Robinson

Sookias, Roland B, 2020, Exploring the effects of character construction and choice, outgroups and analytical method on phylogenetic inference from discrete characters in extant crocodilians, Zoological Journal of the Linnean Society 189 (2), pp. 670-699 : 682

publication ID

https://doi.org/10.1093/zoolinnean/zlz015

persistent identifier

https://treatment.plazi.org/id/DC418788-0921-5C54-1F61-2233923E1BFE

treatment provided by

Plazi

scientific name

Robinson
status

 

Robinson –Foulds index and resolution

The Robinson –Foulds ( RF) index of the inferred trees compared to the DNA-based phylogeny are generally similar (column ‘R–F distance from DNA’ in ‘Phylogenetic analyses undertaken and result.xlsx’ in supplement; minimum 4, maximum 8, mode 4). Using unweighted parsimony, the RF index is lowest using all characters (mean 4.7), then with the reassessed matrix (mean 5), then with only robust characters (mean 6.25). On average, for analyses including two outgroups, neighbour-joining trees have the lowest RF distance (5), followed by implied weights parsimony (5.2), unweighted parsimony (5.3), and Bayesian analyses (5.6). Using a single outgroup, unweighted parsimony slightly outperforms neighbour-joining (4.95 vs. 5).

The number of resolved nodes in the trees ranges from 8 to 15 (mean 13.65). For unweighted parsimony, the number of nodes is highest using all characters (mean 14), followed by the reassessed matrix (mean 13) and just robust characters (mean 12.67), as expected given the decreasing number of characters. The mean number of nodes resolved by unweighted parsimony is highest (12.7), followed by implied weights parsimony (12.5), and Bayesian analysis and neighbour-joining (both 12.4). Using a single outgroup, unweighted parsimony trees have very slightly more nodes than neighbour-joining trees (mean 13.25 versus 13.2).

Kingdom

Animalia

Phylum

Arthropoda

Class

Arachnida

Order

Araneae

Family

Anyphaenidae

Darwin Core Archive (for parent article) View in SIBiLS Plain XML RDF