RESEARCH ARTICLE Open Access
Palaeoecological di e ences unde lie a e
co-occu ence o Miocene Eu opean
p ima es
Daniel DeMiguel
1,2*
, Lau a Domingo
3,4
, Is ael M. Sánchez
2
, Isaac Casano as-Vila
2
, Josep M. Robles
2
and
Da id M. Alba
2
Abs ac
Backg ound: The wo main p ima e g oups eco ded h oughou he Eu opean Miocene, hominoids and
pliopi hecoids, seldom co-occu . Due o bo h hei a i y and insu icien ly unde s ood palaeoecology, i is cu en ly
unclea whe he he in equen co-occu ence o hese g oups is due o sampling bias o e lec s di e en
ecological p e e ences. He e we ely on he densely sampled p ima e-bea ing sequence o Abocado de Can Ma a
(ACM) in Spain o es whe he u no e s in p ima e assemblages a e co ela ed wi h palaeoen i onmen al changes.
We econs uc die a y e olu ion h ough ime (ca. 12.6–11.4 Ma), and hence clima e and habi a , using oo h-wea
pa e ns and ca bon and oxygen iso ope composi ions o enamel o he ubiqui ous musk-dee Mic ome yx.
Resul s: Ou esul s e eal ha p ima e species composi ion is s ongly co ela ed wi h dis inc en i onmen al
phases. La ge-bodied hominoids (d yopi hecines) a e eco ded in humid, densely- o es ed en i onmen s on he
lowe mos po ion o he ACM sequence. In con as , pliopi hecoids inhabi ed less humid, pa chy ecosys ems,
being eplaced by d yopi hecines and he small-bodied Plioba es owa d he op o he se ies in galle y o es s
embedded in mosaic en i onmen s.
Conclusions: These esul s suppo he iew ha pliopi hecoid p ima es p e e ed less humid habi a s han
hominoids, and e eal ha di e ences in beha iou al ecology we e he main ac o unde pinning hei a e co-
occu ence du ing he Eu opean Miocene. Ou indings u he suppo ha ACM hominoids, like Miocene apes as a
whole, inhabi ed mo e seasonal en i onmen s han ex an apes. Finally, his s udy highligh s he impo ance o
high- esolu ion, local in es iga ions o complemen la ge -scale analyses and illus a es ha con inuous and densely
sampled ossili e ous sequences a e essen ial o deciphe ing he complex in e play be ween bio ic and abio ic
ac o s ha shaped pas di e si y.
Keywo ds: Hominoids, Pliopi hecoids, P ima e e olu ion/adap a ion, Palaeodie , S able iso opes, Too h wea ,
Feeding beha iou , Palaeobiology
© The Au ho (s). 2021 Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License,
which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e
app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons licence, and indica e i
changes we e made. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle's C ea i e Commons
licence, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle's C ea i e Commons
licence and you in ended use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain
pe mission di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/licenses/by/4.0/.
The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he
da a made a ailable in his a icle, unless o he wise s a ed in a c edi line o he da a.
* Co espondence: demiguel@uniza .es
1
ARAID ounda ion / Uni e sidad de Za agoza, Depa amen o de Ciencias de
la Tie a, and Ins i u o Uni e si a io de In es igación en Ciencias Ambien ales
de A agón (IUCA), Ped o Ce buna 12, 50009 Za agoza, Spain
2
Ins i u Ca alà de Paleon ologia Miquel C usa on , Uni e si a Au ònoma de
Ba celona, Edi ici ICTA-ICP, C/ Columnes s/n, Campus de la UAB, 08193
Ce danyola del Vallès, Ba celona, Spain
Full lis o au ho in o ma ion is a ailable a he end o he a icle
DeMiguel e al. BMC Biology (2021) 19:6
h ps://doi.o g/10.1186/s12915-020-00939-5
Backg ound
Fossil p ima es om he Miocene o Eu ope a e gene -
ally a e and absen om mos si es, and when eco ded,
di e en p ima e species only seldom co-occu wi hin a
single locali y (s a ig aphic ho izon). As a esul , he e
is an ongoing deba e abou he ac o s unde pinning he
geog aphic and ch onos a ig aphic dis ibu ion o Mio-
cene p ima es in his con inen [1–6]. Be o e he dispe -
sal o ce copi hecoids (Old Wo ld monkeys) in o Eu ope
by he ea ly Tu olian (ca. 8.5 Ma, la e Miocene), wo
main g oups a e eco ded he e: pliopi hecoids, gene ally
conside ed a Eu asian clade o s em ca a hines (i.e. p e-
ceding he ce copi hecoid-hominoid spli [7]), and homi-
noids (c own ca a hines mo e closely ela ed o ex an
apes and humans han o ce copi hecoids [8–10]). Bo h
g oups p esumably dispe sed om A ica o Eu asia ol-
lowing he closu e o he Te hys Seaway du ing he la e
middle Miocene and subsequen ly di e si ied ac oss he
con inen gi ing ise o mul iple gene a and species.
