Variation in modes and rates of evolution in nuclear and mitochondrial ribosomal DNA in the mushroom genus Amanita (Agaricales, Basidiomycota): Phylogenetic implications
Jm. Moncalvo et al., Variation in modes and rates of evolution in nuclear and mitochondrial ribosomal DNA in the mushroom genus Amanita (Agaricales, Basidiomycota): Phylogenetic implications, MOL PHYL EV, 16(1), 2000, pp. 48-63
Modes and rates of molecular evolution, and congruence and combinability fo
r phylogenetic reconstruction, of portions of the nuclear large ribosomal s
ubunit (nLSU-rDNA) and mitochondrial small subunit (mtSSU-rDNA) genes were
investigated in the mushroom genus Amanita. The AT content was higher in th
e mtSSU-rDNA than in the nLSU-rDNA A transition bias in which AT substituti
ons were as frequent as transitions was present in the mtSSU-rDNA but not i
n the nLSU-rDNA. Among-sites rate variation in nucleotide substitutions at
variable sites was present in the nLSU-rDNA but not in the mtSSU-rDNA Likel
ihood ratio tests indicated very different models of evolution for the two
molecules. A molecular clock could be rejected for both data sets. Rates of
molecular evolution in the two molecules were uncoupled: faster evolutiona
ry rates in the mtSSU-rDNA and nLSU-rDNA were not observed for the same tax
a. In separate phylogenetic analyses, the nLSU-rDNA data set had higher phy
logenetic resolution. The partition homogeneity test and statistical bootst
rap support for branches indicated absence of conflict in the phylogenetic
signal in the two data sets; however, tree topologies produced from the sep
arate data sets were not congruent. Heterogeneity in modes and rates of evo
lution in the two molecules pose difficulties for a combined analysis of th
e two data sets: the use of equally weighted parsimony is not fully satisfa
ctory when rate heterogeneity is present, and it is impractical to determin
e a model for maximum-likelihood analysis that fits simultaneously two hete
rogeneous data sets. Overall topologies produced from either the separated
or the combined analyses using various tree reconstruction methods were ide
ntical for nearly all statistically significant branches. (C) 2000 Academic
Press.