FACULTY OF SCIENCES
Department of Biology
Evolutionary Morphology of Vertebrates
Early development of Corydoras aeneus
(Siluriformes, Callichthyidae): a case study for
understanding the evolutionary basis of loricarioid
ontogenetic patterning
Part 1 – Text
Frank Huysentruyt
Thesis submitted to obtain the degree of
Doctor in Sciences (Biology)
Proefschrift voorgedragen tot het bekomen
van de graad Doctor in de Wetenschappen
(Biologie)
Academiejaar 2007-2008
Rector: Prof. Dr. Paul van Cauwenberge
Decaan: Prof. Dr. Herwig Dejonghe
Promotor: Prof. Dr. Dominique Adriaens
Early development of Corydoras aeneus
(Siluriformes, Callichthyidae): a case study
for understanding the evolutionary basis of
loricarioid ontogenetic patterning
Part 1 - Text
Frank Huysentruyt
Thesis submitted to obtain the degree of
Doctor in Sciences (Biology)
Proefschrift voorgedragen tot het bekomen van
de graad Doctor in de Wetenschappen (Biologie)
Academiejaar 2007-2008
Rector: Prof. Dr. Paul van Cauwenberge
Decaan: Prof. Dr. Herwig Dejonghe
Promotor: Prof. Dr. Dominique Adriaens
SAMENSTELLING LEES*- EN EXAMENCOMMISSIE
Prof. Dr. Wim Vyverman
Prof. Dr. Dominique Adriaens
Prof. Dr. Peter Bossier
Prof. Dr. Michel Chardon
Dr. Tom Geerinckx
Prof. Dr. Ann Huysseune
Dr. Isaäc Isbrücker
Prof. Dr. Giorgos Koumoundouros
(Univeristeit
(Universiteit
(Universiteit
(Universiteit
(Universiteit
(Universiteit
(Universiteit
(Universiteit
Gent, voorzitter)
Gent)*
Gent)
Luik)*
Gent)
Gent)
Amsterdam, Nederland)
Patras, Griekenland)*
DANKWOORD
Er zijn mensen die zeggen dat ik soms te veel praat. Ik denk dat dat terecht is.
Veel valt daar niet aan te doen, het is wie ik ben. Maar af en toe heb ik ook echt
iets te zeggen, zoals nu, en dus schrijf ik het beter op, zodat het blijft.
In dit deel wil ik alle mensen bedanken die, elk op hun manier, aan dit werk
hebben bijgedragen, hetzij in een actieve rol, hetzij in een meer passieve rol, de
ene daarom niet minder belangrijk dan de andere.
In eerste instantie wil ik natuurlijk Prof. Dr. D. Adriaens bedanken voor zowat
alles. Vanzelfsprekend in de eerste plaats omdat hij in mij heeft geloofd en me
de kans heeft gegeven om dit tot een goed einde te brengen. Daarnaast is er
niemand die zelfs maar in de buurt komt van de mate waarin hij actief aan dit
werk heeft bijgedragen. Van het lezen en verbeteren van teksten (die in het
begin echt zeer slecht moeten geweest zijn) tot het aanbrengen van ideeën en
het bediscussiëren van allerlei resultaten. Daarnaast wens ik hem zeker en vast
ook te bedanken voor de kansen die hij me heeft gegeven om buitenlandse
congressen bij te wonen waardoor ik niet alleen allerlei leuke locaties heb kunnen
bezoeken, maar waar ik ook onschatbare ervaring in het presenteren en
bediscussiëren van wetenschappelijke data mocht opdoen. Ik weet dat het zijn
bewuste en niet altijd even evidente keuze is om zijn studenten deze kansen te
bieden en ben hem daar, samen met mijn collega-studenten, zeer dankbaar
voor.
Ten tweede wil ik natuurlijk ook mijn collega’s bedanken. Ik kan oprecht zeggen
dat ik altijd graag naar het labo ben gekomen en dat was vooral aan hen te
danken. Ik denk niet dat er mensen tussenzitten met wie ik niet kon opschieten
en sommige zijn hopelijk vrienden voor het leven geworden. In alfabetische
volgorde: Barbara, Celine, Fatemeh, Heleen, Joachim, Marleen, Mimi, Natalie,
Paul, Soheil, Stijn en Tom, bedankt voor een onvergetelijke tijd!
In het bijzonder wil ik de hulp van Barbara, Joachim en Marleen bij al het
praktische werk nog eens extra in de bloemetjes zetten, hun expertise vormt de
hoeksteen van alle wetenschappelijk werk dat in ons labo wordt en zal worden
verricht. Ook bij het analyseren van sommige data in dit werk heb ik de hulp
gekregen zonder welke het simpelweg niet zou zijn gelukt: Bieke, bedankt voor
het vele werk waar ik je te pas en te onpas mee heb lastig gevallen.
In een voorlaatste paragraaf wil ik mijn familie bedanken. Mijn broer,
schoonbroers en schoonzus omdat ze er gewoon altijd zijn als je ze nodig hebt.
Mijn schoonouders daarnaast voor de vele steun en het niet gespeelde
enthousiasme waarmee ze dit doctoraat ondersteund hebben. En natuurlijk
vooral mijn ouders. Jullie hebben mij gedurende mijn ganse studies (en daarmee
bedoel ik mijn volledige schoolcarrière) voortdurend gestimuleerd om het
onderste uit de kan te halen. Iets wat ik aanvankelijk meer niet dan wel heb
gedaan, maar het verstand komt nu eenmaal met de jaren. Het is me nu toch
gelukt en het besef dat er twee mensen waren die op de momenten dat het wat
slabbakte toch in mijn bleven geloven heeft daar een zeer grote rol in gespeeld.
Bedankt.
In laatste instantie wil ik mijn gezin bedanken. Eerst en vooral Jo(ke), om mij te
leren studeren (en dat je dat in blokken van langer dan een half uur moet doen)
en verder gewoon om er nu toch al tien jaar te zijn, elke dag. Ik hou zeer veel
van je! Verder wil ik ook mijn twee zoontjes, Bas en Daan (en bij voorbaat mijn
ongeboren kinderen, zodat er later geen ruzie is omdat de ene wel en de andere
niet in papas doctoraat mocht), die absoluut in geen enkel opzicht ook maar iets
met dit doctoraat te maken hebben, bedanken. Ik wil ze bedanken omdat ze, en
dit kan contradictorisch klinken, ondanks de slapeloze nachten, ziektes, het vele
werk en het constante weg-en-weer-gerij, het afwerken van dit doctoraat zo
gemakkelijk hebben gemaakt. Het hebben van kinderen zorgt nu eenmaal voor
de nodige zin voor relativering en rust die het schrijven van dit werk tot een
plezier heeft gemaakt.
Aan iedereen dus nogmaals bedankt !
Frank.
A hen is only an egg’s way of making another egg.
Samuel Butler
TABLE OF CONTENTS
TABLE OF CONTENTS
1.
General Introduction
1.1.
2.
1
General context
1
Aims of the study
3
Outline of this thesis
6
Material and Methods
2.1.
2.2.
3.
General context and aims
Material
9
Choice of species
9
Material examined
11
Methods
15
In toto clearing and staining
15
Dissections
16
Serial sectioning
17
3D-reconstructions
18
Adult Morphology
3.1.
