EVOLUTIONARY MORPHOLOGY OF THE
EXTREMELY SPECIALIZED FEEDING
APPARATUS IN SEAHORSES A N D
PIPEFISHES (SYNGNATHIDAE)
Part l - Text
Thesis su bm itted to o b ta in th e degree
of D octor in Sciences (Biology)
P ro efsch rift voorgedragen to t h e t
bekom en van de graad van D octor
in de W eten sch ap p en (Biologie)
A cadem iejaar 2010-2011
Rector: Prof. Dr. Paul van C auw enberge
D ecaan: Prof. Dr. H erw ig D ejonghe
Prom otor: Prof. Dr. D om inique A driaens
EVOLUTIONARY MORPHOLOGY OF THE
EXTREMELY SPECIALIZED FEEDING
APPARATUS IN SEAHORSES A N D
PIPEFISHES (SYNGNATHIDAE)
Part l - Text
Heleen Leysen
Thesis subm itted to o b ta in th e degree
of D octor in Sciences (Biology)
P ro efsch rift voorgedragen to t h et
bekom en van de graad van D octor
in de W eten sch ap p en (Biologie)
A cadem iejaar 2010-2011
Rector: Prof. Dr. Paul van C auw enberge
Decaan: Prof. Dr. H erw ig D ejonghe
Prom otor: Prof. Dr. D om inique A driaens
R e a d i n g * a n d e x a m i n a t i o n c o m m it t e e
Prof. Dr. Luc Lens, voorzitter (Universiteit Gent, BE)
Prof. Dr. Dominique Adriaens, promotor (Universiteit Gent, BE)
Prof. Dr. Peter Aerts (Universiteit Antwerpen & Universiteit Gent, BE)*
Prof. Dr. Patricia Hernandez (George W ashington University, USA)*
Dr. Anthony Herrei (Centre National de la Recherche Scientifique, FR)*
Dr. Bruno Frédérich (Université de Liège, BE)
Dr. Tom Geerinckx (Universiteit Gent, BE)
Dankwoord
H et schrijven van dit doctoraat was me niet gelukt zonder de hulp, raad en steun van een
aantal mensen. Een woord van dank is hier dan ook gepast.
Allereerst wil ik Prof. Dr. Dominique Adriaens bedanken voor alles wat hij de afgelopen
jaren voor mij heeft gedaan. Er zijn veel zaken die het goede verloop van een doctoraat
bepalen en de invloed van de prom otor is volgens mij een niet te onderschatten factor. Ik
ben dan ook heel blij met Dominique ais promotor, hij stond letterlijk dag en nacht klaar
met goede raad, suggesties voor nieuwe technieken, nuttige artikels, correcties van
manuscripten en hulp bij statistische verwerking. Ik wil hem bedanken voor zijn
fenomenale kennis en expertise waaruit ik gretig heb kunnen putten en voor de vele
interessante congressen en workshops waar hij me naartoe heeft laten gaan, maar vooral
om me de kans te geven aan dit doctoraat te beginnen en om me te helpen het tot een
goed einde te brengen. Echt heel erg bedankt!
Dit onderzoek maakt deel uit van ‘het zeepaardjesproject’, wat niet alleen financiële
steun van het Fonds voor W etenschappelijk Onderzoek (FWO) betekende, maar ook een
zeer interessante en aangename samenwerking met het Labo voor Functionele
Morfologie van de Universiteit van Antwerpen. Vooral Gert heeft me heel erg geholpen,
zijn boeiende experimenten en onze discussies over hoe het hyoid nu toch precies werkt,
zetten me telkens weer aan het denken. Ook Sam verdient een woord van dank voor zijn
feedback en alle modelleringswonderen die hij heeft verricht. Prof. Dr. Peter Aerts wil ik
bedanken voor zijn constructieve kritiek
en enthousiasme tijdens de
projectvergaderingen en op congres. Dr. Anthony Herrei ben ik dank verschuldigd voor
het snel beantwoorden van mijn mailtjes en voor zijn interesse in het
zeepaardjesonderzoek. Uiteraard was dit doctoraat niet mogelijk zonder zeepaardjes en
zeenaaiden, en dus bedank ik de derde partner van het project, de Koninklijke
Maatschappij voor Dierkunde van Antwerpen en in het bijzonder Philippe Jouk, Zjef
Pereboom en Peter Galbusera voor het bezorgen van de nodige specimens. Maar ook
dank aan de Mariene Biologie voor de staalname van Syngnathus rostellatus en S. acus, Jos
Snoeks van het Koninklijk Museum voor M idden Afrika voor het uitlenen van de
Microphis brachyurus aculeatus specimens, Ronald Vonk van het Zoologisch Museum van
Amsterdam voor toegang tot de collectie en David en Marcel voor hun hulp bij het
kweken van Hippocampus reidi.
H et Instituut voor de Aanmoediging van Innovatie door W etenschap en Technologie in
Vlaanderen (IWT) ben ik erkentelijk voor het toekennen van een doctoraatsbeurs.
I also greatly appreciated the cooperation w ith Prof. Dr. Elizabeth Dumont. Her
experience in the field of finite element greatly inspired me and I thoroughly enjoyed
the F EM workshop at Amherst. I wish to thank her for all her assistance, advice and
motivation. I am indebted to Prof. Patricia Hernandez for her expertise in
immunohistochemistry and her unfailing patience w hen introducing me into the basics
of this technique. Although I wasn’t able to incorporate immuno’s in this dissertation, I
am very grateful for everything she taught me.
Voor alle problemen met ophelderingen, kleuringen of verdunningen, voor histologische
protocols, voor snijwerk, om te weten of mijn afval bij organische of anorganische zuren
hoorde, voor het indienen van onkostennota’s, voor hulp bij procite of endnote, om een
bepaald product terug te vinden, voor de verzorging van het zeepaardj esaquarium en
voor nog veel meer praktisch werk kon ik steeds bij Barbara en Joachim terecht. H un
kennis heeft me heel erg geholpen en onnoemelijk veel uren labowerk uitgespaard.
Ik kon ook steeds rekenen op het bewonderenswaardige geduld van Hilde, Miranda,
Jeroen en M artin voor alle administratieve hulp, practicum-gerelateerde zaken, ITproblemen of onvindbare boeken en artikels. Bedankt!
Voor het maken van de CT-scans ben ik dank verschuldigd aan Prof. Dr. Luc Van
Hoorebeke, Manuel, M atthieu, Loes, Prof. Dr. Nora De Clerck en Andrei.
Gerard wil ik bedanken voor zijn hulp bij de vormgeving. Niets dan lof voor absara!
H et werk van mijn Bachelorstudenten Anouck, Lien en Karen is dit onderzoek zeker ten
goede gekomen en daarvoor wil ik hen graag bedanken. Ook de masterthesis van
Marjolein, Karen en Annelies zijn van onschatbare waarde geweest, waarvoor dank.
Veel dank gaat naar mijn (ex-)collega’s die me steeds met plezier naar de Ledeganck
deden komen. Frank en Stijn, een vrolijker duo zal het lab volgens mij niet gauw meer
kennen. Zij slaagden er zelfs in om een middagpauze in de resto gezellig te maken, wat
toch een hele prestatie is. Natalie bedank ik voor haar bemoedigende woorden, toen ik
nog een eiland met haar deelde en later via mail. Tom wil ik graag extra bedanken voor
zijn niet af latende behulpzaamheid bij het vinden van het juiste engels woord, bij het
regelen van allerlei IWT- of andere administratieve dingen en bij het nalezen van
teksten, maar ook bij het opeten van een tros pitloze druifjes of een fruittaart. Paul,
bedank ik voor zijn hulp bij de morfometrische analyses, voor zijn nuchtere levenskijk en
voor onze weddenschap die me tot lezen aanzet. M et Celine heb ik een lange,
gemeenschappelijke weg afgelegd, samen thesis schrijven, samen IW T voorbereiden,
samen naar congressen en workshops en samen practica begeleiden. Ik ben haar
dankbaar voor haar zin voor relativering en om altijd klaar te staan met een helpende
hand. Voor het oplossen van allerhande computerproblemen of statistische vraagstukken
kon ik rekenen op Tim. Hij en Niels hebben samen, door hun optimisme en honger naar
kennis mijn motivatie echt een stevige duw in de rug gegeven, meer dan ze beseffen. En
Niels, ik kijk uit naar alle toekomstige ‘weetjes van de dag’! Los van al de
werkgerelateerde hulp wil ik Barbara en Joachim, die me geholpen hebben om me van bij
het begin thuis te voelen in het lab, graag bedanken voor hun opgeruimdheid en voor de
fijne babbeltjes tussen het werken door. Catherine wil ik bedanken omdat ze me heeft
aangezet langer op het lab te blijven, voor de ontspanning daarbuiten en voor de
prachtige muziek (La Meglio Gioventù! Elis Regina! Roby Lakatos! Frida! The Tolga
Quartet!...) die het werk zeker heeft vergemakkelijkt. M et Annelies heb ik met veel
plezier een eiland, liefde voor Italië en ontroerende neefjes- en nichtjesverhalen gedeeld.
Ik wil haar bedanken voor haar hulp bij het maken van reconstructies, uiteindelijk lukte
het ons samen om de meest raadselachtige nevenverschijnselen van Amira op te lossen.
Emilie ben ik veel dank verschuldigd voor de gezellige gesprekjes, de geweldig
motiverende aansporingen in de laatste weken en voor onze gedeelde (zij het passieve)
interesse voor fotografie. Eindelijk zal ze rustig kunnen werken zonder dat ik steeds over
haar schouder meekijk. Soheil wist steeds interessante artikels, boeken en software op te
sporen en heeft een niet te onderschatte bijdrage geleverd aan mijn kennis van Iran. Het
beste lamsvlees van mijn leven at ik bij hem thuis, waarvoor dank. Aiso Fatemeh has
made me enthusiastic about the Iranian culture by bringing delicious cakes and cookies
to the lab. I would like to thank her for joining me through the last stressful weeks
before the deadline. I thank Spyros for sharing his great taste in music w ith me and for
never returning to the ARC after a meeting w ith Dominique, w ithout dropping in on
our office. I also own a lot of thanks to Mimi, I’ve always admired her cheerfulness, sense
of perspective and perseverance and I loved attending the M ontreal meeting together. I
want to thank Aslr for being such a good friend and for the great time we had in Torino.
Uiteraard wil ik ook graag alle andere mensen van op ’t derde bedanken, voor de
practicabegeleiding, gezellige middagpauzes, uitstapjes naar de Serli, filmavondjes en
internationale culinaire ontdekkingen. Nu ik het opschrijf besef ik pas hoe hard ik dit
allemaal zal missen!!
Bovendien is hier een hartelijke ‘dankjewel’ voor al mijn mede-bioloogjes op zijn plaats.
Stijn, Patricia, Sasha en Greet wil ik bedanken voor de leuke intermezzo’s en de heerlijke
reizen. Ik hoop dat daar nog vaak een vervolg op komt! Ook Marjolein verdient veel dank
voor het aanbrengen van het zeepaardj esonderwerp in het lab en voor alles wat we samen
gedaan hebben. Vera, Loe & Sam, Stien & Wim, Mimi en de Italianen bedank ik voor de
nodige ontspanning, voor het luisteren naar mijn geklaag, voor de vele etentjes, om af en
toe te vragen hoe het nog met de zeepaardjes is en vooral om er op bepaalde momenten
niet naar te vragen. Hanne en Barbara, de treinritjes Antwerpen-Gent zal ik steeds
associëren met gezelligheid (en ‘keneel’). Een betere woon-werkverplaatsing kan ik me
niet inbeelden, super hard bedankt! Goedele, Aslr, Catherine, Stijn en Freija (en hun
zetels en bedden) bedank ik om me van het Gentse leven na de laatste trein te laten
genieten. M ijn huisgenoten, Nelles, Jasper, Annelies en Anton, verdienen ook een
bedankje omdat ze zo’n aangenaam gezelschap waren om bij thuis te komen na een
werkdag.
Jef, Ona, Bas, Toon, Boris, Renée, Oskar, Dunia, Leyla, Ziya en Sanaa, bedank ik niet
omdat ze iets aan dit doctoraat hebben bijgedragen, maar gewoon omdat ze er zijn. Mijn
broers en (toekomstige) schoonzussen wil ik ongelooflijk hard bedanken om me te
verwennen met heerlijke maaltijden, leuke uitstapjes en reisjes en om me terug kleine
zus te doen voelen. En tenslotte heel erg veel dank aan mijn ouders om me zelf mijn weg
te laten kiezen en om me volledig te steunen bij elke beslissing die ik nam.
Heleen
Er gaat meer boven je pet dan eronder
Toon Hermans
Table
of c o n te n t s
1 I n tr o d u c tio n _______________________________________
1.1. A daptive
evolutio n a n d specialization ____________
1.2. S y n g n a t h i d a e : A
_l
_i
case o f extreme specialization __
_4
1.2.1. W h a t are syn gnathid s?_________________________
_4
1.2.2. T rop hic s p ecializa tio n and c o n str a in ts_________
a n d thesis outline ____________________________
_9
11
1.3.1. C o n t e x t and general aim_______________________
11
1.3.2. T h e sis o u tlin e and s p ecific aims________________
12
1.3. A ims
2 M aterial & m e th o d s _______________________________
17
2.1. M aterial ____________________________________________
.17
2.2. M et h o d s ____________________________________________
20
2.2.1. K eep in g live s p e c im e n s _________________________
20
2.2.2. Sacrificing and fix a tio n _________________________
21
2.2.3. B io m etry_______________________________________
21
2.2.4.
¡n f o f o clearing and staining____________________
22
2.2.5. D is s e c tin g ______________________________________
23
2.2.6. Serial s e c tio n in g ________________________________
23
2.2.7. C o m p u te d t o m o g r a p h y scan n in g______________
23
2.2.8. Graphical 3 D -r e c o n s tr u c tin g ___________________
.24
2.2.9. G e o m e tr ic m o r p h o m e tr ic analysis_____________
25
2.2.10. Finite e l e m e n t analysis________________________
27
2.3. T E RM IN O LOGY _______________________________________
.29
3 M o r p h o lo g ic a l d e s c r ip tio n ________________________
.33
3.1. M o r p h o l o g y
o f the c h o n d r o - a n d o s t e o c r a n iu m .
.33
A b stra ct______________________________________________
.33
3.1.1. In tro d u ctio n ____________________________________
.34
3.1.2. Brief material & m e th o d s _______________________
.35
3.1.3. Results__________________________________________
36
3.1.4. D is c u s s io n ______________________________________
.44
3.2. Implic atio ns
o f s n o u t el o n g a t io n ________________
.53
A b stra ct______________________________________________
.53
3.2.1.
.54
In trod u ction ___________________________________
3.2.2. Brief material & m e th o d s ______________________________________________ 57
3.2.3. Results_________________________________________________________________ 58
3.2.4. D is c u s s io n _____________________________________________________________ 72
4 M o r p h o l o g i c a l v a r i a t i o n ______________________________________________________ 8 3
A b stra ct_____________________________________________________________________ 83
4.1. In tro d u ction _____________________________________________________________ 84
4.2. Brief material & m e th o d s ________________________________________________88
4.3. Results___________________________________________________________________89
4.4. D is c u s s io n _______________________________________________________________ 93
5 F u n c t i o n a l i n t e r p r e t a t i o n ___________________________________________________ 1 0 1
5.1. Kinematics
o f the feeding apparatus ____________________________________
101
A b stract____________________________________________________________________ 101
5.1.1. In trod u ctio n __________________________________________________________ 102
5.1.2. Brief material &
m e t h o d s ___________________________________________ 105
5.1.3. Results________________________________________________________________ 108
5.1.4. D is c u s s io n ____________________________________________________________ 115
5.2. M echanical
stress distribution in the feeding apparatus ______________
121
A b stract____________________________________________________________________ 121
5.2.1. In trod u ctio n __________________________________________________________ 122
5.2.2. Brief material &
m e t h o d s ___________________________________________ 125
5.2.3. Results________________________________________________________________ 127
5.2.4. D is c u s s io n ____________________________________________________________ 129
6 G e n e r a l d i s c u s s i o n ____________________________________________________________1 3 5
6.1. O n t o g en et ic
6.2. Fu n c t io n a l
m o r p h o l o g y _______________________________________________ 135
m o r p h o l o g y _________________________________________________142
6.3. Ev o l u t io n ar y
6.4. G eneral
m o r p h o l o g y ______________________________________________
152
c o n c l u s i o n _____________________________________________________
156
7 S u m m a r y & s a m e n v a t t i n g ____________________________________________________1 5 9
7.1. S u m m ar y __________________________________________________________________ 159
7.2. S a m e n v a t t in g _____________________________________________________________165
8 R e f e r e n c e s _____________________________________________________________________ 1 7 1
P u b lic a tio n lis t
193
Introduction
This d issertatio n deals w ith th e m orphology of th e specialized feeding ap p aratu s
of pipefishes an d seahorses w ith in an ev o lu tio n ary context. T heir rem arkable
head m orphology form s th e base of an in terestin g paradox b etw een exceptionally
sh o rt prey cap tu re tim es on th e one h a n d and severe hydrodynam ic c o n strain ts
on th e other. T h at paradox is elab o rated o n in th e p re sen t ch ap ter, b u t first an
in tro d u c tio n to th e concepts of ‘ev o lu tio n ’ and ‘sp ecializatio n ’ is provided, before
th e y are applied to pipefishes and seahorses.
1.1.
A d a p t iv e e v o l u t i o n a n d s p e c i a l i z a t i o n
Long before th e te rm ‘evolution ary biology’ existed, p h ilosophers an d scientists
trie d to explain th e vast p h en o ty p ic v ariatio n p re sen t in n atu re. M an y great
m inds agonized over th e process th a t accounts fo r th e enorm ous com plexity and
diversity of life. As early as in th e six th c en tu ry BC, an cie n t G reeks like
A naxim ander
an d
Em pedocles
developed
th e
first
th eo ries
involving
tra n sfo rm a tio n of organism s over tim e, w h a t w e n o w call evolution. E v olution is
defined as th e change in characteristics of living organism s b etw een generations
(Ridley, 1993). It explains life’s diversity, w h ich can be regarded as th e response to
2
C h a p te r i - Introduction
e n v iro n m en tal diversity. A t th e sam e tim e, evo lu tio n acco u n ts fo r th e u n ity of
life since all individuals, ex tin c t an d ex tan t, are related th ro u g h th e process of
descent w ith m odification from com m on ancestors.
T he th e o ry of descent w ith m odification, as originally fo rm u lated by C harles
D arw in (1859) an d A lfred Russel W allace (1889), is essentially based o n th re e
concepts. Firstly, b irth rates of several organism s are m u ch h ig h er th a n th e ir
e n v iro n m en t can su p p o rt (K utschera & N iklas, 2004). To illu stra te th is, D arw in
(1859) calculated th a t, startin g w ith one p air of elep h an ts and using a m inim um
rate of n a tu ra l increase, in less th a n five cen tu ries th e re w o u ld be as m any as
fiftee n m illion e lep h an ts alive. H e concluded th a t th e re m u st be a n a tu ra l dow nreg u latin g m echanism for th is ov erp ro d u ctio n an d argued th a t it could be fo u n d
in com petition. M em bers of all species1 are th o u g h t to com pete w ith each o th e r
for access to resources of w h ich th e availability is lim ited. D arw in (1859) referred
to th is p h e n o m en o n as th e struggle fo r existence. Secondly, living organism s
display individual v ariatio n in p h en o ty p ic tra its, w h ich have a genotypic basis as
w as discovered in 1866 by G eorge M endel, to be rediscovered only in 1900 by
H ugo de Vries. D arw in an d W allace b o th n o te d th a t c ertain ch aracter differences
have an effect on an individual's ab ility to survive an d reproduce. A n organism in
possession of such a tra it has b e tte r chances in th e struggle fo r life. A nd thirdly,
th e perceived v ariatio n appeared to be h eritab le and could th u s be passed from
g e n eratio n to g en eratio n leading to sim ilarly ch aracterized offspring.
These th re e argum ents com bined allow ed D arw in an d W allace to fo rm u late th e
hypothesis of n a tu ra l selection as th e driving force fo r evolution. In sh o rt, it
im plies th e increased p ro b ab ility of survival an d re p ro d u c tio n of th e m em bers in
a p o p u la tio n b e st fitte d to deal w ith th e en v iro n m en tal factors, th is is, survival of
th e fittest. T he fitness of an individual in th is co n tex t is defin ed as th e relative
c o n trib u tio n of its genotype to th e n ex t g en eratio n , relative to th e c o n trib u tio n s
of o th e r genotypes (Lawrence, 2005). A ch aracter th a t increases th e fitness of an
individual w ith respect to o th e r individuals in th e p o p u la tio n is an adaptive tra it
(Bock, 2003). H ence, individuals th a t are b e tte r ad ap ted to th e ir e n v iro n m en t
have a b e tte r chance to reproduce an d in tu rn , som e of th e ir offsp rin g w ill likely
possess th e adaptive characters.
1 a s p e c i e s is d e f i n e d a s a g r o u p o f i n t e r b r e e d in g n a t u r a l p o p u l a t i o n s t h a t i s r e p r o d u c t iv e ly
i s o l a t e d f r o m o t h e r s u c h g r o u p s (M a y r , 1 9 9 6)
Chapter i - Introduction
3
O ccasionally an ad ap ta tio n can be so extrem e th a t it su b stan tially changes (part
of) th e m orphology of an anim al, providing it w ith u n iq u e fu n c tio n a l features.
A n exam ple of th is is th e p eculiar bill w ith overlapping u p p er an d low er bill tips
in crossbills (finches of th e genus Loxia), w h ich appears to be an adaptive tra it to
ex tract seeds from conifer cones w ith im proved efficiency (B enkm an, 1987; 1993;
Edelaar et al., 2005). A n o th e r illu stra tio n of an ex trao rd in ary p h en o ty p e is fo u n d
in th e great a n te a te r (Myrmecophaga tridactyla L.), w h ich belongs to th e su b o rd er
V erm ilingua. The a n te rio r p a rt of its cran iu m is elo n g ated and th e anim al has a
long, sticky to n g u e w ith sm all spikes, w h ich can be p ro tru d e d to a great ex ten d
an d serves to catch ants an d te rm ites (N aples, 1999). Scale-eating cichlids (part of
th e trib e Perissodini) also exhib it a rem arkable m orphology. T h eir m o u th is n o t
longer d irected rostrally, b u t is b e n t laterally (eith er le ft or right) due to
asym m etry of th e jaws, w h ich m akes it a very e fficien t to o l fo r rem oving scales of
th e flan k of o th e r fish (Liem & S tew art, 1976; S tew art & A lbertson, 2010).
Such tra its are considered adaptive specializations, provided th a t th e y im prove
th e perform ance of behaviours related to th a t m orphology an d hence g ran t
individuals a fitness advantage in a given ecological settin g (Ferry-G raham et al.,
2002). In a tro p h ic context, adaptive specialists are m ore efficien t a t o b tain in g
th e ir p re fe rre d prey item s th a n generalists. T herefore, crossbills, an teaters and
scale-eating cichlids can be regarded as adaptive specialists. Usually, specialists
occupy only a n arro w ecological niche; th ey are lim ited to a re stric ted range of
th e available resources due to ecological, m echanical or ev o lu tio n ary co n strain ts
(D obzhansky, 1973; Ferry-G raham et al., 2002). T rophic specialists, exploiting a
lim ited dietary b read th , w o uld be p red icted to lack flexibility or v ersatility in
th e ir feeding strategies com pared to g eneralists (R alston & W ain w rig h t, 1997).
T he feeding v ersatility of an individual is in te rp re te d as its ab ility to m odulate
th e kinem atics of feeding stru c tu re s in response to d ifferen t food types
(Sanderson, 1991). In consequence of th e specialist’s stereotypy, species th a t
occupy a sm all niche could be m ore p ro n e to e x tin c tio n since th e y are less
re silien t to changes in en v iro n m en tal conditions. H ence, species appear to face a
trad e-o ff betw een selection for a b ro ad er ecological niche in ord er to cope w ith
unfavourable abiotic factors an d lim ited food availability o n th e one hand, and
pressure tow ards a n arro w niche w ith increased in ta k e p roficiency of a p a rticu lar
food type on th e o th e r hand.
C h a p te r i - Introduction
4
1.2.
SYNGNATHIDAE: A CASE OF EXTREME SPECIALIZATION
1.2.1. W h a t are s yng nat hid s?
M em bers of th e fam ily Syngnathidae (i.e. seahorses, pipefishes, pipehorses and
seadragons) probably are th e w o rld ’s m ost unfish-like fish an d in th e p ast th ey
have caused m any taxonom ists a serious headache. Even L innaeus (1766)
categorized seahorses an d pipefishes in a division of th e A m phibia (to g eth er w ith
e.g. lam preys, sharks, rays, sturgeons and razorfish).
In fact, seahorses, pipefishes and th e ir allies belong to th e T eleostei, a group
c o u n tin g over 26,000 e x ta n t species (Fig. 1.1). This gro u p includes 96% of all
living fishes an d is th e m ost diversified gro u p of vertebrates. Teleosteans occur in
every freshw ater, brackish an d m arine h a b ita t (rivers, lakes an d oceans) and some
can even survive on land. T hey ex h ib it a large v ariety of lo co m o tio n behaviours
(sw im m ing, gliding, w alking or staying im m obile) an d th e te le o st d iet is
extrem ely diverse (algae, Zooplankton, w orm s, snails, in sect larvae, clams,
shrim ps, am phibians, o th e r fish, fry, scales, blood, faeces,...). In sh o rt, th e y have
rad iated in to countless niches (H elfm an et al., 2009). T eleo stean sy n ap o m o rp h ies2
include: th e presence of a ven tro cau d al process a t th e q u ad rate bone, th e
m axillary bones are m obile an d enable u p p er jaw p ro tru sio n , co ro n o id bones are
absent, th e re is a fusion b etw een articu lar, angular an d re tro a rtic u la r bones in th e
low er jaw, th e p a rie tal bones are caudally d istin ctly bro ad er th a n rostrally, th e
urohyal bone is form ed as an ossification w ith in th e stern o h y o id eu s te n d o n and
th e re is a single, m edian vom eral bone th a t form s th e roof of th e m o u th (N elson,
2006; W iley & Johnson, 2010). T he T eleostei can be divided in to fo u r cohorts,
nam ely
th e
O steoglossom orpha,
E lopom orpha,
O stario clu p eo m o rp h a
(=O tocephala) and E uteleostei (Fig. 1.1A; N elson, 2006).
T he E uteleostei, or “tru e ” teleosts, are by far th e largest gro u p w ith in th e
Teleostei. Shared traits, i.e. presence of an adipose fin, n u p tia l b reed in g tubercles
an d a ro stral m em branous c o m p o n en t to th e first u ro n eu ral, are fo u n d in th e
2 a s y n a p o m o r p h y is a h o m o lo g o u s c h a r a cte r c o m m o n t o t w o or m o r e ta x a a n d is t h o u g h t t o h a v e
i t s o r ig i n i n a s i n g l e s t r u c t u r e i n t h e i r c o m m o n a n c e s t o r ( L a w r e n c e , 2 0 0 5 )
Chapter i - Introduction
5
basal euteleosteans an d ap p aren tly have b een lo st in th e m ore derived ones
(Lauder & Liem, 1983). T he E u teleo stei consist of 28 orders an d 346 fam ilies and
com prise, am ong others, th e A can th o m o rp h a (Fig. 1.1B). This group, also called
th e spiny-rayed fishes, is characterized by th e presence of tru e fin spines in th e
dorsal, anal an d pelvic fins in stead of h ard en ed segm ented rays (H elfm an et al.,
2009). E ncom passing approxim ately 16,000 species in over 300 fam ilies (including
cods, tu n as, stonefishes, flatfishes, p u ffers and m u ch more), th ey com prise over
60% of th e e x ta n t teleosts and a b o u t o n e -th ird of all living vertebrates.
A can th o m o rp h m onophyly is su p p o rte d by b o th m orphological an d m olecular
analyses (N elson, 2006). T he phylogeny of th e P ercom orpha, a clade w ith in th e
A canthom orpha, how ever, is p ro b lem atic and c u rre n tly u n d er discussion (e.g.
Lauder & Liem, 1983; Jam ieson, 1991; P a re n ti & Song, 1996; S pringer & O rrell,
2004; S m ith & C raig, 2007; Sm ith, 2010). In th is d issertatio n , th e p ercom orphs are
th o u g h t to encom pass n ine orders as show n in Fig. 1.1C (N elson, 2006). This
group co n tain s over 13,000 m ainly m arine species th a t are ch aracterized by th e
pelvic fin being atta ch e d to th e cleith ru m of th e p ecto ral girdle an d a ro stral
pelvic process th a t is ventrally displaced (N elson, 2006; H elfm an et al., 2009).
T raditionally, pipefishes, seahorses and th e ir allies are encom passed by th e order
G asterosteiform es, b u t th e system atic relatio n sh ip s am ong gastero steifo rm fishes
have been th e subject of m uch discussion (see W ilso n & O rr (subm itted) fo r a
review). A ccording to th e classification of Keivany & N elson (2006) based o n 110
inform ative osteological characters, th e order is divided in to th re e suborders:
firstly th e H ypoptychoidei, com prising only th e fam ily of th e sand eels
(H ypoptychidae), secondly th e
(G asterosteidae)
and
th e
G asterosteoidei, inclu d in g
tu b e sn o u ts
(A ulorhynchidae),
th e
an d
sticklebacks
th ird ly
th e
Syngnathoidei to w h ich fam ilies such as th e seam oths (Pegasidae), arm o u red
sticklebacks (Indostom idae), tru m p etfish e s (A ulostom idae) an d pipefishes and
seahorses (Syngnathidae) belong (Fig. 1.2; Keivany & N elson, 2006). R ecently,
how ever, th e m onophyly of G asterosteiform es has been q u estio n ed based on
m olecular phylogenetic research (C hen et al., 2003; S m ith & W h eeler, 2004;
D e tta i & L ecointre, 2005; K aw ahara et al., 2008). A lth o u g h m ost of th ese
m olecular studies only include a few gastero steifo rm representatives, m aking it
d ifficu lt to evaluate th e relatio n sh ip s of th e ord er as a w hole, K aw ahara et al.
(2008) analyzed m em bers of all eleven families. T hey fo u n d su p p o rt to place th e
6
C h a p te r i - Introduction
Indostom idae, Syngnathoidei (m inus Indostom idae) an d G asterosteoidei to g e th e r
w ith th e H ypoptychoidei in th re e separate clades w ith in th e P ercom orpha and
suggest a basal divergence (Fig. 1.3). A ccording to th is p o in t of view, th e
D actylop teroidei (flying gurnards) are classified as th e sister gro u p of th e
Syngnathoidei. The classification w ith G asterosteoidei as th e closest sister group
of th e Syngnathoidei, as suggested by K eivany & N elson (2006), is follow ed here,
since m orphological analyses provide n o altern ativ e an d th e results of th e
m olecular analyses are n o t co n sisten t w ith each other.
N evertheless, w h e th e r additio n al research confirm s th e tre e pro p o sed by
K aw ahara et al. (2008) or n o t, Gasterosteus aculeatus L. (three-spined stickleback)
rem ains valid as a generalized p erco m o rp h out-group. G astero steifo rm fishes are
ch aracterized by an arm o u r of plates th a t covers th e body, th e m o u th is usually
small, th e pelvic girdle is never directly attach ed to th e cleithra, th e y have only a
lim ited n u m b e r of branchiosteg al rays and th e ir gills are e ith e r reduced or
m odified (Lauder & Liem, 1983; Jam ieson, 1991; N elson, 2006). Besides th a t, m any
species have an uncom m on locom otion, as fo r in stan ce th e shrim pfishes and
razorfishes (family C entriscidae) th a t sw im head-dow n w ith th e dorsal edge of
th e ir razor-like body leading (Fig. 1.2; H elfm an et al., 2009) an d th e cleaner
p ipefish (Doryrhamphus janssi (H erald & Randall)) w h ich is k n o w n to sw im
upside-dow n. W ith in th is diverse order, differences in reproductive behaviour
are nu m ero u s (e.g. W ilso n et al., 2001; K aw ahara et al., 2008). Some species, like
sticklebacks, produce a glue-like substance th a t is used fo r bu ild in g a n e st o u t of
p la n t m aterial w hile m ale seahorses possess a p o u ch (this w ill be elab o rated on
later) (N elson, 2006; K aw ahara et al., 2008). G astero steifo rm body size is also
reasonably variable, ranging fro m less th a n 3 cm stan d a rd le n g th (SL) in
Indostomus species (Indostom idae) to over a m eter SL in th e genus Fistularia
(F istulariidae) (N elson, 2006; K aw ahara et al., 2008). In to tal, th e ord er of
G asterosteiform es covers eleven fam ilies, 71 genera an d 278 species, th e m ajority
of w h ich can be fo u n d in m arine w aters (N elson, 2006).
T he fam ily of Syngnathidae is subdivided in to tw o subfam ilies: Syngnathinae
(pipefishes,
seadragons
and
pipehorses)
an d
H ip pocam pinae
(seahorses).
Seadragons are relatively large fishes w ith a high, laterally fla tte n e d body covered
by leafy appendages (Fig. 1.4D). E ach of th e tw o seadragon genera includes only
one species: Phyllopteryx taeniolatus (Lacepède) (weedy seadragon) and Phycodurus
Chapter i - Introduction
7
eques (G ünther) (leafy seadragon). Pipehorses look like an in term ed iate form
b etw een pipefishes an d seahorses, having a straig h t, h o riz o n ta l body w ith a
slightly tilte d head an d a prehensile tail (Fig. 1.4F; K uiter, 2003; 2004). The genera
Solegnathus and Syngnathoides an d also th e genera Acentronura, Amphelikturus and
Idiotropiscis belong to th e pipehorses, th e la tte r th re e are som etim es referred to as
pygm y pipehorses. All o th e r m em bers of th e Syngnathinae subfam ily are
pipefishes. The Syngnathidae is by far th e m ost speciose fam ily of th e
G asterosteiform es, w ith an estim ated species n u m b e r varying from 232 (N elson,
2006) to over 320 (K uiter, 2003). N u m ero u s m isid en tificatio n s, synonym s and
even spelling m istakes have resu lted in a taxonom ic chaos. Especially am ong
seahorses, species id en tific a tio n is problem atic, since m orphological differences
b etw een species are subtle and th e y are able to change colour an d develop skin
filam en ts on th e ir body in order to cam ouflage them selves (Lourie et al., 1999a;
2004; M u ru g a n et al., 2008; Sanders et al., 2008). The subfam ily H ip pocam pinae
has recently been th e subject of revision by Lourie an d co-w orkers of th e P roject
Seahorse (1999b, 2004). It consists of only one genus (Hippocampus) and P roject
Seahorse
c u rre n tly
recognizes
38
species
(h ttp://seahorse.fisheries.ubc.ca,
c o n su lted on 16/02/2011) versus 54 valid species nam es according to Fishbase
(Froese & Pauly, 2011). W ith in th e syngnathid fam ily a large m orphological
diversity exists: size, body shape, colour p a tte rn , orn am en ts, body p o stu re, fin
arran g em en t and sn o u t ph en o ty p e are all highly variable (Fig. 1.4). N evertheless,
th e re are som e shared features, w h ich w ill be discussed below.
P robably th e m ost d istin c t syngn ath id ch aracteristic is th e in cu b atin g area on th e
v en tral side of th e m ale body. This m orphological s tru c tu re can vary fro m an
u n p ro te c te d p a tch of skin o n to w h ich th e bro o d is loosely attach ed , over one
sim ple or tw o overlapping m em branes o n th e tru n k or tail, to a fully enclosed
pocket-like p o u ch w ith a sm all ap ertu re (H erald, 1959; K ornienko, 2001; W ilso n
et al., 2001; C arcu p in o et al., 2002; K varnem o & Sim m ons, 2004). T he evo lu tio n ary
tre n d tow ards increasing p a re n tal care corresponds fairly w ell to th e prevailing
m olecular-based phylogeny of th e fam ily (Fig. 1.5; W ilso n et al., 2001; 2003;
H elfm an et al., 2009; W ilso n & O rr, subm itted). The in c u b atin g area is associated
w ith th e syngnathid u nique reproductive strategy: th e fem ale deposits th e eggs in
th e p o u ch of th e m ale w here th e y are incubated. T he m ale provides all p o st­
fe rtiliza tio n
p a re n ta l care
(e.g. p ro tectio n , n u tritio n , osm o reg u latio n
and
8
C h a p te r i - Introduction
v en tilatio n ) and fully developed, in d e p e n d e n t young are released fro m th e p o u ch
a fte r a few w eeks (K ornienko, 2001; W ilso n et al., 2001; V in cen t & Giles, 2003;
F oster & V incent, 2004; Lourie et al., 2004).
In stead of scales, th e syngnathid body is covered w ith a com plex of precisely
arranged bony plates to p ro te c t th e m against p red atio n . This arm our, how ever
rigid a ro u n d th e tru n k , does n o t im pede a rtic u la tio n of th e separate segm ents in
th e caudal region. Especially in th e tail of seahorses an d pipehorses, th a t lacks a
caudal fin, flexibility is high. T he do rso v en tral b en d in g ab ility of th e ta il (it can
curl even over 360e) is a ra th e r u n u su al m o d ificatio n th a t m akes it very efficien t
for grasping and holding on to objects (H ale, 1996; B ru n er & B artolino, 2008).
Seahorses n o t only use th e ir prehensile ta il to an ch o r them selves to holdfasts, b u t
it also has a fu n c tio n in c o u rtsh ip (en tw in in g of tails) an d m ale c o m p etitio n (tail
w restling) (Lourie et al., 1999b).
Syngnathids exhibit great cam ouflage capacities. They can have a speckled colour
p a tte rn , w eed-like appendages, skin filam ents or w arts th a t m ake th e m look like
algae, seagrass or sticks (K uiter, 2003). C o m b in ed w ith a slow, m anoeuvring
m otion, enabled by u n d u la tio n of th e ir dorsal an d p ecto ral fins, th ey fade alm ost
com pletely w ith th e surro u n d in g s (C onsi et al., 2001). This favours th e am bush
foraging strategy of syngnathids, th e y sit and w a it u n til a prey, usually a sm all
crustacean, com es close, w h ich is th e n cap tu red by a quick flick of th e head
(F oster & V incent, 2004; K endrick & H yndes, 2005). The d ie t of syngnathids,
being
gape-lim ited
su ctio n
feeders,
com prises
sm all
crustaceans
(m ostly
am phipods an d copepods), o th e r sm all in v erteb rates an d fish fry (Payne et al.,
1998; Lourie et al., 1999b; T eixeira & M usick, 2001; W oods, 2003; F o ster &
V incent, 2004).
A lth o u g h th e y individually have a very sm all hom e range, syngnathids are
ub iq u ito u s (F oster & V incent, 2004). M o st species occur in shallow te m p e rate and
tro p ical w aters, especially in th e Indo-W e st Pacific at th e coast of A ustralia,
S outh-E ast Asia and Japan (Lourie et al., 1999b; K uiter, 2003). Seahorses are
alm ost exclusively m arine (Hippocampus capensis B oulenger lives in estuaries),
w hile pipefishes can be fo u n d in m arine, brackish an d even fresh w ater (Foster &
V incent, 2004). T heir p re fe rre d h a b ita t can consist of seagrasses, algae, corals,
m angroves, etc., w h ich all offer p ro te c tio n from p red ato rs an d a h ig h prey density
(Payne et al., 1998; Lourie et al., 1999b, K endrick & H yndes, 2003).
Chapter i - Introduction
9
D espite th e extensive in te re st in th e rep ro d u ctiv e strategies (e.g. K varnem o et al.,
2000; C arcu p in o et al., 2002; V in cen t & Giles, 2003; P o o rten a a r et al., 2004), life
h isto ry (e.g. K anou & K ohno, 2001; F o ster & V incent, 2004), p o p u la tio n dynam ics
(e.g. Teske et al., 2003; Lourie et al., 2005), aq u acu ltu re (e.g. W oods, 2003;
K oldew ey & M artin -S m ith , 2010) an d con serv atio n (e.g. G offredo et al., 2004;
F oster & V incent, 2004; Scales, 2010) of syngnathids, little is k n o w n a b o u t m ost
aspects of th e ir phenotype. M an y q uestions regarding th e ir feeding behaviour,
head m orphology an d possible adaptive tra its to th e specialized su ctio n feeding
rem ain unansw ered. The few detailed anatom ical studies th a t have b een carried
o u t date from th e first h alf of th e previous c en tu ry or even b efo re (e.g.
M cM u rrich , 1883; Jungersen, 1910). Some of th o se studies do n o t deal w ith
syngnathids exclusively, b u t superficially describe th e cran iu m of a w ide array of
taxa (e.g. G regory, 1933; de Beer, 1937) w hile o th ers focus o n th e developm ent of
th e skull (e.g. M cM u rrich , 1883; Kadam, 1958; 1961). P u b licatio n s o n m yology or
fu n c tio n a l in te rp re ta tio n of th e feeding ap p aratu s are scarce. H ow ever, B ranch
(1966) supplies a reasonable d escrip tio n of th e cranial osteology an d m yology of
Syngnathus acus L. an d he even m akes suggestions a b o u t th e feeding m eth o d
based o n his observations. Basic k in em atical d ata o n su ctio n feeding events of
Hippocampus erectus Perry an d Syngnathus floridae (Jordan & G ilbert) is provided
by B ergert and W a in w rig h t (1997) an d of H. erectus an d H. zosterae Jordan &
G ilb ert by C olson et al. (1998), b o th in co m b in atio n w ith a cursory m orphological
descrip tio n of th e head of th e recorded species. Finally, M u ller an d Osse (1984)
investigated th e hydrodynam ics of su ctio n feeding in several fishes, inclu d in g th e
snake p ipefish (Entelurus aequoreus (L.)). All th ese studies fo rm th e fo u n d a tio n for
th e p re sen t d octoral research o n evo lu tio n ary m orphology of th e sy n g n ath id
feeding apparatus.
1.2.2. Trop hic sp eci al iza ti on and c on s tr ai nt s
T he trad e-o ff described above (C h ap ter 1.1), b etw een selection fo r specialization
on th e one h a n d an d tro p h ic niche en larg em en t o n th e o th er, n atu rally also holds
for m em bers of th e fam ily Syngnathidae.
10
C h a p te r i - Introduction
All syngnathids have an elongated, tu b u la r sn o u t w ith tin y te rm in a l jaws th a t
enables th e m to p e rfo rm su ctio n feeding. The le n g th of th e sn o u t is highly
variable,
ranging
from
very
sh o rt
as
in
th e
m ushroom -coral
p ip efish
(Siokunichthys nigrolineatus D aw son) (Fig. 1.4E) to extrem ely elongate in for
exam ple th e w eedy seadragon (Phyllopteryx taeniolatus) (Fig. 1.4D). H aving a
sn o u t w ith a very sm all diam eter, seahorses and pipefishes are able to feed only
on prey p articles of lim ited size. T h eir m orphology th u s confines th e m to a diet
of only sm all prey item s from a b ro ad range of available prey (T ipton & Bell, 1988;
T eixeira & M usick, 1995; W oods, 2003; K endrick & H yndes, 2005). H ence
syngnathids can be regarded as specialists (Ferry-G raham et al., 2002). T heir
extrem e p h en o ty p e (i.e. th e slender snout) w ill only be selected fo r if it is
associated w ith a fitness advantage (Ridley, 2003). Previous studies have show n
th a t syngnathids ex hibit h igh velocity an d g reat p recisio n w h e n p erfo rm in g
su ctio n feeding (M uller & Osse, 1984; B ergert & W ain w rig h t, 1997; de Lussanet &
M uller, 2007). A general a ctin o p tery g ian su ctio n feeding event involves a rapid
expansion of th e buccal cavity, w h ich resu lts in a pressure d ifference b etw een th e
inside of th e cavity and th e su rro u n d in g w ater. So w h e n th e jaws open, a w ater
flow is g en erated th a t tra n sp o rts th e prey in to th e m o u th (e.g. Lauder, 1985). In
seahorses and pipefishes th is buccal expansion is preceded by a ro ta tio n of th e
head th a t p o sitions th e m o u th close to th e prey, a feeding m echanism k n o w n as
pivot feeding (de L ussanet & M uller, 2007; V an W assen b erg h et al., 2008). Prey
in ta k e tim e can be as little as 6 ms, ran k in g syngnathids am ong th e fastest feeding
teleosts ever recorded (C h ap ter 5.1; M u ller & Osse, 1984; B ergert & W ain w rig h t,
3.997; de L ussanet & M uller, 2007). The elongated sn o u t w ith sm all diam eter
increases th e w a ter flux rate in th e m o u th and th u s enhances su ctio n velocity.
H ow ever, as m e n tio n e d before, th e m in u te m o u th a p ertu re of syngnathids w ill
directly c o n strain th e size of prey th a t can be ta k e n in. M oreover, th e design of
th e head w ill likely in flu en ce th e hydrodynam ics of su ctio n feeding. E longation
of th e sn o u t w ill increase th e m o m en t of in e rtia of th e h ead d u rin g n eu ro cran ial
elevation an d drag forces w ill becom e m ore im p o rta n t w ith dim inishing sn o u t
diam eter. H ence, lo n g-snouted species w ill need to g en erate m ore energy to b o th
reach a n d suck in th e prey.
T he h igh level of cranial specialization is likely associated w ith a reduced
fu n c tio n a l v ersatility an d could th u s resu lt in low er resilience w h en facing
Chapter i - Introduction
il
e n v iro n m en tal changes (R alston & W ain w rig h t, 1997; A driaens & H errei, 2009).
In co m b in atio n w ith an th ro p o g en ic th re a ts (h ab itat loss an d overexploitation)
th is m ight explain th e vulnerable ecological statu s of Syngnathidae. O f th e 71
sy ngnathid species registered in th e IU C N Red List of T h rea te n ed Species, 10 are
critically endangered, endangered or v ulnerable, 20 are n ear-th re aten e d or of least
co n cern an d for all 41 o th ers th e re is in su ffic ie n t data to assess th e risk of
e x tin c tio n (IU C N , 2010). T heir sparse d istrib u tio n , low m obility, sm all hom e
ranges, len g th y p a re n tal care, low fecu n d ity an d m onogam y of som e species
severely reduce th e ir chances to recover fro m ecological changes, m aking th e m
susceptible to e x tin c tio n (Foster & V incent, 2004).
In order to
u n d erstan d th e p o te n tia l evo lu tio n ary origin an d ecological
im plications of th is
u nique p h en o ty p e, th e
extrem ely specialized cranial
m orphology of S yngnathidae is th e focus of th is study.
1.3.
AIMS AND THESIS OUTLINE
1.3.1. C o n t e x t and general aims
This d octoral study form s p a rt of an in teg ral R esearch F o u n d a tio n Flanders
(FW O ) research pro ject e n title d “F u n c tio n a l consequences and ecological
im plications of extrem e m orphological specialization: design an d fu n c tio n of th e
feeding ap p aratu s in seahorses an d pipefishes (S yngnathidae)” (project code
G.0539.07). The p ro ject involves a p a rtn e rsh ip b etw een th e R esearch G roup
E volutionary M orphology of V erteb rates at G h e n t U niversity (U G ent), th e
L aboratory of F u n c tio n a l M orphology a t th e U niversity of A n tw erp (UA) an d th e
Royal Zoological Society of A n tw erp (KMDA).
T he m ain aims of th e pro ject are to investigate th e m orphological an d fu n c tio n a l
co n strain ts
of
th e
syngnathid
tro p h ic
system
related
to
th e
extrem e
specialization, and to explore w h e th e r th ese c o n strain ts m ig h t p artially explain
th e reduced ecological resilience of syngnathids. The core of th is research p ro ject
consists of th e follow ing th re e facets:
12
C h a p te r i - Introduction
• A m orphological analysis of th e cranial diversity in syngnathids,
w h ich I carried o u t an d is th e subject of th is d issertatio n
(UGent).
• A fu n c tio n a l analysis of feeding in syngnathids, p erfo rm ed by
G e rt Roos an d Sam V an W assen b erg h (UA).
• An
o n to g en etic
analysis
of
m orphological
an d
fu n c tio n a l
changes in th e head of a syngnathid, an ongoing stu d y started by
me an d A nnelies G enbrugge, in co llab o ratio n w ith G e rt Roos
an d Sam V an W assenberg h (U G ent an d UA).
As th e th ird p a rty of th e project, th e K M D A c o n trib u te d by supplying anim als,
b o th th ro u g h C IT E S -regulated trad e and established breed in g program s.
T he fu n c tio n a l c o m p o n en t of th e research, being carried o u t a t th e U niversity of
A ntw erp, focuses m ainly on th e dynam ics an d kinem atics of th e specialized
su ctio n feeding, th e fu n c tio n a l flexibility and v ersatility of th e feeding ap p aratu s
an d th e effect of sn o u t le n g th variatio n o n th e perfo rm an ce (Roos, 2010).
This d octoral thesis is th e collectio n of th e results of th e m orphological analyses
as p a rt of th e F W O project. T he im plications of sn o u t elo n g atio n as d ep icted in
th e previous p arag rap h form th e cen tral th em e of th is research. The overall goal
is to resolve th e evolutionary p a tte rn leading to th e extrem e m orphological
specialization of seahorses an d pipefishes.
1.3.2. Thesis o u t li n e and s p e c if ic aims
In to ta l, th is d issertatio n is divided in to seven chapters.
T he first (current) c h ap ter starts w ith a sh o rt in tro d u c tio n to th e concepts of
‘evo lu tio n ’ an d
‘specialization’. A
n ex t
p arag rap h
provides
b ack g ro u n d
in fo rm atio n on th e phylogenetic p o sitio n , general biology an d specialized
feeding ap p aratu s of th e fam ily Syngnathidae as an in tro d u c tio n to these
w o n d e rfu l anim als. The c h ap ter concludes w ith th e overall objectives of th e
research an d th e o u tlin e of th e thesis.
Chapter i - Introduction
13
In th e second c h ap ter th e exam ined specim ens, th e ir origin an d ap p licatio n are
listed. Also th e m ethods utilized are explained here, in ord er to avoid re p e titio n
in th e follow ing chapters. T he ad d itio n al chapters w ill only c o n tain a concise
m aterial & m ethods section w ith reference to th e second chapter. The last
parag rap h of th is ch ap te r includes som e term in o lo g y used in th e rem ain d er of th e
dissertation.
T he first results are given in th e th ird chapter. A detailed m orphological
descrip tio n of th e feeding ap p aratu s of a few syn g n ath id rep resen tativ es is
provided. The stru c tu ra l an d fu n c tio n a l in teractio n s b etw een th e individual
com p o n en ts in th e head are exam ined in ord er to u n d e rstan d h o w seahorses and
pipefishes are able to p e rfo rm such exceptionally fast an d p o w erfu l su ctio n
feeding. Tw o subdivisions are made. The first one focuses o n th e p ecu liar sn o u t
m orphology of Syngnathus rostellatus N ilsson an d Hippocampus capensis in a m ere
o n to g en etical co n tex t an d includes com parison w ith th e generalized p erco m o rp h
Gasterosteus aculeatus (C h ap ter 3.1). In p articu lar, th e follow ing q u estio n s w ere
addressed.
• Is sn o u t elo n g atio n already p re se n t early in ontogeny?
• W h ic h osteological co m p o n en ts fo rm p a rt of th e sn o u t and how
do th e y interact?
• W h a t m orphological changes c o n trib u te to an ev o lu tio n ary sh ift
tow ards sn o u t elongation?
T he second study evaluates th e im plications of sn o u t elo n g atio n o n th e
m usculoskeletal stru c tu re by com paring th e h ead m orphology of several
sy ngnathid species w ith varying sn o u t length. It is h y p o th esized th a t changes in
th e sn o u t le n g th w ill alter th e fu n c tio n a l lever system s as to assure th e same
kin em atical energy o u tp u t (C hap ter 3.2). The follow ing q u estio n s w ere addressed.
• Does sn o u t elo ngation affect th e cranial m usculoskeletal system?
• Is th e syngnathid feeding ap p aratu s a m orphologically conserved
e n tity (for exam ple due to physical constraints)?
• W h a t are th e m ain m orphological differences w ith in th e family?
T he fo u rth ch ap te r focuses on th e m orphological v ariatio n fo u n d w ith in th e
sy ngnathid family. A geom etric m o rp h o m etric analysis is carried o u t to analyze
C h a p te r i - Introduction
14
th e in tra- and interspecific shape v ariatio n in th e h ead of several syngnathid
species. By q u antifying th e a m o u n t of v ariatio n in th e cranium , th e hypothesis
th a t specialization leads to a reduced m orphological p lasticity is tested. In
addition, juvenile Hippocampus reidi G insburg are in clu d ed in th e analysis as to
fu rth e r elaborate on th e developm ental shape changes of th e head. The follow ing
questions w ere addressed.
• W h a t are th e m ost p ro m in e n t differences in th e head shape of
pipefishes versus seahorses, an d are th ese differences related to
relative sn o u t length?
• Does relative sn o u t le n g th have an effect o n th e a m o u n t of
intraspecific
v ariatio n
(i.e.
are
lo n g -sn o u ted
species
m ore
c o n strain ed th a n sh o rt-sn o u ted ones)?
• Is th e developm ental p erio d a fte r release from th e p o u ch in
seahorses characterized by im p o rta n t head shape changes?
In th e fifth chapter, th e results o n th e fu n c tio n a l in te rp re ta tio n of th e tro p h ic
m orphology in syngnathids are presented. The ch ap te r consists of tw o parts.
In th e first p art, a previously described p la n ar fo u r-b ar m odel fo r su ctio n feeding
in syngnathids (Fig. 1.6; M uller, 1987) is evaluated. This in te g ra tio n of fo rm and
fu n c tio n is done in close collab o ratio n w ith th e U niversity of A ntw erp. G ert
Roos and Sam V an W assenbergh p e rfo rm ed th e k in em atical analyses, w hereas I
provided a detailed m orphological analysis of th e individual co m p o n en ts
(linkages, jo in ts an d muscles) involved in th e m odel (C h ap ter 5.1). T he follow ing
questions w ere addressed.
• H ow does th e syngnathid feeding ap p aratu s w o rk an d h o w does
it differ from a typical teleo stean feeding sequence?
• H ow are syngnathids able to achieve th ese incredibly sh o rt prey
cap tu re tim es?
• Does th e proposed th e o re tic al fo u r-b ar m echanism (Fig. 1.6;
M uller, 1987) reflect th e feeding kinem atics?
T he second p a rt concerns th e m echanical stress exerted o n th e skull as a resu lt of
th e fast and p o w erfu l su ctio n feeding. The tro p h ic ap p aratu s needs to cope w ith
h igh forces an d pressures acting o n it d u ring feeding. W h e th e r and h o w th e
le n g th an d th e stru c tu re of th e skeletal feeding ap p aratu s in fluence th e stress
Chapter i - Introduction
15
d istrib u tio n in th e skull is th e objective of th is stu d y (C hapter 5.2). The follow ing
questions w ere addressed.
• W h e re in th e head does th e stress caused by m echanical loading
d u ring su ctio n feeding accum ulate?
• Is th e re
a relatio n sh ip
b etw een
sn o u t
le n g th
an d
stress
d istrib u tio n p a tte rn in th e sy n g n ath id cranium ?
All findings are discussed in ch ap te r six. The o b tain ed results are synthesized and
in te g ra te d w ith data from th e lite ra tu re in an a tte m p t to unravel th e evo lu tio n ary
p a tte rn leading to th e extrem ely specialized cran ial m orphology. In th e first p art,
th e role of th e extensive p a re n ta l care in th e sy n g n ath id ev o lu tio n is considered.
N ext, th e m odifications of b o th th e feeding ap p aratu s and th e feeding m ode are
discussed. Thirdly, evolutionary tra n sitio n s w ith in th e fam ily, such as th e origin
of th e grasping tail and tilte d h ead w ith respect to th e body, are dealt w ith in a
phylogenetic fram ew ork. The last p a rt of th e general discussion includes a
concise conclusion.
A nd finally a synopsis, b o th in English and D u tch , covering all previously
in tro d u ce d chapters, is provided in th e sev en th chapter.
i6
C h a p te r l - Introduction
17
2
2.1.
Material & methods
M aterial
T he m ain p a rt of th is research w as based o n 76 specim ens covering 14 sy n g n ath id
species (Table 2.1, 2.2, Fig. 2.1). Except fo r th e specim ens of Syngnathus acus, m ost
specim ens w ere also used in a geom etric m o rp h o m etric analysis, w h ich included
368 specim ens of 38 d iffe re n t species (Table 2.3). A m ong th o se w as an
o n to g en etic series of Hippocampus reidi, w ith juveniles ranging in age b etw een 1
an d 65 days, used for an o n to g e n etic m o rp h o m etric stu d y in c o m b in atio n w ith
th e adults of H. reidi.
Selection of th e species w as based o n system atic p o sitio n w ith in th e existing
phylogenetic fram ew ork (Fig. 1.5; W ilso n & O rr, subm itted), sn o u t m orphology
(prim arily regarding th e level of sn o u t elongation) an d th e availability of
specim ens.
As visible in table 2.1, m any specim ens w ere o b ta in e d th ro u g h C IT E S-regulated
com m ercial trade, b u t th e specim ens of Hippocampus capensis an d Hippocampus kuda
B leeker w ere o b tain ed from th e A n tw erp Zoo an d th e specim ens of S. rostellatus
an d S. acus w ere caught on th e Belgian c o n tin e n ta l shelf (N o rth Sea) by th e
M arin e Biology research group of th e G h e n t U niversity. O f th e specim ens
analyzed m orphom etrically, m ost belong to th e collection of th e Zoological
M u seu m of A m sterdam , b u t th e specim ens of Microphis brachyurus aculeatus
i8
C h a p t e r 2 - M a te ria l & m e t h o d s
(Bleeker) are p a rt of th e Royal M u seu m of C en tral A frica co llectio n an d some
com m ercially o b ta in e d specim ens w ere included as w ell (Table 2.3). Several
juveniles w ere included in th e research: H. capensis is b red a t th e A n tw erp Zoo,
th e juvenile H. reidi are th e offspring of seahorses held in a breeding facility at
th e research group (as is described u n d er 2.2.1) an d som e of th e cau g h t S.
rostellatus males w ere p re g n a n t (Table 2.2).
T a b l e 2 .1 - A d u l t s p e c i m e n s u s e d f o r t h e m o r p h o l o g i c a l s t u d i e s w i t h r e f e r e n c e t o t h e s p e c i f ic
c h a p te r (s ) ( b e t w e e n b r a c k e t s ) i n w h i c h t h e y a re u t i l iz e d . A b b r e v ia t io n s : A Z , A n t w e r p Z o o ; C ,
c o m m e r c i a l tr a d e ; C T , c o m p u t e d t o m o g r a p h y ; F E A , f i n i t e e l e m e n t a n a ly s is ; H L , h e a d le n g t h ;
N S , N o r t h S e a ; S L , s ta n d a r d le n g t h .
sp ec ies
Corythoichthys intestinalis
Corythoichthys intestinalis
Corythoichthys intestinalis
Corythoichthys intestinalis
Corythoichthys intestinalis
Doryrham phus dactyliophorus
Doryrham phus dactyliophorus
Doryrham phus dactyliophorus
Doryrham phus janssi
Doryrham phus janssi
Doryrham phus janssi
Doryrham phus janssi
Doryrham phus melanopleura
Doryrham phus melanopleura
Dunckerocampus pessuliferus
Dunckerocampus pessuliferus
Syngnathus acus
Syngnathus acus
Syngnathus acus
Syngnathus acus
Syngnathus acus
Syngnathus acus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
Syngnathus rostellatus
H ippocam pus abdominalis
H ippocam pus barbouri
H ippocam pus breviceps
H ippocam pus capensis
H ippocam pus capensis
H ippocam pus capensis
H ippocam pus capensis
H ippocam pus capensis
H ippocam pus kuda
( c o n tin u e d o n n e x t page)
SL (mm)
120.2
102.7
117.6
120.8
136.2
103.5
91.2
105.1
102.7
105.6
81.5
90.8
41.6
56.5
94-5
112.3
330.2
332.4
325.1
350.1
371.6
311.3
126.2
111.1
111.5
97-9
95-7
100.0
109.1
110.8
93.8
105.0
97-5
114.5
126.4
129.1
222.7
141.4
50.7
96.1
97.6
99.0
99.2
133.0
117.4
HL (mm)
16.1
16.9
16.4
16.4
18.2
25.4
23.6
25.0
22.6
24.5
16.4
19-5
11.6
13.2
26.0
30.1
39-4
43-4
42.7
45.8
46.1
38.7
14.8
14.5
13.0
12.2
13.6
12.6
14.0
15-3
13-5
13.0
13.7
14.1
15.1
15-5
30.9
25-5
9.8
15.2
17.7
16.0
16.8
19.4
24.8
o rig in
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
C
C
c
AZ
AZ
AZ
AZ
AZ
AZ
m eth o d
C T -scan n in g (3.2)
clearin g a n d sta in in g (3.2)
clearin g a n d sta in in g (3.2)
d issectin g (3.2)
d issectin g (3.2)
C T -scan n in g , FEA, d isse c tin g (3.2, 5.2)
se rial se ctio n in g , 3 D -re c o n stru c tin g (3.2)
clearin g a n d sta in in g (3.2)
se rial se c tio n in g (3.2)
d issectin g (3.2)
clearin g a n d sta in in g (3.2)
clearin g a n d sta in in g (3.2)
clearin g a n d sta in in g (3.2)
C T -scan n in g , FEA, d isse c tin g (3.2, 5.2)
se rial se c tio n in g (3.2)
d issectin g (3.2)
se rial se c tio n in g
se rial se c tio n in g
clearin g a n d sta in in g
d issectin g
d issectin g
C T -scan n in g
clearin g a n d sta in in g (3.1)
clearin g a n d sta in in g (3.1)
clearin g a n d sta in in g
clearin g a n d sta in in g (3.1)
sta in in g
sta in in g
sta in in g
C T -scan n in g , F E A (5.2)
sta in in g
clearin g a n d sta in in g (3.1)
sta in in g
d issectin g
d issectin g
d issectin g
C T -scan n in g , FEA, d isse c tin g (3.2, 5.2)
C T -scan n in g
C T -scan n in g
clearin g a n d sta in in g (3.1)
clearin g a n d sta in in g (3.1)
clearin g a n d sta in in g (3.1)
C T -scan n in g
clearin g a n d sta in in g
clearin g a n d sta in in g
C h a p t e r 2 - M a te ria l & m e t h o d s
sp ec ies
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
H ippocam pus
T a b le
2 .2
kuda
kuda
kuda
re id
re id
re id
re id
re id
re id
re id
re id
re id
zosterae
zosterae
zosterae
-
L is t
of
19
SL (mm)
HL (mm)
93-3
94.6
125.2
117.2
113.5
103.5
119.0
109.7
113.7
138.1
152.8
123.9
28.3
29.0
32.8
19-5
19.2
23.8
23.8
21.4
24.0
25.2
24.4
23.8
28.3
32.2
ju v e n ile
o rig in
AZ
AZ
AZ
C
C
C
C
C
C
C
C
C
C
C
C
25-5
4-5
5-3
5-3
s y n g n a t h id
sp e c ie s
m eth o d
clearin g a n d sta in in g
clearin g a n d sta in in g
se rial se c tio n in g
clearin g a n d sta in in g (3.1, 3.2, 5.1)
clearin g a n d sta in in g (3.1, 3.2, 5.1)
se rial se ctio n in g , 3 D -re c o n stru c tin g (3.2, 5.1)
C T -scan n in g , F E A (3.1, 5.1, 5.2)
clearin g a n d sta in in g (3.1, 5.1)
d issectin g (3.2)
d issectin g (3.2)
d issectin g (3.2)
C T -scan n in g , F E A (5.2)
C T -scan n in g , se rial se c tio n in g
C T -scan n in g , FEA, se rial se c tio n in g (3.2, 5.2)
clearin g a n d sta in in g (3.2)
u sed
fo r
th e
m o r p h o lo g ic a l
s t u d ie s .
A b b r e v ia t io n s : C T , c o m p u t e d t o m o g r a p h y ; d p , d a y s p o s t - r e l e a s e f r o m p o u c h ; H L , h e a d le n g t h ;
p r , p r e r e le a s e f r o m p o u c h ; S L , s t a n d a r d l e n g t h .
m eth o d
clearin g a n d sta in in g
clearin g a n d sta in in g (3.1)
clearin g a n d sta in in g (3.1)
se rial se ctio n in g , 3 D -re c o n stru c tin g (3.1)
3 dp
4 dp
clearin g a n d sta in in g (3.1)
1 dp
se rial se ctio n in g , 3 D -re c o n stru c tin g (3.1)
1 dp
C T -scan n in g
1 dp
c learin g a n d sta in in g
7-5
1 dp
clearin g a n d sta in in g
6.0
p r1
clearin g a n d sta in in g (3.1)
11.0
p r1
clearin g a n d sta in in g (3.1)
14.5
p r1
clearin g a n d sta in in g (3.1)
11.3
p r1
clearin g a n d sta in in g (3.1)
11.4
1-9
p r1
se rial se c tio n in g (3.1)
13.l 2
2.1
p r1
se rial se c tio n in g
19.3
3-5
p r1
se rial se c tio n in g
4.8
25-5
1 T h e age o f th e sp ecim en s o f S. rostellatus c o u ld n o t be d e te rm in e d since th e y w e re c o lle c te d fro m th e b r o o d a re a o f
th e sa m p le d m ales.
2 S ta n d a rd le n g th w as e s tim a te d b y in te r p o la tio n b ased o n th e h e a d le n g th over sta n d a rd le n g th ra tio o f o th e r
specim ens.
sp ec ies
H ippocam pus capensis
H ippocam pus capensis
H ippocam pus capensis
H ippocam pus capensis
H ippocam pus capensis
H ippocam pus reidi
H ippocam pus reidi
H ippocam pus reidi
H ippocam pus reidi
S yngnathus rostellatus
S yngnathus rostellatus
S yngnathus rostellatus
S yngnathus rostellatus
S yngnathus rostellatus
S yngnathus rostellatus
S yngnathus rostellatus
SL (mm)
HL (mm)
29.4
13.3
13.6
12.8
14.0
7.0
7.2
6.7
2.8
2.9
2.8
3.2
2.0
2.1
2.3
1.8
1.8
2.3
2.2
age
14 dp
1 dp
2 dp
T a b l e 2 .3 - L is t o f a ll 3 8 s y n g n a t h id s p e c i e s p l u s t h e j u v e n i l e s o f H i p p o c a m p u s r e i d i u s e d f o r
t h e m o r p h o m e t r i c a n a ly s is . A b b r e v ia t io n s : A Z , A n t w e r p Z o o ; B , b r e d a t t h e r e s e a r c h g r o u p ; C ,
c o m m e r c i a l tr a d e ; D , d r a w in g s ( s e e c h a p t e r 4); H L , h e a d le n g t h ; N S , N o r t h S e a ; R M C A , R o y a l
M u s e u m o f C e n t r a l A fr ic a ; S n L , s n o u t le n g t h ; s t d e v , s t a n d a r d d e v ia t io n ; Z M A , Z o o l o g i c a l
M u s e u m o f A m s te r d a m .
sp ec ies
Anarchopterus criniger
Anarchopterus tectus
B ryx dunckeri
B ryx randalli
Campichthys tricarinatus
Corythoichthys intestinalis
Cosmocampus balli
Cosmocampus brachycephalus
(c o n tin u e d o n n e x t page)
n u m b er o f sp ec im en s
o r ig in
m ean (Sn L /H L ) ± st d ev
1
1
1
1
1
D
D
D
D
D
C +Z M A
D
D
0.253
0.250
0.265
0.337
0.315
0.46910.019
0.359
0.271
7
1
1
C h a p t e r 2 - M a te ria l & m e t h o d s
20
sp e c ie s
n u m b e r o f sp e c im e n s
o r ig in
m e a n (S n L /H L ) ± s t d e v
1
2
6
1
6
D
ZM A
ZM A
C
C
C
C
ZM A
ZM A
RM CA
D
D
D
ZM A
D
NS
ZM A
C+ZM A
C
C
AZ
ZM A
ZM A
ZM A
C+ZM A
ZM A
C
B
ZM A
ZM A
C
0.328
0.347+0.013
0.372+0.018
0.662
0.572+0.012
0.438+0.023
0.649+0.083
0.539
0.569
0.578+0.012
0.420
0.238
0.180
0.553+0.026
0.278
0.458+0.011
0.483+0.038
0.390+0.024
0.502+0.017
0.424+0.014
0.338+0.022
0.360+0.029
0.328
0.476+0.015
0.428+0.036
0.354+0.032
0.426+0.026
0.442+0.028
0.375+0.028
0.496+0.006
0.315+0.135
Cosmocampus hildebrandi
Doryichthys martensii
Doryichthys retzii
D oryrham phus dactyliophorus
D oryrham phus janssi
D oryrham phus melanopleura
Dunckerocampus pessuliferus
Gasterotokeus biaculeatus
H aliichthys taeniophorus
M icrophis brachyurus aculeatus
M icrophis caudocarinatus
M in yich th ys inusitatus
Siokunichthys bentuviai
Syngnathoides biaculeatus
Syngnathus parvicarinatus
Syngnathus rostellatus
Syngnathus typhle
H ippocam pus abdominalis
H ippocam pus barbouri
H ippocam pus breviceps
H ippocam pus capensis
H ippocam pus guttulatus
H ippocam pus hippocampus
H ippocam pus histrix
H ippocam pus kuda
H ippocam pus ramulosus
H ippocam pus reidi
H ippocam pus reidi juveniles
H ippocam pus spinosissimus
H ippocam pus trim aculatus
H ippocam pus zosterae
2.2.
3
2
1
1
95
1
1
1
18
1
15
15
5
3
2
4
29
1
5
40
18
12
54
3
5
3
M ethods
2.2.1. K eep ing live s p e c i m e n s
Specim ens o b tain ed from com m ercial tra d e (Bassleer Biofish, B elgium an d De
Jong M arinelife, th e N etherlands) w ere k e p t in a 1.15 x 0.55 x 0.65 m aq u ariu m
(4 1 1 1). T he ta n k had a 10 cm sandy b o tto m and live rock, a p ro te in skim m er, tw o
h e atin g elem ents, tw o lig h t bulbs w ith a p h o to p e rio d of 12 h o u rs, a filter an d a
pum p. T em p eratu re was k e p t co n sta n t at 240 C and salin ity a t 32 ppt. The anim als
w ere fed defrosted Artemia or M ysis daily as syngnathids feed m ainly o n sm all
crustaceans (Payne et al., 1998; F o ster & V incent, 2004; K endrick & H yndes, 2005;
Felicio et al., 2006; K arina et al., 2006; K itsos et al., 2008).
T he juvenile Hippocampus reidi specim ens w ere o b ta in e d fro m successful
C h a p t e r 2 - M a te ria l & m e t h o d s
21
breeding in th e lab. Juveniles th a t em erged fro m th e m ale’s p o u ch w ere
im m ediately tra n sfe rre d to a sm aller ta n k 0.35 x 0.20 x 0.25 m (17.5 1) a t a
te m p e ra tu re of 26o C. W a te r w as p artly refresh ed th re e tim es a day an d th e
juveniles w ere fed ro tifers first an d la te r A nem ia, th re e tim es daily. V arious males
gave b irth to in to ta l n ine litte rs of young, w h ich w ere k e p t in separate aquaria.
M o rta lity caused none of th e juveniles to live over 94 days.
2.2.2. Sacrificing and fix ati on
Sacrificing w as achieved by an overdose of M S 222 (tricaine m eth an esu lfo n ate,
Sigma A ldrich, St. Louis, M O ) in accordance w ith th e Belgian law o n th e
p ro te c tio n of lab o rato ry anim als (KB d.d. N ovem ber 14, 1993). F ix atio n w as done
in a 10% b u ffere d and n eu tra liz ed fo rm alin solution.
2.2.3. Biometry
A fter sacrificing, sm all individuals w ere p h o to g ra p h e d using a digital cam era
(C olorV iew 8, S oft Im aging System G m bH M ü n ste r, G erm any) m o u n te d o n an
O lym pus SZX-9 stereoscopic m icroscope (O lym pus G m bH H am burg, G erm any)
an d driven by th e softw are pro g ram analySIS 5.0 (Soft Im aging System G m bH
M ü n ste r, G erm any). P ictures of larger specim ens w ere c ap tu red by m eans of a
m acroscopic digital cam era (Sony DSC-F717 cam era w ith C ari Zeiss VarioS onnar 9-7-48.5mm f/2-2.4 lens, C an o n EOS 400D cam era w ith C an o n EF-S 1855mm fZ3.5-5.60r lens or N ik o n D40X cam era w ith i8-55m m f/3.5-5.6 G il AF-S
DX N IK K O R lens). A scale bar was in clu d ed in each p ic tu re and th e specim ens
w ere o rie n te d in a plane parallel to th e cam era, w ith m axim al overlap of p aired
anatom ical stru c tu re s as to m inim ize o rie n ta tio n a l errors. For all specim ens,
freew are ImageJ 1.43U (R asband W .S., US N atio n al In s titu te s of H ealth ,
M aryland) w as used to m easure stan d ard le n g th (SL), h ead le n g th (HL) an d sn o u t
le n g th (SnL). S tandard le n g th of pipefishes w as m easured as described by
FishBase (Froese & Pauly, 2011), i.e. th e distance b etw een th e tip of th e sn o u t and
th e base of th e caudal fin (Fig. 2.2A). In seahorses th e situ a tio n is d ifferen t since
C h a p t e r 2 - M a te ria l & m e t h o d s
22
th e y lack a caudal fin. H ere stan d ard le n g th w as defined according to th e
m easuring p ro to co l described by Lourie et al. (i999a;b) i.e. as th e sum of head
len g th , tru n k le n g th (TrL) an d ta il le n g th (TaL) (Fig. 2.2B). For all specim ens,
head le n g th w as m easured as th e distance from th e tip of th e u p p er jaw to th e
base of th e cleith ral curvature ju st below th e gili slit, an d sn o u t le n g th as th e
distance betw een th e tip of th e u p p er jaw and th e caudal m argin of th e nostril.
2.2.4. !n f o f o clearing and staining
C learing and stain in g of bone and cartilage w ith alizarin red S an d alcian blue
w as done according to a slightly m odified version of th e p ro to co l of Taylor & V an
Dyke (1985) (Table 2.4). A djustm en ts include om ission of th e degreasing, final
cleaning and guan in e rem oval steps, lim ita tio n of cartilage stain in g to less th a n
12 h o u rs as th e acetic acid could decalcify bone, w h ich m akes alizarin red bone
stain in g ineffective, an d sh o rte n in g of bleaching tim e to a few h o u rs m axim um.
A n O lym pus SZX-7 stereoscopic m icroscope eq u ip p ed w ith an O lym pus SZX-DA
draw ing a tta c h m e n t was used to study and draw th e b o n y an d cartilaginous
elem ents of th e cranium . K O H 5% was used to com pletely d isarticu late th e
Suspensorium of som e ad u lt specim ens, so all bones could be individually
exam ined in detail.
T a b l e 2 .4 - C le a r in g a n d s t a i n i n g p r o t o c o l m o d i f i e d f r o m T a y lo r & V a n D y k e (1 9 8 5 ).
s te p
d e h y d ra tio n
cartilag e sta in in g
n e u tr a liz a tio n
b le a c h in g
clearin g
b o n e sta in in g
f u rth e r clearin g
p re se rv a tio n
s o lu ti o n
50% alco h o l
75% alco h o l
96-100% alco h o l
96-100% alco h o l
9-30 m g a lc ia n b lu e 8G X (Sigm a) i n lo o m l o f 40% g lacial ac e tic acid a n d
60% ab so lu te a lc o h o l
sa tu ra te d b o ra x (N a2B40 7.io H 20 )
10% H 20 2 in 0.5% K O H
en zy m e b u ffer: 0.45 g p u rifie d try p s in in 400 m l o f 30% sa tu ra te d b o ra x
0.5% K O H i n 0.1% a liz a rin re d S (Sigma)
0.5-3% K O H
25% gly cerin e + 75% 0.5% K O H
50% gly cerin e + 50% 0.5% K O H
75% gly cerin e + 25% 0.5% K O H
100% glycerine
d u r a t io n
12h
12h
12h
12h
12h
48h
0.5-6I1
12h-...
24h
12h-...
12h
12h
12h
storage
C h a p t e r 2 - M a te ria l & m e t h o d s
23
2.2.5. D is s e c ti n g
For th e study of so ft tissue stru ctu res, such as ligam ents, ten d o n s and m uscles,
dissections w ere perform ed. V isualization of m uscle fibre o rie n ta tio n , origin and
in sertio n sites was en h an ced by th e use of an iodine so lu tio n (Bock & Shear,
1972).
2.2.6. Serial s e c t i o n i n g
Serial histological cross sections of 14 specim ens w ere m ade in o rd er to stu d y th e
detailed anatom y. H ead or e n tire body w as em bedded in T echnovit 7100 (H eraeus
K ulzer W eh rh eim , G erm any) according to th e T echnovit 7100 U ser In stru c tio n s
(Table 2.5). N ext, sem i-thin sections (2 or 5 pm) w ere c u t using a Leica Polycut
SM 2500 sliding m icrotom e (Leica M icrosystem s G m bH W etzlar, G erm any)
equipped w ith a w olfram carbide co ated knife. Finally, slices w ere stain ed w ith
to lu id in e blue, m o u n te d on slides w ith e ith e r DPX or a xylene based m o u n tin g
m edium an d covered.
T a b l e 2 .5 - T e c h n o v i t 7 1 0 0 U s e r I n s t r u c t io n s e m b e d d i n g p r o t o c o l
s te p
s o lu ti o n
d u r a t io n
v acu u m fix a tio n
w ash in g
d é c a lc ific a tio n
w ash in g
d e h y d ra tio n
10% b u ffe re d fo rm a lin
ta p w a te r
D ecalc (H isto lab )
ta p w a te r
30% alco h o l
50% alco h o l
70% alco h o l
96% a lc o h o l (tw o tim es a lc o h o l renew al)
T e c h n o v it 7100 s o lu tio n A
T e c h n o v it 7100 s o lu tio n A re n e w a l
ad d T e c h n o v it 7100 h a rd e r II
place in deep fre e z e r
place a t ro o m te m p e ra tu re (check progress)
place in o v e n (approx. 40°C)
days-w eeks
8h
36h
e m b ed d in g
p o ly m e riz a tio n
12h
12h
12h
12h
m in. 24 h
m in. 48 h
12h
12h
ap p ro x . 12h
lh
2.2.7. C o m p u t e d t o m o g r a p h y s c an ni ng
C T-scans of one or m ore specim ens of alm ost all species (w ith th e ex cep tio n of
Doryrhamphus janssi, Dunckerocampus pessuliferus Fow ler an d Hippocampus kuda)
C h a p t e r 2 - M a te ria l & m e t h o d s
24
w ere m ade e ith e r at th e M icro CT R esearch G ro u p of th e U niversity of A n tw erp
(M CT; h ttp ://w e b h o i.u a .a c.b e /m c t) or a t th e C en tre fo r X-ray T om ography of
G h e n t U niversity (UGCT; M asschaele et al., 2007; h ttp ://w w w .u g ct.u g en t.b e). The
M C T group uses a SkyScan-1072 com pact d esk to p system fo r non -d estru ctiv e
h igh reso lu tio n X-ray m icroscopy an d m icrotom ography (SkyScan K ontich,
Belgium). A t th e U G CT, all specim ens w ere scanned using th e tran sm issio n tu b e
head, a t 100 kV tu b e voltage (except fo r one of th e H. reidi specim ens, w h ich was
scanned w ith th e d irectio n al tu b e head a t 80 kV tu b e voltage). For each
individual b etw een 500 an d 1900 p ro jectio n s of b etw een 500x500 an d 1096x1096
pixels w ere recorded (1000 p rojectio n s of 748x748 pixels fo r th e H. reidi), w ith an
exposure tim e betw een 1 s an d 1.6 s p er p ro jectio n an d covering 360 degrees. The
raw data w ere processed and reco n stru cte d using th e in-house developed CTsoftw are O ctopus (V lassenbroeck et al., 2007).