The e a e app oxima ely one hund ed known locali ies
eco ding ei he o bo h o hese g oups—almos 20%
co esponding o Abocado de Can Ma a (ACM) in
Spain—al hough hey only co-occu in less han 10% o
hem, wi h pliopi hecoid-bea ing locali ies being sligh ly
mo e abundan (ca. 55 s. 45%) han hominoid-bea ing
ones [4,5]. Eu opean hominoids a e gene ally la ge
han pliopi hecoids and conside ed g ea apes (homi-
nids), excep o he small-bodied Plioba es, in e p e ed
as a s em hominoid [11].
The e a e e y ew si es in he Eu opean Miocene
whe e hominoids and pliopi hecoids co-occu [1,4,5],
and in mos cases, ossils o each g oup come om di -
e en locali ies wi hin he same si e (e.g. di e en ka s ic
issu e illings om La G i e) o i is unce ain whe he
hei emains came om he same s a ig aphic ho izon
(e.g. Cas ell de Ba be à [6]). S ong aphonomic e idence
suppo ing sympa y is only a ailable om Rudabánya in
Hunga y [12] and ACM (locali y ACM/C5-C3 [5]).
These locali ies he e o e o e a unique oppo uni y o
e alua e he palaeoen i onmen al condi ions ha en-
abled he coexis ence o pliopi hecoids and hominoids.
Gi en he a i y o p ima e emains among mammalian
assemblages om he Eu opean Miocene, he in equen
co-occu ence o wo di e en p ima e species a a sin-
gle locali y migh be, a leas in pa , a sampling a e ac
[1]. Howe e , he lack o co-occu ence in many well-
sampled p ima e-bea ing locali ies would a he suppo
he iew ha hei in equen coexis ence is a eal
phenomenon ha equi es an explana ion.
The compe i i e exclusion p inciple [13] p edic s ha
species occupying he same ecological niche canno co-
exis on he long- e m, ul ima ely leading o he p e a-
lence o one o e he o he , o o he p og essi e
di e gence o hei espec i e niches. This explana ion is
unlikely o hold o di e en clades such as pliopi he-
coids and hominoids, cha ac e ised among o he s by di -
e en locomo o adap a ions—leading o he p oposal
ha hese wo g oups p obably had di e en habi a
p e e ences, which only enabled hei coexis ence unde
pa icula ecological condi ions [1]. Ea ly ecomo pholo-
gical analyses based on ungula e hypsodon y (a p oxy o
ege a ion s uc u e also used o in e palaeop ecipi a-
ion) ha e concluded ha bo h hominoid and
pliopi hecoid-bea ing locali ies om he Eu opean Mio-
cene we e mo e humid han hose lacking p ima es [2].
Mo e ecen wo k based on hypsodon y u he showed
ha pliopi hecoids gene ally inhabi ed mo e humid en i-
onmen s (i.e. wi h highe mois u e and/o ain all) han
hominoids, al hough p obably less humid han hose in
which bo h g oups co-occu [4]. Howe e , gi en he
small numbe o ossil locali ies eco ding bo h axa,
such compa isons lack s a is ical powe and may ail o
conside palaeoen i onmen al di e ences ac oss geog-
aphy and ime h oughou he Miocene, especially a
he egional and local scales.
Focusing on aunal elemen s ha accompany p ima es
wi hin a single a ea and o e a es ic ed ime span
would allow us o es whe he u no e s in he p ima e
assemblage a e co ela ed o local changes in palaeo-
en i onmen al condi ions. The composi e s a ig aphic
sequence o ACM, loca ed in he a ea o els Hos ale s de
Pie ola wi hin he Vallès-Penedès Basin (NE Ibe ian
Peninsula [14]) (Fig. 1a–c), and spanning mo e han 1
My (12.6–11.4 Ma [5,15]), o e s an unpa alleled op-
po uni y o es his hypo hesis o se e al easons.
Fi s , he ACM sequence has deli e ed one o he mos
di e se p ima e assemblages om he Eu opean Mio-
cene, including bo h hominoids and pliopi hecoids [5].
Second, he co-occu ence o hominoids and pliopi he-
coids has only been eco ded in one ou o he 19 ACM
p ima e-bea ing locali ies, and he dis ibu ion o each
g oup h oughou he se ies does no appea andom
[5]. Finally, hanks o con inuous palaeon ological su -
eillance du ing he cons uc ion o a land ill, mos o
he ossil inds a e accu a ely da ed based on de ailed
li ho-, bio- and magne os a ig aphic co ela ions [5,16,
17]. This o e s he oppo uni y o es al e na i e expla-
na ions o he a iable empo al dis ibu ion o bo h
p ima e g oups du ing a es ic ed ime span and wi hin
a uni o m deposi ional se ing.
Wi h his aim in mind, he e we p esen a econs uc-
ion o he local clima e and palaeoen i onmen s
h ough he ACM sequence based on oo h wea and
den al enamel s able ca bon and oxygen iso ope alues
( ep esen ed by he no a ion δ
13
Candδ
18
O) o he u-
minan Mic ome yx—(Fig. 1d and Figu e S1)—a ep e-
sen a i e o he amily Moschidae (musk-dee ) [18,19].
Because he die o any plan -ea ing mammal is a di ec
DeMiguel e al. BMC Biology (2021) 19:6 Page 2 o 15
Fig. 1 (See legend on nex page.)
