The diversity of species and traits is the outcome of multiple evolutionary processes operating over millions of years. These processes affect rates of trait evolution, speciation, and extinction. A key problem is identifying the relative importance of distinct processes in driving observed patterns in species traits and phylogenetic trees. Here, we show how two processes, competition and stabilising selection, can act opposingly but still leave identifiable traces in macroevolutionary data. Building on previous simulation studies, we model the joint influence of competition and stabilising selection on rates of trait evolution, speciation completion, and extinction. We find that opposing effects result in nuanced patterns of trait evolution and diversification dynamics that, when considered in isolation, could easily lead to misinterpretation. In particular, while the best fitting likelihood model of trait evolution is typically Brownian motion when competition and stabilising selection act simultaneously, we find stronger phylogenetic signal than expected under BM and trait distributions that are distinctly platykurtic. Under the same modelling scenarios diversification is often best characterised by a pure birth model coupled with strongly suppressed species richness. Taken together, we suggest that by considering multiple simple metrics measuring trait evolution and diversification dynamics it is possible to distinguish between competition or stabilising selection as single dominant processes from the opposing effects of the two processes operating together in macroevolutionary data.

The simulations zipped file contains the outputs of simulations run under the CAMPSITE model. Each .rds file is a list of 100 simulations run under different parameter values - the titles of the different .rds files reflect the different parameters used. Each simulation (out of 100) is an output of the CAMPSITE model, and thus is a list containing the following three lists: 1. all: containing tree (the full phylogeny), tips (a vector of trait values at the tips), trait_mat (a list of vectors, each one the trait values of the lineage in each time step. NA if lineage did not exist at that time step), lin_mat (matrix of data on each lineage, including speciation and extinction times). 2. gsp_fossil: containing tree (the phylogeny with extant and extinct good species), tips (a vector of trait values at the tips). 3. gsp_extant: containing tree (the phylogeny with extant only good species), tips (a vector of trait values at the tips).

The included R package contains all the code necessary to run the CAMPSITE model, as well as to run the simulations and generate the visualisations described in the manuscript. The attached data includes the output from all simulations described in the manuscript.