2
Ica onyc e is index (holo ype; PU 18150) om he G een Ri e Fo ma ion, Wyoming.
CONTENTS
Abs ac ....................................................................... 4
In oduc ion .................................................................... 5
Rela ionships and Classi ica ion o Eocene Ba s: A His o ical O e iew ........... 12
Rela ionships Among Ex an Lineages o Ba s .................................. 31
Goals o he Cu en S udy ................................................... 40
Ma e ials and Me hods ......................................................... 40
Taxonomic Sampling and Monophyly .......................................... 40
Ou g oups .................................................................. 42
The Da a Se ................................................................ 46
Cha ac e s Examined in Fossil Ba s ............................................ 48
Skull and Den i ion ........................................................ 48
An e io Axial Skele on .................................................... 64
Pec o al Gi dle ............................................................ 70
Fo elimb ................................................................. 77
Pos e io Axial Skele on and Pel is .......................................... 81
Hindlimb ................................................................. 84
Comple eness ............................................................... 85
Me hods o Phylogene ic Analysis ............................................. 86
Analysis o Cha ac e T ans o ma ions ......................................... 87
Phylogene ic Resul s ........................................................... 88
Resul s o Analyses .......................................................... 88
Analysis 1: All Cha ac e s, Fossil Taxa Excluded .............................. 88
Analysis 2: All Cha ac e s, All Taxa ......................................... 89
Analysis 3: All Taxa, Cha ac e s Limi ed o Those Sco ed in Fossil Fo ms ....... 92
Summa y ................................................................... 94
Cha ac e E olu ion in Ea ly Chi op e ans ........................................ 94
Cha ac e T ans o ma ions Associa ed wi h Basal Nodes .......................... 94
Fea u es Diagnosing he Mic ochi op e an C own G oup ......................... 99
Cha ac e T ans o ma ions a Basal Nodes: A Func ional Pe spec i e ............. 101
E olu ion o Echoloca ion and Fo aging S a egies ................................ 107
Timing o he O igin o Fligh and Echoloca ion ............................... 108
Vision and he E olu ion O Echoloca ion ..................................... 111
Fo aging Ecology o Eocene Ba s ............................................ 119
E olu ion o Fo aging S a egies: A Phylogene ic Pe spec i e .................... 129
Classi ica ion o Eocene Ba s ................................................... 133
Conclusions and Summa y ..................................................... 139
Acknowledgmen s ............................................................ 143
Re e ences ................................................................... 143
Appendix 1: Specimens Examined .............................................. 170
Appendix 2; Cha ac e Desc ip ions ............................................. 172
Appendix 3: Da a Ma ix ...................................................... 180
4 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
ABSTRACT
The Eocene ossil eco d o ba s (Chi op e a)
includes ou gene a known om ela i ely com-
ple e skele ons: Ica onyc e is,A chaeonyc e is,
Hassianyc e is, and Palaeochi op e yx. Phyloge-
ne ic ela ionships o hese axa o each o he and
o ex an lineages o ba s we e in es iga ed in a
pa simony analysis o 195 mo phological cha -
ac e s, 12 DNA es ic ion si e cha ac e s, and
one cha ac e based on he numbe o R-1 andem
epea s in he m DNA d-loop egion. Resul s in-
dica e ha Ica onyc e is,A chaeonyc e is,Has-
sianyc e is, and Palaeochi op e yx ep esen a se-
ies o consecu i e sis e - axa o ex an mic ochi-
op e an ba s. This conclusion s ands in con as
o p e ious sugges ions ha hese ossil o ms
ep esen ei he a p imi i e g ade ances al o bo h
Megachi op e a and Mic ochi op e a (e.g., Eochi-
op e a) o a sepa a e clade wi hin Mic ochi op-
e a (e.g., Palaeochi op e ygoidea). A new highe -
le el classi ica ion is p oposed o be e e lec hy-
po hesized ela ionships among Eocene ossil ba s
and ex an axa. C i ical ea u es o his classi i-
ca ion include es ic ion o Mic ochi op e a o
he smalles clade ha includes all ex an ba s ha
use sophis ica ed echoloca ion (Emballonu idae
1
Yinochi op e a
1
Yangochi op e a), and o mal
ecogni ion o wo mo e inclusi e clades ha en-
compass Mic ochi op e a plus he ou ossil gen-
e a.
Compa isons o esul s o sepa a e phylogene -
ic analyses including and subsequen ly excluding
he ossil axa indica e ha inclusion o he ossils
changes he esul s in wo ways: (1) al e ing pe -
cei ed ela ionships among ex an o ms a a ew
poo ly suppo ed nodes; and (2) educing pe -
cei ed suppo o some nodes nea he base o
he ee. Inclusion o he ossils a ec s some cha -
ac e pola i ies (hence sligh ly changing ee o-
pology), and also changes he le els a which
ans o ma ions appea o apply (hence al e ing
pe cei ed suppo o some clades). Resul s o an
addi ional phylogene ic analysis in which so - is-
sue and molecula cha ac e s we e excluded om
conside a ion indica e ha hese cha ac e s a e
c i ical o de e mina ion o ela ionships among
ex an lineages.
Ou phylogeny p o ides a basis o e alua ing
p e ious hypo heses on he e olu ion o ligh ,
echoloca ion, and o aging s a egies. We p opose
ha ligh e ol ed be o e echoloca ion, and ha
he i s ba s used ision o o ien a ion in hei
a bo eal/ae ial en i onmen . The e olu ion o
ligh was ollowed by he o igin o low-du y-cy-
cle la yngeal echoloca ion in ea ly membe s o
he mic ochi op e an lineage. This sys em was
mos likely simple a i s , pe mi ing o ien a ion
and obs acle de ec ion bu no de ec ion o ack-
ing o ai bo ne p ey. Owing o he mechanical
coupling o en ila ion and ligh , he ene gy cos s
o echoloca ion o lying ba s we e ela i ely low.
In con as , he bene i s o ae ial insec i o y we e
subs an ial, and a mo e sophis ica ed low-du y-cy-
cle echoloca ion sys em capable o de ec ing,
acking, and assessing ai bo ne p ey subsequen -
ly e ol ed apidly. The need o an inc easingly
de i ed audi o y sys em, oge he wi h limi s on
body size imposed by he mechanics o ligh ,
echoloca ion, and p ey cap u e, may ha e esul ed
in educ ion and simpli ica ion o he isual sys em
as echoloca ion became inc easingly impo an .
Ou analysis con i ms p e ious sugges ions ha
Ica onyc e is,A chaeonyc e is,Hassianyc e is,
and Palaeochi op e yx used echoloca ion. Fo ag-
ing s a egies o hese o ms we e econs uc ed
based on pos c anial os eology and wing o m,
cochlea size, and s omach con en s. In he con-
ex o ou phylogeny, we sugges ha o aging
beha io in he mic ochi op e an lineage e ol ed
in a se ies o s eps: (1) gleaning ood objec s du -
ing sho ligh s om a pe ch using ision o o i-
en a ion and obs acle de ec ion; p ey de ec ion by
passi e means, including ision and/o lis ening
o p ey-gene a ed sounds (no known examples in
ossil eco d); (2) gleaning s a iona y p ey om a
pe ch using echoloca ion and ision o o ien a-
ion and obs acle de ec ion; p ey de ec ion by pas-
si e means (Ica onyc e is,A chaeonyc e is); (3)
pe ch hun ing o bo h s a iona y and lying p ey
using echoloca ion and ision o o ien a ion and
obs acle de ec ion; p ey de ec ion and acking us-
ing echoloca ion o lying p ey and passi e
means o s a iona y p ey (no known example, al-
hough Ica onyc e is and/o A chaeonyc e is may
ha e done his a imes); (4) combined pe ch hun -
ing and con inuous ae ial hawking using echolo-
ca ion and ision o o ien a ion and obs acle de-
ec ion; p ey de ec ion and acking using echo-
loca ion o lying p ey and passi e means o s a-
iona y p ey; calca -suppo ed u opa agium used
o p ey cap u e (common ances o o Hassianyc-
e is and Palaeochi op e yx; e ained in Palaeo-
chi op e yx); (5) exclusi e eliance on con inuous
ae ial hawking using echoloca ion and ision o
o ien a ion and obs acle de ec ion; p ey de ec ion
and acking using echoloca ion (Hassianyc e is;
common ances o o Mic ochi op e a). The an-
si ion o using echoloca ion o de ec and ack
p ey would ha e been di icul in clu e ed en-
ionmen s owing o in e e ence p oduced by
mul iple e u ning echoes. We he e o e p opose
ha his ansi ion occu ed in ba s ha o aged in
o es gaps and along he edges o lakes and i e s
1998 5SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
in si ua ions whe e po en ial pe ch si es we e ad-
jacen o ela i ely clu e - ee open spaces. Ae ial
hawking using echoloca ion o de ec , ack, and
e alu e p ey was appa en ly he p imi i e o aging
s a egy o Mic ochi op e a. This implies ha
gleaning, passi e p ey de ec ion, and pe ch hun -
ing among ex an mic ochi op e ans a e second-
a ily de i ed specializa ions a he han e en ions
o p imi i e habi s. Each o hese habi s has ap-
pa en ly e ol ed mul iple imes.
The e olu ion o con inuous ae ial hawking
may ha e been he ‘‘key inno a ion’’ esponsible
o he bu s o di e si ica ion in mic ochi op e an
ba s ha occu ed du ing he Eocene. Fossils e-
e able o six majo ex an lineages a e known
om Middle–La e Eocene deposi s, and econ-
s uc ion o ghos lineages leads o he conclusion
ha a leas se en mo e ex an lineages we e min-
imally p esen by he end o he Eocene.
Only ex ensi e phylogene ic analysis, based on as many sui es o cha ac e s as pos-
sible, and ca ied ou in conjunc ion wi h adap a ional and ae odynamic s udies, can
o m he basis o econs uc ion o e olu iona y change. Padian (1987: 19)
The beha io o ossil animals and he e olu ion o ligh will p obably always be
a subjec o con en ion, and al hough we may ne e know o ce ain i we ha e
ound he igh answe s, we can always dis inguish he possible om he impossible,
he p obable om he imp obable. No be g (1990: 268)
INTRODUCTION
Ba s i s appea in he ossil eco d in Ea -
ly Eocene deposi s o No h Ame ica, Eu-
ope, A ica, and Aus alia ( able 1). The di-
e si y o Eocene ba s is ema kable—24
gene a a e cu en ly ecognized, and new
species a e desc ibed almos e e y yea . O
he eigh ba gene a ha make hei i s ap-
pea ance in he Ea ly Eocene, almos hal a e
known om spec acula , nea ly comple e
skele ons: Ica onyc e is,A chaeonyc e is,
Hassianyc e is, and Palaeochi op e yx.
Much o wha is known (o hypo hesized)
abou he ea ly e olu ion o ba s is based on
s udies o hese axa (e.g., Jepsen, 1966,
1970; Russell and Sige´, 1970; Rich e and
S o ch, 1980; No acek 1985a, 1987; Habe -
se ze and S o ch, 1987, 1988, 1989; Habe -
se ze e al., 1989, 1992, 1994; No be g,
1989; S o ch, 1989).
Ica onyc e is was desc ibed by Jepsen
(1966) on he basis o a beau i ully p ese ed
skele on om he Ea ly Eocene G een Ri e
Fo ma ion o Fossil Basin, Wyoming (see
F on ispiece). A small ba wi h a long ail,
Ica onyc e is had a wingspan o app oxi-
ma ely 30 cm and p obably weighed 10–16
g (Habe se ze and S o ch, 1987; No be g,
1989). The holo ype was collec ed om beds
ha a e Middle Wasa chian in age, app oxi-
ma ely 53 Ma (Woodbu ne, 1987; Woodbu -
ne and Swishe , 1995). A leas h ee addi-
ional skele ons o Ica onyc e is we e sub-
sequen ly disco e ed in he same deposi s
(No acek, 1985a, 1987; Habe se ze and
S o ch, 1987; see appendix 1). Isola ed ee h
om Cla k o kian deposi s in No h Ame ica
we e e e ed o c . Ica onyc e is sp. by Gin-
ge ich (1987), hus po en ially ex ending he
ange o he genus o he La e Paleocene ( ig.
1). O he agmen a y ma e ial sugges s ha
Ica onyc e is may ha e pe sis ed in No h
Ame ica un il he Ga dene bu ian (McKen-
na and Bell, 1997), an ea ly B idge ian in-
e al ha ended app oxima ely 50 Ma
(Woodbu ne, 1987; Woodbu ne and Swishe ,
1995). Only one species, Ica onyc e is index
Jepsen, 1966, is cu en ly ecognized in
No h Ame ica. Howe e , he geog aphic
ange o Ica onyc e is may ha e ex ended
beyond No h Ame ica o Eu ope. Russell e
al. (1973) ecognized a new species (Ica on-
yc e is?menui) based on a la ge collec ion
o jaw agmen s and isola ed ee h om Ea -
ly Eocene ( ig. 1; MP
1
8–9) deposi s in
1
Mammal Paleogene (MP) e e ence le els a e bio-
s a ig aphic in e als used o subdi ide he Paleogene
o Eu ope. These uni s, which a e o unspeci ied du a-
ion, a e based on e es ial mammalian aunas. The MP
sys em is based on he assump ion ha he ossil eco d
6 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
F ance. Howe e , he a ini ies o his ma e-
ial emain ques ionable because o he ag-
men a y na u e o a ailable specimens and
lack o any diagnos ic apomo phies sha ed
wi h he No h Ame ican o m.
Palaeochi op e yx,A chaeonyc e is, and
Hassianyc e is a e known p incipally om
he amous Ea ly/Middle Eocene ‘‘G ube
Messel’’ deposi s (MP 11) nea Da ms ad ,
Ge many. Hund eds o comple e and pa ial
skele ons o ba s ha e been ound a Messel,
many p ese ed wi h s omach con en s indi-
ca ing ha hey we e insec i o es ha had
been o aging success ully jus p io o dea h
(Smi h e al., 1979; Rich e and S o ch, 1980;
Rich e , 1987; Habe se ze e al., 1992,
1994). This unusual concen a ion o ossil
ba s ha show no e idence o cause o dea h
may ha e esul ed om elease o poisonous
gas ha o e came ba s o aging o e he su -
ace o Lake Messel (Habe se ze and S o ch,
1988; Habe se ze e al., 1992). Ano he pos-
sibili y is ha he ba s died as a esul o
poisoning by oxic alkaloids om a blue-
g een algal bloom (see Pybus e al. [1986]
o a mode n example).
Palaeochi op e yx is by a he mos com-
mon ba a Messel, accoun ing o almos
75% o all ba inds (Habe se ze and S o ch,
1989; Habe se ze e al., 1992). Mo e han
50 skele ons o Palaeochi op e yx a e known
om Messel, and wo species a e cu en ly
consis s o con inuously e ol ing lineages; he e o e,
sepa a ion be ween ances o and descendan species (and
he empo al ange o each species) is comple ely a bi-
a y. In an e o o p e en his subjec i e elemen om
con ounding bios a ig aphy, he bounda ies o he MP
uni s we e no de ined by he i s and las appea ance
o axa; acco dingly, hey a e no ue bios a ig aphic
zones (Schmid -Ki le , 1987). Each MP le el has a e -
e ence locali y whose auna unc ions as he ype o he
e e ence le el (Schmid -Ki le , 1987). The me hod
used o o de MP le els no co ela ed o ma ine s ages
is no clea . I appea s o be based on aunal simila i y
and he e olu iona y s age o he lineages wi hin he
espec i e aunas (Russel e al., 1982; Schmid -Ki le ,
1987). O de ing o he MP le els by adiome ic ages,
magne os a ig aphy, and he supe posi ion o s a a has
no been possible (Schmid -Ki le , 1987). Al hough he
use o aunal e olu iona y s age in de eloping he se-
quence o MP le els is subjec i e and o ques ionable
alidi y, i is cu en ly he only op ion o placing Eu-
opean ossil mammals in geologic ime.
ecognized: Palaeochi op e yx upaiodon
Re illiod, 1917 ( ig. 2) and Palaeochi op e-
yx spiegeli Re illiod, 1917 (Smi h and
S o ch, 1981; Habe se ze and S o ch, 1987).
P. upaiodon was he smalle o he wo spe-
cies, p obably weighing 7–10 g and ha ing
an es ima ed wingspan o 24–28 cm (Habe -
se ze and S o ch, 1987; No be g, 1989). In
con as , P.spiegeli had an es ima ed wing-
span o 26–30 cm and p obably weighed 10–
18 g (Habe se ze and S o ch, 1987; No -
be g, 1989). In addi ion o he Messel ma-
e ial, isola ed agmen s e e able o Palaeo-
chi op e yx ha e also been epo ed om
Spa nacian, Cuisian, Lu e ian, and Ba onian
(MP 10, 11–13, 16) deposi s om elsewhe e
in Eu ope (Russell e al., 1973, 1982; Sa age
and Russell, 1983).
A chaeonyc e is is known om a leas six
skele ons om Messel, and wo species a e
cu en ly ecognized based on his ma e ial:
A chaeonyc e is igonodon Re illiod, 1917b
( ig. 3) and A.pollex S o ch and Habe se ze ,
1988 (Russell and Sige´, 1970; Smi h and
S o ch, 1981; Habe se ze and S o ch, 1987;
S o ch and Habe se ze , 1988). A. igonodon
p obably had a wingspan o 32–37 cm and
may ha e weighed 17–27 g (Habe se ze and
S o ch, 1987; No be g, 1989). A.pollex was
sligh ly la ge , wi h an es ima ed body
weigh o 30–35 g (S o ch and Habse se ze ,
1988). In addi ion o he Messel ma e ial,
agmen a y specimens e e able o A chaeo-
nyc e is ha e been epo ed om Spa nacian,
Cuisian, and possibly Ba onian deposi s in
F ance (MP 8–13, 16; Russell e al, 1973,
1982; Godino , 1981; Sa age and Russell,
1983; Schmid -Ki le , 1987). This ma e ial
includes a leas one addi ional species, A -
chaeonyc e is b ailloni Russell, Louis, and
Sa age, 1973 (MP 8–9; Schmid -Ki le ,
1987). Ano he species, A chaeonyc e is e-
illiodi Russell and Sige´, 1970, is based on
a pa ial den i ion om Messel. Always con-
side ed somewha p oblema ic, his o m was
ans e ed o Hassianyc e ididae as Hassi-
anyc e is? e illiodi by Habe se ze and
S o ch (1987). Smi h and Russell (1992) sub-
sequen ly epo ed disco e y o a skull o e-
illiodi and con i med ha his species
should be placed in Hassianyc e is.
Hassianyc e is is known om a leas 20
skele ons om Messel. Th ee species a e
1998 7SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
cu en ly ecognized based on his ma e ial:
Hassianyc e is messelensis Smi h and
S o ch, 1981; Hassianyc e is magna Smi h
and S o ch, 1981; and Hassianyc e is e il-
liodi Russell and Sige´, 1970 (Smi h and
S o ch, 1981; Habe se ze and S o ch, 1988;
Smi h and Russell, 1992). Wingspans and
weigh s ha e been es ima ed o only hose
species known om ela i ely comple e skel-
e ons (i.e., H.messelensis and H.magna).H.
messelensis p obably had a wingspan o 35–
40 cm and weighed 25–45 g (Habe se ze
and S o ch, 1987; No be g, 1989). H.magna
had a wingspan o 45–50 cm and may ha e
weighed 65 g, making i he la ges known
Eocene ba (Habe se ze and S o ch, 1987;
No be g, 1989). A ou h species, Hassi-
anyc e is joeli Smi h and Russell, 1992, is
known om a pa ial den a y wi h ee h om
uppe Yp esian (Ea ly Eocene) deposi s in
Belgium (Smi h and Russell, 1992).
RELATIONSHIPS AND CLASSIFICATION
OF EOCENE BATS:
A HISTORICAL OVERVIEW
Phylogene ic ela ionships among Ica on-
yc e is,Palaeochi op e yx,A chaeonyc e is,
and Hassianyc e is ha e been he subjec o
conside able deba e, as ha e hei ela ion-
ships o ex an lineages. I has long been
hough ha many Eocene ossils ep esen
ea ly membe s o ex an mic ochi op e an
amilies, bu Ica onyc e is,Palaeochi op e-
yx,A chaeonyc e is, and Hassianyc e is (o -
en e e ed o as he ‘‘a chaic’’ Eocene ba s)
ha e emained enigma ic. Mos wo ke s ha e
ega ded some o all o hese o ms as ep-
esen a i es o an ea ly g ade o chi op e an
e olu ion, bu opinions ha e di e ed con-
ce ning hei ela ionships o each o he ,
o he Eocene axa, Megachi op e a, Mic o-
chi op e a, and o ex an mic ochi op e an
supe amilies. This unce ain y has been
e lec ed in classi ica ions, which ha e a -
ied conside ably o e he las ew decades.
Because ideas conce ning ela ionships o
Ica onyc e is,Palaeochi op e yx,A chaeo-
nyc e is, and Hassianyc e is de eloped in he
con ex o a ich body o li e a u e on ea ly
Te ia y ba s, we e iew he e (in ch onolog-
ical o de ) he his o y o classi ica ion and
phylogene ic hypo heses ega ding all Eo-
cene chi op e an axa, no jus he ou gen-
e a examined in he p esen s udy. The ax-
onomy and spelling o names a e hose o
he o iginal au ho (s) o each s udy. Ques ion
ma ks associa ed wi h axonomic names o
indica e unce ain y (e.g., Ica onyc e is?
menui) also e lec he usage o he o iginal
au ho (s).
Twen y- ou gene a o ba s a e now ec-
ognized om he Eocene, al hough some
we e o iginally desc ibed based on ma e ial
om younge deposi s ( able 1). Se e al ea -
ly Te ia y ba ossils we e epo ed in he
li e a u e p io o 1875, bu mos we e iso-
la ed ee h o agmen s o jaws e e ed o
ei he Chi op e a, Rhinolophus,o Vespe i-
lio, he la e being a was e-baske axon ha
once included mos ba s. Summa ies o he
his o y o ossil ba disco e ies p io o 1875
we e p o ided by Re illiod (1922) and Le-
gend e and Sige´ (1983).
Mos mode n ba classi ica ions a e based
la gely upon ha o Dobson (1875: 345),
who was he i s o p o ide a comp ehensi e
classi ica ion o ex an ba s ‘‘a anged ac-
co ding o hei na u al a ini ies.’’ Dobson
(1875) b ie ly conside ed he o igins o Chi-
op e a, al hough he did no explici ly discuss
any o he ossil ma e ial known a he ime.
Accompanying Dobsons (1875) accoun was
a igu e ‘‘illus a ing he a ini ies o he am-
ilies and gene a o Chi op e a, and p obable
lines o descen om ances al o ms . . .’’
( ig. 5). In his diag am, Dobson (1875) used
Palaeochi op e a as a name o he g oup o
la gely unknown ossil ba s ha he p esumed
we e ances al o all ex an ba lineages.
Schlosse (1887) named Pseudo hinolo-
phus and Vespe ilia us based on ma e ial
om he La e Eocene o Ea ly Oligocene
Que cy phospho i e deposi s in F ance (cu -
en ly e e ed o MP 16–23; C oche e al.,
1981; Sige´ and Legend e, 1983; Schmid -
Ki le , 1987). D awing on ex ensi e com-
pa isons wi h ex an gene a, Schlosse (1887)
sugges ed a close ela ionship be ween Pseu-
do hinolophus and Rhinolophus and no ed
se e al poin s o simila i y be ween Vespe -
ilia us, species o Vespe ilio (many now
placed in o he gene a and e en o he ami-
lies; see below), and Taphozous. He subse-
quen ly concluded ha Pseudo hinolophus
should be g ouped wi h hinolophids and
8 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 1. Cha showing geologic ime scale, s anda d geologic ages based on ma ine s a a, and he
co ela ion o land mammal ages and le els. Da a complied om Fahlbusch (1976), Russell e al. (1982),
Sa age and Russell (1983), Be gg en e al. (1995), Woodbu ne and Swishe (1995), and McKenna and
Bell (1997). Co ela ions o MP le els o ma ine s a a o known age a e based on he ollowing: MP
7 (Hooke , 1991; Woodbu ne and Swishe , 1995); MP 10, 13, 14, and 15 (Russell e al., 1982; Aub y,
1983); and MP 21 (Hooke , 1992). Abb e ia ions: EU., Eu opean; Ma., milleannus, millions o yea s
be o e p esen ; N.A., No h Ame ican. Fo a de ini ion o MP le els, see oo no e 1 on page 5.
a
Ages acco ding o Fahlbusch (1976).
b
Ages acco ding o Sa age and Russell (1983).
c
These MP e e ence le els a e no co ela ed o ma ine s a a; ins ead, hey a e placed in an app ox-
ima e ch onological o de (based on Russell e al. [1982] and Schmid -Ki le [1987]) be ween MP
le els ha can be co ela ed o ma ine s a a.
Vespe ilia us wi h espe ilionids. Schlosse
(1887) also desc ibed agmen a y ma e ial
ha he e e ed o Rhinolophus sp. and o a
la e Te ia y species o Rhinolophus de-
sc ibed by Filhol (1872). Schlosse (1887)
also discussed an isola ed hume us ha he
sugges ed was simila o ha o hinolophids.
Wei ho e (1887) discussed new ma e ial
o Pseudo hinolophus and sugges ed modi-
ica ions o Schlosse ’s (1887) desc ip ion
and alloca ion o his axon. Based on com-
pa isons wi h ex an ma e ial, Wei ho e
(1887) concluded ha Pseudo hinolophus
was mo e closely ela ed o Phyllo hina (
5
Hipposide os) han o Rhinolophus.He e-
e ed he isola ed hume us desc ibed by
Schlosse (1887) o ‘‘Taphozous(?),’’ no ing
ha i was clea ly di e en om he ypical
hinolophid o m. Wei ho e (1887) also de-
sc ibed se e al new axa based on c anioden-
al ma e ial om he Que cy deposi s, in-
cluding Alas o (which he associa ed wi h
Pseudo hinolophus and Phyllo hina), Nec-
oman is (which he concluded was ela ed o
1998 9SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 2. Palaeochi op e yx upaiodon (SMF ME 10) om Messel, Ge many. F om Habe se ze and
S o ch (1987: ig. 2). Pho o by E. Pan ak (Senkenbe gmuseum).
Phyllos omidae), and a ossil species e e ed
o ‘‘Rhinolophus(?).’’ Recen collec ions
ha e sugges ed ha Nec oman is is es ic ed
o he MP 17 auna (Sige´ and Legend e,
1983; Schmid -Ki le , 1987).
Winge (1892 [ ansla ed in o English in
Winge, 1941]) discussed he s a us o Pseu-
do hinolophus and Alas o , no ing ha al-
hough some au ho s had ea ed hese as
gene a sepa a e om Phyllo hina (
5
Hip-
poside os), ew had no ed any signi ican di -
e ences. He concluded ha Palaeophyllo-
pho a clea ly belonged o he same g oup as
Pseudo hinolophus, al hough Palaeophyllo-
pho a is somewha mo e specialized. Winge
(1892) also discussed an undesc ibed, almos
comple e skull o Vespe ilia us om he
collec ions a Copenhagen. Se e al ea u es
o his specimen led him o conclude ha
Vespe ilia us is a ela i ely specialized
membe o Emballonu idae, closely ela ed
o Taphozous. He no ed ha his conclusion
was no su p ising gi en he axonomic his-
o y o species p e iously compa ed wi h
Vespe ilia us. Ci ing Schlosse s (1887)
o iginal desc ip ion o Vespe ilia us, Winge
(1892) obse ed ha he ‘‘ espe ilionid’’
ha Schlosse hough was mos simila o
10 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 3. A chaeonyc e is igonodon (SMF 80/1379) om Messel, Ge many. F om Habe se ze and
S o ch (1987: ig. 3). Pho o by E. Pan ak (Senkenbe gmuseum).
Vespe ilia us—Vespe ilio alec o—was
conside ed by Dobson (1878) o be conspe-
ci ic wi h Emballonu a mon icola. Winge
u he no ed ha he isola ed hume us e-
e ed o ‘‘Taphozous(?)’’ by Wei ho e
(1887) undoub edly also ep esen s Vespe -
ilia us.
Meschinelli (1903) desc ibed A chaeop-
e opus based on a skele on wi h a poo ly
p ese ed skull om la e Oligocene deposi s
1998 17SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Palaeochi op e ygidae (including Palaeochi-
op e yx and ?Cecilionyc e is). Bo h o hese
amilies we e le in ‘‘Supe amily unce -
ain’’ wi hin Mic ochi op e a (Rome , 1966:
382). Pa adoxonyc e is was e e ed o Chi-
op e a ince ae sedis (Rome , 1966).
Jepsen (1966) desc ibed Ica onyc e is
om a single, beau i ully p ese ed skele on
om Fossil Basin in he G een Ri e Fo -
ma ion o Wyoming. He placed his o m in
he new amily Ica onyc e idae wi hin Mi-
c ochi op e a. Jepsen discussed o he pu a-
i e ba s om ea ly Te ia y deposi s in
No h Ame ica, bu did no make any de-
ailed compa isons wi h Eocene ba s om
Eu ope.
Sige´ (1968) desc ibed he ba s o an Ea ly
Miocene auna om Bouzigue, F ance, in-
cluding a new species o Pseudo hinolophus.
He e iewed he con en s and s a ig aphic
ange o Pseudo hinolophus and p esen ed a
e ised diagnosis o his axon, which he
conside ed o ep esen a subgenus o Hip-
poside os. He sugges ed se e al possibili ies
o ela ionships among Pseudo hinolophus,
B achyhipposide os (a subgenus known om
Oligocene and Miocene deposi s), and mod-
e n species o Hipposide os, bu he eached
no de ini i e conclusions o he han ha
hese o ms a e e y closely ela ed.
Jepsen (1970) published de ailed s e eo-
pho og aphs o Ica onyc e is in an essay
conce ning he e olu ion o ba s and pow-
e ed ligh . Howe e , he ex o his publi-
ca ion con ained li le in o ma ion on Ica on-
yc e is ha was no a ailable in he o iginal
desc ip ion (Jepsen, 1966), and no compa i-
sons we e d awn be ween Ica onyc e is and
Eocene ba s om Eu ope. Conce ning he e-
la ionships o Ica onyc e is o o he ba s,
Jepsen (1970: 40) no ed only ha ‘‘Ica on-
yc e is index as a species may ha e been di-
ec ly ances al o all o o some li ing mi-
c oba s o megaba s o o none o ou con-
empo a y chi op s.’’
In a sho pape on he classi ica ion o
ba s, Koopman and Jones (1970) ecognized
h ee amilies o Eocene ba s: Ica onyc e i-
dae (including only Ica onyc e is), A chaeo-
nyc e idae (A chaeonyc e is), and Palaeochi-
op e ygidae (Palaeochi op e yx and Ceci-
lionyc e is, he la e o which was placed in-
ce ae sedis). These amilies we e lis ed
unde he heading ‘‘Fossil Chi op e a o un-
ce ain s a us’’ (Koopman and Jones, 1970:
28). Paleunyc e is and Pa adoxonyc e is
we e e e ed o Chi op e a ince ae sedis,
and se e al Eocene axa we e placed in mod-
e n amilies: Vespe ilia us in Emballonu i-
dae, Nec oman is in Megade ma idae, and
S ehlinia in Vespe ilionidae (Koopman and
Jones, 1970). Palaeophyllopho a and Pa a-
phyllopho a we e placed in hei own ibe
Palaeophyllopho ini in Hipposide inae
(Koopman and Jones, 1970). A chaeop e o-
pus was placed in A chaeop e opodinae
wi hin Megachi op e a.
Slaugh e (1970) discussed e olu ion o
he chi op e an den i ion, d awing conclu-
sions abou he a ini ies o many ossil axa
based on hei den al mo phology. Slaugh e
(1970: 59) obse ed ha he den i ion o Ica-
onyc e is is ‘‘mo e like he ances al con-
di ion han any o he known chi op e an.’’
He wen on o no e ha Cecilionyc e is,Pa-
laeochi op e yx, and A chaeonyc e is ha e
sligh ly mo e de i ed den i ions, bu did no
associa e hem wi h any ex an lineages. He
obse ed ha Nec oman is is clea ly a me-
gade ma id, al hough somewha less de i ed
han ex an membe s o he amily. Slaugh e
placed Pseudo hinolophus on he lineage
leading o mode n Hipposide os, and ecog-
nized Palaeophyllopho inae as close ela i es
o hipposide ines. The ‘‘Eocene Rhinolo-
phus’’ we e placed a he base o he lineage
leading o mode n Rhinolophus, palaeophyl-
lopho ines, and hipposide ines (Slaugh e ,
1970: ig. 5), wi h Nyc e is shown as he sis-
e -g oup o his hinolophid clade. He also
ecognized Vespe ilia us as he oldes mem-
be o Emballonu idae, no ing ha he p o-
o ypic den i ion o he supe amily Embal-
lonu oidea (which he de ined as including
noc ilionids) would be like ha o Vespe i-
lia us excep o he o m o he hypocone.
Finally, Slaugh e no ed ha S ehlinia ap-
pea s o be in e media e be ween ances al
espe ilionids and he lineage leading o all
espe ilionids excep miniop e ines and mu-
inines.
Russell and Sige´ (1970) we e among he
i s o p esen de ailed compa isons o he
Messel ba s wi h ex an axa. Thei s udy
concluded wi h an upda ed classi ica ion o
he a chaic Eocene ba s and emended diag-
18 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 6. Russell, Louis, and Sa age’s hypo hesis o phylogeny and di e gence imes o Eocene ba s
( ed awn om Russell e al., 1973: ig. 11). The o iginal cap ion o his igu e ead ‘‘Hypo he ical
ela ionships be ween he ea lies Eu opean ba s.’’
noses o each axon. They placed he a chaic
Eocene gene a in a new supe amily Palaeo-
chi op e ygoidea wi hin Mic ochi op e a.
Two amilies we e ecognized wi hin his
g oup: Palaeochi op e ygidae and Ica onyc-
e ididae. Ica onyc e ididae was de ined as
including Ica onyc e is and possibly A -
chaeop e opus. In con as , Palaeochi op e -
ygidae was di ided in o wo sub amilies,
Palaeochi op e yginae (including Palaeochi-
op e yx and Cecilionyc e is) and A chaeo-
nyc e idinae (con aining only A chaeonyc-
e is).
Sige´ (1971) desc ibed an isola ed o elimb
om S ampien deposi s in F ance (MP 21–
24; Russell e al., 1982; Schmid -Ki le ,
1987) and e e ed i o Tada ida sp. Al-
hough his specimen may ep esen Cu ie -
imops, lack o associa ed den al ma e ial has
p e en ed a con i med iden i ica ion. Hand
(1990) e e ed his specimen o Tada idinae
inde . (Tada idinae was named by Legend e
[1984] in a e iew o ex an molossids).
Russell e al. (1973) desc ibed agmen-
a y den al emains om he Ea ly Eocene o
F ance (MP 8–9; Schmid -Ki le , 1987) ha
hey e e ed o new species o Ica onyc e -
is? and A chaeonyc e is as well as a new ge-
nus, Ageina. Based on den al mo phology,
hey hypo hesized ha Ica onyc e is?menui
was closely ela ed o Palaeochi op e yx.A -
chaeonyc e is was ecognized as a dis inc
ye ela ed lineage, while Ageina was placed
on a dis an b anch o unce ain a ini ies
( ig. 6; Russell e al., 1973: ig. 11). Al-
hough classi ica ion was no explici ly dis-
cussed, hey placed Ica onyc e is?menui in
Palaeochi op e ygoidea: Ica onyc e idae?,
1998 19SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 7. Smi h’s hypo heses o ba ela ionships ( ed awn om Smi h, 1976: ig. 1). The o iginal
cap ion o his igu e ead ‘‘A, cladog am o he gene ally accep ed iew o chi op e an phylogeny
wi h he Mic ochi op e a de ied om a common emballonu idlike ances y. B, an al e na i e p oposal
o chi op e an e olu ion wi h se e al mic ochi op e an lineages being de i ied, independen ly, om a
wo ld-wide paleochi op e an g ade and he Megachi op e a (P e opodidae) de i ed ei he sepa a ely om
an insec i o ous s ock o om he paleochi op e an g ade. a, Emballonu oidea; b, Rhinolophoidea; c,
Phyllos oma oidea; d, Vespe ilionoidea.’’ No e ha Smi h (1976) used ‘‘Myzapodidae’’ ins ead o ‘‘My-
zopodidae,’’ and ‘‘Phyllos oma oidea’’ ins ead o ‘‘Phyllos omoidea’’ (
5
Noc ilionoidea).
A chaeonyc e is b ailloni in A chaeonyc e -
inae ( amily no speci ied), and Ageina in
‘‘Family unce ain’’ (Russell e al., 1973: 35).
Sige´ (1974) e iewed he a ailable ma e-
ial o S ehlinia and compa ed i o o he os-
sil and ex an axa. Based on his analysis,
Sige´ ecognized Nyc e obius,Paleunyc e is,
and Re illiodia as junio synonyms o S eh-
20 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
linia, and concluded ha S ehlinia was bes
placed in he amily Ke i oulidae wi hin
Vespe ilionoidea. He compa ed S ehlinia
wi h ‘‘les pale´ochi op e´ ygoı¨de´s’’ bu did no
commen on con en s o classi ica ion o his
g oup.
Smi h (1976) discussed Eocene ossil ba s
in a e iew o phylogene ic ela ionships and
pa e ns o e olu iona y adia ion o ex an
ba s ( ig. 7). Appa en ly d awing on Dob-
son’s (1875) concep , Smi h (1976: 52–53)
used he in o mal name ‘‘Paleochi op e a’’ o
collec i ely e e o Ica onyc e is,Palaeo-
chi op e yx,A chaeonyc e is,Cecilionyc e -
is, and Ageina. He pe cei ed his g oup as
ep esen ing ‘‘a wo ld-wide paleochi op e an
g ade in ea ly o middle Eocene imes . . .
[ ha ] was p imi i e and gene alized in mos
espec s.’’ He sugges ed ha ex an Mic o-
chi op e a could be easily de i ed om he
paleochi op e an g ade, and ha his di e -
gence p obably occu ed in he Paleocene.
He clea ly did no hink ha any o he Eo-
cene axa lis ed abo e we e ances al o ex-
an mic ochi op e ans. Smi h u he sug-
ges ed ha A chaeop e opus was p obably
also de i ed om he paleochi op e an g ade,
al hough he ques ioned he megachi op e an
a ini ies o his o m. Smi h (1976: 54) ob-
se ed ha ‘‘The dis inc ness and ma ked de-
pa u e o megachi op e an den i ion om
ha o he Mic ochi op e a, as well as om
known Te ia y ossils, sugges s o me ha
his g oup o ba s had hei o igin much ea -
lie in he paleochi op e an g ade o pe haps
. . . sepa a ely om an insec i o ous ances-
al s ock.’’
Smi h (1976) also discussed o he Eocene
ossils ha he hough could be e e ed o
ex an amilies. These included Vespe ilia-
us ( e e ed o Emballonu idae); Palaeo-
phyllopho a,Pa aphyllopho a,Rhinolophus,
Pseudo hinolophus, and Hipposide os (Rhin-
olophidae); Nec oman is and P o ampy us
(Megade ma idae); and S ehlinia and Nyc e -
obius (Vespe ilionidae). The p esence o
mode n amilies (and e en gene a) in Middle
and La e Eocene deposi s was in e p e ed as
addi ional e idence o an ea ly di e gence
o mic ochi op e ans om he paleochi op-
e an s ock (Smi h, 1976).
Sige´ (1977) discussed he ossil eco d o
Eocene ba s in a e iew o Paleogene mam-
malian aunas o Eu ope. He e e ed o he
Messel ba s as ‘‘chi op e` es pale´ochi op e´ -
ygoı¨de´s,’’ including wi hin his g oup Ica on-
yc e ididae and Palaeochi op e ygidae. The
la e amily con ained wo sub amilies: A -
chaeonyc e idinae and Palaeochi op e ygi-
nae. Vespe ilia us was e e ed o Embal-
lonu idae, S ehlinia o Vespe ilionoidea
(Ke i oulidae), Hipposide os (Pseudo hino-
lophus) and Palaeophyllopho a o Rhinolo-
phoidea (Hipposide idae), and Nec oman is
o Megade ma idae. In he absence o con-
adic o y e idence, Sige´ sugges ed ha pa-
laeochi op e ygines, emballonu oids, espe -
ilionoids, and hinolophoids migh ha e
o igina ed om unspeci ied endemic Eu o-
pean g oups. Based on i s supposed ‘‘den u e
appa emmen insec i o e’’ (insec i o ous
den i ion), Sige´ (1977: 185) in e p e ed A -
chaeop e opus as a o m in e media e be-
ween megachi op e ans and ica onyc e ids.
In a e iew pape conce ning he e olu ion
o ba ligh , Smi h (1977) no ed ha one o
he ‘‘ i e monophyle ic supe amilies’’ o
Mic ochi op e a was Palaeochi op e ygo-
idea, which included Palaeochi op e ygidae
and Ica onyc e idae. While Russell and Sige´
(1970) had en a i ely e e ed A chaeop e -
opus o Ica onyc e idae, Smi h (1977) no ed
ha he had examined he only known spec-
imen o A chaeop e opus and had ound no
jus i ica ion o his a angemen . Acco ding-
ly, he es ic ed Ica onyc e idae o only Ica-
onyc e is. To ou knowledge, Smi h was he
i s o sugges ha Ica onyc e is,Palaeo-
chi op e yx,A chaeonyc e is,Cecilionyc e -
is, and Ageina o med a clade ( a he han a
g ade) dis inc om o he mic ochi op e an
ba s.
Ba ghoo n (1977) desc ibed wo new
skulls o Vespe ilia us om he Que cy
Phospho i es in sou he n F ance. Based on a
cladis ic analysis o c anial cha ac e s in ex-
an and ossil emballonu ids and se e al ou -
g oups, Ba ghoo n concluded ha Vespe i-
lia us is he sis e -g oup o a clade com-
p ised o Taphozous and Saccolaimus.
Bu le (1978) e iewed he e olu iona y
his o y o ba s in A ica and p o ided a clas-
si ica ion o ossil o ms known om ha
con inen . In his con ex , he placed Vampy-
a us in Mic ochi op e a ince ae sedis, ex-
1998 21SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 8. Van Valen’s (1979) hypo hesis o ba ela ionships ( ed awn om Van Valen, 1979: ig. 1).
The o iginal cap ion o his igu e ead ‘‘Phylogeny o he known amilies o ba s . . . In ao de s,
supe amilies, and amilies o he Mic ochi op e a a e sepa a ed by solid, dashed, and do ed lines
espec i ely.’’ The numbe s e e o axa discussed in an appendix; we ep oduce hese he e because
hey a e in o ma i e wi h espec o Van Valen’s uncon en ional concep s o g oup membe ship: 1,
Megachi op e a, P e opodidae; 2, Mic ochi op e a, undisco e ed amily; 3, Vespe ilionia, Vespe ilion-
oidea, Na alidae; 4, Recen Na alidae; 5, Vespe ilionidae; 6, Recen Vespe ilionidae; 7, Mys acinidae;
8, Recen Mys acinidae; 9, Molossidae; 10, Phyllos oma ia, Rhinopoma oidea, undisco e ed Family; 11,
Emballonu id– hinolophoid s em; 12, Rhinolophoidea; 13, Megade ma idae; 14, Nyc e idae; 15, Rhin-
olophidae; 16, Rhinopoma idae; 17, Emballonu id–noc ilionoid s em; 18, Emballonu idae ( amily should
pe haps ex and as a back as 11); 19, Recen Emballonu idae; 20, C aseonyc e ididae; 21, Noc ilion-
oidea, pe haps ea ly Mo moopidae; 22, Noc ilionidae; 23, Mo moopidae; 24, Recen Mo moopidae; 25,
Phyllos oma idae; 26, Recen Phyllos oma idae; 27, Desmodon idae. Spelling o g oup names ollows
hose o he o iginal au ho .
plici ly ejec ing he conclusions o Schlosse
(1911) and Smi h (1976).
Van Valen (1979) published he i s cla-
dis ic assessmen o highe -le el ba ela ion-
ships ( ig. 8). This s udy was based on di-
e se mo phological cha ac e s, al hough
speci ic me hods o analysis we e no de-
sc ibed and a axon–cha ac e ma ix was no
included. The ocus o Van Valen’s s udy was
ex an ba s, bu he also discussed some ea-
u es o he be e known ossil o ms as e-
po ed in he li e a u e. He p oposed he new
subo de Eochi op e a o he a chaic Eocene
ba s, wi hin which he ecognized one amily
(Palaeochi op e ygidae). The la e g oup
was desc ibed as ‘‘including A chaeonyc-
e (id)idae and p obably Ica onyc e (id)idae
and A chaeop e opodinae as sub amilies’’
(Van Valen, 1979: 109). Van Valen indica ed
ha Eochi op e a was ances al o all ex an
ba s including Megachi op e a and Mic o-
chi op e a. He no ed ha ex an megachi op-
e ans we e p obably de i ed om pe sis en
eochi op e ans such as A chaeop e opus. The
22 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
c i e ia used o lump o ms ances al o di -
e en clades in o one g oup (Eochi op e a)
appea s o ha e been pe cei ed adap i e sim-
ila i y and hei sha ed e en ion o nume ous
p imi i e ai s. Van Valen (1979: 109) ac-
knowledged ha many o he axa in his clas-
si ica ion we e pa aphyle ic, which he no ed
‘‘ e lec s he p og essi e e olu ion o g ades
o adap a ion.’’
Sige´ and Russell (1980) p o ided de ailed
desc ip ions and compa isons o Palaeochi-
op e yx upaiodon and Cecilionyc e is, and
desc ibed a new genus Ma hesia based on
den al ma e ial om Geisel al, ‘‘G ube Ce-
cilie’’ (MP 13; F anzen and Haubold, 1987).
All o hese o ms we e placed in Eochi op-
e a: Palaeochi op e ygidae: Palaeochi op e -
yginae by Sige´ and Russell (1980).
Smi h and S o ch (1981) desc ibed he
new genus Hassianyc e is om Messel (MP
12; Schmid -Ki le , 1987). Hassianyc e is
was desc ibed as di e ing signi ican ly om
palaeochi op e ygoids in possessing an ad-
anced deg ee o den al educ ion and o he
de i ed den al and os eological ea u es.
Smi h and S o ch (1981: 164) concluded ha
Hassianyc e is ‘‘is mo e closely associa ed
wi h he emballonu oid/ hinolophoid sec ion
o he Mic ochi op e a han o any o he
g oup.’’ As a esul , hey e e ed Hassianyc-
e is o Mic ochi op e a ince ae sedis a he
han placing i in Palaeochi op e ygoidea.
