ARTICLE

DOI: 10.1038/s41467-017-00827-7 OPEN Mass extinctions drove increased global faunal cosmopolitanism on the supercontinent Pangaea

David J. Button1,4,5, Graeme T. Lloyd 2, Martín D. Ezcurra1,3 & Richard J. Butler1

Mass extinctions have profoundly impacted the evolution of life through not only reducing taxonomic diversity but also reshaping ecosystems and biogeographic patterns. In particular, they are considered to have driven increased biogeographic cosmopolitanism, but quantita- tive tests of this hypothesis are rare and have not explicitly incorporated information on evolutionary relationships. Here we quantify faunal cosmopolitanism using a phylogenetic network approach for 891 terrestrial vertebrate species spanning the late Permian through Early Jurassic. This key interval witnessed the Permian–Triassic and Triassic–Jurassic mass extinctions, the onset of fragmentation of the supercontinent Pangaea, and the origins of dinosaurs and many modern vertebrate groups. Our results recover significant increases in global faunal cosmopolitanism following both mass extinctions, driven mainly by new, widespread taxa, leading to homogenous ‘disaster faunas’. Cosmopolitanism subsequently declines in post-recovery communities. These shared patterns in both biotic crises suggest that mass extinctions have predictable influences on animal distribution and may shed light on biodiversity loss in extant ecosystems.

1 School of Geography, Earth and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK. 2 School of Earth and Environment, Maths/Earth and Environment Building, The University of Leeds, Leeds LS2 9JT, UK. 3 Sección Paleontología de Vertebrados, CONICET−Museo Argentino de Ciencias Naturales “Bernardino Rivadavia”, Avenida Ángel Gallardo 470, Buenos Aires C1405DJR, Argentina. 4Present address: North Carolina Museum of Natural Sciences, Raleigh, NC 27607, USA. 5Present address: Department of Biological Sciences, North Carolina State University, 3510 Thomas Hall, Campus Box 7614, Raleigh, NC 27695, USA. Correspondence and requests for materials should be addressed to D.J.B. (email: [email protected])orto R.J.B. (email: [email protected])

NATURE COMMUNICATIONS | 8: 733 | DOI: 10.1038/s41467-017-00827-7 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-00827-7

arth history has been punctuated by mass extinction (and sampled) within a given locality or not. It does not explicitly events1, biotic crises that fundamentally alter both incorporate information on the supra-specific phylogenetic E 1, 2 biodiversity and biogeographic patterns . A common relationships between taxa, such as could be used to estimate generalisation is that mass extinctions are followed by periods of phylogenetic distance present between different species present at – increased faunal cosmopolitanism1 4. For example, the Early different localities. As such, it may be difficult or impossible to Triassic aftermath of the Permian–Triassic mass extinction, apply to a global fossil record dominated by singletons (species the largest extinction event known5, 6, has been considered occurring at just one locality), as is common for tetrapods. as characterized by a globally homogeneous ‘disaster fauna’ Moreover, the results are potentially sensitive to systematic dominated by a small number of widely distributed and abundant variation in taxonomic practice (i.e., ‘lumping’ vs. ‘splitting’) and – taxa1, 3, 6 8. Similar patterns have been proposed for the differential temporal and spatial sampling. Consequently, it may aftermath of the mass extinction at the end of the Triassic9. be useful to consider how closely related sets of species from pairs However, explicit quantitative tests of changes in cosmopolitan- of localities are on a continuous scale. ism across mass extinctions are rare and have been limited to Here we present a modification of this network model that small geographical regions3 or have not incorporated information addresses these issues by incorporating phylogenetic information from evolutionary relationships (phylogeny)2, 3. into the calculation of BC. Rather than treating links between taxa In order to test the impact of mass extinctions on biogeo- in different geographic regions in a binary fashion, they are graphic patterns, a method for quantifying relative changes in instead inversely weighted in proportion to the phylogenetic cosmopolitanism through time is required. Sidor et al.3 proposed distance between them (Fig. 1a, c). These reweighted links are that the spatial occurrence data can be modelled as a bipartite then used to calculate phylogenetic biogeographic connectedness taxon-locality network, specifying the distribution of fossil taxa (pBC). As with BC, higher levels of pBC equate to more (e.g., species) within defined localities (e.g., geographic areas such cosmopolitan faunas, with less phylogenetic distance between sets as continents or basins). The biogeographic structure of this of species from pairs of localities. By contrast, lower values of pBC network can then be quantified. Faunal heterogeneity (or indicate greater endemism, and increased phylogenetic disparity biogeographic connectedness, BC) can be measured as the between sets of species from pairs of localities. This method was rescaled density of the network—the number of taxa actually applied using an informal supertree (Fig. 2a; Supplementary shared between localities relative to the total possible number of Note 1) and species-level occurrence data set of terrestrial taxon links between them3 (Fig. 1a, b). Higher values of BC amniotes ranging from the late Permian to late Early Jurassic equate to increased cosmopolitanism (i.e., less heterogeneity), (c. 255–175 Ma; see Supplementary Note 2). A k-means cluster whereas decreases in BC indicate increasing faunal endemism or analysis was used to group taxa into ten distinct geographical provinciality (i.e., greater heterogeneity). This approach has been regions based on their occurrence palaeocoordinates (Fig. 2b; previously applied to assess regional changes in cosmopolitanism Supplementary Note 3). The sampled interval includes the within southern Gondwana across the Permian–Triassic mass Permian–Triassic and Triassic–Jurassic mass extinction events, extinction3. Results indicated a decline in BC from the late and the origins of key terrestrial vertebrate clades such as Permian to the Middle Triassic, indicating that cosmopolitanism crocodylomorphs, dinosaurs, lepidosaurs, mammaliaforms, increased following the extinction event. However, this study did pterosaurs, and turtles9. It is of particular biogeographic interest not include the critical immediate post-extinction faunas (earliest due to the presence of the supercontinent Pangaea10, which began Triassic), and it is also unclear whether this regional signal is to break apart by the Early Jurassic. Although barriers to dispersal representative of global biogeographic trends. might be perceived as sparse on a supercontinent, numerous This network method uses only the binary presence–absence studies have suggested faunal provinciality and endemism – data—i.e., information on whether a given species was present on Pangaea, perhaps driven by climatic variation3, 9, 11 13.

a

KombuisiaShansiodonVinceria TetragoniasKannemeyeriaDolichuranusRechnisaurus

bc

Fig. 1 Schematic illustration of network biogeography methods. a Simplified phylogeny of Dicynodontia. b, c Taxon-locality networks. Localities are indicated by the large, pale brown circles, taxa are coloured as in (a). Taxa are connected by brown lines to the locality at which they occur. b Rescaled non-phylogenetic biogeographic connectedness (BC) of Sidor et al.3. A single taxon, Kannemeyeria (yellow), is present at all three localities, resulting in a link of value = 1(solid black line) between each locality. c Phylogenetic biogeographic connectedness (pBC), as proposed here. Links (grey lines) between taxa from different localities are weighted inversely to their phylogenetic relatedness. Line thickness and shade is proportional to the strength of the link (and thus inversely proportional to phylogenetic distance between the two taxa)

2 NATURE COMMUNICATIONS | 8: 733 | DOI: 10.1038/s41467-017-00827-7 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-00827-7 ARTICLE

ab Synapsida

Prom

Scylacosuchus ore

C

i

P i ciegp m Ichibengops P y t

n P P

I o n r r s

c

i o

P o i r schor s r Olivie C t a youhsvanderrieti Lycosuchus s t C i

s G s r C i

t o i Ch h t risterog e o d S r y t e r s e p i e h ro st r g l o o n c Euchamber r Mos Aki a a rdosuc r d Annath e r r P i

u

s e g tho o P y a o g T A hu e

n Moschorh i i s p c n t a s

hy e rosu l i a l erp gn o a n g n a r

S k h d a o e l p a dnognathus par a r a

i g Ic o o a g S u h s c a n nosau n e o i a x r n n s g x e chowhai r i g s at d d i a na p u Theriognat m i t t r o t n c t m n Me a ch r i i AMN h t g g lex o cep s u n Ictido a c s e i d t u doc i s u i I o i h w i i n l i i e chus par e r I cti athoi o s y Mirot h y a at o o n t ct b n v s o s t t s hoid c u I o r s s si i i s o erapsidu t u i c h rv s la st a id a o a h I u o hus c b s e u S i r r ct u e u d ss isa u g p s a h id a t d d d f u r a e s n ti g u i n u s s u c u l spl i o l n r s u u s nyamadziens co i e a h e e Hofme de l o n e e os dos h l ilp Viatk e t s a en n idosuc u Noth aw c a at g s si e osuc Icti g r g r r an s u verus r l a i i i i u s e y u h s i u i s nburgensis d ri h s i s a

amps n i s l u p b HF s u SAM− ur sv hu n i d e n r e sbrevidenses de s r ic n K n n t l g e n p i n e s i mirabisia v e n i i s h hembutoma r v r ic u m n i u Tra o sc SAM i so i s Sca u nus sparv ss a r hus s e g a u e d ap l p d u a a l . g s i t i aty c u a x o i a n h o il nt a r a b c e h n ed a tsi a i l s sfoveatusus e i w o r e m t r r u r c s i ar n n sw k c o m t o nh os f f es n u n e p tu n sm o w l n aits a yn t eu u a e Choeros r a c renites ornament r g sq u u n u f i o i l d w ogomp t h c hoide chus o c satt s u p t e u g r c a e u g a a v hops intermediu avjusc AR ri l s o i r o o r so y i e n s ps e u i l nw c s ue osuc rh a s p n s u michus a s s a sdigitipeses p g a sp suchu d s e N y n p s s u s a n n m a a uc yr s athu s s loposau i a Regisau hus p aplatyc lo p s s g m r s d at s i o h Lycide t o s ov kitc u r l sc l s e o p p e nta r is s c o u s s s Mi s a r r o op a r s h o p t n o i hu i gto h p er inus p a m sl apatrichia er s h n y s k − s p u oy i otho i t o n e s b n u n h c no g a aatavusia o u t s s i c g sk n r n o ru sb e a cidens e r i o Tetracyno PK petri k r n r u or n s a t n n s n c Or B u e o r h s d o u s a o u h i s i g l e o a u r odonto o t o n a a c gc vus o e ar h P o n m l t e u a s t n v e u a y l r j d h u o n u hu e n n e o crog o u u o r t e l l i n Tetracyn o h s i e n n u s u c tr c t r r t n o o d slongicep l n baur r n u a hus orenburgens ni t e c s a o c p a b p i i c h s a s u u nu a e p o k i i c i p i e r s x l o t e a i e u a p t en c l K−K ingi 9527 n u a s do hodo ps ae i o c a u r r s e li u n o a u i l i o a ic n i e y u y e aen l u go − y o o r i n r ca g u cto l a a s i m c sa rimaev s e c go g s m Eri h s s t s s o i r s c s u a m v u u r ca g a r r g o a po i ap j i s s is s g i s d y r c au y c a te i a h u n ops lo y r r r s r i b i n a i w