2.2.8. Graphical 3 D - r e c o n s t r u c t i n g
C o m p u ter-g en erated
3D -reconstructions
w ere
m ade
to
visualize
skeletal
elem ents, using C T-data, and overall to p o g rap h y (i.e. bones, cartilage, muscles,
ligam ents, ten d o n s, etc), using histological sections. Im ages of th e histological
sections w ere cap tu red using a digital cam era (C olorview 8) m o u n te d o n a
R eichert-Jung Polyvar lig h t m icroscope (R eichert D epew , USA), co n tro lled by th e
softw are program analySIS 5.0. The digital im ages of histological data w ere
im p o rted in th e softw are package A m ira (3.1, 4.1.0, 5.1.0 or 5.2.2; Visage Im aging,
Berlin, G erm any). A lignm ent of th e histological sections and tracin g of th e
elem ents was done m anually by su p erim p o sitio n to g et m axim al overlap of all
structures. E ach elem ent w as ren d ered an d separately sm oothed. In som e cases
R hinoceros 3.0 softw are (M cN eel E urope SL Barcelona, Spain) w as used for
m aking images of all stru c tu re s com bined. For th e C T-data, b o th softw are
program s A m ira 5.2.2 an d myVGL 2.0.4 (V olum e G raphics G m bH , H eidelberg,
G erm any) w ere used to g enerate volum e ren d ered as w ell as surface ren d ered
re c o n stru ctio n s of th e cranial skeleton.
C h a p t e r 2 - M a te ria l & m e t h o d s
25
2.2.9. G e o m e t r i c m o r p h o m e t r i c analysis
G eom etric m o rphom etries are freq u en tly used to qualify an d q u an tify anatom ical
shape changes w ith an o ntogenetic, phylogenetic or pathological basis. Especially
landm ark-based analyses provide an accu rate d escrip tio n of th ese shape changes
an d
in
a d d itio n
have
th e
advantage
of
allow ing
a
clear visualization,
in te rp re ta tio n an d com m u n icatio n of th e results (Z elditch et al., 2004).
H ere, th e landm ark-based geom etric m o rp h o m etric analysis involved th e use of
th e T h in Plates Spline freew are (tps; R ohlf, F.J., State U niversity of Stony Brook,
N ew York). F irst tp sU til version 1.46 (Rohlf, 2010a) w as used fo r co n stru ctio n of
th e tps-files. Landm arks (LM) w ere digitized o n th e p ictu res of th e head w ith
tpsD ig2 version 2.16 (Rohlf, 2010b). The follow ing 15 hom ologous landm arks
w ere digitized (Fig. 2.3): (1) dorso ro stral tip of prem axillary bone; (2) distal p o in t
of d en tary bone; (3) v en tro cau d al p o in t of low er jaw below suspensorial
articu latio n ; (4) v e n tra l p o in t w h ere sn o u t is d orsoventrally at its narrow est; (5)
dorsal p o in t w here sn o u t is dorsoventrally a t its n arrow est; (6) base of
m esethm oidal curvature; (7) caudal b o rd er of nostril; (8) o p erculo-hyom andibular
articu latio n ; (9) v e n tro ro stra l tip of sp h en o tic bone; (10) m ost dorsal p o in t of
skull at th e level of th e eye w ith respect to th e line co n n ectin g L M i w ith LM12;
(11) m ost dorsal p o in t of skull a t th e level of th e braincase w ith respect to th e
line co n n ectin g L M i w ith LM12; (12) base of cleith ral cu rv atu re w h ere th e gili
slit is situated; (13) dorsal p o in t of p ecto ral fin base; (14) v e n tra l p o in t of p ecto ral
fin base; and (15) v en tro cau d al tip of p reo p ercu lar bone. For all specim ens, head
le n g th w as calculated as th e distance b etw een L M i an d LM12, sn o u t le n g th as
th e distance betw een L M i and LM 7 an d sn o u t h e ig h t as th e distance b etw een
LM 4 an d LM5. The in te rla n d m a rk distances w ere calculated using PAST version
2.03 (H am m er et al., 2001).
S uperim p o sitio n is used to rem ove all non-shape v ariatio n in th e co n fig u ratio n s
of landm arks
(i.e. v ariatio n
in
p o sitio n , o rie n ta tio n
and
scale). V arious
su p erim p o sitio n m ethods exist, here th e landm arks w ere su b m itte d to a
generalized
P rocrustes
co n fig u ratio n s
using
analysis
least-squares
(GPA),
w h ich
estim ates
fo r
superim poses
tra n sla tio n
and
landm ark
ro ta tio n
p aram eters (Adams et al., 2004). The c en tro id of each c o n fig u ratio n is tra n slated
to th e origin, th e sets of landm arks are ro ta te d to m inim ize th e squared
26
C h a p t e r 2 - M a te ria l & m e t h o d s
differences betw een corresponding landm arks an d co n fig u ratio n s are scaled to a
com m on, u n it size. TpsSm all version 1.20 (Rohlf, 2003) w as used to co n firm th e
fitness of th e dataset for fu rth e r statistical analysis. N ext, fo r each specim en
p a rtia l w arps w ere com p u ted in tpsR elw version 1.49 (Rohlf, 2010c) and
su b m itte d to a prin cip al co m p o n en t analysis (PCA) resu ltin g in relative w arps
(RW s) th a t sum m arize th e variatio n am ong th e specim ens. D efo rm atio n grids
w ere g en erated to visualize th e shape v ariatio n an d relative w arp scores and
c en tro id size w ere saved for fu rth e r analyses.
O ne tps-file com prising all pipefishes and one fo r th e seahorses w ere loaded in
C oordG en6f (Sheets, 2000a) an d T w o G roup6h (Sheets, 2000b), b o th p a rt of th e
In te g ra te d M o rp h o m etries Package freew are (IM P; Sheets, H.D., C anisius
College, Buffalo, N ew York). The (d issim ila rity of pipefishes versus seahorses was
d eterm in ed by m eans of a G oodall’s F test, w h ich exam ines differences in m ean
shape betw een tw o groups relative to th e shape v ariatio n fo u n d w ith in th e
groups. It com pares th e P rocrustes distance b etw een th e m eans of tw o sam ples to
th e a m o u n t of v ariatio n fo u n d in th e samples. The p ro b ab ility th a t th is F-score
could have arisen by chance w as te ste d by a B ootstrap resam pling pro ced u re
(4,900 B ootstraps).
Finally, tpsR egr version 1.37 (Rohlf, 2009) w as used to p e rfo rm regression
analyses betw een th e o b tain ed relative w arp scores o n th e one h a n d and cen tro id
size, log head le n g th or sn o u t le n g th over head le n g th (i.e. relative sn o u t length)
on th e o th e r hand.
A n o th e r tps-file w as made, com prising only Hippocampus reidi specim ens, b o th
ad u lt an d juvenile. Id en tical landm arks w ere placed, a PC analysis was carried o u t
an d d efo rm atio n grids w ere used to visualize shape changes. In o rd er to d etect
im p o rta n t tra n sfo rm a tio n s d u ring developm ent, w e w an ted to p lo t h ead shape
(i.e. relative w arp scores) over age. Since w e did n o t k n o w th e exact age of all
specim ens, head le n g th was used as a proxy (co rrelatio n co efficien t b etw een age
an d head le n g th is 0.93, based on te n H. reidi juveniles w ith age ranging b etw een 1
an d 65 days; Fig. 2.4).
C h a p t e r 2 - M a te ria l & m e t h o d s
27
2 .2.10. Finite e l e m e n t analysis
F in ite elem en t analysis is a tech n iq u e in w h ich m echanical loading of a physical
system is sim ulated in a m odel to investigate th e resu ltin g stress d istrib u tio n . The
fin ite elem ent m eth o d allow s estim atio n of stress d istrib u tio n in objects of
com plex geom etry for w h ich n o exact analytical so lu tio n s exist. It im plies
dividing th e com plex an d irregularly shaped o bject in to a large n u m b er of
discrete, m anageable elem ents co n n ected to each o th e r by nodes. T he m esh, i.e. all
th e elem ents an d nodes com bined, needs to be assigned m aterial p ro p erties th a t
define th e elastic behaviour of th e m odel, such as Y oung’s m odulus an d P o isso n s
ratio. N ext, b o u n d ary co n strain ts are applied to th e m esh to an ch o r th e m odel
an d th e loading co n ditions are defined. Finally, th e m odel is solved u n d er th e
assigned m aterial p roperties, b o u n d ary co n d itio n s and applied forces to o b ta in
nodal displacem ents and th e n th e resu ltin g stresses. A lth o u g h in itially developed
for engineering, th e fin ite elem en t m eth o d has started to fin d its w ay in to
biological analyses as w ell (e.g. D u m o n t et al., 2005; R ichm ond et al., 2005).
T he raw C T -data of H. reidi and H. zosterae w ere processed to g en erate a
c o n tin u o u s, w a te r-tig h t’ surface m odel using th e softw are pro g ram M im ics 12
M aterialise NV (Leuven, Belgium); fo r all o th e r specim ens A m ira 5.1.0 or 5.2.2
w as used. The operculum , b ran ch ial arches an d p o st-cran ial stru c tu re s w ere
rem oved. All of th e m odels w ere sim plified an d in itially sm o o th ed in A m ira 5.1.0.
Ideally th e m odel of a fish head, having m ore degrees of freedom com pared to a
m am m alian skull, w ou ld consist of several separate b u t artic u la tin g bony
elem ents, including cartilaginous stru ctu res, su tu re m orphology, connective
tissue, ligam entous connections, etc. N o t only w ill th e com plexity of such a
system vastly increase th e c o m p u ta tio n al resources req u ired to c o n stru c t th e
fin ite elem ent m odel an d ru n th e analysis, b u t u n fo rtu n ate ly , m any of th e
necessary data is lacking for syngnathids in p a rtic u la r an d even fo r fishes in
general. Since gen eratin g a m odel th a t accurately reflects th e
im possible,
an
exploratory
analysis
was
p e rfo rm ed
by
reality is
m aking
some
sim plifications, i.e. th e c ran iu m w as m odelled as a single u n it, w ith o u t tak in g
su tu res in th e braincase and a rticu latio n s b etw een th e fu n c tio n a l co m p o n en ts in
th e skull in to account. The tw o su sp en so rio -n eu ro cran ial joints, th e suspensoriohyoid artic u la tio n , th e low er jaw, th e m axillary bones an d th e in te ro p e rc u lar
28
C h a p t e r 2 - M a te ria l & m e t h o d s
bone w ere fused an d tre a te d as solid, im m obile elem ents. O bviously, th e results
of these analyses m ust be in te rp re te d cautiously, yet I feei th a t th ey still give a
valuable estim atio n of th e stress d u rin g su ctio n feeding in syn g n ath id skulls.
To reduce th e effects of m odel size, R hinoceros 4.0 so ftw are (M cN eel E urope SL
Barcelona, Spain) w as used to scale th e reco n stru ctio n s to th e same braincase
len g th , defined as th e distance betw een th e ro stral b o rd er of th e p arietal bone
an d th e occipital jo in t of th e basioccipital bone
G eom agic S tudio 10 G m bH (S tu ttg art, G erm any) w as used to fin e -tu n e th e
m odels, p rep are th e m for solid m eshing (i.e. fill artificial holes, rep air in tersectin g
triangles and adjust th e aspect ra tio of th e triangles), an d to reduce trian g le
num ber. The surface m odels w ere ex p o rted to S trandy P ty Ltd (Sydney, A ustralia)
an d tran sfo rm ed in to solid models.
A ssigning m aterial p ro p erties to th e m odels w as n o t straig h tfo rw ard since th e re
is only data of tw o fish species available. E rickson et al. (2002) analyzed th e pelvic
m etapterygia of Polypterus sp. and fo u n d a m ean Y oung’s m odulus of 17.6±7.8
G Pa and in a re c en t pu b licatio n , H o rto n & Sum m ers (2009) d eterm in ed a Y oung’s
m odulus for acellular bone in Myoxocephalus polyacanthocephalus (Pallas) of
6.4810.31 GPa. D ata concerning m aterial p ro p erties of skull bones in fishes is
lacking. Besides th a t, th e re is considerable v ariatio n in featu res like b o n e density,
cortical thickness, d irectio n of loading, m icro stru ctu re, etc. b etw een taxa, bones
an d even sites w ith in bones. So even if th e m aterial p ro p erties of fish skull bones
w o uld have b een exam ined, th e re is n o g u aran ty th a t th e y w o u ld ap p ro x im ate th e
values of syngnathid skull bone pro p erties. For exam ple, P eterso n & D echow
(2003) fo u n d th a t th e elastic p ro p erties of som e cran ial bones in h u m an s are m ore
sim ilar to th o se m easured in long bones th a n to th o se of th e m andible. H ence, an
a rb itra ry Y oung’s m odulus of 20 G Pa w as chosen, w h ich is th e m ean of a range of
m aterial p ro p erties of various bones in birds an d e u th e ria n mammals: 6.7-34.1
G Pa (C urrey, 1999). A ccording to th e sam e line of reasoning, th e m odels w ere
assigned w ith a P oisson’s ra tio of 0.3, based o n various bones in actinopterygians,
birds an d e u th e ria n mammals.
C o n strain ts w ere applied to a few nodes a t th e occipital jo in ts of th e skulls and
su ctio n pressure was sim ulated by loading th e suspensoria and ro stral p a rt of th e
n e u ro c ran iu m w ith a pressure of 20 kP a o rie n te d inw ards, to w ard s th e central,
lo n g itu d in al axis of th e cylindrical sn o u t (Fig. 2.5). Because o u r goal was to
C h a p t e r 2 - M a te ria l & m e t h o d s
29
com pare th e relative m agnitudes and p a tte rn s of stress am ong th e m odels (not
absolute stress m agnitudes), w e applied th e sam e pressure to th e sn o u ts of all
m odels based on m easurem ents fro m w ith in th e buccal cavity of tru m p e tfish (20
kPa, H uskey & Q u in te ro , 2006). V on M ises stress w as used to describe th e effect
of th e m echanical loading on th e models. The V on M ises crite rio n is a form ula
for com bining th e prin cip al stresses (acting in th e x, y, an d z directions) at a given
p o in t in to an equivalent stress, or V on M ises Stress, w h ich is th e n com pared to
th e yield stress of th e m aterial. If th e V on M ises stress exceeds th e yield stress,
th e n th e m aterial is considered to be at th e failure condition. H ence, V on M ises
stress is a good p re d ic to r of failure u n d er ductile frac tu re (N alla et al., 2003). W e
p red icted stress d istrib u tio n s in th e skull m odels by m eans of lin ear fin ite
elem en t analysis.
2.3.
T erminology
A ccording to Balon (1975), th e larval p erio d of fishes begins w ith th e sw itch to
exogenous feeding an d lasts u n til th e m edian fin fold is d ifferen tiated . The
juvenile perio d is th e n defined as startin g w ith com plete d ifferen tia tio n of fins
an d ending w ith th e o n set of gam ete m a tu ra tio n (Balon, 1975). G iven th e fact
th a t th e d ifferen tia tio n of th e fins precedes th e o n set of exogenous feeding in
syngnathids (Azzarello, 1990 an d p erso n al observation), a larval perio d is m issing
an d new ly released young are considered juveniles. M oreover, new b o rn s resem ble
m in ia tu re adults in form (see ch ap ter 3.1 an d 6.1), as is co n firm ed for
Hippocampus kuda (C hoo & Liew, 2006). In th e la tte r species, g ro w th allom etries
a fte r release from th e brood p o u c h reflect typical teleo stean juvenile g ro w th and
n o t larval g ro w th (C hoo & Liew, 2006). T herefore, th e te rm juvenile is used
in stead of larva here. A lth o u g h m any au th o rs refer to a larval stage an d even a
m etam o rp h o sis3 inside th e pouch, th e la tte r c an n o t be excluded w ith th e data
3 d e f i n e d b y L a w r e n c e (2 0 0 5 ) a s t h e t r a n s f o r m a t io n o f o n e s t r u c t u r e i n t o a n o t h e r , m o s t c o m m o n l y
r e f e r r in g t o t h e r a d ic a l c h a n g e i n f o r m a n d s t r u c t u r e u n d e r g o n e b y s o m e a n im a ls b e t w e e n e m b r y o
a n d j u v e n i l e s ta g e
30
C h a p t e r 2 - M a te ria l & m e t h o d s
p re sen te d in th is thesis (e.g. C hoo & Liew, 2006; F orsgren & Lowe, 2006).
T h ro u g h o u t th is dissertation, prey distance refers to th e distance b etw een th e
prey an d th e sn o u t tip a t th e o n set of prey cap tu re, hence before h ead ro ta tio n
(distance 6-8 on Fig. 5.2). Prey cap tu re tim e is defined as th e tim e necessary to
suck in th e prey.
T he term inology of all m uscles follow s th a t of W in te rb o tto m (1974). For th e
m ajor p a rt of th e osteological co m ponents, th e term in o lo g y pro p o sed by
L ekander (1949), H a rrin g to n (1955) an d A rratia & Schultze (1990) is follow ed.
N evertheless, in som e occasions a n o th e r nam e, w h ich seem ed m ore co rrect or
apt, is applied. T he use of th o se term s is acco u n ted fo r below.
T he d en tary and an g u lo articu lar bones in Syngnathus rostellatus and H. capensis
could be a fusio n of several bones. In m ost teleosts, th e d en tary bone com prises
th e p erich o n d ral m entom eckelian, th e derm al splenial an d th e derm al d en tary
bones, and should th u s be nam ed ‘den to -sp len io -m en to m eck eliu m ’ according to
th e no m en clatu re of L ekander (1949). T he a n g u lo articu lar b o n e is th e fu sio n of
th e p erich o n d ral a rticu lar bone, th e derm al splenial bones and th e derm al angular
bone; th e ‘angulo-splenio-articular’. H ow ever, w h e th e r th is is also th e case for
syngnathids is n o t certain, because th e absence of th e p reo p ercu lo -m an d ib u lar
canal m ay indicate th e absence of th e splenial bones. In th e c u rre n t deficiency of
conclusive o n to g en etic evidence to elucidate th is, th e term s ‘d en tary b o n e’ and
‘an g u lo articu lar b o n e’ are used here.
K indred (1924) suggested th e re is a p tery g o id b o n e in Syngnathus fuscus Storer,
w h ich w o u ld be a fusio n of th e ectopterygoid an d th e end o p tery g o id bones.
A ccording to K adam (1961) th e ecto p tery g o id an d th e en dopterygoid bones ossify
separately in Nerophis (species n o t stated), Syngnathus serratus (T em m inck &
Schlegel) an d Hippocampus (species n o t stated). B ergert & W a in w rig h t (1997)
fo u n d b o th an ectopterygoid an d an endopterygoid b o n e in S. floridae, and solely
an ectopterygoid bone in H. erectus. In th e species stu d ied h ere w e fo u n d no
in d ications of an endopterygoid bone. As Kadam (1961) co rrectly p o in te d o u t, th e
bone th a t K indred (1924) describes as th e pterygoid b o n e consists of tw o separate
C h a p t e r 2 - M a te ria l & m e t h o d s
31
elem ents and one of th e m is indeed th e derm al ecto p tery g o id bone. H ow ever, he
did n o t n o tice th a t th e bone he called th e en dopterygoid b o n e is p erich o n d ral,
an d th e re fo re hom ologous to a m etapterygoid bone. B ergert & W a in w rig h t
(3.997) follow ed K indred (1924) in id en tify in g th e m etapterygoid b o n e of S.
floridae as th e endopterygoid bone. In addition, th e y did n o t m e n tio n th e
presence of a sim ilar bone in H. erectus. S w in n e rto n (1902) sta te d th a t in
Gasterosteus aculeatus th e pterygoid b o n e takes up th e p o sitio n of b o th
endopterygoid an d ectopterygoid bones, how ever, only one cen tre of o ssification
is found. A ccording to de Beer (1937) G. aculeatus possesses b o th an ectopterygoid
an d an endopterygoid bone, fused to fo rm w h a t he calls a pterygoid bone. W e
could n o t exclude a fusio n b e tw ee n th e ecto- an d endo p tery g o id b o n e in th e
sy ngnathid species stu d ied here. H ow ever, based o n its to p ography, v e n tro la te ral
to th e au to p alatin e an d th e m etap tery g o id bone, th e sy n g n ath id p tery g o id b o n e is
considered hom ologous to th e ecto p tery g o id bone.
As B ranch (1966) m en tioned, th e hom ology of th e circu m o rb ital bones has been
unclear. K indred (1924) an d de Beer (1937) defin ed th e m etapterygoid b o n e of S.
fuscus as “th e in tram em b ran o u s o ssification dorsal to th e quadrate, ro stral to th e
sym plectic an d excluded from c o n tac t w ith th e m etap tery g o id process of th e
p alato q u ad rate by th e pterygoid”. H ow ever, K adam (1961), B ranch (1966) and
P a tte rso n (1977) p o in te d o u t th a t th is is n o t th e m etap tery g o id bone, b u t th e
lacrim al bone. Jungersen (1910) id en tified th e circu m o rb ital bones as th e
p o sterio r and a n terio r p re o rb ita l bones in Syngnathus typhle L. (w hich he called
Syphonostoma typhle) because of th e ir p o sitio n lateral of th e a d d u cto r m andibulae
m uscle. G regory (1933) sta te d th a t Phyllopteryx possesses “a row of a n to rb ita l
plates on th e side of th e oral tu b e ”, w h ich he labelled as tw o m etapterygoid
bones. As previously m en tioned , K indred (1924) an d de Beer (1937) m a in tain ed
th a t th e lacrim al bone in S. fuscus is th e m etapterygoid bone, a lth o u g h th ey
correctly p o in te d o u t th a t th e second in frao rb ital b o n e is a circu m o rb ital bone.
K adam (1961) described th e tw o bones of th e su b o rb ital chain in Nerophis as an
a n te rio r p re o rb ita l bone an d a p o sterio r su b o rb ital b o n e an d he rem arked th a t in
Syngnathus and Hippocampus th e re are tw o p re o rb ita l bones. The use of th e term s
p re o rb ita l an d su b o rb ital bones should be avoided as th e y only in d icate th e
p o sitio n of th ese bones relative to th e o rb it b u t do n o t say an y th in g a b o u t th e ir
32
C h a p t e r 2 - M a te ria l & m e t h o d s
hom ology (Daget, 1964). T herefo re w e use th e term s a n to rb ita l b o n e and
in frao rb ital bones, as e.g. in L ekander (1949), N elson (1969) an d S chultze (2008).
T he te rm prevom eral bone is used by m any au th o rs (e.g. G regory, 1933; de Beer,
1937; H a rrin g to n , 1955) in stead of vom eral bone. H ere, th e la tte r is used, because
of th e hom ology w ith th e vom eral bone in sarcopterygians (Schultze, 2008). The
use of th e term s p arietal and p o stp arietal bones as su b stitu tes fo r th e fro n tal and
p arie tal bones is also based on th e term in o lo g y of S chultze (2008), w h o provides
evidence for absence of th e m am m alian fro n tal bones in actinopterygians.
33
3
3.1.
Morphological description
M o r p h o lo g y o f th e c h o n d r o - an d osteo cra n iu m
M o d i f i e d fr o m :
L e y s e n H , J o u k P , B r u n a in M , C h r i s t i a e n s J,
A d r ia e n s D .
C r a n ia l a r c h it e c t u r e o f t u b e - s n o u t e d
G a s t e r o s t e i f o r m e s (S y n g n a t h u s r o s te l la t u s a n d
H i p p o c a m p u s capen sis)
J o u r n a l o f M o r p h o l o g y 2 0 1 0 , 2 7 1 :2 5 5 -2 7 0 .
Abstract
T he long sn o u t of pipefishes and seahorses (Syngnathidae, G asterosteiform es) is
form ed as an elo n g atio n of th e eth m o id region. This is in c o n tra st to m any o th e r
teleosts w ith elongate sn outs (e.g. b u tterfly fish es) in w h ich th e sn o u t is form ed as
an extension of th e jaws. Syngnathid fishes p e rfo rm very fast su ctio n feeding,
accom plished by p o w erfu l n eu ro cran ial elevation and hyoid depression. Clearly,
su ctio n th ro u g h a long an d n arro w tu b e an d its hydrodynam ic im plications can
be expected to require certain ad ap tatio n s in th e cranium , especially in
m usculoskeletal elem ents of th e feeding apparatus. N o t m u ch is k n o w n ab o u t
w h ich skeletal elem ents actually su p p o rt th e sn o u t and w h a t th e effect of
34
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
elo n g atio n is on related stru ctu res. H ere, w e give a d etailed m orphological
descrip tio n of th e cartilaginous and b o n y feeding ap p aratu s in b o th juvenile and
ad u lt Syngnathus rostellatus and Hippocampus capensis. O u r results are com pared
to previous m orphological studies of a generalized teleo st, Gasterosteus aculeatus.
W e fo u n d th a t th e e th m o id region is elo n g ated early d u rin g developm ent, w ith
th e eth m o id plate, th e hyosym plectic and th e basihyal cartilage being ex ten d ed in
th e chondrocranium . In th e juveniles of b o th species alm ost all bones are form ed,
a lth o u g h only as a very th in layer. T he elo n g atio n of th e vom eral, m esethm oid,
quadrate, m etapterygoid, sym plectic an d p reo p ercu lar bones is already p resent.
P robably because of th e long an d specialized p a re n ta l care, w h ich enables th e
release of advanced developm ental stages from th e bro o d in g pouch, m orphology
of th e feeding ap p aratu s of juveniles is already very sim ilar to th a t of adults. W e
describe m orphological features related to sn o u t elo n g atio n th a t may be
considered ad ap tatio n s for su ctio n feeding; e.g. th e p ecu liar shape of th e
in terh y al bone and its saddle-shaped a rtic u la tio n w ith th e p o sterio r ceratohyal
bone m ight aid in explosive hyoid depression by reducing th e risk of hyoid
dislocation.
3.1.1. Introduction
U nlike o th e r lo n g-snouted teleosts (e.g., b u tterfly fish es, C haeto d o n tid ae), th e
tu b u la r sn o u t of syngnathids is n o t form ed by th e ex ten sio n of th e jaws, b u t by
an elo ngation of th e region betw een th e au to p alatin e b o n e and th e lateral
e th m o id bone, nam ely th e eth m o id region. This sn o u t w ith tin y te rm in a l jaws
enables pipefishes and seahorses to p e rfo rm fast and p o w erfu l su ctio n feeding.
T hey approach th e ir prey from below and a rapid n eu ro cran ial elevation
p o sitions th e m o u th close to th e prey. N ex t an explosive expansion of th e sn o u t
follow ed by low er jaw depression, causes w a ter to flow in to th e m o u th ap ertu re
(C h ap ter 5.1; M u ller & Osse, 1984; M u ller, 1987; de L ussanet & M u ller, 2007).
S uction feeding in pipefishes an d seahorses is th e fastest ever recorded in
teleosts. M u ller and Osse (1984) fo u n d th a t Entelurus aequoreus cap tu red its prey
in 5 ms, w hile B ergert an d W a in w rig h t (1997) recorded a tim e of 5.8 ms for
Hippocampus erectus an d 7.9 ms fo r Syngnathus floridae. De L ussanet an d M u ller
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
35
(2007) recorded cap tu re tim es of 6-8 ms fo r S. acus an d Roos et al. (chapter 5.1)
recorded 5.77 ms for H. reidi. It was recen tly discovered th a t new b o rn s are even
faster (Van W assenbergh et al., 2009). H ow ever, having a long an d n arro w sn o u t
is n o t w ith o u t hydrodynam ic costs. For exam ple, by increasing th e le n g th of th e
sn o u t, th e m o m en t of in e rtia increases. Secondly, it im plies th a t a large difference
in pressure b etw een th e buccal cavity and th e su rro u n d in g w a ter m u st be created
(Poiseuille’s law). A nd finally, as th e u p p er an d low er jaws closing th e m o u th
ap ertu re are m inute, th e prey size is constrained. H ence, th e hydrodynam ic
im plications of su ctio n feeding th ro u g h a long, n arro w tu b e can be expected to
rely
on
special
adap tatio n s
in
th e
feeding
ap p aratu s,
p articu larly
of
m usculoskeletal com ponents form ing an d acting u p o n th e jaws an d eth m o id
region.
To u n d e rstan d to w h a t degree stru c tu ra l specializations of th e tu b u la r sn o u t can
be related to th is highly p e rfo rm a n t su ctio n feeding, a detailed ex am in atio n of
th e m orphology is needed. T hus far, studies dealing w ith sy n g n ath id m orphology
are scarce or lack great detail (M cM urrich, 1883; de Beer, 1937; Kadam , 1958;
1961; B ranch, 1966). To fill th e gap in c u rre n t know ledge, th is study focuses on
th e detailed anatom y of th e cranial skeletal system of Syngnathus rostellatus
(N ilsson’s pipefish)
an d
Hippocampus
capensis
(Knysna
seahorse).
Special
a tte n tio n is paid to th e sn o u t m orphology to u n d e rstan d w h ich skeletal elem ents
are in fact elongated an d w h a t effect th is elo n g atio n m ay have o n th e cranial
arch itectu re. T he study of juveniles is req u ired fo r a b e tte r co m p reh en sio n of
interspecific differences, as w ell as th e detailed anatom ical n a tu re of sn o u t
elongation. The highly derived sy n g n ath id m orphology is com pared to th a t of a
generalized teleost, nam ely Gasterosteus aculeatus (three-spined stickleback), based
on th e study of A nker (1974).
3.1.2. Brief material & m e t h o d s
T he fo u r adults and five juveniles of Syngnathus rostellatus (126.2, 111.1, 97.9,
105.0, 11.0, 14.5, 11.3, 11.4 an d 13.1 m m SL respectively), th re e adults and fo u r
juveniles of Hippocampus capensis (96.1, 97.6, 99.0, 13.3, 13.6, 12.8 an d 14.0 m m SL
respectively) and tw o adults and one juvenile of H. reidi (117.2, 113.5 an(l 7-0 m m
36
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
SL respectively) stu d ied are p re sen te d in Table 2.1 and 2.2. A d u lt as w ell as
juvenile specim ens of all species (w ith ex cep tio n of a juvenile H. reidi) w ere
cleared an d stain ed according to th e p ro to co l of Taylor an d V an D yke (1985). In
th e juveniles, bone stain in g was n o t very clear, so serial histological cross sections
w ere used, w h ich also enabled m ore precise d e tec tio n of th e skeletal elem ents.
G raphical 3 D -reconstructions of th e ch o n d ro cran iu m of b o th S. rostellatus and H.
capensis w ere generated. T he specim en of S. rostellatus (13.1 m m SL) used fo r serial
sectioning show s th e hyoid in a restin g p o sitio n , w hereas th a t of H. capensis (12.8
m m SL) h ad its hyoid depressed. For m ore in fo rm atio n o n specim ens and
protocols I refer to ch ap ter 2.
3.I.3. Results
A. Ju ven ile cran ium
Syngnathus rostellatus
T he cartilaginous n e u ro c ran iu m consists of tw o p a rts th a t are sep arated by th e
eyes: th e ro stral eth m o id an d th e caudal otic capsule (Fig. 3.1). The eth m o id p late
is long an d n arro w b u t becom es w id er rostrally w h ere it lies v en tral to th e ro stral
cartilage (Fig. 3.iA,B). M o re caudally th e eth m o id p late bears a vertical ridge, i.e.
th e in tern asal septum , con n ected to th e o rb ito n asal lam inae, w h ich enclose th e
o rb ito n asal foram ina (Fig. 3.iA,B). A lth o u g h th e eth m o id p late an d th e sep tu m
are firm ly fixed, histological differences am ong th e cartilaginous elem ents
suggests th a t th e in tern asal sep tu m is n o t form ed as an o u tg ro w th of th e eth m o id
plate. T here is a clear difference in th e size, shape and o rg an izatio n of th e ir
chondrocytes (Fig. 3.2D). The eth m o id p late is c o n tin u o u s w ith th e trab ecu la
com m unis, th a t lies m edial to th e o rb its (Fig. 3.iB,C). V entrally th e otic capsule is
provided w ith
an a rtic u la tio n
facet fo r th e
h y o m an d ib u lar p a rt of th e
hyosym plectic cartilage. M eckel’s cartilage bears a v en tral re tro a rtic u la r process
an d articu lates caudally w ith th e p tery g o q u ad rate p a rt of th e p alato q u ad rate
cartilage, w h ich is roughly L-shaped (Fig. 3.1A). The p alatin e p art, w h ich is
com pletely separated from th e p tery g o q u ad rate p a rt, lies laterally to th e eth m o id
plate (Fig. 3.1A). The largest cartilage elem en t of th e sp lan ch n o cran iu m is th e
hyosym plectic cartilage, w h ich consists of a long, h o riz o n ta l sym plectic p art, and
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
37
a sh o rte r oblique h y o m andibular p a rt (Fig. 3. i A,C). A t th e ven tro cau d al m argin of
th e hyosym plectic cartilage lies th e in terh y al cartilage, artic u la tin g v en trally w ith
th e ceratohyal cartilage (Fig. 3.1C). M edial to th e tw o ceratohyal cartilages lies
one long basihyal and tw o sh o rte r hypohyal cartilages (Fig. 3.1C).
Juveniles of S. rostellatus show th e o n set of ossification in m ost places, how ever,
only a very th in layer of bone was observed (Fig. 3.2). V en tral to th e eth m o id p late
th e derm al p arasphenoid bone has already form ed. This very long b o n e ru n s from
th e e th m o id region up to th e p o sterio r p a rt of th e otic region (Fig. 3.2D,F).
F o rm atio n of th e m esethm oid b o n e begins dorsal to th e e th m o id p late and
a ro u n d th e in tern asal sep tu m (Fig. 3.2C). A th in b o n y sh eet a t th e v en tral end of
th e o rbitonasal lam inae is th e p recu rso r of th e lateral eth m o id bone. A ro u n d th e
m ain p a rt of M eckel’s cartilage, th e d en tary b o n e is form ed (w h eth er th is bone
includes th e m entom eckelian an d splenial bones is u n c ertain due to th e absence
of canals, see ch ap ter 2.3 for a discussion o n th e term inology; Fig. 3.2B). This
bone bears a large v en tral ridge and caudally encloses th e an g u lo articu lar bone
(this could be fused w ith th e splenial bones, b u t again n o canals w ere observed,
see ch ap ter 2.3 for a discussion o n th e term inology), w h ich is still p o orly
developed an d only p re sen t on th e lateral side of M eckel’s cartilage (Fig. 3.2B).
T he re tro a rtic u la r bone is visible as a sm all ossification of th e v en tro cau d al p a rt
of th e M eckel’s cartilage (Fig. 3.2B). In th e u p p er jaw, b o th m axillary and
prem axillary bones have appeared and are already fairly w ell developed. The
form er articu lates w ith th e ro stral cartilage dorsally. T he au to p alatin e bone is
p re sen t b u t does n o t bear a clear m axillary or vom eral a rtic u la tio n facet yet (Fig.
3.2A). V en tral to th e p alato q u ad rate cartilage th e ecto p tery g o id b o n e is form ed
(Fig. 3.2A). This derm al bone show s a sm all h o riz o n ta l p a rt and a longer vertical
one th a t m eets th e dorsal process of th e q u ad rate bone. A t th e dorsal edge of th e
p alato q u ad rate cartilage, th e sm all m etapterygoid b o n e arises (Fig. 3.2C). The
q uadrate bone bears a dorsal process, as w ell as a v en tro m ed ial an d v e n tro lateral
w ing. M ore caudally th ese w ings enclose th e cartilaginous hyosym plectic an d th e
sym plectic bone (Fig. 3.2C,E). T he sym plectic b o n e consists of b o th th e
ossification a ro u n d th e ro stral p a rt of th e hyosym plectic cartilage and a dorsal
crest on to p of th e p erich o n d ral p a rt (Fig. 3.2E). T he h y o m an d ib u lar b o n e is
form ed
caudally
a rticu latio n s
w ith
a ro u n d
th e
th e
hyosym plectic
n e u ro c ran iu m
and
cartilage
o p ercu lar
an d
bears
dorsal
bone
th a t
rem ain
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
38
cartilaginous (Fig. 3.2F). The p reo p ercu lar b o n e consists of b o th a sh o rt an d long
process. T he long one covers th e q u ad rate and sym plectic bones ro strally and is
also provided w ith a large lateral process (Fig. 3.2E). Its sh o rte r oblique b ar covers
th e h y o m andibular bone caudally (Fig. 3.2F). All o th e r elem ents of th e hyoid arch,
i.e., basihyal, hypohyal, ceratohyal an d in terh y al cartilages, show th e presence of a
very th in sheet of bone (Fig. 3.2E,F). T he hypohyal bones bear a v e n tro lateral and
a v entrom edial process, w h ich su rro u n d th e ceratohyal bones v en trally (Fig. 3.2E).
A n te rio r and p o sterio r ceratohyal bones are h ard to d istin g u ish fro m each o th e r
a t th is stage (Fig. 3.2E). W ith in th e te n d o n of th e sternohyoideus m uscle, th e
urohyal bone has also arisen. The o p ercu lar bone is a th in b u t fairly large bony
sheet, bearing a lateral process and a rticu latin g w ith th e h y o m an d ib u lar bone
medially. N one of th e o th e r opercu lar bones (in tero p ercu lar, su b o p ercu lar and
su p rap reo p ercu lar bones) an d n e ith e r th e b ranchiostegal rays are p re sen t yet.
Hippocampus capensis
F or th e c h o n d ro cran iu m of H. capensis (Fig. 3.3), w e re p o rt only th o se featu res
w h ich differ from S. rostellatus.
T he eth m o id plate of th e cartilaginous n e u ro c ran iu m in H. capensis, is sh o rte r and
rostrally n arro w er th a n th a t of S. rostellatus (Fig. 3-3A,B). C audal to th e olfactory
organs, th e eth m o id plate w idens and m eets th e o rb ito n asal lam inae (Fig. 3-3A,B,
3.4D). It is also c o n tin u o u s w ith th e trab ecu la com m unis, b u t in th e seahorse th e
la tte r is m uch sh o rte r an d m ore ro b u st (Fig. 3.3C). The otic capsule has a d istin ct
p o sitio n com pared to th a t in S. rostellatus, nam ely dorsocaudal to th e orbits.
H ence, it does n o t lie on th e sam e level as th e eth m o id plate, b u t a t an angle to
th e la tte r (otic capsule tilte d a b o u t 340 up; Fig. 3.3A). A t th e v en tral surface of th e
otic
capsule,
th e
a rtic u la tio n
facet
of th e
h y o m an d ib u lar
p a rt
of
th e
hyosym plectic cartilage is m uch m ore p ro m in e n t an d it is laterally flan k ed by a
sp h en o -p tero tic ridge (Fig. 3.3A). M eckel’s cartilage is m ore tap ered rostrally
com pared to th a t of S. rostellatus (Fig. 3.3A). The sym plectic p a rt of th e
hyosym plectic cartilage is som ew hat sh o rte r in H. capensis. The hy o m an d ib u lar
p art, how ever, is longer an d m ore vertically o rie n ta te d com pared to th a t of th e
p ipefish (Fig. 3.3A). In th e seahorse, th e sh o rte r basihyal cartilage lies in fro n t of
th e ceratohyal cartilages, w h ich m ay be due to th e hyoid being re tra c ted (Fig.
3.3A,C).
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
39
A lm ost all bones are p re sen t in th e juvenile H. capensis studied, except fo r th e
circu m o rb ital bones (Fig. 3.4). The vom eral bone lies v en tral to th e eth m o id p late
an d becom es covered by th e parasp h en o id bone m ore caudally (Fig. 34A,B,D).
T he la tte r bears tw o ra th e r large lateral w ings th a t reach th e v en tral surface of
th e otic capsule. The d en tary bone ro strally bears a sm all lateral process and has a
w ell developed co ronoid process. The a n g u lo articu lar bone and re tro a rtic u la r
bone are p ro m in e n t an d th e re is a ligam entous c o n n ectio n b etw een th e
re tro a rtic u la r bone an d th e slender in te ro p e rc u lar b o n e th a t co n tin u es to ru n up
to th e p o sterio r ceratohyal bon e (Fig. 3.4C). The dorsal crest of th e sym plectic
bone is larger in H. capensis com pared to S. rostellatus (Fig. 3.4E). T here is a large
spine on th e lateral surface of th e p reo p ercu lar b o n e an d th e ascending b ar is
o rie n te d vertically instead of obliquely as in th e p ip efish (Fig. 3.4F). The bony
sheets a ro u n d th e hypohyal and ceratohyal cartilages are w ell developed (Fig.
3.4F). In addition, th e a n te rio r an d p o sterio r ceratohyal bones are d istin c t from
each other. In th e seahorse, th e urohyal b o n e is m u ch sh o rter. The o p ercu lar bone
has a convex shape an d bears a p ro m in e n t lateral process. Also th e su b o p ercu lar
bone and branchiostegal rays are fairly w ell developed in juvenile H. capensis.
B. A d u lt c ra n iu m
Syngnathus rostellatus
T he m ost distinctive ch aracter of th e skull of Syngnathus rostellatus is th e highly
ex tended tu b e sn o u t (Fig. 3.5). It is form ed by th e elo n g atio n of th e vom eral,
m esethm oid an d th e circu m o rb ital bones of th e n eu ro cran iu m an d of th e
quadrate, m etapterygoid, sym plectic, p reo p ercu lar an d in te ro p e rc u lar bones of
th e sp lan ch n o cran iu m (Fig. 3.5A).
B oth th e m axillary and prem axillary bones are relatively sm all an d to o th less (Fig.
3-5A,B,D,E). T he m axillary bon e bears tw o cartilaginous processes dorsally: a
ro stral prem axillary one and a caudal one fo r th e a rtic u la tio n w ith th e vom eral
bone. Below th e la tte r process th e re is also a cartilaginous a rtic u la tio n surface for
th e au to p alatin e bone. The ro u n d ro stral cartilage is situ a ted m ediocaudal to th e
m axillary bone an d dorsal to th e vom eral bone. V entrally, th e m axillary b o n e is
trian g u larly shaped, covering th e co ro n o id process of th e d en tary b o n e to w h ich
it is ligam entously connected. T he slender prem axillary b o n e is rostrocaudally
40
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
fla tte n e d and tapers ventrally. It is provided w ith a dorsocaudal cartilaginous
a rtic u la tio n head for th e m axillary bone.
T he vom eral bone is a long and n arro w b o n e th a t broadens rostrally, form ing an
a rtic u la tio n w ith th e au to p alatin e bone laterally an d th e m axillary bone rostrally
(Fig. 3-5A,B,D,E). T he h in d p a rt of th e vom eral b o n e reaches th e lateral eth m o id
bones and is covered dorsally by th e m esethm oid bone. M o re caudally, it is
w edged in a fissure of th e p arasph en o id bone. The m esethm oid b o n e covers m ore
th a n half th e le n g th of th e sn o u t and stretch es o u t caudally, up to th e p arietal
bones (Fig. 3-5A,B). T he lateral e th m o id bone is a slim bone th a t separates th e
nasal opening from th e orb its (Fig. 3-5A,B).
T he p arasphenoid bone is p o sitio n ed rostrally b etw een th e dorsal m esethm oid
bone and th e v en tral vom eral bon e (Fig. 3.5A). It bears tw o lateral w ings b eh in d
th e orb its an d fits in to a w edge of th e basioccipital b o n e caudally. In m ost
specim ens studied of S. rostellatus only tw o circu m o rb ital bones are presen t,
w h ich seem to be hom ologous to an a n to rb ito lacrim al an d a second in frao rb ital
bone (C hapter 2.3). O nly one specim en has ju st one b o n e on its rig h t side. In th e
individuals w ith tw o circu m o rb ital bones, th e large an to rb ito lacrim al bone
caudally reaches th e fro n t end of th e nasal opening, and covers a large p a rt of th e
q uadrate bone (Fig. 3-5A,B). V entrally, th e a n to rb ito lacrim al bone show s one or
several sm all in d en tatio n s. The second in frao rb ital b o n e is m u ch sm aller and
borders th e v en tral side of th e nasal opening, as w ell as th e a n terio r side of th e
orb its (Fig. 3-5A,B,C).
T he large den tary bone of th e low er jaw has a w ell developed co ro n o id process
(Fig. 3-5A,C,D). Inside a cavity of th e d en tary bone, th e sm aller an g u lo articu lar
bone fits, w h ich bears a distinctive cartilaginous a rtic u la tio n w ith th e q u ad rate
bone caudally (Fig. 3-5A,C,D). The re tro a rtic u la r bone is very small, w ith a strong
m an d ib u lo -in tero p ercu lar ligam ent co n n ectin g it to th e in te ro p e rc u lar bone (Fig.
3.5A,C,D).
In th e a d u lt stage, th e a u to p alatin e b o n e carries a p ro m in e n t cartilaginous
m axillary process, a sm aller a rtic u la tio n condyle fo r th e vom eral bone an d a
slender cartilaginous process caudally (Fig. 3-5A,B,D,E). T here is n o separate
derm o p alatin e bone and as in m ost ex tan t teleosts, it is probably fused to th e
au to p alatin e bone (A rratia & Schultze, 1991). The ectopterygoid b o n e is roughly
trian g u larly shaped, w ith a vertical p a rt ru n n in g along th e ascending process of
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
41
th e quadrate bone an d a h o riz o n ta l p a rt th a t is covered dorsally by th e vom eral
bone (Fig. 3-5A,B,D,E). This h o riz o n ta l p a rt show s a gap in to w h ich th e
cartilaginous process of th e au to p alatin e bone fits, w ith a firm c o n n ectio n
lin k in g b o th bones. Lateral to th e vom eral bone an d b eh in d th e ectopterygoid
bone lies th e m etapterygoid bo n e w h ich tap ers caudally and is covered by th e
up p er ro stra l m argin of th e lacrim al b o n e (Fig. 3-5A,B). The q u ad rate bone, a long
p e rich o n d ral bone th a t stretch es o u t caudally, is m ostly covered by th e
m etapterygoid bone rostrally an d th e tw o circu m o rb ital bones caudally (Fig.
3-5A,B,C).
T he h y o m andibular bone articu lates dorsally by a double condyle w ith th e
sp h en o tic bone an d w ith th e cartilaginous surface b etw een th e sphenotic, th e
p ro o tic an d th e p te ro tic bones and it also bears a dorsocaudal o p ercu lar process.
T he sym plectic bone is alm ost com pletely covered by th e p reo p ercu lar and
circu m o rb ital bones and form s th e v en tral b o rd er of th e o rbits (Fig. 3.5A). It
bifu rcates rostrally in to tw o processes: a low er h o riz o n ta l p a rt th a t joins th e
q uadrate bone, and a m ore dorsal oblique crest lying b e h in d th e u p p er m argin of
th e second in frao rb ital bone.
T he long h o riz o n ta l process of th e p reo p ercu lar b o n e overlaps w ith th e q u ad rate
bone rostrally w here it tap ers (Fig. 3-5A,C). M edially th e p reo p ercu lar b o n e has
tw o ridges: one su p p o rtin g th e sym plectic b o n e an d one fo r in sertio n of th e
levator arcus p a latin i m uscle, w h ich co n tin u es to ru n along th is ridge and m ore
caudally in a groove of th e h y o m an d ib u lar bone. V en trally th e p reo p ercu lar bone
has a cartilaginous d iffe re n tia tio n w here th e cartilaginous head of th e in terh y al
bone articulates. T here is no a rtic u la tio n b etw een th e in terh y al b o n e and th e
h y o m andibular bone. T he in te ro p e rc u lar bone is covered by th e p reo p ercu lar
bone an d th e quadrate bone, w ith an in tero p ercu lo -h y o id lig am en t co n n ectin g it
to th e p o sterio r ceratohyal bone caudally (Fig. 3.5C). The in terh y al bone, w h ich is
s to u t and small, is ventrally provided w ith a very firm , saddle-shaped jo in t fo r th e
p o sterio r ceratohyal bone (Fig. 3-5A,C). T he p o sterio r ceratohyal b o n e has a small
lateral process, close to th e in terh y al artic u la tio n (Fig. 3.5C). O n to th is process,
th e in teroperculo-hyoid ligam en t attach es ro strally an d a t its caudal base, th e tw o
branchiostegal rays are connected. T here is a firm in te rd ig ita tio n b etw een th e
p o sterio r an d a n te rio r ceratohyal bone. D istally, th e re is a sm all trian g u larly
shaped gap betw een th e le ft and rig h t a n terio r ceratohyal bones, ju st below th e
42
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
very firm cartilaginous symphysis. T he a n terio r ceratohyal bones are co n n ected
to th e urohyal bone by a p aired ceratohyal-urohyal lig am en t (Fig. 3.5C). The
hypohyal bone is a sm all elem ent th a t is firm ly co n n ected to th e m edial face of
th e a n terio r ceratohyal bone. M edial to th e a n te rio r ceratohyal bones and
covered by th e o th e r elem ents of th e hyoid lies th e slender basihyal bone, w h ich
rem ains cartilaginous rostrally. T he urohyal bone is a fairly long an d slender bone
th a t broadens som ew hat rostrally w h ere th e ceratohyal-urohyal ligam ents a tta c h
(Fig. 3.5C).
T he opercular bone is large an d has a convex lateral surface (Fig. 3.5A,B,C). T here
is ju st a tin y gili slit close to th e cleithrum . The su p rap reo p ercu lar b o n e is a sm all
bone lying dorsorostrally to th e o p ercu lar b o n e (Fig. 3-5A,B). The su b opercular
bone is sickle-shaped, covered by th e v en tral edge of th e o p ercu lar bone. T he tw o
branchiostegal rays, w h ich are long and slender, join th e caudal m argin of th e
o percular bone an d reach up to th e gili slit (Fig. 3-5A,C). T here are n o canals for
th e lateral line system p re sen t in any of th e bones studied.
Hippocampus capensis
T he prem axillary and m axillary bones look very sim ilar to th o se in S. rostellatus
(Fig. 3-6A,B,D,E). In H. capensis, how ever, th ey are m ore heavily b u ilt and th e
m axillary b one show s a m ore p ro m in e n t convex curve w h e n view ed rostrally. The
ro stral cartilage has a m ore elliptical shape in stead of being ro u n d (Fig. 3.6E).
T he dorsal p a rt of th e tu b e sn o u t consists of th e vom eral b o n e an d th e
m esethm oid bone (Fig. 3.6B). The la tte r has a slightly b ifu rca te d ro stral end and
covers approxim ately h alf th e sn o u t length. T he lateral e th m o id b o n e is very
d istin c t and has q u ite a large lateral process (Fig. 3.6A,B).
T he parasp h en o id bone stretch es v en trally along th e n e u ro c ran iu m an d bends
som ew hat upw ards in th e otic region (Fig. 3.6A). The n u m b e r of circu m o rb ital
bones in H. capensis is variable. In spite of th is variability, som e of th e m can be
considered as hom ologous (an to rb ital, lacrim al and d erm o sp h en o tic bones) as
in d icated by Schultze (2008). The d erm o sp h en o tic b o n e is co n sisten tly p re sen t in
all specim ens observed. V ariation w as fo u n d a t th e level of all o th e r circu m o rb ital
bones, including le ft rig h t v ariatio n (e.g. one specim en, 97.6 m m SL, has an
additio n al fo u rth circu m o rb ital b o n e o n its rig h t side of w h ich th e hom ology is
less obvious). A n o th e r specim en (99.0 m m SL) also seem ed to have a fused
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
43
a n to rb ito lacrim al bone, w hereas separate bones w ere observed in others. The
m ost com m on p a tte rn observed is w here th e a n to rb ita l bone is th e sm allest,
covering th e q uadrate bone and th e m etapterygoid b o n e (Fig. 3.6A,B). The
lacrim al bone also covers th e qu ad rate bone and is provided w ith a do rso ro stral
gap in to w h ich th e m etapterygoid b o n e fits (Fig. 3.6A,B). Finally, th e second
in frao rb ital bone covers th e quadrate, th e p reo p ercu lar and a large p a rt of th e
sym plectic bones (Fig. 3.6A,B,C). O f th e circu m o rb ital bones, th e m ost a n terio r
one covers th e n ext a t its caudal end, so th e a n to rb ita l bone covers th e lacrim al
bone, w h ich in tu r n covers th e second in frao rb ital bone.
T he d en tary bone is a sh o rt b u t solid b o n e (Fig. 3.6A,C,D). V entrocaudally, th e
an g u lo articu lar bone bears tw o v en tral processes in b etw een w h ich th e sm all
re tro a rtic u la r bone fits (Fig. 3.6A,C,D).
T he au to p alatin e bone is a ra th e r slender b o n e w hereas th e ectopterygoid b o n e is
som ew hat firm er com pared to th e one in S. rostellatus (Fig. 3.6A,B,D,E). The
m etapterygoid bone fits in to a gap of th e lacrim al b o n e caudally (Fig. 3.6A,B).
T he tw o n eu ro cran ial condyles of th e hy o m an d ib u lar b o n e are larger an d m ore
d ista n t from each o th e r in th e seahorse. In addition, th e h y o m an d ib u lar bone is
provided w ith a lateral process th a t is firm ly co n n ected to th e p reo p ercu lar bone.
T he oblique fork of th e sym plectic b o n e p re se n t in th e p ip efish is larger in th e
seahorse and form s a dorsal p late u p o n th e p e rich o n d ral part. O n ly th e caudal
p art, th a t borders th e v e n tro ro stra l m argin of th e orbits, is visible in a lateral view
(Fig. 3.6A). T he p reo p ercu lar bone has a sh o rt ascending process th a t form s th e
p o sterio r m argin of th e orb its (Fig. 3.6A,C). T he in te ro p e rc u lar bone is m uch
sh o rte r com pared to th a t in S. rostellatus (Fig. 3.6A,C). T he interhyal, th e a n terio r
an d p o sterio r ceratohyal, th e hypohyal and basihyal bones resem ble th o se of S.
rostellatus (Fig. 3.6A,C). T he urohyal bone, w h ic h is m ore ro b u st, has a ro stral
b ifu rc a tio n w ith b o th processes co n n ected to th e a n te rio r ceratohyal bone by
ceratohyal-urohyal ligam ents (Fig. 3.6C).
T he o percular bone is h igher an d has a less ro u n d ed dorsocaudal edge (Fig.
3-6A,B,C). T he su p rap reo p ercu lar b o n e is absent. The tw o very th in an d slender
branchiostegal rays reach up to th e caudal edge of th e o p ercu lar bone (Fig.
3-6A,B,C). As in S. rostellatus th e canals fo r th e lateral line are ab sen t in all bones
studied.
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
44
3.1.4. D is cu ssi on
A. A sp e c ts o f s n o u t e lo n g a tio n
As show n in Table 2.2, even th o u g h size ranges are sim ilar, th e re is a difference in
developm ental stage betw een th e juveniles of Syngnathus rostellatus (11.0-14.5 m m
SL) an d Hippocampus capensis (12.8-14.0 m m
SL).
D ue to
th e
d ifferen t
developm ental stages of o u r specim ens (S. rostellatus specim ens h ad n o t le ft th e
bro o d pouch), w e c a n n o t lin k th e m orphological differences b etw een th e tw o
species to differences in th e ir developm ental rate. H ow ever, th is poses no
pro b lem for th e m ain goal of th is study, i.e. to show th e re la tio n b etw een sn o u t
elo n g atio n and cranial m orphology in an early developm ental stage. T herefore,
w e w ill focus on th e differences b etw een b o th species, irrespective of th e ir
d ifferen t developm ental stages.
B oth S. rostellatus and H. capensis have an elo n g ated sn o u t com pared to
Gasterosteus aculeatus. This elo n g atio n is re stric te d to th e e th m o id region
(vomeral, m esethm oid, circum orb ital, qu ad rate, m etapterygoid, p reo p ercu lar,
in te ro p e rc u lar an d sym plectic bones). It appears to occur early in developm ent, as
observed in several Syngnathidae (e.g. Hippocampus antiquorum Leach (Ryder,
1881), Syngnathus peckianus S to rer (M cM urrich, 1883), S. fuscus (K indred, 1921),
Hippocampus (Kadam, 1958) an d Nerophis (Kadam, 1961)). In H. antiquorum an d S.
peckianus, th e eth m o id region is even elo n g ated before th e yolk sac is fully
absorbed (Ryder, 1881; M cM u rrich , 1883).
A sh o rt com parison b etw een som e of th ese elem ents in syngnathids an d th e
stickleback, as a generalized te le o st rep resen tativ e w ith o u t an elongated sn o u t, is
given here in o rder to u n d e rstan d th e im plications of sn o u t elo n g atio n o n cranial
m orphology in syngnathids (Fig. 3.7).
T he vom eral bone stretches up to th e lateral eth m o id b o n e in S. rostellatus and H.
capensis, b u t in Nerophis it does n o t reach th e nasal region (Kadam, 1961).
A ccording to K adam (1961), th is is a difference b etw een th e G astro p h o ri
(syngnathids w ith th e brood p o u ch ro stra l to anal fin: e.g. Nerophis) an d th e
U ro p h o ri (brood p o u ch caudal to anal fin: e.g. Syngnathus and Hippocampus).
R ostrally th e vom eral bone provides an a rtic u la tio n w ith th e m axillary bone, b u t
th e re is no m esethm oid-prem axillary artic u la tio n p re se n t as th e re is in prim itive
teleosts (Gregory, 1933).
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
45
In S. rostellatus and H. capensis th e q u ad rate b o n e consists of a p erich o n d ral
ascending process and a m em branous h o riz o n ta l process. W h e th e r or n o t th is
h o riz o n ta l
process
is
hom ologous
to
th e
one
considered
a
teleo stean
synapom orphy by A rratia an d S chultze (1991), could n o t be co n firm ed here. The
process is m uch sm aller on th e q u ad rate b o n e in G. aculeatus, w h ich is
trian g u larly shaped w ith its apex dorsally (Anker, 1974). T he v e n tro ro stra l co rn er
of th e q uadrate bone provides th e a rtic u la tio n w ith th e low er jaw and
v entrocaudally it bears a cartilaginous exten sio n th a t lies lateral to th e sym plectic
bone (Anker, 1974).
T he p reo p ercu lar bone in S. rostellatus an d H. capensis is L-shaped. In th e fo rm er
th e h o riz o n ta l process is su b stan tially lo n g er th a n th e vertical one, w hile in H.
capensis th e difference is less and in G. aculeatus th e vertical process is th e largest
(Anker, 1974). C audally, th is v ertical process m eets th e o p ercu lar b o n e in
sy ngnathid species (Jungersen, 1910; K indred, 1924; Kadam, 1961; B ranch, 1966),
b u t in G. aculeatus th e y only join each o th e r dorsally (ventrally th ey are separated
by an ascending process of th e su b o p ercu lar bone; S w in n erto n , 1902; A nker,
1974)In G. aculeatus th e in te ro p e rc u lar b o n e covers th e su b o p ercu lar b o n e caudally
(Anker, 1974), b u t b o th lie w ell sep arated from each o th e r in S. rostellatus and H.
capensis.
T he occurrence of an a n to rb ita l b o n e and lacrim al bone, follow ed by six
in frao rb ital bones bo rd erin g th e o rb it (the first, th ird an d six th being th e
lacrim al, jugal an d derm o sp h en o tic bones, respectively), is a prim itiv e fe a tu re of
m ost teleosts (Reno, 1966; N elson, 1969; Schultze, 2008). In th e su b o rd er
Syngnathoidei o th e r circu m o rb ital bones besides th e lacrim al b o n e are usually
ab sen t (N elson, 2006), how ever, in syngnathids th e re are usually tw o to th re e
in frao rb ital bones, w h ich develop late (Kadam, 1961). In S. rostellatus an d H.
capensis th e circu m o rb ital bones are p o sitio n ed in fro n t of th e o rb it in stead of
a ro u n d it. T here is, how ever, a difference b etw een th o se tw o species, as m ost
specim ens of th e seahorse stu d ied have an a n to rb ita l bone, a lacrim al b o n e (= first
in frao rb ital bone) an d a second in frao rb ital bone, w hereas th e re are only tw o
circu m o rb ital bones p re sen t in alm ost all S. rostellatus specim ens studied. H ere,
th e p o sterio r one corresponds to th e second in frao rb ital bone. The a n te rio r one is
th e largest one an d appears to be a fu sio n b etw een th e a n to rb ita l b o n e and th e
46
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
lacrim al bone. This hypothesis is su p p o rted by th e absence of a separate
a n to rb ita l bone, th e bone being as large as an d ta k in g th e place of b o th th e
a n to rb ita l bone and th e lacrim al b o n e in H. capensis. In addition, th e re is a v en tral
in d e n ta tio n th a t could p o in t o u t th e in co m p lete fu sio n b etw een th e a n to rb ita l
bone and th e lacrim al bone. The fo rm atio n of th e a n to rb ito lacrim al bone could
be a stru c tu ra l advantage to stre n g th en th e elongated sn o u t laterally. D u rin g th e
fast elevation of th e snout, large, v en trally o rie n te d forces are expected to be
exerted o n to th e dorsal p a rt of th e snout. In th e case of an u n fu sed a n to rb ita l
bone and lacrim al bone, a possible b en d in g zone b etw een th e tw o bones exists.
T he fo rm atio n of an a n to rb ito lacrim al bone could reduce th e risk of b en d in g and
still allow s lateral expansion of th e snout. In G. aculeatus th e re are th re e separate
circu m o rb ital bones p re sen t (S w innerton, 1902; de Beer, 1937; A nker, 1974).
B. F a s t s u c tio n fe e d in g a d a p ta tio n s
S yngnathid fishes are k n o w n to cap tu re prey by an u n u su al feeding strategy
k n o w n as pivot feeding (de L ussanet & M u ller, 2007). T hey p e rfo rm a rap id
elevation of th e head, w h ich brings th e m o u th quickly close to th e prey (M uller,
1987). T hen, expansion of th e ir long sn o u t generates a fast w a ter flow th a t carries
th e prey in to th e m outh. This increase in buccal volum e is m ainly achieved by a
lateral expansion, in stead of v en tral expansion typical fo r m ost su ctio n feeding
fish (Roos et al., 2009). T he hyoid is k n o w n to play an im p o rta n t role in
Suspensorium ab d u ctio n as w ell as in depression of th e low er jaw (Roos et al.,
2009).
Seahorses and pipefishes are am bush predators; th e y sit and w a it u n til a prey
com es close to th e m o u th (F oster & V incent, 2004). T hey are k n o w n to consum e
m ainly sm all crustaceans such as am phipods an d copepods (Foster & V incent,
2004; K endrick & H yndes, 2005) an d a recen t study by C astro et al. (2008) show ed
th a t nem atodes are also one of th e m ain food item s consum ed in th e wild.
A ccording to K endrick an d H yndes (2005), th e tro p h ic specialization of these
fishes can be explained by th e ir extrem e sn o u t m orphology (length an d gape),
th e ir feeding behaviour an d in th e case of seahorses, th e ir low m obility.
Syngnathids have a very sm all m o u th ap ertu re, severely lim itin g food particle
size. The m axillary an d prem axillary bones of S. rostellatus and H. capensis are
ra th e r small. T eeth, b o th oral and pharyngeal, are ab sen t an d prey is sw allow ed
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
47
w hole (Lourie et al., 1999b). Gasterosteus aculeatus, how ever, has a large, teeth bearing prem axillary bone th a t is p ro tru sib le (de Beer, 1937; A lexander, 1967a;
A nker, 1974; M o tta , 1984; N elson, 2006). U n d er th e assu m p tio n th a t a long
ascending process of th e prem axillary b o n e is associated w ith a great a m o u n t of
p ro tru sio n (Gosline, 1981; M o tta , 1984; W e s tn e a t & W ain w rig h t, 1989; W estn e at,
2004), th e lack of an ascending process in S. rostellatus and H. capensis indicates
th e re is no u p p er jaw p ro tru sio n (Branch, 1966; B ergert & W ain w rig h t, 1997).
T hey do have a sm all ro stral cartilage, d o rso ro stral to th e eth m o id p late and
m edial to th e m axillary bones. This ro stral cartilage is n o t necessarily an
a d ap ta tio n to th e p o w erfu l su ctio n feeding b u t could ra th e r be an ancestral
featu re also fo u n d in Percidae, C ichlidae, A th erin o id ei, G asterosteidae and
others, w here it assists in u p p er jaw p ro tru sio n (Alexander, 1967a; 1967b; M o tta ,
1984). A lternatively, th e ro stral cartilage in syngnathids could be involved in th e
fast ro ta tio n of th e m axillary an d prem axillary bones d u rin g m o u th opening.
D epression of th e low er jaw induces a ro stral sw ing of th e m axillary bone,
because of th e firm prim ordial lig am en t ru n n in g fro m th e co ro n o id process to
th e m axillary bone. As a consequence of th e co n n ec tio n b etw een m axillary and
prem axillary bones, b o th ro ta te rostrally. The m o u th a p ertu re is th e n laterally
enclosed, resu ltin g in a m ore circular gape, hence, a m ore ro strally d irected w ater
flow in to th e m o u th m ight be g en erated as h y p o th esized by Lauder (1979; 1985)
an d experim entally show n by S anford et al. (2009). K indred (1921) an d K adam
(1961) also fo u n d a ro stral cartilage in S. fuscus and Nerophis, w h ich is co n n ected
to th e p alatin e cartilage w ith dense connective tissue. K adam (1958) fu rth e r
m en tio n s a ro stral cartilage artic u la tin g w ith th e prem axillary an d m axillary
bones in Hippocampus.
T he low er jaw of S. rostellatus and H. capensis is sim ilar to th e one in G. aculeatus,
b u t m uch sh o rte r relative to th e ir h ead length. T he a n g u lo articu lar b o n e in th e
sy ngnathid species is m ore tig h tly fixed to th e d en tary bone, im proving th e
rigidity of th e low er jaw. This m ig h t facilitate ab d u ctio n of th e le ft and rig h t
low er jaws, observed d u ring m an ip u la tio n of specim ens (Roos et al., 2009). In th e
stickleback th e re is no fusion betw een th e angular b o n e and a rticu lar bone. The
angular bone also fits in to a cavity of th e d en tary bone, b u t w ith a p o te n tia l
pivoting zone in b etw een th e m (Anker, 1974). T here is a saddle-like jo in t b etw een
th e a rticu lar bone and th e q uadrate bone, as in S. rostellatus and H. capensis.
48
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
T he m etapterygoid bone is a p erich o n d ral ossification of th e m etapterygoid
process of th e p alato q u ad rate cartilage (A rratia & Schultze, 1991). In G. aculeatus,
as in o th e r general teleosts, th e q u ad rate and th e h y o m an d ib u lar bones are
co n n ected by m eans of th e m etapterygoid bone, form ing th e Suspensorium
(Gregory, 1933; A nker, 1974). This is n o t th e case in S. rostellatus an d H. capensis,
w here th e re is no co n n ec tio n b etw een th e sh o rt m etap tery g o id an d th e
h y o m and ibular bones.
ru d im en ta ry
N e ith er
m etapterygoid
is th e re
process
of
a co n n ec tio n
th e
b etw een
p tery g o q u ad rate
p a rt
th e
of
very
th e
p alato q u ad rate cartilage and th e hyosym plectic cartilage in th e p ip efish and
seahorse juveniles.
T he sym plectic p a rt of th e hyosym plectic cartilage in S. rostellatus juveniles is
very long com pared to th e hyom an d ib u lar p art, w ith th e angle b etw een th ese tw o
p a rts being obtuse. In H. capensis, b o th p arts are alm ost equally long an d th e y are
p erp en d icu lar to each other. This arran g em en t looks very m u ch like th e one in G.
aculeatus (S w innerton,
1902; K indred,
1924). K adam
(1961) describes th e
sym plectic bone in Nerophis as a ch o n d ro m em b ran o u s bone w ith a p erich o n d ral
p art, nam ely th e ossification of th e a n te rio r region of th e hyosym plectic cartilage,
an d an in tram em b ran o u s p art, w h ich rises up fro m th e p erich o n d ral part. The
vertical p late bears a d orsorostral process and decreases gradually in h eig h t m ore
caudally. This is also fo u n d in S. rostellatus andiT. capensis.
A t th e 6.3-9.0 m m SL stage of G. aculeatus, w h ere th e re is n o ossification of th e
cranial cartilage yet, th e hyom an d ib u lar p a rt of th e hyosym plectic cartilage
already has th e tw o-headed a rtic u la tio n w ith th e n eu ro cran iu m as seen in adults
(S w innerton, 1902; K indred, 1924; de Beer, 1937; A nker, 1974). The d o rso ro stral
condyle articu lates in a socket form ed by th e sp h en o tic bone, th e dorsocaudal
condyle fits in a socket of th e p te ro tic bone (Anker, 1974). In th e juvenile
syngnathids (S. rostellatus, H. capensis an d H. reidi), th e re is only a single
cartilaginous articu latio n . The hy o m an d ib u lar b o n e in ad u lt S. rostellatus an d H.
capensis is sim ilar to th e one in G. aculeatus-, it also bears a double a rticu lar facet
w ith th e n eu ro cran iu m , as in H. reidi. D issection an d m a n ip u la tio n of th is double
h y o m and ibular a rtic u la tio n in S. rostellatus and H. capensis proved th a t it is very
firm . Strikingly, in S. fuscus (K indred, 1924; de Beer, 1937), Nerophis (Kadam, 1961)
an d S. acus (Branch, 1966) only a single condyle is p resen t, w h ich is th o u g h t to
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
49
increase th e freedom of m ovem ent of th e hy o m an d ib u lar bone (K indred, 1924;
B ranch, 19 66).
T he c o n n ectio n betw een th e Suspensorium and th e hyoid arch is provided by th e
in terh y al bone. The general te le o st a rtic u la tio n is a ball-and-socket jo in t, w ith a
rod-shaped in terh y al bone bearing a ro u n d ed h ead th a t fits in to a facet of th e
Suspensorium , allow ing th e in terh y al b o n e to ro ta te in every d irectio n w ith
respect to th e Suspensorium (Anker, 1989; A erts, 1991). T he c o n fig u ratio n in G.
aculeatus is sim ilar (Anker, 1974). This is n o t tru e fo r S. rostellatus an d H. capensis,
w here th e in terh y al bone articu lates w ith th e p reo p ercu lar b o n e dorsally and
bears tw o a rtic u la tio n heads ventrally, in b etw een w h ich th e p o sterio r ceratohyal
bone articulates. In th a t way, m ovem ent is m ore re stric te d to one in a
rostro cau d al d irection, resu ltin g in a hyoid depression d u rin g th e expansive phase
of th e su ctio n feeding. T he tw o heads of th e hy o m an d ib u lar bone in co m b in atio n
w ith th e ro b u st in terh y al bone can be assum ed to in d irectly reduce th e degrees of
freedom b etw een th e hyoid an d th e n eu ro cran iu m , hence c o n tra ctio n of th e
sternohyoideus m uscle is expected to be tra n slated in a m ore p o w erfu l hyoid
depression. Fast hyoid ro ta tio n is th u s possible w ith a reduced risk of
d isartic u latio n of th e ceratohyal bone. In S. peckianus (M cM urrich, 1883), S. fuscus
(de Beer, 1937), Nerophis (Kadam, 1961), S. acus (Branch, 1966), S. floridae an d H.
erectus (B ergert & W ain w rig h t, 1997) th e in terh y al bone is sim ilar, b u t it is
claim ed to artic u la te w ith th e h y o m an d ib u lar b o n e in stead of th e p reo p ercu lar
bone.
M u ller and Osse (1984) show ed th a t h ig h negative pressures w ill be reached in
th e gili cavity of th e p ipefish Entelurus aequoreus d u rin g prey capture. A ccording
to Osse and M u ller (1980) th e sm all gili slit and th e strongly ossified gili cover are
considered adap tatio n s to th e pivot type of feeding, ch aracterized by a very fast
n eu ro cran ial elevation (M uller & Osse, 1984; M u ller, 1987; de L ussanet & M uller,
2007). T he pressure in th e opercu lar cavities is considered to be h ig h er w ith
increasing sn o u t length, an d a com parison b etw een d ifferen t sy n g n ath id species
show ed th a t increasing sn o u t le n g th also results in a stru c tu ra lly m ore ro b u st
o percular bone (e.g. m ore ridges, g reater cu rv atu re and th ick er; Osse & M uller,
1980; M u ller & O sse, 1984). In b o th , S. rostellatus an d H. capensis, th e gili slits are
nearly closed by a firm sheet of connective tissue covered w ith skin w ith only a
sm all ap ertu re a t th e dorsocaudal tip. T h eir o p ercu lar b o n e is firm an d th ic k and
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
50
has a convex surface w h ich w ill h elp in w ith sta n d in g m edially directed forces.
C om parison betw een th e tw o species reveals th a t th e o p ercu lar b o n e in th e
pipefish, w h ich has a m ore elongated sn o u t, is sm aller, th ic k e r an d has a g reater
curvature, as expected.
T he branchiostegal rays su p p o rt th e b ran ch io steg al m em brane, w h ich closes th e
gili
cavity
ventrally.
A m ong
teleo sts
th e re
can
be
m ore
th a n
tw e n ty
branchiostegal rays, b u t acanthop tery g ian s alm ost never have m ore th a n eight
(G osline, 1967; A rratia & Schultze, 1990). In syngnathids th e n u m b er of
branchiostegal rays varies betw een one an d th re e (M cA llister, 1968). H ere, in S.
rostellatus an d H. capensis, th e re are only tw o b ran ch io steg al rays p re sen t o n each
side. As G osline (1967) p o in te d ou t, th e n u m b er of b ran ch io steg al rays is related
to th e le n g th of th e hyoid bar. In syngnathids, th e hyoid is relatively sm all, w h ich
m ight be associated w ith its lost fu n c tio n as a m o u th b o tto m depressor (Roos et
al., 2009). A lo n g er hyoid w ill have an increased m o m en t of in e rtia resu ltin g in
hyoid depression a t a low er velocity. In addition, th e angle b etw een th e line of
actio n of th e sternohyoideus m uscle and th e hyoid w ill becom e less favourable as
th e hyoid le n g th increases. Thus, th e le n g th of th e hyoid b ar is expected to be
c o n strain ed and consequently, th e re w ill be less available space fo r a tta c h m e n t of
th e branchiostegal rays.
In b o th ad u lt an d juvenile H. capensis th e braincase is tilte d dorsally w ith respect
to th e eth m o id region so it is situ a ted dorsocaudally to th e o rb ita in stead of
caudally as in S. rostellatus, S. fuscus an d Nerophis (all pipefishes; K indred, 1921;
K adam, 1961). In th e H. reidi juvenile (7.01 m m SL) th is dorsal tiltin g of th e otic
capsule was also visible, a lth o u g h th e tilt was less th a n th e one in H. capensis
juvenile (only a b o u t 200 up com pared to 340 in th e latter). In ad u lt H. capensis,
m ore or less th e same tilt is observed (38o up). K adam (1958) described th e
presence of a sp h en o -p tero tic ridge a t th e base of th e taen iae m arginales (w hich
he calls p o sto rb ita l processes) in Hippocampus th a t appears to be m issing in
Nerophis (Kadam, 1961) an d S. fuscus (K indred, 1921). W e also observed th is ridge
in H. capensis and H. reidi, b u t n o t in S. rostellatus. A p art from th a t, th e
hyosym plectic a rtic u la tio n socket m ediocaudally to th is ridge is m ore d istin c t in
th e Hippocampus species studied.
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
51
It is obvious th a t already a t an early developm ental age th e juvenile feeding
ap p aratu s resem bles th a t of a d u lt S. rostellatus an d H. capensis. This m ig h t be th e
re su lt of th e specialized p a re n tal care th a t enables th e p o stp o n in g of release from
th e brooding p o u ch u n til an advanced developm ental state is reached. In th e
seahorse brooding pouch, oxygen is su p p lied th ro u g h su rro u n d in g capillaries and
th e m ale p ro la c tin h o rm o n e is secreted, ind u cin g b reak d o w n of th e ch o rio n to
produce a p lacen tal fluid (Lourie et al., 1999b; C arcu p in o et al., 2002; F o ster &
V incent, 2004). Lack of oxygen an d endogenous energy is pro b ab ly n o lo n g er a
lim itin g facto r an d em ergence fro m th e p o u c h m ay be delayed, as in Galeichthys
feliceps V alenciennes, an ariid m o u th -b ro o d er (Tilney & H ech t, 1993).
52
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
3.2.
53
Im p l i c a t i o n s o f s n o u t e l o n g a t i o n
M o d i f i e d fr o m :
L e y s e n H , C h r i s t i a e n s J ,D e K e g e l B , B o o n e M N ,
V a n H o o r e b e k e L, A d r ia e n s D .
M u s c u lo s k e le t a l str u c tu r e o f t h e fe e d in g sy s te m
a n d im p lic a tio n s o f s n o u t e lo n g a t io n in
H ip p o c a m p u s reidi a n d D o r y r h a m p h u s d a c ty lio p h o ru s
J o u r n a l o f F is h B i o lo g y , i n p r e s s
A bs tract
Syngnathids cap tu re prey by m eans of high-velocity p iv o t feeding w ith
Suspensorium ab d u ctio n an d n eu ro cran ial elevation. The presence of adaptive
evolutionary m odifications in th e cranial m u sculoskeletal system , related to th is
type of feeding, is assum ed. It is h y p o th esized th a t th e sy n g n ath id tro p h ic
apparatus, being highly specialized, w ill show a conserved m orphology am ong
species; how ever, som e stru c tu ra l v ariatio n related to v ariatio n in sn o u t le n g th is
expected. A th o ro u g h m orphological d escrip tio n of th e feeding ap p aratu s in a
lon g -sn o u ted seahorse an d p ip efish (Hippocampus reidi and Doryrhamphus
dactyliophorus) revealed som e specialized featu res th a t m ig h t be associated w ith
54
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
th e fast and p o w erfu l su ctio n feeding, like th e tw o ligam entous co n n ectio n s
b etw een th e low er jaw and th e hyoid, th e saddle jo in t of th e la tte r w ith th e
Suspensorium an d th e v ertebro-p ecto ral fu sio n th a t articu lates o n th re e p o in ts
w ith th e cranium . D espite th e conserved m orphology of th e feeding apparatus,
som e m orphological differences w ere observed b etw een th e tw o species. For
instan ce in H. reidi th e o rie n ta tio n of th e occipital jo in t is v en tro cau d al, th e
sternohyoideus an d epaxial m uscles are m ore bu lk y an d b o th have a sh o rt ten d o n .