Following many ea lie au ho s, Smi h and
S o ch (1981) ecommended emo al o A -
chaeop e opus om Palaeochi op e ygoidea
and subsequen placemen in P e opodidae
based on mo phological e idence ha A -
chaeop e opus is an ea ly megachi op e an
( o de ails o his a gumen , see Habe se ze
and S o ch, 1987). In doing so, Smi h and
S o ch explici ly es ic ed Palaeochi op e -
ygoidea o Palaeochi op e yx,A chaeonyc-
e is,Ica onyc e is,Cecilionyc e is,Ageina,
and Ma hesia. They (1981: 163) no ed he
ollowing conce ning possible ela ionships
o Palaeochi op e ygoidea as hus de ined:
While we would no now go so a as including hese
ba s in he amily Vespe ilionidae G ay 1821, hey
may well sha e a close sis e -g oup ela ionship wi h
he supe amily Vespe ilionoidea o hey may be
placed as he sis e -g oup o a la ge one including
he Phyllos oma oidea and Vespe ilionoidea.
Legend e and Sige´ (1983) wen back o
one o he i s ossil ba s discussed in he
scien i ic li e a u e—Cu ie ’s (1822) ‘‘Ves-
pe ilion de Mon ma e’’—and ound ha
addi ional p epa a ion o he o iginal speci-
men (a pa ial skele on) e ealed ea u es di-
agnos ic o Molossidae. This o m, which
was a iously e e ed o Vespe ilio pa i-
siensis,Se o inoides an iquus,‘‘Vespe ilio’’
se o inoides, and c . Tada ida sp. by p e i-
ous 19 h-cen u y au ho s, was named Cu-
ie imops by Legend e and Sige´ (1983). I is
cu en ly hough o be La e Eocene (P ia-
bonian) in age (MP 19; Russell e al., 1982;
Schmid -Ki le , 1987). Legend e and Sige´
(1983) sugges ed ha Cu ie imops was pos-
sibly ances al o a la ge complex including
all ex an molossids wi h he excep ion o
Tada ida and Mo mop e us.
Sige´ and Legend e (1983) e iewed he
ossil eco d o ba s in di e en deposi ional
en i onmen s and egions in he Medi e a-
nean basin. They e e ed Ica onyc e is,Pa-
laeochi op e yx,Cecilionyc e is,Ma hesia,
A chaeonyc e is, and Ageina o Eochi op-
e a. Hassianyc e is was desc ibed as an ea -
ly mic ochi op e an no clea ly ela ed o any
o he ex an supe amilies. Se e al addi ion-
al Eocene o ms we e e e ed o ex an am-
ilies, including Vespe ilia us,Hipposide os
(Pseudo hinolophus), Palaeophyllopho a,
Nec oman is,Rhinolophus,S ehlinia,Cu-
ie imops, and c . Tada ida. Al hough am-
ily a ini ies we e no speci ied, hese e e -
ences we e clea ly in ended o ollow pas
classi ica ions o hese axa.
Os ande (1983) desc ibed Chad onyc e -
is based on a maxilla y agmen wi h ee h
om wha a e now conside ed o be La e Eo-
cene (Chad onian) deposi s in Neb aska.
Based on compa isons wi h S ehlinia ( e-
e ed o Ke i oulidae by Sige´, 1974) and
‘‘Palaeochi op e ygidae,’’ Os ande e e ed
Chad onyc e is o Vespe ilionoidea: Ke i-
oulidae.
Gup a (1984) discussed ba phylogeny and
ela ionships o Ica onyc e is. Appa en ly
ailing o d aw any dis inc ion be ween sym-
plesiomo phy and synapomo phy, Gup a
sugges ed ha Ica onyc e is may be a me-
gachi op e an based on se e al sha ed p im-
i i e ea u es (e.g., p esence o a claw on he
index inge ). Gup a (1984: 42) also p o-
1998 23SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 9. Cladog am o in e amilial ela ionships in e ed om audi o y cha ac e s ( ed awn om
No acek, 1980a: ig. 1). The o iginal cap ion o his igu e ead ‘‘Wagne ee esul ing om analysis
o 18 audi o y cha ac e s o Recen chi op e an amilies. C aseonyc e idae and Myzopodidae a e omi -
ed. Ho izon al ba s b idging wo o mo e b anches a e synapomo phies; sho black ba s a e indepen-
den ly de i ed cha ac e s unde his a angemen ; s iped ba s ep esen cha ac e s a e e e sals.’’
posed an unusual hypo hesis conce ning he
o igin o ba s:
The close simila i y in he s uc u e o he pa agium
o P e osau s and o ba s, and he p esence o hai in
bo h, can compel us o hink ha he ba s migh ha e
e ol ed om he P e osau s wi h modi ica ions o
mammalian cha ac e s du ing 12 million yea s [o
he] Paleocene, he pe iod du ing which no ossils [o
p e osau s o ba s] ha e so a been epo ed.
None o hese ideas ha e been accep ed by
subsequen au ho s because all published
da a se s indica e ha ba s a e he ian mam-
mals (e.g., No acek, 1980a, 1986, 1990; No-
acek and Wyss, 1986; Simmons, 1993a,
1994, 1995). P e osau s a e almos uni e -
sally ega ded as a chosau ian diapsid ep-
iles ha a e mo e closely ela ed o c oco-
diles and bi ds han o mammals (e.g., Pa-
dian, 1984, 1987; Gau hie , 1986).
Hill and Smi h (1984) p o ided a classi-
ica ion o all chi op e an gene a in hei
book on he na u al his o y o ba s. They ec-
ognized h ee amilies in Palaeochi op e y-
goidea: Palaeochi op e ygidae (including
Palaeochi op e yx,Cecilionyc e is, and
Ma hesia), A chaeonyc e idae (including
A chaeonyc e is and Ageina), and Ica onyc-
e idae (including only Ica onyc e is). Pa -
adoxonyc e is and Hassianyc e is we e le
as ‘‘Family ince ae sedis’’ wi hin Mic ochi-
op e a (Hill and Smi h, 1984: 221). Follow-
ing Smi h and S o ch (1981), Hill and Smi h
(1984) sugges ed ha Palaeochi op e ygo-
idea appea ed o ha e a ini ies wi h Vespe -
ilionoidea, while Hassianyc e is migh be
associa ed wi h Emballonu oidea and Rhin-
olophoidea. Many Eocene and Oligocene
ossils we e e e ed o ex an amilies by
Hill and Smi h (1984). They placed A -
chaeop e opus in i s own sub amily A -
24 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
chaeop e opodinae wi hin P e opodidae, Ves-
pe ilia us in Emballonu inae wi hin Embal-
lonu idae, Nec oman is and P o ampy us in
Megade ma idae, Pseudo hinolophus in Hip-
poside inae wi hin Hipposide idae, Palaeo-
phyllopho a and Pa aphyllopho a in Palaeo-
phyllopho inae wi hin Hipposide idae, and
S ehlinia in Vespe ilioninae wi hin Vespe -
ilionidae.
Legend e (1985) e iewed he ossil ec-
o d o molossids and ee alua ed he a ini-
ies o Wallia, a axon o iginally desc ibed
by S o e (1984) as a p oscalopid insec i o e
om he Uin an (Middle Eocene) Swi
C eek Local Fauna o Saska chewan, Cana-
da. Legend e concluded ha Wallia p obably
ep esen s he oldes known molossid, and he
placed bo h Wallia and Cu ie imops in Tad-
a idinae, he sub amily o which all Te ia y
ossil molossids ha e been e e ed (Hand,
1990).
Sige´ (1985) ocused on ossil ba s om
La e Eocene deposi s o he Fayum a ea,
Egyp , and edesc ibed and discussed he s a-
us o Vampy a us Schlosse , 1910 and P o-
ampy us Schlosse , 1911. He concluded ha
hese wo axa a e objec i e synonyms based
on a single specimen, a well-p ese ed hu-
me us. Sige´ (1985) concluded ha Vampy-
a us was oo poo ly known o be assigned
o any amily, bu sugges ed ha mo phology
o he holo ype indica ed a ini ies wi h ei-
he hipposide ids, na alids, o phyllos om-
ids. Sige´ also desc ibed a new axon based
on c anioden al ma e ial, Philisis, which he
placed in i s own amily, Philisidae. Based
on compa isons wi h ossil and ex an o ms,
Sige´ placed Philisidae wi hin Vespe iliono-
idea and sugges ed ha i was mo e closely
ela ed o Vespe ilionidae sensu la o han o
Molossidae ( ig. 12). Specula ing ha Vam-
py a us and Philisis migh e en ually be
shown o be conspeci ic i mo e comple e
ma e ial we e o be disco e ed, Sige´ (1985)
hypo hesized ha Vampy a idae (which
would eplace Philisidae as he co ec amily
name in his case) would p o e o be mo e
closely ela ed o Na alidae han o Vespe -
ilionidae o Molossidae. In each o Sige´’s
phylogene ic ees ( ig. 12), Eochi op e a was
shown as occupying he basal b anch. I is
no clea , howe e , whe he his was mean
o imply monophyly o Eochi op e a; com-
men s in he ex sugges ha Sige´ included
Eochi op e a o ep esen he basal s ock o
Chi op e a.
Ho a´cek (1986) discussed ela ionships o
S ehlinia and Ke i oula and concluded (con-
a Sige´, 1974) ha hese axa we e no close-
ly ela ed. Ins ead, Ho a´cek sugges ed ha
S ehlinia migh be ela ed o Miniop e inae.
Mein and Tupinie (1986) b ie ly e iewed
he ea ly Te ia y eco d o ba s in a sho
pape on he e olu ion o echoloca ion sys-
ems. They ollowed Van Valen (1979) in
conside ing Ica onyc e is,A chaeonyc e is,
and Palaeochi op e yx as ep esen a i es o
Eochi op e a, while e e ing o he ossil
axa o ex an supe amilies o amilies. Mein
and Tupinie lis ed Vespe ilia us in Embal-
lonu idae, S ehlinia in Vespe ilionoidea, and
e e ed Nec oman is and possibly P o am-
py us o Megade ma idae.
Ginge ich (1987) desc ibed Wyonyc e is
based on a pa ial den a y and isola ed ee h
om La e Paleocene (Cla k o kian) deposi s
o he Willwood Fo ma ion o Wyoming. He
e e ed Wyonyc e is o Palaeochi op e ygi-
dae on he basis o den al simila i ies, bu
subsequen au ho s (e.g., Habe se ze e al.,
1994; Hand e al., 1994) ha e ques ioned he
chi op e an a ini ies o his o m (see oo -
no es o able 1). Ginge ich also desc ibed
ee h e e ed o c . Ica onyc e is om he
Willwood Fo ma ion.
Habe se ze and S o ch (1987) discussed
in de ail he classi ica ion and unc ional
mo phology o Paleogene ba s, concen a ing
on he ossil ba s om Messel. They a gued
agains usage o Palaeochi op e ygoidea and
Eochi op e a, ecognizing ha each was
p obably pa aphyle ic e en i A chaeop e o-
pus was emo ed. Ins ead, hey ecognized
and p o ided e ised diagnoses o h ee
amilies placed in Mic ochi op e a ince ae
sedis: A chaeonyc e ididae (including A -
chaeonyc e is and Ica onyc e is), Palaeochi-
op e ygidae (Palaeochi op e yx), and a new
amily, Hassianyc e ididae (Hassianyc e is,
including ‘‘A chaeonyc e is’’ e illiodi Rus-
sell and Sige´, 1970). The ela ionships o Ce-
cilionyc e is,Ageina, and Ma hesia we e no
add essed.
No acek (1987) ocused on phylogene ic
ela ionships o he bes known Eocene axa
(Ica onyc e is and Palaeochi op e yx) a he
1998 25SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 10. Rela ionships o selec ed axa in-
e ed om ea u es o e al memb ane s uc u e
and de elopmen ( ed awn om Lucke , 1980a:
ig. 5). The o iginal cap ion o his igu e ead
‘‘Cha ac e s a e dis ibu ion o e al memb ane
and ep oduc i e ea u es o Chi op e a . . .’’
han on issues o classi ica ion. Following a
de ailed e iew o a wide a ay o mo pho-
logical ea u es, No acek (1987) concluded
ha bo h Ica onyc e is and Palaeochi op e-
yx a e mo e closely ela ed o ex an mic o-
chi op e ans han o ex an megachi op e ans
( ig. 14). He concluded ha Ica onyc e is
and p obably Palaeochi op e yx a e ou -
g oups o all Recen amilies o Mic ochi-
op e a, bu could no ule ou he possibili y
ha one o bo h o hese Eocene axa migh
ha e special a ini ies wi hin Mic ochi op e a
(i.e., migh be mo e closely ela ed o one o
mo e ex an amilies han o o he s). No a-
cek (1987: 15–16) u he concluded ha
. . . he Palaeochi op e ygoidea seems me ely an a -
i icial con en ion o g oup se e al ea ly mic ochi-
op e ans whose ela ionships wi h mode n amilies
o his subo de emain poo ly known . . . Ica onyc-
e is and Palaeochi op e yx a e mo e accu a ely des-
igna ed as Mic ochi op e a ince ae sedis. Mo eo e ,
he e seems no jus i ica ion o a o mal designa ion
o an ‘‘ances al’’ g oup (e.g., Eochi op e a sensu Van
Valen, 1979) o dis inguish Ica onyc e is,Palaeochi-
op e yx, and o he Eocene ba s om he Recen chi-
op e an subo de s.
In a classi ica ion o ossil e eb a es,
Ca oll (1988) named a new supe amily o
he a chaic Eocene ba s, Ica onyc e oidea,
which he placed wi hin Mic ochi op e a. He
included wo amilies in Ica onyc e oidea:
Ica onyc e idae (Ica onyc e is) and Palaeo-
chi op e ygidae (A chaeonyc e is,Palaeo-
chi op e yx,Cecilionyc e is, and Ma hesia).
In concep and con en s, Ica onyc e oidea is
iden ical o Palaeochi op e ygoidea as ec-
ognized by mos p e ious au ho s. Ca oll
ga e no jus i ica ion o his appa en ly un-
necessa y name change, and no subsequen
au ho s ha e used Ica onyc e oidea. Ca oll
addi ionally p o ided a classi ica ion o o h-
e Eocene ba s, placing Vespe ilia us in Em-
ballonu idae, Nec oman is in Megade ma i-
dae, Rhinolophus in Rhinolophidae, Hippo-
side os,Palaeophyllopho a, and Pa aphyl-
lopho a in Hipposide idae, ?Vampy a us in
Phyllos omidae, and S ehlinia in Vespe i-
lionidae. He also placed A chaeop e opus in
P e opodidae.
Robbins and Sa ich (1988) used p o ein
elec opho esis and immunological dis ance
da a o add ess ela ionships among ex an
Emballonu idae. They also conside ed mo -
phological da a p esen ed by Ba ghoo n
(1977), and included a b ie discussion o he
a ini ies o Vespe ilia us. Robbins and Sa -
ich concluded by assigning Vespe ilia us o
a new ibe, Vespe ilia ini, wi hin he sub-
amily Taphozoinae.
Sige´ (1988) desc ibed he ossil ba s om
he Ma inesian (
5
Robiacian; MP 16) Le
B e ou auna om he Que cy Phospho i es
in F ance. He e e ed Vespe ilia us o Em-
ballonu idae, and Palaeophyllopho a and
Hipposide os (Pseudohipposide os) o Hip-
poside idae. Sige´ (1988: 93) also epo ed he
p esence o a new o m ha he e e ed o
‘‘Rhinolophoidea, sp. inde .’’
S o ch and Habe se ze (1988) desc ibed a
new species o A chaeonyc e is (A.pollex)
based on wo skele ons om Messel, and
compa ed his axon wi h o he Eocene ba s.
26 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
They also epo ed he disco e y o a new
skull ha con i med Habe se ze and S o ch’s
(1987) sugges ion ha ‘‘A chaeonyc e is’’
e illiodi should be e e ed o Hassianyc e -
is a he han o A chaeonyc e is o A chaeo-
nyc e ididae.
Habe se ze and S o ch (1989) discussed
ecology and echoloca ion abili ies o Eocene
ba s om Messel as in e ed om wing and
cochlea mo phology. Based on compa isons
wi h a ious ex inc and ex an ba s, Habe -
se ze and S o ch (1989: 214) no ed ha
A chaeonyc e is igonodon Re illiod 1917 ( amily
A chaeonyc e ididae) is an a chaic species.... Pa-
laeochi op e yx upaiodon Re illiod 1917, and P.
spiegeli Re illiod 1917, ( amily Palaeochi op e ygi-
dae) a e small specialized species....Hassianyc e is
messelensis Smi h and S o ch 1981, H.magna Smi h
and S o ch 1981, and H. e illiodi (Russell and Sige
1970) (Family Hassianyc e ididae) a e he mos ad-
anced species in den al and skele al ea u es.
Habe se ze and S o ch (1989) a gued
agains Van Valen’s (1979) concep o Eochi-
op e a, which was o iginally de ined as in-
cluding all o he axa men ioned in he ex-
ce p abo e. They no ed ha Eochi op e a as
concei ed by Van Valen would be a g oup
o p imi i e and unspecialized species ha
ul ima ely ga e ise o bo h Megachi op e a
and Mic ochi op e a. No ing ha he Messel
ba auna includes ecologically di e se o ms
ha a e all ‘‘comple ely de eloped’’ mic o-
chi op e ans, Habe se ze and S o ch (1989:
231) sugges ed ha Eochi op e a was a mis-
leading concep , a leas when applied o he
Messel ba s.
Sige´ (1990) and Sud e e al. (1990) de-
sc ibed new species o Vespe ilia us om
S ampian (MP 25) and ea ly Ba onian (MP
14) deposi s o F ance, espec i ely. Bo h e-
ained Vespe ilia us in Emballonu idae.
Sige´ addi ionally desc ibed a new species o
Hipposide os (subgenus Pseudo hinolo-
phus), a new species o S ehlinia (which he
e e ed o Na alidae sensu Van Valen, 1979),
and a new genus Vayla sia. All ou o he
new axa desc ibed by Sige´ (1990) we e
based on agmen a y den i ions and isola ed
pos c anial elemen s om he Que cy Phos-
pho i es. Vayla sia was e e ed o Hipposi-
de idae based on den al and hume al cha -
ac e s, bu Sige´ (1990: 1132) no ed ha i
‘‘p obably ep esen s he s em g oup o he
genus Rhinolophus.’’
Sige´ (1991a) la e desc ibed he genus
Dizzya om he Ea ly Eocene o Chambi in
no he n A ica. Following b oad compa i-
sons wi h ex an and ex inc o ms, he e-
e ed Dizzya o Vespe ilionoidea: Philisi-
dae. A specimen e e ed o Rhinolophoidea
gen. and sp. inde . was also desc ibed, and
Sige´ wen on o ex ensi ely discuss he bio-
s a ig aphic, biogeog aphic, and paleoeco-
logical implica ions o he ba auna om he
A ican Eocene. In a inal no e, Sige´ dis-
cussed usage o he name Eochi op e a. Con-
a Habe se ze and S o ch (1987, 1989) and
No acek (1987), he (1991a: 372) a gued ha
his axon ‘‘es p e´ e´ e´ pa logique, commo-
di e´, e e icaci e´a` celui supe - amilial des
Palaeochi op e ygoidea inclus dans les Mi-
c ochi op e a.’’ Sige´ (1991a: 373) a gued
ha ecognizing he ‘‘ axon-g ade’’ Eochi-
op e a is a simple, ope a ional app oach o
classi ying o ganisms o unce ain a ini ies
ha sha e a la ge numbe o p imi i e cha -
ac e s. Exp essing he unusual opinion ha
Mic ochi op e a mus i sel be conside ed
pa aphyle ic (‘‘l’uni e´ sub-o dinale Mic ochi-
op e a, qui . . . doi eˆ e clai men pe c¸ue
comme un assemblemen pa aphyle´ ique,
mais se aleu empi ique. . ..’’; Sige´, 1991a:
373), he emphasized he empi ical alue o
o mally g ouping axa based on hei pe -
cei ed g ade o e olu ion. He a gued ha
such a classi ica ion sys em should pe haps
be main ained o he chi op e an subo de s
( o indica e hei adap i e le el in he con in-
uum o ba e olu ion), e en i i becomes
possible o diagnose s ongly suppo ed,
monophyle ic supe amilies including some
o he a chaic Eocene axa.
Sige´ (1991b) discussed mo phology o he
deciduous den i ion o Eocene ba s and p e-
sen ed a igu e summa izing ela ionships
and s a ig aphic anges ( ig. 15). He indi-
ca ed ha i e ex an lineages (Megachi op-
e a, Phyllos oma oidea [
5
Noc ilionoidea],
Rhinolophoidea, Emballonu ioidea, and Ves-
pe ilionoidea) and h ee ex inc lineages
(Hassianyc e ididae, A chaeonyc e ididae,
and Palaeochi op e ygidae) we e in exis ence
in he Eocene. He placed A chaeonyc e idi-
dae and Hassianyc e ididae in Eochi op e a,
1998 33SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 16. Two al e na i e hypo heses o in e amilial ela ionships sugges ed by No acek ( ed awn
om No acek, 1991: igs. 6, 7). The o iginal cap ion o A ead ‘‘ This cladog am . . . is based on
discussions in Koopman (1984). The posi ion o Palaeochi op e yx . . . ollows No acek (1987). No e
he emo e b anch posi ion o Rhinopoma idae.’’ he cap ion o B ead ‘‘as in [ he p e ious igu e]
modi ied by eloca ion o Rhinopoma idae as a sis e axon o Rhinolophoidea ( ollowing Pie son,
1986).’’
oge he wi h Myzopodidae, Thy op e idae,
and Fu ip e idae, which Van Valen (1979) e-
duced in ank o sub amily le el.
No acek (1980a) and Lucke (1980a)
published cladog ams o ba ela ionships
based on analyses o single o gan sys ems.
No acek ocused on mo phology o he au-
di o y egion; Lucke concen a ed on mo -
phology and de elopmen o e al mem-
b anes. P ima ily because he ee de i ed
om audi o y cha ac e s ( ig. 9) di e ed sig-
ni ican ly om all p e ious hypo heses o
34 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 17. Two al e na i e hypo heses o yino-
chi op e an ela ionships based on hyoid muscu-
loskele al mo phology ( ed awn om G i i hs e
al., 1992: igs. 10, 11). Bo h ees we e equally
pa simonious in he con ex o he hyoid da a p e-
sen ed by G i i hs and Smi h (1991) and G i i hs
e al. (1992), bu hese au ho s p e e ed ee A
on he basis o hypo hesized pa e ns o ans o -
ma ion in wo cha ac e s.
highe -le el ela ionships, No acek (1980a)
wa ned agains using ha cladog am (o any
o he de i ed om a single o gan sys em) as
a basis o a new phylogene ic econs uc ion
o classi ica ion. Lucke ’s (1980a) s udy was
unable o esol e many ela ionships ( ig.
10), bu p o ided some suppo o mono-
phyly o Noc ilionoidea and a close ela ion-
ship be ween Megade ma idae and a clade
con aining Vespe ilionidae and Thy op e i-
dae. Monophyly o Vespe ilionidae sensu
Koopman and Jones (1970) was appa en ly
assumed by bo h No acek (1980a) and Luck-
e (1980a).
Eisenbe g (1981) published a phylogeny
o ba s in his book on mammalian adia ions.
Eisenbe g’s (1981: 147) ee ( ig. 11) di e ed
only sligh ly om ha p esen ed by Smi h
(1976; see ig. 7). Like Smi h (1976), Eisen-
be g depic ed each o he mic ochi op e an
supe amilies as monophyle ic. Howe e , Ei-
senbe g placed Emballonu oidea, Rhinolo-
phoidea, and Phyllos oma oidea (
5
Noc i-
lionoidea) oge he in an un esol ed clade
wi h Vespe ilionoidea as he sis e -g oup,
a he han iden i ying Emballonu oidea and
Rhinolophoidea as sis e - axa as had Smi h
(1976).
Gopalak ishna and Cha i (1983) desc ibed
e al memb ane de elopmen in Miniop e us,
a axon gene ally placed in i s own sub amily
in Vespe ilionidae. Based on hei indings,
which indica ed majo di e ences be ween
miniop e ines and o he espe ilionids, Go-
palak ishna and Cha i ecommended emo al
o Miniop e us o i s own amily, Miniop e -
idae.
Pie son (1986) p oposed a se ies o phy-
logene ic hypo heses based on an analysis o
ans e in immunological dis ance da a ( ig.
13). He esul s di e ed signi ican ly om
hose o p e ious s udies in many a eas, in-
cluding (1) placing Rhinopoma idae wi hin
Rhinolophoidea, (2) associa ing Mys acini-
dae wi h Noc ilionoidea a he han Vespe -
ilionoidea, (3) placing Fu ip e idae and Na-
alidae a he base o he mic ochi op e an
ee, and (4) g ouping Tomopeas and Min-
iop e us wi h Molossidae a he han wi h
Vespe ilioninae. In he con ex o Pie son’s
ees, Yinochi op e a is monophyle ic, bu
Yangochi op e a and he ou supe amilies
a e no . One o he mo e s iking esul s o
Pie son’s (1986) immunological s udy— he
placemen o Mys acinidae—was explo ed in
g ea e dep h in Pie son e al. (1986), which
also included a discussion o some mo pho-
logical ea u es.
Hill and Ha ison (1987) e iewed c anio-
den al mo phology and s uc u e o he bac-
ulum in selec ed espe ilionids, and p o-
posed a new gene ic classi ica ion o hose
o ms adi ionally included in Vespe ilion-
inae (e.g., by Mille , 1907). Hill and Ha i-
son (1987) included Myo ini and An ozoini
1998 35SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 18. Fen on’s (1992) phylogeny o ba amilies wi h Al ingham’s (1996) addi ion o sub amilies
( ed awn om Fen on [1992: p. 10] and Al ingham [1996: ig. 1.12]). Fen on’s (1992) cap ion ead ‘‘A
amily ee o phylogeny shows he p esumed e olu iona y ela ionships be ween he li ing amilies o
ba s. The amilies a e g ouped acco ding o supe amilies and subo de s.’’ Al ingham’s (1996) cap ion
ead ‘‘An e olu iona y ee o he ba s o sub amily le el. Conside able unce ain y su ounds some
pa s o his ee, in pa icula he ela ionships be ween megaba s and mic oba s, and he de ailed
classi ica ion o he Phyllos omidae and Vespe ilionidae.’’
as ibes wi hin Vespe ilioninae, and ec-
ognized Nyc ophininae as a dis inc sub am-
ily.
Tiuno (1989) examined a ia ion in mo -
phology o he ongue and male accesso y
glands (e.g., p os a e, seminal essicles,
Cowpe ’s gland, ampulla y glands) in a num-
be o Old Wo ld species, and discussed he
phylogene ic implica ions o hese da a.
Based on obse ed di e ences, Tiuno con-
cu ed wi h Gopalak ishna and Cha i’s
(1983) ecommenda ion ha Miniop e idae
should be ecognized as a amily dis inc
om Vespe ilionidae. In con as , Tiuno
ound no di e ences be ween hinolophines
and hipposide ines, and he e o e ecom-
mended ha hey be placed in a single am-
ily, Rhinolophidae.
Bake e al. (1991a) analyzed a ia ion in
DNA es ic ion si es in a s udy designed o
36 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 19. Volle h and Helle ’s hypo hesis o ela ionships among espe ilionids ( ed awn om Vol-
le h and Helle , 1984: ig. 7). The o iginal cap ion ead ‘‘Cladog am o Vespe ilionidae based on ka y-
ological ea u es . . . Whe e he names o species a e lacking, only one species was s udied (12 cases).
F om ka yologically he e ogeneous gene a all species s udied a e shown. The do ed line be ween Ep-
esicus and Sco ophilus indica es a second possibili y o he ela ions o Ep escini . . .[ ha ibe] could
be g ouped oge he wi h Sco ophilus ....[Abb e ia ions:] H.
5
Hespe op e nus,Hyps.
5
Hypsugo,
Ny.
5
Nyc alus,P.
5
Pipis ellus,T.
5
Tylonyc e is.’’
1998 37SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
es ba monophyly (see below). Al hough
hei da a se could no esol e subo dinal o
in e amilial ela ionships o ba s, suppo
was ound o se e al g oups, including Yin-
ochi op e a, Noc ilionoidea, and a clade con-
aining Mo moopidae and Noc ilionidae.
In he 1980s and ea ly 1990s, a signi ican
con o e sy a ose in chi op e an sys ema ics
ega ding ba monophyly. The his o y o his
con o e sy was e iewed by Simmons
(1994) and will no be epea ed he e. Al-
hough ba diphyly has been p oposed by
se e al au ho s (e.g., Smi h and Madkou ,
1980; Hill and Smi h, 1984; Pe ig ew, 1986,
1991a, 1991b, 1994, 1995; Pe ig ew and Ja-
mieson, 1987; Pe ig ew e al., 1989; Rayne ,
1991b; Pe ig ew and Ki sch, 1995), a g ow-
ing body o da a p o ides e y s ong sup-
po o ba monophyly. Da a suppo ing chi-
op e an monophyly include mo phological
ea u es om many o gan sys ems (Lucke ,
1980a, 1993; Wible and No acek, 1988;
Ko un, 1989; Thewissen and Babcock,
1991, 1993; Kay e al., 1992; No acek, 1992,
1994; Bea d, 1993; Simmons, 1993a, 1994,
1995; Wible and Ma in, 1993; Simmons and
Quinn, 1994; Miyamo o, 1996), DNA–DNA
hyb idiza ion da a (Ki sch e al., 1995;
Hu cheon and Ki sch, 1996; Ki sch, 1996),
and sequence da a om nume ous mi ochon-
d ial and nuclea genes (Adkins and Honey-
cu , 1991, 1993, 1994; Mindell e al., 1991;
Amme man and Hillis, 1992; Bailey e al.,
1992; S anhope e al., 1992, 1993, 1996; Ho-
neycu and Adkins, 1993; Knigh and Min-
dell, 1993; No acek, 1994; Alla d e al.,
1996; Miyamo o, 1996; Po e e al., 1996).
Ba monophyly now ep esen s one o he
mos s ongly suppo ed phylogene ic hy-
po heses wi hin Mammalia (Simmons, 1994;
Miyamo o, 1996).
No acek (1991) used wo no el phyloge-
nies as a amewo k o discussing e olu ion
o cochlea ea u es in ba s ( ig. 16). These
ees we e no de i ed om an explici cha -
ac e analysis, bu we e ins ead based on
conside a ion o he mo phological cha ac-
e s discussed in Koopman (1984) and No-
acek (1987) as well as Pie son’s (1986) im-
munological esul s (see cap ion o ig. 16).
Bo h o No acek’s (1991) ees indica ed
pa aphyly o Yinochi op e a and Emballon-
u oidea, le monophyly o Yangochi op e a
un esol ed, and sugges ed monophyly o he
emaining h ee supe amilies. Monophyly o
Vespe ilionidae (including ke i oulines, o-
mopea ines, and miniop e ines) was no dis-
cussed.
G i i hs and his colleagues used os eo-
myological cha ac e s o he hyoid egion o
explo e ela ionships among a ious g oups
o yinochi op e an ba s (G i i hs and Smi h,
1991; G i i hs e al., 1992). G i i hs e al.
(1992) p esen ed wo al e na i e phylogenies
( ig. 17), nei he o which suppo ed mono-
phyly o Emballonu oidea o Rhinolophoi-
dea. The only clade ha appea ed in bo h
ees was Rhinolophidae
1
Hipposide idae.
Gopalak ishna and Badwaik (1992) dis-
cussed e al memb ane s uc u e and used
phene ic simila i y o e alua e phylogene ic
ela ionships among ba amilies. They con-
cluded ha
. . . simila i ies be ween Molossidae and P e opodidae
and di e ences be ween Molossidae and Vespe i-
lionidae sugges a close ela ionship be ween P e o-
podidae and Molossidae han be ween Molossidae
and Vespe ilionidae. I is, he e o e, sugges ed on
pu ely emb yological g ounds ha Molossidae be
sepa a ed om he Supe - amily Vespe ilionidae and
be placed somewhe e be ween P e opodidae and Em-
ballonu idae (Gopalak ishna and Badwaik, 1992: 7).
Fen on (1992) p o ided a phylogeny o ba
amilies in his semipopula book on ba s, bu
did no discuss he sou ce o his hypo hesis.
The same opology was ep oduced by Fen-
on (1995). This ee ( ig. 18) was de eloped
o summa ize possible ela ionships o ba s
as e lec ed in nume ous sys ema ic s udies
and classi ica ions (e.g., Hill, 1974; Pie son
e al., 1986); i was no based on any new
da a o da a analyses (Fen on, pe sonal com-
mun.). Al ingham (1996) ep oduced Fen-
on’s ee (again wi h no men ion o sou ce),
modi ying i only by adding b anches o
each sub amily ( ig. 18).
Volle h and Helle (1994) p o ided he
i s cladis ic hypo hesis o ela ionships
among gene a o Vespe ilionidae sensu la o
( ig. 19). Using da a om G-banded ch o-
mosomes, hey iden i ied homologous a ms
and p esen ed a cladog am based on an anal-
ysis o de i ed ch omosomal ea u es (e.g.,
ansloca ions, Robe sonian usions, is-
sions). They concluded ha (1) Miniop e i-
nae is he sis e -g oup o a clade con aining
38 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 20. Resul s o Simmons’ (1998) analysis o highe -le el ela ionships based on mo phology
and DNA es ic ion si es. The o iginal cap ion o his igu e ead ‘‘S ic consensus o wo equally
mos -pa simonious ees (608 s eps each; CI
5
0.410, RI
5
0.592) ound in a heu is ic analysis . . . The
numbe s below in e nal b anches indica e he pe cen age o boo s ap eplica es in which each clade
appea ed; numbe s abo e he b anches a e decay alues ( he minimum numbe o addi ional s eps
equi ed o collapse each clade).’’
Ke i oulinae
1
Mu ininae
1
Vespe ilioni-
nae; (2) Myo ini alls ou side a clade con-
aining he emaining espe ilionines; (3)
Nyc ophilus is a membe o Vespe ilionini;
(4) Vespe ilionini and Pipis ellini a e sis e -
axa; and (5) Ep esicini (including Hespe -
op enus), ‘‘Nyc iceiini,’’ and Pleco ini all
ou side he Vespe ilionini
1
Pipis ellini
clade. An ozous,Tomopeas,Lasiu us, and
o he New Wo ld axa we e omi ed om
Volle h and Helle ’s s udy.
Sudman e al. (1994), d awing in pa
upon he unpublished wo k o Ba kley
(1984), in es iga ed ela ionships o Tomo-
peas ( he only membe o Tomopea inae) o
espe ilionids and molossids using p o ein
elec opho esis, cy och ome bgene sequenc-
es, and mo phological cha ac e s. They con-
cluded ha hese da a suppo a close ela-
ionship be ween Tomopeas and molossids
a he han espe ilionids, and ecommend-
ed ha Tomopea inae be ans e ed om
Vespe ilionidae o Molossidae.
Ki sch and his colleagues (Pe ig ew and
Ki sch, 1995; Hu cheon and Ki sch, 1996;
Ki sch, 1996; Ki sch and Hu cheon, 1997;
Hu cheon e al., in p ess; Ki sch and Pe i-
g ew, in p ess; Pe ig ew and Ki sch, in
p ess) epo ed esul s o a se ies o DNA-
DNA hyb idiza ion s udies o ela ionships
among a se o axa including bo h megachi-
op e ans and mic ochi op e ans. Su p ising-
ly, hese expe imen s sugges ed ha Rhino-
lophoidea and P e opodidae may be sis e -
axa, implying mic ochi op e an pa aphyly.
Howe e , his esul has been ques ioned by
1998 39SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
TABLE 2
Highe -le el Classi ica ion o Recen Ba s
P oposed by Simmons (1998)
O de Chi op e a
Subo de Megachi op e a
Family P e opodidae
Subo de Mic ochi op e a
In ao de ince ae sedis
Supe amily Emballonu oidea
Family Emballonu idae
In ao de Yinochi op e a
Supe amily Rhinopoma oidea
Family C aseonyc e idae
Family Rhinopoma idae
Supe amily Rhinolophoidea
Family Nyc e idae
Family Megade ma idae
Family Rhinolophidae
Sub amily Rhinolophinae
Sub amily Hipposide inae
In ao de Yangochi op e a
Supe amily ince ae sedis
Family Mys acinidae
Supe amily Noc ilionoidea
Family Noc ilionidae
Family Mo moopidae
Family Phyllos omidae
Supe amily Molossoidea
Family An ozoidae
Family Molossidae
Sub amily Tomopea inae
Sub amily Molossinae
Supe amily Vespe ilionoidea
Family Vespe ilionidae
Sub amily Vespe ilioninae
Sub amily Miniop e inae
Sub amily Myo inae
Sub amily Mu ininae
Sub amily Ke i oulinae
Supe amily Na aloidea
Family Myzopodidae
Family Thy op e idae
Family Fu ip e idae
Family Na alidae
he au ho s hemsel es, who ha e sugges ed
ha base composi ional bias migh be e-
sponsible o p oducing a alse phylogene ic
signal linking wo A-T ich clades (P e o-
podidae and Rhinolophoidea; Hu cheon and
Ki sch, 1996; Ki sch, 1996; Ki sch and
Hu cheon, 1997; Hu cheon e al., in e iew;
Ki sch and Pe ig ew, in e iew; Pe ig ew
and Ki sch, in e iew). Al hough opology o
o he pa s o he ee emains suspec , many
eco e ed g oupings a e cong uen wi h
hose iden i ied in p e ious s udies. As no ed
by Hu cheon e al. (in p ess), ‘‘a ee as s a -
ling as ou s ob iously mus be e i ied by
addi ional s udies . . .’’
The mos ecen comp ehensi e a emp o
esol e highe -le el ela ionships among ex-
an amily-le el lineages o ba s was ha o
Simmons (1998). She assembled a da a se
o 192 disc e e cha ac e s including many
new mo phological cha ac e s, all o he el-
e an mo phological da a discussed by mos
p e ious wo ke s (e.g., Van Valen, 1979;
Lucke , 1980a; No acek, 1980a, 1991;
Ba kley, 1984; G i i hs and Smi h, 1991;
G i i hs e al., 1992), and he DNA es ic-
ion si e da a p esen ed by Bake e al.
(1991a). To es espe ilionid monophyly
and e alua e ela ionships o i s sub amilies,
Vespe ilionidae sensu la o was spli in o se -
en subg oups o analysis: Vespe ilioninae,
Myo inae, Miniop e inae, Mu ininae, Ke i-
oulinae, An ozoinae, and Tomopea inae.
Pa simony analyses o he esul ing da a se
p oduced a well- esol ed ee ( ig. 20) in
which many nodes we e s ongly suppo ed.
Majo esul s included he ollowing: (1)
Emballonu idae appea s o be he sis e -
g oup o all o he mic ochi op e ans, he e-
o e Emballonu oidea and Yinochi op e a (as
adi ionally ecognized) a e no monophy-
le ic; (2) Rhinopoma idae and C aseonyc e -
idae a e sis e - axa; (3) Rhinolophoidea,
Noc ilionoidea, Vespe ilionoidea, and Yan-
gochi op e a each appea o be monophyle -
ic; (4) Vespe ilionidae sensu la o is no
monophyle ic; An ozoinae and Tomopea i-
nae a e mo e closely ela ed o Molossidae
han o o he espe ilionids; and (5) Myzo-
podidae, Thy op e idae, Na alidae, and Fu-
ip e idae o m a clade.
Simmons (1998) p oposed a numbe o
nomencla u al changes ( able 2) based on he
phylogene ic esul s, es ic ing Emballonu-
oidea o Emballonu idae, placing Rhino-
poma idae and C aseonyc e idae in Rhino-
poma oidea, es ic ing Yinochi op e a o
Rhinopoma oidea
1
Rhinolophoidea, aising
An ozoinae o amily ank as An ozoidae,
e e ing Tomopea inae o Molossidae, ec-
ognizing An ozoidae
1
Molossidae as a
new supe amily Molossoidea, ecognizing
Myzopodidae
1
Thy op e idae
1
Na alidae
40 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
1
Fu ip e idae as a new supe amily Na a-
loidea, and es ic ing Vespe ilionidae o
Vespe ilioninae
1
Miniop e inae
1
Myo -
inae
1
Mu ininae
1
Ke i oulinae. No ing
ha se e al o he la e sub amilies migh
e en ually be aised o amily ank, Simmons
(1998) es ic ed Vespe ilionoidea o Ves-
pe ilionidae as de ined abo e, pending u -
he s udy.
GOALS OF THE CURRENT STUDY
The da a se de eloped by Simmons
(1998)—which o cou se includes many
cha ac e s o iginally desc ibed by o he s (see
appendix 2)—includes mo e han 80 c anio-
den al and pos c anial os eological cha ac-
e s. This p o ides a unique oppo uni y o
e alua e he e olu iona y ela ionships o he
a chaic Eocene axa in he con ex o a phy-
logeny o ex an lineages, and o in es iga e
me hodological issues in ol ing ossil axa
and missing da a in phylogene ic analyses.
The goals o he cu en s udy a e as ollows:
(1) o de e mine he ela ionships o Ica on-
yc e is,A chaeonyc e is,Hassianyc e is, and
Palaeochi op e yx o each o he and o am-
ily-le el lineages o ex an ba s; (2) o e al-
ua e he e ec s o including ossil gene a on
pe cei ed ela ionships among ex an o ms;
(3) o e alua e he e ec s o including so -
issue and molecula cha ac e s in a s udy in-
cluding bo h ossil and ex an g oups; (4) o
documen pa e ns o cha ac e ans o ma-
ion in he basal pa o he chi op e an ee;
and (5) o conside he implica ions o hese
da a o heo ies conce ning he ea ly e o-
lu ion o echoloca ion and o aging s a egies
in Mic ochi op e a.
MATERIALS AND METHODS
TAXONOMIC SAMPLING AND
MONOPHYLY
The p esen s udy consis s o a phyloge-
ne ic analysis o 24 amily-le el lineages o
ex an ba s, wo o dinal-le el ex an ou g oup
axa (Scanden ia, De mop e a), and ou Eo-
cene ossil ba gene a (Ica onyc e is,A -
chaeonyc e is,Palaeochi op e yx,Hassi-
anyc e is; see appendix 1 o specimens ex-
amined). The ex an lineages a e he same as
hose analyzed by Simmons (1998); axo-
nomic names used o ba clades ollow he
classi ica ion p oposed by Simmons (1998;
able 2). Monophyly o Vespe ilioninae (in-
cluding Nyc ophilinae) was assumed o
p ac ical easons ollowing Volle h and Hell-
e (1994) and Simmons (1998). Monophyly
o each o he o he ex an ba lineages used
as OTUs (ope a ional axonomic uni s) is
well es ablished (see Simmons [1998] and
e e ences ci ed he ein). Each o he ossil
gene a also appea s o be monophyle ic ( a-
ble 3), wi h he possible excep ion o Ica-
onyc e is.
In he absence o apomo phic ai s, mos
wo ke s ha e diagnosed Ica onyc e is on he
basis o p imi i e ea u es ha ha e been
modi ied in all o he ba s (Jepsen, 1966,
1970; No acek, 1987). Al hough Jepsen
(1966, 1970) no ed 14 cha ac e s o Ica on-
yc e is ha he conside ed as ‘‘p imi i e’’ o
‘‘gene alized’’ ela i e o he condi ions seen
in ex an ba s, close examina ion o hese
ea u es by No acek (1987) educed his lis
conside ably. No acek iden i ied only ou
p imi i e ea u es appa en ly ound in Ica-
onyc e is ha ha e been los o modi ied in
all o he known ossil and ex an ba s: (1)
un used s e nal elemen s, (2) ela i ely sho
adius, (3) comple e phalangeal o mula (2-
3-3-3) on he digi s o he wings, and (4)
head and neck o emu se a an angle o he
sha . Ou examina ion o addi ional speci-
mens o ex an ba s plus ma e ial o Ica on-
yc e is ha was no a ailable o Jepsen o
No acek indica e ha wo mo e o hese ea-
u es should be emo ed om he lis .
Ou obse a ions indica e ha s e nal ele-
men s a e ypically un used in ju enile, sub-
adul , and some young adul ba s. No acek
(1987) no ed ha he holo ype o Ica onyc-
e is index (PU 18150) migh ha e un used
s e nal elemen s simply because i was a
young indi idual. This was con i med by ou
examina ion o UW 21481a–b ( ig. 21), a
p e iously undesc ibed specimen o Ica on-
yc e is index om Wyoming.
2
The s e nal el-
2
We iden i ied UW 21481a–b as Ica onyc e is index
1998 41SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
emen s in his indi idual a e used wi h he
excep ion o he join s be ween he manub i-
um and mesos e num and be ween he me-
sos e num and xiphis e num. Su u es a e is-
ible be ween h ee mesos e nal elemen s, bu
he emaining mesos e nal elemen s a e ully
used ( ig. 22). This deg ee o usion is sim-
ila o ha ound in young adul ba s o many
ex an amilies. Because UW 21481a–b and
PU 18150 a e simila in all o he espec s,
we conclude ha PU 18150 was a young
adul Ica onyc e is, whe eas UW 21481a–b
ep esen s a somewha olde indi idual. A
simila pa e n o a ia ion is also seen in
some o he Messel ba s. Fo example, all
mesos e nal elemen s a e used in some spec-
imens o A chaeonyc e is igonodon (e.g.,
SMF Me 80/1379), while all su u es a e s ill
isible in o he s (e.g., SMF Me 963a). Ica-
onyc e is he e o e appea s o ha e had an
on ogene ic pa e n o s e nal usion simila
o ha seen in A chaeonyc e is and many ex-
an ba s.
based on mo phology and collec ion locali y. Based on
compa isons wi h Jepsens o iginal s e eopho og aphs
(many o which emain unpublished), we ound ha UW
21481a–b appea s iden ical o he holo ype o Ica on-
yc e is index PU 18150 in i ually all espec s (al hough
see ex o a discussion o s e nal usion). Fo ea m
leng h in UW 21481a–b is 47.5 mm; o ea m leng h in
PU 18150 is 48.0 mm. Measu emen s o skull leng h,
ibia leng h, and leng h o a ious wing elemen s in UW
21481a–b a e wi hin
6
2 mm o hose epo ed by Jepsen
(1966) o PU 18150. This ange o a ia ion is well
wi hin ha seen in ex an species o simila body size
(e.g., Swanepoel and Genoways, 1979). UW 21481a–b
was collec ed om ea ly Eocene sedimen s o he G een
Ri e Fo ma ion o Wyoming, he same deposi s om
which PU 18150 was collec ed (Jepsen, 1966). Because
de ailed den al compa isons be ween PU 18150 and UW
21481a–b we e p ecluded by p ese a ion o he la e
specimen, i emains possible ha UW 21481a–b ep-
esen s a species o Ica onyc e is dis inc om index;
howe e , we ound no da a o suppo such an in e p e-
a ion. Acco dingly, we e e UW 21481a–b o Ica on-
yc e is index Jepsen, 1966.
G ande (1980: ig. III.26) igu ed a badly p ese ed
ba ossil om Middle Eocene deposi s o he G een
Ri e Fo ma ion o Colo ado ha he no ed ‘‘p obably
ep esen s an undesc ibed species.’’ Un o una ely, his
specimen is in a p i a e collec ion and has ne e been
desc ibed. This eco d sugges s, howe e , ha ba spe-
cies o he han Ica onyc e is index may ha e been p es-
en in wes e n No h Ame ica in he Eocene.
