256 (2001) 51–58
www.elsevier.nl / locate / jembe
Facultatively internal fertilization and anomalous embryonic
development of a non-copulatory sculpin Hemilepidotus
gilberti Jordan and Starks (Scorpaeniformes: Cottidae)
a ,* b
Youichi Hayakawa , Hiroyuki Munehara
a
Department of Life Sciences, Graduate School of Arts and Sciences, University of Tokyo, Komaba 3-8-1, Meguro-ku, Tokyo 153-8902, Japan
b
Usujiri Fisheries Laboratory, Hokkaido University, Minami-Kayabe, Hokkaido 041-1613, Japan
Received 4 August 2000; received in revised form 6 September 2000; accepted 26 September 2000
Abstract
Fertilized residual eggs were observed in the ovaries of spent females of a non-copulatory sculpin Hemilepidotus gilberti Jordan and Starks. Fertilized eggs were present in 23 of 35 females, and approximately 38% of the total residual eggs (n5227) were fertilized. These eggs were thought to be fertilized facultatively with spermatozoa that entered the ovary through ovarian fluid
21
during spawning. The high calcium concentration (1.4260.21 mM kg ) in ovarian fluid, which is beyond the threshold concentration required for fertilization, may allow internal fertilization to occur. Embryos at various developmental stages were observed, but all were deformed and surrounded by unhardened chorions. Since no larvae were observed, all the fertilized residual eggs would have degenerated in the ovary in accordance with other unfertilized residual eggs. These observations suggest that the ovary of the oviparous fish H. gilberti is an unsuitable environment for embryos to develop, possibly because it may be unable to supply developing embryos with needed elements, such as oxygen. 2001 Elsevier Science B.V. All rights reserved.
Keywords: Ovarian fluid; Oviparous; Internal fertilization; Sculpin; Sperm motility
1. Introduction
Marine sculpins are oviparous, but some species copulate, while the others are non-copulatory (Breder and Rosen, 1966; Munehara et al., 1989). Hemilepidotus gilberti is a non-copulatory species, a fact which was confirmed by actual observation of
*Corresponding author. Tel.: 181-3-5454-6640; fax: 181-3-5454-6998. E-mail address: [email protected] (Y. Hayakawa).
spawning behavior (Hayakawa and Munehara, 1996). Although it is unlikely that internal fertilization occurs in non-copulatory fishes, residual eggs have been observed developing in the ovary of a spent female. A rivulus, Rivulus marmoratus and hermaphroditic individuals of the threespine stickleback Gasterosteus aculeatus, have self-fertilized eggs (Greenbank and Nelson, 1959; Kweon et al., 1998). However, internal fertilization in H. gilberti, which is the gonochoristic species, is thought to occur facultatively during spawning, in light of its spawning behavior and the characteristics of its gametes (Hayakawa and Munehara, 1996, 1998). Spawning of H. gilberti continues for 20–50 min, and males emit semen several times during spawning (Hayakawa and Munehara, 1996). Eggs are enclosed with viscous ovarian fluid and isolated from seawater during spawning. The ovarian fluid provides spermatozoa with a suitable extracellular environment for their movement, and eggs are fertilized in the ovarian fluid (Hayakawa and Munehara, 1998).
In the present study, the residual eggs in the ovaries of H. gilberti were examined to confirm how frequently fertilized eggs occur, and the non-organic composition and the osmolality of ovarian fluid were measured to investigate how internal fertilization occurs in this non-copulatory sculpin.
2. Materials and methods
Thirty-five post spawning females of H. gilberti, which were used for observations of residual eggs, were collected with gill nets in the coastal waters (approximately 50–90 m depth) off Usujiri (418579N, 1408589E), southern Hokkaido, Japan, from November to December 1992. Residual eggs were dissected from ovaries and immediately immersed
21
in a 1 / 6 balanced salt solution (BSS; Yamamoto, 1949; NaCl 58.44 g l , KCl 2.09 g
21 21
l , CaCl2 2.52 g l ). The embryonic stages of the eggs were observed under a stereoscopic microscope. Eggs that had developed beyond the two-cell stage were regarded as fertilized. Eggs that had an uncleaved blastodisc or opaque chorion were regarded as unfertilized. Eggs, in which development was abnormal, such as those with loose cell aggregations, were regarded as unidentified.