8516 atrox Rubidgea C j o o n S ac g ou kovi A y r o r s ro S D j us si L y A a ay o v S o s r o NHC R r r i s us Urumchia lii su I y o o r o ri s n i h u m o mphod r e r A r a o rub L o u o a i g L o r r n s g Procyno ioi o ip e a s r i u L i mphodon ol H jarus y L n o s ia e v o ci N u u u n l l a r a h rus d ps L N r e io u e m ev d C i O 10990 ps l l n s ur r i i a n u u u Ant d i a l e nki u e i a s r nsan s G G o y r l t e i g l do m G n l r s sc c Proga azhenia co o des a G a a h a h r e ri en i A G e e o e o y o es wangw E a e k br n l n u a l e haras rd i S o p e u u y e s a l co d odon d d e h u c s k u i yo h i m a in Sa po h d r e y n u n n us A y s r o e h A c b e ye s i x u e osio on ng CLB gei Cyo lo g a u a m s m a a c Cyo o c i usa ur e i i i d h b t i Cy l r a m c a h p e cosu co a cha e i s e u e r r i n s l nstrictu ej a s C g ec m r a i i o erta parv s A a h l eo r r a au r a c e ncheyuensis a b e m i i a C cosuchus oche j icep A J a s e a o s n i a tus c ng h i s s e m m r i ia Cromptodon mamiferoidesl a ten J u h e yer y e c r o e s s b J an nn e u r a h er n d ge s a g b a o o e sp y e i r ynosaurus supp oensi I a ng c i r e e r o e sau uchus delProcynosuchusognathu sp. da i E t a r e p y e o o d nn u u e 4 S S u a d i ann y y c u ar l m k nn d m m ey e o m e g Baur nyo ri s S a nn a n l e ed Ga Bauria robusta hus bo is 4 S a o k a m e y ig n M k o si le ue is s ti P i s a e e m p i Nanicto Galesaurida n ab n o k i r e m Cynognath ru ocynus ilo Z i o ka n m a an Wa n d em i e gi i c i Antarc ca s i o me r ia s N lesau s n a R n a o nn e n s a s Dvini un vi S n l hn hu nn e r ne u r i Tita Pl ia S i h an a nn r a nk a T B si a c a nn y e s loot s v Si r c k k nn a n u mibia C y hrinaxodo o s gi a S li k a e y . n An a cyn R e u a a nne k a a aharpeae gr reus U r a e i k wuh ng e nogomp tycra lotr y r k a m i s nw sha s Diademodt sa rus p b R i ak e sp o a n s i a ac Do a r a k e m o i i tarctica Di ca Cynogna e X r b to d n t id op vi Pa a a e be d on n i ck a hu rgen p P a r n o ilu i urus i n d l p. k c u nie on frere ri lis P a nn on io n en u Zambia Trir sc o s P nn od do n ja s a gi la ein ma a aa d s i y g a cr h i i yn stus P a o a n 2 n a s Tr n l K h o n s i ic ar ej s i h h l nliorh nic s K haa s i tu o r u r n a n d ia v s u K s yr A ira o a selegansu d is S i n r r ss cc a s g S h a o ia ia ed vi th i ndesc Triracho d ter bi e e D h e a u S u a v i yi chodon ademon t na S in i n urus r r l a s s s s Cri on d nsis p t. ha c er r a T m u a n i i on S h o s u sh rr b s p s La thu inus gei s S in c e i sa a ec e i e i Pa Beis c e th n c g H s u s n i n Nan cra R i ni s ru a k c d e Sinognatng odon metaa t us sp o V n ra on u sc u d e i Scal y ra s t V i o N lu u r u sd n tz ol n s n cho odon sp us a u r th g li h ber don berryi g s s V et g u o ia on a c o o ha tidae T ng z s a ru us ma a e h Mand ua don o s p. o s au u r n b k l li Arct enodo g n u A UM a r n il tsc z M o n gia don . An t o a u g i s e s L lgnathus u s O s tr s g u r an u Namibi m od ros o s p o Trave s Kw y s t o sa h nam lu n ra e i e i Bor angwa su m phodonh wildi inde L y s tr don a a b e s tr a da Lua on mod sp . ro o ri o o ic rd p ave e us b L s t tyc d st s t a go n nd Ly y s n a ph u en r e e g olu . p a s od o r u e c B rsodon s angustif y t. L y u f e r ti ogomphodonrs mphoomph n gra o i L r u m c m u u hu c dw l i ni or a gwa o s u a a ir Gomph Mas o po zensi ungi ei s E m a o o a l o mi e Boreogom L T t n s s n oo a t Mass d cil Ji er p o o n sa g lo o o d uangwa sp. la a x to w lu n g do o americ d is P n o qu ri s omphodon n ns Da i a c od d n a a r s eto d rysdarons s v e n o h tr sb s hopsi u p n on ta D l o s u s s S o San e lemmyr k Vi e y n d p a pa li r si sc in s u th caleno do togna hirsc hl ii ic y o e i u a s o gna a D c il c n a a n ep o n eckeri D i ia e c m ra g m ntosuchu ta phodon sp. tt a lli s to b re s ti i a o s t he rid na D a do o o s n nbn u c hu hso n r m s lgin o e h Pro ruzo thus pasc i gei B o u u o d p e al dontoi socharpet d Si t re e d on c b h je on G hb oc o lo e us m s i p i ni lp rp ia l a o od l s u its h Exae tu M do ffersoa Ka k m h a u c e Exaer be Dadadon E te i ia h n h l i m c its enas irus n e k om a r ha u ty n s de nh I i no ep a n p h ta w i a l i re Ex rum b g G e la a e p a b l te s RuberoExa smacd rasil avia u n e sv c s u p t o i G lac g io d en is odonetodoaeretodon m on bes psoni al P u p in ce st op s is n s c i o h h io ru l a ru e n e ab ie e A y c u c us kod e s A retodon s n isalo S itc a h m a r g u leod d ro rale s h c a cy to b a ran am n b o riograndensisn dontes is K a s to s ar b k a ar a R h le o r a 50 nr n b a ng . i Ch Ectenin g nsis iriei i R e o id o b s p nw i on b oych Id d p o s a i slim tra en ral lo d on n ru is ini Chin a O ro t a o n lu ai i il s s Chini Lum tati tin T s tr n d o u n s n u Tr q rac o jo u e o d n b a s n u u wdhustic us r A us d n o o s mi o oo hi s cid Prob odo iq hyrhamphuion lunensis u e d on o ja f r n q u kuia fuzzi ae A d en o n n s sb ro e Pa o u o O u d n od a e e ic id c o ns d Al u e n h b ru d s a v nch T y a don the o rii O d e s u i m si c r i Pr nodoin an n e O u d g ia a lo ti cu k her og Ch odo O u in r s a us rc e c et oz ju kala q e a h l a r p. a ioh i a O s g p a rerent s w i ocynodProtho n na ni nsps y n e h n e o s s erp ri t qu otoninensnoro Da e wi c p af d ss a tr o hus K a te e i aa i e k o g o . s c is i to id snu Ira R er d et r do cus i K i te u is n to e m s io on io o on cargnia je s C is b o n id ili jathe gra n nd n nsp m bu d to fu c i d do bras en se C o o do n a ge am n K om n o o r d C El r ndi s ni . K y n d nmg i i h ium he e i i ic y o do s b n ali lio a od s lie s D ic n o u ru e Di P m t gu a tu n n y s ur n s ar achygenhe re d s i D ic s a o . tu i T thr ini r a ia is D ia s d p ta e a r a ium rn ib n n o lo sow c ub ith og m an ensissis ti N op ss ian i el My ga d p sar n O e na ust kerstd i y o p ondom o i Tr d t e ezi D m d o d ro wt s y s Ol lig o hus blu elo d o b e ka p ig o ith nti s e E y ct ep c e a o Oliky e d monuid ni Emym si rtia sn a yc m k p le ae roo e E o e p elic m at lis to K yp g h d s b o nf l ci s B a hu ok us tris o in Pr o im o on p a hy A ie Bi yen y n d mi s R s d n r at i nt not O s p riconet. o to rod o g . tsi is a eno tath luf hus e E c e od n p i es b rct her ligo e r ii st r onbs ha ri no ica th e nge mial o D ri te do d n e a T iu eriu ri ky is i P is ro io o is al ri T B my u p aj r te th d h tylo ri ien m mwelleh ns or P ris o io a ni B ty us is d th iodon w m la roomiii oc dontlod o un m P n do th n to b ho at o ther na a sp E n iodono un s o i ni Dianz he o n gnum s . E do th d on u m er id tida ium minnense i n io d op s v ize rium es E do th io s lu hu i ho Luf e n do th o gu t en Yun D me m indet E n o ypr te esc fs B in ngia e a o E nd h os ct lo o n n ng x xim r E n n oose Br tuca a eb l i ic us. rac ha de do o a Tri nodo ito ong ad a B nt bo o n itsi tzi sil raithetyl d e num a lo yn do a ol ith o o iro lica L o ic o wh h Br e don n na s C d yn sc ti riu rium bre mtrat ta ic ps s er a as mrio lon v a a Eo d o iru vis i ilo M b iro ra Eo le h job in jev do inicy el ga s li a ec a ta i lli Ade S n gr a e tre G al pus i isa e v u in quadrno an rmino vu le kov nv tzv no ph s s lo o do d s G a e a ya icu u basil co en i G nyu cai n m h tr an Bride no a nmais e mi ria get n en ric G e donngul is V le e a do cc af on t us c e U tsh ni o s s Indot d he a ri O mi m lu icu li w ri ro rig ris no ens eha n h anaum m neyi Su tra d p ilia ovi phu M e dori pto a aju ce orl ya org M EozPacrium prdo ni P iar o nsqn n an organucodone w nt sae T om de do n um u os y ap ri yo odo ic vi ai co trodono pardonanh a Anise s rm e cc do M i ni B ro losy ia gus bi s us n o datai ic tra b s ten nte Boc Morgahe rga vi M s on don eu o a Morga iku nuco v d Au od o on p aco n o ats us i y ph us adamT M n ucodonpen d o S otosya ne s s Br ega odo nuc g ons ni N mp ho eu 249 su s r ach n o ens a p on 0 go ficu to y zos t d p. P PV ru ni ven is zo Br tr am on pe is itanooph s ae ns Bra st a od au o ye T tan GS uru mag pr e B chyrod ch on lip ehl ri Ti R sa us on en sis rachyzost z oninfoyzost ru ens er o ur d um en i B ostr dne is i UFnte sa syo y g ne rachy od ro r A o us vija s o on couparmal sp don ae nteo e ne ga st 2 rod zo s A rcha o ussgi ther et. Kue o str p. A oph aur us sa n n i odo . 2 in os r lu e ind hio hneot Wo nform te S m au ha da ut n m zi le ep li egisc ch he Woers al aior U lemosoc ha m olo t.1 rium ut ia sp U pin ep rus m de DinnetKu p ers mir . 1 Ta inoc au lus in eh ra ia ab os ae yi neot ecursobu ilis Tap oc epha lid le he h tle h toc pha see H riu er r ri P ra ce us ni Na Co a m iumspis Ke no ur we i T kun nd dro ne pi osa s o its is rishu od oro co zor . Ta rm op ha ns s us l on paidon spadium wuium o h sw e si hal oth M osc op pen sis ep Ha eriu k M pscapou ien oc byli ra m as ni oschho sk vle c er Har miyidakot iensi os M sc op uis on nd tris am aens s Mo kr uscsva ros T iyid ein i oschntia nath uru ngi he ae in det s M a og sa o si T rotein .1 Aven hin alo rus l it homasia mooreus det. 2 elp ph sau us s Ha Tho ni D ioce cko luswhaedi fron i rami ma kol Cr ee pha term avi ssis ya Th siaan ai eb ce in es c ple via cl om Ri thio ops oid du e asi tiq i tru ion hal s s V Hu Ind mm ah ua S uth loide rmicu icto ast obaatar zoenseniahni Str ioceppha ia zi Arg riaco econo truth ce rusb rou e ento nodo d S thio au s le ela co on wi fia Stru os alu am nod n ina b e euter eph usp s Dyskron f equ lei D roc anin rox hu A it aria alis Tau c s fe nta ync sfal He odon so apino hu ule latyrh S tom nos indirum T osuc truc us p s s Braz tereos ylos fer cus itan ria l nu i il ter pa us T nke epha ralens osau num tag molu Jo coc usarcaus u ilis Mesrus onic s tyra pyg uch mirab sa tumid us S o s hus E osa npaulo um olybd eno suc uno urus e M temm no lsi i tosa ten nsis Es mme e kson Mil uru uidens te ectes c le s afr Es chyd alusja Br retta ican Pa ph Mill oom rub us llace 4 ilis s eros ia p idge Bu RB b nsi nsis Mille aurus nuferpl i T tia mira bone erge rosa exa NMurne suk eub Mill urusornfield B nitia ne sis E ero i iukse etia s tken udiba psis priatus N aburn a via Bol mu c Par rneti ta os s curso ei obu hiwe x nsis Bolosauaurus st ris Pr de c rda ene Bo riat Len ex mowillod losau rusm us obalop inus ricei rus ajo L horh rus p eps Beleb gran r Lop osau gic i a Belebey eym dis emur n lon son Bele axim L enodo x hop augus bey chengi i Lyca kyla si Be todunensis rpetos martin trai lebey He inus oons Austra ve tidorh rus b Microl lothy grandi Ic osau brinki novi i eter ris s mi s Hipp rus roma Del mckin thi posau oides orhynch zieoru Hip such tener sis Delorh us p m iarmo uchus ouen Acleis ynchus ciriscus B mos shank torhinu felli Biar us da us Colobomycters pte pholer i ranim limbat l Feeserpe otic us Ra odon ferox ni ton Dimetr codon illisto Lanth okla ter hena urus w anosuchushomensi Sp ontosa i Lanth s ecod ilmarth icus N aniscus efrwatsoni S tleriaw saxon Rh yctiphru e Cu saurus i ipae retus a movi telo lei s B osauru cudens Pan us bay anerge s ashkyrol stricuspi Haptod us bo xicanu eter b as dens phosaur novome Nycte hkyricu Eda osaurus galops roleter ine s Edaph urus me Em ptus laucosa kayi Bashky eroleter G saurus stiorum roleter m levis Lupeo us harde t Ma esensis thasaur i M croleter a Ian n schultze ni acroleter gilis Ianthodo s branso Bra poezicus rhynchu i dysaurus Cotylo us romer Brad seeleyi lorhynch ncocki 0 ysa oty ha Nochele urus bain C orhynchus saurus ale i Cotyl romeri Embrithosa xanderi elosaurus Bu urus schwa Ang urus tecton nostegos a rzi Ennatosa ensis kokanensi saurus tex Deltavjatia s Tricha i Honania c vjatkensis asea broili omplicidenta C r dolesorum Welgevo ta Oromycte i nden pareias aleops camp p Parasau aur Oed i Arga rus geinitzi othyris parkey na parieasaur E s scholtzi N indet. 1 Heleosauru ri anoparia luckho nosaurus rome Provelosa ffi Mese giceps urus americanus Elliotsmithia lon Anthodon serr don agilis Pumilio arius Varano pareia pricei Watongia meieri Pareiasuchus nasico s brevirostris Pareias rnis Varanop uchus peringueyi osaurus wellesi Argana par Aer ieasaur indet. 2 Aerosaurus greenleeorum Shihtienfenia permica elcobriensis San Ruthiromia chuansaurus pygmaeus Pyozia mesenensis e Pareiasaurus serridens Archaeovenator hamiltonensis Scutosauruskarpinskii Stereophallodonciscoensis Obirkovia gladiator Ophiacodon mirus Arganacerasvacanti Varanosaurus acutirostris Elginia mirabilis Archaeothyris florensis Owenetta rubidgei Tonganosaurus hei

