip; inner pseudoclasper with symmetrically developed anterior thorn and posterior lobe resembling in shape a socket fitting to small knob on the inner face of the outer pseudoclasper. Isthmus narrow; penis curved, about as long as outer pseudoclaspers. Colour. Live colour unknown. Preserved colour uniformly dark brown. Comparison. Alionematichthys ceylonensis belongs to the species group without an upper preopercular pore together with A. phuketensis sp. nov., A. plicatosurculus sp. nov. and A. shinoharai sp. nov., from which it differs in the peculiar shape of the inner pseudoclasper and the 94 Dinematichthyine fishes of the Indo-west Pacific, Part IV Table 2. Meristic and morphometric characters of Alionematichthys ceylonensis sp. nov. Holotype + Holotype 83 paratypes USNM- 263707 Mean (range) n Standard length in mm 59 53.9(19-72) 84 Meristic characters Dorsal fin rays 79 78.9 (75-83) 66 Caudal fin rays 15 15.9(15-17) 40 Anal fin rays 61 60.5 (57-65) 66 Pectoral fin rays 19 20.9(19-22) 14 Precaudal vertebrae 12 12 . 0 ( 11 - 12 ) 66 Caudal vertebrae 29 30.0 (29-31) 66 Total vertebrae 41 42.0 (41-43) 66 Rakers on anterior gill arch 17 16.6(14-18) 13 Pseudobranchial filaments - 2 13 D/V 6 5.8 (5-6) 66 D/A 21 22.6 (20-26) 66 V/A 13 13.3(13-15) 66 Morphometric characters in % of SL Head length 27.8 26.8 (25.8-27.8) 11 Head width 15.3 14.1 (12.1-16.1) 11 Head height 16.0 16.4(14.8-17.9) 11 Snout length 5.5 6.0 (5.4-6. 8 ) 11 Upper jaw length 14.8 11.8-15.1 (14.0) 11 Diameter of pigmented eye 2.6 2.5 (2.0-2. 8 ) 11 Diameter of pupil 1.1 1.5(1.1-1.7) 11 Interorbital width 7.8 7.3 (6.3-8.0) 11 Posterior maxilla height 4.7 4.4 (3,44.9) 11 Postorbital length 19.9 19.4(18.5-20.7) 11 Preanal length 47.9 47.5(45.8-51.7) 11 Predorsal length 32.4 31.0(30.0-32.5) 11 Body depth at origin of anal 18.7 18.3(14.5-20.1) 11 fin Pectoral fin length 12.6 12.6(11.7-13.2) 11 Pectoral fin base height 6.5 6.2 (5.2-7.2) II Ventral fin length 23.3 22.7(19.2-24.5) 12 Base ventral fin - anal fin origin 28.3 28.3 (26.7-32.8) 11 knob at the inner face of the outer pseudoclasper and the slender otoliths (otolith length to otolith height 2.3-2.4 vs 2.1-2.3). From A. phuketensis sp. nov., it differs further in the rounded snout profile (vs pointed) and the lack of a distinct postdorsal projection of the otolith. From /(. shinoharai sp. nov.. it further differs in the lack of scales above the opercular spine (vs 2), the long sulcus (otolith length to sulcus length 1.5-1.6 vs 1.9-2.0) and the notch at the ventral sulcus margin at the joint of ostium and cauda (vs no notch). Distribution. Alionematichthys ceylonensis occurs only along the shores of the island of Sri Lanka (Fig. 1). Biology. A 53 mm SL gravid female (USNM 394975) contains 80 embryos, length 4.8 mm TL, and 60 orange eggs, diameter 0.7 mm. Etymology. Named after the type locality, Sri Lanka, previously known as Ceylon. It is an adjective. Alionematichthys crassiceps sp. nov. (Figs 1,4-6; Tables 1,3) Material examined. 53 specimens, 34-63 mm SL. HOLOTYPE - BPBM 9310, male, 63 mm SL, Truk, Caroline Islands, Southfield beach, 08°0’N, 147°0’E, inshore, 0-4 m, rock and Halimeda algae, J. E. Randall, 10 July 1969. PARATYPES - BPBM 9217, 1 male, 40 mm SL, 1 female, 42 mm SL, Palau Islands, NE Arakabesan Island, 0.5-1 m, corals, J. E. Randall and E. S. Helfman, 5 June 1968; BPBM 40932, 3 females, 46, 46 and 47 mm SL, 3 males, 34, 36,43 mm SL, same data as for holotype; AMS I. 20547-042, 1 male, 40 mm SL, 2 females, 45 and 48 mm SL, Panasesa reef, Louisiadc Archipelago, Papua New Guinea, 5-8 m, B. Goldman, 24 March 1969; CAS 227282, 1 male, 49 mm SL, 1°4’28”N, 154°42 , 28”E, 0-15 m, lagoon coral heads on inner margin of reef flat at Tewawaelal, extreme W end of atoll; low tide; water clear, sand & coral. Kapingamarangi Atoll, Pohnpei, Caroline Islands, Micronesia, Atta, Kindaro and H. H. Rofen, 14 July 1954; CAS 222543,2 males, 50 and 52 mm SL, Vanua Levu, Fiji; CAS 222563, 2 males, 40 and 47 mm SL, 1 female, 37 mm SL, Suva, Fiji; CAS 227285, 1 female, 52 mm SL, Vanua Levu, Fiji; CAS 227303, 1 female, 30 mm SL, 1 male, 40 mm SL, Suva, Fiji; USNM 223247, 1 male, 55 mm SL, 07°00’30”N, 158°11 , 55”E, Caroline Islands, Pohnpei, off east side of Jokaj Passage, 0-4 m, V. G. Springer and party, 5 Sept. 1980; USNM 394976,3 females, 42, 43 and 56 mm SL, 4 males 40-57 mm SL, 2 juveniles 12 and 22 mm SL, 18°44’S, 174°06’W, Vava’u Islands, Tonga, 0-11 m, J.T. Williams et ai, 17 Nov. 1993; USNM 352743, 7 males, 48-60 mm SL, 5 females, 47-67 mm SL, 18°44’25'’S, 169° 12’41”E, Vanuatu, Erromango, Port Narevin, 0-5 m, J. T. Williams et al., 28 May 1996; USNM 374196, 1 male, 57 mm SL, 01°3ri2”S, 145°01’30”E, Papua New Guinea, Hermit Islands, Jalun Island, V. G. Springer and party, I Nov. 1978. Tentatively assigned specimens. USNM 366687, 3 males, 47,50 and 53 mm SL, Bay SE of K’en Ting, Taiwan, 0-3 m, V. G. Springer and party, 22 April 1968. Additional material. USNM 334123, 1 male, 49 mm SL, Tonga; USNM 374210, 1 male, 45 mm SL, 4 females, 34-50 mm SL, Tongatapu, Tonga. Diagnosis. Vertebrae 10-11+29-31=40-42, dorsal fin rays 72-81, anal fin rays 57-64; eyes small (1.3-2.5% SL); snout blunt, without cirri; scales on cheeks and patch above opercular spine with 3-10 scales; upper preopercular pore present; outer pseudoclasper broad-based, simple flap like; inner pseudoclasper with free, forward-pointing anterior lobe and broad posterior lobe; otolith length to height 1.9-2.0, with rounded dorsal rim, otolith length to sulcus length 1.7-1.8, cauda very short, ostium length to cauda length 4.5-5.0. 95 W. Schwarzhans and P. R. Moller Fig. 4. Alionematichthys crassiceps sp. nov., BPBM 9310, holotype, male, 63 mm SL. Description (Figs 4-6).The principal meristic and morphometric characters are shown in Table 3. Mature at about 45 mm SL. Body stout, with rounded, robust head profile (interorbital width 6.6-8.9 (7.4)% SL). No cirri on snout. Eye size 1.3-2.5 (1.3)% SL. Head with scale patch on cheek containing 6-10(10) vertical rows of scales on upper part and 3-5 vertical rows on lower part. Horizontal diameter of scales on body of holotype 2.1% SL, in about 26 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end strongly expanded. Anterior nostril positioned high, one-third distance from upper lip to anterior margin of eye. Posterior nostril small, about one-sixth the size of eye or less. Opercular spine with free tip, pointed. Anterior gill arch with 16-19 (16) rakers, 3 elongate. Pseudobranchial filaments 1-2 (1). Head sensory pores (Figs 5A-E, 6A,B). Supraorbital pores 2 to 3. Infraorbital pores 6 (3 anterior and 3 posterior), posterior pores about one-third the size of anterior pores. Mandibular pores 6 (3 anterior and 3 posterior): first anterior pore large, tubular, with cirrus. Preopercular pores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view; third nontubular; small upper preopercular pore. [See description of Alionematichthys ceylonensis for position of pores.] Dentition (of holotype). Premaxilla with 6 outer rows of granular teeth and 1 inner row of larger teeth anteriorly. Anteriomiost teeth in inner row up to 3/4 diameter of pupil. Vomer horseshoe-shaped, with 4 outer rows of small teeth and 1 inner row of large teeth up to 1/2 diameter of pupil. Palatine with 4 rows of equally long teeth up to I /2 diameter of pupil. Dentary with 2 outer rows of granular teeth and 2 inner rows of larger teeth anteriorly, up to about size of diameter of pupil. Otolith (Figs 5M-P, 6E). Compact in shape, length to height 1.9-2.0, moderately thin (otolith height to otolith thickness about 2.5). Anterior tips moderately pointed, posterior tip much expanded postero-dorsally. Dorsal rim evenly curved; ventral rim deep, deepest at its middle or slightly anteriorly. Inner face moderately convex, outer face flat, both smooth. Otolith length to sulcus length 1.7-1.8. Sulcus mcdianly positioned, not inclined, with separated colliculi but no notch at ventral rim of sulcus at joint of ostium with cauda. Cauda very small, ostium length to cauda length 4.5-5.0. Ventral furrow distinct, very close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyscs present from vertebrae 6 to 11 (10). Pleural ribs on vertebrae 2 to 10 (9). First anal fin pterygiophore not or only slightly longer than subsequent pterygiophore. Male copulatoiy organ (Figs 5F-L, 6C, D). Two pairs of moderately large pscudoclaspers, outer pair about 50% larger than inner; outer pseudoclasper with widened base and rounded tip, fiat, flap-like; inner pseudoclasper with free, forward-pointing anterior thorn, connected at base to anterior rim of outer pseudoclasper and broad posterior lobe with deep inclined furrow on its inner margin. Isthmus moderately wide; penis curved, somewhat longer than outer pseudoclaspers. Colour. Live colour unknown. Preserved colour uniformly light brown. Remarks. Alionematichthys crassiceps shows a high degree of regional differentiation. Specimens from Vanuatu (Fig. 5C) and Fiji have a generally higher number of scales (7-10) above the opercular spine than those from Tonga (Fig. 5D) or Micronesia (Fig. 5A, B) (5-7) on either side of the distribution range. Those from Tonga show a peculiar pattern in the way that 3-4 larger scales form a straight row with 2-3 small scales above (Fig. 5D). The ones from western Micronesia (Truk and Palau) (Fig. 5A-B) have the highest number of vertical scale rows on the cheeks (mostly 8 vs 6-7). Those from Pohnpei and Kapingamarangi Atoll (Fig. 5E) appear to have slightly larger eyes than those from other locations. All these differences, however, are very subtle and are not supported by other characters so as to warrant a separation into different species. Three specimens from Taiwan (Fig. 6A-E) are tentatively placed in the species (A. afif. crassiceps ) due to some folded, cirri-like dermal flaps on the chin, a generally lower number of scales above the opercular spine (3-6), a rather short anterior thorn of the inner pseudoclasper and its remote location from the other specimens of the species. Comparison. Alionematichthys crassiceps is readily recognised by its big head and the small eyes, both characters 96 Dinematichthyine fishes of the Indo-west Pacific, Part IV Fig. 5 (part). Alionematichthys crassiceps sp. nov.. A, lateral view of head, holotype; B, ventral view of head, holotype; C, lateral view of head, holotype; D, lateral view of head, USNM 394976, male, 49 mm SL; E, lateral view of head, USNM 223247, male, 55 mm SL; F, inclined lateral view of male copulatory organ, holotype; G, ventral view of male copulatory organ, USNM 374196, 58 mm SL; H, view of left pseudoclaspers from inside, holotype; I, view of left pseudoclaspers from inside, USNM 394976, male, 49 mm SL; ,J, view of left pseudoelaspers from inside, USNM 394976, 57 mm SL; K, view of left pseudoclaspers from inside, CAS 222543,50 mm SL; L, view of left pseudoclaspers from inside, USNM 223247, 55 mm SL. 97 W. Schwarzhans and P. R. Moller Fig. 5 (continued). M, median view of right otolith, CAS 222543, male, 50 mm SL; N, median view of right otolith, USNM 352743, female, 67 mm; O. median view of right otolith, USNM 374196, male, 58 mm SL; P, median view of right otolith, USNM 394976, female, 56 mm SL. that connects it with the West Atlantic A. minyomina, from which it is distinguished by the presence of scales above the opercular spine (vs absent) and the compact otolith with an otolith length to otolith height ratio of 1.9-2.0 (vs 2.2-2.3). The only co-occurring Dinematichthyini with big heads and small eyes are those of the genus Diancistms Ogilby of the Diancistms erythraeus species group which, however, shows a completely different pseudoelasper pattern (see Schwarzhans et ctl. 2005), no scales above the opercular spine and an otolith with an undivided sulcus Fig. 6. Alionematichthys aff. crassiceps, USNM 366687, A, lateral view of head; B, ventral view of head; C, inclined lateral view of male copulatory organ; D, view of left pseudoclaspers from inside; E, median view of right otolith. 98 Dinematichthyine fishes of the Indo-west Pacific, Part IV and fused colliculi. Front other co-occurring species of Alionematichthys, it is distinguished by the following additional characters: from A. piger, by the small outer pseudoclasper (vs long), the lack of cirri on the snout (vs present) and the presence of scales above the opercular spine (vs absent); from A. plicatosurculus sp. nov., by the different pseudclasper pattern, the presence of an upper preopercular pore (vs absent), the presence of scales above the opercular spine (vs absent); from A. riukiuensis, through the anteriorly free inner pseudoclasper (vs connected), the compressed otolith (vs elongate) and the lack of cirri on the snout (vs present); and from A. winterbottomi sp. nov., in the shape of the inner pseudoclasper and the lack of cirri on the snout (vs present). Distribution. Alionematichthys crassiceps occurs along oceanic islands of the West Pacific from Micronesia (Palau, Chuuk, Pohnpei, Kapingamarangi Atoll), Hermit Island and Table 3. Meristic and morphometric characters of Alionematichthys crassiceps sp. nov. Holotype BPBM 9310 Holotype + 42 paratypes Mean (range) n Standard length in mm 63 46.6(12-67) 43 Meristic characters Dorsal fin rays 73 76.0(72-81) 38 Caudal fin rays 16 16 10 Anal fin rays 59 60.7 (57-64) 38 Pectoral fin rays 22 21.7(20-23) 9 Precaudal vertebrae ii 11.0(10-11) 39 Caudal vertebrae 30 30.2(29-31) 39 Total vertebrae 41 41.2(40-42) 39 Rakers on anterior gill arch 16 17.4(16-19) 9 Pseudobranchial filaments 1 1.9 (1-2) 8 D/V 6 5.9 (5-7) 39 D/A 20 20.6(18-23) 39 V/A 13 12.7(12-13) 39 Morphometric characters in % of SL Head length 28.1 30.0(28.1-34.3) 9 Head width 17.6 15.6(13.7-18.1) 9 Head height 19.9 18.6(17.3-20.8) 9 Snout length 7.2 6.7 (5.7-8.0) 9 Upper jaw length 14.7 14.5(13.6-16.9) 9 Diameter of pigmented eye 1.3 1.9(1.3-2.5) 9 Diameter of pupil 1.0 1.2 (0.9-1.8) 9 Interorbital width 7.4 7.8 (6.6-8.9) 9 Posterior maxilla height 4.7 4.9 (4.6-5.7) 8 Postorbital length 21.0 22.3 (21.0-26.3) 9 Preanal length 49.1 49.1 (44.0-57.9) 9 Predorsal length 33.1 33.1 (31.6-36.6) 9 Body depth at origin of anal fin 20.6 19.4(17.9-23.1) 9 Pectoral fin length 15.5 15.5(12.6-18.8) 9 Pectoral fin base height 6.6 6.7 (6.0-8.2) 9 Ventral fin length 21.8 26.3(21.8-29.8) 7 Base ventral fin - anal fin origin 25.7 27.7(24.3-32.1) 9 the Kiriwina (Trobriand) Islands of New Guinea, Vanuatu, Fiji to Tonga (Fig. 1). There are a few tentative records from Taiwan. Biology. A 56 mm SL female (USNM 394976) contained 95 embryos, length 4.1 mm TL, and ca. 50 orange eggs, diameter 0.6 mm. Etymology. Named after the big head characteristic for the species; crassus (Latin = thick) and adjective ending derived from caput (Latin = head). It is a noun in apposition. Alionematichthys minyomma (Sedor and Cohen, 1987) (Figs 7, 8; Tables 1,4) Dinematichthys minyomma Sedor and Cohen, 1987: 6; Nielsen et al. 1999: 130; Moller et al. 2004: 149 (as D. minyomma). Species 2. — Sedor 1985. Tabic 4. Meristic and morphometric characters of Alionematichthys minyomma (Sedor and Cohen, 1987) Holotype USNM 280122 Holotype + 10 paratypes and 28 non-types Mean (range) n Standard length in mm 66 53.3 (26-78) 39 Meristic characters Dorsal fin rays 75 74.7 (71-79) 19 Caudal fin rays 16 16 19 Anal fin rays 58 57.8 (55-61) 19 Pectoral fin rays 23 21.7(21-23) 19 Precaudal vertebrae 10 10.8(10-11) 19 Caudal vertebrae 29 29.1 (29-30) 19 Total vertebrae 39 39.8 (39-40) 19 Rakers on anterior gill arch 19 17.5(15-19) 19 Pseudobranchial filaments 2 2 18 D/V 6 5.5 (5-6) 19 D/A 22 22.4 (21-25) 19 V/A 13 13.1 (12-14) 19 Morphometric characters in % of SL Head length 29.0 28.0 (26.6-29.8) 19 Head width 17.0 15.5(13.8-17.5) 18 Head height 23.5 20.3 (18.0-23.7) 18 Snout length - 7.0(6.3-7.9) 15 Upper jaw length 15.5 14.8(13.8-15.9) 19 Diameter of pigmented eye 1.5 2.0(1.6-2.3) 19 Diameter of pupil - 1.2 (0.9-1.6) 16 Interorbital width 8.5 8.1 (7.0-9.1) 19 Posterior maxilla height 6.0 5.0 (4.8-54) 18 Postorbital length 22.0 20.3(19.3-21.5) 18 Preanal length 49.5 49.9 (45.4-56.0) 19 Predorsal length 32.5 32.1 (30.7-34.0) 19 Body depth at origin of anal fin - 20.9(18.9-23.4) 19 Pectoral fin length 13.5 14.8(13.3-16.8) 18 Pectoral fin base height - 7.1 (6.6-8.1) 15 Ventral fin length 28.0 25.6 (20.7-28.5) 19 Base ventral fin - anal fin origin 32.0 29.1 (25.9-32.1) 18 99 W. Schvvarzhans and P. R. Moller 110*W 100*W 90"W 80*W 70*W 60'W 50'W Fig. 7. Sample sites of O Alionematichthys minyomma (Sedor and Cohen, 1987). One symbol may represent several samples. Material examined. (40 specimens, 26-80 mm SL): see Moller et al., 2004. Diagnosis. Vertebrae 10-11+29-30=39-40, dorsal fin rays 71-79. anal fin rays 55-61; eyes small (1.5-2.3% SL); snout rounded, without cirri; scales only on cheeks; upper preopercular pore present; outer pseudoclasper broad- based, simple flap-shaped, about twice as large as inner pseudoclasper; inner pseudoclasper very short, its margin almost straight from short anterior pointed lobe to posterior lobe; isthmus between pseudoclaspers very wide; otolith length to height 2.2-2.3, with regular dorsal rim, otolith length to sulcus length 1.6, ostium length to cauda length about 3.5-4.0. Remarks. For a detailed description and additional illustrations see Moller et al. (2004) (as Dinematichthys minyomma). Comparison. Alionematichthys minyomma belongs to the small species group of the genus characterised by their small eyes and big head, further characterised by the lack of cirri and the presence of an upper preopercular pore. It is distinguished from the only other species of the group (A. crassiceps ) by the lack of scales above the opercular spine, differences in the shape of the inner pseudoclaspers and the slender otoliths (otolih length to otolith height 2.2-23 vs 1.9-2.0). Distribution. Alionematihthys minyomma represents the only species of the genus in the western tropical Atlantic (off Honduras, Atlantic Panama and Columbia, Haiti, Puerto Rico and northern Antilles) (Fig. 7) i.e. outside of the Indo-west Pacific. Alionematichthysphuketensis sp. nov. (Figs 1, 10, 11, Tables 1,5) Material examined. (9 specimens, 25-41 mm SL). HOLOTYPE - ZMUC P771659, male 40 mm SL, Rawai beach, Phuket, Thailand, 07°47’0”N, 98° 19’0”E, J. Nielsen, 24 Nov. 1972. PARATYPES-ZMUC P771660-67, 1 male, 36 mm SL, 6 females, 28^H mm SL, 1 juvenile, 25 mm SL, same data as holotype. Diagnosis. Vertebrae 12+30-31=42^43, dorsal fin rays 81-89, anal fin rays 60-65; eyes moderately large (2.3- 2.9% SL); body slender, snout pointed, without cirri; scales only on cheeks; no upper preopercular pore (occasionally single pore on one side); outer pseudoclasper broad-based, small; inner pseudoclasper small, anteriorly inclined flap; otolith length to height 2.1-2.3. with distinct postdorsal mm Fig. 8. Alionematichthys minyomma (Sedor and Cohen, 1987). A. lateral view of head, USNM 280123, male, 66 mm SL; B, ventral view of head, USNM 280123, female, 61 mm SL; C, view of left pseudoclaspers from inside, USNM 280123, male, 71 mm SL; D, inclined lateral view of male copulatory organ of same specimen; E, median view of right otolith, USNM 280123, male 71 mm SL. 100 Dinematichthyine fishes of the Indo-west Pacific, Part IV Table 5. Meristic and morphometric characters of Alionematichthys phuketensis sp. nov. Holotype ZMUC P771659 Holotype + 8 paratypes Mean (range) n Standard length in mm Meristic characters 40 33.0(25-41) 9 Dorsal fin rays 83 83.3 (81-89) 9 Caudal fin rays 15 15.2(14-16) 6 Anal fin rays 62 62.9 (60-65) 9 Pectoral fin rays 20 19.9(19-21) 7 Precaudal vertebrae 12 12 9 Caudal vertebrae 31 30.8 (30-31) 9 Total vertebrae 43 42.8 (42-43) 9 Rakers on anterior gill arch 18 16.0(15-18) 6 Pseudobranchial filaments 2 2 4 DW 6 6 9 D/A 24 23.1 (21-25) 9 V/A 14 Morphometric characters in % of SL 13.3(13-14) 9 Head length 26.8 26.7(25.1-27.5) 9 Head width 12.5 11.7(11.0-12.5) 9 Head height 15.0 15.1 (14.0-16.1) 9 Snout length 6.2 6.2 (5. 