Supplemen a y Ma e ials o “Whole-genomes illumina e he d i e s o gene ee
disco dance and he empo o inamou di e si ica ion (A es: Tinamidae)”
Au ho s:
Lukas J. Mushe 1, The ese A. Ca anach1, Thomas Valqui2, Robb T. B um ield3, Alexand e
Aleixo4, Ke in P. Johnson5, Jason D. Wecks ein1,6
Au ho a ilia ions:
1 The Academy o Na u al Sciences o D exel Uni e si y, Depa men o O ni hology
Philadelphia, PA, 19103, USA.
2 Facul ad de Ciencias Fo es ales, Uni e sidad Nacional Ag a ia La Molina, Lima, Pe u and
CORBIDI-Cen o de O ni ología y Biodi e sidad, Lima, Pe u.
3 Depa men o Biological Sciences and Museum o Na u al Science, Louisiana S a e
Uni e si y, Ba on Rouge, LA, 70803, USA.
4 Ins i u o Tecnológico Vale - ITV, Belém, B azil.
5 Illinois Na u al His o y Su ey, P ai ie Resea ch Ins i u e, Uni e si y o Illinois U bana-
Champaign, Champaign, IL, USA.
6 Depa men o Biodi e si y, Ea h, and En i onmen al Sciences, D exel Uni e si y,
Philadelphia, PA, 19103, USA.
O e iew
He ein, we p esen supplemen a y Me hods, Resul s, Discussion, Figu es, and Tables.
We expand on and discuss analyses ha sc u inize he co a ia es o gene ee disco dance,
wi h he objec i e o examining le els o s ochas ic e o in ou da a.
Backg ound on gene ee es ima ion e o
High h oughpu da ase s can con ain biased gene alignmen s ha a i icially inc ease
gene ee disco dance, mimicking he e ec s o ILS o in og ession, and hus biasing
phylogene ic es ima ion unde s anda d app oaches (Doyle e al. 2015; Smi h e al. 2018; Zhao
e al. 2023). Fo ins ance, mul i-species coalescen (MSC) me hods ha u ilize gene ees as
inpu may ail i gene ee disco dance is d i en by es ima ion e o a he han coalescen
a ia ion (Xi e al. 2015; Meiklejohn e al. 2016; Sp inge and Ga esy 2016; Simmons and
Ga esy 2021). Sho e alignmen s wi h ela i ely ew pa simony in o ma i e si es (PIS) may
esul in poo ly esol ed gene ees due o a lack o phylogene ic signal in he da a, a p oblem
e e ed o as s ochas ic e o (Xi e al. 2015; Meiklejohn e al. 2016; Kapli e al. 2021).
Con e sely, u ilizing alignmen s o long loci ha con ain many PIS is expec ed o educe
s ochas ic e o . Some s udies ha e shown a nega i e co ela ion be ween he numbe o PIS
and disco dance be ween gene and species ees (Bu b ink e al. 2020), implying ha as PIS
a e added o alignmen s, s ochas ic e o is educed. On he o he hand, inc easing alignmen
leng h also inc eases he p obabili y o sampling mul iple independen ly so ing coalescen
genes (c-genes; genomic segmen s wi h dis inc b anching his o ies), which may ange in size
om ens o base pai s in leng h o deep di e gences o a ew housand base pai s o ecen
di e gences, and his is known o a ec phylogene ic in e ence (Doyle 1995, 1997; Schie up
and Hein 2000; Hobol h e al. 2007; Zhao e al. 2023). Ano he , mo e oublesome sou ce o
gene ee inaccu acy is sys ema ic e o , which is d i en by da a biases ha cause model
iola ions (e.g., GC con en a ia ion) (Rod íguez-Ezpele a e al. 2007). The e o e, dissec ing
biological and a i ac ual sou ces o gene ee disco dance o in e species ees mo e
accu a ely is a majo challenge in he age o phylogenomics, whe e mo e da a does no
au oma ically ansla e o imp o ed in e ences (Meiklejohn e al. 2016; Blom e al. 2017; Smi h
e al. 2023).
Me hods
DNA ex ac ion, lib a y p epa a ion, and sequencing
Fo esh issues, we ex ac ed genomic DNA using he MagA ac High Molecula
Weigh ki om Qiagen (Valencia, Cali o nia). Toe pad ex ac ions o his o ical s udy skins was
ca ied ou in a dedica ed his o ical DNA ex ac ion labo a o y a he Academy o Na u al
Sciences o D exel Uni e si y o educe con amina ion isk. Toe pad samples we e i s washed
in a b ie ba h o E OH o help emo e supe icial con aminan s, and hen soaked in H2O o
h ee hou s o hyd a e he desicca ed lesh o DNA lysis. Samples we e hen diges ed using
180 µL ATL and 30 µL p o einase K o each sample and incuba ed a 56º C o e nigh . DNA
isola ion was hen pe o med using he QiaQuick spin columns and ex ac ion ki om Qiagen
(Valencia, CA).
