Developmentally Incomplete Barb Rami Increased the Morphological Diversity of Early Feathers
preprint
OA: closed
Abstract
Recent studies have significantly advanced our understanding of the evolutionary processes and developmental mechanisms behind feather branching patterns. However, little attention has been given to the tissue differentiation of branches associate with this process. Here, we describe eight feathers preserved in Burmese ambers, characterized by ribbon-like barb rami composed solely of cortex, lacking a medulla. These feathers are categorized into three morphotypes based on the details of their barb rami morphologies. For each morphotype, potential developmental scenarios were proposed through inference of barb development in chickens, suggesting limited tissue differentiation of barb rami in these early feathers. Functional simulations indicate that the configuration of modern barb rami is aerodynamically more stable than those of early feathers, indicating that flexural stiffness may have been a crucial factor driving the evolution of feather branches toward modern configurations. The discovery of developmentally incomplete barb rami in Burmite feathers implies that, while the three-level hierarchical branching pattern of modern feathers had evolved by the Jurassic period, tissue differentiation patterns of feather branches may not have fully stabilized by the Late Cretaceous. This instability likely contributes to the morphological diversity of early feathers, resulting in forms not seen in modern birds.
Full text
84,856 characters
· extracted from
oa-pdf
· click to expand
Developmentally Incomplete Barb Rami Increased the Morphological Diversity
of Early Feathers
Running title: Mesozoic Feathers with Incomplete Barb Rami
Yanyun Zhang
1, Jiawei Tang2, Ying Wang2*, Shuo Wang1*
1 School of Life Sciences, East China Normal University, Shanghai 200241, P. R. China
2 School of Energy and Power Engineering, University of Shanghai for Science and Technology,
Shanghai 200093, P. R. China
* To whom correspondence should be addressed.
Sh uo Wa ng. Email: sw a n g@ b i o. ec n u . e du . cn ( SW) ORCID I D: 0000 - 0002- 407 6- 02 02
Ying W a n g . Em ail: wangyi [email protected] d u.cn (Y W) O RCID I D: 000 0-0 002- 690 6- 944X
Auth o r’ s C ontributions: S .W. desig n e d th e projec t; Y.Z., J.T. Y. W. and S. W. per fo rmed e x periments ;
Y.Z . , J .T . Y.W. C . D . an d S . W . an al y ze d da ta ; and S.W ., Y.Z . a n d Y .W. wrote the pape r.
Ackno w ledgem e nt s : We th ank Dr. Tau t i s Sk orka and Dr. Tea Ja s h as hvili (U n iv ers ity of So uther n
Ca liforni a ) for th e ir a ss is ta nc e in sca nn in g the spec i me n s , and to N . D ., C.C. and Y .S . (E ast Ch i n a Norma l
Unive rsit y ) fo r their discussio n. We a lso a p p rec ia te the invalu abl e help of Y .H . (IVPP) an d R.Y. ( East
Chi n a Norm al U nive rsit y) in pro c essing VG. S .W. is sup ported by the H u ma n Fronti e r Sci ence Pro gram
(LT00 0 728 / 2 01 8 ) , the Zij iang Prog r a m fo r Talented Scho lars at East China Norma l Unive rsit y , and the
Shan ghai Pu jian g Pro g ram (2 3PJ 14 02300) .
Compet ing Inter est s . The a ut hors declare no c o mpeting i ntere st s.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
Abst ract: R e c ent stu die s hav e s ig nificantly adv anced ou r und e r s tandi ng of t he
ev o lut i o nar y pr oce ss es a nd d evel opmental me ch anis m s behind fe ather branching
patterns . How ever , littl e a t t e nt ion ha s been give n t o t he ti ss ue diff er e nt ia t i o n of br anc h e s
ass oc i ate with thi s pr oce ss. H e r e, w e d esc r i be eight f eath er s pre serv e d in Burmes e
amber s, ch ar ac t er ized by r i b b on- like barb rami c ompo s ed s ol e ly of c or t ex, la c king a
medu ll a. Th e s e fe ath er s ar e ca t e g oriz e d into t hre e m orphotyp es ba s ed o n th e d et ail s of
t h e i r ba rb ra m i m o rp h o l og i es. For ea ch morphotype, p otential d ev e lopmenta l s c ena rio s
w e r e prop o sed thr ou gh infer ence of b arb d e v elopm e nt in c hick ens , sug g e sti ng limit e d
tissu e diff erentiati on of barb r a mi in t h ese ea rly f eath er s. Fun c ti onal simu la t i ons indicat e
that the c onfigur ation o f mod ern bar b r ami is a er odyn amic a lly m or e s t a b l e th an tho s e of
ear ly f e ath er s , indi ca t ing that fl exur al s tiffnes s may h ave b ee n a cru cial f actor dr iving t he
ev o lut i o n of fe ath er br a nch e s to w ard mo d ern co nf igur ati ons . Th e dis c ov e r y of
devel opmental ly inc ompl et e bar b r ami in Bur mite fe a t he r s implie s that , whil e th e
t h re e - le ve l h ie r a rc h ic al b ra n c h i n g p a t t ern o f m o de rn f e a t h ers had e vo l ve d b y th e J u ras s ic
perio d, tis sue diff erentiati on patterns o f fea t h e r b ranche s ma y not h ave f ull y st abilize d by
the Lat e Cr eta c e ou s . T his i nsta bility li kely c ont rib ute s t o th e mor phologi ca l dive r s ity of
e a rl y f e a t h e rs , res u l ti n g i n f o rm s n o t se e n i n m o d e r n b i rd s.
Key words: barb ∣ am b e r ∣ La te Creta c e ou s ∣ evo-devo ∣ barbu le∣ m edu lla ∣ f ollicl e
1. I n tr o d u c ti o n
Extant feath er s, with th eir hi erar chical br anc h e s of r achi s , bar bs , an d b arb ul e s , a r e
among th e mo s t c omple x in tegum entary s tr u c t ur e s ever kno w n 1, 2 . A t ypic a l bipinnate
feath er con s ist s of a centr al r ac his ( f e ath e r shaft ) ext e nd ing from t h e pr oxim a l cal amu s
and divid es th e f ea t her into t wo v ane s 2 . The pr ima r y br anc h e s of th e r ac hi s, c a lle d b arb s ,
e a c h ha ve a ce nt r a l r a m us t h a t s u p p or t s tw o r o ws o f b a rb ul es , wh i c h a r e t h e fe a t he r's
sec ondar y br anc h e s 1-3 . T he morphol o gical dive rsity of mo de rn feath er s ar is es l arg el y
fr o m the b r anc hing pattern s of the r a chi s, b arb s, and bar b ul es 1,3 , in additi on t o their
color ation.
Exc e pt for t he c al amu s and b arb ul e s, both the r achi s a nd b a r b s in mod e r n feath er s
ar e ch a r a c te ri z e d by a d ens e p e r iphe r a l c or t ex s urr o unding a spongy interna l medu ll a 1,4 .
This sand wich- like s tr u c t ur e grant s mod ern fe ather s with fle xibility and ligh tness whi le
maintaining the stiffnes s nec essar y fo r flight, bo dy str eamlining, and othe r functions 4, 5 .
Alt ho u g h s t udi e s o v er the p ast d ec ade s h av e gr eat ly adv a nc ed our un d e r s tan ding of th e
ev o lut i o nar y pr oce ss es a nd d evel opmental me ch anis m s behind fe ather branching
patterns 2, 3, 6 -8 , lit tl e att ention ha s b e en give n to th e ev o lution ar y pr o ces s of ti ssu e
differ entiation of fe ath er br a nch e s 4, 5 . A rec ent stud y highlig ht e d t he v entrally ope n
rachi se s d o c um e nt ed in M e so z oic fe ath e r s, w hi ch r esult from d evel opme nt a l
incomple t e ne ss of typic al cylind ric al s ha f t s . Thi s und er scor e s th e r o l e of t i ssu e
differ entiation along th e ra dia l a x i s of t he fo llic le in d riving th e morphol ogi c a l va r iation s
in early f eath er s 5 . Whil e it remains un cl ea r wh e t h e r the se ar e patho logi c a l st ructur es or
simply r e pr es ent failed att empts of nat ure , their pr es e nc e in both non- a vi al an a nd avi al a n
therop od s, and ev e n in th e s umme r plumage o f ex t a nt penguins 5 , s u g ge s t s that
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
incomple t e tiss ue diff er entiation migh t have pla y e d f unc ti onal r ol es in c r e a t in g rachi a l
variati ons . The s e vari ations ma y ha v e h elp e d bir d s a dapt to new ecol ogical nic hes, rai sing
import a nt questi ons ab out ho w t h e ev o lution o f tiss ue dif f e r entiation of f e ath e r br anche s ,
in conjunc ti o n with br a nching patt erns , has cont ribute d t o th e m orphol ogi c al diver sity of
feath er s.
S i n c e th e r a c h ia l ri dg e f o rm s t h r o ug h t h e f us i o n o f ba rb r i dg es 3 , it has b ee n
specu lat e d th at v entr a lly op en bar b r ami ma y have exi s t in ea r ly f eath er s, a lthough non e
had b e e n di scov ere d wh e n ventr a l ly op en rac hise s were fir s t d es c r ib e d 5 . H e r e , w e rep ort
enigma t ic f e a t her s f rom Burm e se am be r s , f e a t uring r i bbon- like b ar b rami a t tache d t o
incomple t e ra c hi se s. Car e f ul examinati on su gge sts th at s ome o f th e s e r i bb on-like b arb s
ar e v entr a lly open, a nd ba r b r ami a r e e ve n split a long th e midline s in one s pe cimen whic h
cre at e uniqu e me s h- lik e c onne ction s of bar b s that h av e not b een do cum ente d i n ext ant or
fos s il bir d s . Func t ional s imu lati ons r e ve a l t hat the a er od ynamic s t ability of ramu s - s plit
barb i s h ig h er t h a n t ha t of b ar b s with s i mple ve nt ra lly open or pl at e-lik e c r oss- section s in
ce r t ain co nd iti on, thoug h s t il l l e s s s ta b le than th e typica l sand wic h - lik e confi gurati on of
mode rn feath er s. Thi s s ugge st s that t hes e earl y fe athe rs wer e like ly not a dapt e d for flight
or b ody str e amlining. The se findin gs enhance o u r u nde r s t anding of ho w t issu e
differ entiation c ontr ibut e d to th e mo r phol ogi c al diver s it y of e arl y fe a t h er s , o ffer ing
insig ht s th at may ins pir e futu r e biomi metic d esign.
2.
M a t e r i al s an d M e t h o d s
Instit ut io n a l Abbrevia tio ns : CN U, Coll eg e of Lif e S cie n ce s , Capit al Nor ma l
Univ er s ity, B eijing, China; E C NU, Sc h ool of Life S c ien ces, E a s t C hina Nor ma l Univer s it y ,
Shangha i , Chi na .