Ano he ea u e ci ed by Jepsen (1966,
1970) and No acek (1987) as a p imi i e
cha ac e o Ica onyc e is is he o ien a ion
o he head and neck o emu , which hey
desc ibed as se a an angle o he long axis
o he sha (al hough no o he ex en seen
in e es ial mammals). No acek (1987: 13)
no ed ha in his ai Ica onyc e is is ‘‘clea -
ly mo e conse a i e han li ing mega- and
mic ochi op e ans, whe e he head is nea ly
aligned wi h he sha and he neck is e y
sho o absen .’’ Howe e , ou compa isons
o he emu o Ica onyc e is wi h hose o
ex an ba s indica e ha No acek was misled
by he angle o p esen a ion o he emu in
PU 18150, which p o ides an oblique iew
o he head and neck ( ig. 23). We ound ha
his iew accen ua es a ela i ely small de-
g ee o o se o he emo al head. Compa -
isons o he emu o PU 18150 and UW
21481a–b wi h hose o ex an megachi op-
e ans and noc ilionoids o simila body size
indica e ha he head and neck o he emu
o Ica onyc e is a e se a he same angle
( ela i e o he sha o he emu ) as seen in
many ex an ba s. Indeed, he p oximal emu
o Rouse us aegyp iacus is i ually iden ical
o ha o Ica onyc e is when iewed om
he same angle. Ica onyc e is hus canno be
conside ed mo e p imi i e han ex an ba s in
e ms o emu mo phology.
Ou su ey hus limi s he lis o p imi i e
ea u es ound in Ica onyc e is (bu modi ied
in all o he known ba s) o he ollowing: (1)
ela i ely sho adius, and (2) comple e pha-
langeal o mula (2-3-3-3) on he digi s o he
wings. No acek (1987) no ed ha ela i e
leng h o he adius had ne e been quan i-
ied, bu his was subsequen ly done by Ha-
be se ze and S o ch (1987). They ound ha
he a io o hume us leng h o adius leng h
was somewha la ge (and he adius he e-
o e ela i ely sho e ) in Ica onyc e is com-
pa ed o o he Eocene ba s, al hough he di -
e ence be ween he ange o alues o Ica-
onyc e is (0.715–0.732) and A chaeonyc e -
is (0.680–0.715) is no g ea .
The diagnosis o Ica onyc e is he e o e
es s on jus wo plesiomo phic ea u es.
Howe e , i seems unlikely ha mul iple lin-
eages a e ep esen ed in his g oup (a leas
among he ou specimens ha we examined)
because he ange o a ia ion in size and
42 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
TABLE 3
Apomo phies
a
Diagnosing Gene a o Eocene Ba s
Ica onyc e is (one species, I. index)
●
No apomo phies
A chaeonyc e is ( h ee species, A. igonodon, A. pollex, A. b ailloni)
●
Ec o lexus deeply e ac ed be ween pa as yle and me as yle on M1 and M2
●
Ven al p ocess o manub ium o s e num o ien ed a app oxima ely 90
8
o axis o body o manubium
●
Pos e io xiphis e num na ow, wi hou la e al la e
●
Do sal ischial ube osi y p esen
Palaeochi op e yx ( wo species, P. upaiodon and P. spiegeli)
●
Me aconule p esen on M1 and M2
●
Leng h o pos pa ac is a equals leng h o p eme ac is a on M3
●
Ven al accesso y p ocess p esen on ce ical e eb a 5
●
In aspinous ossa o scapula ela i ely b oad
●
Ven al p ojec ion p esen on an e omedial lange o scapula
Hassianyc e is ( ou species, H. messelensis, H. magna, H. e illiodi, and H. joeli)
●
P o oconid and hypoconid on mola ee h excep ionally all and obus
●
Las lowe p emola no mola i o m, me aconid lacking and alonid sho
●
Fi s uppe p emola educed o iny peg o absen
●
Mandible deep, hickened do so en ally
●
Radius unusually long and s ongly cu ed
●
T ochi e ex ends well beyond le el o head hume us
●
Me aca pal o digi V ela i ely sho compa ed o me aca pals III and IV
a
No ound in any o he Eocene axon. Membe ship in ex an amilies is p ecluded by documen ed o in e ed
absence o de i ed cha ac e s diagnos ic o hose axa ( o a lis o cha ac e s diagnosing ex an lineages see Simmons,
1998: ables 1, 2). Sou ces: Russell and Sige´, 1970; Smi h and S o ch, 1981; Habe se ze and S o ch, 1987, 1989;
S o ch and Habe se ze , 1988; pe sonal obs.).
shape o skele al elemen s among e e ed in-
di iduals appea s o all wi hin limi s cha -
ac e is ic o ex an species (see oo no e 2).
We ha e he e o e assumed monophyly o
Ica onyc e is o he pu poses o he cu en
s udy, al hough we ecognize ha he Eu o-
pean ma e ial e e ed o ?Ica onyc e is men-
ui and ?Ica onyc e is sp. may ep esen one
o mo e di e en clades.
OUTGROUPS
Only one ou g oup is necessa y o oo a
phylogene ic ee (Nixon and Ca pen e ,
1993), bu a leas wo ou g oups a e usually
included in cladis ic analyses o es ablish
cha ac e pola i ies and o pe mi es ing o
ing oup monophyly (Maddison e al., 1984).
Ideally, ou g oups should comp ise he nea -
es sis e - axa o he ing oup because he
p obabili y o homoplasy inc eases wi h ime
since di e gence om a common ances o .
The iden i y o he sis e -g oup o ba s is s ill
he subjec o conside able deba e, and ecen
cong uence s udies ha e ailed o esol e ex-
is ing con lic s be ween mo phological da a
and se e al molecula da a se s (e.g., Alla d
e al., 1996; Miyamo o, 1996; S anhope e
al., 1996). Ne e heless, p ac ical conside a-
ions equi e ha a choice o ou g oups be
made o acili a e analyses o ela ionships
and cha ac e e olu ion wi hin Chi op e a. A
close ela ionship be ween Chi op e a and
De mop e a (colugos o lying lemu s) is
s ongly suppo ed by mo phological da a
(Wible and No acek, 1988; No acek, 1992,
1994; Simmons, 1993a, 1994, 1995; Szalay
and Lucas, 1993; Wible, 1993; Simmons and
Quinn, 1994; Miyamo o, 1996; G i i hs,
MS), 12S DNA sequences (sampled om
mo e han 150 species ep esen ing 20 o -
de s; V ana, 1994), and combined mo pho-
logical da a and cy och ome oxidase II
(COII) gene sequences (No acek, 1994). Ac-
co dingly, Simmons (1998) used De mop e a
as an ou g oup o in es iga ing in e ela ion-
ships o ba s. Mo phological synapomo -
1998 49SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
maxillae ha lie on he ace adjacen o he
ex e nal na ial opening. The nasal b anches
o he p emaxillae a e ela i ely well de el-
oped in Ica onyc e is,A chaeonyc e is,Has-
sianyc e is, and Palaeochi op e yx, as hey
a e in mos ex an ba s. In con as , he nasal
b anches o he p emaxillae in hinolophoids
a e ei he en i ely absen o a e educed o
iny splin s o bone. The nasal b anches o
he p emaxillae a e well de eloped in bo h
ou g oups, sugges ing ha he s a e seen in
he Eocene ossil ba s is p imi i e.
Cha ac e 11: Pala al b anches o p e-
maxillae well de eloped (0); o educed o
absen (1). The pala al b anches o he p e-
maxillae a e de ined as hose po ions o he
p emaxillae ha con ibu e o he an e io
pala e. The pala al b anches o he p emax-
illae a e ela i ely well de eloped in Ica on-
yc e is and many ex an ba s (Nyc e idae,
Rhinolophidae, Phyllos omidae, Noc ilioni-
dae, Mo moopidae, Mys acinidae, Myzopod-
idae, Thy op e idae, Na alidae, some Molos-
sinae). In con as , he pala al b anches o he
p emaxillae a e ei he educed o iny splin s
o bone o a e en i ely absen in o he ex an
lineages. This educ ion is appa en ly a de-
i ed condi ion since well de eloped pala al
b anches a e p esen in bo h ou g oups and
mos o he mammals.
The condi ion o he an e io pala e in A -
chaeonyc e is and Palaeochi op e yx could
no be de e mined om a ailable specimens.
We ound ha he pala al b anches o he p e-
maxilla ies a e no isible in specimens p e-
pa ed in do sal iew, and ha he mandibula
ami obscu e he pala e in all specimens p e-
pa ed in en al iew. Acco dingly, hese
o ms a e sco ed ‘‘?’’ o his cha ac e .
Smi h and S o ch (1981: 154, 164) no ed
ha in Hassianyc e is he ‘‘pala al b anch [o
he p emaxilla] no well de eloped, p emax-
illa ies no used [a midline],’’ and ha ‘‘ he
appa en shape and educ ion o he p emax-
illa y is qui e eminiscen o he de i ed con-
di ion o his elemen in emballonu ids.’’
Howe e , he sou ce(s) o hese obse a ions
we e no discussed, and we we e unable o
con i m hem du ing ou mic oscopic exam-
ina ion o specimens. Smi h and S o ch ap-
pa en ly desc ibed hese ea u es based on a-
diog aphs, al hough he published e sions
a e oo uzzy o be o use. In bo h published
adiog aphs (Smi h and S o ch, 1981: igs. 1,
2), an image o he an e io den i ion is su-
pe imposed on he an e io pala e. Consid-
e ing he complex na u e o his egion and
he high densi y o he den al images, any
obse a ions o pala al mo phology in hese
specimens mus be iewed wi h cau ion. We
conside ed sco ing Hassianyc e is as ha ing
a educed pala al b anch o he p emaxilla
based on he epo by Smi h and S o ch
(1981), bu subsequen ly no ed ha educ ion
and lack o usion o he pala al b anches
we e omi ed om Habe se ze and S o ch’s
(1987) diagnosis o Hassianyc e ididae
(which con ains only Hassianyc e is). Gi en
he p ominence o hese ea u es in Smi h
and S o ch’s o iginal diagnosis o Hassianyc-
e is, we can only conclude ha he e is con-
side able unce ain y abou he s uc u e o
he p emaxilla in hese ba s. Acco dingly, we
sco e Hassianyc e is as ‘‘?’’ o his cha ac-
e .
Cha ac e 15: Ha d pala e ex ends pos-
e io ly in o in e o bi al egion (0); o e -
mina es ei he a o an e io o le el o zy-
goma ic oo s (1). The ha d pala e, which
o ms a bony sepa a ion be ween he o al and
nasal passages, e mina es pos e io ly a he
mesop e ygoid ossae. Posi ion o he pos-
e io edge o he ha d pala e a ies indepen-
den ly o he pos e io ex en o he mola
oo h ow, and is appa en ly linked o he
s uc u e o he nasal passages and so is-
sues o he pha yngeal egion. The ha d pal-
a e ex ends pos e io ly in o he in e o bi al
egion in A chaeonyc e is,Palaeochi op e-
yx, one o he ou g oups (Scanden ia), and
many ex an ba s (P e opodidae, some Em-
ballonu idae, some Hipposide inae, some
Phyllos omidae, Mo moopidae, Noc ilioni-
dae, Mys acinidae, Myzopodidae, Thy op e -
idae, some Fu ip e idae, Na alidae, An ozo-
idae, Tomopea inae, some Molossidae, and
Vespe ilionidae). In con as , he ha d pala e
e mina es ei he a o an e io o he le el o
he zygoma ic oo s in he o he ou g oup
(De mop e a) and all emaining ex an ba s.
Lack o ag eemen be ween he wo ou -
g oups p ecludes a p io i de e mina ion o
he p imi i e condi ion o his cha ac e . The
pos i ion o he pos e io end o he ha d pal-
a e in Ica onyc e is and Hassianyc e is could
no be de e mined due o he posi ion o he
50 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
mandibula ami in all a ialable specimens,
so hese o ms a e sco ed ‘‘?’’ o his cha -
ac e .
Cha ac e 16: Two uppe inciso s in each
side o jaw (0); o one inciso (1); o uppe
inciso s absen (2). Because o di icul ies
associa ed wi h de e mining homologies o
he an e io den al loci among di e en ba
lineages ( e iewed by Slaugh e , 1970), we
chose o sco e he numbe o ee h in a ious
den al egions (i.e., uppe inciso s, lowe in-
ciso s, uppe p emola s, lowe p emola s)
a he han a emp ing o sco e p esence/ab-
sence o ee h a speci ic loci. Al hough some
po en ially in o ma i e pa e ns may be o e -
looked by his me hod, his app oach p e-
se es ha homology in o ma ion o which
we a e mos con iden and pe mi s sco ing o
all axa (including he ou g oups) o each
cha ac e .
The size o one o mo e ee h in a numbe
o species has been educed o he poin
whe e hese ee h a e conside ed es igial
(Slaugh e , 1970). This is pa icula ly com-
mon in he p emola den i ion. In some in-
s ances he e is wi hin-species (and e en
wi hin-indi idual) polymo phism wi h e-
spec o p esence/absence o es igial ee h
(e.g., he an e io uppe p emola in Rhino-
lophus cli osus, he middle lowe p emola
in Ch o op e us au i us). In ou expe ience,
olde indi iduals equen ly lack such ee h,
while hey a e o en p esen in younge ani-
mals. This pa e n sugges s ha es igial
ee h a e o en los du ing he li e ime o he
indi idual. Consequen ly, we ha e sco ed he
axa in ques ion as ha ing he highe den al
o mula in cases o wi hin-species polymo -
phism in p esence/absence o es igial ee h.
Two uppe inciso s a e p esen on each
side o he jaw in A chaeonyc e is,Hassi-
anyc e is, and Palaeochi op e yx. This is he
same uppe inciso o mula as seen in many
ex an ba s (some P e opodidae, some Em-
ballonu idae, Nyc e idae, some Phyllos omi-
dae, Mo moopidae, Noc ilionidae, Na alo-
idea, some Vespe ilioninae, Miniop e inae,
Myo inae, Mu ininae, and Ke i oulinae). In
con as , only one inciso is p esen on each
side o he uppe jaw in some P e opodidae,
some Emballonu idae, Rhinopoma oidea,
Rhinolophidae, some Phyllos omidae, Mys-
acinidae, Molossoidea, and some Vespe i-
lioninae. The uppe inciso s a e en i ely ab-
sen in some P e opodidae and Megade ma-
idae. Bo h ou g oups ha e wo uppe inci-
so s on each side o he jaw. In his con ex ,
he s a e seen in A chaeonyc e is,Hassianyc-
e is, and Palaeochi op e yx appa en ly ep-
esen s he p imi i e condi ion. We we e un-
able o unambiguously de e mine he numbe
o uppe inciso s in Ica onyc e is, so his ax-
on is sco ed ‘‘?’’ o his cha ac e .
Cha ac e 17: Th ee lowe inciso s in
each side o jaw (0); o wo inciso s (1); o
one inciso (2); o inciso s absen (3). Th ee
lowe inciso s a e p esen on each side o he
jaw in Ica onyc e is,A chaeonyc e is,Has-
sianyc e is, and Palaeochi op e yx. This is
he same lowe inciso o mula seen in many
ex an ba s (some Emballonu idae, Nyc e i-
dae, Na aloidea, some An ozoidae, some
Molossinae, and Vespe ilionidae). Two low-
e inciso s a e p esen on each side o he
lowe jaw in some P e opodidae, some Em-
ballonu idae, Rhinopoma oidea, Megade -
ma idae, Rhinolophidae, some Phyllos omi-
dae, Mo moopidae, some An ozoidae, To-
mopea inae, and some Molossinae. A single
lowe inciso is p esen on each side in some
P e opodidae, some Phyllos omidae, Noc i-
lionidae, Mys acinidae, and some Molossi-
nae. Finally, he lowe inciso s a e en i ely
absen in some P e opodidae and some Phyl-
los omidae. Bo h ou g oups ha e h ee lowe
inciso s on each side. In his con ex , he
s a e seen in he Eocene ba gene a appa -
en ly ep esen s he p imi i e condi ion.
Cha ac e 18: Th ee uppe p emola s in
each side o jaw (0); o wo p emola s (1);
o one p emola (2). Th ee uppe p emola s
a e p esen on each side o he jaw in Ica-
onyc e is,A chaeonyc e is,Hassianyc e is,
and Palaeochi op e yx magna.Hassianyc e -
is messelensis has only wo uppe p emola s
( ig. 24), appa en ly ha ing los he iny peg-
like an e io p emola seen in H.magna.
Among ex an ba s, h ee uppe p emola s
a e p esen in some P e opodidae, some
Phyllos omidae, Myzopodidae, Thy op e i-
dae, Na alidae, some Myo inae, and Ke i ou-
linae. In con as , wo uppe p emola s a e
p esen on each side o he jaw in some P e -
opodidae, Emballonu idae, some Megade -
ma idae, Rhinolophinae, some Hipposide i-
nae, some Phyllos omidae, Mo moopidae,
1998 51SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 24. Hassianyc e is messelensis (SMF
ME 1414a). Closeup la e al iew o skull. No e
he obus den a y and den i ion, and he absence
o he i s uppe p emola . A es igal i s uppe
p emola is p esen in H.magna. AN, angula
p ocess o den a y; VAP, en al accesso y p o-
cess o e eb a (C3?). Scale ba
5
10 mm. F om
Habe se ze and S o ch (1987: ig. 5). Pho o by
E. Pan ak (Senkenbe gmuseum).
Mys acinidae, Fu ip e idae, some Molossi-
nae, some Vespe ilioninae, Miniop e inae,
some Myo inae, and Mu ininae. A single up-
pe p emola is p esen on each side in Rhin-
opoma oidea, Nyc e idae, some Megade -
ma idae, some Hipposide inae, some Phyl-
los omidae, Noc ilionidae, An ozoidae, To-
mopea inae, some Molossinae, and some
Vespe ilioninae. Among he ou g oups,
Scanden ia has h ee uppe p emola s, while
De mop e a has wo. Lack o ag eemen be-
ween he wo ou g oups p ecludes a p io i
de e mina ion o he p imi i e condi ion o
his cha ac e .
Cha ac e 19:Middle uppe p emola
wi h h ee oo s (0); o wi h wo oo s (1);
o wi h one oo (2). Among ba s wi h h ee
uppe p emola s, he middle p emola (usu-
ally designa ed P3) may ha e ei he one, wo,
o h ee oo s. P3 has h ee oo s in Ica on-
yc e is. In con as , his oo h has only wo
oo s in A chaeonyc e is,Hassianyc e is, and
Palaeochi op e yx. Among ex an ba s, P3 is
iple- oo ed only in Na alidae. P3 has wo
oo s in P e opodidae, some Phyllos omidae,
Thy op e idae, and some Ke i oulinae. P3 is
single- oo ed in some Phyllos omidae, My-
zopodidae, Myo inae, and some Ke i ouli-
nae. Th ee oo s a e p esen on P3 in bo h
ou g oups, sugges ing ha he iple- oo ed
condi ion seen in Ica onyc e is is p imi i e.
The double- oo ed condi ion seen in he o h-
e Eocene axa is appa en ly de i ed. Be-
cause i is hough ha P3 is o en he i s
oo h los om he p emola den i ion in spe-
cies wi h a educed numbe o ee h (Slaugh-
e , 1970), we sco ed his cha ac e only in
o ms ha e ain h ee uppe p emola s (s a e
‘‘0’’ o cha ac e 18 abo e). Acco dingly,
Emballonu idae, Yinochi op e a, Mo moopi-
dae, Noc ilionidae, Mys acinidae, Fu ip e i-
dae, Molossoidea, Vespe ilioninae, Miniop-
e inae, and Mu ininae a e sco ed ‘‘-’’ o
his cha ac e .
Cha ac e 20: Th ee lowe p emola s in
each side o jaw (0); o wo p emola s (1).
Th ee p emola s a e p esen on each side o
he lowe jaw in Ica onyc e is,A chaeonyc-
e is,Hassianyc e is and Palaeochi op e yx.
Among ex an ba s, h ee lowe p emola s
a e p esen in Rhinolophinae, some Phyllos-
omidae, Mo moopidae, Na aloidea, some
Vespe ilioninae, Miniop e inae, some My-
o inae, and Ke i oulinae. Only wo lowe
p emola s a e p esen in all o he lineages.
Among he ou g oups, Scanden ia has h ee
lowe p emola s, while De mop e a has wo.
Lack o ag eemen be ween he wo ou -
g oups p ecludes a p io i de e mina ion o
he p imi i e condi ion o his cha ac e .
Cha ac e 21: Lowe i s and second mo-
la s wi h p imi i e ibosphenic a angemen
o cusps and c is ids (0); o nyc alodon (1);
o myo odon (2); o ee h modi ied o ui
and/o nec a o blood eeding, cusps and
c is ids no dis inc (3). Menu and Sige´
(1971) discussed mola mo phology in ba s,
and dis inguished wo o ms o he alonid o
m1 and m2 ha hey e med ‘‘nyc alodon y’’
and ‘‘myo odon y.’’ In bo h mo pho ypes,
he hypoconulid lies adjacen o he en ocon-
id on he lingual edge o he oo h, ha ing
shi ed lingually om he midline posi ion
(whe e i is mo e o less equidis an om he
labial and lingual bo de s o he oo h) cha -
ac e is ic o p imi i e he ian ibosphenic
mola s. Nyc alodon y is de ined by he po-
si ion o he pos c is id, which connec s he
hypoconid wi h he hypoconulid in nyc alo-
don o ms. Myo odon y is dis inguished by
an al e na i e a angemen in which he pos -
52 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
c is id bypasses he hypoconulid o connec
ins ead wi h he en oconid. The pos c is id is
ela i ely high and o ms a sha p shea ing
edge in bo h condi ions, bu no c is id uns
be ween he en oconid and hypoconulid.
Al hough i is easy o classi y mola mo -
phology o mos ex an ba s, Menu and Sige´
(1971) no ed ha p oblems a ise when Eo-
cene ba s a e conside ed because he hypo-
conulid is loca ed in he midline posi ion in
some axa, always connec ed o he hypocon-
id and some imes connec ed o he en oconid
by low c is ids. The pos e io edge o he al-
onid is hus domina ed by a la ge conical
cusp ( he hypoconulid) a he han by a
shea ing c es ( he pos c is id). This mo -
phology, e med ‘‘a chaic’’ by Menu and
Sige´ (1971), appea s in nume ous La e C e-
aceous and Paleocene mammals, including
o ms hough o ep esen he ea lies mem-
be s o se e al eu he ian adia ions (e.g., p i-
ma es, insec i o es; Menu and Sige´, 1971).
This appa en ly ep esen s he p imi i e i-
bosphenic condi ion.
Possible di e ences in unc ion o he
h ee alonid ypes discussed abo e ha e no
ye been examined. Nyc alodon and myo o-
don ee h a e a leas supe icially simila in
e ms o he leng h and heigh o he pos -
c is id, which may indica e an inc eased e-
liance on shea ing as compa ed o he p im-
i i e ibosphenic condi ion. Inc eased shea -
ing has been associa ed wi h specializa ions
o eeding on ha d-shelled bee les (F eeman,
1979) and on so -bodied p ey (S ai , 1993).
F eeman (1979) ound ha molossid ba s ha
eed on bee les ha e ela i ely la ge ee h
wi h longe shea ing cusps han do molossids
ha eed on so -bodied p ey. Al e na i ely,
S ai (1993) ound ha mammalian insec i-
o es ha eed p edominan ly on so -bodied
p ey (e.g., mo hs and ca e pilla s) ha e el-
a i ely longe shea ing c es s (summed
ac oss he en i e mola ) han do close ela-
i es ha ha e mo e gene alized eeding hab-
i s and consume la ge quan i ies o ha d-bod-
ied bee les. Wha e e he ela ionship be-
ween shea ing and p ey ype, hese s udies
sugges ha a ia ion in alonid o m may be
indica i e o di e ences in die a y habi s.
Howe e , he e does no seem o be any ob-
ious co ela ion be ween eeding habi s and
alonid ype in ex an ba s. Fo example,
hinolophids ha eed on so -bodied p ey
(e.g., Rhinolophus blasii) and hose ha eed
on ha d-bodied p ey (e.g., Hipposide os
comme soni) bo h ha e nyc alodon ee h
(S ai , 1993). I he e o e seems p ema u e
o d aw any unc ional conclusions om
ans o ma ions in alonid s uc u e in ba s.
The i s lowe mola s exhibi he p imi-
i e ibosphenic a angemen o he alonid
cusps and c is ids in A chaeonyc e is. In con-
as , he lowe mola s a e nyc alodon in
Hassianyc e is. Bo h nyc alodon and myo o-
don condi ions appea in Palaeochi op e yx,
al hough ne e in he same indi idual. Bo h
nyc alodon and myo odon specimens ha e
been e e ed o Palaeochi op e yx upaio-
don (e.g., by Russel and Sige´, 1970), aising
he possibli y o wi hin-species polymo -
phism. Howe e , we a e unawa e o any
cases o wi hin-species polymo phism in his
cha ac e in ex an ba s. This sugges s ha P.
upaiodon may be a composi e species, a hy-
po hesis ha will equi e es ing wi h addi-
ional da a. Meanwhile, we sco e Palaeochi-
op e yx as exhibi ing bo h nyc alodon y and
myo odon y.
Van Valen (1979) claimed ha he lowe
mola s o Ica onyc e is a e nyc alodon , an
obse a ion ha was subsequen ly ci ed by
o he au ho s (e.g., Ginge ich, 1987). We dis-
ag ee wi h Van Valen’s assessmen . Ou ob-
se a ions indica e ha Ica onyc e is index
had p imi i e ibosphenic lowe mola s. The
Eu opean specimens e e ed o ?Ica onyc-
e is by Russell e al. (1973) include bo h
nyc alodon and p imi i e ibosphenic mo -
pho ypes. Howe e , his ma e ial consis s o
isola ed ee h only, and we conside he a -
ini es o hese specimens o each o he and
o Ica onyc e is index o be unce ain a bes .
Acco dingly, we sco e Ica onyc e is as ex-
hibi ing only he p imi i e ibosphenic con-
di ion.
Among he ex an amilies, he p imi i e
ibosphenic condi ion o he alonid is seen
in ossil membe s o Hipposide inae (i.e.,
Palaeophyllopho a) and Megade ma idae
(i.e., Nec oman is; Bea d e al., 1992). Nyc-
alodon y occu s in Emballonu idae, Rhino-
poma oidea, Nyc e idae, ex an Megade ma-
idae, Rhinolophinae, ex an Hipposide inae,
some Phyllos omidae, Mo moopidae, Fu ip-
e idae, Na alidae, some Molossinae, some
1998 53SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 25. Ica onyc e is index (UW 2244). Ven al iew o he basic anium and pos e io lowe jaws.
F om No acek (1987: ig. 2). An, angula p ocess o den a y; Co, cochlea; Ec , ec o ympanic; Gl,
glenoid ossa; Oa, o bicula apophysis; Sl, s ylohyal; Z, zygoma ic a ch.
Vespe ilioninae, Miniop e inae, some My-
o inae, and Mu ininae. In con as , myo o-
don y is ound in some Phyllos omidae, Noc-
ilionidae, Mys acinidae, Myzopodidae, Thy-
op e idae, An ozoidae, Tomopea inae, some
Molossinae, some Vespe ilioninae, some
Myo inae, and Ke i oulinae. In P e opodidae
and some Phyllos omidae he ee h a e mod-
i ied o ui and/o nec a o blood eeding
and he cusps and c is ids a e no dis inc .
Among he ou g oups, Scanden ia exhibi s
he p imi i e ibosphenic condi ion, bu we
could no sco e De mop e a owing o i s
unique den al mo phology. Al hough pola i y
canno be asce ained unambiguously, he
da a sugges ha he p imi i e ibosphenic
condi ion is p imi i e o ba s.
Cha ac e 22: Lowe jaw wi h elonga e
angula p ocess (0); o wi hou elonga e an-
gula p ocess (1). The angula egion o he
lowe jaw in mammals is he si e o inse ion
o jaw adduc o muscles including m. mas-
se e pa s supe icialis, m. masse e pa s p o-
undus, m. zygoma icomandibula is, m. man-
dibuloau icula is, and m. p e ygoideus in e -
nus (
5
medial p e ygoid; Tu nbull, 1970). In
many ‘‘gene alized’’ eu he ian mammals and
some specialized ca ni o es, a bony ex en-
sion o he angle— e med he angula p o-
cess—lies be ween he inse ions o m. p e -
ygoideus in e nus and m. masse e . These
muscles inse bo h on he angula p ocess
and on a common aphe do sal, pos e io , and
en al o he angula p ocess (Tu nbull,
1970). This pa e n o muscle inse ion is
seen in ba s ha ha e an elonga e angula
p ocess (S o ch, 1968; Kallen and Gans,
1972). Bo h m. masse e and m. p e ygoideus
in e nus ac o close and p o ude he lowe
jaw. M. masse e also pulls he ipsila e al
den a y ( he den a y on he same side o he
skull as he muscle) la e ally, while m. p e -
ygoideus in e nus pulls he ipsila e al den a y
lingually (S o ch, 1968; Kallen and Gans,
1972).
The lowe jaw has an elonga e angula
p ocess in Ica onyc e is ( igs. 25, 26), A -
chaeonyc e is,Hassianyc e is ( ig. 24), and
Palaeochi op e yx ( ig. 27). This is simila o
he condi ion seen in mos ba s. In con as ,
he angula p ocess is ei he e y sho ( el-
a i e o i s do so en al wid h) o is e ec-
i ely absen in P e opodidae, C aseonyc e -
idae, and some Nyc e idae. An elonga e an-
gula p ocess is p esen in Scanden ia bu no
in De mop e a. Lack o ag eemen be ween
he wo ou g oups p ecludes a p io i de e -
mina ion o he p imi i e condi ion o his
cha ac e . Howe e , occu ence o an elon-
ga e angula p ocess in many p esumably
p imi i e eu he ians (e.g., lep ic ids, Asio-
yc es, lipo yphlans) sugges s ha p esence
o an elonga e angula p ocess may be p im-
i i e o Eu he ia.
54 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 26. Ica onyc e is index (UW 21481a–b). S e eopho og aphs o a en al iew o he basic anial
egion ( op) wi h labelled key pho og aph (bo om cen e ). AN, angula p ocess o den a y, BS, basi-
sphenoid; CP, co onoid p ocess o den a y; CTP, caudal p ocess o ympanic; ER, epi ympanic ecess;
PGF, pos glenoid o amen; PR, pe io ic; SL, ylohyal; SQ, squamosal; Z, zygoma ic a ch. Pho og aphs
by C. Ta ka.
Cha ac e 23: Angula p ocess p ojec s a
o below le el o occlusal plane o oo h ow,
well below co onoid p ocess (0); o angula
p ocess p ojec s abo e le el o occlusal plane
o oo h ow, a same le el as he co onoid
p ocess (1). The angula p ocess, when p es-
en , ex ends om he pos e o en al ‘‘co ne ’’
o he lowe jaw. The angula p ocess p ojec s
a o below he le el o he occlusal plane o
he oo h ow (well below he co onoid p o-
cess) in Ica onyc e is ( igs. 25, 26), A chaeo-
nyc e is,Hassianyc e is ( ig. 24), and Palaeo-
chi op e yx ( ig. 27). This is simila o he
condi ion seen in mos ex an ba s and bo h
ou g oups. In con as , he angula p ocess
p ojec s abo e he le el o he occlusal plane
o he oo h ow ( oughly a he same le el as
he co onoid p ocess) in Myzopodidae, Thy-
op e idae, and Fu ip e idae. In his con ex ,
he condi ion in he Eocene ossil ba s appea s
o be p imi i e. This cha ac e canno be e al-
ua ed in axa ha lack an angula p ocess
(e.g., axa sco ed ‘‘1’’ o 22); hese o ms a e
he e o e sco ed ‘‘-’’ o his cha ac e ).
1998 55SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 27. Palaeochi op e yx upaiodon, posi i e p in s o adiog aphs; om No acek (1987: ig. 4).
A. Do so en al iew o he skull o SMF ME 788b. B. La e al iew o SMF ME 1127. An, angula
p ocess o den a y; Co, cochlea; Cp, co onoid p ocess o den a y; Ec , ec o ympanic; Sl, s ylohyal; Z,
zygoma ic a ch.
Cha ac e 24: Pos o bi al p ocess p esen
(0); o absen (1). The pos o bi al p ocess is
a la e ally p ojec ing p ocess o he on al
ha o ms pa o he pos e odo sal im o
he o bi . A pos o bi al p ocess is p esen in
A chaeonyc e is,Hassianyc e is, and Pa-
laeochi op e yx. Among ex an o ms, a
pos o bi al p ocess is p esen only in P e o-
podidae, Emballonu idae, Nyc e idae, and
Megade ma idae. All o he ex an ba s lack a
pos o bi al p ocess. A pos o bi al p ocess is
p esen in bo h ou g oups, sugges ing ha
p esence o his s uc u e ep esen s he p im-
i i e condi ion. We we e no able o asce ain
whe he his p ocess is p esen in Ica onyc-
e is; his axon is he e o e sco ed ‘‘?’’ o
his ea u e.
Cha ac e 25: Pa s cochlea is o pe osal
su u ed o basisphenoid (0); o loosely a -
ached o basisphenoid ia ligamen s and/o
hin splin s o bone (1). The pe osal o all
ex an mic ochi op e ans (wi h he excep ion
o some Emballonu idae) is no used wi h
o su u ed o he basisphenoid. Ins ead, he
pa s cochlea is is loosely a ached o he ba-
sisphenoid ia ligamen s and/o hin splin s
o bone ( ig. 28). This condi ion (some imes
e e ed o as ‘‘cochlea isola ion’’) is
hough o unc ion in educing bone con-
duc ion o la yngeal ib a ions (Henson,
1970; Van Valen, 1979). Al hough he e is
some a ia ion in he deg ee o isola ion o
he cochlea as judged by he ela i e sizes o
he openings ha su ound he pe io ic ( he
an e ola e al py i o m enes a, an e omedial
o medial pa s o he basicochlea issu e,
and pos e omedial jugula o amen; No a-
cek, 1991), his ‘‘loosely a ached’’ condi ion
clea ly di e s om he ypical mammalian
pa e n in which he pe io ic is i mly su u ed
o he su ounding bones o he basic anium
including he basisphenoid.
56 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 28. Nyc alus noc ula (Vespe ilioninae; ed awn om Henson, 1970: ig. 9). Ven al iew o
he basic anium and aud o y egion showing majo os eological ea u es, blood essels, and muscles.
The ec o ympanic and malleus ha e been e lec ed la e ally on he igh side o he skull. APP, an e io
p ocess o pe osal; BF, basicochlea issu e; BO, basioccipi al; BOP, basioccipi al pi ; CO, cochlea; ECT,
ec o ympanic; ETR, epi ympanic ecess; FC, enes a cochleae (
5
ound window); FM, o amen mag-
num; FO, o amen o ale; FV, enes a es ibuli (
5
o al window); GF, glenoid ossa; I, incus; ICA,
in e nal ca o id a e y; M, malleus; MS, m. s apedius; MTT, m. enso ympani; OAM, o bicula apoph-
ysis o malleus; OC, occipi al condyle; PF, py i o m enes a; PGF, pos glenoid o amen; PP, pa occipi al
p ocess; PT, p e ygoid hamulus; SF, s apedial ossa; ST, s apes; STA, s apedial a e y; T, endon o m.
enso ympani; TM, ympanic memb ane; Z, zygoma ic a ch.
Based on obse ed gaps and pa e ns o
b eakage in he basic anium o p ese ed
specimens o A chaeonyc e is and Palaeo-
chi op e yx, i appea s ha he pe io ic was
loosely a ached o he basisphenoid (and he
cochlea ela i ely isola ed) in hese axa. In
con as , he pe io ic is su u ed o he basi-
sphenoid in P e opodidae and some Embal-
lonu idae. A simila condi ion is seen in bo h
ou g oups and mos o he mammals, sug-
ges ing ha he ‘‘su u ed’’ condi ion is p im-
i i e, and ha he ‘‘loosely a ached’’ con-
di ion seen in A chaeonyc e is and Palaeo-
chi op e yx is de i ed. We we e unable o de-
e mine he ype o pe io ic connec ion in
Ica onyc e is and Hassianyc e is; hese axa
a e he o e sco ed ‘‘?’’ o his cha ac e .
Cha ac e 26: Cochlea no enla ged (0);
o mode a ely enla ged (1); o g ea ly en-
la ged (2). Much has been w i en abou
cochlea size and echoloca ion in ba s, wi h
mos au ho s ag eeing ha cochlea enla ge-
men (speci ically enla gemen o he basal
u n) is a specializa ion o pe cei ing he
echoes o high- equency echoloca ion sig-
nals (Henson, 1970; Segall, 1971; B uns e
al., 1983–1984; No acek, 1985a, 1987,
1991; Habe se ze and S o ch, 1992). By
plo ing maximum ex e nal cochlea wid h
agains skull leng h, No acek (1985a, 1987,
1991) ound ha ex an Mic ochi op e a and
Megachi op e a ha e nono e lapping dis i-
bu ions, wi h mic ochi op e ans ha ing con-
sis en ly la ge cochleae. The megachi op e -
an condi ion (i.e., a ela i ely small cochlea)
was in e p e ed as he p imi i e condi ion be-
cause i also occu s in many o he small o
medium-sized mammals (No acek, 1985a,
1987, 1991).
No acek (1985a, 1987, 1991) also es i-
ma ed cochlea wid h and skull leng h o Ica-
onyc e is and Palaeochi op e yx, and ound
ha hese ossil axa ell wi hin he ange o
a ia ion obse ed in ex an Mic ochi op e a.
Because an enla ged cochlea is de i ed and
is associa ed wi h use o sophis ica ed echo-
1998 57SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
loca ion,
3
hese obse a ions led No acek
(1985a, 1987, 1991) o conclude ha Ica on-
yc e is and Palaeochi op e yx we e echolo-
ca ing ba s closely ela ed o ex an Mic o-
chi op e a.
In ano he s udy o ela i e cochlea size
in ba s, Habe se ze and S o ch (1992) sug-
ges ed ha No acek’s (1985a, 1987, 1991)
3
We ecognize wo ca ego ies o echoloca ion sys-
ems based on he in o ma ion ha is ob ained om e-
sul ing echoes. We use he e m ‘‘sophis ica ed echolo-
ca ion’’ o desc ibe sys ems ha can be used o o ien-
a ion in clu e ed en i onmen s (e.g., wi hin ege a ion
o unde he o es canopy) and o de ec ion, acking,
and e alua ion o mo ing objec s including p ey. Calls
associa ed wi h sophis ica ed echoloca ion a e p oduced
in he la ynx and a e o en (bu no always) highly s uc-
u ed in e ms o call leng h, equency, and ype o
modula ion. Call s uc u e, du a ion, and pulse in e al
may be in en ionally a ied by he animal depending on
he ci cums ances (e.g., sea ching o p ey, app oaching
p ey, a acking p ey), and call s uc u e is o en species
speci ic. The e is a s ong co ela ion be ween call ype
and ligh / o aging s a egies, in pa because he in o -
ma ion con en o e u ning echoes a ies depending on
call s uc u e and pulse in e al. Howe e , all sophis i-
ca ed echoloca o s appa en ly eco d and p ocess in o -
ma ion o se e al ypes, including ela i e iming o
pulse and echo (including di e en a i al imes a he
igh and le ea s), di e ences in in ensi y be ween
pulse and echo, and (in some cases) equency shi s
be ween pulse and echo. Sophis ica ed echoloca ion ap-
pa en ly p o ides he animal wi h a de ailed acous ic
map o i s en i onmen , a map ha changes momen by
momen depending on mo emen o bo h he echoloca-
o and he objec s in i s en i onmen . This o m o
echoloca ion occu s in all ex an mic ochi op e ans bu
no in megachi op e ans.
In con as , we use he e m ‘‘p imi i e echoloca ion’’
o e e o less complex sys ems in which echoes o
b oadband clicks p oduced by he animal (by ongue-
clicks, oice, o some o he means) a e used o ob ain
gene al in o ma ion abou he animals su oundings.
Pulses o sound used in p imi i e echoloca ion sys ems
a e always b oadband, low in in ensi y, sho in du a ion,
and ne e exhibi s uc u ed changes in equency o e
ime. The da a used by he animal appea s o consis
only o di e ences in iming o pulse and e u ning
echo. The in o ma ion ob ained using p imi i e echolo-
ca ion is p obably limi ed o de ec ion o ela i ely la ge,
s a iona y obs acles (e.g., ca e walls); his o m o echo-
loca ion is no use ul o de ec ing and acking p ey o
o ien a ion in clu e ed en i onmen s. See discussion un-
de ‘‘E olu ion o Echoloca ion and Fo aging S a e-
gies’’ o e e ences and mo e in o ma ion abou he ax-
onomic dis ibu ion o di e en echoloca ion sys ems.
me hods we e biased by his use o skull
leng h as a measu e o body size, because
skull leng h also e lec s die a y habi s and
s uc u e o he den i ion. Habe se ze and
S o ch alle ia ed his p oblem by using ba-
sic anial wid h (measu ed be ween he ou -
e mos bony ma gins o he semici cula ca-
nals) as an indica o o size, and also adop ed
a mo e p ecise me hod o measu ing cochlea
wid h. Using adiog aphs, hey measu ed
cochlea wid h om he end o he i s hal
u n o he cochlea o he end o he second
hal u n, hus gua an eeing homologous
measu emen s ha excluded he p omon o-
ium and es ibula sys em. Using hese
me hods, Habe se ze and S o ch ob ained
esul s di e en om hose o No acek
(1985a, 1987, 1991). Ra he han ha ing
comple ely nono e lapping dis ibu ions,
hey ound ha Mic ochi op e a ( ep esen ed
by mo e han 200 species) and Megachi op-
e a (mo e han 70 species) exhibi ed a na -
ow zone o o e lap in ela i e cochlea size
( ig. 29). Some mic ochi op e ans wi h un-
usually small cochleae o hei body sizes
(e.g., Phyllos omus has a us,Lep onyc e is
ni alis,Ca ollia pe spicilla a,Megade ma
ly a,Megade ma spasma) ell wi hin o on
he bo de o he smalles polygon con aining
nonecholoca ing megachi op e ans (Habe -
se ze and S o ch, 1992). The obse ed pa -
e n o dis ibu ion ( ig. 29) was in e p e ed
as e idence ha cochlea size in ba s a ies
essen ially con inuously om Megachi op-
e a (wi h he smalles cochleae, lacking so-
phis ica ed echoloca ion) h ough Rhinolo-
phidae (wi h he la ges cochlea and highly
sophis ica ed echoloca ion), wi h cochlea
size mi o ing he unc ional signi icance o
he acous ic sense.
Habe se ze and S o ch (1992) also mea-
su ed cochlea and basic anial wid h in A -
chaeonyc e is igonodon,A.pollex,Palaeo-
chi op e yx upaiodon,P.spiegeli,Hassi-
anyc e is messelensis, and H. e illiodi, and
compa ed he esul ing measu emen s wi h
da a om ex an o ms. They ound ha bo h
species o A chaeonyc e is had a ela i ely
small cochlea, alling in he zone o o e lap
be ween Megachi op e a and Mic ochi op-
e a, close o he poin ep esen ing Megad-
e ma spasma. In con as , bo h species o Pa-
laeochi op e yx and bo h species o Hassi-
58 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 29. Plo o he oblique diame e (wid h) o he cochlea e sus basic anial wid h; see ex o
discussion. This g aph was ed awn om Habe se ze and S o ch (1992: ig. 4) and se e al new da a
poin s we e added. The zone o o e lap be ween Megachi op e a and Mic ochi op e a is enclosed wi h
a dashed line. Numbe ed da a poin s e e o he ollowing axa: 1, Ica onyc e is index;2,A chaeonyc-
e is igonodon;3,A chaeonyc e is pollex;4,Palaeochi op e yx upaiodon;5,Palaeochi op e yx spie-
geli;6,Hassianyc e is e illiodi;7,Hassianyc e is messelensis;8,Rouse us aegyp icus;9,Rouse us
leschenaul i; 10, Rhinopoma mic ophyllum; 11, Rhinopoma ha dwickei; 12, C aseonyc e is honglon-
gyai; 13, Megade ma ly a; 14, Megade ma spasma; 15, Nyc e is g andis; 16, Nyc e is mac o is; 17,
Nyc e is hebaica; 18, Nyc e is hispida; 19, Phyllos omus has a us; 20, T achops ci hosus; 21, Lep o-
nyc e is ni alis; 22, Ca ollia pe spicilla a; 23, Desmodus o undus; 24, Noc ilio albi en is; 25, Mo -
moops megalophylla; 26, Mys acina obus a; 27, Myzopoda au i a; 28, Thy op e a icolo ; 29, An o-
zous pallidus; 30, Tomopeas a us; 31, Ke i oula pelucida; 32, Tupaia glis; 33, Tupaia ja anica. Values
o Cynocephalus a iega us (no shown on g aph): basic anial wid h
5
28.1; cochlea wid h
5
3.85.
1998 65SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
join s a e lacking. T1 and T2 a e no used
in ei he ou g oup, no a e hey used in mos
o he mammals. In his con ex , lack o T1–
T2 usion in Ica onyc e is and Palaeochi op-
e yx appa en ly ep esen s he p imi i e con-
di ion. We we e unable o de e mine he con-
di ion o T1 and T2 in A chaeonyc e is and
Hassianyc e is; hese axa a e he e o e
sco ed ‘‘?’’ o his cha ac e .
Cha ac e 80: An e io ibs no used o
e eb ae (0); o i s ib used o e eb ae
(1); o a leas i s i e ibs used o e e-
b ae (2). The an e io ibs a e no used o
e eb ae in Ica onyc e is,Hassianyc e is,
and Palaeochi op e yx. Among ex an ba s,
a simila condi ion is seen in P e opodidae,
Emballonu idae, Rhinopoma idae, Nyc e i-
dae, Megade ma idae, and Yangochi op e a.
In con as , he i s ib is used o he e -
eb al column in C aseonyc e idae, and a
leas he i s i e ibs a e used o e eb ae
in Rhinolophidae. No usion is seen be ween
ibs and e eb ae in he ou g oups and mos
o he mammalian g oups. This sugges s ha
he lack o usion seen in he Eocene ba s is
p imi i e. We we e unable o de e mine i ib
usion was p esen o absen in A chaeonyc-
e is; his axon is he e o e sco ed ‘‘?’’ o
his cha ac e .
Cha ac e 81: Wid h o i s ib simila o
o he ibs (0); o i s ib a leas wice he
wid h o o he ibs (1). The wid h o he i s
ib is simila o ha o he o he ibs in Ica-
onyc e is and Palaeochi op e yx. This con-
di ion is also seen in mos ex an ba s. In
con as , he i s ib is a leas wice he
wid h o he o he ibs in Rhinolophoidea.
The wid h o he i s ib is simila o ha o
he o he ibs in bo h ou g oups and mos
o he mammals, sugges ing ha his condi-
ion, which is seen in he Eocene ba s, is
p imi i e. We we e unable o de e mine he
wid h o he i s ib in A chaeonyc e is and
Hassianyc e is; hese axa a e he e o e
sco ed ‘‘?’’ o his cha ac e .