6 - 6 .5) 9 Upper jaw length 13.1 12.9 (12.5-13.7) 9 Diameter of pigmented eye 2.9 2.6 (2.3-2.9) 9 Diameter of pupil 1.8 1.6 (1.4-1. 8 ) 9 Interorbital width 7.3 7.3 ( 6 .9-7.9) 9 Posterior maxilla height 4.4 3.7 (3.3-4.4) 9 Postorbital length 18.4 18.6(17.9-19.4) 8 Preanal length 44.9 43.4 (41.0-45.6) 9 Predorsal length 31.0 30.7 (29.2-32.2) 9 Body depth at origin of anal fin 16.9 15.3(13.7-16.9) 9 Pectoral fin length 12.4 13.2(11.9-15.0) 9 Pectoral fin base height 5.2 5.8 (5.2-6.3) 9 Ventral fin length 19.8 23.5 (19.8-24.8) 9 Base ventral fin - anal fin origin 25.2 23.9 (22.4-25.2) 9 projection, otolith length to sulcus length 1.6-1.7, ostium length to cauda length 3.0-3.3. Description (Figs 10, llj.The principal meristic and morphometric characters are shown in Table 5. Mature at about 40 mm SL (largest male at 40 mm SL probably subadult). Body slender, with pointed head profile. No cirri on snout. Eye size 2.3-2.9 (2.9)% SL. Head with scale patch on check containing 7 vertical rows of scales on upper part and 4 vertical rows on lower part. Horizontal diameter of scales on body about 1.5% SL, in about 30 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded. Anterior nostril positioned high, 1/3 distance from upper lip to anterior margin of eye. Posterior nostril small, about 1/8 the size of eye. Opercular spine with free tip, thin and sharply pointed. Anterior gill arch with 15-18 (18) rakers, 3 elongate. In holotype and two paratypes, lower two elongate rakers interrupted by one plate-like raker. In rest of paratypes, three elongate rakers in single row. Pseudobranchial filaments 2. Head sensoiy pores (Fig. 11 A, B). Supraorbital pores 2. Infraorbital pores 6 (3 anterior and 3 posterior), three posterior pores about 1/3 the size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior): first anterior pore large, tubular, without cirri. Preopercular pores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view; third non-tubular; no upper preopercular pore (rarely single pore on one side). [See description of Alionematichthys ceylonensis for position of pores.] Dentition (of holotype). Premaxilla with 5 outer rows of granular teeth and I inner row of larger teeth anteriorly. Anteriormost teeth in inner row up to 1/4 diameter of pupil. Vomer horseshoe-shaped, with I outer row of small teeth and 1 inner row of large teeth up to 1/5 diameter of pupil. Palatine with 3 outer rows of small teeth and 1 inner row of large teeth up to 1/5 diameter of pupil. Dentary with 5 outer rows of granular teeth and 1 inner row of larger teeth anteriorly, up to about 1/3 diameter of pupil. Otolith (Fig. 11E-G). Elongate in shape, length to height 2.1-2.3 (36-41 mm SL) and moderately thin (otolith height to otolith thickness about 1.6). Anterior tip moderately pointed, posterior tip expanded and irregularly ornamented. Dorsal rim shallow, with distinct postdorsal projection and furrow thereafter. Inner face moderately convex, outer face almost flat, both smooth. Otolith length to sulcus length 1.6-1.7. Sulcus slightly supramedially positioned, slightly inclined, with separated colliculi and marked notch at ventral rim of sulcus at joint of ostium with cauda. Ostium length to cauda length 3.0-3.3. Ventral furrow very close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyses present from vertebrae 6 to 12. Pleural ribs on vertebrae 2 to 11. First anal fin pterygiophore not or only slightly longer than subsequent pterygiophore. Male copulatory organ (Fig. 11C, D). Two pairs of small pseudoclaspers, which may not be fully mature in all specimens investigated. Outer pseudoclasper a simple flap with broad base; inner pseudoclasper a forward inclined small flap without significant indentation. Isthmus moderately narrow; penis curved, longer than outer pseudoclaspers. Colour. Live colour unknown. Preserved colour uniformly light brown. Comparison. Alionematichthys phuketensis belongs to the species group without an upper preopercular pore together with A. ceylonensis, A. plicatosurculus sp. nov. and A. shinoharai sp. nov. It differs from all of them in the combination of a slender body shape with a pointed snout and the peculiar postdorsal projection of the otolith rim. From A. plicatosurculus sp. nov., it differs further in the simple inner pseudoclasper morphology and from 101 W. Schwarzhans and P. R. Moller 90°E 100*E 110'E 120'E 130*E 140*E 150*E 160*E 170’E 180* 170’W 160*W 150*W 140*W 130°W 120*W Fig. 9. Sample sites of © Alionematichthys suhiensis sp. nov. and O A. piger (Alcock. 1890). One symbol may represent several samples. A. shinoharai sp. nov., in the lack of scales above the opercular spine (vs two), the longer sulcus (otolith length to sulcus length 1.6-1.7 vs 1.9-2.0) and the notch at the ventral sulcus margin at the joint of ostium and cauda (vs no notch). Alionematichthys phuketensis co-occurs with A. piger, A. nukiuensis and Dinematichthys iluocoeteoides. From the latter two it differs readily in the absence of scales above the opercular spine (vs present or entire operclc with scales in D. iluocoeteoides). From A. piger it di tiers in the lack of cirri on the snout, the lack of an upper prcopcrcular pore and the peculiar dorsal projection of the otolith rim. Distribution. Alionematichthys phuketensis so far is only known from the type locality, Ravai Beach near Phuket, Thailand (Fig. I). Etymology. Named after the type locality, the town of Phuket, Thailand. It is an adjective. Alionematichthys piger (Alcock, 1890) (Figs 9, 12, 13, Tables 1,6) Dinematichthys piger Alcock, 1890: 432 (Great Coco Island, Andaman Islands); Alcock, 1905: plate 37, fig. 3; IBrotulina piger (Alcock, 1890). — Nielsen el al. 1999: 126. Material examined. 2288 specimens, 16-84 mm SL. AMS 1.17094-070, 3 males, 5 females, 11 juveniles, 08°30’S, 151°05’E, Kiriwina Islands (Trobriand), Papua New Guinea; AMS 1.17102-004, 2 specimens, 08°30’S I51°05’E, Kiriwina Islands (Trobriand), Papua New Guinea; AMS 1.17424-012, 1 female, 31 °32’S, 159°04’E, Lord Howe Island; AMS 1.18052-040, 2 females, 28-30 mm SL, 01°44’N, 172°59’E, AbaiangAtoll, Gilbert Islands, Kiribati; AMS 1.20629-001, 3 males, 10 females, 12°N, 165°E, Bikini Atoll, Marshall Islands; AMS 1.27134-040, 1 female, 29°27’2”S, 159°06’8”E, Lord Howe Island, 6-9 m; AMS 1.27138-047, 1 male, 46 mm SL. 2 females, 34-67 mm SL,29°27'8”S, 159°05'E, Middleton Reef, Lord Howe Rise, Tasman Sea, 4-9 m; AMS 1.27148-029, 1 female, 29°57'S, 159°01’2”E, Elizabeth Reef, Tasman Sea, 0-10 m; AMS 1.27891-040, 1 specimen, 29°56’S, I59°01’E, Elizabeth Reef, Tasman Sea, 0-5 m; AMS 1.28950-045, 2 specimens, 17°29’1”S, 149°5r3”W, Moorea, French Polynesia, 8-12 m; AMS 1.33740-026,3 males, 49-59 mm SL and 6 females, 10°09.58”S, 144°34.94”E, Ashmore Reef, Coral Sea; AMS 1.34510-020, 42 specimens, 15°44.8’S, 144°38.9'W,Taiaro Atoll,Tuamotu Islands, 0-2 m; AMS 1.37315-032, 2 males, 65-75 mm SL, 4 females, 40-54 mm SL, 18°44'45”S, 169°12’68”E, Erromango, Vanuatu, 0-6 m; AMS 1.39387-001,12 specimens, Pelagai Island, Guam, 1-2 m; BPBM 7553, 1 male, 4 females, Fanning Island, Line Islands; BPBM 8009, 2 males, 5 females, Marshall Islands; BPBM 8242, 1 male, 2 females, Enewetak Atoll, Marshall Islands; BPBM 8651, 1 male, 5 females, Tahiti; BPBM 8658, I female, Tahiti; BPBM 9311, 1 female, Chuuk, Micronesia; BPBM 10240, 1 specimen, Rangiroa, Tuamotu Islands; BPBM 1 0307, 2 specimens, Rangiroa, Tuamotu Islands; BPBM 10658, 1 male and 9 females, Gilbert Island, Kiribati; BPBM 10830,1 specimen, Aitutaki, Cook Islands; BPBM 11588, 1 female, Tahiti; BPBM 12254, 1 male, 2 juveniles, Ducie, Pitcairn; BPBM 13574, 1 male and 1 female, Mangareva, Gambier Islands, French Polynesia; BPBM 13915, 2 females, Rarotonga, Cook Islands; BPBM 13929, I male. Rarotonga, Cook Islands; BPBM 14161, 1 female. Palmyra Atoll, Line 102 Dinematichthyine fishes of the Indo-west Pacific, Part IV Islands; BPBM 14788, 1 male, 3 females, Lord Howe Island; BPBM 15169, 1 male, 1 female, Wake Island, USA; BPBM 15199, 1 male, 58 mm SL, 6 females, 26-60 mm SL, Kiritimati, Line Islands; BPBM 15323, 1 male, 4 females, Betio, Tarawa Atoll, Kiribati; BPBM 15616, 1 female, Guadalcanal, Solomon Islands; BPBM 16495, 5 males, 7 females, Pitcairn, Polynesia; BPBM 16534, 2 males, 1 female, Oeno, Pitcairn; BPBM 16607, 1 female, Oeno, Pitcairn; BPBM 16912, 2 females, Pitcairn; BPBM 16970, 2 males, 31 and 59 mm SL, 5 females, 28-55 mm SL, Pitcairn; BPBM 16993, 1 female, Pitcairn; BPBM 17042,3 females, Pitcairn; BPBM 17223, 1 male, 1 female, Mangareva, Gambier Islands, French Polynesia; BPBM 17690, I female, Ryukyu Islands, Japan; BPBM 22357, 2 females, Marshall Islands; BPBM 22834,2 males, 1 female, Simiian Island, Thailand; BPBM 23262, 1 male and 3 females, Taiwan; BPBM 23338, 2 males and 2 females, Taiwan; BPBM 33514, 1 male. Chesterfield Bank, Coral Sea; BPBM 37451, 1 female, Kiritimati; BPBM 38268, 1 female, Tonga; BPBM 38773,2 males, Tahiti; CAS 227288, 43 specimens, Ifalik Atoll, Micronesia; CAS 57943, 3 specimens, Raroia.Tuamotu Islands; CAS 65671, 1 female, Madang, Papua New Guinea; CAS 81437, 29 specimens, Raroia, Tuamotu Islands; CAS 227280, 1 juv, 21 mm SL, 24 males 40-63 mm SL, 27 females 42-71 mm SL, Kapingamarangi Atoll, Micronesia; CAS 81446, 16 specimens, Ifalik Atoll, Micronesia; CAS 81451, 4 specimens, Kapingamarangi Atoll, Micronesia; CAS 81468, 52 specimens, Raroia, Tuamotu Islands; CAS 81483, 96 specimens, Kapingamarangi Atoll, Micronesia; CAS 222525,29 specimens, Fiji;CAS 222527,2 specimens, Fiji; CAS 222531, 1 female, Fiji; CAS 222537, 1 female, Fiji; CAS 227283,2 specimens, Fiji; CAS 222549,2 specimens, Fiji; CAS 222553, 2 specimens, Fiji; CAS 222554, 4 specimens, Yadua Island, Fiji; CAS 222556, 5 specimens, Fiji; CAS 222569, 1 specimen, Bua Bay, Fiji; CAS 227286, 2 specimens, Vanua Levu, Fiji; CAS 222572, 1 specimen, Vanua Levu, Fiji; CAS 222582, 5 specimens, Balvu Harbour, Fiji; CAS 227293, 61 specimens, Fiji; CAS 222601, 45 specimens, Budd Reef, Fiji; CAS 222604, 9 specimens, Fiji; CAS 222606, 20 specimens, Suva, Fiji; CAS 222607, 40 specimens, Suva, Fiji; CAS 227301, 1 specimen, Suva, Fiji; CAS 222625,8 specimens, Suva, Fiji; CAS 222626, 13 specimens, Suva, Fiji; CAS 222633, 19 specimens, Suva, Fiji; CAS 222640, 2 specimens, Suva, Fiji; KAUM-I. 2042, 1 female, 77 mm SL, off Okinoerabu Island, Kagoshima, Japan, July 1962; KAUM-I. 10652, 1 female, 76 mm SL, off Okinoerabu Island, Kagoshima, Japan; KAUM-I. 10655, 1 female, 56 mm SL, off Okinoerabu Island, Kagoshima, Japan; KAUM-I. 10658, 1 male, 49 mm SL, off Okinoerabu Island, Kagoshima, Japan; KAUM-I. 10663, I male, 33 mm SL, off Okinoerabu Island, Kagoshima, Japan; KAUM-I. 11482-83, 1 male, 58 mm S Land 1 female, 50 mm SL,30°27'23”N, 130 o 29’59”E, Isso, Yakushima Island, Kagoshima, Japan; MCZ 158557, 1 specimen, 04°27’S, 171°13’W, Maura Atoll, Kiribati; MCZ 162572, 4 specimens, 04°27’S, 171°14’W, Manra Atoll, Kiribati; MCZ 162573, 5 specimens, 04°27’S 171°15’W, Manra Atoll, Kiribati; MNHN 1976-0218, 3 females, 54-56 mm SL, Marshall Islands; MNHN 1980- 0563,2 specimens, New Caledonia; MNHN 1980-0696, 1 female. New Caledonia; MNHN 1980-0961, 1 male, 65 mm SL, 1 female, New Caledonia; NMNZ P.035811, 1 male, 44 mm SL, 1 female, 35 mm SL, Niue, Polynesia; NSMT-P 55637, 5 specimens, Hainan Island, China; NSMT-P 55823,2 males, 3 females and 1 juvenile, Hainan Island, China; ROM 82976, 10 specimens, 18°45’52”S, 178°31 ’13”E, Fiji; ROM 47600,5 specimens, 18°46’30”S, 178°30’28”E, Fiji; ROM 47601,10 specimens, 18°46’13”S, 178°28’05”E, Fiji; ROM 61176,7 specimens, 17 o 31’00”S, 149°55’30”W, Society Islands; ROM 61180, males, 30 and 47 mm SL, female 35 mm SL, 17°36’3r’S, 149°48’42”W, Society Islands; ROM 82979, I female, 22°21’00”S, 166°2r00”E, New Caledonia; ROM 82978, 1 female, 68 mm SL, 20°48’51 ”N, 107°05’ 13”E, Vietnam; RUSI40747, 7 specimens, Kiritimati, Kiribati; SMNS 17837,1 specimen, 41 mm SL, Aitutaki Island, Cook Islands, 18°50’34”S, 1 59°47’28”W, 0.2-3.5 m depth. SMNS 19762, 11 specimens, 21°38’57”S, 166°18’31”E, New Caledonia, 0-8.5 m; SMNS 25171, 4 specimens, 2I°35’45”S, 167°50'06”E, Loyalty Islands, New Caledonia, 2-3.8 in; SMNS 21646, 13 specimens, 20°53’35”S, 167°08’02”E, Loyalty Islands, New Caledonia, 0-2.5 m; SMNS 22732, 1 male, 1 female,20°36’30”S, 164°5r50”E,0-6m; SMNS 22782,1 male, 20 o 3U49”S, 164 o 47’01”E, New Caledonia, 0.2-3.5 m; SMNS 22918, 5 specimens, 20°52’10”S, 167°08’02”E, Loyalty Islands, New Caledonia, 0-3 m; SMNS 22967, 5 specimens, 20°51’45”S, 167°08’10”E, Loyalty Islands, New Caledonia, 0-3.5 m; SMNS 23747, 1 male, 1 female, 20°46’54”S, 167°08’19”E, Loyalty Islands, New Caledonia, 0-3 m; SMNS 23923,8 specimens, 20 o 47 , H”S, 167°07’20”E, Loyalty Islands, New Caledonia, 0-3 m; SMNS 24849,18 specimens, 21 °56’ 13 ”N, 120°47’53”E, Taiwan; SMNS 24864, 2 specimens, 21°56’13”N, 120°47’53”E, Taiwan; SMNS 24866, 1 specimen, 21°56’13”N, 120°47’53”E, Taiwan; UF 42494, 2 males, 1 female, Enewetak Atoll, Marshall Islands; USNM 376214, 2 males, 04°0US, 101°0UE, Sumatra, Mentawai; USNM 384607, 24 specimens, 21°N, 120°E, Taiwan; USNM 224327,34 specimens, 05°51 ’N, 158°20’E, Pohnpei, Micronesia, 0-2 m; USNM 297347,59 specimens, 20°24'45”N, 121°55’02”E, Batanes, Philippines, 0-6 m; USNM 384608, 10 specimens, 18°5UN 121°22’E, Fuga Island, Philippines; USNM 384604, 17 specimens, 19°51 ’36”S, 174°25’06”W,Tonga, 0-2m; USNM 338463, 18 specimens, 18°40’55”S, 174°06’09”W, Vava’u Group, Luamoko Island, Tonga, 0-3 m; USNM 384605, 26 specimens, 18 0 44’3r’S, 174°06’36”W, Vava’u Group, Tonga, 0-11 m; USNM 384593,6 specimens, 20°29’ 12”S, 166°19’18”E, Loyalty Islands, New Caledonia, 3-10 m; USNM 384601,38 specimens, 19°3r33”S, 169°29’50”E, Tanna, Vanuatu, I 5 m; USNM 394979, 3 females, 103 W. Schvvarzhans and P. R. Moller Fig. 10. Alionematichthysphuketensis sp. nov., ZMUC P771659, holotype, male, 40 mm SL. Fig. 11. Alionematichthys phuketensis sp. nov.. A, lateral view of head, ZMUC ZMUC P771661, male, 37 mm SL; B, ventral view of head, ZMUC P771661, male, 36 mm SL; C, view of left pseudoclaspers from inside, holotype; D. inclined lateral view of male copulatory organ, holotype; E, median view of right otolith, holotype; F, ventral view of right otolith, holotype; G, median view of right otolith, ZMUC P771660, female, 41 mm SL. 13°52’25”S, 167°33’20”E, Banks Islands, Vanuatu, 9-13 m; USNM 366488, 3 females, 28-43 mm SL, Phuket, Thailand; USNM 366507, 23 males, 40-84 mm SL, 35 females, 39-71 mm SL, Kiriwina Islands (Trobriand), Papua New Guinea; USNM 366510, 3 males and 19 females, 39-64 mm SL, 05°33’N, 95°09’E, Pulau Boenta, Sumatra; USNM 366569, 2 females, 62-84 mm SL, 23°25’S, 151°55’E, Heron Island, Queensland, Australia; USNM 394980, 1 female, 10°35’05”N, I22°08’30”E; Iloilo, Philippines, 0-7 m; USNM 374192, 1 female, 44 mm SL, 09°06’N, 122°55’E, Negros Island, Philippines; USNM 376177, 1 male, 45 mm SL, Taiwan; USNM 376178,1 female, 18°56’34”S, 168°59’34”E, Erromango, Vanuatu, 0-6 m; USNM 376180, 3 males and 7 females, 46-62 mm SL, 08°29’N, 97°39’E, Thailand; USNM 376187, 1 male, 52 mm SL, Sabah; WAM P.29081-016, 3 females, 15°31’S, 123°09’E, Adele Island, Western Australia, 0.1-2.0 m; WAM P.30633-017, 1 female, 05°09’S, I45°50’E; WAM P.31646-002, 1 male, 13°S 121°E; YCM-P 30026, 1 female, Ishigaki Island, South Ryukyu Islands, Japan; YCM P 34215, 1 female, Kakeroma Island, Amami Island, North Ryukyu Islands, Japan. Additional material. USNM 115392, 1 male, 43 mm SL, Phoenix Islands, Kiribati; USNM 115395, I male, 52 104 Dinematichthyine fishes of the Indo-west Pacific, Part IV Fig. 12. Alionematichthyspiger (Alcock, 1890), AMS I.ex. 37315-032, male, 65 mm SL. mm SL, Phoenix Island, Kiribati; USNM 142010,4 males and 4 females, 28-42 mm SL, Rongcrik Atoll, Marshall Islands; USNM 374219,5 males and 5 females, 25—45 mm SL, Rongelap Atoll, Marshall Islands; USNM 142012, 1 male, 45 mm SL, Kwajalein Atoll, Marshall Islands; USNM 142013, 11 specimens, 40-62 mm SL, Enewetak Atoll, Marshall Islands; USNM 142014, 9 specimens, 16—40 mm SL, Bikini Atoll, Marshall Islands; USNM 142017, 18 females, 27-58 mm SL, Bikini Atoll, Marshall Islands; USNM 374220,46 specimens, 32-57 mm SL, Bikini Atoll, Marshall Islands; USNM 142020, 92 specimens, 28-56 mm SL, Bikini Atoll, Marshall Islands; USNM 374215, 2 males and 1 female, 45-58 mm SL, Amo Atoll, Marshall Islands; USNM 167355, 3 males and 7 females, 40-55 mm SL, Onotoa, Gilbert Islands, Kiribati; USNM 210154, 1 male, 32 mm SL, Seram, Tandjungliang, Indonesia; USNM 376161, 86 specimens, 30-84 mm SL, Taiwan; USNM 263663, 5 males and females, 48-60 mm SL, Tahiti, French Polynesia; USNM 263674, 19 males and 42 females, 25-56 mm SL, 16°00’S, 175°53’W, Tonga; USNM 263690, 30 specimens, 29-57 mm SL, 10°55’N, 121°02’E, Palawan, Philippines; USNM 263711, 1 male, 86 mm SL, 21°55’N, 120°49’E, Taiwan; USNM 300090, 1 male and 2 