3.2.
Adult Osteology
19
Abstract
19
Brief introduction
19
Brief material and methods
20
Results and discussion
20
Adult Myology
35
Abstract
35
Brief introduction
35
Brief material and methods
36
Results
36
Discussion
42
i
TABLE OF CONTENTS
4.
Ontogeny
4.1.
4.2.
4.3.
4.4.
4.5.
4.6.
The Egg
Abstract
49
Introduction
49
Brief material and methods
50
Results and discussion
51
Early Development and Growth
53
Abstract
53
Brief introduction
54
Brief material and methods
54
Results
56
Discussion
59
The Cranium
63
Abstract
63
Brief introduction
64
Brief material and methods
65
Results
65
Discussion of chondrocranial development
73
Discussion of osteocranial development
80
The Cranial Myology
87
Abstract
87
Brief introduction
88
Brief material and methods
89
Results
89
Discussion
91
The Postcranial Skeleton
97
Abstract
97
Brief introduction
97
Brief material and methods
98
Results
98
Discussion
101
Ontogeny Of Corydoras pygmaeus,
a Miniaturized Congeneric
Abstract
ii
49
105
105
TABLE OF CONTENTS
5.
Introduction
105
Brief material and methods
108
Results
108
Discussion
111
General Discussion
5.1.
5.2.
The Ontogeny of Corydoras aeneus
117
The theory of saltatory ontogeny
117
The Ontogeny of Corydoras aeneus
118
Corydoras pygmaeus and Ancistrus cf. triradiatus
121
Thresholds in the development of Loricarioidea
122
The Evolution of Algae Scraping in Loricarioidea
125
Scaridae, another example of adaptive radiation an
6.
algivory in teleosts
125
Three phases of evolution
126
Summary & Samenvatting
6.1.
Summary
137
6.2.
Samenvatting
143
7.
References
149
8.
Publication List
165
iii
TABLE OF CONTENTS
iv
Chapter
1
Introduction
1 GENERAL CONTEXT AND AIMS
1.1 GENERAL CONTEXT AND AIMS
General context
In 1859 Darwin formulated his answer to the one question which had been
troubling biologists over the centuries: 'Why are there so many different
species?’. The answer he proposed explained how species evolved by random
variation with an associated selection of those phenotypes with the highest
fitness. He later added the theory of sexual selection to explain the presence of
some unfavourable traits in extant species (Darwin, 1871). In addition to this,
speciation in sexually reproducing organisms is explained by reproductive
isolation between subpopulations, descended from a single interbreeding parent
population. More recently, various authors have altered the focus point of this
question toward: ‘Why does one group have so many species and another so
few?’ or, if the number of extant species is held as a measure for evolutionary
success, ‘why is one group more successful than another?’. Turner (2007)
formulates this question in terms of rate and asks: ‘What makes some groups
diversify faster and more extensively than others?’. This question becomes even
more relevant if a great discrepancy in evolutionary success is noted between
closely related groups. The process of evolution in which very low constraints
apply, strongly increasing rate and quantity of the speciation process in a single
group is often referred to as adaptive radiation. Various definitions of this term,
however,
exist.
According
to
Simpson
(1953),
it
is
the
simultaneous
diversification of a lineage into numerous sublineages and specializations, while
Seehausen (2004) states it to be the evolution of ecological and phenotypic
diversity within a rapidly multiplying lineage. In conclusion, a general consensus
could be to describe adaptive radiation as an evolutionary process in which both
the rate of evolution as well as the diversity obtained increase significantly. In
this context, the rate of evolution can be viewed as both the rate at which new
characters get fixed in populations as well as the rate at which new species1
arise.
1
Mayr (1996, 2001) defined species as groups of interbreeding natural populations that
are reproductively isolated from other such groups.
1
1 GENERAL CONTEXT AND AIMS
In addition, all species of a radiation constitute a monophyletic group, and they
often share some innovative trait or sets of traits (sometimes called key
innovations (Liem, 1973)) (West-Eberhard, 2003).
Currently, the interest in describing global biodiversity has again stimulated
modern biologists to further contemplate on the evolutionary mechanisms
generating this diversity. In this context, various studies have dealt with groups
showing clear examples of such radiation processes, like Darwin’s finches,
Caribbean anole lizards, reef-dwelling parrotfish, African cichlids,… and studies
on these groups continue to date (e.g., Grant & Grant, 2002; Losos et al., 2003;
Streelman et al., 2002; Albertson et al., 2003, 2005; Turner, 2007).
A similar case of adaptive radiation is found in the evolution of catfishes of the
family Loricariidae. This family is the most speciose of all Siluriformes (ca. 23%
of all species) and the fifth most speciose of all teleosts (ca. 3% of all species),
containing 716 species over 96 genera (Nelson, 2006; Ferraris, 2007) (fig. 1.1).
Members of this family are also characterized by the peculiar trophic niche they
occupy. All loricariids exhibit a large ventral suckermouth, which allows them to
adhere to smooth surfaces and in addition, their trophic morphology has evolved
into a scraping feeding tool (fig. 1.1). The use of this specialized feeding
mechanism for (algae) scraping is, however, not obligatory since a wide variety
of items associated with underwater substrates have been found in the diets of
loricariids (Aranha et al., 1998; Delariva & Agostinho, 2001; Nelson, 2002).
Loricariidae belong to the superfamily of the Loricarioidea (fig. 1.2; 1.3), of
which Schaefer & Lauder (1986, 1996) have argued that the various families
constructing it exhibit a trend of increasing morphological complexity in which
various elements have been decoupled2, creating opportunities for the evolution
of new structures and functions (see 5.2). In members of the more basal
loricarioid family of the Callichthyidae a suckermouth and algae scraping
apparatus is still lacking; but the mouth already has a more ventral position than
most non-loricariid siluriform families, while in the family of the Astroblepidae
such a suckermouth is present, although no algivory is known (fig. 1.3).
2
A decoupling of biological components refers to the unlinking of developmental
pathways, tightly linked functions, aspects of stereotyped behaviour patterns, mechanical
associations of bones, ligaments and muscles, or the reduction of a high genetic
correlation (Lauder et al., 1989). The recognition of ‘new structures’ (sometimes termed
novelties) varies among literature, but in many cases so-called new structures are
actually the result of profound decoupling events (Geerinckx, 2006).
2
1 GENERAL CONTEXT AND AIMS
The Loricarioidea in general and Loricariidae specifically have to date not been
the subject of many studies on the mechanisms of radiation in this superfamily,
perhaps due to their specioseness. As mentioned, Schaefer & Lauder (1986,
1996) have tackled the matter of the decoupling of various cranial elements as
one of the mechanisms facilitating the evolution of such highly specialized
morphologies, but most studies dealing with this group have been limited to a
descriptive morphology of the adult, often in the context of phylogenetic
analyses (e.g. Schaefer, 1990; Arratia & Huaquín, 1995; Reis, 1998; Aquino &
Schaefer, 2002). A more thorough approach, focused on the evolutionary
mechanisms behind the radiation in Loricarioidea could yield valuable information
on the evolution of algae scraping in this group, since it occurred at both the
superfamily and family level, with clear and recognizable steps of increase in
morphological complexity. An algae scraping trophic morphology has also
evolved in the family of the Mochokidae, a catfish group found on the African
continent (fig. 1.4). In contrast to the Loricarioidea, however, algivory in this
family has evolved at an infrafamiliar level. It has not led to a radiation of
comparable diversity, since the family comprises only 193 species, which are
confined to 10 genera, with algivory present in only 57 species belonging to 3
genera (Friel & Vigliotta, 2006; Ng & Bailey, 2006; Wright & Page, 2006;
Ferraris, 2007; Vigliotta, 2007).