Females, used for measurement of inorganic electrolyte concentrations and osmolality of ovarian fluid, were collected from June to September 1992, by angling at the same location. Ovarian fluid was separated from the egg masses by centrifuging at 270g for 15 min after collection of the egg mass by squeezing ripe females. Sodium, potassium, calcium and magnesium in the ovarian fluid were analyzed using a flame spec-trophotometer (Hitachi 170-30). Osmolality was measured by the freezing point method using a Knauer semimicro-osmometer.
3. Results
3.1. Eggs
Table 1
Embryonic stages of residual eggs of H. gilberti (n (%))
Fertilized eggs Unfertilized Unidentified
Stage n eggs
2–4-cell stage 6
16–32-cell stage 11
Morula stage 6
Blastula 7
Forming of embryo 8
Forming of optic vesicle 23 Forming auditory vesicle 12 Showing heart pulse 11 Appearance of melanophores and 3 showing embryonic movement
Total number (%) 86 (37.9) 134 (59) 7 (3.1)
in paired ovaries, except for two females in which eggs were only found in one side of the ovary. As shown in Table 1, 86 developing eggs were obtained from 23 females, which amount to 37.9% of the total residual eggs (n5227). Fertilized eggs were observed from near the genital opening to the anterior region of the ovaries. The developmental stages of fertilized eggs varied from two- to four-cell stage (n56), 16–32-cell stage (11), morula stage (six), blastula stage (7), forming embryo (8), forming of optic vesicle (23), forming of auditory vesicles (12), showing heart pulse (11), to appearance of melanophores and showing embryonic movement (3). Interesting-ly, eggs of a wide range of developmental stage (e.g., embryos forming of optic vesicle, at forming auditory vesicles and at showing heart pulse) were observed to co-occur in the same ovary. In addition, most embryos were deformed (e.g., the tails of the embryos, which showed embryonic movement, were twisted or extremely bent laterally relative to the body axis) and surrounded by unhardened chorions (Fig. 1a,b). In some cases, the head developed abnormally (Fig. 1b). No larvae were observed.
3.2. Ovarian fluid
The concentrations (mean6S.D.) of sodium, potassium, calcium and magnesium in
21 21 21
ovarian fluid were, 12965.7 mM kg , 6.6860.76 mM kg , 1.4260.21 mM kg , and
21 21
1.5660.11 mM kg , respectively (n55). The osmolality was 34063.9 mOsm kg (n55).
4. Discussion
Fertilized eggs were found in the ovary of non-copulatory H. gilberti in the present observation.
exposed to seawater during egg release, because the gelatinous ovarian fluid covers the whole egg mass for a few hours until it dissolves in seawater (Hayakawa and Munehara, 1996). Spermatozoa of H. gilberti show higher motility in the ovarian fluid than in seawater, and eggs can be fertilized in the ovarian fluid extruded from ovarian cavity (Hayakawa and Munehara, 1998). In the present study, it was shown that the ovarian
21
fluid of H. gilberti has an osmolality of about 340 mOsm kg and is rich in sodium. Our previous study showed that spermatozoa of this species exhibit extremely high
21
motility in NaCl solutions at osmolalities between 300 and 400 mOsm kg (Hayakawa and Munehara, 1998). Thus, the ovarian fluid is thought to provide a suitable extracellular environment in which spermatozoa of H. gilberti can move.