Ancien t a n d exta n t feathers : All amber - e mb e dd e d fe ath er sp ec im ens d e s crib e d in
this pap er w e r e d onat ed in 20 16 b y a n anonymo us amat eur col le c tor t o t he Bio lo gy
Mus e u m of Ea s t China N or mal U niver sity (EC N U), Sh anghai, China. N one of t he
specimens w e r e as s o ci ate d w ith Mya nmar Ec onomic Cor po r ation, a milit ary- o wned
cong l omer at e, ensu ring complia nc e with th e mo r a t or ium i ssu ed by the So cie t y o f
Vert e br ate Pa l eontol og y
9 .
For his t ologic al s ec tion s, a reg e ne ra ti n g 5 th prim a r y r e m ex, 4 w eeks into
reg ene r ation, and an ear l y - s t ag e d e v el oping c overt c ontour f e ath er from th e br ea st w e r e
extr act ed from a one- ye a r - ol d W hite Le gh or n c hicken (Ga ll us ga l lus ) t ha t wa s rai s e d at
the animal center of EC N U. Th e prima r y r e m ex h ad re ache d appro ximat e ly o ne fourth of
its full g ro wth, and the por ti on still d e v el oping within the fol licl e w a s s ampl e d. To l ab e l
prolife r a t ing c e l ls , 1% br omo de ox y ur i dine ( Brd U, Sigma B 50 02) was inj ec t e d into th e
chic k en v ein 3 h our s prior t o f e ather ex traction. The pr o c e dur e s f or fe ath e r c o l le c t ion a n d
paraffin s ec tioning w e r e c ond uct e d in a c cor dance with protoc ol s appr o v ed by E as t Ch ina
Normal U niv ersit y (Appr ov al Numb e r ARX M20240 10).
Termino logy : T he terms u s e d to de sc r i be f eath e r stru ctur es and ori entati ons fol lo w
Lucas and St ett enheim (1 9 72)
1 , whil e the d esc r iptions o f the d e v el oping follic l e adh er e t o
Pr um (1999) 3 . Spe c if ical ly, "pro ximal" and " dis t al" r efe r to stru ctur es cl o s er t o or farth er
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
fr o m the cal amus , r e s pe ctively; whil e " centra l" a nd "periph er al" r ef er t o s tr u ct ure s n ear
or a w ay fr om the fo llic le c enter, r espec tively. T hrou gho ut the pap e r , we employ e d t e r ms
like " sand wic h - like," " ribb on- lik e," " p lat e-lik e ," a nd " b eam- lik e" t o de scr ibe diff er ent
configur ation s of f eath er b a r bs. "S and wic h -li k e" r efer s t o b arb r ami typic a l of mod ern
feath er s, wh er e an internal spong y me dul la is sur r o unded by a c ort ex l a y er w hen v ie w e d
in cr o s s- secti on. " Ribb on-like" d e s crib e s bar b r a mi with onl y a do r sal c or t ex, sho wing
so m e diff er entiation of dor sal cort ex but la c king a dis tinct me d ull a and ve ntr a l c or t ex.
"P la t e - lik e" is u se d f o r b a r b r ami t hat cons i st sol ely of an expand ed cort e x, without
distinc t med ul la ; whil e "b e am-like" ind icat e s the mo st prim itive config urati on, fe atur ing
minimal differ entiation of th e d or sa l co rtex a n d no oth er t is su e specialization.
H i s t ol o g i cal se c t i on s an d s t ai ni n g : The gro wing portion of th e reg e ner a t ing
chic k en primary reme x w a s emb edd e d in pa r affin a n d s ec tion ed at 5 mm interval s ,
starting fr om the immatu r e pro ximal end to est abli s h a gr o w th s eri e s. Aft er 4 hour s of
fixation in 4% par a fo r mal de h yd e, sa mple s w er e d ehyd rat ed t hr ough a g r a ded eth anol
ser ies r ang ing from 30% to 1 00 %. P a raffin sec t i o n s with a thickne ss o f 7μ m w er e c ut
perpendi cul a r t o th e f ea t h e r ’s l ong ax i s . H e mato xy lin and E os in ( H&E) s t aining foll o w e d
e s tab lish ed pr oto c ol s by Fisch er et al. (20 0 8)
10 , a n d BrdU a s sa ys w e re p e r fo rm e d
accor ding to Le e et al. ( 200 1) 11 , us i n g an A EC I m m u no h i s t o c he m i s t r y Co l o r D e ve lo p m e n t
Kit (Sangon B iotec h , E67003 1 ) inst ead of t he c hr omo genic metho d 11 . Ma s s on staining
w a s c o nd ucte d at d ev e l opmental st a g e I primary r e m ex f ol lo w ing S uvik (20 12) t o
comple m ent the H&E stanning 12 . Dig i t a l imag e s of histo lo g ic al slid e s w e r e c apture d u sing
a Nikon Di gital Sight 10 C a m er a Sy st em att ach e d to a N ik on Ni- E tr ans mitt ed
micro scope.
Micro CT sc a nning : Computed tom ogr aphy (CT ) s c ans of E CNU A 32 wer e perfor m e d
using a b eam ene r gy of 1 8kV at a r e s ol ution of 4.0 um per pix e l u s in g the Ph oenix
v-tome- x indu stri al CT s c anne r at the Mol ecula r Imaging Ce n te r , Univ er sity of S outh ern
Califor nia . ECN U A 143- 1 was scanned unde r 1 2 kV at a r es oluti on o f 3.9u m per pixel
using the sam e scanner at Shang h ai Yinghua Inspec ti on & Tes ting. Se gm e ntation w a s
ca r ri e d out utilizing Mimics 15 and Amir a softw ar e at the Univ er sity o f South ern
Califor nia .
Fu ncti o nal Simulatio n : The flex ur al s tif fnes s of fe ath ers with differ ent
cro ss- s e ction al s hap e s of b arb r ami w e r e e stimat ed u sing ANSYS FL UENT 12.0 s of t war e 13 .
3D mod el s of EC NU A 32 and an uncate goriz ed cont o ur f e ather of similar siz e from a
one-y ear - old W hit e L e gh orn chic k e n w er e compare d to as ses s th e a ero dynamic
perfor m a nce (Fig. S 1 A , B). Du e to limit e d c omputing power, simplifie d mod e l s w er e u s e d
for a e r od y namic an aly se s of f e ather s with va rio us bar b r am us c r os s- secti ons (Fig. S2) .
Barbu le s ar e omitt ed from t he s implif ie d mo d e l s b ecau s e the y a r e not t he f ocus of t h e
pres e nt stu dy. A ribb on- lik e br anch w it h cr es cent-sh aped cro s s - sec t i o n si mulat e th e
ve nt ral ly open b ar b rami of m o r phot ype I fe a t h e r s (Fig. S2A , mod e l I); a r ibbo n -lik e
branch with pair e d hal f-cr e scents cr o s s - se c tion s imu la t e morph otype II fea ther b arb s ,
wh e r e b a r b ramu s ar e ma de- up b y hal ve s of two adj ac e nt s plit ba r b br anch es (Fig . S2B ,
mode l II ); a br a nch with a 10: 1 fl at e ll iptica l soli d cr o ss - s e c tions simul ate th e plat e-lik e
barb r amu s of morph otype II I feather s (F ig . S2C , mod el III ); a branc h w it h a perf ec t l y
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
round and hol lo w c ro ss- s e ction s s imu l ate id e alize d, non-r eali stic b a r b r amu s (Fig. S2, D ,
mode l I V), and a br anc h with a 5:1 e lli p t ical h oll ow c r oss- se ctions s im ul ate the b arb
r a m us o f m o de rn f l i g h t f e a t he rs ( F ig . S2 E , m o de l V ) . S t r u c t ura l p a r a me t e rs o f e ac h m o de l
ar e given in Fig. S3.
To inve s tigat e wh e the r the aer od y nami c perf ormanc e of a s o lid c ro s s - s ec tion dif fe r s
fr o m that of h oll o w c r o ss- secti on with t he same p aram et ers , simul a t ions als o incl ud ed
branch e s w i th a perf e ct ly ro und soli d cro s s - s ec t ion ( Fig. S2F , mod el VI) a nd a 5: 1
elliptic a l s olid cr o ss - s e c tion (Fig. S2 G , m o d el V II) wer e als o incl ud ed in the s im ul a t io n .
For simplicity, t he r a c hi s in e a ch m od e l wa s simplifie d t o a cylindri ca l s hap e a nd
comple t el y fi xe d, with only th e bar bs fr ee t o mov e . Air fl o w wa s a ssum e d to be ev enly
applied. Sinc e th e bar bs wer e not in te r l oc k ed , no c ompl ete van e was simu l ated in t he
analys es.
Fl ui d -St r uctu r e I n teract io n (FSI ) A n aly s i s : Fluid- s tr u c t ur e inter a ction ( FSI)
a n a l y s is e xa m i n e s how de fo rm a ble s o li d s t r u c t ures b e ha ve u n de r flu i d f low f o r ces a n d
ho w str u c tur al d ispla cement aff e c t f l uid f lo w
14 . D u e to th e mini ma l d ef o rm a t io n of
feath er und er air f lo w, its impac t o n the flo w fi eld di strib ut io n wa s con sid e r ed negligib l e ,
all o w ing for a one- w a y FSI appr o a ch 15 . The FSI analysi s inclu d e: 1 ) m ode ling the ext er na l
flo w field surr oun d the f ea t her ; 2) tr a n sf err ing t he r esult s of th e flo w fi el d ana ly si s t o th e
stru ctur al fi el d; and 3) pe r f orming one -w ay co upled a na ly s i s. Sinc e t he bar bs and r achi s
ar e compos ed o f f eath e r k e r atin, mat e r ial pr op e r tie s wer e s e t b a sed on rea l f e ath e r
keratins , w it h a d ens it y o f 1. 15g/cm
3 16 , a Y oun g ’s mod ul us (a par amet er ch a ract erizing
the mat er ial s tif fnes s a g ain st t he ext er nal for c e ) of 2 .5x 1 0 9 N/M 2 17 , 1 8 , and a Poi s s on' s
ratio of 0.45 19 . E a c h m o d e l f e a t h e r w a s s e t t o a l e n g t h o f 6 0 m m , w i t h 1 0 m m l o n g b a r b s
form ing a 45° angl e with the r ac h is (Fig . S2 A). F or al l mod el s, the c ylindr i c a l r achi s ha d a
diamet er of 2 mm, and th e b arb rami h ad a di amet e r o f 1 mm, w ith a co rte x t hickne s s of
0 . 0 5 m m f o r h o l l ow ra m i (F i g . S 3 ) . T h e v e n t ral l y o pe n b a rb ra m i o f m orp h o t y p e I
f e a t he rs co rre s p o n de d t o h a l f of a h o ll ow ra m us (F i g . S 3 ① ) , whil e each o f the p air ed
ramu s br anch e s of mo rphotyp e II fe at her c or r e s pond e d to a q uar ter of th e h oll o w r amus
(Fig. S3 ② ).