Aims of the study
This evolution of algae scraping trophic specializations in both loricarioids and
mochokids has formed the basis of an ongoing FWO project: ‘Evolutionary
adaptiveness for a highly specialized feeding niche: algae scraping in tropical
catfishes’ led by the Evolutionary Morphology of Vertebrates lab at the Ghent
University (UG). For this project, partnership was found in the Laboratory for
Functional Morphology of the Antwerp University (UA) and the Ichthyology
Department of the Africa Museum at Tervuren (RMCA). The project is interpreted
as a study of morphology and function throughout the ontogeny of both
plesiomorphic and specialized members of both loricarioids and mochokids. This
3
1 GENERAL CONTEXT AND AIMS
way, the extent of convergence in both groups can be estimated, but, in
addition, a comparison of developmental sequences of related taxa within a
phylogenetic
framework
also
serves
the
purpose
of
identifying
possible
heterochronic events (Mabee & Trendler, 1996). As it is risky to infer process
from pattern (Hanken & Wake, 1993), adult morphology alone may be a poor
guide to these processes. Therefore, an adequate assessment of differences in
ontogenetic patterns between the different families will substantially improve my
understanding of morphological divergence (Straus, 1985).
In this context, the major aspects dealt with in this project are:
-
The ontogeny of a loricariid (Ancistrus cf. triradiatus) (T. Geerinckx – UG).
-
The functional morphology of several loricariid fishes (A. cf. triradiatus,
Otocinclus vestitus, Farlowella acus) (T. Geerinckx, K. Nijs, D. Maes – UG).
-
A
biomechanical
study
using
X-rays
and
EMG
of
a
loricariid
(Pterygoplichthys lituratus) (T. Geerinckx – UG, A. Herrel – UA).
-
A kinematic analysis of suction feeding in a callichthyid (Corydoras
splendens) (T. Lieben, A. Herrel – UA).
-
A comparative study of non-algae scraping (Synodontis sp.) and algae
scraping (Atopochilus savorgnani) African Mochokidae (C. Ide, J. De
Puysseleir, F. Huysentruyt, T. Geerinckx – UG).
And, finally, the subject of this thesis:
-
The ontogeny of a basal, non-algae scraping, loricarioid (Corydoras
aeneus).
The analysis of such a basal, non-algae scraping representative allows to obtain
a view on the plesiomorphic situation within loricarioid evolution. Groups that are
situated at a more basal position in a phylogeny generally exhibit characters that
have a more plesiomorphic character (i.e. more resembling the situation found in
the common ancestor). These plesiomorphic characters are evidently also found
throughout ontogeny and are not confined to the adult morphology. Therefore,
the entire ontogeny of these groups can be considered as a plesiomorphic state,
which makes the study of these ontogenetic trajectories highly informative. A
4
1 GENERAL CONTEXT AND AIMS
comparison of these trajectories to those of more specialized groups would
indicate the point(s) and manner(s) of divergence in both groups, yielding
information on the mechanisms behind the evolution of a specialized morphology
in the latter group. It is long known that ontogeny and evolution are intimately
and reciprocally interrelated, since evolutionary changes in morphology require
changes in development that produce relevant structures of interest, whose
variation provides the material for evolution by natural selection (Klingenberg,
1998).
The subject of this research was subsequently divided into five main objectives:
•
To provide a detailed description of the adult cranial and
postcranial morphology, identifying the different structures
and determining their homology in comparison to related
taxa. This allows a more precise comparison of the ontogeny
of those structures in my study species with a loricariid
•
species (see 3.1; 3.2).
To perform an analysis of the early ontogeny of the species
through an analysis of both external morphology as well as of
growth. Such an analysis points out the timing and rate of the
development and the subsequent shifts in developmental
properties throughout ontogeny. These results also provide a
framework onto which the results of the ontogeny of cranial
•
structures can be mapped (see 4.2; 4.7; 5.1).
To
present
description
a
of
detailed
the
(macroscopic
ontogeny
of
the
and
microscopic)
cranial
structures:
chondrocranium, osteocranium and musculature (see 4.3;
•
4.4; 4.5) and of the postcranial skeleton (see 4.6).
To present an overview of the ontogeny of Corydoras
pygmaeus, a miniaturized congener of my subject species, in
order to determine rigidity of the ontogenetic pattern in
Callichthyidae and the effects of heterochronic events acting
•
on it (see 4.7).
To compare all these results with similar results described for
Ancistrus cf. triradiatus by Geerinckx (2006) (see 5.1).
5
1 GENERAL CONTEXT AND AIMS
•
Finally, to perform the implementation of all above results
and comparisons in the context of the evolution of algae
scraping and subsequent radiation in Loricariidae (see 5.2)
As a general hypothesis in this research I expected the main differences found
in the adult morphology of both a specialized versus a basal loricarioid to
originate early in ontogeny.
The evolution of algae scraping has generally been believed to have evolved via
one of two possible evolutionary patterns. A first option consists of an
evolutionary pattern in which the suckermouth attachment develops first,
followed by the development of the scraping feeding mode, while in a second the
order of events is the other way around (Adriaens, 2003; Geerinckx, 2006).
From a phylogenetic perspective, the first option would appear to be the most
plausible, since in the more basal family of the astroblepids such a suckermouth
has evolved without the evolution of an algae scraping feeding apparatus.
Therefore, I hypothesized that a suckermouth has evolved prior to an algae
scraping feeding mode and it was expected that the earliest differences in the
ontogenetic patterns of callichthyids and loricariids would be related to the
formation of such a suckermouth. By studying the early ontogeny in both
families, I further expected to find evidence that corroborates the decoupling
hypothesis of Schaefer & Lauder (1986, 1996). Finally, the possibility existed
that, next to the formation of a suckermouth and/or algae scraping mechanism,
additional differences between both ontogenies were found, which had also
facilitated the evolution of both elements in the more specialized groups.
Outline of this thesis
The body text of this thesis is divided into five main chapters.
In the first chapter, a general introduction is provided, followed by the general
aims of the study itself, in which the central questions of this research are
formulated.
6
1 GENERAL CONTEXT AND AIMS
The second chapter presents an overview of the material used in this study and
of the general methods used in the study of these specimens. This chapter is
then referred to in all chapters presenting my results under a ‘brief material and
methods’ section. Specific methods used only in certain parts of my study are
also commented on in these sections instead of in the general ‘Material and
methods’ chapter.
The third chapter is the first chapter in which results are presented. It deals with
the adult morphology of the species studied and is divided into two parts: one
discussing the skeletal morphology and another dealing with the species’ cranial
and body musculature. In this dissertation I chose to start with a description of
the adult morphology, since ontogeny is best studied with a thorough knowledge
of its endpoint.