DeMiguel e al. BMC Biology (2021) 19:6 Page 3 o 15
link wi h he habi a s in which i li es, we used he die
(i.e. ecology) o his uminan o in o m abou ACM p i-
ma e ecological p e e ences and habi a s. The selec ion
o Mic ome yx is based on he ollowing easons: (1) he
eco d o his axon h oughou he ACM s a ig aphic
sequence, cha ac e ised by abundan isola ed ee h and
den ogna hic agmen s, allows us o cons uc a con-
inuous oo h-wea and iso opic eco d; (2) by ocusing
on a single genus, we can cha ac e ise mo e consis en ly
changes in he ege a ion co e , ood ab asi eness, e c.,
ac oss he selec ed ime in e al, a oiding hus biases
due o di e en physiologies; and (3) Mic ome yx was
ubiqui ous in he Miocene o Ibe ia, inhabi ing a a ied
ange o biomes om mo e o less open sa annas o
(sub) opical o es s [18,20], and exhibi ing an ex ao -
dina y e sa ili y in e ms o exploi a ion o nu ien s
and esou ces. The combina ion o all hese ac o s jus i-
ies he sui abili y o employing Mic ome yx as a case
s udy o in es iga e he en i onmen al and clima ic
shi s ha ook place du ing he la es middle Miocene
in he a ea o els Hos ale s de Pie ola.
Resul s
Too h wea
The ossil ma e ial s udied consis s o den ogna hic e-
mains and isola ed ee h o Mic ome yx. Al hough ini-
ially a single species o Mic ome yx was epo ed om
ACM [15], he cu en ly a ailable den al ma e ial indi-
ca es he p esence o h ee di e en mo pho ypes ha
likely ep esen di e en species (Addi ional ile 1: Sup-
plemen a y in o ma ion, No e 1).
Fo mesowea , we measu ed indi iduals and p o ide
he esul s o he h ee mo pho ypes sepa a ely (Add-
i ional ile 1: Supplemen a y in o ma ion, No e 2). All
Mic ome yx mo pho ypes (Table 1) show occlusal su -
aces wi h p edominance o high elie (pH = 95–100%)
and sha pened cusps (pS = 69–87%), al hough he e is a
conside able p opo ion o ounded apices (pR = 13–
31%). Mo pho ypes do no ha e any incidence o blun
cusps o , excep o Mic ome yx mo pho ype 3 (pL =
5%), low occlusal elie (which ela es o a low heigh
di e ence be ween oo h cusps and alleys). A e age
mesowea sco e (MS) o mo pho ypes anges om 0.19
o 0.31 (Table 1). We do ind signi ican di e ences wi h
he chi-squa e es (χ
2
) bu ma ginally non-signi ican
wi h he Fishe exac es . Fo he chi-squa e es , he e-
sul s show ha Mic ome yx mo pho ype 1 is di e en
om mo pho ype 2 (p= 0.0283), whe eas non-signi ican
di e ences a e be ween mo pho ypes 2 and 3 (p=
0.2255) and be ween mo pho ypes 1 and 3 (p= 0.2019).
On a e age, mesowea esul s indica e a b owsing on
so ege a ion and low le els o ab asi es (endogenous
phy oli h- ich g asses and dico yledonous, and exogen-
ous dus and g i ), al hough Mic ome yx mo pho ype
1—wi h mo e ounded cusps and highe MS (Table 1)—
shows a shi owa d he exploi a ion o oughe and
mo e ab asi e oods han he o he s.
S able iso ope da a
The di e ence be ween ca bona e (δ
18
O
CO3
) and phos-
pha e (δ
18
O
PO4
) oxygen iso opic composi ion can be
used o moni o possible bioapa i e diagene ic al e a ion.
Mic ome yx oo h enamel did no unde go ex ensi e
pos -bu ial al e a ion since he di e ence calcula ed be-
ween δ
18
O
CO3
and δ
18
O
PO4
(Δ
18
O
CO3
-
PO4
=δ
18
O
CO3
−
δ
18
O
PO4
) alues o he whole da ase (8.6 ± 0.8‰)is
wi hin he ange ob ained when conside ing mode n
mammals (~ 8.6–9.1‰[22,23]) (Addi ional ile 1: Sup-
plemen a y in o ma ion, No e 3).
Mic ome yx mo pho ypes (Table 1and Addi ional ile 6:
Table S1) yielded oo h enamel δ
13
C alues indica i e o
woodland o woodland-mesic C
3
g assland condi ions
(see Addi ional ile 1: Supplemen a y in o ma ion, No e
4 o a de ailed explana ion o he calcula ed δ
13
C cu -
o alues among di e en habi a s). Signi ican di e -
ences in δ
13
C alues ha e been only ound be ween mo -
pho ypes 1 and 2 o Mic ome yx ( = 4.250, p< 0.001)
(Addi ional ile 7: Table S2). Too h enamel δ
18
O
CO3
and
δ
18
O
PO4
alues do no show signi ican di e ences
among he h ee mo pho ypes (δ
18
O
CO3
:F= 0.845, d =
2, p= 0.439, and δ
18
O
PO4
:F= 0. 562, d = 2, p= 0.577)
(Addi ional ile 7: Table S2).