In D. dactyliophorus, on th e o th e r hand, th e p ro tra c to r hyoidei m uscle is enclosed
by th e m andibulo-hyoid ligam ent, th e sternohyoideus an d epaxial ten d o n s are
long an d a sesam oid bone is p re sen t in th e latter. These features w ere com pared
to th e c o n d itio n in several o th e r sy n g n ath id species w ith a large range of sn o u t
lengths, in o rder to evaluate th e im plications of sn o u t elo n g atio n o n th e
m usculoskeletal stru c tu re of th e cranium . T he arched p a th of th e ad d u cto r
m andibulae m uscle and th e greater rigidity of th e low er jaw m ig h t be related to
elo n g atio n of th e sn o u t since it yields an increased m echanical advantage of th e
low er jaw system and a reduced to rq u e b etw een th e elem ents of th e low er jaw
durin g p ro tra c to r hyoidei m uscle c o n tra ctio n , respectively. N evertheless, m ost
observed features did n o t seem to be related to sn o u t length.
3.2.1 In troduction
Pipefishes and seahorses are rem arkable creatures. T h eir ex ten d ed body is
covered w ith an arm o u r of bony p lates in stead of scales, th e y have in d ep en d en tly
m oveable eyes an d males have a special reproductive s tru c tu re like a p o u ch or a
brooding area, w h e th e r or n o t covered by overlapping m em branes, to p ro te c t th e
eggs and young (e.g. H erald, 1959; V in cen t & Sadler, 1995; Lourie et al., 1999b;
W ilso n et al., 2001; K uiter, 2003; N elson, 2006). In ad d itio n , seahorses are
ch aracterized by a grasping ta il in stead of a caudal fin, a vertical body axis an d a
tilte d head (e.g. Lourie et al., 1999b; 2004). T he ev o lu tio n to th is u p rig h t body
p o stu re m ight be related to th e Late O ligocene expansion of seagrass h ab itats
since th e vertical leaves serve as excellent cam ouflage fo r seahorses, favouring
b o th am bushing prey an d avoiding p re d a tio n (Teske & Beheregaray, 2009), as w ell
as a v ertical body p o sitio n has been suggested to increase th e distance fro m w h ich
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
55
a prey can be cap tu red d u ring su ctio n feeding in seahorses (Van W assen b erg h et
al., 2011). Besides all this, syngnathids have a tu b u la r sn o u t, form ed by th e
elo n g atio n of th e ethm oid region (chapter 3.1). W ith in th e fam ily a large
diversity in th e sn o u t m orphology exists, ranging fro m very sh o rt-sn o u ted
species, like Siokunichthys breviceps Sm ith, to species w ith an extrem ely elongated
sn o u t, like Phyllopteryx taeniolatus. All sy n g n ath id species stu d ied th u s far are
k n o w n to p e rfo rm fast and p o w erfu l su ctio n feeding w ith th a t sn o u t (e.g. ch ap ter
5.1; Ryer & O rth , 1987; B ergert & W ain w rig h t, 1997; de Lussanet & M uller, 2007;
V an W assenbergh et al., 2008; S to rero & G onzález, 2009). In teleosts, su ctio n
feeding is th e m ost com m on feeding p a tte rn . It involves a fast en larg em en t of th e
oro-branchial cavity resu ltin g in a pressure difference b etw een th e inside of th a t
cavity an d th e su rro u n d in g w ater. W h e n th e jaws open, a w a te r flow is created
w h ich w ill tra n s p o rt th e prey in to th e m outh. T he o ro -b ran ch ial expansion can
be g en erated in v entral, dorsal, lateral an d ro stral directio n s by hyoid depression,
n eu ro cran ial elevation, suspensorial ab d u ctio n an d b o th m axillary p ro tru sio n and
low er jaw depression, respectively (Lauder, 1983). In pipefishes an d seahorses,
how ever, th e jaws are n o t p ro tru sib le (Branch, 1966; B ergert & W ain w rig h t, 1997)
an d th e re is alm ost no v en tral expansion of th e buccal cavity by m eans of hyoid
ro ta tio n (Roos et al., 2009). Instead, th e m ain buccal volum e increase in
syngnathids is achieved by rap id suspensorial ab d u ctio n , in d u ced by th e
depression of th e hyoid an d low er jaw, w here n eu ro cran ial elevation p ositions
th e m o u th close to th e prey (Roos et al., 2009). This type of feeding, w here only
th e head an d n o t th e e n tire body is accelerated to w ard s th e prey, is called pivot
feeding (de L ussanet & M uller, 2007) an d is also fo u n d in som e deepsea fishes and
flatfishes (Tchernavin, 1953; M u ller & O sse, 1984; G ibb, 1995; 1997).
D espite a large variatio n in th e sn o u t m orphology (K uiter, 2003), th e feeding
m echanism of syngnathids seems q u ite conserved. H ence, th e q u estio n is w h a t
th e best m orphology, i.e. long or sh o rt sn o u t, w o u ld be fo r o p tim al perfo rm an ce
durin g th is specialized type of su ctio n feeding. K endrick an d H yndes (2005)
p red icted th a t syngnathids w ith lo n g er sn o u ts w o u ld be able to a tta c k prey from
a greater distance and w ith g reater speed th a n th o se w ith sh o rte r snouts, an d th u s
w o uld be m ore successful at catching relatively fast m oving prey. This p red ictio n
w as confirm ed by a d ietary analysis of b o th sh o rt an d lo n g -sn o u ted species; th e
la tte r h ad consum ed a greater a m o u n t of m obile prey. Also de Lussanet and
56
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
M u ller (2007) argued th a t th e o ptim al sn o u t le n g th is a species-specific trad e-o ff
b etw een a long sn o u t on th e one hand, w h ich decreases th e angle th e h ead needs
to ro ta te an d hence th e tim e to ap p ro ach th e prey, an d a sh o rt sn o u t o n th e o th e r
hand, w h ich decreases th e m o m en t of in e rtia of th e h ead resu ltin g in faster prey
intake. Roos et al. (subm itted a) p e rfo rm ed k in em atical analyses of prey cap tu re
events in b o th a long- an d a sh o rt-sn o u te d pipefish, respectively Doryrhamphus
dactyliophorus (Bleeker) an d Doryrhamphus melanopleura (Bleeker). T hey n o te d
th a t th e to ta l distance travelled by th e sn o u t tip d u rin g p iv o t feeding is larger in
th e lo n g-snouted pipefish. This suggests th a t species w ith a long sn o u t can attack
prey from fu rth e r away, so probab ly m ore elusive prey can be caught. O n th e
o th e r hand, sh o rt-sn o u ted species show ed lo w er prey cap tu re tim es. Besides th a t,
th e y also fo u n d th a t prey catching efficiency w as su b stan tially less in lo n g ­
sn o u ted pipefishes com pared to th e pipefishes w ith a sh o rt snout. Since a correct
p o sitio n in g of th e m o u th w ith resp ect to th e prey is crucial, th e observed
difference in prey catching efficiency m ig h t be related e ith e r to th e prey type
used in th e experim ent, w h ich could have b een to o sm all fo r th e lo n g -sn o u ted D.
dactyliophorus to allow accurate aim ing, or to th e larger distance b etw een th e
sn o u t tip an d th e eyes in th e lo n g -sn o u ted p ip efish (Roos et al., su b m itte d a). A
m ore re c en t study (Roos et al., 2011) show s th a t a large sn o u t d iam eter im proves
volum e increase an d expansion tim e, b u t negatively effects m axim al flow
velocity. A long sn o u t reduces b o th m axim al flow velocity an d su ctio n volum e,
how ever, as m e n tio n e d before, th e tim e to reach th e prey is decreased.
C onsidering th e highly specialized feeding strateg y of syngnathids, i.e. th e highvelocity pivot feeding w ith Suspensorium a b d u ctio n an d n eu ro cran ial elevation,
it can be assum ed th a t th is has b een associated w ith adaptive ev o lu tio n ary
m odifications
in
th e
cranial
m usculoskeletal
system.
In
addition,
m ore
specialized species are generally th o u g h t to show a reduced m orphological
v ersatility since even th e sm allest deviation of one elem en t of th e com plex
in te g ra te d system can have an effect o n th e p erfo rm an ce (Adriaens & H errei,
2009). H ence, th e hypothesis te ste d in th is study is th a t th e sy n g n ath id tro p h ic
apparatus, being highly specialized, w ill show a conserved m orphology to som e
degree b u t w here existing stru c tu ra l v ariatio n can be related to v ariatio n in sn o u t
length. Also, since changes in th e sn o u t geom etry w ill have an im pact on
fu n c tio n a l lever system s (especially at th e level of ten d o n s, ligam ents and th e ir
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
57
attachm ents), differences are expected to be fo u n d th a t reflect com pensations to
su stain kin em atic versus force efficiency in th ese lever systems. For exam ple,
elo n g atio n of th e sn o u t an d w ith it th a t of th e jaw closing a d d u cto r m andibulae
m uscle com plex, is expected to be associated w ith a change in its m usculoskeletal
arch itectu re. So, an a rran g em en t of th e jaw closure ap p aratu s in th e longer
sn o u ted species to assure th e sam e k in em atical efficiency is expected. The
m echanical advantage of th e low er jaw lever system is fo r in stan ce increased by
increasing th e h e ig h t of th e co ro n o id process of th e d en tary b o n e (w here th e
a d d u cto r m andibulae ten d o n s attach), or by a m ore dorsocaudally o rie n ta te d
m uscle’s line of action w ith respect to th e jaw articu latio n . Also, w h en th e
p ro tra c to r hyoidei m uscle co n tracts w hile th e m o u th is k e p t closed, it m ight
induce som e to rq u e b etw een th e elem ents of th e low er jaw. L ong-snouted species
likely g enerate h igher stra in levels in th e m uscle an d th e re fo re th e in te rd ig ita tio n
b etw een th e an g u lo articu lar an d d en tary bones is expected to be m ore firm.
Finally, w h e n sn o u t le n g th increases, th e m o m en t of in e rtia d u rin g h ead ro ta tio n
w ill increase as well, so a larger p o w er o u tp u t is n eeded to accom plish fast
n eu ro cran ial elevation. This m ig h t be realised by a larger epaxial m uscle mass or
an im proved pow er am plifying system , such as a lo n g er epaxial ten d o n .
To validate these hypotheses, a th o ro u g h m orphological d escrip tio n of th e
feeding ap p aratu s in a lo n g-sn o u ted seahorse, Hippocampus reidi, an d an
extrem ely long-snouted pip efish (D. dactyliophorus) is p resen ted , to establish
m orphological tra its th a t could be associated w ith th e specialized feeding type.
These tra its are com pared am ong several o th e r sy n g n ath id species w ith a large
range of sn o u t lengths to determ in e th e im plications of sn o u t elo n g atio n o n th e
m usculoskeletal c o n fig u ra tio n of th e feeding apparatus.
3.2.2. Brief material & m e t h o d s
All anim als w ere o b tain ed from com m ercial trad e in accordance w ith th e CITES
req u irem en ts (Table 2.1).
D issections w ere perfo rm ed on th re e specim ens of H. reidi (113.7 mm, 138.1 m m
an d 152.8 m m SL), one specim en of D. dactyliophorus (103.5 m m SL), one
Hippocampus abdominalis Lesson specim en (222.7 m m SL), one D. melanopleura
58
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
specim en (56.5 m m SL), tw o specim ens of Corythoichthys intestinalis Ramsay
(120.8 m m an d 136.2 m m SL), one Doryrhamphus janssi specim en (105.6 m m SL)
an d one Dunckerocampus pessuliferus specim en (112.3 m m SL).
T hree specim ens of H. reidi (117.2 mm, 113.5 m m a n d 109.7 m m Ls), one of D.
dactyliophorus (105.1 m m SL), tw o of C. intestinalis (102.7 m m a n d 117.6 m m SL),
one of Hippocampus zosterae (32.8 m m SL) one of D. melanopleura (41.6 m m SL)
an d tw o of D. janssi (81.5 m m an d 90.8 m m SL) w ere cleared an d stain ed fo r bone
an d cartilage according to th e p ro to co l of Taylor an d V an D yke (1985).
Serial histological cross sections of th e head of a H. reidi (103.5 m m SL), a D.
dactyliophorus (91.2 m m SL), a H. zosterae (29.0 m m SL), a D. janssi (102.7 m m SL)
an d a D. pessuliferus (94.5 m m SL) specim en w ere made.
C T-scans of a n o th er specim en of H. reidi (119.0 m m SL), th e dissected D.
dactyliophorus specim en (103.5 m m SL), th e dissected H. abdominalis specim en
(222.7 m m SL), a specim en of C. intestinalis (120.2 m m SL), th e dissected D.
melanopleura (56.5 m m SL) an d th e sectioned H. zosterae specim en (29.0 m m SL)
w ere m ade and com p u ter-g en erated 3 D -reconstructions w ere gen erated to
visualize m usculoskeletal to p o g rap h y of th e cranium . The 3 D -reco n stru ctio n of
H. reidi show s a slight d isto rtio n at th e level of th e sn o u t, w h ich is pro b ab ly an
a rte fa ct due to th e alig n m en t (Fig. 3.8). H ow ever, th is does n o t im pair th e
qualitative in te rp re ta tio n s in th is study.
3.2.3. Results
A th o ro u g h descrip tio n of th e head m orphology of H. reidi and D. dactyliophorus
is given as a representative p ipefish and seahorse feeding ap p aratu s, w h ich is
fu rth e r used for a com parison w ith th a t of o th e r syngnathids species. Special
a tte n tio n is paid to th e stru c tu re s th a t are expected to have undergone
m odifications related to differences in sn o u t len g th , as h y p o th esized in th e
in tro d u ctio n . N ext, com parison of th o se stru c tu re s is m ade b etw een several
sy ngnathid species w ith varying sn o u t length.
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
A.
59
M o rp h o lo g ic a l d e s c r ip tio n o f H ip p o ca m p u s r e id i
B oth low er jaw halves of H. reidi consist of M eckel’s cartilage, and th e dentary,
th e an g u lo articu lar an d th e re tro a rtic u la r bones. T he d en tary bone (Fig. 3.8A, 3.9)
is th e largest elem ent. It bears n o te e th and roughly has a trian g u la r shape w h en
view ed laterally. The dorsocaudal p ro m in e n t coronoid process is co n n ected to th e
v e n tro la te ral surface of th e m axillary b o n e by m eans of th e prim o rd ial ligam ent.
Tw o of th e a d d u cto r m andibulae m uscle bundles, nam ely A í an d A2, a tta c h o n to
th e coronoid process (Fig. 3.8A, 3.9, all ad d u cto r m andibulae ten d o n s w ere
reco n stru cte d to g e th e r (t-A) and th e ir d ifferen t a tta c h m e n t sites w ere observed
on dissections). A í attaches by m eans of a slender b ifu rcated ten d o n , of w h ich
th e dorsal p a rt attaches o n to th e dorsom edial surface of th e m axillary b o n e and
th e v en tral p a rt on th e corono id process of th e d en tary bone. The A í m uscle
b u n d le (Fig. 3.9A) is situ a te d in b etw een th e m edial A3 an d th e lateral A2 and
ru n s dorsal to th e la tte r to th e caudal edge of th e sym plectic b o n e w here it
attaches. T he b u n d le A2 (Fig. 3.8B, 3.9A) is th e m ost lateral one, w ith th e longest
te n d o n an d th e largest m uscle mass. Fibres are o rie n ta te d ro strocaudally and
a tta c h on th e p reo p ercu lar an d h y o m an d ib u lar bones. The A3 te n d o n attaches
v entrally on th e m edial face of th e dentary. The sh o rt an d th in A3 m uscle sh eet is
th e m ost m edial b u n d le of th e m uscle com plex an d attach es o n th e ro stral m argin
of th e sym plectic bone. The m uscle b u n d le is in all th re e cases m u ch lo n g er th a n
th e te n d o n in fro n t of th e m uscle and th e y all ru n alm ost in a stra ig h t line. The
d en tary bone is provided w ith a tu b e-lik e cavity in w h ich th e ta p erin g end of th e
an g u lo articu lar bone fits. A t th e m edial side of th e d en tary bone, th e sto u t
m andibulo-hyoid ligam ent (Fig. 3.8C, 3.9) attach es ventrocaudally, w h ich is
co n n ected to th e a n te rio r ceratohyal bone. M o re ro strally an d close to th e
d en tary sym physis, th e re is a sm all a tta c h m e n t site of th e th in p ro tra c to r hyoidei
te n d o n (Fig. 3.8C). The p ro tra c to r hyoidei m uscle (Fig. 3.8A, 3.8C, 3.9A, 3.10A)
ru n s ventrally to th e m andibulo-hyoid lig am en t and attach es to th e p o sterio r
ceratohyal bone (see ch ap ter 5.1 fo r a m ore d etailed d escrip tio n of th is muscle).
T he ro stral te n d o n of th e p ro tra c to r hyoidei is covered dorsally by th e
in term an d ib u laris m uscle (Fig. 3.8C) c o n n ectin g le ft an d rig h t d en tary bones. The
an g u lo articu lar bone (Fig. 3.8A, 3.9) bears a ro stral process enclosed by th e
d en tary cavity and th e angular p a rt of th e coronoid process stretch es along th e
caudal m argin of th e d en tary bone. T he d en tary an d an g u lo articu lar bones are
6o
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
firm ly fixed by connective tissue an d fo rm a single fu n c tio n a l un it. C audally, th e
an g u lo articu lar bone form s a saddle-shaped cartilaginous a rtic u la tio n fo r th e
q uadrate bone su rro u n d e d by stro n g ligam ents. T he re tro a rtic u la r b o n e (Fig.
3.8A, 3.8C, 3.9) is a very sm all bon e th a t is co n n ected to th e in te ro p e rc u lar bone
by m eans of a firm m an d ib u lo -in tero p ercu lar ligam ent.
B oth th e prem axillary and m axillary bones are reduced an d to o th less bones th a t
a rtic u la te w ith each o th e r dorsally. The prem axillary b o n e (Fig. 3.8A, 3.8B, 3.9) is
a slender bone th a t has a som ew hat w id er h ead th a t tap ers ventrally. B oth
prem axillary heads are ligam entously co n n ected an d a sh o rt ligam entous sheet
connects th e prem axillary bone to th e m axillary bones rostrally. The m axillary
bone (Fig. 3.8A, 3.8B, 3.9) is m ore ro b u st w ith a b ro ad en in g v en tral p a rt w h ere th e
p rim ordial ligam ent attaches. D orsally at th e m axillary heads a lig am en t attaches
th a t runs to th e au to p alatin e bone. In b etw een th e le ft an d rig h t m axillary bones,
a ro stral cartilage is situated, w h ich is enclosed by a large lig am en t th a t ru n s to
th e vom eral bone. The m axillary bones a rticu late w ith th e cartilage dorsally to
th e vom eral bone.
T he Suspensorium form s th e lateral w all of th e tu b u la r sn o u t an d m ost elem ents
are th e re fo re elongated. The sm all a u to p alatin e bone (Fig. 3.8A, 3.8B, 3.9B) is a
slender b ar w ith a th ic k e n ed ro stral end, situ a ted in a h o riz o n ta l p lan e lateral of
th e vom eral bone. Its v en tral side fits in to a dorsal groove of th e ectopterygoid
bone an d b o th are firm ly con n ected by m eans of connective tissue. T he ro stral
p a rt of th e a u to p alatin e bone extends from th is groove and articu lates
cartilaginously w ith th e ethm oid cartilage dorsal to th e vom eral bone. This form s
th e ro stral one of th e tw o a rticu latio n s th a t m oveably co n n ect th e Suspensorium
to th e neurocranium . A part from th e h o riz o n ta l p a rt th a t su rro u n d s th e v en tral
p a rt of th e a u to p alatin e bone, th e ectopterygoid b o n e (Fig. 3.8A, 3.8B, 3.9B) bears
a
tap erin g
descending
process.
This
process
ru n s
along
th e
ascending
p e rich o n d ral p a rt of th e quadrate bone to w h ich it is co n n ected th ro u g h a sheet
of connective tissue. The m edial surface of th e ectopterygoid b o n e is co n n ected
to th e lateral side of th e vom eral b o n e by m eans of a ligam ent. The m etapterygoid
bone (Fig. 3.8A, 3.8B, 3.9B) is a plate-sh ap ed b o n e w ith a tap erin g end, w ith its
v en tral edge covered by th e circu m o rb ital bones. R ostrally it su rro u n d s th e
d orsolateral surface of th e q uadrate bone and caudally its v e n tro la te ral end is
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
61
covered by th e lacrim al bone to w h ich it is stro n g ly a ttach ed by m eans of
connective tissue. The quadrate b o n e (Fig. 3.8, 3.9B) consists of tw o parts, i.e. an
ascending p e rich o n d ral one w ith th e q u ad rato -m an d ib u lar a rtic u la tio n a t its base
an d a large h o riz o n ta l plate w ith a tap erin g end. T he m an d ib u lar a rtic u la tio n is
saddle-shaped w ith th e lateral head larger and lying m ore ro strally com pared to
th e m edial one. B oth processes are co n n ected to th e an g u lo articu lar b o n e by
m eans of s to u t ligam ents. Laterally to th e ascending process th e re is a
lo n g itu d in al groove, w h ich is closed o n th e lateral side by th e circu m o rb ital
bones. In th is groove lie th e ten d o n s of th e a d d u cto r m andibulae m uscle com plex
(Fig. 3.10B). The h o riz o n ta l b ra n c h of th e q u ad rate bone reaches th e nasal cavity
an d has a convex shape w h en view ed rostrally. In th e space th u s m edially created,
lie th e hyosym plectic cartilage an d th e sym plectic bone, a ttach ed to th e surface
of th e q uadrate bone by m eans of connective tissue. A lm ost halfw ay th e le n g th of
th e quadrate bone its h o riz o n ta l process starts tap erin g , w ith th e p reo p ercu lar
bone lying ventrally. T here is a slender v en tral gap b etw een th e tw o bones and
th e ven tral edge of th e quadrate bone an d th e dorsal m argin of th e p reo p ercu lar
bone are firm ly co n n ected th ro u g h a ligam ent. T he sym plectic b o n e (Fig. 3.9A) is
rod-shaped an d bears a plate-like, dorsal crest w ith a ro stral process. T he rod
encloses th e hyosym plectic cartilage b u t th e ro stral en d rem ains free, as w ell as
th e caudal p a rt w here it is still fused w ith th e h y o m an d ib u lar p a rt of th e
hyosym plectic cartilage. R ostrally th e p e rich o n d ral p a rt is su rro u n d e d by th e
q uadrate bone and m ore caudally by th e p reo p ercu lar bone. T he plate-like dorsal
crest of th e sym plectic bone closes th e m edial side of th e cavity created by th e
circu m o rb ital bones in w h ich th e a d d u cto r m andibulae m uscle com plex runs.
T he A í an d A3 bundles of th e com plex (Fig. 3.9A) a tta c h o n th e caudal an d ro stral
p a rt of th is dorsal crest of th e sym plectic bone, respectively. The d o rso ro stral end
of th e sym plectic bone is con n ected to th e m esethm oid b o n e th ro u g h a large
ligam ent. C audally th e sym plectic b o n e has a d o rso lateral ridge su p p o rtin g th e
a d d u cto r arcus p a la tin i m uscle (Fig. 3.8B, 3.9A). This m uscle has its origin along
th e v en tro la te ral side of th e parasp h en o id b o n e an d inserts o n th e sym plectic
bone rostrally an d on th e hyom an d ib u lar bone caudally. A t th e caudal tip of th e
p e rich o n d ral rod of th e sym plectic bone, a firm lig am en t attach es m edially th a t
ru n s to one of th e processes of th e in terh y al bone. The v en tral side of th e
p reo p ercu lar bone is con n ected to th e dorsal edge of th e in te ro p e rc u lar b o n e by
62
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
m eans of a long ligam ent, leaving en o u g h space fo r depression of th e
in te ro p e rc u lar bone. B undle A2 of th e ad d u cto r m andibulae m uscle com plex
attaches on th e m edial face of th e p reo p ercu lar an d h y o m an d ib u lar bones. The
h y o m and ibular bone (Fig. 3.9A) bears a d orsorostrally an d a dorsocaudally
o rie n te d a rtic u la tio n head th a t fit respectively in to a socket o n th e v en tral
surface of th e sp h en o tic bone and in th e cartilaginous surface b etw een th e
sphenotic, th e p ro o tic and th e p te ro tic bones. It also has a m edial an d a lateral
w ing form ing a hollow space fo r th e levator arcus p a latin i m uscle an d th e
d ilatato rs o perculi m uscles (Fig. 3.9A). The d ila ta to r o p ercu li m uscle consists of
tw o bundles, a dorsal one an d a v en tral one. The v en tral b u n d le of th e d ila ta to r
o perculi m uscle inserts m edially o n th e do rso ro stral edge of th e o p ercu lar bone
an d has its origin on th e lateral surface of th e m edial w ing of th e hy o m an d ib u lar
bone. It is covered by th e d ila ta to r o p ercu li dorsalis m uscle th a t originates at th e
su tu re of th e lateral w ings of th e p arasp h en o id b o n e w ith th e sp h en o tic b o n e and
in serts on th e dorsal process of th e o p ercu lar b o n e by m eans of a ten d o n . For a
m ore th o ro u g h descrip tio n of th e p reo p ercu lar and h y o m an d ib u lar bones in H.
reidi, see ch ap ter 5.1.
T he hyoid arch consists of th e in terh y al, th e a n te rio r an d p o sterio r ceratohyal,
th e hypohyal and th e basihyal bones. A lth o u g h strictly n o t p a rt of it, th e urohyal
bone is also closely associated w ith th e hyoid. T he in terh y al bone (Fig. 3.8A, 3.8C,
3.9A) provides th e a rtic u la tio n betw een th e Suspensorium and th e o th e r elem ents
of th e hyoid, w h ich form a rigid un it. The p o sterio r ceratohyal b o n e (Fig. 3.8A,
3.8C, 3.9A) has an irregular shape w ith a sm o o th dorsal surface th a t articu lates in
b etw een th e tw o heads of th e in terh y al bone. The very sm all a tta c h m e n t site of
th e te n d o n of th e hyohyoideus ab d u cto r m uscle th a t ru n s to th e first of th e tw o
branchiostegal rays is situ a ted at th e caudal surface of th e p o sterio r ceratohyal
bone. T he bone tap ers ventrally w h ere th e re is a firm synchondrosis w ith th e
a n te rio r ceratohyal bone. The a n te rio r ceratohyal bone (Fig. 3.8A, 3.8C, 3.9A) has
a som ew hat trian g u la r shape w ith a sm all b u t firm sym physis b etw een th e left
an d rig h t bones close to th e ir distal apices. Just dorsal of th e sym physis, th e very
m uch reduced hypohyal bones lay to g e th e r w ith th e sm all basihyal bone. They
are all strongly a ttach ed to th e a n terio r ceratohyal b o n e and are alm ost
in c o rp o ra te d by it (not visible on figures). The urohyal b o n e (Fig. 3.8A, 3.8C, 3.9A)
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
63
is a small, b lu n t bone, m easuring less th a n th e to ta l hyoid length. T he sh o rt
te n d o n of th e sternohyoideus m uscle (Fig. 3.8A, 3.8C, 3.9A) encloses th e urohyal
bone alm ost en tirely a t its caudal tip. It is a very large m uscle, w h ich inco rp o rates
its te n d o n en tirely (Fig. 3.10C). M o re details a b o u t th e hyoid of H. reidi can be
fo u n d in ch ap te r 5.1.
T he e th m o id bones form th e dorsal surface of th e sn o u t an d w ith th e exception
of th e lateral ethm oid, th e y are all elongated. The n arro w vom eral b o n e (Fig. 3.8A,
3.8B, 3.9) broadens rostrally w here it articu lates w ith th e m axillary bones
ventrally. D orsally it is co n n ected to a ro stral p a rt of th e eth m o id cartilage th a t
articu lates w ith th e autopalatin e. C audally it reaches th e lateral e th m o id bone
an d is covered by th e m esethm o id b o n e an d th e p arasp h en o id b o n e dorsally. The
m esethm oid bone (Fig. 3.8A, 3.8B, 3.9A) in te rd ig itate s w ith th e lateral eth m o id
bones caudally. T he p araspheno id b o n e (Fig. 3.8A) co n n ects th e e th m o id region
w ith th e occipital region of th e skull. A long th e m ajor p a rt of its le n g th th e
a d d u cto r arcus p a la tin i m uscle originates (Fig. 3.9A). A t th e caudal m argin of th e
m uscle atta ch m e n t, th e parasp h en o id b o n e bears tw o lateral w ings, ju st b eh in d
th e eyes. From th e re an d caudally, th e bone is b e n t upw ards an d fits in to a cavity
of th e basioccipital bone.
T he lateral w all of th e sn o u t is covered by a series of circu m o rb ital bones (Fig.
3.8A, 3.8B). T hey are very variable in n u m b er an d shape. In th e re c o n stru cte d
specim en th e re are th re e bones o n th e rig h t side, w h ich could be considered
hom ologous to th e a n to rb ital, th e lacrim al and th e second in frao rb ital bones. O n
th e le ft side, how ever, th e re are fo u r circu m o rb ital bones, of w h ich th e hom ology
is som ew hat problem atic. The o rb it is b o rd ered by th e lateral eth m o id bones
rostrally, th e p arietal bone dorsally, th e sp h en o tic b o n e caudally and th e
parasp h en o id bone medially. The p a rie tal bone (Fig. 3.8A, 3.8B, 3.9A) is th e largest
bone
of
th e
n eu ro cran iu m ,
in te rd ig itatin g
w ith
th e
m eseth m o id
bone
m ediorostrally an d th e lateral e th m o id bones laterorostrally. C audal to th e eyes,
it has a su tu re w ith th e p ro o tic bone, th e sp h en o tic bone, th e p te ro tic bone, th e
epioccipital bone an d th e supraoccipital bone. T o g eth er w ith th e la tte r, th e
p arie tal bone covers alm ost th e e n tire n eu ro cran ial roof.
T he otic region com prises th e sphenotic, th e pro o tic, th e p te ro tic and th e
p o sttem p o ral bones. The sphen o tic bone (Fig. 3.8A, 3.8B) has a sm all lateral spine
64
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
an d a w ell developed p o sto rb ita l process. From th e v en tral tip of th is process, th e
levator arcus p a latin i m uscle originates (Fig. 3.9A). T he sp h en o tic b o n e form s
ventrally, a t th e su tu re w ith th e p ro o tic bone, one of th e tw o a rtic u la tio n sockets
for th e h y o m andibular bone. The p ro o tic b o n e is situ a te d v en trally and
in te rd ig itate s w ith all su rro u n d in g bones. A t th e v en tral surface of th e p ro o tic
bone, th e a d d u cto r o perculi m uscle has its origin an d attach es o n th e dorsom edial
crest of th e o percular bone. The levator o p ercu li m uscle (Fig. 3.9A) originates
from th e v en tral surface of th e pro o tic, th e p te ro tic an d b asioccipital bones, and
attaches along th e dorsal edge of th e o p ercu lar bone a t th e base of th e
dorsom edial crest. The p te ro tic bone
(Fig. 3.8A, B) is provided w ith
a
cartilaginous socket in w h ich th e caudal a rtic u la tio n h ead of th e hyom an d ib u lar
bone articulates. A part from th is v en tral co m p o n en t, th e p te ro tic b o n e also has a
large lateral c o m p o n en t th a t fits in b etw een th e sp h en o tic, p arietal, ep ioccipital
an d p o sttem p o ral bones. The p o sttem p o ral b o n e (Fig. 3.8A, 3.9A), w h ich also
bears a lateral spine, covers th e n eu ro cran iu m laterocaudally. T here is no
p o stp arietal bone.
All th e elem ents of th e occipital region are co n n ected to each o th e r by serrated
sutures. T he su praoccipital bone (Fig. 3.8A, 3.8B, 3.9A) is w edged in b etw ee n th e
le ft an d rig h t p arietal bones rostrally an d its caudal tap erin g en d to u ch es th e
corona (i.e. th e first p o stcran ial b o n y plate). It has an ascending p rofile so th e
n e u ro c ran iu m is hig h er th a n w ide w h e n view ed caudally. A t th e cen tre of th e
su praoccipital caudal surface, th e supracarinalis m uscle (Fig. 3.9A, 3.10D) attach es
w h ich ru n s to th e n e u ra l arch of th e first v erteb ra and to th e second p o stcran ial
bony plate. It consists of a le ft and a rig h t m uscle b u n d le w hose fibres are m erged
an d h ence are h a rd to separate. Lateral of th e supracarinalis m uscle a ttach m en t,
th e w ell-developed epaxial m uscle attach es by m eans of tw o large, b u t sh o rt
ten d o n s (Fig. 3.10D). The epioccipital b o n e (Fig. 3.8A, 3.8B) lies m edially to th e
p o sttem p o ral bone an d laterally to th e supraoccipital bone. It is su rro u n d e d by
th e p arietal an d p te ro tic bones ro strally and th e exoccipital caudally. The
exoccipital bone (Fig. 3.8A, 3.8B) flanks th e occipital jo in t an d th e foram en
m agnum laterally. It covers th e largest p a rt of th e caudal n eu ro cran ial surface
an d is ligam entously con n ected to th e d o rso ro stral tip of th e cleithrum . The
basioccipital bone (Fig. 3.9A) bears th e occipital jo in t an d stretch es o u t in
b etw een th e le ft an d rig h t exoccipital an d p te ro tic bones v en trally w h ere it
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
65
in te rd ig itate s w ith th e parasp h en o id bone. The occipital jo in t has a ven tro cau d al
o rie n ta tio n , c o n trib u tin g to th e angle b etw een th e head and th e body in th e
seahorse. Synchondroses rem ain p resen t, fo r instance in b etw een th e p ro o tic and
sp h en o tic bone an d at th e b o rd er of th e basioccipital bone.
T he in te ro p e rc u lar bone (Fig. 3.8A, 3.8C, 3.9A) is a plate-like bone th a t lies
v entrally in th e snout, m edially to th e p reo p ercu lar bone. It is co n n ected to th e
la tte r, th e low er jaw an d th e hyoid by m eans of ligam ents described earlier. The
su bopercular bone (Fig. 3.8A, C) is a sm all and very th in b o n e ven tro m ed ial to th e
o percular bone to w h ich it is co n n ected by m eans of connective tissue. The
o percular bone (Fig. 3.8A, B, C) is a large, convexly shaped b o n e th a t tap ers at its
dorsocaudal edge. Its v en trocau d al m argin is ro u n d ed w ith th e tw o slender
branchiostegal rays ru n n in g along th e edge. In b etw een th e in sertio n s of th e
d ila ta to r o perculi an d a d d u cto r o p ercu li m uscles, th e re is a sm all cartilaginous
h y o m andibular a rtic u la tio n laterally. The a tta c h m e n t of th e levator operculi
m uscle (Fig. 3.9A) can be fo u n d along th e dorsal m argin of th e o p ercu lar bone.
This stra ig h t dorsal edge follow s th e lateral ridge of th e n eu ro cran ial floor and
caudally th e tap erin g en d covers th e m edial cleith ru m w h ere th e re is a sm all gili
slit.
B. M o r p h o lo g ic a l d e sc r ip tio n o f D oryrham phus dactyliophorus
T he descrip tio n of th e cran iu m of D. dactyliophorus w ill be lim ited to those
features in w h ich it differs from H. reidi.
T he den tary bone (Fig. 3.11A, 3.11C, 3.12) has a som ew hat m ore com plex shape. It
is longer th a n w ide an d is provided w ith a p ro m in e n t lateral crest th a t gives th e
ro stral end a concave profile w h e n view ed frontally. The co ro n o id process is less
developed. The a d d u cto r m andibulae m uscle com plex (Fig. 3.11A, 3.12) consists of
only tw o m uscle bundles; a v en tral one, th o u g h t to be hom ologous to th e A 24,
an d a dorsal one, probably hom ologous to A 35, th a t a tta c h b o th w ith a very long
te n d o n o n to th e coronoid process of th e d en tary bone. T he A í b u n d le was n o t
4 H o m o l o g y is b a s e d o n s im ila r a t t a c h m e n t s i t e s a s t h e A 2 b u n d le i n H i p p o c a m p u s re id i a n d o n t h e
l a c k o f a d i v i s i o n i n s e r t i n g o n t h e m a x illa r y b o n e s , c h a r a c t e r is t ic f o r A í .
5 H o m o l o g y is b a s e d o n t h e i d e n t i c a l o r ig i n a s t h e A 3 b u n d le i n H i p p o c a m p u s r e i d i a n d o n t h e
v e n t r a l p o s i t i o n o f t h e m u s c l e w i t h r e s p e c t t o A 2 . H o w e v e r , t h e i n s e r t i o n o n t h e c o r o n o id p r o c e s s
o f t h e d e n ta r y b o n e , u n c o m m o n fo r A 3 , m ig h t p o in t to a se c o n d a r y d iv is io n o f A 2 in ste a d .
66
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
found. The fusion w ith th e an g u lo articu lar bone (Fig. 3.11 A, 3.11C, 3.12B) seems
to be m ore firm com pared to th e seahorse, as w ell as it has a larger process
form ing th e p o sterio r p a rt of th e co ro n o id process. The te n d o n of th e p ro tra c to r
hyoidei m uscle (Fig. 3.11C, 3.12A) attach es fu rth e r aw ay fro m th e sym physis
b etw een th e le ft and rig h t d en tary bone. The m uscle is in te rru p te d by dense
connective tissue th re e tim es. Strikingly, th e fibres of th e p ro tra c to r hyoidei
m uscle are enclosed by th e heavily b u ilt m andibulo-hyoid lig am en t (Fig. 3.10E,
3.11C), leaving only th e ro stral and caudal b ifu rcatin g ends free.
T he prem axillary and m axillary bones are id en tical to th e ones in H. reidi, except
for th e dorsom edial process of th e
m axillary bone being sm aller in D.
dactyliophorus (Fig. 3.11A, 3.11B, 3.12).
T he suspensorio-neurocranial a rtic u la tio n is a hinge jo in t consisting of tw o sm all
p rojections
of th e
au to p alatin e
bone
(Fig.
3.11A,
3.11B, 3.12B) w ith
a
cartilaginous p a rt of th e vom eral b o n e in betw een. T he h o riz o n ta l p a rt of th e
ectopterygoid bone (Fig. 3.11A, 3.11B, 3.12B) in D. dactyliophorus is provided w ith
a large la te ro ro stra l process th a t is ab sen t in H. reidi. C audally th e re is a slender
groove in to w h ich th e p alato q u ad rate cartilage an d th e p e rich o n d ral p a rt of th e
q uadrate bone are situ a ted and th e ectopterygoid b o n e m eets th e m etapterygoid
bone dorsocaudally. As all o th e r bones of th e Suspensorium , th e m etapterygoid
bone (Fig. 3.11A, 3.11B) in th is p ip efish is m u ch m ore elo n g ated com pared to H.
reidi. Besides th a t, th e m ost p ro m in e n t difference is th a t th e m etapterygoid bone
in th e form er has a dorsal p a rt an d a lateral p a rt a t an alm ost p erp en d icu lar angle
to each o th e r so th a t th e cross section of th e sn o u t has a rectan g u lar appearance,
com pared to a m ore circular shape in th e seahorse (Fig. 3.10F). U nlike th e
q uadrate bone in H. reidi, th e one in D. dactyliophorus has n o clear d ifferen tia tio n
b etw een th e ascending p erich o n d ral p a rt and th e h o riz o n ta l p late since b o th
p a rts are alm ost sim ilar in h e ig h t (Fig. 3.11). T he te n d o n s of th e ad d u cto r
m andibulae m uscle com plex lie ro strally in a fu rro w of th e q u ad rate bone th a t is
laterally closed by th e lacrim al b o n e and th e m etapterygoid b o n e (Fig. 3.10F).
M ore caudally th e ten d o n s ru n dorsally in th e snout. S om ew hat m ore th a n
halfw ay its length, th e q uadrate b o n e starts to ta p e r an d th e caudal th ird covers
th e hyosym plectic cartilage an d th e sym plectic b o n e laterally w hereas it is
covered itself by th e p reo p ercu lar bone. The h o rizo n tal, m edial groove of th e
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
67
p reo p ercu lar bone (Fig. 3.11), in w h ich th e caudal en d of th e q u ad rate bone and
th e sym plectic bone lies, bulges ou tw ard s o n its lateral side. The A2 m uscle of th e
a d d u cto r m andibulae com plex (Fig. 3.12A) attach es o n th e m edial surface of th e
p reo p ercu lar bone an d caudally o n th e dorsal crest of th e sym plectic bone.
C audally th e dorsal m argin of th e p reo p ercu lar b o n e curves outw ards, form ing
th e ro u n d ed b o rd er of th e o rb it an d in te rd ig ita tin g w ith th e v en tro cau d al p a rt of
th e second in frao rb ital bone. T here is n o p reo p ercu lar spine. The p erich o n d ral
p a rt of th e sym plectic bone (Fig. 3.11A, 3.12A) in D. dactyliophorus is m u ch longer
w ith respect to th e plate-like p art, com pared to th e situ a tio n in th e seahorse.
F u rth erm o re, th e vertical plate is provided w ith a p ro m in e n t d o rso lateral crest
th a t in terd ig itates w ith th e dorsom edial p a rt of th e second in frao rb ital b o n e and
th a t separates th e v en tral a d d u cto r m andibulae m uscle com plex fro m th e dorsal
a d d u cto r arcus p a latin i muscle. B undle A3 of th e a d d u cto r m andibulae m uscle
(Fig. 3.12A) attach es rostrally on th e dorsal crest as in H. reidi. The hyo m an d ib u lar
bone (Fig. 3.12A) is b lu n te r th a n th e one in H. reidi. The m edial and lateral w ings
of th e seahorse, in b etw een w h ich th e levator arcus p a latin i m uscle an d d ila ta to r
o perculi m uscles run, are here obliquely o rien tated , resu ltin g in a ra th e r
ve n tro la te ral an d a dorsom edial w ing. The co n n ec tio n w ith th e p reo p ercu lar
bone is very tig h t as th e hyom an d ib u lar b o n e is provided w ith several lateral
p ro tru sio n s th a t in te rd ig itate w ith th e p reo p ercu lar bone. A p art fro m th a t, th e
h y o m andibular bone and its lig am en t and m uscle a tta ch m e n ts are sim ilar to
th o se in H. reidi.
T he shape an d ligam ent co n fig u ratio n of th e in terh y al bone (Fig. 3.11 A, 3.11C) is
sim ilar in th e tw o species. The p o sterio r ceratohyal b o n e (Fig. 3.11 A, 3.11C,
3.12A) in D. dactyliophorus show s a lo t of in d e n ta tio n s an d p ro tru sio n s. The
m edial ones form th e in te rd ig itatin g p arts w ith th e a n te rio r ceratohyal bone,
en suring a firm connection. Laterally th e re is a knob-like process th a t is also
p re sen t in th e seahorse b u t it is larger in th e pipefish. The in tero p ercu lo -h y o id
ligam ent an d th e te n d o n of th e p ro tra c to r hyoidei m uscle attach , respectively,
rostrally and ju st ventrally to th is process. A t th e caudal surface, v en tral to its
cartilaginous head, th e te n d o n of th e hyohyoideus ab d u cto r m uscle (Fig. 3.11C,
3.12A) originates. This m uscle ru n s along w ith th e b ran ch io steg al rays before it
attaches, so it does n o t a tta c h o n th e ro stral tip of th e b ranchiostegals b u t
68
som ew hat
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
m ore
caudally.
R ostrally
th e
fused
b ran ch io steg al
rays
are
ligam entously a ttach ed to th e p o sterio r ceratohyal bone. The a n te rio r ceratohyal
bone (Fig. 3.11A, 3.11C) has th e same, som ew hat trian g u lar, shape as in H. reidi
b u t is spinier. T he apices of th e le ft and rig h t bones do n o t diverge very m u ch and
th e sym physis is located at th e level of th e apices, so m ore distal com pared to th e
seahorse. The hypohyal bones are q u ite large an d situ a ted a t th e m ost dorsal half
of th e a n terio r ceratohyal bone, alm ost at th e level of th e m andibulo-hyoid
ligam ent a tta c h m e n t (not visible o n figures). In H. reidi th e basihyal bone is really
tiny, in D. dactyliophorus it is a long and slender o ssification a ro u n d th e even
longer cartilaginous basihyal rostrally. The urohyal bone (Fig. 3.11 A, 3.11C,
3.12A) is long; m ore th a n th re e tim es th e hyoid length. The sternohyoideus
m uscle (Fig. 3.11A, 3.11C, 3.12A) is n o t as large as th a t in H. reidi, and th e w ell
developed te n d o n is n o t in co rp o rated by th e m uscle (Fig. 3.10G).
As m e n tio n e d before, th e e th m o id region of D. dactyliophorus is m u ch m ore
elongated com pared to th a t in H. reidi. The vom eral bone consists of a cen tral rod
w ith tw o lateral w ings (Fig. 3.11A, 3.11B). O nce covered by th e m esethm oid bone
caudally, th e w ings disappear w hereas th e cen tral p a rt of th e vom eral b o n e runs
up to th e en d of th e lateral eth m o id bone. R ostrally th e m esethm oid b o n e (Fig.
3.11A, 3.11B) tap ers w hereas in H. reidi its ro stral end bifurcates. B oth vom eral
an d m esethm oid bones show a low dorsal crest along th e ir e n tire len g th , in th e
seahorse th is crest is only p re sen t in th e m esethm oid b o n e a t th e level of th e
nasal cavities. R ostrally, at th e level of th e p arietal bones, th e p arasp h en o id bone
(Fig. 3.11A) lies in b etw een th e dorsal m esethm oid b o n e an d th e v en tral vom eral
bone. T he caudal end of th e p arasp h en o id b o n e is n o t b e n t dorsally as in th e
seahorse, b u t it does also fit in a cavity of th e basioccipital bone.
In D. dactyliophorus th e re are only tw o circu m o rb ital bones: a very sm all lacrim al
bone and a large second in frao rb ital b o n e (Fig. 3.11A, 3.11B). The lacrim al bone
covers th e ten d o n s of th e a d d u cto r m andibulae m uscle com plex an d th e
ascending process of th e quadrate. D orsally it is p artially enclosed by th e
m etapterygoid bone th a t separates th e tw o in frao rb ital bones fro m each other.
T he second in frao rb ital bone (Fig. 3.10F) has an alm ost h o riz o n ta l p a rt and a
vertical one dorsally sim ilar to th e m etapterygoid bone. It covers th e ad d u cto r
m andibulae ten d o n s and th e m etapterygoid b o n e rostrally an d th e ad d u cto r
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
69
m andibulae m uscle caudally. T he h o riz o n ta l p a rt in terd ig itates b o th w ith th e
sym plectic bone an d w ith th e p reo p ercu lar bone caudally w h ere it form s p a rt of
th e b o rd er of th e orbit. In all D. dactyliophorus specim ens studied, th e n u m b er and
shape of th e circu m o rb ital bones is th e same. The p re o rb ita l p a rt of th e p arietal
bone (Fig. 3.11A, 3.11B, 3.12A) is larger in th e p ip efish th a n in th e seahorse and
th e form er does n o t have a spine. The spine of th e sp h en o tic b o n e (Fig. 3.11A) is
also ab sen t b u t a p art from th a t, th e re are alm ost n o differences. The ad d u cto r
o perculi m uscle does n o t o rigin ate fro m th e p ro o tic b o n e b u t from th e p te ro tic
bone (Fig. 3.12A). The p o sttem p o ral bone (Fig. 3.11A, 3.11B) is spineless b u t
show s a p ro m in e n t ridge. In D. dactyliophorus th e p o stp arietal b o n e is also
missing.
T he supraoccipital bone (Fig. 3.10H, 3.11A, 3.11B, 3.12A) has a biconcave profile
in cross section w here it is provided w ith clear a tta c h m e n t sites fo r th e
p o stcran ial muscles. The supracarinalis m uscle consists of a separate le ft and
rig h t b u n d le th a t a tta c h directly in th e concave in d e n ta tio n s of th e supraoccipital
bone. These bundles ru n to th e n eu ral arch of th e first v erteb ra (not
reconstructed) w here th e y a tta c h by m eans of a very sm all ten d o n . T o g eth er w ith
th e epioccipital bone (Fig. 3.11A, 3.11B), th e su p rao ccip ital b o n e encloses a cavity
for th e very solid te n d o n of th e slender epaxial muscle. In terestin g ly , w ith in th e
te n d o n , a very long bony rod is em bedded. The ro stral p a rt of th e ten d in o u s tissue
is in fact a short, b u t firm ligam en t th a t connects th e supraoccipital b o n e to th is
sesam oid bone. M ore caudally th e v e n tro lateral surface of th e sesam oid b o n e is
covered by th e epaxial m uscle an d its ten d o n . The dorsal co m p o n en ts of b o th th e
epioccipital bone an d th e exoccipital b o n e (Fig. 3.11 A, 3.11B) are larger in th e
pipefish, le ft and rig h t bones even in terd ig itate w ith each o th e r m edially, v en tral
to th e su praoccipital bone. The occipital jo in t, form ed by th e basioccipital bone
(Fig. 3.12A), is o rie n te d caudally so th a t th e cran iu m and v erteb ral colum n lie on
th e same axis.
C. O th e r s y n g n a th id sp e c ie s
For purposes of com parison w ith o th e r pipefishes and seahorses only th o se
features th a t m ight be related to sn o u t elo n g atio n are treated.
Hippocampus zosterae (Fig. 3.13A) is a sh o rt-sn o u ted seahorse. The artic u la tio n
b etw een au to p alatin e an d vom eral b o n e involves less cartilage th a n th e one in H.
7o
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
reidi. The d en tary an d an g u lo articu lar bones in te rd ig itate ra th e r loosely and th e
coronoid process is low w ith resp ect to th e le n g th of th e low er jaw long axis. The
a d d u cto r m andibulae ten d o n s are m u ch sh o rte r th a n th e m uscle b u n d les and th e
p ro tra c to r hyoidei m uscle, w h ich is n o t enclosed by th e m andibulo-hyoid
ligam ent, attaches m ediorostrally o n th e dentary. The hyoid is a very solid
stru c tu re w ith a firm ceratohyal symphysis. As is th e case in H. reidi, th e urohyal
le n g th is less th a n th e hyoid le n g th and th e sternohyoideus m uscle is of
su b sta n tia l size. T here is no sesam oid b o n e in th e sh o rt epaxial te n d o n and th e
epaxial m uscle is w ell developed.
In H. abdominalis (Fig. 3.13B), a seahorse w ith in term ed iate sn o u t len g th , th e
an g u lo articu lar and d en tary bones are n o t very firm ly in te rd ig itate d and th e
d en tary bone is provided w ith a h ig h co ro n o id process. The a d d u cto r m andibulae
m uscle com plex ru n s approxim ately in a h o riz o n ta l line caudally and only ab o u t
a fo u rth of its le n g th consists of ju st ten d o n . The a n te rio r p a rt of th e vom eral
bone broadens som ew hat and at its ro stral end it articu lates w ith th e
au to p alatin e bone by m eans of a p o in t articu latio n . Like th e o th e r seahorses, th e
urohyal bone of H. abdominalis is sh o rte r th a n th e hyoid an d n o epaxial sesam oid
bone is present. H ow ever, th e p ro tra c to r hyoidei m uscle is enclosed by th e firm
m andibulo-hyoid ligam ent for alm ost its e n tire len g th , sim ilar to th e situ a tio n in
D. dactyliophorus. The p ro tra c to r hyoidei te n d o n attach es o n th e v entrom edial
surface of th e d en tary bone, ro stral to th e m andibulo-hyoid lig am en t a tta c h m e n t
site. S om ew hat m ore caudally, th e lig am en t in co rp o rates th e m uscle an d only th e
caudal th ird of th e p ro tra c to r hyoidei m uscle is le ft free. H ere, th e le ft and rig h t
bundles diverge an d ru n ven tral to th e lig am en t to respectively th e p o sterio r and
a n te rio r ceratohyal bone, w here th ey attach. The sternohyoideus an d epaxial
m uscles are b o th w ell developed an d have sh o rt tendons.
Corythoichthys intestinalis (Fig. 3.13C) has a sh o rt b u t very slender snout. Its skull
bones are covered w ith m any sm all in d en tatio n s, especially a t th e level of th e
braincase, th e opercular bone and caudally o n th e p reo p ercu lar bone. The
vom eral bone is very slim an d provides a sim ple a rtic u la tio n w ith th e
au to p alatin e bone rostrally. Also th e d en tary b one is slender an d its ro stral p a rt is
b e n t upw ards to alm ost th e same h eig h t as th e w ell developed co ro n o id process.
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
71
T he an g u lo articu lar bone is fairly w ell atta ch e d to th e d en tary bone. The
p ro tra c to r hyoidei m uscle an d m andibulo-hyoid lig am en t ru n separately to th e
p o sterio r an d a n te rio r ceratohyal bones, respectively. T he le n g th of th e ad d u cto r
m andibulae ten d o n s is sm all in com parison to th e m uscle length. The sym physis
b etw een le ft and rig h t a n terio r ceratohyal bones is large, tak in g up alm ost h alf of
th e to ta l hyoid length. The urohyal b o n e is long, being m ore th a n tw o tim es th e
hyoid len g th , an d has a sm all b ifu rca tio n caudally. The stern o h y o id eu s and
epaxial m uscles are sm aller th a n th e ones in H. reidi, b u t th e ir ten d o n s are n o t as
long as in D. dactyliophorus. N evertheless, C. intestinalis has a long an d slender
ossified rod w ith in th e te n d o n of th e epaxial m uscle. H ow ever, th e ligam entous
p a rt th a t connects th e su praoccip ital b o n e w ith th is rod is sh o rte r th a n in D.
dactyliophorus. Like all pipefishes, C. intestinalis has a caudally o rie n te d occipital
join t, w ith no angle betw een th e head and th e body.
In D. melanopleura (Fig. 3.13D), also a sh o rt-sn o u ted pipefish, th e sn o u t and
m axillary bones are covered w ith sm all p rotuberances. The occipital jo in t, w h ich
has a caudal o rie n ta tio n , is elliptically shaped an d q u ite broad. The low er jaw in
D. melanopleura has a fairly solid co n n ec tio n b etw een th e separate elem ents an d a
relatively h igh coronoid process. T he p ro tra c to r hyoidei m uscle is n o t enclosed
by th e m andibulo-hyoid lig am en t in D. melanopleura. The te n d o n s of th e
sternohyoideus and th e epaxial m uscle are all relatively long. The tw o m uscles
also have a sm aller d iam eter th a n th e m uscles of th e seahorses. Doryrhamphus
melanopleura, like D. dactyliophorus an d C. intestinalis, has a long and slender
sesam oid bone w ith in th e te n d o n of th e epaxial m uscle. The urohyal b o n e is long,
exceeding th e hyoid th re e tim es in len g th , and th e caudal b ifu rca tio n is large,
alm ost reaching th e cleithrum .
T he cran iu m of th e lo n g-snouted D. janssi and D. pessuliferus is q u ite sim ilar to
th e one in D. dactyliophorus. T he low er jaw of D. janssi looks like th e one in D.
dactyliophorus, w ith a low coro n o id process an d ro b u st c o n n ectio n b etw een
d en tary and an g u lo articu lar bone. H ow ever, th e m andibulo-hyoid lig am en t does
n o t enclose th e p ro tra c to r hyoidei m uscle. In D. pessuliferus th e low er jaw
in te rd ig ita tio n is also firm an d in addition, th e p ro tra c to r hyoidei m uscle is
enclosed by th e m andibulo-hyoid ligam ent. In b o th species, th e p a rt of th e
72
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
a d d u cto r m andibulae com plex th a t consists of solely te n d o n is approxim ately
h alf th e to ta l le n g th and th e m uscle ru n s dorsally in th e snout. The h in g e-jo in t
b etw een th e au to p alatin e and th e vom eral bone fo u n d in D. dactyliophorus is a
sim ple p o in t-jo in t in th ese tw o species here. All th re e species have a very long
urohyal bone, b u t in b o th D. pessuliferus and D. janssi, it is a forked, very slender
rod. In th e la tte r th e urohyal bone keeps ru n n in g in b etw een th e hypaxial m uscle
bundles even w ell beyond th e p ecto ral fin, reaching a le n g th of over six tim es th e
hyoid len g th , w h ich is alm ost as long as th e e n tire snout. In D. pessuliferus th e
urohyal le n g th is only som ew hat m ore th a n tw o tim es th e hyoid length. T heir
sternohyoideus m uscle, w h ich is n o t very large, has a w ell developed ten d o n ,
sim ilar to th e situ a tio n fo u n d in D. dactyliophorus. B oth D. pessuliferus an d D.
janssi possess a long an d slim sesam oid b o n e in th e long epaxial ten d o n .
3.2.4. D is cu ssi on
Pipefishes and seahorses are pivot feeders; th ey cap tu re prey by m eans of a rapid
d o rso ro ta tio n of th e head in co m b in atio n w ith an equally fast expansion of th e
buccal cavity. S yngnathid pivot feeding starts w ith hyoid ro ta tio n , w h ich is
follow ed by n eu ro cran ial elevation (b o th m ovem ents are coupled in a fo u r-b ar
m echanism ; ch ap ter 5.1; M uller, 1987; V an W assen b erg h et al., 2008). N ext, th e
m o u th is opened, th e sn o u t is expanded and th e prey is sucked in a fte r only 6 ms
approxim ately (B ergert & W ain w rig h t, 1997; de Lussanet & M u ller, 2007; Roos et
al., 2009). The specialized feeding strategy is h y p o th esized to be associated w ith
m orphological m odifications an d in n o v atio n s of th e feeding apparatus, especially
w ith respect to th e snout. T herefore, th e m orphology of th e m ain cranial
elem ents w ill be discussed in a fu n c tio n a l co n tex t first. In th e n ex t p a rt of th e
discussion, th e stru c tu ra l v ariatio n in th e cran iu m of all stu d ied syngnathids is
considered in an a tte m p t to assess th e im p licatio n of v ariatio n in sn o u t le n g th on
th e head m orphology.
A. S p e c ia liz e d fe e d in g s tr a te g y
T he studied sy ngnathid species all have a fused d en tary and a n g u lo articu lar bone,
as w ell as a saddle-shaped q u ad rato -m an d ib u lar jo in t an d a firm b u t m obile
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
73
d en tary symphysis. The hyoid c o n fig u ra tio n resem bles th e one of th e low er jaw
in having a solid synchondrosis b etw een th e a n te rio r and p o sterio r ceratohyal
bones, a saddle-shaped interhyo-cerato h y al artic u la tio n an d a stro n g an d flexible
ceratohyal symphysis. Besides th e special low er jaw an d hyoid m orphology, th e
m axillary bones are m uch reduced in size and th e p ecto ral girdle is im m ovably
co n n ected to th e vertebrae. A lth o u g h th ese tra its are n o t necessarily syngnathid
synapom orphic characters, th ey m ig h t be related to th e pivot feeding. These
features an d th e ir assum ed role d u rin g su ctio n feeding are elab o rated on, based
on th e m orphology of H. reidi an d D. dactyliophorus.
All th re e elem ents of each low er jaw h alf (dentary, an g u lo articu lar and
re tro a rtic u la r bones) fu n c tio n as a single u n it th a t articu lates by m eans of a
saddle jo in t w ith th e Suspensorium . This type of a rtic u la tio n p erm its only
ro ta tio n
in
a
sagittal
plane,
resu ltin g
in
m o u th
o p en in g
an d
closing.
T heoretically, low er jaw depression can be accom plished in th re e d ifferen t ways:
(1) th e m andibulo-hyoid ligam en t tran slates hyoid ro ta tio n in to m o u th opening,
(2) th e m an d ib u lo -in tero p ercu lar and in tero p ercu lar-h y o id ligam ents fo rm a
second coupling b etw een hyoid ro ta tio n and low er jaw depression, and (3)
c o n tra c tio n of th e p ro tra c to r hyoidei m uscle w ill also o p en th e m o u th (note th a t
an o percular four-bar m echanism fo r depressing th e low er jaw (e.g. A nker, 1974;
W estn e at, 1990; H u n t von H erb in g et al., 1996b; V an W assen b erg h et al., 2005) is
n o t fu n c tio n a l in syngnathids since th e re is n o c o n n ectio n b etw een th e o p ercu lar
an d in te ro p e rc u lar bones). In w h a t follow s, th ese th re e low er jaw depressing
m echanism s are discussed in m ore detail.
(1) The very w ell developed m an d ibulo-hyoid lig am en t attaches m edially o n th e
d en tary bone, v e n tro ro stra l to th e q u ad rato -m an d ib u lar a rtic u la tio n (so below
th e axis of ro tatio n ) and runs to th e a n te rio r ceratohyal bone. These a tta ch m e n t
sites differ from th e general a ctin o p tery g ian co n fig u ratio n of such a m andibulohyoid ligam ent, w here it connects th e re tro a rtic u la r process of th e low er jaw
w ith th e p o sterio r ceratohyal bo n e (e.g. Oncorhynchus mykiss (W albaum ) (Verraes,
3.977), Polypterus seneyalus Cuvier, Lepisosteus oculatus W in ch e ll an d Am ia calva L.
(Lauder, 1980a; Lauder & N o rto n , 1980), Ateleopus japonicus Bleeker (Sasaki et al.,
2006)). V erraes (1977) suggested th a t in Gasterosteus aculeatus (w hich belongs to
th e same order as th e Syngnathidae; N elson, 2006) th e h y alo-m andibular lig am ent
described by A nker (1974) is n o t hom ologous to th e m andibulo-hyoid lig am en t of
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
74
o th e r fishes. The m andibulo-hyoid lig am en t is th e n ab sen t or it could have
becom e sp lit up in a separate m an d ib u lo -in tero p ercu lar an d in tero p ercu lo -h y o id
ligam ent (Stiassny, 1996). This m ig h t also be tru e fo r syngnathids, since th e
a tta c h m e n t sites of th e m andibulo-hyoid lig am en t resem ble th e ones of th e
hyalo-m andibular ligam ent in
G. aculeatus. A n arg u m en t su p p o rtin g this
hypothesis is th e above m e n tio n e d m a n d ib u lo -in tero p ercu lar and in tero p ercu lo hyoid ligam ents being also p re sen t in syngnathids.
(2) The m andibulo-opercular ligam en t co n n ects th e re tro a rtic u la r b o n e to th e
in te ro p e rc u lar bone, w h ich is in its tu r n co n n ected to th e ceratohyal p o sterio r by
m eans of th e in teroperculo-hyoid ligam ent. O bviously th ese th re e ligam ents
(m andibulo-hyoid,
m an d ib u lo -in tero p ercu lar
an d
interopercular-hyoid)
w ill
couple hyoid to low er jaw depression.
(3) T here is also th e possibility of jaw depression th ro u g h c o n tra ctio n of th e
p ro tra c to r hyoidei m uscle. W h e n th e m o u th of H. reidi is closed, th e line of
actio n of th e p ro tra c to r hyoidei m uscle lies dorsal to th e q u ad rato -m an d ib u lar
join t, so c o n tra ctio n w ill n o t resu lt in low er jaw depression, b u t in stead th e
m uscle w ill probably be stra in ed an d elastic energy w ill be sto red (sim ilar to th e
c a ta p u lt m echanism in th e epaxial m uscle of syngnathids (Van W assen b erg h et
al., 2008)). B oth H. reidi an d D. dactyliophorus specim ens used fo r th e 3Dre c o n stru ctio n s have th e ir hyoid an d low er jaw depressed, b u t th e m andibulohyoid lig am en t does n o t seem to be fully extended, as it is curved v en trally along
th e hyoid. This could suggest th a t th e lig am en t c o n trib u te s only to th e in itial
phase of m o u th opening, resu ltin g in a low ering of th e d en tary (and th u s also of
th e p ro tra c to r hyoidei te n d o n a tta c h m e n t site) u n til th e m uscle line of action
shifts to below th e low er jaw articu latio n . F u rth e r low er jaw depression could
th e n be th e consequence of a sudden release of th e stra in energy sto red in th e
p ro tra c to r hyoidei m u scle-tendo n system. The m andible w o u ld th e n ro ta te
fu rth e r v entrally due to its ow n in ertia, causing th e slack statu s of th e
m andibulo-hyoid ligam ent. The te n d in o u s in te rru p tio n s of th e p ro tra c to r
hyoidei m uscle w o u ld be especially beneficial since ten d o n s are b e tte r su ited for
sto rin g elastic strain energy th a n m uscle tissue is (A lexander & B ennet-C lark,
1977)T he m o u th is closed again th ro u g h th e c o n tra ctio n of th e a d d u cto r m andibulae
com plex, in co m b in atio n w ith th e ro ta tio n of th e hyoid to its restin g p o sitio n
C h a p t e r 3 - Morphological description
75
an d th e relaxation of th e m an d ib u lar ligam ents an d p ro tra c to r hyoidei muscle.
Low er
jaw
a d d u ctio n
is
also
accom plished
by
c o n tra ctio n
of
th e
in term an d ib u laris m uscle th a t attaches close to th e d en tary sym physis and
perhaps in d irectly by c o n tra ctio n of th e a d d u cto r arcus p a latin i m uscle, w h ich
w ill ad d u ct th e suspensoria (the la tte r m uscle m ig h t also be responsible fo r th e
in itia l locking of th e cranial fo u r-b ar system p rio r to n eu ro cran ial elevation
(Roos et al., su b m itte d b)).
Syngnathids do n o t have a typical tele o ste a n hyoid c o n fig u ra tio n w ith a ball and
socket jo in t b etw een th e in terh y al and th e h y o m an d ib u lar bones. Instead, th e
in terh y al bone has a ro u n d ed dorsal head th a t articu lates w ith th e p reo p ercu lar
bone an d tw o large v en tral processes th a t fo rm a saddle-shaped a rtic u la tio n w ith
th e p o sterio r ceratohyal bone. As in th e low er jaw, th is type of jo in t w ill lim it th e
tra n slatio n al an d ro ta tio n a l degrees of freedom of th e a rtic u la tio n an d may
facilitate m uscle actio n to be tra n sfe rre d strictly u n to hyoid depression. This may
reduce th e risk of dislocation d u rin g fast an d p o w erfu l hyoid depression,
pow ered by th e c o n tra c tio n of th e stern o h y o id eu s m uscle. A t first, all elem ents of
th e hyoid w ill ro ta te as a single rigid u n it w ith respect to th e p reo p ercu lar bone.
H ow ever, a t a c ertain angle, th is ro ta tio n seems to becom e restrain ed and from
th a t p o in t on, th e in terh y al bon e rem ains im m obile an d th e p o sterio r ceratohyal
bone starts artic u la tin g w ith respect to it. The angle over w h ich th e hyoid tu rn s
durin g su ctio n feeding is m ore th a n 900. H ence, once b eyond th e p o in t of 900,
th e re is no d irect c o n trib u tio n of hyoid depression to buccal volum e increase
th ro u g h a low ering of th e buccal floor (Roos et al., 2009). Indirectly, how ever,
hyoid depression does c o n trib u te to oral expansion via suspensorial abd u ctio n ,
w h ich is th e m ain cause of buccal volum e increase (Roos et al., 2009). W h e n th e
hyoid is p ro trac te d , th e tw o a n te rio r ceratohyal bones lay at an angle w ith respect
to each o th e r in a w ay th a t b o th hyoid bars and th e sym physeal hinge axis do n o t
fall in th e sam e plane. C o n tra c tio n of th e sternohyoideus m uscle w ill th e re fo re
n o t only resu lt in hyoid depression, b u t also in ro ta tio n of th e hyoid bars a t th e ir
sym physis (Aerts, 1991). This ro ta tio n w ill move le ft an d rig h t hyoid o utw ards
along th e ir long axis, hence, th e suspensoria w ill be abducted. This m ig h t be
aided by c o n tra ctio n of th e levator arcus p a latin i muscle.
In all species studied, th e
m axillary bones are very m u ch reduced and
prem axillary bones are toothless. T h eir c o n trib u tio n to th e buccal volum e
76
C h a p t e r 3 - Mor phological description
increase d u ring su ctio n w ill n o t be su b stan tial, n o r are th e y used fo r grasping th e
prey since prey is exclusively cau g h t by m eans of suction. G iven th e absence of an
ascending process of th e prem axillary bone, no real p ro tru sio n of th e u p p er jaw
seems possible. H ow ever, th e co n fig u ratio n does allow ro ta tio n of b o th m axillary
bones w ith respect to th e vom eral bone. W h e n th e low er jaw depresses, th e
m axillary bones sw ing rostrally due to th e p rim o rd ial lig am en t th a t co n n ects th e
v en tral surface of th e m axillary bo n e to th e co ro n o id process of th e d en tary bone.
This ro stral sw ing could h elp closing th e m o u th op en in g laterally, giving it a
m ore circular gape, w h ich creates a m ore rostrocaudally o rie n te d w a te r flow
(A lexander, 1967c; Lauder, 1979; 1985; A erts & V erraes, 1987). The m edially
situ a ted ro stral cartilage, also fo u n d in Gasterosteus aculeatus (Anker, 1974) and
for exam ple in c y p rin o d o n tifo rm fishes (H ern an d ez et al., 2009), could be a
rem in iscen t featu re th a t tran slates th e ro ta tio n of th e m axillary bones in to
m ovem ent of th e prem axillary bones, sim ilar to th e k in e th m o id in cyprinid fishes
(H arrin g to n , 1955; A lexander, 1966; 1967a; M o tta , 1984; H ern an d ez et al., 2007).
T here is a sm all overlap b etw een th e cleith ru m of th e p ecto ral girdle o n th e one
h a n d an d b o th exoccipital an d p o stte m p o ra l bones o n th e o th e r h a n d (a
su p ra cle ith ru m is absent). Since th e p ecto ral girdle is firm ly a ttach ed to th e
v erteb ral colum n, th is overlap form s tw o extra a rticu latio n s (left and right)
b etw een th e cran iu m an d th e v erteb ro -p ecto ral com plex, besides th e occipital
joint. These th re e a rtic u la tio n p o in ts all lay o n th e same axis, so fast n eu ro cran ial
elevation w ith a reduced risk of dislocation is possible. In m ost actinopterygians,
th e p ecto ral girdle consists of a relatively sm all scapulacoracoid (or a scapula and
coracoid) th a t su p p o rts th e fin an d is attach ed to th e cleith ru m , w h ich is
co n n ected to a series of o th e r derm al bones inclu d in g th e p o sttem p o ral one
(G osline, 1977; Starck,
1979; M cG onnell, 2001). U nlike th e
situ a tio n
in
syngnathids, in m ost actinopteryg ian s th e re is n o c o n n ectio n b etw een th e girdle
an d th e vertebral colum n an d m ost elem ents of th e sh o u ld er girdle articu late
w ith each o ther, resu ltin g in a very m obile stru c tu re (Gosline, 1977). In G.
aculeatus, th e re
is
a su p ra cle ith ru m
th a t
is
m ovably
co n n ected
to
th e
p o sttem p o ral bone rostrally and to th e cleith ru m caudally (Anker, 1974). In
addition, a sm all costa an d connective tissu e provides th e co n n ex io n b e tw ee n th e
cleith ru m an d th e transverse process of th e first v erteb ra (Anker, 1974). This
situ a tio n , i.e. th e reduced m obility com pared to th e typical actin o p tery g ian
C h a p t e r 3 - Morphological description
77
co n fig u ratio n , may be regarded as th e plesiom orphic co n d itio n to th e syngnathid
configu ration. O n th e supraoccip ital bone, th e supracarinalis m uscle an d th e
te n d o n of th e epaxial m uscle attach. The epaxial m uscle an d its te n d o n are
involved in a pow er-am plifier system , w h ich m akes th e extrem ely high-velocity
ro ta tio n of th e head in syngnathids possible (Van W assen b erg h et al., 2008). The
supracarinalis m uscle is con n ected to th e first v erteb ra an d m ig h t also be
associated w ith n eu ro cran ial elevation d u rin g su ctio n feeding.
B. V a ria tio n rela ted to sn o u t le n g th
T he m orphology of th e low er jaw an d hyoid, as previously described, is very
sim ilar in all stu d ied species, regardless of th e ir sn o u t phen o ty p e. This
conservative
m orphology
is precisely w h a t w as
expected
in
th ese
high
perform ance su ctio n feeders, as even a tin y change in th e com plex in teg rated
system could have a negative effect on th e feeding perfo rm an ce (Adriaens &
H errei, 2009). H ow ever som e v ariatio n related to differences in sn o u t le n g th was
expected, since an elo n g atio n of th e sn o u t is th o u g h t to in flu en ce th e efficiency
of th e m echanics of th e lever system s in th e head.
Increase in sn o u t le n g th im plies th e re has to be an elo n g atio n of e ith e r th e
a d d u cto r m andibulae te n d o n or m uscle complex. Since p o w erfu l jaw closure is
n o t req u ired in su ctio n feeding, elo n g atio n of th e te n d o n in stead of th e m uscle is
energetically m ore efficient. A n ad d itio n al advantage of te n d o n over m uscle
elo n g atio n is th a t te n d in o u s tissu e occupies less space. This is reflected in th e jaw
m uscle m orphology of th e species w ith th e longest snouts: in D. dactyliophorus, D.
janssi an d D. pessuliferus th e ten d o n s of th e ad d u cto r m andibulae are all very long,
w ith approxim ately h alf of th e distance b etw een origin an d in sertio n being
covered by te n d o n alone. Besides th a t, elo n g atio n of th e a d d u cto r m andibulae
com plex is expected to be associated w ith changes in th e co ro n o id h eig h t, o n to
w h ich th e m uscle attaches, or in th e m uscle line of actio n to m a in ta in kin em atic
efficiency. The angle b etw een th e axis co n n ectin g th e jaw jo in t w ith th e
a tta c h m e n t site of th e a d d u cto r m andibulae m uscle o n th e co ro n o id process (i.e.
th e in-lever arm) an d th e line of actio n of th e m uscle w ill decrease w ith
increasing sn o u t length. This w ill reduce th e m echanical advantage of th e lever
system. The m echanical advantage increases w ith elo n g atio n of th e in-lever arm,
hence, a h igher coronoid process is h y p o th esized to be p re se n t in lo n g -sn o u ted
78
C h a p t e r 3 - Mor phological description
species. The coronoid process is low w ith respect to th e le n g th of th e low er jaw
in b o th short- and lo n g-snouted species (D. dactyliophorus, D. janssi, C. intestinalis
an d H. zosterae), w hereas o thers (H. reidi, H. abdominalis, D. melanopleura and D.
pessuliferus) have a relatively h ig h co ro n o id process. T herefore, th e hypothesis
does n o t h old tru e. H ow ever, th e p a th of th e ad d u cto r m andibulae m uscle is m ore
arched in th e lo n g-snouted pipefishes (D. dactyliophorus an d D. pessuliferus), and
th e m uscle is a t th e sn o u t tip v en trally su p p o rte d by a ridge of th e q u ad rate th a t
probably prevents it from low ering d u rin g co n tractio n . H ence, th e m uscles line
of actio n a t th e level of th e a tta c h m e n t site o n th e d en tary is o rie n ta te d m ore
dorsally. This w ill likely resu lt in an increased angle w ith th e in-lever arm and
could hence increase th e m echanical advantage of th e low er jaw lever system.
If th e p ro tra c to r hyoidei m uscle is stra in ed p rio r to m o u th opening, as suggested
in th e first p a rt of th e discussion and based on k in em atic data (Roos et al., 2009),
to rq u e m ight be induced b etw ee n th e separate elem ents of th e low er jaw. This
could be m ore of an issue in lo n g-sn o u ted species, since m ore elastic stra in energy
can be stored in th e long p ro tra c to r hyoidei m uscle and its ten d o n . T herefore, an
increase in th e am o u n t of in te rd ig ita tio n and ligam entous c o n n ectio n b etw een
th e an g u lo articu lar an d d en tary bones w ith increasing sn o u t le n g th was
hypothesized. In seahorses, th e d en tary -an g u lo articu lar in te rd ig ita tio n ranged
from loosely (H. zosterae) over in term ed iate (H. abdominalis) to ra th e r firm (H.
reidi) w ith increasing sn o u t leng th , su p p o rtin g th e hypothesis. Also, all lo n g ­
sn o u ted pipefishes seem to have a tig h te r c o n n ectio n b etw een th e elem ents of
th e low er jaw th a n som e sh o rte r sn o u ted species do, like C. intestinalis and
Synynathus rostellatus (chapter 3.1). H ow ever, a n o th e r sh o rt-sn o u te d pipefish, D.
melanopleura, has q u ite a firm ly in te rd ig itate d low er jaw as well.
T he sternohyoideus m uscle is c o n flu e n t w ith th e hypaxial m uscle, w h ich plays an
im p o rta n t role in th e coupling of hyoid ro ta tio n and n eu ro cran ial elevation
(chapter 5.1; M uller, 1987; V an W assen b erg h et al., 2008). B efore prey cap tu re, th e
epaxial an d hypaxial m uscles c o n tra ct an d ten sio n is b u ilt up in b o th ten d o n s
(Van W assen b erg h et al., 2008). This energy is suddenly released w h e n th e fourb ar m echanism is unlocked, resu ltin g in p o w erfu l m ovem ents of th e hyoid and
head. S n o u t elongation w ill increase th e m o m en t of in e rtia of th e head, hence, a
larger pow er o u tp u t is needed to overcom e th is d u rin g head ro tatio n . It is
th e re fo re h ypothesized th a t th e epaxial an d stern o h y o id eu s m uscles of a lo n g ­
C h a p t e r 3 - Morphological description
79
sn o u ted syngnathid w ill e ith e r have a larger cross section or a lo n g er te n d o n to,
respectively, generate extra force and store extra energy fo r head ro tatio n . In all
seahorses, regardless of th e ir sn o u t len g th , b o th epaxial an d sternohyoideus
m uscle have a sh o rt te n d o n an d a large diam eter. O n th e o th e r hand, all
pipefishes, w ith exception of C. intestinalis, have slender b u t very long
sternohyoideus an d epaxial te n d o n s an d m uscles. Thus, n o tre n d of e ith e r
increase or decrease in te n d o n le n g th or m uscle d iam eter w ith increasing sn o u t
le n g th was observed. Yet, th e re appears to be a p h y logenetic p a tte rn (pipefishes
versus seahorses) in head ro ta tio n tech n iq u e. The long te n d o n an d sm all diam eter
of th e sternohyoideus m uscle in th e pipefishes seems beneficial fo r g en eratin g
th e req u ired pow er. In th e seahorses, th e large m uscle cross section m ay indicate
th a t th e m ain pow er for head ro ta tio n is g en erated by m uscle co n tra ctio n (as
opposed to release of sto red stra in energy). P o ten tially , th e b e n t neck in th e
seahorses prevents th e optim al fu n c tio n in g of th e elastic recoil m echanism in th e
sternohyoideus and hypaxial m uscles as fo u n d in pipefishes. H ow ever, C.
intestinalis seems to have an in term ed iate co n fig u ratio n , w ith ra th e r sh o rt te n d o n
lengths an d larger epaxial and stern o h y o id eu s muscles. The sim ilarity w ith
seahorse m uscle m orphology m ig h t be re lated to th e ir tro p h ic behaviour, sit-andw a it feeding strategy ra th e r th a n active p u rsu it of th e prey, as in o th e r pipefishes.
