To investigate whether the rates of phenotypic change differ among different life-history traits, we fitted the following model to data from the empirical and modelling studies separately (Models 2 and 3): (2) Both models were fitted by maximum-likelihood method and reduced stepwise using likelihood-ratio comparisons with χ statistics as suggested by Crawley (2007).
Firstly, to assess the overall rates of phenotypic change, their dependence on F, and differences between empirical and modelling studies we fitted a model (Model 1) (1) where α is the random effect of the study ID, ε is the normally distributed error term and × indicates two-way interaction between F and SS.
However, the intercepts differed between empirical and modelling studies, which means that, for a given value of F, the magnitude of change was larger in empirical than in modelling studies (the coefficient for modelling studies was = −1.44) (Table 1).
However, few empirical or modelling studies have addressed the long-term consequences of ocean acidification for marine ecosystems5,6,7.
Building upon empirical and modelling studies of Caribbean reefs16, 17, 48, our findings show that not only is the identity of constituent corals an important driver of habitat structural complexity, but this occurs across scales.
There are at least four reasons that might explain why experimental rates are so high and very different from findings in empirical and modelling studies.
Further empirical and modelling studies are necessary to confirm or refute some of the assertions made here, and to provide further mechanistic detail.
The same could apply to the comparison of empirical and modelling studies, although modelling studies typically assumed constant values of genetic variance and therefore modelled evolutionary rates were rather constant over time.
As detailed in Text S5, previous empirical and modelling studies suggest a relative increase of the transmissibility Δ R0/ R0 ∈ [0, 1] as well as an immune escape σ ∈ [0, 0.5] for a post-pandemic variant.
The majority of both empirical and modelling studies have focused only on the first two steps and have concluded that for many crop species, transgene escape is inevitable (Jenczewski et al. 2003; Papa and Gepts 2003; Wilkinson et al. 2003b).