Sho gun sequencing lib a ies we e p epa ed o each ex ac using he Hype lib a y
cons uc ion ki (Kapa Biosys ems). These lib a ies we e sequenced using 150 bp pai ed-end
eads on an S4 lane o an Illumina No aSeq 6000. These lib a ies we e pooled and agged wi h
unique dual-end adap o s, and pooling consis ed o be ween 13 and 18 samples pe lane aimed
o achie e a leas 30X co e age o he nuclea genome. Adap e s we e immed and
demul iplexed using bcl2 as q .2.20. We deposi ed aw eads in he NCBI SRA (Table S1).
A e mapping eads o e e ence genomes using BWA (see main ex ), we used ‘-
doFas a 2’ lag in ANGSD o ensu e ha he consensus nucleo ide was w i en o each
polymo phic si e.
In an ea lie d a o his manusc ip we used “C yUnd” (NCBI GCA_013389825.1) as
e e ence o se e al samples(Mushe e al. 2024). Du ing e isions we ound ha samples
mapped o his genome we e mo e likely o clus e wi h each o he han wi h o he mo e likely
ela i es in ce ain phylogene ic analyses. This led us o emap hese samples o a di e en
e e ence, C. s igulosus (LSUMNS B-9577), in he cu en e sion.
Assessing genome comple eness
To assess genome quali y, comple eness, and edundancy o each assembly, we used
Benchma king Uni e sal Single Copy O hologs (BUSCO) e sion 5.3.0(Simão e al. 2015).
BUSCO sea ches genome assemblies and iden i ies genes p esen in single copy using a
da abase o known single-copy o hologs om a clade-speci ic da abase o genes. We used he
‘a es_odb10’ lineage da ase , which u ilizes 8,338 genes in he chicken genome. We used he ‘-
- augus us’ lag o ob ain nucleo ide sequences o genes. This se ing uses augus us e sion
3.2(Ho e al. 2019) o anno a e each assembly, and ou pu s ull nucleo ide gene sequences o
all comple e single-copy o hologs in as a o ma . This s ep was necessa y o ob ain ou coding
gene da ase o phylogenomic analysis. We also used sam ools o es ima e mean and s anda d
de ia ion o sequence co e age o each genome.
Alignmen imming
An ea lie e sion o his s udy ound ha BUSCOs wi h mo e Pa simony In o ma i e
Si es did no signi ican ly imp o e gene ee accu acy, implying a s ong signal o sys ema ic
e o in his da ase . We examined a hand ul o BUSCO alignmen s by eye and ound long
sequences unique o a small numbe o indi iduals in each alignmen . We suspec ed ha hese
we e d i en by misanno a ion in he BUSCO pipeline, and hus immed all loci using T imAL o
emo e si es in each alignmen wi h missing da a o mo e han wo samples. This seemed o
alle ia e he p oblem as we ound a igh e co ela ion be ween he numbe o pa simony
in o ma i e si es and RF dis ance be ween gene and species ees a e imming.
Phyluce ha es ing o UCEs om whole genomes and addi ional phylogenomic de ails
We ha es ed UCEs h ee imes o cons uc h ee UCE da ase s ha each a ied in he
leng h o he lanking egion a ound he UCE co e. These UCE da ase s included 100 bp
(hence o h UCE100 da ase ), 300 bp (UCE300 da ase ), and 1,000 bp (UCE1000 da ase ; in
he main ex o his s udy, we discuss only his UCE se ) o lanking egion on bo h 5’ and 3’
ends o each conse ed co e. These h ee UCE se s a e no subse s o one ano he , bu
independen ex ac ions o he same UCE’s wi h di e en lank sizes. To examine di e ences
among sex-linked and au osomal ma ke s, we also c ea ed wo da ase s o long UCEs (1,000
bp lanking egions), one o hose ha mapped o he Z-ch omosome (an a ian sex
ch omosome wi h idealized Ne o 0.75 au osomal Ne) and one ha mapped o au osomes (see
Supplemen a y Ma e ials).
To gene a e Z-linked and au osomal da ase s, we i s used he Phyluce sc ip s o
ha es all UCEs om each pseudo-ch omosome-le el genome. We hen made a lis o UCE
loci mapping o he Z-ch omosome by using “g ep” in UNIX o ex ac lines in he ‘.las z’ ile
con aining UCEs on he Z-ch omosome in each genome and “g ep - ” o exclude lines on he Z-
ch omosome (au osomal da ase ). This gene a ed new ‘.las z’ iles o each sample which we e
hen used in he Phyluce pipeline.