In chapter four, the ontogenetic component of the thesis is dealt with. It is
divided into seven parts, of which the first six deal with Corydoras aeneus itself.
The first part describes the species’ egg morphology while the second part deals
with early development and growth. In parts three to six results on the ontogeny
of the species’ morphology are presented, dealing with respectively: the
chondrocranium, osteocranium, cranial musculature and appendicular skeleton.
Finally, in the last part, an overview of the ontogeny of a miniaturized congeneric
(Corydoras pygmaeus) is presented.
In the final chapter, these results are discussed. In the first part of this chapter,
an overview of ontogeny of Corydoras aeneus is presented and important
thresholds in the development of this species are recognized. This is also done
for C. pygmaeus and Ancistrus cf. triradiatus and the occurrence of these
thresholds in all three species are placed in an evolutionary perspective. In a
second part of this chapter, all data gathered in this study, combined with
additional data from literature are used to create an overview of the evolution of
algae scraping in the superfamily of the Loricarioidea in general. Finally, in the
last part of this chapter general conclusions of this thesis are formulated.
7
1 GENERAL CONTEXT AND AIMS
8
Chapter
2
Material and Methods
2 MATERIAL AND METHODS - MATERIAL
2.1 MATERIAL
Choice of species
The general aim of this study was to perform a detailed microscopical
anatomical study of the development of the cranial elements in a basal loricarioid
representative. To obtain this level of detail (using histological serial sections)
through an entire ontogenetic series for both soft and hard tissues, time did not
allow the study of multiple species and therefore, the research needed to be
limited to a single species. Subsequently, the choice of the species became
highly important, since it needed to reflect the plesiomorphic loricarioid condition
as well as possible. This would allow the use of results as a basal model for
further comparative research on members of this group. Therefore, as a first
option, members of the group Nematogenyidae + Trichomycteridae, the
sistergroup to all other loricarioids (fig. 1.2), were taken into consideration.
However, none of the species in both families showed the possibility for
generating a reliable ontogenetic series, mainly due to the scarcity of material
and the absence of good breeding protocols. In addition, those species which
were more readily available were the ones showing a high degree of
specialization. This way, e.g. the nematogenyid family is monotypic, with its only
extant species Nematogenys inermis appearing only in Chile where it is very rare
(Ferraris, 2007; de Pinna, 2003). Trichomycterids, on the other hand, are very
speciose holding 207 species in 41 genera (Ferraris, 2007), but very few data
exist on the reproductive behaviour of the various species (needed to create
breeding protocols). In addition, a high number of species which belong to this
family are highly specialized (e.g. parasitic members in Vandellinae and
Stegophilinae (de Pinna & Wosiacki, 2003)), which also affects breeding as well
as their suitability as a plesiomorphic model.
The Callichthyidae, on the other hand, the sister family to Scoloplacidae +
Astroblepidae + Loricariidae (fig. 1.2), does include several species suited for
breeding in captivity. The family comprises 194 species distributed over 8 genera
(Ferraris,
2007):
Callichthys,
Dianema,
Hoplosternum,
Lepthoplosternum,
Megalechis, Aspidoras, Scleromystax, and Corydoras. The latter genus is the
9
2 MATERIAL AND METHODS - MATERIAL
most speciose genus of the family (152 species (Ferraris, 2007)), widespread in
South America (Gosline, 1940; Nijssen, 1970; Kramer & Braun, 1983) and well
known among aquarists for its many ornamental species (Burgess & Quinn,
1992). Given this and given the fact that several studies on members of the
genus in general have already been published (Strauss, 1985; Howes & Teugels,
1989; Reis, 1998, Olivera et al., 1992, 1993), it was made my genus of choice.
In addition, most of the studies done on the genus deal with the species
Corydoras paleatus and C. aeneus (fig. 2.1). Therefore, the choice of one of both
was logical. In comparison to the studies on C. paleatus, those published on C.
aeneus showed a wider variety of subjects dealing with both morphology,
physiology (e.g. Kramer & McClure, 1980, 1981; Kramer & Braun, 1983; Oliveira
et al., 1992, 1993; Shiba et al., 1982; Sire & Huysseune, 1996; Huysseune &
Sire, 1997) as well as reproductive biology (Kohda et al., 1995, 2002; Pruzsinsky
& Ladich, 1998). In addition, C. aeneus is also very popular in the trade of
freshwater ornamental fish and it is annually bred and shipped in large quantities
all over the world (Tamaru et al., 1997). As a result, both specimens and
breeding protocols (e.g. Fuller, 2001) for this species are easily available. Given
this, C. aeneus (Gill, 1858) was chosen as the species of interest in this study.
However, despite it being commercially bred, almost nothing is known about its
ontogeny. Some attention has been paid to the early ontogeny of some aspects
of the head in other callichthyids (Hoedeman, 1960c), but still a lot of relevant
information is lacking. The same accounts for the adult morphology: a complete
overview of the cranial and postcranial morphology is absent, despite of its
relevance for ongoing phylogenetic research on Loricarioidea, to which these
callichthyids belong (Reis, 1998; Britto & Castro, 2002). In studies by Reis
(1998) and Britto (2003), morphological data provided a phylogenetic framework
for both the family (Callichthyidae), and subfamily (Corydoradinae) and
Shimabukuro-Dias et al. (2004) used molecular data to investigate the
phylogenetic affinities within the family. The phylogeny of the genus itself,
however, remains unresolved. (Schaefer, 1990; Reis, 1998; Aquino & Schaefer,
2002), as well as the generic relationships within the callichthyids (Reis, 1997,
1998) and very little information exists on the phylogeny of the highly diverse
genus Corydoras.
In this context, even the species of choice in this research still raises several
questions in a taxonomical perspective. Where most Corydoras species have a
10
2 MATERIAL AND METHODS - MATERIAL
distribution that is very often confined to a single basin, the distribution of
Corydoras aeneus, covers almost the entire South American tropics. Therefore, it
has been argued that the current definition of the species probably comprises
several separate nominate species. At this point, several geographically distinct
populations of Corydoras aeneus can be discerned by body coloration, a
characteristic
which
disappears
in
preserved
specimens
(Isbrücker, pers.
comm.). However, since no major additional morphological differences were
identified to date and since the focus of this research is situated at a
morphological level, this does not affect the results presented here. To allow a
future identification of the species used, five additional specimens from the same
strain were deposited at the Zoology Museum of the Ghent University (UGMD
175375-379).
Material examined
Breeding
Specimens were commercially obtained from an aquarium shop (Poisson d’Or Belgium). Since origin of the adult specimens is unknown and since long-term
breeding under artificial conditions could affect early development, a projection
of the results of this study onto the whole species should be considered with
caution. Still, since my main interest is to describe the general patterns of early
development of the species (determined by size and age) the use of this brood
stock seems justifiable, apart from the practical reasons (ability to obtain and
breed them). A total of 35 specimens (25 were males), were put together in a
1.0 x 0.5 x 0.6 m tank. The tank had a 3 cm sand layer, a box filter, a heating
apparatus
and
was
heavily
furnished
with
plant
material
(Microsorum,
Echinodorus and Anubias-species). A temperature of 24-26°C, pH-level of 8-8.5
and hardness of 9-12° was maintained. The specimens were fed on a diet of
commercial food flakes (TetraMin). In the tank, dry season conditions were
imitated by lowering the water level to 10 cm and lowering light intensity for a
period of 20 to 30 days. This was followed by an imitation of rain season
conditions (daily adding of fresh, colder water and sufficient aeration) (Fuller,
2001). About a week after the start of this artificial rain season, several egg
11
2 MATERIAL AND METHODS - MATERIAL
clutches (maximum 10) were found on the glass walls of the tank for several
consecutive days.