Du e to the signific ant diff er ences in el as t i c ity b etw een the cor t ex a nd m e d u lla, th e
fle xur al s t iffne ss of fe ath e r b arb s i s pr imarily d et e r mined by th e geom etry of t he
cort ex 20- 22 . Simu lati ons subj ecte d each morphotype t o airfl o w e it he r p e r pen dicular t o t he
feath er v anes or p a r a ll e l to th e f eath er r a chi s from t h e calam us, exploring ae rodyn amic
perfor m a nce s und er l ater al an d he a d wind conditions , r e s pe c t iv e ly. Sinc e t he primar y
function of c over t c o n tour f e ather s i s t o s t r e a mlin e t he bo dy surf ac e by a llo w ing a ir flo w
to pass smoot hly a long the fe ath e r fr om t he calamu s , w e believ e t he h ea d w i nd c ondition
clos el y r eplicat e the a er odyna m ic perf or manc e of c ontour f e ath er s in their natural stat e .
While a b s ol ut e val ues may not r efl e c t re al- wor l d co nd iti ons , th e r e l ativ e flex i bilit y
(displ acem ent from t h e stati c stat e) and th e s t a t ic str e ss r e spo ns e s t o stand ar dized f or ce s
offe r insight into the a ero dynamic p erf o r mance of each f e a t he r t y pe.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
2. Result s:
Morpho ty p es :
Eight feathe rs pr es erv e d in sev en pie c e s of a m ber ar e inc lu ded i n this s t ud y. W hil e
rachi al mor p h ol ogy v ari es both amon g fe athe rs and along t h e ir l ength s, all specimens
show e vid ence o f a ventr all y open ra c hi s, alth oug h the ext ent varie s. For example , in
ECNU A 143- 1 (F ig . 1 A) , th e ventr all y open r egion ext end s thr ough out t h e ent ire ra chis
s i m i l a r to th a t o f CN U A 00 0 5 5 , whe re a s in ECNU A 147, it is re stri cte d to t he pr o xima l
half (F ig. 1J). Sinc e thi s stu dy f ocu se s primarily on th e b ar bs r a t he r than the r a chi s a nd
barb ul e s , th e feath er s ar e c at egor iz e d i nto t hre e m orphotyp es b a sed on the morpholo gy
of the ba r b rami:
Morphot ype I (Figs . 1- 2) is repr es ent e d by fou r ambe r s pe cime n s: ECNU A49,
A143, A 147 , and A 1 48. ECN U A14 7 a nd A 1 48 ea ch c ontain a ne ar ly complete i s olat ed
feath er w ith sy m m e tr ically di s tr ib ute d b ar b s and b arb ul e s. EC NU A 143 includ e s t w o
isolat ed f e ath er s and ar e l a b el e d as E CNU A143- 1 (Fig. 1A- D ) and A 143-2 ( Fig. 2A- D ),
re spective ly . EC N U A49 includ es only a s i ng le b arb (Fig. 2E- G ), thu s little i s kn o wn a bo ut
the r est of the f eath er. All m o r photyp e I f e ather s ar e c h ar act eriz ed by pro x imodist all y
bro ad, rib bon-lik e bar b rami w it h bar bul e s al ong their proxima l and di s t al edge s. CT
images r ev e a l that th es e r i bbo n - like b a rb r ami consi s t s ol e l y of a thin dor s a l c orte x that i s
cre sc ent-sh a pe d in c r o ss s e ction, with out pr esenting a me du lla o r a v entral cort e x ( Fig.
S4). Th is ab senc e r e s ult s in ventr ally o pen rami t ha t have not be en o bserv e d in modern
bird s. Ho wev e r , the ribb on- like r ami l ac k a midlin e r id ge and l ater al flang e s s e en in th e
previo usly r eport ed ventr all y o p en ra chise s 5 , s ug gesting th ey ar e me cha ni cal ly mor e
fle xibl e r e lativ e t o the d evel opmentally incompl e t e ventra lly o p en ra c hises ( se e
Aerodyn amic performance of d ifferent b arb rami for mor e infor ma t ion ).
Ribbon- lik e bar b r ami ar e found in specific r egi ons of indiv idu al b a r b s or a cr o ss
differ ent bar b s along a f e athe r sh a f t . For inst anc e , in ECNU A 148 (Fi g. 1L, bl u e
arr o w h e ad s ) , r ibbon-lik e r e g ions alt e r nate w ith s olid b eam section s, whi le in A143- 1 ( Fig .
1A, O ) , mo st b arb r ami ar e ribb on-like proximal ly an d tr a nsiti on into soli d b e a ms di stal ly .
This v ariati on r ef l e ct s th e dif fer enti al e x pansi on of t he d or sa l cor t ex during t h e
devel opment of a s ingle b arb . Ther ef o re, morpho type I f e ath er s ar e c h ar a cteriz ed by
ribbon- like, v entr all y ope n bar b r ami, with bar b ul es pr es ent along their pr oxima l and
dist al ed ge s .
Morphot ype I I ( Fig . 3) is r epr esent ed b y a s ingle i so lat e d f e a t he r ( ECN U A32),
which f eatur es ne arly sy mmetr i c a lly distribut e d ba r bs and a r elativ ely thi ck rachi s ( Fig.
3A- I). C T image s sh ow that the ra c hi s has a v entr al cl eft with a V - sha p ed cr o s s - se c t io n ,
rathe r than c ir c u lar or c r escent- s h ape d one ( Fi g. 3B , K-M) . Th e bar bs al ong th e pr o xima l
and di stal q uar ter s of th e r achi s a r e b eam-li k e, but tho se in the middle f or m forks soon
after bran c hing off fr om the r a chi s (Fi g. 3J and J ’) . Th e fork br anc h e s ar e r i bbon-lik e ,
each cor re spond ing t o h alf of the cr e s c ent-sh aped dor s al cor te x of a b arb r amus . D i sta lly,
the branch es from adj acent split bar bs f us e t o f o r m a s olid b eam, r esulting in a distinctiv e
mes h- like str uc tu r e ( Fig. 3Q ). N ota bly , bar bul e s ar e abs e n t a long th e edge s of the bar b
rami but present on t he f ork branch es, whe r e th ey ar e or i e nt e d to war d t he ce nter of t he
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
fork (Fig. 3J and J’). Thi s sugg est s that the f ork br anc h e s r epre s e nt halve s of a dja c ent
barb r ami th at split along the midline (F i g . 3 N - Q ), a configur a t ion not ob s e r v ed in any
other f o ss i l or mod e r n f eath er s . I n c r oss- s e ction vi ew, the dist al fo rk branch c o nsi stentl y
lie s ventra lly to the pro ximal br anc h w her e they mer ge into the ra chi s (F ig . 3L- M) , giving
the impr e s s ion t ha t ba r bs t w i st immediate ly a ft e r branching from t he f eath er shaft ( Fig .
3J-K) . Thus, mo r p h o t y p e I I f e ather s are di stingui s h ed by th e pr ese n ce of f or k e d bar b
rami with b arb ul e s a long t h e for k br anche s , o r ie nt ed to w ar d the c ent er of th e f ork.
Morphot ype III (Fi g. 4) i s r epr es e nt ed by t w o amb e r s p ecime n s, ECN U A 10 1 ( Fig.
4A- E) and A 149 (Fi g. 4F-I ), each c ont a i ning a sing l e f e ather with appa r e nt ly sy mme t rical
distrib ution of b arb s and b ar bul e s . Un like other f eath er morph otype s , th e b a r b rami of
Mor photy p e I II f e ather s are c ompre ss ed pr oximo di stally and b r o a d do rsov e ntrally ( Fi g .
4D, I) . Ther e i s no m ed ull a pr es e n t, a nd bar bul e s are di stribut e d a long b ot h the di s t a l
and proxim al surf ac es of t h e b arb rami. Thu s , morpho type III fea t h e rs a r e
d i s t i n g u is he d b y t he a bs e n c e o f a m e du l la a n d t h e p res e n c e o f p ro x i m o d is t al ly
compre ss e d b a r b r ami with bar bul es a ttac h ed t o both sur f ace s, setting t h e m apart fr om
morphot y pe I fe ath ers.
Aer o dyn amic performance of d ifferent b arb rami:
Our s imu lati on resul t s s h o w t ha t , among all te s t s, f eath er barb s with a perfect l y
round cr os s - se c t ion exhibit the high e s t str ength, whil e t h ose w ith elliptic al c r o ss- s e ction s
demon s tr at e r el a t iv el y lo wer str e ngth ( Figs . 5-6) . This s ug gest s t hat the fl e x ur al s tif fn e s s
of feath er b arb s i s pr imarily d eter mi ne d by cr os s - s e c tional geometr y, r ath er than t h e
materi al pr ope r ti e s of fe ath er ker atins 4, 18, 23 . I n b o t h s e ts o f te s t s , b a rb ra m i w i th s o l i d
cro ss- s e ction s c ons i stently s h o w l ess def or ma t ion compar e d t o tho s e wit h holl o w
cro ss- s e ction s ( Fi g s . 5 - 6) , in dicating t hat a s o lid s tr uc t ur e , giv e n th e sam e sha p e a nd
diamet er , s ignific antly enhanc es mecha nica l strength.
Spec ifi ca lly, f eather bar bs w it h a pe rf ec tly round and hol lo w c ro ss- section s rep re sent
the most s t a b le configur a t i o ns w h en a ir fl o w pa s s es a l ong the f e athe r va n e s fr om t he
ca l amu s (Fig. 6D) . Ho w e v er, wh en airf lo w is perp endicul ar to the f eath er v a nes , f or ked
barb s w it h pa ir ed hal f- c r esc e nt c r oss- secti on exhibit g r e at er s tr e ngth (Fig s . 5 B, 7A) . Thi s
advant a g e ari s e s no t fr om incr ea sed r esis t a nc e of th e pair ed h alf- cr es cent shape it self ,
but b ec aus e th e se fo rk br anc he s c onne ct to f or m a m esh-lik e stru ctur e, pr o vi ding b etter
stabili ty und er pe r pendi c u lar airflo w compar ed t o oth er c r o ss- s e ctiona l sh apes . W he n
su b j ect ed to lat er al airf l ow, b arb r ami with a 1: 1 0 f la t elliptical cro s s - sec t i on ex p eri enc e
le ss displ a c em ent than t hos e w ith a 1 :5 ellipt ic a l c r o ss- secti on (Fig. 5 H) , du e to t he
sma l le r wind war d ar ea o f t he f o r m e r. H o weve r , w h en airf lo w is dir ec t e d fr o m t h e
ca l amu s, b arb s w it h a 1:10 fla t elliptical c r oss- se ction e xhibit the gr e ate s t di s pla c e ment
( F i g . 6 H) , a s th e ir la rg e r w i n dw a rd ar ea ca us es m o re de fle c t io n c o m p a r ed t o the 1 : 5
elliptic a l cro s s - s ec ti on.
The s e finding s indicat e that b ar b r ami w it h a 1:5 el liptical cr os s - se c t io n , d e sign e d t o
simul a t e mod e r n f e ath er bar b s , ar e among the mo s t un stabl e str u c tu r e s (F igs . 5 H a n d
6H). N o neth el e ss, mod er n f e ath er barb s o v erc ome th is s t r uct ura l lim itation by
interconne c t ing into feath er v ane s thr o u g h h ook l et s, for ming an inter loc kin g stru ctur e
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
that is f unc t ionally mor e a dv a nce d th an the m es h - like connection s o bser ve d in ECNU
A32. This inter l o cking enhance s m echanical strength, all o w in g mod er n fea t h e r s t o
with s t a nd airf l ow for c es whi le minimizing s e lf- w eight.