females, 40—45 mm SL, 20°25’N, 121°57’E, Batanes Islands, Philippines; USNM 300093, 7 males and 8 females, 38-75 mm SL, 20°20’N, 121°49’E, Batanes Islands, Philippines; USNM 300098, 1 female, 47 mm SL, 20°28’N, 121°58’E, Batanes Islands, Philippines; USNM 300102,1 male and 2 females, 21°07’N, 121°56’E, Batanes Islands, Philippines; USNM 319899,7 males and females, 45-73 mm SL, Ouvea, Loyalty Islands; USNM 329746, 17 males and 12 females, 31-57 mm SL, 21°18’S, 174°26’W, Tonga; USNM 333251,3 males and females, 29—41 mm SL, Tongatapu, Tonga; USNM 336507, 2 females, 55-60 mm SL, 21°25'S, 174°56’W, Tonga; USNM 374214, 6 males and 3 females, 35-55 mm SL, 19°16’S, 174°22’W, Tonga; USNM 338465, 35 specimens, 32-61 mm SL, Vavau, Tonga; USNM 338966, 7 males and 2 females, 30-45 mm SL, 18°42’S, 174°02’W, Tonga; USNM 338967, 9 males and 14 females, 30-50 mm SL, 18°38’S, 174°04'W, Tonga; USNM 352073, 1 female, 47 mm SL, 16°35’S, 168°06’E, Lamen Island, Vanuatu; USNM 352742, 123 specimens, 38-80 mm SL, 18°44’S, 169°12’E, Erromango, Vanuatu; USNM 358502, 18 males and females, 24-71 mm SL, Tanna, Vanuatu; USNM 361266, 7 females, 35-80 mm SL, 13°44’S, 167°24’E, Banks Islands, Vanuatu; USNM 376169,1 female, 48 mm SL, 15°00’S, 168°03’E, Maewo, Vanuatu; USNM 366200, 5 males and females, 17-65 mm SL, Tonga; USNM 366222, 5 males and females, 38-61 mm SL, Line Islands; USNM 366223, 1 male, 51 mm SL, Starbuck Atoll, Line Islands; USNM 366225, 1 male, 2 females, Bora Bora, Society Islands; USNM 366489, 2 females, 34-49 mm SL, Taka Atoll, Marshall Islands; USNM 376218, 1 female, 44 mm SL, 04°14’S, 152°10’E, Bismarck Archipelago, Papua New Guinea; USNM 366511, 2 females, 36 mm SL, Amami, Kikai Island, Ryukyu Islands; USNM 376213, 2 females, 38-63 mm SL, Kiriwina Islands (Trobriand); USNM 366556, 1 female, 38 mm SL, Taiwan; USNM 366560, 1 female, 49 mm SL, 25°12’N, 121°4UE, Taiwan; USNM 366561, 12 specimens, 43-78 mm SL, 25°12’N, 121 °41 ’E, Taiwan; USNM 366574, 2 males and 8 females, 44-51 mm SL, Kwajalein Atoll, Marshall Islands; USNM 366575, 1 male, 35 mm SL, Phoenix Atoll, McKean Island; USNM 374221, 1 female, 52 mm SL, Kwajalein Atoll, Marshall Islands; USNM 366719, 1 male and 2 females, 51-67 mm SL, Jarvis Island, Line Islands; USNM 366725,2 males and 1 female, 40-51 mm SL, Bora Bora, Society Islands; USNM 366843, 1 male and 1 female, 54-55 mm SL, 20°37’S, 178°40’E, Fiji; USNM 376211, 1 male and 4 females, 30-48 mm SL, 18°57’S, 178°17’E, Fiji; USNM 366847, 1 male, 49 mm SL, 18°56’S, 178°21 ’W, Fiji; USNM 366848,2 males and 4 females, 32-45 mm SL, 17°47’S, 177°13'E, Fiji. Diagnosis. Vertebrae 11-12+29-33=40-45, dorsal fin rays 72-88, anal fin rays 55-70; eyes moderately large (1.7- 2.8% SL); body slender, snout rounded, with many cirri; scales only on cheeks (rarely 1-2 scales above opercular spine); upper preopercular pore present; outer pseudoclasper broad-based, large, often extending beyond hood in resting position; inner pseudoclasper small, anterior and posterior lobes of equal size; otolith length to height 1.9-2.1, with gently curved dorsal rim, otolith length to sulcus length 1.5-1.6, ostium length to cauda length 3.5-4.5. 105 W. Schwarzhans and P. R. Moller Fig. 13 (part). Alionematichthys piger (Alcock, 1890), A, lateral view of head, ROM 82976, male, 5 1 mm SL; B, lateral view of head, USNM 384707, male, 61 mm SL; C, lateral view of head, USNM 366507, male, 54 mm SL; 1), ventral view of head, USNM 384707, male, 61 mm SL; E, lateral view of head, CAS 227288, female, 51 mm SL; F, inclined lateral view of male copulatory organ, USNM 376 1 80, male, 57 mm SL; G, view ot left pseudoclaspers from inside, USNM 376180, male, 57 mm SL; H, view of left pseudoclaspers from outside, USNM 376180, male, 44 mm SL; !, ventral view of male copulatory organ, USNM 366507, male, 54 mm SL; J, view oflcft pseudoclaspers from inside, SMNS 1 7837,41 mm SL; K, view of left pseudoclaspers from inside, ROM 82976,43 mm SL; L, view of left pseudoclaspers from inside, MNI IN 1980-0961,65 mm SL; M, view of left pseudoclaspers from inside, CAS 227288, male, 59 mm SL; N, view of left pseudoclaspers from inside, BPBM 8009, male, 43 mm SL; O, ventral view of female copulatory organ, SMNS 19762, 85 mm SL; P. inclined lateral view of female copulatory organ, SMNS 19762, 85 mm SL. 106 Dinematichthyine fishes of the Indo-west Pacific, Part IV Fig. 13 (continued). Q, median view of right otolith, USNM 297347, male, 67 mm SL; R, ventral view of right otolith, USNM 297347, male, 67 mm SL; S. median view of right otolith, USNM 384605, male, 52 mm SL; T, median view of right otolith, AMS 1.23381-001, isolated otolith; U, median view of right otolith, MNHN 1976-0218, female, 56 mm SL. Description (Figs 12, 13). The principal meristic and morphometric characters are shown in Table 5. Mature at about 40 mm SL. Body slender, with rounded head profile. Many cirri on snout. Eye size 1.7-2.8 (2.6)% SL. Head with scale patch on cheek containing 6-9 vertical rows of scales on upper part and 3^1 vertical rows on lower part. Horizontal diameter of scales on body of a 65 mm SL male (AMS 1.37315-032) about 1.8% SL, in about 27 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded. Anterior nostril positioned high, 1/2.5 distance from upper lip to anterior margin of eye. Posterior nostril small, about one-fifth to one-sixth the size of eye. Opercular spine with free tip, thin and sharply pointed. Anterior gill arch with 13-19 rakers, 3 elongate, sometimes with short plate-like raker between the two lower elongate rakers. Pseudobranchial filaments 2. Head sensory’pores (Fig. 13A-E). Supraorbital pores 2 to 3. Infraorbital pores 6 (3 anterior and 3 posterior), three posterior pores about 1/3 the size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior): first anterior pore large, tubular, without cirri. Preopercularpores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view; third nontubular; upper preopercular pore present. [See description of Alionematichthys ceylonensis for position of pores.] Dentition (of a 59 mm SL specimen - USNM 376180). Premaxilla with 5 outer rows of granular teeth and 2 inner row of larger teeth anteriorly. Anteriormosl teeth in inner row up to 1/2 diameter of pupil. Vomer horseshoe-shaped, with 2 outer rows of small teeth and I inner row of larger teeth up to 1/3 diameter of pupil. Palatine with 2 outer rows of small teeth and 1 inner row of larger teeth up to 1/3 diameter of pupil. Dentary with 5 outer rows of granular teeth and 1 inner row of larger teeth anteriorly, up to about 3/4 diameter of pupil. Otolith (Fig. 13Q-U). Moderately elongate in shape, length to height 1.9-2.1 and thin (otolith height to otolith thickness about 2.3). Anterior tip broadly rounded, posterior tip expanded and irregularly ornamented. Dorsal rim gently curved, without distinct angles. Inner face moderately convex, outer face almost flat, both smooth. Otolith length to sulcus length 1.5-1.6. Sulcus slightly supramedially positioned, slightly inclined, with separated colliculi and marked notch at ventral rim of sulcus at joint of ostium with cauda. Ostium length to cauda length 3.5-4.5. Ventral furrow very distinct and close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyses present from vertebrae 6 to II (12). Pleural ribs on vertebrae 2 to 10(11). First anal fin plcrygiophore not or slightly longer than subsequent pterygiophore. Male copulatory organ (Fig. 13G-N). Two pairs of pseudoclaspers, outer usually very large and extending beyond hood in resting position. Outer pseudoclasper a large, flat flap with broad base; inner pseudoclasper short, 107 W. Schwarzhans and P. R. Moller sometimes compressed, with anterior thorn and posterior lobe being equal in size. Isthmus narrow; penis curved, about as long as outer pseudoclaspers. Female copulatory organ (Fig. 130, P). Few female specimens of A. piger show soft dermal flaps around the vagina which may reach a quite considerable size, as in the figured specimens but do not exhibit diagnostic valuable morphology. Colour. According to Alcock (1890:433), the live colour is uniform dark brown, almost black. The preserved colour is similar. Variability. Alionematichthyspiger is remarkable for an unusual wide variation in pseudoclasper morphology that does not follow a geographic pattern. The majority of the male specimens show a relatively large outer pseudoclasper for a member of this genus, extending beyond the hood in resting position and about twice as long as the inner pseudoclasper. In extreme forms the inner pseudoclasper is depressed (Fig. 13J, K.). On the other extreme, the inner pseudoclasper is rather stubby and the outer pseudoclasper less than half the length of the inner pseudoclasper and not extruding in resting position (Fig. 13M, N). The very large variation is also reflected in the meristic counts (Tables 1, 5) and variations in the amount of cirri on the snout (Fig. 13A-E) which may represent geographic variances. Otherwise, heads (Fig. 13A-E) and otoliths (Fig. 13Q-U) of fishes with such diverse pseudoclaspers do not document matching variation. Remarks. Alcock (1890) described Dinematichthys piger from Great Coco Island, Andaman Islands. The unique holotype of D. piger (ZSI F12939) was not available for review (see also Schwarzhans and Moller 2007: 66). According to Alcock’s description it is a 61 mm long specimen with 75 dorsal fin rays, 55 anal fin rays and 90 scales along the lateral line and his figure shows a fish with a rather high position of the anterior nostril at 1 /3 the distance from upper lip to aggregate distance to anterior margin of eye and no scales on the opercle. The latter two observations agree with a species of the genus Alionematichthys. There are two species of the genus Alionematichthys observed along the shores of the Andaman Sea, one without scales above the opercle, the other with many scales above the opercle. The latter has been attributed to a species described from the Ryukyu Islands, namely Dinematichthys riukiuensis, now placed in the genus Alionematichthys (see below). Another species described above from the Thailand coast, A. phuketensis, across the Andaman Sea from Alcock’s type locality, does not seem to attain such size and has higher dorsal (81-89) and anal (60-65) fin ray counts. Finally, Alcock’s drawing is detailed enough to reveal that the holotype must be a female. With all these facts in mind it has to be considered that the specimens attributed to A. piger in the following and the diagnosis of the species and description are not fully satisfactory until the holotype has been re-studied. Comparison. Alionematichthys piger belongs to the species group with an upper preopercular pore, large eyes and no (rarely 1 or 2) scales above the opercular spine comprising also A. samoaensis sp. nov. and A. suluensis sp. nov., which both have a much more restricted distribution. It differs from A. samoaensis sp. nov. by the lack of a knob at the distal inner end of the outer pseudoclasper and the simple flap at the anterior margin of the posterior nostril (vs strongly elevated funnel shaped). From A. suluensis, it differs in the anterior lobe of the inner pseudoclasper being equally long as its posterior lobe (vs very long anterior lobe, twice as long as posterior lobe) and many cirri on the snout (vs no cirri). Distribution. Alionematichthys piger is one of the most common and most widely distributed Dinematichthyini (Fig. 9). Its distribution ranges from the Andaman Islands to Sumatra, Thailand, Vietnam, Taiwan and the Ryukyu Table 6. Meristic and morphometric characters of Alionematichthys piger ( Alcock, 1890) Holotype * Holotype + 340 non-types Mean (range) n Standard length in mm Meristic characters 61 50.6 (26-86) 99 Dorsal fin rays 75 79.7 (72-88) 305 Caudal fin rays - 16.0(14-18) 82 Anal fin rays 55 61.8(55-70) 305 Pectoral fin rays - 20.6(19-22) 9 Precaudal vertebrae - 11.1 (11-12) 340 Caudal vertebrae - 31.1 (29-34) 337 Total vertebrae 42.1 (40-45) 337 Rakers on anterior gill arch - 15.8(13-19) 31 Pseudobranchial filaments - 2 2 D/V - 5.7(5-7) 338 D/A - 22.7 (19-28) 336 V/A Morphometric characters in % of SL 13.0(12-16) 337 Head length 25.1 25.7(24.0-27.5) 10 Head width - 14.0(11.8-15.0) 9 Head height 16.7 15.9(13.6-16.8) 10 Snout length - 5.8(54-6.6) 9 Upper jaw length - 13.5 (12.5-14.4) 9 Diameter of pigmented eye 2.6 2.3(1.7-2.8) 10 Diameter of pupil - 1.4 (0.9-1.6) 9 Interorbital width - 7.1 (6.5-7.9) 9 Posterior maxilla height - 4.5 (4.0-5.2) 9 Postorbital length 17.8 18.6(17.6-19.5) 10 Preanal length 50.4 48.0 (44.4-50.9) 10 Predorsal length 31.2 30.1 (27.4-31.2) 10 Body depth at origin of anal fin 16.6 17.8(14.8-19.5) 10 Pectoral fin length 13.9 13.5(11.9-14.7) 10 Pectoral fin base height ■ 5.7 6.1 (5.7-6.9) 10 Ventral fin length - 24.0(21.2-27.6) 9 Base ventral fin - anal fin origin - 29.7(25.3-34.5) 9 * Data taken from Alcock (1890 and 1905). 108 Dinematichthyine fishes of the Indo-west Pacific, Part IV Islands in the north, to Australia and as far south as Lord Howe Island and eastwards to Micronesia, the Line Islands, the Tuamotu Archipelago and Pitcairn Group. Together with Diancistrus katrineae Schwarzhans, Moller and Nielsen, it represents the species farthest to the east in the West Pacific at Ducie Atoll. The species is found in holes and crevices of coralline and volcanic rock. Alionematichthysplicatosurculus sp. nov. (Figs 1, 14, 15, Tables 1,7) Material examined. (152 specimens, 28-79 mm SL). HOLOTYPE - USNM 384198, male, 79 mm SL, 10°34’45”N, 122°30’30”E, 0-4 m, small cove lined with volcanic rock and with rocky boulders in cove, volcanic rock boulders and sandy channels, coral over part of area; Guimaras Island, Pulang Duta, Sinabsapan, Panay, Philippines, J. T. Williams, D. G. Smith, K. E. Carpenter, C. Carpenter, N. Minsalan and T. Minsalan, 24 Sept. 1995. PARATYPES - USNM 209801. 1 male, 53 mm SL, Indonesia, Ambon, off Tandjung Suli, shallow coral reef, 0-2 m, V. G. Springer and M. F. Gomon, II Jan. 1973; CAS 227311, 1 female, 45 mm SL, 05°14’S, 145°47’E, Papua New Guinea, Madang, Nagada Harbour, 1-7 m, S. G. Poss, D. Catania et al., 5 December 1987; USNM 365822, 2 males, 57 and 62 mm SL, Papua New Guinea, Papua, Harvey Bay, east of Oro Bay, 0-10 m, T. R. Roberts, 6-7 August 1975; USNM 365832, 1 male, 55 mm SL, 2 females, 46 and 58 mm SL, 05°11’S, I45°50’E, Papua New Guinea, Kranket Island, mangrove right up to coral, 0-1 m, V. G. Springer etal., 7 Nov. 1978; USNM 366567, 1 male, 50 mm SL, Solomon Islands, Bougainville, Kieta Harbor, coral reef offmain market, 1-4 m, D. M. Cohen and party, 10 Mar. 1965; USNM 366596, 1 female, 44 mm SL, 09°06’30”N, 122°55’24'’E, Philippines, ncarGiligaon, N ofMaloh, Negros, 0-2 m, L. W. Knapp etal. 26 April 1979; USNM 376183, I male, 47 mm SL, 04°14'S, 152°10’E, Papua New Guinea, Bismarck Archipelago, Rabaul, New Britain, W. Davis et al., 27 Feb. 1965; USNM 394981, 3 females, 40, 52 and 62 mm SL, 1 male 48 mm SL, same data as holotype; WAM P.30635-001, 1 male, 45 mm SL, Papua New Guinea. Madang, 0-4 m, G. R. Allen et al., 10 Feb. 1993; WAM P.29624-044, 5 females, 37-52 mm SL, Table 7. Meristic and morphometric characters of Alionematichthys plicatosurculus sp. nov. Holotype Holotype + 29 paratypes n USNM 384198 Mean (range) Standard length in mm Meristic characters 79 48.9 (28-79) 30 Dorsal fin rays 77 80.3 (76-84) 28 Caudal fin rays 16 16 7 Anal fin rays 64 62.6 (60-67) 28 Pectoral fin rays 20 20.2(19-21) 11 Precaudal vertebrae 12 12.0(11-12) 27 Caudal vertebrae 31 31.1 (30-33) 27 Total vertebrae 43 43.1 (42-45) 27 Rakers on anterior gill arch 15 16.3(14-18) 11 Pseudobranchial filaments 2 2 11 D/V 6 5.9 (5-6) 27 D/A 20 22.3 (19.0-25.0) 27 VIA 14 13.4(11.0-19.0) 27 Morphometric characters in % of SL Head length 24.7 26.0(24.7-28.1) 11 Head width 14.5 13.6(11.7-15.6) 11 Head height 16.6 16.0(13.8-17.7) 11 Snout length 5.3 5.9(5.1-7.3) 11 Upper jaw length 13.3 13.4(12.5-14.5) 11 Diameter of pigmented eye 2.2 2.6 (2.2-3.0) 11 Diameter of pupil 1.2 1.6(1.2-2.2) 11 Interorbital width 7.5 7.2 (6.7-8.1) 11 Posterior maxilla height 4.5 4.4 (3.8-4.8) 11 Postorbital length 17.8 18.5(17.6-19.2) 11 Preanal length 45.9 46.1 (43.5-47.7) 10 Predorsal length 30.4 31.3 (29.8-33.0) 11 Body depth at origin of anal fin 18.2 18.4(17.1-19.4) 11 Pectoral fin length 13.6 13.9(11.9-16.0) 11 Pectoral fin base height 6.0 5.9 (5.2-6.7) 11 Ventral fin length 22.0 23.5(19.4-25.8) 14 Base ventral fin - anal fin origin 22.3 26.6 (22.3-28.5) 11 9 males, 28-52 mm SL, 05°09’S, 145°48’E, Papua New Guinea, Madang, G. R. Allen, 15 Oct. 1987. Fig. 14. Alionematichthys plicatosurculus sp. nov., USNM 394198, holotype, male, 79 mm SL. 109 W. Schwarzhans and P. R. Moller Fig. 15. Alionemalichthys plicatosurcttlus sp. nov.. A, lateral view of head, holotype; B. ventral view of head, holotype; C, lateral view of head, paratypc, USNM 394981, female, 52 mm SL; D, inclined lateral view ofmalecopulatory organ, holotype; E, view of left pscudoclaspcrs from inside, holotype; F, ventral view of left pseudoclaspers, holotype; G, ventral view of male copulatory organ, holotype; H, view of left pseudoclaspers from inside, USNM 366567, 50 mm SL; I, median view of right otolith, holotype; J, ventral view of right otolith, holotype. 