Fo all da ase s we used MAFFT 7.453 (Ka oh and S andley 2013) o align o hologous
UCE loci o downs eam phylogenomic analysis and il e ed hese alignmen s in o 75%
comple e (≥75% o samples ep esen ed in each gene alignmen ) and 100% comple e (all
samples ep esen ed in each gene alignmen ) alignmen s. Howe e , o he au osomal and Z-
ch omosome UCE da ase s, we only used 75% comple e alignmen s o maximize alignmen
inclusion in hese wo da ase s. This esul ed in en da ase s (Table S4). All alignmen s we e
immed using T imAL .1.4 (Capella-Gu ie ez e al. 2009) wi h a gap h eshold o 97.5% o im
ou po en ial BUSCO mis-anno a ions unique o a small numbe o samples (see Supplemen a y
Ma e ials). Each da ase was analyzed using he same phylogene ic me hods as ou lined in he
main ex , which esul ed in species and gene ees o each o he en da ase s.
Assessing o holog quali y and gene ee e o
To e alua e a ia ion in phylogene ic in o ma ion con en and s ochas ic e o among
da ase s, we es ima ed he numbe and p opo ion o pa simony in o ma i e si es (PIS) o each
alignmen wi hin each da ase o 100% comple e alignmen s (downs eam analyses equi e ull
ep esen a ion o all loci and gene ees). PIS we e de ined as a iable si es in he alignmen
whe ein each a ian nucleo ide is ep esen ed by a leas wo samples. To examine po en ial
e ec s o sys ema ic e o , we also es ima ed GC con en a ia ion ( a iance in GC con en
among all si es in each alignmen ) o each alignmen . To examine le els o gene ee
disco dance o each da ase , we hen measu ed he Robinson-Foulds dis ances (RF dis ance)
(Robinson and Foulds 1981) be ween each gene ee and he in e ed species ee (using he
in e ed species ee o each da ase , a he han one species ee o all da ase s). RF
dis ances we e quan i ied using he ‘RF.dis ’ unc ion in phango n 2.11.1 (Schliep 2011). To
es o di e ences in hese measu es among da ase s, we used K uskal-Wallis es s in base R
(R Co e Team 2023) ollowed by a pos -hoc Dunn Tes using he FSA package (Ogle e al.
2023). We expec ha , i all gene ee disco dance is associa ed wi h biological p ocesses, he
mean and a iance in RF dis ances among da ase s will be ela i ely cons an . I he mean RF
dis ance is highe o some da ase s, i could mean ha he e is inc eased e o in ha da ase .
S ill, as alignmen leng h inc eases, so does he numbe o PIS, bu also he po en ial inclusion
o mul iple c-genes. So lowe mean RF dis ances o one da ase could indica e con e gence o
gene ee opologies on he conca ena ed opology, a he han a educ ion in e o .
Gene ee disco dance a ises om mul iple sou ces, including ILS, in og ession, and
po en ial s ochas ic o sys ema ic e o . One way o examine whe he he measu ed
disco dance is biological o e oneous is o see how gene ee disco dance co a ies wi h
alignmen in o ma ion con en (an in es iga ion o po en ial s ochas ic e o ) o GC con en
a ia ion (an in es iga ion o po en ial sys ema ic e o ). S ochas ic e o a ises because he e is
no enough in o ma ion o econs uc a gene his o y esul ing in ambiguously o inco ec ly
esol ed gene ees. I he e is no impac o s ochas ic e o , one would expec no s a is ical
ela ionship be ween hese PIS and gene ee disco dance. The e o e, his analysis gi es us a
baseline unde s anding o a es o e o and which da ase s ha e educed e o . Mo eo e , i
allows us o look o de ia ions om his ela ionship (e.g., be ween Z-linked and au osomal
ma ke s) o see i one o hese da ase s has lowe a es o disco dance (RF dis ance) gi en he
same alues o in o ma ion con en . Thus, o in es iga e co a ia ion be ween gene ee
disco dance, PIS, and GC con en a ia ion, we hen an gene alized linea models using
in o ma ion con en (PIS) o GC con en a ia ion as he independen a iables and RF dis ance.
We compa ed AIC alues o linea and loga i hmic i s o each eg ession and chose he model
wi h he lowes AIC o each da ase . These eg ession analyses we e done once o each
da ase , and hen, o be e in es iga e b oade pa e ns o sys ema ic e o gi en bo h sho e
and longe alignmen s, once wi h he h ee UCE da ase s combined. To a oid pseudo-
eplica ion in he combined eg ession, we andomly emo ed eplica ed UCE’s om he da a-
ame such ha each UCE was only ep esen ed once (ei he by he UCE100, UCE300, o
UCE1000 alignmen ).
