Advances in Marine Biology (eBook)

8 Reproduction

8.1 Sexual reproduction

8.1.1 Reproductive periods

Reproductive periods for hexactinellids vary depending on the species and possibly even the population. Records are scant and most stem from early collection expeditions in which scientists were focusing specifically on morphology and thus carried fixatives to preserve tissues and larvae (Schulze, 1880, 1887, 1899; Ijima, 1901, 1904; Okada, 1928). There are three recent studies (Boury‐Esnault and Vacelet, 1994; Leys and Lauzon, 1998; Leys et al., 2006).

Ijima (1901) made a special search at different seasons for reproductive cells in Euplectella marshalli but found neither developmental stages nor larvae. In subsequent collections of rossellid hexactinellids (Ijima, 1904), he still only found developmental stages and larvae in very few specimens. In Vitrollula fertilis, he found larvae in April, larvae and various developmental stages in July but only archaeocyte congeries in November, and from this presumed that the main active reproductive period is in early summer.

Okada (1928) probably carried out the most extensive survey of reproduction in a single population by collecting specimens of F. sollasii (Farreidae, Hexactinosida) monthly from the Nakanoyodomi (about 600 m deep) in the Sagami Sea. After Ijima's work, he was surprised to find reproductive specimens every month, and suggested that the breeding season was likely year‐long because of the fairly constant temperature and uniform environmental conditions that exist at that depth. Other surveys have not had the same luck as Okada, however. Although deep sea expeditions do occasionally report finding a reproductive glass sponge, a survey of tissue collected monthly from the NE Pacific rossellid species Rhabdocalyptus dawsoni failed to turn up anything but archaeocyte congeries (Leys and Lauzon, 1998). However, a study of samples of Heterochone calyx collected in October 1982 by one of us (H.M.R.) has revealed one specimen with numerous spermatocysts (accounting for approximately half of the congeries), some of which appeared to have emptied their contents, thus possibly having spawned, and many early embryos in various stages of cleavage. Of the many specimens of the NE Pacific reef‐building sponge Aphrocallistes vastus that have been collected since the early 1980s, developing embryos were only found in one specimen collected in November 1995. Despite extensive work on glass sponges in Antarctic waters, there are no reports of sexually reproductive individuals. Thus, it is likely that shallower populations (those in the NE Pacific and Antarctica) are affected by seasonality of the surface waters more than deep‐water populations. The hint from the two Heterochone and Aphrocallistes individuals is that reproductive period occurs in the autumn.

8.1.2 Gametogenesis

The bulk of our knowledge on reproduction in glass sponges stems from studies on two animals: F. sollasii (Okada, 1928) and Oopsacas minuta (Boury‐Esnault and Vacelet, 1994; Boury‐Esnault et al., 1999; Leys et al., 2006). This section of development will draw heavily from these accounts and from that information provided in Ijima's description (1904) of larvae in V. fertilis. The genera used below refer to the species used in these studies.

Gametes—both sperm and eggs—arise within archaeocyte congeries that are suspended within the trabecular reticulum between flagellated chambers (Okada, 1928; Boury‐Esnault et al., 1999). Spermatocysts are first identifiable as dense groupings of archaeocytes up to 30 μm by 23 μm in Oopsacas minuta surrounded by a thin (0.5 μm) layer of the trabecular reticulum (Figure 50A and B). In young spermatocysts, cells are larger in the centre than at the periphery of the cyst, ranging from 2.7 to 5.3 μm in diameter with a nucleus 1.6–2.6 μm and a nucleolus 0.5–1 μm (Boury‐Esnault et al., 1999). Each cell has a flagellum that coils around the cell. All spermatocytes cells are connected by plugged cytoplasmic bridges, and cells at the periphery of the cyst are connected to the surrounding trabecular envelope by plugged cytoplasmic bridges. At this point, the characteristics of free sperm remain unknown.

Oogenesis also occurs within archaeocyte congeries. The first oocyte is identifiable as a large cell (10 μm in diameter) within the congerie that has begun to accumulate yolk and lipid inclusions (Okada, 1928; Boury‐Esnault et al., 1999) (Figure 50C and D). Archaeocytes are connected to one‐another by plugged cytoplasmic bridges and are suggested to act as nurse cells providing the lipid and yolk to the developing oocyte. But at some point the oocyte presumably breaks this connection because the mature oocyte is a completely independent ovoid cell 100–120 μm in diameter in Oopsacas minuta (Boury‐Esnault et al., 1999) and 70–130 μm in F. sollasii (Okada, 1928), with large lipid inclusions (3.2‐ to 6.7‐μm diameter) at the periphery and membrane‐bound yolk inclusions (1.3–2.7 μm) and very small vacuoles more centrally. Boury‐Esnault et al. (1999) describe a 45 to 50 μm diameter nucleus with a 10 μm nucleolus within the oocyte, but although a subsequent study (Leys et al., 2006) found identifiable nuclear regions at the light microscope level, a nuclear membrane was not visible in any thin section studied by transmission electron microscopy. This may be what Okada referred to as a ‘vesicular’ nucleus (Okada, 1928, p. 3). The presumed nuclear material forms a dense osmiophilic region that occupies the centre of the cell and radiates out into the peripheral regions of the cytoplasm, not unlike the chromatin in the nucleoids of some bacteria (Robinow and Kellenberger, 1994). The surface of the oocyte has numerous short pseudopodia. Okada found most oocytes at the outer trabecular layer of F. sollasii and suggested that after fertilisation they migrate in to the inner trabecular layer to lie beside a flagellated chamber. In Oopsacas minuta, however, oocytes and developing embryos can be found throughout the body wall, from just under the dermal membrane to just under the atrial membrane, where they lie adjacent to flagellated chambers.

8.1.3 Embryogenesis

Cleavage is total and equal, but asynchronous, for the first five cycles until the embryo has approximately 32 cells (Figures 51 and 54). The embryo remains of the same size during these divisions, partitioning cytoplasm, yolk and lipid into daughter blastomeres. Early blastomeres retain all the characteristics of the oocyte, with pseudopodia extending from their surface, a dense nuclear region, large lipid inclusions at the periphery and membrane‐bound yolk inclusions more centrally. It is not until the 32‐cell stage (blastula) that a distinct nucleus can be seen in individual blastomeres (Leys et al., 2006). This feature may reflect the fact that divisions are rapid, leaving little time for re‐assembly of the nuclear membrane between cycles. If so, the appearance of nuclei at this stage—concurrent with the change to unequal cleavage—could reflect slowing of the cell cycle.