r Barasaurus besairiei Amygdalodon patagonicus Ruhuhuaria reiszi Barapasaurustagorei Saurodektesrogersorum Kotasaurus y amanpalliensis Candelaria barbouri Isanosaurus attavipachi Owenetta kitchingorum Vulcanodon karibaensis Coletta seca Vulcanodontidae indet. O Sauropareion anoplus hmdenosaurus liasicus Tazouda Kitchingnathus untabeni saurus naimi magnidentis Pulanesaura eocollum Pintosaurus Gong anthosaurus ignatjevi xianosaurus shibeien Pha us Blikanasau sis

Phaanthosaurus sim rus cromptoni a C edectes perforatus hinshakiangosauru Thel gensis Lamp s chunghoensis lodon dongshen lughsaura dharm Eumetabo a vjatkensis Camelotia bore aramensis Tichvinski Les alis inskia jugensis semsaurus sa Tichv tatus Antetonitru uropoides ophon rariden s ingenipes

Timan us beltanae Melanorosa

Lasasaur s Melanor urus thabane r ycephalu osaurus re nsis olodon bath enis Sefap adi Eumetab phon spinig L anosaurus Terato ressus eonerasau zastronens rpeton opp Aar rus taque is Palaeolatitude Region hele eps donyx trensi T on trigonic X celestae s Procoloph ianus ixiposauru rus ordos s Mussa s suni taedrusau n asiaticu urus pata Pen ocolopho don Anchisa gonicus Neopr onia bolo Sei urus poly Halig onia sp. taad rues zelus Halig m Chuxio si dutoitoru ngosauru honodus odon Yunnan s lufeng P glyphan Ji osaurus ensis

oloparia armatus ngshano huangi

Sc saurus um Grypo saurus a lero rtin nyx xinw Sc ron lace i Mass african aensis ptopleu s shedd ospond us Le ognathu on Masso ylus ca Lib a caliod Pra spondy rinatus oturni s sp. dhania lus kaa S gnathu ri Leye gracili lae Hypso s fenne saurus s gnathu tritus Adeop maray Hypso ncon Lu appos ensis elepho sterni fengos aurus Th n lilien jcii Colora aurus mogna oiodo on kre Gl disau huene i Anom moiod sae acialis rus bre i Ano a tere Sarah aurush vis rocolin bruma I sauru amm P urgia tica gnavu s auri eri Orenb nigma Gy sauru fontan