Cha ac e 82: Fi s cos al ca ilage no
ossi ied o used wi h manub ium o i s ib
(0); o i s cos al ca ilage ossi ied and
used o manub ium (whe e i appea s o
o m a winglike la e al p ocess o he ma-
nub ium) and used o i s ib (1). The ma-
nub ium is he an e io mos elemen o he
s e num. I lies a he junc ion o he p oxi-
mal cla icles (wi h which i a icula es an-
e ola e ally), he cos al ca ilages o ib 1
(wi h which i a icula es pos e ola e ally),
and he mesos e num (wi h which i a icu-
la es pos e io ly). The manub ium ex ends
la e ally only o he le el o he cla icula
join in Ica onyc e is ( ig. 22), A chaeonyc-
e is ( ig. 3), Hassianyc e is, and Palaeochi-
op e yx, and he e is no e idence o usion
o he i s cos al ca ilage o he manub ium
o i s ib. This is simila o he condi ion
seen in mos ex an ba s. In con as , he ma-
nub ium has a winglike p ojec ion on each
side ha ex ends la e ally well beyond he
cla icula join in Rhinopoma oidea, Megad-
e ma idae, and Rhinolophidae. In hese
o ms, his winglike la e al p ojec ion is
used di ec ly o he i s ib. I seems likely
ha hese manub ial ‘‘wings’’ a e o med by
ossi ica ion and usion o he i s cos al ca -
ilage (which is ypically b oad and has a
simila shape) wi h he body o he manu-
b ium. Manub ial wings a e lacking and he
i s ib is no used o he manub ium o ibs
in bo h ou g oups and mos mammals, sug-
ges ing ha he condi ion seen in he Eocene
ba s is p imi i e.
Cha ac e 83: Second cos al ca ilage a -
icula es wi h s e num a manub ium–mesos-
e num join (0); o second ib a icula es
wi h manub ium, no con ac be ween ib (o
cos al ca ilage) and mesos e num (1). The
second cos al ca ilage a icula es wi h he
s e num a he manub ium–mesos e num
join in Ica onyc e is ( igs. 21, 22), A chaeo-
nyc e is ( ig. 3) Hassianyc e is, and Palaeo-
chi op e yx. This condi ion is simila o ha
seen in mos ex an ba s. In con as , in Rhin-
olophoidea he second ib a icula es wi h he
manub ium and he e is no con ac be ween
ib (o cos al ca ilage) and mesos e num.
The second cos al ca ilage a icula es wi h
he s e num a he manub ium–mesos e num
join in bo h ou g oups. This sugges s ha
he condi ion seen in he Eocene ba s is el-
a i ely p imi i e.
Cha ac e 84: Second ib a icula es wi h
s e num ia cos al ca ilage (0); o second
ib used o s e num, cos al ca ilage absen
o ossi ied (1). The second ib a icula es
wi h he s e num ia a cos al ca ilage in Ica-
onyc e is ( igs. 21, 22), A chaeonyc e is
( ig. 3), Hassianyc e is, and Palaeochi op-
66 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 30. Ica onyc e is index (holo ype; PU 18150). S e eopho og aphs o close-up do sal iew o
he shoulde egion. F om Jepsen (1970: ig. 9); ep in ed om o iginal nega i es.
e yx. This condi ion is simila o ha seen
in mos ex an ba s. In con as , in Rhinolo-
phidae he second ib is used o he s e num
and he cos al ca ilage is absen o ossi ied.
In hese o ms, an ex ensi e shee o hin
bone uns be ween he second ib and he
la e al p ocess o he manub ium. The second
ib a icula es wi h he s e num ia a cos al
ca ilage in bo h ou g oups and mos o he
mammals, sugges ing ha he condi ion seen
in he Eocene ba s is ela i ely p imi i e.
Cha ac e 85: Mesos e num a icula es
wi h a leas i e cos al ca ilages pos e io
o second ib (0); o a icula es wi h ou
cos al ca ilages pos e io o second ib (1);
o a icula es wi h only h ee cos al ca i-
lages pos e io o second ib (2). The me-
sos e num a icula es wi h a leas i e cos al
ca ilages pos e io o he second ib in Ica-
onyc e is ( igs. 21, 22), A chaeonyc e is
( ig. 3), Hassianyc e is, and Palaeochi op-
e yx. Among ex an lineages, a simila con-
di ion is seen in some P e opodidae, some
Phyllos omidae, Mo moopidae, Noc ilioni-
dae, Mys acinidae, Myzopodidae, Na alidae,
Molossidae, some Vespe ilioninae, and
some Myo inae. In con as , he mesos e num
a icula es wi h only ou cos al ca ilages
pos e io o he second ib in some P e opod-
idae, C aseonyc e idae, Rhinolophoidea, and
some Phyllos omidae. This numbe is u he
educed o h ee cos al ca ilages in some
Vespe ilioninae, Mu ininae, and Ke i ouli-
nae. The mesos e num a icula es wi h i e
o mo e cos al ca ilages in bo h ou g oups,
sugges ing ha he condi ion seen in he Eo-
cene ba s is ela i ely p imi i e.
Cha ac e 86: Ribs wi h no an e io lam-
inae (0); o an e io laminae p esen (1).
An e io laminae a e hin pla es o bone ha
un along he leading edges o ibs an e io
o he main body o he ib in some ba s.
These s uc u es, which a e o en nea ly
anspa en , appea o p o ide an inc eased
a ea o muscle a achmen . An e io laminae
a e absen in Ica onyc e is ( igs. 22, 30), bu
a e p esen in Palaeochi op e yx. Among ex-
an o ms, an e io laminae a e absen in
some Emballonu idae, Rhinolophidae, some
Phyllos omidae, Mys acinidae, An ozoidae,
1998 67SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 30 (con inued). Key o s e opho og aphs
on p eceding page. AB, axilla y bo de o scap-
ula; AC, ac omion p ocess o scapula; CL, cla -
icle; HE, head o hume us; HU, hume us; IF, in-
aspinous ossa o scapula; SF, sup aspinous os-
sa o scapula; SN, sup ascapula no ch o scapula;
T, ochi e .
Tomopea inae, and some Molossinae. An e-
io laminae a e p esen in some P e opodi-
dae, some Emballonu idae, Rhinopoma o-
idea, Nyc e idae, Megade ma idae, some
Phyllos omidae, Mo moopidae, Noc ilioni-
dae, Na aloidea, some Molossinae, and Ves-
pe ilionidae. Bo h ou g oups lack an e io
laminae, sugges ing ha absence o an e io
laminae in Ica onyc e is is ela i ely p imi-
i e, while p esence in Palaeochi op e yx is
ela i ely de i ed. We we e unable o ade-
qua ely de e mine p esence/absence o hese
s uc u es in A chaeonyc e is and Hassianyc-
e is; hese axa a e he e o e sco ed ‘‘?’’ o
his cha ac e .
Cha ac e 87: An e io laminae on ibs
na ow, lamina wid h less han ha o main
body o ib (0); o an e io laminae wide,
equal o o wide han main body o ib (1).
The an e io laminae on he ibs o ba s may
be ei he ela i ely na ow (lamina wid h less
han ha o he main body o he ib) o wide
(equal o o wide han he main body). The
an e io laminae a e na ow in Palaeochi op-
e yx. Among ex an ba s, na ow an e io
laminae a e ound in P e opodidae, Embal-
lonu idae, C aseonyc e idae, Phyllos omidae,
Myzopodidae, Molossidae, Vespe ilioninae,
Myo inae, Mu ininae, and Ke i oulinae.
Wide an e io laminae a e seen in Rhinopom-
a idae, Nyc e idae, Megade ma idae, Mo -
moopidae, Noc ilionidae, Thy op e idae, Fu-
ip e idae, Na alidae, and Miniop e inae.
This cha ac e canno be e alua ed o axa
ha lack an e io laminae (s a e ‘‘0’’ o cha -
ac e 86 abo e); hese o ms (including Ica-
onyc e is) a e sco ed ‘‘-’’ o his cha ac e .
Because bo h ou g oups lack an e io lami-
nae, he pola i y o his cha ac e canno be
de e mined a p io i. As no ed abo e, we we e
unable o adequa ely de e mine p esence/ab-
sence o an e io laminae in A chaeonyc e is
and Hassianyc e is; hese axa a e he e o e
sco ed ‘‘?’’ o his cha ac e .
Cha ac e 88: Ribs wi h no pos e io lam-
inae (0); o pos e io laminae p esen (1).
Pos e io laminae a e hin pla es o bone ha
un along he ailing edges o ibs pos e io
o he main body o he ib. Like he an e io
laminae, hese s uc u es may be nea ly
anspa en , and hey appea o p o ide an
inc eased a ea o muscle a achmen . Pos-
e io laminae a e p esen in Ica onyc e is
( igs. 22, 30), Hassianyc e is ( ig. 4), and Pa-
laeochi op e yx ( ig. 2), and hey a e also
p esen in mos ex an ba s. Pos e io laminae
a e absen in Mys acinidae and some Molos-
sinae. Pos e io laminae a e also absen in
bo h ou g oups, sugges ing ha hei p es-
ence in Ica onyc e is and Palaeochi op e yx
is de i ed. We we e unable o adequa ely de-
e mine p esence/absence o hese s uc u es
in A chaeonyc e is; his axon is he e o e
sco ed ‘‘?’’ o his cha ac e .
Cha ac e 89: Pos e io laminae on ibs
na ow, lamina wid h less han ha o main
body o ib (0); o pos e io laminae wide,
equal o o wide han main body o ib (1).
Like he an e io laminae, pos e io laminae
may be ei he ela i ely na ow (lamina
wid h less han ha o he main body o he
ib) o wide (equal o o wide han he main
body). The pos e io laminae a e na ow in
68 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Ica onyc e is ( igs. 22, 30) bu a e wide in
Hassianyc e is and Palaeochi op e yx.
Among ex an ba s, na ow pos e io laminae
a e ound in P e opodidae, some Emballon-
u idae, C aseonyc e idae, Phyllos omidae,
Noc ilionidae, Na alidae, Molossoidea, and
some Vespe ilioninae. Wide pos e io lami-
nae a e seen in some Emballonu idae, Rhin-
opoma idae, Rhinolophoidea, Mo moopidae,
Myzopodidae, Thy op e idae, Fu ip e idae,
some Vespe ilioninae, Miniop e inae, My-
o inae, Mu ininae, and Ke i oulinae. This
cha ac e canno be e alua ed o axa ha
lack pos e io laminae (s a e ‘‘0’’ o cha ac e
88 abo e); hese o ms a e sco ed ‘‘-’’ o
his cha ac e . Because bo h ou g oups lack
pos e io laminae, pola i y o his cha ac e
canno be de e mined a p io i. As no ed
abo e, we we e unable o adequa ely de e -
mine p esence/absence o an e io laminae in
A chaeonyc e is and Hassianyc e is; hese
axa a e sco ed ‘‘?’’ o his cha ac e .
Cha ac e 90: An e io ace o manub i-
um small (0); o b oad, de ined by ele a ed
idges (1). In addi ion o i s ole in connec -
ing he body o he s e num o he cla icles
and an e io ibs, he manub ium is also an
impo an si e o muscle o igin o pa s o
he m. pec o alis complex, which p o ides
he majo i y o he powe o he downs oke
o he wings (Vaughan, 1959, 1970b; No -
be g, 1970; S ickle , 1978; He manson and
Al enbach, 1983, 1985). The an e io ace o
he manub ium p o ides he poin o o igin
o pa o he an e io di ision o m. pec-
o alis, which o a es he hume us and helps
o pull i downwa d and o wa d and du ing
he downs oke when he wing is p o ac ed
agains he o ce o he ai s eam (Vaughan,
1959, 1970b; No be g, 1970; S ickle , 1978;
He manson and Al enbach, 1983, 1985). The
an e io ace o he manub ium o he s e -
num is ela i ely small and poo ly de ined in
A chaeonyc e is,Hassianyc e is, and Pa-
laeochi op e yx. This condi ion is simila o
ha seen in mos ex an ba s. In con as , he
an e io ace o he manub ium is a b oad,
iangula su ace ha ex ends on o he la -
e al p ocesses and is de ined by h ee ele-
a ed idges in some P e opodidae, Megad-
e ma idae, Rhinolophidae, some Phyllos om-
idae, Na alidae, and some Molossinae.
Among he ou g oups, he an e io ace o
he manub ium is small in Scanden ia bu is
b oad and iangula in De mop e a. Lack o
ag eemen be ween he ou g oups p ecludes
a p io i de e mina ion o he p imi i e con-
di ion o his cha ac e . We we e unable o
de e mine he o m o he an e io ace o he
manub ium in Ica onyc e is; his axon is
he e o e sco ed ‘‘?’’ o his cha ac e .
Cha ac e 91: Ven al p ocess o manu-
b ium absen (0); o en al p ocess p esen ,
dis al ip blun o ounded (1); o en al
p ocess p esen , dis al ip la e ally com-
p essed (2). The en al p ocess o he ma-
nub ium p o ides he an e io a achmen
poin o a se ies o ligamen ous shee s ha
un down he midline o he s e num
(Vaughan, 1959, 1970b; No be g, 1970,
1972a; S ickle , 1978; He manson and Al-
enbach, 1983, 1985). The en al p ocess,
oge he wi h hese ligamen ous shee s,
o ms he o igin o he m. pec o alis com-
plex (Vaughan, 1959, 1970b; No be g, 1970;
S ickle , 1978; He manson and Al enbach,
1983, 1985). The dis al ip o he en al p o-
cess o he manub ium is blun and some-
wha ounded in Ica onyc e is ( igs. 21, 22)
and A chaeonyc e is ( ig. 3). The en al p o-
cess in hese o ms is iangula in c oss sec-
ion and poin s somewha pos e io ly. In con-
as , he ip o he en al p ocess is la e ally
comp essed (keel-like) in Hassianyc e is and
Palaeochi op e yx. In c oss sec ion, he en-
al p ocess in hese o ms is lens-shaped
wi h he long axis unning an e opos e io ly.
Among ex an ba s, a en al p ocess wi h a
blun o ounded ip is ound in Rhinopom-
a idae, Megade ma idae, Rhinolophinae,
some Hipposide inae, some Phyllos ominae,
Mys acinidae, An ozoidae, Tomopea inae,
some Molossinae, and some Vespe ilioni-
nae. A la e ally comp essed en al p ocess
is seen in P e opodidae, Emballonu idae,
C aseonyc e idae, Nyc e idae, some Hippos-
ide inae, some Phyllos ominae, Mo moopi-
dae, Noc ilionidae, Na aloidea, some Molos-
sinae, some Vespe ilioninae, Miniop e inae,
Myo inae, Mu ininae, and Ke i oulinae.
Among he ou g oups, a blun en al p o-
cess occu s in De mop e a, bu he en al
p ocess is absen in Scanden ia. Al hough he
ou g oup e idence is ambiguous, his dis i-
bu ion sugges s ha a en al p ocess wi h a
blun , ounded ip (as in Ica onyc e is and
1998 69SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
A chaeonyc e is) may be p imi i e, while he
la e ally comp essed, keel-like condi ion
seen in Hassianyc e is and Palaeochi op e-
yx appea s o be de i ed.
Cha ac e 92: Angle be ween axis o en-
al p ocess and body o manub ium acu e
(0); o app oxima ely 90
8
(1); o ob use (2);
o en al p ocess bilobed wi h one acu e
and one ob use p ocess (3). The axis o he
en al p ocess is de ined as he long axis o
he hickened base and cen al body o he
p ocess. The o ien a ion o he axis o he
en al p ocess (as seen in la e al iew) a -
ies among ba s. In some axa, he angle be-
ween he en al p ocess and he body o he
manub ium is acu e, and he en al p ocess
appea s o p ojec pos e o en ally. In o he
o ms, he angle is app oxima ely 90
8
(so ha
he en al p ocess p ojec s en ally) o ob-
use ( en al p ocesses p ojec s an e o en-
ally). Al hough a ull ange o a ia ion be-
ween hese condi ions is heo e ically pos-
sible, we ound ha mos species could be
easily placed in one o he ca ego ies de ined
abo e. The only excep ions we e a ew axa
ha ha e a bilobed en al p ocess wi h one
acu e and one ob use lobe.
Ou examina ions o he Eocene ossils in-
dica ed ha he en al p ocess is ob use in
Ica onyc e is ( igs. 21, 22) and Palaeochi-
op e yx, and is o ien ed a app oxima ely
90
8
in A chaeonyc e is ( ig. 3). Among ex-
an o ms, an ob use en al p ocess is seen
in some P e opodidae, some Emballonu idae,
Noc ilionidae, some Mo moopidae, some
Mys acinidae, Myzopodidae, Molossoidea,
and Vespe ilioninae. The en al p ocess is
o ien ed a app oxima ely 90
8
in some P e -
opodidae, some Emballonu idae, Rhinopom-
a oidea, some Phyllos omidae, some Mo -
moopidae, Thy op e idae, Myo inae, Mu i-
ninae, and Ke i oulinae. An acu e en al
p ocess occu s in Megade ma idae, Rhino-
lophidae, some Phyllos omidae, and Na ali-
dae. The en al p ocess is bilobed in Nyc-
e idae, Fu ip e idae, and Miniop e inae.
Among he ou g oups, an ob use en al p o-
cess occu s in De mop e a. This ea u e can-
no be e alua ed in axa ha lack a en al
p ocess on he manub ium (s a e ‘‘0’’ o
cha ac e 91 abo e). Scanden ia, whose
membe s lack a en al p ocess, is he e o e
sco ed ‘‘-’’ o his cha ac e . Al hough he
ou g oup e idence is ambiguous, p esence o
an ob use en al p ocess in De mop e a sug-
ges s ha his condi ion, which is also seen
in Ica onyc e is and Palaeochi op e yx, may
be p imi i e. The app oxima e 90
8
angle o
he en al p ocess in A chaeonyc e is would
hus be in e p e ed as a de i ed condi ion.
We we e unable o de e mine he angle o he
en al p ocess in Hassianyc e is; his axon
is he e o e sco ed ‘‘?’’ o his cha ac e .
Cha ac e 93: Leng h o manub ium pos-
e io o la e al p ocesses
.
2.5 imes ans-
e se wid h (0); o leng h
,
2 imes ans-
e se wid h (1). The leng h o he manub ium
pos e io o he la e al p ocesses is less han
wice he ans e se wid h o his po ion o
he manub ium in Ica onyc e is ( igs. 21,
22), A chaeonyc e is ( ig. 3), Hassianyc e is,
and Palaeochi op e yx. This condi ion is
simila o ha seen in mos ex an ba s. In
con as , he manub ium is ela i ely elon-
ga ed (pos e io po ion
.
2.5 imes ans-
e se wid h) in Nyc e idae, some Phyllos-
omidae, and Molossidae. Bo h condi ions
occu among he ou g oups— he manub ium
is ela i ely sho in De mop e a bu is elon-
ga e in Scanden ia. Acco dingly, he p imi-
i e condi ion o his ea u e canno be de-
e mined a p io i.
Cha ac e 94: Mesos e num na ow,
mean wid h less han hal he dis ance be-
ween cla icles a s e nocla icula join (0);
o mesos e num b oad, mean wid h g ea e
han h ee- ou hs he dis ance be ween cla -
icles (1). The mesos e num (
5
body o s e -
num) a icula es wi h he manub ium an e i-
o ly and wi h he xiphis e num pos e io ly.
The ela i e wid h o he mesos e num in a
ba can be es ima ed by compa ing mean me-
sos e nal wid h wi h he ans e se dis ance
be ween he igh and le cla icles a hei
join s wi h he manub ium.
4
The mesos e -
num is ela i ely na ow (mesos e nal wid h
less han hal he in e cla icula dis ance) in
Ica onyc e is ( igs. 21, 22), A chaeonyc e is,
4
Modi ica ions o he manub ium and ibs in a num-
be o axa (e.g., hinolophoids) p eclude meaning ul
size compa isons o he mesos e num wi h hese ele-
men s. Acco dingly, we chose o compa e wid h o he
mesos e num wi h he in e cla icula dis ance because
ou obse a ions sugges ha he dis ance be ween he
cla icles is co ela ed p incipally wi h body size.
70 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Hassianyc e is, and Palaeochi op e yx. This
is simila o he condi ion seen in P e opod-
idae, Emballonu idae, Rhinopoma oidea,
Megade ma idae, Rhinolophidae, Noc ilion-
oidea, Mys acinidae, Myzopodidae, Molos-
soidea, and some Vespe ilionidae. In con-
as , he mesos e num is b oad (wid h g ea -
e han h ee- ou hs he in e cla icula dis-
ance) in Nyc e idae, Thy op e idae,
Fu ip e idae, Na alidae, some Vespe ilioni-
nae, Miniop e inae, Myo inae, Mu ininae,
and Ke i oulinae. The mesos e num is ela-
i ely na ow in bo h ou g oups, sugges ing
ha he na ow condi ion seen in he Eocene
ba s is ela i ely p imi i e.
Cha ac e 95: Xiphis e num wi hou keel
(0); o wi h p ominen median keel (1). The
xiphis e num lacks a median longi udinal
keel on i s en al su ace in Ica onyc e is
( ig. 22), A chaeonyc e is, and Palaeochi-
op e yx. This condi ion is seen among ex-
an ba s in P e opodidae, Rhinopoma idae,
Hipposide inae, some Phyllos omidae, Noc-
ilionidae, Myzopodidae, Molossoidea, and
some Vespe ilioninae. In con as , a p omi-
nen , en ally p ojec ing median keel is
p esen on he xiphis e num in Emballonu -
idae, C aseonyc e idae, Nyc e idae, Megad-
e ma idae, Rhinolophidae, some Phyllos om-
idae, Mo moopidae, Mys acinidae, Thy op-
e idae, Fu ip e idae, Na alidae, some Ves-
pe ilioninae, Miniop e inae, Myo inae,
Mu ininae, and Ke i oulinae. A xiphis e nal
keel is absen in bo h ou g oups, sugges ing
ha absence o a keel is ela i ely p imi i e.
We we e unable o de e mine i a xiphis e nal
keel is p esen in Hassianyc e is due o dis-
o ion and la ening o he s e nal egion in
all a ailable specimens; his axon is he e-
o e sco ed ‘‘?’’ o his cha ac e .
Cha ac e 96: Pos e io xiphis e num
wi h wide la e al la e (0); o no la e ally
la ed (1). The pos e io xiphis e num has a
wide la e al la e in Ica onyc e is ( ig. 22),
Hassianyc e is, and Palaeochi op e yx. This
la e is p oduced by a s eady inc ease in
wid h o he xiphis e num om he an e io
o he pos e io end o he elemen . In con-
as , he xiphis e num is no la ed (and
wid h o he pos e io end is app oxima ely
equal o he wid h o he an e io end) in A -
chaeonyc e is. Among ex an o ms, a la e al
la e is p esen in some P e opodidae, some
Emballonu idae, Rhinopoma oidea, Nyc e i-
dae, Megade ma idae, Hipposide inae, Noc-
ilionoidea, Mys acinidae, Thy op e idae,
Molossoidea, and Vespe ilionidae. The e is
no la e al la e in some P e opodidae, some
Emballonu idae, Rhinolophinae, Myzopodi-
dae, and Fu ip e idae. Bo h ou g oups ha e
a xiphis e num wi h a wide la e al la e. This
pa e n sugges s ha he la ed mo phology
seen in Ica onyc e is,Hassianyc e is, and
Palaeochi op e yx may be ela i ely p imi-
i e, while absence o a xiphis e nal la e in
A chaeonyc e is may ep esen a ela i ely
de i ed condi ion.
P
ECTORAL
G
IRDLE
Cha ac e 97: Ac omion p ocess wi hou
medial shel (0); o wi h shel ha p ojec s
medially o e sup aspinous ossa o medial
base o ac omion p ocess (1). The dis al hal
o he medial su ace o he ac omion p ocess
se es as he a achmen poin o one end o
he ans e se scapula ligamen (
5
do sal
scapula ligamen ), which s e ches be ween
he ac omion and he an e omedial im o he
scapula do sal o m. sup aspina us (Vaughan,
1959, 1970b; No be g, 1970; S ickle , 1978;
He manson and Al enbach, 1983, 1985). In
a ious ba s he ans e se scapula ligamen
p o ides an inc eased a achmen a ea o m.
sup aspina us (which ele a es, ex ends, and
o a es he hume us), m. ac omiodel oideus
(which ele a es he hume us), and m. spi-
nodel oideus (which ele a es and lexes he
hume us; Vaughan, 1959, 1970b; No be g,
1970; S ickle , 1978; He manson and Al en-
bach, 1983, 1985). P esence o a medial shel
on he dis al ac omion is p obably co ela ed
wi h changes in he ex en and o ien a ion o
he ans e se scapula ligamen and in he
o igins o hose pa s o m. sup aspina us and
m. ac omiodel oideus ha o igina e di ec ly
om he ac omion p ocess.
The ac omion p ocess o he scapula lacks
a medial shel in Ica onyc e is,A chaeonyc-
e is,Hassianyc e is, and Palaeochi op e yx.
This is simila o he condi ion seen in mos
ex an ba s. In con as , he ac omion p ocess
has a shel like p ojec ion ha ex ends medi-
ally o e he sup aspinous ossa o medial
base o he ac omion p ocess in Thy op e i-
dae, Fu ip e idae, Na alidae, some Vespe i-
1998 71SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
lioninae, Miniop e inae, and Ke i oulinae.
No medial shel is seen in ei he ou g oup,
sugges ing ha absence o a medial shel on
he ac omion p ocess is he p imi i e condi-
ion.
Cha ac e 98: Tip o ac omion p ocess
wi hou an e io p ojec ion (0); o wi h i-
angula an e io p ojec ion (1). The ip o
he ac omion p ocess lacks an an e io p o-
jec ion in Ica onyc e is,Hassianyc e is, and
Palaeochi op e yx. The ac omion simila ly
lacks such a p ojec ion in P e opodidae, Em-
ballonu idae, Yinochi op e a, Phyllos omi-
dae, Na alidae, and Mu ininae. In con as , a
iangula an e io p ojec ion is p esen jus
en al and medial o he ip o he ac omion
p ocess in Mo moopidae, Noc ilionidae,
Mys acinidae, Myzopodidae, Thy op e idae,
Fu ip e idae, Molossoidea, Vespe ilioninae,
Miniop e inae, Myo inae, and Ke i oulinae.
This p ojec ion cu es sligh ly en ally in
some axa. No an e io p ojec ion is p esen
on he ac omion p ocess in he ou g oups;
his sugges s ha absence o an an e o en al
p ojec ion in he Eocene o ms may be p im-
i i e. We we e no able o de e mine i an
an e o en al p ocess is p esen in A chaeo-
nyc e is; his axon is he e o e sco ed ‘‘?’’
o his cha ac e .
Cha ac e 99: Dis al ac omion p ocess
wi hou pos e ola e al p ojec ion (0); o wi h
iangula pos e ola e al p ojec ion (1). The
en i e la e al su ace o he ac omion p ocess
se es as pa o he o igin o m. ac omiod-
el oideus, a muscle ha ele a es and o a es
he hume us and p o ides impo an con ol
du ing he ups oke o he wing (Vaughan,
1959, 1970b; S ickle , 1978). P esence o a
pos e ola e al p ojec ion on he dis al ac o-
mion may se e o inc ease he a ea o a -
achmen o m. ac omiodel oideus, and may
a ec he momen a m o a leas some o
he ibe s in his muscle.
The ac omion p ocess lacks a pos e ola -
e al p ojec ion in Ica onyc e is,A chaeonyc-
e is,Palaeochi op e yx, and some speci-
mens o Hassianyc e is (e.g., SMF Me
1500). This condi ion is simila o ha seen
in mos ex an ba s. In con as , a iangula
pos e ola e al p ojec ion is p esen on he
dis al ac omion p ocess in ex an Molossidae
and some specimens o Hassianyc e is (e.g.,
SMF Me 1540a). Bo h specimens o Hassi-
anyc e is a e simila in size; acco dingly, his
a ia ion may ep esen ei he wi hin-species
polymo phism (wi hin H.messelensis)o
axonomic polymo phism (i may di e en i-
a e H.messelensis om a p e iously uniden-
i ied specimen o H. e illiodi). We a e un-
awa e o any wi hin-species polymo phism
in his cha ac e in ex an ba species, so we
a o he la e in e p e a ion.
No pos e ola e al p ojec ion is p esen on
he ac omion p ocess in he ou g oups, sug-
ges ing ha absence o a pos e ola e al p o-
jec ion in he Eocene ba s is p imi i e. P es-
ence o a iangula pos e ola e al p ojec ion
on he ac omion in some Hassianyc e is is
appa en ly a de i ed condi ion.
Cha ac e 100: Do sal a icula ace ( o
ochi e o hume us) absen om scapula
(0); o p esen (1). A seconda y a icula ion
be ween he hume us and scapula occu s in
many ba s when he hume us is abduc ed and
an enla ged ochi e (
5
g ea e ube osi y)
con ac s a do sal a icula ace on he scap-
ula. As discussed by Vaughan (1959, 1970b),
Hill (1974), S ickle (1978), Hill and Smi h
(1984), Al enbach and He manson (1987),
and Schlosse -S um and Schleimann (1995),
his seconda y a icula ion o ms a c i ical
pa o a ‘‘locking mechanism’’ in he shoul-
de . Func ional explana ions o he shoulde -
locking mechanism a e di e se and, in some
cases, con adic o y (see discussion in
Schlosse -S um and Schleimann, 1995). One
possibili y is ha his mechanism unc ions
in pa o a es he ups oke o he wing, hus
acili a ing long- ange, as ligh (1) by p o-
iding a longe eco e y pe iod o c i ical
ligh muscles be ween wing bea s, and (2)
by educing he amoun o o ce ha mus be
exe ed by many o he same muscles
(Vaughan, 1959, 1970b; Hill, 1974; S ickle ,
1978; Hill and Smi h, 1984). An al e na i e
iew based on new mo phological and ex-
pe imen al e idence sugges s ha he sec-
onda y shoulde join se es o inc ease he
momen a m o m. pec o alis and educe
p ona o y mo emen s o he abduc ed o e-
a m du ing he downs oke (Al enbach and
He manson, 1987; Schlosse -S um and
Schliemann, 1995).
The do sal a icula ace is si ua ed on he
do sal su ace o he scapula immedia ely an-
e omedial o he glenoid ossa. Ica onyc e is
72 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
lacks a do sal a icula ace . In con as , a
do sal a icula ace is p esen in Hassianyc-
e is and Palaeochi op e yx. Mos ex an
ba s ha e a do sal a icula ace ; excep ions
include P e opodidae, some Emballonu idae,
Rhinopoma idae, and Noc ilionidae. Bo h
ou g oups and mos o he mammals do no
ha e a seconda y a icula ion be ween he
hume us and scapula, and he e o e lack a
do sal a icula ace on he scapula. This
sugges s ha absence o a do sal a icula
ace in Ica onyc e is ep esen s a ela i ely
p imi i e condi ion. We we e unable o de-
e mine i a do sal a icula ace occu s in
A chaeonyc e is; his axon is he e o e
sco ed ‘‘?’’ o his cha ac e .
Cha ac e 101: Do sal a icula ace
aces do sola e ally and consis s o small
g oo e on an e omedial im o glenoid ossa
(0); o aces do sola e ally and consis s o
an o al ace on an e omedial im o glenoid
ossa (1); o aces do sally and consis s o a
la ge, la ace clea ly sepa a ed om gle-
noid ossa (2).InPalaeochi op e yx, he do -
sal a icula ace consis s o a small, do so-
la e ally acing g oo e on he an e omedial
im o he glenoid ossa o he scapula. In
con as , he do sal a icula ace consis s o
an o al, do sola e ally acing ace on he an-
e omedial im o he glenoid ossa in Has-
sianyc e is. Among ex an ba s, he ‘‘small
g oo e’’ condi ion is seen in Emballonu i-
dae, Nyc e idae, and some Hipposide inae.
An o al, do sola e ally acing do sal a icula
ace on he an e omedial im o he glenoid
ossa is ound in C aseonyc e idae, some
Megade ma idae, some Phyllos omidae,
Mo moopidae, Thy op e idae, Fu ip e idae,
and Na alidae. In con as , he do sal a icula
ace is a la ge, la ace ha aces do sally
and is clea ly sepa a ed om he glenoid os-
sa in some Megade ma idae, Rhinolophinae,
some Hipposide inae, some Phyllos omidae,
Mys acinidae, Myzopodidae, Molossoidea,
and Vespe ilionidae. This cha ac e canno
be sco ed in axa ha lack a do sal a icula
ace o hose in which p esence/absence o
a do sal a icula ace has no been de e -
mined. Bo h ou g oups, P e opodidae, and
Ica onyc e is lack a do sal a icula ace ;
hese axa a e sco ed ‘‘-’’ o his cha ac e .
A chaeonyc e is is sco ed ‘‘?’’ o e lec in-
su icien in o ma ion o his axon. This
cha ac e canno be pola ized a p io i be-
cause bo h ou g oups lack a do sal a icula
ace .
Cha ac e 102: In aspinous ossa o
scapula na ow, leng h
$
2 imes wid h (0);
o wide, leng h
#
1.5 imes wid h (1). The
in aspinous ossa p o ides he si e o o igin
o m. in aspina us and m. e es majo , mus-
cles ha ac o lex, o a e, and (in he case
o m. in aspina us) abduc he hume us
(Vaughan, 1959, 1970b; No be g, 1970;
S ickle , 1978; He manson and Al enbach,
1983, 1985). The wid h o he in aspinous
ossa is measu ed om he junc ion o he
scapula spine and he e eb al bo de o he
scapula o he axilla y bo de in a line pe -
pendicula o he long axis o he ossa
(which is de ined by he line o maximum
leng h). The in aspinous ossa is na ow
(leng h g ea e han o equal o wice he
wid h) in Ica onyc e is ( igs. 22, 30), A -
chaeonyc e is, and Hassianyc e is. In con-
as , he in aspinous ossa is ela i ely wide
(leng h
#
1.5 imes wid h) in Palaeochi op-
e yx ( ig. 2). Among ex an o ms, a na ow
in aspinous ossa is seen in P e opodidae,
Rhinopoma oidea, Rhinolophoidea, some
Mo moopidae, Myzopodidae, Thy op e idae,
some Molossinae, and Mu ininae. A wide in-
aspinous ossa occu s in Phyllos omidae,
Noc ilionidae, some Mo moopidae, Mys a-
cinidae, Fu ip e idae, Na alidae, An ozo-
idae, Tomopea inae, some Molossinae, Ves-
pe ilioninae, Miniop e inae, Myo inae, and
Ke i oulinae. The in aspinous ossa in bo h
ou g oups is na ow, sugges ing ha a na -
ow ossa is ela i ely p imi i e and ha he
wide in aspinous ossa in Palaeochi op e yx
ep esen s a de i ed condi ion.
Cha ac e 103: In aspinous ossa wi h
one ace (0); o wo ace s (2); o h ee ac-
e s (2). Face ing o he in aspinous ossa,
which e ec i ely se es o inc ease he su -
ace a ea wi hou inc easing he ou line di-
mensions o he ossa, is hough o unc ion
o inc ease he a ea o o igin o m. in as-
pina us and m. subscapula is (Vaughan,
1959; see discussion o m. subscapula is
unc ion below unde cha ac e 106). Face -
ing may also e lec compa men aliza ion o
subdi ision o ibe s o hese muscles in o
uni s wi h dis inc unc ions, al hough his
has ye o be in es iga ed.
1998 73SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
The in aspinous ossa has wo ace s in
Ica onyc e is,A chaeonyc e is,Hassianyc-
e is, and Palaeochi op e yx. The smalle o
hese ( he medial ace ) lies pos e io o he
scapula spine and aces pos e ola e ally; he
la ge la e al ace , which is con ex, lies
mo e pos e iola e ally and aces do sally.
Among ex an ba s, wo ace s a e ound in
some P e opodidae, Rhinopoma oidea, Nyc-
e idae, some Phyllos omidae, Mys acinidae,
Fu ip e idae, Na alidae, An ozoidae, Tomo-
pea inae, some Vespe ilioninae, and Myo i-
nae. In con as , he in aspinous ossa has
h ee ace s in some P e opodidae, Emballon-
u idae, Megade ma idae, Rhinolophidae,
some Phyllos omidae, Mo moopidae, Noc i-
lionidae, Myzopodidae, Thy op e idae, Mo-
lossinae, some Vespe ilioninae, Myo inae,
Mu ininae, and Ke i oulinae. In hese o ms,
he ace jus pos e io o he scapula spine
is called he pos e omedial ace ; his appea s
o be homologous o he medial ace ound
in he same posi ion in ba s ha ha e only
wo ace s. The la ge la e al ace is u he
subdi ided in o an in e media e ace and a
pos e ola e al ace . The pos e ola e al ace
lies along he la e al edge o he scapula and
aces pos e ola e ally; he in e media e ace
lies be ween he pos e medial and pos e o-
la e al ace s and aces an e omedially. The
in aspinous ossa is no subdi ided in he
ou g oups. In hese o ms, he e is only a sin-
gle, la ge con ex ace in he in aspionous
ossa. Acco dingly, i is no possible o de-
e mine a p io i he p imi i e condi ion o
ba s ( wo ace s o h ee ace s).
Cha ac e 105: La e al o pos e ola e al
ace o in aspinous ossa es ic ed, does
no ex end in o in aglenoid egion an e io ly
o w ap a ound in e media e ace a pos e-
io (caudal) angle o scapula (0); o pos-
e ola e al ace mo e ex ensi e, ex ends in o
in aglenoid egion and w aps a ound cau-
dal end o in e media e ace (1).InIca on-
yc e is,A chaeonyc e is,Hassianyc e is, and
Palaeochi op e yx, he ex en o he la e al
ace o he in aspinous ossa is ela i ely
es ic ed; his ace does no ex end in o he
in aglenoid egion o w ap a ound he me-
dial ace a he pos e io (caudal) angle o
he scapula. A simila condi ion is seen in
mos ex an ba s ega dless o he numbe o
ace s in he in aspinous ossa. In con as ,
he pos e ola e al ace is mo e ex ensi e in
Megade ma idae and Rhinolophidae. In hese
o ms (which ha e h ee ace s), he pos e o-
la e al ace ex ends in o he in aglenoid e-
gion an e ola e ally, and i w aps a ound he
caudal end o he in e media e ossa pos e-
io ly. This cha ac e canno be e alua ed in
axa ha ha e only one in aspinous ace ,
which p ecludes use ul sco ing o his cha -
ac e in he ou g oups; hese axa a e sco ed
‘‘-’’ o his ea u e. Acco dingly, he p imi-
i e condi ion o ba s canno be econs uc -
ed a p io i.
Cha ac e 106: Thick lip p esen along
axilla y bo de o scapula (0); o hick lip
wi h bladelike la e al edge p esen (1); o
hick lip absen , axilla y bo de la o sligh -
ly up u ned (2). The axilla y bo de o he
scapula p o ides pa o he si e o o igin o
m. subscapula is, which adduc s and ex ends
he hume us (al hough a mo e impo an
unc ion may be o s abilize and p o ide ine
con ol o he wing), and m. e es mino ,
which is a weak lexo o he hume us
(Vaughan, 1959, 1970b; No be g, 1970;
S ickle , 1978; He manson and Al enbach,
1983, 1985). Mo phology o he edge o he
scapula along i s axilla y (la e al/pos e ola -
e al) bo de a ies among ba s. A hick lip is
p esen along he axilla y bo de o he scap-
ula in Ica onyc e is ( ig. 30), A chaeonyc e -
is,Hassianyc e is, and Palaeochi op e yx.
The bone i sel is ai ly hin; he appea ance
o a hickened lip is p oduced by an ab up
old nea he axilla y bo de . This condi ion
is simila o ha seen among ex an P e o-
podidae, some Emballonu idae, Mys acini-
dae, Thy op e idae, Molossoidea, Vespe i-
lioninae, Miniop e inae, Myo inae, and Ke -
i oulinae. In con as , he axilla y bo de o
he scapula is cha ac e ized by a hick lip
wi h a bladelike la e al edge in some Em-
ballonu idae, Noc ilionoidea, and Myzopod-
idae. No hickening is seen along he axilla y
bo de in he emaining lineages o ba s; in-
s ead, he edge o he scapula is la o sligh -
ly up u ned, and he ab up old is absen .
Bo h ou g oups exhibi a simple, hick lip
(s a e ‘‘0’’) on he axilla y bo de , sugges ing
ha his is he p imi i e condi ion.
Cha ac e 107: Pi o a achmen o cla-
icula ligamen absen om scapula (0); o
p esen an e io and medial o glenoid ossa
74 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
(1). The cla icula ligamen ex ends be ween
he base o he co acoid p ocess and he cla -
icle. The e is no e idence o a pi o a ach-
men o he cla icula ligamen on he scap-
ula in Ica onyc e is and Hassianyc e is, and
he same is ue in mos ex an ba lineages.
In con as , a dis inc pi o he cla icula
ligamen is p esen an e io and medial o he
glenoid ossa in some Emballonu idae, Rhin-
olophoidea, some Phyllos omidae, and Mo -
moopidae. When a do sal a icula ace is
p esen , he pi o he cla icula ligamen is
loca ed medial o his s uc u e. The pi o
he cla icula ligamen is absen in bo h ou -
g oups, sugges ing ha absence o his s uc-
u e ep esen s he p imi i e condi ion. We
we e unable o de e mine i a pi o he cla-
icula ligamen is p esen in A chaeonyc e -
is and Palaeochi op e yx; hese axa a e
he e o e sco ed ‘‘?’’ o his cha ac e .
Cha ac e 108: An e omedial edge o
scapula wi hou p ojec ions o langes (0); o
wi h iangula an e omedial lange (1). The
an e omedial lange is a oughly iangula
lange ha p ojec s en ally om he an e -
omedial edge o he scapula (medial o he
sup ascapula no ch) in some ba s (Vaughan,
1959, 1970a; No be g, 1970; S ickle , 1978;
He manson and Al enbach, 1983, 1985). The
do sal aspec o he scapula edge in his e-
gion se es as he o igin o he ans e se
scapula ligamen , while pa o he o igins
o m. subscapula is and he an e io di ision
o m. se a us an e io occupy i s en al su -
ace (Vaughan, 1959, 1970b; No be g, 1970;
S ickle , 1978; He manson and Al enbach,
1983, 1985). P ojec ions om he bo de o
he scapula in his egion appa en ly simul-
aneously se e o p o ide an inc eased a ea
o a achmen o he ans e se scapula lig-
amen and an inc eased a ea o muscle o i-
gins. P esence o a la ge, iangula an e o-
medial lange is co ela ed wi h ela i ely
la ge size o m. subscapula is (Vaughan,
1959), which is one o he p incipal adduc-
o s and ex enso s o he hume us. M. sub-
scapula is plays an impo an ole in he
downs oke du ing ligh and also helps o
suppo he weigh o he an e io pa o he
body in e es ial locomo ion (Vaughan,
1959, 1970b; S ickle , 1978).
The an e omedial im o he scapula lacks
p ojec ions o langes in Ica onyc e is,A -
chaeonyc e is, and Hassianyc e is. In con-
as , a iangula an e omedial lange is p es-
en on he medial supe io (an e io ) bo de
o he scapula in Palaeochi op e yx. Among
ex an ba s, he an e omedial lange is absen
in P e opodidae, Emballonu idae, Rhino-
poma oidea, Yinochi op e a, Noc ilionoidea,
Mys acinidae, Myzopodidae, Thy op e idae,
and Fu ip e idae. A la ge, iangula an e o-
medial lange is p esen in Myzopodidae, Na-
alidae, Molossoidea, and Vespe ilionoidea.
Absence o his lange in bo h ou g oups
(and mos o he mammals) indica es ha ab-
sence o his s uc u e is he p imi i e con-
di ion.
Cha ac e 109: Co acoid p ocess s ou
and o mode a e leng h (0); o e y long and
hin (1). The co acoid p ocess o he scapula
is he si e o o igin o he co acoid head (
5
sho head) o m. biceps b achii. M. biceps
b achii is a wo-pa muscle ha lexes and
o a es he o ea m, adduc s he wing, and
helps o hold he o ea m igidly ou s e ched
agains he he opposing ac ion o m. iceps
du ing much o he downs oke (Vaughan,
1959, 1970b; No be g, 1970; S ickle , 1978;
He manson and Al enbach, 1983, 1985).
Leng h o he co acoid p ocess plays a ole
in unc ion o m. biceps b achii because a
long co acoid p ocess enables he co acoid
head o he biceps o ac as an adduc o o
he wing by placing i s o igin below he line
o he long axis o he hume us; he g ea e
he displacemen om his axis, he g ea e
he mechanical ad an age o he biceps as an
adduc o (Vaughan, 1959, 1966).
The co acoid p ocess o he scapula is
s ou (wide han he cla icle) and o mod-
e a e leng h in Ica onyc e is,A chaeonyc e -
is,Hassianyc e is, and Palaeochi op e yx.
This condi ion is simila o ha seen in mos
ex an ba s. In con as , he co acoid is e y
long and hin (dis al hal no as wide as cla -
icle) in Fu ip e idae, Na alidae, Tomopea i-
nae, and some Molossinae. Bo h ou g oups
exhibi he ‘‘sho and s ou ’’ mo phology,
sugges ing ha his is he p imi i e condi ion
o he co acoid p ocess.
Cha ac e 110: Co acoid p ocess cu es
en ola e ally (0); o cu es en ally (1);
o cu es en omedially (2). The di ec ion
o cu a u e o he co acoid p ocess a ies
conside ably among ba s ( ig. 31). O ien a-
1998 81SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
i II is ossi ied in bo h ou g oups, sugges ing
ha he condi ion seen in he Eocene ba s is
p imi i e.
Cha ac e 150: Wing digi II wi h ossi ied
second phalanx (0); o second phalanx unos-
si ied o absen (1). The second phalanx is
de ined by i s a icula ion wi h he dis al end
o he i s phalanx o a pa icula digi . The
second phalanx in wing digi II is ossi ied in
Ica onyc e is,A chaeonyc e is,Hassianyc-
e is, and Palaeochi op e yx. Among ex an
ba s, he second phalanx is p esen and os-
si ied only in P e opodidae and Rhinopoma-
idae; his elemen is unossi ied o absen in
all o he lineages. Digi II o he manus has
an ossi ied second phalanx in bo h ou -
g oups, sugges ing ha he condi ion seen in
he Eocene ba s is p imi i e.