Hatching
After hatching, which took place at about 3 days after fertilization, the
hatchlings were moved to a smaller tank with similar water conditions (24-26°C
and pH 8-8.5) and a photoperiod of 12 hours of darkness/light. From three days
up to three weeks after hatching the larvae were fed on a diet of decapsulated
Artemia-nauplii. After that, the diet was changed to crushed flakes and two
weeks later to the adult diet of whole flakes (TetraMin).
Collection of specimens
Specimens from various nests were removed post-hatching at different ages,
sedated and killed with an overdose of MS-222 (3-aminobenzoic acid ethyl ester,
Sigma). Small specimens were photographed with the use of a ColorView digital
camera, mounted onto an Olympus SZX9 binocular microscope. Afterwards,
these specimens were measured to the nearest 0.1 mm using AnalySIS 5.0
software. Larger specimens were measured to the nearest 0.5 mm. In the
smallest specimens in which the caudal fin had not yet developed and in which
notochord flexion did not yet occur, notochord length (NL) was used as an
alternate measurement to SL (figure 4.3B). The combined use of NL and SL in
the analysis of growth is a method previously used in other ichthyological studies
(see Kavanagh & Alford, 2003). After measurement, all specimens were given a
collection number (AA, AB,…) (table 2.1) and preserved in a buffered 4%
paraformaldehyde fixative (pH 7.0).
Additional species
In order to determine the importance and extent to which heterochronic events
have played in callichthyid evolution, in chapter 4.7, a brief comparative study on
a miniaturized congener was done. Species of choice for this was Corydoras
pygmaeus (for a motivation on this choice see 4.7). Specimens were kept, bred
and collected as described for C. aeneus. Specimens were also given a collection
number (Cp 01, Cp 02,…) (table 2.2).
12
2 MATERIAL AND METHODS - MATERIAL
TABLE 2.1 Specimens of Corydoras aeneus used in this study, excluding specimens that
were only measured for use in chapter 4.2. (AB, Alcian Blue; AR, Alizarin Red; T,
Toluidine Blue).
No.
Coll. No.
NL/SL (mm)
Age (Ph)
Method
Staining
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
BC
BG
AU
Y
AM
AV
BN
AA
AX
AN
BP
AW
BR
AD
AY
AJ
AP
AZ
AG
BT
BW
BA
BK
AR
BB
BZ
BJ
CA
BQ
BM
CC
BO
CD
BS
CF
CI
CH
BV
BY
CK
CB
CE
CM
CO
CQ
CS
CL
CG
CJ
CN
CT
3.3
4.0
4.0
4.4
4.8
4.8
4.9
5.1
5.2
5.3
5.3
5.4
5.6
5.7
5.7
5.7
5.9
6.2
6.2
6.3
6.4
6.6
6.6
7.2
7.2
7.3
7.7
7.8
7.9
8.2
8.6
9.1
9.3
9.7
9.8
10.7
11.1
11.2
11.7
12.4
12.5
13.1
13.7
13.9
15.0
15.5
15.6
15.6
16.0
16.9
18.0
0h
17 h
1d
3d
9d
2d
3d
4d
4d
9d
4d
4d
5d
5d
5d
7d
11 d
6d
7d
6d
8d
7d
10 d
13 d
8d
10 d
9d
11 d
13 d
11 d
12 d
12 d
13 d
14 d
14 d
18 d
16 d
16 d
18 d
20 d
20 d
22 d
22 d
25 d
29 d
33 d
30 d
24 d
27 d
33 d
37 d
Sections (2µm)
Sections (2µm)
Clearing
Sections (2µm)
Clearing
Clearing
Sections (2µm)
Clearing
Clearing
Clearing
Sections (2µm)
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Sections (2µm)
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Sections (2µm)
Clearing
Clearing
Clearing + CT
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Clearing
Sections (5µm)
Clearing
Clearing
Clearing
Clearing
Clearing
T
T
AB + AR
T
AR
AB + AR
T
AR
AB + AR
AB + AR
T
AB + AR
AR
AB + AR
AR
AR
AB + AR
AR
T
AR
AB + AR
AB + AR
AR
AB + AR
AR
AB + AR
AB + AR
AB+ AR
AB +AR
AR
AB + AR
T
AB + AR
AR
AR
AR
AB +AR
AB + AR
AR
AB + AR
AB + AR
AR
AR
AR
T
AB + AR
AB + AR
AB + AR
AB + AR
AR
13
2 MATERIAL AND METHODS - MATERIAL
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
CR
CP
CZ
CW
DE
DH
DJ
DF
DW
EI
EL
DV
EJ
EH
EE
EK
DT
DS
DU
EA
EB
DZ
ED
EC
DY
18.5
19.0
19.5
20.5
22.0
25.0
26.0
30.0
35.0
35.0
36.0
36.0
37.0
37.0
39.0
40.0
41.0
43.0
43.0
43.0
44.0
45.0
47.0
48.0
50.0
41 d
37 d
50 d
50 d
70 d
80 d
90 d
70 d
-
Clearing
Clearing
Sections (5µm)
Clearing
Clearing
Clearing
Sections (5µm)
Clearing
Clearing
Dissection
Clearing
Clearing
Clearing
Dissection
Sections (5µm)
Clearing
Clearing
Clearing
Clearing
Dissection
Dissection
Dissection
Dissection
Dissection
Sections (5µm)
AB +AR
AB + AR
T
AB + AR
AB + AR
AB + AR
T
AB + AR
AR
AB + AR
AR
AB + AR
T
AB + AR
AR
AR
AR
T
In addition to table 2.1, figure 2.2 presents a growth curve showing the different
specimens of Corydoras aeneus used, depicting their specific treatment.
TABLE 2.2 Specimens of Corydoras pygmaeus used in this study, excluding specimens
that were only used for measurements in chapter 4.7. (AB, Alcian Blue; AR, Alizarin Red;
T, Toluidine Blue). (*indicates the use of an adjusted clearing and staining procedure,
see 2.2)
No.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
14
Coll. No.