rgia e ulae posa s rach alis

renbu jmesc us Eucn urus elis O s ma aen E emes sinens P ape s am toni ucne aur is K Kape ben Rio mes us fo apes nsis jasau auru rtis K omie t. Jak rus i s ent pes k inde P lapa ncert axoni Ka idae lis lateo llisau us s orhin ustra . Pla saur rus as Capt tus a s sp P teosa us e ymm uroric rhinu s late urus rlenbe etric Sa apto ticep U osau eng rgie a 1 la e n 2 C s na ru lh si hinu gnus R ysaur s gra ardti s aptor ma uti uehle us t cilis C rhinus s ag Plat ia olen apto hinu ndis N eosa bedh tinoi C aptor gra t. am urav eim C urus inde Efr balia us c ensis disa nae vi A aas royc ullin Mora auri veje sylo ia mi how gwor radis s ka nsis T sa nor dhu thi Mo phiu ane ta heco urus rii gom ush us Ar don yal cato ingto s rob na P cus tos ensi Ge us q iscu dia s anty auru auru s rhin thian s aca nsi C dra s p s an su Ro thyri nse ilis S hrom co c ereir tiquu Gan leo s ka grac atu og adu abd s Pa auru lis um P rna isau cus alor cos sce yor am lia tu rus um ola raeo r ma oi Eo pad pin no Petr A ato nenkkeli rap rom iqui vasi ven akh jae Pan tor aeu m Oro a iv vus nsis E pha lune s b olani ura live sha gia nsis arbe an osa s e cus i Cr no pro ren anth elur ravu ati ndr S yolo sau tos ai L Co au blem exa hi inos pho rus uros pro al inic Be au sa de el rus ania m D rbe rus urus guc Co au auti ania koi ilo ros tria e hiia pitis R uti hen s Ind pho au ssic llioti nus Wa Ra anc naio is ian sa rus us ukh lim ns Dra D uru lias sb or ene is Z co ilop s w sicu uru nect s c ens upa ven hos ethe s sa uro uru zin tilis L ys ato au ril gyz yp sa zor ua p. ilien auru r re r li Kyr H alle us q ss Lop ste s ge V aur usa ru ae G ho rnu rou nti lles saur osauenn s ojira stro s li gie Va bro nc nd nsi Po sa ph lien ri ola ala se ne p. P reb uru eus ste D eg uru reo roc a T s q a rni 3 us s ir 4 M a r s P o h u elen os usp au atu s odo mp ero ay anc aur nos ud nsi Sh k so po i sis gal os pa ica pe C ak es gna da e nc re gu ca ui i a e − au th ind M ala D na tea er C mp N rush ust et eg usungi ive au i oe os −B o ria . M aur oun s p ielh ini Co lop au ak lyo ss os Y uru a v ba tei C nn hys rusae t ke icu an sa m on hi oe ec is axon nsi s rep to ga rn a w ica C lop tic rh rizo s D on ea ste an lon ri oe hy ut C od ne od ala ro gu po ke FM lo sis o es ns ros P au ali a ief t. Pa N phy s elo ien is ce S P lin s nde n H si p. ph sis A me rus e i us C gu C s b ys Pa au da im s D oe ra UP au is ros uri tiss atu ra lop ptor 2 ri sp. ca sa la sl is Se co hy l 08 I eo s ru ns ri E gi ra sis ufe 9 hn chu au vie tle o sa pto k ng −2 ue su s o ch is Ta dro ur r h ay en 09 K eo eo rac a ns w m us a en si 0 −50 w s Da a a n t s hn ehn aclla ne u h eu ha iga ak ue Ku et ire a dic s C em all s lli n ata K hin h ebb in nsi A hin o ae m i e op ac r us de is lw de no urp S eg ma ur bri on P a sa sa hi M ga isa s av ni Sao lke u ur ta s ru s a ri reba ria ru us ra ika au ru erm se i S nj m s b c ha a os au b fra as ta ua H a ry hau B P r os us s e H ur ns err le an li hy nt ur ru luc a i err iko au e rie sm od ep do sa au rus on in G e s ru ras ns a us G hy er os u s ja fi G uai ras au s a is lli ip ak n sa bin s e u b a ru go ur D hit lla lo ro uru te is N aib as ur s rd ida eo ha s sa la cil r B ya a a us pr ill e W al ru a us ra au ie sa sa ur isc ice oi in P cep au ur g s ri Ta no s u ida i de no os sa or o e tis s au ru e hig t. la al ebbanri nd lev ri Sc a au ru s in u P h R va o C uell lo i A e ur ru s ca de ala p d oc m m tay ird i riz lid us s pa nd t. s oce Go n ku n n a g E o os o lu rr e ten n he c do o b an lli S m na a e fe in lar s 5 a d s i a u h n g i 6 l p u w g c u S r l g t ien s P S Do ino ur s c M ut s c us er e on s phenorh a ru m lus P a e au e h i ns i is ys hy os au s ti s e nid llo ru lid a is ol c ev s ru pe si A g e sa s osarri P ra l vo au s en llis br om ns u e so B C le ru ili a r N ic as c ru rn ur ni C vos u s nv ino A H to o s st i le sa ra o m r rg M sa ta n la i C vo b c pe 0 K e U u x do wl le us us us m 9 i N ay nt K ru af ren er C ur ur e 78 n H e ini R s ric i a aur a s so L M nt an 141 co a sis os os tum d yi H yc U a n nu v vos v c MKhu jo s et o K he H 6 s s le le e e A s tu la L e rh te et 1 or C C Cl s S us y s a es rod in R r er s us s ur u m s E o u U od od r u jimmrob ig F oc th o s A o o au aur sa ra s nsi s S M u o nt a 1 n n s vo e la la oe N rs sa o ng 0 to to vo os e ip a a id ns i L to H o u sau u 0 sa sa e ev Cl v ym m ej lensie s a rm rp ru s u u no c ig tid P q be C a s ru tid ri ri Cl Cl e le S on ha a ye i is ui U r d s e d da h e d c l le n ri S a n rg P vu i tu ns e P ta za s n to A ac n ta ia s ag ck i Sp ru ita ne i I g i os sa 23 n nd pa hui ot ring e an gn n sa a u da 3 o i e a n sau r sp s T o o u u ra n 8 st t. Z o o en a hm si i e to s ru ru g ic d v h p s le n ni E c s ph s s ve e u le p s ru re a s D ag hno a i m n rs s no C S eu s a ss i ti u ty ag e e h he h sau ru s a ns is R io le s ru s u r z oe p p o u u b e r ii O uhuhd B a s c d tii ue ki s ro or sa jim s n lf o u f ro o l no st u obo i P tis ru a ru ra m e ae y rot iro s m ua s d llo S e s si s g h ns C ma P a e e uy u an e id tr C u s n s ili a i ig s id n f m p n e p. L ile ifi h S c S m s lle s n e rto s e to h s u s e a i y il a n e J e roc u n s p s is E tu a d lkSle to e l s A c c m oc ru us ro lo n i A u n u F s b sa li is la r u r t e hy in s c g ru o a ra o Ma c a u ys h g t. L i o u s r ile ur c uri r ocne a is a n c ta e S e lis e ta e s id h d P r M rd is a ra saa d is w a lo l o st a io c a rd yt lo v in s M a isu u p i s po S u n a b a n e e n s D a lto ru h i l i rid M o b rry Ta h ia a e uu s i r p chu s y twe e les g o a T c n d ld q uu s D ro a u si n 78 n g u ra ia i e ti s s k s n s au a n yt r e tf n ic aa D ro m u s o b s is L a Q n h ro a p h L ro m o c e n a is ri L n a bu p s icus la a m o m h lg a g l d e T ta ro us n s d e ii L g m er u in dm w d d st rus o u r h W a e o s e e w y us y u he c e a c ss s g r m e o l n ix in a n a p s h b a u i D a e p ero ro n ill s tus i H ug a s o u p o tr ro s C o le rp e n g o i A to tr e ro g o a n m ri r e to r e s s h t n h a re u s C a y s n n gi e n ATo y p s lo ic l a r ne u i D a m g r to id g g s S m n o ny l s o tt s i m p n h n rom o is A tr s h e b tt D im y a a a a r s a u ru ic u eano e i P i p l t m c e s ii rota y T u askw b c ll n i m o y o hus ha in e P n he riadoa g e as a o w n E a o rp lo g p ri a s a n c j c w u ra r h g n h n det Ta op l h d y rus s o ti S d p p o n a b o a r e o a r u s u s r d E MU i s h d a th a rh re . st m d m h u r sbrob ar s A u m i o o t o nt y n y a n n sa h u hu p d o c d h i n s n s c a c hu a va e s r A rc r ep o n o d he c a P ze ru to u s u c si ic u B t im 691 p ma id e h is T e ho s o s ol t si a i u ic h h n s ns i A p p o h o u e r n y i C e s o o o we e d au r o n p s w w a e os y r tr rp a c sz li i i s e ril i te o o s t le tt R a g ia d 2 d lu r a s n h t p ilo o s n h k u i P a v a a h 5 o s in o s d o ra T S r es H ru ru a c c k joi i m d c od n n itte ic e d e T N ri o a ie s A e e ra ra t pu tra y u t n T M au a yn t n v d i i P us t ti o y o ra x li s . au h s o a o s S e c m d c lu n rd ub ze s s r d n r n re in o u a o s n o o e b e e C c tr d s r z o i A h s ncer l o ia o a c d c o ii hy c es p x id e fa s M a e n s schty a r s n o hus m a y s e u s D l r d d a c l l o en i n ar d hu n v h i i e s o a a u ty u la m E y c nchus c o x sp t s y E ianc s o m c l sw v n an n r y o a n e ni O y c t ru u e e pt fe y h rhy so y e n x n o o a o ty y s s Rh h h g en e kl P p e suchuc lu s cc o ld r C lo r to ny g ki do p. P rt n hun u h lu f a il n i o Me o n o g o ti oc r i ro h s s z ili p d h e in e c t o nei y I P r o e c lu a s l i i l Lan s l n s g ns e e H o to s u r st b c m u a a lu s nau B od u o o l S r t ma g o r r n i a e mm hu s n x . P emo o s u o c u is b e en s r t F d n d G p . P to s u c o s a ff t e n B A e o e . la c hu s y a a S ita do d nhu hu p . K i s u to a dy v l r t ll s s ync p e o p s p L la ty p uc c h a o i u i i i h a a e ap n s ns T i a o r hu u sss u l r ni s s s r b d p d o ns si ta y ty o sh u r iai ll i i o ap d e n is a D e e o g t hu c u ice l m o a o d do o ns s e D rg n o sr sl r d d r p e o e d r s i ib rr n g to hus n u e o a d nen o i R i e o t n a s smi a p Isa y r ro p p e a d en e s m S b o a t u ic rm u d e p pe d a a p i o t s s s a h chu u f e s e e y o pe e r n o a D p e h t u t h g e T p p r d d a p a fun o i S d t ri u h us h b c t. Hy y H o d nju ie x r h h ro s c c u c a h e o n y e ro da a m l o s i He a o o r u h r to g H p o d a a br u i H m e o s hu ss slma d rg l H e y e er r sa r r . l n n c u n b e 9 10 ro n e t er P t s u so i p p e ro ta ad rg e p e o i d hu sl su i er n H y e p n c s r i Ca s p c o a u c sw pe e c s bw p e u fe s s p o s c hu a n t a H y o do e riassicus nd C T o v sge i i is y Hy d pri ow r e p e s u e hu l i b H yp e c s a ss us i i s p a c i H a e p a x uan is e T M a r u e r u c n s m i eo aal us t ic i P e n d r o u h se torh l ie g p H a r r r le s n P le u o s c c a a r k ot o ua parhu sg v i o r h hu u ac e s2 Zi in a d o u g c chu n e e is P a 7 s f s u hu to ff c d o ka n R o l e e sac . y r m r B a Pr ja u hu us y oss h s L o o to 337 te x s u c s r la il i S o sa s su r ne n hi a s c hu sc s i n o r e u c so T u s t n s sr p q p s c m o u o ur a e l c F u o c hu g p a y e p i i y ro hus a r s shu tsc i us i p tos o a e u h . h v p y s a p n o r F k is s s 2 a s ra u a d n e e c te u Dag lem er u o r on r u o W y t o s a d nus a is u r o s t y o H a T u o ji h s sis s D u u u ru m o a l u s H to so e u ta P s u o c c s l s e s N Ka s er c wa u i ssic n is A c ch li un g ne a t a auru r fr ag n a e o c s h hu ss e n u n sm Pr he u o e s s B r in hu i to Pro ro o s hu re r a n Yo r o c a h u u r e c li s a r o s m r u S g la g u i a s ct r i te au ia ai c hu ba a a c u s u c s s n hu s T P sm h s i u ens n i 54 S a s st s u m e us yo uc s n s P a t r u F s h s a c fr h c g d m s H ung r nas e i u k s e r h u s a e gu t e i s a u a a u a i sj e n hu i u a r i S r c c j r ir ir il v s c o o h a tu e r u ro c o s o o s l d s Pro A t c or hu s 417 s E e hu a i k e i Ch u C chus e i s s M t ro ho s u hus r s c a a u hu c m m e n i i a S a p s o d m d p s s s E s i S f s u u c f o u cha s n uc P o s s i N o f t s r e a i p m i su uy s ia p e hu i go a to s u su c e a n a s l o ssh ia ia e n is L il g t c h R t r k u n − p n p Po e u o r c n r e gu to u v i am s e H l i s s c h hus us s a i c a V chus a n n e e P ot o a m t i t c c tos han i a y s Xi vo n n u c hu chu u a e a t e u l u a h e s a M m B C u u e e K a o ni i s k o hu u k i Sa c h s rj ic n ha C o os o c c us s n n s s s r e F aur a emov en t si h A l p u o i m a A i H r o C s s h s q ax i i s m u osu w r o p k hu t s s hu s ac m f u n c Q t os c a p s d a lis arja q r T r e u u u s c u s c S hus o mo sa y i m o au ee u hr o G e e G z h u i g k u e a c as t a l c k s aca r nsis di T Y i n s sa c s s c u hus G eri s o o l i a p L s s r ha l ba gi D c o a u rus S a u t y h u ri ia o s D u s u hu u l rt u go s ra k s e voi s A o r s o r B f p i e h r i c at n l e i a n u s c l n p s r ve d a f s L i n g chus u hu f c i i a z C C z b od ei li i i S es rf n n e s u e . l ti r i Ch s u hu ro r r L c e o r l en f d E U a a e s i u c a r s 3 s a h e e c e o hu t te s p A Go a o au u l a en s c ja n ue omeri s Sc g l o r n c deccan l s o i aen s s o i e r n r r Ca e c i o s u 4 i r i i r o h c xmi i s o i N n l s s v n n u l ss n A m h i v a n e ny k ze s a i aci u S d n m u n si up a a a l on r a s r a s o lo s c u s a b r on n S o u iq s c k on s e i P r e h r e d e i n i r P e eoa a c g i T p i P u r g a uc i l s Dor lta a s isl u s T t i suchu a u s is u E in s n f us s A s c a sc a xp a n T a g t s i g e u l i i T t l h i a rr s a L o m a a s l Sha a Aspe astensi a rr t h w ii a t y t os a t c a k e a p c n i R ec a o t r e u hu a as l chu kami s R a o u h n ss ia chu hu y a e u c ra e i C f g o u u n G h a rv d s A R y a A C l us u s R p s a o r g s hu e u p o a odon A S N r g R e e n N u g arntnagracilis Parringtonia b u u u E a c E rx s ls r c i p h wa V u m p i r t c s t n s r l s e a o s s zh b us g a e p is l u s i et t o u S s un ii w c a u u rv r p o u ci n fi s o si uc elli lar i o s s s ct mu e o ua s s a 4 s hus i o i s n u eu e o s s o h i id s 0 e t a u b i o u e o n c r t e i d p e r s i a an a deri s u sc g d a s c c sst s bon a y c o a t l on s eo g a nsi e n mi b n t a e v i s d a su o 7 r y r os a c l i is s w t u xi i p a a n p t v a 2 h t pf i p o well n c u i n m l ch t t ol l i x h r o f k s w a u v ho b hu n w po n t u h sis i i o h nP a hu k g s che o o shun e a ag t e h o n p u s 0 e h m s o y i r a h s s D a 1 h n t ra hi a r u h o s xt a e e s o u egal e a e i p e n a x e us a en e u c u rfecta t s n r ch or s e r b e ne s s l i a m a n o au c u p d a l u o us z eyeri e s o M i Ha och an o u r os m or e en tosau s b egor e i o r o t c c ta i e s i n − e l t d s d i t r ki hu sm r m li t d t s c s u r orum s i e e r hus r e b O o u p n u a p h nus b 10 a t h s o i i i r d c s o suc p r y i a a r r a s x Ya Ja p n i s g m m ho u p mp r s s D m a us gr c doughtyi u r b a hus n l A 3 D m a s r c o p s a ty i sis s g u c n k an r l a i i suc u s k Gu i a o te t m u x r hu tt chus u ti r d sk r g a e s b g a x s e h s u i a o a n e t c a P su o sp s b u o n e r shortus a a s u u i anen 7 ra hu o s i r r e s b o l l o p y r o h s u us m sm n tto n e e il a r a u er u n r t s u r h me n s uc i r r A a a ax ea a p a dino al R i i g s u − c m T h s s s a ol id i i 1 u d c eorhinus im t s su n s W o C c en m urus k o o n r o ci s i g s t a w s t u l i c u w c c u n s o ru l cc i c n b h s o y c a roso l u s o a e hus u i x Pr s e l 1 e Pa e a s e a r chus m s d a e e f a i s och uchu e e u h n h p a r e co r u b h c re hus t e d ie h f p u c cc e u d T b sbe p i r s t u e m t c s u sl u r v k l u 3 q s a d u c s n hos u us g c o l li n storh o o ey a a s u h i e su u f on u hu y e ll Pa ja p a p asu p e h n s i n ns n hus u Rha u r ter s u e ro os ho os h u h p k n ampsa isc r c i p r e r s s t s s h t e n a P Paleor s sauru s p i i s Proto d u s u b pu s u h r i a d u s y c hus c o e u s n ae p o u u sto h u . n e suchu c a e h s e r h t i i sr o M s a i usuc u TU n u o ro pt u o cr a ss u so r sa i o o s i r h s re r i m to c s s n n r a i o s s m i s c a o s s tt s u s e e n u g s l i g v p g s a t y s T h Pa ch s h s u r o u c t h P h r e s o Angi M ha o g s s p u a och u p uti o s ck r ss l o a Par N o r c s u c ni hu p . c i Le o r c h t c r r a osu r o r Ch u o i o st a Ebra ra c t p i i a s u c T . k s h j o m a e li e s p s n e An hus li o a u R Nicro p u h p u r u r Ph p s i ud p tos . e e Pra so n i o d e a a gi o o c s B c u s a r c Smil l e s e g s s o i s r i a in o n i l p o o r n e opr r t r t x s t h u e i suc opr r a ro r u t i Ma r h u p l n h o i a m se r s i t t s h L i s e i r j su s a l An n e i p a s m e t r o e a i u e e r i s Le i r a p s P o o u o t a a n p t o aer d i l i a r p a i n r a y o e h n .

