Goals of the project - exploration of the effect of incomplete sampling (spatial vs non-spatial, temporal) on measures of disparity
- exploration of that combined with effect of extinction selectivity
Tree simulation - B,D rates fixed & homogeneous
- constant rates
- ME w/optional selectivity (no, trait-based, geo-based, correlated trait-geo)
- radiation -> test magnitude of the ME (low/high) + only keep one for the correlated sims -> magnitude of the correlation (low/high) & magnitude of radiation (low/high)
Trait simulation => simulation of traits on the complete tree - low/high rate of evolution => keep 2 traits but test one dataset with varying rate of evolution for each trait => decide how to simulate geographically-correlated traits -> OU w/different optima ? BM with different speeds ?
Geographical simulation
- joined trees so defined by the tree simulation
- 1 vs 5 migration events => to see how this can be adapted for the ME setup - need to sample the attachment points before the ME
Fossil sampling
- sampling variations (constant, temporal, spatial, spatial in 2nd bin only) => test high and low sampling rates => trait value based on the position of the fossil specimen for each fossil => control dataset is high constant fossil sampling
Output metrics look at increase/decrease in disparity - relative change between bins (with sign of change)
- look at change compared to the benchmark (almost complete sampling)