K inem atical data of th e p a th travelled by th e m o u th d u rin g feeding also show s a
convergence betw een seahorses and Corythoichthys intestinalis (Van W assen b erg h
et al., 2011). Besides th a t, a recen t phylogeny (Teske & Beheregaray, 2009) suggests
a closer relatio n sh ip b etw een Corythoichthys an d Hippocampus species th a n for
instan ce b etw een th e la tte r an d th e genus Synynathus.
Finally, th e a rtic u la tio n betw een th e a u to p alatin e b o n e an d th e eth m o id cartilage
a t th e level of th e vom eral bone is m erely a p o in t a rtic u la tio n in H. reidi.
H ow ever, in D. dactyliophorus th e a u to p alatin e b o n e is provided w ith tw o sm all
lateral processes in betw een w h ich th e vom eral b o n e articu lates, th u s form ing a
hinge joint. This co n fig u ratio n p erm its m o tio n w ith only one ro ta tio n a l degree
of freedom , w h ich m ig h t reduce th e risk of dislocation d u rin g suspensorial
abduction. The extrem ely long sn o u t in D. dactyliophorus m ay th e n explain th e
necessity for such a stro n g er join t, as to rq u e forces exerted on th e sn o u t may be
m ore su b stan tial d u ring sn o u t abduction. N o n e of th e o th e r syngnathids (H.
8o
C h a p t e r 3 - Mor phological description
zosterae, H. abdominalis, D. melanopleura, C. intestinalis, D. janssi and D. pessuliferus)
have th e same hinge-like a rtic u la tio n b etw een au to p alatin e and vom eral bones.
A p art from th e h ypothesized differences, som e o th e r featu res are w o rth
m entioning. O ne of th o se is th e m andibulo-hyoid lig am en t th a t encloses th e
p ro tra c to r hyoidei m uscle entirely, th u s form ing a sh ea th fo r th e muscle. This
strik in g c o n fig u ra tio n is fo u n d in b o th long- an d sh o rt-sn o u ted pipefishes (D.
dactyliophorus, D. pessuliferus and C. intestinalis) and in H. abdominalis, b u t it is
ab sen t in th e o thers (D. janssi, D. melanopleura, H. reidi an d H. zosterae). H ence, it
is n o t related to sn o u t length, n e ith e r is it ch aracteristic fo r pipefishes. The
fu n c tio n a l advantage of th is co n fig u ratio n , how ever, rem ains unclear.
A n in te re stin g finding related to th e epaxial m uscle is th e presence of a long
sesam oid bone em bedded in to its very firm te n d o n in all pipefishes studied.
Usually, sesam oid bones are form ed in te n d o n s th a t pass over a jo in t w h ere th ey
p ro te c t th e te n d o n from dam age an d w h ere th ey could im prove th e m echanical
efficiency of th e related m uscle (Sarin et al., 1999; H all, 2005). M in eralisa tio n of
an energy-storing te n d o n w o uld be a bad th in g since it w o u ld significantly
reduce th e elasticity (C urrey, 2010). This loss of p o te n tia l strain energy storage
capacity w ill negatively affect th e epaxial c ata p u lt perform ance. P otentially, th e
elastic energy necessary for th e p o w er en h an cem en t is sto red in th e sm all b u t
firm epaxial ligam ent co n necting th e supraoccipital b o n e w ith th e te n d o n bone.
T he fact th a t th is ligam ent is sh o rte r in th e tw o p ipefishes w ith a sh o rt sn o u t
su p p o rts th is hypothesis since less stra in energy is req u ired fo r pivot feeding in
sh o rt-sn o u te d species as th e m o m en t of in e rtia is low er. The epaxial sesam oid
bone is fo u n d in none of th e seahorses. Again, th is m ig h t be explained by th e
seahorse’s tilte d head, due to w h ich th e epaxial te n d o n is curved. The presence of
a long an d rigid elem en t w ith in th is te n d o n w o u ld pro b ab ly im pede force
tran sfer, since th e curved line of actio n of th e epaxial m uscle w o u ld n o t
co rrespond to th e stra ig h t long axis of th e sesam oid bone. Still, th e epaxial
sesam oid bone is n o t a pip efish specific ch aracter, since it w as n o t fo u n d in S.
rostellatus (chapter 3.1). M ore com parative m orphological an d k in em atic analyses
w ill be req u ired before any conclusions can be m ade o n th e fu n c tio n a l
significance of th e presence or absence of th is sesam oid bone.
Also, a p ro m in e n t phylogenetic difference in urohyal le n g th w as observed. The
seahorses studied here all have a sh o rt urohyal bone, w ith a le n g th of less th a n
C h a p t e r 3 - Morphological description
81
th e hyoid le n g th an d lacking a b ifu rcatio n . In th e pipefishes th e urohyal b o n e is
very m uch elongated w ith respect to th e hyoid le n g th (to even over six tim es as
long in D. janssi). O f all pipefishes, th e urohyal b o n e w as sh o rte st in S. rostellatus,
w ith a le n g th of a b o u t tw ice th a t of th e hyoid (chapter 3.1). T he urohyal le n g th
does n o t seem to be co rrelated w ith sn o u t le n g th as b o th sh o rt-sn o u te d and lo n g ­
sn o u ted pipefishes have a very long urohyal bone. W ith th e ex ception of D.
dactyliophorus and S. rostellatus, th e p ip efish urohyal b o n e has a b ifu rcatin g caudal
end, w h ich can be small, as in C. intestinalis, or w ell developed, as in D.
melanopleura an d D. pessuliferus, or it can consist of tw o exceptionally long and
slender diverging rods, as in D. janssi. A nalogous to th e epaxial sesam oid b one, th e
absence of an elongated urohyal in seahorses m ay be related to th e tilte d p o sitio n
of th e head w ith respect to th e body (m aking th e distance along a stra ig h t line
b etw een th e hyoid an d th e pecto ral girdle to o short).
In conclusion, som e v ariatio n in low er jaw rigidity an d a d d u cto r m andibulae
m uscle line of actio n was found, w h ich could be sn o u t le n g th specific. H ow ever,
m ost strik in g
features, like th e
m andibulo-hyoid lig am en t enclosing th e
p ro tra c to r hyoidei m uscle, th e presence of an epaxial sesam oid b o n e an d th e
elo n g atio n of th e urohyal bone in som e pipefishes or th e w ell developed
sternohyoideus m uscle in seahorses, could n o t be explained by v ariatio n in sn o u t
length. These tra its may ra th e r be associated w ith d iffe re n t p ow er g en eratin g
strategies, nam ely storage an d release of elastic energy versus p u re m uscle force.
H ow ever,
sim ilarity
of
c ertain
features
m ig h t
also
be
associated
w ith
phylogenetic relatedness of th e species; possibly n o t all shared tra its have a pu rely
fu n c tio n a l base (see also discussion of ch ap te r 4).
82
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
83
4
■
Morphological variation
M o d i f i e d fr o m :
L e y s e n H , R o o s G , A d r ia e n s D .
M o r p h o l o g i c a l v a r ia t i o n i n p i p e f i s h a n d s e a h o r s e h e a d
s h a p e i n r e l a t i o n t o s n o u t l e n g t h a n d d e v e lo p m e n t a l
g r o w th
J o u r n a l o f M o r p h o lo g y , s u b m itte d
A bs tract
T he feeding apparatus of S yngnathidae, w ith its elongate tu b u la r sn o u t and tiny,
to o th less jaws, is highly specialized fo r p e rfo rm in g fast an d p o w erfu l pivot
feeding. In addition, th e prolong ed sy n g n ath id p a re n tal care probably enables th e
juveniles to be provided w ith a feeding a p p aratu s th a t resem bles th e one in
adults, b o th in m orphology an d fu n ctio n . In th is stu d y a landm ark-based
geom etric m o rp h o m etric analysis w as carried o u t o n th e h ead of syngnathid
representatives in order to (l) exam ine to w h a t degree p ip efish shape v ariatio n is
d ifferen t from th a t of seahorses; (2) d eterm in e w h e th e r th e h ig h level of
specialization reduces th e a m o u n t of in trasp ecific m orphological v ariatio n fo u n d
w ith in th e family; an d (3) elucidate w h e th e r or n o t im p o rta n t shape changes
occur in th e seahorse head durin g post-release ontogeny. W e fo u n d th a t (1) th e re
is a sig nificant shape difference b etw een p ip efish an d seahorse head shape: th e
m ain differences concern sn o u t le n g th and heig h t, p o sitio n an d o rie n ta tio n of
84
C h a p t e r 4 - Morphological variation
th e p ecto ral fin base and h e ig h t of th e head and o p ercu lar bone. W e hypothesize
th a t th is m ight be related to d iffe re n t prey cap tu re k in em atics (long sn o u t w ith
little head ro ta tio n versus sh o rt sn o u t w ith large h ead ro tatio n ) and to d ifferen t
body p o stu re (in line w ith th e head versus vertical w ith a tilte d head) in pipefishes
an d seahorses; (2) b o th pipefishes and seahorses show ed an inverse relatio n
b etw een relative sn o u t le n g th and in trasp ecific v ariatio n an d alth o u g h pipefishes
show a large diversity in relative sn o u t elongation, th ey are m ore co n strain ed in
term s of head shape; an d (3) th e head of juvenile Hippocampus reidi specim ens still
undergoes gradual shape changes a fte r being expelled fro m th e b ro o d pouch.
O n to g en etic changes include low ering of th e sn o u t and h ead b u t also differences
in o rie n ta tio n of th e p reo p ercu lar b o n e an d low ering of th e sn o u t tip.
4.1 In troduction
T he fam ily Syngnathidae consists of pipefishes, seadragons, pipehorses and
seahorses, all having a tu b u la r sn o u t w ith tiny, term in al, to o th less jaws. The
sn o u t can be extrem ely elongate, up to a sn o u t le n g th of alm ost tw o th ird s of th e
to ta l head le n g th as in th e w eedy seadragon (Phyllopteryx taeniolatus). In m ost
anim als jaw m orphology, m o u th shape an d d e n titio n type can provide a q u ite
accurate idea a b o u t th e p referred prey type and feeding strategy. T he shape of a
fish skull, consisting of over 30 m oveable b o n y elem ents and m ore th a n 50
m uscles (Lauder, 1983), reflects th e individual’s d iet an d prey cap tu re m echanism
to a great e x te n t (G erking, 1994; D elariva & A gostinho, 2001; Ferry-G raham et al.,
200ia; Palm a & A ndrade, 2002). D ietary influences o n head m orphology (or vice
versa) are p articu larly evident in tro p h ic specialists, w h o exploit a lim ited d ietary
b re a d th w ith respect to th e available prey types in th e ir h a b ita t (Sanderson, 1991).
These species have highly specific feeding dem ands an d th e ir h ead usually is
ch aracterized by specific ad ap tatio n s w h ich m ake th e feeding ap p aratu s wellsu ited for th e ir diet (e.g. jaw asym m etry and specialized d en tal m orphology in
scale-eating cichlids (Liem & S tew art, 1976; T akahashi et al., 2007; S tew art &
A lbertson, 2010)).
T he syng nathid feeding a p p aratu s and feeding strateg y are highly specialized (see
ch ap te r 3.2 an d 5.1). Seahorses an d pipefishes ap p ro ach th e ir prey ventrally, or
C h a p t e r 4 - Morphological variation
85
th e y sit and w a it u n til a prey passes above th e snout. T h an th e head rapidly
ro ta te s dorsally to p o sitio n th e m o u th close to th e prey an d alm ost a t th e same
tim e, a fast buccal expansion creates a w a ter flow th a t tra n sp o rts th e prey in to
th e m o u th (Van W assenbergh et al., 2008). It is sh o w n th a t a sy n g n ath id su ctio n
feeding event is p erfo rm ed w ith g reat precisio n an d at h ig h speed (M uller & Osse,
1984; B ergert & W ain w rig h t, 1997; de L ussanet & M uller, 2007). The ra te of th e
incom ing w a te r is increased by th e sm all d iam eter of th e sn o u t, hence su ctio n
velocity is enhanced. Prey in tak e tim e is as little as 6 ms, m aking syngnathids
am ong th e fastest feeding teleosts ever recorded (C h ap ter 5.1; M u ller & Osse,
1984; B ergert & W ain w rig h t, 1997; de L ussanet & M uller, 2007). D ue to th e tin y
m o u th ap ertu re and n arro w sn o u t, th e sy n g n ath id d iet consists solely of sm all
prey item s (m ostly crustaceans) (T ipton & Bell, 1988; Teixeira & M usick, 1995;
W oods, 2003; F oster & V incent, 2004; K endrick & H yndes, 2005). W e already
n o te d th e sim ilarities betw een pipefishes and seahorses (the long an d slender
sn o u t w ith sm all to o th less jaws at th e tip), b u t th e re are also som e n oticeable
differences. Pipefishes (subfam ily Syngnathinae) are ch aracterized by a long,
cylindrical body an d a n arro w head w ith o u t spines, w hereas seahorses (subfam ily
H ippocam pinae) have an u p rig h t body axis w ith a p reh en sile tail an d a high,
tilte d head usually w ith spines and a corona (i.e. th e first n u ch al p late w ith a
crow n-like process on top). In th is stu d y w e are in te re sted in th e v ariatio n in
sy ngnathid head shape and w e address th re e d ifferen t questions. First, w e
investigated w h e th e r pipefishes and seahorses occupy a d istin c t p a rt of th e
sy ngnathid m orphospace, or w h e th e r th e ir head shape is spread over a c o n tin u u m
of shapes w ith tra n sitio n a l form s in b etw een th e typical p ip efish an d seahorse
m orphotypes.
A n o th e r in te re stin g aspect is th e hypothesis th a t th e tro p h ic ap p aratu s in
specialists is w ell-adapted to th e prevailing circum stances. H ence, only a small
deviation of one elem en t of th e specialized an d com plexly in te g ra te d feeding
system m ay resu lt in reduced fu n c tio n a l perfo rm an ce of th e w hole system
(Adriaens & H errei, 2009). C o n seq u en tly it is expected th a t a specialized organism
w ill show a less versatile feeding system m orphology an d hence a reduced
a m o u n t of intraspecific variation. So o u r second q u estio n is w h e th e r th is can be
generalized for syngnathids: does th e h ig h level of specialization in fluence th e
a m o u n t of m orphological v ariatio n fo u n d w ith in th e family? L ong-snouted
86
C h a p t e r 4 - Morphological variation
species are expected to be able to catch a prey fro m a g reater distance an d have
sh o rte r prey cap tu re tim es (M uller & Osse, 1984; K endrick & H yndes, 2005; de
Lussanet & M uller, 2007). A nalysis of th e sy n g n ath id sto m ach c o n te n t has
d em o n strated th a t long-snouted species consum e m ore prey item s th a t are
relatively m ore m obile, com pared to sh o rt-sn o u ted species (K endrick & H yndes,
2005). H ence, lo n g-snouted syngnathids can be regarded as tro p h ically m ore
specialized an d th e y are th e re fo re expected to show a re d u c tio n in m orphological
variatio n w ith respect to sh o rte r sn o u ted syngnathids.
Thirdly, n o t only th e ir feeding ap p aratu s and strategy m akes syngnathids an
in te re stin g study object, th e y are p articu larly w ell k n o w n fo r th e ir rem arkable
p a re n ta l care. N orm ally, new ly h a tch e d larval fishes are n o t fully developed yet,
how ever th e ir feeding ap p aratu s m u st be o p eratio n al by th e tim e th e yolk sac is
resorbed. O fte n th is critical perio d is accom panied by a d ietary sh ift d u ring
developm ent so th a t in early stages feeding can occur in a d ifferen t w ay an d on
d ifferen t prey item s th a n after th e tra n s itio n to juvenile an d ad u lt feeding. This
can involve in ten se m orphological tra n sfo rm a tio n s d u rin g o n to g en y (e.g. th e
v en tral sh ift of th e m o u th in th e su ck erm o u th catfish Ancistrus cf. triradiatus
described by G eerinckx et al. (2005) or th e radical tra n sfo rm a tio n fro m ta p e ta d to
w h alefish as recently discovered by Jo h n so n et al. (2009), w h ic h consists of a
m ajor increase in jaw le n g th am ong o th e r things). In syngnathids, how ever, th is
m etam orphosis does n o t take place (at least n o t after release from th e pouch):
new ly b o rn seahorses are k n o w n to be provided w ith an ad u lt feeding ap p aratu s
an d prey cap tu re kinem atics (Van W assenbergh et al., 2009). This sim ilarity is
fa c ilitated by th e prolonged p a re n ta l care in seahorses an d pipefishes; em bryos
h a tc h in a special bro o d area on th e m ale body w h ere th e y are n o u rish ed and
p ro te c te d (Lourie et al., 1999b; C arcu p in o et al., 2002; F oster & V incent, 2004).
Release of th e young in to th e free w o rld is delayed u n til an advanced
developm ental stage has been reached; likely th e incredible in v estm en t of
seahorse p aren ts in th e ir offsprin g has ex ten d ed th e developm ental phase in th e
p o u ch (C hapter 3.1 an d Foster & V incent, 2004; D hanya et al., 2005; Van
W assenbergh et al., 2009). All elem ents of th e ex tern al sk eleto n in juvenile
pipefishes and seahorses are p re sen t and th e y are capable of in d e p en d e n t feeding
a t th e m o m en t of release from th e p o u ch (K ornienko, 2001; F oster & V incent,
2004). T herefore it is plausible th a t all great m orphological tra n sfo rm a tio n s w ill
C h a p t e r 4 - Morphological variation
87
tak e place durin g th e perio d in th e b ro o d pouch. T he only changes th a t are likely
to occur a fte r being expelled fro m th e p o u ch w ill be th e re su lt of d ifferen tial
relative g ro w th p a tte rn s (allom etries) since th e relative sn o u t le n g th of new ly
released seahorses is m uch sm aller com pared to th e one of adults. T rad itio n al
m o rp h o m etric
research
has
d em o n strated
th a t
th e re
are
som e
gradual
o n to g en etic changes in relative h ead an d sn o u t dim ensions. C hoo an d Liew (2006)
n o te d a decrease in head length , h ead h eig h t, sn o u t le n g th and sn o u t h e ig h t in
re la tio n to stan d ard le n g th in Hippocampus kuda, w ith sn o u t le n g th an d h eig h t
increasing relatively to head le n g th an d height. O th e r allom etries w ere observed
by Roos et al. (2010) in H. reidi, w h ere sn o u t le n g th increased an d sn o u t h eig h t
decreased relative to head length. H ow ever, p o te n tia l su b tle differences in head
shape (e.g. p o sitio n of th e corona, o rie n ta tio n of th e o p ercu lar bone, steepness of
th e m esethm oidal curvature) ra th e r th a n changes in head d im ension could n o t be
discovered w ith th e applied lin ear m easurem ents. D etectio n of th ese shape
changes w ith in a co ordinate system requires a d etailed geom etric m o rp h o m etric
analysis. The final q uestion of th is research is w h e th e r or n o t juvenile seahorses
still show a perio d w ith su b stan tial m orphological tra n sfo rm a tio n s after release
from th e bro o d pouch.
T he
previously
m e n tio n e d
q uestions w ere
addressed by q u an tify in g
th e
m orphological v ariatio n in th e c ran iu m of a b ro ad range of syngnathid
representatives. F irst w e exam ined to w h a t degree p ip efish shape v ariatio n is
d ifferen t from th a t of seahorses. W e p erfo rm ed a landm ark-based geom etric
m o rp h o m etric analysis on th e ex tern al h ead of a large n u m b e r of syngnathid
representatives an d expected it to yield a clear sep aratio n of th e tw o subfam ilies,
due to th e p ro m in e n t differences in head h e ig h t and tiltin g am ong o th e r things.
Secondly, w e exam ined w h e th e r sy n g n ath id shape v ariatio n is c o n strain ed in
re la tio n to relative sn o u t length, e ith e r a t an in ter- or in trasp ecific level. For th is
p art, o u r pred ictio n s are based o n th e previously fo rm u lated assu m p tio n th a t a
h igh level of specialization is associated w ith a re d u c tio n of m orphological
variety. If th is hypothesis holds tru e , a lo n g -sn o u ted p ip efish or seahorse species,
being m ore specialized th a n a sh o rt-sn o u te d one, w ill show a reduced level of
intraspecific m orphological variation. For th e same reason, at an interspecific
level, w e expected pipefishes, w h ich can have extrem ely elongated sn o u ts and
th u s are tro p h ically m ore specialized, to be m orphologically m ore co n strain ed
88
C h a p t e r 4 - Morphological variation
th a n seahorses. To te st th is hypothesis at an in trasp ecific level, a regression
analysis betw een th e relative average sn o u t le n g th of a species an d th e variance of
th e shape variables of th a t species as a proxy of in trasp ecific v ariatio n w as carried
out. For th e hypothesis a t an intersp ecific level, w e com pared b o th subfam ilies in
term s of th e ir variance. Finally, w e investigated th e hypothesis th a t syngnathids
do n o t experience extensive o n to g en etic shape changes fro m th e m o m en t th ey
are released from th e p o u ch because im p o rta n t m orphological tra n sitio n s have
likely occurred d u ring th e perio d of p a re n ta l care.
4.2. Brief material & m e t h o d s
In to ta l 368 specim ens of 38 d iffe re n t species w ere analyzed m orp h o m etrically
(Table 2.3). This includes 54 juvenile specim ens of Hippocampus reidi, ranging in
age betw een 1 and 65 days, w h ich w ere used fo r th e o n to g en etic study in
co m b in atio n w ith th e adults of H. reidi. A nim als w ere sacrificed an d fixed as
described in ch ap ter 2.2.2 an d th e rig h t side of th e head of all anim als was
p h o to g rap h ed (see ch ap ter 2.2.3).
T he geom etric m o rp h o m etric analysis involved th e use of th e T h in Plates Spline
freew are (tps; R ohlf, F.J., State U niversity of Stony Brook, N ew York) as
explained in ch ap te r 2.2.9 a n d figure 2.3.
T he range of sn o u t lengths in o u r sam ple did n o t adequately reflect th e diversity
in relative sn o u t le n g th k n o w n to exist in th e p ip efish subfam ily; it lacked sh o rt­
sn o u ted pipefishes. For exam ple, species like Siokunichthys bentuviai C lark and
Anarchopterus (Dawson) have a relative sn o u t le n g th of respectively a b o u t 0.180
an d 0.250 (m easurem ents m ade on draw ings by D aw son in K uiter, 2003), w h ich is
even sh o rte r th a n th e relative sn o u t le n g th of th e seahorse species w ith th e
sh o rte st sn o u t in o u r analysis, Hippocampus zosterae. T herefore, draw ings of th e
lateral side of th e head of 11 sh o rt-sn o u te d species and one lo n g -sn o u ted species
(Microphis caudocarinatus (W eber)) m ade by D aw son (1978; 1983; K uiter, 2003)
w ere added to th e original d ataset (Table 2.3). These draw ings did n o t include a
scale bar, hence for analyses relying o n size factors (e.g. regression b etw een
c en tro id size an d relative w arp scores) th e draw ings w ere le ft out. T he use of th e
d ataset w ith o u t th e draw ings is clearly m e n tio n e d th ro u g h o u t th e results.
C h a p t e r 4 - Morphological variation
89
W e carried o u t th e same analyses w ith o u t LM2, because its p o sitio n (on th e distal
p o in t of th e d en tary bone) is expected to be d ep en d en t o n th e gape (Fig. 2.3). In
specim ens w ith a depressed low er jaw LM 2 w ill be lo cated m ore v en tro ro strally
com pared to its p o sitio n in fishes w ith a closed m outh. O m ittin g LM 2 allow ed us
to assess th e consequences of th is p o te n tia l error.
A n o th e r tps-file was m ade for th e o n to g en etical analysis (see ch ap ter 2.2.9).
For tho se species rep resen ted in th e analysis by m ore th a n te n specim ens, th e
species variance was calculated using th e follow ing form ula:
y
A
d ij2
i(ni -
!)
w ith d;j th e E uclidean distance of all R W scores th a t to g e th e r explain over 95% of
th e to ta l variatio n of th e jth specim en in species i to th e R W scores of th e species
consensus, an d n¡ th e n u m b e r of specim ens in species i. The species variance is a
m easure of h ow m uch each specim en w ith in a species deviates fro m th e species
consensus; th e larger th e variance, th e m ore divers th a t species is. H ence, it is a
reflectio n of th e m orphological in trasp ecific v ariatio n of a species.
T he same form ula w as used to d eterm in e a m easure fo r in tersp ecific v ariatio n for
pipefishes an d for seahorses respectively, only th e overall subfam ily consensus
w as used in stead of th e species consensus an d all species variances w ere sum m ed
for each subfam ily. The previously m en tio n e d fo rm u la is calculated analogous to
th e p a rtia l disparity described by Foote (1993), b u t it has th e n u m b er of
specim ens in a subgroup (nr i) in th e d e n o m in a to r ra th e r th a n th e to ta l n u m b er
of specim ens of all subgroups com bined (N r i). In th is way, th e proxy for
intraspecific v ariatio n is less susceptible to differences in su b group size.
4.3. Results
A. O verall sh ap e v a r ia tio n
Firstly, no differences betw een th e results w ith or w ith o u t LM 2 w ere observed,
hence th e d ataset including LM 2 w as used because it allow ed us to in te rp re t
low er jaw le n g th (distance b etw een LM3 an d LM2).
90
C h a p t e r 4 - Morphological variation
In addition, w e could prove th a t sim p lificatio n of a p ic tu re by a draw ing did n o t
affect th e placing of landm arks since th e results a fte r adding th e draw ings w ere
as expected: th e sh o rt-sn o u ted species all clu ster to g e th e r n ear th e Doryichthys
species, w h ich are also sh o rt-sno u ted , an d th e lo n g -sn o u ted p ip efish Microphis
caudocarinatus could be fo u n d closer to th e o th e r lo n g -sn o u ted species (Fig. 4.1).
As expected, th e G oodall’s F te st show ed th a t th e re is a sig n ifican t difference
b etw een pipefishes and seahorses (F-score = 713.03; df = 26.00, 7800.00; p-level <
0.0001; B ootstrap p-level = 0.0002).
T he first and second relative w arp scores are p lo tte d to visualize th e shape
differences (Fig. 4.1). The first axis of th e relative w arp p lo t (R W i), w h ich
explains 75.08% of th e to ta l variation, separates pipefishes fro m seahorses.
Positive scores of th e R W i (i.e. p ip efish m orphotype) m ainly reflect a vast
decrease in head h e ig h t w ith respect to th e consensus (ventral sh ift of LM10 and
LM11) an d a dorsal displacem ent of th e sh o rte n ed p ecto ral fin base (dorsal sh ift
of LM13 an d LM14). In addition, th e y reflect a do rso v en tral sh o rte n in g of th e
o percular bone (shortening of th e distance b etw een LM15 an d LM8), an increase
in sn o u t le n g th (caudal displacem ent of LM6 an d LM7) an d a dorsoventral
low ering of th e sn o u t an d low er jaw (dorsal sh ift of LM2, LM3 an d LM4).
Specim ens a t th e o th e r extrem e (i.e. seahorse m orphotype) have a large an d very
h igh head (dorsorostral sh ift of LM10 and dorsocaudal sh ift of LM11), a
dorsoventrally enlarged o percular b o n e (dorsal sh ift of LM8 in c o m b in atio n w ith
th e v en tral displacem ent of LM15), a relatively sh o rt b u t h ig h sn o u t (convergence
of LM1-5 w ith LM6-7 to g e th e r w ith a v en tral displacem ent of LM3, LM 4 and
LM15) an d an elongated and v en trally displaced p ecto ral fin base (v entrorostral
sh ift of LM13 an d even m ore so fo r LM14) com pared to th e consensus
configuration.
T he shape changes rep resen ted by th e second relative w arp axis, w h ich explains
10.02% of th e to ta l shape variation, also reflects v ariatio n in sn o u t h eig h t
(dorsoventral sh ift of L M i an d LM5), sn o u t le n g th (divergence or convergence of
LM5 an d LM6) an d head le n g th (rostrocaudal m ovem ent of LM8 an d LM15
to g e th e r w ith rostrocaudal m ovem ent of b o th LM13 an d LM14), b u t all to a
lesser degree com pared to R W i. Besides th a t, R W 2 rep resen ts v ariatio n in a
rostro cau d al pecto ral fin base displacem ent (shifts in LM13 an d LM14 p o sitio n
C h a p t e r 4 - Morphological variation
91
along th e long axis of th e head) an d in th e o rie n ta tio n of th e o p ercu lar bone
(rostrocaudal m ovem ent of LM8 w ith resp ect to LM15).
B. S h a p e v a r ia tio n a n d s n o u t d im e n s io n s
T he regression analysis (of th e d ataset w ith o u t th e draw ings) show ed th a t less
th a n 3.5% of th e to ta l shape variatio n w as explained by e ith e r cen tro id size or log
head length, indicatin g th a t th e o b tain ed R W scores are in d e p en d e n t of size. O n
th e o th e r hand, 65.32% of th e overall v ariatio n is explained by differences in
relative sn o u t length. O f all relative w arps, th e first is th e only one th a t is
co rrelated w ith relative sn o u t le n g th (a co rrelatio n co efficien t of 0.89). H ence,
these results co n firm th a t one of th e m ain differences b etw een th e p ip efish and
seahorse m orphotype is th e ir relative sn o u t len g th , as was already evident from
th e com parison of d efo rm atio n grids along R W 1. Also, th e first relative w arp axis
is significantly negatively co rrelated w ith relative sn o u t h eig h t (a co rrelatio n
coefficien t of -0.92).
Four pip efish an d fo u r seahorse species are rep resen ted in th e analysis by m ore
th a n te n specim ens and th u s th e ir species variance was calculated as a proxy for
intraspecific shape v ariatio n as described earlier (Table 4.1). B oth p ip efish and
seahorse subfam ilies seem ed to show an inverse re la tio n b etw een relative sn o u t
le n g th and intraspecific v ariation , so th e m orphological v ariatio n of lo n g -sn o u ted
species is likely to be m ore co n strain ed th a n th a t of sh o rt-sn o u ted species (Fig.
4.2). H ow ever, w ith only fo u r data p o in ts in each group, th e re was n o t su fficien t
statistical su p p o rt for th is trend . A n o th er th in g th a t can clearly be n o te d is th a t
alm ost each seahorse species show s a h ig h er a m o u n t of in trasp ecific v ariatio n
th a n th e pipefishes. This was also fo u n d o n an in tersp ecific level, overall
v ariatio n w as low er for pipefishes (4.86 io"3) th a n fo r seahorses (7.78 io '3).
T a b l e 4 .1 - L is t o f t h e f o u r p i p e f i s h e s a n d f o u r s e a h o r s e s f o r w h i c h t h e i n t r a s p e c i f i c v a r ia t i o n
w a s c a lc u la t e d
sp e c ie s
Syngnathus rostellatus
Syngnathus typhle
Syngnathoides biaculeatus
M icrophis brachyurus aculeatus
H ippocam pus reidi
H ippocam pus ramulosus
H ippocam pus guttulatus
H ippocam pus kuda
n u m b er of
sp e c im e n s
m e a n (S n L /H L )
± st dev
in tr a s p e c if ic
v a r i a ti o n
15
15
18
0.45810.011
0.48310.038
0.55310.026
0.57810.012
0.42610.026
0.35410.032
0.36010.029
0.42810.036
1.46E-03
3.55E-03
1.59E-03
0.96E-03
3.39E-03
8.33E-03
8.03E-03
6.94E-03
95
12
18
29
40
92
C h a p t e r 4 - Morphological variation
It is obvious from th e p lo t of R W i versus R W 2 scores (Fig. 4.1) th a t th e seahorse
specim ens are scattered all aro u n d th e g rap h w hereas th e space occupied by th e
pipefishes is m ore stre tch e d along an axis as if n o t all possible p o sitio n s in th e
sy ngnathid m orphospace could be occupied. This d ifference was q u a n tifie d by
perfo rm in g a lin ear regression b etw een R W i an d R W 2 (accounting fo r th e m ost
p ro m in e n t shape changes), resu ltin g in a co rrelatio n co efficien t of R2=o.6i (plevelco.oooi) for pipefishes com pared to R2=o.o8 (p-levelco.oooi) fo r seahorses.
This proves again th a t seahorses show a hig h er level of in tersp ecific v ariatio n and
th a t pipefishes all follow a sim ilar m orphological p a tte rn . The p ip efish
m orphospace is b e st described as an ellipse w ith th e long axis prim arily
explaining differences in relative sn o u t len g th , b u t also in sn o u t height,
rostro cau d al p o sitio n of th e p ecto ral girdle, h ead h e ig h t an d o p ercu lar height.
This tre n d probably depends greatly o n th e selection and n u m b er of specim ens
an d species in th e analysis.
C. O n to g e n e tic s c a lin g
T he a d u lt Hippocampus reidi specim ens are w ell sep arated fro m th e o th e r age
classes by a c o m b in atio n of th e first an d second relative w arp axis, explaining,
40.25% an d 16.75%, respectively, of th e to ta l shape v ariatio n (Fig. 4.3). A pparently,
even a fte r 65 days, changes in head shape still occur. O n to g en etic shape
tra n sitio n s from a one day old stage to a d u lt involve a re d u ctio n of th e obtuse
angle betw een th e ro stral edge of th e o p ercu lar b o n e an d th e sn o u t to an alm ost
rig h t angle (dorsorostral sh ift of LM8 and to a lesser degree of LM15),
dorsoven tral low ering of th e sn o u t tip (ventral displacem ent of L M i w ith respect
to LM5) an d tiltin g of th e p ecto ral fin base (dorsocaudal sh ift of LM13 an d even
m ore of LM14), besides a dorsoven tral low ering of th e h ead w ith respect to its
le n g th (ventral displacem ent of LM10 an d LM11) an d a do rso v en tral com pression
of th e sn o u t (dorsal sh ift of LM3, LM 4 an d LM15) (Fig. 4.3). To evaluate th e
presence of d ifferen t stages or inflex io n p o in ts in th e ontogeny, w e p lo tte d head
le n g th (as a m easure of age) over relative w arp scores. A lth o u g h th e re are changes
in head shape d u ring post-release developm ent, n o clear tre n d in head shape over
tim e could be established (Fig. 4.4).
C h a p t e r 4 - Morphological variation
93
4.4. D is cu ssi on
A n im p o rta n t rem ark is th a t som e of th e v ariatio n fo u n d m ig h t be caused by
phylogenetic non-independency. R elated species m ig h t share th e same tra its due
to shared ancestry. To co n tro l fo r p h ylogenetic relatedness in com parative
analyses, a fully resolved phylogeny is required. U n fo rtu n a te ly a tre e in cluding all
p ipefish an d seahorse genera is c u rre n tly n o t available. H ence, th e results of th is
geom etric m o rp h o m etric analysis need to be in te rp re te d carefully.
A.
P ip e fis h versus s e a h o rs e s h a p e v a r ia tio n
B oth subfam ilies are m orphologically clearly separated fro m
each other,
suggesting a bim odal d istrib u tio n in head shape o n subfam ily level in stead of a
m orphological c o n tin u u m (Fig. 4.1). The m ain differences b etw een th e p ip efish
an d seahorse head shape are sn o u t le n g th and h eig h t, p o sitio n an d o rie n ta tio n of
th e p ecto ral fin base and h e ig h t of th e head an d o p ercu lar bone. T he sh o rt­
sn o u ted pipefishes cluster at th e le ft h a n d side of th e p ip efish m orphospace, as
expected since th e first relative w arp axis, separating pipefishes (+RWT) from
seahorses (-R W i), is highly co rrelated w ith sn o u t le n g th (Fig. 4.1). H ow ever, th ey
do n o t occur m ore to th e le ft in th e p lo t th a n Hippocampus zosterae, y et th e
relative sn o u t le n g th of som e of th e m is smaller. This confirm s th a t R W 1 reflects
m ore th a n ju st relative sn o u t length, as elab o rated o n in th e results.
T he first relative w arp axis is co rrelated positively w ith sn o u t le n g th and
negatively w ith sn o u t height. W h e n relative sn o u t le n g th is p lo tte d over relative
sn o u t h e ig h t for all syngnathids studied, lin ear regression yields a negative slope
(slope=-o.3794, R2=o.8i, p-levelco.oooi). H en ce an increase in relative sn o u t
le n g th results in a decrease of relative sn o u t h e ig h t and th is re la tio n is m ore
p ro n o u n c e d in pipefishes (slope^o.3134, R2=o.67, p-levelco.oooi) com pared to
seahorses (slope=-o.i982, R2=o.36, p-levelco.oooi). In pipefishes, ab so lu te sn o u t
h e ig h t does n o t vary m uch over a w ide range of sn o u t lengths, so lo n g -sn o u ted
pipefishes te n d to have a relatively n arro w snout. De Lussanet an d M u ller (2007)
have d eterm in ed th a t th e o ptim al sn o u t le n g th of syngnathids is inversely related
w ith sn o u t cross-section. If it can be assum ed th a t n a tu ra l selection w o u ld favour
an o ptim al sn o u t len g th , o u r results co n firm th e hypothesis of de L ussanet and
M u ller (2007). T hey sta te d th a t a long sn o u t is advantageous since it m inim izes
C h a p t e r 4 - Morphological variation
94
th e angle over w h ich th e head m u st be tu rn e d an d hence reduces th e tim e to
reach th e prey (de Lussanet an d M uller, 2007). H ow ever, a lo n g er sn o u t im plicates
a larger strik in g distance, w h ich m ay reduce strik in g success (Roos et al.,
su b m itte d a). M oreover, an increased sn o u t le n g th im plies a h ig h er to ta l m o m en t
of in ertia, w h ich negatively influences th e angular acceleration of h ead ro tatio n .
R eduction of th e cross-sectional area of th e sn o u t decreases th e to ta l m o m en t of
in e rtia and com pensates for th e lo n g er snout. This is co n firm ed by o u r analysis.
M u ller & O sse (1984) reasoned th a t th e velocity an d acceleratio n of th e w ater
relative to th e m o u th opening increases w ith increasing sn o u t len g th , hence
faster prey can be caught. Roos et al. (2010) estab lish ed th is experim entally since
th e y fo u n d th a t th e sn o u t le n g th in b o th juvenile an d ad u lt H. reidi approaches
th e o ptim al le n g th for a given w id th , in ord er to m inim ize th e tim e to reach prey.
A pparently,
lo n g-snouted
species
only
require
lim ited
h ead
ro ta tio n
in
com parison w ith a sh o rt-sn o u ted species th a t requires a large excursion of th e
head to overcom e th e same distance. In ad d itio n , th e increase in cranial ro ta tio n
perform ance by sn o u t elo ngatio n is opposed to th e associated re d u ctio n in
su ctio n perform ance (Roos et al., 2011). Im p ro v em en t of th e efficiency of th e tw o
d ifferen t aspects of pivot feeding, i.e. h ead ro ta tio n versus su ctio n perfo rm an ce
m ight be reflected in p ipefish versus seahorse m orphology. Seahorses have an
u p rig h t p o stu re an d in rest th e ir h ead is a t an acu te angle w ith th e ir body. D u rin g
head ro ta tio n of a feeding strike, th e supraoccipital b o n e slides u n d er th e corona
(thus producing th e ch aracteristic clicking sou n d (C olson et al., 1998)). Pipefishes,
on th e o th e r hand, have th e ir head in line w ith th e ir body a t th e b eg in n in g of a
prey cap tu re
event. M a n ip u la tio n
of cleared and stain ed specim ens
(D.
dactyliophorus, S. rostellatus and C. intestinalis) show ed th a t from a certain angle,
fu rth e r head ro ta tio n is m echanically im peded. E ith er th e cranio-vertebral jo in t
m orphology lim its n eu ro cran ial elevation or th e su p rao ccip ital b o n e or th e
epaxial bones (very long and slender ossifications w ith in th e te n d o n of th e
epaxial m uscle in som e pipefishes, see ch ap ter 3.2) are p u sh ed against th e first
n u ch al plate (no sliding w as observed in pipefishes). M u ller & O sse (1984) suggest
a m echanical lim ita tio n of head ro ta tio n by th e lig am en t co n n ectin g th e
su praoccipital bone w ith th e cleith ru m (w hich th e y refer to as th e p late form ed
by th e fused p ecto ral radiais an d dorsal v erteb ral processes). W e presum e th a t th e
seahorse co n fig u ratio n m ight allow fo r a g reater ro ta tio n of th e head w ith
C h a p t e r 4 - Morphological variation
95
respect to th e body com pared to th e p ip efish situ atio n . If th is hypothesis holds,
seahorses do n o t need to have an extrem ely elo n g ated sn o u t to overcom e th e
p rey -m o u th distance as a vast a m o u n t of cranial ro ta tio n w o u ld suffice to reach
th e prey. A re c e n t study by Van W assenbergh et al. (2011) proved th a t th e b e n t
tru n k o rie n ta tio n of seahorses im proves feeding p erfo rm an ce by increasing th e
strik e distance. A typical p ipefish body o rie n ta tio n , o n th e o th e r hand, appears to
be beneficial by increasing strik e velocity. A n o th er reason to assum e im proved
head ro ta tio n perform ance w ith an increase in tiltin g of th e head is th e large
in sertio n site for th e w ell-developed epaxial m uscle, w h ich plays a crucial role in
n eu ro cran ial
elevation,
on
th e
E xperim ental su p p o rt for th e
h ig h
supraoccipital
hypothesis
is fo u n d
bone
in
in
seahorses.
k in em atical data:
n eu ro cran ial elevation in a d u lt Hippocampus reidi reaches 310 (C h ap ter 5.1) and
even as m uch as 400 in juvenile H. reidi (Van W assenbergh et al., 2009), versus a
ro ta tio n of only a b o u t 200 in Syngnathus leptorhynchus G irard (Van W assenbergh
et al., 2008) an d ro ta tio n does n o t exceed io° in Doryrhamphus dactyliophorus or D.
melanopleura (Roos et al., su b m itte d a). H ow ever, B ergert an d W ain w rig h t (1997)
fo u n d a n eu ro cran ial elevation of 290 in S. floridae, equal to w h a t th ey recorded
for Hippocampus erectus, w h ic h show s th a t o u r hypothesis c an n o t be generalized
for all syngnathids. A com parative stu d y of th e su p rao ccip ital-n u ch al p late
m orphology an d m echanics th a t w ill h elp resolve th is is called for.
O th e r m orphological differences b etw een p ip efish and seahorse heads, i.e. v en tral
p o sitio n of th e pecto ral fin base an d o p ercu lar b o n e h ig h er th a n long in th e latter,
m ight also be explained as a consequence of th e head tiltin g in seahorses. The
p ecto ral girdle is a ttach ed to th e v erteb ral colum n so th e v en tral displacem ent of
th e p ecto ral fin base observed in o u r d efo rm atio n grids, actually rep resen ts a sh ift
of th e e n tire body from a caudal p o sitio n to a v en tro cau d al one w ith respect to
th e head. At th e same tim e, by tiltin g th e head w ith th e p ecto ral girdle
im m ovably fixed to th e verteb ral colum n, th e distance b etw een th e caudal m argin
of th e p reo p ercu lar bone an d th e ro stral tip of th e cleith ru m is reduced leaving
only little space for th e gills in ro stro cau d al d irection. This m ig h t have caused th e
b ran ch ial cham ber in seahorses to expand m ore in a do rso v en tral d irectio n hence
th e h eig h t of th e opercular bon e exceeds its le n g th (the opposite is th e case in
pipefishes).
96
C h a p t e r 4 - Morphological variation
W e d em o n strated th e existence of several differences in th e head shape of
pipefishes and seahorses, yet th e y evolved from th e sam e ancestor. T he oldest
fossils of th e fam ily date 50 M yr back an d m orphologically resem ble m odern
pipefishes (Teske et al., 2004; Teske & Beheregaray, 2009). It w o u ld be in te re stin g
to see w here in th e syngnathid m orphospace th e m ost recen t com m on ancestor,
from w h ich pipefishes and seahorses diverged fro m each other, w ill plot.
A ccording to a re c e n t p u b lic a tio n by Teske and B eheregaray (2009) pipehorses of
th e genus Idiotropiscis can be regarded as th e e x ta n t ev o lu tio n ary lin k b etw een
seahorses an d pipefishes. These seahorse-like pipefishes sw im h o rizo n tally b u t
th e ir head is tilte d ventrally over an angle of a b o u t 250 w ith resp ect to th e body
(K uiter, 2004). Based on pictures, th e head of Idiotropiscis species seems to stan d
m idw ay b e tw ee n th e one of pipefishes and seahorses in term s of sn o u t an d head
dim ension. H ence, it is expected to show o th e r tra n sitio n a l features as w ell, for
instan ce in th e p o sitio n of th e p ecto ral fin an d th e shape of th e o p ercu lar bone. A
fu tu re
geom etric m o rp h o m etric analysis th a t includes one of th e th re e
Idiotropiscis species m ight yield elu cid atin g results w ith respect to syngnathid
evolution.
B. S h a p e v a r ia tio n a n d s n o u t le n g th
W e h y pothesized th a t th e re w o u ld be a g reater a m o u n t of m orphological
v ariatio n w ith in a sh o rt-sn o u ted species versus w ith in a lo n g -sn o u ted one
(intraspecifically) an d also in seahorses com pared to pipefishes (interspecifically).
As explained before, th e ex p ectatio n th a t lo n g -sn o u ted species w o u ld be
m orphologically m ore c o n strain ed follow s fro m th e experim entally based
assum ption th a t th ey are m ore specialized. S u p p o rt fo r syngnathids w ith a long
sn o u t being m ore specialized was fo u n d in g u t co n ten ts analyses. The d iet of
lon g -sn o u ted syngnathids consists of a h ig h er n u m b e r of m obile prey com pared
to w h a t is ingested by sh o rt-sn o u ted species (Pranzoi et al., 1993; K endrick &
H yndes, 2005). This suggests th a t a long sn o u t enables a species to cap tu re prey
from a greater distance an d in a sh o rte r tim e, as argued by M u ller and Osse (1984)
an d de Lussanet & M u ller (2007), an d w as la te r d em o n strated by Roos et al.
(subm itted a). H ence lo n g-snouted syngnathids can be c o n fid en tly regarded as
tro p h ic specialists.
C h a p t e r 4 - Morphological variation
97
O u r prelim in ary results on intrasp ecific v ariatio n are w ith in o u r expectations:
for b o th pipefishes an d seahorses th e re is an in d icatio n of a negative relatio n sh ip
b etw een relative sn o u t le n g th and degree of in trasp ecific v ariation. In o th e r
w ords, th e m ore specialized a species (longer snout), th e m ore m orphologically
c o n strain ed it is (reduced variation). H ow ever w e m u st be careful w h en draw ing
conclusions from these results as w e only looked a t fo u r d ifferen t species for
each subfam ily and also because b o th th e co rrelatio n co efficien t and th e slope in
pipefishes are sm all (R2=o.26, slope=-o.oi02). A dding a n o th e r species m ight
change th e observed relatio n sh ip b etw een pipefishes an d seahorses. For fu tu re
research, it w ill be in te re stin g to include o th e r species, b o th sh o rt-sn o u te d ones
(for instan ce H. zosterae and Siokunichthys bentuviai) an d lo n g -sn o u ted species (D.
dactyliophorus an d H. barbouri Jordan & R ichardson) to see if th e p re sen t inverse
re la tio n still stands.
O n an interspecific level, seahorses show a h ig h er degree of v ariatio n com pared
to pipefishes. H ow ever, w e m ust n o te th a t th e average relative sn o u t le n g th of
th e fo u r pip efish species is hig h er th a n th a t of th e fo u r seahorse species. Since w e
saw th a t long-snouted species are m orphologically m ore co n strain ed th a n sh o rt­
sn o u ted ones, th e difference in in tersp ecific v ariatio n (i.e. pipefishes having less
variatio n th a n seahorses) is possibly th e resu lt of o u r selectio n of lo n g -sn o u ted
pipefishes versus sh o rte r sn o u ted seahorses. N evertheless, if w e look a t th e
d istrib u tio n of all specim ens, b o th long- an d sh o rt-sn o u te d ones, in th e p lo t of
R W i versus R W 2 scores (Fig. 4.1), w e see th a t th e m orphospace occupied by
seahorses is sm aller com pared to th e p ip efish m orphospace. H ence, th e diversity
in relative sn o u t shape is larger fo r pipefishes th a n fo r seahorses. Yet, th e fact th a t
th e pip efish specim ens have an elliptical d istrib u tio n in th e p lo t suggests th a t
th e y are m ore co n strain ed in term s of cranial m orphology since all specim ens
follow a sim ilar allom etric p a tte rn . A lth o u g h som e fu rth e r in v estig atio n is
needed, for n ow w e can conclude th a t th e level of in tersp ecific v ariatio n in
seahorses probably is hig h er th a n in pipefishes.
Studies dealing w ith th e effect of specialization o n m orphological diversity are
scarce.
H olliday
&
S tep p an
(2004)
in v estig ated
how
hypercarnivory
(specialization to a m eat-only diet) affects m orphological v ariatio n in m am m alian
carnivores.
T hey
fo u n d
th a t
hypercarnivores
show
significantly
low er
m orphological diversity th a n th e ir p rim itively non-h y p ercarn iv o ro u s sister
98
C h a p t e r 4 - Morphological variation
groups, w h ich probably is related to fu n c tio n a l co n strain t. These findings
su p p o rt o u r observation w ith in syngnathids.
A m ethodological co n sid eratio n m ig h t be th e fact th a t in th ese analyses
specim ens belonging to th e ‘species’ H. guttulatus Cuvier, H. kuda an d H. ramulosus
Leach w ere used. These ‘species’ are k n o w n to be species com plexes an d th ey
encom pass various m orphotypes. Lourie et al. (1999b) do n o t even recognize H.
ramulosus as a separate species b u t ra th e r as a synonym of H. guttulatus. C urrently,
th e re is no alternative to using th ese species com plexes as th e y have n o t been
revised yet. The use of th e invalid species nam e H. ramulosus should be avoided,
how ever, th e seahorse id en tific a tio n key (Lourie et al., 2004) o fte n does n o t yield
one unam biguous species nam e b u t ra th e r various possible solutions. H ence, here
th e assignm ent of specim ens to th e species com plexes is p re fe rre d over th e use of
u n id en tifiab le
id e n tific a tio n
specim ens.
F u rth erm o re,
does n o t tak e
th e
th is
com plication
edge of th e
results
of
of th e
seahorse
geom etric
m o rp h o m etric analysis. For th e calcu latio n of th e R W scores, th e v ariatio n
am ong specim ens (and n o t groups) is calculated hence w h e th e r a specim en
belongs to species A or species B w ill n o t change its p o sitio n o n th e R W 1 versus
R W 2 plot. O n an intraspecific level, it m ig h t be so th a t w h a t w e consider n o w as
one species in fact are several, w h ich could reduce th e a m o u n t of intraspecific
v ariatio n for th e separated species com pared to w h a t w e have fo u n d here.
H ow ever, th e re is no reason to assum e th a t break in g dow n th e species com plex
in to separate species w ill affect th e in trasp ecific v ariatio n d ifferen tly fo r longversus sh o rt-sn o u ted species. H ence, in th e absence of taxonom ical clarification,
w e do n o t consider th e use of species com plexes an issue.
C. Shape v a r ia tio n and o n to g e n e tic sca lin g
No evidence of saltato ry developm ent w as fo u n d th a t could c o n tra d ict th e
hypothesis th a t in syngnathids com plex m orphological changes and o n to g en etic
m odifications takes place d u ring th e extended p erio d of p a re n ta l care in th e
bro o d p o u c h an d n o t afterw ards. In m any o th e r fish, freshly h a tch e d em bryos
possess a yolk sac for endogenous n u tritio n . The yolk sac w ill be d ep leted a t th e
tra n s itio n to active feeding; a p erio d ch aracterized by th e developm ent of th e jaw
C h a p t e r 4 - Morphological variation
99
ap p aratu s and hence changes in head shape. In syngnathids, th e re is n o larval
p eriod an d m etam orphosis is ab sen t or occurs in th e b ro o d pouch, so th e feeding
ap p aratu s is already fully developed b efo re em erging (K ornienko, 2001; C hoo &
Liew, 2006). This is co n siste n t w ith th e findings fo r Syngnathoides biaculeatus
(Bloch), Hippocampus kuda an d H. reidi, w here th e new b o rn s resem ble m in iatu re
adults w ith a com pletely fu n c tio n a l feeding ap p aratu s an d a fully absorbed yolk
sac (D hanya et al., 2005; C hoo & Liew, 2006; Roos et al., 2011). A n o th e r reason to
expect little change is th e fact th a t prey cap tu re events in juvenile seahorses are
observed to be even faster th a n in adults, 2.5 ms com pared to 5 ms (Van
W assenbergh et al., 2009 an d ch ap te r 5.1). To ensure th e success of th e incredibly
rapid pivot feeding, th e com bined m ovem ent of all elem ents involved in th e
feeding apparatus (low er jaw, hyoid, n eu ro cran iu m , Suspensorium and operculum )
is required. The slightest deviation of one elem en t fro m its n o rm al p a th m ight
cause a re d u ctio n in perform ance. So, juvenile seahorses are expected to be highly
constrained, n o t only m orphologically b u t also kinem atically.
Still th e juveniles differ from th e adults; am ong th e m ain allom etric shape
changes are low ering of th e sn o u t and h ead w ith respect to sn o u t len g th , b u t also
a d ifferen t o rie n ta tio n of th e p reo p ercu lar b o n e an d of th e sn o u t tip. Roos et al.
(2010) n o ticed a tra n s itio n from sh o rt and b ro ad sn o u ts to long and n arro w ones
durin g th e developm ent of H. reidi. This su p p o rts o u r findings th a t th e head of
juvenile seahorses still undergoes gradual shape changes after release fro m th e
bro o d pouch. R eynolds n u m b e r of th e juvenile’s e n v iro n m en t is relatively low
an d th e re fo re th e escape response of th e (even sm aller) prey item s w ill be lim ited
(Larsen et al., 2008; Roos, 2010). H igh-velocity h ead ro ta tio n fo r a fast ap p ro ach of
th e prey is th u s no necessity; on th e o th e r hand, p o w erfu l su ctio n is req u ired to
set th e viscous fluid in to m otion. A feeding ap p aratu s w ith a sh o rt an d b ro ad
sn o u t is beneficial for optim izin g su ctio n p erfo rm an ce an d is hence typically
fo u n d in juvenile syngnathids.
P robably th e allom etric shifts in buccal cavity shape, changes in th e o rie n ta tio n
of th e opercular bone an d low ering of th e sn o u t tip w ill have an effect o n th e
hydrodynam ics of su ctio n feeding. The h ig h efficiency of th e feeding m echanism
in juvenile H. reidi can m ost likely only be assured if along w ith th ese
m orphological changes th e kinem atics of prey cap tu re w ere also altered to a
c ertain degree. Indeed, o n to g en etic changes in th e sn o u t dim ension of H. reidi
100
C h a p t e r 4 - Morphological variation
appear to have an influence on su ctio n feeding capacity and o n cranial ro ta tio n
perform ance (Roos et al., 2011).
101
5
5 .1 .
Functional interpretation
K inem atics o f t h e fe e d in g a p p a ra tu s
M o d i f i e d fr o m :
R o o s G *, L e y s e n H *, V a n W a s s e n b e r g h S , H e r r e i A ,
J a c o b s P , D i e r i c k M , A e r t s P , A d r ia e n s D .
L in k in g m o r p h o l o g y a n d m o t i o n : A t e s t o f a fo u r -b a r
m e c h a n ism in sea h o r se s.
P h y s i o l o g i c a l a n d B i o c h e m i c a l Z o o l o g y 2 0 0 8 , 8 2 :7 -1 9 .
* a u th o r s w it h e q u a l c o n tr ib u tio n
A bs tract
S yngnathid fishes possess a highly m odified c ran iu m ch aracterized by a long and
tu b u la r sn o u t w ith m in u te jaws a t its end. Previous studies in d icated th a t these
species are extrem ely fast su ctio n feeders w ith th e ir feeding strik e ch aracterized
by a rapid elevation of th e head accom panied by a ro ta tio n of th e hyoid. A p lan ar
fo u r-b ar m odel is proposed to explain th e coupled m o tio n of th e n e u ro c ran iu m
an d th e hyoid. Because b o th n eu ro cran ial elevation an d hyoid ro ta tio n are crucial
for th e feeding m echanism in previously stu d ied Syngnathidae, a detailed
evaluation of th is m odel is needed. In th is study, w e p re se n t k in em atic d ata of th e
feeding strike in th e seahorse Hippocampus reidi. W e com bined these data w ith a
102
C h a p t e r 5 - Functional interpr et ati on
detailed m orphological analysis of th e im p o rta n t linkages an d jo in ts involved in
ro ta tio n of th e n e u ro c ran iu m an d th e hyoid an d w e com pared th e kin em atic
m easurem ents w ith o u tp u t of a th e o re tic al fo u r-b ar model. The kin em atic
analysis show s th a t n eu ro cran ial ro ta tio n never precedes hyoid ro ta tio n , th u s
indicatin g th a t hyoid ro ta tio n triggers th e explosive feeding strike. O u r data
suggest th a t w hile n e u ro c ran iu m and hyoid in itially (first 1.5 ms) behave as
p red icted by th e fo u r-b ar m odel, eventually, th e hyoid ro ta tio n is u n d erestim ated
by th e model. S hortening, or a p o sterio r d isplacem ent of th e sternohyoideus
m uscle (of w h ich th e p o sterio r en d is co n flu e n t w ith th e hypaxial m uscles in H.
reidi), probably explains th e discrepancy b etw een th e m odel an d o u r kin em atic
m easurem ents. As a result, w hile fo u r-b ar m odelling indicates a clear coupling
b etw een hyoid ro ta tio n and n eu ro cran ial elevation, th e detailed m orphological
d e term in a tio n of th e linkages and jo in ts of th is fo u r-b ar m odel rem ain crucial in
order to fully u n d e rstan d th is m echanism in seahorse feeding.
5.1.1. In troduction
S uction feeding is p re sen t in m ost v erteb rate groups, inclu d in g frogs (Dean,
2003), salam anders (Elw ood & C undall, 1994), tu rtle s (Van D am m e & A erts, 1997)
an d even m am m als (B loodw orth & M arshall, 2005). H ow ever, su ctio n feeding is
ub iq u ito u s in som e groups, such as te le o st fish (M uller & Osse, 1984; Lauder,
1985) an d elasm obranchs (W ilga et al., 2007). The k in em atics of su ctio n feeding
have been studied an d com pared b etw een a w ide v ariety of fish (e.g. C arro ll &
W ain w rig h t, 2003; G ibb & Ferry-G raham , 2005; V an W assen b erg h et al., 2005).
A m ong th e fish species studied, a strik in g v ariatio n in head m orphology can be
found, ranging from th e asym m etrical m orphology fo u n d in fla tfish (Gibb, 1997)
to th e m ore general laterally com pressed p erco m o rp h m orphology in larg em o u th
bass (Svanbäck et al., 2002) to th e extrem ely elongated u p p er an d low er jaw fo u n d
in long-jaw ed b u tte rfly fish (Ferry- G raham et al., 2001a). H ow ever, despite these
g reat differences in cranial m orphology, all th ese species have converged
behaviourally to p e rfo rm su ctio n feeding.
A ccording to Lauder (1985), prey cap tu re th ro u g h su ctio n is ch aracterized by fo u r
phases:
p rep aratio n ,
expansion,
com pression
an d
recovery.
D u rin g
th e
C h a p t e r 5 - Functional interpr et ati on
103
p re p a ratio n phase, th e buccal cavity is com pressed an d co n seq u en tly th e buccal
volum e is decreased. In th e su b seq u en t expansion phase, th e m o u th opens
quickly th ro u g h hyoid ro ta tio n or th ro u g h th e coupling b e tw ee n th e low er jaw,
th e
in te ro p e rc u lar bone
an d th e
operculum . A fter m o u th
opening, th e
Suspensorium abducts laterally an d th e cran iu m ro tates dorsally. Because of this
rapid expansion of th e head, th e volum e in th e buccal cavity increases rapidly,
subsequently sucking in th e prey w ith th e su rro u n d in g w ater. To ensure a
c o n tin u o u s p o sterio r flow of w ater, th e expansion m ovem ent m u st proceed from
th e fro n t to th e back of th e head. This is o fte n term ed th e ro stro cau d al expansion
sequence an d is observed in all suction-feeding teleo sts stu d ied to date (e.g.
Lauder, 1985; C arroll & W ain w rig h t, 2003; G ibb & Ferry-G raham , 2005; V an
W assen b erg h et al., 2005). D u rin g th e com pression phase, th e m o u th of th e fish
closes, th e Suspensorium adducts an d th e hyoid elevates, to g e th e r decreasing th e
volum e of th e buccal cavity. Finally, th e o p ercu lar an d b ran ch io steg al valves
open, allow ing th e sucked w a ter to flow o u t of th e buccopharyngeal cavity
th ro u g h th e gili arches. In th e last phase, th e recovery phase, all th e skeletal
elem ents re tu rn to th e ir original position.
O ne of th e m ost extrem e cranial m orphologies in te le o st fish is u n d o u b ted ly
fo u n d in syngnathid fishes. These fish are ch aracterized by an elongated tu b u la r
sn o u t w ith m in u te jaws at its end. D espite th is p ecu liar feeding m orphology, it is
unclear w h e th e r th e y have converged o n th e general te le o st suction-feeding
p a tte rn as m e n tio n e d above. H ow ever, it is k n o w n th a t seahorses and p ip efish
use su ctio n to cap tu re th e ir prey (M uller, 1987; B ergert & W ain w rig h t, 1997; de
L ussanet & M uller, 2007; V an W assen b erg h et al., 2008). M oreover, th e ir feeding
strik e is characterized by a very fast ro ta tio n of th e hyoid an d a rap id elevation of
th e head. Prey cap tu re tim es w ith in 6 ms are recorded, m aking th e m am ong th e
fastest suction-feeding v ertebrates (B ergert & W ain w rig h t, 1997; de L ussanet &
M uller, 2007; V an W assen b erg h et al., 2008). To execute such extrem ely rapid
m ovem ents, M u ller (1987) suggested a need fo r p o w er am plification. M ore
recently, V an W assenbergh et al. (2008) verified th a t th e rap id elevation of th e
head is probably accom plished by elastic recoil of th e ten d o n s of th e epaxial
muscles.
M u ller (1987) also p o stu la te d a m echanical linkage b etw een th e elevation of th e
n e u ro c ran iu m an d th e ro ta tio n of th e hyoid an d described a fo u r-b ar linkage
104
C h a p t e r 5 - Functional interpr et ati on
m odel by w h ich th e rapid dorsal ro ta tio n of th e n e u ro c ran iu m could p o w er th e
expansion of th e oropharyngeal cavity an d cause explosive suction. The pro p o sed
fo u r-b ar m odel consists of th e ceratohyal-interhyal com plex, th e sternohyoideusurohyal com plex, th e neuro cran iu m -su sp en so riu m com plex an d th e p ecto ral
girdle (Fig. 1.6). If th is fo u r-b ar linkage is k e p t in a locked p o sitio n and
accom panied by active m uscle c o n tractio n , elastic energy could p o te n tia lly be
sto red and la te r released. In pipefish, it w as originally h y p o th esized an d later
d em o n strated (M uller, 1987; V an W assen b erg h et al., 2008, respectively) th a t th e
hyoid is k e p t in a locked p o sitio n w hile th e epaxial an d hypaxial m uscles
contract. T h ro u g h sh o rte n in g of th e m uscles, th e ten d o n s of th e epaxial m uscles
le n g th e n and could store elastic energy. A sm all deviation of th e hyoid fro m its
stable p o sitio n could th e n trig g er th e release of previously sto red elastic energy,
resu ltin g in very fast m ovem ents of th e head and th e hyoid.
In spite of these hypotheses, o u r u n d erstan d in g of th ese extrem ely fast
m ovem ents is lim ited because of th e relative low tem p o ral re so lu tio n of th e
recordings of th e su ctio n event in previous studies (200-400 fram es p er second
[fps], B ergert & W ain w rig h t, 1997; 1,000 fps, de L ussanet & M u ller, 2007; b u t
2,000 fps, V an W assenbergh et al., 2008). H ow ever, to te st th e p roposed fo u r-b ar
system , a q u a n tita tiv e analysis of th e m ovem ents of th e cranial elem ents d u ring
th e strike is essential. A dditionally, a detailed stu d y of th e sp atial to p o g rap h y and
m orphology of th e join ts an d linkages of th e pro p o sed fo u r-b ar system is crucial
for o u r u n d erstan d in g of th e fu n c tio n in g of th e system.
In th is study, w e exam ined th e feeding kinem atics and cran ial m orphology in th e
seahorse species Hippocampus reidi. Q u a n tific atio n of th e m ovem ents of th e
cranial elem ents d u ring prey cap tu re com bined w ith a d etailed 3D-analysis of th e
m orphology of th e linkages in th e pro p o sed fo u r-b ar system based o n serial
histological sections and C T -scanning allow s us to te st th e validity of th e m odel
proposed by M u ller (1987). M oreover, th ese data can provide in sig h t o n h o w th e
extrem e su ctio n perform ance of th ese anim als is achieved.
C h a p t e r 5 - Functional interpr et ati on
105
5.1.2. Brief material & m e t h o d s
D iffere n t Hippocampus reidi specim ens w ere used fo r th e m orphological analyses
an d th e collection of kin em atic data.
A. M o rp h o lo g y
T hree specim ens (117.2, 113.5 a n d 109.7 m m SL; Table 2.1) w ere cleared and
stain ed according to th e p ro to co l of Taylor and V an D yke (1985). O n one of them ,
dissections w ere perfo rm ed on th e hyoid fo r a clearer view of th e artic u la tio n
facets. Serial histological cross sections (5 pm) of th e h ead of a fo u rth specim en
(103.5 m m SL) w ere made. The fifth H. reidi (119.0 m m SL) w as scanned a t th e
m odular m icro-C T setu p of G h e n t U niversity using th e d irectio n al tu b e head, at
80 kV tu b e voltage. N ext, com p u ter-g en erated 3 D -reco n stru ctio n s w ere m ade
based o n C T -data and histological sections to visualize overall m orphology. For
m ore details on m ethods, procedures or protocols, see ch ap ter 2.2.
O nly th o se m uscles, ten d o n s, ligam ents an d cartilaginous elem ents related to th e
hyoid w ere reco n stru cte d (see c h ap ter 3.2 fo r th e com plete reconstructions). The
3D -reconstruction, using th e histological data, show s a slight d isto rtio n a t th e
level of th e sn o u t (Fig. 5.1C), w h ich is p ro b ab ly an arte fa ct due to th e alignm ent.
O nly serial sections of th e cranial sk eleto n w ere used, so th e c leith ru m w as n o t
fully reco n stru cte d (see Fig. 3.8 fo r a com plete reco n stru ctio n ).
B. K in e m a tic s
Tw o days before a p la n n ed film ing session, w e sto p p ed feeding th e anim als.
D uring th e recording sessions, anim als w ere placed in a sm aller ta n k (0.3 m x 0.2
m X 0.05 m) w ith a scale bar atta ch e d to th e tank. A t th e tim e of th e experim ents,
five anim als w ere fed sm all crustaceans a ttach ed to th e o u tflo w of a p ip ette.
M ovem ents of several cranial elem ents w ere q u a n tifie d d u rin g prey cap tu re by
m eans of lateral view high-speed video recordings. The video recordings w ere
m ade using a Redlake Im aging M o tio n P ro digital cam era (Redlake, Tucson). F our
arrays of eight each red u ltra b rig h t LEDs w ere used fo r illu m in atio n . Sequences
w ere film ed a t 2,000 fps w ith a s h u tte r tim e of 0.2 ms. O n ly videos in w h ich th e
head w as o rie n te d p erp en d icu lar to th e lens of th e cam era w ere re ta in ed for
fu rth e r analysis, th u s elim in atin g c o rrectio n for parallax. A to ta l of 14 recordings
io 6
C h a p t e r 5 - Functional interpr et ati on
from five individuals w ere used, nam ely th re e videos of fo u r individuals an d tw o
of one individual.
E ight landm arks w ere digitized (Fig. 5.2) fram e by fram e using Didge softw are (A.
C ullum , C re ig h to n U niversity, O m aha, NE): (1) th e tip of th e nose spine, just
a n te rio r of th e eye; (2) th e tip of th e sn o u t p o sterio r to th e m axillary bones; (3)
th e sym physis of th e ceratohyal bones; (4) th e d istal tip of th e v e n tro lateral spine
on th e p ecto ral girdle; (5) th e tip of th e process o n th e second n u ch al plate; (6) th e
proxim al tip of th e prem axillary bone; (7) th e distal tip of th e d en tary bone; and
(8) a d istin c t p o in t on th e prey th a t could be track ed dow n d u rin g th e w hole
sequence (e.g. th e eye).
Based on th e X an d Y coordinates of these landm arks, five variables w ere
calculated: (a) m o u th opening (distance 6-7); (b) hyoid angle (angle b etw een 1-2
an d 3-p, p being th e coordinate of th e cen tre of th e in terh y al bone, of w h ich th e
p o sitio n w as first d eterm in ed based o n a C T-scan an d su bsequently recalculated
for each tim e step based on th e p o sitio n of th e n e u ro c ran iu m landm arks 1-2); (c)
n eu ro cran ial elevation (angle betw een 1-2 an d 4-5); (d) cleith ru m to hyoid
distance (distance 3-4); an d (e) prey distance (distance 6-8).
In order to reduce d ig itatio n noise, th e k in em atic profiles w ere filtered using a
fo u rth -o rd e r low -pass zero phase sh ift B u tte rw o rth filte r w ith a c u to ff frequency
of 500 H z an d velocities an d accelerations w ere calculated th ro u g h n u m erical
d iffe re n tia tio n
of
th e
sm ooth ed
displacem ent
profiles.
The
maxim al
displacem ents, velocities an d accelerations of each variable w ere m easured as w ell
as th e tim e needed to reach m axim al displacem ents. Tim e o was defined as th e
im age before th e first visible m ovem ent.
C.
F o u r-B a r M o d e l
T he coupling betw een th e ro ta tio n of th e n e u ro c ran iu m and th e hyoid was
evaluated by calculating th e o u tp u t of th e p lan ar fo u r-b ar system described by
M u ller (1987). T he follow ing landm arks w ere defin ed as th e jo in ts in th e four-bar
system (Fig. 5.2): (p) th e cen tre of th e in terh y al bone; (q) th e p o in t of artic u la tio n
of th e n e u ro c ran iu m an d th e v erteb ral c o lu m n -p e cto ral girdle complex; (r) th e
m ediorostral tip on th e cleith ru m n ear th e origin of th e sternohyoideus m uscle
(s); an d th e sym physis of th e ceratohyal bones. C T-scan im age reco n stru ctio n s
w ere used to d eterm in e th e p o sitio n s of th ese fo u r-b ar jo in ts relative to th e
C h a p t e r 5 - Functional interpr et ati on
107
digitized landm arks on th e high-speed video im ages (Fig. 5.2). T he co o rd in ates of
p an d q w ere calculated w ith respect to a fram e of reference defined by tw o
p o in ts m oving w ith th e n e u ro c ran iu m (landm arks 1-2; Fig. 5.2). Sim ilarly, th e
co o rdin ate of r is calculated w ith respect to landm arks 4-5 (Fig. 5.2). Finally,
because th e coordinate of s at tim e t = o corresponds to a digitized co o rd in ate
(landm ark 3; Fig. 5.2), th e in itia l co n fig u ratio n of th e fo u r-b ar system at t = o is
set.
From th is m o m en t on, th e distances of pq, ps, rs an d qr are k e p t co n stan t. The
neu ro cran iu m -su sp en so riu m b ar qp was chosen as th e fixed b ar (i.e. th e frame).
T he tim e-d ep en d en t change in th e angle b etw een p q and qr equalled th e change
in th e m easured n eu ro cran ial elevation angle an d was used as in p u t m otion. The
form ulas p resen ted in A erts an d V erraes (1984) w ere used to calculate th e
re su lta n t ro ta tio n of th e hyoid w ith resp ect to th e fram e u n d er th e fo u r-b ar
co n ditions described above. Because th e in vivo m easured hyoid angle (angle 1-2
to 3-p, Fig. 5.2) is also available, th e fo u r-b ar m odel p re d ic tio n can be com pared
w ith th is angle for each tim e step sam pled in th e k in em atic analysis.
In order to visualize th e spatial co n fig u ratio n of th e m odel co m p o n en ts before
an d a fte r n eu ro cran ial elevation, th e follow ing pro ced u re w as executed. Based on
th e 3D -reco n stru ctio n of th e C T-data, 3D -coordinates (bilateral w here possible)
w ere retrieved using A m ira 4.1.0 so ftw are from a selectio n of 23 a rtic u la tio n
p o in ts an d skeletal reference p o in ts (Fig. 5.3). The co o rd in ates w ere th e n
im p o rted in to R hinoceros 3.0 so ftw are to c o n stru c t a sim plified m odel of th o se
stru c tu re s involved in th e hyoid ro ta tio n an d n eu ro cran ial elevation: (1)
m andibulo-quadrate articu latio n ; (2) m an d ib u lar symphysis; (3) prem axillary
symphysis; (4) ro stral tip of th e vom eral bone; (5) distal tip of th e nose spine; (6)
a rtic u la tio n b etw een th e supraoccipital b o n e an d th e co rona (i.e. th e first n u ch al
plate); (7) distal tip of th e dorsocaudal process o n th e corona; (8) d istal tip of th e
process on th e second n u ch al plate; (9) a rtic u la tio n b etw een th e n eu ro cran iu m
an d th e first vertebra; (10) a rtic u la tio n b etw een th e n e u ro c ran iu m and th e
p ecto ral girdle; (11) distal tip of th e v e n tro la te ral spine o n th e p ecto ral girdle; (12)
tip of th e angle form ed b e tw ee n th e dorsal h o riz o n ta l arm of th e p ecto ral girdle
an d th e vertical arm; (13) ro stra l p o in t of th e p ecto ral com plex (form ed by th e
cleith ru m an d several plates of th e b o n y arm our); (14) caudal tip of th e urohyal
bone; (15) sym physis of th e ceratohyal bones, (16) a rtic u la tio n b etw een p o sterio r
io 8
C h a p t e r 5 - Functional interpr et ati on
ceratohyal bone and in terh y al bone; and (17) a rtic u la tio n b etw een in terh y al and
p reo p ercu lar bones.
D. S ta tis tic s
In order to statistically determ ine w h e th e r a difference in tim in g exists b etw een
cranial ro ta tio n , hyoid ro ta tio n an d m o u th opening, a co n tin g en cy tab le (Fisher’s
exact test) was used. D ata from feeding sequences in w h ich m ovem ents occur at
th e same tim e fram e w ere o m itte d fro m th e co n tin g en cy table. In case a tem p o ral
d istin c tio n occurs, th e contingency tab le tests w h e th e r or n o t th is difference in
tim in g is significant. N ote, how ever, th a t th is statistical te st does n o t rule o u t
th a t th e actual m ovem ents m ig h t occur at exactly th e sam e in stan t. This
sep aratio n of th e data set was used only fo r th is specific statistical analysis.
To com pare o u r m easurem ents w ith th e p red icted values of th e m odel, a p aired tte st was used at each tim e fram e to d eterm in e a t w h ich in sta n t th e p red icted and
th e m easured values statistically differ. In addition, a least squares regression
analysis was used to investigate w h e th e r a stro n g co rrelatio n exists in o ur
kin em atic data betw een th e angles of hyoid ro ta tio n an d cranial ro ta tio n d u ring
th e feeding sequence.
5.I.3. Results
A. M o rp h o lo g y
T he follow ing descrip tio n based o n th e histological d ata was co n firm ed o n in toto
cleared an d stained specim ens.
T he h y o m andibular bone bears th re e cartilaginous a rtic u la tio n heads: tw o
dorsally for th e sp h en o tic bone and th e cartilaginous surface b etw een th e
sphenotic, th e p ro o tic an d th e p te ro tic bones respectively and a laterocaudal
ro u n d ed condyle for th e opercular b o n e (Fig. 5. i A,C). B oth th e o p ercu lar and
n eu ro cran ial processes are su rro u n d e d by ligam ents. A lateral an d a m edial w ing
on th e hyom andibular bone form a groove in w h ich th e levator arcus p alatin e
m uscle lies dorsally an d th e p reo p ercu lar b o n e in te rd ig itate s w ith th e v en tral side
of th e lateral w ing. The hyom an d ib u lo -sp h en o tic lig am en t is co n n ected to th e
distal tip of th e lateral w ing and a lig am en t co n n ected to th e p arasp h en o id bone
C h a p t e r 5 - Functional interpr et ati on
109
attaches on th e m edial wing. A long its m edial len g th , th e hy o m an d ib u lar bone is
provided w ith a ten d in o u s a tta c h m e n t of th e a d d u cto r arcus p alatin e muscle.
V entrorostrally, th e re is a synchondrosis b etw een th e hy o m an d ib u lar bone and
th e sym plectic bone (Fig. 5.1A). A firm lig am en t ru n s b etw een th e hy o m an d ib u lar
an d th e in terh y al bones. The hy o m an d ib u lar b o n e in Hippocampus reidi is n o t
co n n ected to th e m etapterygoid or in te rh y a l bones, as it usually is in teleosts
(Gregory, 1933; H a rrin g to n , 1955; Rojo, 1991). A ccording to Ju n g ersen (1910) th e
c o n ta c t b etw een th e m etapterygoid an d th e h y o m an d ib u lar bones is lo st in th e
fam ilies C entriscidae, A ulostom idae, Solenostom idae an d S yngnathidae, w hereas
A nker (1974) show ed th a t it is still p re se n t in Gasterosteus aculeatus.
T he p reo p ercu lar bone is a very long L-shaped derm al b o n e (Fig. 5.1B). Its
h o riz o n ta l b ra n c h stretches o u t ro strally an d is atta ch e d to th e q u ad rate and th e
in te ro p e rc u lar bones th ro u g h ligam ents. C audally, th e v ertical b ran ch is tig h tly
co n n ected
to
th e
lateral
face
of th e
hy o m an d ib u lar
bone
th ro u g h
an
in te rd ig ita tio n and connective tissue. M edially, it bears a ridge th a t su p p o rts th e
sym plectic bone along its le n g th an d th e levator arcus p a la tin i m uscle fits in to a
lo n g itu d in al groove (Fig. 5.1A). Laterally, it bears a spine. The levator arcus
p a la tin i m uscle inserts ten d in o u sly m ediocaudally o n to th e p reo p ercu lar bone.
