Phenotypic variation and transgenerational responses of plants to different types of precipitation predictability

Abstract

BACKGROUND Current climate change alters mean climatic conditions, weather variability, and environmental predictability. Climatic models forecast strong increases in weather variability and decreases of environmental predictability, i.e. in the level of temporal autocorrelation of weather events. This suggests that in the future, organisms will face great challenges. However, little experimental evidence for these predictions exists, and the scarce experimental studies simulated droughts and/or extreme events, both being detrimental for many organisms. Therefore, the effects of changes in environmental predictability remain largely untested. Environmental predictability is suggested to affect functional traits, phenotypic expression and physiology, especially in sessile organisms like plants. More predictable environments may pose fewer challenges, and theoretic studies suggest that they lead to lower phenotypic variation. In contrast, less predictable environments are expected to favour advances in phenology, potentially altering reproductive success, survival, and population growth. Since plants derive important resources through their roots, root investment strategies and their plasticity might be key to cope with changes in environmental predictability. In addition, lower environmental predictability may entail greater intra-population and intra-individual variability, and greater physiological changes, which potentially affects ecosystem services. Finally, environmental predictability may also affect the speed and direction of adaptive evolution, and the shape and strength of selective regimes. Theoretic models further forecast that higher environmental predictability favours adaptive evolution, while lower predictability rather leads to increased phenotypic plasticity. However, these theoretical predictions remain largely untested. OBJECTIVES The general objective of this doctoral thesis is to experimentally test whether and how lower precipitation predictability affects phenotypic and transgenerational responses of two plant species: the common sainfoin Onobrychis viciifolia and the common poppy Papaver rhoeas. The specific objectives are: First, investigating whether less predictable precipitation leads to phenological changes, and whether they affect reproduction, survival and population growth (Chapter 1 and 3). Second, testing whether and how less predictable precipitation affects the selective regime, root functional traits, root investment strategies, and whether observed responses may affect plant performance (Chapter 2). Third, testing whether environmental predictability mainly affects a given life-stage or whether it affects all life-stages similarly (Chapter 1-5). Fourth, testing whether the speed and direction of transgenerational responses depend on precipitation predictability (Chapter 1-3). Fifth, testing whether rapid transgenerational responses induced by differences in precipitation predictability are congruent with adaptive evolution and/or with an increase in phenotypic plasticity (Chapter 1). Sixth, testing whether less predictable precipitation modulates intra-population and intra-individual variation, and reproduction, and whether observed changes are transmitted to the next generation (Chapter 4). Seventh, investigating whether precipitation predictability changes physiological response, and whether such changes may affect ecosystem services (Chapter 5). METHODOLOGY Seeds of Onobrychis viciifolia (Fabaceae) and Papaver rhoeas (Papaveraceae) were sown during four consecutive years in experimental field plots situated at an experimental research station located near Jaca (Huesca, Spain). More and less predictable precipitation was simulated at daily and inter-seasonal time-scales, using an automatic irrigation system. To test for treatment differences in phenology and reproduction, many functional traits of both species were weekly measured. At the end of the annual cycle, offspring seeds were individually collected, measured, and stored over the winter. To test whether differences in precipitation predictability affected ecosystem services, lignin, carbon and nitrogen content of the plant’s above- and below-ground parts were determined. To test whether precipitation predictability affects the speed and direction of transgenerational responses, three generations of stored offspring seeds (descendants) were sown in the subsequent year. In each of the three years, descendants were sown in the same plot, under the same precipitation conditions, and at the same time as the ancestral seeds (i.e. seeds that were naive with regard to the simulated predictability conditions), and their development trajectories and traits were compared. To disentangle between evolutionary change and evolution of phenotypic plasticity, descendant seeds of the second and third generation were planted in all treatment combinations. RESULTS Chapter 1 shows that lower short-term predictability of precipitation leads to phenological advance and to an increase in reproductive success and population growth. Both species exhibited rapid transgenerational responses that mitigated treatment differences observed in ancestors. The detected transgenerational responses were in line with increased phenotypic plasticity, rather than a shift in trait expression. Moreover, transgenerational responses mainly existed with respect to early growth conditions. Chapter 2 demonstrates that, in both species, root investment depends on precipitation predictability, and that investment changes lead to increased performance. Moreover, precipitation predictability induced differences in the strength of selection but not in the type of the selective regime. Chapter 3 shows that decreased precipitation predictability enhances seedling emergence, survival and reproductive rates in both species. Mainly treatment during early growth affected survival and reproduction, while treatment during late growth let to smaller differences. Descendants of P. rhoeas exhibited increased viability compared to ancestors (in both treatments), and this difference was greater when exposed to less predictable precipitation. Chapter 4 evidences that precipitation predictability significantly affected intra-population and intra-individual variability, both being greater when exposed to less predictable precipitation. Intra-individual variability in reproductive traits is under stabilizing selection, and precipitation predictability affected the location of the optima, optimal reproduction, as well as the strength of selection, but not type of selection (i.e. stabilizing vs directional selection, etc). Chapter 5 demonstrates that less predictable precipitation led to lower lignin content, higher ANPP (Aboveground Net Primary Productivity), more aboveground biomass, and higher nitrogen content in roots and leaves. This shows that less predictable precipitation altered plant physiology, which positively affected ecosystem services. CONCLUSIONS In contrast to the expected negative effects of decreased environmental predictability, our study shows that at least some plant species may benefit from a lower precipitation predictability, and that this positively affected ecosystem services. Earlier reproduction observed in less predictable precipitation is in line with theory, but led to increased reproductive success, survival, and population growth, rather than affecting them negatively. This difference most likely exists, because the theoretical models assume that earlier reproduction will be associated with shorter reproductive periods, while in this study it was associated with a longer reproductive period. The positive responses are also congruent with rapid and plastic responses exhibited by root functional traits, and are in line with recent studies showing that plants were affected by the predictability of soil nutrients. Less predictable precipitation increased intra-population variability but also intra-individual variability, which demonstrates that intra individual variability is a highly plastic trait that is under natural selection, rather than phenotypic noise. These results also suggests that higher intra-individual variance may affect population dynamics and biodiversity. Moreover, under decreased precipitation predictability, plants exhibited rapid transgenerational responses, which were congruent with increases in phenotypic plasticity, and mainly existed with respect to precipitation predictability during early growth. In contrast, reproduction of ancestors was chiefly affected by late predictability. This suggests that it is important to consider the developmental stage at which organisms are exposed to differences in environmental predictability, in evolutionary and ecological theories. The findings reveal that many herbaceous plants may be able to cope with increasing environmental uncertainties, thanks to investment changes, existing, and evolving phenotypic plasticity. Therefore, to anticipate the impact of climate change, theoretical models and mitigation efforts should take into account that plants may rapidly respond to changes in climatic predictability, and that selective regimes may change as well. This suggests that lower climatic predictability may induce less biodiversity losses than expected by climatic models.