BULLETIN OF MARINE SCIENCE, 65(3): 745–754, 1999 DOES EARLY DEVELOPMENT OF THE CHILEAN TUNICATE PYURA PRAEPUTIALIS (HELLER, 1878) EXPLAIN THE RESTRICTED DISTRIBUTION OF THE SPECIES? Marcela Clarke, Verónica Ortiz and Juan Carlos Castilla ABSTRACT Relationships between the rate of larval development and dispersal of a species can have a large influence on observed patterns of its distribution and abundance. Distribu- tion patterns of the dense belt-forming rocky intertidal ascidians Pyura praeputialis and P. stolonifera (Urochordata: Ascidiacea: Pyuridae) show drastic differences. P. stolonifera extends along the coast of South Africa for 1000s of kilometers and P. praeputialis shows a similar range in Australia (New South Wales, Victoria and Tasmania). On the contrary, in Chile P. praeputialis is characterized by a population which occurs along less than 70 km of coastline in and around the Bay of Antofagasta. The larval and development char- acteristics of the South African and Australian species have been studied, while those of the population from Chile has not been described. This study describes the early embry- ology and larval development of the Chilean species and shows that the development times and the free-swimming larval stage of P. praeputialis in Chile and Australia are similar. Therefore, it is suggested that the differences in the extension of their range of distribution in Australia and Chile cannot be accounted for due to differences in the de- velopmental biology, but it may be due to a recent invasion of the species to Chile. Com- parison between the development biology of P. praeputialis and P. chilensis are also made. Geographic patterns of the distribution of species result from the combination of his- toric and regional environmental factors and local processes, such as perturbation, com- petition and predation, which determine the survival of a species in a given habitat (Ricklefs, 1987; Cornell and Lawton, 1992; Ricklefs and Schulter, 1993). On the other hand, devel- opment, larval biology and dispersal capabilities of a given species play important roles in determining distribution patterns over geographic scales; whereby understanding a species development and larval biology may provide insights into how species occupy wide-spread, restricted, or disjunct geographic patterns. Knowledge of the larval develop- ment and dispersal capabilities of organisms is also important to determine population dynamics and has implications for the management of natural populations, including the design of nature and fisheries reserves and forecasting the response of populations to regional environmental changes. Biogeographic patterns in the Southern Hemisphere have revealed the importance of historic and tectonic processes in determining the distribution of both marine and terres- trial species, indicated by the existence of common genera among the southern continents which were once part of Gondwana (Keast ,1971; Knox, 1980; Stevens, 1980). One ma- rine example is the ascidian fauna, which in the southern hemisphere is dominated by the genus Pyura (Kott, 1985). Although several of these species form dense intertidal belts along rocky shores of the southern continents (Fairweather, 1991; Paine and Suchanek, 1983; Castilla, 1998), they are still taxonomically poorly understood. In South Africa, Pyura stolonifera (Heller, 1878) extends along 1000s of kilometers from Cape Town to Natal (Day, 1974), and P. praeputialis (see Fairweather, 1991) [= P. stolonifera, see Kott, 1985] in Australia is also found along 1000s of kilometers along the coast of New South 745 746 BULLETIN OF MARINE SCIENCE, VOL. 65, NO. 3, 1999 Wales, Victoria and Tasmania (Kott, 1985). On the contrary, in South America P. praeputialis (Heller, 1878) [= P. bradleyi Van Name, 1931] shows a restricted geographic distribution, where it is found exclusively along approximately 60–70 km of coastline in and around the Bay of Antofagasta, Chile (Cerda and Oliva, 1997; Castilla 1997, 1998). Nevertheless, Van Name (1945), based on samples collected for the Peabody Museum of Yale University, reported the presence of P. bradleyi (used as a synonym of P. praeputialis, Kott, 1985) in Peru, at Zorritos (3°40'S, 80°40'W). These ascidians inhabit exposed rocky shores where they form beds of densely aggre- gated individuals permanently attached to the rocks, at mid and low intertidal fringes and extend to the shallow subtidal (Guiler, 1959; Paine and Suchanek, 1983; Fairweather, 1991; Fielding et al., 1994; Castilla, 1998). The population of P. praeputialis in Antofagasta Bay forms extensive intertidal beds up to 10–12 m wide (Castilla 1997, 1998), with indi- viduals reaching up to 35–40 cm in height. The species attains a pseudo-coloniality, which imparts upon them distinct competitive advantages (Paine and Suchanek, 1983). The embryonic development, metamorphosis and post-metamorphosis have been reported for P. praeputialis from Sydney, Australia (Anderson et al., 1975), P. stolonifera, from Cape Town, South Africa (Griffiths, 1976), and for P. chilensis from Concepción, Chile (Cea, 