Rela ionship be ween mesowea and δ
13
C alues
A sca e plo showing he co ela ion be ween mean
MS and mean δ
13
C(‰VPDB) among Mic ome yx mo -
pho ypes by locali ies was cons uc ed (Fig. 2). Niche do-
mains a e isually p esen ed o each a iable. This
app oach (desc ibed in Addi ional ile 1: Supplemen a y
in o ma ion, No es 2 and 4) allowed us o con ex ualise
(See igu e on p e ious page.)
Fig. 1 Abocado de Can Ma a (ACM) and he moschid Mic ome yx.aGeog aphical si ua ion and gene al geological con ex o he Vallès-Penedès
Basin. bDe ailed geological map o he basin and he sequence o Abocado de Can Ma a (ACM) (black do ). cCo ela ion o he composi e local
magne os a ig aphy o ACM se ies wi h he Geomagne ic Pola i y Time Scale (modi ied om Alba e al. [5]). Eu opean Land Mammal Ages,
Mammal Neogene (MN) uni s and local biozones o he Vallès-Penedès Basin a e shown on he le . The shadowed egion indica es an
unsampled in e al o he Vallès-Penedès eco d. The s a ig aphic posi ions o he ACM locali ies s udied in his wo k a e shown o he igh on
he composi e li hos a ig aphic column. No e ha he bo om bounda y o he lowe mos local biozone is unknown. dLi e econs uc ion o a
Mic ome yx azanzae male. A by I.M.S
DeMiguel e al. BMC Biology (2021) 19:6 Page 4 o 15
Table 1 Summa y o mesowea and iso opic alues o Mic ome yx om he ACM sequence acco ding o mo pho ypes and en i onmen al phases
Mo pho ypes #MpS pR pH MS #Cδ
13
CSD
δ
13
C
δ
18
O
CO3
SD
δ
18
O
CO3
#Pδ
18
O
PO4
SD δ
18
O
PO4
Δδ
18
O
CO3
-
δ
18
O
PO4
Mo pho ype 1 8 69.2 30.8 100 0.31 8 −10.5 0.7 27.9 1.0 6 19.6 1.4 8.4
Mo pho ype 2 24 80.6 19.4 100 0.19 21 −11.8 0.8 27.7 1.5 19 19.1 1.9 8.7
Mo pho ype 3 11 86.7 13.3 95 0.21 6 −11.0 1.2 27.0 0.6 5 18.5 1.1 8.3
En i onmen al
phases
#MpS pR pHMS#Cδ
13
CSD
δ
13
C
δ
18
O
CO3
SD
δ
18
O
CO3
#Pδ
18
O
PO4
SD
δ
18
O
PO4
Δδ
18
O
CO3
-
δ
18
O
PO4
δ
13
C
die , mequ
MAP
a
(mm/
yea )
MAP
b
(mm/
yea )
MAT
(°C)
Phase III
(11.70–11.60 Ma)
23 71.9 28.1 97.3 0.32 19 −11.1 1.2 27.2 1.2 19 18.5 1.6 8.7 −27.2 801 608 17.1
Phase II
(11.90–11.79 Ma)
15 83.3 16.7 100 0.16 13 −11.4 1.2 27.7 1.2 6 19.8 1.7 8.2 −27.5 992 765 20.3
Phase I
(12.33–11.95 Ma)
7 90.9 9.1 100 0.09 5 −12.0 0.3 28.9 0.8 5 20.4 0.9 8.5 −28.1 1190 928 21.8
#M (numbe o samples o mesowea ); pe cen age o specimens wi h sha p (pS) and ounded (pR) cusps; pe cen age o specimens wi h high (pH) occlusal elie ; mesowea sco e (MS); #C (numbe o samples o
s able iso ope analyses on he ca bona e ac ion); mean δ
13
C(‰VPDB); s anda d de ia ion (SD) δ
13
C(‰VPDB); mean δ
18
O
CO3
(‰VSMOW); s anda d de ia ion (SD) δ
18
O
CO3
(‰VSMOW); #P (numbe o samples o
s able iso ope analyses on he phospha e ac ion); mean δ
18
O
PO4
(‰VSMOW); s anda d de ia ion (SD) δ
18
O
PO4
(‰VSMOW); Δδ
18
O
CO3
−δ
18
O
PO4
, mean δ
13
C
die , mequ
(‰VPDB); in e ed mean MAP (mm/yea ) ( om
Kohn [21]) wi hou (es ima ed MAP
a
) and wi h (es ima ed MAP
b
) al i ude and la i ude co ec ion; and in e ed mean MAT (°C)
DeMiguel e al. BMC Biology (2021) 19:6 Page 5 o 15

he niche occupa ion pe mo pho ype and locali y gi en
he a iables in es iga ed. MS and δ
13
C alues poin o a
equen inges ion o C
3
plan s in woodland o mesic C
3
g asslands. Mean MS o some indi iduals o mo pho-
ypes 1 and 2 om a ew locali ies ( hose wi h MS = 0.33
o 0.5) is also compa ible wi h egula consump ion o
C
4
ege a ion. C
4
plan s ha e ne e been documen ed as
an impo an componen o plan communi ies in he
Ibe ian Neogene (despi e being eco ded he e since he
Oligocene) [24,25]. Howe e , hey may well ha e been
p esen o some ex en in some a eas and/o ime in e -
als (e.g. ACM/C4-C1 and ACM/C5-C2 a ~ 11.8 Ma).