Th e m ak ing o f a t y pic a l fea t her barb :
A bar b, th e primary br a nch of a f e athe r, con s ist s of a r amu s ( main s h aft) lined w it h
ro ws of b ar bu l e s on eith e r sid e 2 . Bar b s f rom differ ent f eathe rs or vari ou s po s i tions al ong
a sing l e f eath er (e .g . , c ov e r t c ontour ) ca n exhibit diver s e cro s s - s ec tion al sh a pe s 1 , w hic h
may a l s o va r y al ong a s ingl e bar b 24 . Alt hough mo st known ext ant fe ather bar b s di s pla y a
proximo dist all y co m pre ss e d , sa nd wi c h-like cr os s - s e c tion with an out er c or tex and a n
internal med ull a 4 , not al l ba r b s or a ll secti ons of a s ingl e bar b cont a in a med ull a ,
dep e ndi ng on the s p eci es a nd f e ather ty pe.
The d ev e l opment of typi cal f e ath er bar bs f ol lo ws a high ly or ganize d proc e s s
involv in g elong ation of ba r b filam e nt s and c e l lul ar d iff e r entiation al o n g thes e
filament s 2, 25 . Init ia lly , a fe w b arb r i dg es f orm on eith er si de of th e r achi a l r i dg e , f oll o w e d
by ne w rid ge s ari sing a s id e all th e w a y do w n t o th e n ew bar b l oc us, whi c h is lo c a t ed
oppos it e, tho ugh not str i c tly, to th e r ac hial ridg e (Fig . 8B, St a g e I). Unlike th e s tationa r y
rachi al ridg e , the bar b r idg es move bo th axia l ly ( par all e l to the l ong axi s of the fol lic l e )
and tang e nt ia l ly ( p e r pe n d ic u lar t o the long a xis of the f olli c l e) d uring gro wth as e xis t ing
ridg es ar e c a r ri ed a way 1, 3 .
A x i a ll y , i n a c r o ss -s e c t io n tr u l y p e rpe n d i c u la r t o t he fo l l ic le ’ s l o n g ax is , al l b a r b
ridg es a ppe ar at ro ugh ly the same d evel opmental s t a g e s in ce they gro w a t an e qu al r at e
(Fig. 8B) . This unif o r mity ena bl e s sp atiotempor al c h ang es in the d evel oping fea t he r t o b e
cha r ac t er ize d 1 . M ean w hil e, t ang e nt ia l mov ement c aus e s th e mor pho logy of the ra m i a n d
barb ul e s to vary along a sing l e bar b due to th e c onc entrati on gradi ents o f sever al
morphog ens acro s s t he pulp pe r ipher y 4 (F ig . 8 B, Stages II and I II). The c ombination of
the a x ia l and tang e nt ia l mov em e n ts r es ult in a he li ca l di s pla c e ment of b arb r i d g e s to ward
the r a c hia l r idg e 1 ( a s se en in seri al cr os s - sec t i ons in Fig. 8) , u ltimate ly inte grating th e
proximal end s of th e b a r b ri dg es into the ra c h ial r id g e ( Fig. 8B, Stag e I V).
His t o l ogic al as s ay s, i nc l uding H& E , Ma sson, and Br dU staining of ser ial
cro ss- s e ction s fr om d evel oping chi c ke n r emex f olli c l e s , reve al t ha t at th e c e llul ar l evel,
ba s il ar c e l l s in t he c e ntr al pa r t of th e e pid er mal c oll a r ar e initially sur r o unde d b y a
thicke r l ayer of in ter medi a t e c ell s (Fi g. S5 A- C). The se inter mediat e c e l ls soon reo rganiz e
into ba r b r id g e s a long th e r adii of t h e fol lic l e, s ep a r at ed fr o m one anoth er b y s ha ll o w
fur ro ws inter c e pt e d by pulp b ulg es ( Fig. S5 D-I) . Sh or tl y aft er th eir for m ation, b arb
ridg es b egin to diff er entiate into the va riou s c ompon ents of b ar b s . Wit hin e a c h bar b r idg e,
a s ingle l ayer of basi lar c e l ls appear s o n either sid e of th e int er me d iat e c e l l s, c r e a t ing t h e
marginal plat e s (Fig. S5 D -I ). The re ma i ning in termedi at e c e ll s a ggr egate i nt o an axial
plate in th e m idd le, f l a nked by bar bul e pl ate s on e a ch si de , an d terminat e s c entr ally at
the ramu s ( Fi g. S5 F , I) .
Cell pr olif er a t ion in t he bar b r i dg e s is not l oc alized in c l ust er s, as Br d U - positiv e c e ll s
ar e scatt er ed thr ougho ut t he epid ermal ba s a l la y er , appe aring not only at the t ips of b arb
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
ridg es bu t a l s o within the f ur r o ws (F ig . 8 E-F . St ag e I). A s th e b arb r idg e s mov e
tangential ly t ow ard t h e r a c hia l r idge, t he b arbu l e plat es r apidl y dif f e r entiat e a nd ar e t he
fir s t t o und e r g o k er a t iniz ation d ur ing fea t h e r morph og enesi s ( Fig. 8, St ag e I ). Apoptosi s
of th e ax i al pl at e r el e a s e s m atur e b arb ul es , whil e ap opt osi s of th e ma r ginal plat e s
separ ate s t he b arb r i dges fr om one anot her . K er atiniza t io n s oon extend s f r om the bar bu le
plate s to th e peripher al w all s ( eq uival ent t o t he dor s a l cort e x of ma t ur e b arb r ami) of t he
devel oping barb r ami (Fig. 8, Stag e I I). A t this s tag e, a high-pr o lifer ati on z one marked b y
more B r d U-po s i tive c e ll s a ppea rs a lon g the epider mal ba sal la y er ( Fig. 8E-F , Sta ge I I),
indicating that the b arb ri dg e s ar e g ro wing to w ar d the f olli c l e center , si milar to t he
rachi al r i dg e 5 .
As t he expanding m ed ull a e pu s h th e bar b rami a ll th e w ay t o w ar d th e per ipher al
feath er sh e ath , ker atinization spr ead s a long the lat e r al cor t ex t o th e ce ntr a l wall s
(eq uival e nt t o t he ventr al c o r te x of mat ur e b arb r a mi) of the dev e lo pin g bar b r a mi ( Fig. 8,
Stag e II I). At t his point, cell prolif er ation d e c r e as es, a n d t he me du ll ae b e c om e fu lly
keratinized once th e p eripher al wa ll s of t h e r ami a t t a ch to th e f eath er sh e a th (Fig. 8 ,
Stag e I V). Th e l a s t c el l s within the b a r b ridg e to ker atin iz e ar e th e b a silar cel ls adj acent t o
t h e base m e n t m e m b ra n e, fo rm i n g t h e v e n t ra l c o rte x o f the m a tu r e ba rb r am i ( F i g . 8 ,
Stag e IV ). In this stage, a ne w zon e of hig hl y prolif er atin g cell s r e a ppe ar s along t he
thicke ne d bas e m ent membrane th at s epa r ate s mat ur e ridg e s from the pu lp (F ig . 8E-F ,
Stag e I V).
While th e cell pro life ration and k e r a tiniz ation pr oc e ss es in the d e v el oping bar b
ridg es fo llo w a simil ar spatiot empora l pattern to th o s e of th e r achi a l r idg e 4, 5 —occurr ing
fr o m the peripher y to wa rd th e cent er along th e r adial axi s of th e pulp—our his t o l ogi cal
section s r e v e al t wo key diff er ences b et ween the dev el opme nt of ba rb ridg e s and th e
rachi al r i dg e :
1. In a devel oping ba r b ridg e, t h e fir s t str uc tur e s to ke r atiniz e ar e the b a r b ul e plate s ,
wh e r e a s in d ev el op i ng r a chi al rid g e, it i s the do rs a l cort e x;
2. The dor s al c ort ex of th e d eveloping r a c hial ri dg e m aint a ins co n stant c ontac t with
t h e fe a t he r s he a t h t h r ou g ho u t de ve lo p m e n t . In c o nt r as t, du r i n g ba r b rid ge
devel opment, th e periph era l w a ll s ( eq u iv al ent to the d ors al c o r tex of matur e b arb
rami) of th e d evel oping barb r ami a r e initially lo ca t ed centr al ly to th e b a r bul e
plate s and gr a du all y move o ut w a rd unt il they appro ach th e f eath er sh e a t h, dr i ve n
by me d ul lar y e xpa n sion.
3. Discussion
Ta ph o nomy eff ec t s:
Despite th e id entif ic ation o f f eath er s w it h v entral ly open r a chi ses in sever al
ca r boniz e d c ompr es sed f os sil s a nd Bu r mese amb er s pe cime ns 5, 26 , v e nt ral ly open b arb
rami hav e not b een doc umente d in a ny know n avia la n or non- a via lan f e ath er ed dino saur s .
This r aise s th e q uesti on of wh ethe r t he s e unu s u al ba r b stru ctur es r epre se nt genuine
morpholo gie s or a r e s imply pr ese r v a t i o n a l ar tifa c t s. Similar t o th e ra c hi s, b a rb r a mi a nd
barb ul e s ar e c ompo s e d of feath er ker atin, whi c h is kno wn for it s re si s tant to wat er,
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
org anic s o lvent s , and m e ch anical alt er ations inclu ding th e po s t mo r t em de gra dation 27 , 2 8 .
This d ur ability is evi d e nt in the pr eser vation of d etai l ed bar b and b ar b ul e st r uc tu r e s in
amber -emb e dd ed f e ath er s 4 , 5 , 29, 30 . G ive n this resili e nc e, it is unlike ly that ba r bul es, b eing
more d elic ate th an bar b rami, wo ul d be mor e r esis t a nt to d eg r ad a t ion tha n the barb
medu ll a and v e nt ral c ort ex. The a b sence of th e m edu ll a and ventr al c ort e x i n th es e
fos s iliz ed ba r bs, simila r to the miss ing r ac hia l med ull a of some ear ly fe ath er s 5 , may be
du e to het ero chr onic de v el opmental tr uncation rath er than taphonomic (pre ser vational)
fact ors . In other w or d s , the mis s ing str uc t ur e s may r esu lt fr om diff e r e nce s in feather
devel opment in the s e anci e nt s p eci es compa r e d t o mo de rn one s, r ath er tha n being lo s t
thr oug h fo ssilization pr oce sse s . T his sugge st s t hat ev o lution ar y c h anges in f e ath er
devel opmental pr oce s s e s, r a t h e r t ha n pr e s erv ation, c ou ld a c coun t f or the ab s e nce of
ce r t ain s tr uc t ur e s lik e the m ed ull a a n d ventral cort e x in f o s silize d feath er s .