To examine i di e ences be ween UCE and BUSCO eg ession esul s could be d i en
by sys ema ic e o , we explo ed hese models again o BUSCO’s a e il e ing he BUSCO
da ase o GC con en a ia ion (by emo ing genes wi h GC con en a ia ion g ea e han wo
s anda d de ia ions om he mean) and o BUSCO’s esembling UCE da ase s ( emo ing
genes g ea e han wo s anda d de ia ions abo e he mean p opo ion o PIS in he UCE1000
da ase and below he mean o he UCE100 da ase ). The idea he e is o a emp o eplica e a
da ase ha looks like he combined UCEs in e ms o PIS. Finally, we combined bo h p e ious
il e s o make a s ic ly il e ed da ase .
Es ima ing he empo o inamou di e si ica ion
Fi s , we il e ed he au osomal da ase o include only a single membe o each axon
(mono ypic species o subspecies) and c ea ed a new 100% comple e au osomal da ase . Nex ,
we gene a ed new gene ees o hese alignmen s and adap ed sc ip s om a p io s udy
(Mushe e al. 2019) o quan i y he RF-dis ance be ween gene and species ee and a measu e
o clock-likeness o each alignmen in he au osomal UCE da ase . Speci ically, we es ima ed
he likelihood a io o clocklike o non-clocklike models o each gene; lowe a ios indica e
molecula e olu ion ha is mo e clocklike. The p o ocol we ollowed was iden ical o ha in
Mushe e al. (2019). A e es ima ing likelihood a ios and RF-dis ances o all alignmen s and
gene ees, we selec ed only hose loci wi h RF dis ances below he mean dis ance, and wi h
likelihood a ios ha we e wo s anda d de ia ions below he mean a io. This allowed us o
include a ela i ely small numbe (22 in o al) o genes wi h ela i ely low a iance in molecula
subs i u ion a es among b anches, bu we e ela i ely in o ma i e o phylogene ic
econs uc ion (Figu e S27).
To es ima e he ime ame o inamou di e si ica ion, we used a ossilized bi h-dea h
model wi h six c own Tinamidae ip- ossil calib a ions using an op imized elaxed clock and
ossilized bi h-dea h model in Beas 2.7.7 wi h he Sampled Ances o s and Op imized Relaxed
Clock packages (D ummond e al. 2006; Ga yushkina e al. 2014; Bouckae e al. 2019;
Zhang and D ummond 2020; Douglas e al. 2021), and condi ioning on he oo and ho
pa ame e s. We hen used he il e ed au osomal UCE1000 alignmen s desc ibed abo e,
gene a ing lines o ques ion ma ks in each alignmen o six ossil axa (see Table S2). We
applied ip da es measu ed as yea s be o e he p esen using ini ial da es lis ed in Table S2,
and uni o m p io s on ip ages using he uppe and lowe bounds o hei es ima ed ages om
he ossil eco d. We linked all clock and ee models bu pa i ioned si e models by locus (22 in
all), wi h a GTR + G model o each pa i ion. De aul p io s and se ings we e used excep
whe e speci ied in Table S3.
Because we used a ela i ely small numbe o DNA sequences, we cons ained 14
nodes in he ee using MRCA monophyle ic node cons ain s. Fi s , we cons ained all nodes
wi h ossils o known posi ion based on p e ious mo phological wo k (Be elli e al. 2014).
Second, a e p elimina y analysis wi hou u he node cons ain s, we cons ained nodes ha
did no con e ge on he opologies known om he main phylogene ic analyses. The speci ic
de ails o he monophyly cons ain s a e as ollows (1) c own Tinamidae, (2) c own No hu inae,
(3) c own Tinaminae, (4) “Eud omia” sp. + Eud omia + Tinamo is, (5) Eud omia + Tinamo is, (6)
Eud omia, (7) No hop oc a + Rhyncho us + No hu a, (8) No hop oc a + Rhyncho us, (9)
No hoce cus, (10) Tinamus + C yp u ellus, (11) C yp u ellus including ossils Macn-sc-3610 and
Macn-sc-3613, (12) C yp u ellus soui, C. a iega us, C. ba le i, C. b e i os is, + C. casaquia e,
(13) C yp u ellus Clade A, and (14) C. bouca di, C. ke iae, C. e y h opus, C. s igulosus, + C.
duidae. The inal cons ain was used o ma ch opology T1, he mos equen ly eco e ed
opology in ou phylogene ic analyses o he mos di icul node in he phylogeny. We an he
MCMC o 150,000,000 gene a ions wi h sampling e e y 50,000 gene a ions. The ee was
summa ized as a maximum clade c edibili y ee wi h median alues o node heigh s, and a
bu n-in o 20% in T eeAnno a o 2.7.7(Bouckae e al. 2019). Con e gence was assessed
using T ace 1.7.1(Rambau e al. 2018).