Discussion
Mic ome yx die a ACM
Ou esul s indica e ha he bulk o Mic ome yx die a
ACM consis ed o oliage wi h a pa icula emphasis on
o bs, dico s and woody lea es. The a ious mo pho-
ypes gene ally main ain sha pened, high- elie cusp api-
ces and low MS—a signal ha in o ms ha oods we e
o ela i ely low ab asion, as seen in ex an o es -
dwelle b owse s [26]. We ule ou a egula consump-
ion o ui s and/o seeds in ACM, as ee h show no
signs o s ong ounding o blun apices—and ugi o -
ous axa ha e signi ican pe cen ages o ounded and
blun cusps because o ip-c ushing [27]. These esul s
con as wi h some p e ious da a o middle and la e
Miocene Mic ome yx om elsewhe e in Eu ope, which
appea s ic ly ugi o ous [28–30]. Howe e , hese e-
sul s a e in ag eemen wi h he lea b owsing in e ed
o o he Mic ome yx [29,31]. The e o e, i seems ha
Mic ome yx, since i s oldes occu ences in he middle
Miocene o Eu asia, was capable o eeding al e na i ely
on ui s, seeds and so lea es, depending on habi a -
speci ic ci cums ances (e.g. ecologic niche pa i ioning
o ood a ailabili y). In he case o ACM, he unusual
seconda y c es s o he uppe mola s o Mic ome yx a e
compa ible wi h an adap a ion o hea ie eliance on
lea es and s ems, as seen in o he mammalian g oups
[32]—an ana omical ai ha is consis en wi h he
Fig. 2 Sca e plo o mesowea and δ
13
C alues (‰VPDB) o Mic ome yx. S ippled a eas show he ansi ion be ween C
3
-domina ed die s, mixed
C
3
-C
4
die s and C
4
-domina ed die s. Colou in o ms abou Mic ome yx mo pho ype (g een o Mic ome yx mo pho ype 1, blue o Mic ome yx
mo pho ype 2 and yellow o Mic ome yx mo pho ype 3), and symbol e e s o he empo al ange wi hin he sedimen a y sequence (diamond
o 12.38–11.95 Ma, ci cle o 11.90–11.79 Ma and squa e o 11.70–11.60 Ma). See Addi ional ile 1: Supplemen a y in o ma ion, No es 2 and 3 o
u he de ails
DeMiguel e al. BMC Biology (2021) 19:6 Page 6 o 15
oli o ous signal e ie ed om mesowea analyses. The
Mic ome yx om ACM a e he e o e he only ones in
which hese ea u es a e ecognised, p obably showing a
egional adap a ion associa ed wi h he pa icula i y o
hese en i onmen s. Mo eo e , Mic ome yx had a wide
die a y plas ici y han mode n Moschus, whose die com-
p ises mainly a bo eal lichens (a esou ce a ely
exploi ed by o he uminan s), o bs and woody lea es
[33].
Wi hin such a gene alised so , lea y b owsing, he e
a e di e ences among mo pho ypes in oo h wea and
iso opic alues h ough ime (Addi ional ile 6: Table
S1). In o he wo ds, he same mo pho ype beha es di -
e en ly when he empo al g adien is conside ed. The
less sha p and mo e ounded cusps o Mic ome yx mo -
pho ypes 1 and 2 eco ded om 11.81 Ma onwa ds (and
a signal o b owse-domina ed mixed eeding o some
indi iduals) e lec a mo e p onounced ab asion han in
olde specimens and indica e ha mo e ab asi e b owse
and/o some dus /g i -in es ed oliage was ea en—as ex-
an b owse s ha eed on lea y, so oods gene ally
main ain sha pened/high elie cusps [27,31] (Addi ional
ile 6: Table S1). Tha is, ab asi e b owse and enc oach-
men s by exogenous dus /g i con en was only sligh o
e en absen in he die o olde Mic ome yx, whe eas i
signi ican ly inc eased in la e o ms along he ACM se-
quence. δ
13
C da a suppo die a y in e ences based on
mesowea , as all Mic ome yx mo pho ypes depic alues
ha a e wi hin woodland o woodland-mesic C
3
g ass-
land condi ions, he eby indica ing a consump ion o
bo h so lea es and mo e ab asi e g asses.
Tempo al pa e ns in Mic ome yx oo h wea and s able
iso opes along he ACM sequence
The p ima e assemblage eco ded a ACM [5] includes
h ee g ea ape (d yopi hecine) species om di e en
gene a (Pie olapi hecus ca alaunicus,Anoiapi hecus b e-
i os is and D yopi hecus on ani)[5,10,16,34], he
pliopi hecoid Pliopi hecus canma ensis [35], p obably a
second pliopi hecoid unassigned o species [36]anda
pu a i e s em hominoid (Plioba es ca aloniae)[11], al-
e na i ely in e p e ed as a pliopi hecoid [37]. The dis i-
bu ion o hese axa is no homogeneous along he ACM
sequence, wi h Pliopi hecus pos da ing mos g ea ape
inds bu p eceding Plioba es [5] (Fig. 3a).
