This study investigated whether undifferentiated germ and/or somatic stem cells remain in the differentiated ovary of a species that does not undergo sex changes under natural conditions and retain their sexual plasticity. The effect of aromatase inhibitor (AI)-treatment on sexually mature female zebrafish was examined. A 5-month AI treatment caused retraction of the ovaries after which testes-like organs appeared, and cyst structures filled with spermatozoa-like cells were observed in sections of these tissues. Electron microscopic observations revealed that these cells appeared as large sperm heads without tails. Sperm formation was re-examined after changing the diet to an AI-free food. A large number of normal sperm were obtained after eight weeks, and no formation of ovarian tissue was observed. Artificial fertilization using sperm from the sex-changed females was successful. These results demonstrated that sex plasticity remains in the mature ovaries of this species.

Although natural populations consist of individuals with different traits, and the degree of phenotypic variation varies among populations, the impact of phenotypic variation on ecological interactions has received little attention, because traditional approaches to community ecology assume homogeneity of individuals within a population. Stage structure, which is a common way of generating size and developmental variation within predator populations, can drive cannibalistic interactions, which can affect the strength of predatory effects on the predator's heterospecific prey. Studies have shown that predator cannibalism weakens predatory effects on heterospecific prey by reducing the size of the predator population and by inducing less feeding activity of noncannibal predators. We predict, however, that predator cannibalism, by promoting rapid growth of the cannibals, can also intensify predation pressure on heterospecific prey, because large predators have large resource requirements and may utilize a wider variety of prey species. To test this hypothesis, we conducted an experiment in which we created carnivorous salamander (Hynobius retardatus) populations with different stage structures by manipulating the salamander's hatch timing (i.e., populations with large or small variation in the timing of hatching), and explored the resultant impacts on the abundance, behavior, morphology, and life history of the salamander's large heterospecific prey, Rana pirica frog tadpoles. Cannibalism was rare in salamander populations having small hatch-timing variation, but was frequent in those having large hatch-timing variation. Thus, giant salamander cannibals occurred only in the latter. We clearly showed that salamander giants exerted strong predation pressure on frog tadpoles, which induced large behavioral and morphological defenses in the tadpoles and caused them to metamorphose late at large size. Hence, predator cannibalism arising from large variation in the timing of hatching can strengthen predatory effects on heterospecific prey and can have impacts on various, traits of both predator and prey. Because animals commonly broaden their diet as they grow, such negative impacts of predator cannibalism on the heterospecific prey may be common in interactions between predators and prey species of similar size.

No abstract

Summary 1.In most animal species, the hatchling stage is a highly vulnerable life-history stage. In many fish and amphibian species, hatchling abundance varies substantially among sites and years, with the result that selection strength in conspecific interactions such as cannibalism is also variable. The variability of selection leads species to evolve phenotypic plasticity, and adaptive trait changes in hatchlings that depend on the density of conspecifics can therefore be expected. However, plasticity strategies in response to cannibalistic interactions in this vulnerable life-his tory stage have received little attention. 2. Because cannibalism success is dependent on the size balance between predator capturing organ and prey body, and also on the prey-capturing ability of the cannibal and the escape ability of its prey, we hypothesize that hatchlings will exhibit faster growth and development in response to conspecific interactions. We performed a series of experiments to test this hypothesis, using pre-feeding larvae of a cannibalistic salamander (Hynobius retardatus). 3. Many traits of the hatchlings reared with conspecifics were larger than the same traits in those reared alone, because the former hatchlings grew faster and were more advanced developmentally. The time to the beginning of feeding was shorter, and swimming speed as an indicator of escape performance was faster, for hatchlings reared with conspecifics. 4. Hatchlings reared with conspecifics more successfully cannibalized small hatchlings and were also highly resistant to being cannibalized by large conspecifics, compared with hatchlings reared alone. These differences in cannibalism success and avoidance of cannibalization were well explained by differences in gape and head size, respectively, between the interacting hatchlings. Therefore, acceleration of growth and development of hatchlings in the proximate presence of conspecifics may confer significant fitness advantages on hatchlings in a high-density cannibalistic situation. 5. Developmental and growth plasticity may thus be a powerful adaptive mechanism that allows individuals to respond to the dynamic character of salamander populations in nature.

No abstract