We also modeled specia ion and ex inc ion a es using he R-package ‘TESS’ (Höhna e
al. 2016). TESS uses a jMCMC algo i hm o es ima e he iming o specia ion and ex inc ion
a e shi s unde an episodic bi h-dea h model. To do so, we p uned ossil ips and ou g oups
ou o he beas ee, and an he MCMC o 106 gene a ions, assuming 85% comple e
axonomic sampling o he clade.
Resul s
Assembly me ics, comple eness, and o holog me ics
We success ully assembled 68 o he 69 newly sequenced genomes and ex ac ed
BUSCO and UCE a ge s om hese assemblies plus 10 publicly a ailable inamou and 2
publicly a ailable ou g oup whole genome assemblies (Table S1). Genome-wide sequence
co e age a ied wi hin and among samples, bu was gene ally high; he mean o a e age
Supplemen al Tables and Figu es:
Table S1: A lis o samples used in his s udy including s a is ics o sequence co e age and
genome comple eness (supplemen al ile only, no shown)
Fossil
S a ig aphic
le el
Es ima ed
age (Ma)
Phylogene ic
posi ion jus i ica ion
P io De ails
Ci a ions
Diogeno nis
agillis
I abo aian-
Casamayo an
53–48.6
Ini ial ip
da e: NA
Mul iple
synapomo phies wi h
Rheidae place his
ossil in Rhei o mes
Logno mal MRCA
cons ain on age o
"Highe a i es"
O se =55, M=7,
S=1
Al a enga
1983, Almeida
e al. 2022
Undesc ibed
ossil Macn-
sc-3610
Mon e
obse acion,
San a C uz
Fo ma ion
18–15.2
Ini ial ip
da e: 16.6
Phylogene ic analysis
o mo phological da a
placed his ossil as
sis e o c own
C yp u ellus
Cons ained
monophyly o
C yp u ellus + s em
ossils, hus allowing
his ossil o eely
a y wi hin s em and
c own C yp u ellus
Be elli and
Chiappe 2005,
Be elli e al.
2014, Almeida
e al. 2022
Undesc ibed
ossil Macn-
sc-3613
Mon e
obse acion,
San a C uz
Fo ma ion
18–15.2
Ini ial ip
da e: 16.6
Phylogene ic analysis
o mo phological da a
placed his ossil as
sis e o c own
C yp u ellus
Cons ained
monophyly o
C yp u ellus + s em
ossils, hus allowing
his ossil o eely
a y wi hin s em and
c own C yp u ellus
Be elli and
Chiappe 2005,
Be elli e al.
2014, Almeida
e al. 2022
C yp u ellus
eai
Cañadón de las
Vacas, San a
C uz Fo ma ion
17.5–16.3
Ini ial ip
da e: 16.9
P e ious phylogene ic
analysis was
equi ocal, placing his
specimen ei he in
s em o c own
C yp u ellus
Cons ained
monophyly o
C yp u ellus + s em
ossils, hus allowing
his ossil o eely
a y wi hin s em and
c own C yp u ellus
Be elli e al.
2014, Almeida
e al. 2022
"Eud omia"
sp.
Ce o Azul
Fo ma ion a
Salinas G andes
7.2–6
Ini ial ip
da e: 6.6
P e ious phylogene ic
analysis placed his
ossil as sis e o
Eud omia + Tinamo is
Cons ained
phylogene ic
posi ion o
"Eud omia sp." hus
cons aining he
posi ion sis e o
Tinamo is+Eud omi
a
Be elli e al.
2014, Almeida
e al. 2022
Eud omia
olsoni
Fa ola Mon e
He moso
Mon ehe mosan
5–4.5
Ini ial ip
da e: 4.75
P e ious phylogene ic
analysis placed his
ossil in he genus
Eud omia
Cons ained
monophyly o ex an
Eud omia + ossil
allowing he ossil o
eely a y wi hin
c own Eud omia
Be elli e al.
2014, Almeida
e al. 2022
Table S2: A lis o all ip ossils used in he Beas analysis along wi h jus i ica ion o hei cons ained
posi ions, p io dis ibu ions and ci a ions.