s . milo s t h d e e e o h is h r d e h c r i mi a R y M n n S r u c s c V e c a s h a

O S i a o u m s t a p h a c M . r

M i c . Ang Mac a p N s M M M a

o

M r olmscu gonzalezdiazi Koilamasuchus M e

a

h

c

a M −100 −50 0 50 100 150

Palaeolongitude

E

u

r

e p t i l i a

Fig. 2 Phylogenetic framework and biogeographic regions employed in this study. a Informal supertree of amniotes used in the analyses. b Triassic palaeogeography, drawn using the ‘paleoMap’ R package63 with additional reference to10, 30,with the geographic regions used as localities for the network analysis indicated as follows. (1) Western USA, British Columbia, Mexico, Venezuela; (2) Eastern USA, Eastern Canada, Morocco, Algeria; (3) Europe, Greenland; (4) Russia; (5) China, Thailand, Kyrgyzstan; (6) Argentina; (7) Brazil, Uruguay, Namibia; (8) South Africa, Lesotho, Zimbabwe; (9) Tanzania, Zambia, Madagascar, India, Israel, Saudi Arabia; (10) Antarctica, southeast Australia

Our methodological approach allows patterns of global opportunistic ‘disaster taxa’6, or both. In order to test which provincialism to be quantified, and the impact of mass of these processes drove observed increases in global pBC, we extinctions on faunal cosmopolitanism tested, within an explicit carried out additional analyses on subsets of our data. The first phylogenetic context. The results demonstrate the evolution set of comparisons was restricted to those less inclusive clades of relatively cosmopolitan ‘disaster faunas’ following both that exhibit high levels of survivorship across each extinction the Permian–Triassic and Triassic–Jurassic mass extinctions, event, thereby removing the influence of preferential extinction suggesting that mass extinctions may have common biogeo- and focusing on patterns for clades established prior to graphical consequences. the extinction. Among these taxa, a significant change in pBC is no longer observed across the Permian–Triassic boundary – Results (Fig. 4a), although the increase across the Triassic Jurassic mass extinction remains significant (Fig. 4b). The second set Global phylogenetic network biogeography results. A marked of comparisons focused on novel, recently-diverging clades, and significant increase in global pBC is observed across the and demonstrates very high levels of pBC for these taxa in Permian–Triassic mass extinction (Fig. 3). A gentle, non-sig- both the Early Triassic and the earliest Jurassic, significantly nificant, decrease occurs from the Early Triassic to the Middle greater than total pBC in both these and the preceding time bins Triassic. This is followed by a strong, significant decrease to (Fig. 4a, b). Comparison of recently diverging clades in all time minimum pBC values (and so maximum provincialism) in the bins recovers the same signal as that from the total data Late Triassic. A significant increase in pBC is then observed after set (Supplementary Note 5, Supplementary Fig. 5), indicating the Triassic–Jurassic mass extinction, in the early Early Jurassic, that variation in pBC is not a result of differences in average clade although pBC does not reach the levels seen in the Early Triassic. age in each time bin. pBC declines to levels similar to those seen in the Late Triassic by the end of the Early Jurassic. These results show no correlation with the number of taxa or regions sampled in each time Geographically localized analyses. To compare hemispherical bin (Supplementary Note 4, Supplementary Figs 1–3) and trends in biogeographic connectedness, pBC was also calculated appear robust to variance in time bin length (Supplementary for Laurasia and Gondwana separately. The signal from Laurasian Figs 3d and 4). occurrences matches very closely with the global pattern (Fig. 5a). Results for non-phylogenetic network biogeographic connect- By contrast, patterns in Gondwana diverge markedly from global edness (non-phylogenetic BC) of the global data set significantly trends in the latest Triassic, where pBC abruptly rises, and then differ from the phylogenetic results (Fig. 3). An overall decline in gradually declines through the Early Jurassic (Fig. 5a). non-phylogenetic BC is still observed through the Triassic, but In addition, pBC analysis was implemented on terrestrial differences between the Lopingian, Early Triassic, and Middle amniote occurrences from the southern Gondwanan data set Triassic time bins are not significant. In addition, no increase in of Sidor et al.3. This data set groups taxa at a geological basin, non-phylogenetic BC is observed over the Triassic–Jurassic rather than broader regional, level; as a consequence, this boundary. analysis indicates how pBC differs at geographically smaller scales. Biogeographic connectedness is lower in the Middle Global analysis of taxon subsets. An increase in global pBC Triassic than in the late Permian under both phylogenetic and across a mass extinction boundary may result from preferential non-phylogenetic treatments of these data (Fig. 5b); however, – survivorship of cosmopolitan lineages8, 14 17, radiation of the result is not significant for phylogenetic BC.