Cha ac e 151: Wing digi II wi h ossi ied
hi d phalanx (0); o hi d phalanx unossi ied
o absen (1). The hi d phalanx (
5
ungual
phalanx) is de ined by i s a icula ion wi h
he dis al end o he second phalanx, and in
mammals i ypically bea s a claw o ho-
mologous s uc u e (nail o hoo ). The hi d
phalanx o wing digi II is p esen and ossi-
ied in Ica onyc e is and A chaeonyc e is.In
bo h axa, he hi d phalanx o his digi
clea ly bo e a claw. Indeed, he speci ic ep-
i he ‘‘index’’ o he ype species o Ica on-
yc e is e e s o p esence o a claw on he
index inge (digi II; Jepsen, 1966). In con-
as , he hi d phalanx o digi II is unossi-
ied o absen in Hassianyc e is and Palaeo-
chi op e yx. Among ex an ba lineages, he
hi d phalanx is p esen and ossi ied only in
P e opodidae; i is unossi ied o absen in all
o he g oups. Digi II o he manus has an
ossi ied hi d phalanx in bo h ou g oups, sug-
ges ing ha he condi ion seen in Ica onyc-
e is and A chaeonyc e is is p imi i e. The
s a e seen in Hassianyc e is and Palaeochi-
op e yx is appa en ly de i ed.
Cha ac e 152: Wing digi III wi h hi d
phalanx comple ely ossi ied (0); o hi d pha-
lanx ossi ied only a he base (1); o hi d
phalanx unossi ied o absen (2). The hi d
phalanx o wing digi III is small bu ully
ossi ied in Ica onyc e is. In con as , he
hi d phalanx o his digi is ei he unossi ied
o absen in A chaeonyc e is,Hassianyc e is,
and Palaeochi op e yx. A small bu ully os-
si ied hi d phalanx is ound among ex an
Phyllos omidae, Mo moopidae, Mys acini-
dae, Myzopodidae, and Thy op e idae. The
hi d phalanx is ossi ied only a he base (and
he dis al po ion o he elemen emains ca -
ilaginous) in Fu ip e idae, Na alidae, Molos-
soidea, and some Vespe ilionidae. The hi d
phalanx is comple ely unossi ied o absen in
P e opodidae, Emballonu idae, Yinochi op-
e a, Noc ilionidae, and some Vespe ilioni-
dae. The hi d phalanx in digi III o he ma-
nus is ully ossi ied in bo h ou g oups and
mos o he mammals, sugges ing ha he
condi ion seen in Ica onyc e is is p imi i e.
Lack o an ossi ied hi d phalanx in A chaeo-
nyc e is,Hassianyc e is, and Palaeochi op-
e yx is ela i ely de i ed.
P
OSTERIOR
A
XIAL
S
KELETON AND
P
ELVIS
Cha ac e 158: No e eb al usion in
pos e io ho acic and lumba se ies (0); o
a leas h ee e eb ae used (1). The e -
eb al column o mos mammals con ains
pos e io ho acic and lumba e eb ae as
sepa a e elemen s ha can mo e ela i e o
one ano he . They a icula e a a ious hy-
pophyseal join s, bu a e no used. This is
also ue o he pos e io ho acic and lumba
se ies in Ica onyc e is ( on ispiece, ig. 21),
A chaeonyc e is,Hassianyc e is ( ig. 4), and
Palaeochi op e yx ( ig. 3), all o which ex-
hibi no e idence o e eb al usion. Among
ex an ba s, e eb al usion in his a ea is
simila ly absen in P e opodidae, Emballon-
u idae, Rhinopoma idae, Nyc e idae, some
Megade ma idae, Rhinolophidae, some Hip-
poside inae, Phyllos omidae, Noc ilionidae,
Mys acinidae, Myzopodidae, Thy op e idae,
Molossoidea, and Vespe ilionidae. In con-
as , a leas h ee e eb ae a e used o-
ge he in he pos e io ho acic and lumba
egion in C aseonyc e idae, some Megade -
ma idae, some Hipposide inae, Mo moopi-
dae, Fu ip e idae, and Na alidae. Bo h ou -
g oups lack e eb al usion in he pos e io
ho acic and lumba egions, sugges ing ha
he condi ion seen in he Eocene ba s is el-
a i ely p imi i e.
Cha ac e 159: Sac um e mina es an e-
io o ace abulum (0); o ex ends pos e i-
o ly o a leas he midpoin o he ace abu-
lum (1). The sac um (as de ined in his s udy)
includes all e eb ae ha a icula e wi h he
82 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
pel is o a e used wi h hose ha do. The
pos e io end o he sac um e mina es an e-
io o he ace abulum in Hassianyc e is.In
con as , he sac um ex ends pos e io ly o a
leas he midpoin o he ace abulum in Ica-
onyc e is ( ig. 23) and Palaeochi op e yx.
7
Among ex an ba s, he sac um e mina es
an e io o he ace abulum in Nyc e idae and
Rhinolophidae; i ex ends pos e io ly o a
leas he midpoin o he ace abulum in all
o he lineages. The sac um e mina es an e-
io o he ace abulum in bo h ou g oups,
sugges ing ha he condi ion seen in Hassi-
anyc e is is p imi i e. The condi ion seen in
Ica onyc e is and Palaeochi op e yx appea s
de i ed. We we e unable o de e mine he ex-
en o he sac um in A chaeonyc e is; his
axon is sco ed ‘‘?’’ o his cha ac e .
Cha ac e 160: Sac al laminae na ow o
absen , e eb a wid h (including laminae)
less han o equal o h ee- ou hs e eb al
body leng h (0); o laminae b oad, e eb a
wid h equal o o g ea e han e eb al
leng h. Sac al laminae a e hin pla es o bone
ha ex end la e ally om he sac al e eb ae
pos e io o he iliosac al join . These s uc-
u es, which a e homologous wi h he ans-
e se p ocesses o lumba e eb ae, do no
a icula e wi h he pel is. Ins ead, hey ypi-
cally o m join s wi h one ano he ha may
o may no be ully used. Gaps a e equen -
ly p esen be ween successi e laminae, and
la e al enclosu e o such gaps may p oduce
a se ies o sac al o amina be ween succes-
si e laminae.
The sac al laminae a e b oad in Ica on-
yc e is ( ig. 23) and Palaeochi op e yx.
B oad sac al laminae a e also p esen in Nyc-
e idae, Myzopodidae, Thy op e idae, Mo-
lossoidea, and Vespe ilionidae. In hese
o ms, he wid h o each pos ischial sac al
e eb a (including he sac al laminae) is
subequal o o g ea e han he leng h o he
7
Ou obse a ions do no ag ee wi h hose o Russell
and Sige´ (1970), who epo ed ha he sac um in Pa-
laeochi op e yx was e y small and included only a sin-
gle e eb a. We a ibu e his disc epency o an inco -
ec econs uc ion o his egion by Russell and Sige´
(1970), who we e wo king wi h ela i ely poo ly p e-
se ed ma e ial. Disco e y o be e p ese ed speci-
mens since hei publica ion (e.g., HLMD Me 15025)
demons a e ha Palaeochi op e yx had a long sac um
composed o mul iple e eb ae.
e eb al body. In cons as , he sac al lami-
nae a e na ow o absen in all o he axa,
ex ending only a sho dis ance la e ally om
he bodies o he e eb ae. Acco dingly, e -
eb al wid h is less han o equal o h ee-
ou hs he leng h o he body o he e eb a.
Among ex an ba s, he sac al laminae a e
na ow o a e absen in P e opodidae, Em-
ballonu idae, Rhinopoma oidea, Megade -
ma idae, Rhinolophidae, Noc ilionoidea,
Mys acinidae, Fu ip e idae, and Na alidae.
Among he ou g oups, he sac al laminae a e
b oad in Scanden ia bu a e absen in De -
mop e a. Acco dingly, his cha ac e canno
be pola ized a p io i. Mo phology o he sa-
c al laminae in unknown in A chaeonyc e is
and Hassianyc e is; hese o ms a e he e o e
sco ed ‘‘?’’ o his cha ac e .
Cha ac e 161: Do somedial edge o as-
cending p ocess o ilium up u ned, la es
do sally abo e he le el o iliosac al a icu-
la ion, iliac ossa la ge and well de ined (0);
o do somedial edge no up u ned, does no
ex end do sally beyond he le el o he ilios-
ac al a icula ion, iliac ossa no la ge o
well de ined (1). The iliosac al a icula ion in
mammals is o med be ween he ascending
p ocess o he ilium and he sac al e eb ae.
The do somedial edge o he ascending p o-
cess o ms he o igin o m. enso ascia la-
ae, which lexes and abduc s he emu
(Vaughan, 1959, 1970b). M. glu eus medius
(which lexes, abduc s, and o a es he emu )
also o igina es om he do somedial edge,
al hough he iliac ossa on he do sola e al
su ace o he ascending p ocess o ms he
p incipal o igin o his muscle (Vaughan,
1959, 1970b).
The do somedial edge o he ascending
p ocess o he ilium is up u ned and la es
do sally abo e he le el o he iliosac al a -
icula ion in Ica onyc e is,Hassianyc e is,
and Palaeochi op e yx. Among ex an ba s,
a do somedially la ed ascending p ocess is
seen only in Rhinolophidae. In all axa his
la e is associa ed wi h a la ge, well-de ined
iliac ossa. In con as , he do somedial edge
o he ascending p ocess is no up u ned and
he la e is absen in all o he ba s. The do -
somedial edge e mina es a he le el o he
iliosac al a icula ion, and he iliac ossa is
ela i ely small and poo ly de ined. The up-
u ned edge and well-de ined iliac ossa a e
1998 83SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
p esen in Scanden ia bu absen in De mop-
e a; acco dingly, his cha ac e canno be
pola ized a p io i. We we e unable o de e -
mine he s a e o he ilium in A chaeonyc-
e is; his axon is he e o e sco ed ‘‘?’’ o
his cha ac e .
Cha ac e 162: Ischium wi h la ge ischial
ube osi y ha p ojec s do sally om pos e-
io ho izon al amus (0); o ischial ube -
osi y small o absen , does no p ojec do -
sally beyond le el o amus (1). The do sal
ischial ube osi y is a p ojec ion ha ex ends
do sally and/o pos e omedially om he
pos e io end o he ischium, o igina ing
om he ‘‘co ne ’’ o he pel is a he junc-
ion be ween he ho izon al and e ical ami
o he ischium (Vaughan, 1959, 1970a). Two
muscles may o igina e om he ischial u-
be osi y: m. semi endinosus and m. semi-
memb anosus, bo h o which ac o ex end
he emu and lex he lowe leg (Vaughan,
1959, 1970b). Vaughan (1959) ecognized an
ischial ube osi y in all o he ba s ha he
dissec ed, bu no ed ha in some o ms (e.g.,
Eumops) he ischial ube osi y is la ge, p o-
jec s do sally abo e he le el o he ho izon-
al amus, and se es as he si e o o igin o
bo h muscles. In o he axa (e.g., Myo is,Ma-
c o us), he ube osi y is smalle , p ojec s
somewha medially (no do sally), and se es
as he si e o o igin o jus m. semi endi-
nosus. In he la e axa, m. semimemb ano-
sus o igina es om he la e al su ace o he
caudal bo de o he ischium (Vaughan,
1959). Ou examina ion o skele ons om
nume ous amilies e ealed a dicho omy be-
ween o ms wi h a la ge, do sally p ojec ing
ischial ube osi y and hose in which he is-
chial ube osi y is small o absen . Al hough
we ha e no new muscle da a, we in e ha
mo phology o he pel is is co ela ed wi h
di e ences in muscle o igins as obse ed by
Vaughan (1959).
The e is no e idence o a la ge, do sally
p ojec ing ischial ube osi y on he pel is o
Ica onyc e is ( ig. 23), Hassianyc e is, and
Palaeochi op e yx. In con as , a la ge, do -
sally p ojec ing ischial ube osi y is p esen
in A chaeonyc e is. Among ex an ba s, a
la ge do sal ischial ube osi y is p esen only
in Mys acinidae and Molossidae. A la ge
do sally p ojec ing ischial ube osi y is p es-
en in one ou g oup (Scanden ia) bu no in
he o he (De mop e a); acco dingly, his
cha ac e can no be pola ized a p io i.
Cha ac e 163: Pubic spine absen (0); o
s aigh (1); o ip o pubic spine ben sha p-
ly do sally (2). The pubic spine is an elon-
ga e p ojec ion om he an e o en al co ne
o he pubis (Vaughan, 1959; Wal on and
Wal on, 1970; Simmons, 1994). P esence o
a pubic spine is a synapomo phy o Chi op-
e a (Simmons, 1994).
Two muscles o igina e om he pubic
spine in ba s: m. pec ineus om he base and
la e al su ace o he spine, m. g acilis om
he en i e en ola e al su ace o he spine
o he ip (Vaughan, 1959, 1970). M. pec i-
neus is an adduc o and lexo o he emu ,
while m. g acilis is a lexo o he lowe leg
and an adduc o o he hind limb (Vaughan,
1959, 1970b). Addi ionally, he pubic spine
is he si e o inse ion o m. psoas mino ,
which o igina es om he lumba e eb ae
(Vaughan, 1959, 1970b). Con ac ion o m.
psoas mino pulls he en al pa o he pel-
is o wa d, he eby a ching he pos e io
lumba sec ion o he e eb al column.
Vaughan (1959) no ed ha his ac ion may
be use ul in doubling up he body when he
ba is g ooming i sel while hanging, o when
pinning an insec o he u opa agium while
adjus ing he g ip o he jaws on he p ey.
Ac ion o his muscle may also help o b ace
he e eb al column when he ba lands, and
also while he ba is lying (i.e., agains he
shock o he ai s eam hi ing he u opa a-
gium du ing sudden maneu e s; Vaughan,
1959).
The pubic spine is ela i ely s aigh and
poin s an e odo sally in Ica onyc e is ( ig.
23), A chaeonyc e is, and Palaeochi op e-
yx. Among ex an ba s, his condi ion is seen
in P e opodidae, Emballonu idae, Rhino-
poma oidea, Nyc e idae, and Yangochi op-
e a. In hese o ms, he cou se o he pubic
spine is essen ially pa allel o ha o he il-
ium, wi h he long axis o he spine di ec ed
a he an e io lumba e eb ae. In con as ,
he ip o he pubic spine is sha ply up u ned
in Megade ma idae, Rhinolophinae, and Hip-
poside inae. The ip o he pubic spine in
hese axa poin s do sally owa d he c anial
end o he ilium. In Hipposide inae, a bony
ex ension om he pubic spine con ac s he
an e io ilium, he eby enclosing a p eace a-
84 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
bula o amen. The pubic spine is absen in
bo h ou g oups, so he o m o he pubic
spine (s aigh o wi h ip up u ned) canno
be pola ized a p io i. We could no de e mine
he o m o he pubic spine in Hassianyc e -
is, so his axon is sco ed ‘‘?’’ o his cha -
ac e .
Cha ac e 165: Ob u a o o amen no -
mal, im well de ined (0); o o amen pa -
ially in illed wi h hin, bony shee along pos-
e o en al im (1). The ob u a o o amen o
he pel is is enclosed do sally and pos e i-
o ly by he ami o he ischium and an e io ly
and en ally by he pubis. In mos mam-
mals, he im o he ob u a o o amen is well
de ined and ounded in c oss sec ion. Mus-
cles o igina ing a ound he im o he ob u-
a o o amen include m. adduc o b e is (ex-
enso and/o adduc o o emu ), m. adduc-
o magnus (ex enso and o a o o emu ),
and m. ob u a o ex e nus (ex enso and ad-
duc o o emu ; Vaughan, 1959, 1970b). M.
ob u a o in e nus is absen in ba s, p obably
as a esul o hip modi ica ions associa ed
wi h 90
8
o a ion o he hindlimbs (Simmons,
1994).
The ob u a o o amen in Ica onyc e is,
A chaeonyc e is,Hassianyc e is, and Pa-
laeochi op e yx is o he no mal mammalian
mo phology wi h a well-de ined im. This is
simila o he condi ion seen in mos ex an
ba s. In con as , he ob u a o o amen is
pa ially in illed by a hin, bony shee along
he pos e o en al im in Rhinopoma idae,
Rhinolophinae, Hipposide inae, and Thy op-
e idae. Such in illing is absen in bo h ou -
g oups, sugges ing ha he condi ion seen in
he Eocene ba s is p imi i e.
H
INDLIMB
Cha ac e 169: Sha o emu s aigh
(0); o wi h bend ha di ec s dis al sha do -
sally (1). Al hough all ba s sha e a complex
se ies o modi ica ions o he hip and p oxi-
mal emu ha se e o o a e he hindlimbs
app oxima ely 90
8
om he ypical mam-
malian condi ion (Simmons, 1994), he em-
o al sha and dis al emu emain ela i ely
unmodi ied. As in mos o he mammals, he
sha o he emu is s aigh in Ica onyc e is
( on ispiece, ig. 21), A chaeonyc e is ( ig.
3), Hassianyc e is ( ig. 4), and Palaeochi op-
e yx ( ig. 2). This condi ion is also seen in
mos ex an ba s. In some ba s, howe e , he
sha o he emu is ben somewha so ha
he dis al end is di ec ed mo e do sally ( he
equi alen o a la e al bend i he emu we e
in he posi ion ypical o non olan mam-
mals). In ba s, which al eady ha e a emu
ha p ojec s la e ally om he hip, his bend
se es o aise he knee join e en highe ,
well abo e he hip join in some o ms. A
bend in he sha o he emu is seen in
Rhinolophoidea and some Phyllos omidae.
The sha o he emu is s aigh in bo h ou -
g oups, indica ing ha he ‘‘s aigh ’’ condi-
ion is p imi i e.
Cha ac e 170: Fibula comple e and well
de eloped (0); o hin and h eadlike (1); o
absen o en i ely unossi ied (2). The ibula
in Ica onyc e is ( on ispiece) and A chaeo-
nyc e is is well de eloped ( ela i ely obus )
and comple e (ossi ied om knee o ankle).
In con as , he ibula in Hassianyc e is and
Palaeochi op e yx is ela i ely much hinne ,
almos h eadlike. Among ex an ba s, a com-
ple e, well de eloped ibula is ound only in
P e opodidae, some Phyllos omidae, Mys a-
cinidae, and Molossidae. The ibula is hin,
h eadlike, and o en only pa ially ossi ied
in Emballonu idae, Rhinopoma oidea, Me-
gade ma idae, Rhinolophidae, some Phyllos-
omidae, Mo moopidae, Noc ilionidae, Na-
aloidea, An ozoidae, and Vespe ilionidae.
The ibula is absen o en i ely unossi ied in
Nyc e idae. Bo h ou g oups ha e a comple e,
well de eloped ibula, sugges ing ha he
condi ion seen in Ica onyc e is and A chaeo-
nyc e is is ela i ely p imi i e. The hin,
h eadlike ibula seen in Hassianyc e is and
Palaeochi op e yx appa en ly ep esen s a
de i ed condi ion.
Cha ac e 171: Calca absen (0); o
p esen (1). The calca is a bony and/o ca -
ilaginous od ha ex ends om he ankle o
suppo he ailing edge o he u opa agium.
P esence o a calca (and m. dep esso ossis
s yli o mis, which adduc s he calca owa d
he lowe leg) is conside ed o be a synapo-
mo phy o Chi op e a (Simmons, 1994,
1995). Howe e , he e is no e idence ha a
calca was p esen in Ica onyc e is and A -
chaeonyc e is. A calca is lacking in e e y
known specimen, and ou examina ion o he
calcaneum sugges s ha no calca ace is
1998 85SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
p esen on he calcaneum o he hese o ms,
leading us o conclude ha a calca was ab-
sen in hese axa. In con as , a calca is
p esen in Hassianyc e is and Palaeochi op-
e yx. In bo h axa, he appea ance o he cal-
ca sugges s ha i may ha e been la gely
ca ilaginous. Among ex an ba s, a calca is
p esen in mos axa; i is absen only in some
P e opodidae, Rhinopoma oidea, and some
phyllos omids. Bo h ou g oups (and all o he
mammals) lack a calca , indica ing ha ab-
sence o he calca is p imi i e. P esence o
a calca in Hassianyc e is and Palaeochi op-
e yx ep esen s a de i ed condi ion.
Cha ac e 172: Digi s II–V o oo wi h
h ee phalanges (0); o wo phalanges (1).
Digi s II–V o he oo each ha e h ee pha-
langes in Ica onyc e is ( on ispiece, ig. 23),
A chaeonyc e is ( ig. 3), Hassianyc e is ( ig.
4), and Palaeochi op e yx ( ig. 2). This con-
di ion is also seen in mos ex an ba s. In
con as , only wo phalanges a e p esen in
digi s II–V in Hipposide inae, Myzopodidae,
and Thy op e idae. Bo h ou g oups (and
mos o he mammals) ha e h ee phalanges
in digi s II–V o he oo , sugges ing ha he
condi ion seen in he Eocene ba s is p imi-
i e.
COMPLETENESS
One p oblem ha plagues phylogene ic
analyses ha include ossil axa is ha o
comple eness (o lack o comple eness) o
a ailable da a. Nume ous s udies ha e illus-
a ed he impo ance o including ossils in
phylogene ic analyses, demons a ing ha
ossils some imes p ese e in o ma ion (in
he o m o unique combina ions o p imi i e
and de i ed cha ac e s a es) ha may be c u-
cial o esol ing ela ionships among ex an
lineages (e.g., Gau hie e al., 1988; Dono-
ghue e al., 1989; No acek, 1992, 1994).
Fossils a e also impo an because hey e-
quen ly a ec cha ac e op imiza ions, which
in u n may a ec conclusions conce ning
he deg ee o suppo o a ious monophy-
le ic g oups, axonomic diagnoses, cha ac e
independence, homoplasy, and he ela i e
iming o a ious e olu iona y e en s (Sim-
mons, 1993a). I is also ob ious ha ossils
mus be included in any s udy ha seeks o
de e mine he ela ionships o ossils o ex-
an g oups. Howe e , inclusion o ela i ely
incomple e axa—be hey ossils o poo ly
known ex an g oups—can some imes d a-
ma ically educe he deg ee o phylogene ic
esolu ion ob ained by inc easing he numbe
o equally pa simonious opologies (Rowe,
1988; Simmons, 1993a, 1993c).
Comple eness o a axon may be de ined
as he pe cen age o cha ac e s o which i
can be sco ed in a gi en analysis (Simmons,
1993a, 1993c). In p ac ice, paleon ologis s
ha e o en sough o maximize comple eness
by ocusing la gely on os eological cha ac-
e s, omi ing om conside a ion many so -
issue o molecula cha ac e s ha canno be
sco ed in ex inc o ganisms (e.g., Bea d,
1993). Howe e , inclusion o ossil OTUs in
an analysis wi h many so - issue cha ac e s
does no necessa ily lead o dec eased eso-
lu ion (Gau hie e al., 1988; No acek,
1992). In some cases, inclusion o ossils can
ac ually inc ease esolu ion by educing he
numbe o equally pa simonious ees, as was
he case wi h No acek’s (1992) hy acoid ex-
ample. When he opposi e occu s and inclu-
sion o ossils inc eases he numbe o op i-
mal ees and dec eases esolu ion, ech-
niques such as Adams consensus may be
used o iden i y ela ionships ha emain s a-
ble in all mos -pa simonious ees (Simmons,
1993a). The ela i e s abili y o opological
placemen o a ossil axon in phylogene ic
ees appea s o depend mo e on he pa ic-
ula combina ion o cha ac e s a es ha i
exhibi s han on i s comple eness. Fo ex-
ample, in an analysis o ela ionships among
a chon an mammals, Simmons (1993a)
ound ha one ossil axon ha was only
19% comple e could be placed unambigu-
ously ela i e o ex an lineages, while ela-
ionships o ano he ossil axon ha was
53% comple e could no be de e mined un-
ambiguously.
In ou iew, he bene i s o including os-
sils and so issue and molecula cha ac e s
in a single analysis a ou weigh he possible
p oblems (see discussion below). Ne e he-
less, we calcula ed he pe cen comple eness
o each OTU in ou da a se in o de o p o-
ide a basis o a pos e io i conside a ions o
he e ec s o comple eness ( able 6). Ex an
lineages in ou da a se we e 62.0–100%
86 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
TABLE 6
S a is ics on Cha ac e Codings o Cu en Da a Se
Te minal axon
Cha ac e s
sco ed wi h
single s a e
a
Cha ac e s
sco ed wi h
wo o mo e
s a es
b
Cha ac e s
sco ed ‘‘
2
’’ Cha ac e s
sco ed ‘‘?’’ Pe cen
comple eness
c
Scanden ia
De mop e a
P e opodidae
Ica onyc e is
A chaeonyc e is
Hassianyc e is
Palaeochi op e yx
Emballonu idae
161 (77.4%)
172 (82.6%)
176 (84.6%)
71 (34.1%)
59 (28.4%)
61 (29.3%)
75 (36.1%)
166 (79.8%)
3 (1.4%)
0
26 (12.5%)
0
0
2 (1.0%)
1 (0.5%)
26 (12.5%)
19 (9.1%)
18 (8.7%)
6 (2.9%)
3 (1.4%)
1 (0.5%)
1 (0.5%)
1 (0.5%)
6 (2.9%)
25 (12.0%)
18 (8.7%)
0
134 (64.4%)
148 (71.2%)
144 (69.2%)
131 (63.0%)
10 (4.8%)
88.0%
91.3%
100%
35.6%
28.9%
30.8%
37.1%
95.2%
Rhinopoma idae
C aseonyc e idae
Nyc e idae
Megade ma idae
Rhinolophinae
Hipposide inae
Phyllos omidae
Mo moopidae
Noc ilionidae
Mys acinidae
Myzopodidae
Thy op e idae
Fu ip e idae
Na alidae
An ozoidae
Tomopea inae
Molossinae
185 (89.0%)
125 (60.1%)
178 (85.6%)
183 (88.0%)
185 (89.0%)
170 (81.8%)
146 (70.0%)
172 (82.7%)
192 (92.3%)
132 (63.5%)
129 (62.0%)
174 (83.7%)
147 (70.7%)
181 (87.0%)
148 (71.2%)
125 (60.1%)
182 (87.5%)
1 (0.5%)
0
6 (2.9%)
9 (4.3%)
5 (2.4%)
13 (6.2%)
57 (27.6%)
13 (6.2%)
0
0
0
1 (0.5%)
1 (0.5%)
0
1 (0.5%)
0
21 (10.1%)
5 (2.4%)
5 (2.4%)
7 (3.4%)
2 (1.0%)
5 (2.4%)
6 (2.9%)
5 (2.4%)
5 (2.4%)
7 (3.4%)
6 (2.9%)
3 (1.4%)
3 (1.4%)
3 (1.4%)
2 (1.0%)
3 (1.4%)
4 (1.9%)
2 (1.0%)
17 (8.1%)
78 (37.5%)
17 (8.1%)
14 (6.7%)
13 (6.2%)
19 (9.1%)
0
18 (8.7%)
9 (4.3%)
70 (33.6%)
76 (36.6%)
30 (13.4%)
57 (27.4%)
25 (12.0%)
56 (26.9%)
79 (38.0%)
3 (1.4%)
91.9%
62.5%
91.9%
93.3%
93.8%
90.9%
100%
91.3%
95.7%
66.4%
63.4%
85.6%
72.6%
88.0%
73.1%
62.0%
98.6%
Vespe ilioninae
Miniop e inae
Myo inae
Mu ininae
Ke i oulinae
167 (80.3%)
145 (69.7%)
187 (89.9%)
132 (63.5%)
145 (69.7%)
22 (10.6%)
0
9 (4.3%)
0
1 (0.5%)
3 (1.4%)
5 (2.4%)
2 (1.0%)
4 (1.9%)
3 (1.4%)
16 (7.7%)
58 (27.9%)
10 (4.8%)
72 (34.6%)
59 (28.4%)
92.3%
72.1%
95.2%
65.4%
71.6%
To al 4471 (71.7%) 218 (3.5%) 145 (2.3%) 1406 (22.5%) 77.5%
a
Includes only cha ac e s s a es 0, 1, 2, e c.; does no include cha ac e s sco ed as inapplicable (‘‘
2
’’) o missing
(‘‘?’’). Pe cen age e lec s he pe cen o he o al cha ac e s (208) used in he analysis.
b
Includes bo h cases o unce ain y and axonomic polymo phism.
c
Comple eness is de ined as he pe cen age o cha ac e s o which a axon can be sco ed based on a ailable da a;
i was calcula ed by sub ac ing he pe cen age o cha ac e s sco ed ‘‘?’’ om 100%.
comple e; he ossils anged om 28.9 o
37.1% comple e.
METHODS OF PHYLOGENETIC
ANALYSIS
The p esen s udy ollows Simmons
(1998) in adop ing a cha ac e consensus
(‘‘ o al e idence’’) app oach o phylogeny e-
cons uc ion. In ecen yea s he e has been
much discussion o he ela i e me i s o
cha ac e consensus e sus axonomic con-
sensus me hods o esol ing sys ema ic
p oblems (Kluge, 1989; Ba e e al., 1991,
1993; Swo o d, 1991; Bull e al., 1993; de
Quei oz, 1993; Ee nisse and Kluge, 1993;
Kluge and Wol , 1993; Nelson, 1993; Chip-
pendale and Weins, 1994; Hulsenbeck e al.,
1994; de Quei oz e al., 1995; Fa is e al.,
1995; Miyamo o and Fi ch, 1995). As dis-
cussed by Simmons (1993a, 1998), cha ac e
cong uence is he mos sensible app oach o
he cu en da a se because he da a consis
1998 87SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
p incipally o disc e e-s a e mo phological
cha ac e s ha canno be easonably pa i-
ioned. A ailable es ic ion-si e da a a e no
capable o esol ing in e amilial ela ion-
ships in he absence o o he da a (Bake e
al., 1991a), hence axonomic cong uence is
no a use ul op ion.
The 208 disc e e cha ac e s used in he
p esen s udy (see appendix 2) we e sco ed
o phylogene ic analysis, and he esul ing
da a ma ix (appendix 3) was analyzed using
PAUP e sion 3.1.1 (Swo o d, 1993). All
ans o ma ions we e uno de ed. A heu is ic
sea ch wi h a andom-addi ion sequence and
1000 epe i ions was used o ind mos -pa -
simonious ees. Nea -mos -pa simonious
ees (one o six s eps longe ) we e iden i ied
in subsequen heu is ic sea ches using he
same pa ame e s, and a decay analysis was
pe o med ollowing he me hods o B eme
(1988). Decay alues o s ongly suppo ed
clades we e ob ained by using cons ained
heu is ic analyses o iden i y he sho es
ees ha did no include a pa icula clade.
A boo s ap analysis using heu is ic me hods
( andom-addi ion sequence, 10 epe i ions
o each o 1000 boo s ap eplica es) was
also used o e alua e he ela i e suppo o
a ious g oupings. MacClade e sion 3.0
(Maddison and Maddison, 1992) was used
o da a en y and examina ion o cha ac e -
s a e dis ibu ions.
Th ee comple e se s o phylogene ic anal-
yses we e conduc ed: (1) an analysis includ-
ing all cha ac e s bu excluding he ossil
axa; (2) an analysis including all cha ac e s
and all axa; and (3) an analysis including all
axa bu excluding hose cha ac e s ha
could no be sco ed in any o he ossil
o ms. The i s analysis was designed o p o-
ide a s a ing poin by e alua ing ela ion-
ships o ex an lineages in he con ex o he
e ised da a ma ix; his e ec i ely ep e-
sen s an upda ed e sion o Simmons’ (1998)
analysis. Ou second analysis ep esen s he
p incipal goal o he p ojec , a cha ac e con-
g uence s udy including bo h Eocene ossil
gene a and ex an lineages. The hi d and i-
nal analysis was designed o e alua e he e -
ec s o so - issue and molecula cha ac e s
on he ou come o a phylogene ic analysis
including ossil o ms ha canno be sco ed
o hese ea u es.
ANALYSIS OF CHARACTER
TRANSFORMATIONS
Cha ac e ans o ma ions we e analyzed
by mapping cha ac e -s a e dis ibu ions on o
he sho es ees de i ed om he second se
o phylogene ic analyses desc ibed abo e
(i.e., hose ob ained using all cha ac e s and
all axa). Op imiza ions we e calcula ed us-
ing bo h he ACTRAN (accele a ed ans o -
ma ion op imiza ion) and DELTRAN (de-
layed ans o ma ion op imiza ion) op ions o
PAUP e sion 3.1.1 (Swo o d, 1993).
MacClade e sion 3.0 (Maddison and Mad-
dison, 1992) was used o isualize he esul s
o cha ac e mapping.
As discussed by Simmons (1993a), wo
kinds o cha ac e ans o ma ions may be
ecognized du ing he op imiza ion p ocess:
unequi ocal ans o ma ions, which ha e
only one pa simonious placemen on he op-
imal ee(s), and equi ocal ans o ma ions,
which can be pa simoniously a anged in wo
o mo e ways. Bo h ACTRAN and DEL-
TRAN place unequi ocal ans o ma ions on
a gi en ee in he same manne , bu hey
ea equi ocal ans o ma ions di e en ly.
ACTRAN o ces ans o ma ions o he low-
es possible poin s on he ee, and hus a-
o s hypo heses o e e sal o e hypo heses
o con e gence. Con e sely, DELTRAN
o ces ans o ma ions o he highes possible
poin s on a ee, hus a o ing hypo heses o
con e gence o e e e sal. DELTRAN is o -
en a o ed in s udies in ol ing ossils be-
cause i places ans o ma ions a he mini-
mal le el a which hey can be obse ed (i.e.,
in he ace o missing da a i does no con-
clude ha de i ed s a es exis below he le el
a which hey can be demons a ed).
In his s udy, we ocused only on ans-
o ma ions ha apply a and below he nodes
whe e ossil axa join he ee. We compa ed
he esul s o ACTRAN and DELTRAN op-
imiza ions in o de o iden i y all equally
pa simonious a angemen s o equi ocal
ans o ma ions, and in e p e ed ou obse -
a ions in he con ex o wha is known abou
he unc ion o a ious s uc u es and s uc-
u al complexes. Resul s o hese analyses
a e p esen ed below unde ‘‘Cha ac e E o-
lu ion in Ea ly Chi op e ans.’’
88 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
PHYLOGENETIC RESULTS
Fig. 35. Resul s o Analysis 1, which included all cha ac e s bu omi ed he ossil axa (see ex
o discussion). Pa simony analysis esul ed in a single mos -pa simonious ee (662 s eps; CI
5
0.403
and RI
5
0.578), which is shown he e. The numbe s below in e nal b anches indica e he pe cen age
o boo s ap eplica es in which each clade appea ed; numbe s abo e he b anches a e decay alues ( he
minimum numbe o addi ional s eps equi ed o collapse each clade). The boo s ap analysis was con-
s ained o conside only ees in which Chi op e a was monophyle ic; his was done because mos o
he cha ac e s ha suppo ba monophyly ( able 5) we e omi ed om his s udy.
RESULTS OF ANALYSES
Analysis 1: All cha ac e s, ossil axa ex-
cluded ( ig. 35). As no ed abo e, his anal-
ysis was designed o p o ide a s a ing poin
by e alua ing ela ionships o ex an lineages
in he con ex o he e ised da a ma ix. This
e ec i ely ep esen s an upda ed e sion o
Simmons’ (1998) analysis, and illus a es he
e ec s o changes ha we made in he da a
ma ix (e.g., co ec ions, new cha ac e s, in-
clusion o hyoid da a o An ozoidae; see
abo e). Analysis o he e ised da a se e-
sul ed in a single mos pa simonious ee
(661 s eps; CI
5
0.404, RI
5
0.579) shown
in igu e 35.
Compa isons o he esul s o ou Analysis
1 ( ig. 35) wi h Simmons (1998) ee ( ig. 20)
e eals se e al opological di e ences, all
wi hin he yangochi op e an pa o he ee.
Fi s , ela ionships wi hin Na aloidea a e ul-
ly esol ed in ou Analysis 1 ee, wi h Thy-
op e idae unambiguously placed as he sis-
e -g oup o he Fu ip e idae
1
Na alidae
clade. This ep esen s a i ial change, since
his opology occu ed in one o he wo mos
pa simonious ees ound by Simmons
(1998) and ecei ed highe boo s ap suppo
han did any o he al e na i es in ha s udy.
A po en ially mo e signi ican change in ou
Analysis 1 ee is placemen o An ozoidae
1998 89SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
as he sis e - axon o Vespe ilionidae (i.e.,
Vespe ilioninae
1
Miniop e inae
1
Myo i-
nae
1
Mu ininae
1
Ke i oulinae) a he han
as he sis e - axon o Molossidae. This sug-
ges s ha Molossoidea as de ined by Sim-
mons (1998) may no be monophyle ic, and
ha An ozoidae may o m a clade wi h es-
pe ilionids as adi ionally hough (e.g., by
Koopman, 1993, 1994). Howe e , suppo
o he An ozoidae
1
Vespe ilionidae clade
was e y low in Analysis 1 (boo s ap alue
5
12%; minimum o one addi ional s ep o
collapse clade), indica ing conside able un-
ce ain y ega ding his g ouping. The same
is also ue o o he clades ha ep esen
changes om Simmons (1998) opology: (1)
placemen o Molossidae (Tomopea inae
1
Molossinae) as he sis e -g oup o he An o-
zoidae
1
Vespe ilionidae clade (boo s ap
alue
5
33%; minimum o one addi ional
s ep o collapse clade); and (2) placemen o
Mys acinidae as he sis e - axon o Molossi-
dae
1
An ozoidae
1
Vespe ilionidae (boo -
s ap alue
5
20%; minimum o one addi-
ional s ep o collapse clade). In essence, all
o hese changes ep esen ea angemen s o
ela ionships a nodes ha we e poo ly sup-
po ed in Simmons (1998) analysis and e-
main poo ly suppo ed in ou analyses o he
upda ed da a se .
Analysis 2: All cha ac e s, all axa ( ig.
36). This analysis ep esen s he p incipal
goal o ou p ojec , a cha ac e -cong uence
s udy including bo h Eocene ossil gene a
and ex an lineages. Encou agingly, pa si-
mony analyses o he comple e da a se in
Analysis 2 p oduced a single mos -pa simo-
nious ee ( ig. 36; 681 s eps; CI
5
0.392, RI
5
0.587). Rela ionships o Ica onyc e is,A -
chaeonyc e is,Hassianyc e is, and Palaeo-
chi op e yx we e ully esol ed, and place-
men o each was mode a ely o s ongly sup-
po ed in he boo s ap and decay analyses
(boo s ap alues
5
54–93%; minimum o
one o six addi ional s eps o collapse clade).
These le els o suppo a e compa able o
hose ound o many ex an clades in Anal-
ysis 2. Indeed, boo s ap alues o b anches
associa ed wi h many ex an axa (e.g., Mys-
acinidae, Vespe ilioninae, Miniop e inae,
Myo inae, Mu ininae, Ke i oulinae) a e
much lowe han hose associa ed wi h he
ossil b anches despi e he ac ha he ex an
lineages a e ep esen ed by wo o h ee
imes mo e da a ( able 6). The absence o
so - issue and molecula cha ac e da a
clea ly does no p eclude ela i ely secu e
placemen o he ossil o ms in his s udy.
Resul s o ou analysis indica e ha Ica-
onyc e is,A chaeonyc e is,Hassianyc e is,
and Palaeochi op e yx ep esen a se ies o
consecu i e sis e - axa o he mic ochi op e -
an c own g oup (i.e., he g oup comp ised o
all ex an lineages). They do no o m a pa a-
phyle ic g oup ances al o bo h Megachi op-
e a and Mic ochi op e a (e.g., Eochi op e a
sensu Van Valen, 1979), a monophyle ic
g oup wi hin Mic ochi op e a (e.g., Palaeo-
chi op e ygoidea sensu Smi h, 1977), o ha e
special a ini ies wi h a ious ex an mic o-
chi op e an supe amilies (as sugges ed by
Smi h and S o ch, 1981). O he ou Eocene
gene a, Palaeochi op e yx sha es he mos
de i ed ai s wi h ex an mic ochi op e ans.
Hassianyc e is and A chaeonyc e is a e con-
secu i e sis e - axa o he clade including Pa-
laeochi op e yx and he mic ochi op e an
c own g oup, and Ica onyc e is occupies he
basalmos b anch in he mic ochi op e an
pa o he ee.
Compa isons o he esul s o Analyses 1
and 2 ( igs. 35, 36) demons a e ha inclu-
sion o Ica onyc e is,A chaeonyc e is,Has-
sianyc e is, and Palaeochi op e yx in he
analysis p oduced mino changes in opology
o he ee, wi h he changes again cen e ed
on hose pa s o he ee ha we e weakly
suppo ed in p e ious analyses. Speci ically,
inclusion o he ossils changed he posi ion
o An ozoidae (which in Analysis 2 now
o ms a clade wi h Molossidae, suppo ing
monophyly o Molossoidea) and Mys acini-
dae (which now appea s as he sis e - axon
o Na aloidea
1
Molossoidea
1
Vespe ilion-
oidea). In e es ingly, all o hese ela ionships
appea ed in Simmons’ (1998) ee ( ig. 20),
al hough no in he esul s o ou Analysis 1
( ig. 35).
Inclusion o he ossils a ec ed he pe -
cei ed suppo o many clades e en when
he e was no e ec on ee opology. This
e ec was mos no iceable nea he base o
ee. Monophyly o he mic ochi op e an
c own g oup (Emballonu idae
1
Yinochi op-
e a
1
Yangochi op e a) ecei ed only mod-
e a e suppo in Analysis 2 (boo s ap alue
90 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 36. Resul s o Analysis 2, which included all cha ac e s and all axa (see ex o discussion).
Pa simony analysis esul ed in a single mos -pa simonious ee (680 s eps; CI
5
0.393; RI
5
0.587),
which is shown he e. The numbe s below in e nal b anches a e boo s ap alues; numbe s abo e he
b anches a e decay alues. The boo s ap analysis was cons ained o conside only ees in which
Chi op e a was monophyle ic.
5
79%; minimum o wo addi ional s ep o
collapse clade), whe eas i ecei ed ex eme-
ly high suppo in Analysis 1 (boo s ap al-
ue
5
100%; minimum o 16 addi ional s eps
o collapse clade). Clea ly, many o he de-
i ed ai s ha diagnose ex an Mic ochi op-
e a e ol ed in a sequen ial pa e n o e
ime; inclusion o he ossil axa ‘‘sp eads
ou ’’ hese synapomo phies o e a la ge pa
o he ee, hus educing pe cei ed suppo
o any single b anch. Ne e heless, mono-
phyly o he mic ochi ope an c own g oup
emains s ongly suppo ed in con as o o h-
e hypo heses. Al e na i e opologies, includ-
ing hose sugges ed by Smi h and S o ch
(1981), appea unlikely gi en esul s o he
boo s ap analysis. Fo example, Hassianyc-
e is o med a clade wi h Yangochi op e a in
only 6% o he boo s ap eplica es. Palaeo-
chi op e yx g ouped wi h Yangochi op e a in
only 5% o he boo s ap eplica es, and
o med a clade wi h some o all Vespe i-
lionoidea in ewe han 1% o he boo s ap
eplica es.
As in Simmons (1998) and Analysis 1 ( ig.
20, 35), esul s o Analysis 2 placed Embal-
lonu idae as he sis e -g oup o he clade con-
aining all o he ex an mic ochi op e an lin-
1998 97SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
TABLE 7
A Re ised Summa y o Mo phological Synapomo phies o Chi op e a
1) Deciduous den i ion does no esemble adul den i ion; deciduous ee h wi h long, sha p, ecu ed cusps
2) Pala al p ocess o p emaxilla educed; le and igh incisi e o amina used in midsaggi al plane
3) Pos pala ine o us absen
4) Jugal educed and jugolac imal con ac los
5) Two en o ympanic elemen s in he loo o he middle-ea ca i y: a la ge caudal elemen and a small os al
elemen associa ed wi h he in e nal ca o id a e y
6) Tegmen ympani ape s o an elonga e p ocess ha p ojec s in o he middle-ea ca i y medial o he epi ympanic
ecess
7) P oximal s apedial a e y en e s c anial ca i y medial o he egmen ympani; amus in e io passes an e io ly
do sal o he egmen ympani
8) Enla ged enes a o undum
9) Vome onasal epi helial ube absen
10) Accesso y ol ac o y bulb absen
11) Pos e io laminae p esen on ibs
12) Modi ica ion o scapula: eo ien a ion o scapula spine and modi ica ion o shape o scapula ossae; educ ion
in o heigh o spine; p esence o a well-de eloped ans e se scapula ligamen ; p esence o a leas wo ace s
in in aspinous ossa
13) Modi ica ion o elbow: educ ion o olec anon p ocess and hume al a icula su ace on ulna; p esence o ulna
pa ella; absence o olec anon ossa on hume us
14) Absence o supina o idge on hume us
15) Absence o en epicondyla o amen in hume us
16) Occipi opollicalis muscle and cephalic ein p esen in leading edge o p opa agium
17) Digi s II–V o o elimb elonga ed wi h complex ca pome aca pal and in e me aca pal join s, suppo enla ged
in e digi al ligh memb anes (pa agia); digi s III–V lack claws
18) Modi ica ion o hip join : 90
8
o a ion o hindlimbs e ec ed by eo ien a ion o ace abulum and sha o emu ;
neck o emu educed; ischium il ed do sola e ally; an e io pubes widely la ed and pubic spine p esen ; absence
o m. ob u a o in e nus
19) Sac um e mina es pos e io o midpoin o ace abulum
20) Absence o m. glu eus minimus
21) Absence o m. sa o ius
22) Vas us muscle complex no di e en ia ed
23) Modi ica ion o ankle join : eo ien a ion o uppe ankle join ace s on calcaneum and as agalus; ochlea o
as agalus con ex, lacks medial and la e al guiding idges; ube o calcaneum p ojec s in plan ola e al di ec ion
away om ankle and oo ; pe oneal p ocess absen ; sus en acula p ocess o calcaneum educed, calcaneoas a-
gala and sus en acula ace s on calcaneum and as agalus coalesced; absence o g oo e on as agalus o endon
o m. lexo digi o um ibula is
24) En ocunei o m p oximodis ally sho ened, wi h la , iangula dis al ace
25) Elonga ion o p oximal phalanx o digi I o oo
26) Emb yonic disc o ien ed owa d ubo–u e ine junc ion a ime o implan a ion
27) Di e en ia ion o a ee, glandlike yolk sac
28) P eplacen a and ea ly cho ioallan oic placen a di use o ho seshoe-shaped, wi h de ini i e placen a educed o a
mo e localized discoidal s uc u e
29) De ini i e cho ioallan oic placen a endo heliocho ial
30) Baculum p esen
31) Le cen al lobe o li e sepa a e om o he lobes o pa ially used wi h igh cen al lobe
32) Caecum absen
33) Co ical soma osenso y ep esen a ion o o elimb e e se o ha in o he mammals
e is,Hassianyc e is,Palaeochi op e yx, and
mos membe s o he mic ochi op e an
c own g oup. Because he ype o pe io ic
a achmen could no be sco ed in a ailable
specimens o Ica onyc e is, we canno be
su e i his cha ac e ans o ma ion occu ed
be o e o a e di e en ia ion o Ica onyc e -
is om he s ock leading o ex an mic ochi-
op e ans.
