Figure4.Phylogenetically informative characters (nucleotide composition at variable sites) for four gene regions studied. Numbers
above the sequences refer to base pair position from the first base in the amplified fragments. Thefirst position for the amplified fragment
for 12S (No. 35) corresponds to No. 570 in the published 12S sequence for Latimeria chalumnae (Hedges et al. 1992), in 16S(No.30)
corresponds to No. 176 for L. chalumnae (Hedges et al. 1992), in 18S(No. 33) corresponds to No. 682 for L. chalumnae (Stock et al. 1991b),
and in cytochrome bgene (No. 6) corresponds to No. 25 for L. chalumnae (Normark et al. 1991).
uniting Acipenser as monophyletic. In the 18S
rDNA tree, there was again no evidence of mono-
phyly ofAcipenser.Finally, the cytochromebtree
groupedA.medirostris,A.transmontanus,A.ruthe-
nusandA.baeriiwithHuso dauricus. Acipenser me-
dirostrisandA.transmontanusshared apomorphies
in the 18S rDNA and cytochrome btrees, and such a
grouping was not ruled out by either of the other
two trees.PseudoscaphirhynchusandScaphirhyn-
chus emerged as monophyletic in only one tree
(16S, Figure 3).
Combined molecular tree and comparison to exist-
ing morphological hypothesis
The DNA sequence characters for the four gene re-
gions were combined into a data matrix and each
character was given equal weight. One parsimony
tree was obtained using this combined molecular
data matrix (Figure 5, Table 6).
Findeis (1993,1997) used osteological characters
for constructing a morphological phylogenetic hy-
pothesis which focused on generic relationships
among Acipenseriformes. Only a single parsimony
tree resulted from his data set. He examined six spe-
cies of acipenseriforms not included in our molecul-
ar survey (Scaphirhynchus platorynchus,A.brevi-