Evolutionary and ecological physiology of mangrove seedlings: Correlates, costs, and consequences of viviparous reproduction

I address evolutionary correlates and ecological consequences of viviparous reproduction in plants, examining mangroves--woody tropical halophytes--as a model system. More than two hundred plant species produce desiccation-intolerant embryos that germinate precociously, quickly lose viability after drying, and are "recalcitrant" to storage. This trait occurs disproportionately frequently among tropical taxa of wetland and wet-forest habitats, and may result secondarily from physiological adaptations to flooding. Integrative physiological and ecological studies of vivipary are critical to devising seed-handling strategies, predicting seedling performance, and understanding evolution of reproductive traits in general.;I quantified levels and dynamics of hormones, ion solutes and carbohydrates in vegetative, maternal (fruit) and embryonic tissues in pairs of related viviparous mangrove and non-viviparous non-mangrove species belonging to four phylogenetically independent families: Rhizophoraceae, Myrsinaceae, Plumbaginaceae, and Arecaceae. Levels of abscisic acid (ABA), the phytohormone coordinating both seed dormancy and plant responses to flooding and salinity, are consistently lower in embryos of mangrove species compared to upland sister taxa. Cytokinin levels are slightly higher in all tissues of mangroves, while sucrose, monosaccharide, and calcium ion contents are lower in viviparous embryos. Abscisic acid and compatible solutes often function to maintain osmotic homeostasis in plant tissues subjected to water stress. Hence, these data indicate that parallel evolutionary shifts in control of water relations at the maternal-embryo interface have occurred in diverse viviparous lineages, likely reinforced by selection pressures generated by water stress common to wetland habitats.;Focusing on the predominant neotropical mangrove, Rhizophora mangle, I demonstrate negative consequences of vivipary for early seedling survival and growth--contrary to the assumption that vivipary confers a selective advantage on seedlings. For example, pre-dispersal seed predators damage approximately thirty percent of viviparous seedlings on mangrove species world-wide, and double the rate of seedling abortion on R. mangle in Belize. Moreover, viviparous seedlings take unusually long to overcome maternal effects in adjusting to altered tidal levels and elevated atmospheric carbon dioxide. By contrast, R. mangle seedlings show more flexible responses to a range of light conditions than do saplings or mature trees.;Long-term experimental studies like these must be employed to predict the performance of viviparous seedlings under environmental stress. The adaptive values of vivipary in mangroves have been well articulated: early germination may facilitate prompt rooting during establishment, maintain buoyancy during sea dispersal, permit transfer of maternal nutrients to the seedling, and maintain osmotic equilibrium of the embryo. However, maternal costs associated with vivipary are potentially significant: numerous attached seedlings may constitute both a substantial carbon sink and a concentrated, apparent resource for herbivores. While post hoc evidence describes why vivipary may enhance seedling success in marine habitats, and thus why such traits may be fixed selectively in evolution, it does not reveal the mechanisms by which vivipary has arisen coincidentally in so many unrelated mangrove taxa (and conversely why such a seemingly advantageous strategy has not proliferated among non-woody, temperate halophytes, and other angiosperms). I propose that optimality theorists should thoroughly explore trade-offs in whole-plant physiology that contribute to the evolution of this and other complex life-history traits.