The nex highes node in he mic ochi op-
98 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
e an ee uni es A chaeonyc e is wi h Has-
sianyc e is,Palaeochi op e yx, and he mi-
c ochi op e an c own g oup. As no ed abo e,
loosening o he connec ion be ween he pe -
io ic and basisphenoid (cha ac e 25) mini-
mally applies a his le el in he ee. O he
de i ed cha ac e s ha diagnose his clade
include (1) educ ion o he numbe o oo s
on P3 o wo o ewe (cha ac e 19; P3 wi h
h ee oo s in bo h ou g oups, P e opodidae,
and Ica onyc e is); (2) p esence o a en al
accesso y p ocess on ce ical e eb a 4
(cha ac e 76; accesso y p ocess absen om
C4 in bo h ou g oups, P e opodidae, and Ica-
onyc e is); and (3) absence o ossi ied hi d
phalanx in wing digi III (cha ac e 152; ully
ossi ied hi d phalanx p esen in bo h ou -
g oups and Ica onyc e is).
O hese cha ac e s, educ ion o he num-
be o oo s on P3 and p esence o a en al
accesso y p ocess on C4 a e synapomo phies
ha unequi ocally diagnose he clade com-
p ising A chaeonyc e is and he lineage lead-
ing o he mic ochi op e an c own g oup.
The absence o an ossi ied hi d phalanx on
wing digi III ep esen s an equi ocal ans-
o ma ion because his condi ion also cha -
ac e izes P e opodidae. Al e na i e explana-
ions o he obse ed pa e n include (1) in-
dependen loss o ossi ica ion o his phalanx
in p e opodids and he lineage leading o ex-
an mic ochi op e ans; o (2) loss o ossi i-
ca ion o he hi d phalanx in he common
ances o o ba s, and seconda y acquisi ion
o ossi ica ion o his elemen in Ica onyc-
e is. Bo h hypo heses a e equally pa simo-
nious, and hey also seem equally likely gi -
en he complex pa e n seen elsewhe e in he
ee (e.g., h ee pa e ns occu wi hin Yan-
gochi op e a— ull ossi ica ion [seconda ily
acqui ed once], ossi ica ion only a phalanx
base [acqui ed wice], and no ossi ica ion
[seconda ily acqui ed once]).
O he cha ac e ans o ma ions ha occu
somewhe e nea he base o he mic ochi op-
e an ee (we canno de e mine exac ly
whe e because o missing da a) in ol e e o-
lu ion o he phane ocochlea condi ion
(cha ac e 27), a deep, cons ic ed s apedial
ossa (cha ac e 31), p esence o an e io
laminae on he ibs (cha ac e 86), b oad pos-
e io laminae on he ibs (cha ac e 89), and
p esence o a do sal a icula ace on he
scapula (cha ac e 100). The c yp ocochlea
condi ion is seen in bo h ou g oups and P e -
opodidae; he phane ocochlea condi ion oc-
cu s in Hassianyc e is,Palaeochi op e yx,
and p imi i ely in he mic ochi op e an
c own g oup. The s apedial ossa in he ou -
g oups and P e opodidae is indis inc o shal-
low and b oad; his con as s sha ply wi h he
deep, cons ic ed s apedial ossae seen in Pa-
laeochi op e yx and p imi i ely in he mic o-
chi op e an c own g oup. Because ex e nal
cochlea mo phology and o m o he s ape-
dial ossa could no be sco ed in a ailable
specimens o ei he Ica onyc e is o A -
chaeonyc e is, we canno be su e whe e on
he ee be ween P e opodidae and Palaeo-
chi op e yx he ans o ma ions o a phane -
ocochlea condi ion and a deep, cons ic ed
s apedial ossa occu ed. No can we ell i
hese changes we e coinciden o occu ed a
di e en poin s in he ee. In e p e i e di -
icul ies simila ly occu wi h o he cha ac e
ans o ma ions as a esul o missing da a.
An e io laminae a e absen om he ibs o
bo h ou g oups, some P e opodidae, and Ica-
onyc e is, while an e io laminae a e p esen
in Palaeochi op e yx and p imi i ely wi hin
he mic ochi op e an c own g oup. Pos e io
laminae a e na ow in P e opodidae and Ica-
onyc e is, bu a e b oad in Hassianyc e is,
Palaeochi op e yx, and p imi i ely wi hin
he mic ochi op e an c own g oup. A do sal
a icula ace is absen om he scapula in
he ou g oups, P e opodidae, and Ica onyc-
e is; howe e , i is p esen in Hassianyc e is,
Palaeochi op e yx, and p imi i ely in he mi-
c ochi op e an c own g oup. Because he
condi ions o hese cha ac e s in A chaeo-
nyc e is could no be de e mined, we canno
be su e i he ans o ma ions in hese s uc-
u es occu ed be o e o a e di e en ia ion
o A chaeonyc e is om he lineage leading
o he mic ochi op e an c own g oup.
Mo ing up he mic ochi op e an pa o
he ee, he nex node uni es Hassianyc e is,
Palaeochi op e yx, and he mic ochi op e an
c own g oup. As no ed abo e, p esence o
he phane ocochlea condi ion (cha ac e 27),
a deep, cons ic ed s apedial ossa (cha ac e
31), an e io laminae on he ibs (cha ac e
86), b oad pos e io laminae on he ibs
(cha ac e 89), and a do sal a icula ace on
he scapula (cha ac e 100) may apply a his
1998 99SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
le el in he ee. De i ed cha ac e s ha un-
ambiguously diagnose he clade comp ising
Hassianyc e is,Palaeochi op e yx, and he
mic ochi op e an c own g oup include: (1)
nyc alodon y (cha ac e 21; p imi i e ibos-
phenic condi ion in Scanden ia, Ica onyc e -
is, and A chaeonyc e is); (2) a g ea ly en-
la ged cochlea (cha ac e 26; ei he no en-
la ged o only mode a ely enla ged in he
ou g oups, P e opodidae, Ica onyc e is and
A chaeonyc e is); (3) dis al ip o en al
p ocess o manub ium la e ally comp essed
(cha ac e 91; ip blun o ounded in De -
mop e a, Ica onyc e is, and A chaeonyc e -
is); (4) absence o an ossi ied hi d phalanx
(claw) on wing digi II (index inge ; cha -
ac e 151; ossi ied claw p esen in bo h ou -
g oups, P e opodidae, Ica onyc e is, and A -
chaeonyc e is); (5) ibula hin and h eadlike
(cha ac e 170; ibula well de eloped in bo h
ou g oups, P e opodidae, Ica onyc e is, and
A chaeonyc e is); (6) calca p esen (cha -
ac e 171; calca absen in bo h ou g oups,
Ica onyc e is, and A chaeonyc e is). Op i-
miza ion o he la e cha ac e ( ig. 38) is
somewha su p ising gi en p e ious hypo h-
eses (e.g., Simmons, 1994, 1995; see able 5)
ha p esence o a calca is a synapomo phy
o Chi op e a. Ou analysis sugges s o he -
wise. A calca seems o ha e been absen in
he mos ecen common ances o o Mega-
chi op e a and Mic ochi op e a, and appa -
en ly e ol ed independen ly in P e opodidae
and in he lineage leading o ex an mic o-
chi op e ans.
The las clade di ec ly in ol ing he ossil
axa analyzed in his s udy is ha comp ising
Palaeochi op e yx plus he mic ochi op e an
c own g oup. Only a single de i ed cha ac e
unequi ocally diagnoses his clade: p esence
o a en al accesso y p ocess on ce ical
e eb a 5 (cha ac e 77; absen in bo h ou -
g oups, P e opodidae, Ica onyc e is,A -
chaeonyc e is, and Hassianyc e is). Two o h-
e cha ac e s minimially apply a his le el,
bu lack o da a o Hassianyc e is and A -
chaeonyc e is make i impossible o de e -
mine he poin o ans o ma ion on he ee.
These cha ac e s include (1) a deep, con-
s ic ed s apedial ossa (cha ac e 31; ossa
indis inc in bo h ou g oups, shallow and
b oad in P e opodidae), and (2) p esence o
an e io laminae on he ibs (cha ac e 86;
laminae absen in bo h ou g oups, some P e -
ododidae, and Ica onyc e is).
FEATURES DIAGNOSING THE
MICROCHIROPTERAN CROWN GROUP
Cha ac e ans o ma ions ha appea o
diagnose he mic ochi op e an c own g oup
all in o wo b oad ca ego ies: (1) ea u es
ha can be posi i ely a ibu ed o his pa -
icula node (by i ue o ha ing been sco ed
wi h a di e en s a e in Hassianyc e is and
mo e dis al ou g oups); and (2) de i ed ea-
u es ha minimally diagnose he c own
g oup bu could no be sco ed in he ossils,
lea ing open he possibili y ha hey may
ha e e ol ed ea lie (close o he basal
node) in he ee. T ans o ma ions o he la -
e so may be unambiguous synapomo -
phies, bu i is no clea a wha le el hey
apply.
In he i s ca ego y, we ind ha only
h ee ans o ma ions in ha d- issue cha ac-
e s appea o diagnose he mic ochi op e an
c own g oup: (1) ee p emaxilla (cha ac e
9; su u ed p emaxilla in bo h ou g oups,
P e opodidae, and all ou Eocene gene a);
(2) educ ion o wo lowe p emola s on each
side o he jaw (cha ac e 20; h ee lowe p e-
mola s in P e opodidae and all ou Eocene
gene a); and (3) a xiphis e num wi h p omi-
nen median keel (cha ac e 95; keel absen
in bo h ou g oups, P e opodidae, Ica onyc-
e is,A chaeonyc e is, and Palaeochi op e-
yx).
In compa ison o he lis s o ea u es di-
agnosing mo e inclusi e clades (see abo e),
hese synapomo phies by hemsel es a e no
compelling. Fo example, ew wo ke s would
ag ee ha a ee p emaxilla is likely p imi-
i e o all ex an mic ochi op e ans (as sug-
ges ed by op imiza ion o his cha ac e on
ou ee; ig. 39) because yangochi op e an
ba s lack his complex specializa ion. In-
s ead, yangochi op e ans ha e a p emaxilla
ha is i mly used o he maxilla in he
adul . The ee p emaxilla is a ea u e unique
among mammals, and we ag ee ha i is ha d
o imagine ei he i s loss o independen o -
igin in wo di e en g oups (which a e, o
cou se, he wo mos -pa simonious hypo he-
ses gi en he opology o ou op imal ee).
Reduc ion om h ee o wo lowe p emola s
100 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 39. P emaxilla–maxilla a icula ions (cha ac e 9) mapped on he phylogeny om igu e 36.
P esence o a ee p emaxilla appea s o be a synapomo phy o he mic ochi op e an c own g oup
(al hough see ex o discussion). Fusion o he p emaxilla o he maxilla is a synapomo phy o Yan-
gochi op e a. The condi ion o P e opodidae is depic ed as ‘‘unce ain’’ because wo condi ions occu
wi hin his g oup: he p emaxilla is su u ed o he maxilla in mos p e opodids, bu hese elemen s a e
used in o he axa. This op imiza ion sugges s ha a su u e join be ween he p emaxilla and maxilla
is p imi i e o P e opodidae.
on each side is a cha ac e ha is simila ly
oublesome, because p esence o h ee lowe
p emola s cha ac e izes many g oups o ex-
an mic ochi op e ans ha nes well up wi h-
in he c own g oup (e.g., Rhinolophidae,
Mo moopidae, Na aloidea, many espe ili-
onids). This is also ue o he xiphis e nal
keel. Al hough p esence o a keel appea s o
diagnose he mic ochi op e an c own g oup,
many lineages wi hin ha clade lack a keel
on he xiphis e num (e.g., Rhinopoma idae,
Hipposide inae, some Phyllos omidae, Noc-
ilionidae, Myzopodidae, Molossoidea, some
Vespe ilioninae).
Despi e he ela i e weakness o hese
da a, monophyly o he mic ochi op e an
c own g oup is indi ec ly suppo ed by a
b oad ange o o he cha ac e s ha un o -
una ely ha e somewha ambiguous dis i-
bu ions. These ans o ma ions minimally di-
agnose he mic ochi op e an c own g oup,
bu migh ha e e ol ed ea lie in he ee;
we could no sco e hem in ossil sis e - axa
o he mic ochi op e an c own g oup. Such
ea u es a e nume ous and include he ol-
lowing: (1) p esence o a agus (cha ac e 2;
absen in ou g oups and P e opodidae); (2)
aquaeduc us cochleae small o absen (cha -
ac e 33; la ge in De mop e a and P e opod-
idae); (3) p esence o sophis ica ed echolo-
ca ion (cha ac e 36; absen in ou g oups and
P e opodidae); (4) m. s yloglossus o igina es
om en al su ace o midpoin o s ylohyal
(cha ac e 58; o igina es om expanded ip
o s ylohyal and/o adjacen su ace o skull
in ou g oups and P e opodidae); (5) cla icle
a icula es wi h co acoid p ocess o scapula
(cha ac e 113; cla icle a icula es wi h ac o-
mion in bo h ou g oups, P e opodidae, and
Ica onyc e is); (6) m. spino apezius clea ly
di e en ia ed om apezius complex (cha -
ac e 124; m. spino apezius no di e en i-
a ed om apezius complex in bo h ou -
g oups and P e opodidae); (7) o igin o m.
ac omiodel oideus does no include ho acic
e eb a 6 (cha ac e 127; o igin does include
T6 in bo h ou g oups and P e opodidae); (8)
spinal co d wi h angle be ween do sal ho ns
1998 101SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
o 0–25
8
(cha ac e 190; angle be ween do -
sal ho ns 35–50
8
in bo h ou g oups, 70–80
8
in P e opodidae); and (9) in e io colliculus
la ge han supe io colliculus (cha ac e 191;
in e io colliculus signi ican ly smalle han
supe io colliculus in bo h ou g oups and
P e opodidae). One o he ans o ma ion may
diagnose he mic ochi op e an c own g oup,
al hough i has been documen ed only in
Yinochi op e a and Yangochi op e a (no da a
o Emballonu idae): m. lexo digi o um
p o undus does no inse on digi II o wing
(cha ac e 154; muscle does inse on digi II
in bo h ou g oups and P e opodidae). E en i
hese ea u es e ol ed in a s epwise ashion
up he ee, a leas a ew p obably ep esen
ue synapomo phies o he mic ochi op e an
c own g oup.
CHARACTER TRANSFORMATIONS
AT BASAL NODES:
A FUNCTIONAL PERSPECTIVE
In he p e ious sec ion, we desc ibed he
cha ac e ans o ma ions ha seem o apply
a each node in he basal pa o he mic o-
chi op e an ee. An in e es ing pa e n also
eme ges when hese ans o ma ions a e
iewed om a unc ional pe spec i e. Ra he
han changes in each o gan sys em being
concen a ed a one o wo nodes, we see a
pa e n o s epwise changes in mul iple unc-
ional sys ems as we mo e up he mic ochi-
op e an ee. This sugges s a complex, mo-
saic pa e n o e olu ion in which se e al
o gan sys ems we e being e ined simul a-
neously.
O e all, he changes associa ed wi h he
acial egion and mas ica o y appa a us a e
ela i ely mino compa ed o hose seen in
o he sys ems. In o de o hei appea ance
along he backbone o he mic ochi op e an
ee (beginning a he base, p io o di e -
gence o Ica onyc e is), hese ans o ma-
ions include (1) e olu ion o an elonga e an-
gula p ocess ( hough his may ul ima ely
p o e o be plesiomo phic o ba s), (2) e-
duc ion o he numbe o oo s on P3, (3)
e olu ion o nyc alodon y, and (4) educ ion
in he numbe o lowe p emola s and mod-
i ica ion o he simple su u ed connec ion be-
ween he p emaxilla and maxilla.
Func ional implica ions o hese changes
emain obscu e, al hough Slaugh e (1970)
no ed ha educ ion and simpli ica ion o he
p emola den i ion in ea ly ba s appea s co -
ela ed wi h sho ening o he ace (b achy-
cephaly). I is possible ha hese ends may
ha e been ela ed o he need o ocus he
ea s an e io ly and educe in e e ence wi h
e u ning echoloca ion calls— he audi o y
equi alen o he ype o acial sho ening
gene ally associa ed wi h e olu ion o bin-
ocula ision. Ano he possibili y is ha loss
o p emola s and educ ion in he numbe o
oo s on he emaining ee h ep esen mech-
anisms o mass educ ion. Flying animals
mus gene a e adequa e li o emain ai -
bo ne, and li equi emen s inc ease wi h in-
c easing body mass (No be g, 1985, 1986a,
1987, 1990; Rayne , 1986; No be g and Ray-
ne , 1987). Dis ibu ion o body mass is also
impo an because i a ec s he posi ion o
he cen e o mass (
5
cen e o g a i y),
which in u n in luences ligh e iciency. The
head is he hea ies pa o he body in mos
e eb a es in pa because o he densi y o
he ee h. In addi ion o he e ec ha a
hea y head may ha e on o al body mass and
he loca ion o he cen e o mass, head mass
also a ec he size o neck muscles needed
o suppo he head and esis o que (Bu¨hle ,
1992). Den al educ ion in bi ds (mos o
which lack ee h en i ely) is widely ega ded
as a specializa ion ha inc eased ligh e i-
ciency because i simul aneously educed o-
al mass and concen a ed mo e o he body
mass nea he cen e o g a i y be ween he
wings (e.g., Wel y, 1955; S ahl, 1985; Bu¨hl-
e , 1992; Feduccia, 1996). Den al educ ion
in p e osau s has also been no ed as a pos-
sible adap a ion o educing body mass and
inc easing ligh e iciency (e.g., S ahl, 1985;
Bu¨hle , 1992). Because ba s a e mammals
ha ely on hei den i ion o ood p ocess-
ing, ex eme den al educ ion is a e (limi ied
mos ly o nec a i o ous axa). Howe e , e en
small educ ions in mass may con ibu e o
ligh e icency.
The ee p emaxilla p esen in Emballon-
u idae and Yinochi op e a (and pe haps in
he mos ecen common ances o o ex an
mic ochi op e ans; ig. 39) p o ides an un-
usual deg ee o mobili y in he snou . U ili y
o his ea u e has no been adequa ely in-
es iga ed, bu ou expe ience in handling
102 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
li e emballonu ids sugges s ha do si lexion
o he snou is unde olun a y con ol. Do -
si lexion o he snou e ec i ely inc eases
he gape o he mou h, which may be im-
po an in eeding and/o emission o echo-
loca ion calls. Wi h espec o he la e , i is
in e es ing o no e ha he mode o emission
o echoloca ion calls—ei he h ough he
mou h (o al emission) o h ough he nasal
passages (nasal emission)— a ies among
ba s ha ha e a ee p emaxilla. Emballon-
u idae, C aseonyc e idae, and pe haps Rhin-
opoma idae a e o al emi e s, while Rhino-
lophoidea and pe haps Rhinopomaidae a e
nasal emi e s (Pede son, 1993). Gi en o-
pology o ou ee ( ig. 36), i appea s ha
o al emission is p imi i e and nasal emission
de i ed as sugges ed by Van Valen (1979)
and Pede son (1993). The ex en o which
do si lexion o he snou may ha e played a
ole in he e olu ion o nasal emission in yin-
ochi op e an ba s has no been explo ed.
F ee mo emen o he p emaxilla may also
be impo an in p ey manipula ion, pa icu-
la ly in hose axa ha ake la ge a h opod
o small e eb a e p ey. Fen on (pe sonal
commun.) obse ed a Nyc e is g andis ea -
ing a Nyc e is hebiaca, and no ed ha he
o me used i s uppe lips ex ensi ely and
manipula ed i s p ey in an almos closed-
mou h ashion. The ex en o which a ee
p emaxilla may acili a e he capaci y o han-
dle la ge p ey i ems has ye o be in es iga -
ed.
Changes in he basic anium and ea egion
also appea o ha e e ol ed in a s epwise
ashion. The i s modi ica ions o appea
(p io o he di e gence o Ica onyc e is
om he mic ochi op e an s em s ock) in-
cluded a mode a ely enla ged cochlea, en-
la ged o bicula apophysis on he malleus,
and an expanded c anial ip on he s ylohyal.
All o hese ea u es a e p obably ela ed o
he e olu ion o echoloca ion. As no ed ea -
lie , p esence o a la ge o bicula apophysis
may imp o e he abili y o he middle ea
ossicles o ansmi high- equency sounds
wi h a minimum ime delay, o may play a
ole in he a oidance o sel -dea ening (see
discussion unde cha ac e 35 abo e). The
ac ha all known ex an mic ochi op e ans
ha e an enla ged o bicula apophysis—and
all use echoloca ion—leads us o suspec ha
wha e e he unc ion(s) o his s uc u e, an
enla ged o bicula apophysis e ol ed o a-
cili a e e icien unc ion o he ea in echo-
loca ion. Likewise, he expanded ip on he
s ylohyal may ep esen a e inemen o he
sys em o p oducing echoloca ion calls. The
mic ochi op e an echoloca ion sys em de-
pends upon calls p oduced in he la ynx,
which is suppo ed in pa by muscles asso-
cia ed wi h he hyoid appa a us. The expand-
ed ip on he s ylohyal—which is i mly a -
ached o he bulla—se es o ancho he hy-
oid appa a us o he skull, hus p o iding a
s able a achmen si e o muscles o he
h oa . This a achmen may be impo an in
suppo ing he la ynx du ing p oduc ion o
echoloca ion calls.
Enla gemen o he cochlea (speci ically
he basal u n) in mic ochi op e ans appea s
o be a specializa ion o inc eased sensi i i y
o high- equency sounds (
.
20 kHz), such as
he e u ning echoes om ocaliza ions used
in echoloca ion (Henson, 1970; No acek,
1985a, 1987, 1991; Habe se ze and S o ch,
1992). In all mammals, he basal u n o he
cochlea is he egion whe e high- equency
sounds a e pe cei ed (Henson, 1970; Dallos,
1973; B uns, 1979; B uns e al., 1983–1984;
Ha is and Dallos, 1984). Enla gemen o
his egion, which e ec i ely leng hens he
basal po ion o he basila memb ane, ap-
pa en ly inc eases sensi i i y o high- e-
quency sounds and sligh equency shi s in
his ange (Henson, 1970; Dallos, 1973;
B uns, 1979; B uns e al., 1983–1984; Ha is
and Dallos, 1984).
In he lineage leading o ex an Mic ochi-
op e a, cochlea enla gemen seems o ha e
e ol ed in a ela i ely con inuous ashion
ha we chose o sco e as a se ies o s eps
( ig. 40). The p imi i e condi ion o ba s
was he p esence o an unenla ged cochlea,
one compa able in size o hose seen in o he
mammals o simila body size. A mode a ely
enla ged cochlea e ol ed p io o he di e -
gence o Ica onyc e is, o be ollowed sub-
sequen ly by e en g ea e enla gemen be o e
he di e gence o Hassianyc e is. Mode a e
enla gemen o he cochlea does no by i sel
indica e ha a ba could echoloca e, because
some nonecholoca ing p e opodids ha e a
mode a ely enla ged cochlea (see discussion
unde cha ac e 26 abo e). Howe e , he
1998 103SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 40. Cochlea enla gemen (cha ac e 26) mapped on he phylogeny om igu e 36. In he lineage
leading o ex an mic ochi op e ans, a ‘‘mode a ely enla ged’’ cochlea e ol ed p io o he di e gence
o Ica onyc e is; a ‘‘g ea ly enla ged’’ cochlea e ol ed p io o he di e gence o Hassianyc e is. Re-
e sals o he o me condi ion appa en ly occu ed independen ly in Mys acinidae and wi hin Megad-
e ma idae and Phyllos omidae (no e ha he la e axa a e ma ked as ‘‘unce ain’’ as a esul o axo-
nomic polymo phism). This op imiza ion also sugges s ha a mode a ely enla ged cochlea e ol ed
independen ly wi hin P e opodidae (see ex o discussion).
combina ion o a mode a ely enla ged coch-
lea, enla ged o bicula apophysis on he mal-
leus, and an expanded c anial ip on he s y-
lohyal a e seen in ex an ba s only in o ms
ha use echoloca ion. We he e o e ollow
p e ious au ho s (e.g., No acek, 1985a,
1987, 1991; Habe se ze and S o ch, 1992)
in in e ing ha Ica onyc e is,A chaeonyc-
e is,Hassianyc e is, and Palaeochi op e yx
used echoloca ion. We ag ee wi h Pye (1968:
797), who obse ed ha
The Eocene b ough mammals mean
And ba s began o sing;
Thei ood hey ound by ul asound
And chased i on he wing.
The implica ions o cochlea mo phology o
econs uc ing o aging s a egies is dis-
cussed in dep h below unde ‘‘Fo aging ecol-
ogy o Eocene ba s.’’
Loose a achmen o he pe io ic o he
basic anium e ol ed p io o he di e gence
o A chaeonyc e is, al hough we canno de-
e mine he exac le el in he phylogeny be-
cause his cha ac e could no be sco ed in
Ica onyc e is. Based on he p esumed unc-
ion o his loose a achmen —cochlea iso-
la ion, hough o unc ion in educing bone
conduc ion o la yngeal ib a ions (Henson,
1970; Van Valen, 1979)—we su mise ha
his ea u e e ol ed in conce wi h he ea ly
s ages o cochlea enla gemen and acili a -
ed he e olu ion o echoloca ion.
Ano he ans o ma ion ha ook place in
he mic ochi op e an lineage p io o he di-
e gence o Hassianyc e is was e olu ion o
a phane ocochlea cochlea. Un o una ely,
we canno de e mine he le el o his ans-
o ma ion in he phylogene ic ee because
his cha ac e could no be sco ed in Ica on-
yc e is o A chaeonyc e is. The phane o-
cochlea condi ion was explained by No a-
cek (1991: 84) as ‘‘an accommoda ion o he
p oblem o ‘‘packing’’ middle ea s uc u es
in a space cons ained by he expanded coch-
lea.’’ As discussed abo e unde cha ac e 27,
his ea u e is p obably linked o cochlea ex-
pansion, al hough we canno be su e a wha
poin he phane ocochlea condi ion e ol ed
ela i e o changes in cochlea size.
104 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
P io o he di e gence o Palaeochi op-
e yx, ye ano he ans o ma ion ook place:
e olu ion o a deep, cons ic ed ossa o m.
s apedius. Again, we canno de e mine he
le el o his ans o ma ion in he phyloge-
ne ic ee because his cha ac e could no be
sco ed in Ica onyc e is,A chaeonyc e is,o
Hassianyc e is. The unc ion o his ans o -
ma ion emains somewha obscu e, al hough
i is possible ha i is in some way linked o
he sys em o a oiding sel -dea ening ha is
used by many echoloca ing ba s (Henson,
1964, 1965, 1966, 1967a, 1979; Jen and
Suga, 1976; Fen on e al., 1995; see discus-
sion below).
T ans o ma ions in ea u es o he pos c a-
nial skele on also seem o ha e e ol ed in a
se ies o s eps up he ee. The i s change
in he pos c anium was p oximal ex ension
o he ochi e up o he le el o he head o
he hume us, which occu ed p io o he di-
e gence o Ica onyc e is. Se e al o he de-
i ed ai s e ol ed subsequen ly, al hough
we canno be su e a wha le el because hey
could no be sco ed in A chaeonyc e is.
These ai s include (1) p esence o a do sal
a icula ace on he scapula, (2) p esence o
an e io laminae on he ibs, and (3) p esence
o b oad pos e io laminae on he ibs. T ans-
o ma ions ha occu ed subsequen o he
di e gence o A chaeonyc e is bu p io o
he di e gence o Hassianyc e is include (1)
la e ally comp essed en al p ocess o he
manub ium, (2) loss o an ossi ied hi d pha-
lanx in digi II o he wing, (3) educ ion o
ibula o a hin, h eadlike elemen , and (4)
p esence o a calca .
Di iding hese ea u es by ana omical e-
gion, we see p og essi e changes occu ing
in di e en unc ional uni s. Modi ica ions o
he shoulde egion began wi h enla gemen
o he ochi e , which p eceded e olu ion o
a do sal a icula ace on he scapula. A
unc ional shoulde -locking mechanism,
which equi es a seconda y a icula ion be-
ween he ochi e and do sal a icula ace ,
was clea ly p esen by he ime ha Hassi-
anyc e is di e ged om he lineage leading
o ex an mic ochi op e ans. In e ms o mo -
phology, he p imi i e shoulde -lock appa -
en ly consis ed o a ochi e ha ex ended
up o bu no beyond he hume al head, and
a do sal a icula ace in he o m o a small
g oo e o o al si ua ed on he an e omedial
im o he glenoid ossa.
In he o elimb skele on, we see p og es-
si e educ ion in he numbe o phalanges in
digi II o he wing ( he index inge ) begin-
ning a e he onse o e olu iona y changes
in he shoulde egion ( ig. 41). Because he
humb is ela i ely small in all ba s (a leas
in compa ison o he o he digi s o he
hand), digi II o ms he leading edge o he
dac ylopa agium (‘‘hand wing’’) nea he
w is . The dac ylopa agium plays an impo -
an ole in bo h li gene a ion and s ee ing
du ing ligh (Vaughan, 1959; No be g, 1969,
1970, 1972b, 1976; Hill and Smi h, 1984).
S i ness o leading edge o he dac ylopa-
agium—pa icula ly he sec ion known as
he dac ylopa agium minus, which ex ends
be ween he second and hi d digi s—is c i -
ical o he abili y o he wing o esis bend-
ing and wis ing o ces (No be g, 1969,
1970, 1972b). This unc ional uni is also im-
po an in main aining wing cambe
(Vaughan, 1959, 1970c; No be g, 1970,
1972b, 1976).
In ba s lacking he dis al phalanges on dig-
i II, a ligamen ous connec ion uns be ween
he end o i s phalanx o digi II and he
base o he second phalanx in digi III (No -
be g, 1969, 1970, 1972b). This ligamen is
kep unde con inuous ension by he s uc-
u e o digi III, which is ben pos e odo sally
due o he s uc u e o he me aca pophal-
angeal join (No be g, 1969, 1970, 1972b).
This a angemen esul s in a con ex ame
wi hin which he dac ylopa agium minus is
kep s e ched unde ension, hus o ming a
uni ha is s i in he plane o he memb ane
(No be g, 1969, 1970, 1972b). Pho og aphs
o lying ba s show ha he ai s eam p o-
duces li le de o ma ion o he dac ylopa a-
gium minus o i s suppo ing elemen s du ing
ligh (No be g, 1969, 1970). The ensile
s eng h o he second digi is c i ical o he
un ion o his uni , since he second digi
mus esis he o ces placed upon he dac-
ylopa agium minus du ing ligh . Reduc ion
o he numbe o phalanges in digi II ap-
pa en ly ook place in a sequen ial ashion in
ba s, wi h comple e loss o ossi ied phalanges
e ol ing independen ly a leas h ee imes in
he mic ochi op e an c own g oup ( ig. 41).
I seems likely ha loss o phalanges se ed
1998 105SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Fig. 41. Changes in he numbe o phalanges on wing digi II ( he index inge ; cha ac e s 149, 150,
and 151) mapped on he phylogeny om igu e 36. P esence o h ee ossi ied phalanges on digi II is
p imi i e o ba s. Reduc ion o wo ossi ied phalanges occu ed jus p io o he di e gence o Hassi-
anyc e is om he lineage leading o he mic ochi op e an c own g oup. The mic ochi op e an c own
g oup is diagnosed by u he educ ion o only one ossi ied phalanx. Compe e phalangeal educ ion (no
ossi ied phalanges on wing digi II) appa en ly e ol ed independen ly a leas h ee imes in he c own
g oup—in Emballonu idae, Na aloidea, and ei he in he common ances o o Rhinolophoidea o inde-
penden ly in Nyc e idae and Rhinolophidae. P esence o wo ossi ied phalanges in Rhinopoma idae
appea s o be a e e sal a he han e en ion o he p imi i e condi ion.
o inc ease he ensile s eng h pe uni mass
o he dis al second digi by eplacing a ic-
ula ed phalanges (which oge he ha e a el-
a i ely low ensile s eng h pe uni mass)
wi h a con inuous ligamen ha has a highe
ensile s eng h pe uni mass. Mass educ-
ion, pa icula ly nea he dis al end o he
wing, con ibu es signi ican ly o e icien
ligh pe o mance (Swa z, 1997). Loss o
he dis al phalanges may ep esen a mech-
anism o dis al mass educ ion as con-
s ained by he need o p o ide he ensile
s eng h necessa y o main ain a s i dac y-
lopa agium minus du ing ligh .
Modi ica ions in he axial skele on in he
basal pa o he mic ochi op e an ee in-
cluded e olu ion o an e io laminae and
b oad pos e io laminae on he ibs. B oad
pos e io laminae e ol ed p io o he di e -
gence o Hassianyc e is, al hough we canno
be su e o he le el because we could no
sco e his cha ac e in A chaeonyc e is. Sim-
ila ly, an e io laminae e ol ed a some poin
p io o he di e gence o Palaeochi op e yx,
al hough we canno be su e o he le el be-
cause we could no sco e his cha ac e in
A chaeonyc e is o Hassianyc e is. In any
case, hese modi ica ions a e absen in Ica-
onyc e is, so hey mus ha e occu ed wi h-
in he basal mic ochi op e an lineage. We ex-
pec ha unc ion o ib laminae is wo old:
o s i en he ibcage and o p o ide la ge
a eas o muscle a achmen . Muscle g oups
ha o igina e di ec ly om he ibcage in-
clude hose o he m. se a us an e io com-
plex, which is a c i ical componen o he
ligh muscula u e. The pos e io di ision o
m. se a us an e io con ibu es o he down-
s oke, and he an e io di ision se es o an-
cho he medial edge o he scapula and may
also help o ini ia e he ups oke o he wing
(Vaughan, 1959, 1970b; No be g, 1970,
1972a; S ickle , 1978; He manson and Al-
enbach, 1983, 1985).
106 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
A close associa ion has been demons a ed
be ween wingbea and sound emission in mi-
c ochi op e ans unde labo a o y condi ions
(Schni zle , 1968, 1970a, 1970b, 1971, 1973;
Su he s e al., 1972; Schni zle and Henson,
1980; Joe man and Schmid , 1981; Heblich,
1986; Lancas e e al., 1992) and in ee- ly-
ing ba s o aging in na u e (Kalko, 1994).
Recen s udies o ligh , espi a ion, ene gy
expendi u e, and echoloca ion ha e indica ed
ha while p oduc ion o echoloca ion calls is
ex emely cos ly in es ing ba s, he e is li le
addi ional cos o echoloca ion in lying ba s
(a leas o sea ch-phase calls; Kalko, 1994)
because he same muscles ha lap he wings
also en ila e he lungs and p oduce he puls-
es o b ea h used o gene a e echoloca ion
calls (Speakman e al., 1989; Rayne , 1991a,
1991b; Speakman and Racey, 1991; Speak-
man, 1993). A c i ical link in his sys em is
m. se a us an e io , which ies he ibcage o
he ligh mechanism. The inc eased a ach-
men a ea o his muscle complex p o ided
by an e io and pos e io laminae on he
ibs—and concomi an inc ease in s i ness
o he ibcage ha we hypo hesize is c ea ed
by hese laminae—may simul aneously in-
c ease e iciency o he ligh mechanism,
en ila ion, and echoloca ion sys em. M. se -
a us an e io plays an impo an ole in ligh
du ing he downs oke and ea ly s ages o he
ups oke. A la ge a ea o o igin may p o ide
o imp o ed pe o mance o his muscle
complex in ligh and may also acili a e en-
ila ion o he lungs bo h by inc easing he
a ea o he connec ion be ween he ligh ap-
pa a us and ibcage. Inc eased s i ness o he
ibcage caused by p esence o ib laminae
could also acilia e exhaling and p oduc ion
o echoloca ion calls by inc easing he o ce
o elas ic ecoil o he ibcage upon elaxa-
ion o m. se a us an e io . We no e ha he
echoloca ing ba species s udied by Speak-
man and his colleagues (Phyllos omus has-
a us,Pleco us au i us,Pipis ellus pipis el-
lus, and Pleco us au i us) all ha e some de-
g ee o de elopmen o ib laminae. I would
be in e es ing o de e mine i he ew mic o-
chi op e ans ha lack ib laminae (e.g., Mo-
lossus molossus) a e as ene gy-e icien
while lying and echoloca ing as a e axa
wi h ib laminae. I ib laminae ha e been
seconda ily los in hese o ms (as we in e
om ou phylogeny), we would expec ha
any loss in e iciency ela ed o ibcage s uc-
u e would ha e been compensa ed o by
o he mechanical changes in he ligh , en-
ila ion, and echoloca ion sys ems.
Ano he change in he axial skele on in-
ol ed e olu ion o a la e ally comp essed
en al p ocess on he manub ium, which ap-
pea ed p io o he di e gence o Hassianyc-
e is. Changes in his s uc u e, which o ms
a keel when la e ally comp essed, a e likely
o be unc ionally ela ed o he ligh mus-
cles ha o igina e om he manub ium, spe-
ci ically m. pec o alis. M. pec o alis p o ides
mos o he powe o he downs oke o he
wings (Vaughan, 1959, 1970b; S ickle ,
1978). Mo e an e io ly loca ed ibe s in his
complex (e.g., hose o igina ing om he ma-
nub ium a he han om he body o he
s e num) d aw he hume us downwa d and
sha ply o wa d, whe eas he pos e io ibe s
pull he hume us downwa d and backwa d
(Vaughan, 1959). Modi ica ion o he en al
p ocess o he manub ium may e lec
changes in he ela i e size, momen a m,
and unc ional impo ance o he an e io
po ion o m. pec o alis.
Mo ing o he hindlimbs, we ind he i s
case in which known modi ica ions o an an-
a omical egion seemingly e ol ed in a sin-
gle segmen o he ee a he han in a s ep-
wise manne . In his ins ance, all changes oc-
cu ed a e he di e gence o A chaeonyc-
e is and p io o he di e gence o
Hassianyc e is. Two ans o ma ions oc-
cu ed a his le el: educ ion o he ibula o
a hin, h eadlike elemen , and e olu ion o
a calca . The unc ional implica ions o he
o me a e no clea . Se e al au ho s ha e
no ed ibula educ ion in ex an mic ochi-
op e ans, and ha e implied ha educ ion o
he ibula is somehow associa ed wi h lack
o a need o a obus ibula in ba s ha ha-
bi ually hang unde b anches and do no use
ypical quad upedal locomo ion (Vaughan,
1959, 1970a; Wal on and Wal on, 1968,
1970; Howell and Pylka, 1977; Hill and
Smi h, 1984). This sugges s ha he ibula
may ha e been educed by de aul when i
was no longe needed o suppo comp essi e
o ces associa ed wi h quad upedial loco-
mo ion. Howe e , he co ela ion be ween
hanging beha io and a hin ibula is no pe -
1998 113SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
chi op e ans, in many species i is compa a-
ble o ha o noc u nal mu oid oden s (e.g.,
Ra us,Pe omyscus) and ma supials (e.g.,
Didelphis; Su he s, 1966; Rahmann, 1967;
Su he s e al., 1969; Chase, 1972; Manske
and Schmid , 1976; Bell, 1982b; Fen on,
1985; Bell and Fen on, 1986; Pe ig ew,
1988; Pe ig ew e al., 1988). In bo h Mic o-
chi op e a and Megachi op e a, isual acui y
de e io a es wi h dec easing ligh le els mo e
slowly han in humans, so ba s can usually
see be e han humans in dim ligh ; h esh-
olds a e simila o hose epo ed o owls
(Neuweile , 1967; Su he s, 1970; Manske
and Schmid , 1976; Fen on, 1985; Bell and
Fen on, 1986). We su mise ha he ea lies
ba s had a leas he same isual capabili ies
as noc u nal oden s and ma supials, and may
ha e been much be e equipped (i.e., i hey
possessed he de i ed isual sys ems seen in
p ima es, de mop e ans, and megachi op e -
ans; o a summa y see Pe ig ew e al.,
1989). Reliance on he isual sys em alone
o o ien a ion p obably p ecluded ae ial in-
sec i o y and ligh wi hin clu e ed en i on-
men s (e.g., wi hin dense ege a ion and
many subcanopy habi a s). Ne e heless,
hese ea ly ba s would ha e bene i ed om
many o he ad an ages o powe ed ligh —
apid, ene gy-e icien anspo a ion o e
ela i ely long dis ances and an inc eased
o aging adius (Smi h, 1977; No be g,
1986a, 1989, 1994; Rayne , 1986; Scholey,
1986; No be g and Rayne , 1987; Thomas,
1987), as well as easy a oidance o p eda o s
(Pome oy, 1990).
This hypo hesis aises he linked ques ions
o how and why one ully unc ional o ien-
a ion sys em ( ision-based o ien a ion)
would be exchanged du ing e olu ion o an-
o he (echoloca ion). While such a ans o -
ma ion is admi edly unlikely i one pos u-
la es a simple one-s ep p ocess, i makes
conside ably mo e sense when iewed as a
mul is ep p ocess ha allowed he mic o-
chi op e an lineage o success ully in ade an
eno mous ye emp y se o ecological niches
o noc u nal ae ial insec i o es. In a a ely
ci ed pape on he e olu ion o eeding s a -
egies in ba s, Gille e (1975) a gued ha ma-
jo adap i e shi s migh ha e e ol ed
h ough wha he e med ‘‘duali y’’—a pe iod
du ing which bo h he p imi i e and de i ed
beha io s we e p esen simul aneously. In
he case o a ans o ma ion om isual o i-
en a ion o echoloca ion, we sugges ha he
ea ly mic ochi op e an lineage passed h ough
a pe iod du ing which hey o ien ed p inci-
pally using ision, bu also used echoloca ion
o p o ide supplemen a y in o ma ion abou
hei su oundings. This possibili y was i s
sugges ed by Pe ig ew (1988: 649), al hough
in he con ex o ba diphyly:
I . . . hink i unlikely ha sona would ha e been
su icien ly de eloped in he i s mic oba o enable
much use ul guidance owa d an ai bo ne insec . The
sho ange o ul asound pulses makes i unlikely ha
i was an echo om an insec ha i s en iced ou
ea ly mic oba o i s b anch, as does he equi ed
neu al p ocessing which seems unlikely o ha e been
su icien ly sophis ica ed on he i s y. Vision seems
a mo e likely candida e o p o ide he app op ia e
esolu ion and ange, jus as i does in many li ing
mic oba s oday. The p ima y ole o sona would
hen ha e been in he de ec ion o obs acles, pa ic-
ula ly in he ca e oos whe e he e is no al e na i e
senso y channel and whe e h ee o he lying e e-
b a es ha e sough shel e , independen ly in en ing
sona on each occasion (swi le s, oilbi ds and ou-
se e megaba s). I his is co ec , hen he use o so-
na o insec cap u e occu ed as a modi ica ion o
he a oidance sys em a e ligh was achie ed.
Indeed, adequa e ision may ha e been a
necessa y p e equisi e o he e olu ion o
echoloca ion in ba s. The senso y ange o
echoloca ion is ela i ely sho (e.g., Su he s,
1970, 1978; G i in, 1971; Fen on, 1980,
1982a, 1984, 1994a; Kick, 1982; Law ence
and Simmons, 1982), and quickly mo ing,
lying o ganisms may equi e mo e in o -
ma ion abou dis an obs acles han can be
ob ained om echoloca ion alone (Su he s
and Wallis, 1970; Chase, 1972; Fen on,
1985; Pe ig ew e al., 1988). This may be
especially ue in he wa m, mois ai o he
opics, whe e a mosphe ic a enua ion o ul-
asonic equencies is g ea es (G i in,
1971; Su he s, 1978; Law ence and Sim-
mons, 1982). Pa icula ly when o ien ing
o e long dis ances (e.g., when commu ing
o and om o aging a eas), ision appea s
o be c ucial o obs acle a oidance and land-
ma k ecogni ion (Da is, 1966; Muelle ,
1966, 1968; Williams e al., 1966; Williams
and Williams, 1967, 1970; Su he s, 1970).
We ind i ha d o imagine how echoloca-
ion—which is essen ially a sho - ange sen-
so y sys em—could ha e e ol ed in a g oup
114 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
o lying o ganisms in he absence o ision
o longe - ange obs acle de ec ion.
Speakman (1993) a gued ha an ea ly ba
which elied on a combina ion o ision and
echoloca ion would be a ‘‘senso y gene al-
is ’’ ha migh be a a selec i e disad an age
compa ed wi h senso y specialis s ha used
ei he ision o echoloca ion. This hypo hesis
was o e ed as a possible explana ion o he
absence o echoloca ion in Megachi op e a,
bu migh also apply o he ea lies mic o-
chi op e ans. Howe e , we do no ind his
hypo hesis compelling, a leas as an a gu-
men agains a ans o ma ion om ision o
echoloca ion as he p ima y senso y sys em
in ea ly Mic ochi op e a. Ea ly membe s o
he mic ochi op e an lineage would ha e
aced no compe i ion om senso y special-
is s in echoloca ion, because specialized ae-
ial echoloca o s— hei descendan s—had
no ye e ol ed. Ra he han being a a se-
lec i e disad an age when compa ed wi h i-
sion specialis s (i.e., megachi op e ans), we
ha e e e y eason o belie e ha ea ly mi-
c ochi op e ans would ha e immedia ely
gained se e al ad an ages when hey i s be-
gan o e ol e echoloca ion, including an in-
c eased abili y o de ec obs acles a sho
ange (pa icula ly in low-ligh si ua ions)
and an inc eased abili y o u ilize ca es as
oos ing places.
P e ious au ho s ha e sugges ed ha he
p imi i e echoloca ion sys em wi hin Mic o-
chi op e a may ha e consis ed o sho
b oadband o mul iha monic clicks p oduced
in he la ynx (Pye, 1980; Simmons, 1980;
Simmons and S ein, 1980; Fen on, 1984;
No be g, 1994; Fen on e al., 1995; A i a and
Fen on, 1997). Such a sys em p obably
e ol ed om ocaliza ions used o o he
pu poses, mos likely in aspeci ic commu-
nica ion (Fen on, 1984, 1985; No acek,
1985a). This hypo hesis is suppo ed by ob-
se a ions ha some ex an mic ochi op e -
ans use echoloca ion calls simul aneously o
o ien a ion and communica ion, while o he s
ha e dis inc calls ha a e appa en ly used
only in a social con ex (Mo¨h es, 1967a,
1967b; Habe se ze , 1981; Mille and Degan,
1981; B own e al., 1983; Fen on, 1984,
1985, 1994b; Guppy e al., 1985). Megachi-
op e an ba s use sho b oadband o mul i-
ha monic FM calls in a a ie y o social si -
ua ions, including emale con ac , p ecopu-
la ion, and hos ile male–male in e ac ions
(Nelson, 1964; Fen on, 1985). Simila calls
a e some imes used by mic ochi op e an
ba s, o en in compa able social si ua ions
(e.g., hos ile in e ac ions be ween males;
Po e , 1979; Fen on, 1985). Young ba s o
many species (bo h megachi op e ans and
mic ochi op e ans) appa en ly use simila
calls when hey a e isola ed om hei mo h-
e s (Fen on, 1985). Toge he , hese obse a-
ions sugges ha b oadband o mul iha -
monic clicks o buzzes we e p esen in he
ocal epe oi e o he ea lies ba s, and hus
we e a ailable as a beha io al subs a e om
which echoloca ion calls could e ol e.