Pa ame e Name
P io
De ails
ORCRa es
Log No mal
Ini ial=0.5, M=1.0, S=0.2, O se =0.0
ORCSigma
Exponen ial
Ini ial=0.2, M=0.3337,O se =0.0
ORCucldMean
Exponen ial
Ini ial=0.001, M=10.0, O se =0.0
hoFBD
Uni o m
Ini ial=0.75, Lowe =0.5, Uppe =1.0, O se =0.0
SamplingP opo ionFBD
Exponen ial
Ini ial=0.25, M=0.25, O se =0.0
MRCA
No p io
Cons ained monophyly o 14 clades o imp o e
accu acy o he phylogene ic esul s gi en he ela i ely
small numbe o loci used. These a e ma ked wi h
whi e ci cles in Figu e 1 o he main ex .
Table S3: Non-de aul p io s used in he beas analysis
No hu a
pa ula
Chapadmalal
Fo ma ion
5–4.5
Ini ial ip
da e: 4.75
P e ious phylogene ic
analysis was
equi ocal, placing his
species ei he in
No hu a o as sis e o
No hu a +
No hop oc a +
Rhyncho us.
Lack o
synapomo phies wi h
No hop oc a +
Rhyncho us sugges i
is a s em o c own
No hu a.
Cons ained
monophyly o
No hop oc a +
Rhyncho us, hus
allowing his ossil o
eely a y wi hin
s em and c own
No hu a, and s em
No hop oc a +
Rhyncho us
Be elli e al.
2014; Almeida
e al. 2022
Da ase
Numbe o loci
To al conca ena ed leng h
UCE100 comple e
2,712
879,985
UCE100 75% comple e
4,631
1,482,732
UCE300 comple e
2,702
1,864,316
UCE300 75% comple e
4,621
3,144,625
UCE1000 comple e
2,628
4,643,253
UCE1000 75% comple e
4,514
7,846,363
UCE1000 au osomes 75% comple e
4,244
7,384,066
UCE1000 Z-ch omosome 75% comple e
304
677,088
BUSCOs comple e
2,274
3,103,102
BUSCOs 75% comple e
7,414
10,083,122
Table S4: Numbe o alignmen s and o al conca ena ed leng h o each da ase .
Da ase
Model
Adj. R-squa ed
slope
P
AIC
BUSCOs
linea (#PIS)
0.3387
-0.0305809
<0.0001
19291
loga i hmic (#PIS)
0.497
-20.6024
<0.0001
18669
linea (%PIS)
0.001863
-7.36
0.0396
20228
loga i hmic (%PIS)
0.01044
-6.684
<0.0001
20208
UCE100Flank
linea (#PIS)
0.7797
-0.759178
<0.0001
19897
loga i hmic (#PIS)
0.6933
-25.8442
<0.0001
20794
linea (%PIS)
0.7747
-237.6429
<0.0001
19959
loga i hmic (%PIS)
0.6933
-25.8442
<0.0001
20843
UCE300Flank
linea (#PIS)
0.636
-0.254330
<0.0001
21092
loga i hmic (#PIS)
0.7253
-37.7134
<0.0001
20331
linea (%PIS)
0.6039
-160.9276
<0.0001
21320
loga i hmic (%PIS)
0.705
-35.7484
<0.0001
20523
UCE1000Flank
linea (#PIS)
0.24
-0.028939
<0.0001
18683
loga i hmic (#PIS)
0.2895
-18.5317
<0.0001
18506
linea (%PIS)
0.1312
-39.8117
<0.0001
19035
loga i hmic (%PIS)
0.2287
-9.0358
<0.0001
18927
UCE Combined
linea (#PIS)
0.6814
-0.108918
<0.0001
24575
loga i hmic (#PIS)
0.9197
-28.6239
<0.0001
20828
linea (%PIS)
0.7765
-248.5049
<0.0001
23611
loga i hmic (%PIS)
0.791
-48.0636
<0.0001
23428
Table S5: Model es s o log e sus linea i s o all gene alized linea models. This model es was done
wice o each da ase , once using he numbe o pa simony in o ma i e si es pe locus (#PIS) and once
using he pe cen age o pa simony in o ma i e si es pe locus (%PIS). Rows highligh ed in g ay indica e
he bes model o each pai ed model es .
Table S6: Clade ages om his and o he s udies (supplemen al ile only, no shown)
Figu e S1: Conca ena ed Phylogeny o inamous based on UCEs wi h 1,000 bp o lanking
sequence. All nodes ha e boo s ap alue o 100% excep whe e no ed. No e ela i ely sho in e nodal
b anch leng hs in Clade A. Tinamou illus a ions by TAC.
Figu e S2: BUSCOs 75% comple e As al phylogeny.
