A plethora of studies have documented the negative impact of ACC on plants’ physiological and functional traits for the survival of plant populations in diverse climate conditions3,7,8,9,22. Our research provides experimental and theoretical evidence that this also applies to diaspore-dispersal traits and the dispersal potential of annual species that have evolved under the Mediterranean climate. The increased aridity predicted in these areas21 could restrict plants’ ability to move to more favorable ecological environments where they can thrive and reproduce under current and future climates. Our findings challenge the main assumption of the informed dispersal hypothesis, which states that traits related to dispersal should improve dispersal in stressful environmental conditions16,19. In particular, we have demonstrated that ACC’s drivers affect the expression of the diaspore’s features, such as the height at which diaspores are released from maternal plants and the period that diaspores are exposed to the wind flow. As a result, in warmer, drier CO2-enriched climates, our results predict that the dispersal distance of Mediterranean annual species will be reduced by about 40% on average. Finally, we have shown that abscission height and terminal velocity are reliable estimators of diaspore dispersal across all evaluated climatic scenarios.
Trait plasticity is the major way by which plants cope with different environments24. The analysis of expression for dispersal-related and functional traits performed here suggests that the differential effects of temperature, enriched-CO2 ambient, and drought on abscission height, diaspore traits, and terminal velocity limited the effect of trait plasticity to enhance dispersal. While terminal velocity decreased in climate change and water-limited conditions, which would favour dispersal distance, the lowered abscission height observed for plants grown under these conditions cancelled out the potential benefit for dispersal of producing smaller and lighter diaspores. Notably, whereas the effect of ACC on terminal velocity was variable across species (Fig. S8), the reduction in the abscission height was pervasive in most analyzed species (Fig. S10). This suggests that the contrasting plastic response observed in these two traits precluded plasticity as an effective mechanism for ensuring Mediterranean annual plants' dispersal in predicted stressful environments, which concurs with theoretical predictions of the existence of an intrinsic limit to plasticity that can constrain the adaptive potential of plants to rapid and extensive environmental changes, such as those resulting from ACC24,25. Similarly, our findings highlight the importance of considering multivariate trait plasticity across environments when evaluating organisms' functional responses26. Thus, PCA results suggest that co-expression of traits (i.e., phenotypic integration) occurs for diaspore mass, diaspore area, abscission height, wind loading, terminal velocity, and C/N ratio since all of them covary positively to PC1 across climatic scenarios; meanwhile, holding time varies orthogonally to them (Fig. S11). Hence, while correlated selection would be expected on traits that, in an integrated manner, enhance fitness-related dispersal27, integration should constrain plasticity because the more closely linked a trait's expression is to other traits, the harder the independent change of that trait in response to the environment26, as is the case here for abscission height, wing loading components, and terminal velocity. The fact that trait covariation remained constant across different climatic scenarios supports this idea (Fig. S12).
It is interesting to note that the analysis carried out on individuals revealed that plant species that produce large infructescences and/or heavy diaspores, such as A. sativa, B. hybridum, or M. polymorpha, positively contributed to trait differentiation in PC1 (as shown in Fig. S12). Conversely, plant species that produce small and light diaspores, and develop shorter infructescences, such as P. rhoeas, A. thaliana, and C. bursa-pastoris, negative contributed more to such trait differentiation (as shown in Fig. S12). This indicates that there is a differential integration of dispersal traits between small and large plant species. Small plant species minimize wing loading to increase the holding time, while larger plant species depend on a complex phenotypic integration for dispersal.
Our study predicts an important additive impact of ACC factors on diaspore dispersal (Fig. 2). In particular, the reduction in the dispersal distance is expected to be about 40% in ambient with elevated temperature, enriched-CO2, and soil water limitation, and ca. 22–23% in the other climatic scenarios compared to the most optimal climatic one (Fig. 2). These findings contrast with previous research that found similar or enhanced dispersal in stressful environments14,16,17,18. For example,, warming increased plant height by 9% in Carduus nutans, positively affecting the spread rate14. In the same species, a different study showed that drought reduced plant height yet predicted dispersal distances of plants affected by water stress was similar or longer than those of plants growing with abundant water17. Three non-exclusive explanations are possible to explain discrepancy of our results with other studies. First, the effect of the environment on dispersal is species-specific. For example, in our study, the dispersal distance of Papaver rhoeas seeds was enhanced by ca. 24% under climate change and drought compared to current ambient and humid conditions (Fig. S13) because the abscission height was unaffected by climate change or drought (Fig. S8), being the exception to the general pattern described here. Second, in the same vein, our research has adopted an approach that includes multiple species in our models and experiments instead of focusing solely on one or a few related species. Third, previous studies have not considered the impact of CO2 enrichment on dispersal, which is known to have net effects on plant physiology apart from temperature2, although our approach does not allow us to test this later effect directly. Our findings therefore strongly support the notion that ACC factors have a significant influence on limiting the dispersal distance of diaspores of annual plants. This is primarily due to a marked decrease in abscission height – in relative terms, the most important predictor of diaspore dispersal28 – compared to a more moderate reduction in terminal velocity (Fig. 1, Fig. 3).
We can draw a few ecological conclusions and predictions based on our findings. Except for one case, our research suggests that the dispersal of diaspores may be significantly limited in the near future in the Mediterranean areas, where increased climate aridity is forecasted in the Mediterranean areas20,23. Because diaspore dispersal is a crucial factor in the population spatial spread 14,29, most species' migration ability is expected to be restricted in the future climate forecasted for the Mediterranean areas. This, in turn, is likely to limit the plant species' capacity to track advantageous ecological niches spatially. However, dispersal involves more than just the initial movement of diaspores30,31,32. Assessing the demographic effects of limited initial dispersal on germination, seedling emergence, and survival in a community-level and density-dependent context would still be necessary to fully understand the impact on population spread of shortened dispersal driven by ACC. 29,33.
In addition, our results have highlighted an important implication: ACC's effects on dispersal traits are species-specific. This, in turn, affects the dispersal distances of plant species in different directions, leading to a community with many losers and few winners (P. roheas, for example). This pattern may eventually result in the colonization of new favorable niches and drive the future assembly of annual Mediterranean plant communities.
This study highlights the limitations of the adaptability of plants to rapid and significant environmental changes. It suggests that intrinsic limits to trait plasticity can restrict plants' adaptive potential. However, it must be noted that our study only employed a single or a minimal number of accessions to analyze trait expression (as required for analyzing plasticity). Therefore, genetic variation for plasticity in functional and life-history traits may exist, leading to populations exhibiting higher levels of plasticity in dispersal traits. This could allow for greater potential for dispersal and colonization in warmer and drier environments. Future studies should explore genetic variation in nature to understand the plastic responses of dispersal-related traits better.
Finally, determining the importance of dispersal-competition trade-offs under climate change35 will require more research to understand the future dynamics of the assembly of plant communities inhabiting ecosystems under changing Mediterranean-type climates.