Ea ly membe s o he mic ochi op e an
lineage, lying and using ision as well as a
p imi i e echoloca ion sys em such as de-
sc ibed abo e, would ha e had access o
many habi a s and ood sou ces bu would
ha e been poo ly equipped o ly in clu e ed
spaces o cap u e lying insec s (No be g,
1994). Pe ec ion o echoloca ion o de ec -
ing, acking, and assessing ai bo ne p ey
would ha e inc eased he o aging op ions
open o hese ba s, especially gi en he ab-
sence o compe i ion (Fen on, 1974a, 1974b,
1980, 1982a, 1984, 1994a; No be g, 1994;
Speakman, 1995). Noc u nal lying insec s
o e an abundan ood supply, so much so
ha he e is li le e idence ha p ey abun-
dance is a limi ing esou ce o mos mic o-
chi op e an ae ial insec i o es,
11
a leas
11
Possible excep ions o his gene al pa e n may in-
clude ae ial insec i o es ha summe in high la i ude
habi a s, whe e he nigh s a e e y sho and peak ae ial
insec a ailabili y occu s be o e i ge s da k (Rydell,
1992; Speakman, 1995). I has ecen ly been shown ha
some ba s li ing unde hese condi ions equen ly ail
o mee hei ene gy equi emen s, ins ead using o po
o balance hei daily ene gy budge s (Kunz, 1980; Ku a
e al., 1987, 1989; Speakman and Racey, 1987; Aude
and Fen on, 1988). Speakman (1995) sugges ed ha i
is noc u nali y (and lack o adequa e p ey esou ces a
nigh ) ha o ces he use o o po in hese animals.
Al hough his may be ue, we sugges ha i is he a ail-
abili y o o po ha has pe mi ed hese species o ex-
pand hei summe anges beyond he limi o mos o he
ba s. In opical and sub opical en i onmen s (whe e
mos insec i o ous mic ochi op eans li e), nigh s a e
longe and p ey densi ies a e gene ally high, hus p ey
a ailabili y is p obably no a limi ing ac o unde no -
mal condi ions (e.g., Fen on e al., in p ess).
1998 115SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
hose in opical and sub opical habi a s
(Fen on, 1980; Fen on e al., in p ess). The
po en ial ad an ages o e ol ing adap a ions
o exploi noc u nal lying insec s as a ood
sou ce a e ob ious.
Fen on e al. (1995) no ed ha ans o -
ma ion o he sophis ica ed echoloca ion sys-
em used by ex an mic ochi op e ans o de-
ec , ack, and e alua e lying insec s e-
qui ed p oduc ion o longe , s onge (high-
in ensi y) onal signals and concomi an
e olu ion o a sys em o a oid sel -dea en-
ing. Among ex an ba s, wo such sys ems
a e known: sepa a ion o he pulse and echo
in ime ( he ‘‘low-du y-cycle’’ app oach),
and sepa a ion o he pulse and echo in e-
quency a he han ime ( he ‘‘high-du y-cy-
cle’’ app oach; Fen on, 1994a, 1995; Fen on
e al., 1995). Low-du y-cycle echoloca ion
in ol es sho signal pulses wi h ela i ely
long gaps be ween hem; high-du y-cycle
echoloca ion in ol es longe pulses and
sho e gaps, wi h pulses longe han he gaps
be ween hem (Fen on, 1994a, 1995; Fen on
e al., 1995). Low-du y-cycle ba s p e en
sel -dea ening by eezing mo emen o he
middle-ea ossicles h ough con ac ion o
he middle-ea muscles du ing pulse emis-
sion and by educing audi o y sensi i i y in
he inne ea h ough changes in he senso y
cells along he basila memb ane (Henson,
1964, 1965, 1966, 1967a, 1967b, 1970; Jen
and Suga, 1976). Re u ning echoes a e e-
cei ed in he gaps be ween pulse emissions,
when he middle-ea muscles elax and au-
di o y sensi i i y is maximized. High-du y-
cycle ba s p oduce long cons an equency
(CF) echoloca ion signals ha o e lap wi h
e u ning echoes. These ba s u ilize he
Dopple e ec , which shi s he equency o
e u ning echoes o a equency di e en
om ha o he o iginal pulses (Schni zle ,
1970b, 1973, 1987; Schulle e al., 1975;
Schulle and Pollack, 1976; Schulle , 1977;
Neuweile e al., 1980; Schni zle and Hen-
son, 1980; Simmons and S ein, 1980; Emde
and Schni zle , 1986, 1990; Neuweile , 1989,
1990; G innell, 1995). Sel -dea ening is e-
duced because he emi ed pulse is domina -
ed by equencies ou side he acous ic o ea
(zone o maximum hea ing sensi i i y),
while bo h he ex e nal and inne ea s a e
sha ply uned o he equencies o he e-
u ning echoes (Schulle and Pollack, 1976;
B uns, 1979; Neuweile e al., 1980; Schni z-
le and Henson, 1980; Va e e al., 1985;
Emde and Schni zle , 1986, 1990; Schni zle ,
1987; Va e , 1987; Neuweile , 1990; Ob is
e al., 1993; Fen on, 1994a; Fen on e al.,
1995).
Mos ex an mic ochi op e ans a e low-
du y-cycle echoloca o s; high-du y-cycle
echoloca ion is used only by hinolophines,
hipposide ines, and one mo moopid (P e -
ono us pa nellii; No ick, 1958a, 1962,
1963a, 1963b, 1977; No ick and Vaisnys,
1964; Schni zle , 1970b, 1973, 1987; Fen on,
1974a, 1974b, 1980, 1982a, 1982b, 1984,
1990, 1994a, 1995; Schulle e al., 1975;
Schulle and Pollack, 1976; Schulle , 1977;
Neuweile e al., 1980; Schni zle and Hen-
son, 1980; Simmons and S ein, 1980; Fen on
and Bell, 1981; Neuweile , 1984, 1989,
1990; Neuwiele and Fen on, 1988; Kalko
and Schni zle , 1989, 1993; Lancas e e al.,
1992; Su lykke e al., 1993; Fen on e al.,
1995; G innell, 1995). I seems clea ha
low-du y-cycle echoloca ion is p imi i e o
ex an Mic ochi op e a based on op imiza-
ion o echoloca ion s a egies on ou phy-
logene ic ee ( ig. 36). Op imiza ion indi-
ca es ha he high-du y-cycle app oach
e ol ed wice, once in he lineage leading o
Rhinolophidae and once wi hin he genus
P e ono us. This con i ms p e ious hypo h-
eses (e.g., Pye, 1980; Simmons, 1980; Sim-
mons and S ein, 1980; Fen on e al., 1995)
ha sugges ed independen o igins o he
use o long CF signals and Dopple compen-
sa ion in hese wo g oups.
The ans o ma ion om p imi i e o so-
phis ica ed low-du y-cycle la yngeal echolo-
ca ion likely ook place in se e al s ages
(Fen on, 1984), p obably acili a ed by he
mechanical coupling o ligh and en ila ion
discussed p e iously (Speakman e al., 1989;
Rayne , 1991b; Speakman and Racey, 1991;
Speakman, 1993). I he e was indeed ‘‘no
cos o echoloca ion o ba s in ligh ’’
(Speakman and Racey, 1991: 421), o (mo e
ealis ically) a ela i ely low cos , i is easy
o imagine how his sys em migh ha e
e ol ed quickly. Howe e , e inemen o his
sys em—and inc eased eliance on ae ial in-
sec i o y—appa en ly b ough wi h i an im-
116 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
po an se ies o e olu iona y cons ain s, in-
cluding limi a ions on body size and ep o-
duc ion (Ba clay and B igham, 1991; Ba -
clay, 1994, 1995; Jones, 1994; A i a and
Fen on, 1997).
Maximum body size may be limi ed in mi-
c ochi op e an ae ial insec i o es o he se -
e al easons, including (1) he small size o
a ailable p ey (McNab, 1969; Black, 1974);
(2) he me abolic equi emen s o sus ained
ligh , pa icula ly a nigh when he mal
gliding and soa ing a e no possible (Rayne ,
1981); (3) he mechanics o ligh and ae ial
p ey cap u e (No be g, 1986a, 1994; No be g
and Rayne , 1987); (4) he e ec i e ange o
echoloca ion calls ( ela i ely sho ) and he
concomi an need o be maneu e able and
agile enough o ca ch small p ey a sho
ange (Ba clay and B igham, 1991); and (5)
he coupling o ligh and echoloca ion,
which may limi he abili y o la ge ba s
(which ha e lowe wing-bea equencies and
hus lowe call- epe i ion a es) o de ec
small lying p ey (Jones, 1994, al hough see
Helle , 1995). As a p obable esul o hese
cons ain s, mic ochi op e an ae ial insec i-
o es a e ypically e y small (
,
30 g adul
body weigh ), wi h only ou ex an species
weighing mo e han 100 g
12
(McNab, 1969;
Black, 1974; Fen on and Fleming, 1976;
K azanowski, 1977; Ba clay and B igham,
1991; A i a and Fen on, 1997). Wi hin his
size ange, mos la ge ae ial insec i o es
seem o be limi ed o ela i ely la ge insec
p ey (e.g., mo hs and la ge bee les; Vaughan,
1977), while smalle ba s can exploi ei he
la ge o small p ey (Ba clay and B igham,
1991; A i a and Fen on, 1977).
Re u ning o he p oblem o explaining he
e olu ion o sophis ica ed echoloca ion—
speci ically ou hypo hesis ha his o m o
o ien a ion eplaced a ision-based sys em in
mic ochi op e an ba s—we no e ha small
body size may ha e p ecluded he e en ion
o de i ed s uc u es o he isual sys em
(such as hose seen in P ima es, De mop e a,
and Megachi op e a) once neu al modi ica-
ions associa ed wi h echoloca ion began o
e ol e. In a e iew a icle on encephaliza ion
12
Taphozous peli (Emballonu idae), Hipposide os
comme soni (Rhinolophidae), Sco ophilus nig i a (Ves-
pe ilionidae), and Chei omeles o qua us (Molossidae).
in e eb a es, Ha ey and K ebs (1990: 145)
no ed ha
I o e all size [o he b ain] is cons ained . . . i may
well be ha specialized enla gemen o one egion
has o be associa ed wi h educ ion in size o ano he
....Ino he wo ds, he e may be ade-o s in he
e olu iona y specializa ion o he b ain.
Coope e al. (1993a: 340) simila ly conclud-
ed ha
. . . non- isual sys ems may ‘‘compe e’’ wi h neu ons
o he isual sys em o a ailable me aboli es ....
Al hough he use o he exp ession compe i ion o
desc ibe hese e olu iona y e en s is somewha unex-
plici , he ela i e expansion o one ce eb al s uc u e,
es ained wi hin he con ines o a b aincase o lim-
i ed olume, may depend upon concomi an dec ease
elsewhe e, no wi hs anding he p oblem o main ain-
ing he mos e icien and adap i e neu onal popula-
ions.
Small body size places limi a ions on he size
o neu on popula ions and may cons ain he
me abolic ene gy a ailable o b ain unc-
ions (Ma in, 1981; A ms ong, 1983; Wil-
liams and He up, 1988; Deacon, 1990a,
1990b; Coope e al., 1993a). Neu ons ha e
high ene gy equi emen s; he b ain can con-
sume up o 20% o ci cula ing oxygen and
glucose e en hough he b ain cons i u es
only a small ac ion o o al body weigh
(Ke y and Schmid , 1948; Ma in, 1981;
A ms ong, 1983; Williams and He up,
1988; Coope e al., 1993a). Re iewing e-
la ionships be ween body size, b ain size,
and me abolic a es in small mammals, Coo-
pe e al. (1993a: 339) summa ized hei ind-
ings as ollows:
These a gumen s emphasize he conclusion ha in
small animals, such as moles and ba s, he ac ion o
me abolism de o ed o neu ons is g ea and bioene -
ge ic limi s become c i ical. Supplying he b ain wi h
su icien oxygen is a challenge o he body o su -
i al. Supe luous neu ons . . . a e hus s ongly se-
lec ed agains and hei educ ion can con ibu e o
he animal’s i ness by imp o ing me abolic e iciency
(Rickle s and Ma ks, 1984).
In he ea ly e olu ion o mic ochi op e an
ba s, i may no ha e been e icien o ea-
sible o e ain de i ed complex s uc u es in
he isual sys em (e.g., lamina ed do sal la -
e al genicula e nucleus, la ge supe io collic-
ulus; Sande son, 1986; Pe ig ew e al., 1989)
and simul aneously p o ide he neu ons and
me abolic ene gy necessa y o p ocessing
inc easingly complex audi o y in o ma ion
1998 117SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
(e.g., in a la ge in e io colliculus; Ai kin,
1986; Casseday and Pollak, 1988; Pollack
and Casseday, 1988; Pollak and Pa k, 1995).
F om a dual o ien a ion sys em employing
bo h ision and limi ed echoloca ion, we su -
mise ha he audi o y sys em o mic ochi-
op e an ba s simply ‘‘ou compe ed’’ he i-
sual sys em o he esou ces a ailable.
E olu iona y educ ion in b ain cen e s e-
sponsible o p ocessing isual in o ma ion
could ha e occu ed h ough a a ie y o on-
ogene ic changes, including unca ion o de-
elopmen o isual cen e s a inc easingly
ea ly s ages. This hypo hesis is suppo ed by
obse ed simila i ies be ween he ela i ely
simple adul do sal la e al genicula e nucleus
(dLGN) in mic ochi op e an ba s (Pe ig ew
e al., 1989) and ea ly de elopmen al s ages
o he dLGN in Tupaia (Casag ande and
B unso-Bech old, 1985). I de elopmen o
he dLGN in Tupaia was a es ed a an ea ly
s age, he adul dLGN would esemble ha o
a mic ochi op e an ba —and eigh o he de-
i ed ai s de ined by Pe ig ew e al. (1989:
512) would be e e sed as he esul o a sin-
gle on ogene ic change. The same would p e-
sumably hold ue o he dLGN o megachi-
op e an ba s, which is simila o ha o Tu-
paia in mos espec s (Pe ig ew e al., 1989).
Ano he mechanism ha migh con ibu e
o educ ion in isual cen e s in he b ain is
changes in he magni ude o neu on dea h
du ing ea ly de elopmen . Neu on dea h is a
no mal pa o b ain de elopmen in many
e eb a es, including all mammals (Finlay e
al., 1987; Williams and He up, 1988). No
su p isingly, he magni ude o neu on dea h
wi hin homologous neu on popula ions a -
ies phylogene ically (Finlay e al., 1987; Wil-
liams and He up, 1988). Fo example, no -
mal de elopmen includes no dea h o e inal
ganglion cells in ish and amphibians (Wil-
son, 1971; Eas e e al., 1981), bu 40% o
e inal ganglion cells no mally die in chick-
ens (Rage , 1980), 60–70% die in a s and
p ima es (Rakic and Riley, 1983; C espo e
al., 1985; P o is e al., 1985), and 80% die
in ca s (Williams e al., 1986). Neu on dea h
is known o be a no mal pa o de elopmen
o he dLGN in p ima es (Williams and Rak-
ic, 1988) and has been shown o change he
p opo ion o con ala e al (c ossed) e sus
ipsila e al (unc ossed) e inal p ojec ions in
ca s (Jacobs e al., 1984). E olu iona y
changes in axon-speci ic a es o neu on
dea h in componen s o he isual sys em
du ing de elopmen may hus a ec he size
and in e connec ions o isual cen e s in
adul s. As audi o y in o ma ion became in-
c easingly impo an o ea ly membe s o he
mic ochi op e an lineage, we suspec ha
b ain cen e s in ol ed in p ocessing isual
in o ma ion we e educed and/o simpli ied
h ough de elopmen al changes such as hese.
The hypo hesized e olu iona y educ ion
o he isual sys em in mic ochi op e an ba s
is no a unique e en ; simila ans o ma ions
a e pos ula ed o ha e occu ed independen -
ly in o he mammalian g oups ha depend
li le on ision, including moles (Johnson,
1954; Lund and Lund, 1965, 1966; Su he s
and B ad o d, 1980; Kudo e al., 1988,
1991), mole a s (B onch i e al., 1991; Coo-
pe e al., 1993a, 1993b), and odon oce e
whales (Jacobs e al., 1975; Sande son, 1986;
Deacon, 1990b). In his con ex , i is in e -
es ing o no e ha odon oce es a e he only
o he g oup o mammals known o use so-
phis ica ed echoloca ion compa able o ha
o mic ochi op e ans (e.g., No is e al.,
1961; No is, 1968; Pu es and Pille i, 1983;
Pille i, 1983, 1990). Ex ensi e e e sals in
he isual sys em a e also commonly accep -
ed as explana ions o he mo phology o
blind snakes, gymnophionan amphibians,
and blind ca e popula ions o ishes (e.g.,
Wilkens, 1971; Halpe n, 1973; Repe´ an e
al., 1987; Voneida and Sliga , 1976; Clai -
ambaul e al., 1980; F i zsch e al., 1985;
Hims ed and Man eu el, 1985). Simila
e olu iona y educ ions ha e aken place in
o he senso y sys ems in some mammals,
such as educ ion o he ol ac o y sys em in
whales (Oelschla¨ge and Buhl, 1985; Oelsch-
la¨ge , 1989) and he ome onasal sys em in
ca a hine p ima es (Me edi h, 1991). As no -
ed abo e, hese so s o changes appea o
ep esen modi ica ions o e icien use o
cellula and me abolic esou ces. When a
sys em is no longe o c i ical impo ance,
eg ession o educ ion—-which may be in-
e p e ed as e e sal in a phylogene ic con-
ex — akes place o a oid was ing cells and
me abolic ene gy ha may be be e spen
elsewhe e (Coope e al., 1993a).
The e olu ion o ligh and echoloca ion
118 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
equi ed modi ica ions o many ana omical
and beha io al sys ems, and his p ocess un-
doub edly ook place in se e al s ages. To
summa ize, we hypo hesize ha ligh
e ol ed i s , p io o he di e gence o Me-
gachi op e a and Mic ochi op e a. The i s
ba s mos likely used ision o o ien a ion
and obs acle de ec ion in hei a bo eal/ae ial
en i onmen . The e olu ion o ligh was la -
e ollowed by he o igin o low-du y-cycle
echoloca ion in basal membe s o he mic o-
chi op e an lineage. This sys em, which was
p obably de i ed om ocaliza ions o igi-
nally used o in aspeci ic communica ion,
was mos likely simple a i s , pe mi ing
o ien a ion and de ec ion o obs acles bu no
de ec ion o acking o ai bo ne p ey. How-
e e , due o he mechanical coupling o en-
ila ion and ligh , ene gy cos s o echoloca-
ion o lying ba s we e low, and he bene i s
o ae ial insec i o y quickly led o de elop-
men o a mo e sophis ica ed low-du y-cycle
echoloca ion sys em capable o de ec ing,
acking, and assessing ai bo ne p ey. The
need o an inc easingly de i ed audi o y
sys em, combined wi h limi s on body size
imposed by he mechanics o ligh , echolo-
ca ion, and p ey cap u e, may ha e esul ed
in educ ion and simpli ica ion o he isual
sys em as echoloca ion became inc easingly
impo an .
The heo y p esen ed abo e is consis en
wi h (and is indeed based on) ba monophyly
and he phylogeny p oposed in he cu en
s udy. When conside ed in a phylogene ic
con ex , he Eocene ossil ba s Ica onyc e is,
A chaeonyc e is,Hassianyc e is, and Pa-
laeochi op e yx a e no in o ma i e abou he
o igin o ligh (which p eceded di e si ica-
ion o he en i e mic ochi op e an lineage),
bu hey do p o ide some in o ma ion con-
ce ning ea ly s eps in he acquisi ion o so-
phis ica ed echoloca ion. As no ed abo e,
some o he ea lies mo phological ans o -
ma ions in he lineage leading o ex an Mi-
c ochi op e a ( hose ha ook place be o e
he di e gence o Ica onyc e is) included
changes in ea u es associa ed wi h he p o-
duc ion o echoloca ion calls (enla gemen o
he c anial ip o he s ylohyal), ansmission
o sounds h ough he middle ea (enla ge-
men o he o bicula apophysis), and some
ine- uning o he inne ea (ini ial enla ge-
men o he basal u n o he cochlea). Loos-
ening o he a achmen o he pe io ic o he
basic anium e ol ed a his le el o jus sub-
sequen ly (p io o he di e gence o A -
chaeonyc e is). I is in e es ing o no e ha
all o hese modi ica ions occu ed be o e
achie emen o wha we desc ibed as ‘‘g ea
enla gemen o he cochlea’’ (i.e., o he size
ange o mos ex an mic ochi op e ans),
which e ol ed somewha la e , jus p io o
he di e gence o Hassianyc e is ( ig. 40).
Pos c anial modi ica ions ha may ha e in-
c eased e iciency o he en ila ion sys em
(e.g., inc eased b ead h o pos e io laminae
on he ibs) e ol ed in he mic ochi op e an
lineage some ime be ween he di e gence o
Ica onyc e is and he di e gence o Palaeo-
chi op e x. O he de i ed ea u es po en ially
ela ed o he echoloca ion sys em also
e ol ed a some poin in he ea ly mic ochi-
op e an lineage. These include a phane o-
cochlea cochlea (p obably associa ed in
some ashion wi h cochlea expansion) and a
deep, cons ic ed ossa o m. s apedius (pe -
haps ela ed o he sys em o a oidance o
sel -dea ening). The o me was minimally
p esen be o e he di e gence o Hassianyc-
e is, he la e be o e he di e gence o Pa-
laeochi op e yx. As no ed abo e, missing
da a o cochlea ossi ica ion in Ica onyc e is
and A chaeonyc e is and o s apedial ossa
o m in A cha onyc e is and Hassianyc e is
make i impossible o exac ly place hese
ans o ma ions in he phylogene ic ee.
E en gi en unce ain y abou he ela i e
iming o some ans o ma ions, ou obse -
a ions con i m he hypo hesis ha sophis i-
ca ed echoloca ion e ol ed in a s epwise
ashion in he ea ly mic ochi op e an lineage.
Howe e , e olu iona y changes in his sys-
em ce ainly did no cease wi h he o igin o
he mic ochi op e an c own g oup. Consid-
e able cochlea size a ia ion exis s among
li ing clades o Mic ochi op e a, wi h mos
ex an mic ochi op e ans (and all high-du y-
cycle echoloca o s) ha ing a cochlea e en
la ge han hose seen in Palaeochi op e yx
and Hassianyc e is (Habse s eze and S o ch,
1992). Reduc ion in cochlea size (back o
he ‘‘mode a ely enla ged’’ condi ion) has
appa en ly occu ed in some lineages—in
Mys acinidae, wi hin Phyllos omidae, and
wi hin Megade ma idae (Fig. 40). Signi ican
1998 119SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
changes (some associa ed wi h cochlea en-
la gmen ) ha e also occu ed in some lin-
eages wi h espec o (1) leng h, wid h, and
hickness o he basila memb ane, (2) leng h
and heigh o he spi al ligamen , (3) o al
numbe o cochlea neu ons, (4) size and dis-
ibu ions o hai cell popula ions, (5) size
and pa e n o luid spaces in he cochlea, and
(6) o m o he equency map and ‘‘acous ic
o ea’’ (Pye, 1966a, 1966b, 1967; Henson,
1970; B uns, 1979; B uns e al., 1981, 1983–
1984; Bu da and Ulehlo a, 1983; Ru¨bsamen
e al., 1988; Neuweile , 1990). Mos o all o
hese changes may e lec modi ica ions o
he audi o y sys em associa ed wi h di e en
echoloca ion and o aging s a egies (B uns
e al., 1981, 1983–1984; Ru¨bsamen e al.,
1988; Neuweile , 1990).
FORAGING ECOLOGY OF EOCENE
BATS
One aspec o mo phology and ecology
no men ioned in he p eceding discussion is
he ela ionship be ween wing shape and o -
aging ecology. The e is an ex ensi e li e a-
u e on his opic (e.g., Re illiod, 1916; Be z,
1958; Vaughan, 1959, 1966; S u hsake ,
1961; Ha man, 1963; Fa ney and Fleha y,
1969; Fen on, 1972; Findley e al., 1972;
Kopka, 1973; Lawlo , 1973; Pi lo , 1977;
Smi h and S a e , 1979; No be g, 1981,
1986a, 1986b, 1987, 1994; Findley and Wil-
son, 1982; Findley and Black, 1983; Al-
d idge, 1986; Baagøe, 1987), wi h many e-
cen con ibu ions in eg a ing da a on echo-
loca ion call s uc u e as well (e.g., Simmons
e al., 1979; Neuweile , 1984, 1989, 1990;
Habe se ze , 1986; Ald idge and Rau enbach,
1987; No be g and Rayne , 1987; Neuweile
and Fen on, 1988; No be g, 1989, 1990,
1994; Fen on, 1990; A i a and Fen on, 1997).
One ou come o his esea ch has been iden-
i ica ion o a se ies o ea u es o wing de-
sign ha a ec ligh pe o mance and a e
co ela ed wi h o aging s a egies. Among
he mos impo an measu es o hese a e
wing loading, aspec a io, and a a ie y o
wing ip indices designed o measu e he size
and shape o he dac ylopa agium (e.g., Fin-
dley e al., 1972; No be g and Rayne , 1987).
Wing loading (
5
body weigh /wing a ea)
p o ides a measu e o he ela i e size o he
wings and o he minimum weigh pe uni
a ea ha he wings mus suppo du ing ligh
(Findley e al., 1972; No be g and Rayne ,
1987; A i a and Fen on, 1997). Inc eased
wing loading equi es g ea e ligh speeds o
gene a e enough li o emain ai bo ne;
ligh speed is p opo ional o he squa e oo
o wing loading (Findley e al., 1972; No -
be g and Rayne , 1987). High wing loading
is ypical o as lie s; slow- lying, mo e ma-
neu e able animals usually ha e lowe wing
loading, wi h ei he lowe body weigh an-
d/o inc eased wing a ea (Findley e al.,
1972; No be g and Rayne , 1987).
Aspec a io (
5
wing span
2
/wing a ea;
No be g and Rayne , 1987) desc ibes he
o e all shape o he wings by quan i ying
hei leng h ela i e o hei cho d (Findley
e al., 1972; No be g and Rayne , 1987; A i a
and Fen on, 1997). Fo any gi en body
weigh and wing loading, high aspec a io
wings a e subjec o less d ag and hus acil-
i a e g ea e ligh speeds han do lowe as-
pec a io wings (Findley e al., 1972). How-
e e , high aspec a io wings also gene a e
less li han do low aspec a io wings, and
hus ba s wi h high aspec a io wings may
equi e g ea e wingbea equencies and ai -
speeds o emain ai bo ne (Findley e al.,
1972). Some ba s wi h e y high aspec a io
wings (e.g., Molossus,Eumops) canno gen-
e a e enough li o become ai bo ne om a
la su ace (pe sonal obs.), and hus usually
d op om hei ele a ed oos s o build up
enough speed and li o ini ia e ligh
(Vaughan, 1959). Low aspec a io wings
gene a e conside able d ag a highe speeds,
bu maximize li a low speeds (Findley e
al., 1972).
A a ie y o di e en measu es o he el-
a i e size and shape o he wing ip ha e
been p oposed, including he ip index (Fin-
dley e al., 1972), alpha angle (Smi h and
S a e , 1979), ip leng h a io (No be g and
Rayne , 1987), ip a ea a io (No be g and
Rayne , 1987), and ip shape index (No be g
and Rayne , 1987). All o hese quan i y a -
ious aspec s o he ela i e size and shape o
he dac ylopa agium ( ha po ion o he wing
dis al o digi V), which p o ides much o
he p opulsion gene a ed by he wing du ing
ligh (Findley e al., 1972).
The mos comp ehensi e s udy o wing
120 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
design and o aging s a egies in ba s was
ha o No be g and Rayne (1987), who
summa ized mo phome ic and beha io al
da a o mo e han 250 ba species. In addi-
ion o he a ios men ioned abo e, hey also
conside ed o al body weigh (mass), wing-
span, and wing a ea. They in e p e ed hese
da a in he con ex o mechanical and ae o-
dynamic models o di e en modes o ligh .
No be g and Rayne (1987: 337) summa-
ized hei esul s as ollows:
Some adap i e ends in ba wing mo phology a e
clea om his analysis. Insec i o es hun in a ange
o di e en ways, which a e e lec ed in hei mo -
phology. Ba s hawking high- lying insec s ha e small,
poin ed wings which gi e good agili y, high ligh
speeds and low cos o anspo . Ba s hun ing o
insec s among ege a ion, and pe haps gleaning, ha e
e y sho and ounded wing ips, and o en ela i ely
sho , b oad wings, gi ing good maneu e abili y a
low ligh speeds. Many insec i o ous species o age
by ‘ lyca ching’ (pe ching while seeking p ey) and
ha e somewha simila mo phology o gleane s. In-
sec i o ous species o aging in mo e open habi a s
usually ha e sligh ly longe wings, and hence lowe
cos o anspo . Pisci o es o age o e open s e ch-
es o wa e , and ha e e y long wings gi ing low
ligh powe and cos o anspo , and usually long,
ounded ips o con ol and s abili y in ligh . Ca -
ni o es mus ca y hea y loads, and hus ha e ela-
i ely la ge wing a eas; hei o aging s a egies con-
sis o pe ching, hun ing and gleaning, and wing
s uc u e is simila o ha o insec i o ous species
wi h simila beha io . Pe ching and ho e ing nec a -
i o es bo h ha e a ela i ely small wing a ea; his
su p ising esul may esul om en i onmen al p es-
su e o sho wingspan o om he ad an age o high
speed du ing commu ing ligh ; he la ge wing ips o
hese ba s a e aluable o li gene a ion in slow
ligh .
Habe se ze and S o ch (1987, 1989) and
Habe se ze e al. (1994) capi alized on he
obse ed ela ionships be ween wing o m
and ligh beha io in ex an ba s o econ-
s uc he possible habi s o Ica onyc e is,
A chaeonyc e is,Hassianyc e is, and Pa-
laeochi op e yx. Habe se ze and S o ch
(1987) es ima ed wing loading, aspec a io,
and ip index o a se ies o ossil specimens,
and compa ed hese alues wi h simila da a
om an ex an ba auna om opical India
(i.e., Habe se ze , 1986). They concluded ha
Ica onyc e is and A chaeonyc e is ha e a
ela i ely high wing loading and low aspec
a io (Habe se ze and S o ch, 1987, 1989;
Habe se ze e al., 1994). Based on compa -
isons wi h ex an o ms, hey ound ha Ica-
onyc e is and A chaeonyc e is ha e wings
ha esemble hose o la ge hinolophoids o
he mean o espe ilionids (Habe se ze and
S o ch, 1987, 1989). Conce ning he combi-
na ion o low aspec a io and high wing
loading, Habe se ze and S o ch (1989: 216)
no ed ha ‘‘ his pa ame e combina ion is
mos likely o unspecialized ligh cha ac-
e is ics as can be ound wi hin he majo i y
o he espe ilionids (e.g., Myo is myo is).’’
In con as , Palaeochi op e yx was ound
o ha e a somewha mo e specialized wing
mo phology (Habe se ze and S o ch, 1987,
1989; Habe se ze e al., 1994). Palaeochi-
op e yx upaiodon was cha ac e ized as ha -
ing e y low aspec a io and e y low wing
loading, while bo h alues we e somewha
highe in P.spegeli (Habe se ze and S o ch,
1987, 1989). On he basis o hese alues,
bo h axa we e ound o be e y simila o
ex an hinolophine and hipposide ine spe-
cies (Habe se ze and S o ch, 1987; Habe -
se ze e al., 1994). This simila i y led Ha-
be se ze and S o ch (1987, 1989) and Ha-
be se ze e al. (1994) o conclude ha Pa-
laeochi op e yx was cha ac e ized by slow,
highly maneu e able ligh close o he
g ound. In he case o P. upaiodon, ae ial
capabili ies may ha e included ligh close o
and e en wi hin oliage wi h po en ially
long-las ing ho e ing phases, as is seen oday
in Hipposide os bicolo , a species wi h a
compa able combina ion o aspec a io and
wing loading alues (Habe se ze and S o ch,
1987, 1989; Habe se ze e al., 1994). Wing
pa ame e s o Palaeochi op e yx spiegali
mo e closely esemble hose o Hipposide os
speo is, a o m ha hun s nea obs acles bu
always s ays in open ai space, ne e lying
amongs oliage (Habe se ze and S o ch,
1987, 1989; Habe se ze e al., 1994).
Habe se ze and S o ch (1987) ound ha
bo h species o Hassianyc e is a e cha ac e -
ized by a high aspec a io and high wing
loading. On he basis o compa isons wi h
Habe se ze ’s (1986) da a om Indian ba s,
Habe se ze and S o ch (1987, 1989) and Ha-
be se ze e al. (1994) obse ed ha Hassi-
anyc e is species a e e en mo e specialized
han high- lying opical espe ilionids and
hinopoma ids, ins ead showing a g ea e e-
semblance o ex an high- lying molossids
and emballonu ids. As a esul , hey con-
1998 121SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
cluded ha Hassianyc e is messelensis and
H.magna ‘‘can be conside ed o be dis inc ly
adap ed o ligh in ee spaces’’ (Habe se ze
e al., 1994: 238).
Habe se ze and S o ch (1987, 1989) and
Habe se ze e al. (1994) no ed ha all ou
Eocene gene a a e cha ac e ized by a sho
wing ip and a small dac ylopa agium ha
accoun s o only abou 37% o he en i e
wing a ea. Among ex an ba s, hey no ed
ha his mo phology is seen in hinolo-
phoids. Habe se ze e al. (1994) concluded
ha he Messel ba s had succeeded in occu-
pying a di e se se o ecological niches by
e ol ing a ia ions o he hinolophoid wing
ype, and sugges ed ha his wing ype (i.e,
low aspec a io wi h sho wing ip) ep e-
sen s he p imi i e wing o m o ba s.
To u he in es iga e he co ela ions de-
sc ibed abo e, Habe se ze and S o ch
(1989) used mul i a ia e me hods de eloped
by No be g and Rayne (1987) o emo e he
e ec s o size ( he i s p incipal componen ),
esul ing in a plo o no malized wing load-
ing (second p incipal componen ) e sus no -
malized aspec a io ( hi d p incipal compo-
nen ). Al hough hey used No be g and Ray-
ne ’s (1987) me hod, Habe se ze and S o ch
(1989) did no use hei da a, bu ins ead
compa ed he ossil o ms o Habe se ze ’s
(1986) e e ence auna (wi h known ligh
and o aging habi s) om opical India. Ha-
be se ze e al. (1989) in e p e ed he esul s
o his mul i a ia e analysis as suppo ing
hei p e ious conclusions abou he ligh
beha io o he ossil ba s (see abo e). How-
e e , Ica onyc e is,A chaeonyc e is igon-
odon, and Palaeochi op e yx upaiodon ap-
pea ed o ha e e y simila wing pa ame e s
based on he size-no malized analysis (a
leas in ou judgmen ). This was no explic-
i ly discussed by Habe se ze e al. (1989);
howe e , hey cau ioned ha hei compa i-
sons migh be biased by he ac ha hey
calcula ed body weigh o he ossil o ms
using a measu e o wingspan, which in o-
duced ci cula i y in o he analysis and migh
ha e esul ed in he obse ed clus e ing.
No be g (1989) eanalyzed wing mo phol-
ogy o he Eocene ossil o ms in he con ex
o he much la ge da a se collec ed by No -
be g and Rayne (1987). No be g (1989)
sugges ed ha he es ima es o body mass
calcula ed by Habe se ze and S o ch (1987)
may ha e been oo high o se e al easons
(e.g., use o wingspan leng h eg ession
equa ions, s aigh a he han cu ed wing-
ips in he ossil econs uc ions). Using e-
ised wing econs uc ions ( ig. 42) and e-
ised es ima es o body weigh based on e-
g ession equa ions o adius leng h agains
o al mass in ecen ba s, No be g (1989) e-
calcula ed wing loading and aspec a io o
each o he ossil species, and addi ionally
calcula ed wing ip leng h a io, wing ip a ea
a io, and wing ip shape index o each ax-
on. Compa isons wi h ecen ba s ( igs. 43,
44; da a om No be g and Rayne , 1987) led
No be g (1989: 204–205) o he ollowing
conclusions ega ding Ica onyc e is,A -
chaeonyc e is,Hassianyc e is, and Palaeo-
chi op e yx:
Based on my econs uc ions, he ancien ba s ex-
amined he e . . . had low aspec a io and high o
a e age wing loading . . . ( ha is, high o a e age
wing loadings in ela ion o body size), as compa ed
wi h ecen ba s. This indica es ha hey had expen-
si e and a e age o as ligh . Thei wing ips we e
ex emely sho and wi h ai ly la ge a ea, which a e
adap a ions o maneu e able ligh . Wing ip leng h
a ios T
1
(handwing leng h/a mwing leng h) a e all
be ween 0.93 and 1.09 ....T
1
,
1 is a he unusual
among mode n ba s.
The wing shape o he ossil species a e simila o
se e al ecen p e opodids and phyllos omids and
some hinolophids (Rhinolophus hipposide os and R.
e umequinum) and espe ilionids (Ba bas ella ba -
bas ellus,Lasiu us bo ealis,Rhogeesa umida and
some Myo is,Ep esicus and Pipis ellus species). Pa-
laeochi op e yx and Hassianyc e is we e also simila
o Rhinopoma ha dwickei ..., bo h in wing shape
and wi h hei ex emely sho wing ips. Rhinopom-
a ids a e ound mainly in dese s and s eppes. They
ha e been no ed o o age o insec s in open coun y
as well as in open spaces a ound ee canopies, and
hei ligh has been desc ibed as swi , as and un-
dula ing (al e na ing lu e s and glides)(see No be g
& Rayne 1987 o e e ences) . . . Hassianyc e is
messelensis and H.magna had highe wing loadings
and sligh ly highe aspec a io han he o he ossil
species, and we e hus as e lie s. They had hough
much lowe aspec a io han molossids and embal-
lonu ids. They may ha e been o aging in a he open
spaces, like hinopoma ids, o along ege a ion like
noc ules and lasiu ines. The o he ancien ba s in-
cluded he e ha e lowe aspec a io and lowe wing
loading. Because hei sho wings and low aspec
a io hese ancien ba s p obably o aged o li ed
among ege a ion (which was also sugges ed by Ha-
be se ze & S o ch 1987) and may ha e been pe ch
hun e s.
Pe ch hun ing—making sho ligh s ou o
122 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
Fig. 42. Recons uc ions o wing shapes o ossil ba s; ed awn om No be g (1989: ig. 3). No be g
(1989) modi ied econs uc ions published by Habse se ze and S o ch (1987) by adding cu ed wing ips
such as hose seen in ex an ba s, changing he angles be ween he phalanges o he ou h digi in
A chaeop e opus, and changing he angles be ween he hi d, ou h, and i h digi s in mos species.
cap u e p ey de ec ed om a ixed pe ch—is
widely ega ded as a beha io ha educes
he ene gy equi ed o success ul o aging
(No be g and Rayne , 1987; Fen en, 1990;
Fen on e al., 1990; No be g, 1994). Ex an
ba s ha o age nea o wi hin ege a ion
may use pe ches because hei sho , low as-
pec a io wings make ligh ela i ely ex-
pensi e (No be g and Rayne , 1987; No -
be g, 1994). Because ligh s o cap u e p ey
a e in equen and o sho du a ion, less en-
e gy is appa en ly spen han would be e-
qui ed by sus ained hawking (No be g and
Rayne , 1987; Fen on, 1990; Fen on e al.,
1990; No be g, 1994). Ano he possible ad-
an age o pe ch hun ing is ha i may
b oaden he a ailable p ey spec um by a-
cili a ing he cap u e and handling o la ge
1998 129SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
sec s while in slow ligh . Palaeochi op e yx
appa en ly specialized in p eying on small-
scaled insec s (i.e., small mo hs and caddis
lies), and i s echoloca ion calls we e mos
likely b oadband calls o mode a e o high
equency (30–90 kHz). The e is no e idence
ha any o he Messl ba s used high-du y-
cycle CF echoloca ion.
Wing mo phology o Ica onyc e is and
A chaeonyc e is sugges s ha hese axa
p obably o aged close o he g ound and
close o ege a ion, as did Palaeochi op e yx
(Habe se ze and S o ch, 1987, 1989, 1992;
No be g, 1989; and Habe se ze e al., 1992,
1994). Howe e , p esence o only mode a e
enla gemen o he cochlea— oge he wi h
absence o a calca —sugges s ha hese
o ms may ha e been pe ch hun e s ha spe-
cialized in gleaning hei p ey om su aces
a he han ca ching i on he wing. A chaeo-
nyc e is may ha e been a bee le specialis ;
we ha e no eco d o he p e e ed p ey o
Ica onyc e is. Echoloca ion calls in bo h axa
we e p obably sho (
#
2 msec) b oadband
FM signals o mode a e o high equency
(30–90 kHz) o sho , mul iha monic calls.
In ei he case, hese echoloca ion calls may
ha e been o low in ensi y, and may ha e
been ‘‘ u ned o ’’ a imes o acili a e lo-
ca ion o p ey by passi e means (e.g., lis en-
ing o p ey-gene a ed sounds o looking o
p ey mo emen s). I is unlikely ha Ica on-
yc e is and A chaeonyc e is used echoloca-
ion o de ec ion, acking, o e alua ion o
p ey. Echoloca ion was p oblably used only
o o ien a ion and obs acle de ec ion, while
p ey de ec ion and acking we e accom-
plished by passi e means.
EVOLUTION OF FORAGING
STRATEGIES:
A PHYLOGENETIC PERSPECTIVE
The phylogeny gene a ed in ou s udy ( ig.
36) p o ides a amewo k o in e p e ing he
mo phological da a and beha io al in e -
ences p esen ed abo e. Gi en he opology
o his ee, we sugges ha he ea lies mem-
be s o he mic ochi op e an lineage ( o ms
cu en ly unknown om ossils) p obably
used ision o o ien a ion and obs acle de-
ec ion in hei a bo eal/ae ial en i onmen ,
and p obably o aged by gleaning insec s and
pe haps some ui s and o he edible i ems
om oliage, ba k, and pe haps he g ound.
Loca ion o po en ial ood i ems was p oba-
bly accomplished by a combina ion o ision
and lis ening o p ey-gene a ed sounds. Like
A chaeop e opus,Ica onyc e is,A chaeo-
nyc e is, and Palaeochi op e yx, hese ba s
p obably had low aspec a io wings, mod-
e a e wing loading, and ela i ely la ge wing
ips, all o which sugges ha hey habi ually
lew close o he g ound and nea ege a ion.
Low-du y-cycle echoloca ion e ol ed sub-
sequen ly, p obably om communica ion
calls ha inciden ally p oduced in o ma i e
echoes. As we sugges ed ea lie , his sys em
was p obably simple a i s , pe mi ing only
o ien a ion and obs acle a oidance bu no
de ec ion, acking, o e alua ion o ai bo ne
p ey. Basic anial modi ica ions p esumed o
be associa ed wi h inc eased e icacy o
echoloca ion began p io o he di e gence o
Ica onyc e is om he mic ochi op e an lin-
eage. Howe e , wing mo phology, a mode -
a ely enla ged cochlea, and absence o a cal-
ca in his axon sugges o us ha Ica on-
yc e is was a pe ch-hun ing gleane a he
han a p eda o on ae ial insec s. This o -
aging s a egy would ha e had he ad an age
o being ela i ely ene gy-e icien while a
he same ime equi ing only mode a e au-
di o y da a-p ocessing capabili ies o suc-
cess ully so he in o ma ion om e u ning
echoes. Echoloca ion calls we e mos likely
sho (
#
2 msec) b oadband FM signals o
mul iha monic calls, p obably o ela i ely
low in ensi y. These calls may ha e been
‘‘ u ned o ’’ a imes o acili a e passi e
p ey localiza ion. Indeed, p ey de ec ion,
acking, and e alua ion we e p obably no
done wi h echoloca ion, bu a he by ision
o lis ening o p ey-gene a ed sounds. Pas-
si e acous ic cues may ha e been pa icula -
ly impo an i hese ba s we e s ic ly noc-
u nal. In essence, he only majo change in
o aging me hod a his le el would ha e
been he addi ion o echoloca ion as a ool
o o ien a ion and obs acle de ec ion. The
basic o aging s a egy—gleaning om a
pe ch—would ha e been he same as seen in
he nonecholoca ing chi op e an ances o s o
Ica onyc e is.
The de i ed mo phological ans o ma-
ions ha diagnose he node linking A chaeo-
130 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
nyc e is wi h Hassianyc e is,Palaeochi op-
e yx, and he mic ochi op e an c own g oup
(e.g., educ ion o he numbe o oo s on P3,
p esence o a en al accesso y p ocess on
C4, absence o an ossi ied hi d phalanx on
wing digi III) a e ela i ely small changes
ha do no indica e any majo shi s in o -
aging ecology. This pa e n sugges s o us
ha A chaeonyc e is e ained much he same
o aging s a egy as Ica onyc e is.A chaeo-
nyc e is and Ica onyc e is sha e a simila
wing mo phology, mode a ely enla ged
cochlea, and lack o a calca , ea u es we in-
e p e as indica ing ha hese ba s o aged
by gleaning p ey ha we e de ec ed om a
pe ch using passi e means a he han echo-
loca ion.
The nex node as one mo es up he ee—
ha which links Hassianyc e is wi h Palaeo-
chi op e yx and he mic ochi op e an c own
g oup—is associa ed wi h a much mo e ex-
ensi e sui e o mo phological changes (see
discussion abo e unde ‘‘Cha ac e T ans o -
ma ions a Basal Nodes’’). Mos no able
among hese a e: (1) a g ea ly enla ged coch-
lea, (2) a do sal a icula ace on he scapula
(al hough his may ha e e ol ed ea lie ), (3)
a la e ally comp essed en al p ocess on he
manub ium o he s e num, (4) inc eased de-
elopmen o ib laminae ( his may ha e
e ol ed somewha ea lie ), (5) a h eadlike
ibula, and (6) p esence o a calca . Taken
oge he , hese ea u es sugges ha a majo
shi in o aging s a egy occu ed in he mi-
c ochi op e an lineage jus p io o he di-
e gence o Hassianyc e is: he e olu ion o
ae ial hawking. This beha io in ol es using
echoloca ion o de ec , ack, and assess p ey,
and he use o he u opa agium (suppo ed
and con olled by a calca ) o cap u e p ey
on he wing. Changes in he pos c anial skel-
e on, pa icula ly he pec o al gi dle, sugges
ha some ‘‘ ine uning’’ o he ligh mech-
anism may ha e accompanied he beha io al
change o ae ial hawking. This seems ea-
sonable gi en he demands o ae ial o aging
beha io , pa icula ly i hese ba s occasion-
ally o aged in g oup si ua ions (e.g., when
caddis lies we e ha ching along he sho e o
Lake Messel) whe e complex cap u e and
e asi e maneu e s may ha e been equi ed.
