can be assessed. Challenges center around establishing Ramphocelus tanagers are distributed throughout homology of characters (alignment procedures), disthe Neotropical lowlands. In this paper, mitochroncovering and dealing with saturation of changes (multidrial DNA (mtDNA) sequence data from cytochrome b ple substitutions at a particular base position), and and ND2 genes are used to estimate relationships whether or not different data sets should be combined among seven of nine species of the avian genus into larger data sets for phylogeny reconstruction.
Ramphocelus. Genetic differentiation is high between
For protein-coding genes, there is often little problem Ramphocelus passerinii passerinii and Ramphocelus with sequence alignment; thus, homology of sites is passerinii costaricensis, and the two subspecies are digenerally easily established (Mindell, 1991). Problems agnosable and distinct from one another both morphoof saturation of changes are not so easily handled. For logically and genetically. Thus, elevation to species example, empirical research suggests that changes at status is recommended. Three clades are supported by third positions of codons (particularly transitions) occur both gene sequences; one clade contains R. passerinii, more frequently than changes at first or second positions R. costaricensis, and R. icteronotus; the second clade because they are the least likely to result in a change contains Ramphocelus carbo, Ramphocelus bresilius, in the amino acid composition of a protein. Thus, those and Ramphocelus nigrogularis; the third clade concharacters that are the most likely to be saturated (third tains Ramphocelus sanguinolentus. The degree of satuposition changes) comprise the majority of informative ration was assessed for both genes and saturation of characters in sequence data sets. Also, controversy exthird position of codons occurs by 10-12% uncorrected ists regarding the conditions under which data sets pairwise sequence divergence. The general area cladoshould be combinedinto single larger data sets for analygram suggests the following area relationships: Pacific sis and whether character weighting should be used (see and Caribbean Central America are sister areas, Choco Β΄is the sister to the Central American area, and discussions in Bull et al., 1993; Chippendale and Wiens, Amazonia/southeastern Brazil is the outgroup area 1994; Eernisse and Kluge, 1993; Huelsenbeck et al., to the Choco Β΄/Central American clade.