T he a rtic u la tio n w ith th e in terh y al b o n e is situ a ted v en trally w here th e
h o riz o n ta l an d vertical branches m eet. Ligam ents a ttach ed to th e v en tral surface
of th e p reo p ercu lar bone enclose th e jo in t fo r th e in terh y al bone. The
p reo p ercu lar bone su p p o rts th e sn o u t laterally an d form s a large p a rt of th e
Suspensorium .
T he p erich o n d ral in terh y al bone is small, solid an d bears a lateral an d a m edial
head ventrally (Fig. 5.4). D orsocaudally, th e ro u n d ed cartilaginous surface form s a
condyle th a t fits in to th e socket jo in t of th e p reo p ercu lar bone. In b etw een th e
lateral and m edial process of th e in te rh y a l bone, a saddle jo in t is fo rm ed w h ere
th e p o sterio r ceratohyal bone articulates. The lateral process bears ligam ents th a t
ru n to th e p o sterio r ceratohyal an d p reo p ercu lar bones, w hile massive ligam ents
from th e sym plectic and th e h y o m an d ib u lar bones a tta c h o n to th e m edial head
dorsally an d ventrally, respectively. Ligam ents fro m b o th processes th u s enclose
th e jo in t cavity betw een th e in terh y al an d th e p o sterio r ceratohyal bones. The
plane of th e a rtic u la tio n facet b etw een th e p reo p ercu lar b o n e and th e in terh y al
110
C h a p t e r 5 - Functional interpr et ati on
bone is o rie n te d obliquely. T he o rie n ta tio n of th e long axis of th e in terh y al bone
ru n s from ventrom edially to slightly dorsolaterally (Fig. 5.1A).
T he p erich o n d ral p o sterio r ceratohyal b o n e is very irregularly shaped. A p art from
th e caudal saddle-shaped facet, w h ich articu lates in b etw een th e tw o processes of
th e in terh y al bone, tw o heads are p re sen t o n th e proxim al p a rt of th e p o sterio r
ceratohyal bone: a lateral one and a v en tral cartilaginous one (Fig. 5.4). The
intero p ercu lo -h y o id ligam ent, co n n ectin g th e in te ro p e rc u lar b o n e w ith th e
p o sterio r ceratohyal bone, attach es ro stral to th is lateral process. The slender
te n d o n of th e p ro tra c to r hyoidei m uscle ru n s v en tral to th e in tero p ercu lo -h y o id
ligam ent an d attaches on th e lateral side of th e p o sterio r ceratohyal b o n e (Fig.
5.1B, 5.4A). Its lateral b o rd er su p p o rts th e tw o slender b ran ch io steg al rays. A t its
distal end, th e p o sterio r ceratohyal b o n e bears a th in , ta p erin g process th a t
in te rd ig itate s w ith th e a n terio r ceratohyal bone. A p art fro m th is in te rd ig itatio n ,
th e re is a firm synchondrosis and a ligam entous co n n ec tio n b etw een th e a n terio r
an d p o sterio r ceratohyal bones.
T he a n te rio r ceratohyal bone th a t in co rp o rates th e sm all hypohyal b o n e is larger
th a n th e p o sterio r one and is m ore or less trian g u larly shaped (Fig. 5.4). The apex
of th e trian g le is situ a ted rostrally, w h ere th e re is a sm all b u t very firm m edial
sym physis com posed of connective tissue b etw een th e distal en d of th e le ft and
rig h t a n te rio r ceratohyal bones. M o re proxim ally, a lig am en t connects th e la tte r
w ith th e very sm all cartilaginous basihyal bone. V entrally, a t th e apex, a lig am en t
ru n n in g to th e urohyal bone attach es (Fig. 5.1, 5.4A). The m andibulo-hyoid
ligam ent runs from th e d en tary bone as a single u n it along th e snout. W h e n it
reaches th e hyoid, it separates in to th re e parts. All p a rts a tta c h to g e th e r m edially
in a cavity of th e a n te rio r ceratohyal bone. C audally, th e a n terio r ceratohyal bone
bears several slender processes in b etw een w h ich th e long rod of th e p o sterio r
ceratohyal bone fits.
T he urohyal bone is a small, m ore or less cylindrical bone. R ostrally, it bears a
th in dorsom edial ridge, giving th e urohyal b o n e a trian g u la r shape at its fro n t in
cross section (Fig. 5.4). A sh o rt and sto u t lig am en t attach es on b o th sides of th e
ridge an d ru n s to th e a n terio r ceratohyal bone. The te n d o n of th e sternohyoideus
m uscle encloses th e urohyal bone caudally (Fig. 5.1).
T he large cleith ru m form s th e m ajority of th e p ecto ral girdle. In lateral view, it is
roughly com m a-shaped, w ith a bigger dorsal p a rt and a ro strally b e n t v en tral p a rt
C h a p t e r 5 - Functional interpr et ati on
i l l
(Fig. 5.1B). The dorsal c o m p o n en t is fla tte n e d laterally except fo r a m edial b ran ch
th a t in terd ig itates firm ly w ith th e processus transversus of th e first v erteb ra (Fig.
5.1A). Laterally, th is p a rt is m ostly covered by th e dorsocaudal edge of th e
o percular bone. T here is also a dorsocaudal in te rd ig ita tio n w ith th e second
n u ch al plate, w h ich lies p o sterio r to th e corona. R ostrally, a solid lig am en t ru n s
from th e ro stral tip of th e c leith ru m to th e caudal en d of th e exoccipital bone.
This tip fits b etw een th e caudal en d of th e p o sttem p o ral bone laterally and th e
exoccipital bone m edially, form ing th e a rtic u la tio n b etw een th e p ecto ral girdle
an d th e neurocranium . A ligam en to u s co n n ectio n w ith th e p o stte m p o ra l bone,
w h ich form s p a rt of th e n e u ro c ran iu m in H. reidi, is lacking. The v en tral p a rt of
th e cleith ru m is bifu rcated , w ith a m edial b ra n c h c o n tac tin g th e scapulacoracoid
an d a lateral b ra n c h ru n n in g ventrally, w here it in te rd ig itate s w ith th e bony
arm o u r form ing som e so rt of p ecto ral com plex (Fig. 5.1C). As such, th e m edial
an d lateral branches enclose a heart-sh ap ed cavity in to w h ich th e sternohyoideus
m uscle and p ecto ral fin m usculatu re lie. These tw o branches fuse m ore caudally
an d form a bony sheet dorsal to th e muscles.
T he p ro tra c to r hyoidei m uscle is generally considered to be em bryologically
form ed by th e fusion of th e in term an d ib u laris p o sterio r and th e a n terio r
interh y o id eu s m uscle (W in te rb o tto m 1974), b u t see G eerinckx an d A driaens
(2007) for a discussion of possible m isin te rp re ta tio n s o n th e tru e n a tu re of th is
m uscle. In H. reidi, how ever, b o th p arts can still be clearly d istin g u ish ed from
each o th e r because th e y are w ell separated by a te n d in o u s c o n n ectio n (Fig. 5.1C).
R ostrally, tw o sh o rt te n d o n s are co n n ected to th e m edial face of th e le ft and rig h t
d en tary bones, ven tral to th e a n te rio r in term an d ib u laris m uscle (connecting th e
tw o d en tary bone). The p o sterio r in term an d ib u laris p a rt starts as tw o individual
bundles of fibres fusing alm ost halfw ay along th e ir le n g th w h en reaching th e
te n d in o u s connection. Fibres of b o th th e p o sterio r in term an d ib u laris and th e
a n te rio r interh y o id eu s m uscle ru n in a lo n g itu d in al d irection. A lth o u g h n o t
visible on th e reco n stru ctio n , th e serial sections show th a t along its e n tire length,
th e a n terio r interh y o id eu s consists of tw o b u n d les separated in th e m iddle by a
th in layer of connective tissue. M o re caudally, th e bundles diverge, giving th e
m uscle its typical X shape. A tta ch m en t by a long an d slender te n d o n occurs on
th e lateral face of th e p o sterio r ceratohyal, ju st v en tral to a tta c h m e n t of th e
intero p ercu lo -h y o id ligam ent (Fig. 5.1B, 5.4A).
112
C h a p t e r 5 - Functional interpr et ati on
T he sternohyoideus m uscle inserts ro strally o n th e urohyal b o n e by a firm te n d o n
th a t caudally splits in to tw o an d co n tin u es to ru n in b etw een th e fibres (Fig. 5.1).
T he m uscle is enclosed by ven tral bony p lates of th e body and p artly by th e
lateral and m edial b ra n c h of th e cleith ru m nearly along its e n tire length. M ore
caudally, th e bony arm o u r extends laterally, w ith th e p ecto ral fin m u scu latu re
being situ a ted dorsally. The site of origin of th e stern o h y o id eu s m uscle is th e
cleithrum , th e scapulacoracoid an d th e v en tral b o n y p lates o n to w h ich th e
m uscle attaches th ro u g h a layer of connective tissue. T he tw o large and firm
bundles of th e sternohyoideus m uscle course v e n tro ro stra lly and u n ite rostrally.
C audally th ey join th e fibres of th e hypaxial muscle.
B. K in e m a tic s
D uring th e p rep arato ry phase, th e anim al approaches its prey relatively slowly.
N ext, th e expansion phase is characterized by a fast ro ta tio n of th e hyoid an d a
ro ta tio n of th e head and snout. T he m o u th starts to o p en a few m illiseconds later
(Table 5.1; Fig. 5.5). C ran ial elevation causes th e m o u th o p en in g to be p o sitio n ed
in close proxim ity to an d directed to w ard s th e prey, w h ich is finally sucked in to
th e m outh. A fter prey capture, a relatively long p erio d is n eeded to resto re th e
original c o n fig u ra tio n of th e skeletal elem ents (Fig. 5.5).
T he cran iu m starts to elevate 0.14 ± 0.08 ms (m ean ± SE) after th e o n set of hyoid
ro ta tio n (Table 5.1; Fig. 5.6A). In eig h t of th e 14 analyzed sequences, feeding
began w ith hyoid ro ta tio n . In six sequences, hyoid ro ta tio n an d n eu ro cran ial
elevation started at th e same video image. T herefore, th e o n set of n eu ro cran ial
elevation did n o t precede th e o n set of hyoid ro ta tio n in any of th e sequences.
C onsequently, in th e eight sequences w h ere a clear tem p o ral d istin c tio n b etw een
b o th m ovem ents is p resent, th e o n set of hyoid ro ta tio n is sig n ifican tly earlier
th a n th e o n set of n eu ro cran ial elevation (Fisher’s exact test, P = 0.04). This resu lt
w as su p p o rte d by prelim in ary data based o n videos recorded a t 8,000 fps.
T he o n set of m o u th opening takes place 0.11 ± 0.01 ms a fte r th e o n set of
n eu ro cran ial elevation (and th u s 0.25 ± 0.07 ms a fte r th e o n set of hyoid rotation).
T he o n set of cranial elevation preceded th e o n set of m o u th o p en in g in 10 of th e
14 sequences. In th e fo u r rem aining sequences, th e onsets of n eu ro cran ial
elevation an d m o u th opening occurred a t th e sam e video image. C onsequently,
th e o n set of n eu ro cran ial elevation is sig n ifican tly earlier o n average th a n th e
C h a p t e r 5 - Functional interpr et ati on
113
o n set of m o u th opening (Fisher s exact test, P = 0.02) in 10 videos w h ere a clear
tem p o ral d istin c tio n w as apparent. M o u th o p en in g reaches a peak value of 0.27 ±
0.16 cm 3.50 ± 0.28 ms a fte r th e b eg in n in g of hyoid ro tatio n .
T he prey is sucked in 2.00 ± 0.16 ms a fte r th e m o u th reaches its m axim um gape
(5.50 ± 0.44 ms a fte r th e o n set of hyoid ro tatio n ). N ext, th e hyoid reaches its peak
excursion of 68.45o ± 4-32° 17-29 ± !-58 ms a fte r th e o n set of hyoid ro ta tio n . The
n e u ro c ran iu m is th e last elem en t reaching its peak elevation of 31.100 ± 2.130
18.46 ± 1.59 ms a fte r th e o nset of hyoid ro ta tio n . D u rin g th e e n tire feeding strike,
th e re exists a stro n g co rrelatio n b etw een th e cran ial elevation an d th e hyoid
ro ta tio n (correlation test, N = 51, R2 = 0.99, P < 0.01).
T he d u ra tio n of th e expansion phase (w hich begins w ith th e o n set of hyoid
ro ta tio n an d ends w ith th e beg in n in g of m o u th closing) is 5.77 ± 0.66 ms. The
com pression phase (w hich is defin ed here as startin g at th e o n set of m o u th
closing an d ending w h en th e m o u th is fully closed) lasts m u ch longer, w ith a
d u ra tio n of 415.74 ± 48.20 ms. T he final recovery phase (w hich begins a t th e
m o m en t th e jaws are fully closed an d ends a t th e tim e all skeletal elem ents have
acquired th e ir original position) has a d u ra tio n of 168.23 ± 36.27 ms. The to ta l
d u ra tio n of th e feeding sequence (from one fram e before th e o n set of hyoid
ro ta tio n u n til th e end of th e recovery phase) is 598.73 ± 25.12 ms, o n average.
T a b l e 5 .1 - S u m m a r y k i n e m a t i c s o f t h e p r e y c a p t u r e e v e n t i n H i p p o c a m p u s r e i d i ( N = 14). S E ,
s ta n d a r d e rr o r.
V ariable
P e a k h y o id r o ta tio n Q 1
P e a k v elo city h y o id r o ta tio n (Vs)
P e a k a c c e le ra tio n h y o id r o ta tio n (Vs2)
O n s e t h y o id r o ta tio n (ms)
T im e to p e a k h y o id r o ta tio n (ms)
T im e to p e a k v elo city h y o id r o ta tio n (ms)
T im e to p e a k a c c e le ra tio n h y o id r o ta tio n (ms)
P e a k c ra n ia l e le v a tio n (°) 1
P e a k v elo city c ra n ia l e le v a tio n (Vs)
P e a k a c c e le ra tio n c ra n ia l e le v a tio n (Vs2)
O n s e t c ra n ia l e le v a tio n (ms)
T im e to p e a k c ra n ia l e le v a tio n (ms)
T im e to p e a k v elo city c ra n ia l e le v a tio n (ms)
T im e to p e a k a c c e le ra tio n c ra n ia l e le v a tio n (ms)
P e a k m o u th o p e n in g (mm) 1
P e a k v elo city m o u th o p e n in g (m/s)
P e a k a c c e le ra tio n m o u th o p e n in g (m /s2)
O n s e t o f m o u th o p e n in g (ms)
T im e to p e a k m o u th o p e n in g (ms)
T im e to p e a k v elo city m o u th o p e n in g (ms)
T im e to p e a k a c c e le ra tio n m o u th o p e n in g (ms)
( c o n tin u e d o n n e x t page)
M ea n
68.45
29.85 X i o 3
23.22 X i o 6
0.50
17.29
1.29
0.50
31.10
13.88 X i o 3
10.36 X i o 6
0.64
18.46
1.36
0.50
2.70
1.58
1.22 X i o 3
0.75
3.50
1.21
0.46
SB
4.32
2.42 X i o 3
1.84 X i o 6
0.00
1.58
0.07
0.00
2.13
976.77
709.71 X i o 3
0.08
1-59
0.06
0.00
0.16
0.12
116.53
O.O7
O.28
O.O9
O.O4
C h a p t e r 5 - Functional interpr et ati on
114
V ariable
P e a k sh o rte n in g p e c to ra l girdle - c e ra to h y a l tip (mm) 1
P e a k v elo city s h o rte n in g p e c to ra l gird le - c e ra to h y a l tip (m/s)
P e a k a c c e le ra tio n sh o rte n in g p e c to ra l g ird le - c e ra to h y a l tip (m /s2)
O n s e t o f sh o rte n in g p e c to ra l girdle - c e ra to h y a l (ms)
T im e to p e a k s h o rte n in g p e c to ra l girdle - c e ra to h y a l tip (ms)
T im e to p e a k v elo city sh o rte n in g p e c to ra l gird le - c e ra to h y a l tip (ms)
T im e to p e a k a c c e le ra tio n s h o rte n in g p e c to ra l girdle - c e ra to h y a l tip (ms)
D u r a tio n o f p re y c a p tu re (ms)
P re y d istan c e a t tim e o (mm)
P e a k v elo city p re y (m/s)
P e a k a c c e le ra tio n p re y (m /s2)
T im e to p e a k v elo city p re y (ms)
T im e to p e a k a c c e le ra tio n p re y (m /s2)
1 C a lc u la te d as th e d iffe re n c e b e tw e e n th e value a t tim e o m s a n d th e m ax im u m value.
C.
M ea n
1.05
55.12
70.24 X i o 3
0.82
12.80
0.54
0.20
5.50
6.71
0.27
190.37
5.10
4.00
SE
0.18
8.45
7.80 X i o 3
0.19
0.25
0.18
0.05
0.44
0.53
0.06
57.62
0.86
O.9I
F o u r-B a r M o d e l
T he hyoid angle calculated th ro u g h th e fo u r-b ar m odel reaches a m axim um peak
of 88.99o ± 5-29° (m ean ± SE) a fte r 18.46 ± 1.49 ms. The m easured hyoid angle is on
average 10.99o larger, w hile th e tim in g of peak excursion differs only by 1.17 ms
(Fig. 5.6B). D uring th e first 1.50 ms, th e m easured hyoid angle statistically shows
no difference from th e p red icted hyoid angle (t-test, t = 1.514, df = 13, P = 0.154).
A fter th is in itia l phase, th e p red icted angle started to u n d erestim ate th e
m easured hyoid angle significantly (t-test, t = 3.034, df = 13, P < 0.01 a fte r tim e =
2.00 ms) an d finally ends at 91.08% of th e m easured angle (Fig. 5.6B). This p a tte rn
is co n sisten tly observed for each individual feeding sequence analyzed.
T he
m orphological
analysis
confirm s
th a t
th e
n e u ro c ran iu m
an d
th e
Suspensorium can be rep resen ted as a single u n it (as assum ed in th e four-bar
model) artic u la tin g w ith th e follow ing stru ctu res: th e p ecto ral girdle, th e
v erteb ral colum n, th e ceratohyal bones (th ro u g h th e in te rh y a l bone) and th e
low er jaws (Fig. 5.7). T he u p p er jaws articu late w ith th e n eu ro cran iu m , b u t w ith
respect to th e four-bar linkage system , th e y can be considered as m echanically
coupled to th e n eu ro cran ial unit. The degrees of freedom b etw een th e
n eu ro cran iu m , p ecto ral girdle and v erteb ral colum n are reduced because th e
v erteb ral colum n is im m ovably co n n ected to th e p ecto ral girdle. As such, th e
p ecto ral girdle can be considered to be p a rt of th e p o stcran ial skeletal un it. This
also im plies th a t th e p o stcranial skeletal u n it articu lates w ith th e n eu ro cran iu m
w ith th re e joints: tw o for th e pecto ral girdle an d one fo r th e first vertebra. All
th re e join ts show a top o g rap h y w h ere th e y are w ell aligned (so th e v erteb ral
a rtic u la tio n is lin in g up w ith th e axis b etw een th e le ft and rig h t p ecto ral
articulation).
C h a p t e r 5 - Functional interpr et ati on
115
As n eu ro cran ial ro ta tio n m u st occur a ro u n d th e axis going th ro u g h th e th re e
n eu ro cran ial joints, its elevation is coupled to a p o sterio r displacem ent of th e
su praoccipital crest w ith respect to th e nu ch al plates. As th e second n u ch al p late
is strongly co n n ected to b o th th e p ecto ral girdle an d th e n eu ral spine of th e first
vertebra, a p o sterio r ro ta tio n of th e su p rao ccip ital crest m u st be com pensated for
by a sliding or tiltin g actio n of th e corona, w ith resp ect to th e second nu ch al
plate a n d /o r th e neurocranium . A t th is p o in t, it c an n o t be discerned w h e th e r th is
m ovem ent involves one or b o th actions, b u t th e m odel does suggest th a t a
fo rw ard m ovem ent of th e corona w ith respect to th e su p rao ccip ital crest occurs
(indicated by th e sh o rten ed distance of th e b ar co n n ectin g th e supraoccipital
crest and th e distal tip of th e process o n th e corona; Fig. 5.7B).
U sing th e im age of a feeding H. reidi a t m axim al n eu ro cran ial elevation, w here
th e depressed hyoid bars could clearly be observed, th e m odel could only be fitte d
p roperly if th e in te rh y a l bones w ere ro ta te d fo rw ard (thus c o n tra ry to w h a t
w o uld be expected for hyoid rotatio n ). O bservations of dissected specim ens also
in d icated th e in terh y al bone to be directed ro strally w h en th e hyoid is depressed.
W h e th e r or n o t th e in terh y al b o n e is also d irected ro strally a t th e o n set of
n eu ro cran ial elevation could n o t be d eterm in ed based o n th e m odel (Fig. 5.7).
5.1.4. D is cu ssi on
A. C o m p a ris o n w ith O th e r T e le o s ts
A com parison of th e k inem atics of th e feeding strik e of Hippocampus reidi w ith
th e rostrocaudal expansion sequence of typical su ctio n feeding teleo sts reveals
differences in th e tim in g of th e m ovem ents of th e skeletal com ponents. W h e n
th e re was a clear tem p o ral d istin ctio n , th e hyoid (and n o t th e m o u th , as observed
for typical te le o st su ctio n feeders; see, e.g. Lauder, 1985; C arro ll & W ain w rig h t,
2003; G ibb & Ferry-G raham , 2005; V an W assen b erg h et al., 2005) in H. reidi is th e
first c o m p o n en t of th e cranial system th a t is set in m otion. H ow ever, in six of a
to ta l of 14 sequences, hyoid ro ta tio n an d cranial elevation occur a t th e sam e tim e
fram e; th erefo re, w e c an n o t statistically ju stify w h e th e r or n o t a difference in
tim in g exists. In those cases w here o u r tem p o ral reso lu tio n was adequate and th e
tw o behaviours show ed tem p o ral separation, hyoid m ovem ents always preceded
C h a p t e r 5 - Functional interpr et ati on
ii6
cranial ro tatio n . C onsidering th e very fast m ovem ents, th is a p p aren t overlap in
tim in g could be caused by th e low tem p o ral reso lu tio n of th e recordings.
T herefore, fu rth e r analyses using h ig h er tem p o ral re so lu tio n are needed to
co n firm th is finding an d te st its generality. The m o u th is d irected to th e prey via
cranial elevation, w h ich occurs ju st a fte r th e o n set of hyoid ro ta tio n . C ran ial
elevation th u s appears to be used fo r p o sitio n in g th e m o u th o p en in g tow ards th e
prey ra th e r th a n c o n trib u tin g to m o u th opening (also fo u n d in clariid catfish;
V an W assen b erg h et al., 2006a), w h ich also differs fro m th e typical te le o st p a tte rn
(e.g. Lauder, 1985). Prey cap tu re occurred w ith in 5-6 ms, w h ich m akes H. reidi
one of th e fastest su ctio n feeders to date, com parable w ith th e re p o rte d prey
cap tu re tim es of a n ten n a riid anglerfish (G robecker & Pietsch, 1979).
B.
K in e m a tic s a n d M o rp h o lo g y R e la te d to th e F o u r-B a r S y ste m
T he four-bar linkage proposed by M u ller (1987) consists of th e n eu ro cran iu m suspensorium com plex, th e hyoid, th e stern ohyoideus-urohyal com plex an d th e
v erteb ral c o lu m n -p e cto ral girdle com plex. In H. reidi, th e a rtic u la tio n b etw een
th e Suspensorium an d th e hyoid is provided by th e in te rh y a l bone. In a
generalized teleost, th is sm all bo n e is usually rod-shaped an d bears a ro u n d ed
head th a t fits in to a facet of th e Suspensorium , form ing a ball-and-socket jo in t
(Anker, 1989; A erts, 1991). This c o n fig u ra tio n p erm its th e in terh y al b o n e to
ro ta te in every d irectio n w ith resp ect to th e Suspensorium (unless re stric ted by
th e Suspensorium itself). In H. reidi, how ever, th e in terh y al bone is reduced to
only th e ball of th e ball-and-socket jo in t, w h ich bears tw o d istin c t v en tral
processes (Fig. 5.4). The p o sterio r ceratohyal bone, form ing a rigid u n it w ith th e
a n te rio r ceratohyal bone, articu lates in b etw een th ese heads. The tw o heads of
th e in terh y al bone reduce th e degrees of freedom b etw een th e hyoid an d th e
Suspensorium . Thus, m ovem ents are largely co n fin ed to a sagittal plane going
th ro u g h th e long axis of th e in terh y al bone. This m ovem ent is very im p o rta n t for
hyoid ro ta tio n d u ring th e expansion phase of su ctio n feeding (Aerts, 1991). The
m orphological specialization of th e in terh y al b o n e does suggest th a t th is bone
an d its m ovem ents play an im p o rta n t role in th e suctio n -feed in g m echanism and
c an n o t be neglected.
T he hyoid b ar an d th e sternohyoideus-urohyal com plex are co n n ected th ro u g h a
ligam ent th a t runs from th e ven tral apex of th e a n terio r ceratohyal b o n e to th e
C h a p t e r 5 - Functional interpr et ati on
117
dorsom edial ridge of th e urohyal b o n e (Fig. 5.1, 5.4A). T he a tta c h m e n t of th is
ligam ent is th e p o sitio n w here th e force is applied d u rin g sternohyoideus
co n tractio n . As th e distance betw een th e a tta c h m e n t and th e ro ta tio n cen tre (the
a rtic u la tio n w ith th e in terh y al bone) is m axim al, th e m o m en t arm and, as a
consequence, th e m o m en t of force w ill be large as th e hyoid becom es depressed
(and th u s th e angle betw een th e hyoid an d th e sternohyoideus m uscle becom es
m ore favourable). C audally, th e te n d o n of th e stern o h y o id eu s m uscle encloses th e
urohyal bone. Surprisingly, several figures in M u ller (1987) show a w o rk in g line
of th e sternohyoideus m uscle form ing an angle w ith th e urohyal bone. This is in
c o n tra d ictio n to his previous d e fin itio n of th e stern ohyoideus-urohyal com plex,
w h ich com prises th e sternohyoid eu s m uscle and th e urohyal bone as a single bar.
Also, anatom ically, th is is im possible, because th e urohyal b o n e in teleosts
typically ossifies w ith in th e te n d o n of th e stern o h y o id eu s m uscle an d w ill
th e re fo re always be p o sitio n ed in th e w o rk in g line of th is m uscle (and th u s in th e
line th ro u g h th e ten d o n ; de Beer, 1937; P atterso n , 1977).
O u r com parison b etw een th e results of th e fo u r-b ar system (M uller, 1987) and
th e observed kinem atics show good ag reem en t d u rin g th e first 1.5 ms of th e
feeding strik e (Fig. 5.6B). A fter th e in itia l 1.5 ms, th e trajecto ries sta rt to diverge
an d a clear difference betw een th e tw o m ovem ent profiles becom es apparent.
G iven th e above m orphological considerations, th is is m ost likely caused by
sh o rte n in g of th e sternohyoideu s-u ro h y al “b ar” (rs in Fig. 5.2). Indeed, calculating
th e distance betw een th e hyoid an d th e tip of th e cleith ru m show s an average
sh o rte n in g of 1.54 mm. A second possible ex p lan atio n of th e discrepancy b etw een
th e four-bar m odel and th e m easured kinem atics (Fig. 5.6B) is ab d u ctio n of th e
h y o m andibular bone; because th is im plies a lateral (outw ard) d isplacem ent of th e
in terh y al join t, th e p rojected le n g th of th e ceratohyal bones o n th e dorsoventral
plane (i.e., th e four-bar plane) w ill decrease. H ow ever, calculating th e distance
b etw een th e hyoid tip an d th e in terh y al bone in th e dorso v en tral p lan e never
show ed a sig nificant decrease d u rin g th e phase of rap id n eu ro cran ial elevation.
T herefore, it appears th a t defining th e stern ohyoideus-urohyal com plex as a bar
of c o n sta n t le n g th an d n o t Suspensorium ab d u ctio n is responsible fo r th e
u n d ere stim atio n of th e actual ro ta tio n of th e hyoid. It should also be n o te d th a t
given th e fu n c tio n a l c o n tin u u m of th e sternohyoideus an d th e hypaxial m uscles
C h a p t e r 5 - Functional interpr et ati on
ii8
in H. reidi (chapter 3.2), hypaxial m uscle sh o rte n in g m ay c o n trib u te to th e
observed decrease in le n g th b etw een th e hyoid tip an d th e c leith ru m tip.
D ue to th e im m obile co n n ec tio n b etw een th e p ecto ral girdle an d th e vertebrae,
th e pecto ral girdle an d v erteb ral colum n should be considered a single
m echanical u n it (at least w ith respect to th e fo u r-b ar linkage system). This
sim plification has also b een m ade by M u ller (1987); he elim in ates th e m ovable
co n n ec tio n betw een th e pecto ral girdle and n eu ro cran iu m and co n tin u es to w o rk
w ith th e verteb ra-n eu ro cran iu m articu latio n . This co n fig u ratio n seems to
indicate a stre n g th en e d b u t m obile co n n ec tio n b etw een n e u ro c ran iu m and
v erteb ral system by th is p ectoral-v erteb ral a n k y lo sis6, as th e neu ro cran io v erteb ral a rtic u la tio n lies o n to th e axis co n n ectin g th e le ft an d rig h t p ecto ral
a rticu latio n s w ith th e neurocran iu m . F rom a m echanical p o in t of view, any
deviation from th is axis w ou ld be p ro n e to cause a dislocation d u rin g extensive
n eu ro cran ial elevation.
A ccording to B ergert an d W a in w rig h t (1997), th e co m p o n en ts of th e M u ller fourb ar linkage m odel are th e h y o m an d ib u lar bone, th e ceratohyal bone, th e
sternohyoideus m uscle to g e th e r w ith th e urohyal b o n e and th e p ecto ral girdle.
T hey give a d escription of th e Hippocampus erectus skull, w ith th e hy o m an d ib u lar
bone covering th e p reo p ercu lar b o n e and artic u la tin g w ith th e in terh y al bone
(B ergert & W ain w rig h t, 1997). H ow ever, th is c o n fig u ra tio n c an n o t be fo u n d in
H. reidi, Hippocampus capensis, or Hippocampus kuda.
C. F in a l C o n c lu s io n
In conclusion, o u r m orphological an d k in em atic d ata clearly show an im p o rta n t
coupling b e tw ee n th e hyoid ro ta tio n an d n eu ro cran ial elevation in H. reidi, as
suggested by M u lle r’s fo u r-b ar m echanism (1987). H ow ever, o u r results do reveal
a sig nificant difference betw een th e m odel and th e observed k in em atics n ear th e
en d of th e n eu ro cran ial elevation phase, w h ich p o in ts o u t th a t th e m odel does
n o t en tirely describe th e actual m ovem ents d u rin g a feeding strik e in H. reidi.
Because th e sternohyoideus m uscle ru n s th ro u g h th e p ecto ral girdle an d is
c o n flu e n t w ith th e hypaxial m usculature, m odelling th e sternohyoideus-urohyal
com plex as a b ar of c o n sta n t le n g th is n o t ap p ro p riate. O u r data show th a t th e
6 a n a n k y lo sis is t h e u n io n o f t w o o r m o r e b o n e s th u s fo r m in g a s in g le u n it
C h a p t e r 5 - Functional interpr et ati on
119
extrem e perform ance of th e feeding system in seahorses is m ade possible by a
series of m orphological specializations of th e cranial system.
120
C h a p t e r 5 - Functional interpr et ati on
5.2.
121
M e c h a n i c a l s t r e s s d i s t r i b u t i o n in t h e f e e d i n g a p p a r a t u s
M o d i f i e d fr o m :
L e y s e n H , D u m o n t E R , B r a b a n t L, V a n H o o r e b e k e L, A d r ia e n s D .
D e a lin g w i t h str e ss in t h e fe e d in g a p p a ra tu s o f se a h o r se s a n d
p ip e fis h e s
Z o o l o g i c a l J o u r n a l o f t h e L in n e a n S o c i e t y , s u b m i t t e d
A bs tract
Pipefishes an d seahorses (Syngnathidae) are extrem ely fast su ctio n feeders, w ith
prey cap tu re tim es of less th a n 6 ms. These fast strikes are likely to resu lt in large
an d rapid pressure drops in th e buccal cavity. The h ig h pressure grad ien ts th a t
occur d u ring su ctio n feeding im ply heavy m echanical loading o n th e cranium ,
th u s th e feeding ap p aratu s is th o u g h t to experience h ig h am o u n ts of stress. By
m eans of a fin ite elem en t analysis (PEA), w e in v estig ated w h ere stress
accum ulates u n d er stro n g su ctio n pressure and w h e th e r th e re is a difference in
th e d istrib u tio n of craniofacial stress b etw een long- an d sh o rt-sn o u te d species.
W e p red icted th a t h igh stress co n cen tratio n s w o u ld occur a t th e a rticu latio n s
an d in th e cartilaginous regions of th e cranium . W e also p red icted th a t, given th e
sam e pressure, th e skulls of lo n g -sn o u ted species ex h ib it low er stress levels th a n
122
C h a p t e r 5 - Functional interpr et ati on
th e skulls of sh o rt-sn o u ted species, since th e fo rm er are th o u g h t to be
stru c tu ra lly b e tte r ad ap ted to deal w ith high su ctio n pressures. T he results
p artially su p p o rt o u r first hypothesis: except fo r Doryrhamphus dactyliophorus
(whose un ch aracteristic stress p a tte rn w e could n o t adequately explain), all
m odels show peak stress co n cen tratio n s at th e a rticu latio n s (e.g. suspensorion eu ro cran ial articu latio n ) an d cartilaginous regions (e.g. m esethm oid-parietal
tran sitio n ). O n th e o th e r hand, th e relatio n sh ip b etw een sn o u t le n g th and
p a tte rn s of stress d istrib u tio n w as m ore d ifficu lt to assess. W e did n o t fin d a
sim ple re la tio n sh ip b etw een sn o u t le n g th and th e m agnitudes of stress p red icted
by th e PEA. In an a tte m p t to explain th is, w e evaluated o u r m ethodology by
assessing th e effect of hyoid p o sitio n an d m odel c o n stru c tio n o n th e stress
d istrib u tio n .
5.2.1. In troduction
A m ong teleosts, th e m ost w idespread m eth o d of prey cap tu re is su ctio n feeding
(Lauder, 1980b; 1983; M u ller & O sse, 1984; Lauder, 1985; C arro ll et al., 2004).
S uction feeding involves a p o w erfu l expansion of th e oral cavity th a t generates a
negative pressure w ith in th e head relative to th e a m b ien t pressure. This pressure
g rad ien t draw s w a ter w ith th e prey in to th e m o u th to fill th e added space and
equalize th e pressure (e.g. A lexander, 1969; Lauder, 1985; S anford & W ain w rig h t,
2002). Buccal expansion can be accom plished by m axillary p ro tru sio n , depression
of th e low er jaw, la te ral ab d u ctio n of th e suspensoria, n eu ro cran ial elevation and
depression of th e hyoid (Alexander, 1969; Lauder, 1983).
Pipefishes an d seahorses (Syngnathidae) are extrem ely fast su ctio n feeders, w ith
prey cap tu re tim es of less th a n 6 ms (C h ap ter 5.1; M u ller & Osse, 1984; B ergert &
W ain w rig h t, 1997; de Lussanet & M u ller, 2007). Explosive cran ial kinem atics,
such as extrem ely rapid accelerations of skeletal elem ents an d h ig h su ctio n
forces, are generated d u ring th e p o w erfu l su ctio n feeding of th ese fishes. A
sy ngnathid feeding strik e is characterized by a very fast depression of th e hyoid
an d an alm ost sim ultaneous n eu ro cran ial elevation. M o u th o p en in g starts only
a fte r th e hyoid has ro ta te d 8o°. B oth lo w er jaw halves an d hyoid bars p u sh th e
suspensoria outw ards w hile depressing, th u s pro d u cin g th e m ain buccal volum e
C h a p t e r 5 - Functional interpr et ati on
123
increase in syngnathids (C hapter 5.1; Roos et al., 2009). Svanbäck et al. (2002)
d em o n strated th a t an increase in velocity of buccal expansion results in an
increasing m agnitude of su ctio n pressure in larg em o u th bass. These large
pressure gradients may resu lt in heavy m echanical loading of th e cran iu m d u ring
su ctio n feeding. T he fast strikes in syngnathids pro b ab ly also produce large and
fast pressure drops. S uction pressure has n o t y et b een m easured in syngnathids,
b u t M u ller an d Osse (1984) d e m o n strated th a t pipefishes like Entelurus aequoreus
(snake pipefish) p e rfo rm su ctio n feeding w ith alm ost p erm an en tly closed
o percular valves (the gili slits are very small), th e re fo re h ig h negative pressure is
b u ilt up inside th e ir buccal cavity. O sse and M u ller (1980) argued th a t th e dom ed
o percular bone, th e lobed gills, th e sm all gili slits and reduced n u m b er of
branchiostegal rays are related to th e h ig h negative pressure experienced d u ring
su ctio n feeding. Because of th is pressure, a great a m o u n t of m echanical force is
th o u g h t to be exerted o n to th e skull bones, especially o n to th o se com prising th e
snout; these are actively ab d u cted an d th u s create th e su ctio n pressure. This force
can be assum ed to resu lt in hig h levels of stress in th e feeding apparatus.
O ne of th e goals of th is study is to d eterm in e w h ere stress accum ulates u n d er th e
assum ed h igh p ressure g en erated d u rin g su ctio n feeding in seahorses and
pipefishes. G iven th e fact th a t cartilaginous tissue has a h ig h flu id c o n te n t, it is
capable of resisting hig h com pressive pressure an d is m ore co m p lian t th a n bone
(e.g. H erring, 1993; L ieberm an et al., 2001; H all, 2005). A pplying force to a n o n ­
fixed elem en t w ill cause it to move in stead of being stressed. T he a rticu latin g
bones in th e skull in syngnathids w ill probably first tra n slate force in to m o tio n
w h e n exposed to su ctio n pressure an d hence w ill experience less m echanical
stress. In addition, presence of a rtic u la r cartilage in jo in ts is believed to p rev en t
load co n ce n tra tio n s (M ori et al., 2010). W e m odelled th e c ran iu m as a single,
im m obile elem en t and assigned it th e m aterial p ro p erties of bone, inclu d in g th e
cartilaginous zones. Precisely because o u r m odels do n o t c o n tain cartilage to
dissipate stress an d do n o t allow m ovem ent of artic u la tin g elem ents, w e p red ict
stress to be co n ce n tra te d a t th o se regions of th e m odels w h ere stress w o u ld be
reduced in th e real skulls (i.e. articu latio n s, sym physes and cartilaginous zones). A
fin ite elem en t analysis (PEA) o n 3D-m odels based o n th e cran iu m of several
sy ngnathid species was perfo rm ed to elucidate this.
C h a p t e r 5 - Functional interpr et ati on
124
W e also investigated a hypothesis based o n Poiseuille’s law, w h ich states th a t th e
v olum etric flow rate of w ater th ro u g h a tu b e is p ro p o rtio n al to th e radius of th e
tu b e to th e fo u rth pow er and to th e pressure difference b etw een th e tw o ends of
th e tu b e, b u t inversely p ro p o rtio n a l to th e le n g th of th e tu b e an d th e dynam ic
fluid viscosity. In o th e r w ords, to o b ta in th e sam e w a ter flow th ro u g h a short,
bro ad tu b e as th ro u g h a long, n arro w one, a larger pressure d ifference is needed
in th e latter. T ranslating th is to sy n g n ath id snouts, w e expect species w ith a long
a n d /o r n arro w sn o u t having to g en erate a larger pressure difference by m eans of
buccal expansion to achieve th e same w a ter flow as sh o rt-sn o u te d species. Roos et
al. (subm itted a) observed th a t th e distance to th e prey w h en th e head reaches its
m axim al ro ta tio n (i.e. th e distance th a t needs to be overcom e by suction) is m uch
longer in a long-snouted species (D. dactyliophorus) com pared to a sh o rt-sn o u ted
one (D. melanopleura). T herefore, long- an d n arro w -sn o u ted species w ill have to
w ith sta n d m ore stress because of th e h ig h er forces and faster accelerations of
cranial elem ents needed to realize th e h ig h su ctio n pressure. W e w an ted to assess
th e im plications of sn o u t elo n g atio n o n th e d istrib u tio n of stress in th e skull by
com paring
th e
stress
p a tte rn s
b etw een
lo n g -sn o u ted
and
sh o rt-sn o u ted
syngnathids. W e hypothesize th a t, w h en exposed to th e same pressure, th e
lo catio n of peak stress w ill be sim ilar am ong species, b u t th a t th e lo n g -sn o u ted
species (for exam ple Doryrhamphus dactyliophorus and Hippocampus reidi) w ill
ex hibit low er levels of stress th a n th e sh o rt-sn o u ted ones (for exam ple D.
melanopleura an d H. zosterae). This is based o n th e assu m p tio n th a t th e skulls of
lon g -sn o u ted species exhibit m orphological a d ap tatio n s th a t enable th e m to
w ith sta n d relatively hig h er su ctio n pressures. So, w h e n th e sam e pressure is
applied to th e c ran iu m of a short- an d a lo n g -sn o u ted species, th e skull of th e
la tte r is expected to be less stressed since its stru c tu re w o u ld be stronger.
By com paring d iffe re n t species, n o t only th e effect of sn o u t le n g th b u t also
differences in sn o u t m orphology have to be ta k e n in to account. Factors such as
variatio n in bone thickness, presence or absence of in d e n ta tio n s or spines,
overlap b etw ee n bony elem ents and p u re b o n e geom etry could in flu en ce th e
p a tte rn s of stress d istrib u tio n . W e trie d to exclude th ese factors by co n stru ctin g
tw o artificial seahorse models. W e started w ith a lo n g -sn o u ted m odel (H. reidi)
an d sh o rte n ed th e sn o u t u n til it h ad th e same le n g th as th e sn o u t in a sh o rt­
sn o u ted seahorse (H. zosterae) in th e first m odel, an d th e sn o u t le n g th of a species
C h a p t e r 5 - Functional interpr et ati on
125
w ith an in term ed iate sn o u t le n g th (H . abdominalis) in th e second. In th is w ay w e
c o n stru cte d th re e m odels (the n a tu ra l H. reidi, th e artificial H. zosterae an d th e
artificial H. abdominalis models) w ith th e same h ead m orphology th a t differ only
in sn o u t length. As a consequence, any m orphological ad ap tatio n s in th e h ead of
th e lo n g-snouted H. reidi m odel fo r w ith sta n d in g h ig h su ctio n pressure are
p re sen t in b o th artificial m odels as well. H ence, w h e n applying th e sam e a m o u n t
of p ressure to th e m odels, w e expect to see id en tical stress p a tte rn s, except for
slightly hig h er co n ce n tra tio n s in th e lo n g er-sn o u ted m odel due to th e larger
surface area u n to w h ich th e pressure is applied.
5.2.2. Brief material & m e t h o d s
A. S p e c im e n s
W e used th re e p ipefish species (genus Doryrhamphus an d Syngnathus) and th re e
seahorse species (genus Hippocampus) in th is study. A d u lt specim ens of D.
dactyliophorus (103.5 m m SL), D. melanopleura (56.5 m m SL), H. abdominalis (222.7
m m SL), H. reidi (119.0 mm, 123.9 m m SL) an d H. zosterae (29.0 m m SL) w ere
o b ta in e d from com m ercial tra d e (Table 2.1, 5.2). T he specim en of S. rostellatus
(110.8 m m SL) w as cau g h t on th e Belgian c o n tin e n ta l shelf (N o rth Sea).
T a b l e 5 .2 - O v e r v i e w o f t h e c o n s t r u c t e d m o d e l s , t h e i r s u r f a c e a r e a a n d s o m e b i o m e t r i c s o f t h e
s p e c i m e n s s c a n n e d . T h e H . r e i d i d e p r e s s e d a n d p r o t r a c t e d m o d e l s a re g e n e r a t e d w i t h C T -d a ta
o f t w o d i f f e r e n t s p e c i m e n s . H L , h e a d le n g t h ; S L , s t a n d a r d le n g t h ; S n L , s n o u t le n g t h .
m odel
D oryrham phus dactyliophorus
D oryrham phus melanopleura
Syngnathus rostellatus
H ippocam pus abdominalis
H ippocam pus reidi d ep ressed
H ippocam pus reidi p r o tra c te d
H ippocam pus zosterae
H ippocam pus abdominalis a rtific ia l1
H ippocam pus zosterae a rtific ia l1
s u rfa c e a re a
(m m 2)
298.3
263.5
224.8
358.2
340.0
384.7
300.0
315.2
276.3
1 m o d e l g e n e ra te d w ith C T -d ata o f H. reidi d ep ressed
SL
(m m )
103.5
56.5
110.8
222.7
119.0
123.9
29.0
-
HL
(m m )
25.4
13.2
SnL
(m m )
16.6
5.8
15-3
30.9
25.2
7-4
11.8
11.0
10.4
1.8
8.8
25-5
5-3
-
5-9
H L /S n L
1-5
2.3
2.1
2.6
2.3
2-5
2.9
-
126
C h a p t e r 5 - Functional interpr et ati on
B. C o n s tr u c tin g m o d e ls
In to ta l, n in e m odels w ere m ade fo r th is analysis: a m odel of each of th e six
species, tw o artificial seahorse m odels and a H. reidi m odel w ith th e hyoid
p ro tra c te d (Table 5.2). For th e H. reidi m odels, C T -data of tw o H. reidi specim ens
w ere used: one w ith th e hyoid p ro tra c te d (used fo r th e H. reidi p ro tra c te d m odel
an d th e tw o artificial models) an d one w ith a depressed hyoid (for th e H. reidi
depressed model). C T-scans an d m odels w ere m ade as explained in ch ap ter 2.2.7
an d 2.2.10.
D u m o n t et al. (2009) d e m o n strated th a t if th e goal of a com parative FEA study is
to id en tify differences in stress m agnitudes th a t stem solely fro m differences in
m odel shape, th e n th e ratios of applied force to m odel surface area m u st be held
constant. In th is analysis w e chose n o t to com pletely elim in ate th e effect of size,
since one of th e aim s is to determ in e th e relatio n sh ip b etw een relative sn o u t
le n g th and stress. In stead w e scaled th e m odels to equal braincase le n g th (see
ch ap te r 2.2.10).
In a d d itio n to m odelling d iffe re n t species, w e created tw o artificial m odels w ith
d ifferen t levels of sn o u t elo ngatio n in R hinoceros 3.0 M cN eel E urope (Barcelona,
Spain). S tartin g w ith th e original H. reidi m odel, th e p a rt of th e sn o u t b etw een
th e ectop terygoid bone an d th e lateral eth m o id b o n e w as com pressed u n til it had
th e same le n g th as th e sh o rt sn o u t in H. zosterae (Fig. 5.8). This m odel w ill be
called th e ‘H. zosterae artificial m odel’. N ext, w e c o n stru cte d a n o th e r m odel w ith
th e same h in d p a rt of th e skull an d th e same jaws as th e original H. reidi m odel,
b u t th is tim e w e decreased sn o u t le n g th to approxim ate th e in term ed iate le n g th
of th e sn o u t in H. abdominalis. This m odel w ill be referred to as th e ‘H.
abdominalis artificial m odel’.
W e also reasoned th a t th e outcom e of th e FEA m ig h t be in flu en ced by th e
o rie n ta tio n of th e hyoid since n o t all m odels h ad th e hyoid in th e same position.
H ence, scans of a second H. reidi specim en, w ith its hyoid p ro tracted , w ere made
an d a m odel w as co n stru cte d w h ich w e could th e n com pare w ith th e original H.
reidi m odel w ith its hyoid depressed. To d istin g u ish b etw een b o th H. reidi m odels
one is called th e ‘H. reidi p ro tra c te d m odel’, th e o th e r th e ‘H. reidi depressed
m odel’.
C h a p t e r 5 - Functional interpr et ati on
C.
127
F in ite e le m e n t a n a ly sis
T he fin ite elem ent analysis was carried o u t as described in ch ap te r 2.2.10 and
figure 2.4. Since th e stress m agn itu d e differs vastly b etw een th e various m odels,
d ifferen t colour scales are used in figures 5.9 an d 5.10 in ord er to avoid a lim ited
range of colouring in th e m odels w ith h ig h est an d lo w est stress levels. This way,
em phasis is laid on th e (dis-)sim ilarity in th e stress d istrib u tio n and com parison
b etw een th e m odels is facilitated.
5.2.3. Results
A. S tre ss d i s t r ib u tio n in p ip e fis h e s a n d s e a h o rs e s
Figure 5.9 illu strates th e von M ises stress d u rin g sim u lated su ctio n feeding. Peak
stress values occur a t th e occipital jo in t w h ere b o u n d ary co n strain ts (b o th against
displacem ent an d ro tatio n ) w ere applied in ord er to an ch o r th e m odel, th is
a rte fa ct can be ignored. In each of th e seahorse m odels stress is c o n ce n tra te d at
th e a rtic u la tio n betw een th e au to p alatin e bone an d th e n eu ro cran iu m , in th e
hyoid an d in th e lateral w all of th e snout. Some elevated stress is also fo u n d at
th e m esethm oid-parietal border. Low est stress is p re se n t in th e braincase, a t th e
su praoccipital bone an d at th e spines o n th e lateral ethm oid, p reo p ercu lar,
parietal, sp h en o tic an d p o sttem p o ral bones. In th e H. reidi m odel h ig h levels of
stress are also p re sen t a t th e h y o m an d ib u lo -sp h en o tic jo in t, w hereas th e
m axillary bones show very little stress (Fig. 5.9A). Stress is c o n ce n tra te d in th e H.
abdominalis m odel a t th e v en tral b o rd er of th e q u ad rate bone (w here th e re is an
in d en tatio n ) an d a t th e tra n s itio n of th e very slender vom eral b o n e to th e
m esethm oid bone (Fig. 5.9B). T here is also stress in th e lateral p a rt of th e sn o u t
an d a t th e tw o suspensorio-neu ro cran ial a rticu latio n s (i.e. th e hyom andibulosp h en o tic and autopalatine-vom eral joint). H ow ever, in th e H. zosterae m odel,
little stress accum ulates at th e level of th e h y o m an d ib u lo -sp h en o tic a rticu latio n ,
especially com pared to th e a m o u n t of stress in th e lateral w all of th e sn o u t (Fig.
5.9C). O n th e o th e r hand, th e co n tac t zone b etw een th e hyoid an d th e
Suspensorium does experience a lo t of stress. W ith in th e seahorses, th e H.
abdominalis m odel, w ith an in term ed iate sn o u t len g th , exhibits h ig h est levels of
stress, even up to five tim es th e stress m agnitude in th e o th e r tw o species. Stress
128
C h a p t e r 5 - Functional interpr et ati on
values are low er in th e m odel of th e lo n g -sn o u ted H. reidi and lo w est in th e sh o rt­
sn o u ted H. zosterae m odel (m ark th e d ifferen t colour scales in figure 5.9).
T he pip efish m odels show m ore diverse p a tte rn s in stress values. The p a tte rn of
stress d istrib u tio n in th e S. rostellatus an d D. melanopleura m odels is q u ite sim ilar
b u t differs from th e situ a tio n in th e D. dactyliophorus model. In th e fo rm er tw o,
hig h est stress c o n cen tratio n s are fo u n d a t th e low er jaw sym physis and ro stral to
th e nostrils, b u t also a t th e caudal su sp en so rio -n eu ro cran ial a rtic u la tio n an d at
th e v en tral b o rd er of th e o rb it (Fig. 5-9E,F). The dorsal an d lateral surface of th e
braincase show s th e low est stress levels. The m odel of S. rostellatus also exhibit
h igh stress a t th e a tta c h m e n t of th e hyoid to th e suspensoria, and alm ost n o stress
in th e hyoid sym physis or in th e p arasp h en o id bone. T he stress in th e D.
melanopleura m odel is c o n ce n tra te d a t th e
hyoid sym physis an d a t th e
autopalatine-vom eral articu latio n . C onversely, in th e D. dactyliophorus m odel, th e
stress accum ulates in th e roof, floor an d lateral w alls of th e braincase, th e
h y o m and ibulo-sphenotic a rtic u la tio n and aro u n d th e o rb it (Fig. 5.9D). T he hyoid
an d th e ro stral p a rt of th e sn o u t w ith th e jaws ex h ib it very low stress. C om paring
th e th re e p ipefish species, stress levels are low est in th e S. rostellatus m odel w h ich
has an in term ed iate sn o u t length. Stress m agnitudes are o n average alm ost double
in th e m odel of sh o rt-sn o u te d D. melanopleura an d by far th e h ig h est in th e
extrem ely long-snouted D. dactyliophorus m odel (m ark th e d ifferen t colour scales
in figure 5.9).
B.
S tre s s d i s t r ib u ti o n in th e a r tif i c i a l m o d e ls
T he stress d istrib u tio n p a tte rn s of th e H. reidi depressed an d th e artificial m odels
are very sim ilar, w ith peak stress co n cen tratio n s at th e hyoid sym physis, th e
h y o m and ibulo-sphenotic
jo in t
an d
th e
ro stra l
a rtic u la tio n
b etw een
th e
Suspensorium an d n e u ro c ran iu m (Fig. 5.10). Also th e low er jaw and o th e r p arts of
th e hyoid besides th e sym physis are p ro n e to som e stress, y et th e m odels exhibit
alm ost no stress a t th e m axillary bones an d th e braincase. In addition, th e stress
in th e sn o u t is tra n sm itte d in an alm ost identical m anner; all th re e m odels show
h igh am o u n ts of stress laterocaudally in th e sn o u t and at th e d o rso ro stral edge at
th e autopalatine-vom eral articu latio n . T he la te ro ro stra l surface of th e sn o u t and
v e n tro ro stra l edge of th e o rb it experience low est stress. A m in o r difference
b etw een th e m odels involves th e stress c o n ce n tra tio n a t th e m esethm oid-parietal
C h a p t e r 5 - Functional interpr et ati on
129
b o rd er in th e H. reidi and H. zosterae artificial m odels th a t seems to be m issing in
th e H. abdominalis artificial model. This h ig h sim ilarity w as expected since th e
th re e m odels all share th e sam e m orphology.
A gain according to o u r expectations, th e overall stress level in th e m odels
increases w ith increasing sn o u t length. The artificial m odel of H. zosterae
experiences very sm all am o u n ts of stress, w hereas th e H. reidi m odel is subject to
m uch hig h er levels an d th e H. abdominalis m odel show s in term ed iate stress values
(m ark th e d ifferen t colour scales in figure 5.10).
5.2.4. D is cu ssi on
A. S tre ss d i s t r ib u tio n in p ip e fis h e s a n d s e a h o rs e s
W e w a n te d to analyze w here in th e skull th e stress w o u ld be co n cen trated , w h en
loaded as expected d u ring su ctio n feeding. O u r p re d ic tio n w as th a t th e largest
stress c o n cen tratio n s w o uld occur a t th e level of th e articu latio n s, sym physes and
cartilaginous zones in th e cranium . This hypothesis is based o n th e assu m p tio n
th a t these areas disperse or absorb th e stress in th e actu al cranium . Since th e
skull w as m odelled as a single, im m ovable elem en t w ith m aterial p ro p erties of
bone, th e assum ed stress ab so rp tio n by th e a rticu latio n s and cartilaginous regions
as such c an n o t be d etected in o u r FEA results. H ence, th e stress peaks th a t m ight
be fo u n d a t th e articu latio n s, sym physes an d cartilaginous regions of th e m odel
w ill probably be m uch low er in reality. This hypothesis is m ore or less su p p o rted
by th e ob tain ed results: ap art from th e m odels of D. dactyliophorus an d H. zosterae,
all m odels show elevated levels of stress a t b o th su sp en so rio -n eu ro cran ial
articulations. A gain w ith th e
ex ception of D. dactyliophorus, peak stress
c o n ce n tra tio n s w ere p re sen t a t th e low er jaw sym physis and hyoid sym physis, as
expected.
H ow ever,
in
H. zosterae
an d
S. rostellatus n o
elevated
stress
c o n ce n tra tio n w as fo u n d in th e hyoid sym physis, w h ich m ig h t be explained by
th e p ro tra c tio n of th e hyoid, as w e w ill explain fu rth er. M o st m odels also have a
stress accum ulation at th e m esethm oid-parietal tra n s itio n ju st b efo re th e eye,
w h ich is a region com prising a lo t of cartilage.
T he cran iu m of D. dactyliophorus experiences m u ch h ig h er stress levels th a n th e
o th e r skulls an d also show s an in v erted stress p a tte rn w ith peak co n cen tratio n s
130
C h a p t e r 5 - Functional interpr et ati on
in th e braincase and low est stress at th e ro stral p a rt of th e snout. W e could n o t
adequately explain these findings.
W e can also com pare th e differences in stress d istrib u tio n b etw een a typical
p ipefish c ran iu m and a typical seahorse. The m ost p ro m in e n t difference b etw een
th e m is th e deeper braincase of seahorses, b u t as w e only applied loads to th e
Suspensorium , n o t m uch stress was exerted o n th e braincase itself. H ow ever th e re
are o th e r differences betw een th e tw o groups. In pipefishes th e braincase is in
line w ith th e long axis of th e sn o u t (and of th e e n tire body) w hereas seahorses
have a tilte d braincase w ith respect to th e sn o u t (C h ap ter 3.1). This tilt can be
q u ite large. A n angle of aro u n d 40° b e tw ee n th e long axis of th e sn o u t and th e
dorsal profile of th e braincase is p re sen t in th e seahorse species studied, w hereas
in these pipefishes th is angle ranges b etw een 150 an d 250 (see also Fig. 5.9). Stress
ten d s to increase at a b ru p t geom etric tra n sitio n s in th e m odel, like a t th e level of
th e acute angle betw een th e m esethm oid and p arie tal bones. W e expected th is
effect to be m ore p ro n o u n c e d in seahorses th a n in pipefishes. Yet w e did n o t fin d
su b stan tially hig h er stress levels in th e m eseth m o id -p arietal tra n sitio n of
seahorses. Perhaps th is can be acco u n ted fo r by specific featu res of th e geom etry
of th e skull. W e fo u n d th a t th e m esethm oid b o n e bears a dorsom edial ridge in
seahorses, th u s allow ing a b e tte r tran sm issio n of th e stress w ith in th e bone. This
could explain th e low er th a n expected stress levels a t th e m esethm oid-parietal
tra n sitio n . A n o th er difference betw een pipefishes and seahorses is th e presence
of spines an d a hig h er su praoccipital b o n e in th e latter. It has b een suggested th a t
th e spines m ake seahorses less attractiv e as prey (Lourie et al., 1999b). W e n o te d
th a t th e spines and distal p a rt of th e su p rao ccip ital b o n e experience alm ost no
stress in any of th e seahorses; hence, stress re d u c tio n m ig h t be an ad d itio n al
advantage of th e spines.
B. E f fe c t o f s n o u t le n g th o n s tre s s d is tr i b u ti o n
W e w ere also in te re sted in th e effect of sn o u t elo n g atio n o n th e cranial stress
d istrib u tio n in seahorses an d pipefishes. W e expected th a t fo r a given m echanical
loading, th e longer th e snout, th e low er th e stress w o u ld be, as th e skull of lo n g ­
sn o u ted species is th o u g h t to be stru c tu ra lly b e tte r ad ap ted to cope w ith hig h er
m echanical loads. H ow ever, th e m odel w ith th e lo n g est sn o u t w as n o t th e least
stressed, n e ith e r am ong th e seahorses n o r am ong th e pipefishes. M oreover, we
C h a p t e r 5 - Functional interpr et ati on
131
did n o t fin d a tre n d in stress c o n ce n tra tio n (increasing or decreasing) fro m lo n g ­
sn o u ted species th ro u g h in term ed iate form s to sh o rt-sn o u te d species. This
suggests th a t th e re is no sim ple re la tio n b etw een sn o u t le n g th an d a m o u n t of
stress expected to be caused by suction. A pro b ab le ex p lan atio n is th a t th e
observed stress level is th e resu lt of th e c o m b in atio n of sn o u t le n g th w ith th e
geom etry of th e bones form ing th e snout. T he previously m en tio n e d dorsom edial
crest on th e m esethm oid bone in seahorses could be an example. In th e seahorses
w e also fo u n d th a t th e lateral bones of th e sn o u t of H. abdominalis are m uch
th in n e r on a cross section relative to th e sn o u t diam eter, w h en com pared to H.
reidi and H. zosterae. This m igh t acco u n t fo r th e accu m u latio n of stress in th e
lateral w all of th e sn o u t in th e fo rm er and could explain w hy th e level of overall
stress exceeds th a t of th e lo n g-sn o u ted H. reidi.
In order to separate th e effect of sn o u t elo n g atio n fro m th e in fluence of sn o u t
m orphology (i.e. bone geom etry), w e com pared th e results of th e tw o artificial
m odels w ith th o se of th e n a tu ra l m odel of H. reidi. T he in itia l hypothesis w as th a t
these th re e m odels w o uld be sim ilar in stress p a tte rn since th e y vary only in
sn o u t length; any m orphological a d ap ta tio n th a t enables th e m to w ith sta n d high
pressure is p re sen t in all th re e of them . H ow ever, w e expected to see a gradual
increase in stress levels w ith increasing sn o u t len g th , because th e surface area to
w h ich th e load is applied increases as well. All m odels seem to tra n sm it th e loads
in a sim ilar way: th e stress d istrib u tio n p a tte rn is alm ost id en tical an d w h en
sh o rte n in g th e sn o u t of H. reidi to th e sn o u t le n g th of H. abdominalis
(interm ediate sn o u t length) th e stress levels w ere reduced and even m ore so fo r a
sn o u t le n g th of H. zosterae (short sn o u t length) (Fig. 5.10). So, in th e absence of
m orphological differences, w e fo u n d changes in stress levels related to relative
sn o u t length.
Based on th e first analysis, w e could n o t co n firm o u r hypothesis th a t a lo n g ­
sn o u ted species en co u n ters less stress because it has a stro n g er stru ctu re. Three
reasons com e to m ind w hy th is m ig h t be. First, lo n g -sn o u ted species m ig h t n o t
have to gen erate hig h er su ctio n pressure d u rin g feeding th a n sh o rt-sn o u ted
species and h ence no stru c tu ra l rein fo rcem en t is needed. This can be rejected
w ith fair certain ty based on Poiseuille’s law and M u ller an d O sse’s (1984) finding
th a t pressure in th e o percular cavity increases w ith increasing sn o u t length. A
second possibility is th a t lo n g -sn o u ted species do experience h ig h levels of
132
C h a p t e r 5 - Functional interpr et ati on
pressure b u t th ey do n o t ex hibit m orphological a d ap tatio n s to w ith sta n d th e
increased stress. If th is w ere tru e , th e lo n g -sn o u ted syngnathids in o u r analysis
should be m ore stressed th a n th e sh o rt-sn o u ted ones. W e did n o t observe this
p a tte rn in com parisons am ong species, b u t it w as p re sen t in th e com parison of
th e artificial models. The fact th a t th e artificial m odels varied only in sn o u t
le n g th clearly d em onstrates th a t th e m orphology of th e cran iu m does have an
in flu en ce on h ow stress is tran sm itted . T hird, lo n g -sn o u ted species m ig h t have a
c ran iu m th a t is ad ap ted to dim in ish stress d u ring su ctio n feeding in a w ay th a t
w e could n o t d e tec t w ith o u r m ethodology (e.g. p erh ap s th e re is m ore cartilage in
highly stressed regions). A lth o u g h a recen t m orphological stu d y fo u n d no
p ro m in e n t features th a t could be related to v ariatio n in relative sn o u t le n g th
(C h ap ter 3.2), it m ight be an in te re stin g avenue to p u rsu e th is m ore th o ro u g h ly
on a histological level in th e future.
C. B io lo g ic a l re le v a n c e o f th e r e s u lts
V alidation of th e results of a fin ite elem en t analysis requires a com parison
b etw een th e results of th e FEA an d in vivo or in vitro data o n b o n e strain in th e
m odelled stru c tu re s (Ross, 2005). U n fo rtu n ately , th e re are n o ex p erim en tal data
on bone stra in in th e cran iu m of Syngnathidae, n o r is it c u rre n tly feasible given
th e size of these skulls. In absence of validation, w e consider o u r com parison of
stress p a tte rn s betw een th e d ifferen t m odels to be reliable, b u t absolute stress
values are not. Still, w e id en tified a few possible m ethodological lim itatio n s of
o u r setup. M o st notably, th e p o sitio n of th e hyoid an d sim p lificatio n of th e
m odels an d th e forces acting u p o n th e m du rin g su ctio n m ig h t in fluence th e FE
outcom e.
A possible source of inaccuracy could be th e p o sitio n of th e hyoid in th e
c o n stru cte d models: in th e H. zosterae, S. rostellatus an d D. melanopleura m odels
th e hyoid is p ro tra c te d w hereas th e o th e r m odels show a depressed hyoid. In
order to te st w h e th e r th e p o sitio n of th e hyoid m ig h t in flu en ce th e stress
d istrib u tio n p a tte rn , w e com pared th e H. reidi depressed m odel w ith th e H. reidi
p ro tra c te d one (Fig. 5.11). T he p a tte rn of stress d istrib u tio n is q u ite sim ilar in
b o th m odels, only th e m esethm o id bone, th e lateral surface of th e sn o u t and
hyoid-suspensorium a rtic u la tio n are subject to m ore stress in th e m odel w ith th e
hyoid p ro tracted . This finding falls w ith in o u r ex p ectatio n since th e p ro trac te d
C h a p t e r 5 - Functional interpr et ati on
133
hyoid does n o t have only one co n tac t p o in t at th e a rtic u la tio n w ith th e
Suspensorium , b u t is m odelled co n tin u o u sly w ith th e re st of th e skull along
alm ost th e e n tire hyoid bars. Thus, a m ore firm co n n ec tio n b etw een le ft and rig h t
Suspensorium is form ed in th e m odel w ith its hyoid p ro trac te d , w h ich reduces
th e flexibility of th e suspensoria and could explain th e observed increase of stress
in th e sn o u t an d m esethm oid b o n e w h e n applying an in w ard pressure o n th e
suspensoria in th e model. O n th e o th e r hand, th e h ig h er levels of stress at th e
hyoid sym physis observed in th e m odel w ith th e depressed hyoid m ake sense,
since at th is p o in t th e pressure of le ft an d rig h t side converges. The outcom e of
th e hyoid te st corresponds w ith o u r previous observations th a t H. zosterae (hyoid
p ro tracted ) experiences m ore stress in th e sn o u t w hereas H. reidi and H.
abdominalis (hyoid depressed) have h ig h stress accu m u latio n s at th e hyoid
sym physis (Fig. 5.9A-C). H ence, th e differences b etw een th e th re e seahorse
m odels m ight be p artially related to changes in th e p o sitio n of th e ir hyoid.
By c o n stru ctin g th e skull as a single u n it, w e ignored th e presence of su tu res and
articulations. This is especially im p o rta n t since su tu res are k n o w n to be less stiff
an d experience hig h er su tu ra l strains th a n b o n e (H errin g & Teng, 2000; Rayfield,
2005; M cH e n ry et al., 2006). In th is context, previous research has sh o w n th a t th e
m orphology of su tu res is related to th e type of d efo rm atio n to w h ich th e y are
m ost resistant. As such, in te rd ig itate d su tu res appear to be co rrelated w ith
com pression w hereas a b u ttin g su tu res usually are subject to ten sile stress
(H erring & Teng, 2000; M ark ey et al., 2006; M ark ey & M arshall, 2007).
A rticu latin g elem ents tra n sm it m echanical loading d ifferen tly th a n
fused
elem ents, since cartilaginous a rticu latio n s can fu n c tio n as stress absorbers.
U n fo rtu n ately , analyzing th e cran iu m as a solid u n it in stead of as an assem bly of
individual p a rts is necessary to reduce th e com plexity of co n stru ctin g th e m odel
an d to b rin g th e processing of th e analysis to a m anageable level. By neglecting
th e effects of m aterial an isotrop y an d h etero g en eity , m uscle force, n eu ro cran ial
elevation an d su tu re m orphology, w e assum e th a t th ese factors affect all th e
species of o u r study in th e same w ay (M cH enry et al., 2006). As te ch n iq u es of
m odelling m uscle forces, bony a rticu latio n s and kinem atics are being developed
for fin ite elem ent analyses, w e look fo rw ard to adding th o se tech n iq u es to th e
analyses described here. In th e p re sen t study, overall h ead shape and relative size
134
C h a p t e r 5 - Functional interpr et ati on
are th e only su b stan tial differences am ong th e m odels an d th e re fo re explain all
th e differences w e observe am ong them .
135
ß
General discussion
A th o ro u g h in sig h t in to an anim al’s m orphology is vital fo r u n d erstan d in g
com plex
processes
locom otion,
associated
resp iratio n ,
w ith
feeding,
etc.)
its
developm ent,
an d
ev o lu tio n ary
perfo rm an ce
history.
In
(e.g.
th is
d issertation, a th o ro u g h m orphological d escrip tio n of th e m usculoskeletal head
in Syngnathidae is provided first. Secondly, th e shape v ariatio n in th e syngnathid
head an d th e fu n c tio n a l im plications of th e feeding ap p aratu s are studied. In th is
chapter, these results are discussed in an o n to g en etic, fu n c tio n a l (w ith respect to
feeding) an d evolutionary context. The role of th e com plex p a re n ta l care, th e
m odifications of th e feeding ap p aratu s and th e ev o lu tio n ary p a tte rn s tow ards
an d w ith in th e fam ily are dealt w ith.
6 .1 .
O n t o g e n e t ic m o r p h o l o g y
Syngnathids are k n o w n for th e ir extrem e p a re n tal care. The eggs, dep o sited by
th e fem ale, are a ttach ed to th e m ale’s body w h ere th e y are fertilized (e.g. H erald,
3.959; V in cen t et al., 1995; D rozdov et al., 1997; P ham T chi M i et al., 1998;
K ornienko, 2001; W ilso n et al., 2001; V in cen t & Giles, 2003; F o ster & V incent,
136
C h a p t e r 6 - G e n e ra l d is c u ssio n
2004; Lourie et al., 2004; Ripley & Foran, 2009). In th e in cu b atin g area of th e male,
th e eggs can be loosely atta ch e d to th e v en tral surface of th e body (e.g. Entelurus,
Nerophis), separated from each o th e r by a spongy mass (e.g. Solegnathus,
Doryrhamphus), covered by a m em brane e ith e r in co m b in atio n w ith p ro tectiv e
plates or n o t (e.g. Syngnathus, Microphis) or enclosed by a sealed b ro o d p o u ch (e.g.
Hippocampus) (Fig. 1.5; H erald, 1959; K ornienko, 2001; W ilso n et al., 2001;
C arcu p in o et al., 2002; K varnem o & Sim m ons, 2004). In th e seahorse’s pouch,
sm all cups are form ed on th e in cu b atin g area in w h ich th e eggs, and la te r th e
em bryos, are enveloped by ep ith elial tissue (Foster & V incent, 2004). The m ale
provides th e em bryos w ith all p o st-fe rtiliza tio n p a re n ta l care u n til th e yolk is
com pletely absorbed. W ith increasing ev o lu tio n ary com plexity of b ro o d area, th e
a m o u n t of p a re n tal in v estm en t in term s of n u tritio n and oxygenation is th o u g h t
to increase (C arcupino et al., 2002). The b ro o d p o u ch of seahorses can be th o u g h t
of as a pseudo-placenta for th e em bryos, having an osm oregulatory, aerative and
n u tritiv e fu n c tio n (K ornienko, 2001; F o ster & V in cen t, 2004; S tö ltin g & W ilson,
2007). Also in Syngnathus pipefishes (e.g. S. fuscus, S. floridae, S. scovelli (E verm ann
& K endall) an d S. abaster Risso) p a re n ta l n u trie n t provision, oxygenation and
osm otic p ro te c tio n has been d e m o n strated (D rozdov et al., 1997; K ornienko, 2001;
C arcu p in o et al., 2002; Ripley & Foran, 2009). The supply of oxygen and
endogenous energy enables a delay of em ergence fro m th e brooding stru ctu re.
From th e m o m en t of p a rtu ritio n (i.e. release fro m th e bro o d in g area), th e
sy ngnathid young are com pletely in d ep en d e n t and receive n o assistance or
p ro te c tio n from th e ir p aren ts (V incent et al., 1995; K ornienko, 2001; F o ster &
V incent, 2004; V an W assenbergh et al., 2009). The size and relative degree of
developm ent of n e w b o rn syngnathids m ight be related to th e degree of
d iffe re n tia tio n of th e in c u b atio n area ex h ib ited by th e male. Species provided
w ith a pouch-like s tru c tu re (e.g. Syngnathus acus, S. typhle, S. acusimilis G ü n th e r
an d S. abaster) appear to have larger young of a m ore advanced developm ental
stage com pared to species th a t sim ply a tta c h th e em bryos to th e ir body surface
(Nerophis lumbriciformes (Jenyns), N.
ophidion
L. an d Entelurus aequoreus)
(M o n teiro et al., 2003; Silva et al., 2006). The release of w ell developed juveniles
w as also seen in o th e r fish w ith u n u su al p a re n tal care. For exam ple, in th e w h ite
barbel (Galeichthys feliceps), an ariid m o u th-brooder, em bryos are v e n tilated and
n o u rish ed in th e a d u lt buccal cavity. H ence, release fro m th e m o u th is postp o n ed ,
C h a p t e r 6 - G e n e ra l d is c u ssio n
137
resu ltin g in a to ta l in c u b atio n p erio d of alm ost double th e p re-h atch in g perio d
(Tilney & H ech t, 1993). A sim ilar strateg y is used in m ost h ap lo ch ro m in e cichlids
(G oodw in et al., 1998; Salzburger et al., 2005).
O ne of th e consequences of th e long an d in ten se p a re n tal care and associated
release of offspring a t advanced developm ental stages fro m th e in c u b atin g area, is
th e great sim ilarity of th e head m orphology of new ly released young to th a t of
adults (e.g. A zzarello, 1990; D hanya et al., 2005; C hoo & Liew, 2006; Silva et al.,
2006). The osteology of early juveniles w as elab o rated o n in ch ap ter 3.1. F rom th is
study, it appears th a t in th e new b o rn s of S. rostellatus, H. capensis and H. reidi
alm ost all bones are form ed, alth o u g h only as very th in layers. The bones th a t
w ere n o t yet com pletely form ed are th e an to rb ita l, tw o or th re e in frao rb ital
bones (for discussion on circu m o rb ital b o n e hom ology, see ch ap ter 2.3) an d some
o percular bones in S. rostellatus. In H. reidi, th e a fo rem en tio n ed bones only sta rt
to develop a fte r 29 days (Reyserhove, 2008). This corresponds to th e observed late
o n set of developm ent of th e circu m o rb ital bones in o th e r syngnathids (Kadam,
1961), w h ic h is also th e case in o th e r teleosts such as som e Perciform es (Sparus
aurata L.; F au stin o & Pow er, 2001) an d som e Siluriform es (Clarias gariepinus
(Burchell), Ancistrus cf. triradiatus E igenm ann an d Corydoras aeneus (Gili);
A driaens & V erraes, 1998; G eerinckx et al., 2007; H u y se n tru y t et al., 2011). The
absence of th e inter-, sub- an d su p rap reo p ercu lar bones in S. rostellatus can be
related to th e fact th a t th e stu d ied specim en h ad n o t y et reached th e free-living
stage. H ow ever, in H. reidi, th e in te ro p e rc u lar bone, as w ell as th e vom eral and
re tro a rtic u la r bones, are also n o t fully developed u n til 25 days post-release
(Reyserhove, 2008). In H. kuda som e m in o r developm ental changes w ere observed
a fte r being expelled from th e p o u ch as w ell (Choo & Liew, 2006). O n ly a fte r fo u r
days did th e cheek spines sta rt to appear an d co ro n ary spines w ere n o t
com pletely developed u n til th e n in th day p o st-p a rtu ritio n . A n o th e r im p o rta n t
difference b etw een early juvenile an d a d u lt m orphology co n cern s th e single
cartilaginous h y om andibulo-neu ro cran ial a rtic u la tio n in juvenile syngnathids,
w hereas in adults th e hyom andib u lar bone bears a double a rticu lar facet w ith th e
n e u ro c ran iu m (C hapter 3.1).
O verall, how ever, these osteological d ifferences b etw een early juveniles and
adults are small. V an W assenb erg h et al. (2009) did show th a t n o t only th e
138
C h a p t e r 6 - G e n e ra l d is c u ssio n
skeletal elem ents b u t also all th e m uscles, ligam ents an d ten d o n s involved in
pivot feeding are p re sen t in n ew b o rn H. reidi juveniles, w h ich seem ed to be tru e
for H. capensis as w ell (Thielem ans, 2007). P erhaps even m ore im p o rtan tly , w e saw
th a t th e elo ngation of th e sn o u t occurs early in developm ent of S. rostellatus, H.
capensis an d H. reidi (C hapter 3.1), w h ich is also tru e fo r several o th e r syngnathids
(e.g. H. antiquorum (Ryder, 1881), S. peckianus (M cM urrich, 1883), S. fuscus
(K indred, 1921), Hippocampus (species n o t stated, Kadam, 1958), Nerophis (species
n o t stated, Kadam, 1961), S. scovelli a n d # , zosterae (Azzarello, 1990)). Also som e of
th e specialized m orphological featu res in th e feeding ap p aratu s th a t m ig h t be
considered adap tatio n s to th e rem arkable p iv o t feeding in syngnathids are
already p re se n t in juveniles. A n exam ple can be fo u n d in th e cartilaginous
interhyal, w h ich is very sim ilar in fo rm to th e b o n y elem en t in adults. B oth have a
double a rtic u la tio n head for th e ceratohyal bone and a ro u n d ed surface fo r th e
a rtic u la tio n w ith th e p reo p ercu lar bone.
In som e fishes, such as eels (e.g. Tesch, 2003; van G in n ek en & M aes, 2005) and
flatfishes (e.g. G effen et al., 2007), th e free-living larva is shaped u nlike th e ad u lt
an d undergoes a p ro fo u n d m etam orphosis to produce th e juvenile. This
su b sta n tia l change in m orphology is o fte n associated w ith a tra n s itio n from
fresh w ater to m arine e n v iro n m en t or w ith th e settlem e n t of p la n k to n ic larvae
w h e n th e y develop in to b e n th ic juveniles in m arine fishes (Rose & Reiss, 1993). In
m any cases m etam orphosis is accom panied by developm ental alteratio n s in
feeding behaviour an d h a b ita t (H elfm an et al., 2009).
