Aspects of Spawning Behaviour in Five Gobiids of the Genus Coryphop- terus (Pisces: Gobiidae) in the Caribbean Sea

Sand dwelling species of Coryphopterus live in shared habitats and were reported to have a similar ecology. Species lineages within the western Atlantic Coryphopterus have recently been reconciled; fish which according to the common keys had formerly been identified as Coryphopterus glaucofraenum could in fact be C. glaucofraenum, C. tortugae or C. venezuelae and can genetically be clearly separated into different clades. Based on the new taxonomic key, ecological data for five Coryphopterus species are provided; size-sex distribution, number of eggs deposited, inter spawning intervals and courtship behaviour of Coryphopterus dicrus Böhlke and Robins, Coryphopterus eidolon Böhlke and Robins, Coryphopterus thrix Böhlke and Robins, Coryphopterus tortugae Jordan, and Coryphopterus venezuelae Cervigón, were examined in Curacao, Netherland Antilles. For the first time ecological data on C. tortugae and C. venezuelae are provided, which had been previously questioned as distinct species from Coryphopterus glaucofraenum. In all species males reached a greater total length than females. Clutch sizes varied between 423 – 5872 eggs and interspawning intervals were between 5 – 14 days; no preferences for spawning at a particular lunar phase was found. Gobiids have evolved many different ways of reproduction, for example monogamy, hermaphroditism, sneaking into other nests or bi-directional sex change [1-11]. Number of eggs and fecundity varies widely among and within species, ranging from less than 100 eggs in the small tropical marine goby Eviota lacrimae to over 500 000 eggs in the large tropical freshwater goby Awaous guamensis, most of these observations have been based on ovary dissections with only a few studies conducted in the field [12, 13]. The advantage of experiments conducted in aquaria is that clutch sizes can be documented from a single female, as opposed to natural conditions where the clutch may be produced by several females and be in different stages of development [14]. Coryphopterus species have been reported to have a sequential (protogynous) hermaphroditic life history [12]. Detailed studies in the genus Coryphopterus have suggested that most species in that genus, including C. eidolon, C. dicrus and C. thrix are protogynous hermaphrodites, no data on reproduction have been available for C. tortugae or C. venezuelae [5, 6, 13]. However, some of the previously recorded data might not be accurate as the identification of western Atlantic Coryphopterus species has historically been difficult because of morphological similarity among some species and uncertainty regarding the number of valid species. Baldwin et al. [15] have recently reconciled western Atlantic Coryphopterus species with genetic lineages to determine the number of valid species and provided a revised taxonomic key. *Address correspondence to these authors at the Department of Organismic Biology, University of Salzburg, Hellbrunnerstr. 34, 5020 Salzburg, Austria; Tel: +43 662 8044 5619; Fax: +43 662 8044 5698; E-mail: [email protected] The five investigated species in this study are of similar adult size, and are found in coral reef environments; C. eidolon, C. glaucofraenum, C. tortugae and C. venezuelae live associated with sand patches in between coral outcrops, whereas C. thrix and C. dicrus are more associated with algae covered coral rocks at the edge of the sandy areas. The aim of this study was to investigate their strategies in reproduction, describe in more detail their ecology, and contribute to the determination of specific factors causing variations within these closely related species. MATERIALS AND METHODOLOGY All studies were performed between April and November 2005, in Curacao, Netherland Antilles. The sample location was the reef south of the Curacao Sea Aquarium (GPS coordinates + 12.08371 68.89645). Fish were collected with Quinaldine (2% in alcohol) and hand nets. Species were identified according to Baldwin et al. [15]. In total 198 fish of different sizes of the five species were collected; 36 C. dicrus; 56 C. eidolon, 38 C. thrix, 7 C. tortugae, and 67 C. venezuelae. The fish were measured (total length, TL) with a dial calliper (Draper 4817P) to the nearest 0.1 cm and sex determined by examining genital papillae using a binocular microscope [3, 4, 16]. The fish were placed into aquaria (200 litres) in pairs for reproductive studies. The tanks were supplied with a continuous flow of fresh sea water and kept at an ambient temperature between 28° and 32°C. They were exposed to natural light on the outside of the Curacao Sea Aquarium. The bottom of the aquarium was filled with sand and segments of polyvinyl chloride (PVC) pipes between 3 and 5 cm in diameter and 8 to 10 cm in length, which provided shelter and breeding substrate for the fish. Five pairs of Coryphopterus eidolon, four of C. thrix and three of C. Spawning Behaviour in Coryphopterus spp. The Open Fish Science Journal, 2009, Volume 2 51 dicrus were used in the study, as well as nine pairs of Coryphopterus cf glaucofraenum; either C. glaucofraenum, C. tortugae or C. venezuelae (see results for species identification problem). Females used in the spawning experiments were 3.0 4.4 cm in total length. Spawning was observed in all species. Each PVC pipe was lined with an acetate sheet to allow easy removal of the eggs and clutch size determination [14]. The fish were fed live brine shrimps three times a day and aquaria checked for eggs. After the eggs were laid, they were removed from the PVC pipe, on the acetate sheet, photographed with a digital camera (Olympus C-5060) and counted on the photograph. Statistical analyses were performed with SPSS 16.0 (SPSS Inc., Chicago). Paired t-tests were used to determine significant differences in number of eggs deposited by species and between body sizes of males and females. For differences in body size of males and females the p value was considered significant if less than 0.05, the Bonferroni correction was used for correcting the values of multiple paired t-tests for the number of eggs between the four species and therefore results considered significant at p-values less than 0.0125.