It cannot be denied that the presence of fertilized eggs in ovaries may have resulted from the retracting of eggs into the ovaries upon closure of the gonopore, when the egg mass separated from the female at the conclusion of spawning. However, fertilized eggs were located not only near the gonopore but even at in the anterior part of the ovary. In general, fertilization of fishes that eggs fertilize externally occurs soon after gametes are released, as eggs lose their capacity for fertilization due to auto-activation induced by exposure to conditions which are anisotonic to body fluid (Yamamoto, 1961; Helfman et al., 1997). In this situation, it is difficult for spermatozoa to approach into the ovary. By contrast, spawning of H. gilberti continues for a long period without interruption, and males emit semen several times toward the emerging eggs. As stated above, spermatozoa of H. gilberti can swim actively in the gelatinous ovarian fluid (Hayakawa and Munehara, 1998). Since eggs are very slowly released with the ovarian fluid, the outward flow of eggs is not so strong that spermatozoa can swim against it. In addition, female H. gilberti remain on egg masses from several to 10 min after spawning (Hayakawa and Munehara, 1996). Therefore, we conclude that spermatozoa probably enter the ovary through the ovarian fluid during spawning, and consequently fertilize unspawned eggs in the ovaries.
Reason for occurrence of embryos in the ovary also may include the possibility of parthenogenesis. It is known that animal eggs sometimes begin to develop by artificial stimulation such as heat, chemical, or physical shocks without sperm (Gilbert, 1997; Piliger, 1997). In addition, gynogenesis occurs in some fishes (reviewed in Helfman et al., 1997). However, eggs of H. gilberti, as stated above, are not directly exposed to seawater during spawning, by being covered with ovarian fluid (Hayakawa and Munehara, 1996). Thus, eggs of H. gilberti are thought to be protected from stimulation causing abnormal activation in natural conditions, because ovarian fluid provides eggs with moderate conditions during spawning. On the other hand, eggs of gynogenetic fish are activated by sperm from other species, but no sperm material is incorporated (Helfman et al., 1997). Therefore, it is not thought that parthenogenesis occurs in H. gilberti.
Females of copulatory sculpins have no fertilized eggs in their ovaries, because their eggs only associate with spermatozoa in the ovary and fertilization occurs externally after spawning (Munehara et al., 1989, 1991, 1994, 1997; Koya et al., 1993). The inhibition of internal fertilization in copulating sculpins is thought to be due to a deficiency of calcium ions in the ovarian fluid, which is slightly below the threshold
21
21
mM kg in Blepsias cirrhosus; Koya et al., 1993). Fertilization is initiated by a rise in calcium concentrations of the surrounding seawater after egg deposition (Munehara et al., 1994, 1997). The mean calcium concentration in the ovarian fluid of H. gilberti was
21
1.42 mM kg , which is approximately 3–4 times greater than that of copulating sculpins. Therefore, the high calcium concentration may allow fertilization to occur in the ovarian fluid.
Fertilized eggs in the ovaries of H. gilberti developed abnormally and were surrounded by unhardened chorions. Abnormal development has also been reported in the residual eggs of the wolffish Anarhichas lupus (Pavlov, 1994). In A. lupus, males and females contact their sexual openings before spawning, and spermatozoa are released directly into the oviduct (Johannessen et al., 1993). Therefore, fertilization occurs internally. Such reproductive behavior presumably allows for the most efficient use of a limited quantity of spermatozoa (Pavlov, 1994). Eggs must be released within a few hours before cleavage begins, or the embryos develop abnormally and die (Pavlov, 1994; Pavlov and Moskness, 1994).