Syngnathids do experience a dietary shift. Prey size increases ontogenetically,
likely since larger prey yield an advantage in term s of energy supply and
subsequently grow th. This o n to g en etic sh ift in d iet is re la ted to a developm ental
increase in m o u th gape, as is th e case in o th e r fish (e.g. W e rn e r & G illiam , 1984;
C ook, 1996; H u n t van H erbing, 1996a). In ad d itio n , a t least som e syngnathids also
undergo a tra n s itio n from a pelagic to a b e n th ic e n v iro n m e n t (M o n teiro et al.,
2003; F oster & V incent, 2004). D espite th ese behavioural changes d u ring
developm ent, no real m etam orphosis takes place. O n ly an increase in g ro w th rate
a t th e tim e of th ese shifts was observed in H. kuda (C hoo & Liew, 2006). D hanya
et al. (2005) also fo u n d th a t n ew b o rn Syngnathoides biaculeatus only differ from
adults in relative sn o u t le n g th and colour. C hoo & Liew (2006) p e rfo rm ed several
body m easurem ents on H. kuda juveniles and n o ticed th a t h ead len g th , sn o u t
C h a p t e r 6 - G e n e ra l d is c u ssio n
le n g th
and
sn o u t
h e ig h t
139
(w hich
th e y
call
sn o u t
depth)
increase
over
developm ental tim e. H ow ever, w h e n com paring th e increase in h ead le n g th w ith
th e one in sn o u t length, th e increase in sn o u t le n g th is relatively larger com pared
to overall head grow th. A sim ilar com parison, b u t th is tim e b etw een head le n g th
an d sn o u t height, show ed th a t sn o u t h eig h t increases m ore slow ly th a n head
length. In o th e r w ords, new ly b o rn H. kuda young have a sh o rte r an d relatively
less h igh sn o u t th a n older juveniles. A stu d y o n H. reidi yielded th e sam e results,
juveniles w ith sh o rt and b ro ad sn o u ts develop in to long an d n arro w sn o u ted
adults (Roos et al., 2010). T he geom etric m o rp h o m etric analysis of ch ap ter 4
revealed th a t th e onto g en y of th e same species also involves allom etric changes of
th e sn o u t h e ig h t an d head h eig h t w ith respect to sn o u t length, as w ell as changes
in th e o rie n ta tio n of th e p reo p ercu lar bone an d low ering of th e sn o u t tip. All
these results show th a t syngnath id p o st-p a rtu ritio n developm ent is characterized
by gradual shape changes, and no m ajor tra n sfo rm a tio n s occur.
As stated, new ly released syngnathids closely resem ble m in iatu re adults. The
m edian fin fold, w h ich is a larval ch aracteristic according to B alon (1975), has
d ifferen tia te d in to th e dorsal, anal an d caudal fins long b efo re p a rtu ritio n and
th u s before th e tra n s itio n to exogenous feeding (chapter 2.3; A zzarello, 1990).
H ence before being expelled fro m th e pouch, p ip efish an d seahorse young have
already tran sfo rm ed in to juveniles (w h eth er or n o t by m eans of a m etam orphosis)
an d m ost m orphological d iffe re n tia tio n takes place inside th e p o u ch (C hoo &
Liew, 2006). As in o th e r fish, th e absence of a tru e larval stage can be related to
th e extrem e reproductive strategy in syngnathids (Balon, 1979; 1984; K olm &
A hnesjö; 2005). E lim in atio n of th e larval stage is generally accepted as an
im p o rta n t ecological an d evolu tio n ary p h e n o m en o n (Balon, 1984) an d may
th e re fo re have played a crucial role in ev o lu tio n to w ard s a specialized feeding
ap p aratu s (H uysentruyt, 2008).
Prolonging th e em bryonic perio d in syngnathids increases th e available tim e for
anatom ical stru c tu re s in th e tro p h ic ap p aratu s to develop, as active feeding
becom es necessary only later in ontogeny. Pipefishes an d seahorses give b irth to
young w ith a feeding system a p p ro p riately con fig u red fo r self-sufficiency (e.g.
L in to n & Soloff, 1964; A zzarello, 1990; P ham T chi M i et al., 1998; K ornienko,
2001; F oster & V incent, 2004; Teske et al., 2005; V an W assen b erg h et al., 2009).
140
C h a p t e r 6 - G e n e ra l d is c u ssio n
N o t only is th e tro p h ic system com pletely developed a t b irth , as is a p p aren t from
th e previous paragraphs, b u t it is also fully fu n ctio n al, resem bling th e ad u lt
feeding ap p aratu s b o th in form an d fu n ctio n . H ence, th e p o stp o n ed o n set of
exogenous feeding in syngnathids enables th e d irect developm ent of th e ad u lt
p h en o ty p e w ith o u t a tra n sitio n a l larval stage (Balon, 1979; 1986). This d irect
developm ent saves th e costs of ad d itio n al fu n c tio n a l rem odelling of th e tro p h ic
ap p aratu s and in ad d itio n it increases th e survival chances of th e young because
th e vulnerable free-living larval stage, ch aracterized by h ig h m ortality, is avoided
(Balon, 1986).
Juvenile p ivot feeding strikes are kinem atically sim ilar to th o se in ad u lts and
sta rt w ith a very fast ro ta tio n of th e hyoid follow ed by n eu ro cran ial elevation
an d finally m o u th opening (Van W assen b erg h et al., 2009). Prey cap tu re in new ly
released H. reidi juveniles occurs tw ice as fast as an ad u lt feeding sequence (2.5 ms
com pared to 5.5 ms) an d th e velocity of head ro ta tio n is over tw o tim es as high
(exceeding 30,ooo°s1 as opposed to less th a n i 4 , o o o ° s ( C h a p t e r 5.1; V an
W assen b erg h et al., 2009). T he p erio d of steady im p ro v em en t of prey cap tu re
perform ance d u ring ontogeny, as typically fo u n d in o th e r fishes (e.g. D rost, 1987;
C oughlin, 1994; R eiriz et al., 1998; W a rb u rto n , 2003), appears to be ab sen t in H.
reidi new borns (Van W assenbergh et al., 2009). H ow ever, prey cap tu re success is
low in n e w b o rn seahorses relative to adults, w h ich m ig h t be re la ted to th e
lo catio n of th e centre of ro ta tio n of th e head. In adults th e cen tre of ro ta tio n is
p o sitio n ed close to th e eyes, hence th e eyes w ill rem ain p ractically im m obile
durin g n eu ro cran ial elevation, facilitatin g visual p red atio n . Juveniles o n th e
o th e r hand, have th e ir cen tre of ro ta tio n fu rth e r aw ay fro m th e eye, w h ich m ight
negatively affect visual acuity (Roos et al., 2010). T herefore, th e gradual changes
in head and sn o u t dim ension d u rin g developm ent of th e juveniles probably helps
increasing prey cap tu re success.
In m any cases, significant o n to g en etic alte ratio n in shape an d p ro p o rtio n alm ost
certainly yields fu n c tio n a l change as w ell (E m erson & Bram ble, 1993). A clear
exam ple of th e effect of allom etrical g ro w th o n fu n c tio n is fo u n d in th e m outhopening m echanism in th e generalized cichlid Haplochromis elegans Trew avas
(O tten , 1982; Liem, 1991; E m erson & Bram ble, 1993). In early stages, m o u th
opening is caused by c o n tra ctio n of th e p ro tra c to r hyoidei m uscle (geniohyoideus
in O tte n , 1982). H ow ever, d u ring dev elo p m en t th e a rtic u la tio n b etw een low er
C h a p t e r 6 - G e n e ra l d is c u ssio n
141
jaw an d Suspensorium becom es displaced so th a t th e m uscle line of actio n ru n s
dorsal to th e articu latio n . H ence, co n tra ctio n of th e p ro tra c to r hyoidei m uscle no
longer results in m o u th opening. C o n cu rre n tly w ith th e change in p o sitio n of th e
suspensorial a rticu latio n , an altern ativ e low er jaw depression m echanism is
developed, i.e. th e o percular m o u th o p en in g system (O tten , 1982).
So far, no such k inem atical tra n s itio n or sh ift in th e sequence of m ovem ents of
skeletal elem ents was observed in post-release developm ent of syngnathids.
N e ith e r is th e re a perio d of ram feeding preceding th e definitive a d u lt su ctio n
feeding, as fo u n d in th e young (b o th larvae and juveniles) of m any o th e r fish (e.g.
Liem, 1991; C oughlin, 1994; C ook, 1996). T he fin d in g th a t th e fu n c tio n of th e
cranial stru c tu re s rem ains sim ilar th ro u g h o u t sy n g n ath id o n to g en y is likely
related to th e lack of su b stan tial shape changes (C hoo & Liew, 2006).
H ow ever, in all fishes th e allom etric increase in h ead an d sn o u t dim ensions w ill
also change th e size and shape of th e buccal cavity. In a su ctio n feeding event, it
is th e sudden increase in buccal volum e th a t generates su ctio n pressure, w h ich
eventually w ill cause th e prey to be tra n sp o rte d in to th e m outh. H ence th e
dim ension and shape of th e buccal cavity u n d o u b ted ly plays an im p o rta n t role in
su ctio n feeding hydrodynam ics. For exam ple, flow p a tte rn s d u rin g expansion are
reasoned to be su b stantially affected by th e size or shape of th e buccal cavity
(D rost et al., 1988; V an W assen b erg h et al., 2006b; Roos et al., 2011).
T he size an d shape of th e m o u th o p en in g is also th o u g h t to play a crucial role in
prey capture. S uction feeders w ith a sm all gape are able to produce a highvelocity w a te r flow and prey can th u s be sucked in fro m a g reater distance
(A lexander, 1967c; C ook, 1996). A d u lt peak gape size is sm aller th a n th e one in
juveniles w ith respect to th e ir sn o u t len g th , so it can be reasoned th a t ad u lts have
a gape m orphology th a t is hydrodynam ically advantageous fo r g en eratin g su ctio n
(Cook, 1996). The relatively larger gape in juveniles m ig h t be related to reducing
viscosity, w h ich plays a m ore im p o rta n t role in th e larval fish e n v iro n m en t th a n
in th e ad u lt one (chapter 4.4; W eb b & W eihs, 1986; H u n t von H erb in g & K eating,
2003). Roos et al. (2011) fo u n d th a t th e re is a difference in feeding p erform ance
b etw een sh o rt an d b ro ad sn o u ted juveniles o n th e one h a n d and long an d n arro w
sn o u ted adults on th e other. W h ile juveniles are able to draw a h ig h a m o u n t of
w a te r tow ards an d inside th e m o u th an d have a sh o rt expansion tim e, adults
p e rfo rm best in term s of cranial ro ta tio n , th u s m inim izing th e tim e to reach prey
142
C h a p t e r 6 - G e n e ra l d is c u ssio n
(Roos et al., 2011). This difference reflects a trad e-o ff th a t w ill be elab o rated u p o n
in ch ap te r 6.2.
In short, th e evolution of th e highly specialized feeding ap p aratu s likely has
b e n efited from th e extended p erio d of p a re n ta l care, since tim e available for
developm ent of th e rem arkable m orphology is th e re b y increased.
6.2.
Fu n c t i o n a l m o r p h o l o g y
T h ro u g h o u t th is d issertation, it has b een established th a t th e sy n g n ath id head
looks q u ite unlike th e head of a generalized teleost. A m ong th e m ost su b stan tial
differences are: (1) th e elo ngation of th o se suspensorial an d n eu ro cran ial bones
form ing
th e
tu b u la r
sn o u t
(i.e.
th e
q uadrate,
p reo p ercu lar,
sym plectic,
m etapterygoid, vom eral an d m esethm oid bones), (2) th e reduced, to o th less
prem axillary an d m axillary bones, (3) th e bones of each low er jaw h alf being fixed
to each other, (4) th e sh o rt hyoid, (5) th e p ecu liar shape of th e in terh y al bone,
w ith its tw o a rtic u la tio n heads ven trally form ing a saddle jo in t, (6) th e loss of th e
co n n ec tio n betw een low er jaw an d o p ercu lar bone, (7) th e large, stro n g ly ossified
o percular bone, and (8) th e rein fo rced cranio-vertebral a rtic u la tio n involving th e
p ecto ral girdle being fixed to th e v erteb ral colum n. It is n o t su rp risin g th a t these
m orphological featu res affect th e kinem atics of su ctio n feeding in syngnathids.
For exam ple, due to th e sm all m o u th o p en in g an d th e long buccal cavity, only a
sm all expansion of th e sn o u t is req u ired to g en erate a high-velocity w a ter flow
(Osse & M uller, 1980; M u ller & O sse, 1984; B ergert & W ain w rig h t, 1997; Roos et
al., 2009).
Syngnathids approach th e ir prey by m eans of a very fast d o rso ro ta tio n of th e
head, a feeding tech n iq u e k n o w n as pivot feeding (de L ussanet & M u ller, 2007;
Roos et al., 2009; su b m itte d a; su b m itte d b). In o rd er to investigate w h e th e r th e
m odifications of th e feeding ap p aratu s can be related to th e specialized pivot
feeding, a th o ro u g h descrip tio n of a typical sy n g n ath id feeding sequence and th e
role of th e m uscles involved in th is com plex feeding system w ill be given, based
C h a p t e r 6 - G e n e ra l d is c u ssio n
143
on observations of live anim als, results from ch ap ters 3.1, 3.2, 5.1 an d th e
p u b lic a tio n by Roos et al. (2009). All tim es su b seq u en tly m e n tio n e d are expressed
w ith respect to hyoid ro ta tio n , since tim e = o ms is defined as one video recording
fram e p rio r to th e o nset of its m ovem ent (see Fig. 5.5).
As in m ost su ctio n feeding teleosts, a syn g n ath id prey cap tu re event consists of
fo u r phases: th e p rep aratio n , expansion, com pression an d recovery phases
(Lauder, 1985).
In th e p re p a rato ry phase, th e prey is slow ly p u rsu e d e ith e r by sw im m ing up to it
or, for syngnathids w ith a prehensile tail, by stretch in g th e body to w ard s th e prey
w hile grasping o n to a holdfast. P rio r to a feeding strike, th e h ead is p o sitio n ed
suitably for prey cap tu re and tilte d ventrally, w ith o u t losing visual focus o n th e
prey.
The
head
depression
is
pro b ab ly
caused
by
c o n tra ctio n
of
th e
sternohyoideus m uscle th a t inserts o n th e urohyal b o n e (and indirectly, to th e
ceratohyal bones) an d is p a rtly co n flu e n t w ith th e hypaxial m uscles. A ctivity of
th e p ro tra c to r hyoidei m uscle (connecting th e d en tary b o n e w ith th e ceratohyal
p o sterio r bone) is req u ired to m ake sure th e e n tire cran iu m an d n o t only th e
hyoid rotates. To c o u n te r th e actio n of th e p ro tra c to r hyoidei m uscle o n th e
low er jaw an d hence p rev en t th e m o u th fro m opening, th e a d d u cto r m andibulae
m uscle, w h ich originates from th e d en tary and m axillary bones and in serts o n to
th e p reopercular, h y o m andibular an d sym plectic bones, pro b ab ly co n tracts as
well.
T he expansive phase starts w ith hyoid ro ta tio n and n o t m o u th o p en in g as usually
is th e case in teleosts (Lauder, 1985). N eu ro cran ial elevation follow s sh o rtly after
(0.4 ms) as b o th m ovem ents (neu ro cran iu m and hyoid ro tatio n ) are coupled (Fig.
5.5). M u lle r (1987) described a fo u r-b ar linkage, consisting of th e hyoid, th e
neu ro cran iu m -su sp en so riu m com plex, th e p ecto ral girdle and th e urohyal bone
to g e th e r w ith th e sternohyoideus m uscle (Fig. 1.6). A lth o u g h M u lle r’s m odel did
n o t en tirely fit th e k inem atical data of Roos et al. (chapter 5.1) due to sh o rten in g
of th e sternohyoideus-urohyal bar, a biom echanical coupling b etw een th e
m ovem ents of th e hyoid an d n e u ro c ran iu m does exist. P rio r to th e feeding strike,
th e hyoid is p ro tra c te d lying in b etw een th e left and rig h t suspensoria an d in line
w ith th e urohyal-sternohyoideus bar. T he epaxial an d hypaxial m uscles c o n tra c t
w hile th e hyoid is k e p t in place, th u s strain in g th e ir long tendons. T he elastic
C h a p t e r 6 - G e n e ra l d is c u ssio n
144
energy stored in these ten d o n s w ill th e n be rapidly released once th e hyoid
deviates from its locking p osition , resu ltin g in a sudden n eu ro cran ial elevation.
By m eans of such elastic recoil, th e lim itatio n s of m uscle perfo rm an ce can be
circum vented (e.g. A lexander & B ennet-C lark, 1977; A erts, 1998; Z ack et al., 2009).
A ccelerations are achieved th a t exceed th e o u tp u t of n o rm al m uscle co n tra ctio n
an d pow er is am plified, provided th a t th e energy is released in a sh o rte r tim e
span. The presence of th is pow er am p lificatio n m echanism was experim entally
proven for Syngnathus leptorhynchus (Van W assen b erg h et al., 2008). M u ller (1987)
proposed th e p ro tra c to r hyoidei to be involved in trig g erin g th e system. G iven
th e a tta c h m e n t of th e p ro tra c to r hyoidei m uscle o n th e hyoid v e n tra l to th e
a rtic u la tio n w ith th e Suspensorium , th is w o u ld im ply c o n tin u o u s c o n tra ctio n of
th e m uscle d u ring th e sto rin g of elastic energy to keep th e hyoid im m obile.
R elaxation of th e m uscle w ou ld cause hyoid depression an d hence dorsal ro ta tio n
of th e head. R ecently, Roos et al. (su b m itted b) suggested suspensorial a b d u ctio n
to trigger th e m echanism , based o n th e observation th a t hyoid ro ta tio n is only
possible if th e lateral sn o u t w alls are ab d u cted (de L ussanet & M u ller, 2007). It is
th e re fo re reasoned th a t co n tra ctio n of th e a d d u cto r arcus p a la tin i m uscle, w h ich
originates from th e p araspheno id b o n e an d in serts o n th e sym plectic and
h y o m and ibular bones, keeps
th e
suspensoria
ad d u cted
an d th e
four-bar
m echanism locked. W h e n th e m uscle relaxes, a sm all w id en in g of th e sn o u t
enables hyoid ro ta tio n , w h ich con seq u en tly releases th e sto red energy.
C om parison of pipefishes w ith seahorses show ed th a t th e sternohyoideus,
hypaxial an d epaxial m uscles are m u ch m ore massive in seahorses w hereas th e
epaxial an d hypaxial ten d o n s are m u ch lo n g er in pipefishes (C h ap ter 3.2). I
h ypothesized th a t in pipefishes th e elastic energy storage m echanism is favoured
over m uscle co n tra ctio n for g en eratin g th e p o w er needed to elevate th e
n eu ro cran iu m , w hereas in seahorses th e w ell developed m uscle b u n d les indicate
th e opposite (C hapter 3.2). A kin em atical analysis of pivot feeding in n ew b o rn
seahorses (Hippocampus reidi) carried o u t by V an W assen b erg h et al. (2009) show s
th a t th e angular velocity of b o th head and hyoid ro ta tio n is incredibly fast. T heir
inverse dynam ic analysis suggests th a t th e elastic recoil system m u st already be
p re sen t to pow er these prey cap tu re m ovem ents (Van W assen b erg h et al., 2009).
Pipefishes an d seahorses are b o th pivot feeders, w ith q u ite sim ilar feeding
behaviour an d th e same pow er am plifier system , how ever, th e m uscles th o u g h t to
C h a p t e r 6 - G e n e ra l d is c u ssio n
145
deliver th e hig h est frac tio n of to ta l su ctio n p o w er (i.e. epaxial an d hypaxial
muscles) are very d ifferen t in m orphology. In seahorses, th e sh o rt ten d o n s and
bulky epaxial an d sternohyoideu s m uscles seem n o t ideal fo r storage and release
of elastic energy. E lectrom yographical (EMG) research to analyze m uscle activity
p rio r to visible m ovem ent m igh t h elp d eterm in e h o w th ese m uscles generate th e
req u ired pow er. The long te n d o n s an d relatively sm all epaxial and hypaxial
m uscle d iam eter of pipefishes clearly reflect th e elastic recoil m echanism . Yet,
elongate sesam oid bones w ere fo u n d in th e epaxial te n d o n of som e pipefishes
(Doryrhamphus
dactyliophorus,
D.
melanopleura,
D.
janssi,
Dunckerocampus
pessuliferus an d Corythoichthys intestinalis) (C h ap ter 3.2). T h eir presence seems
q u ite puzzling since a bone inside a te n d o n is n o t likely to increase th e elastic
p ro p erties of th e ten d o n . H ow ever, if th e p o w er am p lificatio n is n o t im peded,
th e n it m ight p rev en t th e te n d o n fro m ru p tu rin g d u rin g c o n tra c tio n of th e
muscles.
T he m ain fu n c tio n of n eu ro cran ial ro ta tio n is to p o sitio n th e m o u th close to th e
prey, in a d d itio n to expanding th e buccal cavity ventrocaudally. As th e
n e u ro c ran iu m articu lates w ith th e v erteb ro -p ecto ral com plex, ro ta tio n w ill move
th e n eu ro cran iu m aw ay from th e fixed p ecto ral girdle.
H yoid depression w ill n o t su b stan tially c o n trib u te to a v en tral expansion of th e
buccal cavity, unlike in o th e r fish, fo r tw o reasons. First, th e hyoid is only
covered by a th in layer of skin th a t does n o t expand m u ch w h e n th e hyoid
retracts. T he second reason is th a t th e angle over w h ich th e hyoid tu rn s d u ring
su ctio n exceeds 900, h ence from 900 on, th e buccal floor is elevated in stead of
depressed. This w as confirm ed em pirically by tak in g X-ray images of an expanded
Hippocampus reidi head w ith a b ariu m so lu tio n in th e m o u th cavity (Roos et al.,
2009). N o flu id w as fo u n d in th e v en tral space created by hyoid depression.
T herefore th e hyoid is th o u g h t to have lost its role in expanding th e buccal cavity
ventrally. H ow ever, hyoid ro ta tio n plays an im p o rta n t role in low er jaw
depression an d suspensorial abduction. W h e n th e
sternohyoideus m uscle
con tracts, th e hyoid n o t only ro tates caudally, b u t also pushes th e p reo p ercu lar
bones outw ards, th u s w idening th e snout. T he exact m echanism of th is a b d u ctio n
rem ains unclear, b u t it m ight be th e resu lt of a tw ist of th e hyoid bars along th e ir
long axis. The shape an d o rie n ta tio n of th e hyoid sym physis allow s an increase of
th e angle b etw een th e long axes of th e hyoid bars d u rin g hyoid ro ta tio n , hence
146
C h a p t e r 6 - G e n e ra l d is c u ssio n
th e suspensoria w ill be p u sh ed outw ards. This k in d of tw ist w ill only be possible
if th e sym physis fu n c tio n s as a hinge jo in t an d if th e hyoid tu rn s as a w hole
(including th e in terh y al bone) a b o u t its p reo p ercu lar a rticu latio n , since th e
saddle-shaped jo in t b etw een in terh y al and ceratohyal p o sterio r b o n e restricts
m ovem ents o th e r th a n th o se in a ro stro cau d al p lan e (see ch ap ter 3.1).
A t 0.7 ms, m o u th opening starts (Fig. 5.5). The m o u th can be o p en ed in th re e
ways. First, as m e n tio n e d briefly before, co n tra ctio n of th e p ro tra c to r hyoidei
m uscle w ill resu lt in low er jaw depression. A second and a th ird m o u th opening
m echanism can be fo u n d in th e ligam entous coupling b etw ee n th e low er jaw and
th e hyoid: th e m andibulo-hyoid lig am en t o n th e one h a n d an d th e m andibuloin te ro p e rc u lar an d in teroperculo -h y o id ligam ents o n th e o th e r hand. D ue to
these linkages, ro ta tio n of th e hyoid w ill re su lt in op en in g of th e m outh. A
ligam entous c o n n ectio n betw een th e low er jaw an d th e m axillary bones
(prim ordial ligam ent) tran slates low er jaw depression in to a ro stral sw ing of th e
m axillary bones. This w ay th e m o u th is laterally closed resu ltin g in a m ore
circular gape w h ich is k n o w n to reduce drag (Alexander, 1967c; Lauder, 1979;
1985; B ergert & W ain w rig h t, 1997).
W h e n th e m o u th is closed in H. reidi, th e line of actio n of th e p ro tra c to r hyoidei
m uscle lies dorsal to th e low er jaw artic u la tio n (C h ap ter 3.2). C o n tra c tio n w ill
th e re fo re elevate th e low er jaw in stead of depressing it, in d icatin g th a t th e
m uscle does n o t play a crucial role in th e in itial phase of m o u th opening. B ranch
(1966) goes one step fu rth e r and states th a t th e p ro tra c to r hyoidei m uscle is
superfluous in th e m o u th opening m echanism , b u t is needed fo r p ro tra c tio n of
th e hyoid. H e based his statem e n t o n th e observation th a t c u ttin g th e p ro tra c to r
hyoidei m uscle in a sacrificed Syngnathus acus specim en h ad n o effect on
depression of th e low er jaw, as c o n tra ctio n of th e stern o h y o id eu s m uscle is
tra n sm itte d by th e ligam ents. I fo u n d th a t th e p ro tra c to r hyoidei m uscle is
enclosed by th e m andibulo-hyoid lig am en t in D. dactyliophorus, D. pessuliferus, C.
intestinalis an d H. abdominalis (C h ap ter 3.2). C o n tra c tio n of th e m uscle and th e
ligam ent is likely n o t coupled as th e y are sep arated fro m each o th e r by a small
circular cavity.
K inem atical data show ed th a t low er jaw depression is only in itia te d a fte r th e
hyoid has ro ta te d over 8o° (Roos et al., 2009). This signifies e ith e r th a t th e
ligam ents can only generate su fficien t te n sio n to o p en th e m o u th after
C h a p t e r 6 - G e n e ra l d is c u ssio n
147
su b sta n tia l hyoid ro ta tio n , or th a t m o u th o p en in g is co u n terac te d d u rin g th e first
8o° of hyoid ro ta tio n . If so, th e n th is c o u n teractin g is pro b ab ly achieved by m eans
of a d d u cto r m andibulae m uscle co n tractio n . The delay of m o u th o p en in g can be
considered an advantage, as it avoids in itia tio n of su ctio n feeding w h e n th e prey
is n o t y et w ith in reach.
A n o th e r m o u th opening m echanism o fte n fo u n d in teleo sts is th e opercular
linkage, w h ich tran slates dorsocaudal m ovem ent of th e o p ercu lar bone, caused by
c o n tra c tio n of th e levator o percu li m uscle (attach in g v en trally o n th e braincase),
th ro u g h re tra c tio n of th e in te ro p e rc u lar b o n e an d th e m an d ib u lo -in tero p ercu lar
ligam ent to depression of th e low er jaw (e.g. A nker, 1974; Lauder, 1985; W estn e at,
1990; H u n t von H erb in g et al., 1996a; W e stn e a t, 2004; V an W assen b erg h , 2005).
This system is n o t fu n c tio n a l in syngnathids since a m echanical lin k b etw een th e
o percular an d in te ro p e rc u lar bones is missing. T he h ig h su b -am b ien t pressures in
th e opercular cavity d u ring su ctio n feeding m ig h t be associated w ith re d u c tio n of
th e size of th e gili slit and to stre n g th en in g of th e o p ercu lu m (Osse & M uller,
1980; M u ller & Osse, 1984). This re in fo rce m e n t likely also im plies a lim ita tio n of
o percular m obility; hence th e loss of o p ercu lar elevation and coupled low er jaw
depression seems beneficial. T he absence of a levator o p ercu li m uscle in some
pipefishes, e.g. D. dactyliophorus (C h ap ter 3.2) an d Syngnathus acus (Branch, 1966)
is very likely related to th e decoupling of th e low er jaw an d th e o p ercu lar bone.
D uring th e first m illiseconds of th e expansive phase, th e volum e of th e buccal
cavity is actually reduced by m eans of a lateral com pression of th e head. This is
probably a consequence of c o n tra ctio n of th e a d d u cto r arcus p a la tin i and th e
a d d u cto r o perculi m uscles (connecting th e o p ercu lar bone to th e p ro o tic bone),
resu ltin g in a m edial m ovem ent of th e Suspensorium an d operculum , respectively.
A t 1.5 ms, th e sm all gili slits o p en and next, a ro u n d 2 ms, th e sn o u t starts to
expand at th e level of th e low er jaws, w h ich is b ro u g h t a b o u t by th e abducing
force c o m p o n en t of low er jaw depression th a t pushes th e q u ad rate bones
outw ards. This is rapidly (2.5 ms) follow ed by a d ila ta tio n of th e m ore p o sterio r
p a rt of th e sn o u t, caused by hyoid depression (over 90e) an d p o te n tia lly aided by
c o n tra c tio n of th e levator arcus p a latin i m uscle th a t has its origin o n th e
sp h en o tic bone an d its in sertio n o n th e hy o m an d ib u lar and p reo p ercu lar bones.
T he resu ltin g suspensorial a b d u ctio n causes th e m ain expansion of th e buccal
cavity.
148
C h a p t e r 6 - G e n e ra l d is c u ssio n
A t 2.6 ms after th e sta rt of th e expansive phase, peak gape is achieved (Fig. 5.5),
follow ed by th e n e u ro c ran iu m reaching its maxim al ro ta tio n (3.6 ms). V ery soon
a fte r th a t, th e prey starts to move in th e d irectio n of th e m o u th op en in g (4.0 ms).
A t 4.8 ms th e opercula are w id en ed by m eans of th e d ila ta to r o p ercu li (in sertio n
on th e o percular bone, origin on th e hyom andibular, p arasp h en o id an d sp h en o tic
bones) an d hyohyoideus ab d u cto r m uscles (originating fro m th e ceratohyal
p o sterio r bone and in sertin g on th e first b ran ch io steg al ray). T he gili slits are
sealed by th e branchiostegal m em brane an d th e tw o slender b ran ch io steg al rays
ju st p rio r to th e prey being sucked in (5.7 ms).
In syngnathids, th e com pression phase is ch aracterized by th e sta rt of m o u th
closure, w h ich is accom plished by co n tra ctio n of th e a d d u cto r m andibulae
m uscle complex. T he d iffe re n t bundles of th is m uscle elevate th e low er jaw and
also re tra c t th e m axillary bones to th e ir m ore caudal position. T he m axim al
buccal expansion is generated, m oving th e prey fu rth e r th ro u g h th e sn o u t (14.3
ms). A t 19.6 ms, peak hyoid ro ta tio n is reached, a fte r w h ich th e hyoid is
p ro tra c te d again by co n tra ctio n of th e p ro tra c to r hyoidei m uscle. Also th e
Suspensorium starts to adduce by co n tra ctio n of th e a d d u cto r arcus p a latin i
m uscle, aided by co n tra ctio n of th e in term an d ib u laris m uscle (connecting left
an d rig h t d en tary bones) th a t pulls b o th low er jaw halves to g eth er. The m o u th is
n o t fully closed u n til a b o u t 420 ms. D ue to c o n tra ctio n of th e a d d u cto r o p ercu li
m uscle, w h ich causes an in w ard m ovem ent of th e o p ercu lar bone an d pulls th e
branchiostegal rays to g eth er, w a ter is passed over th e gills an d expelled th ro u g h
th e tin y b ran ch ial openings.
D uring th e recovery phase all elem ents re tu rn to th e ir original position. The
n e u ro c ran iu m is com pletely depressed and th e hyoid is p ro tra c te d again. This
phase lasts u n til approxim ately 600 ms a fte r in itia tio n of prey cap tu re, som etim es
longer.
C om parison b etw een th e k inem atics of th e sy n g n ath id p iv o t feeding an d th o se of
a general te le o st su ctio n feeding event reveals th e follow ing: (1) m ovem ent of
cranial elem ents in syngnathids does n o t p roceed in a ro stro cau d al sequence
startin g w ith m o u th opening, follow ed by hyoid depression an d finally head
ro ta tio n . H ow ever, lateral expansion does follow th e ro stro cau d al w ave typical
for m any aquatic su ctio n feeders (see B ishop et al., 2008 an d references w ithin).
C h a p t e r 6 - G e n e ra l d is c u ssio n
149
T he low er jaws abduct first, n ex t th e hyoid bars p u sh th e suspensoria ou tw ard s
an d lastly th e opercula w iden. (2) Buccal volum e increase is m ainly caused by
lateral sn o u t abduction, th e re is a c ertain am o u n t of caudal expansion due to
n eu ro cran ial elevation an d no ro stral or v en tral expansion (absence of m axillary
p ro tru sio n and low ering of th e buccal floor). (3) N eu ro cran ial elevation is
su b sta n tia l an d is used to p o sitio n th e m o u th close to th e prey. (4) A n o p ercu lar
m o u th opening m echanism is absent.
So th e sy ngnathid tro p h ic ap p aratu s does n o t only diverge fro m a g eneral teleo st
in its m orphology, b u t also in fu n ctio n . The pivot feeding tech n iq u e yields
exceptionally sh o rt prey cap tu re tim es, w ith large forces, velocities and
accelerations involved. H ence, th e cranial m orphology is likely d eterm in ed by
m echanical dem ands associated w ith th is type of feeding. The fin ite elem ent
analysis th a t sim ulated su ctio n pressure in th e cran iu m of several seahorses and
pipefishes confirm ed this. H ig h est stress m agnitudes w ere fo u n d in th o se regions
expected to be able to w ith sta n d h ig h am o u n ts of com pressive stress (i.e. cartilage
an d articulations). Also, head m orphology is fo u n d to
affect th e
stress
d istrib u tio n g en erated durin g feeding. H ow ever, n o sim ple re la tio n w as fo u n d
w ith relative sn o u t length. N evertheless, sn o u t le n g th and d iam eter w ill probably
in flu en ce prey cap tu re perform ance. The fam ily of S yngnathidae consists of
species w ith m any d ifferen t degrees of sn o u t dim ension, yet all syngnathids
recorded so far exhibit th e same pivot feeding strategy. This raises th e qu estio n
w h a t th e m ost favourable sn o u t le n g th is fo r carrying o u t pivot feeding, or m ore
specifically, w h e th e r sn o u t elo n g atio n is related to an increase in feeding
perform ance.
In theory, an elongate sn o u t is beneficial because it decreases th e head ro ta tio n
angle to bridge a certain distance b etw een th e sn o u t tip an d th e prey (K endrick &
H yndes, 2005; de L ussanet & M u ller, 2007). Prey can th e re fo re be a ttack ed from
fu rth e r away and, as th e p re d a to r does n o t need to ap p ro ach its prey closely to
in itia te a feeding strike, it w ill be less likely to be d etected an d has a g reater
chance to cap tu re th e prey (Flynn & R itz, 1999; K endrick & H yndes, 2005). Also,
lon g -sn o u ted species should be able to achieve a h ig h er angular velocity of th e
head, provided th a t th e mass to be ro ta te d is m inim ized (M uller & O sse, 1984).
C onsequently, species w ith a long sn o u t are able to cap tu re m ore rapidly
150
C h a p t e r 6 - G e n e ra l d is c u ssio n
sw im m ing prey (P ranzoi et al., 1993; K endrick & H yndes, 2005; O liveira et al.,
2007). Besides th a t, a sm all gape allow s th e fish to be m ore selective in th e u p tak e
of prey (Branch, 1966). O n th e o th e r hand, feeding th ro u g h a long and slender
tu b e is th o u g h t to en tail several physical constraints. Firstly, w ith increasing
sn o u t length, th e m o m en t of in e rtia d u rin g h ead ro ta tio n increases as w ell (de
Lussanet & M uller, 2007). To com pensate fo r th is, th e cross-sectional area of th e
sn o u t m u st be m inim ized, how ever a sm all gape is unfav o u rab le fo r th e cap tu re
of large prey an d th u s a tin y m o u th o p en in g lim its syngnathids to consum e only
sm all prey item s. Secondly, a long sn o u t also increases th e distance b etw een th e
eye and th e prey and given th a t syngnathids are visual p red ato rs, th is m ight
reduce th e perform ance in term s of accurate aim ing an d p o sitio n in g of th e sn o u t
tip relative to th e prey (Roos et al., su b m itte d a). Thirdly, th e lo n g er th e sn o u t, th e
larger th e sub-am bient pressure in th e buccal cavity m u st be to create th e same
flow velocity as in sh o rte r sn o u ted species (Poiseuille’s law). Finally, th e to ta l
buccal volum e increase is expected to be sm aller and expansion tim e lo n g er in
lon g -sn o u ted species, w h ich w ill negatively affect su ctio n perfo rm an ce (Roos et
al., 2011).
Some ex p erim en tal analyses on th e re la tio n b etw een sn o u t le n g th and prey type
or prey cap tu re p erform ance have b een carried out. A com parison b etw een th e
feeding kinem atics of th e long-sn o u ted p ip efish D. dactyliophorus and th e sh o rt­
sn o u ted D. melanopleura show ed th a t b o th species cap tu re prey by m eans of th e
sam e stereotypical pivot feeding kinem atics, b u t differences in certain aspects of
th e feeding strike could be observed (Roos et al., su b m itte d a). T he lo n g-snouted
species tu rn s its head over a larger angle to b rin g its m o u th close to th e prey,
hence a larger distance is covered. In addition, D. dactyliophorus is able to generate
a hig h er velocity of th e m o u th d u rin g h ead ro ta tio n , th u s m inim izing th e tim e to
reach th e prey. A dietary study show ed th a t lo n g -sn o u ted syngnathids consum e
m ore elusive prey from a w ide range of available prey types th a n sh o rt-sn o u ted
species, confirm ing th e sh o rt tim es to reach prey (K endrick & H yndes, 2005). The
sh o rt-sn o u te d D. melanopleura, how ever, cap tu red its prey in less tim e and h ad a
h igher success rate, likely due to th e sm aller distance b etw een eyes and m o u th
(Roos et al., su b m itte d a). Successful prey cap tu re appears to d ep en d greatly on
adequate p o sitio n in g of th e sn o u t tip relative to th e prey.
C h a p t e r 6 - G e n e ra l d is c u ssio n
151
A resem blance w ith b u tterfly fish es (C haetodontidae) can be observed as th is
fam ily also encom passes b o th sh o rt- an d long-jaw ed species. T h eir rem arkable
feeding apparatus consists of elongated prem axillary bones an d low er jaw th a t
are n o t fused to form a tu b e, w h ich differs from syngnathids w here th e bones of
th e eth m o id region form an elo n g ated snout. A lth o u g h b earin g te eth , th e longjaw ed b u tterfly fish es (Chelmon rostratus (L.), Forcipiger flavissimus Jordan &
M cG regor an d Forcipiger longirostris (Broussonet)) cap tu re th e ir prey by m eans of
a c o m b in atio n of su ctio n an d ram feeding (F erry-G raham et al., 2001a). T he prey
is approached by sw im m ing or by p ro tru d in g th e long an d slender jaws tow ards
it. F erry-G raham et al. (2001a) d em o n strated th a t th e c h ae to d o n tid w ith th e
longest jaws (F. longirostris) in itia te d its feeding strik e fro m a g reater distance
com pared to o th e r (shorter jawed) b u tterfly fish es, in d icatin g th a t a lo t of ram was
p erfo rm ed to catch th e prey. Besides th a t, th e d iet of th is high-perform ance
species is k n o w n to be exclusively com prised of elusive prey. A n o th er study, in
w h ich kin em atical data of five c h ae to d o n tid species was analyzed, revealed th a t
th e tw o short-jaw ed species never m issed a strik e w hereas several failed atte m p ts
a t prey cap tu re w ere observed in th e long-jaw ed b u tterfly fish es (Ferry-G raham et
al., 200lb). These findings are sim ilar to th e observations in lo n g -sn o u ted
syngnathids, th a t ro ta te th e ir head over a g reater angle (i.e. in itia te a prey cap tu re
event from fu rth e r away), consum e m ore highly m obile prey and have low er
feeding success rate com pared to sh o rt-sn o u ted species. B oth exam ples su p p o rt
th e statem e n t by G ibb & F erry-G raham (2005) th a t as individuals of a species
becom e m ore specialized in catching elusive prey, th is w ill be reflected in
m odification of th e ir su ctio n feeding ability, e ith e r by changing th e ir prey
cap tu re kinem atics or th e ir feeding ap p aratu s m orphology.
All th is suggests th a t no o ptim al sn o u t le n g th exists fo r p erfo rm in g pivot feeding.
R ather, th e re seems to be a trade-o ff b etw een favouring th e p erfo rm an ce of head
ro ta tio n and m axim izing su ctio n perfo rm an ce (Roos et al., 2011). These d ifferen t
strategies are reflected in e ith e r a long sn o u t w ith a sm all d iam eter and a high
angular velocity to decrease th e tim e to reach prey o n th e one hand, or sh o rten in g
of th e sn o u t, th u s increasing th e su ctio n volum e an d th e velocity of th e w ater
flow on th e o th e r hand. T he developm ental sh ift fro m short- an d b ro ad -sn o u ted
juveniles to longer an d m ore slender sn o u ted adults could be related to th is (see
152
C h a p t e r 6 - G e n e ra l d is c u ssio n
ch ap te r 6.1 and Roos et al., 2010). T he trad e-o ff m ig h t explain th e large
m orphological diversity in relative sn o u t dim ension p re sen t in Syngnathidae.
6.3.
E v o lu tio n a ry m o rp h o lo g y
A sig nificant a m o u n t of research has b een carried o u t to resolve th e phylogeny of
th e P ercom orpha in general and th e G asterosteiform es in p a rtic u la r (e.g. Jo h n so n
& P atterso n , 1993; C h en et al., 2003; M iya et al., 2003; S m ith & W h eeler, 2004;
D e tta i & L ecointre, 2005; M iya et al., 2005; N elson, 2006; K eivany & N elson, 2006;
K aw ahara et al., 2008; W ilso n & O rr, subm itted). In spite of th is, affin ities are still
n o t very w ell understood. As m e n tio n e d in th e in tro d u c tio n (C h ap ter 1), th e
m onophyly
of
th e
order
G asterosteiform es,
as
established
based
on
m orphological studies (Fig. 1.2; Jo h n so n & P atterso n , 1993; N elson, 2006; Keivany
& N elson, 2006), is recen tly d isp u ted by m olecular d ata (Fig. 1.3; C h en et al., 2003;
S m ith & W h eeler, 2004; D e tta i & L ecointre, 2005; K aw ahara et al., 2008).
A ccording to th e la tte r, th e o rder should be divided in to th re e groups, w h ich are
th o u g h t to have diverged early in P erco m o rp h evolution. The first gro u p consists
of th e Indostom idae, w hose phylogenetic p o sitio n has always b een controversial
(M cA llister, 1968; Joh n so n & P atterso n , 1993; B ritz & Johnson, 2002; N elson,
2006; K aw ahara et al., 2008) being placed w ith in th e G asterosteoidei (Britz &
Johnson,
2002; T akata
& Sasaki, 2005;
N elson,
2006), in
th e
su b o rd er
Syngnathoidei (Keivany & N elson, 2006) or n ested w ith in a n o th er order,
S ynbranchiform es (K aw ahara et al., 2008). T he second clade com prises th e fo rm er
suborder G asterosteoidei an d th e
th ird group encom passes th e
su b o rd er
Syngnathoidei (K aw ahara et al., 2008). A ccording to th is vision, th e S yngnathoidei
clu ster
to g e th e r
w ith
th e
D actylopteridae
(flying
gurnards),
how ever
m orphological data does n o t su p p o rt th is sister tax o n relatio n sh ip (Johnson &
P atterso n , 1993).
A lth o u g h th e G asterosteiform es rem ain a p ro b lem atic clade an d fu rth e r analyses,
b o th
m orphological
an d
m olecular
(perhaps
using
n u clear
in stead
of
m ito ch o n d rial gene data), are needed to resolve th e in te rre la tio n s w ith in th e
C h a p t e r 6 - G e n e ra l d is c u ssio n
153
clade, th e previously described in consistency does n o t affect th e results of th is
dissertation. Gasterosteus aculeatus (three-spined stickleback), used as rep resen tin g
th e ancestral c o n d itio n for a com parison w ith Syngnathus rostellatus and
Hippocampus capensis in ch ap te r 3.1, m ig h t n o t be a sister gro u p species of th e
S yngnathidae, b u t it rem ains a generalized p erco m o rp h in b o th p o in ts of view,
an d is th e re fo re still suitable as an out-group. T he results of th is com parison are
already briefly discussed in ch ap ter 6.2.
T he first evidence of Syngnathidae in th e fossil reco rd dates back to th e M iddle
Eocene (ca. 50 M ya) (P atterson, 1993; Teske et al., 2004). These ancestors resem ble
e x ta n t pipefishes, b o th having th e head in line w ith th e body, a tu b u la r sn o u t
w ith a sm all m o u th an d bony plates covering th e body (Teske & Beheregaray,
2009). T he tra n s itio n from p ipefish to seahorse m orphology (i.e. th e u p rig h t body
axis w ith th e tilte d head in co m b in atio n w ith th e preh en sile tail) is n o t
rep resen ted in th e fossil record th a t is hence in ad eq u ate to elucidate all
evolutionary p a tte rn s w ith in th e family. T he only k n o w n seahorse fossils w ere
fo u n d in M arecchia in Italy (ca. 3 Mya) and recen tly in Tunjice in Slovenia (ca. 13
M ya) (Teske et al., 2004; Z alohar et al., 2009). D ue to th ese findings an d th e fact
th a t seahorses have a cosm opolitan d istrib u tio n , m eaning th a t divergence of th e
genus Hippocampus from th e o th e r syngnathids should have o ccu rred b efo re th e
com plete closure of th e T ethyan seaway, th e origin of seahorses is th o u g h t to be
a t least 20 M ya (Casey et al., 2004; Z alo h ar et al., 2009). A close affin ity b etw een
seahorses an d p ipefish of th e genus Syngnathus has been established based on
m olecular data (W ilson et al., 2001; W ilso n & Rouse, 2010). The o b tain ed
phylogeny corresponds surprisingly w ell w ith th e relatio n sh ip s pro p o sed by
H erald (1959), w h ich he in fe rred fro m studying b ro o d p o u ch m orphology w ith in
th e Syngnathidae (Fig. 1.5). W ilso n et al. (2001; 2003) state th a t th e extrem ely
rapid diversification of th e brood in g stru c tu re s m ig h t be th e driving force b eh in d
th e evolution of th e family. T he a tta c h m e n t of eggs and em bryos to th e m ale
body is th o u g h t to be derived fro m n e st b u ild in g and guarding behaviour of m ale
ancestors (Baylis, 1981). From gluing organic m aterial to g e th e r to b u ild a nest, as
observed in th e th ree-sp in ed stickleback, it m ay be only a sm all step to gluing
em bryos on th e body surface.
C h a p t e r 6 - G e n e ra l d is c u ssio n
154
R ecently, a n o th e r m olecular phylogeny (Teske & B eheregaray, 2009) co n trad icts
th e close relatio n sh ip b etw een Syngnathus pipefishes and seahorses, in d icatin g
Idiotropiscis pygm y pipehorses, i.e. seahorse-like pipefishes th a t sw im h o rizo n tally
(K uiter, 2004), to be th e sister tax o n of seahorses. These pipehorses can be
considered as th e living evolu tio n ary lin k b etw een pipefishes and seahorses.
S eparation b etw een th e pipehorse and th e seahorse lineages occurred d u rin g th e
Late O ligocene (ca. 25-28 M ya; hen ce m u ch earlier th a n th e first k n o w n seahorse
fossil) an d is suggested to be b ro u g h t a b o u t by h a b ita t association (Teske &
Beheregaray, 2009). T ectonic events in th e W e s te rn Pacific O cean d u rin g th is
geological p eriod re su lted in th e increase of shallow -w ater areas d o m in ated by
seagrass beds. Fragm ents of seagrasses w ere fo u n d to g e th e r w ith th e seahorse
fossils in th e Tunjice H ills, giving an in d icatio n of th e an cestral h a b ita t (Z alohar
et al., 2009). The seahorses’ vertical body p o stu re greatly b e n e fitte d by this
vegetatio n since seagrasses provide b e tte r cam ouflage com pared to algal reefs
(Teske & Beheregaray, 2009). Also p resen tly , seahorses occur a b u n d an tly in th is
h a b ita t because seagrasses form adequate holdfasts an d are usually associated
w ith h igh co n ce n tra tio n s of copepods an d o th e r im p o rta n t sy n g n ath id prey item s
(T ipton & Bell, 1988; F o ster & V in cen t, 2004; V izzini & M azzola, 2004; C u rtis &
V incent, 2005; M u ru g a n et al., 2008; Scales, 2010). Sim ilar to th e u p rig h t seahorse
body, th e selective pressure tow ard s a m ore cryptic m orphology m ig h t also have
led to th e ev olution of fleshy appendages in seadragons (W ilson & Rouse, 2010).
A p art from th e highly specialized b ro o d p o u c h as a key in n o v atio n of th e fam ily
or th e ecologically based selection fo r ev o lu tio n fro m p ip efish to seahorse, th e
biom echanics of prey cap tu re m ig h t also have b een a driving force. In ch ap te r 4,
th e geom etric m o rp h o m etric analysis yielded som e differences in general head
m orphology
b etw een
pipefishes
and
seahorses.
T he
m ost
im p o rta n t
m odifications th a t characterize th e tra n sitio n fro m a p ip efish to a seahorse
include a m ore ven tral p o sitio n of th e p ecto ral fin, a decrease in sn o u t le n g th and
an increase in sn o u t, head and o p ercu lar heights. I reasoned th a t all these
differences could be related to th e tilte d head of seahorses, e ith e r directly (e.g.
displacem ent of th e p ecto ral fin along w ith th e body) or in d irectly (as a
consequence of head rotation). The tiltin g of th e h ead w ith resp ect to th e body is
suggested to allow a greater degree of head ro ta tio n in seahorses (C h ap ter 4).
H ence, a larger distance to prey could be sp an n ed w ith o u t having to increase th e
C h a p t e r 6 - G e n e ra l d is c u ssio n
155
sn o u t length. If th e prey is in fro n t an d above th e m o u th , ro ta tio n of th e head
over a large angle w ill b rin g th e tip of th e sn o u t closer to th e prey along th e
dorsoventral axis, how ever, it increases th e distance to th e prey along th e
rostro cau d al axis. So, in order to suck in th e prey, a larger w a ter volum e has to be
p u t in to m otion, w h ich requires m ore energy. To overcom e th is problem , th e
sn o u t tip can be p o sitio n ed som ew hat v e n tro ro stra l to th e prey before th e strike,
as m odelled in Fig. 5.7B (note th e p o sitio n of th e m o u th w ith resp ect to th e
o u tflo w of th e pipette). H ow ever, th is m ig h t increase th e risk of d e tec tio n by th e
prey an d subsequently trig g er an escape response. A lternatively, th e e n tire head
can be m oved a n terio rly d u ring ro ta tio n , w h ich is su b tly visible in Fig. 5.7A (also
n o te th e m ore p o sterio r p o sitio n of th e m o u th a t th e o n set of n eu ro cran ial
elevation in th e video fram e com pared to th e model). This a n terio r m ovem ent of
th e m o u th w as also observed by V an W assen b erg h et al. (2011). H ig h speed
recordings of prey in ta k e in c o m b in atio n w ith m odelling of th e h ead ro ta tio n
show ed th a t a m ore p ro n o u n c e d angle b etw een h ead an d body, as in seahorses, is
associated w ith an a n te rio r tra n sla tio n of th e head, in a d d itio n to its ro tatio n .
This way, th e m o u th moves tow ard s th e prey, w h ich can th u s be c ap tu red from
fu rth e r away. O n th e o th e r hand, m athem atical m odelling suggests th a t a head in
line w ith th e body, as in pipefishes, favours h ig h er velocities of th e m o u th d u ring
head ro ta tio n (Van W assenberg h et al., 2011). The tilte d h ead of seahorses is
considered to act in th e same w ay as th e elo n g atio n of th e sn o u t observed in
m any syngnathids. T hey b o th increase th e prey distance, enlarge th e space in
w h ich prey can be c ap tu red an d th e re fo re a larger n u m b er of food item s m ig h t be
cau g h t (Van W assen b erg h et al., 2011). Roos et al. (2011) fo u n d th a t th e tim e to
reach prey is decreased by increasing sn o u t le n g th d u rin g ontogeny. A pparently,
tim e to reach prey can also be m inim ized by h ead ro ta tio n in co m b in atio n w ith
a n te rio r m ovem ent of th e m o u th as in seahorses. Perhaps th is explains th e
absence of extrem ely lo n g-sno u ted seahorses. The seahorse w ith th e longest
sn o u t is Hippocampus histrix K aup (Lourie et al., 2004), w h ic h has a sn o u t le n g th
of a b o u t h alf its head length. In terestin g ly , th e sn o u t of th e tw o fossil seahorse
species fo u n d in Slovenia (H. sarmaticus an d H. slovenicus), also m easures half th e
le n g th of th e head (Z alohar et al., 2009). Pipefishes such as Doryrhamphus
dactyliophorus an d Dunckerocampus pessuliferus, how ever, have a m u ch m ore
elongate snout, m easuring over 65% of th e head length.
C h a p t e r 6 - G e n e ra l d is c u ssio n
156
D uring a seahorse feeding strike, th e a n terio r disp lacem en t of th e m o u th is n o t
always accom plished by sw im m ing to w ard s th e prey, b u t by stretch in g th e body
tow ards it, w hile keeping th e tail w rap p ed a ro u n d a holdfast. H ence, th e
ev olution to an increasingly b e n t head m ig h t be associated w ith th e loss of th e
caudal fin and th e ta il becom ing prehensile. C hoo & Liew (2006) observed some
vestigial caudal fin rays in new ly b o rn Hippocampus kuda th a t disappear a fte r a
few days, confirm ing th e secondary loss of th e caudal fin d u rin g evolution. G iven
th e pipehorse m orphology, being a pipefish-like body w ith a seahorse-like tail, it
is likely th a t th e origin of th e grasping tail an d th e associated sit-and-w ait feeding
behaviour preceded th a t of th e tilte d head in c o m b in atio n w ith a n terio r
tra n sla tio n of th e m o u th (Van W assen b erg h et al., 2011). K inem atical analysis of
prey cap tu re in Idiotropiscis pipehorses m ight yield in te re stin g results for
u n d erstan d in g th e evolutionary p a tte rn th a t has led to th e seahorse m orphology
an d feeding strategy.
6 .4 .
G e n e ra l c o n c lu s io n
This d issertatio n focuses on th e rem arkably specialized feeding ap p aratu s of th e
Syngnathidae. As in m any evolu tio n ary m orphological studies, th e research
started w ith questions such as ‘W h a t does it consist of?’, ‘H o w does it w ork?’, and
‘W h y does it look th e w ay it looks?’. These q uestions seem ed especially
in te re stin g for syngnathids since th e ir head looks so u n u su al w ith th e tu b u la r
sn o u t an d sm all jaws. Prey cap tu re is incredibly fast an d y et th e advantages of th e
m orphology are n o t im m ediately a p p aren t given th e hydrodynam ic lim itatio n s.
H ence, a research pro ject was in itia te d w ith c en tral aim to elucidate th e
evolutionary p a tte rn th a t has led to th e extrem e m orphological specialization of
Syngnathidae.
In th e first place, th e results of th is d issertatio n yield a n u m b e r of stru c tu ra l
features th a t probably are related to th e pivot feeding. A n exam ple of th is is th e
in terh y al bone th a t deviates greatly fro m th e typical teleo stean rod-shape in
bearing tw o a rtic u la tio n heads fo r th e hyoid. T he saddle jo in t th u s form ed,
C h a p t e r 6 - G e n e ra l d is c u ssio n
157
con strain s hyoid m ovem ent o th e r th a n ro ta tio n in th e sagittal plane. A p art from
th is an d o th e r m odifications, th e syngnathid h ead is also ch aracterized by
coupling events. The m ost im p o rta n t of th ese is fo u n d in th e fo u r-b ar system th a t
links hyoid ro ta tio n to n eu ro cran ial elevation and at th e same tim e w orks as a
pow er am plifier. Also a decoupling is observed as th e o p ercu lar m o u th opening
m echanism is lost. These m orphological tra its are highly conserved am ong th e
family: th e y could be d em o n strated in all stu d ied species.
T he fam ily is characterized by a large m orphological v ariatio n w ith regard to
sn o u t dim ensions. The in itia l assu m p tio n w as th a t species w ith a longer sn o u t are
m ore specialized for catching elusive prey, w h ich w as d e m o n strated by a dietary
analysis (K endrick & H yndes, 2005). By m eans of a p relim in ary geom etric
m o rp h o m etric analysis it was established th a t lo n g -sn o u ted syngnathids have a
reduced intraspecific v ariatio n in term s of th e ir h ead shape, as expected fo r a
specialized m orphology. H ow ever, a k in em atical study proved th a t p ip efish w ith
a long sn o u t do n o t necessarily p erfo rm faster pivot feeding th a n sh o rt-sn o u ted
species (Roos et al., su b m itte d a). M oreover, w h e n looking a t th e syngnathid
phylogeny (Fig. 1.5), th e re is no ev o lu tio n to w ard s a m ore elongate sn o u t since
th e longest-snouted species are fo u n d in th e Phyllopteryx and Doryrhamphus
genera and n o t am ong seahorses (Hippocampus), w h ich are considered as
phylogenetically th e m ost derived syngnathids. This suggests th e presence of a
trad e-o ff ra th e r th a n a straig h tfo rw ard ev o lu tio n ary trend. A long sn o u t is
favoured to increase th e angular velocity of h ead ro ta tio n and to m inim ize th e
tim e to reach prey, w hereas a sh o rt sn o u t is advantageous fo r increasing su ctio n
volum e an d velocity (Roos et al., 2011).
Finally, th e exceptional reproductive strateg y an d pro lo n g ed p a re n ta l care very
likely played an im p o rta n t role in th e ev o lu tio n of th e feeding apparatus, since
th e y allow th e release of fully in d e p e n d e n t juveniles th a t resem ble adults b o th in
form an d function.
Perhaps m ost im p o rtan tly , th e results of th is d issertatio n h ig h lig h t th a t
Syngnathidae, w ith th e ir highly specialized feeding apparatus, extrao rd in ary
developm ent and u nequalled perform ance, establish conclusive p ro o f fo r th e
incredible capabilities of n a tu ra l selection an d evolution.
158
159
Summary & samenvatting
7.1.
S u m m ary
This PhD thesis deals w ith th e ev o lu tio n of th e specialized h ead m orphology in
Syngnathidae, a fam ily am ong th e b o n y fishes th a t com prises pipefishes,
seahorses, seadragons and pipehorses. A detailed m orphological d escrip tio n of th e
feeding ap p aratu s of several sy n g n ath id rep resen tativ es form s th e core of th is
research. Besides th a t, o th e r aspects like m orphological v ariatio n w ith in th e
fam ily, perform ance of th e tro p h ic system and m echanical loading d u rin g prey
in ta k e are th o ro u g h ly studied. The follow ing sum m ary provides a clear overview
of th e m ain results and conclusions of each c h ap ter an d p resen ts answ ers to th e
research questions as fo rm u lated in c h ap ter 1.3.2.
In th e first ch ap te r som e general concepts like ‘ev o lu tio n ’, ‘n a tu ra l selectio n ’ and
‘specialization’ are illustrated. A n in tro d u c tio n to th e biology of Syngnathidae
follows, in w h ich th e ir taxonom ical p o sitio n , rep ro d u ctio n , com plex p a re n tal
care and external m orphology are discussed briefly. N ext, th e scien tific p ro b lem
of th e study is described. The feeding ap p aratu s is extrem ely specialized in this
family. All representatives have a tu b u la r sn o u t w ith small, te rm in a l jaws. Yet, th e
le n g th and diam eter of th e sn o u t is variable am ong species. Prey is cap tu red by
i 6o
C h a p te r 7 - S u m m ary & S am en v attin g
m eans of head ro ta tio n , w h ich p o sitio n s th e m o u th close to th e prey, follow ed by
an increase of th e buccal cavity volum e. This expansion generates a sub-am bient
pressure an d hence a w a ter flow tow ards th e m o u th is created. This type of
feeding is k n o w n as pivot feeding. Prey in tak e is very fast (prey cap tu re tim es are
less th a n 6 ms), how ever, su ctio n feeding by m eans of a long an d slender sn o u t
also im plies hydrodynam ic c o n strain ts (such as m ore drag due to su ctio n th ro u g h
a n arro w tu b e an d a h igh m o m en t of in e rtia caused by ro ta tio n of a h ead w ith a
long snout). This paradox renders th e tro p h ic ap p aratu s in Syngnathidae a very
in te re stin g research topic. T he objective of th is d o cto ral research is to unravel th e
evolutionary
p a tte rn
b e h in d
th e
extrem e
m orphological
an d
fu n c tio n a l
specialization.
C h ap ter tw o provides an overview of th e stu d ied m aterial an d th e applied
m ethods. First, a list of th e chosen species is given w ith referen ce to th e origin of
th e specim ens an d w h a t th ey w ere used for. In a second p art, th e tech n iq u es th a t
are used th ro u g h o u t th e research are explained in depth. D etailed p rotocols for
th e clearing an d stain in g of th e specim ens and fo r th e histological sectio n in g are
p resented, as w ell as th e applied m ethods fo r m easuring, dissecting, C T -scanning
an d graphical 3D -reconstructioning. The geom etric m o rp h o m etric an d fin ite
elem en t analyses are elaborated o n a little m ore. The ch ap te r closes w ith a
clarificatio n of various controversial term s.
T he th ird ch ap te r deals w ith th e m orphology of th e feeding ap p aratu s in
Syngnathidae and is divided in tw o parts.
In th e first p art, th e stru c tu re of th e cran iu m of a p ip efish (Syngnathus rostellatus)
an d th a t of a seahorse (Hippocampus reidi) is placed in to an ev o lu tio n ary co n tex t
by com parison w ith th e head m orphology of a stickleback (Gasterosteus aculeatus).
In addition, m orphology of th e juvenile h ead a ro u n d th e tim e of release fro m th e
p o u ch is described. E longation of th e sn o u t in th e cartilaginous skull is situ a ted
a t th e level of th e eth m o id region, w ith th e e th m o id plate, th e hyosym plectic
cartilage and th e basihyal cartilage being extended. W ith ex ception of th e
circu m o rb ital an d som e opercular bones, all cranial bones are already form ed in
th e juveniles, alb eit solely as a th in layer of bone. This close resem blance w ith th e
ad u lt head is probably associated w ith th e ex ten d ed bro o d p erio d an d p a re n tal
C h a p te r 7 - Sum m ary & sam en v attin g
161
care. D ue to this, em ergence of th e young from th e p o u ch can be delayed u n til an
advanced developm ental stage is reached. A p art fro m th a t, th e results show th a t
th e ad u lt sn o u t is m ade up by th e elongated vom eral, m esethm oid, q uadrate,
m etapterygoid, sym plectic an d p reo p ercu lar bones. Some m orphological featu res
th a t can be regarded as evolutio n ary ad ap tatio n s to th e specialized feeding type
are th e reduced m axillary and to o th less prem axillary bones, th e firm ly co n n ected
bones of th e low er jaw, th e ro b u st o p ercu lu m an d th e saddle-shaped jo in t
b etw een th e in terh y al an d ceratohyal p o sterio r bone.
T he second p a rt of th is ch ap ter focuses o n th e effect of sn o u t elo n g atio n w ith in
th e family, b o th on an osteological an d a m uscular level. A n exhaustive
m orphological study of th e head of b o th a lo n g -sn o u ted seahorse (H. reidi) an d a
p ipefish w ith
an extrem ely elongate sn o u t (Doryrhamphus dactyliophorus),
com pared to several syngnathids w ith b o th sh o rt and in term ed iate sn o u t lengths
(H. zosterae, H. abdominalis, Corythoichthys intestinalis and D. melanopleura),
confirm ed th e previous results. Also, a few specialized m uscle, te n d o n and
ligam ent con fig u ratio n s are discovered, such as th e sternohyoideus m uscle, w h ich
is very w ell developed in seahorses, th e p ro tra c to r hyoidei m uscle th a t is enclosed
by th e m andibulo-hyoid lig am en t in som e syngnathids an d th e long, rod-shaped
sesam oid bone in th e epaxial ten d o n s of several pipefishes. These observations
seem to reveal d ifferen t pow er g en eratin g strategies d u rin g pivot feeding,
favouring storage an d release of elastic energy in th e epaxial and hypaxial
ten d o n s over p lain m uscle force in pipefishes and vice versa in seahorses.
A lth o u g h th e syngnathid m usculoskeletal stru c tu re consists of som e elem ents
shared by all representatives, th e re are also n o tab le m orphological differences
am ong species. H ow ever, only a few of these differences could be related to
variatio n in relative sn o u t length.
T he m orphological v ariatio n w ith in th e fam ily is q u a n tifie d in ch ap te r fo u r by
m eans of a geom etric m o rp h o m etric analysis carried o u t on a large n u m b er of
sy ngnathid representatives. T h ro u g h th is m ethod, it is established th a t th e m ost
im p o rta n t differences betw een th e p ip efish an d seahorse head m orphology are
related to changes in sn o u t and head dim ension, p o sitio n of th e p ecto ral girdle
an d h e ig h t of th e opercular bone. This v ariatio n is likely associated w ith th e
head-body axis (body in line or at an angle w ith th e head) an d w ith th e feeding
1Ö2
C h a p te r 7 - S u m m ary & S am en v attin g
k inem atics (long sn o u t w ith little head ro ta tio n or sh o rt sn o u t w ith large
rotation). D ue to th e hig h degree of specialization in th e feeding ap p aratu s of
Syngnathidae, a sm all deviation of th e m ovem ents of or in te ra ctio n s b etw een th e
elem ents in th e head, m ight resu lt in reduced fu n c tio n a l perform ance. This leads
to th e hypothesis th a t each species of th e fam ily is ch aracterized by lim ited
intraspecific m orphological variation. M orphological v ariatio n w ith in species
w ith a long sn o u t, w h ich are regarded as being m ore specialized com pared to
sh o rt-sn o u te d species, w ill be dim inished. The p relim in ary results of th is study
co n firm o u r expectations. Finally, it w as investigated w h e th e r juvenile seahorses
(H. reidi) still undergo im p o rta n t o n to g en etic tra n sfo rm a tio n s a fte r leaving th e
pouch. G radual shifts in shape, like th e dorsoventral n arro w in g of th e sn o u t and
head and th e re o rie n ta tio n of th e p reo p ercu lar bone, characterize juvenile
developm ent. H ow ever, th e period of m ajor developm ental changes seems to have
ta k e n place in th e pouch.
In th e fifth chapter, som e fu n c tio n a l aspects of th e specialized feeding ap p aratu s
are addressed.
T he first p a rt of th is research was carried o u t in co llab o ratio n w ith G ert Roos
an d Sam van W assenbergh (U niversity of A ntw erp) an d it evaluates a previously
described p lan ar fo u r-b ar m echanism (M uller, 1987) th a t couples hyoid ro ta tio n
to n eu ro cran ial elevation. B oth th e m orphological d escrip tio n of all linkages and
join ts involved in th e system , as th e k in em atical recordings of a feeding strik e in
a seahorse (H. reidi), d em onstrate th e incredible specializations of th e tro p h ical
system. N o t only does th e syngnath id pivot feeding differs fro m su ctio n feeding
in a general te le o st in term s of velocity (prey cap tu re takes less th a n 6 ms), b u t
also in sequence of m ovem ents; th e expansion phase starts w ith hyoid ro ta tio n
an d n o t w ith m o u th opening as is usually th e case (i.e. ro stro cau d al wave). Besides
th a t, th e results provide clear p ro o f fo r th e coupling b e tw ee n hyoid and
neurocranium . H ow ever, a discrepancy is discovered b etw een th e p red icted
m ovem ents (based on four-bar m odelling) an d th e observed m ovem ents (based on
video recordings). This is m ost likely due to m odelling th e b ar th a t consists of th e
urohyal bone to g e th e r w ith th e stern o h y o id eu s m uscle, as co n sta n t in length,
w hereas it sh ortens due to m uscle co n tractio n .
C h a p te r 7 - Sum m ary & sam en v attin g
163
T he second p a rt of th is ch ap te r explores w h ere in th e skull m ost m echanical
stress is exerted as a consequence of th e large pressure differences d u rin g feeding.
This w as done by m eans of a fin ite elem en t analysis o n th e m odelled skulls of
th re e seahorse and th re e p ipefish species w ith various relative sn o u t le n g th (H.
reidi, H.
abdominalis, H. zosterae, D.
dactyliophorus,
S. rostellatus an d D.
melanopleura). In accordance w ith o u r expectations, in all m odels (except fo r D.
dactyliophorus) peak stress accum ulations w h ere observed a t th e level of
a rticu latio n s
an d
cartilaginous
regions
in
th e
m odelled
crania.
These
c o n ce n tra tio n s of stress w ill be m u ch lo w er in th e real skulls since b o th
cartilaginous tissue an d artic u la tin g elem ents are th o u g h t to resist com pressive
stress well. A second aim w as to analyze w h e th e r or n o t a difference in stress
d istrib u tio n p a tte rn
exists betw een sh o rt- an d lo n g -sn o u ted species. The
p re d ic tio n w as th a t species w ith a long sn o u t need to g en erate larger pressure
differences, and hence should be able to w ith sta n d h ig h er am o u n ts of stress
th ro u g h c ertain m orphological ad ap tatio n s. W h e n applying th e sam e pressure to
b o th short- and lo n g-snouted m odels, th e m odels w ith a long sn o u t are th o u g h t
to experience less stress. This hypothesis could n o t be co n firm ed by th e results,
how ever, th e geom etry of th e bones in th e sn o u t (bone thickness, a m o u n t of
overlap b etw een d ifferen t bones, etc) has a clear in flu en ce o n th e stress
d istrib u tio n .
In c h ap ter six, th e o b tain ed results are discussed in an o n to g en etic, fu n c tio n a l
an d evolutionary context, follow ed by a general conclusion.
First, th e rem arkable developm ent of Syngnathidae is dealt w ith. The p aren tal
care is long an d extensive; th e m ale attaches th e eggs and em bryos o n to its body
or keeps th e m in a sealed pouch, even a fte r hatching. The degree of p a re n tal
in v estm en t increases w ith increasing com plexity of th e b ro o d area. Even th e m ost
basal syngnathids p ro te c t th e young against p re d a tio n and som e m ore derived
species osm oregulate, n o u rish an d aerate th e young as well. D ue to th e
p ro lo n g a tio n of th e b ro o d period, o n set of exogenous feeding is delayed and m ore
tim e for developm ent of th e young is available. H ence, new ly b o rn syngnathids
are alm ost fully developed (the larval phase and re lated m etam orphosis are ab sen t
or occur in th e pouch) an d juveniles are able to cap tu re prey by m eans of pivot
164
C h a p te r 7 - S u m m ary & S am en v attin g
feeding. In sum , it can be stated th a t th e far-reaching p a re n tal care pro b ab ly has
aided in th e ev olution of th e specialized syn g n ath id h ead m orphology.
T he second p a rt of th e discussion h ighlights th e fu n c tio n in g of th e syngnathid
feeding system. T he m ost im p o rta n t stru c tu ra l an d fu n c tio n a l m odifications in
com parison w ith th e tro p h ic ap p aratu s of a g eneralized te le o st are repeated.
N ext, a d etailed descrip tio n of th e m ovem ents an d activity of all separate
elem ents in th e head d u ring feeding in tak e is given. Finally, th e advantages and
disadvantages related to sn o u t elo n g atio n are listed. From th is it w as deduced
th a t th e m orphological v ariatio n associated w ith sn o u t dim ension p re sen t in th e
fam ily, is probably th e resu lt of a trade-off. To optim ize h ead ro ta tio n
perform ance, a long sn o u t is favoured, w hereas a sh o rt sn o u t is beneficial for
im proving su ctio n capacity. This m ig h t a cc o u n t fo r th e large m orphological
diversity seen w ith in Syngnathidae.
P a rt th re e considers th e results in a p h y logenetic and ev o lu tio n ary fram ew ork.
First, argum ents for com parison of th e sy n g n ath id head w ith th e head of th e
th ree-sp in ed stickleback (Gasterosteus aculeatus), even th o u g h th e y m ig h t n o t be
sister taxa, are provided. T hen, th e ev o lu tio n ary tre n d s th a t m ig h t have led to th e
seahorse m orphology (i.e. th e preh en sile tail, tilte d h ead an d v ertical body
posture) are described. T he genus Idiotropiscis (pygmy pipehorses) is regarded as
th e ex tan t tra n s itio n from pipefishes to seahorses. D ivergence b etw een seahorses
an d pipehorses is th o u g h t to have b e n efitte d fro m th e increase of h ab itats
do m in ated by seagrasses, w h ich provide b e tte r p ro te c tio n to organism s w ith an
u p rig h t body axis. In addition, biom echanics of pivot feeding m ig h t have played a
role in th e evolution tow ards seahorses. The tilte d head-body axis likely p ro m o tes
head ro ta tio n and th e grasping ta il allow s stre tch in g of th e body in th e d irectio n
of th e prey. The a n terio r directed m ovem ent of th e m o u th reduces th e tim e
needed to reach th e prey. H ence, th e selective advantage of sn o u t elongation,
w h ich is th o u g h t to sim ilarly m inim ize tim e to reach prey, is reduced in
seahorses. The absence of extrem ely lo n g -sn o u ted seahorses could th u s be
explained.
Finally, these findings are concisely recap itu lated in th e general conclusion.
C h a p te r 7 - Sum m ary & sam en v attin g
7.2.
D it
165
Sam envatting
d o cto raatsp ro efsch rift
b eh an d elt
de
evolutie
van
de
gespecialiseerde
kopm orfologie in S yngnathidae, een fam ilie b in n e n de beenvissen w aartoe
zeenaaiden, zeepaardjes, zeedraken en ‘pipehorses’ behoren. E en gedetailleerde
m orfologische beschrijving van h e t voedselopnam eapparaat bij enkele syngnathe
vertegenw oordigers vo rm t de k e rn van d it onderzoek. D aarnaast w o rd en andere
aspecten zoals m orfologische variatie b in n e n de fam ilie, w erk in g van h e t tro fisc h
systeem
en
m echanische
belastin g
tijdens
de
voedselopnam e
uitvoerig
bestudeerd. De volgende sam envatting geeft een duidelijk o v erzich t van de
belangrijkste re su ltaten en conclusies p er h o o fd stu k en b ie d t een an tw o o rd op de
onderzoeksvragen zoals geform uleerd in h o o fd stu k 1.3.2.
In h e t eerste h o o fd stu k w o rd en enkele algem ene b eg rip p en zoals ‘evolutie’,
‘n atu u rlijk e selectie’ en ‘specialisatie’ toegelicht. D aarop volgt een inleiding to t de
biologie van Syngnathidae, w aarbij de taxonom ische positie, v o o rtp lan tin g ,
com plexe broedzorg en uitw endige m orfologie aan b od kom en. V ervolgens w o rd t
de w etenschappelijke probleem stellin g geschetst. H e t voedselopnam e-appararaat
is in deze fam ilie extreem gespecialiseerd. Alle vertegenw oordigers h eb b en een
buisvorm ige sn u it m et kleine, eindstandige kaken. De lengte en d iam eter van de
sn u it is ech ter variabel over verschillende soorten. V oedselopnam e (‘pivot
feeding’) g eb eu rt door m iddel van k o p ro ta tie w aard o o r de s n u ittip n aar de p ro oi
geb rach t w o rd t, gevolgd door een volum etoenam e in de buccale holte. Deze
expansie creëert een o n d e rd ru k w aard o o r een w aterstro o m in de ric h tin g van de
m ond o n tstaat. H e t vangen van een p ro o i g e b eu rt heel snel (m inder dan 6 ms),
m aar voedselopnam e m et b eh u lp van een lange, sm alle sn u it h o u d t ook een aan tal
hydrodynam ische b ep erk in g en in (zoals een hoge w e ersta n d bij h e t zuigen van
w a te r door een d u n n e buis en een g ro o t in e rtie m o m e n t tijd en s ro ta tie van een
kop m et lange snuit). D eze paradox m aak t h e t voedselopnam e-apparaat bij
Sygnathidae to t een zeer in te re ssa n t onderzoeksobject. De doelstelling van d it
d o cto raatso n d erzo ek is dan ook h e t o n tra fe le n van h e t ev o lu tio n aire p a tro o n dat
aan de basis ligt van deze extrem e m orfologische en fu n ctio n ele specialisatie.
i66
C h a p te r 7 - S u m m ary & S am en v attin g
H o o fd stu k tw ee geeft een overzicht van h e t bestu d eerd e m ateriaal en de
toegepaste m ethodes. E erst w o rd t een opsom m ing gegeven van de gekozen
so o rte n m et verm elding van de h erk o m st van de specim ens en w aarvoor ze
g eb ru ik t w erden. In een tw eede deel w o rd en de te ch n iek e n die d o o rh een h e t
onderzoek toegepast w erden, uitg eb reid uitgelegd. G edetailleerde p rotocols voor
h e t o p helderen en k leu ren van specim ens en voor h e t m aken van histologische
coupes w o rd en gegeven, alsook de g eh an teerd e m eth o d e bij h e t m eten, u itv o eren
van dissecties, nem en van CT-scans en m aken van grafische 3D -reconstructies.
Iets m eer aan d ach t w o rd t besteed aan de geom etrisch m orfo m etrisch e en de
eindige e lem en ten (finite elem ent) analyses. T en slo tte volgt in de laatste
paragraaf een v erduidelijking van een a an tal controversiële begrippen.
H e t derde h o o fd stu k h an d elt over de m orfologie van h e t voedselopnam eap p araat bij Syngnathidae en is onderverdeeld in tw ee delen.
In h e t eerste deel w o rd t de bo u w van h e t cran iu m van een zeenaald (Syngnathus
rostellatus) en van een zeepaard ('Hippocampus capensis) in een evolutieve co n tex t
g eplaatst
door
vergelijking
m et
de
kopm orfologie
van
een
stekelbaars
(Gasterosteus aculeatus). O ok de m orfologie van de juveniele k op ro n d h e t tijd stip
van vrijkom en u it de buidel, k o m t aan bod. De sn u itv erlen g in g van de
kraakbenige schedel situ e e rt zich te r h o o g te van de eth m o id regio en de
eth m oidplaat, h e t hyosym plecticum en h e t basihyale zijn verlengd. In de
juvenielen zijn alle craniale beenderen, m et u itzo n d erin g van de circu m o rb italia
en enkele operculaire beenderen, al gevorm d, zij h e t enkel ais een zeer d u n laagje.
D eze grote gelijkenis m et de adulte kop is w aarschijnlijk te w ijte n aan de lange
broedperiode en uitgebreide ouderlijke zorg. H ierd o o r k a n h e t v rijlaten van de
jongen
u it
de
buidel
uitg esteld
w o rd e n
to t
ze
een
vergevorderd
ontw ik k elin g sstad iu m b e re ik t hebben. V erder to n e n de re su lta te n aan d at de
ad ulte sn u it gevorm d w o rd t door een verlenging van h e t vom er, m esethm oid,
quadratum ,
m etapterygoid,
sym plecticum
en
preoperculare.