Aerodyn amic sig nifica nces of vent r a l ly op en f eather b arbs: U nlike th e
ve nt ral ly op e n fea t h e r r achi s , w hich h as b een d ocum ented not only in the f os s il r e c or d
but a lso in the sum me r plum age of exta nt penguins 5 , th er e i s no ev id enc e of f e ather s w ith
ve nt ral ly open b arb s in either f o ssil th erop od s or mo der n bir d s . T hi s ab s e nc e mak es it
imposs i bl e to unc over thei r d ev e lopm e ntal mech a nisms ( e .g ., t r anscript omic dif fe r ence s
compa r e d t o the s a n dwich- like config urati on of typica l e xtant c ontour f e ather s) or t o
infer t he ir pot ential f unction s . Th e lac k of h ookl e t s in t h e s e ear l y f e ath er s, a s pr ev io u sly
identifie d in o t h e r Burmit e fe ath er s 4, 5 , c o m b i n e wi t h t h e p r e s en c e of v e n t r a l l y o p en
feath er b ar b s , s ugg est l imited a e r o dyn a mic s t ability . C ons i d ering th eir act ual size a nd
stru ctur al w e a kne ss, t h e s e might b e c o nt our f e athe rs that w er e cover ed by oth e r cov ert
c o n t o ur s o r f l i g h t f ea t he rs , r e nd e r i ng t h e i r un s ta b le ae r o dy n a m i c p e r fo rm a n c e le s s
sig nif icant.
When air f lo w pas s e s al o ng th e f eath e r from th e c al amu s, the insta b i lity of f e ather s
with ve nt ra lly op en barb r a m i ( Fig. 6H) s ugg e s t s that they may hav e r esonat e d a t c ert a in
airfl o w s p e ed s . Thi s f urther indi c at e s t hat t hes e ear ly f e athe r s w er e e it her not design e d t o
str eamline the body s urf ac e or th at th e animal s with su c h f eather s like ly co ul d not fl y o r
glid e. The clo s e d van e porti ons of c ov e rt c ont our f e ath e r s in mo de r n bird s ar e eq uipped
with h ookl e t s , whi le the op e n vane s wi thout h ookl et s a l so do not po s s e ss v e n t rally op e n
barb r a mi 1 . Ev e n among mo d ern bir d s th a t ha v e s e c onda r i ly l os t t h e ir ability to fly ( e. g. ,
o s tr ic he s a n d e m us ) , n o fe a t he r s w i th v e n t r al l y o pe n ba r bs ha ve bee n fo u nd 31 , f urther
demon s tr ating that thi s incomple t e d evel opme nt of ba r b rami has b e en ab a ndone d by
mode rn bird s.
Ho wev e r , o ur functional simul ation r esu lts off er ins ight s that c o ul d ins pir e futur e
biomimetic d e s igns. The s plit bar b r am i enha nce their s t ruct ura l st ability b y formin g a
mes h- like s t ru ctur e, sugg esting a s trate gy fo r d e sig ni ng materi al s tha t nee d t o
sig nif icantly re du ce we i ght w hil e maint a ining mec h a nic al s tr e ngth .
Pu ta t ive d evelopment al mechan i sm s fo r bi zarre f eath er barbs:
The g ene r al d evel opmental pr oc ess es of ear ly f e ather s, inclu ding cel lul ar be havior s
in early f eath e r b arb r i dge s, a ppe ar t o be similar to t h ose in ext ant feath e r s 4, 5 . H o w ever ,
whil e alt eri ng t he e xpre ssion s of r e g ul a tor y g ene s can glob al ly a f fect f eath er st ructur es 32 ,
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
repr odu cing t h e exact ba rb mo rphol o gie s o f ea rly fe ath er s in mod e r n g r ow ing f eather
folli cl es r emains c h all enging at pre s e n t. This difficu lty is c ompound ed by the v ar i a bil ity
in bar b ramu s morphol ogi e s , whi ch can differ not only w it hin a sing le f e ath er but al s o
among vari ou s fe a t h e r types. Th e l ack o f e ff e ctive meth od s to pr ec i s ely contro l the sp a t ia l
and tempor al expr es s ion of gen es in devel oping feather f o l licl e s fu r ther compli c at e s th es e
effo rts. I n ad dition, the g en e re gul a t o ry netw ork s that ma y h a v e pr od uce d the uniqu e
morpholo gie s of ear ly f ea t h e r s m a y h a ve d ecay ed o r b e e n entirely l o s t 4 . Co ns eq ue nt l y ,
the c e ll ul ar a nd mol ec ul a r events l eadi ng to t he dev el o pment of v ar io u s ribb o n- lik e barb
rami can only b e inferr ed from t he ge neral dev elopmenta l pr oce sses ob ser v e d in e xtant
feath er s, an appro ac h t r adition all y emp loy e d by P al eo-E v o - Devo stu di es 33 .
In morphotype I f e ather s, ba rb ra mi ca n be e it he r ri b b on- like o r be am-like,
dep e ndi ng on their s p ec i fi c positi ons . Our find ing s s u gg e s t t ha t the a bs e nc e o f the ramu s
medu ll a and v entral c o r te x in t he ventral ly op e n r e gi ons ma y r es ult fro m an ea r l y
tr unc ation of c el l prolif er ation, a fai lur e of diff er entiate d me du ll ary c e ll s to k er atiniz e, or
both, simil a r to previo u s ly r eport e d ve ntrally op en rac hi se s . Cons equ ently, the bar b r a mi
ca n only incr ea se s tr eng t h by e xpa n ding the d or sal c or t ex and may be the ov e r a ll
diamet er (F igs. 5 -6). If a w el l-ker atinize d m e d ull a c oul d b e form ed, th e s oli d be am-lik e
barb r ami wo ul d not be ne c e s sary. Thus, th e r e is no re a s on to a ssum e th a t t is s ue
differ entiation i s mor e adv a nc ed in the beam- like r e gion s. W e th er efo re propo s e th a t th e
solid b e am c on sis t s s o l ely of th e non- ex panded d or s al c o r te x (F ig . 9A , B ii) .
Unlike morphot ype I feath ers, the mo r phogenic pr ocess of morph otype I I fe ath er
ECNU A32 i s inferr ed to have be en m ore sophisti cat ed. Like m or photyp e I feath e r s , t he
medu ll a and ve nt r al c or t ex of bar b r ami a r e abs ent. But th e s plit d or sal cor t e x s ugge st s
that th e axi al plat e , whic h separ ate s th e proxima l a nd di s ta l bar bul e plat e s a nd typically
terminate s upon r eaching the d ev e l oping barb ramu s, ext ends a ll the wa y to the apica l
points of th e d e v e loping b ar b r i dg es th rough out t h e m or phogenic pro ce ss (F ig. 9B v ) . Th e
apopto s is of the far - centr a l ly- ext end e d axia l plate s not only r e sult s in the s pli tting of t he
barb r a mi b ut als o ca us es th e b a s ila r c ell s at th e ap exe s o f a singl e d eve lo pin g bar b r idg e
to d iv id e int o two gr oup s (F ig . 9Bv). M e an whil e , th e marginal plat e s, l ocat ed bet w e e n
adj acent d eveloping b ar b rid ge s and u su all y ext ending to the a pe x es of the b a r b rid g es,
terminate b ef ore r e aching t he apica l basil ar c e ll s (Fig. 9Bv) . Thi s le a ds t o a fai lur e in t he
s e p a ra t i o n o f t h e ba rb ra m i a t th e a p i c a l e n ds o f ba r b rid g es , re su l t i n g i n t h e fus i o n o f
halv es f r om adj acent s plit r a mi at th e b ar b ti p s (Fig . 9Bv). Thi s out com e is f urthe r
facilit ated by th e agg re ga t i on of th e ba s il ar cell s fr om the a djac ent d e v el oping bar b
ridg es.
In morphotype III fea t her s, no tr a c e of a me d ull a has be en identifie d in th e bar b
rami, sugg esting that medu ll ar y c e ll s did not dif f e r entiat e d uring th e mor phog e ne si s of
barb rid ge s . W hil e t h e entir e b arb r a m i ar e c ompose d s o le ly of c ort ex , th ey do not e xpa nd
tr a nsv e r s e ly a s se en in mor pho t y p e I feath er s , but e xp and s d or sov e nt ra lly in s t ea d ( Fig .
9Bii) .
Sever al r egulat or y ge n es involv ed in the norma l d ev e l opmental pr o cess e s of e xt ant
feath er s hav e b e en ide nt ifie d ac c o r ding t o th e w e l l- a c ce pt ed mo de l of f eathe r
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
morphog ene s i s 4, 34, 35 . Fo r e x ample , f ea ther r egen erati on re lie s o n coll ar bul g e r egi on 36 ,
and a molec ul ar gr adi ent of Wnt 3a al o n g t h e r a c h i s- new b ar b lo cu s dir ecti on (eq uivalent
to the ante rior to po st erior dir e ction of Pr um 3 ) play s a r ol e in the v ane fo rmati on in flight
f e a t he rs b u t n o t i n do w n y fe a t he rs 37 . Wnt inhibitors a l s o r e gu lat e f e ather r e ge ne rati on
and a xis f ormation 38 . T h e p a t t e r n i n g o f b a r b r i d g e s m a y b e i n f l u e n c ed b y Ephrin B1 ,
which is expr e ssed in the marginal plat e epithe lum 39 . I n a b i l a t e ra l ly s y m m e tri c fe a t he r,
the top ol ogi e s of the r achi s and th e b ar b gr o wth z o n e ar e pr imaril y r eg ul ate d by GDF 1 0
and GREM 1 , w i t h th e r e t i n o ic a c id s i g n a li n g la nd s ca p e c los el y rel a te d t o f e a t he r
asymmetr y 40 . Sh h ( sonic hedg eho g ) a nd Wnt 3a a r e r eq u i r ed f o r t h e s p ec i f i c at i o n o f
marginal and ax i al plat e cel l s, re s p ecti ve ly 4, 3 4, 3 5 . Whil e Bmp4 and Bmp 2 ar e e s s entia l for
the dif f e r e n tiation of ba r bul e pl ates, Sh h e xpr e s s is minimal in t he se a re a s 34 . Bmp 2 is
tr a nsi ently e x pr e s sed in th e periph era l marginal pl ate s during ear ly r amog enesis be f or e
switching to the b arb ul e plat e epithe li a , indica t ing its r ol e in spec if y ing m a r g i na l pla t e
cell fat e
41 . In ad dition, the e xpre ssion o f MMP2 in the marginal pl ate s of fea the r fo llicle in
ad ult chic k en is inv olv e d in t his pr oc e ss 42 . Ov e r expr es s ion of Bm p4 inc r ea ses ba r b fu s i o n
by pr e v enting the a popto si s of mar g i nal pla t e c e ll s, wh e r eas s uppre ssion o f Shh coul d
produ c e w e b- lik e bar b f usi ons simila r t o th e fo r ked b arb r ami s e en in mor photype II
feath er s 43 .