Eggs of teleost fishes need certain extracellular properties for hardening of the chorion and successful embryonic development to occur (Kusa, 1949a,b; Zotin, 1958; Davis, 1975; Lønning et al., 1984). In Cylopterus lumpus, the chorion does not harden in a medium lacking calcium and magnesium (Lønning et al., 1984). Zotin (1958) demon-strated from an experiment using salmonid fish eggs that extracellular calcium is necessary for activation of the hardening enzyme secreted from the cortical layer. Calcium and magnesium are also indispensable to embryonic development (Lønning et al., 1984). It is well known that a sufficient oxygen supply is necessary for the survival of embryos, and body deformities are a common response of fish embryos to an oxygen deficit (Hishida and Nakano, 1954; Keckeis et al., 1996). Among viviparous fishes (e.g., genus Sebastes), gestating fish regulate oxygen dissolved in ovarian fluid according to the growth of the embryos (Boehlert and Yoklavich, 1984; Boehlert et al., 1986, 1991). The ovary of the oviparous H. gilberti is considered to be an unsuitable environment for embryonic development, partly because it probably lacks both a supply of oxygen and the ions needed to support the survival of embryos. Deformed embryos that occur in several fishes under unfavorable conditions, such as low oxygen level, die either before or soon after hatching (Keckeis et al., 1996). Presumably, deformed embryos of H. gilberti also die either during development or soon after hatching, and degenerate in the ovary together with any other unfertilized residual eggs. Most spawned eggs are thought to develop normally by receiving sufficient oxygen and ions from seawater because the ovarian fluid completely disappears within several hours of spawning.
frequently in H. gilberti. These suggest that the copulation of marine sculpins might have evolved from the facultative insemination of non-copulatory species, depending on physiological characteristics of spermatozoa adapted for ovarian fluid. On the other hand, all the internally fertilized eggs developed abnormally. Therefore, it is possible that the deficiency of calcium ions in the ovarian fluid, which inhibits internal fertilization in copulatory marine sculpins, might have evolved after the establishment of copulatory behavior, and might have developed to preserve eggs as an alternate to viviparity.
Acknowledgements
We thank Drs. K. Shimazaki, H. Ogi and Y. Sakurai, Faculty of Fisheries, Hokkaido University, for their good suggestions, and Dr. S. Plaistow, National Institute of Environmental Studies and J. Bower, Hokkaido University for critically reading the manuscript. We are also indebted to the staff of Usujiri Fisheries Laboratory, Hokkaido University, and the Usujiri Fisheries Cooperative Society, for collecting the specimens. This study was supported in part by a Research Fellowship of Japan Society for Promotion of Science for Young Scientists. This paper is contribution number 136 from the Usujiri Fisheries Laboratory, Faculty of Fisheries, Hokkaido University. [SS]
References
Boehlert, G.W., Yoklavich, M., 1984. Reproduction, embryonic energetics and the maternal-fetal relationship in viviparous genus Sebastes (Pisces: Scorpaenidae). Biol. Bull. 167, 354–370.
Boehlert, G.W., Kusakari, M., Shimizu, M., Yamada, J., 1986. Energetics during embryonic development in kurosoi, Sebastes schlegeli Hilgendorf. J. Exp. Mar. Biol. Ecol. 101 (3), 239–256.
Boehlert, G.W., Kusakari, M., Yamada, J., 1991. Oxygen consumption of gestating female Sebastes schlegeli: estimating the reproductive costs of livebearing. Environ. Biol. Fish. 30 (1–2), 81–89.
Breder, C.M., Rosen, D.E., 1966. In: Modes of Reproduction in Fishes. The American Museum of Natural History, New York.
Davis, J.C., 1975. Minimal dissolved oxygen requirements of aquatic life with emphasis on Canadian species: a review. J. Fish. Res. Board Can. 32, 2295–2332.
Gilbert, S.F., 1997. In: Developmental Biology. Sinauer Associates, Sunderland, MA.
Greenbank, J., Nelson, P.R., 1959. Life history of the threespine stickleback Gasterosteus linnaeus Linnaeus in Karluk Lake and Bare Lake Kodiak island. Alaska Fish. Bull. 153 (59), 537–559.
Helfman, G.S., Collette B.B Facey, D.E., 1997. In: The Diversity of Fishes. Blackwell Science, MA. Hayakawa, Y., Munehara, H., 1996. Non-copulatory spawning and female participation during early egg care in
a marine sculpin Hemilepidotus gilberti. Ichthyol. Res. 43 (1), 73–78.
