Case studies: species delimitation
To test the performance of species delimitation algorithms applied on
different USCO datasets, we delineated species with tr2 (Fujisawa et
al., 2016) and SODA v. 1.0.2 (Rabiee et al., 2020) based on the results
of the three different extraction approaches in all four case studies,
using the programs’ default parameters and not providing a guide tree.
Tr2 conducts species delimitation based on the topological variation
between rooted gene trees using the distribution of triplets (i.e.,
topologies with three individuals). As input, we used the gene trees
generated by IQ-TREE, with the gene trees being re-rooted with
nw_reroot (part of Newick Utilities 1.6.0; Junier & Zdobnov, 2010) on
the outgroup of the respective study case. We included only trees that
contained all specimens, as gene trees lacking specimens cannot be
handled by the program. SODA uses the distribution of topologies of
quartets of individuals in unrooted gene trees to infer species
boundaries (Rabiee et al., 2020). We used all gene trees as input for
SODA.