Enkele
m orfologische asp ecten die ais ev o lu tio n aire ad ap taties aan de gespecialiseerde
voedselopnam e k u n n e n beschouw d w o rd en , zijn de gereduceerde en tan d lo ze
m axillaire beenderen, de stevige verb in d in g tu sse n de verschillende b een d eren
van de onderkaak, h e t ro b u u ste kieuw deksel en h e t zadelvorm ig g ew rich t tu ssen
interh y ale en ceratohyale posterior.
C h a p te r 7 - Sum m ary & sam en v attin g
167
In h e t tw eede deel van d it h o o fd stu k ligt de n ad ru k op h e t effect van
snuitverlenging b in n e n de fam ilie, zow el op osteologisch ais m yologisch vlak.
E en gedetailleerde m orfologische studie van de kop van zow el een lang-snuitige
zeepaard (H. reidi) ais een zeenaald m et extreem lange sn u it (Doryrhamphys
dactyliophorus), in vergelijking m et een aan tal sy n g n ath en m et k o rte en
in term ed iaire sn u it (H. zosterae, H. abdominalis, Corythoichthys intestinalis en D.
melanopleura), bevestigt de re su ltaten u it h e t voorgaande deel. B ovendien w o rd en
enkele gespecialiseerde spier-, pees- en lig am en tco n fig u raties aan h e t lich t
gebracht. Zo is de sternohyoideus spier heel goed o n tw ik k eld bij zeepaardjes,
w o rd t de p ro tra c to r hyoidei spier bij enkele sy n g n ath en om geven door h e t
m andibulo-hyoid lig am en t en w o rd t in de epaxiale p ezen van een aan tal
zeenaaiden een lang, staafvorm ig sesam oid b een gevonden. D eze w aarn em in g en
suggereren de aanw ezigheid van verschillende verm ogen v ersterk en d e strateg ieën
voor k o p ro ta tie tijdens de voedselopnam e. H e t v rijlaten van elastische energie,
opgeslagen in de epaxiale en hypaxiale pezen, lijk t beg u n stig d in zeenaaiden
terw ijl sp ierk rach t verkozen w o rd t in zeepaardjes. H oew el h e t syngnathe
m usculoskeletale systeem u it een aan tal e lem en ten b estaat die gedeeld w o rd en
door alle vertegenw oordigers, zijn
er dus to c h
een
aan tal opm erkelijke
m orfologische verschillen tu sse n soorten. Slechts w einig van deze verschillen
k u n n e n ech ter gerelateerd w o rd en aan variatie in relatieve snuitlengte.
A an slu iten d bij h e t voorgaande, w o rd t in h o o fd stu k vier de m orfologische
variatie b in n e n de fam ilie g ek w an tificeerd door m iddel van een geom etrisch
m orfom etrische analyse bij een g ro o t aan tal syngnathe vertegenw oordigers. Via
deze m ethode k an vastgesteld w o rd en d at de belangrijkste verschillen in
kopm orfologie
tu sse n
zeenaaiden
en
zeepaardjes
gerelaterd
zijn
aan
v eran d erin g en in sn u it- en kopdim ensie, de positie van de p ecto rale vin en de
hoogte van h e t kieuw deksel. D eze variatie h e e ft w aarschijnlijk te m aken m et de
o rië n ta tie van de lichaam sas te n o p zich te van de k op (lichaam in h e t verlengde of
o n d er een hoek m et de kop) en aan de k in em atica van de voedselopnam e (lange
sn u it en b ep erk te k o p ro ta tie of k o rte sn u it en veel rotatie). G ezien de hoge graad
van specialisatie in h e t voedselopnam e-apparaat bij Syngnathidae zo u een kleine
afw ijking in de in te ra ctie tu ssen de verschillende elem en ten in de kop al to t een
verm inderde p restatie k u n n e n leiden. D it doet verm oeden dat elke so o rt b in n e n
i6 8
C h a p te r 7 - S u m m ary & S am en v attin g
de fam ilie g ek en m erk t w o rd t door een b ep erk te in trasp ecifiek e m orfologische
variatie. V ooral de m orfologische p la sticiteit van so o rten m et een lange sn u it, die
ais m eer gespecialiseerd beschouw d w o rd en dan k o rt-sn u itig e soorten, zal
gereduceerd zijn. De prelim inaire re su ltaten van deze stu d ie bevestigen onze
verw achtingen. T en slo tte w o rd t o n d erzo ch t of er nog b elangrijke o n to g en etisch e
tran sfo rm aties o p tre d e n bij juveniele zeepaardjes (H. reidi) n a h e t vrijkom en u it
de buidel. G raduele vorm verschuivingen, zoals h e t dorsoventraal v ern au w en van
de sn u it en kop en de h e ro rië n ta tie van h e t p reo p ercu lair been, k en m erk en de
juveniele ontw ikkeling. M aar, de periode van ingrijpende veran d erin g en lijk t
reeds doorlopen te zijn in de buidel.
In
het
vijfde
h o o fd stu k
kom en
enkele
fu n c tio n e le
aspecten
van
het
gespecialiseerde voedselopnam e-apparaat aan bod.
H e t eerste deelonderzoek w erd uitgevoerd in sam enw erking m et G ert Roos en
Sam van W assenbergh (U niversiteit A ntw erpen) en evalueert een eerder
beschreven p lan air vierstangensysteem (M uller, 1987) d at ro ta tie van h e t hyoid
k o p p e lt aan n eu ro cran iale elevatie. Z ow el de m orfologische beschrijving van alle
stangen en g ew rich ten b e tro k k e n bij h e t systeem , als de k in em atisch e opnam es
van de voedselopnam e bij een zeepaardje (H. reidi) to n e n de ongelooflijke
specialisatie van h e t tro fisc h ap p araat aan. N ie t alleen v ersch ilt de syngnathe
‘pivot feeding’ van zuigvoeding bij een gegeneralizeerde teleo st in snelheid (duur
van een prooiopnam e is m inder dan 6 ms), m aar ook de volgorde van bew eging is
anders. De expansiefase b eg in t m et h y o id ro tatie en n ie t m et m o n d o p en in g zoals
gebruikelijk (rostrocaudale golf). B ovendien leveren de re su ltaten een duidelijk
bew ijs voor de koppeling tu sse n hyoid en n eu ro cran iu m . Er w o rd t ech ter een
discrepantie
tu ssen
de
voorspelde
bew egingen
(op
basis
van
het
vierstangenm odel) en de w aargenom en bew egingen (op basis van de videoopnam es) aangetoond. D it is hoogst w aarschijnlijk te w ijte n aan h e t m odelleren
van de stang bestaande u it h e t urohyale en de stern o h y o id eu s spier ais
onveranderlijk in lengte, terw ijl deze v e rk o rt bij sp ierco n tractie.
In h e t tw eede deel w o rd t o n d erzo ch t w aar in de schedel de m eeste m echanische
stress
o p tre e d t
ais
gevolg
van
de
g ro te
d ru k v eran d erin g en
tijdens
de
voedselopnam e. H iervoor w o rd t een eindige elem en ten analyse uitgevoerd op de
gem odelleerde
schedels van
drie
zeepaard-
en
drie
zeen aald so o rten
m et
C h a p te r 7 - Sum m ary & sam en v attin g
v ariërende
relatieve
169
sn u itlen g te
(H . reidi, H.
abdominalis, H. zosterae, D.
dactyliophorus, S. rostellatus en D. melanopleura). In overeenstem m ing m et de
verw ach tin g en
w aargenom en
(uitgezonderd
te r
hoogte
in
van
D.
dactyliophorus)
articu laties
en
w o rd en
kraakbenige
stresspieken
zones
in
de
gem odelleerde schedels. D eze co n cen traties aan stress zu llen in de w erkelijke
schedels veel lager liggen aangezien zow el k raak b een ais articu leren d e elem en ten
vero n d ersteld w o rd en stress goed te w eerstaan. Een tw eede doei is nagaan of er
een verschil is in stre ssd istrib u tie p a tro o n tu sse n k o rt- en langsnuitige soorten.
De hypothese is d at so o rten m et een lange sn u it g ro tere drukverschillen m o eten
creeëren, hogere stress m oeten k u n n e n w eerstaan en dus bepaalde m orfologische
adaptaties zu llen h eb b en die d it to elaten . W a n n e e r k o rt- en langsnuitige
m odellen blootg esteld w o rd en aan eenzelfde druk, zu llen de m odellen m et een
lange sn u it de m inste stress ervaren. D eze h y p o th ese w o rd t n ie t bevestigd door de
resu ltaten , m aar de geom etrie van de b een d eren in de sn u it (dikte van h e t been,
m ate van overlap tu ssen de been d eren , enz.) h e e ft een duidelijke invloed op de
stress d istributie.
In
h o o fd stu k
zes w o rd en
de behaalde
re su ltaten
bediscussieerd
in
een
o ntogenetische, fu n ctio n ele en ev o lu tio n aire context, gevolgd door een algem ene
conclusie.
T en eerste k o m t de bijzondere o n tw ik k elin g bij Syngnathidae aan bod. De
broedzorg d u u rt lang en is complex; h e t m an n etje h e ch t de eitjes en em bryos aan
h e t lichaam vast of b ew aart ze in een afgesloten buidel, zelfs n a d at ze u it h e t ei
gekom en zijn. De m ate van ouderlijke in v esterin g n eem t to e m et to en em en d e
com plexiteit van de b ro e d stru c tu u r. Zelfs de m eest basale sy n g n ath en b ieden de
jongen bescherm ing teg en predatie, m eer geëvolueerde so o rte n zorgen ook voor
osm oregulatie,
voeding
en
zuu rsto fto ev o er.
D oor
de
verlenging
van
de
broedperiode w o rd t de sta rt van exogene voedselopnam e u itg esteld en is er m eer
tijd voor de o n tw ik k elin g van de jongen. Pasgeboren sy n g n ath en zijn dan ook
bijna volledig o n tw ik k eld (de larvale fase en b ijh o ren d e m etam orfose zijn afw ezig
of v inden plaats in de buidel) en ju v en ielen k u n n e n zelfstandig p ro o ien vangen
door m iddel van de speciale ‘pivot feeding’. Sam envattend k an gesteld w o rd en dat
de ingrijpende broedzorg w aarschijnlijk de evolutie van de gespecialiseerde
syngnathe kopm orfologie in de h a n d h e eft gew erkt.
C h a p te r 7 - S u m m ary & S am en v attin g
In h e t tw eede deel van de discussie w o rd t de n a d ru k gelegd op de w erk in g van de
syngnathe kop. De belangrijkste stru c tu re le en fu n ctio n ele m odificaties in
vergelijking m et h e t tro fisch e ap p araat van een gegeneraliseerde te le o st w o rd en
herhaald. D aarna volgt een gedetailleerde beschrijving van de bew eging van en
in te ra ctie tu sse n de afzonderlijke elem en ten in de kop tijd en s de voedselopnam e.
T en slo tte w o rd en de voor- en nadelen gerelateerd aan snuitv erlen g in g opgesom d.
H ie ru it b lijk t d at de m orfologische variatie in te rm e n van sn u itd im en sie in de
fam ilie w aarschijnlijk h e t gevolg is van een trade-off. V oor de o p tim alisatie van
k o p ro ta tie is een lange sn u it voordelig, terw ijl voor h e t g en ereren van een grote
zu ig k rach t een k o rte sn u it g unstig er is. De grote m orfologische d iv ersiteit b in n e n
de Syngnathidae zo u zo verklaard k u n n e n w orden.
H e t derde deel b eh an d elt de re su ltaten in een fylogenetisch en ev o lu tio n air
kader. Er w o rd t eerst k o rt b earg u m en teerd w aarom vergelijking van de syngnathe
kop m et die van de driedoornige stekelbaars (Gasterosteus aculeatus) nog steeds
n u ttig is, ook ais ze n ie t eikaars zu stertax a b lijken te zijn. V ervolgens w o rd en de
evolutionaire
tre n d s besp ro k en
die to t
de zeepaardm orfologie
(d.w.z. de
grijpstaart, de gekantelde kop en de verticale lichaam shouding) zo u d en h eb b en
k u n n e n leiden. H e t genus Idiotropiscis (‘pygm y p ip eh o rses3) w o rd t b eschouw d ais
de nog levende tran sitiev o rm tu sse n zeenaaiden en zeepaardjes. De divergentie
van pipehorses en zeepaarden zo u bevorderd zijn
door de opm ars van
zeegrashabitats die m eer bescherm ing bied en aan organism en m et een rechte
lichaam sas. O ok de biom echanica van de voedselopnam e zo u een rol gespeeld
k u n n e n h eb b en in de evolutie van zeepaardjes. De k a n telin g van de kop te n
opzichte van de lichaam sas zo u k o p ro ta tie k u n n e n bevoordelen en de g rijp staart
la at een strek k in g van h e t lichaam in de ric h tin g van de p ro o i toe. Deze
voorw aartse bew eging van de m o n d zo u de tijd die nodig is om de p ro o i te
b ereik en reduceren, w a t precies h e t effect van sn u itv erlen g in g is. Ais een grote
k o p ro ta tie in com binatie m et een voorw aartse bew eging in de ric h tin g van de
p ro o i h etzelfd e selectief voordeel b ie d t ais een lange sn u it, zo u d at een verklaring
k u n n e n zijn voor h e t o n tb re k e n van zeepaardjes m et u itzo n d erlijk lange snuit.
T en slo tte w o rd en deze bevinding en k o rt en bondig sam engevat in de algem ene
conclusie.
C h a p t e r 8 - R e fe re n c e s
171
8
ADAMS DC, ROHLF FJ, Slice DE. 2004. G eom etric m orphom etries: Ten years of
progress follow ing th e ‘rev o lu tio n ’. Ita lia n Jo u rn al of Zoology 71:5-16.
ADRIAENS D, H errel A. 2009. F u n ctio n al consequences of extrem e m orphologies
in th e craniate tro p h ic system . Physiological an d B iochem ical Zoology 82:1-6.
ADRIAENS D, V erraes W . 1998. O n to g en y of th e o steo cran iu m in th e A frican
catfish, Clarias gariepinus (Burchell, 1822) (Siluriform es: Clariidae):
O ssification sequence as a response to fu n c tio n a l dem ands. Jo u rn al of
M orphology 235:183-237.
A erts P. 1991. H yoid m orphology an d m ovem ents relative to abducing forces
d u ring feeding in A stato tilap ia elegans (Teleostei: Cichlidae). Jo u rn al of
M orphology 208:323-345.
AERTS P. 1998. V ertical jum ping in Galago senegalensis: The q u est fo r an obligate
m echanical pow er am plifier. P hilosophical T ran sactio n s of th e Royal Society
of London, Biological Sciences 353:1607-1620.
AERTS P, V erraes W . 1984. T h eo retical analysis of a p lan ar fo u r-b ar linkage in the
teleo stean skull: The use of m athem atics in biom echanics. A nnales de la
Société Royal Z oologique de Belgique 114:273-290.
AERTS P, V erraes W . 1987. Do inertial effects explain the maxim al rotation of the
maxilla in th e rainbow tro u t (Salmo gairdneri) during feeding? Annales de la
Société royale zoologique de Belgique 117:221-235.
ALEXANDER R M cN . 1966. T he fu n ctio n s and m echanism s of th e p ro tru sib le
u p p er jaws of tw o species of cyprinid fish. Jo u rn al of Zoology (London)
149:288-296.
ALEXANDER R M cN . 1967a. The fu n ctio n s and m echanism s of th e p ro tru sib le
172
C h a p t e r 8 - R e f e re n c e s
u p p er jaws of som e acan th o p tery g ian fish. Jo u rn al of Zoology (London)
151:43-64.
ALEXANDER R M cN . 1967b. M echanism s of th e jaws of som e ath erin ifo rm fish.
Jo u rn al of Zoology (London) 151:233-255.
ALEXANDER R M cN . 1967c. F u n c tio n a l design in fishes. H u tc h in so n U niversity
Library, London, p. 160.
ALEXANDER R M cN . 1969. M echanics of th e feeding actio n of a cyprinid fish.
Jo u rn al of Zoology (London) 159:1-15.
ALEXANDER R M cN , Ben n et -C lark HC. 1977. Storage of elastic stra in energy in
m uscle and o th e r tissues. N a tu re 265:114-117.
ANKER G C h. 1974. M orphology an d k in etics of th e h ead of th e stickleback,
Gasterosteus aculeatus. T ransactio n s of th e Zoological Society of L ondon
32:311-416.
ANKER G C h. 1989. The m orphology of jo in ts and ligam ents of a generalized
Halpochromis species: H. elegans Traw avas 1933 (Teleostei, Cichlidae). III. The
hyoid an d th e branchiostegal apparatus, th e b ran ch ial ap p aratu s and th e
shoulder girdle apparatus. N eth erlan d s Journal of Zoology 39:1-40.
A rratia G, SCHULTZE H P. 1990. The urohyal: D evelopm ent and hom ology w ith in
osteichthyans. Jo u rn al of M orphology 203:247-282.
ARRATIA G, SCHULTZE H P. 1991. P alato q u ad rate an d its ossifications:
D evelopm ent and hom ology w ith in osteichthyans. Jo u rn al of M orphology
208:1-81.
A zzarello MY. 1990. A com parative study of the developm ental osteology of
Syngnathus scovelli and Hippocampus zosterae (Pisces: Syngnathidae) and its
phylogenetic implications. Evolutionary M onographs 12:1-90.
BALON EK. 1975. T erm inology of intervals in fish developm ent. Jo u rn al of
Fisheries R esearch Board of C anada 32:1663-1670.
BALON EK. 1979. T he th e o ry of saltatio n an d its ap p licatio n in th e o n to g en y of
fishes: steps and thresholds. E n v iro n m en tal Biology of Fishes 4:97-101.
BALON EK. 1984. R eflections on som e decisive events in th e early life of fishes.
T ransactions of th e A m erican Fisheries Society 113:178-185.
BALON EK. 1986. Types of feeding in th e o n to g en y of fishes and th e lif e-history
model. E n v iro n m en tal Biology of Fishes 16:11-24.
Baylis JR. 1981. The evolution of parental care in fishes, w ith reference to
D arw in’s rule of male sexual selection. Environm ental Biology of Fishes
6:223-251.
Benkm an CW . 1987. Crossbill foraging behavior, bill structure, and patterns of
food profitability. The W ilson Bulletin 99:351-368.
C h a p t e r 8 - R e fe re n c e s
173
B e n k m a n C W . 1993. A d ap tatio n to single resources an d th e ev o lu tio n of crossbill
(Loxia) diversity. Ecological M o n o g rap h s 63:305-325.
BERGERI BA, W a in w rig h t PC. 1997. M orphology and kinem atics of prey capture
in th e syngnathid fishes Hippocampus erectus and Syngnathus floridae. M arine
Biology 127:563-570.
Bisho p KL, W a in w rig h t PC, H olzman R. 2008. A n terio r-to -p o sterio r w ave of
buccal expansion in su ctio n feeding fishes is critical fo r o p tim izin g flu id flow
velocity profile. Jo u rn al of th e Royal Society In terface 5:1309-1316.
Bloodw orth B, MARSHALL CD. 2005. Feeding kinem atics of Kogia and Tursiops
(Odontoceti: Cetacea): characterization of suction and ram feeding. The
Journal of Experim ental Biology 208:3721-3730.
BOCK W J. 2003. Ecological aspects of th e ev o lu tio n ary processes. Zoological
Science 20:279-289.
BOCK W J, Shear RC. 1972. A stain in g m eth o d fo r gross dissection of v erteb rate
m uscles. A natom ischer A nzeiger 130:222-227.
GM . 1966. C o n trib u tio n s to th e fu n c tio n a l m orphology of fishes. P art
III. T he feeding m echanism of Syngnathus acus Linnaeus. Zoologica A fricana
2:69-89.
BRANCH
BRITZ R, JOHNSON D. 2002. “Paradox lo st”: Skeletal o n to g en y of Indostomus
paradoxus an d its significance fo r th e p h ylogenetic relatio n sh ip s of th e
Indostom idae (Teleostei, G asterosteiform es). A m erican M u seu m N o vitates
3383:2-42.
BRUNER E, Bartolino V. 2008. M orphological v ariatio n in th e seahorse v erteb ral
system. In te rn a tio n a l Jo u rn al of M orphology 26:247-262.
C arcupino M, Baldacci A, M a zzin i M, P ranzoi P. 2002. Functional
significance of th e male brood pouch in th e reproductive strategies of
pipefishes and seahorses: A m orphological and ultrastructural comparative
study on three anatom ically different pouches. Journal of Fish Biology
61:1465-1480.
CARROLL AM , WAINWRIGHT PC. 2003. F u n c tio n a l m orphology of prey cap tu re in
th e sturgeon, Scaphirhynchus albus. Jo u rn al of M orp h o lo g y 256:270-284.
C arroll AM, W a in w rig h t PC, H uskey SH, C ollar DC, T uringan RG. 2004.
M orphology predicts suction feeding perform ance in centrarchid fishes. The
Journal of Experim ental Biology 207:3873-3881.
C asey SP, H all HJ, Stanley HF, V in c en t ACJ. 2004. The origin and evolution of
seahorses (genus Hippocampus ): A phylogenetic study using cytochrom e b
gene of m itochondrial DNA. M olecular Phylogenetics and Evolution 30:261272.
C h a p t e r 8 - R e f e re n c e s
174
C astro ALD, D in iz AD, M a rtin s IZ, V en d el AL, de O liveira TPR, R osa IM D .
2008. A ssessing diet com positio n of seahorses in th e w ild using a n o n
destructive m ethod: Hippocampus reidi (Teleostei: Syngnathidae) as a studycase. N eotropical Ichthyology 6:637-644.
CHEN W , B o n illo C, Le c o in tr e G. 2003. R epeatability of clades as a crite rio n of
reliability: A case study fo r m olecular phylogeny of A can th o m o rp h a
(Teleostei) w ith larger n u m b e r of taxa. M o lecu lar Phylogenetics and
E volution 26:262-288.
C
CK, L iew HC. 2006. M orphological dev elo p m en t an d allom etric g ro w th
p a tte rn s in juvenile seahorse. Jo u rn al of Fish Biology 69:426-445.
h o o
C o l s o n DJ, P a t e r SN, B r a in e r d EL, L ew is SM. 1998. Sound p ro d u c tio n d u ring
feeding in Hippocampus seahorses (Syngnathidae). E n v iro n m en tal Biology of
Fishes 51:221-229.
CoNSi TR, Seifert PA, T riantafyllou M S, E delman ER. 2001. T he dorsal fin
engine of th e seahorse (Hippocampus sp.). Jo u rn al of M orphology 248:80-97.
COORA. 1996. O n to g en y of feeding m orphology an d kinem atics in juvenile fishes:
A case study of th e c o ttid fish Clinocottus analis. T he Jo u rn al of E xperim ental
Biology 199:1961-1971.
C o ugh lin DJ. 1994. S uction prey cap tu re by clow nfish larvae (Amphiprion
perideraion). C opeia 1:242-246.
CURRBY JD. 1999. T he design of m ineralised h a rd tissues fo r th e ir m echanical
fu nctions. The Journal of E xperim ental Biology 202:3285-3294.
JD. 2010. M echanical p ro p erties and ad ap tatio n s of som e less fam iliar
b ony tissues. Jo u rn al of th e M echanical Behavior of Biom edical M aterials
3:357-372.
CURRBY
CURTIS JM R, V in c en t ACJ. 2005. D istrib u tio n of sym patric seahorse species
along a g rad ien t of h a b ita t com plexity in a seagrass-dom inated com m unity.
M arin e Ecology Progress Series 291:81-91.
D a get J. 1964. Le crâne des téléostéens. M ém oires d u M u séu m d’H isto ire
N a tu re lle Serie A. Zoologie 31:163-341.
DARWIN C. 1859. O n th e origin of species by m eans of n a tu ra l selection, or
preserv atio n of favored races in th e struggle fo r life. Jo h n M urray: London
(rep rin t e d itio n 1998: W o rd sw o rth E ditions Lim ited, H erfo rd sh ire. p. 392.
DAWSON CE. 1978. Syngnathus parvicarinatus a new A u stralian pipefish, w ith
no tes on S. sauvagei (W hitley) and Leptonotus caretta (K lunzinger). C opeia
2:288-293.
DAWSON CE. 1983. W este rn A tlan tic occurrence of th e genus M inyichthys (Pisces:
Syngnathidae), w ith a d escriptio n of M inusitatus n. sp. C opeia 3:774-776.
C h a p t e r 8 - R e fe re n c e s
175
DB Beer GR. 1937. T he developm ent of th e v erteb rate skull. U niversity Press,
O xford, p. 552.
DB L u s s a n e t M H E , MULLER M. 2007. T he sm aller y o u r m o u th , th e lo n g er your
snout: P redicting th e sn o u t le n g th of Syngnathus acus, Centriscus scutatus and
o th e r p ip e tte feeders. Journal of th e Royal Society In terface 4:561-573.
DEAN M N . 2003. S u ctio n feeding in th e p ip id frog, Hymenochirus boettgeri:
K inem atic an d behavioral considerations. C opeia 4:879-886.
D elariva RL, A g o s t in h o AA. 2001. R elationship b e tw ee n m orphology an d diets
of six n eo tro p ical loricariids. Jo u rn al of Fish Biology 58:832-847.
D ettai A, Lec o in tr e G. 2005. F u rth e r su p p o rt fo r th e clades o b ta in e d by
m ultiple m olecular phylogenies in th e acan th o m o rp h bush. C o m p tu s R endus
Biologies 328:674-689.
D h a n y a S, R a ja g o p a l S, A jm a l K h a n S, B a la s u b r a m a n ia n T. 2005. Em bryonic
developm ent in alligator pipefish, Syngnathoides biaculeatus (Bloch, 1785).
C u rre n t Science 88:178-181.
DOBZHANSKY T. 1973. N o th in g in biology m akes sense except in th e lig h t of
evolution. A m erican Biology T eacher 35:125-129.
DROST M R. 1987. R elation betw een aim ing an d catch success in larval fishes.
C anadian Journal of Fisheries an d A quatic Sciences 44:304-315.
DROST M R, M uller M , O ssb JW M . 1988. A q u a n tita tiv e hydrodynam ic m odel of
su ctio n feeing in larval fishes: The role of frictio n al forces. Proceedings of
th e Royal Society of L ondon Series B Biological Sciences 234:263-281.
D r o z d o v AL, K o r n ie n k o ES, K r a s n o lu ts k y . 1997. R ep ro d u ctio n and
developm ent of Syngnathus acusimilis. R ussian Jo u rn al of M arin e Biology
23:265-268.
DUMONT ER, P iccirillo J, G rosse IR. 2005. F in ite-elem en t analysis of b itin g
behavior an d bone stress in th e facial skeletons of bats. T he A natom ical
R ecord P a rt A 283A:3i9-330.
DUMONT ER, G rosse IR, S later GJ. 2009. R equirem ents fo r com paring th e
perform ance of fin ite elem en t m odels of biological stru ctu res. Jo u rn al of
T heoretical Biology 256:96-103.
E d e l a a r P, POSTMA E, K n o p s P, P h illip s R. 2005. N o su p p o rt fo r a genetic basis
of m andible crossing directio n in crossbills (Loxia spp.). The A uk 122:1-7.
E l w o o d JRL, CUNDALL D. 1994. M orp h o lo g y an d behavior of th e feeding
ap p aratu s in Cryptobranchus alleganiensis (Caudata). Jo u rn al of M orphology
220:47-70.
C h a p t e r 8 - R e f e re n c e s
176
EMERSON SB, Bramble DM . 1993. Scaling, allom etry, an d skull design. In:
H a n k en J, H aii BK, editors. The skull volum e 3: F u n c tio n a l and evo lu tio n ary
m echanism s. T he U niversity of C hicago Press, Chicago, pp. 384-421.
E rickson GM, C atanese J, K eaveny TM. 2002. Evolution of the biom echanical
m aterial properties of the femur. Anatomical Record 268:115-124.
FAUSTINO M , P ow er DM . 2001. O steological developm ent of th e viscerocranial
sk eleto n in sea bream: A lternative ossification strategies in te le o st fish.
Jo u rn al of Fish Biology 58:537-572.
F el Icio AKC, R osa IL, Souto A, F reitas RHA. 2006. Feeding behavior of the
longsnout seahorse Hippocampus reidi Ginsburg, 1933. Journal of Ethology
24:219-225.
F erry-G raham LA, W a in w rig h t PC, Bellw ood DR. 2001a. Prey capture in
long-jawed butterflyfishes (Chaetodontidae): The functional basis of novel
feeding habits. Journal of Experim ental M arine Biology and Ecology
256:167-184.
F erry-G raham LA, W a in w rig h t PC, H ulsey CD , Bellw ood DR. 2001b.
E volution an d m echanics of long jaws in b u tterfly fish es (Fam ily
C haetodontidae). Jo u rn al of M orphology 248:120-143.
F erry-G raham LA, Bolnick DI, W a in w r ig h t PC. 2002. Using functional
m orphology to examine the ecology and evolution of specialization.
Integrative and Com parative Biology 42:265-277.
FLYNN AJ, R itz DA. 1999. E ffect of h a b ita t com plexity an d p red ato ry style o n th e
cap tu re of fish feeding on aggregated prey. Jo u rn al of th e M arin e Biology
A ssociation of th e U n ited K ingdom 79:487-494.
M. 1993. C o n trib u tio n s of individual taxa to overall m orphological
disparity. Paleobiology 19:403-419.
FOOTE
POSTER SJ, V in c en t ACJ. 2004. Life h isto ry an d ecology of seahorses:
Im plications for conservation an d m anagem ent. Jo u rn al of Fish Biology 64:161.
P ranzoi P, M accagnani R, R ossi R, C eccherelli VU. 1993. Life cycles and
feeding habits of Syngnathus taenionotus and S. abaster (Pisces,
Syngnathidae) in a brackish bay of the Po river delta (Adriatic Sea). M arine
Ecology Progress Series 97:71-81.
F R O E SE R, PAULY D. 2011. Fishbase. W o rld W id e W eb electro n ic p ublication.
V ersion 05/2010. w w w .fishbase.org. D ow nloaded o n 23 A ugust 2010.
G eerinckx T, ADRIAENS D. 2007. O n to g en y of th e in term a n d ib u la r an d hyoid
m usculature in th e su ck e rm o u th arm o u red catfish Ancistrus cf. triradiatus
(Loricariidae, Siluriform es). A nim al Biology 57:339-357•
C h a p t e r 8 - R e fe re n c e s
177
G e e rin c k x T, B ru n a in M , A d ria e n s D. 2005. D evelopm ent of th e
ch o n d ro cran iu m in th e su ck e rm o u th arm ored catfish A n c istru s cf. tr ir a d ia tu s
(Loricariidae, Siluriform es). Jo u rn al of M orp h o lo g y 266:331-355.
G e e rin c k x T, B ru n a in M , A d ria e n s D. 2007. D evelopm ent of th e o steo cran iu m
in th e su ck e rm o u th arm ored catfish A n c istru s cf. tr ir a d ia tu s (Loricariidae,
Siluriform es. Jo u rn al of M orphology 268:254-274.
G effen AJ, VAN der V eer H W , N ash RDM . 2007. The cost of m etam orphosis in
flatfishes. Journal of Sea R esearch 58:35-45.
G erking SD. 1994. Feeding ecology of fish. A cadem ic Press, C alifornia, p. 416.
GlBB AC. 1995. K inem atics of prey cap tu re in a flatfish, P leu ro n ic h th ys verticalis.
Jo u rn al of E xperim ental Biology 198,1173-1183.
GlBB AC. 1997. D o fla tfish feed like o th e r fishes? A com parative stu d y of
p erco m o rp h prey-capture kinem atics. T he Jo u rn al of E xperim ental Biology
200:2841-2859.
GlBB AC, FERRY-GRAHAM LA. 2005. C ran ial m ovem ents d u rin g su ctio n feeding in
te le o st fishes: A re th ey m odified to enhance su ctio n p ro d u ctio n ? Zoology
108:141-153.
G offredo S, PlCCINBTTI C, Z accanti P. 2004. V o lu n teers in m arine conservation
m onitoring: A study of th e d istrib u tio n of seahorses carried o u t in
collab o ratio n w ith recreatio n al scuba divers. C o n serv atio n Biology 18:14921503.
G o o d w in NB, Balshinb -Earn S, R eynolds JD. 1998. Evolutionary transitions in
parental care in cichlid fish. Proceedings of the Royal Society of London
Series B Biological Sciences 265:2265-2272.
G osline W A. 1967. R eduction in b ran ch io steg al ray num ber. C opeia 1:237-239.
GOSLINE W A. 1977. The stru c tu re and fu n c tio n of th e derm al p ecto ral girdle in
bony fishes w ith p a rtic u la r reference to ostariophysines. Jo u rn al of Zoology
(London) 183, 329-338.
GOSLINE W A. 1981. The evo lu tio n of th e prem axillary p ro tru sio n system in some
teleo stean fish groups. Journ al of Zoology (London) 193:11-23.
GREGORY W K . 1933. Fish skull: A stu d y of th e ev o lu tio n of n a tu ra l m echanism s.
T ransactions of th e A m erican P hilosophical Society 32:75-481.
GROBECKER DB, PlBTSCH TW . 1979. H igh-speed cinem atographic evidence for
u ltrafa st feeding in a n te n n a riid anglerfishes. Science 205:1161-1162.
HALB M E. 1996. F u n c tio n a l m orphology of v e n tra l ta il b en d in g an d prehensile
abilities of th e seahorse, H ip p o c a m p u s kuda. Jo u rn al of M orphology 227:51-65.
HALL BK. 2005. Bones and cartilage: D evelopm ental an d ev o lu tio n ary skeletal
biology. Elsevier A cadem ic Press, London, p. 760.
i
78
C h a p t e r 8 - R e f e re n c e s
HAMMER 0 , H a r p e r DAT, Ryan PD. 2001. PAST: Paleontological Statistics
softw are package for ed u catio n and d ata analysis. Palaeontologia E lectrónica
4. p.9.
HARRINGTON R W Jr. 1955. The o steo cran iu m of th e A m erican cyprinid fish,
N o t r o p i s bifren atus, w ith an a n n o ta te d synonym y of te le o st skull bones.
C opeia 4:267-291.
H elfman GS, C ollette BB, F acby DE, Bo w en BW . 2009. The diversity of fishes
2nd edition. Biology, evolution, an d ecology. W iley-B lackw ell, C hichester, p.
720.
HERALD ES. 1959. From pipefish to seahorse: A study of phylogenetic
relationships. Proceedings of the California Academy of Sciences 29:465-473.
HERNANDEZ LP, B ird N C , Staab KL. 2007. U sing zeb rafish to investigate
cypriniform
evolutionary
novelties:
fu n c tio n a l
developm ent
and
evolutionary diversification of th e kinethm oid. Jo u rn al of E xperim ental
Zoology (M ol D ev Evol) 3086:625-641.
H ernandez LP, G ibb AG, F erry-G raham L. 2009. Trophic apparatus in
cyprinodontiform fishes: Functional specializations for picking and scraping
behaviors. Journal of M orphology 270:645-661.
HERRING SW. 1993. E pigenetic an d fu n c tio n a l influences o n skull grow th. In:
H a n k en J, H all BK, editors. The skull volum e 1: D evelopm ent. T he U niversity
of C hicago Press, Chicago, pp. 153-206.
HERRING SW, T eng S. 2000. Strain in the braincase and its sutures during
function. Am erican Journal of Physical A nthropology 112:575-593.
HOLLIDAY JA, Stbppan SJ. 2004. Evolution of hypercarnivory: The effect of
specialization on morphological and taxonom ic diversity. Paleobiology
30:108-128.
HORTON JM, Summers AP. 2009. The m aterial properties of acellular bone in a
teleost fish. The Journal of Experim ental Biology 212:1413-1420.
HUNT von H brbing I, Keating K. 2003. T em p eratu re-in d u ced changes in
viscosity an d its effects on sw im m ing speed in larval haddock. In: B row m an
HI, Skiftesvik AB, editors. T he Big F ish Bang. Proceedings of th e 26the
A n n u al Larval Fish C onference. In s titu te of M arin e Research, Bergen, pp. 2334H u n t v o n H b rb in g I, M iyake T, H a l l BK, B o u tilie r RG. 1996a. O ntogeny of
feeding and respiration in larval A tlantic cod G a du s m orhua (Teleostei,
Gadiformes): I. M orphology. Journal of M orphology 227:15-35.
H u n t v o n H b rb in g I, M iyake T, H a l l BK, B o u tilie r RG. 1996b. O ntogeny of
feeding and respiration in larval A tlantic cod G a du s m orhua (Teleostei,
Gadiformes): II. Function. Journal of M orphology 227:37-50.
C h a p t e r 8 - R e fe re n c e s
179
HUSKEY S, Q uintero R. 2006. W h y th e long face? Prey cap tu re in tru m p etfish es.
In teg rativ e and C om parative Biology 46:E65-E65.
HUYSENTRUYT F. 2008. Early developm ent of C o r y d o ra s aeneus (Siluriform es,
C allichthyidae): A case stud y fo r u n d erstan d in g th e ev o lu tio n ary basis of
loricarioid o n to g e n etic p attern in g . U n p u b lish ed D o cto ral Thesis. G h e n t
U niversity.
HUYSENTRUYT F, GEERINCKX T, BRUNAIN M , ADRIAENS D. 2011. D evelopm ent of
th e o steocranium in C o ryd o ra s aeneus (Gili, 1858) (C allichthyidae,
Siluriform es). Jo u rn al of M orphology.
IU C N . 2010. IU C N Red List of T h reaten ed Species. W o rld W id e W eb electro n ic
publication.
V ersion 2010.4.
w w w .iu cn red list.o rg.
D ow nloaded on
15 D ecem ber 2010.
JAMIESON BGM. 1991. Fish evo lu tio n an d system atic: Evidence fro m sperm atozoa.
W ith a survey of lo p h o p h o ra te , ech in o d erm and p ro to c h o rd a te sperm an d an
a cco u n t of gam ete cryopreservation. C am bridge U niversity Press, C am bridge,
p. 319.
JOHNSON GD, Patterson C. 1993. Percom orph phylogeny: A survey of
acanthom orphs and a new proposal. Bulletin of M arine Science 52:554-626.
Jo h n so n GD, Paxton JR, Sutton TT, Satoh TP, Sado T, N ishida M, M iya M.
2009. D eep-sea m ystery solved: asto n ish in g larval tra n sfo rm a tio n s and
extrem e sexual dim orphism u n ite th re e fish families. Biology L etters 5:235239.
JUNGERSEN H FE. 1910. Ich th y o to m ical c o n trib u tio n s. II. The s tru c tu re of th e
A ulostom idae, Syngnathidae and Solenostom idae. D et K ongelige D anske
V idenskabernes Selskabs S k rifter 8:267-364.
K ad am KM. 1958. The developm ent of th e ch o n d ro cran iu m in th e sea-horse,
H ip p o c a m p u s (Lophobranchii). Zoological Jo u rn al of th e L innean Society
43:557-573-
KADAM KM. 1961. T he developm ent of th e skull in N e ro p h is (Lophobranchii).
A cta Zoologica 42:257-298.
K a n o u K, K o h n o H. 2001. Early life h isto ry of a seahorse, H ip p o c a m p u s m ohnikei,
in Tokyo Bay, Japan. Ichtyological R esearch 48:361-368.
K a rin a A, F e lic io C, R o sa IL, S o u to A, F r e ita s RHA. 2006. Feeding behavior of
th e longsnout seahorse H ip p o c a m p u s reidi Ginsburg, 1933. Journal of
Ethology 24:219-225.
Kawahara R, M iya M , M abuchi K, Lavoué S, I noue JG, Satoh TP, Kaw aguchi
A, NlSHIDA M. 2008. In te rrela tio n sh ip s of th e 11 g astero steifo rm fam ilies
(sticklebacks, pipefishes, an d th e ir relatives): A n ew perspective based on
w hole m itogenom e sequences fro m 75 h ig h er teleosts. M olecular
i8 o
C h a p t e r 8 - R e f e re n c e s
P h y lo g e n e t ic s a n d E v o lu t io n 4 6 :2 2 4 -2 3 6 .
K bivany Y, NELSON JS. 2006. In te rrela tio n sh ip s of G asterosteiform es
(A ctinopterygii, Percom orpha). Jo u rn al of Ichthyology 46:884-896.
KENDRICK AJ, H yndbs GA. 2003. P a tte rn s in th e abundance an d size-d istrib u tio n
of syngnathid fishes am ong h a b itats in a seagrass-dom inated m arine
environm ent. E stuarine, C oastal an d Shelf Science 57:631-640.
KENDRICK AJ, H yndbs GA. 2005. V ariations in th e d ietary com positions of
m orphologically diverse syng n ath id fishes. E n v iro n m en tal Biology of Fishes
72:415-427.
KINDRED JE. 1921. The ch o n d ro cran iu m
M orphology 35:425-456.
of S y n g n a th u s fuscus. Jo u rn al of
KINDRED JE. 1924. A n in te rm e d ia te stage in th e developm ent of th e skull of
S y n g n a th u s fuscus. A m erican Jo u rn al of A natom y 33:421-447.
K its o s M S, T z o m o s T, A n a g n o s t o p o u l o u L, K o u k o u r a s A. 2008. D iet
com position of th e seahorses, H ip p o c a m p u s g u tt u l a tu s C uvier, 1829 and
H ip p o c a m p u s h ip p o c a m p u s (L., 1758) (Teleostei, Syngnathidae) in th e A egean
Sea. Journal of Fish Biology 72:1259-1267.
K old ew ey HJ, M a r t in -Sm it h KM. 2010. A
aquaculture. A quaculture 302:131-152.
global
review
of
seahorse
KOLM N, A h n e sjö I. 2005. Do egg size an d p a re n ta l care coevolve in fishes?
Jo u rn al of Fish Biology 66:1499-1515.
KORNIENKO ES. 2001. R epro d u ctio n an d developm ent in som e genera of p ip efish
an d seahorses of th e fam ily Syngnathidae. R ussian Jo u rn al of M arin e Biology
27:515-526.
K u iter RH. 2003. Seahorses, pipefishes an d th e ir relatives. A com prehensive
guide to S yngnathiform es revised edition. TM C Publishing, C horleyw ood. p.
237.
KUITER RH. 2004. A n ew pygm y p ipehorse (Pisces: Syngnathidae: Idiotropiscis)
from eastern A ustralia. Records of th e A u stralian M u seu m 56:163-165.
KUTSCHERA U, NIKLAS KJ. 2004. The m o d ern th e o ry of biological evolution: an
expanded synthesis. N atu rw issen sch aften 91:255-276.
Kv a r n em o C, SIMMONS LW . 2004. Testes in v estm en t and spaw ning m ode in
pipefishes and seahorses (Syngnathidae). Biological Jo u rn al of th e L innean
Society 83:369-376.
K v a rn e m o C, M o o r b G I, J o n e s AG, N e ls o n W S, A vise JC. 2000. M onogam ous
pair bonds and m ate sw itching in th e w e ste rn A u stralian seahorse
H ip p o c a m p u s subelongatus. Jo u rn al of E v o lutionary Biology 13:882-888.
LARSEN PS, M a d sen CV, R iisg â RD. 2008. E ffect of te m p e ra tu re an d viscosity on
C h a p t e r 8 - R e fe re n c e s
181
sw im m ing velocity of th e copepod Acartia tonsa, b rin e shrim p A nem ia salina
an d ro tife r Brachionus plicatilis. A quatic Biology 4:47-54.
LAUDER GV, Liem K R 1983. The ev o lu tio n an d in terrelatio n sh ip s of th e
a ctin o p tery g ian fishes. B ulletin of T he M u seu m of C om parative Zoology
150:95-197.
LAUDER GV. 1979. Feeding mechanics in prim itive teleosts and in the
halecom orph fish Am ia calva. Journal of Zoology (London) 187:543-578.
LAUDER GV. 1980a. E volutio n of th e feeding m echanism in prim itive
a ctin o p tery g ian fishes: A fu n c tio n a l anatom ical analysis of Polypterus,
Lepisosteus and Amia. Journal of M orphology 163:283-317.
LAUDER GV. 1980b. The su ctio n feeding m echanism in sunfishes (Lepomis): A n
experim ental analysis. The Jo u rn al of E xperim ental Biology 88:49-72.
LAUDER GV. 1983. Food capture. In: W ebb PW , W eihs D, editors. Fish
B iom echanics. Praeger Publishers, N ew York. pp. 280-311.
LAUDER GV. 1985. A quatic feeding in lo w er vertebrates. In: H ild eb ran d M ,
B ram ble D M , Liem KP, W ak e DB, editors. F u n c tio n a l v erteb rate
m orphology. The B elknap Press, C am bridge, pp. 210-229.
LAUDER GV, N orton SP. 1980. A sym m etrical m uscle activity d u rin g feeding in
th e gar, Lepisosteus oculatus. Jo u rn al of E xperim ental Biology 84:17-32.
LAWRENCE E. 2005. H en d erso n ’s d ictio n ary of biology 13th edition. Pearson
E du catio n Lim ited, Essex, p. 748.
LEKANDER B. 1949. The sensory line system and th e canal bones in th e head of
som e O stariophysi. A cta Zoologica 30:1-131.
Lieberman DE, D evlin MJ, P earson OM. 2001. A rticular area responses to
m echanical loading: Effects of exercise, age and skeletal location. Am erican
Journal of Physical A nthropology 116:266-277.
LlEM KF. 1991. A fu n c tio n a l app ro ach to th e developm ent of th e head of teleosts:
Im plications on c o n stru ctio n al m orphology an d co n strain ts. In: Schm idtK ittle r N, Vogel K, editors. C o n stru ctio n al M orphology an d E volution.
Springer-Verlag, Berlin, pp. 231-249.
LlEM KF, Stewart DJ. 1976. E volution of th e scale-eating cichlid fishes of Lake
Tanganyika: A generic revision w ith a d escrip tio n of a new species. B ulletin
M u seu m of C om parative Zoology 147:319-350.
LINNAEUS C. 1766. System a n a tu ra e p er régna tria n atu rae, secundum classes,
ordines, genera, species, cum characteribus, d ifferen tiis, synonym is, locis.
Tornus 1 (R egnum anim ale). E ditio duodécim a, reform ata. L au ren tii Salvii,
H olm iae. p. 532.
LINTON JR, Soloff BL. 1964. The physiology of the brood pouch of the male sea
C h a p t e r 8 - R e f e re n c e s
182
horse H ippocam pus erectus. B u lletin of M arin e Science of th e G ulf and
C aribbean 14:45-61.
Lourie SA, P ritchard JC, C asey SP, T ruong SK, H all HJ, V in c en t ACJ. 1999a.
The taxonom y of V ietnam ’s exploited seahorses (family Syngnathidae).
Biological Jo u rn al of th e L innean Society 66:231-256.
LOURIE SA, V in c en t ACJ, H all HJ. 1999b. Seahorses. A n id e n tific a tio n guide to
th e w o rld ’s species and th e ir conservation. P roject Seahorse, London, p. 214.
Lo u r ie SA, F oster SJ, C o o pe r E W T, V in c e n t ACJ. 2004. A guide to th e
id e n tific a tio n of seahorses. P roject Seahorse an d TRAFFIC N o rth America.
U niversity of B ritish C olum bia an d W o rld W ild life F und, W ash in g to n DC.
p. 114.
LOURIE SA, G reen D M , V in c en t ACJ. 2005. D ispersal, h a b ita t differences, and
com parative phylogeography of S o u th east A sian seahorses (S yngnathidae :
Hippocampus). M olecular Ecology 14:1073-1094.
M a r k e y M J, M A R S H A L L CR. 2007. Linking form an d fu n c tio n of th e fibrous joints
in th e skull: A new q u a n tific a tio n schem e fo r cran ial su tu res using th e ex tan t
fish Polypterus endlicherii. Journ al of M orp h o lo g y 268:89-102.
M arkey M J, M ain RP, M arshall CR. 2006. In vivo cranial su tu re fu n c tio n and
su tu re m orphology in th e ex tan t fish Polypterus: im plications fo r in ferrin g
skull fu n c tio n in living an d fossil fish. The Jo u rn al of E xperim ental Biology
209:2085-2102.
M asschaele BC, C n u d d e V, D ier ic k M , Jacobs P, V a n H oorebeke L,
V lassenbroeck J. 2007. UGCT: N ew X-ray radiography an d tom o g rap h y
facility. N uclear In stru m e n ts & M eth o d s in Physics R esearch S ection A:
A ccelerators, S pectrom eters, D etecto rs an d A ssociated E q u ip m en t 580:266269.
MAYR E. 1996. W h a t is a species an d w h a t is not? P hilosophy of Science 63:262277.
DE. 1968. T he evo lu tio n of branchiostegals and associated opercular,
guiar, and hyoid bones and th e classification of teleo sto m e fishes, living and
fossil. B ulletin of th e N atio n al M u seu m of C anada (Biological Series) 221:1239.
M C A L L IS T E R
M cG O N N E L L IM .
2001. The evo lu tio n of th e p e cto ral girdle. Jo u rn al of A natom y
199:189-194.
M c H enry CR, C lausen PD, D an iel W JT , M eers M B, P en d h a rk a r A. 2006.
B iom echanics of th e ro stru m in crocodilians: A com parative analysis using
fin ite-elem en t m odeling. The A natom ical R ecord P a rt A 288A:827-849.
JPM A. 1883. O n th e osteology an d developm ent of Synynathus
peckianus (Storer). T he Q u arterly Jo u rn al of M icroscopical Science 23:623-
M cM U R R IC H
C h a p t e r 8 - R e fe re n c e s
183
650.
M iya M , T akeshima H , E n d o H , I shiguro NB, I noue JG, M ukai T, Satoh TP,
Yamaguchi M , Kaw aguchi A, M abuchi K, Shirai SM, N ishida M. 2003.
M ajor p a tte rn s of hig h er teleo stean phylogenies: A new perspective based on
100 com plete m ito ch o n d rial D N A sequences. M o lecu lar Phylogenetics and
E volution 26:121-138.
M iya M , Sa t o h TP, N ish id a M . 2005. The p h ylogenetic p o sitio n of toadfishes
(order B atrachoidiform es) in th e h ig h er ray-finned fish as in fe rred from
p a rtitio n e d Bayesian analysis of 102 w h o le m ito ch o n d rial genom e sequences.
Biological Jo u rn al of th e L innean Society 85:289-306.
N M , A lm a d a VC, V ie ira M N . 2003. Early life h isto ry of th e p ip efish
Nerophis lumbriciformis (Pisces: Syngnathidae). Jo u rn al of th e M arin e
Biological A ssociation of th e U n ited K ingdom 83:1179-1182.
M O N T E IR O
MOTTA PJ. 1984. M echanics and fu n c tio n of jaw p ro tru sio n in te le o st fishes: A
review. C opeia 1:1-18.
MULLER M. 1987. O p tim iz a tio n p rinciples applied to th e m echanism s of
n e u ro c ran iu m levation an d m o u th b o tto m depression in b o n y fishes
(H alecostom i). Jo u rn al of T h eo retical Biology 126:343-368.
MULLER M , O sse JW M . 1984. H ydrodynam ics of su ctio n feeding in fish.
T ransactions of th e Zoological Society of L ondon 37:51-135.
M urugan A, D hanya S, Rajagopal S, Balasubramanian T. 2008. Seahorses and
pipefishes of th e Tamil N adu coast. C urrent Science 95:253-260.
N alla RK, KINNEY JH , R itch ie RO. 2003. M ech an istic failure criteria fo r th e
failure of h u m a n cortical bone. N a tu re M aterials 2:164-168.
NAPLES VL. 1999. M orphology, ev o lu tio n an d fu n c tio n of feeding in th e g ian t
a n te a te r (Myrmecophaga tridactyla). Jo u rn al of Zoology L ondon 249:19-41.
NELSON GJ. 1969. In frao rb ita l bones and th e ir b earing o n th e phylogeny and
geography of osteoglossom orph fishes. A m erican M u seu m N ovitates 2394:137NELSON JS. 2006. Fishes of th e w o rld 4th edition. Jo h n W iley & Sons, N ew Jersey,
p. 601.
OLIVEIRA F, E r zin i K, G onçalves JMS. 2007. Feeding h ab its of th e deep-snouted
pipefish Syngnathus typhle in a tem p erate coastal lagoon. E stuarine, C oastal
an d Shelf Science 72:337-347.
OSSE JW M , MULLER M . 1980. A m odel of su ctio n feeding in te le o st fishes w ith
som e im plications for v en tilatio n . In: Ali M A, editor. E n v iro n m en tal
physiology of fishes. P lenum P ublishing C o rp o ratio n , N ew York. pp. 335-352.
O tten E. 1982. The developm ent of a m o u th -o p en in g m echanism in a generalized
C h a p t e r 8 - R e f e re n c e s
184
species: H. elegans Trew avas
N eth erlan d s Journal of Zoology 32:31-48.
H a p lo ch ro m is
1933
(Pisces,
Chichlidae).
PALMA J, A n d r a d e JP 2002. M orphological stu d y of D ip lo d u s sargus, D ip lo d u s
p u n ta z z o , an d L ith o g n a th u s m orm g ru s (Sparidae) in th e E astern A tlan tic and
M ed iterran e a n Sea. Fisheries R esearch 57:1-8.
LR, S o n g J. 1996. P hylogenetic significance of th e pectoral-pelvic fin
association in acan th o m o rp h fishes: A reassessm ent using com parative
neuroanatom y. In: Stiassny M LJ, P a re n ti LR, Jo h n so n GD, editors.
In te rrela tio n sh ip s of fishes. A cadem ic Press, C alifornia, pp. 427-444.
PARENTI
C. 1977. C artilage bones, derm al bones and m em brane bones, or th e
exoskeleton versus th e endoskeleton. In: A ndrew s SM, M iles RS, W alk e r AD,
editors. Problem s in V ertebrate Evolution. A cadem ic Press, London, pp. 77121.
PATTERSO N
C. 1993. A n overview of th e early fossil reco rd of A canthom orphs.
B ulletin of M arin e Science 52:29-59.
PATTERSO N
P a y n e M F, R ip p in g a le RJ, L o n g m o re RB. 1998. G ro w th an d survival of juvenile
pipefish (S tig m a to p o r a argus) fed live copepods w ith h ig h an d low H U FA
co n ten t. A quaculture 167:237-245.
P h a m T c h i M i, K o r n ie n k o ES, D r o z d o v AL. 1998. E m bryonic an d larval
developm ent of th e seahorse H ip p o c a m p u s kuda. R ussian Jo u rn al of M arin e
Biology 24:325-329.
PETERSON J, D e c h o w PC. 2003. M ate ria l p ro p erties of th e h u m an cranial vault
an d zygoma. The A natom ical R ecord P a rt A
C W , W O O D S C M C , J A M E S J P , G lA M B A R T O L O M E I FM , L O K M A N PM.
2004. R eproductive biology of fem ale big-bellied seahorses. Jo u rn al of Fish
Biology 64:717-725.
POORTENAAR
RALSTON KR, W a in w r ig h t PC. 1997. F u n c tio n a l consequences of tro p h ic
specialization in pufferfish. F u n ctio n al Ecology 11:43-52.
R ayfield EJ. 2005. U sing fin ite-elem en t analysis to investigate su tu re
m orphology: A case study using large carnivorous dinosaurs. T he A natom ical
R ecord P a rt A 283A:349-365.
R e ir iz L, N ic ie z a AG, BRAÑA F. 1998. Prey selection by experienced an d naive
juvenile A tlan tic salm on. Journ al of Fish Biology 53:100-114.
RENO H W . 1966. The in frao rb ital canal, its lateral-line ossicles an d neurom asts, in
th e m innow s N o tr o p is volucellus and N . buchanani. C opeia 3:403-413.
R e y s e rh o v e L. 2008. Liggen m orfologische tran sfo rm aties in h e t cran iu m van
larvale zeepaardjes (H i p p o c a m p u s reidi) aan de basis van de m o rtaliteitsp iek in
h u n ontw ikkeling? U npu b lish ed B achelor Thesis. G h e n t U niversity.
C h a p t e r 8 - R e fe re n c e s
185
R ic h m o n d BG, W right BW, G rosse I, D ec h o w PC, Ross CP, Spencer MA,
STRAIT DS. 2005. F in ite elem en t analysis in fu n c tio n a l m orphology. The
A natom ical R ecord P a rt A 283A:259-274.
RIDLEY M. 1993. Evolution. O xford: B lackw ell S cientific P ublications, p. 670.
RIDLEY M. 2003. Evolution. O xford U niversity Press, O xford, p. 472.
JL, F O R A N CM . 2009. D irect evidence for em bryonic u p tak e of paternallyderived n u trie n ts in tw o pipefishes (Syngnathidae: Syngnathus spp.). Jo u rn al
of C om parative Physiology B-Biochem ical System ic an d E n v iro n m en tal
Physiology 179:325-333.
R IP L E Y
ROHLF FJ. 2003. TpsSmall: T h in P late Spline Sm all V ariation (version 1.20). Stony
B rook U niversity, N ew York.
ROHLF FJ. 2009. TpsRegr: T h in P late Spline R egression (version 1.39). Stony
B rook U niversity, N ew York.
ROHLF FJ. 2010a. TpsUtil: T h in Plate Spline U tility (version 1.46). Stony Brook
U niversity, N ew York.
ROHLF FJ. 2010b. TpsDig2: T h in P late Spline D igitise (version 2.16). Stony Brook
U niversity, N ew York.
ROHLF FJ.
1 .4 9 ).
TpsRelw: T h in P late Spline R elative W arps Analysis (version
Stony B rook U niversity, N ew York.
2010C .
ROJO AL. 1991. D ictio n ary of evo lu tio n ary fish osteology. CRC Press, Boca Raton,
p. 273.
ROOS G. 2010. Specializations of extrem e head m orphology: A d etailed fu n c tio n a l
m orphological study of feeding in seahorses an d pipefishes (Syngnathidae).
U npu b lish ed D octoral Thesis. U niversity of A ntw erp.
Roos G, VAN W assbnbbrgh S, H brrbl A, A erts P. 2009. Kinematics of suction
feeding in th e seahorse Hippocampus reidi. The Journal of Experim ental
Biology 212:3490-3498.
Roos G, Van W assbnbbrgh S, H brrbl A, A driaens D, A brts P. 2010. Snout
allom etry in seahorses: Insights on optim ization of pivot feeding
perform ance during ontogeny. The Journal of Experim ental Biology
213:2184-2193.
Roos G, Van W assbnbbrgh S, A erts P, H errel A, A driaens D. 2011. Effects of
snout dim ensions on th e hydrodynamics of suction feeding in juvenile and
adult seahorses. Journal of Theoretical Biology 269:307-317.
Roos G, Van W assbnbbrgh S, A brts P, H errel A, A driaens D. subm itted a.
W hy th e long face? A comparative study of feeding kinem atics of tw o
pipefish w ith different snout lengths. Journal of Fish Biology.
i86
C h a p t e r 8 - R e f e re n c e s
Roos G, Lbysbn H, Van W assbnbbrgh S, A brts P, H brrbl A, A driaens D.
subm itted b. Locking th e pivot-feeding mechanism in Syngnathidae: The role
of snout adduction. Animal Biology.
ROSE CS, Reiss JO. 1993. M etam o rp h o sis and th e v erteb rate skull: O n to g en etic
p a tte rn s and developm ent m echanism s. In: H a n k en J, H aii BK, editors. The
skull volum e 1: D evelopm ent. The U niversity of C hicago Press, Chicago, pp.
289-346.
ROSS CP. 2005. F in ite elem en t analysis in v erteb rate biom echanics. The
A natom ical R ecord P a rt A 283A:253-258.
RYDER JA. 1881. A c o n trib u tio n to th e developm ent an d m orphology of th e
lo p h o b ran ch iates (H ip p o c a m p u s a n tiqu oru m , th e sea-horse). B ulletin of th e
U n ited States Fish C om m ission 1882:191-199.
R y e r C H , ORTH RJ. 1987. Feeding ecology of th e n o rth e rn pipefish, S yn g n a th u s
fu scu s, in a seagrass com m u n ity of th e lo w er C hesapeake Bay. E stuaries
10:330-336.
Salzburger W , M ack T, V erheyen E, M bybr A. 2005. O u t of Tanganyika:
Genesis, explosive spéciation, key-innovations and phylogeography of the
haplochrom ine cichlid fishes. BMC Evolutionary Biology 5:17.
Sanders JG, C ribbs JE, F ibnbbrg HG, H ulburd GC, Katz LS, Palumbi SR. 2008.
The tip of th e tail: M olecular identification of seahorses for sale in
apothecary shops and curio stores in California. Conservation Genetics 9:6571 SANDERSON, S. L. 1991. F u n c tio n a l stereotypy an d feeding perfo rm an ce correlated
in a tro p h ic specialist. F u n ctio n al Ecology 5:795-803.
SANFORD CPJ, W a in w rig h t PC. 2002. Use of so n o m icro m etry d em o n strates th e
lin k b etw een prey cap tu re kinem atics an d su ctio n pressure in larg em o u th
bass. The Journal of E xperim ental Biology 205:3445-3457.
SANFORD CPJ, DAY S, K in o w N. 2009. The role of m o u th shape o n th e
hydrodynam ics of su ctio n feeding in fishes. In teg rativ e an d C om parative
Biology 49 (suppl.i):ei49.
Sarin VK, E rickson GM, G iori NJ, Bergman AG, C arter DR. 1999. C oincident
developm ent of sesamoid bones and clues to their evolution. Anatomical
Record 257:174-180.
SASAKI K, T a n a k a Y, T a k a ta Y. 2006. C ran ial m orphology of A te le o p u s ja p o n ic u s
(A teleopodidae, A teleopodiform es), w ith a discussion o n m etam orphic
m o u th m ig ratio n an d lam p rid ifo rm affinities. Ichthyological R esearch
53:254-263.
SCALES H. 2010. A dvances in th e ecology, biogeography an d conservation of
seahorses (genus H ip p o c a m p u s ). Progress in Physical G eography 34:443-458.
C h a p t e r 8 - R e fe re n c e s
187
SCHULTZE H P. 2008. N o m en clatu re and h o m o lo g izatio n of cranial bones in
actinopterygians. In: A rratia G, S chultze H P, W ilso n W V H , editors.
M esozoic fishes 4 - H om ology an d phylogeny. V erlag Dr. F ried erich Pfeil.,
M ü n ch en , pp. 23-48.
H D . 2000a. C oordG en6f: In te g ra te d M o rp h o m etries Package C o o rd in ate
G en eratin g
U tility.
C anisius
College,
B uffalo,
N ew
York,
h ttp ://w w w 3 .can isiu s.ed u /~ sh eets/m o rp h so ft.h tm l.
SH EETS
H D . 2000b. T w oG roup6h: In te g ra te d M o rp h o m etries Package. C anisius
College,
Buffalo,
N ew
York,
h ttp ://w w w 3 .can isiu s.ed u /~ sh eets/m o rp h so ft.h tm l.
SH EETS
S ilv a K, M o n t e i r o N M , A lm a d a VC, V ie ira M N . 2006. Early life h isto ry of
S y n g n a th u s abaster. Jo u rn al of Fish Biology 68:80-86.
SMITH W mL. 2010. P ro m o tin g re so lu tio n of th e p erco m o rp h bush: A reply to
M ooi an d Gili. C opeia 3:520-524.
SMITH W m L, W h eeler W C . 2004. Polyphyly of th e m ail-cheeked fishes
(Teleostei: Scorpaeniform es): Evidence from th e m ito ch o n d rial an d n uclear
sequence data. M olecular Phylogenetics and E volution 32:627-646.
SMITH W m L, C raig M T. 2007. C astin g th e p erco m o rp h n e t widely: The
im p o rtan ce of b ro ad taxonom ic sam pling in th e search fo r th e placem en t of
serranid and p ercid fishes. C opeia 1:35-55.
SPRINGER VG, O r r e l l TM . 2004. A ppendix: Phylogenetic analysis of 147 fam ilies
of a ca n th o m o rp h fishes based o n dorsal gill-arch m uscles and skeleton.
B ulletin of th e Biological Society of W ash in g to n 11:237-260.
S tarck D. 1979. D er S ch u lterg ü rtel der Fische. In: Starck D, editors.
V ergleichende A natom ie der W irb eltiere. Springer-V erlag, Berlin, pp. 467472.
STEWART TA, A lbertson RC. 2010. E v o lu tio n of a u n iq u e p red ato ry feeding
apparatus: fu n c tio n a l anatom y, developm ent an d a genetic locus fo r jaw
la te rality in Lake Tanganyika scale-eating cichlids. BMC Biology 8:11.
S tia s s n y
MLJ.
1996.
Basal
ctenosquam ate
relatio n sh ip s
and
th e
in te rre la tio n sh ip s of th e m y cto p h ifo rm (Scopelom orph) fishes. In: Stiassny
M LJ, P a re n ti LR, Jo hnson G D, editors. I n te rrela tio n sh ip s o f fish es. Academ ic
Press, San Diego, pp. 405-426.
Stö LTING KN, WILSON AB. 2007. M ale preg n an cy in seahorses and pipefish:
Beyond th e m am m alian model. BioEssays 29:884-896.
S t o r e r o LP, GONZÁLEZ RA. 2009. Prey selectivity an d tro p h ic behavior of th e
P atagonian seahorse, H ip p o c a m p u s p a ta g o n ic u s, in captivity. Jo u rn al of th e
W o rld A quacu ltu re Society 40:394-401.
i8 8
C h a p t e r 8 - R e f e re n c e s
SVANBÄCK R, WAIN WRIGHT PC, F erry -G rah am LA. 2002. Linking cranial
kinem atics, buccal pressure, and su ctio n feeding perfo rm an ce in larg em o u th
bass. Physiological an d B iochem ical Zoology 75:532-543.
SWINNERTON H H . 1902. A c o n trib u tio n to th e m orphology of th e te leo stean head
skeleton, based u p o n a stu d y of developing skull of th e th ree-sp in ed
stickleback (Gasterosteus aculeatus). Q u arterly Jo u rn al of M icroscopical
Sciences 45:503-597.
Taka h a sh i R, M o r iw a k i T, H o ri M. 2007. Foraging beh av io u r an d fu n c tio n a l
m orphology of tw o scale-eating cichlids fro m Lake Tanganyika. Jo u rn al of
Fish Biology 70:1458-1469.
T akata J, SASAKI K. 2005. B ranchial stru c tu re s in th e G asterosteiform es, w ith
special reference to m yology an d p h y logenetic im plications. Ichthyological
R esearch 52:33-49.
TAYLOR W R , VAN D yke GC. 1985. Revised p rocedures fo r stain in g an d clearing
sm all fishes an d o th e r v ertebrates fo r bone an d cartilage study. C ybium 9:107119.
W . 1953. T he feeding m echanism s of a deep sea fish. B ritish
M u seu m (N atu ral H istory), London, p. 101.
T C H E R N A V IN
T eixeira RL, M u sick JA. 1995. T rophic ecology of 2 congeneric pipefishes
(Syngnathidae) of th e low er York River, V irginia. E n v iro n m en tal Biology of
Fishes 43:295-309.
T esc h FW . 2003. T he eel 3rd edition. B lackw ell Science, O xford, p.408.
T eske PR, B eheregaray LB. 2009. E v o lu tio n of seahorses’ u p rig h t p o sitio n was
lin k ed to O ligocene expansion of seagrass h abitats. Biology L etters 5:521-523.
T eske PR, C h e r r y M I, M a t t h e e CA. 2003. P o p u la tio n genetics of th e
endangered K nysna seahorse, H ip p o c a m p u s capensis. M o lecu lar Ecology
12:1703-1715.
TESKE PR, C h e r r y M I, M a t t h e e CA. 2004. T he ev o lu tio n ary h isto ry of seahorses
(Syngnathidae: H ip p o c a m p u s ): M o lecu lar data suggest a W e st Pacific origin
an d tw o invasions of th e A tlan tic O cean. M o lecu lar Phylogenetics and
E volution 30:273-286.
T eske PR, H a m il to n H, P alsboll PJ, C h o o CK, G abr H , Lourib SA, Sa n to s M ,
S rbbpada A, CHERRY M I, M a tth ee CA. 2005. M o lecu lar evidence fo r lo n g ­
distance co lo n izatio n in an Indo-Pacific seahorse lineage. M arin e Ecology
Progress Series 286:249-260.
T h ie le m a n s A. 2007. Studie n a ar h e t m usculo-skeletale ap p araat bij een larvaal
Kaaps zeepaardje ('H ippocam pu s capensis) aan de h a n d van een grafische 3Dreco n stru ctie. U npu b lish ed B achelor Thesis. G h e n t U niversity.
C h a p t e r 8 - R e fe re n c e s
189
T iln e y RL, HECHT T. 1993. Early o n to g en y of G a leich th ys fe lic e p s fro m th e so u th
east coast of S o u th Africa. Jo u rn al of Fish Biology 43:183-212.
TIPTON K, Bell SS. 1988. Foraging p a tte rn s of tw o syn g n ath id fishes: Im p o rtan ce
of harp actico id copepods. M arin e Ecology Progress Series 47:31-43.
VAN DAMME J, A e r ts P. 1997. K inem atics an d fu n c tio n a l m orphology of aquatic
feeding in A ustralian snake-necked tu rtle s (Pleurodira; C helodina ). Jo u rn al of
M orphology 233:113-125.
VAN GlNNEKEN VJT, M a e s GE. 2005. T he E u ro p ean eel (A n gu illa a n g u illa ,
Linnaeus), its lifecycle, ev o lu tio n an d rep roduction: a lite ra tu re review.
Reviews of Fish Biology an d Fisheries 15:367-398.
V a n W a ssen berg h S, H errel A, A dria en s D, A erts P. 2005. A te st of m outhopening an d hyoid-depression m echanism s du rin g prey cap tu re in a catfish
using high-speed cineradiography. The Journal of E xperim ental Biology 208:
4627-4639.
V a n W a ssen berg h S, H errel A, A dria en s D, A erts P. 2006a. M o d u latio n and
variability of prey cap tu re k in em atics in clariid catfishes. Jo u rn al of
E xperim ental Zoology 305A:559-569.
V a n W a ssen berg h S, A erts P, H errel A. 2006b. H ydrodynam ic m odelling of
aquatic su ctio n perform ance an d in tra-o ral pressures: lim itatio n s for
com parative studies. Journal of th e Royal Society In terface 3:507-514.
V a n W a ssen berg h S, St r o t h er JA, F la m m ang B, F erry -G rah am LA, A erts P.
2008. E xtrem ely fast prey cap tu re in p ip efish is p o w ered by elastic recoil.
Jo u rn al of th e Royal Society In terface 5:285-296.
V a n W a ssen berg h S, R oos G, G en bru g g e A, Leysen H , A erts P, A dria en s D,
HERREL A. 2009. S uction is kids play: Extrem ely fast su ctio n in n ew b o rn
seahorses. Biology L etters 5:200-203.
V a n W a ssen berg h S, R oos G, F erry L. 2011. A n adaptive ex p lan atio n fo r th e
horse-like shape of seahorses. N a tu re C o m m u n icatio n s 2.
V e r r a e s W . 1977. P ostem bryonic o n to g en y an d fu n c tio n a l anatom y of th e
lig am en tu m m andibulo-hyoideum an d th e lig am en tu m in tero p ercu lo m andibulare, w ith notes o n th e o p ercu lar bones and som e o th e r cranial
elem ents in Salm o g a ir d n e r i R ichardson, 1836 (Teleostei: Salm onidae). Jo u rn al
of M orphology 151:111-119.
VINCENT ACJ, S a d l e r LM. 1995. F a ith fu l p air bonds in w ild seahorses,
H ip p o c a m p u s w hitei. A nim al B ehaviour 50:1557-1569.
VINCENT ACJ, G ile s BG. 2003. C o rrelates of rep roductive success in a w ild
p o p u la tio n of H ip p o c a m p u s w h itei. Jo u rn al of Fish Biology 63:344-355.
VINCENT ACJ, B erg lu n d A, A h n e sjö I. 1995. R eproductive ecology of five
p ipefish species in one eelgrass m eadow. E n v iro n m en tal Biology of Fishes
190
C h a p t e r 8 - R e f e re n c e s
4 4 :3 4 7 -3 6 1 .
VlZZINI S, M azzo la M . 2004. T he tro p h ic s tru c tu re of th e p ip efish co m m u n ity
(Pisces: Syngnathidae) from a w e ste rn M ed iterran e a n seagrass m eadow based
on stable isotope analysis. E stuaries 27:325-333.
V la s s e n b r o e c k J, D ie r ic k M , M a s s c h a e le B, C n u d d e V, V a n H o o re b e k e L,
JACOBS P. 2007. S oftw are tools fo r q u a n tific a tio n of X-ray m icrotom ography
a t th e UGCT. N uclear In stru m e n ts & M eth o d s in Physics R esearch Section
A: A ccelerators, S pectrom eters, D etecto rs an d A ssociated E q u ip m en t 580:442445WALLACE AR. 1889. D arw inism : A n ex position of th e th e o ry of n a tu ra l selection.
W ith som e of its applications. M acM illian, London, p. 494.
W arburton K. 2003. L earning of foraging skills by fish. Fish and Fisheries 4:203215.
WEBB P W , W e ih s D. 1986. F u n c tio n a l lo co m o to r m orphology of early life history
stages in fishes. T ransactions of th e A m erican Fisheries Society 115:115-127.
WERNER EE, GILLIAM JF. 1984. The o n to g en etic niche an d species in teractio n s in
size-stru ctu red populations. A n n u al Review of Ecology, E v o lu tio n and
System atics 15:393-425.
W estn ea t M W . 1990. Feeding m echanics of te le o st fishes (Labridae;
Perciform es): A te st of four-b ar linkage models. Jo u rn al of M orphology
205:269-295.
WESTNEAT M W . 2004. E volution of levers an d linkages in th e feeding
m echanism s of fishes. In tegrativ e and C om parative Biology 44:378-389.
WESTNEAT M W , W a i n w r i g h t PC. 1989. Feeding m echanism of Epibulus
insidator (Labridae; Teleostei): ev o lu tio n of a novel fu n c tio n a l system. Jo u rn al
of M orphology 202:129-150.
WlLEY EO, J o h n s o n DG. 2010. A te le o st classification based o n m onophyletic
groups. In: N elson JS, S chultze H P, W ilso n M V H , editors. O rig in and
p hylogenetic in te rre la tio n sh ip s of teleosts. H o n o rin g G loria A rratia. V erlag
Dr. F riedrich Pfeil, M ünchen, pp. 123-182.
WlLGA CD , MOTTA PJ, Sanford CP. 2007. E v olution an d ecology of feeding in
elasm obranchs. In tegrative and C om parative Biology 47: 55-69.
WILSON N G , R ouse G W . 2010. C o n v erg en t cam ouflage and th e non-m onophyly
of ‘seadragons’ (Syngnathidae: Teleostei): Suggestions fo r a revised taxonom y
of syngnathids. Zoologica S cripta 39:551-558.
WILSON AB, ORR JW . subm itted. The ev o lutionary origins of sy n g n ath id fishes.
Jo u rn al of Fish Biology.
W ilso n AB, V in c e n t A, A h n e s j O I, M eyer A 2001. M ale p regnancy in seahorses
C h a p t e r 8 - R e fe re n c e s
191
an d pipefishes (fam ily Syngnathidae): Rapid diversification of p a tern al bro o d
p o u ch m orphology in fe rred fro m a m olecular phylogeny. The Jo u rn al of
H ered ity 92:159-166.
W ilso n AB, A h n b sj O I, V in c e n t ACJ, M eyer A. 2003. T he dynam ics of male
brooding, m ating p a tte rn s, an d sex roles in pipefishes and seahorses (family
Syngnathidae). E volution 57:1374-1386.
WlNTBRBOTTOM R. 1974. A descriptive synonym y of th e stria te d m uscles of th e
Teleostei. Proceedings of th e A cadem y of N a tu ra l Sciences of Philadelphia
125:225-317.
WOODS CM C . 2003. G ro w th an d survival of th e juvenile seahorse Hippocampus
abdominalis reared on live, fro zen and artificial foods. A q u acu ltu re 220:287298.
ZACK TI, C lavbrib T, P ater SN. 2009. Elastic energy storage in th e m antis
shrim p’s fast p red ato ry strike. The Jo u rn al of E xperim ental Biology 212:40024009.
Z a l o h a r J, H itij T, K r i z n a r M . 2009. Tw o n ew species of seahorses
(Syngnathidae, Hippocampus) fro m th e M iddle M io cen e (Sarm atian)
C opro litic H o riz o n in T unjice H ills, Slovenia: The oldest fossil reco rd of
seahorses. A nnales de P aléontologie 92:71-96.
Z elditch M L, Sw idbrski DL, Sheets H D , F in k W L. 2004. G eom etric
m orphom etries for biologists: A prim er. Elsevier A cadem ic Press, L ondon p.
443-
192
C h a p t e r 8 - R e f e re n c e s
C h a p t e r 8 - R e fe re n c e s
193
Publication list
Roos G, Leysen H , V a n W assen berg h S, H errel A, Jacobs P, D ier ic k M , A erts
P, ADRIAENS D. 2009. Linking m orphology an d m otion: A te st of a four-bar
m echanism in seahorses. Physiological an d B iochem ical Zoology 82( i ):7-19.
V a n W a ssen berg h S, R oos G, G enbrug ge A, Leysen H , A erts P, A d ria en s D,
HERREL A. 2009. S uction is kids play: Fast su ctio n in n ew b o rn seahorses.
Biology L etters 5:200-203.
L eysen H , J o u k P, B r u n a in M , C h r i s t i a e n s J, A d r ia e n s A. 2010. C ran ial
a rc h itec tu re of tu b e -sn o u ted G asterosteiform es (S yn g n a th u s ro stellatu s and
H ip p o c a m p u s capensis). Jo u rn al of M orphology 27i(3):255-270.
L eysen H , C h r i s t i a e n s
ADRIAENS D. in press.
im plications of sn o u t
da ctylio p h o ru s. Jo u rn al
J, D e K e g e l B, B o o n e M N , V a n H o o re b e k e L,
M usculoskeletal stru c tu re of th e feeding system and
elong atio n in H ip p o c a m p u s reidi an d D u n ckeroca m p u s
of Fish Biology,
Roos G, Leysen H , V an W a ssen berg h S, A erts P, H errel A, A d ria en s D.
subm itted. Locking th e pivot-feeding m echanism in Syngnathidae: T he role
of sn o u t adduction. A nim al Biology.
LEYSEN H , R oos G, A dria en s D. sub m itted . M orphological v ariatio n in p ip efish
an d seahorse head shape in re la tio n to sn o u t le n g th and developm ental
grow th. Jo u rn al of M orphology.
Leysen H , D u m o n t ER, Brabant L, V a n H oorebeke L, A d ria en s D. subm itted.