We show mesowea and iso opic da a o Mic ome yx
ee h in ch onological o de o he locali ies o a ime
span anging om 12.38 o 11.63 Ma (Fig. 3), o com-
pa e hei ch onos a ig aphic dis ibu ion wi h palaeo-
en i onmen al changes h ough ime. While emaining
wi hin a b owsing die a y ca ego y based on C
3
plan s,
di e ences obse ed in bo h mesowea and iso opic
alues among Mic ome yx specimens p o ide insigh on
he en i onmen al changes along he ACM empo al se-
quence du ing he la es middle Miocene. This is pa -
icula ly ele an gi en ha he lack o o he
palaeoen i onmen al p oxies om his a ea (such as
pollen o mac oplan emains) hinde s a mo e p ecise
econs uc ion o he ege a ion s uc u e and o he
cha ac e is ics o he a ious habi a s occupied by
p ima es.
O e all, he da a epo ed he e indica e ha ACM
habi a s became p og essi ely less humid and mo e
Fig. 3 Co ela ion o mesowea and iso opic alues o Mic ome yx a anged empo ally along he ACM sequence. aEn i onmen al phases,
sepa a ed by dashed lines, and s a ig aphic anges o he p ima es eco e ed a ACM based on occu ence in locali ies. bA e age mesowea
sco es (MS) by locali y o Mic ome yx mo pho ypes. cMic ome yx oo h enamel aw and mean δ
13
C(‰VPDB). Calcula ed a e age mode n
equi alen o die composi ion (δ
13
C
die , meq
,‰VPDB) and es ima ed MAP (mm/yea ) alues (wi hou [MAP
a
] and wi h [MAP
b
] al i ude and
la i ude co ec ion) a e gi en in pa en heses. dMic ome yx oo h enamel aw and mean δ
18
O
CO3
and δ
18
O
PO4
(‰VSMOW) alues. Calcula ed
MAT (°C) alues a e gi en in pa en heses. Locally weigh ed polynomial lines a e i ed o iso opic da a. Calcula ions u he include he 95%
con idence egion. Colou symbols a e o aw da a and g ey symbols a e o mean alues. See Addi ional ile 1: Supplemen a y in o ma ion, No e
4 o u he de ails
DeMiguel e al. BMC Biology (2021) 19:6 Page 7 o 15
he e ogeneous (o , a leas , a e cha ac e ised by a g adi-
en owa d less dense canopy s uc u e and mo e open
pa ches), as e lec ed by (i) a end owa d highe alues
o mesowea — om he sha pes cusps and low MS o 0
a 12.38 Ma (ACM/C1-Ee) o mo e in e media e (mo e
ounded) cusp mo phologies and highe MS a ound 0.5
a 11.60 Ma (ACM/C5-D1) (Fig. 3(b)); (ii) an inc ease in
mean Mic ome yx δ
13
C alues ( om −12.4‰in ACM/
C1-Ee o −11.1‰in ACM/C5-D1; Fig. 3(c)); and (iii)a
high a iabili y in he ype o ege a ion consumed— e-
ealed by a wide ange o δ
13
C alues in he younges
locali ies and uelled by he coexis ence o he h ee
mo pho ypes (see a change in s anda d de ia ion alues
in Table 1). The inc ease in mean δ
13
C alues may ha e
been d i en by wo phenomena: a shi owa d d ie hab-
i a s, including non- o es pa ches, o , al e na i ely, a
g ea e eliance on ui s h ough ime. A change owa d
a mo e ugi o ous die would ha e led o a sligh in-
c ease in Mic ome yx oo h enamel δ
13
C, as a signi ican
consump ion o ui s ul ima ely esul s in highe bioa-
pa i e δ
13
C alues [38,39]. The e is, howe e , li le ea-
son o suppo s ong ugi o y o ACM Mic ome yx,as
his is con adic ed by hei a i ion-domina ed meso-
wea pa e ns. This does no mean ha ui s we e un-
a ailable a ACM. In ac , all o he p ima e species
eco ded elied on ugi o y o a la ge ex en , e en i
wi h a di e en emphasis on ha d-objec eeding de-
pending on he species [11,40,41]. Howe e , a d op in
he es ima ed mean annual p ecipi a ion (MAP) alues
h oughou he sequence ( om ~ 1395 o ~ 762 mm/
yea o om ~ 1097 o ~ 575 mm/yea wi h al i ude and
la i ude co ec ion; Table 1, Fig. 3(c), Addi ional ile 6:
Table S1) suppo s he ac ha ui s we e p e e en ially
exploi ed by a bo eal, o a leas semi e es ial, species
ins ead o e es ial axa such as Mic ome yx.