Figu e S3: BUSCOs 75% iq ee phylogeny
Figu e S4: BUSCOs 100% As al phylogeny
Figu e S5: BUSCOs 100% iq ee phylogeny
Figu e S12: UCE300 100% As al phylogeny
Figu e S13: UCE300 100% iq ee phylogeny
Figu e S14: UCE1000 au osomes 75% As al phylogeny
Figu e S15: UCE1000 au osomes 75% iq ee phylogeny
Figu e S16: UCE1000 Z ch omosome 75% As al phylogeny
Figu e S17: UCE1000 Z ch omosome 75% iq ee phylogeny
Figu e S18: UCE1000 75% whole da ase As al phylogeny
Figu e S19: UCE1000 75% whole da ase iq ee phylogeny
Figu e S20: UCE1000 100% whole da ase As al phylogeny
Figu e S21: UCE1000 100% whole da ase iq ee phylogeny
ances o ees o epidemiology and ossil calib a ion. PLoS Compu . Biol. 10:e1003919.
Hobol h A., Ch is ensen O.F., Mailund T., Schie up M.H. 2007. Genomic ela ionships and
specia ion imes o human, chimpanzee, and go illa in e ed om a coalescen hidden
Ma ko model. PLoS Gene . 3:e7.
Ho K.J., Lomsadze A., Bo odo sky M., S anke M. 2019. Whole-Genome Anno a ion wi h
BRAKER. Me hods Mol. Biol. 1962:65–95.
Höhna S., May M.R., Moo e B.R. 2016. TESS: an R package o e icien ly simula ing
phylogene ic ees and pe o ming Bayesian in e ence o lineage di e si ica ion a es.
Bioin o ma ics. 32:789–791.
Kapli P., Flou i T., Tel o d M.J. 2021. Sys ema ic e o s in phylogene ic ees. Cu . Biol.
31:R59–R64.
Ka oh K., S andley D.M. 2013. MAFFT mul iple sequence alignmen so wa e e sion 7:
imp o emen s in pe o mance and usabili y. Mol. Biol. E ol. 30:772–780.
Maddison W.P. 1997. Gene T ees in Species T ees. Sys . Biol. 46:523–536.
McCo mack J.E., Tsai W.L.E., Fai clo h B.C. 2016. Sequence cap u e o ul aconse ed
elemen s om bi d museum specimens. Mol. Ecol. Resou . 16:1189–1203.
Mclean B.S., Bell K.C., Allen J.M., Helgen K.M., Cook J.A. 2019. Impac s o In e ence Me hod
and Da a se Fil e ing on Phylogenomic Resolu ion in a Rapid Radia ion o G ound
Squi els (Xe inae: Ma mo ini). Sys . Biol. 68:298–316.
Meiklejohn K.A., Fai clo h B.C., Glenn T.C., Kimball R.T., B aun E.L. 2016. Analysis o a Rapid
E olu iona y Radia ion Using Ul aconse ed Elemen s: E idence o a Bias in Some
Mul ispecies Coalescen Me hods. Sys . Biol. 65:612–627.
Mushe L.J., Ca anach T.A., Valqui T., B um ield R.T., Aleixo A., Johnson K.P., Wecks ein J.
2024. Whole-genome phylogenomics o he inamous (A es: Tinamidae): compa ing gene
ee es ima ion e o be ween BUSCOs and UCEs illumina es apid di e gence wi h
in og ession. bioRxi .
Mushe L.J., C ac a J. 2018. Phylogenomics and species delimi a ion o a complex adia ion o
Neo opical suboscine bi ds (Pachy amphus). Mol. Phylogene . E ol. 118:204–221.
Mushe L.J., Fe ei a M., Aue bach A.L., C ac a J. 2019. Why is Amazonia a “sou ce”o
biodi e si y? Clima e-media ed dispe sal and synch onous specia ion ac oss he Andes in
an a ian g oup (Ti y inae). P oc. Roy. Soc. B.
Ogle D.H., Doll J.C., Wheele A.P., Dinno A. 2023. FSA: simple ishe ies s ock assessmen
me hods. R package e sion 0.9.
Pamilo P., Nei M. 1988. Rela ionships be ween gene ees and species ees. Mol. Biol. E ol.
5:568–583.
Rambau A., D ummond A.J., Xie D., Baele G., Sucha d M.A. 2018. Pos e io Summa iza ion in
Bayesian Phylogene ics Using T ace 1.7. Sys . Biol. 67:901–904.
Reddy S., Kimball R.T., Pandey A., Hosne P.A., B aun M.J., Hacke S.J., Han K.-L., Ha shman
J., Huddles on C.J., Kings on S., Ma ks B.D., Miglia K.J., Moo e W.S., Sheldon F.H., Wi
C.C., Yu i T., B aun E.L. 2017. Why Do Phylogenomic Da a Se s Yield Con lic ing T ees?