NATURE COMMUNICATIONS | 8: 733 | DOI: 10.1038/s41467-017-00827-7 | www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-00827-7

0.25 PTB TJB = BC

= pBC 0.20

0.15

0.10

0.05 Biogeographic connectedness

0.00 Lopingian E.Tr. Anisian Ladinian early Late Triassic late Late Triassic early Early Jurassic late Early Jurassic

Fig. 3 Results from BC analysis of Lopingian-Early Jurassic terrestrial amniotes. Results from both non-phylogenetic (BC, red) and phylogenetic (pBC, blue) analyses of global biogeographic connectedness are shown. Shaded polygons represent 95% confidence intervals (calculated from jackknifing with 10,000 replicates) for both the BC and pBC analyses. The Permian–Triassic boundary (PTB) and Triassic–Jurassic boundary (TJB) extinction events are indicated by dotted lines. E. Tr.: Early Triassic

ab 0.35 = Complete data = Continuous clades 0.30 = Novel clades

0.25

0.20

0.15

0.10 Biogeographic connectedness

0.05

0.00

Lopingian Early Triassic late Late Triassic early Early Jurassic

Fig. 4 Results from BC analysis of taxonomic subsets. Comparison of results for the data subsets across the Permian–Triassic a and Triassic–Jurassic b mass extinctions. Results for the entire data set are in black, those for less inclusive clades showing high survivorship in red, and those for the most recently diverging taxa in purple. Ninety-five percent confidence intervals, calculated from jackknifing with 10,000 replicates, are indicated

Discussion and variable time bin length supports their interpretation as real The Triassic represents an important time in the evolution of biogeographical signals. vertebrate life on land. It witnessed a series of turnover events Our taxon subset analyses were explicitly aimed at that resulted in a major faunal transition from Palaeozoic disentangling the impact of alternative mechanisms that could communities, dominated by non-mammalian synapsids and lead to this pattern of increased post-extinction pBC. Novel parareptiles, to more modern faunas, including clades such as clades, those diverging immediately prior to or immediately after crocodylomorphs, dinosaurs, lepidosaurs, mammaliaforms, and each mass extinction, were analysed separately and exhibit turtles9, 18. Our novel phylogenetic network approach helps to relatively high levels of pBC (i.e., increased cosmopolitanism place these major faunal transitions of the Triassic within a global relative to the preceding time bin) in both the Early Triassic and biogeographical context by allowing changes in faunal earliest Jurassic (Fig. 4a, b). By contrast, surviving clades, those connectivity to be quantified within an explicit evolutionary well-established prior to the extinction and extending through it, framework. exhibit no increase across the Permian–Triassic boundary and Our results demonstrate an overall decrease in pBC from the only a moderate increase across the Triassic–Jurassic boundary Lopingian to the Early Jurassic, but punctuated by significant (Fig. 4b). This indicates that the increases in pBC following each increases across both the Permian–Triassic and Triassic–Jurassic extinction were primarily driven by the opportunistic radiation of mass extinction events. This provides quantitative support for novel taxa to generate cosmopolitan ‘disaster faunas’, rather classically held hypotheses about the presence of a global than being due to preferential extinction of endemic taxa19. cosmopolitan fauna in the aftermath of and in response to these Recently-diverging clades in other time bins do not exhibit events2, 3. The robustness of these results to sampling variation elevated pBC (Supplementary Note 5) and there is no correlation

4 NATURE COMMUNICATIONS | 8: 733 | DOI: 10.1038/s41467-017-00827-7 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-00827-7 ARTICLE

ab PTB TJB = Gondwanan pBC 0.4 0.35 = Laurasian pBC = pBC = BC 0.30 0.3 0.25 0.20 0.2 0.15

0.10 0.1 0.05 Biogeographic connectedness 0.00 Biogeographic connectedness 0.0 late Permian Middle Triassic E.Tr. Anisian Ladinian Lopingian late E. Jur. early E. Jur.

late Late Triassic early Late Triassic

Fig. 5 Results from BC analysis of geographically localized areas. a Comparison of pBC trends during the Lopingian-Early Jurassic from Gondwanan localities (in green) against those for Laurasia (in purple). Ninety-five percent confidence intervals, calculated from jackknifing with 10,000 replicates, are indicated. b Results from analysis of basin-level terrestrial amniote occurrences from the late Permian and Middle Triassic of southern Pangaea, from the data set of Sidor et al.3. Phylogenetic BC results are given in blue, non-phylogenetic BC in red. Ninety-five percent confidence intervals, calculated from jackknifing with 1000 replicates, are indicated. Abbreviations as in Fig. 3; E. Jur.: Early Jurassic between pBC and average branch length in each time bin were dominated by subtropical desert28. Interestingly, whereas (Supplementary Note 6, Supplementary Fig. 6), indicating that this biome was more fragmented by seasonally wet conditions this result is due to abnormal conditions following each mass through into the Jurassic within Laurasia, it remained relatively extinction as opposed to being a property of clade age. stable in Gondwana26, 28. It is possible that this stability may have The global biogeographic restructuring of biological commu- contributed to the evolution of a distinct fauna in the southern nities associated with these mass extinction events hence provides hemisphere. Alternatively, however, this distinct Gondwanan evidence of the release of biotic constraints3, which would pattern may be a sampling artefact. Although the inclusion of have facilitated the radiation of new or previously marginal phylogenetic information allows the approach used here to groups, such as archosauromorphs following the Permian–Triassic incorporate more data than previous methods, sampling of latest mass extinction3, and dinosaurs and mammaliaforms during the Triassic and earliest Jurassic Gondwanan localities is relatively Early Jurassic20, 21. This highlights the importance of historical poor and uneven, leading to the low statistical power of results contingency in the history of life, where unique events such as within these time bins. In the earliest Jurassic, in particular, over mass extinctions have exerted strong influences on the subsequent 80% of Gondwanan tetrapod occurrences are from the upper – macroevolutionary patterns observed in deep time22 24. Elliot and Clarens formations of South Africa. Further evaluation The global pBC pattern recovered here differs from the more of this possible signal will require sampling of new Late Triassic geographically focused and temporally limited non-phylogenetic and Early Jurassic Gondwanan localities, particularly from India study of Sidor et al.3, which found Middle Triassic levels of BC in and Antarctica. southern Pangaea to be lower than those seen in the late Permian. Under our non-phylogenetic network analysis of the global Reanalysis of the amniote occurrences from the basin-level data data set, no increase in BC is observed across the Triassic–Jurassic set of Sidor et al.3 demonstrates that pBC also declines between boundary; indeed, no significant differences are observed between these time bins, although not significantly (Fig. 5b). Looking any consecutive time bins (Fig. 3). This highlights the importance more broadly, pBC trends in Gondwana differ from those seen in of including phylogenetic information in global analyses Laurasia (Fig. 5a). This is particularly evident in the Late Triassic such as that conducted here; without the incorporation of and Early Jurassic, in which a significant increase and decrease in phylogeny, aspects of biogeographic signal may be obscured. pBC is seen in Laurasia for each time bin, respectively, but not in The decline of pBC to minimal levels towards the end of Gondwana (Fig. 5a). the Triassic supports hypotheses of strong faunal provinciality These results suggest that localized biogeographic patterns and increased endemism within Pangaea during the early within Gondwana may have been decoupled from those seen Mesozoic3, 9, 12, 13, 29. The distribution of Late Triassic tetrapods – elsewhere in the northern hemisphere. This would corroborate varies with latitude9, 11 13, a pattern also observed in terrestrial previous work, suggesting the evolution of a distinct fauna, that floras9, 27. This is somewhat unexpected, given that oceanic includes massopodan sauropodomorphs, ornithischians, basal barriers to dispersal were scant30 and the latitudinal temperature saurischians, and prozostrodontian cynodonts as relatively gradient was weak28 in Pangaea during the Late Triassic. Instead, common taxa in South America and Africa during the the ‘mega-monsoonal’ climate of Late Triassic Pangaea28 would Late Triassic11. The occurrences of guaibasaurids25 and floral have driven provinciality of faunas through strong latitudinal and similarities26, 27 provide some links between South American seasonal variation in precipitation12, 13. Patterns of endemism communities and the upper Maleri Formation of India, although farther back into the Palaeozoic are presently unclear because the latter assemblage remains relatively poorly-known and the Lopingian was preceded by a poorly-understood period of sampled. The Triassic–Jurassic mass extinction was a global taxonomic turnover during the Guadalupian31. Analysis of older event19 and it is unclear why decoupling of biogeographic trends Palaeozoic time bins will be required to elucidate changes in within Gondwana should occur. Sampling within Gondwana endemism during the earlier history of Pangaea. during this interval is uneven, with the bulk of occurrences This background trend of increasing endemism contrasts coming from palaeolatitudes between 30–60°S (see Supplemen- sharply with the increase in pBC immediately following each tary Note 4). During the Late Triassic the 30–60° latitudinal belts mass extinction. This highlights the unique macroevolutionary