Al hough he oxygen iso ope composi ion does no
a y signi ican ly h ough ime, he e is a sligh dec ease
in bo h δ
18
O
CO3
and δ
18
O
PO4
alues ( om −28.8‰and
20.3‰in ACM/C1-Ee o 27.4‰and 18.7‰in ACM/
C5-D1; Fig. 3(c, d)). Oxygen iso ope composi ion o ca -
bona e (δ
18
O
CO3
) and phospha e (δ
18
O
PO4
) ac ions o
oo h enamel e lec s δ
18
O o body wa e (δ
18
O
bw
)[42,
43]. Changes associa ed wi h δ
18
O
bw
alue mi o a ia-
ions in he iso opic composi ion o inges ed wa e , ei-
he h ough d inking o plan wa e (in he case o
he bi o es). Mic ome yx δ
18
O alues a e less likely o
a y acco ding o physiological ac o s like ac iona ed
wa e loss h ough he lungs o skin, since his is a ela-
i ely mesic en i onmen o e all, and he species a e
closely ela ed o e a na ow ime window. When con-
side ing ex inc mammals such as Mic ome yx, i is di i-
cul o assess he ype o wa e economy hey may ha e
had, due o he lack o mode n analogues. The ex an
sis e g oup o Mic ome yx wi hin he Moschidae is he
genus Moschus (musk dee ) [44], which inhabi s o es
and moun ainous pa s o Asia [33,45] and has a b ows-
ing die , al hough i also has he abili y o cope wi h
poo e , less nu i ious oods when high-quali y o age is
in sho supply, such as in win e [33,46]. Independen ly
om i s die a y beha iou , Moschus has been obse ed
o d ink wa e on a daily basis (P ikhod’ko, pe s.
comm.); he e o e, i s oo h enamel δ
18
O signal will
la gely be dependen on d inking wa e δ
18
O alues. On
he assump ion ha he wa e eliance o Moschus is ap-
plicable o Mic ome yx, mean annual empe a u e
(MAT) alues ha e been es ima ed based on Mic ome yx
oo h enamel δ
18
O
PO4
alues. They show a dec easing
end along he ACM sequence om 21.4 °C in ACM/
C1-Ee o 17.4 °C in ACM/C5-D1 (Table 1, Fig. 3(d),
Addi ional ile 6: Table S1). This end owa d lowe
empe a u es may be amed wi hin he g adual cooling
ha s a ed by 14 Ma a e he Mid-Miocene Clima ic
Op imum [47]. In he Ibe ian Peninsula, his long in e -
al, which coincided wi h he expansion o meso he mic
deciduous ege a ion and he ex inc ion o signi ican
dec ease in abundance o he mophilous e e g een
plan s [48,49], wi nessed an inc ease in he di e si y o
moschids [18].
P ima e assemblage composi ion in ela ion o
palaeoen i onmen al changes
Ou analyses u he show a luc ua ion in die compos-
i ion o Mic ome yx indi iduals, e ealing he exis ence
o h ee dis inc en i onmen al phases in ACM (Table 1,
Fig. 3(a–d)), wi h empo al pa e ns in p ecipi a ion,
empe a u e and a idi y ha ela e o changes in p ima e
assemblage composi ion.
Phase I
A i s phase anges om he beginning o he sequence
(12.38 Ma, ACM/C1-Ee) o ~ 11.95 Ma (ACM/C4-Cp),
whe e only Mic ome yx mo pho ype 2 is eco ded.
O e all, Mic ome yx main ained sha p apices, high- elie
cusp and low a e age MS o 0.1, and oo h enamel δ
13
C
alues o −12.0 ± 0.3‰(VPDB) (Table 1) ha poin o
he consump ion o plan esou ces om ela i ely dense
wooded a eas (see Addi ional ile 1: Supplemen a y in-
o ma ion, No e 4 o explana ion o he calcula ed δ
13
C
cu -o alues among di e en habi a s). Es ima ed MAP
anges om 928 o 1190 mm/yea (depending on
whe he a co ec ion o al i ude and la i ude is applied
o no ) (Table 1). In phase I, Mic ome yx oo h enamel
δ
18
O
CO3
and δ
18
O
PO4
alues a e he highes among he
h ee en i onmen al phases, wi h calcula ed MAT alues
eaching 21.8 °C (Table 1). Acco ding o Whi ake ’s
biome classi ica ion [50], es ima ed MAP and MAT o
phase I would co espond o hose o a opical seasonal
o es /sa anna (Addi ional ile 5: Figu e S4). This ag ees
DeMiguel e al. BMC Biology (2021) 19:6 Page 8 o 15
wi h he so -lea y b owsing die in e ed om mesowea
and indica es a humid clima e wi h ain all seasonali y
(al hough no ma ked) and he de elopmen o long-
s anding o es s wi h bushy and woody ege a ion [23,
31,51]. This ype o en i onmen a he beginning o he
ACM se ies, cha ac e ised by humid and wa m o es s
wi h a dense uppe canopy, is somewha mo e seasonal
han p e ious in e ences o ACM as a whole [52] and
would be sui able o he mul iple la ge-bodied homi-
noids—A. b e i os is,D. on ani and P. ca alaunicus—
eco ded du ing phase I. The abundance o ees may
ha e allowed hominoids o ea a di e se a ay o ege a-
ion, anging om lea es and so ui s (Anoiapi hecus
and D yopi hecus) o ha de and b i le ui s (Pie ola-
pi hecus)[11,41]. This also i s wi h he pos c anial
mo phology o Pie olapi hecus, which indica es an o -
hog ade bodyplan wi h adap a ions o a bo eal e ical
climbing [10,21,34,53]. Only a single Mic ome yx spe-
cimen om ACM/C2-A3 (IPS29396) displays mo e
ounded cusps and sligh ly highe δ
13
C and δ
18
O alues.