Accor ding to the w e ll - acc e pte d mo del 34 , the bal a n ce of Nog gin and Bmp4
det ermine s the numbe r , s iz e, and spacing of bar b rid ges , w hi le Shh i s ne gative ly
reg ul ate d b y Bmp s 34, 35 . Ov e r ex pr essi on of Bmp s s u p p res se s Shh e xpr essi o n and t h e
su b s e qu e nt for ma t ion of th e mar gi nal plat es 34 . Conver s e ly, suppr essi on of Bm p s
promote s br a nching, lik el y by enhanc i ng the ri dge-f or ming a ctiv it y of t h e b a si lar cell s
and s pe cifying th e f a t e of m a r g i na l pl at e c e ll s 34 . In ad dition, th e e xp r e ssi on of Ski is a ls o
linked to t h e l os s of the m e d ull a in e x t ant f e ath er s 4 .
B a se d o n c u r r e n t k n o w led g e , ri b b o n -l ik e b a rb ra m i m a y re su l t f ro m a n i m b a la n c e i n
the e x pr e s si o n of Bm p s r e l ativ e to th eir antagoni s t s 43 . T he s p l i t b a rb ra m i o f mo rp h o t y pe
II fe athe r s cou ld h av e forme d thro ugh t he het er otopic expr e s s ion of Shh at the axi al plat e
and Wnt3a at the mar ginal plat e s , which may hav e s w appe d the i dentit ie s of the s e
stru ctur e s in c er t a in d e v el opme nt a l sta ges. Fr om th is per sp e ctiv e, both he t e r ot opic
(spati al) and h e t er ochr o ni c (t empor al ) expre s s ion s o f th e s e re g u lat or y g e nes dur ing
feath er morph og e ne si s coul d h av e c o ntr ibuted t o the f o r mati on of the s pli t, ve nt ral ly
open ba r b rami. The a b s ence of a m edu lla in morphotype III fea t he r s might be attri b u ted
to Ski ex pr e ssi o n 4 . Colle c tiv ely, var iou s evo- d e v o mec h anism s may acco unt for t he
unus ual b arb mor pho logie s ob s e r v e d in e arl y f e ath er s , a waiting v e r ific a t ion as tec hniq ue s
be c om e avail abl e .
Tempora l devel opmental sh i fts d rive morph olog ic a l vari a tio n s a long barb
rami:
Since th e di s ta l por ti on of a f eath er m atur es fir s t w hil e the pr o x imal c ounter par t i s
still und erg oing mor phogene s i s, any tempor a l phy s iol ogi c a l and/ or e n v i r onme nt a l
c h an g e s t h at t r i g g er a p h e n o t y p i c r e s p o n se co u l d m a n i f e s t al o n g t h e p r o x i m a l ( i m m a t u r e )
to dist al (matur e ) axi s of a d e v eloping fe ath er 3, 5 . This phenom enon is wel l d e mo n s tr at e d
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
by regi onal morph olo gical c h anges ind uc ed thr o ug h th e tempora l e xpre ssion of ectopi c
genes by r e plic ation-c ompe t e nt a vian sar coma- le uko s i s v ir u s ( RCA S) 4, 4 4 , whic h has b e e n
instrumental in stu dying f e athe r m o r phoge n e sis. F rom t hi s p e r s pe ctive, an d b ase d on
cellu lar and mol ecul ar me ch anis m s inferr ed f rom mod ern fe ath e r re gene r a t ion, t he
tempora l d evelopm ental s hift s of bar b r ami for e ach fe ath er morphotyp e can be de sc r i bed
as f ol lo w s :
Morphot ype I feat her b a rbs (Fig . 9Bii): The se bar b s ar e ch ara c t er ize d by a
solid dist al b e a m that transiti ons into a pr oxim a l r ibb on-lik e, ventr all y op en ramu s. T wo
devel opmental s c en ario s ar e pr op ose d : ( i) A t the onset of ba r b r a m i form ation, dor s a l
cort ex differ entiati on be gins , prod ucin g a s olid di st a l b ea m w ith minimal ex pa nsion of
t h e do rs a l c or t ex a n d n o d i f fe re n ti a tio n o f o t he r ba rb r a m us t i ss ues (F i g . 9 Bi i , t p 1 ) ; a n d ( i i )
The d or s al c or t ex e xpand s in w idth , bu t med ul la diff er entiation is abs ent or present b ut
fail s t o k e rati niz e , r esu lting in ventra l ly open b arb r ami (Fig. 9Bi i, tp2). Th is i s likel y
eq uiv al ent to s t age II dev e l oping s a nd w ich-like bar b r ami of e xtant fe ath e rs ( F i g . 9 A ) .
Morphot ype II f eath er bar bs ( F ig. 9Bii i and v) : The d e v el opmental pr o c e s se s
of mor photype I I feather bar bs can b e ex pl aine d by thr ee po ss i bl e sc enario s : (i) Bas il ar
cell s ke r atinized without fur th e r prolif e ration and diffe rentiati on, for ming the so lid di s t al
be a m ( Fig. 9B iii, tp1 ) ; (ii) Th e d or sa l c o rt e x e xpa nd s in w idt h , b u t m e du ll a diff er entiation
is ab sent or pre sent but fa i ls to ker at inize (Fig. 9Biii, tp2); (iii) Th e exp a n ded dor s a l
cort ex split a long t he mid line (Fig. 9 Biii, t p3); and (iv) E ach br anch of th e split b ar b
r a m u s f u s e d t o b r a n c h o f t h e a d j a c e n t b a r b r a m u s i m m e d i a t e l y b e f o r e m e r g i n g i n t o t h e
rachi s ( Fig. 9Biii , t p 4) .
Morphot ype I II feat her ba rbs (Fig. 9 B i v ) : T hes e ba r bs ar e like ly f o r m ed s o le l y
by the cort e x w it h out pro ximo dist all y exp and e d . Alth ou gh no med ul la i s pro duc ed, thi s
su g g es t s that the b arb r ami c an be f ully k e r a t iniz e d a l l the w a y t o the v e ntr al ape x .
Ther efor e, w e explain t hi s a lte r n ative d evel opmental p ath wa y is m or e advanc e d
compa r e d to the soli d b eam and ventr a l ly op en b arb r a mi of morph otype s I and II
feath er s.
Taken t ogethe r, the app e ar a nc e o f r ibbon-like b ar b r ami o f fe r s thr e e key
ev o lut i o nar y ins ig hts: (i) Widespread absence of med ul la i n early f eathers: Th e
lac k of a me du lla points t o limite d ti ssu e diff e r e nt ia t ion within the r a c his and b a r b s
duri ng th e ea r ly s t age s of f eath er ev ol ution. This is suppor te d b y o bserv atio ns that th e
rachi s i s no thicker th an th e f eath er b arb s in s e ver al B urmite fe ath er s. Th e inability t o
form well -ker atinize d, br anching , sand wic h -li k e stru ctur es like ly m eant t ha t mec h anica l
str engthening c o uld only be achi eve d by expanding the d or sal cort e x a nd possibly
increa s ing th e o v era ll diam eter , r e s u l ting in ventrally open r achi s e s and barb r a mi.
Ho wev e r , confirming t his c oncl us iv el y remains ch all eng ing d ue t o the p oor l y pr eser ved
det ail e d str uct ur es in car boniz e d f o ssils.; (ii) Development al plasti city enrich es
feat her morph ology: The pr e s ence of ventr ally open r e g ions inter s p er sed w it h solid
be a m s e cti ons al ong a singl e ba rb r a mus sugge st s that s pati otempor al de v el opme nt a l
plas t icity played a sig n ific ant rol e in e ar ly feath e r s . This plasti city likel y all o w e d for
cons i d era bl e va r i ability in fe ather stru ctur e , contr ibuting to t he m o r pho lo g ic al div er s i ty
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
ob s e rve d in e arly f e athe rs . ; and (i i i ) Lat er ev o lution ar y em ergence o f a medu lla:
The c onsi s tent pre senc e of a me du ll a within rac hi s es and b arb rami appe ar s t o hav e
devel op ed l at e r in fe ath er ev o lut ion, ind icating that f eath er br a nching evolv ed be f or e th e
comple t e stabili zation of tiss u e diff er entiation w it hin the r ac hi s and b arb s.
Co nc lu s i o ns:
Our findings s ug ge s t s th a t the tissu e differ entiation w i thin th e rachi s e s an d barb
rami of e arly f e ather s w as qu ite limited . I n some fe a t h e r s th at eith er fail e d to form a
medu ll a or f orm ed one b ut f ail e d t o ker atinize, r achi s and b arb r ami c ompensat ed f or
stru ctur al w eakne s s by e xpanding t h e dor s al c or t e x . Thi s l ed t o t h e v ent rall y op e n
rachi se s a nd bar b s , whic h a r e r a re ly o b s erv e d in mo d e r n f eath e r s . T he a bs e nc e of a f ull y
form ed medu ll a a nd ventr al cort ex inc r eas e d t he de vel o pmenta l pla s t ic ity of r achise s a nd
barb s in e ar ly fe a t h e r s, w hi ch likel y g iving rise to numero u s fe a t her mor photypes not
seen in mod ern bir d s. V ari a t ions in d or sa l cort e x e xp ansi on and de v e lopme nt al pla sticity
of me du ll a r y t i ssu es c ontr ib ute d t o t he mo r phol og ic al diver sity of ear ly f eath er s, in
ad dition to th e br a nching pa t t erns. A s the me d ul la and v entr a l co rte x f or med mo r e
cons i st ently ove r time, th e sand wich -like config urati on of mo d ern f eath e r br anc h e s
stabili zed, a nd th e d evel opmental pl a sti city of t h e r achi s and bar bs g r a du a l ly di minishe d.
References
1 Luc as, A . M . & Ste tt e nhe im , P . R . A vian anatomy : In te gu ment . 1-750 (U nited S t ates
Depart ment of Agr icultu re , 19 7 2) .
2 Prum , R . O. & Dyck, J . A hier arch i ca l model o f p lu m age: Mor p ho logy, deve l op m ent ,
and e volutio n . J o ur n al of E x p er ime n t al Zo o l o g y P a rt B : Mol e cu l a r a n d De v el o p men t a l
Evolution 298B , 7 3-90 , doi :10 .10 02 /je z .b . 27 (2003) .
3 Prum , R. O . Dev e lop me n t and ev o lut i ona ry or i g in o f feat hers. J ou r n al of E x p er i men t al
Zoolo gy 285, 2 91-306 (1999) .
4 Chang, W .-L . e t a l . The mak i ng of a f l i g ht feathe r: Bi o- ar c h ite c tu ral p ri nc i ple s and
adap ta tion. Ce l l 179 , 1 409 -1423.e1417 , d oi:1 0. 101 6/ j.cel l.2019 .11 .0 08 (2019).
5 Wang, S. e t a l . Va r iat i ons of Mes o zoic f eathe r s : Ins ig h ts fro m the morph ogene s i s o f
ext an t fea th er ra chi s es . Evolu tio n, 2121 - 2 133 ( 2020) .
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
6 Xu, X . in The evo l u tio n o f fe at hers : F ro m their or ig in to th e prese nt ( e d s C h r i s t i a n
Foth & O l iver W . M. Rauhu t) 67 -78 (Spr ing er Inte rna ti o nal Publi shi n g, 202 0 ).
7 Xu, X . & G uo, Y . The o r igi n an d ea r l ie r e volutio n of fe a ther s: I nsi g hts from rec en t
pal eon to lo gi cal and ne on to lo gi cal da ta . V ertebr ata Pa lAs ia t ica 47, 3 11-32 9 (2009) .
