A review and meta-analysis of intraspecific differences in phenotypic plasticity: Implications to forecast plant responses to climate change

Aim Many studies use differences among plant populations to infer future plant responses, but these predictions will provide meaningful insights only if patterns of plasticity among populations are similar (i.e., in the absence of population-by-environment interaction, P x E). In this study, we tested whether P x E is considered in climate change studies. Specifically, we evaluated whether population differentiation varies across environments and whether P x E is determined by aspects of the study system and experimental design. Location Global. Methods We conducted a literature search in the Thomson Reuters Web of Science database to identify studies assessing population differentiation in a climate change context. We quantified the occurrence of P x E and performed a meta-analysis to calculate the percentage of traits showing P x E in the study cases. Results We identified 309 study cases (from 237 published articles) assessing population differentiation in 172 plant species, of which 64% included more than one test environment and tested P x E. In 77% of these studies, P x E was significant for at least one functional trait. The overall proportion of traits showing P x E was 33.4% (95% confidence interval 27.7-39.3). These results were generally consistent across life-forms, ecoregions and type of experiment. Furthermore, population differentiation varied across test environments in 76% of cases. The overall proportion of traits showing environment-dependent population differentiation was 53.7% (95% confidence interval 37.9-69.3). Conclusions Our findings revealed that differences in phenotypic plasticity among populations are common but are usually neglected in order to forecast population responses to climate change. Future studies should assess population differentiation in many test environments (accounting for P x E) that realistically reflect future environmental conditions, assessing climate change drivers that are rarely considered (e.g., multifactor experiments incorporating higher CO2 levels). Our review also revealed the predominant focus of population studies on trees from temperate climates, identifying underexplored life-forms (shrubs, annuals), phylogenetic groups (ferns, ancient gymnosperms) and ecoregions (tropical, arctic) that should receive more attention in future.