110 Dinematichthyine fishes of the Indo-west Pacific, Part IV Additional material. BMNH 1907.2.28.11,1 specimen, (bad condition), Neu-Lauenburg/ Duke of York Islands, Bismarck Archipelago, Papua New Guinea; USNM 263681,33 males and 52 females, 30-70 mm SL, 09°06’N, 122°55’E, Philippines, Negros; USNM 263683,4 males and 9 females, 33-71 mmSL, 09°03’N, 122°59’E, Philippines, Negros; USNM 377193, 1 male and 3 females, 68-72 mm SL, 09°10’N, 123°26’E, Philippines, Siquijor Island; USNM 377202, 8 specimens, 48-71 mm SL, 10°35’N, 122°08’E, Philippines, Panay Island; USNM 345618, 5 specimens, 11°25’N, 123°15’E, Panay, Philippines; USNM 365818,20 specimens, Papua New Guinea, Bagabag Island; USNM 366480, 1 female, 52 mm SL, 05°14’S, I45°47’E, Papua New Guinea, Madang; USNM 3766222, 1 male, 66 mm SL, 05°14'S, 145°47’E, Papua New Guinea, Madang, USNM 366508, 3 females, 40-60 mm SL, 03°23’S, I43°40’E, Papua New Guinea, Mushu Island. Diagnosis. Vertebrae 11-12+31-33=42-45, dorsal fin rays 76-84, anal fin rays 60-67; eyes moderately large (2.2—3.0% SL); body slender, snout rounded, without cirri; scales only on cheeks; no upper prcopercular pore; outer pseudoclasper broad-based, large, sometimes extending beyond hood in resting position, with thickened tip, particularly on the inner face; inner pseudoclasper large, thick, with posterior lobe folded over anterior pointed lobe; otolith length to height 2.1 -2.3, with gently curved dorsal rim, otolith length to sulcus length 1.6-1.7. ostium length to cauda length 3.5=4.5. Description (Figs 14-15).The principal meristic and morphometric characters are shown in Table 7. Mature at about 40 mm SL. Body slender, with rounded head profile. No cirri on snout. Eye size 2.2-3.0 (2.2)% SL. Head with scale patch on cheek containing 5-6 (6) vertical rows of scales on upper part and 2-3 vertical rows on lower part. Horizontal diameter of scales on body of holotype about 1.7% SL, in 32 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded, with small knob. Anterior nostril positioned high, 1/2.5—1/3 distance from upper lip to anterior margin of eye. Posterior nostril small, about 1/5 to 1/6 the size of eye. Opercular spine with free tip, thin and sharply pointed. Anterior gill arch with 14-18(15) rakers, 3 elongate in row. Pseudobranchial filaments 2. Head sensory pores (Fig. 15 A-C). Supraorbital pores 3. Infraorbital pores 6 (3 anterior and 3 posterior), three posterior pores about one-quarter the size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior); first anterior pore large, tubular, without cirri. Prcopercular pores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view; third tubular; upper preopercular pore absent. [See description of Alionematichthys ceylonensis for position of pores.] Dentition (of holotype). Premaxilla with 7 outer rows of granular teeth and 1 inner row of larger teeth anteriorly. Anteriormost teeth in inner row up to 1/4 diameter of pupil. Vomer horseshoe-shaped, with 2 outer row of small teeth and 1 inner row of large teeth up to 1/4 diameter of pupil. Palatine with 3 outer rows of small teeth and 1 inner row of long teeth up to 1/3 diameter of pupil. Dentary with 5 outer rows of granular teeth and 1 inner row of larger teeth anteriorly, up to about 1/2 diameter of pupil. Otolith (Fig. 151, J). Moderately elongate in shape, length to height 2.1-2.3 and thin (otolith height to otolith thickness about 3). Anterior tip broadly rounded to slightly angular, posterior tip much expanded and slightly ornamented. Dorsal rim gently curved, without distinct angles. Inner face moderately convex, outer face concave to almost flat, both smooth. Otolith length to sulcus length 1.6-1.7. Sulcus slightly supramedially positioned, slightly inclined, with separated colliculi and marked notch at ventral rim of sulcus at joint of ostium with cauda. Ostium length to cauda length 3.5-4.5. Ventral furrow feeble and close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyses present from vertebrae 6 to 11 (12). Pleural ribs on vertebrae 2 to 10(11). First anal fin pterygiophore slightly longer than subsequent pterygiophore. Male copulatory organ (Fig. 15D-H). Two pairs of moderately large pseudoclaspers. Outer pseudoclasper large, sometimes extending beyond hood in resting position, with thickened tip (particularly in large specimens) expressed as massive knob on inner face; inner pseudoclasper large, thick, with posterior lobe very' large and folded over anterior pointed lobe. Isthmus moderately narrow; penis curved, about as long as outer pseudoclaspers. Colour. Live colour unknown. Preserved colour uniformly brownish. Comparison. Alionematichthys plicatosurculus belongs to the species group without an upper preopercular pore together with A. ceylonensis, A. phuketensis and A. shinoharai sp. nov., from which it differs in the peculiar shape of the thick inner pseudoclasper with the posterior lobe folded over the anterior lobe and the massive tip of the inner face of the outer pseudoclasper. For further distinction from A. ceylonensis and A. phuketensis sec respective descriptions. From A. shinoharai sp. nov. it differs further in the lack of scales above the opercular spine (vs 2), the long sulcus (otolith length to sulcus length 1.5-1.6 vs 1.9-2.0) and the notch at the ventral sulcus margin at the joint of ostium and cauda (vs no notch). Distribution. Alionematichthys plicatosurculus is known from the Philippines south of 12°N, Ambon and along the northern coast of New Guinea to the Bismarck Archipelago and Solomon Islands (Fig. 1). Etymology. Named after the characteristic folded inner pseudoclasper by a combination ofplicatus (Latin = folded) and surcutus (Latin = grapevine tendril) used for the term pseudoclasper. It is a noun in apposition. W. Schvvarzhans and P. R. Mollcr Fig. 16. Sample sites of O Alionematichthys riukiuensis (Aoyagi, 1954), * Alionematichthys aff. riukiuensis 1. * Alionematichthys aff. riukiuensis 2, © A. samoaensis sp. nov., © A. winlerbottomi sp. nov., © Alionematichthys sp. 1, © Alionematichthys sp. 2. One symbol may represent several samples. Fig. 17. Alionematichthys riukiuensis (Aoyagi, 1954), A, fresh dead, WAM P.30909-002 [exact specimen and size not known], B, NMST-P 55637, male,74 mm SL. 112 Dinematichthyine fishes of the Indo-west Pacific, Part IV Alionematichthys riukiuensis (Aoyagi, 1954) (Figs 16-19, Tables 1,8) Dinematichthys riukiuensis Aoyagi, 1954: 235; Machida 1992: 270; Machida 1994: 461; Hayashi 1995: 11; Nielsen et al. 1999: 130. Brotulinafusca. — Matsubara 1955:800 (not Diancistrus fuscus Fowler, 1946). Dinematichthys megasoma Machida, 1994:459; Nielsen et al. 1999: 130; Schwarzhans and Mollcr2005: 78. Material examined. 536 specimens, 19-150 mm SL. PARALECTOTYPE - YCM-P 30001, female, 72 mm, Okinawa Island, Ryukyu Islands; NON-TYPES - AMS 1.37399-076, male, 73 mm SL, Vanuatu; AMS 1.18739-037, 1 male and 2 females, 14°42’S, 145°27’E, Lizard Island, Queensland, Australia; AMS 1.19108-032, 2 males and 3 females, 14°40’S, I45°28’E, Lizard Island, Great Barrier Table 8. Meristic and morphometric characters of Alionematichthys riukiuensis (after Aoyagi 1954). Holotype YCM-P 30023* Holotype + non-types Mean (range) n Standard length in mm Meristic characters 90 72.3(20-148) 121 Dorsal fin rays 79 83.9 (77-92) 107 Caudal fin rays 16 15.7(14-16) 40 Anal fin rays 61 65.1 (59-72) 107 Pectoral fin rays 22 20.5(19-23) 11 Precaudal vertebrae 11 11.3(11-12) 121 Caudal vertebrae 31 31.5(29-33) 120 Total vertebrae 42 42.8 (41-44) 120 Rakers on anterior gill arch - 19.1 (15-21) 10 Pseudobranchial filaments - 1.9(0-3) 9 D/V - 5.7 (5-7) 111 D/A - 23.7(19-27) 112 VIA Morphometric characters in % of SL 13.1 (12-14) 112 Head length 29.3 26.8 (25.7-29.3) 12 Head width - 14.8(11.6-18.6) 11 Head height - 17.1 (14.3-20.5) 11 Snout length 7.3 5.8 (4.7-7.3) 12 Upper jaw length 14.3 13.9(12.4-15.1) 11 Diameter of pigmented eye 3.0 2.6 (2.5-3.0) 12 Diameter of pupil - 1.4 (1.2-1.6) 10 Interorbital width 7.7 7.3 (6.2-8.0) 11 Posterior maxilla height 5.3 4.6(3.8-5.3) 11 Postorbital length 20.6 19.5(19.0-20.6) 12 Prcanal length 49.8 47.1 (45.7-49.8) 11 Predorsal length 33.0 31.2 (29.9-33.0) 11 Body depth at origin of anal fin 21.4 19.9(16.8-23.4) 11 Pectoral fin length 15.7 14.4(12.8-16.1) 10 Pectoral fin base height 7.5 6.1 (4.4-7.5) 11 Ventral fin length 20.4 21.6(20.0-24.8) 12 Base ventral fin - anal fin origin - 26.5(21.8-30.9) 10 * Data from Machida (1994) included. Reef, Australia, 1-10 in, D. F. Hoese and party, 17 Nov. 1975; AMS 1.20770-015,42 specimens, 11 °55’S, 143°27’E, Sir Charles Hardy Island, Queensland, Australia; AMS 1.20770-073,11 specimens, 11 °55’S, 143°27’E, Sir Charles Hardy Island, Queensland, Australia; AMS 1.21422-029,5 specimens, 14°S, 145°E, Coral Sea; AMS 1.33740-026, 1 male, 10°S 144°E, Coral Sea; AMS 1.34311-010, 1 male, 5 females, 22°14’S, 150°19’E, Australia, Queensland, Cannibal Group; AMS 1.37315-032,3 males and 4 females, 18°44’S, 169°12’E, Vanuatu, Erromango; AMS 1.37322- 009,4 females, 19°31’S, 169°29’E, Vanuatu, Tanna; AMS 1.37339-076, 1 male and 1 female, 16°47’S, 168°21 ’E, Vanuatu, Epi; AMS 1.37922-021, 1 male, and 1 female, 13°52’S, 167°32’E, Vanuatu. Banks Islands; BPBM 5921, 1 female, Vanuatu, Efate; BPBM 11465, 2 females, New Caledonia; BPBM 40934, 1 male, 2 females, Ryukyu Islands; BPBM 40936,2 females, Similan Island Thailand; CAS 65664, I male, Papua New Guinea, Madang; KAUM-I. 11459,1 male, 61 mmSL,30°27’23”N, 130°29’59”E, Isso, Yakushima Island, Kagoshima, Japan; KAUM-I. 10661, 1 male, 48 mm SL, off Okinoerabu Island, Kagoshima, Japan; KAUM-I. 11484, 1 male, 55 mm SL. 30°27’23”N, 130°29’59”E, Isso, Yakushima Island, Kagoshima, Japan; KSHS 3650 (otolith only); MNHN 1980-0851, 3 females. New Caledonia; NSMT-P 55636, female 74 mm SL, China, Hainan Island, China; NSMT-P 55637, 1 male, 74 mm SL, Hainan Island, China; NTM S. 13425-007, 1 female, 11°58’S, 123°2 HE, Ashmore Islands, Australia; ROM 71841, 9 specimens, Vietnam; ROM 71843, 11 specimens, Vietnam; ROM 71850, 5 specimens, 20°N 107°E; SMNS 26429, 5 specimens, eastsoutheast ofThio, New Caledonia, 21°38’57”S, 166°18’31”E, 0-8.5 m; SMNS 26430, 6 specimens, Mare, Cap Wabao, Loyalty Islands, 21°35’45 ,, S 167°50’06”E, 2-3.8 m; SMNS 22705, 6 specimens, northnorthwest of Ouegoa, New Caledonia, 20°15’20”S, 164°24’10”E, 0.5-3.5 m; SMNS 22804, 5 specimens, east-south-east of Ponerihuen, New Caledonia, 21°04’55”S 165°28'10”E, 0.3-2.8 m; SMNS 26431, 2 specimens, north-west of Hienghene, New Caledonia, 20°31’49”S 164°47’0r’E, 0.2-3.5 m; USNM 199541, 9 males and 14 females, 42-66 mm SL, 04°01’S, 101 °01 ’E, Mentawai, Sumatra; USNM 394978, 5 specimens, 21°N 120°E; EX USNM 345618, 7 males, 29-65 mm SL, 1 female, 40 mm SL, juvenile 20 mm SL, 11°25’N, 123°15’E, Panay, Philippines; USNM 394982,31 specimens, 16°47’S, 168°2UE, Epiu, Vanuatu; USNM 355385, 38 specimens, Tanna, Vanuau; USNM 394984, 1 female, 75 mm SL, 17°31’S, 168°19\ Efate, Vanuatu; USNM 362434, 1 male, 64 ram SL, 3 females, 4 juveniles, 14°S 167°E, Banks Islands, Vanuatu; USNM 366490, 3 males, 72-78 mm SL, 3 females, 57-72 mm SL, 08°29’N, 97°39’E, Similan Islands, Thailand; USNM 366600, 1 male, 62 mm SL, 09°19’N, 123° 18’E, Negros, Philippines; USNM 376188, 1 male, 150 mm SL, 2 females, 116-119 mm SL, Taiwan; WAM P.30305-037, 1 male, 2 females, 13°52’S, 126°56’E, Sir Graham Moore Islands, Western Australia; 113 W. Schwarzhans and P. R. Moller Fig. 18. Alionematichthys riukiuensis (Aoyagi, 1954), A, lateral view of head, ROM 71841, male, 73 mm SL; B, lateral view of head, WAM P.30308-001, male. 112 mm SL; C, ventral view of head, WAM P.30308-001, male, 112 mm SL; D, laieral view of head, AMS 1.34311- 010, female, 69 mm SL; E, inclined lateral view of male copulatory organ, USNM 394982, 92 mm SL; F, view of left pseudoclaspers from inside, USNM 394982, 92 mm SL; G, ventral view of male copulatory organ, SMNS 26430, 130 mm SL; H, view of left pseudoclaspers from inside, USNM 376188, 150 mm SL; I. ventral view of left pseudoclaspers, WAM P.30308-001, 112 mm SL; J, median view of right otolith, WAM P.30308-001, 112 mm SL. 114 Dinematichthyine fishes of the Indo-west Pacific, Part IV WAM P.30308-001,7 specimens, 13°S, 125°E, Timor Sea; WAM P.30909-002,8 specimens, 16°S, 123°E, North West Australia; YCM-P44101, 1 male and 2 females, Ishigaki, Ryukyu Islands; YCM-P44102, 3 specimens, Kakeroma Island, Ryukyu Islands. Additional material. USNM 151418,1 male, 59mm SL, Philippines; USNM 151420,1 male, 74 mm SL, Philippines; USNM 151422, 1 female 73 mm SL, Philippines; USNM 151423, 1 female, 82 mm SL, Philippines; USNM 151424, 1 male, 84 mm SL, Philippines; USNM 151425, 1 female, 102 mm SL, Philippines; USNM 151426, 1 female, 50 mm SL, Philippines; USNM 164589, 1 female, 78 mm SL, Espiritu Santo, Vanuatu; USNM 376160, 1 male and 3 females, 62-70 mm SL, Taiwan; USNM 377258, 15 specimens, 38-83 mm SL, 02°35’S, 150°47’E, Kavieng, New Ireland, Papua New Guinea; USNM 263666, 5 males and 9 females, 34-83 mm SL, 16°25’N, 119°54’E, Luzon, Philippines; USNM 263680, 2 males, 59-109 mm SL, 23°29’S, 152°05’E, One Tree Island, Queensland, Australia; USNM 377214, 25 specimens, 47-109 mm SL, 09°06’N, 122°55’E, Negros, Philippines; USNM 376347, I female, 61 mm SL, 10°52’N, 120°56’E, Palawan, Philippines; USNM 376225, 1 female, 49 mmSL, 10°55’N, 121°02’E, Palawan, Philippines; USNM 263700, 1 male, 68 mm SL, 09°04’N, 123°16’E, Apo Island, Philippines; USNM 263743, 5 specimens, 34-76 mm SL, Milne Bay, Papua New Guinea; USNM 263755, 15 specimens, 31-119 mm SL, Taiwan; USNM 377208, 1 male and 3 females, 79-104 mm SL, 10°35’N, 122°08’E, Panay, Philippines; USNM 376163, 6 females, 68-92 mm SL, 18°44’S, 169°12’E, Erromango, Vanuatu; USNM 355152, 1 female, 100 mm SL, 17°03’S, 168°21’E, Emae, Vanuatu; USNM 356612, 10 specimens, 19-95 mm SL, Tomotu, Santa Cruz Islands, Solomon Islands; USNM 376167, 3 females, 35-51 mm SL, 13°32’S, 167°20\ Banks Islands, Vanuatu; USNM 363192, 1 male and 1 female, 33-75 mm SL, Banks Islands, Vanuatu; USNM 376166, 3 females, 30-88 mm SL, 13°04’S, 167°39’E, Banks Islands, Vanuatu; USNM 365819,7 specimens, 44-72 mm SL, Bagabag Island, Papua New Guinea; USNM 376223,2 males and 4 females, 55-80 mm SL, 05°14'S, I45°47’E, Madang, Papua New Guinea; USNM 366482, 16 males and 39 females, 32-75 mm SL, 10°47’S, 152°24’E, Louisade Archipelago, Deboyne Atoll, Papua New Guinea; USNM 376202, 1 female, 55 mm SL, 05°14’S, 145°47’E, Madang, Papua New Guinea; USNM 376190,1 female, 65 mmSL,05°14’S, 145°47’E, Madang, Papua New Guinea; USNM 366521,10 specimens, 32-80 mm SL, 05°10’S, 145°51’E, Kranket Island, Papua New Guinea; USNM 366522,2 males and I female, 40 mm SL, 01°31’S, 145°01 ’E, Hermit Islands, Papua New Guinea; USNM 366524, 1 male, 67 mm SL, 05°14’S, 145°47’E, Madang, Papua New Guinea; USNM 366557, 1 male and 1 female, 51-73 mm SL, 21°55’N, 120°44’E, Taiwan; USNM 366559, 1 female, 56 mm SL, Taiwan; USNM 374226, 6 specimens, 35-103 mm SL, Taiwan; USNM 366576, 1 female, 52 mm SL,01°12'S, 144°16’E, Ninigo Atoll, Papua New Guinea; USNM 376173, 3 males and 3 females, 64-104 mm SL, 23°29’S, 152°05’E, One Tree Island, Queensland, Australia; USNM 376171, 1 male and 1 female, 63 mm SL, 23°25’S, 151°55’E, Heron Island, Queensland, Australia; USNM 366682, 1 female, 82 mm SL, 23°25’S, 151°55’E, Heron Island, Queensland, Australia; USNM 366688, 1 male, 38 mm SL, Taiwan; USNM 366693, 7 specimens, 21-82 mm SL, Taiwan. Tentatively assigned specimens (A. aff. riukiuensis). BPBM 40938, 5 males and 2 females,Taiwan; USNM 309643, 3 males and 3 females, 40-60 mm SL, 04°52’N, 119°26’E, Sibutu Island, Philippines. Diagnosis. Vertebrae 11-12+30-32=41=44. dorsal fin rays 77-92 (average 84), anal fin rays 59-72 (average 65); eyes moderately large (2.5-3.0% SL); body moderately slender, massive, snout with many cirri; scales on cheeks and large scale patch above opercular spine (6-17), large specimens with few scales also below opercular spine; upper prcopercular pore present; outer pseudoclaspcr broad- based, not very large, not extending beyond hood in resting position; inner pseudoclaspcr not much smaller than outer pseudoclasper, anterior and posterior lobes of equal size, anterior lobe broadly connected to anterior rim of outer pseudoclasper; otolith slender, length to height 2.2-2.4, with gently curved dorsal rim, otolith length to sulcus length 1.6-1.7, ostium length to cauda length 3.5^1.0. Description (Figs 17-19).The principal meristic and morphometric characters are shown in Table 8. One of the largest species attaining up to 150 mm SL; mature at about 50-55 mm SL. Body moderately slender, massive, with moderately pointed head profile. Many cirri on snout. Eye size 2.5-3.0 (3.0)% SL. Head with scale patch on cheek containing 8-13 vertical rows of scales on upper part and 4-5 vertical rows on lower part, scale patch above opercular spine with 6-17 scales, in adults, scaled patch below opercular spine with up to 8 scales. Horizontal diameter of scales on body of a 74 mm SL male, about 1.5% SL, in about 40 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded with small knob. Anterior nostril positioned high, 1/2-2.5 distance from upper lip to anterior margin of eye. Posterior nostril small, about 1/6 to 1/8 the size of eye. Opercular spine with free tip, pointed. Anterior gill arch with 15-21 rakers, 3 elongate. Small plate-like raker between lower two elongate rakers in most specimens. Pseudobranchial filaments 0-3 (usually 2). Head sensory pores (Fig. 18A-D). Supraorbital pores 3. Infraorbital pores 6 (3 anterior and 3 posterior), three posterior pores about one-third size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior): first anterior pore large, tubular, with cirri. Preopercular pores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view; third tubular; upper preopercular pore present. [See description of Alionematichthys ceylonensis for position of pores.] 