Cp
Cp
Cp
Cp
Cp
Cp
Cp
Cp
Cp
Cp
Cp
Cp
Cp
Cp
Cp
01
02
08
09
04
10
16
06
17
22
47
21
31
25
56
NL/SL (mm)
Age (Ph)
Method
3.7
4.3
4.8
5.4
5.6
5.9
6.6
7.5
7.9
8.5
9.1
9.2
10.8
10.0
11.0
0h
1d
3d
4d
3d
7d
12d
14d
13d
22d
16d
22d
35d
29d
-
Clearing
Clearing
Clearing
Clearing*
Clearing
Clearing*
Clearing
Clearing
Clearing*
Clearing*
Clearing
Clearing*
Clearing*
Clearing*
Clearing
Staining
AB
AB
AB
AB
AB
AB
AB
AB
AB
AB
AB
AB
AB
AB
AB
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
AR
AR
AR
AR
AR
AR
AR
AR
AR
AR
AR
AR
AR
AR
AR
2 MATERIAL AND METHODS - METHODS
2.2 METHODS
In toto clearing and staining
As mentioned in table 2.1 and 2.2 and in figure 2.2, several specimens were
subjected to a clearing and staining procedure. For this purpose, a slightly
modified version of the protocol as described by Taylor & Van Dyke (1985) was
used (table 2.3). Their method differs from Hanken & Wassersug’s (1981)
method in, a.o., avoiding extensive damage from the acid cartilage stain on the
bony structures by the alcohol series and the use of borax, and by using KOH
instead of trypsin for the clearing. Small specimens often benefit from trypsin
instead of KOH though, and therefore, in these cases, KOH was replaced by
trypsin in the first clearing step of the protocol (table 2.3). Since the acetic acid
used in the cartilage staining step could decalcify bone, this step was usually not
exceeded over 12h and for control, this step was omitted in several specimens
(table 2.1; 2.3).
TABLE 2.3 Clearing and staining procedure adapted from Taylor & Van Dycke (1985).
Step
Dehydration
Solution/Action
Duration
50% alcohol
12h
75% alcohol
12h
96-100% alcohol
12h
96-100% alcohol
12h
Neutralization
alcian blue (8GX, Sigma): 0.009-0.03% in (40ml glacial
acetic acid/60ml 96-100% alcohol)
saturated borax solution (Na2B4O7.10H2O)
Bleaching
3-10% H2O2 in 0.5% KOH
0.5h-...
Clearing
1-4% KOH / 0.15% trypsin in 30% NaBO3
12h-...
Bone staining
alizarin red (Sigma): 0.5% KOH in 0.1% alizarin red,
until colour switches from deep red to purple (stir)
24h
Further clearing
0.5-4% KOH
12h...
Preservation
25% glycerin + 75% 0.5% KOH
12h
50% glycerin + 50% 0.5% KOH
12h
75% glycerin + 25% 0.5% KOH
12h
100% glycerin
storage
Cartilage staining
8-24h
48h
15
2 MATERIAL AND METHODS - METHODS
The method described in table 2.3, however, proved unreliable in the clearing
and staining of Corydoras pygmaeus specimens, where several specimens were
lost in the procedure due to overstaining with alcian blue. To counter this, the
adapted method by Gavaia et al. (2000) was used as described in table 2.4 for
the specimens marked with * in table 2.2.
TABLE 2.4 Clearing and staining procedure adapted from Gavaia et al. (2000).
Step
Washing
Solution/Action
Duration
TBST (50 mM Tris, pH 7.4; 150 mM NaCl; 0.1% Triton
X-100)
0.25h
aq. dest.
0.25h
Neutralization
alcian blue (8GX, Sigma): 0.009-0.03% in (40ml glacial
acetic acid/60ml 96-100% alcohol, filtered) (pH ca. 1)
0.1% 1% KOH in alcohol
Hydration
80% alcohol
2h
70% alcohol
2h
50% alcohol
2h
30% alcohol
2h
aq. dest.
2h
Clearing
0.5-5% KOH
4h-...
Bone staining
alizarin red (Sigma): 0.5% KOH in 0.1% alizarin red,
until colour switches from deep red to purple (stir)
0.5h-2h
Final clearing
0.5-4% KOH with a drop of H2O2
0.5h-4h
Preservation
25% glycerin + 75% 0.5% KOH
2h
50% glycerin + 50% 0.5% KOH
2h
75% glycerin + 25% 0.5% KOH
2h
100% glycerin
storage
Cartilage staining
Washing
0.2-0.5h
48h
Dissections
For the study of soft tissue, such as ligaments and muscles, dissections were
performed on otherwise untreated fixed specimens (marked as ‘dissection’ in the
method section in table 2.1). Visualization of muscle fiber arrangement was
enhanced by the use of an iodine solution (Bock & Shear, 1972). For the study of
cartilaginous and bony elements dissections were also performed on cleared and
stained specimens. This allowed the study of separate elements and of those
regions that were invisible in an overall dorsal, ventral and lateral view (like e.g.
16
2 MATERIAL AND METHODS - METHODS
a detailed view of the gill arches, a ventral view of the neurocranium floor, a
median view of the lower jaw,…). Study of specimens was done using an
Olympus SZX7 stereomicroscope, equipped with a camera lucida for generating
line drawings. Digital images of specimens (and eggs, see 4.1) were taken using
an Olympus SZX9 stereomicroscope, equipped with a ColorView 8 digital camera
driven by AnalySIS 5.0 software (Olympus).
Serial sectioning
In order to study the detailed anatomy of the earliest stages in ontogeny, serial
sections were used. For this purpose, several specimens (table 2.1) were
embedded in a Technovit 7100 medium and 2 or 5 µm thick sections were made
using a Reichert-Jung Polycut microtome. After this, these sections were
mounted onto glass slides, stained with toluidine blue and covered. Examination
and digital imaging of these sections was done on a Reichert-Jung Polyvar light
microscope, also equipped with a ColorView 8 camera driven by AnalySIS 5.0
software (Olympus).
TABLE 2.5 Technovit 7100 embedding procedure.
Step
Solution/Action
Duration
Vacuum fixation
4% buffered formalin
days to weeks
Washing
Tap water
8h
Decalcification
Decalc (Histolab)
36h
Washing
tap water
5h
Dehydration
30% alcohol
12h
50% alcohol
12h
70% alcohol
12h
96% alcohol (two times alcohol renewal)
12h
Technovit 7100 solution A
min. 24h
Technovit 7100 solution A renewal
min. 48h
add Technovit 7100 Harder II
12h
place in deepfreeze
12h
place at room temperature (check progress)
approx. 2h
place in oven
1h
Embedding
Polymerization
17
2 MATERIAL AND METHODS - METHODS
3D-reconstructions
For the 3-dimensional visualization of the detailed anatomical structures
studied through the use of serial sections, graphical reconstructions were
generated. To do this, a series of digital images of the serial sections were made
at fixed distance intervals between the slides photographed. This way, e.g., an
image was taken every 4 slides of a 2 µm thick series, leaving a fixed 8 µm
distance interval between all images. These intervals were then recalculated into
a voxel coordinate system (3D-pixels), after which this value was assigned to
images when uploading them into the Amira 3.1.1 (T.G.S.) software. There,
images were aligned and the different anatomical structures of interest were
manually traced. After this, surface models of each of these structures were
generated, and digitally smoothed. 3D-meshes of these structures obtained this
way were exported using the .dxf format, in which the object’s coordinates are
included, allowing an automatic repositioning of all structures in relation to each
other in other software packages used. In a next stage, these .dxf meshes were
imported into the Rhinoceros (McNeel) software package. This software allowed
rendering of different surfaces onto the 3D-meshes as well as the ability to
obtain bitmap (.tiff) images taken from different angles of view. Afterwards, the
reconstructions obtained were compared to cleared and stained specimens of
similar size. Several comparisons made this way showed no substantial
differences in the size and shape of comparable elements (hard tissue) between
both methods.