Da a Type In luences he A ian T ee o Li e mo e han Taxon Sampling. Sys . Biol. 66:857–
879.
Robinson D.F., Foulds L.R. 1981. Compa ison o phylogene ic ees. Ma h. Biosci. 53:131–147.
Rod íguez-Ezpele a N., B inkmann H., Rou e B., La illo N., Lang B.F., Philippe H. 2007.
De ec ing and o e coming sys ema ic e o s in genome-scale phylogenies. Sys . Biol.
56:389–399.
Schie up M.H., Hein J. 2000. Recombina ion and he molecula clock. Mol. Biol. E ol. 17:1578–
1579.
Schliep K.P. 2011. phango n: phylogene ic analysis in R. Bioin o ma ics. 27:592–593.
Simão F.A., Wa e house R.M., Ioannidis P., K i en se a E.V., Zdobno E.M. 2015. BUSCO:
assessing genome assembly and anno a ion comple eness wi h single-copy o hologs.
Bioin o ma ics. 31:3210–3212.
Simmons M.P., Ga esy J. 2021. Collapsing dubiously esol ed gene- ee b anches in
phylogenomic coalescen analyses. Mol. Phylogene . E ol. 158:107092.
Smi h B.T., Ha ey M.G., Fai clo h B.C., Glenn T.C., B um ield R.T. 2014. Ta ge cap u e and
massi ely pa allel sequencing o ul aconse ed elemen s o compa a i e s udies a
shallow e olu iona y ime scales. Sys . Biol. 63:83–95.
Smi h B.T., Mauck W.M., Benz B., Ande sen M.J. 2018. Une en missing da a skews
phylogenomic ela ionships wi hin he lo ies and lo ikee s [In e ne ]. .
Smi h B.T., Me win J., P o os K.L., Thom G., B um ield R.T., Fe ei a M., Mauck W.M., Moyle
R.G., W igh T.F., Joseph L. 2023. Phylogenomic Analysis o he Pa o s o he Wo ld
Dis inguishes A i ac ual om Biological Sou ces o Gene T ee Disco dance. Sys . Biol.
72:228–241.
Sp inge M.S., Ga esy J. 2016. The gene ee delusion. Mol. Phylogene . E ol. 94:1–33.
S ille J., Feng S., Chowdhu y A.-A., Ri as-González I., Duchêne D.A., Fang Q., Deng Y.,
Kozlo A., S ama akis A., Cla amun S., Nguyen J.M.T., Ho S.Y.W., Fai clo h B.C., Haag J.,
Houde P., C ac a J., Balaban M., Mai U., Chen G., Gao R., Zhou C., Xie Y., Huang Z.,
Cao Z., Yan Z., Ogil ie H.A., Nakhleh L., Lindow B., Mo el B., Fjeldså J., Hosne P.A., da
Fonseca R.R., Pe e sen B., Tobias J.A., Székely T., Kennedy J.D., Ree e A.H., Like A.,
S e ande M., An unes A., Tie ze D.T., Be elsen M.F., Lei F., Rahbek C., G a es G.R.,
Schie up M.H., Wa now T., B aun E.L., Gilbe M.T.P., Ja is E.D., Mi a ab S., Zhang G.
2024. Complexi y o a ian e olu ion e ealed by amily-le el genomes. Na u e.
Tajima F. 1983. E olu iona y ela ionship o DNA sequences in ini e popula ions. Gene ics.
105:437–460.
Tea Y.-K., Xu X., DiBa is a J.D., Lo N., Cowman P.F., Ho S.Y.W. 2021. Phylogenomic Analysis
o Conca ena ed Ul aconse ed Elemen s Re eals he Recen E olu iona y Radia ion o
he Fai y W asses (Teleos ei: Lab idae: Ci hilab us). Sys . Biol. 71:1–12.
Xi Z., Liu L., Da is C.C. 2015. Genes wi h minimal phylogene ic in o ma ion a e p oblema ic o
coalescen analyses when gene ee es ima ion is biased. Mol. Phylogene . E ol. 92:63–71.
Zhang R., D ummond A. 2020. Imp o ing he pe o mance o Bayesian phylogene ic in e ence
unde elaxed clock models. BMC E ol. Biol. 20:54.
Zhao M., Ku is S.M., Whi e N.D., Monc ie A.E., Lei e R.N., B um ield R.T., B aun E.L., Kimball
R.T. 2023. Explo ing Con lic s in Whole Genome Phylogene ics: A Case S udy Wi hin
Manakins (A es: Pip idae). Sys . Biol. 72:161–178.