115 W. Schwarzhans and P. R. Moller Fig. 19. Alionematichthys a(T. riukiuensis (Aoyagi, 1954), A, lateral view ol head. BPBM 40938, male, 62 mm SL; B, ventral view ot head, BPBM 40938, male, 62 mm SL; C, lateral view of head, USNM 309643, male, 58 mm SL; 1). inclined lateral view of male copulatory organ, BPBM 40938, male, 80 mm SL; E, ventral view of male copulatory organ, BPBM 40938, male, 80 mm SL; F, view of left pseudoclaspers from inside, USNM 309643,58 mm SL; G, view of left pseudoclaspers from inside, BPBM 40938, male, 62 mm SL; H, view ofleft pseudoclaspers from inside, BPBM 40938, male, 80 mm SL; 1, median view of right otolith, USNM 309643, male, 58 mm SL. 116 Dinematichthyine fishes of the Indo-west Pacific, Part IV Dentition (of a 74 mm SL male, NSMT-P 55637). Premaxilla with 6 outer rows of granular teeth and 2 inner rows of larger teeth anteriorly. Anteriormost teeth in inner row up to 1/2 diameter of pupil. Vomer horseshoe-shaped, with 6 outer rows of small teeth and 1 inner row of large teeth up to 1/4 diameter of pupil. Palatine with 4 outer rows of small teeth and 1 inner row of long teeth up to 1/4 diameter of pupil. Dentary with 6 outer rows of granular teeth and 1 inner row of larger teeth anteriorly, up to about 1/2 diameter of pupil. Otolith (Fig. 18J). Elongate in shape, length to height 2.2-2.4 and thin (otolith height to otolith thickness about 3). Anterior tip slightly pointed, posterior tip pointed, expanded and irregularly ornamented. Dorsal rim gently curved, without distinct angles. Inner face moderately convex, outer face slightly concave to almost flat, both smooth. Otolith length to sulcus length 1.6-1.7. Sulcus medianly positioned, not inclined, with separated colliculi and marked notch at ventral rim of sulcus at joint of ostium with cauda. Ostium length to cauda length 3.5-4.0. Ventral furrow distinct and close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyses present from vertebrae 6 to 11(12). Pleural ribs on vertebrae 2 to 11 (12). First anal fin pterygiophore not or only slightly longer than subsequent pterygiophore. Male copulatory organ (Fig. 18E-I). Two pairs of rather small pseudoclaspers, not extending beyond hood in resting position. Outer pseudoclasper a broad based flap; inner pseudoclasper slightly shorter, very thick (best seen in ventral view: Fig. 18G), with anterior pointed lobe broadly connected by ligament to anterior rim of outer pseudoclasper, posterior lobe broad, about equal in size. Isthmus moderately narrow; penis curved, slightly longer than outer pseudoclaspers. Colour. Live colour brownish-red or orange-brown (SMNS 22804), sometimes with yellowish vertical fins (Fig. 17A). Preserved colour medium to dark brown. Variability. Alioneniatichthys riukiuensis shows some variation in the number of scales on the opercle above and below the opercular spine and a certain variation of the dorsal and anal fin ray counts, both of which does not seem to follow a geographic pattern. Remarks. Machida (1994) described Dinematichthys megasoma from Australia. These specimens agree well with the variability observed in A. riukiuensis, including the characters regarded by Machida as distinctive from Dinematichthys riukiuensis, i.e. the short tubular posterior nostril, the widely scaled opercle and the lateral scale row counts. We have, however, observed two specific lots which somewhat depart from the general range of variability (see Fig. 19). They are both notable for few or no cirri on the snout and few scales above the opercular spine (2-6) (Fig. 19A-C). They vary amongst themselves in meristics (see Table 1, 8). Their pseudoclaspers resemble those of A. riukiuensis (Fig. 19D-H) perfectly as do their otoliths (Fig. 191). Due to the limited amount of specimens available and the high degree of similarity with A. riukiuensis they are here referred to as Alioneniatichthys aff. riukiuensis. Comparison. Alioneniatichthys riukiuensis belongs to the species group with an upper preopercular pore, large eyes and (many) scales above the opercular spine and cirri on the snout (comprising also A. winterbottomi sp. nov.). It differs mainly in the characters of the pseudoclaspers and the presence of scales below the opercular spine in adults (vs always absent in A. winterbottomi sp. nov.). It differs from A. piger and A. samoaensis sp. nov., which likewise show many cirri on the snout in the presence of many scales above the opercular spine (vs none or rarely 1-2) and the pseudoclasper morphology. Also, A. riukiuensis is notable for the large size it can attain and at which it matures, greater than in all other species of Alioneniatichthys. Distribution. Alioneniatichthys riukiuensis is widely distributed in the Indo-west Pacific from the Ryukyu Islands, Hainan Island and the west coast ofThailand in the north to the northern shores of Australia, New Caledonia and Vanuatu in the south (Fig. 16). Alioneniatichthys samoaensis sp. nov. (Figs 16, 20, 21, Tables 1,8) Material examined. (36 specimens, 22-59 mm SL). HOLOTYPE - CAS 81486, male, 46 mm SL, ca. 11 °03'S, 171°04'W, seaward face of leeward reef on Swains Island, American Samoa, coll. L. R. Taylor Jr., 23 May 1967. PARATYPES-AMS IB. 6513, male, 54 mm SL, 14°27'S, 178°05'W,Alofi, Niue, W.N.McDowall, May 1963; BPBM 24122, 2 females, 56 and 59 mm SL, Samoa, Tutuila, R. C. Wass, 1976-1977; CAS 227305,20 females, 22-53 mm SL, 11 males, 25-47 mm SL, juvenile, 22 mm SL, same data as holotype. Diagnosis. Vertebrae 11-12+29-31=40-43, dorsal fin rays 73-80, anal fin rays 57-64; eyes moderately large (2.2-2.8% SL); body slender, snout with many cirri; scales on cheeks, no scales above opercular spine; upper preopercular pore present; posterior nostril funnel-shaped; outer pseudoclasper broad-based, not extending beyond hood in resting position, with thick distal knob on inner face; inner pseudoclasper about half size of outer pseudoclasper. anterior thorn slightly bent, posterior lobe broader than anterior thorn. Description (Figs 20, 21).The principal meristic and morphometric characters are shown in Table 9. Body moderately slender, with moderately pointed head profile. Many cirri on snout. Eye size 2.2-2.8 (2.2)% SL. Head with scale patch on cheeks containing 6-7 vertical rows of scales on upper part and 2-3 vertical rows on lower part, no scale patch on opercle. Horizontal diameter of scales on body in holotype 1.7% SL, in about 24 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded. Anterior nostril positioned high, one-half distance from upper lip to anterior margin of eye. Posterior nostril with 117 W. Schwarzhans and P. R. Moller Fig. 20. Alionematichthys samoaensis sp. nov., CAS 81486, holotype, male, 46 mm SL. Fig. 21. Alionematichthys samoaensis sp. nov., A. lateral view of head, holotype; B, ventral view of head, holotype; C, inclined lateral view of male copulatory organ. CAS 81486. male, 39 mm SL; D, ventral view of male copulatory organ, holotype; E, view of left pscudoclaspers from inside, holotype; F, view ofleft pseudoclaspers from inside, CAS 81486, 39 mm SL. broad funnel-shaped rim, about one-eighth the size of eye. Opercular spine with free tip, pointed. Anterior gill arch with 14-18 (15) rakers, 3 elongate. In holotype and a few paratypes lower two elongate rakers interrupted by one plate-like raker. In other paratypes, three elongate rakers in row. Pseudobranchial filaments 1-2 (usually 2). Head sensory pores (Fig. 21 A, B). Supraorbital pores 3. Infraorbital pores 6 (3 anterior and 3 posterior), three posterior pores about one-third the size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior); first anterior pore large, tubular, with cirri. Preopercular pores: 3 lower, first and second with joined opening, covered by 118 Dinematichthyine fishes of the Indo-west Pacific, Part IV Table 9. Meristic and morphometric characters of Alionematichthys samoaensis sp. nov. Holotype CAS 81486 Holotype + 35 non-types Mean (range) n Standard length in mm 46 37.3 (22-59) 36 Meristic characters Dorsal fin rays 76 76.6 (73-80) 33 Caudal fin rays 16 16.1 (16-17) 24 Anal fin rays 59 59.4 (57-64) 33 Pectoral fin rays 22 21.6(20-22) 12 Precaudal vertebrae 12 11.2(11-12) 33 Caudal vertebrae 30 30.7(29-31) 33 Total vertebrae 42 41.9(40-43) 33 Rakers on anterior gill arch 15 15.6(14-18) 12 Pseudobranchial filaments 2 1.9 (1-2) 10 D/V 5 5.6 (5-6) 33 D/A 23 22.6 (20-26) 33 V/A 13 13.2(13-14) 33 Morphometric characters in % of SL Head length 27.2 27.2 (26.2-28.9) 10 Head width 14.2 13.8(11.8-16.2) 10 Head height 16.4 16.6(15.3-17.9) 10 Snout length 5.5 6.1 (5.5-6.5) 9 Upper jaw length 14.7 14.3(13.3-15.6) 10 Diameter of pigmented eye 2.2 2.5 (2.2-2.8) 9 Diameter of pupil 1.4 1.4 (1.2-1.5) 9 Interorbital width 7.8 7.6 (7.2-8.0) 10 Posterior maxilla height 4.7 4.7 (3.9-5.3) 10 Postorbilal length 19.8 19.4(18.5-21.2) 10 Preanal length 48.7 47.7 (44.8-50.4) 10 Predorsal length 30.7 31.1 (28.9-32.5) 10 Body depth at origin of anal fin 19.4 18.8(16.3-20.4) 10 Pectoral fin length 13.4 13.5(12.7-15.7) 10 Pectoral fin base height 6.5 6.6 (6.2-6.8) 10 Ventral fin length 26.0 25.5 (22.5-27.1) 10 Base ventral fin - anal fin origin 29.6 28.3(25.8-31.6) 10 dermal flap in lateral view; third tubular; upper preopercular pore present. [See description of Alionematichthys ceylonensis for position of pores.] Dentition (of holotype). Premaxilla with 5 outer rows of granular teeth and 1 inner row of larger teeth anteriorly. Anteriormost teeth in inner row up to 1/3 diameter of pupil. Vomer horseshoe-shaped, with 2 outer rows of small teeth and 1 inner row of large teeth up to 1/3 diameter of pupil. Palatine with 1 outer row of small teeth and 1 inner row of long teeth up to 1/3 diameter of pupil. Dentary with 5 outer rows of granular teeth and 1 inner row of larger teeth anteriorly, up to about 2/3 diameter of pupil. Otolith (not known). Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyscs present from vertebrae 6 to 11 (12). Pleural ribs on vertebrae 2 to 10-11 (12). First anal fin pterygiophore not or only slightly longer than subsequent pterygiophore in females, much longer in males, but not reaching last precaudal parapophysis. Male copulatory organ (Fig. 21C-F). Two pairs of rather small pseudoclaspers, not extending beyond hood in resting position. Outer pseudoclasper a broad based flap, with thick knob distally on inner face, its supporter somewhat bent backward at tip; inner pseudoclasper thick, about half the size of outer pseudoclasper, anterior lobe slightly bent, posterior lobe broader than anterior thorn. Isthmus narrow; penis curved, slightly longer than outer pseudoclaspers. Colour. Live colour unknown. Preserved colour uniformly light brown. Comparison. Alionematichthys samoaensis belongs to the species group with an upper preopercular pore, large eyes and no scales above the opercular spine, comprising also A. piger and A. suluensis sp. nov. Alionematichthys samoaensis obviously is closely related to A. piger, which it replaces in region of Samoa, differing mainly in the distinct knob on the inner face of the outer pseudoclasper (vs thin outer pseudoclasper without knob) and the funnel- shaped posterior nostril (vs anterior rim elevated only). From A. suluensis, it differs readily in the presence of many delicate cirri on the snout (vs absent) and the anterior lobe of the inner pseudocasper being as long as its posterior lobe (vs twice as long). Fig. 22. Alionematichthys shinoharai sp. nov., NSMT-P 34895, holotype, male, 34 mm SL. 119 W. Schwarzhans and P. R. Moller Fig. 23. Alionematichthys shinoharai sp. nov., A, lateral view ofhcad, holotype; B, ventral view of head, holotype; C, view of left pseudoclaspers from inside, holotype; D, inclined lateral view of male copulatory organ, holotype; E, median view of right otolith, holotype; F, ventral view of right otolith, holotype; G, median view of right otolith, YCM-P36416, female, 42 mm SL. Distribution. Alionematichthyssamoaensis is endemic to the Islands of the Samoa group, and is in fact the only endemic Dinematichthyini restricted to that region (Fig. 16). Biology. A 46 mm SL female (CAS 81486) contains ca. 100 embryos, length 4.6 mm TL, Each embryo with three short pigmented rows on the posterior part of the body. Etymology. Named after the type locality, American Samoa. It is an adjective. Alionematichthys shinoharai sp. nov. (Figs 1,22,23, Tables 1, 10) Material examined. (2 specimens, 34^12 mm SL). HOLOTYPE - NSMT-P 34895, male, 34 mm SL, 28°H’2”N, 129°16’E, Sakinome beach, Amami-Oshima Island, Ryukyu Islands, Japan, 5 m, K. Matsuura and M. Aizawa, 14 June 1991. PARATYPES - YCM-P 36416, female, 42 mm SL, Amami Islands, Kakeroma Island, Ryukyu Islands, Japan, 24 Aug. 1995. Diagnosis. Vertebrae 11+32-33=43-44, dorsal fin rays 83-85, anal fin rays 67; eyes large (3.0-3.3% SL); body slender, snout pointed, without cirri; scales on cheeks, two scales above opercular spine; upper preopcrcular pore absent; outer pseudoclasper broad-based simple flap; inner pseudoclasper very small, somewhat depressed; otolith moderately elongate, length to height 2.1 -2.2, with regularly curved dorsal rim, sulcus small, otolith length to sulcus length 1.9-2.0, cauda very small, ostium length to cauda length 5.5-7, ostium height to cauda height 2 . 2 - 2 . 7 . Description (Figs 22-23).The principal meristic and morphometric characters are shown in Table 10. Mature at about 35-40 mm SL (male of 34 mm SL probably subadult). Body slender, with pointed head profile. No cirri on snout. Eye size 3.3 (3.0)% SL. Head with scale patch on cheeks containing 7 vertical rows of scales on upper part and 4 vertical rows on lower part, 2 scales on opercle above opercular spine. Horizontal diameter of scales on body of holotype 1.5% SL, in about 24 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded with small knob. Anterior nostril positioned high, one-third distance from upper lip to anterior margin of eye. Posterior nostril with narrow elevated rim, about one-eighth the size of eye. Opercular spine with free tip, sharply pointed. Anterior gill arch with 15(16) rakers, 3 elongate. Pseudobranchial filaments 2. Head sensory pores (Fig. 23 A, B). Supraorbital pores 3. Infraorbital pores 6 (3 anterior and 3 posterior), three posterior pores about one-third the size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior): first anterior pore large, tubular, with single cirrus anteriorly. Preopercular pores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view; third tubular; upper preopercular pore absent. [See description of Alionematichthys ceyionensis for position of pores.] Dentition (of holotype). Premaxilla with 5 outer rows of granular teeth and 1 inner row of larger teeth anteriorly. Anteriormost teeth in inner row up to 1 /4 diameter of pupil. Vomer horseshoe-shaped, with 1 outer row of small teeth and 1 inner row of larger teeth up to 1/5 diameter of pupil. Palatine with 1 outer row of small teeth and 1 inner row of longer teeth up to 1/4 diameter of pupil. Dentary with 5 120 Dinematichthyine fishes of the Indo-west Pacific, Part IV Table 10. Meristic and morphometric characters of Alionematichthys shinoharai sp. nov. Holotype NSMT- P34895 Paratype YCM- P36416 Standard length in mm Meristic characters 33 42 Dorsal fin rays 83 85 Caudal fin rays 14 14 Anal fin rays 67 67 Pectoral fin rays 22 22 Precaudal vertebrae 11 11 Caudal vertebrae 33 32 Total vertebrae 44 43 Rakers on anterior gill arch 16 15 Pseudobranchial filaments 2 2 D/V 6 5 D/A 20 24 VIA 13 14 Morphometric characters in % of SL Head length 27.2 26.7 Head width 11.9 12.8 Head height 15.9 16.2 Snout length 5.6 6.5 Upper jaw length 13.8 13.8 Diameter of pigmented eye 3.0 3.3 Diameter of pupil 2.0 2.1 Interorbital width 6.5 7.0 Posterior maxilla height 4.0 4.0 Postorbital length 19.8 18.0 Preanal length 44.8 44.9 Predorsal length 32.0 32.6 Body depth at origin of anal fin 16.1 18.1 Pectoral fin length 14.1 15.1 Pectoral fin base height 6.6 5.5 Ventral fin length 24.7 22.6 Base ventral fin - anal fin origin 25.2 24.1 outer rows of granular teeth and 1 inner row of larger teeth anteriorly, up to about 1/3 diameter of pupil. Otolith (Fig. 23E-G). Moderately elongate in shape, length to height 2.1-2.2 (34-40 mm SL), rather thick (otolith height to otolith thickness about 2). Anterior tip sharply pointed, posterior tip pointed, slightly expanded. Dorsal rim gently curved, with postdorsal angle in male specimen (Fig. 23E) and absent in female (Fig. 23G), exhibiting slight sexual dimorphism. Inner and outer face moderately convex, smooth. Otolith length to sulcus length 1.9-2.0. Sulcus medianly positioned, slightly inclined, with separated colliculi, without marked notch at ventral rim of sulcus at joint of ostium with cauda. Cauda very small, ostium length to cauda length 5.5-7, ostium height to cauda height 2.2-2.7. Ventral furrow feeble, close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyses present from vertebrae 6 to 11. Pleural ribs on vertebrae 2 to 10. First anal fin pterygiophore only slightly longer than subsequent pterygiophore. Male copulatory organ (Fig. 23C, D). Two pairs of small pseudoclaspers, which may not be fully mature in single male specimen investigated. Outer pseudoclasper a simple flap with broad base; inner pseudoclasper a small depressed flap without significant indentation. Isthmus moderately narrow; penis slightly curved, longer than outer pseudoclaspers. Colour. Live colour unknown. Preserved colour uniformly light brown. Comparison. Alionematichthys shinoharai belongs to the species group without an upper preopercular pore together with A. ceylonensis, A. phuketensis and A. plicatosurculus. It differs from all these in the presence of two scales above the opercular spine (vs none), the somewhat less highly positioned anterior nostril with one- third distance from upper lip to anterior margin of eye (vs 1/2-2.5) and the peculiar otolith with its short sulcus and the very small cauda. The combination of these characters is so distinctive that they could indicate a separate genus, which, however, will be subject to review of a larger male specimen in order to fully evaluate the pseudoclasper pattern. Distribution. Alionematichthys shinoharai has so far only been found at the northern islands of the Ryukyu Island chain of Japan, i.e. Amami-Oshima (Fig. I). Etymology. Named in honour of Gento Shinohara, NSMT, Tokyo, in recognition of his many contributions to ichthyology and for his kind support of the present revision. It is a noun in apposition. Fig. 24. Alionematichthys suluensis sp. nov., AMS 1.40149-028, holotype, male, 40 mm SL. 121 W. Schwarzhans and P. R. Moller Alionematichthys suluensis sp. nov. (Figs 9,24, 25, Tables 1, 11) Material examined. (57 specimens, 26-49 mm SL). HOLOTYPE - AMS 1.40149-028, male, 40 mm SL, 12°21.70N, 121°27.57E, 0-20 m, rocky surge area at base of cliffs, S tip of Buyamao Island off SE Mindoro Island, Philippines, coll. MIN team, rotenone, 30 May 2000. PARATYPES - AMS 1.40149-028), 1 male, 30 mm SL, 5 females, 30-42 mm SL, 3 juveniles, 13-17 mm SL, same data as holotype; USNM 263691, 1 male, 43 mm SL and 5 females, 41-45 mm SL, 08°5r42”N, 123°24’36”E, Solino Island, Zamboanga, Mindanao, Philippines, 0-5 m, A. Alcala et al., 4 May 1979; USNM 263694, 6 males and 12 females, 27-49 mm SL, 10°52’N, 121°12’E, Palawan, Philippines; USNM 376216,4 males and 1 female, 29-45 mm SL, 09°04'N, 123°08'E, Negros, Philippines; USNM 366601,1 male, 46 mm SL, 09°04’N, 123°16’E,Apo Island, Philippines; USNM 366602,1 female, 34 mm SL, 09°23’N, 123°15’E, Negros, Philippines; USNM 374183, 1 male, 31 mm SL, 10°35’05”N, 122°08’30”E, rocky tidepool, Talisayan Point, San Joaquin, Lawigan, Iloilo Province, Philippines, 0-7 m, J. T. Williams et al., 25 Sep. 1995; USNM 376179, 3 males and 6 females, 26-48 mm SL, 09°03’N, 122°59’E, Negros, Philippines; USNM 376181 ’ 3 males, 37-44 mm SL, 10°55’05”N, 121°02’03”E, Putic Island, Palawan, Philippines, 0-4.6 m, V. Springer et al. t 22 May 1978; USNM 376182,2 males, 34 and 44 mm SL,’ 09°06’30”N, 122°55’24”E, near Giligaon, north of Maloh, Negros. Philippines, 0-2 m, L. W. Knapp et al., 26 April 1979; USNM 376186, 2 males, 32 and 34 mm SL and ] female, 41 mm SL, 08°51’24”N, 123°24’36”E, Solino Island, Zamboanga, Mindanao, Philippines, 0-5 m, L. W. Knapp et al., 3 May 1979. Diagnosis. Vertebrae 1 1-12+30-32=41-43, dorsal fin rays 72-82, anal fin rays 56-64; eyes moderately large (2.1—2.5% SL); body slender, snout without cirri; scales only on cheeks, no scales above opercular spine; upper preopercularpore present; outer pseudoclasper moderately broadbascd, long, sometimes extending beyond hood in resting position; inner pseudoclasper half as long as outer pseudoclasper, anterior lobe very long, twice as long as posterior lobe; otolith length to height 2.1-2.2, with shallow Fig. 25. Alionematichthys suluensis sp. nov.. A, lateral view of head, holotype; B, ventral view of head, holotype; C, view of left pseudoclaspers from inside, holotype; D, inclined lateral view of male copulatory organ, holotype; E, median view of right otolith, AMS 1.40149-028, female, 40 mm SL; F, ventral view of right otolith, AMS 1.40149-028, female, 40 mm SL. 122 Dinematichthyine fishes of the Indo-west Pacific, Part IV Table 11. Meristic and morphometric characters of Alionematichthys suluensis sp. nov. Holotype AMS I. 40149- 028 Holotype + 56 paratypes Mean (range) n Standard length in mm 40 36.7(13-48) 57 Meristic characters Dorsal fin rays 80 77.5 (72-82) 28 Caudal fin rays 16 15.8(15-16) 21 Anal fin rays 63 60.2 (56-64) 28 Pectoral fin rays 20 20.2(19-22) 12 Precaudal vertebrae 11 11.5(11-12) 28 Caudal vertebrae 32 30.7 (30-32) 28 Total vertebrae 43 42.1 (41-43) 28 Rakers on anterior gill arch 14 14.5(12-16) 11 Pseudobranchial filaments 2 2 11 D/V 6 6.1 (6-7) 28 D/A 23 22.3 (20-25) 28 V/A 13 13.6(13-15) 28 Morphometric characters in % of SL Head length 24.4 25.2 (23.9-27.0) 11 Head width 12.2 11.9(10.2-13.6) 11 Head height 14.5 15.1 (14.3-16.6) 11 Snout length 5.7 5.5 (4.9-6.1) 11 Upper jaw length 12.3 13.0(12.3-13.9) 11 Diameter of pigmented eye 2.2 2.2 (2.1-2.5) 11 Diameter of pupil 1.2 14(1.2-1.7) 11 Interorbital width 6.3 6.6 (6.0-7.4) 11 Posterior maxilla height 4.0 4.2 (3.7-4.7) 11 Postorbilal length 16.9 18.0 (16.9-18.9) 11 Preanal length 46.0 46.5 (43.6-48.6) 11 Predorsal length ? 