8 Prum , R ich a rd O . & B rus h , A lan H . Th e evo lu t iona ry o ri gin and dive rs i f icat ion of
feathers . The Qua rte r ly R eview o f B iolog y 77 , 261 -295, d oi :1 0 .1086 /3 41993 (2 0 02 ).
9 Haug, C . et a l . Comme n t o n th e l et t er of th e So ci et y of V e rt e br a t e Pa l e ont ol og y (SV P )
dated Ap ri l 21, 202 0 regar di ng “Fo s s ils fro m conf lic t zone s and r e p roduc i b il ity o f
fo s s il -ba se d s c ien t i f ic dat a ” : the importa nc e of p ri vate c ol le ct io ns. Pa l Z 94, 41 3- 429 ,
doi : 10.10 07 /s 1254 2- 020-0 052 2- x (2020 ).
10 Fis che r, A . H ., J acob s on, K . A ., R ose, J . & Zel l e r, R . He ma to xy lin and eosin s ta in ing
of t is sue and cel l s ec ti o ns. Co ld Spr ing Har bor Pr o toc ols 2008 , pdb.pr ot49 86 ,
doi : 10.11 01 /pd b.p rot 4 986 (2008) .
11 Lee, S . H ., Fu, K. K . , Hu i, J. N . & R i ch m a n, J . M . Nogg in a nd ret ino i c acid trans f o r m
the i den t it y of avian fa ci al pr o m ine nc es . Na t u r e 414, d oi :10. 1 038/ 4149 09a (20 01 ) .
12 Suvik, A . & E ffend y , A . The us e o f mod i f ied Masson ’s t ri chro me sta ining in colla gen
eva lua t i on in wound hea li ng s tud y . Mal J Vet Res 3 , 39 -47 (20 1 2) .
13 AN SYS FLU ENT 12.0 U ser's G u ide ,
(200 9).
14 Bektaş, M., G u le r, M. & Ku rtu lu s , D . O ne - way FS I anal y s is of bio-insp i r e d f l ap ping
wings. 6 , 1 7 2-194 , d oi :10 .1 504 /I JSA.20 2 0.10034459 (2020 ).
15 Wu, K. , Ye, Z ., Ye , K . & Hong , Z. Mecha n ical chara c te ri s t ics o f the defor ma tion o f bi r d
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
feathers in a ir flow . Chine se J ou rna l o f Th eoret i ca l a nd App li ed M ec han ic s 55 , 874 - 884 ,
doi : 10.60 52 /04 59-1879-22 -520 (2023) .
16 Herte l, H. St r uc t ur e, f o rm, mov e men t . 1-251 (R e i nhold Pub lish i ng Corpo ra t io n ,
1966 ) .
17 Macleod , G. D . Me cha ni c al prope r ti es of c on tour fe a the rs. The J o urna l o f
Exper i men ta l B i o logy 87, 65- 71 (198 0) .
18 Bonse r , R . & P urslow, P . Th e Young's modu l us o f fe a ther kera ti n . The Jo urnal of
Exper i men ta l B i o logy 198, 10 2 9-103 3 ( 1 9 95).
19 Sull i van , T . N. , Wang , B . , Esp inosa , H . D. & Mey ers , M. A . Extre me ligh twei gh t
stru c tu res : av i an fea the r s and bones. Ma te rials Today 20 , 377 - 391 ,
doi : ht tps ://doi.org/10.10 16/j. ma tt o d.20 17. 0 2.004 (2017) .
20 Bachm ann , T. , E m me r lich, J., Bau mga rtner, W. , Schneide r, J. M . & Wagner , H.
Flex u ral sti ff nes s of feathe r s h af ts : ge o m etry ru l es ov e r ma ter i a l p roper ties . J Exp Biol
215, 4 05-41 5, d oi:10 .1 242 / je b.0 59451 (2 012).
21 Bonse r , R . & Pur slow , P . The Young's modu lu s of fe ath er ke rat in. J Exp B iol 198 ,
1029 - 1 033 , doi :10 .124 2 /je b.1 98 .4 .10 29 ( 1995).
22 Pursl ow , P. & V in cen t, J. Mechanica l p r o per t ies of pr i mary feather s f rom the pige on .
Jou r n a l o f Ex pe ri mental B iology 72, 251 - 260 (1978) .
23 Wang, B. & Mey er s, M . A . Sea gu l l f eather s ha ft : C orr ela ti on betw ee n st ruct ure a nd
mec han ic a l response. Ac ta Bio m ateria li a 48 , 270 - 288 ,
doi : ht tps ://doi.org/10.10 16/j. actbio.2016.1 1.006 ( 2 017 ) .
24 Muelle r, J . & G ib so n, L . J. S tru c tu re a nd mec han ics of w ater -hol din g fea t her s of
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
Namaqua s andgr ous e ( Pte rocles n a m aqua). J R Soc I n ter face 20 , 20 220 878 ,
doi : 10.10 98 / rs if .20 22 .0878 ( 2 0 23) .
25 Lin , G. -W ., L i, A. & C h uong , C .- M . in The Ev o lu ti o n o f F e a t he rs : Fr om Th eir O ri gin t o
the Pres en t (ed s Chr istian Foth & Ol iv e r W . M. Rauhut) 13 -26 (Sp ri n g er
Inte rnational P u b li s hin g, 2 020 ).
26 O' C on n or , J. K . , C h i appe , L. M. , Ch uon g, C.- M ., Bott je r, D . J . & You, H. Ho mol og y a n d
potentia l cel lul a r and molecu l a r mechan isms for the d ev e lopment of unique fea t h er
morp h olo gies in ear ly b i r ds. Geos c ien c es 2 , 157 - 177 ,
doi : 10.33 90 /ge oscienc es 203 0157 (20 1 2) .
27 Moy e r, A. E ., Zheng , W . & Sch w e itzer , M . H . Ke ra ti n du rab i l i ty has imp li c at ion s for the
fos s il recor d : Resul ts f ro m a 1 0 year fe athe r degr a da t ion ex pe ri me n t. P l oS one 11 ,
e015 7699, doi: 10.1371/ jour n al .pone.015 769 9 ( 2016) .
28 Mc C oy, V. E . e t a l . A n c ient a min o a c i ds from fo s s il fea th er s in a mber. Sci ent if ic
Reports 9 , 6 420, d oi:1 0.1 03 8/s41598-0 1 9 - 4 2938 -9 (201 9 ).
29 Xing, L., C oc kx , P. & McKe l l ar , R . C . D is a ssociated fe ath ers i n B ur mese am b er s hed
new li g ht on m id-C reta c eous dinos a urs an d av ifauna . Go ndwana R es ea rch,
doi : ht tps ://doi.org/10.10 16/j. gr .2019.12. 0 17 ( 2 0 20) .
3 0 X i n g , L. , Coc kx , P ., Mc Ke l l a r, R. C. & O’ Con n o r, J. Or n ame nt a l f e a t h e r s i n C r e ta c eou s
Burme s e a mbe r: Resolv in g the e ni g ma o f r ac h is-d o m ina te d fea ther struc tu re. Jou r nal
of Pa la eoge ography 7 , 13, doi :1 0.118 6 /s 4250 1-018 -0014 -2 ( 2 018 ).
3 1 G i l l , F . B . , P r u m , R . O . & R ob i n s on , S . K . Orn it h olog y . 4 th edn , (W . H . F reema n a nd
Company, 2019) .
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
32 Ng, C. S . & L i, W .- H . Genet ic a nd m olec u lar b as is o f f e athe r d i ve rs i ty in b i rds .
Ge no me Bio l ogy a nd Evo lu t ion 10 , 257 2- 25 86, doi:10 .1093/gbe /ev y18 0 ( 2 018 ).
3 3 T h e wi s s e n, J . G. M ., Coop e r, L . N. & B e hr i n g er, R. R. Dev e l o p m en t al b i o l o gy e nri ch e s
pal eon to lo gy . Journa l o f Ver te brate P a l eont ol og y 32, 1 223-1 234 ,
doi : 10.10 80 /02 7246 34 .2012.707717 (201 2).
34 Yu, M., Wu, P ., W ide litz , R . B . & Chu ong, C. -M . The m or p ho g ene s is o f feat h ers .
Nature 420 , 30 8, doi:10 .103 8 /na ture 0 119 6
http s : //ww w .natu re .c om/a rt i c les/na ture0 1196#s upp lementa ry -i n forma t i o n (2002 ).
3 5 H a r r i s , M . P . , F a l l on , J . F . & P r u m , R . O . Shh - Bm p 2 signa ling modu le and the
evo lu ti o nary orig in and div e r sif ic a t ion of fea the rs . Jo u rna l o f Exper imental Zo o logy
294, 1 60-17 6, d oi:10 .1 002 / je z. 1015 7 (20 02).
36 Yue, Z. , Jia n g, T. -X. , W i de l i tz , R . B. & C hu ong, C. - M. M apping ste m c e ll act iv i t ie s in
the fe a ther foll icle. Natu re 438 , 1026 -1029, doi : 10. 1038/natu re04222 (2005 ).
37 Yue, Z . C ., J ian g, T . X ., W i de li tz, R . B . & C huong , C . M. W nt3a gr adie n t c o nv erts
radial to bila tera l feather s y m me t ry vi a t opol og ical ar rangem ent of epi t he li a. P Na t l
Acad Sci U SA 103 , 951 -955, d oi :DO I 1 0. 1073/p na s.05 0 689410 3 (20 06) .
38 Chu, Q. e t a l . Dk k 2/F rzb in the de rm al pap illae regula tes feath er rege ne rat i on .
Develo p men tal B i o logy 387, 167 -178 , d oi : h t tps : / /doi .or g /1 0.101 6 / j. ydb i o.2014.01 . 010
(2014) .
39 Suks aw ea ng , S ., Jiang, T . X. , Ro ybal, P., C huo ng , C. M. & W idel i t z, R . Roles of
EphB3/ephr i n -B 1 i n feathe r morpho genesi s . I n t J De v Bi ol 56, 719 -7 28 ,
doi : 10.13 87 / i jdb.120021rw ( 2012) .
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
40 Li, A . et a l . Di vers e fea ther shape e v olut i on enabled b y cou p l ing an isotrop ic signal lin g
modules wi th se lf-o rgan izing bra nc h in g p rogra mme . N atu re Com mun ic a t ion s 8 ,
nco m ms1 41 39, doi:10 .10 38 /ncom ms14 1 39 (20 17 ) .
41 Harr is, M. P. , Fal l on , J. F. & Pru m , R. O. Shh-B mp2 signa ling m odu le and the
evolu ti onary o r igi n and di ve r si fi c at io n o f fea the rs . J Ex p Z o ol 294, 16 0- 176 ,
doi : 10.10 02 / jez .1 0157 (2 00 2) .
4 2 J i a n g, T .- X., T u a n , T. L ., W u , P. , Wid e l it z , R. B. & Ch u o n g , C.-M . Fr om b u d s t o f o lli c l e s :
Mat r i x me talloprot e inas es in deve l o p mental tissue remo de l i ng dur i ng fea th er
morp h ogenesis . D i f fe rent iat ion 81, 307 -3 14 ,
doi : ht tps ://doi.org/10.10 16/j. di ff .2 011 .03 . 0 04 (20 11 ) .
