Physiological And Genetic Compensation Mechanisms For Ovoviviparous Reproductive Mode Selection Disadvantages In Snakes

Oviparity is a predominant mode of reproduction in reptiles,and many female reptiles lay eggs with embryos that have completed nearly one-third of the total embryonic development.About 20%of the squamate reptiles are viviparous,and viviparous females give birth to fully developed young.Viviparity has evolved independently at least 115 times in squamate reptiles,far more frequently than all other taxa combined.Squamates reptiles are therefore the ideal model systems to study the evolution of viviparity.Since the entire embryonic development stage of viviparous is in maternal body,the embryos in the oviduct will face the problem of insufficient oxygen.In addition,the eggs of some oviparous species will still face a low oxygen environment after being out of females,such as underground lair and flooded nest when rainy season begins.We speculate that the embryo can tolerate hypoxic environment to a certain extent and we incubate the eggs of the tiger keelback snake(Rhabdophis tigrinus)under the different oxygen concentrations.Our data showed that:(1)hypoxia strongly restricted embryonic development;(2)the short-term hypoxia exposure for the first 5 days of incubation had no significant effect on embryonic development,suggesting that embryos of oviparous reptiles could be retained in maternal body for a short period of time;and(3)oxygen demand in early stages of embryonic development was lower than that in late stages,suggesting that there should be some physiological mechanisms in viviparous reptiles to meet demand of embryos for oxygen in late developmental stages.Previous studies have suggested that the residual yolk in hatchlings abdomen plays a buffered role in embryonic development when the incubation environment is unpredictable.However,the excessive egg yolk will increase the risk of survival for pregnant females.Therefore,we calculate that there are differences in yolk allocation strategies between oviparous and viviparous reptiles,that is,viviparous reptiles will allocate less yolk to their offspring than oviparous.We collected the data about the residual yolk of 20 species in snakes(12 oviparous and 8 viviparous snakes)and find that the mean of the residual yolk dry mass relative to total dry mass in viviparous species is less than that in oviparous ones;but the proportion of the residual yolk in viviparous species is not always less than that in oviparous.For example,the relative size of residual yolk in a viviparous snake Sinonatrix annularis is larger than that in five oviparous species(Coelognathus radiatus,Ptyas korros,P.mucosus,Xenochrophis piscator and Zaocys dhumnades;16%versus 12-15%of the body dry mass).In addition,we find that both carcass and fat body dry mass are positively correlated with the residual yolk dry mass and this relation is phylogenetically dependent,which suggests that the amounts of residual yolks in snakes is more likely to be phylogenetically dependent than to be associated with parity mode.Because of differences in maternal gestation period between different reproductive mode,the maternal metabolic levels have significant differences.We compare the levels of respiratory metabolism before and after reproduction in two snakes(Dinodon rufozonatum versus Sinonatrix annularis).And we analysis the correlation between maternal reproductive output and metabolic cost during pregnancy.The results show that the prenatal metabolic levels of the two snakes are higher than postpartum metabolic level.The metabolic level of the viviparous species is significantly higher than that of oviparous,which is probably due to the secretion activity of the shell gland and the synthesis of the eggshell.We use general regression model to analysis the relationship between female reproductive output and reproductive metabolic growth rate,and we find that the relative reproduction output is better to explain the maternal metabolic changes during pregnancy according the minimum AIC value.Because of the long pregnancy,the frequency of mating and production in viviparous species is significantly lower than that in oviparous.Therefore,we assume that the viviparous species should have a higher incidence of multi-paternity in order to improve the genetic diversity of the offspring.We analyze the multi-parental results among 18 species of snakes(11 oviparous species versus 7 viviparous species).The data shows that the parity mode in snakes has no effects on multiple paternity rates and the number of sires per clutch.However,there is a significant positive correlation between the number of sires per clutch and the clutch size.This suggests that the multiple paternity in single clutch is a common phenomenon in snakes,and this pattern,to a certain extent,reflects the development trend of the population.One of two factors that lead to the phenomenon of multiple paternity in reptiles is sperm storage,the other is the decoupling between mating and ovulation in time.During the period between mating and ovulation,the female snakes can make multiple mating which is beneficial to female post-copulatory sexual selection.Therefore,the multi-paternal pattern will be affected by cryptic female choice.In this study,to understand the sperm use pattern in female snakes,we record the position of embryos in the oviduct by dissecting the viviparous water snake(Sinonatrix annularis),and analyze the number of sires.Our result reflects that before using the sperms in oviduct,the females have mixed the sperms from multiple males,it is beneficial for sperm competition possibly.Besides,the high parental skew indicates the cryptic female choice,which also can enhance the quality of the sperm used for insemination.Multiple paternity is common in snakes,but the cause for this phenomenon is not yet clear.We test the growth rate of the king ratsnake(Elaphe carinata)offspring by a common garden experiment.We tried to explore the ecological adaptation of multiple paternity from the aspect of offspring fitness.The experimental data showed that:(1)there was no significant difference in the morphology of offspring between the single-sired and multiple-sired clutches;(2)there was also no significant difference between the hatchlings from dominant paternity and others;and(3)the number of sires and offspring fitness were negatively correlated,which suggests that the increased sires by multiple mating could enhance the average fitness of offspring.