Hayakawa, Y., Munehara, H., 1998. The environment for fertilization of non-copulating marine sculpin, Hemilepidotus gilberti. Environ. Biol. Fish. 52 (1–3), 151–186.
Hishida, T., Nakano, E., 1954. Respiratory metabolism during fish development. Embryologia 2, 67–79. Johannessen, T., Gjøsæter, J., Moksness, E., 1993. Reproduction, spawning behaviour and captive breeding of
the common wolffish Anarhichas lupus L. Aquaculture 115 (1–2), 41–51.
Koya, Y., Munehara, H., Takano, K., Takahashi, H., 1993. Effects of extracellular environments on the motility of spermatozoa in several marine sculpins with internal gametic association. Comp. Biochem. Physiol. 106A (1), 25–29.
Kusa, M., 1949a. Hardening of the chorion of salmon eggs. Cytologia 15, 131.
Kusa, M., 1949b. Further notes on the hardening of chorion salmon eggs. Cytologia 15, 145.
Kweon, H.S., Park, E.H., Peters, N., 1998. Spermatozoon ultrastructure in the internally self-fertilizing hermaphroditic teleost, Rivulus marmoratus (Cyprinodontiformes, Rivulidae). Copeia 1998 (4), 1101–1106. Lønning, S., Kjorsvik, E., Davenport, J., 1984. The hardening process of the egg chorion of the cod, Gadus
morhua L., and lumpsucker, Cyclopterus lumpus L. J. Fish Biol. 24, 505–522.
Munehara, H., Takano, K., Koya, Y., 1989. Internal gametic association and external fertilization in the elkhorn sculpin, Alcichthys alcicornis. Copeia 1989 (3), 673–678.
Munehara, H., Koya, Y., Takano, K., 1991. The little dragon sculpins Blepsias cirrhosus, another case of internal gametic association and external fertilization. Jpn. J. Ichthyol. 37 (4), 391–397.
Munehara, H., Koya, Y., Takano, K., 1994. Conditions for initiation of fertilization of eggs in the copulating elkhorn sculpin. J. Fish Biol. 45, 1105–1111.
Munehara, H., Koya, Y., Hayakawa, Y., Takano, K., 1997. Extracellular environments for the initiation of external fertilization and micropylar plug formation in a cottid species, Hemitripterus villosus (Pallas) (Scorpaeniformes) with internal insemination. J. Exp. Mar. Biol. Ecol. 211 (2), 279–289.
Pavlov, D.A., 1994. Fertilization in the Wolffish, Anarhichas lupus: external or internal? J. Ichthyol. 34, 140–151.
Pavlov, D.A., Moskness, E., 1994. Production and quality of eggs obtained from wolffish (Anarichas lupus L.) reared in captivity. Aquaculture 122 (4), 295–312.
Piliger, J.F., 1997. Bilateral animal phyla: protostome coelomates. Sipunculans and echiurans. In: Gilberti, S.E., Raunio, A.M. (Eds.), Embryology, Constructing the Organism. Sinauer Associates, Sunderland, MA, pp. 167–188.
Wourms, J.P., 1981. Viviparity: the maternal-fetal relationship in fishes. Am. Zool. 21, 473–515.
Yabe, M., 1985. Comparative osteology and myology of the superfamily Cottoidae (Pisces: Scorpaeniformes), and its phylogenetic classification. Memoirs Fac. Fish. Hokkaido Univ. 32, 1–30.
Yamamoto, T., 1949. On the influence of salinity on the fertilization of the flounder’s egg. Seibutu 4, 46–51. Yamamoto, T., 1961. The physiology of fertilization in the fish egg. Int. Rev. Cytol. 12, 361–405. Zotin, A., 1958. The mechanism of hardening of the salmonid egg membrane after fertilization of spontaneous