43 Yu, M. , Wu, P. , Wid e l it z , R. B . & C h u o ng, C . M. T h e mo r p hogenesis of fea th ers .
Nature 420 , 30 8-312 , doi :10.1038 /nature 01196 ( 2 002 ).
44 Ng, C . S . et al . The chicken fr izzle fe a the r is due to an α-k e r atin ( KR T7 5 ) mu ta ti on t ha t
cau ses a de fect ive rac h is . PLoS Gen e t ics 8 , e1 00 2748,
doi : 10.13 71 / jou rnal.pgen .10 02 748 (20 12 ) .
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
Figure 1∣ Morphotype I feathers showing ventrally open barb rami.
Left (A-L): Photographs of isolated feathers showing ventrally open barb rami: (A-D) images of
ECNU A143-1, (E-J) images of ECNU A147, (K-N) images of ECNU A148; red boxes in (A)
indicate the areas shown in close-up images (B-D); red boxes in (E) and (F) indicate the areas
shown in close-up images (F, H, I, J) and (G), respectively; red boxes in (K) indicate the areas
shown in close-up images (L-N). Paired black arrowheads in (D) mark the position where barb
ramus is ventrally opened, whereas paired blue arrowheads in (N) highlight where solid barb
ramus is present. Right (O-Q), Schematic drawings illustrating general morphologies (right) and
cross-sections of rachis and barb rami (left) at various positions along the corresponding feather.
Red dashed lines indicate where cross-sections were taken, blue silhouettes represent solid (circle)
or ventrally open (crescent-shaped) rachial cross-sections, while green silhouettes indicate solid
(circle) or ventrally open (crescent-shaped) barb rami cross-sections on either side of the rachis
(not to scale). Scale bar: 0.5mm for (A), (E), (F), and (M); 0.2mm for (B-D), (G-J), and (N-L); and
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
1.5mm for (K). (O-Q) are not to scale.
Figure 2∣ Morphotype I feathers showing ventrally open barb rami.
Left (A-G): Photographs of isolated feathers showing ventrally open barb rami: (A-D) images of
ECNU A143-2, (E-F) images of ECNU A49. Red boxes in (A) indicate the areas shown in
close-up images (B-D); while red boxes in (E) highlight areas shown in close-up images (F-G).
Paired black arrowheads in (C) mark the position of the ventrally open barb ramus. Right (H-I),
Schematic drawings illustrating general morphologies (right) and cross-sections of rachis and barb
rami (left) at various positions along the corresponding feather (H) or barb (I). Red dashed lines
denote the locations from where cross-sections were taken, blue silhouettes represent ventrally
open (crescent-shaped) rachial cross-sections, and green silhouettes indicate solid (circle) or
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
ventrally open (crescent-shaped) barb rami cross-sections (not to scale). Scale bar: 0.5mm for (A)
and (E); 0.2mm for (B), (D) and (G), 0.1mm for (C) and (F). (H-I) are not to scale.
Figure 3∣ Morphotype II feathers showing the split ventrally open barb rami.
Photograph (A), 3D segmentation (B) and schematic drawing (C) of the isolated feather ECNU
A32. Red boxes in (A) highlight areas shown in close-up images (D-H), while red boxes in (F)
indicate areas shown in close-up images (G) and (I). Paired black arrowheads in (G) and (I)
indicate positions where the ventrally open barb rami are present. Representative sections (J-P)
from the 3D segmentation of ECNU A32 highlight the split barb rami (red arrowheads) in dorsal
(J) and ventral (J’) views. Red lines in (K) mark the positions where the cross-sections shown in
(L) and (M) were taken. (N-P) each highlights a barb fork, which connect to form a unique
mesh-like structure (Q, not to scale), with red lines marking the positions of cross-sections of the
barb rami shown as green silhouettes. Scale bar: 1.25mm for (A-C) and (J-P); 0.2mm for (D-I’).
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
Figure 4∣ Morphotype III feathers showing plate-like barb rami without medulla.
Left (A-I): Photographs of isolated feathers showing plate-like barb rami: (A-E) images of ECNU
A101, (F-I) images of ECNU A149. Red boxes in (A) indicate the areas shown in close-up images
(B-E); while red boxes in (F) highlight areas shown in close-up images (G-J). Right (K-L),
Schematic drawings illustrating general morphologies (right) and cross-sections of rachis and barb
rami (left) at various positions along the corresponding feather. Red dashed lines mark the
locations where cross-sections were taken, blue silhouettes represent solid (circle) or ventrally
open (crescent-shaped) rachial cross-sections, and green silhouettes indicate plate-like barb rami
cross-sections on either side of the rachis (not to scale). Scale bar: 0.5mm for (A) and (G-I); 5mm
for (F); 0.2mm for (B-E). (K-L) not to scale.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
Figure 5∣ Simulations illustrating the maximum displacements of feather barbs with various
cross-sectional shapes under airflow perpendicular to the feather vanes.
The models simulate: (A) morphotype I feather barb rami with crescent-shaped cross-sections; (B)
morphotype II feather barb rami with paired half-crescents cross-sections; (C) morphotype III
feather barb rami with 10:1 flat elliptical solid cross-sections; (D) idealized, non-realistic tubular
barb rami with perfectly round and hollow cross-sections; (E) modern flight feather with 5:1
elliptical hollow barb rami cross-sections; (F) idealized, non-realistic tubular barb rami with
perfectly round and solid cross-sections; and (G) idealized, non-realistic barb rami with 5:1
elliptical solid cross-sections. Resultant displacement (URES) is expressed in millimeters, with the
symbol next to each model indicating the shape of barb ramus cross-sections. A comparison of
aerodynamic performance of each simulated model is shown in (H), with the x-axis representing
airflow speed and the y -axis representing displacement. The results demonstrate when airflow is
perpendicular to the feather vanes, barbs with paired half-crescent cross-section exhibit the
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
greatest strength (B, H), whereas barbs with a 10:1 flat elliptical solid cross-section exhibit the
least strength (C , H). This highlights the significant impact of cross-section shape on barb
flexibility and deformation under perpendicular airflow.
Figure 6∣ Simulations illustrating the maximum displacements of feather barbs with various
cross-sectional shapes under airflow passing along feather vanes from the calamus.
The models simulate: (A) morphotype I feather barb rami with crescent-shaped cross-sections; (B)
morphotype II feather barb rami with paired half-crescents cross-sections; (C) morphotype III
feather barb rami with 10:1 flat elliptical cross-sections; (D) idealized, non-realistic tubular barb
rami with perfectly round and hollow cross-sections; (E) modern flight feather with 5:1 elliptical
hollow barb rami cross-sections; (F) idealized, non-realistic tubular barb rami with perfectly round
and solid cross-sections; and (G) idealized, non-realistic barb rami with 5:1 elliptical solid
cross-sections. Resultant displacement (URES) is expressed in millimeters, with the symbol next
to each model indicating the shape of barb ramus cross-sections. A comparison of aerodynamic
performance of each simulated model is shown in (H), with the x-axis representing airflow speed
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
and the y -axis representing displacement. The results demonstrate that when airflow passes along
feather vanes from the calam us, barbs with a hollow cross-section are the most stable
configuration (D, H), whereas barbs with a 5:1 flat elliptical hollow cross-section exhibit the least
strength (E, H). This highlights the significant impact of cross-section shape on barb flexibility
and deformation under parallel airflow.
Figure 7 | Aerodynamic performance of ECNU A32 (A) and an extant contour feather of
similar size (B) under increased airflow, showing the different in displacement cause by
structural variations.
The x-axis represents airflow speed, and the y -axis represents displacement. The results show that
the basic structure of the extant contour feather is completely destroyed when the airflow speed
reaches 6 m/s, leading to an unrealistic displacement of 550 mm. In contrast, no apparent
structural destrusction is observed in the ECNU A32 under a maximum airflow speed of 10m/s,
demonstrating that the amber-embedded feather with forked barbs exhibits greater mechanism
strength.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
Figure 8 ∣ Tissue differentiation (H&E) and cell proliferation (BrdU) of a regenerating 5 th
primary remex from a one-year old chicken. The p hoto of the regenerating remex is shown in
(A), with the red box indicating an enlarged image of the sampled developing portion within the
follicle, and the yellow lines mark positions of slices shown in (B-F). (B-D) H&E staining,
illustrating barb morphologies at different levels of the regenerating remex; and (E-F) BrdU
staining, highlighting cell proliferation at each level of a regenerating remex. Yellow boxes mark
the areas shown in the corresponding close-up images. Scale bar: 2mm for (A); 1mm for (B);
0.2mm for (C) and (E); 0.1mm for (D) and (F)
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
Figure 9 ∣ Schematic illustration showing the diverse barb morphologies resulting from
incomplete developed barb rami. (A) Cross-sections of a developing extant flight feather
demonstrating the morphological changes of barbs during maturation (Left), alongside
cross-sections of mature barb rami in three Mesozoic feather morphotypes (I, II, and III),
highlighting the diverse barb morphologies observed in Mesozoic feathers correspond to various
stages of incomplete development in extant feathers (right). The red dashed lines indicate the
positions where cross-sections are taken. (B) Schematic depiction of temporal developmental
shifts along barb rami of each morphotype of early feather. (i) a model of growing feather with the
sheath and pulp omitted, highlighting barb arrangement within feather follicle (modified from
Lucas and Stettenheim
1). The black barbs indicate those depicted in (ii-iv), illustrating that the
distal tip of the barb forms first at the new barb locus while the proximal part continues to grow as
they approach to the rachis (time points 1-4 are indicated as tp 1-4). (ii-iv) morphological
variations along the barb rami (left) produced by temporal developmental shifts (right) for each
feather morphotype. The cross-sections show only the relative positions of the barb ridges at a
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
given time point when the corresponding part of the barb rami is formed; this does not imply that
all depicted barb ridge morphologies are present in the cross-section at a given time. (v) enlarged
view of barb ridges in a developing morphotype II feather, highlighting the cellular and tissue
features associated with morphological changes along the barb rami.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted December 12, 2024. ; https://doi.org/10.1101/2024.12.07.627359doi: bioRxiv preprint
Text is read by the "Ask this paper" AI Q&A widget below. Extraction quality varies by source — PMC NXML preserves structure cleanly, OA-HTML may include some navigation residue, and OA-PDF can have broken hyphenation. The publisher copy (via DOI) is the canonical version.
My notes (saved in your browser only)
Ask this paper
Answers must be backed by verbatim quotes from this paper's full text. Hallucinated quotes are dropped automatically; if no verbatim passage answers the question, we say so. How this works
Citation neighborhood (no data yet)
We don't have any in-corpus citations linked to this paper yet. This is a recent paper (2024) — citers typically take a year or two to land, and the OpenAlex reference graph may still be filling in.
Source provenance
- europepmc
- last seen: 2026-05-20T01:45:00.602351+00:00
- unpaywall
- last seen: 2026-06-13T06:42:57.164913+00:00