In e es ingly, he mo phological and be-
ha io al changes associa ed wi h he e olu-
ion o ae ial hawking do no seem o ha e
been co ela ed wi h any app eciable modi-
ica ions in wing shape o wing loading
(based on he esul s o No be g,1989; see
igs. 43, 44). Simila i ies in aspec a io, wing
ip indices, and wing loading in Palaeochi-
op e yx,A chaeonyc e is,Ica onyc e is, and
A chaeop e opus lead us o conclude ha he
body size and wing o m o Palaeochi op-
e yx is plesiomo phic, and ha his mo -
phology was p esen in he mos ecen com-
mon ances o o Palaeochi op e yx and Has-
sianyc e is. Acco dingly, he body size and
wing mo phology seen in Hassianyc e is
(discussed below) a e au apomo phic ea-
u es.
Like Ica onyc e is and A chaeonyc e is,
Palaeochi op e yx had low aspec a io
wings, mode a e wing loading, and ela i ely
la ge wing ips, all o which indica e ha
hese ba s habi ually lew close o he g ound
and nea ege a ion. Howe e , he g ea ly en-
la ged cochlea and well de eloped calca
sugges ha Palaeochi op e yx was ully ca-
pable o ae ial hawking. This hypo hesis is
consis en wi h he analyses o s omach con-
en s and aphonomy o Palaeochi op e yx a
Lake Messel. Echoloca ion signals used by
Palaeochi op e yx we e p obably sho - o
mode a e-leng h b oadband FM calls.
As poin ed ou by Fen on (1990) and Fen-
on e al. (1990), many ex an ba s ha hun
om pe ches also hawk insec s in con inuous
ligh , wi h alloca ion o ime be ween hese
wo o aging s a egies e lec ing p ey a ail-
abili y. Gi en ha gleaning om a pe ch
p obably ep esen s he p imi i e o aging
s a egy o he mic ochi op e an lineage, we
suspec ha Palaeochi op e x may ha e used
a combina ion o pe ch hun ing (including
lyca ching) and slow ae ial hawking o cap-
u e i s p ey, much like mode n nyc e ids.
This would be consis en wi h cochlea size
in Palaeochi op e x (as es ima ed by Habe -
se ze and S o ch, 1992; ig. 29), which plac-
es his genus jus a he lowe end o he
ange o a ia ion seen among o ms ha a e
con inuous ae ial hawke s (e.g., mos es-
pe ilionids) and a he uppe end o he
ange o a ia ion in ex an pe ch-hun ing
o ms ha use low-du y-cycle echoloca ion
(e.g., nyc e ids and megade ma ids).
The ansi ion om gleaning s a iona y
1998 131SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
p ey de ec ed by passi e means ( ision o lis-
ening o p ey-gene a ed sounds) o ae ial
hawking using echoloca ion o de ec and
ack p ey mus ha e been a complex p ocess
ha in ol ed in e media e s eps. I seems un-
likely ha echoloca ion was used o de ec -
ing s a iona y p ey du ing his ansi ion, be-
cause de ec ion o ha d a ge s es ing on
ha d o i egula backg ounds is one o he
mo e di icul asks aced by echoloca o s
(Simmons e al., 1980; Fen on, 1990, 1994a,
1995). Ins ead, we hypo hesize ha ba s ha
egula ly o aged by gleaning s a iona y p ey
(de ec ed by passi e means) inc easingly
came o hun mo ing insec s, pe haps leaping
a e p ey ha had been s a led by mo e-
men s o a acks by he ba . A his poin ,
echoloca ion (which hey we e al eady using
o o ien a ion and obs acle a oidance) could
p o ide mo e in o ma ion abou p ey loca-
ion and mo emen han could be ob ained
by passi e means. To ully exploi his o m
o da a collec ion, he animals would ha e
had o inc ease signal s eng h in o de o
maximize ange and p o ide he necessa y
ime o ack and e alua e a ge s. Howe e ,
en i onmen al clu e (which p oduces many
dis ac ing echoes) would ha e p esen ed a
majo impedimen o he e olu ion o e ec-
i e lyca ching beha io . I he e o e seems
likely ha lyca ching using echoloca ion o
de ec and ack p ey p obably e ol ed in
ba s ha equen ed o es gaps and he edges
o o es s along lakes and i e s, places
whe e ege a ion (wi h po en ial pe ches) lies
adjacen o ela i ely clu e - ee open spac-
es. Once lyca ching om pe ches was well
es ablished, i is easy o imagine a p og es-
si e ansi ion o spending mo e ime on he
wing, ul ima ely leading o he e olu ion o
axa ha elied exclusi ely on o aging by
con inuous ae ial hawking. I is no clea
when ba s began o use he u opa agium o
p ey cap u e, bu e olu ion o he calca ap-
pa en ly p eceded o was coinciden wi h he
e olu ion o ae ial hawking.
Ou hypo hesis ha he e olu iona y an-
si ion om gleaning (using passi e p ey de-
ec ion) o ae ial hawking (using echoloca-
ion) ook place in habi a s associa ed wi h
o es gaps o o es edges along bodies o
wa e canno be es ed gi en he spa se ossil
eco d o ba s. Howe e , i is in e es ing o
no e ha all ou o he ossil ba s conside ed
he e (Ica onyc e is,A chaeonyc e is,Hassi-
anyc e is, and Palaeochi op e yx) ha e been
collec ed om lake deposi s ha a e p e-
sumed o ha e been su ound by sub opical
o opical o es s a he ime o deposi ion
(MacGini ie, 1969; G ande, 1980;
Schaa schmid , 1992). The excellen p ese -
a ion o many o he ba ossils ound a
Messel and Fossil Basin indica es ha hese
animals we e no anspo ed a a e dea h,
sugges ing ha hey may ha e o aged
a ound he edges and o e he lakes in which
hey we e ul ima ely p ese ed.
Species o Hassianyc e is a e la ge in
many dimensions han mos o he o he Eo-
cene ba s, and analyses o wing mo phology
by No be g (1989) indica ed ha Hassianyc-
e is was cha ac e ized by signi ican ly high-
e wing loading han we e Palaeochi op e-
yx,A chaeonyc e is,o Ica onyc e is. In-
deed, Hassianyc e is appa en ly had a highe
wing loading han mos ex an ba s ( ig. 43).
These obse a ions, aken oge he wi h a
easonably high aspec a io, a g ea ly en-
la ged cochlea, and p esence o a calca , sug-
ges ha Hassianyc e is o aged by as ae ial
hawking, mos likely well abo e he g ound
in o es gaps o abo e he canopy. Hassi-
anyc e is hus ep esen s ano he shi in o -
aging s a egy, away om combined pe ch
hun ing and ae ial hawking nea he g ound
o ege a ion o a as - lying, con inuous ae-
ial hawking o aging s a egy simila o ha
seen in ex an hinopoma ids and some es-
pe ilionids. As no ed abo e, howe e , his
shi appa en ly occu ed a e he lineage
leading o Hassianyc e is di e ged om he
lineage leading o he mic ochi op e an
c own g oup.
Con inuing o mo e up he phylogene ic
ee, we ind ha he clade comp ising Pa-
laeochi op e yx plus he mic ochi op e an
c own g oup is diagnosed by only one un-
ambiguous synapomo phy, he p esence o a
en al accesso y p ocess on C5. This ea u e
is no indica i e o any change in o aging
habi s, bu a he a con inua ion o neck
modi ica ions associa ed wi h oos ing be-
ha io (see discussion unde cha ac e 77).
T ans o ma ions ha diagnose he mic ochi-
op e an c own g oup (e.g., modi ica ion o
he p emaxilla a icula ion, educ ion in he
132 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
numbe o lowe p emola s) indica e changes
in he mas ica o y appa a us, bu no majo
beha io al shi s.
In summa y, we p opose ha o aging be-
ha io in he mic ochi op e an lineage
e ol ed in a se ies o s eps: (1) gleaning ood
objec s du ing sho ligh s om a pe ch us-
ing ision o o ien a ion and obs acle de ec-
ion; p ey de ec ion by passi e means, in-
cluding ision and/o lis ening o p ey-gen-
e a ed sounds (no known examples in ossil
eco d); (2) gleaning s a iona y p ey om a
pe ch using echoloca ion and ision o o i-
en a ion and obs acle de ec ion; p ey de ec-
ion by passi e means (Ica onyc e is;A -
chaeonyc e is); (3) pe ch hun ing o bo h
s a iona y and lying p ey using echoloca ion
and ision o o ien a ion and obs acle de ec-
ion; p ey de ec ion and acking using echo-
loca ion o lying p ey and passi e means
o s a iona y p ey (no known example, al-
hough Ica onyc e is and/o A chaeonyc e is
may ha e done his a imes); (4) combined
pe ch hun ing and con inuous ae ial hawking
using echoloca ion and ision o o ien a ion
and obs acle de ec ion; p ey de ec ion and
acking using echoloca ion o lying p ey
and passi e means o s a iona y p ey; cal-
ca -suppo ed u opa agium used o p ey
cap u e (common ances o o Hassianyc e is
and Palaeochi op e yx; e ained in Palaeo-
chi op e yx); and (5) exclusi e eliance on
con inuous ae ial hawking using echoloca-
ion and ision o o ien a ion and obs acle
de ec ion; p ey de ec ion and acking using
echoloca ion (Hassianyc e is; common an-
ces o o mic ochi op e an c own g oup).
Gi en he opology o he ee we de i ed
ea lie ( ig. 36), i seems mos likely ha he
la e o aging s a egy— eliance on con in-
uous ae ial hawking—was p imi i e o he
mic ochi op e an c own g oup. This is con-
sis en wi h op imiza ion o o aging s a e-
gies bo h wi hin he mic ochi op e an c own
g oup and among he ossil s em g oup
o ms.
The conclusion ha ae ial hawking is he
p imi i e o aging s a egy o he mic ochi-
op e an c own g oup sugges s ha gleaning,
passi e p ey de ec ion, and pe ch hun ing
among ex an axa ep esen seconda ily de-
i ed specializa ions a he han e en ions o
p imi i e habi s. I so, opology o ou ee
( ig. 36) indica es ha passi e p ey de ec ion
e ol ed independen ly a leas i e o six
imes—in phyllos omids, mys acinids, an o-
zoids, espe ilionids, and ei he indepen-
den ly in nyc e ids and megade ma ids o in
he common ances o o Rhinolophoidae.
Simila ly, gleaning mus ha e e ol ed inde-
penden ly a leas ou imes wi hin Mic o-
chi op e a—in phyllos omids, an ozoids,
espe ilionids, and hinolophoids. Pe ch
hun ing appa en ly e ol ed a leas h ee
imes—in phyllos omids, espe ilionids, and
hinolophoids. Al hough passi e p ey de ec-
ion, gleaning, and pe ch hun ing a e linked
in some axa (e.g., mos megade ma ids and
nyc e ids), hey a e decoupled in o he o ms.
Fo example, Mys acina appa en ly uses pas-
si e p ey de ec ion bu does no glean o hun
om pe ches; ins ead, i app oaches i s p ey
‘‘on oo ’’ (B. Lloyd, pe sonal commun.).
Rhinolophids some imes hun om pe ches
and glean, bu appa en ly do no use passi e
cues o de ec hei p ey (No be g and Ray-
ne , 1987). La ia ons hun s om pe ches
bu appa en ly does no glean o use passi e
p ey de ec ion (Vaughan and Vaughan,
1986). These obse a ions sugges ha pas-
si e p ey de ec ion, gleaning, and pe ch
hun ing e ol ed in di e en ways in di e en
mic ochi op e an lineages.
The ela i e iming o he e olu iona y
and beha io al changes in he lineage leading
o ex an Mic ochi op e a can be es ima ed
om he ossil eco d. Taxa ep esen ing
s ages 2, 4, and 5 as de ined abo e we e ap-
pa en ly p esen simul aneously a Messel,
sugges ing ha e olu iona y ans o ma ions
in o aging s a egies may ha e occu ed
e y apidly in ea ly membe s o he mic o-
chi op e an lineage. This was p obably acil-
i a ed by he mechanical coupling o en i-
la ion and ligh , which mean ha he ene gy
cos s o echoloca ion o lying ba s we e low,
pa icula ly in compa ison o he ene ge ic
bene i s o ae ial hawking o insec s in he
absence o any compe i o s.
The e olu ion o con inuous ae ial hawk-
ing may ha e been he ‘‘key inno a ion’’
(sensu Liem, 1973) esponsible o he bu s
o di e si ica ion in mic ochi op e an ba s
ha occu ed in he Eocene. Fossils e e able
o six majo ex an lineages a e known om
Middle–La e Eocene deposi s ( able 1): (1)
1998 133SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
Emballonu idae, (2) Megade ma idae, (3)
Hipposide inae, (4) Rhinolophinae, (5) Na-
aloidea, and (6) Molossinae. Recons uc ion
o ghos lineages ollowing he me hods o
No ell (1992) leads o he conclusion ha
many mo e ex an lineages we e minimally
p esen by he end o he Eocene, including
(7) Rhinopoma oidea, (8) Nyc e idae, (9)
Noc ilionoidea, (10) Mys acinidae, (11) An-
ozoidae, (12) Tomopea inae, and (13) Ves-
pe ilionidae. All o hese mus ha e di-
e ged du ing he Eocene gi en he Eocene
ages o hei sis e - axa. I hus appea s ha
Ica onyc e is and he Messel ba s p o ide an
unp eceden ed iew o he s eps leading up
o a majo adap i e adia ion o mammals.
CLASSIFICATION OF EOCENE BATS
The phylogene ic esul s o his s udy in-
dica e ha many g oupings o Eocene axa
p e iously ecognized in o mal classi ica-
ions (e.g., Eochi op e a, Palaeochi op e y-
goidea, A chaeonyc e ididae [including Ica-
onyc e is]) a e no monophyle ic. Beginning
a he lowes axonomic le el, i seems mos
app op ia e o place each o he ou gene a
conside ed in his s udy in i s own monophy-
le ic amily
13
:Ica onyc e is in Ica onyc e i-
13
The e is conside able disag eemen in he li e a u e
conce ning he o ma ion o amily-g oup names based
on gene ic names ending in -nyc e is. The amily name
in common usage o ex an Nyc e is is Nyc e idae
(Koopman and Jones, 1970; Hill and Smi h, 1984;
Koopman, 1984, 1993, 1994). Re illiod (1917b) used a
similia o ma ion when he named A chaeonyc e idae,
as did Jepsen (1966) when he named Ica onyc e idae.
Howe e , Russel and Sige´ (1970) a gued ha he p ope
o ma ion o amily-g oup names om -nyc e is e-
qui ed he spelling -nyc e ididae. They hus changed he
spelling o A chaeonyc e idae o A chaeonyc e ididae.
Some au ho s ha e ollowed his usage (e.g., Habe se z-
e and S o ch, 1987, 1989; Habse se ze e al., 1992,
1994), bu many ha e con inued o use A chaeonyc e -
idae (e.g., Hill and Smi h, 1984; Hand e al., 1994; Mc-
Kenna and Bell, 1997). Habse se ze and S o ch (1987)
ollowed Russel and Sige´’s (1970) usage when hey
named Hassianyc e ididae, and subsequen au ho s ha e
used his spelling (e.g., Habe se ze e al., 1992; Mc-
Kenna and Bell, 1997).
Despi e Russell and Sige´’s (1970) a gumen abou he
na u e o he g eek oo o -nyc e is, we hink ha i is
coun e p oduc i e o spell some amily-g oup names in
one ashion and o he s di e en ly. Because Nyc e idae
is widely accep ed as he spelling o he amily-g oup
name based on Nyc e is, we a gue ha all amily-g oup
names based on gene a ending in -nyc e is should be
spelled in he same ashion. Acco dingly, we ecognize
he Ica onyc e idae, A chaeonyc e idae, and Hassianyc-
e idae as he mos app op ia e spellings o amily-
g oup names based on Ica onyc e is,A chaeonyc e is,
and Hassianyc e is.
dae Jepsen, 1966; A chaeonyc e is in A -
chaeonyc e idae Re illiod, 1917b; Hassi-
anyc e is in Hassiancy e idae Habe se ze
and S o ch, 1987; and Palaeochi op e yx in
Palaeochi op e ygidae Re illiod, 1917b. This
a angemen p ese es monophyly o highe
axonomic g oups and se es o highligh he
mo phological and p esumed beha io al di -
e ences among gene a.
As we iew hem, Ica onyc e idae, A -
chaeonyc e idae, and Hassiancy e idae each
cu en ly con ain only he nomina e genus.
The si ua ion is sligh ly mo e complex in he
case o Palaeochi op e ygidae. Two addi ion-
al gene a, Cecilionyc e is and Ma hesia,
ha e been e e ed o Palaeochi op e ygidae
by p e ious au ho s based on den al ea u es
(e.g., Sige´ and Russell, 1980; Hill and Smi h,
1984; Bea d e al., 1992). In he absence o
addi ional da a, we p o isionally accep his
assessmen , al hough we no e ha disco e y
o mo e comple e ma e ial may ul ima ely
demons a e ha Palaeochi op e ygidae as
hus de ined (including Palaeochi op e yx,
Cecilionyc e is, and Ma hesia) is a pa aphy-
le ic assemblage.
Ageina, which is known only om den al
agmen s, has been e e ed a iously o
‘‘ amily unce ain’’ (Russell e al., 1973), Pa-
laeochi op e ygoidea (Smi h, 1977; Smi h
and S o ch, 1981), Eochi op e a (Sige´ and
Legend e, 1983), A chaeonyc e idae (Hill
and Smi h, 1984), and possibly Na alidae
sensu Van Valen (Bea d e al., 1992). We
conside he assessmen o Bea d e al.
(1992), which is based on he mos ecen
e idence (including compa isons wi h he
new axon Hon o i s), o ep esen he bes
cu en wo king hypo hesis. Acco dingly, we
ollow sugges ions made by Bea d e al.
134 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
(1992) and e e bo h Ageina and Hon o i s
o Na aloidea ince ae sedis.
Aus alonyc e is, which is also poo ly
known, was placed in Mic ochi op e a, am-
ily ince ae sedis by Hand e al. (1994). They
no ed ha his o m has a den i ion simila
o ha o A chaeonyc e is (a leas in e ms
o he absence o de i ed ai s), al hough he
pe osal e e ed o Aus alonyc e is appea s
ela i ely mo e de i ed and is somewha es-
pe ilionoid-like (Hand e al., 1994). These
and o he obse a ions sugges o us ha
Aus alonyc e is p obably i s in he ee
somewhe e be ween A chaeonyc e is and he
mic ochi op e an c own g oup. Gi en his
hypo hesis, any decision abou how o clas-
si y Aus alonyc e is equi es conside a ion
o b oade issues su ounding he classi ica-
ion o s em-g oup o ms.
The p incipal nomencla u al p oblem
aced in his s udy conce ns highe -le el
classi ica ion o he Eocene g oups ha all
ou side he mic ochi op e an c own g oup—
in o he wo ds, whe e o pu Ica onyc e idae,
A chaeonyc e idae, Hassiancy e idae, and
Palaeochi op e ygidae. Deba es ha e aged
in he sys ema ic li e a u e o decades con-
ce ning he ela i e p os and cons o di e en
app oaches o de ining and naming axonom-
ic g oups (e.g., Ghiselin, 1966, 1984; Nelson,
1972, 1974; Bock, 1974; Duncan and Es a-
b ook, 1976; Es ab ook, 1978, 1986; Je e -
ies, 1979; Wiley, 1979, 1981; Duncan, 1980;
Phillips, 1984; Rowe, 1987, 1988; Gau hie
e al., 1988; Heywood, 1988; de Quie oz and
Gau hie , 1990, 1992, 1994; Minelli, 1991;
Lucas, 1992; Meie and Rich e , 1992; Lucas
and Luo, 1993; Wyss and Flynn, 1993; B y-
an , 1994, 1996; de Quei oz, 1994; Smi h,
1994; Wyss and Meng, 1996; McKenna and
Bell, 1997). Fou p incipal issues ha e been
deba ed: (1) whe he axa should be ecog-
nized on he basis o sha ed cha ac e s (i.e.,
diagnoses) o de ined phylogene ically (i.e.,
on he basis o hei ela ionships); (2)
whe he all named g oups abo e he species
le el mus be monophyle ic (some wo ke s
ha e a gued ha con ex pa aphyle ic g oups
may be use ully employed in classi ica ions);
(3) he p os and cons o ecognizing o mal
anks (e.g., ‘‘o de ,’’ ‘‘ amily’’) in classi i-
ca ions; and (4) he ela i e me i s o node-
based, c own-clade- es ic ed, s em-based,
s em-modi ied node-based, and o he ypes
o phylogene ic de ini ions. We concu wi h
mos ecen au ho s in concluding ha phy-
logene ic de ini ions a e essen ial and ha all
named axa should be monophyle ic (o a
leas po en ially monophyle ic). We conside
issues ela ed o ank o be o seconda y im-
po ance compa ed wi h he de ini ion o axa
(see discussion below). The cen al issue, in
ou opinion, is he me hod(s) used o de ine
he limi s o axa, pa icula ly hose o which
we may apply well-known axonomic names
(e.g., Mic ochi op e a).
Gi en he phylogene ic ela ionships hy-
po hesized he e, h ee gene al op ions o de-
ining Mic ochi op e a seem app op ia e. Us-
ing a node-based de ini ion ha emphasizes
he impo ance o he ex an c own g oup
(i.e., a c own-clade- es ic ed de ini ion), we
migh de ine Mic ochi op e a as he clade
s emming om he mos ecen common an-
ces o o Emballonu idae and Yangochi op-
e a. I de ined his way, Yinochi op e a
would be included wi hin Mic ochi op e a
ega dless o whe he Emballonu idae alls
ou side Yinochi op e a (as sugges ed by his
s udy) o inside Yinochi op e a (as sugges ed
by Koopman, 1985, 1994). A less explici
c own-clade de ini ion migh de ine Mic o-
chi op e a as he clade s emming om he
mos ecen common ances o o all ex an
ba s ha use sophis ica ed echoloca ion.
When de ined ei he way, Mic ochi op e a
would be equi alen o he mic ochi op e an
c own g oup as discussed ea lie , and would
he e o e exclude Palaeochi op e ygidae,
Hassiancy e idae, A chaeonyc e idae, and
Ica onyc e idae. The main ad an ages o
hese de ini ions a e ha hey a e cong uen
wi h gene al usage o he name Mic ochi op-
e a by mos biologis s, and ha hey can be
expec ed o emain ela i ely s able (e.g., dis-
co e y o new ossils will no a ec he lim-
i s o Mic ochi op e a; ossil species would
ei he all inside o ou side his clade). The
la e op ion (de ining Mic ochi op e a as he
clade s emming om he mos ecen com-
mon ances o o all ex an echoloca ing ba s)
depends on pe sis ence o ex an o ms, and
hus migh be conside ed po en ially uns able
because majo ex inc ions could a ec g oup
con en s in he u u e (see discussion o his
p oblem in Lucas [1992] and Lucas and Luo
1998 135SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
[1993]). Howe e , we conside his o be o
mino impo ance in he case o ba s since
he basal lineages (e.g., Emballonu idae,
Yinochi op e a, and Yangochi op e a) a e all
speciose.
The p incipal disad an age o using a
c own-clade- es ic ed de ini ion is ha i e -
ec i ely excludes some Eocene axa om
Mic ochi op e a simply because hey a e ex-
inc . Gi en hei mo phology, ela ionships,
and p esumed ecologies (see abo e), i is
likely ha Ica onyc e is,A chaeonyc e is,
Hassianyc e is, and Palaeochi op e yx
would be conside ed mic ochi op e ans we e
hey ali e oday. Wi hou a o mal name
linking hese o ms o ex an Mic ochi op-
e a, conside able phylogene ic (and in e ed
beha io al and ecological) in o ma ion
would be los om any classi ica ion scheme.
An al e na i e node-based de ini ion o
Mic ochi op e a could include he Eocene
g oups (Ica onyc e idae, A chaeonyc e idae,
Hassianyc e idae, and Palaeochi op e ygi-
dae) in Mic ochi op e a i he la e was de-
ined as he clade s emming om he mos
ecen common ances o o Ica onyc e idae
and Yangochi op e a (o some o he ex an
clade). Ad an ages o his de ini ion include
concu ence wi h cu en usage as applied by
mos paleon ologis s (e.g., Habe se ze and
S o ch, 1987, 1988, 1992; No acek, 1987,
1991; Ca oll, 1988; S o ch and Habe se ze ,
1988; Hand e al., 1994; McKenna and Bell,
1997), and cong uence o his de ini ion wi h
ideas abou he ‘‘key cha ac e ’’ o Mic o-
chi op e a—sophis ica ed la yngeal echolo-
ca ion, which we in e was p esen in Ica-
onyc e is,A chaeonyc e is,Hassianyc e is,
and Palaeochi op e yx. Howe e , his de i-
ni ion lea es us wi hou a o mal name o
he mic ochi op e an c own g oup, which is
he clade o cen al in e es o mos biolo-
gis s. Mo eo e , his de ini ion migh p o e
p oblema ic i addi ional ba ossils a e ound
ha all ou side he Ica onyc e idae
1
Yan-
gochi op e a clade, ye s em om he b anch
leading o Mic ochi op e a a he han ha
leading o Megachi op e a.
This p oblem aises ye ano he possibili y,
namely a s em-based app oach. Mic ochi op-
e a migh be de ined as all chi op e ans sha -
ing a mo e ecen common ances o wi h
Yangochi op e a han wi h P e opodidae.
This de ini ion would elimina e he p oblem
o classi ying u u e s em-g oup ossils, bu
s ill lea es us wi hou a o mal name o he
mic ochi op e an c own g oup. Such a de i-
ni ion is also p oblema ic because i blu s he
dis inc ion be ween Megachi op e a and Mi-
c ochi op e a. Sophis ica ed la yngeal echo-
loca ion, he single ai mos o en associ-
a ed wi h li ing mic ochi op e ans, was ap-
pa en ly absen in he mos ecen common
ances o o Megachi op e a and Mic ochi-
op e a. This implies ha echoloca ion was
also absen in he e y ea lies membe s o
he mic ochi op e an s em g oup. Use o a
s em-g oup de ini ion o Mic ochi op e a
would hus o ce us o include nonecholo-
ca ing ela i es o ex an mic ochi op e ans
in Mic ochi op e a, a solu ion ha many
wo ke s would no ind accep able.
S ill ano he issue in ol es wha names (i
any) should be applied o nodes along he
backbone o he ee below he mic ochi op-
e an c own g oup. Fo example, he e a e
conside able mo phological (and in e ed
ecological) di e ences be ween he basal
b anches (e.g., Ica onyc e idae and A chaeo-
nyc e idae) and he mo e de i ed clade in-
cluding Hassianyc e idae, Palaeochi op e y-
gidae, and he mic ochi op e an c own
g oup. Monophyly o he la e clade is well
suppo ed, and a o mal name o his g oup
would d aw a en ion o i s exis ence and a-
cili a e discussion o his g oup, objec i es
ha we conside desi able. Al hough p o id-
ing names o e e y node in a ee is exces-
si e, some p e iously unnamed clades do, in
ou iew, equi e o mal ecogni ion.
Keeping all o hese issues in mind, we p o-
pose a new highe -le el classi ica ion o ba s
( able 8) ha combines bo h node-based and
s em-based de ini ions in a way ha we hink
maximizes bo h u ili y and s abili y. This so-
lu ion is simila o ha p oposed by Wyss and
Meng (1996) o gli i o m mammals ( oden s,
lagomo phs, and hei ex inc ela i es). In ou
classi ica ion, we apply s em-based names o
he wo p incipal b anches o he chi op e an
ee. Mic ochi op e amo pha is de ined as all
chi op e ans sha ing a mo e ecen common
ances o wi h Mic ochi op e a han wi h Me-
gachi op e a; Megachi op e amo pha is de-
ined as all chi op e ans sha ing a mo e ecen
common ances o wi h Megachi op e a han
136 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
TABLE 8
A Highe -le el Classi ica ion o Ba s Including
Selec ed Fossil Gene a Discussed in he Tex
a
O de Chi op e a
Megachi op e amo pha, new axon
†A chaeop e opus
Subo de Megachi op e a
Family P e opodidae
Mic ochi op e amo pha, new axon
†Aus alonyc e is
†Family Ica onyc e idae
†Ica onyc e is
†Family A chaeonyc e idae
†A chaeonyc e is
Mic ochi op e a o mes, new axon
†Eppsinyc e is
†Family Palaeochi op e ygidae
†Palaeochi op e yx, †Ma hesia, †Cecilionyc e is
†Family Hassianyc e idae
†Hassianyc e is
Subo de Mic ochi op e a
†Vampy a us
Supe amily Emballonu oidea
Family Emballonu idae
Sub amily Taphozoinae
†Vespe ilia us
Sub amily Emballonu inae
In ao de Yinochi op e a
Supe amily Rhinopoma oidea
Family C aseonyc e idae
Family Rhinopoma idae
Supe amily hinolophoidea
Family Nyc e idae
Family Megade ma idae
†Nec oman is
Family Rhinolophidae
†Vayla sia
Sub amily Rhinolophinae
Sub amily Hipposide inae
†Hipposide os (Pseudo hinolophus),
†Palaeophyllopho a, †Pa aphyllopho a
In ao de Yangochi op e a
†Family Philididae
†Philisis, †Dizzya
Family Mys acinidae
Supe amily Noc ilionoidea
Family Phyllos omidae
Family Mo omoopidae
Family Noc ilionidae
Supe amily Na aloidea
†Hon o i s
†Aegina
†S ehlinia
†Chad onyc e is
Family Myzopodidae
Family Fu ip e idae
TABLE 8—(Con inued)
Family Thy op e idae
Family Na alidae
Supe amily Molossoidea
Family An ozoidae
Family Molossidae
Sub amily Tomopea inae
Sub amily Molossinae
†Wallia, †Cu ie imops
Supe amily Vespe ilionoidea
Family Vespe ilionidae
Sub amily Vespe ilioninae
Sub amily Miniop e inae
Sub amily Myo inae
Sub amily Mu ininae
Sub amily Ke i oulinae
a
See ex o a discussion o classi ica ion below he
le el o sub amily. Gene a ha a e no placed in a spe-
ci ic amily o sub amily a e conside ed ince ae sedis
wi hin he nex highe -le el axon lis ed abo e (e.g.,
S ehlinia, which is lis ed unde supe amily Na aloidea
wi hou e e ence o amily o sub amily, is conside ed
ince ae sedis wi hin Na aloidea).
wi h Mic ochi op e a. As so de ined, Mic o-
chi op e amo pha includes Ica onyc e idae,
A chaeonyc e idae, Hassiancy e idae, Palaeo-
chi op e ygidae, and Mic ochi op e a, as well
as a numbe o poo ly known ossil o ms
(e.g., Aus alonyc e is) ha a e mo e closely
ela ed o ex an Mic ochi op e a han o Me-
gachi op e a.
Wi hin Mic ochi op e amo pha, Mic ochi-
op e a o mes is de ined as he clade s em-
ming om he mos ecen common ances o
o Hassianyc e idae, Palaeochi op e ygidae,
and Mic ochi op e a; his is a node-based de -
ini ion. Any ossil o ms ha nes wi hin
hese g oups o all be ween hem on he ee
would be placed in Mic ochi op e a o mes.
Aus alonyc e is may someday be shown o
belong o his g oup, bu o now i mus be
conside ed Mic ochi op e amo pha ince ae
sedis.
Finally, we de ine (and he eby es ic )
Mic ochi op e a o he clade s emming om
he mos ecen common ances o o Embal-
lonu idae and Yangochi op e a. This node-
based, c own-g oup- es ic ed de ini ion
equa es Mic ochi op e a wi h wha we ha e
hus a in his pape called he ‘‘mic ochi-
op e an c own g oup’’— he smalles clade
ha comp ises all o he ex an lineages o
1998 137SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
mic ochi op e an ba s. As no ed abo e, he
ad an ages o his de ini ion o Mic ochi op-
e a a e ha i is cong uen wi h gene al us-
age o he name by biologis s, and i p o ides
s abili y because u u e ossil disco e ies o
ex inc ions will no a ec he limi s o his
axon. The p incipal disad an age o his de -
ini ion—exclusion o he Eocene ossil axa
la gely because hey a e ex inc —is mi iga -
ed by ou ecogni ion o Mic ochi op e a-
mo pha and Mic ochi op e a o mes, which
ocuses a en ion on he mic ochi op e an a -
ini ies o hese axa wi hou squeezing hem
in o Mic ochi op e a i sel .
This sys em o classi ica ion, which ec-
ognizes wo named g oups mo e inclusi e
han Mic ochi op e a and less inclusi e han
Chi op e a, p esen s p oblems om he poin
o iew o anks, since no anks a e ecog-
nized o exis be ween o de and subo de .
This so o p oblem has haun ed sys ema-
is s o decades, mos pa icula ly since he
ad en o cladis ic me hods o phylogene ic
analysis. Some wo ke s ha e ad oca ed
elimina ion o all anks in a o o a simple
hie a chical sys em (e.g., de Quei oz and
Gau hie , 1992), while o he s ha e main-
ained ully anked classi ica ions despi e
me hodological p oblems ha a ise om oo
many names and oo ew anks (e.g., Mc-
Kenna, 1975; Wilson and Reede , 1993; Mc-
Kenna and Bell, 1997). We adop an in e -
media e posi ion. While ecognizing ha o -
mal anks a e ex emely use ul o commu-
nica ion, a angemen o sys ema ic
collec ions, and o he bookkeeping cho es
(e.g., bibliog aphic p ojec s), we also hink
ha he o mal anks a ailable (e en he el-
a i ely la ge numbe ecognized by McKen-
na and Bell, 1997) canno su ice o desc ibe
he complex hie a chies now being eco e ed
om phylogene ic analyses. Acco dingly, we
choose o ecognize bo h anked and un an-
ked names. We e ain anks o many axo-
nomic names, pa icula ly hose ha apply
p incipally o ex an axa and a e in b oad
use in he neon ological li e a u e (e.g., O de
Chi op e a, Subo de Mic ochi op e a, Fam-
ily Phyllos omidae). Fo o he clade names,
pa icula ly hose ha we an icipa e will be
used p incipally o discussions o ossils
(e.g., Mic ochi op e amo pha, Mic ochi op-
e a o mes), we do no p opose o mal anks.
This comp omise allows he o mal ecog-
ni ion o names o clades a nume ous le els
in he ee, while a he same ime p omo ing
s abili y by e aining mos amilia names a
amilia anks (e.g., Mic ochi op e a does no
ha e o be educed o he ank o supe amily
o accommoda e ecogni ion o mo e inclu-
si e axa wi hin Chi op e a). Al hough his
app oach will p obably be c i icized by ad-
he en s o bo h he ‘‘no ank’’ and ‘‘always
ank’’ schools o classi ica ion, we hink ha
i is he only easonable solu ion o he ax-
onomic p oblems p esen ed by Chi op e a.
To a oid p oli e a ion o edundan names,
we cu en ly limi he use o adi ional in-
ao dinal and supe amilial ank names o
wi hin Mic ochi op e a. We ollow Simmons
(1998) in ecognizing se en monophyle ic
supe amilies and wo in ao de s wi hin Mi-
c ochi op e a. All o hese axa a e ecog-
nized on he basis o node-based de ini ions.
The supe amilies we ecognize a e Embal-
lonu oidea (Emballonu idae), Rhinopoma o-
idea (Rhinopoma idae
1
C aseo yc e idae),
Rhinolophoidea (Nyc e idae
1
Megade ma-
idae
1
Rhinolophidae), Noc ilionoidea
(Noc ilionidae
1
Mo moopidae
1
Phyllos-
omidae), Na aloidea (Myzopodidae
1
Thy-
op e idae
1
Fu ip e idae
1
Na alidae), Mo-
lossoidea (An ozoidae
1
Molossidae), and
Vespe ilionoidea (Vespe ilionidae). Yino-
chi op e a con ains Rhinopoma oidea
1
Rhinolophoidea; Yangochi op e a con ains
Mys acinidae
1
Noc ilionoidea
1
Na aloidea
1
Molossoidea
1
Vespe ilionoidea. No all
amilies a e e e ed o supe amily-le el
axa (e.g., Mys acinidae is le ince ae sedis
in Yangochi op e a), and no all supe ami-
lies a e e e ed o in ao de s (e.g., Embal-
lonu oidea) because we see no pu pose in
p o iding edundan names. The eason ha
we ecognize wo mono ypic supe amilies
(Emballonu oidea and Vespe ilionoidea) is
ha we an icipa e ha u u e wo ke s may
subdi ide hese g oups in o mul iple amilies
as phylogene ic esolu ion inc eases in hese
pa s o he ee.
Classi ica ion o mos o he emaining
Eocene axa is ela i ely s aigh o wa d gi -
en wha is known abou hese o ms and he
classi ica ion p oposed abo e. Following he
conclusions o p e ious au ho s (see he In-
oduc ion), we place Vespe ilia us in Em-
138 NO. 235BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY
ballonu idae: Taphozoinae, and e e Nec o-
man is o Megade ma idae ( able 8). Simila ly,
we ollow he consensus ha Hipposide os
(Pseudo hinolophus), Palaeophyllopho a,
and Pa aphyllopho a should be classi ied as
Rhinolophidae: Hipposide inae. Ra he han
placing Vayla sia in Hipposide inae as did
Sige´ (1990), we ollow McKenna and Bell
(1997) and e e Vayla sia o Rhinolophidae
ince ae sedis, al hough we no e ha his ax-
on may subsequen ly be shown o ep esen
a basal membe o Rhinolophinae (see dis-
cussion in Hand and Ki sch, 1998).
As no ed by Simmons (1998), ecogni ion
o Hipposide inae and Rhinolphinae as sub-
amily-le el axa ( a he han as dis inc am-
ilies) is p e e ed because i acili a es dis-
cussion o he la ge clade comp ising hese
axa (i.e., Rhinolophidae). We do no ag ee
wi h McKenna and Bell (1997) ha he name
Hipposide inae should be eplaced wi h Rhi-
nonyc e inae. Al hough Rhinonyc e inae (
5
Rhinonyc e ina G ay, 1866) has p io i y o e
Hipposide inae Flowe and Lydekke , 1891
as a amily-g oup name, no au ho o he han
G ay (1866) used he o me name un il i
was esu ec ed by McKenna and Bell
(1997). Mille (1907) used he name Hip-
poside idae o his g oup in his in luen ial
monog aph on he amilies and gene a o
ba s because Hipposide os G ay, 1831 has
p io i y o e Rhinonyc e is G ay, 1866 (
5
Rhinonic e is G ay, 1847). All subsequen
au ho s ha e ollowed Mille ’s (1907) usage
o Hipposide idae/-inae, and we hink ha
li le would be gained by eplacing i wi h an
unknown name. Resolu ion o his p oblem
will equi e a pe i ion o he In e na ional
Commission on Zoological Nomencla u e;
meanwhile, we e ain Hipposide inae o he
sake o s abili y.
Al hough we ag ee wi h Hooke (1996)
ha Eppsinyc e is is a ba , we do no ag ee
ha i should be e e ed o Emballonu idae.
Hooke (1996) lis ed wo synampo phies
linking Eppsinyc e is o emballonu ids: p o-
g essi e mesial elonga ion o he igonid
om m3 o m1, and biloba ion o p4. We ind
nei he o hese o be con incing e idence o
emballonu id a ini ies. Published pho o-
g aphs o Eppsinyc e is (Hooke , 1996: pls.
1 and 2) illus a e ha mesial elonga ion o
he igonid is ba ely dis inguishable in his
o m, and ou su ey o den al a ia ion sug-
ges s ha a bilobed p4 is common in many
amilies o ba s. In ou opinion, no ea u e
o Eppsinyc e is (as i is cu en ly known)
clea ly link i wi h any ex an amily. How-
e e , posi ion o he hypoconulid on he low-
e mola s indica es ha Eppsinyc e is was ei-
he nyc alodon o myo odon ( he ee h a e
oo wo n o de e mine which mo phology is
p esen ), indica ing ha his genus p obably
belongs somewhe e wi hin he mic ochi op-
e a o m clade. On his basis, we place Epps-
inyc e is in Mic ochi op e a o mes ince ae
sedis.
In he yangochi op e an pa o he ee,
we e e Philisis and Dizzya o he ex inc
amily Philisidae ollowing Sige´ (1985,
1991a), which we in u n place ince ae sedis
wi hin Yangochi op e a pending addi ional
s udy. As no ed by Sige´ (1985), Philisidae
may e en ually p o e o be closely ela ed o
Na aloidea i Vampy a us is shown o be a
synonym o Philisis. As discussed p e ious-
ly, we e e Hon o i s and Aegina o Na a-
loidea ince ae sedis ollowing sugges ions
made by Bea d e al. (1992). We also e e
S ehlinia and Chad onyc e is o Na aloidea
ince ae sedis based on ecommenda ions o
he same au ho s. We place Wallia and Cu-
ie imops in Molossidae: Molossinae ince -
ae sedis ollowing McKenna and Bell
(1997) based on a gumen s de eloped in Le-
gend e and Sige´ (1983) and Legend e
(1985). Finally, we ollow Sige´ (1985) and
McKenna and Bell (1997) in e e ing Vam-
py a us o Mic ochi op e a ince ae sedis un-
il mo e ma e ial o his o m is disco e ed.
Because ela ionships among ex an mega-
chi op e ans a e s ill ela i ely poo ly unde -
s ood—in pa due o incong uence among
phylogenies de i ed om di e en da a se s
(Hood, 1989; Ki sch e al., 1995; Colgan and
Flanne y, 1995; Sp inge e al., 1995; Holla
and Sp inge , 1997)—we de ine Megachi op-
e a using a s em-modi ied node-based de i-
ni ion as ecommended by Wyss and Meng
(1996). Megachi op e a is hus de ined as he
clade s emming om he mos ecen com-
mon ances o o P e opus and all Recen
mammals mo e closely ela ed o P e opus
han o Mic ochi op e a o any o he mam-
malian o de o subo de . Megachi op e a as
hus de ined includes all ex an P e opodidae,
1998 145SIMMONS AND GEISLER: RELATIONSHIPS OF EOCENE BATS
ma idae) while hun ing o ogs. Can.
J. Zool. 59: 750–753.
Ba ghoo n, S. F.
1977. New ma e ial o Vespe ilia us Schlos-
se (Mammalia, Chi op e a) and sug-
ges ed ela ionships o emballonu id
ba s based on c anial mo phology. Am.
Mus. No i a es 2618: 29 pp.
Ba kley, L. J.
1984. E olu iona y ela ionships and na u al
his o y o Tomopeas a us (Mammalia:
Chi op e a). Mas e ’s hesis, Louisiana
S a e Uni ., Ba on Rouge, Lousiana.
Ba on, G., H. S ephan, and H. D. F ahm
1996a. Compa a i e neu obiology in Chi op-
e a. Vol. 1: Mac omo phology, b ain
s uc u e, ables, and a lases. Basel:
Bi kha¨use Ve lag.
1996b. Compa a i e neu obiology in Chi op-
e a. Vol. 2: B ain cha ac e is ics in ax-
onomic uni s. Basel: Bi kha¨use Ve lag.
1996c. Compa a i e neu obiology in Chi op-
e a. Vol. 3: B ain cha ac e is ics in
unc ional sys ems, ecoe hological ad-
ap a ion, adap i e adia ion and e olu-
ion. Basel: Bi kha¨use Ve lag.
Ba e , M., M. J. Donoghue, and E. Sobe
1991. Agains consensus. Sys . Zool. 40:
486–493.
1993. C usade? A eply o Nelson. Sys . Biol.
42: 216–217.
Bea d, K. C.
1993. Phylogene ic sys ema ics o P ima o-
mo pha, wi h special e e ence o De -
mop e a. In F. S. Szalay, M. J. No a-
cek, and M. C. McKenna (eds.), Mam-
mal phylogeny: Placen als, pp. 129–
150. New Yo k: Sp inge -Ve lag.
Bea d, K. C., B. Sige´, and L. K ish alka
1992. A p imi i e espe ilionoid ba om
he ea ly Eocene o cen al Wyoming.
C. R. Acad. Sci. Pa is 314: 735–741.
Bell, G. P.
1982a. Beha io al and ecological aspec s o
gleaning by a dese insec i o ous ba ,
An ozous palllidus (Chi op e a: Ves-
pe ilionidae). Beha . Ecol. Sociobiol.
10: 217–223.
1982b. P ey loca ion and senso y ecology o
wo species o gleaning, insec i o ous
ba s, An ozous pallidus (Vespe ilioni-
dae) and Mac o us cali o nicus (Phyl-
los oma idae). Ph.D. disse a ion,
Ca le on Uni ., O awa, Canada.
1985. The senso y basis o p ey loca ion by
he Cali o nia lea -nosed ba Mac o us
cali o nicus (Chi op e a: Phyllos omi-
dae). Beha . Ecol. Sociobiol. 16: 343–
347.
Bell, G. P., and M. B. Fen on
1986. Visual acui y, sensi i i y and binocula -
i y in a gleaning insec i o ous ba , Ma-
c o us cali o nicus (Chi op e a: Phyl-
los omidae). Anim. Beha . 34: 409–
414.
Belwood, J. J., and G. K. Mo is
1987. Ba p eda ion and i s in luence on call-
ing beha io in Neo opical ka ydids.
Science 238: 64–67.
Ben on, M. J., and R. Hi chin
1997. Cong uence be ween phylogene ic and
s a ig aphic da a on he his o y o li e.
P oc. R. Soc. London B 264: 885–890.
Be gg en, W. A., D. V. Ken , J. J. Flynn, and J.
A. Van Cou e ing
1985. Cenezoic geoch onology. Geol. Soc.
Am. Bull. 96: 1407–1418.
Be gg en, W. A., D. V. Ken , C. C. Swishe , and
M.-P. Aub y
1995. A e ised Cenozoic geoch onology and
ch onos a ig aphy. In: W. A. Be gg en,
D. V. Ken , M. -P. Aub y, and J. Ha -
denbol (eds.), Geoch onology, ime
scales and global s a ig aphic co ela-
ion. Tulsa, Oklahoma: Socie y o Sed-
imen a y Geology Special Publica ion
54: 129–212.
Be z, E.
1958. Un e suchungen u¨be die Ko ela ion
de Flugmechanismus bei den Chi op-
e en. Zool. Jah b. Ab . Ana . 77: 491–
526.
Bha naga , K. P., and J. R. Wible
1994. Obse a ions on he ome onasal o gan
o he colugo Cynocephalus (Mamma-
lia, De mop e a). Ac a Ana . 151: 43–
48.
Bhiwgade, D. A., A. B. Singh, A. P. Maneka , and
S. N. Menon
1992. Ul as uc u al de elopmen o cho io-
allan oic placen a in he Indian miniop-
e us ba , Miniop e us sch eibe sii u-
liginosus (Hodgson). Ac a Ana . 145:
248–264.
Black, H. L.
1974. A no h empe a e ba communi y:
s uc u e and p ey popula ions. J. Mam-
mal. 55: 138–157.
Blood, B. R.
1987. Con e gen hind limb mo phology and
he e olu ion o ish-ca ching in he
ba s Noc ilio lepo inus and Myo is (Pi-
zonyx) i esi (Mammalia, Chi op e a).
Ph.D. disse a ion, Uni . o Sou he n
Cali o nia.
[Document text truncated for crawler view.]