NATURE COMMUNICATIONS | 8: 733 | DOI: 10.1038/s41467-017-00827-7 | www.nature.com/naturecommunications 5 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-00827-7 regimes associated with mass extinctions24, 32, with post-extinc- continents as input areas10, 11, 13, 15. This traditional approach is potentially tion ‘disaster faunas’ being the result of the abnormal selective problematic on a supercontinent where, for example, eastern North American and north-western African localities were much closer to each other than to localities conditions operating in the wake of these crises. An increase fi ‘ ’ in southwestern North America or southern Africa. Instead, we de ned our in global cosmopolitanism, with a prevalence of disaster taxa , geographic areas on the basis of k-means clustering of the palaeocoordinate data has also been observed in marine invertebrates across for 2144 terrestrial fossil occurrences from the relevant time span, obtained mostly the Ordovician-Silurian33, 34, Permian–Triassic35, 36, and from the Paleobiology Database (Supplementary Note 3). Importantly, this — Cretaceous-Palaeogene14 mass extinctions, although these studies approach does not require or use any information on taxonomy or phylogeny it is solely designed to find geographically-discrete clusters of fossil localities—and thus, have not explicitly incorporated phylogenetic data. Similarly, it is fully independent from the subsequent network analyses. more generalized insect-plant associations show higher The data were binned at epoch level, with each epoch analysed separately to survivorship across the Cretaceous-Tertiary mass extinction37 avoid confusion arising from continental movements. K-means clustering was and, on the smaller scale, Pleistocene-Holocene warming resulted performed within R, varying the value of k from 5–15. For each value of k, the in a greater unevenness of small mammal faunas in northern analysis was repeated with ten random starts, with 100 replicates). Performance of 38 different analyses was then compared on the basis of the percentage of variance California . Our demonstration of a similar signal in terrestrial explained, and results were compared with palaeogeographic reconstructions communities in the latest Palaeozoic and early Mesozoic suggests through this interval10, 54 (Supplementary Table 3; full results are given as that mass extinctions exert predictable biogeographical influ- Supplementary Data 3). This resulted in the designation of ten discrete – – palaeogeographic regions that each represent localities for the network analyses ences. However, the Permian Triassic and Triassic Jurassic (Fig. 1b). Taxa were assigned to one or more regions as appropriate, yielding a events may be unique amongst terrestrial mass extinctions due to taxon-locality matrix for each time bin (Supplementary Data 4). the presence of Pangaea, where the perceived reduction in barriers to overland dispersal might have facilitated the devel- Phylogenetic network biogeography analyses. Non-phylogenetic biogeographic opment of high levels of terrestrial cosmopolitanism. Extending connectedness(BC) was previously quantified3 as the rescaled density of a taxon- the methodology employed here to other extinction events, such locality matrix, calculated as follows: – as for terrestrial faunas across the Cretaceous Palaeogene O À N BC ¼ : ð1Þ boundary, will provide further tests of generalizable biogeo- ðLÃNÞÀN graphic trends across different mass extinction events. These common trends observed in the fossil record have the In this formula, O = the number of links in the network (the sum of all values in a taxon-locality matrix, which will equal the number of occurrences in a potential to inform modern conservation efforts, given that the non-phylogenetic analysis), N = the number of taxa, and L = the number of current biodiversity crisis is acknowledged as representing localities. This gives the ratio between the number of taxa present beyond a single another mass extinction event39. Global homogenisation due to locality and the maximum possible number of occurrences (i.e., every taxon present human activities, such as landscape simplification40, ecosystem at every locality). Aside from whether a taxon is identical or not, no further – — disruption40 42, anthropogenic climate change4, 38, 42, and phylogenetic information is included using this method links are only considered 42–44 where an individual taxon is shared between different localities, and are all equally introduction of exotic species , represents a principal threat weighted. to contemporary biodiversity43, 45. Ongoing extinction will Herein, this method was modified to include phylogenetic information (pBC) exacerbate this42, 43 with a shift towards a more generalized by weighting links between taxa as inversely proportional to the phylogenetic ‘disaster’ fauna projected on the basis of current trends4, 46. The distances between them. Phylogenetic distances between taxa were measured by summing the branch lengths in millions of years representing the shortest distance observation of global collapse in biogeographic structure between two taxa. This was then scaled against the maximum possible phylogenetic accompanying previous mass extinctions, as documented here, distance (i.e., the total distance of the summed branch lengths between the two corroborates this and is of key importance in forecasting the most distantly related taxa). This scaled value was then subtracted from one to biological repercussions of the current biodiversity crisis. yield the weight of each link: the values of links between taxa hence vary between one (co-occurrence of the same species in two separate localities) and zero (when comparing the two most distantly related taxa in the taxon-locality matrix). Methods The sum of the reweighted taxon-locality matrix was then substituted for O in Eq. 1 Phylogeny. An informal supertree of 1046 early amniote species ranging from to yield a value of pBC. This method has been made available as the “BC” function 315–170 Ma was constructed from pre-existing phylogenies (Fig. 2a; Supplemen- within the R package dispeRse55 (available at github.com/laurasoul/dispeRse): tary Note 1, Supplementary Data 1). We used an informal supertree approach example analysis scripts are given as Supplementary Data 5 and Supplementary rather than a formal supertree in order to maximize taxonomic sampling, including Data 6. It should be noted that a given value of pBC will be a non-unique solution: species that have not been included in quantitative phylogenetic analyses. In the same value could theoretically be generated by many links between distantly- addition to the taxa included in the biogeographic connectedness analyses, this related taxa or by fewer links between more closely-related species. Disentangling sample included some stratigraphically older taxa in order to more accurately these possibilities is difficult. However, comparison of results with measured date deeper nodes. In order to account for phylogenetic uncertainty, 100 time- phylogenetic distances and number of taxa in each time bin indicates that pBC calibrated trees, with random resolution of polytomies, were produced from results are not merely driven by differences in the relatedness of sampled taxa, and this supertree utilizing the ‘timePaleoPhy’ function of the paleotree package47 instead reflect genuine biogeographical signal (see supplementary information). in R (version 3.2.3;48). Trees were dated according to first occurrence dates, Analysis of a simulated null (stochastically generated) data set indicated a with a minimum branch length of 1 Myr. predictable and systematic pattern of increasing pBC through time. This is due to the increasing distance from a persistent root to the tips through time, resulting in phylogenetic branch lengths between nearest relative terminal taxa becoming Taxon occurrences and ages. A global occurrence database of 891 terrestrial proportionately shorter. In order to compare pBC between different time bins, it is amniote species was assembled, primarily from the Paleobiology Database49, with therefore necessary to remove this tendency for pBC to increase in later time bins. the addition of some occurrences from the literature (see Supplementary Note 2, We achieved this through the introduction of a constant, μ, which collapses all Supplementary Data 2). Taxa were dated at stage level. They were then placed in branches below a fixed “depth” such that root age is equal to μ million years before the following time bins for analysis: Lopingian, Early Triassic (Induan and the tips. The introduction of this constant also alleviates problems of temporal Olenekian), Anisian, Ladinian, early Late Triassic (Carnian–early Norian), late Late superimposition of biogeographic signals that may otherwise occur. It means that Triassic (late Norian–Rhaetian), early Early Jurassic (Hettangian, Sinemurian), and pBC results reported for each time bin reflect patterns generated by biogeographic late Early Jurassic (Pliensbachian, Toarcian). The Late Triassic was not split into its – processes in the preceding μ million years, preventing these recent biogeographic constituent stages due to the disproportionately long Norian stage:50 53 rock units signals of interest from being swamped by those from deeper time intervals. from this epoch were instead assigned to either the early Norian or the late Norian A μ value of 15 was chosen based on the results of sensitivity analyses varying the (Supplementary Tables 1, 2). value of μ from 5–25 Myr in 1 Myr increments (Supplementary Note 7, Supplementary Fig. 7). Geographic areas. In order to conduct network and many other palaeobiogeo- This method was applied to the taxon-region matrix for each time bin, and the graphic analyses, it is necessary to identify a series of geographically discrete areas 100 time-calibrated supertrees, pruning taxa not present within the bin of interest (the localities of the taxon-locality network in the network methodology). These (effectively making each tree ultrametric) to calculate pBC. Jackknifing, with 10,000 areas are typically defined solely on the basis of geography (rather than shared flora replicates, was used to calculate 95% confidence intervals. This analysis was then or fauna) because the aim is to test faunal similarity between geographically distinct repeated without phylogenetic information to gauge the importance of phylogeny regions of the world. For example, previous analyses have commonly used modern on observed patterns.

6 NATURE COMMUNICATIONS | 8: 733 | DOI: 10.1038/s41467-017-00827-7 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-00827-7 ARTICLE

