The role of epigenetics in life history evolution.

Adaptations and responses of organisms to cope with variation in environmental conditions have interested biologists for over a century. In this study, we focus on the annual killifish genus Nothobranchius. All species are endemic to temporary ponds native to the African continent. These fishes have evolved a wide array of adaptations to survive in this harsh environment. When their habitat is inundated, dormant eggs buried in the sediment layer hatch. The larvae rapidly grow to maturity and from then onwards reproduce on a daily basis until death or until the habitat vanishes. The drought resistant eggs subsequently bridge the dry season, going through a series of developmental arrests. Environmental conditions, however, are variable between inundations of a given site and between sites. Given their limited dispersal capacity, gene flow is limited and environmental variability has to be handled in situ.

Both the length of inundations and the presence of riverine fish predators differs in time. Laboratory exposures to either a desiccation or a predation risk induced a phenotypic plastic response. In both instances, adults doubled the number of eggs they produced when exposed, indicating a flexibility in life history to respond to the variability of their habitat. This flexibility however appeared to be restricted to the adult stage, as early life history variables remained unaltered under the same risks. Potentially, the inflexible juvenile stages constitute a conservative bet hedging strategy, where an all-round phenotype is produced that performs decently under most common circumstances. Speeding up life history might be physiologically impossible at that point or too costly should those cues prove to be unreliable. Delaying life history in turn might equally be too costly given their occurrence in a rapidly vanishing habitat. Standardized follow-up on embryonic development of four different N. furzeri populations revealed variation within and between populations in the developmental trajectory. This further supports the existence of a diversified bet hedging strategy, where variability in the speed of embryonic development ensures that some individuals will end up in a good inundation.

Aside from embryonic development, populations differed furthermore in life history traits such as the age at maturation and egg deposition. Populations originating from a dry region where inundations are typically short exhibited a faster life history pace than populations from a more humid region where inundations typically last longer. In light of life history theory, these life histories appear to be fit to reproduce as much as possible given the time available. The difference in male and female lifespan in our experimental setup suggests a trade-off between reproduction and survival, a trade-off which is supported by the negative correlation between egg deposition and lifespan in N. wattersi females. Furthermore, data obtained in a common garden experiment indicate that fast maturation times are associated with smaller adult body sizes and higher fecundities. In addition, egg deposition (as a proxy for pace of life) appeared to be linked with certain mating behaviours. In a population where females deposit a large number of eggs, males attempted to spawn a lot. Those females spent little time on mate assessment and did not resist coercive spawnings by non-preferred males. In turn, in a population where egg deposition was lower, males attempted to spawn less frequently. Females spent more time assessing potential mates and lowered the number of eggs they deposited when coerced into spawning by a non-preferred male. These behaviours fit predications postulated by the pace-of-life syndrome hypothesis, as a higher pace of life appears to be associated with a more aggressive mating system compared to a lower a pace of life.

These mating behaviours also shed new light on sexual selection and reproductive isolation in Nothobranchius. We found that the level of sexual size dimorphism was larger in a slow paced population compared to a fast paced population, suggesting that there are more opportunities for female choice and sexual selection in the former. Female choice in itself however appears to be an ineffective prezygotic isolation barrier, as it is largely overruled by male coercion. Potentially, selection against hybrids played a role in the origin of reproductive isolation in Nothobranchius fishes.