1969/1970). Nevertheless, P. praeputialis embryonic development, from Antofagasta, has not been described. The disjunct geographical distribution of these species make inter- continental comparative studies attractive, since the species of these locations could be characterized by different larval developments. Furthermore, it has been hypothesized that the local population of P. praeputialis at Antofagasta could represent a relic group of the fauna of Gondwanaland (Kott, 1985). The aims of this paper, that form part of a long-term study on the biology and ecology of the Chilean P. praeputialis, are to describe its embryonic, metamorphic and post-meta- morphic stages, and develop comparisons with the other southern hemisphere pyurid species, in particular with P. praeputialis from Australia. MATERIALS AND METHODS Adult specimens of P. praeputialis were collected during low tides, from May 1997 to January 1998, from two mid-intertidal rocky platforms at Quebrada El Way (23º45'S, 70º26'W), approxi- mately 10 km south from the city of Antofagasta, and at Las Conchitas (23º31'S, 70º32'W), ap- proximately 40 km north of Antofagasta (Fig. 1). Specimens were knife-detached from the base of their stolon, transported to the laboratory and kept in an open sea water circulating system at the University of Antofagasta. Temperatures ranged between 19–21ºC. Each specimen was removed from its test and opened to expose the gonads. The state of maturity was assessed by visual estimates following the scale proposed by Gutiérrez and Lay (1965) and confirmed with microscopic observations. These authors reported that in P. praeputialis (wrongly identified in the paper as P. chilensis) the gonadal lobes are packed with sperm and ova from March to August. However, unpublished results (J. C. Castilla) indicate that this period extends from March to December–January. Olive-green turgescent or distended ovaries, and mammillated while-yel- lowish testes were considered as evidence for mature stages. The ovaries and testes were removed with a scalpel to separate off pieces of egg and sperm slices, which were then lightly pressed between two microscope slides. Eggs and sperm were suspended with filtered sea water (1 µm AMETEX filter) and washed three times with filtered sea water. Eggs were dispensed into glass Petri dishes (9 cm diam) in sufficient number to provide an evenly spaced distribution on the bottom. The sperm suspension was prepared in a similar dish with filtered sea CLARKE ET AL.: EARLY DEVELOPMENT OF PYURA PRAEPUTIALIS 747 Figure 1. Map of the Mejillones Peninsula and Bay of Antofagasta, showing in black the extension of the intertidal Pyura praeputialis beds. Main sites of collection are shown. water until a white suspension was obtained. Finally, 500 cc of solution of eggs and 200 cc of sperm were mixed in a 1-L glass beaker under continuous aeration provided by a diaphragmatic pump (1.4 W), at 19–21ºC (according to Stern and Holland, 1993, ascidians resist polyspermy under these conditions). After 30 min the water was changed three times to remove excess sperm. Following the decanting of fertilized eggs, the supernatant was carefully eliminated. No mesh sieve was used in the operation. Thereafter, filtered sea water was changed every 24 h. Twenty six artificial cross fertilizations, from individual specimens (n = 52), were conducted. Because Pyura feeding occurs after metamorphosis, embryonic development stages advance with- out providing food. The development was checked several times during the day by collecting em- bryonic stages with a 10 cc pipette and examining them under a Zeiss stereomicroscope and an Olympus CH2 microscope. The drawings of the different stages were made by direct observation of the embryos and the use of photographs and video recording. To facilitate the observation of tad- pole larvae they were transferred to a 1-L glass flask (18 cm in diameter) with four Petri dishes (4 cm in diameter) placed on the bottom. Embryos were cultured up to 17 d. RESULTS Normal development was observed in 21 of the laboratory fertilizations. Five fertiliza- tions produced larvae which did not metamorphose and died before hatching. In these cases the Pyura specimens were collected after a storm in August 1997, and showed mature, although flaccid, gonads. Gamets.—Mature eggs were spherical, olive-green with reddish pigments scattered over the surface and 240–300 µm in diameter, including the inner follicle cells (Fig. 2A). The egg pigment crescent is absent. After immersion in sea water an external vitelline coat (chorion) lifts off with small inner follicle cells in the outer surface and test cells 748 BULLETIN OF MARINE SCIENCE, VOL. 65, NO. 3, 1999 Figure 2. Embryonic development of Pyura praeputialis from Antofagasta, Chile: A = Unfertilized egg. B = 2 cell stage. C, D, E = 4, 8 and 16 cell stages. F = Later gastrula. G = Early neurula. H = Later neurula. I = Early embryo. J = Later embryo. (opposite page) K = Metamorphosis: tadpole larvae showing the trunk, tail and leaf-like caudal fin. L, M = Tadpole attachment and tail resorption. N, O = Ampullae development.