This specimen migh e lec he exploi a ion o mo e
ab asi e elemen s p obably loca ed along less humid—
bu s ill o es ed—pa ches in speci ic locali ies. I is no e-
wo hy ha only one pliopi hecoid is eco ded in his
i s phase (Fig. 3(a)) (Pliopi hecoidea inde . om
ACM/C3-B2 a 12.06 Ma). This immedia ely p ecedes a
i s sho pulse o dec eased humidi y as documen ed
by ACM/C2-A3, which migh explain he lack o g ea
ape eco d be ween hei i s appea ance in he se-
quence a 12.4–12.3 and hei mo e abundan eco d a
12.0–11.9 [5]. This shi ing clima ic pa e n a 11.98 Ma
owa d less humid condi ions migh ha e also in luenced
(p e e ed) ood a ailabili y (as seen in Mic ome yx) and
impelled hominoids o exploi al e na i e sou ces, espe-
cially as allback oods [41], no consumed be o e.
Phase II
ACM locali ies expe ienced a di e en en i onmen al
phase om ~ 11.90 o 11.79 Ma. The e was a apid in-
c ease in Mic ome yx pheno ypic di e si y and popula-
ion abundance a e ACM/C3-Ak (11.88 Ma), wi h he
i s co-occu ence o all mo pho ypes (a leas h ee)
being eco ded a ACM/C4-C1 (Fig. 3(b–d)). Compa ed
o phase I, om 11.88 Ma onwa ds he less sha p and
mo e ounded cusp shapes, highe a e age MS o 0.16 o
Mic ome yx and he inc ease in he mean δ
13
C(−11.4 ±
1.2‰, VPDB) (Table 1, Fig. 3(b, c)) a e consis en wi h
less humid and mo e open a eas du ing his pa o he
ACM sequence. The b oade ange o δ
13
C obse ed in
phase II (Fig. 3(c)) is cong uen wi h a phase o in-
c eased habi a he e ogenei y. Es ima ed MAP anges
om 765 o 992 mm/yea , whe eas es ima ed MAT is
20.3 °C (Table 1). The biomes o phase II would be in
he domains o opical seasonal o es /sa anna and
sub opical dese [50] (Addi ional ile 5: Figu e S4). In
he ligh o he auna eco ded a ACM [15], we conside
he la e in e ence as un ealis ic and mos likely a ib-
u able o a p ese a ional bias owa d d ie ecosys ems
[54]. Al e na i ely, highe CO
2
le els du ing he Miocene
migh p oduce a simila bias in biome econs uc ions,
gi en hei documen ed ela ionship no only wi h
highe empe a u es bu also enhanced wa e -use e i-
ciency and lea -le el p oduc i i y [55]. This “ o es
e iliza ion e ec ”, esul ing om highe CO
2
le els in
he Miocene, migh ha e esul ed in mo e o es ed en i-
onmen s han indica ed by es ima ed MAP and TAP
based on cu en s anda ds. Disce ning whe he such
po en ial biases apply uni o mly o he whole ACM se-
quence would equi e a axonomically b oade iso opic
sampling in selec ed ACM locali ies—as a e aging alues
om mul iple axa om he same locali y would a gu-
ably p o ide mo e obus MAP es ima es [54]. Ne e he-
less, we conside ha he palaeoen i omen al changes
eco ded by Mic ome yx iso opic alues h ough ime
a e a leas alid in ela i e e ms, e en i hei exac in-
e p e a ion in e ms o ex an biomes should be subjec
o u he sc u iny. The de elopmen o mosaic en i on-
men s (wi h he ea lie o es ed habi a s con aining o
he i s ime pa ial clea ing as new open pa ches) in
ACM migh ha e allowed he local e olu ion o new
Mic ome yx mo pho ypes (i.e. species) adap ed o mo e
open landscapes and wi h di e en die a y p e e ences
(e.g. mo e ab asi e o bs, sh ubs and o he ligneous
ege a ion ich in phy oli hs, and e en some g ass). Such
an in e p e a ion is ein o ced by he eco d a ACM/
C3-Ak o he bo id Te hy agus, a common aunal elem-
en in he mo e open and a id palaeoen i onmen s om
inne Ibe ia ha is o he wise no documen ed om he
Vallès-Penedès Basin [56]. Ou esul s suppo g ea e
habi a he e ogenei y, a he han a comple e change in
he palaeoen i onmen compa ed o he p e ious phase.
On he one hand, Mic ome yx mo pho ype 2 (wi h a in-
i y o humid condi ions) pe sis s in phase II wi h li le
a ia ion in MS and δ
13
C—wi h highe alues a he end
o phase II (a ACM/C4-C1, 11.81 Ma) likely indica ing
ha mo e ab asi e oods and/o some g i loaded oliage
was ea en a his ime. On he o he hand, he new mo -
pho ypes 1 and 3 appea o he i s ime wi h highe
mean MS and δ
13
C alues (Fig. 3(b, c)).
These changes owa d habi a (canopy) agmen a ion,
leading o a mosaic o o es pa ches in e up ed by
mo e open woodlands and maybe e en sh ublands,
would ha e ep esen ed a challenge o he ugi o ous
and p esumably a bo eal g ea apes om ACM—espe-
cially in die a y e ms (gi en he impossibili y o main-
aining a yea - ound supply o ipe ui s), and pe haps
also om a locomo o iewpoin (a leas o he highly
a bo eal Pie olapi hecus, gi en he need o a el ac oss
DeMiguel e al. BMC Biology (2021) 19:6 Page 9 o 15