Am. Paterson et al., PENGUINS, PETRELS, AND PARSIMONY - DOES CLADISTIC-ANALYSIS OF BEHAVIOR REFLECT SEABIRD PHYLOGENY, Evolution, 49(5), 1995, pp. 974-989
Whether or not behavior accurately reflects evolutionary relationships
(phylogeny) has been hotly debated by ethologists and comparative psy
chologists. Previous studies attempting to resolve this question have
generally lacked a quantitative, phylogenetic approach. In this study
we used behavior and life-history (BLH) information (72 characters) to
generate phylogenetic trees for 18 seabird species (albatrosses, petr
els, and penguins). We compared these trees with trees obtained from i
sozyme electrophoretic analysis of blood proteins (15 loci and 98 elec
tromorphs) and partial mitochondrial 12S ribosomal DNA sequences (381
base pairs). Cladistic analysis of the BLH data set generated three MP
trees (tree length = 243, CI = 0.52, RI = 0.57) with significant clad
istic structure. The BLH characters were classified into four types (f
oraging, agonistic, reproductive, and life history) and levels of homo
plasy for each type were measured. No significant differences were fou
nd among these categories, The BLH trees were shown to be significantl
y more congruent with the electrophoretic and 12S sequence trees than
expected by chance. This indicates that seabird BLH data contains phyl
ogenetic signal. Areas of incongruence between BLH trees and a phyloge
ny generated by combining the data sets were predicted to result from
ecological constraints that did not covary with phylogeny. These predi
ctions were supported by the results of a concentrated changes test. T
his study found that this BLH data set was no more homoplasious than m
olecular data and that BLH trees were significantly congruent with mol
ecular trees.