18
Chapter
3
Adult Morphology
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
3.1. ADULT OSTEOLOGY
Modified from:
Huysentruyt F & Adriaens D
Descriptive osteology of Corydoras aeneus
(Gill, 1858) (Siluriformes: Callichthyidae)
Cybium 2005, 29(3): 261-273.
Abstract
Both the cranial and postcranial osteology of Corydoras aeneus is described based on
the study of five adult specimens. The results were compared to the condition described
for other loricarioid fishes. Most results corresponded with the expected pattern based on
studies dealing with parts of the callichthyid osteology, although some differences were
observed. One of these differences was the presence of a suprapreopercular bone, a
bone that was previously unmentioned in any callichthyid species. Further, several
homologies were resolved and mainly confirmed existing hypotheses. This way, for
example, based on the presence of several branches of the lateral line system, the
compound dorsolateral bone of the otic region was identified as the posttemporopterotico-supracleithrum. Further, the presence of several otoliths in the compound bone
forming the neurocranial floor in the occipital region, confirmed the fact that the bone
comprized out of a fusion of both basi- and exoccipital bones. This study, however, failed
to
resolve
some
other
questions
regarding
homologies
(i.e.
lacrymo-antorbital,
suprapreopercle) pointing out the need for further ontogenetic research. In this light, this
study provides the basis for this further comparative and ontogenetic research on
callichthyids specifically and on loricarioids in general.
Brief introduction
The aim of this part of the study is to provide a full description of the osteology
of the species, as a basis for further ontogenetic research on this species, as well
as to contribute to future phylogenetic studies.
19
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
Brief material and methods
The specimens examined are presented in table 2.1. For clearing and staining
procedures, see 2.2. For the nomenclature of the skeletal elements I followed
Schaefer (1990) and Reis (1998). The homology of the autopalatine with the
dorsal part of a premandibular arch follows Daget (1964) and Jarvik (1980).
Results and discussion
In Corydoras aeneus the neurocranium is pyriform, with a small ethmoid and
orbital region, broadening at the temporal region into a large occipital region (fig
3.1B).
Ethmoid region (figs. 3.1, 3.2)
The mesethmoid in C. aeneus is narrow anteriorly and broadens posteriorly
(see also Fink & Fink, 1996; Arratia, 2003a). It lacks cornua and broadens
substantially toward its posterior margin. Although a general trend toward a
reduction of the cornua is present in all catfishes (Lundberg, 1982; Schaefer,
1987), a total lack of these cornua is only present within the Callichthyidae, with
the exception of the species of the former genus Brochis, in which extremely
reduced cornua are still present (Schaefer, 1990; Reis, 1998). Hoedeman
(1960d) suggested an initial formation of these cornua in Hoplosternum and
Callichthys, but without further ossification, implying a reduction (presumably as
a result of allometric growth). Further ontogenetic research will have to reveal
whether this also holds for C. aeneus. On its posterior margin the mesethmoid
contacts the frontals with a V-shaped suture dorsally and a W-shaped wedge
with the prevomer ventrally. The posterior, V-shaped suture with the frontals in
C. aeneus is also found in other callichthyids, in Nematogenys inermis and in the
Scoloplacidae, but not in Loricariidae, Astroblepidae and Trichomycteridae, which
reflects the plesiomorphic condition in the Loricarioidea (Schaefer, 1990).
Laterally, the mesethmoid contacts the lateral ethmoids and the autopalatine
ventroposteriorly. In addition, it is connected by a ligament to both the reduced
premaxilla and the maxilla. The presence of such ligaments is a condition which
20
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
all Callichthyidae share with the Scoloplacidae, Loricariidae and Astroblepidae
(Schaefer, 1990). The well developed lateral ethmoids, together with the
mesethmoid, the frontal, the autopalatine and the lacrymo-antorbital, surround
the nasal cavity, which is different from the situation in the Callichthyinae, where
a large depression in the lateral ethmoid forms the total nasal capsule (Reis,
1998). In this cavity the free, tube-like, nasal bone encloses the anterior part of
the supraorbital canal. This canal directly enters the frontal bone at the posterior
margin of the nasal (fig. 3.3) in contrast to the situation in the Callichthyinae
where the supraorbital canal consequently first enters the lateral ethmoid (Reis,
1998). The nasal bone has the typical catfish, tube-like shape, although it only
bears two pores, in contrast to the three pores found in most diplomystids and
primitive loricarioids (Arratia & Huaquín, 1995). The toothless prevomer is dropshaped and forms an elongated V-shaped suture with the parasphenoid
posteriorly. An independent prevomer is, within the Loricarioidea, present in all
families except the Scoloplacidae (Schaefer, 1990; Arratia, 2003a).
Orbital region (figs. 3.1, 3.2B)
The first bone of the infraorbital series of C. aeneus, the lacrymo-antorbital, is
a large, plate-like bone, which forms most of the ventral border of the orbita.
This fusion of the first infraorbital bone, the lacrimal, with the antorbital bone is
common among catfishes (Schaefer, 1990), but in C. aeneus some questions
regarding the true nature and origin of the infraorbital bones remain. A first
question is whether this bone really is the lacrymo-antorbital, merely an
expansion of the lacrimal bone or whether both lacrimal and/or antorbital are
totally absent (Schaefer, 1990; Reis, 1998; Arratia, 2003a). Subsequently, the
homologies of the remaining infraorbital bones and of possible fusions within this
series remain unclear. Regardless, the first two infraorbitals have become platelike in all Callichthyidae, a condition which they share with more primitive nonsiluriform
teleosts, and some other Siluriformes (e.g. Clarias
gariepinus
(Adriaens et al., 1997)). In general, in Siluriphysi, the infraorbital series is
reduced to tube-like bones bearing the infraorbital canal. Plate-like infraorbitals
are therefore believed to be secondarily derived (Fink & Fink, 1996). The
lacrymo-antorbital bears the first part of the infraorbital canal with two of its
sensory pores. This canal further continues through the smaller second
infraorbital bone (fig. 3.3). Furthermore, the anterior extension of this
21
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
infraorbital canal into the first bone of the series occurs in Corydoras, Aspidoras
and Scleromystax but is not present in other callichthyid species (Schaefer,
1990; Britto, 1998; Reis, 1998). On its dorso-posterior margin the second
infraorbital broadens and connects both the sphenotic, of which the dermal part
is in fact the last infraorbital bone (Gosline, 1975), and the posttemporopterotico-supracleithrum. The orbital skull roof is formed by the two large frontal
bones, separated posteriorly by the anterior cranial fontanel. This fontanel is
divided into two openings by the ossified epiphyseal bridge and is elongated
posteriorly. As in the genera Hoplosternum, Megalechis, Lepthoplosternum,
Dianema and members of the former genus Brochis, the anterior fontanel enters
the parieto-supraoccipital bone in C. aeneus (Reis, 1998). The fontanel itself is
minute, in contrast to that in other callichthyids, where a larger fontanel is
present
(Schaefer,
1990;
Reis,
1998).