31.0(29.4-32.5) 10 Body depth at origin of anal fin 15.7 15.8(13.5-18.4) 11 Pectoral fin length 12.1 12.9(10.8-14.5) 11 Pectoral fin base height 4.4 5.3 (44-5.8) 11 Ventral fin length 25.6 24.0(21.2-25.9) 11 Base ventral fin - anal fin origin 28.6 28.4(26.0-31.6) 11 dorsal rim and postdorsal angle, otolith length to sulcus length 1.7-1.8, ostium length to cauda length 3.3-3.7. Description (Figs 24, 25).Thc principal meristic and morphometric characters are shown in Table 11. One of the smallest species in the genus; mature at about 30 mm SL, maximum size about 50 mm SL. Body slender, with moderately pointed head profile. No cirri on snout. Eye size 2.1-2.5 (2.2)% SL. Head with scale patch on cheek containing 5-7 (5) vertical rows of scales on upper part and 3 vertical rows on lower part, no scales on opercle. Horizontal diameter of scales on body about 2.0% SL, in 24 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded with little knob. Anterior nostril positioned high, one-third distance from upper lip to anterior margin of eye. Posterior nostril small, about one- sixth the size of eye. Opercular spine with free tip, pointed. Anterior gill arch with 12-16 (14) rakers, 3 elongate in a row. Pseudobranchial filaments 2. Head sensory ’ pores (Fig. 25A, B). Supraorbital pores 3. Infraorbital pores 6 (3 anterior and 3 posterior), three posterior pores about half the size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior): first anterior pore large, tubular, with single cirrus anteriorly. Preopercular pores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view; third tubular; upper preopercular pore present. [See description of Alionematichthys ceylonensis for position of pores.] Dentition (of holotypc). Premaxilla with 4 outer rows of granular teeth and 1 inner row of larger teeth anteriorly. Anteriormost teeth in inner row up to 3/4 diameter of pupil. Vomer horseshoe-shaped, with 1 outer row of small teeth and 1 inner row of larger teeth up to 1/3 diameter of pupil. Palatine with an outer row of small teeth and an inner row of larger teeth up to 1/3 diameter of pupil. Dentary with 3 outer rows of granular teeth and 1 inner row of larger teeth anteriorly, merging into 1 row of larger teeth posteriorly. Large dentary teeth up to about 3/4 diameter of pupil. Otolith (Fig. 25E, F). Moderately elongate in shape, length to height 2.1-2.2 (30-49 mm SL) and moderately thick (otolith height to otolith thickness about 1.5). Anterior tip slightly pointed, posterior tip pointed, expanded and irregularly ornamented. Dorsal rim shallow, with notch above anterior tip and behind marked postdorsal angle. Inner face convex, outer face slightly concave to almost flat, both smooth. Otolith length to sulcus length 1.7-1.8. Sulcus slightly supramedially positioned and slightly inclined, with separated colliculi and marked notch at ventral rim of sulcus at joint of ostium with cauda. Ostium length to cauda length 3.3—3.7. Ventral furrow very close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyses present from vertebrae 6 to 11 (12). Pleural ribs on vertebrae 2 to 11 (12). First anal fin pterygiophore markedly longer than subsequent pterygiophore, reaching tip of last precaudal parapophysis in males. Male copulatory organ (Fig. 25C, D). Two pairs of pseudoclaspers, outer sometimes extending beyond hood in resting position. Outer pseudoclasper moderately broad based with thick supporter bent backwards at tip; inner pseudoclasper about half the length, thin, with very long anterior inclined lobe about twice as long as small postci ioi lobe. Isthmus narrow; penis curved, slightly shorter than outer pseudoclaspers. Colour. Live colour unknown. Preserved colour uniformly brown. Comparison. Alionematichthys suluensis belongs to the species group with an upper preopercular pore, large eyes and no scales above the opercular spine as A. piger and A. samoaensis. It differs from both in the lack of cini on the snout (except for a single cirrus associated with the 123 W. Schwarzhans and P. R. Moller Fig. 26. Alionematichthys winterbottomi sp. nov., ROM 47597, holotype, male, 54 mm SL. Fig. 27. Alionematichthys winterbottomi sp. nov.. A, lateral view of head, ROM 47597, holotype, male, 54 mm SL; B, ventral view of head, holotype; C, ventral view of male copulatory organ, ROM 82975, paratype, 58 mm SL; I), inclined lateral view of male copulatory organ, ROM 82975, paratype, 58 mm SL; E, view of left pseudoclaspers from inside, ROM 82975, paratype, 65 mm SL; F, view of left pseudoclaspers from inside, holotype; G, median view of right otolith, CAS 227284, male, 76 mm SL. 124 Dinematichthyine fishes of the Indo-west Pacific, Part IV first anterior mandibular pore) and the very peculiar shape of the inner pseudoclasper with its long anterior thorn. Distribution. Alionematichthys suluensis appears to be restricted in distribution to the Philippines between 13° and 8° N, mainly in the Sulu Sea, where it co-occurs with A. piger, A. plicatosurculus and A. riukiuensis (Fig. 16). Biology. A 39 mm SL female (USNM 376179), contains 325 embryos, length 3.7 mm TL. Etymology. Named after the type locality, the Sulu Sea of the Philippines. It is an adjective. Alionematichthys winterhottomi sp. nov. (Figs 16, 26, 27, Tables 1, 12) Material examined. (217 specimens, 10-88 mm SL). HOLOTYPE - ROM 47597, male, 80 mm SL, 18°45 , 52”S, 178°31’ 13 ’’ E, shallow reef off first black rock 300 m S of University of the South Pacific Research Station, Dravuni Island, Fiji, coll. R. Winterbottom, A. R. Emery, F. Emery and R. McKinnon, 20 March 1983. PARATYPES - CAS 222530, 2 males, 56 and 57 mm SL and 1 female, 42 mm SL, Vatumatau Bay, Fiji, D. W. Greenfield et al., 12 Nov. 2002; CAS 227291, 1 male, 65 mm SL and 1 female, 57 mm SL, Viti Levu north, Fiji, D. W. Greenfield et al., 31 March 2002; CAS 222532, 1 male, 48 mm SL, 1 female, 49 mm SL and 1 juvenile, 29 mm SL, Fiji, D. W. Greenfield et al., 14 Nov. 2002; CAS 222540, 1 female, 52 mm SL, Fiji, 3 April 2002; CAS 222544, 2 males, 42 and 48 mm SL and 2 females, 45 and 55 mm SL, Fiji, D. W. Greenfield et al., 14 March 1982; CAS 222552, 4 males, 51-70 mm SL, 2 females, 48-54 mm SL, and 2 juveniles, 22-23 mm SL, Great Sea reef, Kia Island, Fiji, D. W. Greenfield et al. 31 March 2002; CAS 222558, 1 male, 63 mm SL, and 1 juvenile, 35 mm SL, Fiji, D. W. Greenfield et al. 15 Nov. 2002; CAS 227284,7 males, 54-75 mm SL and 5 females, 41-64, Fiji, reef NE ofYaqaga Island and NW of Ovatova reef, D. W. Greenfield et al. 25 March 2002; CAS 227307, 3 males, 42-57 mm SL, 3 females, 40-59 mm SL, Bua Bay, Fiji, D. W. Greenfield et al. 24 March 2002; CAS 222586, 2 males, 42-44 mm SL and 3 females, 43, 50 and 50 mm SL, Nasau Bay, Vanua Levu, Fiji, D. W. Greenfield et al., 25 May 2003; CAS 222589, 1 female, 57 mm SL, Fiji, D. W. Greenfield et al. 25 May 2003; CAS 222590, 2 males, 51 and 65 mm SL, 1 juvenile, 25 mm SL; Fiji east shore, D. W. Greenfield et al. 15 March 2002; CAS 222592, 1 male, 46 mm SL, 2 females, 31 and 44 mm SL, Budd Reef, Fiji, D. W. Greenfield et al. 23 May 2003; CAS 227294, 1 male, 57 mm SL, Fiji, D. W. Greenfield et al., 20 May 2003; CAS 227295, 4 males, 45-57 mm SL, 12 females, 29-58 mm SL, Budd Reef, Fiji, D. W. Greenfield et al., 22 May 2003; CAS 227297, 1 male, 53 mm SL and 3 females, 35-52 mm SL, Fiji, D. W. Greenfield et al, 23 May 2003; CAS 222605,2 females, 44 and 73 mm SL, Fiji, D. W. Greenfield et al., 5 January 2003; CAS 227298, 1 female, 88 mm SL, Fiji, D. W. Greenfield et al, 28 May 1999; CAS 227299, 3 males, 55-75 mm SL and 1 female, 53 mm SL, Fiji, D. W. Greenfield et al, 29 January 2002; CAS 222608, 3 males, Table 12. Meristic and morphometric characters of Alionematichthys winterhottomi sp. nov. Holotype ROM 47597 Holotype + 198 paratypes Mean (range) n Standard length in mm 80 48.9(10-88) 199 Meristic characters Dorsal fin rays 81 82.5 (80-86) 24 Caudal fin rays 16 16 24 Anal fin rays 62 63.9 (61-65) 24 Pectoral fin rays 22 21.3 (20-22) 11 Precaudal vertebrae 12 11.0(11-12) 26 Caudal vertebrae 31 31.0(29-32) 26 Total vertebrae 43 42.0 (40-43) 26 Rakers on anterior gill arch 17 18.3(17-20) 11 Pseudobranchial filaments 2 1.9 (1-2) 11 D/V 5 5.7 (5-6) 26 D/A 22 23.1 (21-24) 26 VIA 13 13 26 Morphometric characters in % of SL Head length 27.2 27.3 (26.3-27.8) 11 Head width 15.9 15.5(14.0-17.2) 11 I lead height 16.4 16.9(15.9-18.3) 11 Snout length 5.2 6.0 (5.2-6.5) 11 Upper jaw length 14.4 14.4(13.6-15.1) 11 Diameter of pigmented eye 2.2 2.5 (2.2-3.0) 11 Diameter of pupil 1.4 1.5 (1.2-1.7) 11 Intcrorbital width 6.8 7.1 (6.2-7.9) 11 Posterior maxilla height 5.0 4.9 (4.5-5.4) 11 Postorbital length 20.4 20.0(19.5-20.6) 11 Prcanal length 47.1 46.9 (44.7-48.6) 11 Predorsal length 30.7 31.5(30.4-32.5) 11 Body depth at origin of anal fin 19.7 18.2(17.2-19.7) 11 Pectoral fin length 15.3 14.6(13.2-15.5) 11 Pectoral fin base height 6.7 6.6 (6.3-7.2) 11 Ventral fin length 22.8 23.0(21.4-24.4) 11 Base ventral fin - anal fin origin 26.3 26.7 (23.4-29.5) 11 38-44 mm SL, Fiji, D. W. Greenfield et al., 7 February 2002; CAS 222613, 3 males, 43-60 mm SL, Barrier reef of Suva, Fiji, D. W. Greenfield et al. 27 Jan. 2002; CAS 222619, 1 male, 64 mm SL, Fiji, D. W. Greenfield et al, 31 May 1999; CAS 222621, male, 46 mm SL, female, 50 mm SL, Nukaiau Island, off Suva, D. W. Greenfield et al., 28 Jan. 200?; ROM 82975, 26 males, 34-78 mm SL, 57 females, 22-75 mm SL, 16 juveniles, 10-34 mm SL, same data as holotype; ROM 82977,5 females 45-70 mm SL, 2 males, 60 and 77 mm SL, 18°46’3(TS, I78°30’28”E, off Yanu-Yanu-Sau Island, south of Dravuni, Fiji, A. R. Emery and party, 30 Mar. 1983. Additional material. USNM 377255, 15 specimens, Vuro, Fiji; USNM 378376, 4 specimens, 27-50 mm SL, Malolo, Fiji; USNM 257649, 1 male and 2 females, 32-58 mm SL, 17°ITS, 176°54’E, YasawaGroup, Viwa Island, Fiji; USNM 263659, 2 males and 2 females, 125 W. Schwarzhans and P. R. Moller 30-72 mm SL, 18°52’S, 178°30’E, Kandavu Island, Fiji; USNM 374211, 1 male, 65 mm SL, 21°20’S, 174°58’W, Eua, Tonga; USNM 366566, 9 specimens, 36-77 mm SL, 18°50’S, 178°32’E, Kandavu Island, Fiji; USNM 376196, 2 females, 39-48 mm SL, 18°56’S, 178°21’W, Lau Group, Yagasa Island, Fiji; USNM 376209, 1 male and 1 female, 45-49 mm SL, 17°06’S, 177°13’E, Yasawa Group, Naviti Island, Fiji. Diagnosis. Vertebrae ll(-12)+29-32=40-43, dorsal fin rays 80-86, anal fin rays 61-65; eyes moderately large (2.2-3.0% SL); body moderately slender, snout with many cirri; scales on cheeks, patch of scales above opercular spine with 5-9 scales; upper preopercular pore present; outer pseudoclasper broad-based, not extending beyond hood in resting position; inner pseudoclasper short, thin, free from outer pseudoclasper, bifurcate, anterior pointed lobe short, posterior lobe much expanded posteriorly; otolith length to height 2.0-2.2, with shallow dorsal rim and weak postdorsal angle, otolith length to sulcus length 1.6-1.7, ostium length to cauda length 3.CM.5. Description (Figs 26, 27).The principal meristic and morphometric characters are shown in Table 12. Mature at about 45 mm SL. Body moderately slender, with pointed head profile. Many cirri on snout. Eye size 2.2-3.0 (2.2)% SL. Head with scale patch on cheek containing 8-10 (9) vertical rows of scales on upper part and 4 vertical rows on lower part, 5-9 (9) scales on opercle above opercular spine. Horizontal diameter of scales on body in holotypc about 1.9% SL, in 30 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded. Anterior nostril positioned high, 1/2.5 distance from upper lip to anterior margin of eye. Posterior nostril small, about one-quarter the size of eye. Opercular spine with free tip, pointed. Anterior gill arch with 17-20 (17) rakers, Upper branch with one knob-like raker and 3-5 (3) plate-shaped rakers and lower branch with 13-15 (13) rakers, all plate-shaped except for knob-shaped first and third. Pseudobranchial filaments 1-2 (usually 2). Head sensory pores (Fig. 27A, B). Supraorbital pores 3. Infraorbital pores 6 (3 anterior and 3 posterior), three posterior pores about one-third size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior): first anterior pore large, tubular. Preopercular pores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view; third tubular; upper preopercular pore present. [See description of Alionematichthys ceylonensis for position of pores.] Dentition (of holotype). Premaxilla with 7 outer rows of granular teeth and 2 inner rows of larger teeth anteriorly. Anteriormost teeth in inner row up to 1/3 diameter of pupil. Vomer horseshoe-shaped, with 3 outer rows of small teeth and 1 inner row of large teeth up to 1/3 diameter of pupil. Palatine with 3 outer rows of small teeth and 1 inner row of 8 large teeth up to 1/3 diameter of pupil. Dentary with 5 outer rows of granular teeth and 1 inner row of larger teeth anteriorly, merging into 1 row of larger teeth posteriorly. Thirteen large dentary teeth, up to about 2/3 diameter of pupil. Otolith (Fig. 27G). Moderately elongate in shape, length to height 2.0-2.2 and moderately thin (otolith height to otolith thickness about 2.2-2.8). Anterior tip slightly pointed, posterior tip pointed, expanded and irregularly ornamented. Dorsal rim shallow, with notch above anterior tip and with indistinct postdorsal angle. Inner face convex, outer face slightly concave to flat, both smooth. Otolith length to sulcus length 1.6-1.7. Sulcus slightly supramedially positioned and slightly inclined, with separated colliculi and marked notch at ventral rim of sulcus at joint of ostium with cauda. Ostium length to cauda length 3.0-4.0 (4.5). Ventral furrow weak, very close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyses present from vertebrae 6 to 11 (12). Pleural ribs on vertebrae 2 to 11 (12). First anal fin ptcrygiophorc slightly to markedly longer than subsequent pterygiophore, but not reaching tip of last precaudal parapophysis. Male copulatory organ (Fig. 27C-F). Two pairs of pseudoclaspers, outer not extending beyond hood in resting position. Outer pseudoclasper broad based, nearly triangular in shape, somewhat thickened towards lip; inner pseudoclasper less half the length, thin, bifurcate, anterior pointed lobe short, anteriorly free from outer pseudoclasper halfway up from base, posterior lobe much expanded posteriorly. Isthmus narrow; penis curved, slightly longer than outer pseudoclaspers. Colour. After preservation the specimens are uniformly brownish. Comparison. Alionematichthys winterbottomi belongs to the species group with an upper preopercular pore, large eyes, many cirri on the snout and within that group to the subgroup with 5 or more scales above the opercular spine together with A. riukiuensis, from which it differs in the specific shape of the thin, free and small inner pseudoclasper with its bifurcate shape and the lack of scales below the opercular spine. Both species arc obviously is closely related to each other. Distribution. Alionematichthys winterbottomi is restricted in distribution to the Fiji and Tonga Islands, where it replaces A. riukiuensis (Fig. 16). Etymology. Named in honour of Richard Winterbottom, ROM, Toronto, Canada, in recognition of his many contributions to ichthyology and for his great support of the present revison. It is a noun in apposition. Alionematichthys sp. 1 (Figs 16, 28, Table 1) Material examined. (2 specimens, 75-87 mm SL). AMS 19108-032, 1 male, 75 mm SL, 14°40’S, 145°28’E, Lizard Island, Great Barrier Reef, Australia, 1-10 m, D. F. Hoese and party, 17 Nov. 1975; WAM R30305-036, 1 126 Dinematichthyine fishes of the Indo-west Pacific, Part IV female, 87 mmSL, 13°52’S, 126°56’E, Sir Graham Moore Islands, Western Australia, depth 0-1 m, G. R. Allen, 15 Aug. 1991. Remarks. Two specimens from northern Australia are characterised by the presence of a uniform scale patch on the opercle reaching from above the opercular spine to well below it. The single male shows an inner pseudoclasper resembling that of A. riukiuensis except for the anterior rim of the inner pseudoclasper being free of a connection to the outer pseudoclasper. All other characters, such as cirri on the snout, head pores, otolith measures, meristics and morphometries perfectly fit within the variation of A. riukiuensis. More material needs to be studied before it can be concluded if these specimens represent an extreme variant of A. riukiuensis or a new species. Alionematichthys sp. 2 (Figs 16, 29, Table 1) Material examined. (1 specimen, 83 mm SL). WAM P.25111-047, 1 female, 83 mm SL,20°28’30”S, 116°32’E, Australia, Kendrew Island, Dampier Arch., G. R. Allen and R. Steene, 3. Nov. 1974. Remarks. A single female specimen from NW Australia likely represents an undescribed species which is characterised by the absence of cirri on the snout, the absence of an upper preopercular pore (transformed to a wart) and the presence of two scales above the opercular spine. A similar combination of characters is found in A. shinoharai, from which it differs in the more stout and massive body and head and the longer sulcus on the inner face of the otolith. The scale patch on the cheek is remarkable for its uniform width with six vertical rows of scales on the upper part and five on the lower part. The otolith is characterised by a poor distinction ot ostium and cauda and the small size of the cauda as found in A. crassiceps. Geographically, this record is just outside the distribution range of other species of Alionematichthys. We have refrained from establishing a new species due to the lack of a male specimen for investigation of the pseudoclasper morphology. Fig. 28. Alionematichthys sp. I, A, lateral view of head, AMS 1. 