ANGIOSPERM DIVERGENCE 153Th
e f
os
sil
- b
as
ed
es
tim
ates
a
re
, c
ont
ra
ry
to t
hei
r us
e,
a
ss
um
ed to p
rov
id
e a
m
ini
m
um
a
ge
fo
r^
th
e s
pli
t b
etw
ee
n th
e t
axo
n i
n q
ue
st
ion a
nd i
ts s
ist
er g
rou
p.
Column^1listst
he taxaaccordingto their usa
ge, and columns2–5 list our ageestimates(
node num
berscorrespond to thoseon thechronogra
ms,Figures8.3-^8
.6). T
hea
ge span giv
eninc
ol
umn 2 resultsfromthe threec
onsecutiveanalyses using ACCTRA
N
and DELTRANoptimization and maxim
umlikelihood forca
lc
ulatingbra
nch^le
ngths.^The^specific ages, resultingfromeac
h analy
sisarereported in columns3–- C
olumn^6 indicatest
hebootstrapes
timat
esofstandarderror, and col
umns7–8 listfo
ss
il-based estimatesextra
ct
ed fromMagallónet^al. (
1 999)and Magallónand Sanders
on (2001).The specifi
cfossil-
base
d age(column 7)c
orrespondsto that given by Magallón
et^al.(^1999) and Magallón andS
anderson(^20 01)a
nda
lsoindi
ca
tes on what taxon theageestimat
e was
b
as
ed.Column^8lists insta
nces
whenthe topology impl
ie
s that some^
otherfo
ss
il provid
esa betterminim
umage estimate fora^ node. In addition,furt
her
comparisons^
ar
e made wheneverthe fossil- b
as
edestimat
e was
bas
ed ona
less
inclusiv
e taxon.Theagefo
r theAristolochiaceae- L
actorid
aceaeclade,for exam
pl
e,w
as
b
as
ed on Lactoridaceaepollen(
Zavada1987).We have^th
erefo
rein
cludedcom
pari
sons
withbotht
he Ari
stolochi
ac
eae-Lactoridaceaea
ndLact
orid
aceae clades
.a The fossil- b
as
ed ageconcernsthe stem- g
roup and notthe crown-gr
oup as forremainingtaxa(Magallón and Sanderson 2001).b We^
ar
e unabl
e^
toc
ompare
the fos
sil- b
as
ede
stim
atefor^ the les
s inclu
sive
ta
xon^
w
itha
mol
ecul
arestimatesince
the taxon was
notincl
uded in the molecula
rdata
set.c Magal
lónet^al.(
1999)
indicated anAptian (124.5–112 Ma)
occurrence^
forPopulus potomace
ns
is,butthe orig
in
al^documentation ofthis taxonindi
cate
dtheAlbi
an(^112- 97 M
a; Doyle
and Hic
key1976).