ABSTRACT

Eggs Gobiid fishes lay their eggs in a dense patch of a single layer on the underside of a solid substrate, e.g. stones, rocks or shells (Miller, 1984), or on seagrass rhizomes (Scaggiante et al., 1999). Egg deposition time appears to be related to body size, varying from less than an hour in smaller species to several hours in larger ones (Mazzoldi, 1999; Mazzoldi et al., 2000; Mazzoldi and Rasotto, 2002) and eggs remain fertilizable for several hours (up to 40 hours in the grass goby, Zosterisessor ophiocephalus, Scaggiante et al., 1999). Eggs are attached by laments and hang downwards. eir shape varies inter-speci cally, being elliptical, pyriform, cylindrical, close to spherical, etc. (Miller, 1984). e function of this variability has not been investigated. Di erent shapes may, however, be based on phylogenetic relationships as well as the ecological characteristics of the nest site. Indeed, eggs of di erent shapes may be di erently packed in the egg patch, with a consequent variable water ow from fanning and possibly promoting sediment shed (Miller, 1984). Moreover, egg shape determines the surface/ volume ratio, therefore in uencing gas exchange. Hence it is rational to expect that variation in shape confers di erent performance to eggs, in terms of gas exchange and consequently developmental time, resistance to mechanical stress, and tendency to sediment deposition. Eggs may also be di erently tightly packed within the layer regardless of size. Indeed in some species eggs are laid compactly, egg density is a function of egg size and egg number is a function of nest size (Mazzoldi, 1999; Mazzoldi et al., 2002). On the contrary in the lagoon goby, Knipowitschia panizzae, spawning on anoxic muddy bottoms, nests are only partially covered by eggs that are loosely packed (Massironi et al., 2005). As hypothesized for egg shape, such di erences in deposition pattern could also be related to di erences in nesting environment, with species spawning in anoxic environments limiting egg number and egg density for oxygen provisioning (Massironi et al., 2005). Gobiid eggs are known to vary in size both within and among species (Miller, 1984; Mazzoldi et al., 2002; Tamada and Iwata, 2005). Egg diameter may vary from less than 0.4 mm to more than 2 mm for the major diameter (Katoh and Nishida, 1994; Mazzoldi et al., 2002; Massironi et al., 2005; Yamasaki and Takihara, 2007). Inter-speci c variability in egg size could be related, as is egg shape, to phylogenetic relationships and nesting environment, however these ideas have not been investigated. Variability in egg size has been reported for the marbled goby Pomatoschistus marmoratus and the common goby P. microps (Bouchereau et al., 1991; Mazzoldi et al., 2002) at the intra-speci c level, with egg size decreasing with the season and

increasing with female size (Mazzoldi et al., 2002). e in uence of seasonal progression on egg size in temperate species is widely reported (Chambers, 1997). However, maternal size is the major factor a ecting egg size in the marbled goby, and this nding is correlated with the high environmental variability of the lagoons and shallow waters where this species lives. e unpredictable variation in temperature and salinity sh experience in these waters could select for a wide reaction norm of o spring size, consequently promoting variability in egg size (Mazzoldi et al., 2002). Correlation between egg and female size has been observed also in Rhinogobius sp., and in this case has been related to di erent spawning habitats. Larger females are known to spawn in the upper reaches of rivers where starvation can represent a consistent risk for larvae. Given the relationship between egg and larval size, the production of larger eggs insures higher survival to the o spring of larger females (Tamada and Iwata, 2005). A similar relationship between egg size and spawning habitat has been documented in the Rhinogobius brunneus complex (Katoh and Nishida, 1994). e mature goby oocytes have a lament adhesion structure at the animal pole (Riehl, 1978; Miller, 1984; Giulianini and Ferrero, 2001; Giulianini et al., 2001; Herler et al., 2006; Kramer and Patzner, 2008) (Fig. 3.3.2A, B). Filaments are arranged around the micropylar pole providing the mode of attachment for these demersal eggs (Miller, 1984). Giulianini et al. (2001) have demonstrated that spermatozoa can pass through these nets of laments to reach the micropyle and fertilize the egg, even a er their attachment to the substrate. Kramer and Patzner (2008) have investigated and compared the mode of egg attachment of the adhesive laments among four Caribbean Coryphopterus species. A similar pattern of egg attachment is observed in all these species. e attaching structure forms a circular, net-like pattern around the micropylar pole (Fig. 3.3.2C, D), which divides distally into single laments approximately 2 µm in diameter (Fig. 3.3.2C, D). is net-like structure, which attaches the eggs to the substrate, is small in C. thrix and C. dicrus, and large and well-developed in C. venezuelae and C. eidolon (Kramer and Patzner, 2008). The eggshell (zona radiata or chorion) varies in thickness, surface structure and adhesive filaments. Its main function is protecting the embryo from mechanical stress. Its thickness varies, on average, from 5 to 15 μm, which is common in teleosts. e thickness and the structure of the zona radiata o en re ect adaptations to the reproductive ecology of the sh. Previous studies have come to the conclusion that sh whose eggs do not need protection from mechanical forces have a relatively thin zona radiata (Riehl and Greven, 1993; Riehl, 1995). e eggs of some gobies (e.g.