19108-032, male, 77 mm SL; B, ventral view of head, AMS 1.19108-032, male, 77 mm SL; C, view of left pseudoclaspers from inside, AMS 1.19108-032, male, 77 mm SL; D. inclined lateral view of male copulatory organ, AMS 1.19108-032, male, 77 mm SL; E, median view of right otolith, WAM P.30305-036, female, 87 mm SL. 127 W. Schwarzhans and P. R. Moller Fig. 29. Alionematichthys sp.2, WAM P.251 1 1-047, female, 83 mm SL, A, lateral view of head, B, ventral view of head; C, median view of right otolith. Dinematichthys Bleeker, 1855 Type species: Dinematichthys iluoecoeteoides Bleeker, 1855 (type locality Kepulauan Batu, Sumatera Barat Province, Indonesia). Gender masculine. Dinematichthys Bleeker, 1855 in part. — Cohen and Nielsen 1978; Cohen and Hutchins 1982; Machida 1994; Nielsen et al. 1999. Diagnosis. Genus of Dinematichthyini with following combination of characters: anterior nostril placed high on snout at about 1/2-1/2.5 distance from lip to anterior rim of eye; head completely and continuously covered with scales on cheeks, opercle and occiput; cirri on snout; male copulatory organ with two pairs of small pseudoclaspcrs; outer pseudoclasper a simple flap, inner pseudoclasper a flap of about half size of outer, with two or three lobes; size up to about 110 mm SL; prccaudal vertebrae usually 11 (10-12), total vertebrae 41-45, dorsal fin rays 75-92, anal fin rays 59-71, V in D 2.1-2.4; upper preopercular pore absent; sulcus of otolith with ostium at least two times as long as cauda; maxilla expanded postero-ventrally. Comparison. Dinematichthys is readily recognised by its uniform and continuous head squamation including cheeks, opercle and occiput, usually visible without removal of mucus. The only other genus sharing this character is Porocephalichthys gen. nov., which does not overlap in distribution with Dinematichthys and which differs in many other characters (see Porocephalichthys gen. nov.). Even when in doubt, the combination of the absence of the upper preopercular pore, the small and simple shaped inner and outer pseudoclaspers, and the otolith with the sulcus divided into ostium and cauda, usually allow a reliable identification. Species. Dinematichthys contains two species, the widely distributed D. iluocoeteoides known throughout the Indo-west Pacific and D. trilobatus sp. nov., endemic to the isolated Christmas and Cocos Islands of the Indian Ocean. Remarks. The holotype was originally part of Bleeker’s collection, who described the first ever dinematichthyine in 1855, Dinematichthys iluocoeteoides, from a single 128 Dinematichthyine fishes of the Indo-west Pacific, Part IV specimen from Batu Island (now Kcpulauan Batu, Sumatra Barat Province), off western Sumatra, Indonesia. Bleeker’s holotype has been lost (see extensive discussion in Cohen and Nielsen (1978)). A redefinition of the genus Dinematichthys and the species D. iluocoeteoides is therefore necessary. Cohen and Nielsen (1978) discussed the nature of a specimen from Bleeker’s original collection (assigned by Gunther (1862)) kept in London (BMNH 1868.2.28.65), but concluded that it could not be the original type (though it was later referred to as the holotype by Eschmeyer (1998) (also Online Catalog of Fishes, version 19 Sept. 2008, http://www.calacademy.org/research/ichthyology/catalog/ fishcatmain.asp). It is a male specimen and, although dried and shrivelled, is here designated as neotype since it resembles Bleeker’s original description in all important aspects, and comes from the type locality and Bleeker’s original collection. Dinematichthys iluocoeteoides Bleeker, 1855 (Figs 30-32, Tables 1, 13) Dinematichthys iluocoeteoides Bleeker, 1855: 319; Gunther 1862; Wounnes and Bayne 1973: 32; Cohen and Nielsen 1978:58; Dor 1984: 59; Winterbottom etal. 1989: 14; Nielsen et al. 1999: 130. Dinematichthys indicus Machida, 1994:451; Nielsen et al. 1999: 130; Schwarzhans and Moller2005: 78. Dinematichthys randalli Machida, 1994: 456; Nielsen etal. 1999: 130. Material examined. (1207 specimens, 11-112 mm SL). NEOTYPE - BMNH 1862.2.28.65, male, 61 mm SL, Batu Island, Indonesia, purchased from P. Bleeker. PARATYPES of D. indicus : ROM 37813-2, 3 males 45, 56 and 92 mm SL, 11 females, 49-77 mm SL, Chagos Archipelago, Indian Ocean, 7°17’30”N, 7°23’56”E, depth 1-2 m, ROM 58269, 5 males, 48-81 mm SL, 3 females, 36-64 mm SL. 12°23’52”S, 43°30’00”E, Chissioua Dzaha, Comoros, Indian Ocean. Additional specimens. AMS I. 17491-006, 1 male, 56 mm SL, 1 female, 59 mm SL, 9°09’S, 159°48’E, Savo Island, Solomon Islands, 0-10 m; AMS 1. 17492-015, 2 females, 41-48 mm SL, 09°0’S, 160°06’E, Florida Island, Solomon Islands; AMS I. 20775-092, 1 male, 69 mm SL, Raine Island, Great Barrier Reef, Queensland, Australia; AMS I. 22612-010, 1 male 56 mm SL and 2 females, 48 and 55 mm SL, 15°49’S, 145°50’E, Escape Reef, Great Barrier Reef, Queensland, Australia; AMS I. 33693-008, 1 male, 66 mm SL, 11°42’45”S, 144°04’E, Great Detached Reef, Queensland, Australia, 6-19 m, AMS I. 33708-062, 1 female, 50 mm SL, 10°59’98”S, 144°01’22”E, Reef 10-418, Queensland, Australia, 2-9 m, AMS I. 37936-070, 1 female, 71 mm SL, 15°00’43”S, 168°03’30”E, Maewo Island, Vanuatu; BMNH 2000.4.19.370-382, 12 specimens, 57-103 mm SL, Abu Dhabi, Persian Gulf; BPBM 8023, 1 male, 86 mm SL, Marshall Islands; BPBM 8072,2 females, 49-65 mm SL, Palau; BPBM 40933, 2 females, 36 and 59 mm SL, Chuuk, Caroline Islands, 0-4 m; BPBM 10816, 1 female, 59 mm SL, Guadalcanal, Solomon Isis; BPBM 11358, 1 male, 107 mm SL and 1 female, 93 mm SL, Fiji; BPBM 40935, 1 male, 47 mm SL, Ryukyu Islands, Japan; BPBM 17704, 1 female, 53 mm SL, Caroline Islands; BPBM 21795, 1 female, 72 mm SL, Mauritius; BPBM 40937, 1 female, 55 mm SL, Similan Island, Thailand; BPBM 40939, 2 females, 32 and 47 mm SL, Seychelles; BPBM 28591, 1 female, 73 mm SL, Philippines; BPBM 29149, 2 females, 55 and 67 mm SL, Marshall Islands; BPBM 30491, I female, Flores, Indonesia; BPBM 30860, 1 male, 49 mm SL, Persian Gulf; BPBM 30861, 1 female, 54 mm SL, Persian Gulf; BPBM 35199, 1 female, 80 mm SL, 27° 04’N, 142°12’30”E, Chichi-jima, Ogasawara Islands, Japan; BPBM 35276,1 male, 74 mm SL, 1 juvenile. Table 13. Meristic and morphometric characters of Dinematichthys iluocoeteoides Bleeker, 1855. Neotype + n Neotype BMNH 578 non-types 1862.2.28.65 Mean (range) Standard length in mm Meristic characters 61 55.7(11-112) 579 Dorsal fin rays 84 83.1 (75-92) 137 Caudal fin rays 16 16.0(15-16) 86 Anal fin rays 69 65.3(59-71) 141 Pectoral fin rays 22 22.7 (21-24) 15 Prccaudal vertebrae 11 11.0(10-12) 151 Caudal vertebrae 31 31.5(30-34) 153 Total vertebrae 42 42.6(41-45) 153 Rakers on anterior gill arch 15 15.4(13-19) 20 Pseudobranchial filaments - 2 12 D/V 6 6.0 (5-7) 145 D/A 22 22.7(19-26) 144 V/A 13 13.1 (12-14) 145 Morphometric characters in % of SL Head length 26.2 26.4(25.8-27.1) 11 Head width 15.2 15.1 (12.2-20.9) 11 Head height 11.4 16.7(11.4-18.9) 11 Snout length 4.9 5.9 (4.9-6.5) 11 Upper jaw length 13.5 13.4(12.7-13.8) 11 Diameter of pigmented eye 2.8 2.7 (2.2-3.3) 11 Diameter of pupil 1.2 1.6 (1.2-2.0) 11 Interorbital width 4.9 6.5 (4.9-7.5) 11 Posterior maxilla height 6.2 4.6 (4.1-6.2) 11 Postorbital length 18.8 18.6(17.8-19.2) 10 Preanal length 45.7 47.2 (44.3-50.6) 11 Predorsal length 31.6 30.8(29.2-31.8) 11 Body depth at origin of anal fin 19.4 19.3(16.3-21.4) 11 Pectoral fin length 10.5 14.6(10.5-16.2) 11 Pectoral fin base height 6.1 6.0 (5.2-6.9) 11 Ventral fin length 17.7 23.5(17.7-25.6) 11 Base ventral fin - anal fin origin 27.4 27.1 (24.1-37.0) 10 129 W. Schwarzhans and P. R. Moller Fig. 30. Sample sites of O Dinematichthys iluocoeteoides Bleeker, 1855, O* holotype of D. iluocoeteoides , O* holotype of D. indicus, O* holotype of D. randalli, © D. trilobatus sp. nov. and © Porocephalichthys dasyrhynchus (Cohen and Hutchins, 1982). One symbol may represent several samples. Fig. 31. Dinematichthys iluocoeteoides Bleeker, 1855, A, fresh dead, BPBM 35594, female, 80 mm SL; B, neotype, BMNH 1862.2.28.65, male, 61 mm SL. 21 mm SL, Ogasawara Islands, Japan; BPBM 35594, 1 female, 80 mm SL, Seychelles; BPBM 38149,3 males, 66, 71 and 102 mm SL and 2 females, 56 and 65 mm SL, Tonga; CAS 14255,6 males, 41-83 mm SL, 6 females 50-102 mm SL, 7°14’32”N, 144°27’12”E, Ifalik Atoll, Micronesia, 0-4 m; CAS 35340,13 female, 34-72 mm SL, 10 males, 28-60 mm SL, Maldives; CAS 35341,19 males, 32-66 mm SL, 31 females, 29-66 mm SL, 1 juvenile, 23 mm SL, Seychelles; CAS 227304, 1 male, 70 mm SL, Madang, Papua New Guinea; CAS 65665, 4 males, 50-75 mm SL, 1 female, 55 mm SL, Madang, Papua New Guinea; CAS 65669, 1 male, 67 mm SL and 2 females, 53 and 63 mm SL, Madang, Papua New Guinea; CAS 227289. 1 female, 76 mm SL, Madang, Papua New Guinea; CAS 227281, 1 male, 59 mm SLand 1 female, 64 mm SL, Kapingamarangi Atoll, Micronesia; CAS 227290, 1 female, 46 mm SL, Viti 130 Dinematichthyine fishes of the Indo-west Pacific, Part IV Fig. 32. Dinematichthys iluocoeteoides Bleekcr, 1855. A, lateral view of head, BPBM 40937. female, 55 nun SL; B. lateral view of head, USNM 384592, male, 72 mm SL; C, dorsal view of head, USNM 384592, male, 72 mm SL; D, ventral view of head, USNM 384592, male 72 mm SL; E. lateral view of head, USNM 358342. male, 36 mm SL; F, view of left pseudoclaspers from inside, holotype; G, ventral view of male copulatory organ, BPBM 35276,74 mm SL; H, inclined lateral view of male copulatory organ, BPBM 35276, 74 mm SL; I, view of left pseudoclaspers from inside, BPBM 35276, 74 mm SL; J, view of left pseudoclaspers from inside, USNM 319902, male, 71 mm SL; K, view of left pseudoclaspers from inside, SMNS 22544, 52 mm SL; L, median view of right otolith, USNM 199680, male, 78 mm SL; M, ventral view of right otolith, USNM 199680, male, 78 mm SL; N, median view of right otolith, WAM P.31213-007, female, 52 mm SL. 131 W. Schwarzhans and P. R. Moller Levu north, Fiji; CAS 222534, 2 females, 58 and 63 mm SL, Fiji; CAS 222535, 2 females, 24 and 48 mm SL, Fiji; CAS 222539,4 females, 55-68 mm SL, Fiji; CAS 222541, female 70 mm SL, Fiji; CAS 227292, 1 female 65 mm SL, Fiji; CAS 222546, 2 males, 41 and 64 mm SL, 1 female, 43 mm SL, Fiji; CAS 222548, 2 males, 52 and 58 mm SL, 4 females, 41-70 mm SL, 4 juveniles, 15-26 mm SL, Fiji; CAS 222549, I male, 60 mm SL, 4 females, 43-65 mm SL, Fiji; CAS 222557, 1 male, 74 mm SL, Fiji; CAS 222560, 1 male, 58 mm SL, 1 female, 55 mm SL, Fiji; CAS 222562,2 males, 47 and 76 mm SL, 1 female with embryos, 50 mm SL, Fiji; CAS 222564,1 male, 49 mm SL, Fiji; CAS 222568, 1 female, 45 mm SL, Fiji; CAS 227306, 1 male, 60 mm SL, Fiji; CAS 227287, 3 males, 38, 45 and 75 mm SL, 11 female, 40-73 mm SL, 2 juveniles, 19 and 23 mm SL, Fiji; CAS 222573, 2 females, 58 and 74 mm SL, Fiji; CAS 222574, 1 female, 38 mm SL, Fiji ; CAS 222575, 1 male, 52 mm SL, 2 females, 41 and 73 mm SL, Fiji; CAS 222577, 3 females, 30-72 mm SL, 2 juveniles, 18 and 22 mm SL, Fiji; CAS 222578, male, 47 mm SL, Fiji; CAS 222579, 1 male, 53 mm SL, Fiji; CAS 222580, 1 female, 50 mm SL, Fiji; CAS 222581, 1 female, 52 mm SL, Fiji; CAS 222587, 1 male, 78 mm SL, 1 female, 32 mm SL, Fiji; CAS 222591, female, 86 mm SL, Fiji; CAS 222591, female, 86 mm SL, Fiji; CAS 227308, 2 females, 57 and 59 mm SL, Fiji; CAS 222593, 2 males, 36-52 mm SL, 3 females, 32-51 mm SL, Fiji; CAS 222597, 8 females, 35-82 mm SL, Fiji; CAS 222599, 4 males, 48-75 mm SL, 3 females, 65-75 mm SL, Fiji; CAS 222600, 2 males, 78 and 87 mm SL, 1 female, 78 mm SL, Fiji; CAS 227296, 1 male, 48 mm SL, Fiji; CAS 222602,4 females, 61-80 mm SL, Fiji; CAS 222603, 1 male, 81 mm SL, 6 females, 65-90 mm SL, Fiji; CAS 227309, 4 males, 51-83 mm SL, 2 females, 65-73 mm SL, Fiji; CAS 227300, 1 female, 82 mm SL, Fiji; CAS 222609, 2 females, 61 and 80 mm SL, 8 juveniles, 14-33 mm SL, Fiji; CAS 222611, 4 females, 35-72 mm SL, 2 males, 44-86 mm SL, 4 juveniles, 20-25 mm SL, Fiji; CAS 227310, 1 male, 53 mm SL, 1 female, 73 mm SL; CAS 222615, 2 females, 60 and 80 mm SL, Fiji; CAS 222617, 2 females, 49-75 mm SL, 2 males, 44 and 50 mm SL, Fiji; CAS 222620, 2 females, 39 and 75 mm SL, Fiji; CAS 227302, 2 females, 65 and 75 mm SL, Fiji; CAS 222624, 1 male, 75 mm SL, Fiji; CAS 222627, 1 male, 81 mm SL, Fiji; CAS 222634, 2 males, 59 and 85 mm SL, 1 juvenile, 19 mm SL; CAS 222639, 1 female, 60 mm SL, 1 male, 58 mm SL, Fiji; KAUM-I. 10660, 1 male, 42 mm SL, Okinoerabu Island, Kagoshima, Japan; MNHN 1965-0431, 2 females, 54 and 70 mm SL, 23°20’0”S, 43°31 ’0”E, Madagascar; MNHN 1977-0893, 1 female, 42 mm SL, 29°15’0”N, 34°45’0”E, Gulf of Aqaba, Israel; MNHN 1980-0243, 1 female, 90 mm SL, 22°26’0”S, 166°26’0”E, New Caledonia; MNHN 1980-0563, I male, 112 mm SL, 17°30’0”S, 167°30’0”E, Pleiades Nord, Loyalty Islands, 15-17 m; EX MNF1N 1980-0851, 1 male, 100 mm SL, 1 female, 66 mm SL, 17°30’0“S, 167°30’0”E, lie Solitaire, ofTNoumea, New Caledonia; NTM S. 13676-031,2 females. 58 and 70 mm SL, 05°10’S, 145°50’3”E, Tab Island, Madang, Papua New Guinea, 15-24 m; ROM 37811, 3 males, 50, 50 and 60 mm SL, 05°25’2r’S, 071°46’52”E, Isle du Coin, Chagos Archipelago; ROM 37812,22 males, 30-81 mm SL, 18 females, 33-68 mm SL, 3 juveniles, 20-24 mm SL, 05°25 W’S, 071 °46’00”E, Isle du Anglaise, Chagos Archipelago; ROM 42314,1 female, 42 mm SL, 1 male, 48 mm SL, 09°28’00”S, 159°49 ! 00”E, Honiara, Solomon Islands; ROM 50299, female, 42 mm SL, male 40 mm SL, 09°28’00”S, 159°42’00”E, 10 km W of Honiara, Solomon Islands; ROM 55144, 63 male, 09°12’16”S, 123°27’15”E, Tonga Point. Philippines; ROM 58267, 6 females, 32-53 mm SL, 5 males, 32-58 mm SL, 2 juveniles, 25 and 25 mm SL, 12°23’52'’S, 43°30’00”E, Point Chongochahari, Comoros; ROM 68122,4 females, 28-65 mm SL, 3 juveniles, 12-22 mm SL, 12°23’52”S, 43°30'00”E, Chissioua Dzaha, Comoros; SAIAB 35108, male, 63 mm SL, female, 54 mm SL, 22°00'N, 120°45’E, Wanlitong, Taiwan; SAIAB 53381, 2 females, 27 and 41 mm SL, 16°18’N. 119°54’E, Luzon, Bolinao, Philippine Islands; SMNS 13691, 2 females, 61 and 72 mm SL, 27°4I’30”N, 34°08’10”E, Ras Mohammed, Red Sea, 0.5-4.5 m; SMNS 17126,1 female, 28 mm, SL, 19°40’23”S, 63°25’58”E, Rodrigues Island, Mascarcne Islands, 0.3-0.9 m; SMNS 17206,4 females, 40-48 mm SL, 3 males, 39-87 mm SL, 19°40’25”S, 63°25’59’'E, Rodrigues Island, Mascarene Islands, 0-2 m; SMNS 18674, 2 males, 59-62 mm SL, 1 female, 80 mm SL, 08°22’06”S, 116°04’46”E, Lombok, Indonesia, 1.5-3.8 m; SMNS 22544, 1 male, 54 mm SL, 29°09'44”S, 34°41’34”E, Gulf of Aqaba, Red Sea, 0-3 m; TAU 11692, 2 females, 46 and 51 mm SL, 4 males, 42-56 mm SL, Seychelles; TAU 11808, I male, 70 mm SL, I female, 52 mm SL, Seychelles; TAU 11830, 3 specimens, Seychelles; UF 173089, 1 females, 81 mm SL, 1 male, 55 mm SL, Enewetak Atoll, Marshall Islands; USNM 376199, 1 female, 53 mm SL, Buru Island, Indonesia; USNM 99066, 1 female, 55 mm SL, Cagayanes, Philippines; USNM 99223, 1 female, 34 mm SL, Ragay Gulf, Philippines; USNM 99226, 1 male, 63 mm SL, Dalaganem, Philippines; USNM 133865, 1 male, 80 mm SL, Tahiti; USNM 142011, 2 females, 62-68 mm SL, Rongelap Atoll, Marshall Islands; USNM 142016, 1 male, 82 mm SL, Bikini Atoll, Marshall Islands; USNM 374223, 1 male and 1 female, 46-85 mm SL. Bikini Atoll, Marshall Islands; USNM 142019, 4 specimens, 89-108 mm SL, Bikini Atoll, Marshall Islands; USNM 374224, 1 female, 91 mm SL, Bikini Atoll, Marshall Islands; USNM 166807, 2 males and 2 females, 60-82 mm SL, Amo Atoll, Marshall Islands; USNM 376193,3 females, 52-80 mm SL, Onotoa, Gilbert Islands, Kiribati; USNM 199680,5 females, 52-65 mm SL, 5 males, 40-83 mm SL, 12°10’S, 44°23’E, Comoros; USNM 206304, 25 specimens, 34-83mm SL, Saint Anne Island, Seychelles; USNM 201252, 1 male, 42 mm SL, Ambon, Maluku Islands, Indonesia; USNM 394973,3 females, 31-71 mm SL, Tutuila, Samoa; USNM 223324,2 specimens, 22- 38 mm SL, Pohnpei, Micronesia; 132 Dinematichthyine fishes of the Indo-west Pacific, Part IV USNM 223412, 6 specimens, 37-62 mm SL, Pohnpei, Micronesia; USNM 223426, 1 female, 63 mm SL, Pohnpei, Micronesia; USNM 223508, 14 specimens, 25-81 mm SL, Pohnpei, Micronesia; USNM 223558,3 specimens, 60-80 mm SL, Pohnpei, Micronesia; USNM 224330,5 specimens, 28-75 mm SL, Pohnpei, Micronesia; USNM 224331, 5 specimens, 23-77 mm SL, Pohnpei, Micronesia; USNM 244012, 1 male, 77 mm SL, 1 female, 32mm SL, 17°45’S, 177°04’W, Malolo Island, Fiji; USNM 244013, 1 male and 4 females, 24-63 mm SL, Malolo Island, Fiji; USNM 263660, 21 specimens, 20-75 mm SL, Lemus, Kavieng, Papua New Guinea; USNM 263667, 5 specimens, 30-65 mm SL, Red Sea; USNM 263668, 1 female, 50 mm SL, Marshall Islands; USNM 263673,7 specimens, 40-86 mm SL, 16°21’S, 43°59’E, Madagascar; USNM 377198, 15 specimens, 45-83 mm SL, 01°33’S, 144°59’E, Papua New Guinea, 0-15 m; USNM 377209, 1 male and 1 female, 53-64 mm SL, 10°52’N, 120°56’E, Palawan, Philippines, 0-14 m; USNM 377191, 1 female, 58 mm SL, 08°51’N, 123°24’E, Zamboanga, Philippines; USNM 263692, 15 specimens, 51-80 mm SL, 09°10’N, I23°26’E, Negros, Philippines, 0-3 m; USNM 377200,3 males and 4 females, 38-82 mm SL, Weligama, Sri