Evolutionary responses to 17 years of simulated climate change in a species-rich grassland ecosystem
We know little about the extent to which adaptation by natural selection will allow populations to persist in situ through climate change. These adaptive responses are potentially important, as they may play a key role in mediating the resistance of ecological communities to environmental change. At Buxton climate change impacts laboratory (BCCIL), semi-natural grassland has been exposed to more than 17 years of simulated climate change including drought, watering, and warming treatments. The grassland has proven to be remarkably resistant to these manipulations, changing little in species composition. We asked whether populations of four coexisting, perennial plant species have evolved in response to simulated climate change (summer drought treatment) within grassland plots at BCCIL. Populations of established plants were collected from long-term drought-treated and control plots at BCCIL, propagated clonally, and used to set up a common environment experiment. Plant phenotypes were monitored for three years, and molecular markers were used to verify distinct plant genotypes (all species), and to test for genomic signatures of selection (two species).
Plant phenotypes showed significant among-clone variation (broad sense heritability = 21–48 %). Three of the four study species showed significant divergence in phenotype between populations exposed to long-term drought treatment and those from control plots (ambient climatic conditions). Altered phenotypes included reproductive effort and timing, leaf morphology, and changes in clonal architecture. No phenotypic measure varied consistently among climatic environments in all study species. Outlier analyses revealed a genomic signature of climatic selection in one of the two tested study species. Our results indicate that grassland populations at BCCIL have evolved in response to 17 years of simulated summer drought, and suggest that a significant proportion of coexisting plant species may be capable of such responses. These genetic changes may have contributed to the stability of grassland community structure at BCCIL through 17 years of climatic perturbation. However, climatic selection did not favour a consistent set of phenotypes in different species, indicating that the nature of evolutionary responses to climate change may be difficult to predict, even within a single ecological community.