D ealing w ith stress in th e feeding ap p aratu s of seahorses an d pipefishes.
Zoological Jo u rn al of th e L innean Society.
© A l l r ig h t s r e s e r v e d . T h is t h e s i s c o n t a i n s c o n f i d e n t i a l i n f o r m a t i o n a n d c o n f i d e n t i a l r e s e a r c h
r e s u l t s t h a t a re p r o p e r t y t o t h e U G e n t ( R e s e a r c h g r o u p E v o l u t i o n a r y M o r p h o l o g y o f V e r t e b r a t e s ,
D e p a r t m e n t o f B io lo g y ) . T h e c o n t e n t s o f t h i s m a s t e r t h e s i s m a y u n d e r n o c ir c u m s t a n c e s b e m a d e
p u b l i c , n o r c o m p l e t e o r p a r t ia l, w i t h o u t t h e e x p l i c i t a n d p r e c e d i n g p e r m is s i o n o f t h e U G e n t
r e p r e s e n t a t i v e , i.e . t h e s u p e r v is o r . T h e t h e s i s m a y u n d e r n o c ir c u m s t a n c e s b e c o p i e d o r d u p l i c a t e d
i n a n y fo r m , u n l e s s p e r m is s i o n g r a n t e d i n w r i t t e n fo r m . A n y v i o l a t i o n o f t h e c o n f i d e n t i a l n a t u r e
o f t h i s t h e s i s m a y i m p o s e ir r e p a r a b le d a m a g e t o t h e U G e n t . I n c a s e o f a d is p u t e t h a t m a y a r is e
w i t h i n t h e c o n t e x t o f t h i s d e c la r a t io n , t h e J u d ic ia l C o u r t o f G e n t o n l y is c o m p e t e n t t o b e
n o tifie d .
EVOLUTIONARY MORPHOLOGY OF THE
EXTREMELY SPECIALIZED FEEDING
APPARATUS IN SEAHORSES AND
PIPEFISHES (SYNGNATHIDAE)
Part
2
- Figures
Heleen Leysen
Thesis su bm itted to o b ta in th e degree
of D octor in Sciences (Biology)
P ro efsch rift voorgedragen to t h e t
bekom en van de graad van D octor
in de W eten sch ap p en (Biologie)
A cadem iejaar 2010-2011
Rector: Prof. Dr. Paul van C auw enberge
D ecaan: Prof. Dr. H erw ig D ejonghe
Prom otor: Prof. Dr. D om inique A driaens
EVOLUTIONARY MORPHOLOGY OF THE
EXTREMELY SPECIALIZED FEEDING
APPARATUS IN SEAHORSES A N D
PIPEFISHES (SYNGNATHIDAE)
Part 2 - Figures
Heleen Leysen
Thesis subm itted to o b ta in th e degree
of D octor in Sciences (Biology)
P ro efsch rift voorgedragen to t h et
bekom en van de graad van D octor
in de W eten sch ap p en (Biologie)
A cadem iejaar 2010-2011
Rector: Prof. Dr. Paul van C auw enberge
Decaan: Prof. Dr. H erw ig D ejonghe
Prom otor: Prof. Dr. D om inique A driaens
Table
of c o n te n t s
1 I n tr o d u c tio n _______________________________________________________________l
2 M a teria l & m e th o d s_______________________________________________________ 12
2.1 M aterial _________________________________________________________________________12
2.2 M et h o d s _________________________________________________________________________14
3 M o r p h o lo g ic a l d e sc r ip tio n ________________________________________________ 22
3.1 M o r p h o l o g y
3.2 Implic atio ns
o f c h o n d r o - a n d o s t e o c r a n iu m ________________________________
22
o f s n o u t e l o n g a t io n _____________________________________________ 36
4 M o r p h o lo g ic a l v a r ia tio n ___________________________________________________________4 8
5 F u n c tio n a l in te r p r e ta tio n _________________________________________________ 56
5.1 K inematics
o f the feeding apparatus ___________________________________________56
5.2 M echanical
stress distribution in the feeding apparatus _____________________ 70
C h a p te r i - Introduction
i
Osteoglossomorpha
Elopomorpha
Elopocephala
Clupeocephala|
Ostarioclupeomorpha
Euteleostei *
Protacanthopterygii
Stenopterygii
Ateleopodomorpha
Neoteleostei
Cyclosquamata
Eurypterygii
Clenosquamata
B
€
Scopelomorpha
Acanthomorpha °
Lampriomorpha
Polymixiomorpha
Euacanlhomorpha
lolacanthogter^gM
r
Paracanthopterygii
Acanthopteryqii
Mugilomorpha
Atherinomorpha
^ e rc o m o rg h ^ ^
Stephanoberyciformes
Beryciformes
Zeiformes
Gasterosteiformes
Synbranchiformes
Scorpaeniformes
Perciformes
Pleuronectiformes
Tetraodontiformes
F ig . l . l - P h y l o g e n e t i c r e l a t io n s h ip s o f A T e le o s t e i; B E u t e l e o s t e i ; C A c a n t h o m o r p h a ( N e l s o n ,
2 0 0 6).
C h a p te r i - In tro d u c tio n
C h a p te r i - Introduction
Hypoptychoidei
Hypoptychidae
Aulorhynchidae
Gasterosteoidei I
«s1—-cH—Q
Gasterosteidae
Indostomoida
Indostomidae
Pegasoida
-< 3 ^ 3 =
O
Pegasidae
Syngnathoidei
Aulostomoidea
Centriscoidea
Syngnathoida
<3
C
C
r
Syngnathoidea |
Fistulariidae
Aulostomidae
Macroramphosidae
Centriscidae
Solenostomidae
Syngnathidae
Syngnathinae
Hippocampinae
F ig . 1 .2 - P h y l o g e n e t i c r e l a t io n s h ip s o f G a s t e r o s t e i f o r m e s ( K e iv a n y & N e l s o n , 2 0 0 6 ). N o t e t h e
d i v i s i o n i n t o t h r e e s u b o r d e r s : H y p o p t y c h o i d e i , G a s t e r o s t e o i d e i a n d S y n g n a t h o id e i . ( A ll p i c t u r e s
f r o m N e l s o n (2 0 0 6 ) e x c e p t f o r H y p o p t y c h i d a e f r o m F is h B a s e ( F r o e s e & P a u ly , 2 0 1 0 )).
C h a p te r i - In tro d u c tio n
C h a p te r i - Introduction
O p h id iifo rm es
S y n b ra n c h ifo rm e s
Indostom idae
M ugiliform es
B le n n iid a e /G o b ie so c id a e
A th erin ifo rm es
C y p rin o d o n tifo rm e s
B elo n ifo rm es
G obioiidei
Syngnathoidei
D acty lo p tero id ei
P le u ro n e c tifo rm e s
C a ra n g id a e
S c o m b rid a e
P e rc id a e
S c o rp a e n o id e i
C ottoidei
Z o arc o id e i
Hypoptychoidei
G asterosteoidei
P erco id ei
L ophiiform es
T e tra o d o n tifo rm e s
F ig . 1 .3 - M o le c u l a r p h y l o g e n y o f P e r c o m o r p h a p r o p o s e d b y K a w a h a r a e t al. (2 0 0 8 ). T h e c la d e s
b e l o n g i n g t o t h e o r d e r G a s t e r o s t e i f o r m e s sensu K e iv a n y & N e l s o n (2 0 0 6 ) a r e i n b o ld .
C h a p te r i - In tro d u c tio n
C h a p te r i - Introduction
F ig .
1 .4
-
Som e
sp e c ie s
o f th e
f a m i ly
S y n g n a t h id a e
(A -C
s u b f a m i ly
H ip p o c a m p i n a e , D -F
s u b f a m i ly S y n g n a t h in a e ) s h o w i n g t h e m o r p h o l o g i c a l d iv e r s it y . A H i p p o c a m p u s abdom in alis-, B H .
b a rg ib a n ti; C H . histrix ; D
lu m nitzeri.
P h y l l o p t e r y x taeniolatus-, E S i o k u n i c h t h y s nigrolineatu s-, F I d i o t r o p i s c i s
8
C h a p te r i - Introduction
Fig. 1.5 - Legend
u n p r o tec ted eggs
■LflÜ LM w a
i n d iv i d u a l e g g c o m p a r t m e n t s
r u d im e n t a r y p o u c h w i t h p la t e s a n d / o r s k in - f o ld s
e n c l o s e d p o u c h ( s e m i- in v e r t e d )
e n c l o s e d p o u c h (in v e r t e d )
e n c l o s e d p o u c h (e v e r te d )
f u l l y e n c lo s e d p o u c h
C h a p te r i - Introduction
Hippocampus (54)
Syngnathus (32)
Enneacampus (2)
Urocampus (2)
Hippichthys (6)
Festucalex (7)
Pugnaso (1)
Vanacampus (4)
Kaupus (1)
Hypselognathus (2)
Haliichthys (1 )
Urophori
Solegnathus (5)
Phyllopteryx (1 )
Syngnathoides (1 )
Stigmatopora (4)
Corythoichthys (11 )
Microphis (21 )
Doryrhamphus (14)
Maroubra (2)
Gastrophori
Entelurus (1)
-----------
Nerophis (3)
F ig . 1 .5 - P h y l o g e n e t i c r e l a t io n s h ip s o f s o m e s y n g n a t h id g e n e r a w i t h s c h e m a t i c c r o s s s e c t i o n s
t h r o u g h t h e m a le b r o o d p o u c h s h o w i n g t h e i n c r e a s e i n
p o u c h e n c lo s u r e . F ig u r e s b e t w e e n
b r a c k e t s i n d ic a t e t h e n u m b e r o f s p e c i e s c o n s i d e r e d v a h d a c c o r d in g t o F is h b a s e (F r o e s e & P a u ly ,
2 0 1 1 ). G e n e r a b e l o w t h e d a s h e d l i n e a re a b d o m in a l b r o o d e r s (G a s t r o p h o r i) , g e n e r a a b o v e t h e l in e
a r e t a i l b r o o d e r s ( U r o p h o r i) . T h e g e n u s H i p p o c a m p u s b e l o n g s t o t h e s u b f a m i ly H ip p o c a m p i n a e , a ll
o t h e r g e n e r a a re p a r t o f t h e s u b f a m i ly S y n g n a t h in a e . M o d i f i e d f r o m W i l s o n & O r r ( s u b m it t e d ) ,
b a s e d o n c o m p l e t e c y t o c h r o m e b a n d p a r t ia l 1 2 S r D N A a n d 1 6 S r D N A s e q u e n c e d a ta .
10
C h a p te r i - Introduction
Fig. 1.6 - Legend
h
h y o id
n
n e u r o c r a n iu m - s u s p e n s o r i u m c o m p l e x
Pg
u
p e c t o r a l g ir d le
V
v e r t e b r a l c o lu m n
p
c o n n e c t i o n b e t w e e n t h e n e u r o c r a n iu m a n d v e r t e b r a l c o lu m n
Q
c o n n e c t i o n b e t w e e n t h e p e c t o r a l g ir d le a n d t h e n e u r o c r a n iu m
u r o h y a l b o n e a n d ste r n o h y o id e u s m u s c le
m -ep a x
e p a x ia l m u s c le
m -h y p a x
h y p a x ia l m u s c le
m -st
s te r n o h y o id e u s m u s c le
C h a p te r i - Introduction
A
m -ep ax
m -e p a x
P=Q
re a c tio n
fo rc e
m -h y p a x
m -h y p ax
m -s t
B
C
F ig . 1 .6 - T h e p la n a r f o u r - b a r c h a i n a s d e s c r ib e d b y M u l le r ( 1 9 8 7 ). T h e s y s t e m is s t r e s s e d b y t h e
e p a x i a l a n d h y p a x i a l m u s c l e s w h e r e a s t h e s t e r n o h y o i d e u s m u s c l e i s t h o u g h t t o a c t a s a t r ig g e r
t h u s i n i t i a t i n g f a s t h e a d a n d h y o i d m o v e m e n t s . A H e a d o f E n t e l u r u s a e q u o r e u s i n r e s t ( le f t ) a n d
e x p a n d e d (r ig h t) s i t u a t i o n ; B D ia g r a m o f t h e b a r s, j o i n t s a n d m u s c l e s o f t h e fo u r -b a r s y s te m ; C
S c h e m a t i c d r a w in g o f t h e r e s u l t a n t f o r c e s . S c a le b a r = 5 m m .
12
C h a p te r 2 - M a te ria l & m e th o d s
C h a p t e r 2 - Ma t er i a l & m e t h o d s
F ig . 2 .1 - R ig h t l a t e r a l s id e o f t h e h e a d s o f t h e 1 4 s y n g n a t h id s p e c i e s s t u d ie d . P ic t u r e s o f f i x e d
s p e c i m e n s s c a le d b a s e d o n b r a in c a s e l e n g t h a n d s o r t e d a c c o r d in g t o s n o u t l e n g t h (A -G p i p e f i s h e s ,
H - N s e a h o r s e s ) . A D o r y r h a m p h u s d a c t y l i o p h o r u s ; B D u n c k e r o c a m p u s p e ssu life ru s ; C S y n g n a t h u s acus;
D
D o r y r h a m p h u s j a n s s i ; E S y n g n a t h u s r o s te lla tu s ; F C o r y t h o i c h t h y s in t e s ti n a l i s ; G D o r y r h a m p h u s
m e l a n o p l e u r a ; H H i p p o c a m p u s b a r b o u r i; I H i p p o c a m p u s k u d a ; J H i p p o c a m p u s reidi; K H i p p o c a m p u s
bre v ic ep s ; L H i p p o c a m p u s a b d o m in a li s ; M H i p p o c a m p u s c ap en sis; N H i p p o c a m p u s z o s te r a e . S c a le b a r =
5 mm.
C h a p t e r 2 - Ma t er i a l & m e t h o d s
14
Fig. 2.2 - Legend
HL
h e a d le n g t h ; f r o m t i p o f u p p e r j a w t o b a s e o f c le i t h r a l c u r v a tu r e
SL
s ta n d a r d le n g t h ; b e t w e e n t i p o f s n o u t a n d b a s e o f c a u d a l f i n
SnL
s n o u t le n g t h ; b e t w e e n t i p o f u p p e r j a w a n d c a u d a l m a r g in o f n o s t r i l
T aL
t a i l le n g t h ; f r o m m i d - p o i n t o f l a s t t r u n k r in g t o t a i l t i p
T rL
t r u n k le n g t h ; f r o m b a s e o f c le i t h r a l c u r v a t u r e t o m i d - p o i n t o f l a s t t r u n k r in g
C h a p t e r 2 - Ma t er i a l & m e t h o d s
F ig . 2 .2 - M e a s u r e m e n t s a s p e r f o r m e d o n A a p i p e f i s h ( M i c r o p h i s b r a c h y u r u s a c u l e a t u s ) , a n d B a
s e a h o r s e (H i p p o c a m p u s reidi). S e e t e x t f o r a d e f i n i t i o n o f t h e l e n g t h s . S c a le b a r = i o m m .
i6
C h a p te r 2 - M a te ria l & m e th o d s
C h a p t e r 2 - Ma t er i a l & m e t h o d s
F ig . 2 .3 - R ig h t l a t e r a l s id e o f t h e h e a d o f a M i c r o p h i s b r a c h y u r u s a c u l e a t u s s p e c i m e n s h o w i n g t h e
15 la n d m a r k s u s e d f o r t h e g e o m e t r ic m o r p h o m e t r i c a l a n a ly s is . T h e t h i n b la c k l i n e , c o n n e c t i n g
la n d m a r k 1 w i t h la n d m a r k 12, m e a s u r e s h e a d l e n g t h a n d w a s u s e d t o d e f i n e la n d m a r k s 1 0 a n d 11.
S e e t e x t f o r a n a n a t o m i c a l d e s c r ip t io n o f t h e la n d m a r k s . S c a le b a r = 2 m m .
i8
C h a p te r 2 - M a te ria l & m e th o d s
C h a p t e r 2 - Ma t er i a l & m e t h o d s
5,0
4,5
4,0
y = 0,0399x + 2,2357
R2 = 0,8636
3,5
_l
X
3,0
2,5
2,0
0
10
20
30
40
50
60
70
age
F ig . 2 .4 - R e l a t io n b e t w e e n a g e a n d h e a d l e n g t h (H L ) o f t e n H i p p o c a m p u s r e i d i j u v e n i l e s w i t h
lin e a r r e g r e s s i o n t r e n d l in e a n d e q u a t i o n . C o r r e l a t io n c o e f f i c i e n t i s 0 .9 3 .
20
C h a p te r 2 - M a te ria l & m e th o d s
C h a p t e r 2 - Ma t er i a l & m e t h o d s
F ig . 2 .5 - H i p p o c a m p u s r e i d i m o d e l f o r t h e f i n i t e e l e m e n t a n a ly s is s h o w i n g t h e a p p l ie d p r e s s u r e
( b lu e ) , t h e c o n s t r a i n t s a t t h e o c c i p i t a l p a r t o f t h e s k u l l (r ed ) a n d t h e c y li n d r i c a l c o o r d in a t e s y s t e m
w i t h t h e o r ig i n a t t h e t i p o f t h e s n o u t ( g r e e n ). A r ig h t la t e r a l v ie w ; B r o s t r a l v i e w .
22
C h a p t e r 3 - M o r p h o l o g i c a l d e s c r i p t i on
Fig. 3.1 - Legend
bh
b a s ih y a l c a r tila g e
ch
c e r a t o h y a l c a r t ila g e
c-M e ck
M e c k e l ’s c a r t ila g e
c -ro st
r o s t r a l c a r t ila g e
e t -p
e t h m o i d p la t e
fn -h y p
fe n e str a h y p o p h y se a
f-o n
o r b ito n a s a l fo r a m e n
hh
h y p o h y a l c a r tila g e
hs
h y o s y m p l e c t i c c a r t ila g e
ih
i n t e r h y a l c a r t ila g e
lm -o n
o r b i t o n a s a l l a m in a
o t-ca p
o t i c c a p s u le
P-P
p a l a t i n e p a r t o f p a la t o q u a d r a t e c a r t ila g e
p -q
p t e r y g o q u a d r a t e p a r t o f p a la t o q u a d r a t e c a r t ila g e
p l-o c
p ila o c c ip ita lis
s - in
in te r n a s a l se p tu m
tn -m
t a e n i a e m a r g in a lis
tn -t-m -p
t a e n i a e t e c t i m e d i a li s p o s t e r io r
tr -c
t r a b e c u la c o m m u n i s
tt-p
t e c t u m p o s te r iu s
tt-sy n
t e c t u m s y n o tic u m
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
tt-p
tt-syn
tn-m tn-t-m-p
f-on
s-in et-p p-p c-rost
p-q
tt-syn
B
tt-p ot-cap
ot-cap
C
tn-t-m-p
tn-m
tr-c
ih
fn-hyp
Im-on
hh
ch
hs
s-in
tr-c
p-q
et-p
hs
bh
c-M eck
c-rost
p-p
c-Meck
p-p
p-q
c-Meck
F ig . 3 .1 - 3 D - r e c o n s t r u c t i o n o f t h e j u v e n i l e c h o n d r o c r a n i u m o f S y n g n a t h u s r o s te l la t u s (1 3 .1 m m
SL ) b a s e d o n h i s t o l o g i c a l s e c t i o n s . A l a t e r a l v i e w o f t h e r ig h t sid e ; B d o r s a l v ie w ; C v e n t r a l v i e w .
S c a le b a r = 0 .2 m m .
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
24
F ig . 3 .2
- Legend
bh
b a s ih y a l c a r tila g e
c-M e ck
M e c k e l ’s c a r t ila g e
ch
c e r a t o h y a l c a r t ila g e
e t -p
e t h m o i d p la t e
hh
h y p o h y a l c a r tila g e
hs
h y o s y m p l e c t i c c a r t ila g e
ih
i n t e r h y a l c a r t ila g e
lm -o n
o r b i t o n a s a l l a m in a
o -a n g
a n g u lo a r t ic u la r b o n e
o -a p a l
a u t o p a la t i n e b o n e
o -b h
b a s ih y a l b o n e
o -c h
c er a to h y a l b o n e
o -d e n
d e n ta r y b o n e
o -e cp
e c to p te r y g o id b o n e
o -h h
hypohyal bone
o -h m
h y o m a n d i b u la r b o n e
o -ih
in te r h y a l b o n e
o -m eth
m e s e th m o id b o n e
o -m p
m e ta p te r y g o id b o n e
o -m x
m a x illa r y b o n e
o -p a r
p a r ie t a l b o n e
o -p a r a
p a r a s p h e n o id b o n e
o -p o p
p r e o p e r c u la r b o n e
o -q
q u a d r a te b o n e
o -r a r t
r e t r o a r t ic u la r b o n e
o -sy m
s y m p le c tic b o n e
o t-ca p
o t i c c a p s u le
p -p
p a l a t i n e p a r t o f p a la t o q u a d r a t e c a r t ila g e
p -q
p t e r y g o q u a d r a t e p a r t o f p a la t o q u a d r a t e c a r t ila g e
s - in
in te r n a s a l se p tu m .
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
o -m e th
»
3?
F ig . 3 .2 - H i s t o l o g i c a l c r o s s s e c t i o n s o f t h e j u v e n i l e c r a n iu m o f S y n g n a t h u s r o s t e l l a t u s (1 3 .1 m m
S L ). A d o r s o r o s t r a l p a r t o f t h e s n o u t a t t h e l e v e l o f t h e a u t o p a la t i n e b o n e ; B l o w e r ja w ; C d o r s a l
p a r t o f e t h m o i d r e g io n ; D i n t e r n a s a l r e g io n ; E h y o id ; F r ig h t h y o i d - s u s p e n s o r i u m a r t ic u la t io n .
S c a le b a r s = 5 0 p m .
26
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
Pig- 3-3 - Legend
bh
b a s ih y a l c a r tila g e
c-M e ck
M e c k e l ’s c a r t ila g e
c -ro st
r o s t r a l c a r t ila g e
ch
c e r a t o h y a l c a r t ila g e
c p -a
a n t e r io r c o p u la
e t -p
e t h m o i d p la t e
hh
h y p o h y a l c a r tila g e
hs
h y o s y m p l e c t i c c a r t ila g e
ih
i n t e r h y a l c a r t ila g e
lm -o n
o r b i t o n a s a l l a m in a
o t-ca p
o t i c c a p s u le
p -p
p a l a t i n e p a r t o f p a la t o q u a d r a t e c a r t ila g e
p -q
p t e r y g o q u a d r a t e p a r t o f p a la t o q u a d r a t e c a r t ila g e
p l-o c
p ila o c c ip ita lis
r -s - p t
s p h e n o - p t e r o t i c r id g e
s - in
in te r n a s a l se p tu m
tn -m
t a e n i a m a r g in a lis
tn -t-m -p
t a e n i a t e c t i m e d i a li s p o s t e r i o r
tr -c
t r a b e c u la c o m m u n i s
tt-sy n
t e c t u m s y n o t ic u m .
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
pl-oc o t-cap
A
r-s-pt
tt-syn
cp-a
tn-t-m -p
ch
tt-syn
B
ot-cap
hh ih bh h s Im-on p-q
tn-m
tn-t-m -p
ot-cap cp -a ih
C
ch
hh
tn-m
tr-c
Im-on
tr-c
bh
h s p-q
s-in
c-M eck
et-p
p-p c-rost
et-p
hs
p-q
c-M eck
c-M eck
p-p
P ig - 3-3 " 3 D - r e c o n s t r u c t i o n o f t h e j u v e n i l e c h o n d r o c r a n i u m o f H i p p o c a m p u s c a p e n s i s (1 2 .8 m m
S L ), w i t h h y o i d d e p r e s s e d , b a s e d o n h i s t o l o g i c a l s e c t i o n s . A la t e r a l v i e w o f t h e r ig h t sid e ; B d o r s a l
v ie w ; C v e n t r a l v i e w . S c a le b a r = 0 .2 m m .
28
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
Fig. 3.4 - Legend
bh
b a s ih y a l c a r tila g e
c-M e ck
M e c k e l ’s c a r t ila g e
ch
c e r a t o h y a l c a r t ila g e
e t -p
e t h m o i d p la t e
hh
h y p o h y a l c a r tila g e
hs
h y o s y m p l e c t i c c a r t ila g e
ih
i n t e r h y a l c a r t ila g e
lm -o n
o r b i t o n a s a l l a m in a
o -a n g
a n g u lo a r t ic u la r b o n e
o -a p a l
a u t o p a la t i n e b o n e
o -c h -p
p o s te r io r c e r a to h y a l b o n e
o -e cp
e c to p te r y g o id b o n e
o -h h
hypohyal bone
o -h m
h y o m a n d i b u la r b o n e
o -ih
in te r h y a l b o n e
o -le th
la t e r a l e t h m o i d b o n e
o -m eth
m e s e th m o id b o n e
o -m p
m e ta p te r y g o id b o n e
o -p a r
p a r ie t a l b o n e
o -p a r a
p a r a s p h e n o id b o n e
o -p o p
p r e o p e r c u la r b o n e
o -q
q u a d r a te b o n e
o -r a r t
r e t r o a r t ic u la r b o n e
o -sy m
s y m p le c tic b o n e
o -v m
vom eral bon e
p -p
p a l a t i n e p a r t o f p a la t o q u a d r a t e c a r t ila g e
p -q
p t e r y g o q u a d r a t e p a r t o f p a la t o q u a d r a t e c a r t ila g e
s - in
in te r n a s a l se p tu m
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
o-meth
F ig . 3.4 - H i s t o l o g i c a l c r o s s s e c t i o n s o f t h e j u v e n i l e c r a n iu m o f H i p p o c a m p u s c a p e n s i s (12.8 m m
S L ). A d o r s o r o s t r a l p a r t o f t h e s n o u t a t t h e l e v e l o f t h e a u t o p a la t i n e b o n e ; B d o r s a l p a r t o f
e t h m o i d r e g io n ; C l e f t l o w e r ja w ; D i n t e r n a s a l r e g io n ; E p a r t o f r ig h t S u s p e n s o r iu m ; F h y o id s u s p e n s o r iu m a r t ic u l a t i o n . S c a le b a r s = 5 0 p m .
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
30
Pig- 3-5 - Legend
c -ro st
r o s t r a l c a r t ila g e
o -a n g
a n g u lo a r t ic u la r b o n e
o - a n t - la c
a n to r b ito la c r im a l b o n e
o -a p a l
a u t o p a la t i n e b o n e
o -c h -a
a n t e r io r c e r a t o h y a l b o n e
o -c h -p
p o s te r io r c e r a to h y a l b o n e
o -c l
c le i t h r u m
o -d e n
d e n ta r y b o n e
o -e cp
e c to p te r y g o id b o n e
o -h m
h y o m a n d i b u la r b o n e
o -ih
in te r h y a l b o n e
o -io -II
s e c o n d in f r a o r b it a l b o n e
o -io p
in t e r o p e r c u la r b o n e
o -le th
la t e r a l e t h m o i d b o n e
o -m eth
m e s e th m o id b o n e
o -m p
m e ta p te r y g o id b o n e
o -m x
m a x illa r y b o n e
o -o p
o p e r c u la r b o n e
o -p a r
p a r ie t a l b o n e
o -p a r a
p a r a s p h e n o id b o n e
o -p o p
p r e o p e r c u la r b o n e
o -p o stp
p o s t p a r ie t a l b o n e
o -p o stt
p o sttem p o ra l b o n e
o -p r m x
p r e m a x illa r y b o n e
o -p t
p te r o tic b o n e
o -q
q u a d r a te b o n e
o -r a r t
r e t r o a r t ic u la r b o n e
o -so c
s u p r a o c c ip ita l b o n e
o -sp h
s p h e n o tic b o n e
o -sp o p
s u p r a p r e o p e r c u la r b o n e
o -sy m
s y m p le c tic b o n e
o -u h
urohyal bone
o -v m
vom eral bon e
r -b r
b r a n c h io s te g a ! ray
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
o-cl
o-postt o-postp o-soc o-spop o-sph o-par o-leth
o-melh o-ant-lac
o-vm
o-mp
o-ecp
o-apal
r-br
o-para o-sym o-io-ll
o-cl
o-postt o-postp o-spop o-par
o-pop
o-leth
o-q
o-meth
o-rart
o-mp o-q
o-den
o-prmx
o-apal
o-op o-soc o-pt
r-br
o-sph
o-ant-lac
o-io-ll
o-lh o-ch-a
o-lo-ll
o-vm o-ecp
o-q
o-ang
o-prmx
~~nr>---
—V
o-mx
o-op
o-ecp
o-rart
o-apal
o-ang
o-mx
o-uh
o-ch-p
o-pop
o-iop
o-prmx
o-den
o-vm
E
o-ecp
o-rart
o-ang
c-rost
o-apal o-mx o-prmx
o-den
P ig - 3-5 " A d u l t o s t e o c r a n i u m o f S y n g n a t h u s r o s te l la t u s (1 1 1 .1 m m S L ). A la t e r a l v i e w o f t h e r ig h t
s id e ; B d o r s a l v ie w ; C v e n t r a l v ie w ; D l a t e r a l v i e w o f s n o u t tip ; E d o r s a l v i e w o f s n o u t t i p . S c a le
bars = 1 m m .
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
32
F ig . 3 .6
- Legend
c -ro st
r o s t r a l c a r t ila g e
cr
coron a
o -a n g
a n g u lo a r t ic u la r b o n e
o -a n t
a n to r b ita l b o n e
o -a p a l
a u t o p a la t i n e b o n e
o -c h -a
a n t e r io r c e r a t o h y a l b o n e
o -c h -p
p o s te r io r c e r a to h y a l b o n e
o -c l
c le i t h r u m
o -d e n
d e n ta r y b o n e
o -e cp
e c to p te r y g o id b o n e
o -e p o c
e p io c c ip ita l b o n e
o -h m
h y o m a n d i b u la r b o n e
o -ih
in te r h y a l b o n e
o -io -II
s e c o n d in f r a o r b it a l b o n e
o -io p
in t e r o p e r c u la r b o n e
o -le th
la t e r a l e t h m o i d b o n e
o - la c
la c r im a l b o n e
o -m eth
m e s e th m o id b o n e
o -m p
m e ta p te r y g o id b o n e
o -m x
m a x illa r y b o n e
o -o p
o p e r c u la r b o n e
o -p a r
p a r ie t a l b o n e
o -p a r a
p a r a s p h e n o id b o n e
o -p o p
p r e o p e r c u la r b o n e
o -p o stt
p o sttem p o ra l b o n e
o -p r m x
p r e m a x illa r y b o n e
o -p t
p te r o tic b o n e
o -q
q u a d r a te b o n e
o -r a r t
r e t r o a r t ic u la r b o n e
o -so c
s u p r a o c c ip ita l b o n e
o -so p
s u b o p e r c u la r b o n e
o -sp h
s p h e n o tic b o n e
o -sy m
s y m p le c tic b o n e
o -u h
urohyal bone
o -v m
vom eral bon e
r -b r
b r a n c h io s te g a ! ray
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
o-cl
cr
o-postt o -so c o -ep o c o-pt
o -sp h o-par
o -p ara o-leth o-m eth
o-vm o-m p o -ecp
o-apal
o-mx
o-prm x
o-den
o-ang
r-br
r-br
o -so p
o -op
o-op
o-hm o-ih
o-pop o-sym o-io-ll o-iop o-lac o-anl o-q o-rart
o-postt o -ep o c o-sph
o-leth
o-lac
o-vm o-m p o -den
c-rost
o-prm x
o-apal
o -d
o -so c o-pt
o-op
o-ih o-uh
o-io-ll o-m eth o -an t o-q o-ecp
o-pop
o-ang
o-io-ll
_
r-br
o-vm
o-ch-p
o-ch-a
o-iop
o-rart
o -ecp o-apal
o-ecp
o-•rart
o-m x
o-prm x
o-ang
F ig . 3 .6 - A d u l t o s t e o c r a n i u m o f H i p p o c a m p u s c a p e n s is (9 6 .1 m m S L ). A l a t e r a l v i e w o f t h e r ig h t
s id e ; B d o r s a l v ie w ; C v e n t r a l v ie w ; D l a t e r a l v i e w o f s n o u t tip ; E d o r s a l v i e w o f s n o u t t i p . S c a le
bars = 1 m m .
34
C h a p te r 3 - M o rp h o lo g ic a l d e s c r ip tio n
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
B
D
P ig - 3-7 - C r a n ia , s c a le d t o s a m e h e a d l e n g t h ( f r o m f r o n t o f e t h m o i d p l a t e ( j u v e n ile s ) o r v o m e r a l
b o n e (a d u lts ) t o b a c k o f o c c i p i t a l r e g io n ) , o f j u v e n i l e s ( b e ig e ) a n d a d u lt s (g r e y ). A
G a s t e r o s t e u s a c u l e a t u s (1 6 .0 m m
SL ) ( a f t e r
S w in n e r t o n , 1 9 0 2 ); B
a d u lt
j u v e n ile
G. a c u l e a t u s (S L n o t
m e n t i o n e d ) ( a f t e r A n k e r , 1 9 7 4 ); C j u v e n i l e c h o n d r o c r a n i u m S y n g n a t h u s r o s te l la t u s (1 3 .1 m m SL ); D
a d u lt S. r o s te l la t u s (1 1 1 .1 m m SL); E j u v e n i l e c h o n d r o c r a n i u m H i p p o c a m p u s c a p e n s i s (1 2 .8 m m SL);
F a d u lt H . c a p e n s i s (9 6 .1 m m SL ). B la c k a r r o w s i n d ic a t e r o s t r a l b o r d e r o f o r b it a , w h i t e a r r o w s
i n d ic a t e f r o n t o f t h e a u t o p a la t i n e b o n e . A r e a i n b e t w e e n t h e a r r o w s c o r r e s p o n d s t o t h e e t h m o i d
r e g i o n . B a r s s h o w l e n g t h o f e t h m o i d r e g i o n r e la t iv e t o h e a d l e n g t h o f j u v e n i l e s ( b e ig e ) a n d a d u lt s
(g r e y ). N o t e t h e e l o n g a t i o n i n b o t h s y n g n a t h id s p e c ie s .
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
36
Fig. 3.8 - Legend
cr
coron a
c -ro st
r o s t r a l c a r t ila g e
1 -c h -u h
c e r a t o h y a l - u r o h y a l l ig a m e n t
1 -io p -h
i n t e r o p e r c u l o - h y o i d l ig a m e n t
1 -m n d -h
m a n d ib u l o - h y o i d l ig a m e n t
1 -m n d -io p
m a n d ib u l o - i n t e r o p e r c u la r l ig a m e n t
m -A 2
s e c o n d b u n d le o f t h e a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x
m -a a p
a d d u c to r a rcu s p a la tin i m u s c le
m -ep a x
e p a x ia l m u s c le
m -h - a b d
a b d u c to r h y o h y o id e u s m u s c le
m - im
in t e r m a n d i b u la r is m u s c l e
m -la p
le v a t o r a r c u s p a l a t i n i m u s c l e
m -p r - h
p r o tr a c to r h y o id e i m u s c le
m -st
s te r n o h y o id e u s m u s c le
o -a n g
a n g u lo a r t ic u la r b o n e
o -a n t
a n to r b ita l b o n e
o -a p a l
a u t o p a la t i n e b o n e
o -c h -a
a n t e r io r c e r a t o h y a l b o n e
o -c h -p
p o s te r io r c e r a to h y a l b o n e
o -c l
c le i t h r u m
o -d e n
d e n ta r y b o n e
o -e cp
e c to p te r y g o id b o n e
o -e p o c
e p io c c ip ita l b o n e
o -e x o c
e x o c c ip ita l b o n e
o -h m
h y o m a n d i b u la r b o n e
o -ih
in te r h y a l b o n e
o -io -II
s e c o n d in f r a o r b it a l b o n e
o -io p
in t e r o p e r c u la r b o n e
o - la c
la c r im a l b o n e
o -le th
la t e r a l e t h m o i d b o n e
o -m eth
m e s e th m o id b o n e
o -m p
m e ta p te r y g o id b o n e
o -m x
m a x illa r y b o n e
o -o p
o p e r c u la r b o n e
o -p a r
p a r ie t a l b o n e
o -p a r a
p a r a s p h e n o id b o n e
o -p o p
p r e o p e r c u la r b o n e
o -p o stt
p o sttem p o ra l b o n e
o -p r m x
p r e m a x illa r y b o n e
o -p t
p te r o tic b o n e
o -q
q u a d r a te b o n e
o -r a r t
r e t r o a r t ic u la r b o n e
o -so c
s u p r a o c c ip ita l b o n e
o -so p
s u b o p e r c u la r b o n e
o -sp h
s p h e n o tic b o n e
o -u h
urohyal bone
o -v m
vom eral bon e
r -b r
b r a n c h io s te g a l ray
t-e p a x
e p a x ia l te n d o n
t-p r -h
p r o tr a c to r h y o id e i te n d o n
t-st
s te r n o h y o id e u s te n d o n
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
o-prmx
o-mx
B
o-ecp o-ant o-q
c-rost o-apal o-vm o-mp
o-den
F ig . 3 .8 -
o-lac o-io-ll
o-meth
o-rart l-mnd-iop l-mnd-h
o-leth
m-A2
o-pt
m-aap o-pop o-par m-lap o-sph
o-pop t-pr-h o-ch-a l-ch-uh o-uh t-st
3 D - r e c o n s t r u c t i o n o f t h e c r a n iu m
o-epoc
o-soc
o-postt o-op cr o-cl
o-exoc
t-epax m-epax
m-st
o f H i p p o c a m p u s re id i ( 1 0 3 .5 m m
o-cl
SL ) b a s e d o n
h i s t o l o g i c a l s e c t i o n s . A la t e r a l v i e w o f t h e l e f t sid e ; B d o r s a l v ie w ; C v e n t r a l v i e w . B e ig e i n d ic a t e s
b o n e , k h a k i i n d ic a t e s c a r t ila g e , o r a n g e i n d ic a t e s t e n d o n a n d l ig a m e n t , r e d i n d ic a t e s m u s c le . S c a le
bar = 2 m m .
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
38
Fig. 3.9 - Legend
c -ro st
r o s t r a l c a r t ila g e
l- c h - u h
c e r a t o h y a l - u r o h y a l l ig a m e n t
1- i o p - h
i n t e r o p e r c u l o - h y o i d l ig a m e n t
l-m n d -h
m a n d ib u l o - h y o i d l ig a m e n t
l- m n d - i o p
m a n d ib u l o - i n t e r o p e r c u la r l ig a m e n t
m -A i
f i r s t b u n d le o f t h e a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x
m -A 2
s e c o n d b u n d le o f t h e a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x
m -A 3
t h i r d b u n d le o f t h e a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x
m -a a p
a d d u c to r a rcu s p a la tin i m u s c le
m -a d d -o p
a d d u c t o r o p e r c u li m u s c l e
m -d o d
d i l a t a t o r o p e r c u li d o r s a h s m u s c l e
m -ep a x
e p a x ia l m u s c le
m -h - a b d
a b d u c to r h y o h y o id e u s m u s c le
m -la p
le v a t o r a r c u s p a l a t i n i m u s c l e
m - l- o p
le v a t o r o p e r c u h m u s c l e
m -p r - h
p r o tr a c to r h y o id e i m u s c le
m -s -c
s u p r a c a r in a lis m u s c l e
m -st
s te r n o h y o id e u s m u s c le
o -a n g
a n g u lo a r t ic u la r b o n e
o -a n t
a n to r b ita l b o n e
o -a p a l
a u t o p a la t i n e b o n e
o -b o c
b a s io c c ip ita l b o n e
o -c h -a
a n t e r io r c e r a t o h y a l b o n e
o -c h -p
p o s te r io r c e r a to h y a l b o n e
o -d e n
d e n ta r y b o n e
o -e cp
e c to p te r y g o id b o n e
o -h m
h y o m a n d i b u la r b o n e
o -io p
in t e r o p e r c u la r b o n e
o -m p
m e ta p te r y g o id b o n e
o -m x
m a x illa r y b o n e
o -p r m x
p r e m a x illa r y b o n e
o -p t
p te r o tic b o n e
o -q
q u a d r a te b o n e
o -r a r t
r e t r o a r t ic u la r b o n e
o -so c
s u p r a o c c ip ita l b o n e
o -sy m
s y m p le c tic b o n e
o -u h
urohyal bone
o -v m
vom eral bon e
p r -c
c o r o n o id p r o c e ss
r -b r
b r a n c h io s te g a l ray
t-A
a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x t e n d o n
t-e p a x
e p a x ia l te n d o n
t-p r -h
p r o tr a c to r h y o id e i te n d o n
t-st
s te r n o h y o id e u s te n d o n
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
o-prmx
P ig - 3-9 " 3 D - r e c o n s t r u c t i o n o f t h e c r a n iu m
o f H i p p o c a m p u s re i d i (1 0 3 .5 m m
SL ) b a s e d o n
h i s t o l o g i c a l s e c t i o n s . A la t e r a l v i e w o f t h e l e f t s id e w i t h s e v e r a l b o n e s r e m o v e d t o s h o w m u s c l e s
u n d e r n e a t h ; B la t e r a l v i e w o f t h e l e f t s id e o f t h e s n o u t t i p . B e ig e i n d ic a t e s b o n e , k h a k i i n d ic a t e s
c a r t ila g e , o r a n g e i n d ic a t e s t e n d o n a n d h g a m e n t , r e d i n d i c a t e s m u s c le . S c a le b a r A = 2 m m ; s c a le
bar B = 1 m m .
40
F ig . 3 .1 0
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
- Legend
l-m n d -h
m a n d ib u l o - h y o i d l ig a m e n t
m -ep a x
e p a x ia l m u s c le
m -p r - h
p r o tr a c to r h y o id e i m u s c le
m -s -c
s u p r a c a r in a lis m u s c l e
m -st
s te r n o h y o id e u s m u s c le
o -a n t
a n to r b ita l b o n e
o -e p a x
s e s a m o i d b o n e i n e p a x ia l t e n d o n
o -io -II
s e c o n d in f r a o r b it a l b o n e
o -io p
in t e r o p e r c u la r b o n e
o -m p
m e ta p te r y g o id b o n e
o -p o p
p r e o p e r c u la r b o n e
o -p a r a
p a r a s p h e n o id b o n e
o -q
q u a d r a te b o n e
o -so c
s u p r a o c c ip ita l b o n e
o -v m
vom eral bon e
t-A
a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x t e n d o n
t-e p a x
e p a x ia l te n d o n
t-st
s te r n o h y o id e u s te n d o n
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
t-epax
m -s-c
t-epax
m-epax
F ig . 3 .1 0 - H i s t o l o g i c a l c r o s s s e c t i o n s o f t h e c r a n iu m o f H i p p o c a m p u s re id i (1 0 3 .5 m m SL ) (A -D )
a n d D o r y r h a m p h u s d a c t y l i o p h o r u s (9 1 .2 m m SL ) (E -H ) s t a i n e d w i t h t o l u i d i n e b lu e . A ,E t h e n a s a l
r e g i o n s h o w i n g t h e m a n d ib u l o - h y o i d h g a m e n t a n d t h e p r o t r a c t o r h y o i d e i m u s c l e o r g a n is a t io n ;
B ,F t h e s n o u t r o s t r a lly w i t h t h e a d d u c t o r m a n d ib u l a e t e n d o n ; C ,G t h e s t e r n o h y o i d e u s m u s c l e a n d
it s te n d o n ; D ,H t h e o c c ip it a l r e g io n w i t h t h e su p r a o c c ip ita l b o n e a n d its a ss o c ia te d t e n d o n s a n d
m u s c l e s , n o t e t h e s e s a m o i d b o n e i n D . d a c t y l i o p h o r u s . S c a le b a r s A - D = 5 0 0 pm ; E - H = 2 0 0 p m .
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
42
Fig. 3.11 - Legend
c-rost
1-epax
1-iop-h
l-m n d -h
l-m nd -iop
1-prim
m -A 2
m -aap
m -dod
m -epax
m -h-abd
m -im
m -lap
m -p r-h
m-s-c
m -st
o-ang
o-apal
o -b h
o-ch-a
o-ch-p
o-cl
o -d en
o-ecp
o-epax
o-epoc
o -ih
o-io-II
o-iop
o-lac
o -le th
o -m eth
o-m p
o-m x
o-op
o-par
o-para
o-pop
o -p o stt
o -prm x
o -p t
o-q
o -ra rt
o-soc
o-sop
o-sph
o-spop
o-sym
o -u h
o-vm
r-br
t-ep ax
t-h -ab d
t-p r-h
t-st
ro s tra l cartilag e
ep ax ia l lig am e n t
in te ro p e rc u lo -h y o id lig am en t
m a n d ib u lo -h y o id lig am en t
m a n d ib u lo -in te ro p e rc u la r lig a m e n t
p rim o rd ia l lig a m e n t
seco n d b u n d le o f th e a d d u c to r m an d ib u la e m uscle co m p lex
a d d u c to r arcus p a la tin i m uscle
d ila ta to r o p e rc u li dorsalis m uscle
ep ax ia l m uscle
a b d u c to r h y o h y o id eu s m uscle
in te rm a n d ib u la ris m uscle
le v a to r arcus p a la tin i m uscle
p ro tra c to r h y o id ei m uscle
su p ra ca rin a lis m uscle
ste rn o h y o id e u s m uscle
a n g u lo a rtic u la r b o n e
a u to p a la tin e b o n e
b a sih y al b o n e
a n te rio r ce ra to h y a l b o n e
p o s te rio r c e ra to h y al b o n e
c le ith ru m
d e n ta ry b o n e
e cto p te ry g o id b o n e
sesam o id b o n e in ep ax ial te n d o n
e p io c c ip ita l b o n e
in te rh y a l b o n e
seco n d in fra o rb ita l b o n e
in te ro p e rc u la r b o n e
lac rim a l b o n e
la te ra l e th m o id b o n e
m e se th m o id b o n e
m e ta p te ry g o id b o n e
m ax illary b o n e
o p e rc u la r b o n e
p a rie ta l b o n e
p a ra sp h e n o id b o n e
p re o p e rc u la r b o n e
p o stte m p o ra l b o n e
p rem a x illa ry b o n e
p te ro tic b o n e
q u a d ra te b o n e
re tro a rtic u la r b o n e
su p ra o c c ip ita l b o n e
su b o p e rc u la r b o n e
sp h e n o tic b o n e
su p ra p re o p e rc u la r b o n e
sy m p lectic b o n e
u ro h y a l b o n e
v o m e ra l b o n e
b ra n c h io ste g a l ray
ep ax ia l te n d o n
h y o h y o id eu s a b d u c to r te n d o n
p ro tra c to r h y o id ei te n d o n
ste rn o h y o id e u s te n d o n
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
F ig . 3 .1 1 - 3 D - r e c o n s t r u c t i o n o f t h e c r a n iu m o f D o r y r h a m p h u s d a c t y l i o p h o r u s (9 1 .2 m m SL ) b a s e d
o n h i s t o l o g i c a l s e c t i o n s . A l a t e r a l v i e w o f t h e l e f t sid e ; B d o r s a l v ie w ; C v e n t r a l v i e w . B e ig e
i n d i c a t e s b o n e , k h a k i i n d ic a t e s c a r t ila g e , o r a n g e i n d i c a t e s t e n d o n a n d h g a m e n t , r e d i n d ic a t e s
m u s c le . S c a le b a r = 2 m m .
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
44
Fig. 3.12 - Legend
l- c h - u h
c e r a t o h y a l - u r o h y a l l ig a m e n t
1-e p a x
1- i o p - h
i n t e r o p e r c u l o - h y o i d l ig a m e n t
l-m n d -h
m a n d ib u l o - h y o i d l ig a m e n t
l- m n d - i o p
m a n d ib u l o - i n t e r o p e r c u la r l ig a m e n t
1-p r im
p r im o r d ia l l ig a m e n t
e p a x ia l h g a m e n t
m -A 2
s e c o n d b u n d le o f t h e a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x
m -A 3
t h i r d b u n d le o f t h e a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x
m -a a p
a d d u c to r a rcu s p a la tin i m u s c le
m -a d d -o p
a d d u c t o r o p e r c u li m u s c l e
m -d o d
d i l a t a t o r o p e r c u li d o r s a h s m u s c l e
m -d o v
d i l a t a t o r o p e r c u li v e n t r a li s m u s c l e
m -ep a x
e p a x ia l m u s c le
m -h - a b d
a b d u c to r h y o h y o id e u s m u s c le
m -la p
le v a t o r a r c u s p a l a t i n i m u s c l e
m -p r - h
p r o tr a c to r h y o id e i m u s c le
m -s -c
s u p r a c a r in a lis m u s c l e
m -st
s te r n o h y o id e u s m u s c le
o -a n g
a n g u lo a r t ic u la r b o n e
o -a p a l
a u t o p a la t i n e b o n e
o -b o c
b a s io c c ip ita l b o n e
o -c h -p
p o s te r io r c e r a to h y a l b o n e
o -d e n
d e n ta r y b o n e
o -e cp
e c to p te r y g o id b o n e
o -e p a x
s e s a m o i d b o n e i n e p a x ia l t e n d o n
o -h m
h y o m a n d i b u la r b o n e
o -io p
in t e r o p e r c u la r b o n e
o - la c
la c r im a l b o n e
o -m p
m e ta p te r y g o id b o n e
o -m x
m a x illa r y b o n e
o -p r m x
p r e m a x illa r y b o n e
o -p t
p te r o tic b o n e
o -q
q u a d r a te b o n e
o -r a r t
r e t r o a r t ic u la r b o n e
o -so c
s u p r a o c c ip ita l b o n e
o -sy m
s y m p le c tic b o n e
o -u h
urohyal bone
o -v m
vom eral bon e
p r -c
c o r o n o id p r o c e ss
r -b r
b r a n c h io s te g a l ray
t-A 2
s e c o n d t e n d o n o f t h e a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x
t-A 3
t h i r d t e n d o n o f t h e a d d u c t o r m a n d ib u la e m u s c l e c o m p l e x
t-e p a x
e p a x ia l te n d o n
t-p r -h
p r o tr a c to r h y o id e i te n d o n
t-st
s te r n o h y o id e u s te n d o n
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
X
E
C
CL
X
E
I
a
CN
<'
POCDCt:
=
¿. r£o . T£3 5‘r
Q .Q
—.
±
O
1
O
ao
m
F ig . 3 .1 2 - 3 D - r e c o n s t r u c t i o n o f t h e c r a n iu m o f D o r y r h a m p h u s d a c t y l i o p h o r u s (9 1 .2 m m SL ) b a s e d
o n h i s t o l o g i c a l s e c t i o n s . A l a t e r a l v i e w o f t h e l e f t s id e w i t h s e v e r a l b o n e s r e m o v e d t o s h o w
m u s c l e s u n d e r n e a t h ; B la t e r a l v i e w o f t h e l e f t s id e o f t h e s n o u t t i p . B e ig e i n d ic a t e s b o n e , k h a k i
i n d i c a t e s c a r t ila g e , o r a n g e i n d ic a t e s t e n d o n a n d h g a m e n t , r e d i n d ic a t e s m u s c le . S c a le b a r A = 2
m m ; s c a le b a r B = 1 m m .
46
F ig . 3 .1 3
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
- Legend
cr
coron a
o -a n g
a n g u lo a r t ic u la r b o n e
o -c h -a
a n t e r io r c e r a t o h y a l b o n e
o -d e n
d e n ta r y b o n e
o -e cp
e c to p te r y g o id b o n e
o -e p a x
s e s a m o i d b o n e i n e p a x ia l t e n d o n
o -ih
in te r h y a l b o n e
o -io p
in t e r o p e r c u la r b o n e
o -m eth
m e s e th m o id b o n e
o -m x
m a x illa r y b o n e
o -o p
o p e r c u la r b o n e
o -p a r
p a r ie t a l b o n e
o -p a r a
p a r a s p h e n o id b o n e
o -p o p
p r e o p e r c u la r b o n e
o -q
q u a d r a te b o n e
o -so c
s u p r a o c c ip ita l b o n e
o -sp h
s p h e n o tic b o n e
o -sy m
s y m p le c tic b o n e
o -u h
urohyal bone
o -v m
vom eral bon e
C h a p t e r 3 - M o r p h o lo g ic a l d e s c r ip ti o n
o-ecp o-meth
o-mx 1 o-vm
o-den
o-ang
o-q
o-sym
o-ih o-pop o-op
o-den
ih o-ch-a o-uh o-op
o-mx o-ecp o-vm o-meth
o-den o-ang o-q o-iop
o-den o-ang
F ig . 3 .1 3 -
o-q
o-sym o-pop
o-ch-a o-ih o-uh
o-sym o-pop o-ch-a o-ih o-uh
3 D -r e c o n s tr u c tio n s o f t h e o s te o c r a n iu m
o-op
o-op
b a s e d o n C T - d a ta i n l a t e r a l v i e w
of A
H i p p o c a m p u s z o s t e r a e (2 9 .0 m m SL ); B H i p p o c a m p u s a b d o m i n a l i s ( 2 2 2 .7 m m SL ); C C o r y t h o i c h t h y s
i n t e s t i n a l i s (1 2 0 .2 m m SL ); D D o r y r h a m p h u s m e l a n o p l e u r a (5 6 .5 m m S L ). S c a le b a r s = 1 m m .
48
C h a p te r 4 - M o rp h o lo g ic a l v a ria tio n
C h a p t e r 4 - M o r p h o lo g ic a l v a ria tio n
occ
in
S 52s
E5
cLt>
(j
9. 5
5, a)]
F ig . 4 .1 - S c a t t e r p l o t o f r e la t iv e w a r p l a n d r e la t iv e w a r p 2, t o g e t h e r e x p l a in in g 90.38% o f t h e
t o t a l s h a p e v a r ia t io n . T h e p o s i t i o n o f t h e m e a n s u b f a m i ly c o n f i g u r a t i o n is d e p i c t e d b y a V (r ig h t
f o r p i p e f i s h e s , l e f t f o r s e a h o r s e ) a n d t h e c o r r e s p o n d i n g d e f o r m a t i o n g r id s w i t h r e s p e c t t o t h e
c o n s e n s u s a r e i llu s t r a t e d .
50
Chapter 4 - Morphological variation
C h a p t e r 4 - M o r p h o lo g ic a l v a ria tio n
0,010
0 seahorse
H. ramulosus
0,008
Intraspecific variation
• pipefish
O
O
H. guttulatus
H. kuda
O
0,006
0,004
S. typhle
H. reidi Q
•
0,002
S. rostellatus
•
S. biaculeatus
•
M. brachyurus •
aculeatus
0,000
0,30
0,35
0,40
0,45
0,50
0,55
0,60
Mean (SnL/HL)
F ig . 4 .2 - S c a t t e r p l o t o f t h e m e a n r e la t iv e s n o u t l e n g t h v e r s u s t h e i n t r a s p e c i f i c s h a p e v a r ia t i o n o f
f o u r p i p e f i s h s p e c i e s (b la c k d o ts ) a n d f o u r s e a h o r s e s p e c i e s ( w h i t e d o ts ) . H L , h e a d le n g t h ; S n L ,
s n o u t le n g t h .
52
Chapter 4 - Morphological variation
C h a p t e r 4 - Mor phological variation
RW2 (16.75%)
P ig - 4-3 -
S c a t t e r p l o t o f r e l a t iv e w a r p
1 a n d r e la t iv e w a r p 2 o f t h e o n t o g e n e t i c s e r ie s o f
H i p p o c a m p u s re id i, t o g e t h e r e x p l a in in g 57% o f t h e t o t a l s h a p e v a r ia t io n . B la c k d o t s r e p r e s e n t
a d u lts , w h i t e d o t s a r e j u v e n ile s . T h e d e f o r m a t i o n g r id s o f t w o s p e c i m e n s , a o n e d a y o l d ( b e lo w )
a n d a n a d u lt (a b o v e ), a r e v is u a liz e d .
54
C h a p t e r 4 - M o r p h o lo g ic a l v a ria tio n
C h a p t e r 4 - Mor phological variation
0,08
0,06
0
0
• •
0 0
0,04
_
0
!•
0
I
0,02
ST
RW
-
o
0,02
0
9000 A r
*
-0,04
Or
c**«o
%
•
0
• %
0
•
O
0
O
8
€>
0
%
O
0,12
•
•< ••
*
o
•
o Üp
o
o
•
-
o
* ° o
0
0
*
0)
•
•
•°
'•o °
0
Q
0
o
-0,08
0,10
0
qg
-0,06
-
O
1
Q
0,00
• 0
° 4
0O
• 0
o.
o
•
<X>
é>
o
0RWI
0
o
• RW2
o
• RW3
ORW4
-0,14
0,20
0,30
0,40
0,50
0,60
0,70
Log (HL)
F ig . 4 .4 - G r o w t h c u r v e o f H i p p o c a m p u s r e i d i j u v e n ile s , w i t h l o g h e a d l e n g t h (L o g (H L )) a s a p r o x y
f o r a g e a n d t h e r e la t iv e w a r p ( R W ) s c o r e s r e p r e s e n t i n g h e a d s h a p e c h a n g e s . N o c le a r t r e n d c o u l d
b e ob served .
C h a p t e r 5 - Functional interpr et ati on
56
F ig . 5 .1 - L e g e n d
hs
h y o s y m p l e c t i c c a r t ila g e
l- c h - u h
c e r a t o h y a l - u r o h y a l l ig a m e n t
1 -io p -h
i n t e r o p e r c u l o - h y o i d l ig a m e n t
l-m n d -h
m a n d ib u l o - h y o i d l ig a m e n t
l- m n d - i o p
m a n d ib u l o - i n t e r o p e r c u la r l ig a m e n t
m - im -p
p o s t e r i o r i n t e r m a n d i b u la r is p a r t o f p r o t r a c t o r h y o i d e i m u s c l e
m - in t - a
a n t e r io r i n t e r h y o id e u s p a r t o f p r o t r a c t o r h y o i d e i m u s c le
m -st
s te r n o h y o id e u s m u s c le
o -c h -a
a n t e r io r c e r a t o h y a l b o n e
o -c h -p
p o s te r io r c e r a to h y a l b o n e
o -c l
c le i t h r u m
o -c o r b
c ir c u m o r b it a l b o n e s
o -d e n
d e n ta r y b o n e
o -e cp
e c to p te r y g o id b o n e
o -h m
h y o m a n d i b u la r b o n e
o -ih
in te r h y a l b o n e
o -io p
in t e r o p e r c u la r b o n e
o -m p
m e ta p te r y g o id b o n e
o -m x
m a x illa r y b o n e
o -o p
o p e r c u la r b o n e
o -p o p
p r e o p e r c u la r b o n e
o -p r m x
p r e m a x illa r y b o n e
o -q
q u a d r a te b o n e
o -sy m
s y m p le c tic b o n e
o -u h
urohyal bone
o -v m
vom eral bon e
t-in t-a
t e n d o n o f a n t e r io r i n t e r h y o id e u s p a r t o f p r o t r a c t o r h y o i d e i m u s c l e
t-st
s te r n o h y o id e u s te n d o n
C h a p t e r 5 - F u n c t i on a l i n t e r p r e t a t i o n
o-hm o-pop
o-mp
o-ecp
o -p rm x
i-mx
o-den
m-int-a
l-mnd-iop
o-ch-a
o-prmx
c
F ig . 5 .1 - G r a p h ic 3 D - r e c o n s t r u c t i o n o f t h e s k u l l o f H i p p o c a m p u s r e i d i b a s e d o n h i s t o l o g i c a l
s e c t i o n s ( n e u r o c r a n iu m a n d s e v e r a l m u s c l e s a n d h g a m e n t s n o t r e c o n s t r u c t e d f o r c la r ity ). A m e d ia l
v i e w o f l e f t s id e ; B l a t e r a l v i e w o f l e f t sid e ; C d o r s a l v i e w . B e ig e i n d ic a t e s b o n e , k h a k i i n d ic a t e s
c a r t ila g e , o r a n g e i n d ic a t e s t e n d o n a n d l ig a m e n t , r e d i n d ic a t e s m u s c le . S c a le b a r = 2 m m .
C h a p t e r 5 - Functional interpr et ati on
58
F ig . 5 .2 - L e g e n d
1
t i p o f t h e n o s e s p in e , j u s t a n t e r io r o f t h e e y e
2
t i p o f t h e s n o u t p o s t e r i o r t o t h e m a x illa r y b o n e s
3
s y m p h y s is o f t h e c e r a t o h y a l b o n e s
4
d i s t a l t i p o f t h e v e n t r o la t e r a l s p in e o n t h e p e c t o r a l g ir d le
5
t i p o f t h e p r o c e s s o n t h e s e c o n d n u c h a l p la t e
6
p r o x i m a l t i p o f t h e p r e m a x illa r y b o n e
7
d is ta l tip o f t h e d e n ta r y b o n e
8
d is tin c t p o in t o n th e p rey
Oh
c e n tr e o f th e in te r h y a l b o n e
O -1
r o t a t i o n p o i n t o f n e u r o c r a n iu m w i t h v e r t e b r a l c o lu m n - p e c t o r a l g ir d le c o m p l e x
r
m e d i o r o s t r a l la n d m a r k o n t h e c le i t h r u m
s
s y m p h y s is o f t h e c e r a t o h y a l b o n e s
C h a p t e r 5 - Functional interpr et ati on
F ig .
5 .2 -
G r a p h ic 3 D - r e c o n s t r u c t i o n
o f th e
s k u l l o f H i p p o c a m p u s re i d i b a s e d o n
r e p r e s e n t i n g t h e d i g i t i z e d la n d m a r k s ( y e ll o w ) a n d t h e c a lc u l a t e d fo u r -b a r s y s t e m (r ed ).
C T -d a ta ,
6o
Pig-
C h a p t e r 5 - Functional interpr et ati on
5-3
_ Legend
1
m a n d ib u lo - q u a d r a t e a r t ic u l a t i o n
2
m a n d ib u la r s y m p h y s is
3
p r e m a x illa r y s y m p h y s is
4
r o stra l tip o f th e v o m e r a l b o n e
5
d i s t a l t i p o f t h e n o s e s p in e
6
a r t ic u l a t i o n b e t w e e n t h e s u p r a o c c i p i t a l b o n e a n d t h e c o r o n a
7
d is ta l tip o f t h e d o r so c a u d a l p r o c e s s o n t h e c o r o n a
8
d i s t a l t i p o f t h e p r o c e s s o n t h e s e c o n d n u c h a l p la t e
9
a r t ic u l a t i o n b e t w e e n t h e n e u r o c r a n iu m a n d t h e f i r s t v e r t e b r a ( n o t illu s t r a t e d )
10
a r t ic u l a t i o n b e t w e e n t h e n e u r o c r a n iu m a n d t h e p e c t o r a l g ir d le
11
d i s t a l t i p o f t h e v e n t r o la t e r a l s p in e o n t h e p e c t o r a l g ir d le
12
t i p o f t h e a n g le f o r m e d b y t h e h o r i z o n t a l a n d v e r t i c a l a r m o f t h e p e c t o r a l g ir d le
13
r o str a l p o in t o f th e p e c to r a l c o m p le x
14
c a u d a l tip o f th e u ro h y a l b o n e
15
s y m p h y s is o f t h e c e r a t o h y a l b o n e s
16
a r t ic u l a t i o n b e t w e e n p o s t e r i o r c e r a t o h y a l a n d i n t e r h y a l b o n e s
17
a r t ic u l a t i o n b e t w e e n i n t e r h y a l a n d p r e o p e r c u la r b o n e s
cr
coron a
n p -II
s e c o n d n u c h a l p la t e
o -c h -a
a n t e r io r c e r a t o h y a l b o n e
o -c l
c le i t h r u m
o -d e n
d e n ta r y b o n e
o -h m
h y o m a n d i b u la r b o n e
o -io p
in t e r o p e r c u la r b o n e
o -le th
la t e r a l e t h m o i d b o n e
o -m e th
m e s e th m o id b o n e
o -m x
m a x illa r y b o n e
o -p a r
p a r ie t a l b o n e
o -p a r a
p a r a s p h e n o id b o n e
o -p o p
p r e o p e r c u la r b o n e
o -p o stt
p o sttem p o ra l b o n e
o -p r m x
p r e m a x illa r y b o n e
o -q
q u a d r a te b o n e
o -so c
s u p r a o c c ip ita l b o n e
o -sp h
s p h e n o tic b o n e
o -u h
urohyal bone
o -v m
vom eral bon e
vi
f i r s t v e r te b r a
C h a p t e r 5 - Functional interpr et ati on
o-pop
2nd nuchal plate
corona
neurocranlal
unit
pectoral
unit \
upper jaw
urohyal
interhyal
lower jaw
P ig - 5-3 " S k u l l m o r p h o l o g y o f a d u lt H i p p o c a m p u s re id i a n d 3 D - m o d e l . A 3 D - r e c o n s t r u c t i o n o f t h e
s k u l l b a s e d o n C T -d a ta ; B
3 D - r e c o n s t r u c t i o n i n d ic a t in g t h e p o s i t i o n o f t h e la n d m a r k s u s e d
c o n s t r u c t i n g t h e m o d e l ( p o i n t 9 n o t illu s t r a t e d ) ; C m o d e l d r a w n b a s e d o n t h e la n d m a r k s a n d
su p e r im p o s e d o n a 3 D -r e c o n s tr u c tio n ; D m o d e l u s e d fo r v is u a liz in g m e c h a n ic a l u n its a n d th e
c h a n g e s i n r e la t iv e p o s i t i o n d u r in g f e e d i n g s tr ik e . S c a le b a r = 5 m m .
62
C h a p t e r 5 - Functional interpr et ati on
F ig . 5 .4 - L e g e n d
a f-ih
a r t ic u l a t i o n f a c e t o f i n t e r h y a l b o n e w i t h p r e o p e r c u la r b o n e
ch
c e r a t o h y a l c a r t ila g e
1 -c h -u h
c e r a t o h y a l - u r o h y a l l ig a m e n t
lh -c h -p
la t e r a l h e a d o f p o s t e r i o r c e r a t o h y a l b o n e
lh -ih
la t e r a l h e a d o f i n t e r h y a l b o n e
1 -io p -h
i n t e r o p e r c u l o - h y o i d l ig a m e n t
1 -m n d -h
m a n d ib u l o - h y o i d l ig a m e n t
m h -ih
m e d ia l h e a d o f in te r h y a l b o n e
m - in t - a
a n t e r io r i n t e r h y o id e u s p a r t o f p r o t r a c t o r h y o i d e i m u s c le
o -c h -a
a n t e r io r c e r a t o h y a l b o n e
o -c h -p
p o s te r io r c e r a to h y a l b o n e
o -ih
in te r h y a l b o n e
o -io p
in t e r o p e r c u la r b o n e
o -p o p
p r e o p e r c u la r b o n e
o -u h
urohyal bone
t-in t-a
t e n d o n o f a n t e r io r i n t e r h y o id e u s p a r t o f p r o t r a c t o r h y o i d e i m u s c l e
v h -c h -p
v e n tr a l h e a d o f p o s te r io r c e r a to h y a l b o n e
C h a p t e r 5 - Functional interpr et ati on
0 -p o p
o-ih
o-iop
1-iop-h
o-ch-p
ch
m-int-a
0-ch -a
t-int-a
1-ch-uh
l-mnd-h
o-uh
A
B
F ig . 5 .4 - D e t a i l o f t h e 3 D - r e c o n s t r u c t i o n o f t h e r e t r a c t e d h y o i d c o m p l e x i n H i p p o c a m p u s re id i
b a s e d o n h i s t o l o g i c a l s e c t i o n s ( s o m e h g a m e n t s n o t r e c o n s t r u c t e d f o r c la r ity ). A l a t e r a l v i e w o f l e f t
s id e ; B c a u d a l v i e w ( o t h e r s t r u c t u r e s n o t r e c o n s t r u c t e d ) . B e ig e i n d ic a t e s b o n e , k h a k i i n d ic a t e s
c a r t ila g e , o r a n g e i n d ic a t e s t e n d o n a n d l ig a m e n t , r e d i n d ic a t e s m u s c le . S c a le b a r s = 1 m m .
64
C h a p t e r 5 - F u n c ti o n a l i n t e r p r e t a t i o n
C h a p t e r 5 - Functional interpr et ati on
0 ms
0.5 ms
ms
545.5 ms
660.5 ms
695.5 ms
P ig - 5-5 " E x a m p le o f a (fa ile d ) f e e d in g s e q u e n c e o f H i p p o c a m p u s r e i d i f e e d i n g o n a m y s id s h r im p
i n u p p e r l e f t c o r n e r . I m p o r t a n t e v e n t s a re i n d ic a t e d b y a n a r r o w . A im a g e o f o n e fr a m e b e f o r e t h e
f i r s t v i s i b l e m o v e m e n t ; B o n s e t o f h y o i d r o t a t i o n ; C o n s e t o f c r a n ia l r o t a t i o n a n d m o u t h o p e n in g ;
D m o u t h r e a c h e s m a x im a l o p e n in g ; E h y o i d r e a c h e s m a x i m a l r o t a t io n ; F n e u r o c r a n iu m r e a c h e s
m a x i m a l r o t a t i o n ; G m o u t h f u l l y c lo s e d ; H n e u r o c r a n iu m r e s t o r e d t o i t s o r ig i n a l p o s it i o n ; I h y o i d
f u lly p r o tr a cte d .
66
C h a p t e r 5 - F u n c ti o n a l i n t e r p r e t a t i o n
C h a p t e r 5 - Functional interpr et ati on
70
60
50
-
2
-
1
40
o
an
30
20
10
0
0
0
3
2
1
4
Time (ms)
100 -
o
0)
80
O
)
C
A3
■g
o
X
0
B
2
4
6
8
10
12
14
Time (ms)
F ig . 5 .6 - A T im e a v e r a g e d k i n e m a t i c p r o f i le s ( N = 14) o f t h e m e a n h y o i d r o t a t i o n ( b la c k d o t s ) , t h e
m e a n n e u r o c r a n ia l r o t a t i o n (g r a y d o t s ) a n d t h e m e a n m o u t h o p e n i n g ( w h i t e d o t s ) o f a ll 1 4
s e q u e n c e s . T im e = o m s w a s c o n s e q u e n t l y d e f i n e d f o r e a c h s e q u e n c e a s t h e fr a m e b e f o r e t h e f ir s t
v i s i b l e m o v e m e n t . P r e y c a p t u r e o c c u r s o n a v e r a g e 5.5 m s a fte r t h e o n s e t o f h y o i d r o t a t i o n . B
C o m p a r i s o n o f t h e a v e r a g e m e a s u r e d h y o i d a n g le ( b la c k d o ts ) a n d t h e a v e r a g e h y o i d a n g le
p r e d i c t e d b y t h e fo u r -b a r m o d e l ( w h i t e d o ts ; N
= 14 ). I n t h e f ir s t 1.5 m s t h e g r a p h s d o n o t
68
C h a p t e r 5 - F u n c ti o n a l i n t e r p r e t a t i o n
C h a p t e r 5 - Functional interpr et ati on
P ig - 5-7 " M o d e l s h o w i n g e x t e n s i v e n e u r o c r a n ia l e l e v a t i o n o n H i p p o c a m p u s reidi. A v i d e o fr a m e s
u s e d f o r p o s i t i o n i n g t h e c o m p o n e n t s o f t h e m o d e l a t t h e o n s e t o f n e u r o c r a n ia l e l e v a t i o n ( le f t ) a n d
a t m a x im a l e l e v a t i o n (r ig h t). B m o d e l a t t h e o n s e t o f e l e v a t i o n a n d t h e p o s i t i o n o f t h e s k u l l at
m a x i m a l e le v a t io n , s u p e r i m p o s e d a n d a l ig n e d b a s e d o n t h e p o s i t i o n o f t h e p e c t o r a l g ir d le .
C h a p t e r 5 - Functional interpr et ati on
70
F ig . 5 .8 - L e g e n d
o -a p a l
a u t o p a la t i n e b o n e
o -b o c
b a s io c c ip ita l b o n e
o -d e n
d e n ta r y b o n e
o -h m
h y o m a n d i b u la r b o n e
o -io p
in t e r o p e r c u la r b o n e
o -le th
la t e r a l e t h m o i d b o n e
o -m eth
m e s e th m o id b o n e
o -m x
m a x illa r y b o n e
o -p a r a
p a r a s p h e n o id b o n e
o -p a r
p a r ie t a l b o n e
o -p o p
p r e o p e r c u la r b o n e
o -p o stt
p o sttem p o ra l b o n e
o -p r m x
p r e m a x illa r y b o n e
o -so c
s u p r a o c c ip ita l b o n e
o -sp h
s p h e n o tic b o n e
o -v m
vom eral bon e
C h a p t e r 5 - Functional interpr et ati on
O -S O C
o-apal
o-mx
o-prm*
F ig . 5 .8 - C o n s t r u c t i o n o f t h e H i p p o c a m p u s z o s t e r a e a r t if i c i a l m o d e l b a s e d o n C T - d a ta s h o w i n g t h e
o r ig i n a l H . z o s t e r a e ( b lu e ) , t h e H . r e i d i d e p r e s s e d m o d e l b e i n g m o d i f i e d (d a r k g r e y ) a n d t h e
c o m p r e s s e d p a r t o f t h e s n o u t ( lig h t g r e y ). N o t e t h a t t h e m o d e l s a r e s c a le d b a s e d o n b r a in c a s e
le n g th . A
u n a l t e r e d m o d e l s o f H . re id i a n d H . z o s te r a e , w i t h
in d ic a t io n
o f so m e b on es; B
c o m p r e s s i o n o f t h e s n o u t ; C a t t a c h m e n t o f t h e s n o u t t i p a n d ja w s . T h e c o m p r e s s e d s n o u t w a s
t h e n s m o o t h e d a n d s i m p h f i e d u n t i l t h e t r ia n g le s h a d t h e s a m e s i z e a s t h e o n e s i n t h e r e s t o f t h e
c r a n iu m .
C h a p t e r 5 - Functional interpr et ati on
72
P ig - 5-9 - L e g e n d
1
a u t o p a la t i n e - v o m e r a l a r t ic u l a t i o n
2
h y o id
3
4
5
la t e r a l w a l l o f t h e s n o u t
6
v o m e r o -m e s e th m o id tr a n s itio n
7
l o w e r ja w s y m p h y s is
m e s e t h m o id - p a r i e t a l b o r d e r
h y o m a n d ib u lo -s p h e n o tic jo in t
8
r o stra l b o rd er o f th e n o s tr il
9
v e n t r a l b o r d e r o f t h e o r b it
10
b r a in c a s e r o o f
C h a p t e r 5 - Functional interpr et ati on
P ig - 5-9 - V o n M i s e s s t r e s s d i s t r i b u t i o n p a t t e r n i n s e a h o r s e s (A -C ) a n d p i p e f i s h e s (D -F ) w i t h l o n g
(A , D ) , i n t e r m e d i a t e (B , E) o r s h o r t (C , F) s n o u t l e n g t h . A H i p p o c a m p u s r e i d i d e p r e s s e d ; B H .
a b d o m in a li s ;
C
H.
z o s te r a e ;
D
D oryrham phus
d a c ty l i o p h o r u s ;
E
Syngn athus
r o s te lla tu s ;
F
D.
m e l a n o p l e u r a . D i f f e r e n t c o lo r s c a le s a re u s e d f o r e a c h m o d e l t o f a c i h t a t e c o m p a r is o n o f t h e s tr e s s
c o n c e n t r a t i o n p a t t e r n . S tr e s s v a l u e s e x p r e s s e d i n M P a .
C h a p t e r 5 - Functional interpr et ati on
74
F ig . 5 .1 0 - L e g e n d
1
h y o i d s y m p h y s is
2
h y o m a n d ib u lo -s p h e n o tic jo in t
3
4
5
l o w e r ja w
6
m a x illa r y b o n e s
7
b r a in c a s e
8
la t e r o c a u d a l l y i n t h e s n o u t
a u t o p a la t i n e - v o m e r a l a r t ic u l a t i o n
h y o id
9
la t e r o r o s t r a l l y i n t h e s n o u t
10
v e n t r o r o s t r a l b o r d e r o f t h e o r b it
C h a p t e r 5 - Functional interpr et ati on
4
F ig . 5 .1 0 - V o n M i s e s s t r e s s d i s t r i b u t i o n p a t t e r n i n t h e a r t if i c i a l ly c o n s t r u c t e d m o d e l s b a s e d o n
H i p p o c a m p u s r e i d i m o r p h o l o g y . A H . r e i d i d e p r e s s e d m o d e l (a s i n F ig . 5.9 A ); B H . a b d o m i n a l i s
a r t if i c i a l m o d e l; C H . z o s t e r a e a r t if i c i a l m o d e l. S t r e s s v a l u e s e x p r e s s e d i n M P a , c o lo r s c a le s are
d i f f e r e n t f o r a ll m o d e ls .
76
C h a p t e r 5 - Functional interpr et ati on
F ig . 5 .1 1 - L e g e n d
1
m e s e t h m o id - p a r i e t a l b o r d e r
2
la t e r o r o s t r a l l y i n t h e s n o u t
3
h y o i d - s u s p e n s o r i u m a r t ic u l a t i o n
C h a p t e r 5 - Functional interpr et ati on
F ig . 5 .1 1 - V o n M i s e s s t r e s s d i s t r i b u t i o n p a t t e r n i n t w o m o d e l s o f H i p p o c a m p u s re i d i s h o w i n g t h e
e f f e c t o f t h e p o s i t i o n o f t h e h y o id . A H . r e i d i p r o t r a c t e d m o d e l; B H . re id i d e p r e s s e d m o d e l. S t r e s s
v a lu e s e x p r e s se d in M P a .
© A l l r ig h t s r e s e r v e d . T h is t h e s i s c o n t a i n s c o n f i d e n t i a l i n f o r m a t i o n a n d c o n f i d e n t i a l r e s e a r c h
r e s u l t s t h a t a re p r o p e r t y t o t h e U G e n t ( R e s e a r c h g r o u p E v o l u t i o n a r y M o r p h o l o g y o f V e r t e b r a t e s ,
D e p a r t m e n t o f B io lo g y ) . T h e c o n t e n t s o f t h i s m a s t e r t h e s i s m a y u n d e r n o c ir c u m s t a n c e s b e m a d e
p u b l i c , n o r c o m p l e t e o r p a r t ia l, w i t h o u t t h e e x p l i c i t a n d p r e c e d i n g p e r m is s i o n o f t h e U G e n t
r e p r e s e n t a t i v e , i.e . t h e s u p e r v is o r . T h e t h e s i s m a y u n d e r n o c ir c u m s t a n c e s b e c o p i e d o r d u p l i c a t e d
i n a n y fo r m , u n l e s s p e r m is s i o n g r a n t e d i n w r i t t e n fo r m . A n y v i o l a t i o n o f t h e c o n f i d e n t i a l n a t u r e
o f t h i s t h e s i s m a y i m p o s e ir r e p a r a b le d a m a g e t o t h e U G e n t . I n c a s e o f a d is p u t e t h a t m a y a r is e
w i t h i n t h e c o n t e x t o f t h i s d e c la r a t io n , t h e J u d ic ia l C o u r t o f G e n t o n l y is c o m p e t e n t t o b e
n o tifie d .