Lanka; USNM 263708, 1 specimen, 56 mm SL, Vuro Island, Fiji; USNM 263712, 27 specimens, 27-75 mm SL, 10°30’S, 44°21 ’E, Comores; USNM 263750, 12 specimens, 22-71 mm SL, Kiriwina, Papua New Guinea; USNM 263756, 18 specimens, 39-83 mm SL, Bougainville, Solomon Islands; USNM 267192, 4 specimens, 17-58 mm SL, Seychelles; USNM 394974, 4 males, 27-36 mm SL, 2 females, 42 and 43 mm SL, Mahe, Seychelles; USNM 300092, 2 males and 1 female, 44-69 mm SL, 20°24’N, 121°55’E, Batanes, Philippines; USNM 300094,1 female, 21 °07’N, 121 °56’E, Batanes, Philippines; USNM 300100, 1 female, 20°17’N, 121°50’E, Batanes, Philippines; USNM 300103, I male, 50 mm SL, 22°54’N, 121°54’E, Batanes, Philippines; USNM 374228, 1 male, 57 mm SL,04°52’N, 119°26’E, Sibutu, Philippines; USNM 319898, 1 female, 78 mm SL, 20°34’S 166°14’E, Loyalty Islands, New Caledonia, 0-5 m; USNM 319902, 2 males 50 and 71 mm SL, 5 females, 20°34’S, I66°14'E, Loyalty Islands, New Caledonia; USNM 334123, 20 specimens, 26-89 mm SL, Tongatapu, Tonga; USNM 334125, 9 specimens, 28-75 mm SL, Tongatapu, Tonga; USNM 334126, 2 specimens, 23-93, mm SL, Tongatapu, Tonga; USNM 334127, 13 specimens, Tongatapu, Tonga; USNM 374209,11 specimens, 35-79 mm SL, 21 °20’S, 174°58’W, Tonga; USNM 336510, 9 specimens, 19°16’S, 174°22’W, Tonga; USNM 338465, 4 females, 31-98 mm SL, Vava’u Group, Tonga; USNM 394977, 6 specimens, 54-67 mm SL, 18°44'31”S, 174°06’36”W, Vava’u Group, Tonga; USNM 377199, 5 specimens, 42-75 mm SL, 10°35’N, 122°08’E, Panay, Philippines; USNM 376156, 1 female, 47 mm SL, 10°28’N, 122°28’E, Guimaraes, Philippines; USNM 394983, 5 females, 43-72 mm SL, 1 male, 58 mm SL, 16°47’13”S, 168°2r36”E, Epi, Vanuatu, 1-10 m; USNM 352643, 6 specimens, 52-55 mm SL, Ryukyu Islands; USNM 355820, 1 male and I female, 40-66 mm SL, Papua New Guinea; USNM 357210,1 juvenile, 23 mm SL, 10°40’S 165°47’E, Solomon Islands; USNM 358342, 1 male, 36 mm SL, 10°16’S, 166°18’E, Solomon Islands; USNM 384592, 1 female, 63 mm SL, 1 male, 72 mm SL, Vanuatu, Efate; USNM 363675, 1 female, 79 mm SL, Vanuatu; USNM 363908, 2 females, 60-88 mm SL, 15°00’S, 168°03’E, Maewo, Vanuatu; USNM 365322, 1 male and I female, 80-102 mm SL, 08°23’S, 162°5UE, Stewart Island, Solomon Islands; USNM 365613, 4 specimens, 42-57 mm SL, 13 o 21’30”S, 176°10’10”W, Wallis Island, Wallis and Futuna Islands, 1-13 m; USNM 377236, 2 males, 41-80 mm SL, Bugabag, Papua New Guinea; USNM 365823, 4 specimens, 42-75 mm SL, Hermit Island, Papua New Guinea; USNM 365826, 7 specimens, 58-66 mm SL, Hermit Island, Papua New Guinea; USNM 365829, 1 female, 37 mm SL, Massas, Papua New Guinea; USNM 366216, 6 specimens, 51-83 mm SL, Chagos Archipelago; USNM 366227, 3 females, 25-50 mm SL, 27°16’N, 33°47’E, Egypt; USNM 366228, 1 male, 45 mm SL, 26°08’N, 34°16’E, Egypt; USNM 366229, 1 female, 68 mm SL, 27°18’N, 33°47'E, Egypt; USNM 366230, 1 male and 1 female, 40-60 mm SL, Mauritius; USNM 366231, 2 juveniles, 17-26 mm SL, 16°25’S, 59°36’E, Cargados Carajos, Mascarene Islands; USNM 366232, 3 females, 38-68 mm SL, 16°43’S, 59°35’E, Cargados Carajos, Mascarene Islands; USNM 366233, 7 specimens, 35-65 mm SL, 16°27’S, 59°36’E, Cargados Carajos, Mascarene Islands; USNM 366234, 9 males and 24 females, 31-70 mm SL, 10°I9’S, 56°35’E, Agalega, Mauritius; USNM 366457,2 specimens, Agalega, Mauritius; USNM 366459, 8 males and 4 females, 32-86 mm SL, 07°15’S, 72°22’E, Chagos Archipelago; USNM 366460, I male, 54 mm SL, 07°15’S, 72°22’E, Chagos Archipelago; USNM 366461,5 males and 8 females, 42-84 mm SL, 07°15’S, 72°22’E, Chagos Archipelago; USNM 366462, 1 female, 45 mm SL, 13°49’S, 72°24’E, Chagos Archipelago; USNM 366463, 2 females, 57-60 mm SL, 07°14’S,72°23’E, Chagos Archipelago; USNM 366464,3 males and 2 females, 39-69 mm SL, 07°20’S, 72°27’E, Chagos Archipelago; USNM 366465,5 males and 1 female, 40-59 mm SL, Marsa Mokrakh, Egypt; USNM 366466,4 females, 27—41 mm SL, Egypt; USNM 366467, 1 female, 68 mm SL, Hikkaduwa, Sri Lanka; USNM 366468, 6 specimens, 40-82 mm SL, Hikkaduwa, Sri Lanka; USNM 366469,1 male, 72 mm SL, Hikkaduwa, Sri Lanka; USNM 366472, 33 specimens, 22-92 mm SL, Egypt; USNM 376219, 6 males and 1 female, 21-65 mm SL, 05°52’S, 112°37’E, Java, Bawaean Island; USNM 366476, 47 specimens, 16°28’S, 59°40’E, Cargados Carajos, Mascarene Islands; USNM 366477, I male, 71 mm SL, Hikkaduwa, Sri Lanka; USNM 366478, 1 male and 1 female, 71-84 mm SL, Galle, Sri Lanka; USNM 366479, 1 female, 60 mm SL, Hikkaduwa, Sri Lanka; USNM 366483, 1 female, 58 mm SL. 07°14’S, 72°23’E, Chagos Archipelago; USNM 366484, 2 females, 42-57 mm SL, 07°20’S, 72°27’E, 133 W. Schwarzhans and P. R. Muller Chagos Archipelago; USNM 366485, 2 males and 6 females, 44-67 mm SL, 07°13’S, 72°25’E, Chagos Archipelago; USNM 366486, 1 female, 84 mm SL, 07°14’S, 72°23’E, Chagos Archipelago; USNM 366487, 1 male and 4 females, 44-69 mm SL, 07°16’S, 72°28’E, Chagos Archipelago; USNM 366491, 1 female, 41 mm SL, 02°50’N, 95°56’E, Sumatra; USNM 366492, 3 males, 35-67 mm SL, 04°14’S, 152°10’E, Bismarck Archipelago, Papua New Guinea; USNM 376220, 1 male, 65 mm SL, 03°47’S, 128°06’E, Maluku, Indonesia; USNM 376200, 1 female, 39 mm SL, 05°52’S, 110°25’E, Karimundjawa, Java, Indonesia; USNM 366500, 1 male and 7 females, 39-75 mm SL, 05°51 ’S, 106°34’E, Pulau Seribu, Indonesia; USNM 366501, 19 specimens, 16-48 mm SL, 01°33’S, 144°59’E, Papua New Guinea; USNM 376205,5 specimens, 48-74 mm SL, Madang, Papua New Guinea; USNM 366504,3 specimens, Madang, Papua New Guinea; USNM 366506, 1 male and 3 females, 68-87 mm SL, Briwadi Island, Kiriwina Islands (Trobriand), Papua New Guinea; USNM 376162, 31 specimens, 07°15’S, 72°22’E, Chagos Archipelago; USNM 366513,1 juvenile, 29 mm SL, Buroa, Egypt; USNM 366514, 1 female, 61 mm SL, 15°32’N, 40°00’E, Ethiopia; USNM 366515, 14 specimens, 15-67 mm SL, Gulf of Aqaba, Egypt; USNM 366516, 1 female, 74 mm SL, Buroa, Egypt; USNM 366517, 2 males and 1 female, 43-71 mm SL, Sinai, Egypt; USNM 366519, 4 females, 41-58 mm SL, 16°28'S, 59°37’E, Cargados Carajos, Mascarcne Islands; USNM 366520,15 specimens, 34-67 mm SL, North Island, Agalega, Mauritius; USNM 376201, lOspecimens, 34-69 mm SL,05°10’S, 145°5UE, Papua New Guinea; USNM 366523, 1 female, 61 mm SL, Madang, Papua New Guinea; USNM 376191,1 female, 45 mm SL, Madang, Papua New Guinea; USNM 366526, 13 specimens, 39-61 mm SL, Kiriwina Islands (Trobriand), Papua New Guinea; USNM 366527, 11 specimens, 59-91 mm SL, 12°46’N, 44°59’E, Aden; USNM 366529, 3 specimens, 12°53’S, 45°16’E, Comores; USNM 366534, 4 specimens, 07°56’N, 81 °34’E, Sri Lanka; USNM 366535, 21 specimens, 26-75 mm SL, 05°24’S, 53° 13’E, Comores; USNM 366551,5 specimens, 21-41 mm SL, Israel; USNM 366552, 22 specimens, 11-98 mm SL, Gulf of Aqaba, Egypt; USNM 366553, 13 specimens, 51-72 mm SL, 15°30’N, 39°54’E, Ethiopia; USNM 366571, 1 male, 54 mm SL, Guadalcanal, Solomon Islands; USNM 366573, 1 male, 73 mm SL, New Georgia, Solomon Islands, USNM 366590, 20 specimens, 30-80 mm SL, 16°36’S, 59°31’E, Cargados Carajos, Mascarene Islands; USNM 366591, 31 specimesn, 28-75 mm SL, 16°26’S, 59°36’E, Cargados Carajos, Mascarene Islands; USNM 366593,30 specimens, 28-74 mm SL, 16°25’S, 59°36’E, Cargados Carajos, Mascarene Islands; USNM 366594, 5 specimens, 35.81 mm SL, 16°15’S, 59°33’E, Cargados Carajos, Mascarene Islands; USNM 374229, 1 female, 51 mm SL, 09°13’N, 123°28’E, Negros, Philippines; USNM 374225, 1 male, 66 mm SL, 09°31’N, 123°40’E, Philippines; USNM 366679, 1 female, 86 mm SL, Kwajalein Atoll, Marshall Islands; USNM 366692, 2 females, 68-72 mm SL, 21°55’N, 120°50’E, Taiwan; USNM 366693, 2 specimens, 48-64 mm SL, 21°55’N, 120°48’E, Taiwan; USNM 366697, 20 specimens, 37-74 mm SL, Red Sea; USNM 366701, 1 specimen. 47 mm SL, 09°03’N, 123°07’E, Philippines; USNM 366702, 4 specimens, 27-43 mm SL, Bararin, Philippines; USNM 366704, 1 male, 49 mm SL, Tagauyan, Philippines; USNM 366720, 1 male, 55 mm SL, Balabac, Philippines; USNM 366721. I female, 44 mm SL, 11°42’S, 167°51 ’E, Vanikolo Islands, Santa Cruz Islands, Solomon Islands,; USNM 366839, 7 specimens, 59-80 mm SL, 18°08’S, 178°24’E, Fiji; USNM 366840, 1 male, 78 mm SL, 19°09’S, 179°45’E, Fiji; USNM 366842,6 specimens, 45-77 mm SL, 18°58’S, 179°52’W, Fiji; USNM 366844, 4 specimens, 55-76 mm SL, 21°38’S, 178°45'W, Fiji; USNM 366845, 6 males and 1 female, 44-81 mm SL, 18°55’S, 178°33’W, Fiji; USNM 366846, 6 specimens, 45-75 mm SL, 18°57'S, 178° 17’E, Fiji; USNM 376206, 6 specimens, 40-111 mm SL, 17°06’S, 177° 13’E, Fiji; USNM 374208, 3 females, 32-60 mm SL, 13°23’S, 176° 11 ’ W, Wallis Island, Wallis and Futuna Islands; USNM 263698, 1 female, 57 mm SL, 01°33’S, 144°59’E, Papua New Guinea; WAM P.27469-005, 5 males, 65-70 mm SL, 15°50’S, 145°50’E, Escape Reef, Queensland, Australia; WAM P.27663-003, 2 males 65 and 65 mm SL, 1 female, 35 mm SL, NW Australia; WAM P.27825-031,2 males, 35 and 42 mm SL, 2 females, 39 and 44 mm SL, Bismarck Sea, Papua New Guinea; WAM P.28191-005, 1 male, 52 mm SL, Coral Sea; WAM P.29627-048, 2 females, Coral Sea; WAM P.29642-010, 2 females, 42-53 mm SL, Coral Sea; WAM P.30618-002, 1 male and 3 females, 67-79 mm SL, Papua New Guinea; WAM P.30633-023, 1 male, 58 mm SL, Papua New Guinea; WAM P.30719-036,1 female, 42 mm SL, Flores Sea, Indonesia; WAM P.30791-002, 1 female, 50 mm SL, New Caledonia; WAM P.30844-050, 3 specimens, Timor Sea; WAM P.31213-058. 1 female, 52 mm SL, Papua New Guinea; WAM P.31437-043, 1 male, 60 mm SL, Timor Sea; YCM-P 34028, 1 male, 54 mm SL, Kakeroma Island, northern Ryukyu Islands, Japan; ZMUC P 771654-56, 2 males, 66 and 85 mm SL and 1 female, 77 mm SL, Seychelles; ZMUC P 771657-58, 1 male, 61 mm SL, and 1 female, 56 mm SL, Mauritius. Diagnosis. Vertebrae 10-12+30-34=41-45, dorsal fin rays 75-92, anal fin rays 59-71; eyes large (2.2-3.3% SL); body moderately slender, snout with many cirri; head continuously covered with scales, ventrally terminating behind maxilla; upper preopercular pore absent; outer pseudoclasper broad-based, small, not extending beyond hood in resting position; inner pseudoclasper small, with two lobes; otolith length to height 2.1-2.3 with regularly curved dorsal rim, otolith length to sulcus length 1.6—1.7, ostium length to cauda length 3.8^4.8. Description (Figs 31, 32).The principal meristic and morphometric characters are shown in Table 13. Mature at about 45 mm SL. Body slender, with moderately pointed head profile. Many cirri on snout. Eye size 2.2-3.3 (2.8)% 134 Dinematichthyine fishes of the Indo-west Pacific, Part IV SL. Head continuously covered with scales on cheeks, opercle and occiput, ventrally terminating behind maxilla. Horizontal diameter of scales on body of 85 mm SL male (ZMUC P771656) about 2.0% SL, in about 28 horizontal rows. Maxillary ending far behind eye, dorsal margin of maxillary covered by upper lip dermal lobe, posterior end expanded with little knob. Anterior nostril positioned high, 1/2—1/2.5 distance from upper lip to anterior margin of eye. Posterior nostril small, about one-fifth to one-sixth the size of eye. Opercular spine with free tip, pointed. Anterior gill arch with 13-19 (15) rakers, 3 elongate in a row. Pseudobranchial filaments 2. Head sensory pores (Fig. 32A-E). Supraorbital pores 2-3. Infraorbital pores 6 (3 anterior and 3 posterior), three posterior pores about one-third the size of three anterior pores. Mandibular pores 6 (3 anterior and 3 posterior): first anterior pore large, tubular, with cirrus anteriorly. Preopercular pores: 3 lower, first and second with joined opening, covered by dermal flap in lateral view, third tubular; no upper preopercular pore. [See description of Alionematichthys ceylonensis for position of pores.] Dentition (of holotype). Premaxilla with 7 outer rows of granular teeth and 2 inner rows of larger teeth. Anteriormost teeth in inner row up to 3/4 diameter of pupil. Vomer horseshoe-shaped, with 3 outer rows of small teeth and 1 inner row of larger teeth up to 2/3 diameter of pupil. Palatine with 2 outer rows of small teeth and 1 inner row of larger teeth up to 1 /2 diameter of pupil. Dentary with 5 outer rows of granular teeth and 1 inner row of larger teeth, up to about size of diameter of pupil. Otolith (Fig. 32L-N). Elongate in shape, length to height 2.1-2.3 and moderately thin (otolith height to otolith thickness about 2.5). Anterior tip rounded, posterior tip slightly pointed, expanded. Dorsal rim shallow, gently curved. Inner face convex, outer face slightly concave to flat, both smooth. Otolith length to sulcus length 1.6-1.7. Sulcus medianly positioned, slightly inclined, with separated colliculi and small notch at ventral rim of sulcus at joint of ostium with cauda. Ostium length to cauda length 3.8^1.8. Ventral furrow weak, moderately close to ventral rim of otolith. Axial skeleton. Neural spine of vertebra 4 inclined and 5-8 depressed. Parapophyses present from vertebrae 6 to 11 (10-12). Pleural ribs on vertebrae 2 to 11 (10-12). First anal fin pterygiophore not longer than subsequent pterygiophore. Male copulatory organ (Fig. 32G-K). Two pairs of small pseudoclaspcrs, outer not extending beyond hood in resting position. Outer pseudoclasper broad based, simple flap-shaped; inner pseudoclasper about half length of outer pseudoclasper, with two lobes. Isthmus moderately narrow; penis curved, slightly longer than outer pseudoclaspers. Colour. Live colour reported as to yellow or bright to red (Fig. 31 A). Preserved colour mostly dark brown. Variability and ontogeny. Despite the wide distribution of D. iluocoeteoides there are few, if any, regional variations to be observed. For instance, in life, colouration of specimens from the Western Indian Ocean is usually described as bright red while a record from the most north¬ eastern location at the Japanese Bonin Islands was reported as yellow (personal communication by J. E. Randall). Another character showing some variation is the amount of cirri on the snout, which appears to be denser in western Indian Ocean specimens than in western Pacific specimens. Specimens from furthest to the southwest (Vanuatu, Fiji, Tonga) are special for their generally somewhat higher meristic counts, both in vertebrae (43-45) as well as dorsal fin rays (84-92). Ontogenetic changes are most strongly expressed in the squamation of the head, which is continuous on cheeks, opercle and occiput in adults and subadults from about 40 mm SL. Smallerjuvenile and larval specimens show some interruption of the scale coverage on the head, particularly between cheeks and opercle, while already complete between cheeks and occiput (Fig. 32E). The density of cirri on the snout also increases with growth, i.e. is absent in small specimens below 40 mm SL. Remarks. Machida (1994) described D. indicus from the Chagos Archipelago and Comores Islands, i.e. from the central and western Indian Ocean and D. randalli from Kosrae, Micronesia. He distinguished D. indicus from D. randalli by the presence of cirri on the top and sides of the head behind the eye (vs no cirri on head) and by lower scale row counts (85-100 vs 107-113). In our review of more than 1000 specimens from an almost continuous geographical range, the main diagnostic character, i.e. presence or absence of cirri on the head could not be verified. It is true that western Pacific specimens appear to have generally fewer cirri, but we regard this character as too weak and too irregularly distributed to be of diagnostic value. Similarly, the scale row count is considered by us as a not sufficiently reliable character. Distinction of these two species from D. iluocoeteoides was based on comparative data from Bleekcr (1855) and Sedor and Cohen (1987), with D. iluocoeteoides and was described as having larger eyes (a little over 5 in head length / eye diameter vs 7.8-9.4) and having a sheathed maxillary (vs unsheathed). Again, in our evaluation the eye size differences are well within the variations observed, while we failed to recognise a difference of sheathed versus unsheathed maxillary in any of the specimens other than possibly caused by preservation. In consequence, all three nominal species arc here regarded as representing a single valid species for which D. iluocoeteoides is the senior synonym. Comparison. Dinematichthys iluocoeteoides is closely related to D. trilohatus sp. nov. (see below for differentiation of the two species). The head squamation continuing on the cheeks, opercle and occiput, which usually is easily visible without the removal of mucus, distinguishes D. iluocoeteoides readily from any other co¬ occurring dinematichthyine species. Confusion may only be possible in the case of juveniles, which show some gaps 135 W. Schwarzhans and P. R. M 0 ller in head squamation between the cheeks and the opercle (Fig. 32E). Also, the absence of the upper preopercular pore and connection of the cheek squamation with the occiput squamation still allow a secure identification in most instances. Distribution. Dinematichthys iluocoeteoides is the most widely distributed species of all Dinematichthyini. ranging from the western Indian Ocean (Red Sea, Arabian Gulf, East Africa) to the Ryukyu and Ogasawara Islands in the north-western Pacific and Marshall and Kiribati Islands, Fiji, Tonga and Samoa in the central and south¬ western Pacific (Fig. 30). The reason for its unusually wide distribution is unknown, but due to the low degree of regional variation (except for the sympatric D. trilobatus sp. nov. of the Christmas and Cocos Islands; see below) is likely to be of recent origin. Despite this wide distribution, D. iluocoeteoides usually does not occur in large numbers at most sampled locations, except for some areas in the western and central Indian Ocean, e.g. the Comores or the Chagos Archipelago. Dinematichthys iluocoeteoides is also notable for its strict association with true reef environments and particularly so with outer reefs and reef cores. Dinematichthys trilobatus sp. nov. (Figs 30, 33, 34, Tables 1, 14) Material examined. (16 specimens, 37-58 mm SL). HOLOTYPE - WAM P.29927-005, male, 46 mm SL, 12°05’S, 096°53’E, Direction Island, Cocos-Keeling Islands, 0.1-2.0 m, G. R. Allen, 24 Feb. 1989. PARATYPES - ANSP 162877, 49 mm SL, I2°06’30”S, 096°49’35”E; Turk Reef, Cocos-Keeling Islands, 45-49 m, W.F. Smith- Vaniz and P. L. Colin, 8 March 1974; WAM P.26106-012, male, 40 mm SL, 7 females, 45-57 mm SL, 10°29’S, 105°40'E, Rhonda Beach, Christmas Island, 10-12 m, G. R. Allen and R. Steene, 2 June 1978; WAM P.29002-021, males, 42, and 50