Taxon subset analyses. In order to investigate the processes giving rise to 14. Jablonski, D. & Raup, D. M. Selectivity of end-Cretaceous marine bivalve observed changes in cosmopolitanism over mass extinction events, analyses were extinctions. Science 268, 389–391 (1995). also performed on two taxonomic subsets. The first reanalysed time bins either side 15. Raup, D. M. & Jablonski, D. Geography of end-Cretaceous marine bivalve of each mass extinction (the Lopingian and Early Triassic and late Late Triassic and extinctions. Science 260, 971–973 (1995). early Early Jurassic), including only small clades exhibiting high survivorship 16. Jablonski, D. Geographic variation in the molluscan recovery from the end- < ≥ ( 20 species, with 20% of lineages crossing the extinction boundary). This Cretaceous Extinction. Science 279, 1327–1330 (1998). fl was intended to minimize the in uence of possible preferential extinction of 17. Vilhena, D. A. et al. Bivalve network reveals latitudinal selectivity gradient at geographically-restricted taxa. the end-Cretaceous mass extinction. Sci. Rep. 3, 1790 (2013). The removal of taxa during mass extinctions opens new vacancies in ecospace, 18. Sues, H.-D. & Fraser, N. C. Triassic Life on land: the Great Transition promoting adaptive radiations in surviving, often previously marginal, clades56, 57. – (Columbia University Press, 2010). For example, the Permian Triassic mass extinction is seen as a causal factor in the 19. Dunhill, A. M. & Wills, M. A. Geographic range did not confer resilience to succeeding radiation of epicynodonts58 and archosaurs3, 59, 60, and the extinction in terrestrial vertebrates at the end-Triassic crisis. Nat. Commun. 6, Triassic–Jurassic radiation as pivotal in the diversification of crocodylomorph61 7980 (2015). and dinosaur clades20, 62. ‘Disaster faunas’ will hence be expected to be composed 20. Brusatte, S. L., Benton, M. J., Ruta, M. & Lloyd, G. T. Superiority, competition, of relatively recently diverging clades, as surviving taxa diversify into broader 59 fi and opportunism in the evolutionary radiation of dinosaurs. Science 321, geographic ranges (e.g., ref. ). To test the signi cance of this, we reanalysed the – time bins immediately following each mass extinction, including only clades that 1485 1488 (2008). fi branched <2 Myr prior to or after the boundary. In order to ensure that the results 21. Luo, Z.-X. Transformation and diversi cation in early mammal evolution. – of this analysis reflected differences in the post-extinction bins as opposed to an Nature 450, 1011 1019 (2007). fi artefact of clade age, also performed analyses applying this filter to the other time 22. Gould, S. J. The paradox of the rst tier: an agenda for paleobiology. bins (see Supplementary Note 6). Paleobiology 11,2–12 (1985). 23. Gould, S. J. Wonderful Life. (W. W. Norton, 1989). 24. Jablonski, D. Mass extinctions and macroevolution. Paleobiology 31, 192–210 Geographically localized analyses. To atomise global pBC signals into (2005). hemispheric trends, pBC was re-calculated for Laurasian and Gondwanan areas 25. Novas, F. E., Ezcurra, M. D., Chatterjee, S. & Kutty, T. S. New dinosaur species separately following an identical procedure to that for global analyses. Finally, to from the upper Triassic upper Maleri and lower Dharmaram formations of compare global results obtained from this new method with the more localized Central India. Earth Environ. Sci. Trans. R. Soc. Edinburgh 101, 333–349 3 analysis of Sidor et al. , another set of analyses was performed following the (2010). taxonomic sampling of the latter. Terrestrial amniote occurrences from the late 26. Bernardes-de-Oliveira, M. E. et al. Mesophytic Gondwanan paleofloras from Permian and Middle Triassic of the Karoo Basin of South Africa; Luangwa Basin of Brazil and India: composition and paleoclimatical approach. Paleontol. Zambia; Chiweta beds of Malawi; Ruhuhu Basin of Tanzania, and the Beacon Cenários Vida 4,93–105 (2011). Basin of Antarctica were taken from the data set of Sidor et al.3. These data and the 27. Césari, S. N. & Colombi, C. E. A new Late Triasssic phytogeographical scenario 100 time-calibrated trees described above were then used to calculate BC and pBC in westernmost Gondwana. Nat. Commun. 4, 1889 (2013). between these basins for each of the sampled time bins. 28. Sellwood, B. W. & Valdes, P. J. Mesozoic climates: General circulation models and the rock record. Sediment. Geol. 190, 269–287 (2006). Data availability. All the data analysed in this study and example code are 29. Shubin, N. H. & Sues, H.-D. Biogeography of early Mesozoic continental available in the supplementary data files. tetrapods: patterns and implications. Paleobiology 17, 214–230 (1991). 30. Golonka, J. Late Triassic and Early Jurassic palaeogeography of the world. Palaeogeogr. Palaeoclimatol. Palaeoecol. 244, 297–307 (2007). Received: 20 March 2017 Accepted: 28 July 2017 31. Bambach, R. K. Phanerozoic biodiversity mass extinctions. Annu. Rev. Earth Planet. Sci. 34, 127–155 (2006). 32. Jablonski, D. Background and mass extinctions The alternation of macroevolutionary regimes. Science 231, 129–133 (1986). 33. Brenchley, P. J., Marshall, J. D. & Underwood, C. J. Do all mass extinctions represent an ecological crisis? Evidence from the Late Ordovician. Geol. J. 36, References 329–340 (2001). 1. Erwin, D. H. Lessons from the past: biotic recoveries from mass extinctions. 34. Sheehan, P. M. The late ordovician mass extinction event. Annu. Rev. Earth – Proc. Natl. Acad. Sci. USA 98, 5399–5403 (2001). Planet Sci. 29, 331 364 (2001). 2. Sahney, S. & Benton, M. J. Recovery from the most profound mass extinction of 35. Schubert, J. K. & Bottjer, D. J. Aftermath of the Permian-Triassic mass all time. Proc. R. Soc. B Biol. Sci. 275, 759–65 (2008). extinction event: paleoecology of Lower Triassic carbonates in the western USA. – 3. Sidor, C. A. et al. Provincialization of terrestrial faunas following the end- Palaeogeogr. Palaeoclimatol. Palaeoecol 116,1 39 (1995). Permian mass extinction. Proc. Natl. Acad. Sci. USA 110, 8129–33 (2013). 36. Rodland, D. L. & Bottjer, D. J. Biotic recovery from the end-Permian mass 4. Blois, J. L., Zarnetzke, P. L., Fitzpatrick, M. C. & Finnegan, S. Climate change extinction; Behavior of the inarticulate brachiopod Lingula as a disaster taxon. and the past, present, and future of biotic interactions. Science 341, 499–504 Palaios 16,95–101 (2001). (2013). 37. Labandeira, C. C., Johnson, K. R. & Wilf, P. Impact of the terminal cretaceous 5. Raup, D. M. Size of the Permo-Triassic bottleneck and its evolutionary event on plant-insect associations. Proc. Natl. Acad. Sci. USA 99, 2061–2066 implications. Science 206, 217–218 (1979). (2002). 6. Chen, Z.-Q. & Benton, M. J. The timing and pattern of biotic recovery following 38. Blois, J. L., McGuire, J. L. & Hadly, E. A. Small mammal diversity the end-Permian mass extinction. Nat. Geosci 5, 375–383 (2012). loss in response to late-Pleistocene climatic change. Nature 465, 771–774 7. Roopnarine, P. D., Angielczyk, K. D., Wang, S. C. & Hertog, R. Trophic (2010). network models explain instability of Early Triassic terrestrial communities. 39. Barnosky, A. D. et al. Has the Earth’s sixth mass extinction already arrived? Proc. R. Soc. B Biol. Sci. 274, 2077–86 (2007). Nature 471,51–57 (2011). 8. Payne, J. L. et al. Early and middle triassic trends in diversity, evenness, and size 40. Western, D. Human-modified ecosystems and future evolution. Proc. Natl. of foraminifers on a carbonate platform in south China: implications for tempo Acad. Sci. USA. 98, 5458–5465 (2001). and mode of biotic recovery from the end-Permian mass extinction. 41. Vitousek, P. M., Mooney, H. A., Lubchenco, J. & Melillo, J. M. Human Paleobiology 37, 409–425 (2011). domination of Earth’s ecosystems. Science 277, 494–499 (1997). 9. Fraser, N. C. & Sues, H.-D. The beginning of the ‘age of dinosaurs’: a brief 42. McKinney, M. L. & Lockwood, J. L. Biotic homogenization: A few winners overview of terrestrial biotic changes during the Triassic. Earth Environ. Sci. replacing many losers in the next mass extinction. Trends Ecol. Evol. 14, Trans. R. Soc. Edinburgh 101, 189–200 (2010). 450–453 (1999). 10. Scotese, C. R. A continental drift flipbook. J. Geol. 11, 729–741 (2004). 43. Olden, J. D., Poff, N. L., Douglas, M. R., Douglas, M. E. & Fausch, K. D. 11. Ezcurra, M. D. Biogeography of Triassic tetrapods: Evidence for provincialism Ecological and evolutionary consequences of biotic homogenization. and driven sympatric cladogenesis in the early evolution of modern tetrapod Trends Ecol. Evol. 19,18–24 (2004). lineages. Proc. Biol. Sci. 277, 2547–2552 (2010). 44. Sax, D. F. & Gaines, S. D. Species invasions and extinction: the future of 12. Whiteside, J. H., Grogan, D. S., Olsen, P. E. & Kent, D. V. Climatically driven native biodiversity on islands. Proc. Natl. Acad. Sci. 105, 11490–11497 (2008). biogeographic provinces of Late Triassic tropical Pangea. Proc. Natl. Acad. Sci. 45. Pereira, H. M. et al. Scenarios for global biodiversity in the 21st century. Science USA. 108, 8972–8977 (2011). 330, 1496–1501 (2010). 13. Whiteside, J. H. et al. Extreme ecosystem instability suppressed tropical 46. Jackson, J. B. C. Colloquium paper: ecological extinction and dinosaur dominance for 30 million years. Proc. Natl. Acad. Sci. 112, 7909–7913 evolution in the brave new ocean. Proc. Natl. Acad. Sci. USA 105, 11458–11465 (2015). (2008).

NATURE COMMUNICATIONS | 8: 733 | DOI: 10.1038/s41467-017-00827-7 | www.nature.com/naturecommunications 7 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-00827-7

47. Bapst, D. W. Paleotree: an R package for paleontological and phylogenetic 63. Rothkugel, S. & Varela, S. paleoMap: spatial Paleobiodiversity Combined with analyses of evolution. Methods Ecol. Evol. 3, 803–807 (2012). Paleogeography. Available at: https://github.com/NonaR/paleoMap (2015). 48. R Development Core Team. R: A Language and Environment for Statistical Computing ISBN 3-900 (R Foundation for Statistical Computing, 2008). Acknowledgements 49. Carrano, M. et al. Taxonomic occurrences of Lopingian to Toarcian recorded in the Paleobiology Database. https://paleobiodb.org (2016). We thank R. Benson, R. Close, D. Cashmore, and E. Dunne for discussion. This research 50. Muttoni, G. et al. Tethyan magnetostratigraphy from Pizzi Mondello (Sicily) received funding from the Marie Curie Actions (grant 630123 to R.J.B.), an ERC Starting Grant (grant 637483 to R.J.B.), and a Discovery Early Career Researcher Award and correlation to the Late Triassic Newark astrochronological polarity time fi scale Tethyan magnetostratigraphy from Pizzo Mondello (Sicily) and (grant DE140101879 to G.T.L.). This is Paleobiology Database of cial publication 289. correlation to the Late Triassic Newark astrochron. Geol. Soc. Am. Bull 116, 1043–1058 (2004). Author contributions 51. Irmis, R. B., Martz, J. W., Parker, W. G. & Nesbitt, S. J. Re-evaluating G.T.L., R.J.B. and M.D.E.: Conceived the research. G.T.L. and R.J.B.: Wrote new the correlation between Late Triassic terrestrial vertebrate functions as required for these analyses. D.J.B.: Compiled the data, performed the biostratigraphy and the GSSP-defined marine stages. Albertiana 38,40–52 analyses and prepared the figures. All authors discussed results and contributed to (2009). writing the manuscript. 52. Olsen, P. E., Kent, D. V. & Whiteside, J. H. Implications of the Newark Supergroup-based astrochronology and geomagnetic polarity time scale (Newark-APTS) for the tempo and mode of the early diversification of the Additional information Dinosauria. Earth Environ. Sci. Trans. R. Soc. Edinburgh 101, 201–229 (2011). Supplementary Information accompanies this paper at doi:10.1038/s41467-017-00827-7. 53. Kent, D. V., Malnis, P. S., Colombi, C. E., Alcober, O. A. & Martinez, R. N. Age fi constraints on the dispersal of dinosaurs in the Late Triassic from Competing interests: The authors declare no competing nancial interests. magnetochronology of the Los Colorados Formation (Argentina). Proc. Natl. Acad. Sci. 111, 7958–7963 (2014). Reprints and permission information is available online at http://npg.nature.com/ 54. Scotese, C. R. Atlas of Earth history. PALEOMAP Proj. Arlingt. Dept. Geol. reprintsandpermissions/ Univ. Texas Arlingt. 1,1–52 (2001). 55. Soul, L. C. & Lloyd, G. T. dispeRse: Models For Paleobiogeography. Available Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in fi at: https://github.com/laurasoul/dispeRse published maps and institutional af liations. 56. Schluter, D. The Ecology of Adaptive Radiation. (Oxford University Press, 2000). 57. Yoder, J. B. et al. Ecological opportunity and the origin of adaptive radiations. J. Evol. Biol. 23, 1581–1596 (2010). Open Access This article is licensed under a Creative Commons 58. Ruta, M., Botha-Brink, J., Mitchell, S. A. & Benton, M. J. The radiation of Attribution 4.0 International License, which permits use, sharing, cynodonts and the ground plan of mammalian morphological diversity. Proc. adaptation, distribution and reproduction in any medium or format, as long as you give Biol. Sci. 280, 20131865 (2013). appropriate credit to the original author(s) and the source, provide a link to the Creative 59. Butler, R. J. et al. The sail-backed reptile Ctenosauriscus from the latest Early Commons license, and indicate if changes were made. The images or other third party Triassic of Germany and the timing and biogeography of the early archosaur material in this article are included in the article’s Creative Commons license, unless radiation. PLoS ONE 6, e25693 (2011). indicated otherwise in a credit line to the material. If material is not included in the 60. Pinheiro, F. L., França, M. A. G., Lacerda, M. B., Butler, R. J. & Schultz, C. L. An article’s Creative Commons license and your intended use is not permitted by statutory exceptional fossil skull from South America and the origins of the regulation or exceeds the permitted use, you will need to obtain permission directly from archosauriform radiation. Sci. Rep. 6, 22817 (2016). the copyright holder. To view a copy of this license, visit http://creativecommons.org/ 61. Toljagic, O. & Butler, R. J. Triassic-Jurassic mass extinction as trigger licenses/by/4.0/. for the Mesozoic radiation of crocodylomorphs. Biol. Lett. 9, 20130095 (2013). 62. Benton, M. J., Forth, J. & Langer, M. C. Models for the rise of the dinosaurs. Curr. Biol. 24, R87–R95 (2014). © The Author(s) 2017

8 NATURE COMMUNICATIONS | 8: 733 | DOI: 10.1038/s41467-017-00827-7 | www.nature.com/naturecommunications