In
astroblepids,
scoloplacids
and
loricariids, however, no open cranial fontanels are found. The frontals further
contact the sphenotics laterally and the orbito- and pterosphenoid ventrally. The
frontals, as in other teleosts, bear the supraorbital canal, but in C. aeneus an
additional central pore is present (fig. 3.3). According to Reis (1998) this pore
represents the parietal branch of that canal in other Siluriformes. Arratia &
Huaquín (1995), however, report the absence of a parietal branch as a loricarioid
synapomorphy, but, they, on the other hand, do report the presence of an
epiphyseal branch in several loricarioids. Therefore, and based on the position
and orientation of this pore I believe it to be homologous with this epiphyseal
branch. The wall of the orbital region is formed by the anterior orbito- and the
posterior pterosphenoid, which both ventrally contact the orbital floor at the level
of the parasphenoid. The orbito-, and pterosphenoid in C. aeneus all posses the
typical shape found in other Siluriformes (Schaefer, 1990; Reis, 1998; Arratia,
2003a). The orbitosphenoid is hour-glass-shaped in ventral view and holds a
large foramen. The parasphenoid is fairly narrow anteriorly and bears two
elongated, anterior processes (in between lies the prevomer). Posteriorly, it
broadens widely at the level of the prootics, further ending in a narrow, sharp
region. Anteriorly, the bone shows a strong, midline ridge. The posterior ‘wings’
of the parasphenoid suture with both prootic bones and the posterior tip
connects to the occipital bone complex. Further, the bone is much broader than
what is described for all Corydoradinae in Reis (1998). In between the
orbitosphenoid,
22
pterosphenoid,
parasphenoid
and
prootics,
as
in
most
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
siluriformes, the combined foramen for the fasciculus opticus and the trigeminofacial
nerve
is
situated.
A
connection
between
the
parasphenoid
and
pterosphenoid, thereby splitting the foramen in an orbital and trigemino-facial
fenestra, is absent.
Otic region (figs. 3.1, 3.2B)
The floor of this region is formed by the posterior tip of the parasphenoid,
flanked by the prootic bones. These square-shaped prootics further connect to
the sphenotics and posttemporo-pterotico-supracleithrum laterodorsally and to
the occipital complex posteriorly. The prootics enclose the utriculus with the
lapillus and furthermore bear a large foramen. These foramina are not
homologous to the auditory foramen (Schaefer, 1987), although their true
purpose remains unknown. The lateral margins of the otic region are made up by
the sphenotics that are also square and make additional contact with the parietosupraoccipital dorsally and the posttemporo-pterotico-supracleithrum posteriorly.
In the sphenotics, both the supraorbital and infraorbital canal come together into
the otic canal, which exits the sphenotic at its posterior margin and continues
into the posttemporo-pterotico-supracleithrum (fig. 3.3). The latter bone has
often been described as the fusion of both supracleithrum and pterotic with the
absence of the posttemporal (Regan, 1911; Lundberg, 1975), while other
authors described it as the fusion between a compound supracleithrum
(supracleithrum, posttemporal and an ossified Baudelot’s ligament) and pterotic
(Schaefer & Lauder, 1986; Schaefer, 1987, 1990; de Pinna, 1993; Reis, 1998). A
compound nature of this bone is also supported by the presence of several
segments and branches of the lateral line system (Schaefer & Aquino, 2000). At
about one-third of the bone length, the preopercular canal branches off and from
that point on the otic canal continues in the postotic canal, from which, at about
two-thirds of the bone length, a first postotic branch, the pterotic branch,
branches off. The canal then continues through the remaining one-third of the
bone into the posttemporal branch and leaves the bone at its posterior margin
(fig. 3.3). The presence of both the preopercular and pterotic branch indeed
confirm the fact that the pterotic bone is included in this bony complex since
both branches are generally inherent to the pterotic (Schaefer & Aquino, 2000).
This situation, in which only one postotic branch (the pterotic branch) is present,
occurs in all corydoradinae. In other callichthyid species two postotic branches
23
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
are present (Schaefer & Aquino, 2000). Furthermore, the presence of a
posttemporal branch confirms the presence of the posttemporal bone in the
complex. Further, in C. aeneus, the posttemporo-pterotico-supracleithrum also
bears a large articular cavity for the articulation of the pectoral girdle with the
skull. Finally, it remains unclear whether the epiotic bones have disappeared
during the development of C. aeneus or whether they have become incorporated
within other bone complexes (Arratia, 2003a).
Occipital region (figs. 3.1, 3.2B)
The skull roof in this region is formed by one, large, compound bone, the
parieto-supraoccipital. The fusion of the dermal parietal bones and perichondral
supraoccipital bones during ontogeny is typical for Siluriphysi (Bamford, 1948;
Lundberg, 1975; Fink & Fink, 1996). Here, the bone neither bears a posterior
fontanel nor latero-sensory canals and has a large posterior process which
contacts the nuchal plate and the first pair of laterodorsal bony scutes. The
neurocranium floor in this region is formed by the compound occipital bone, a
fusion between the basi- and exoccipitals. Within the Loricarioidea, a similar
fusion only occurs in Scoloplacidae and Callichthyidae (Reis, 1998). Evidence for
the presence of the basioccipital bone within the complex is found in the position
of the bone contacting the parasphenoid anteriorly and articulating with the
centre of the first vertebrae posteriorly, where it forms the posterior rim of the
neurocranium (Rojo, 1991). Another indication is the fact that the bone
encapsulates the asteriscus. Similar, the encapsulation of the sagitta confirms
the presence of the exoccipitals within the bone complex. Further, this bone
contacts the complex vertebral centre and its outgrowths on its posterior side
and the posttemporo-pterotico-supracleithrum through the ossified Baudelot’s or
transscapular ligament on its dorsolateral side (for a nomenclature on this
structure, see Lundberg (1975), Fink & Fink (1981), Schaefer (1987) and Reis
(1998)).
Maxillary bones (figs. 3.1, 3.2)
The highly reduced premaxillary bone is toothless in adult specimens of C.
aeneus and bears a small dorsal process. The absence of teeth on the premaxilla
is common to all callichthyids, although teeth are present in the early stages of
C. aeneus (Machado-Allison & Garcia, 1986; Huysseune & Sire, 1997). The
24
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
maxillary bone is also reduced to a small bone lacking dentition and supporting
the maxillary barbel. In C. aeneus the bone is comma-shaped and bears a small
process on its postero-lateral face. The bone articulates with the autopalatine
through two articular facets which creates a hinge-joint. Both premaxillary and
maxillary bones are ligamentously connected to the mesethmoid and next to that
another ligament connects the maxillary bone to the palatine. A similar highly
mobile and reduced premaxilla is present in all Callichthyidae. The fact that this
increased mobility is caused by a ligamentous junction with the mesethmoid is a
character shared with Astroblepidae and Loricariidae, but not with Scoloplacidae
(Schaefer & Lauder, 1986). The shape and function of the maxillary bone (small,
toothless and supporting the maxillary barbel) is the same as in all Siluriformes,
with the exception of the Diplomystidae and †Hypsidoridae (Grande, 1987; Fink
& Fink, 1996; Grande & de Pinna, 1998; Arratia, 2003a). Also, the presence of a
pair of palatine condyles on the bone is common to all Siluriformes, except
Astroblepidae and Helogenes-species (de Pinna, 1993). The bone’s posterolateral process serves as an insertion site for the retractor tentaculi muscle.
Premandibular arch (figs. 3.1, 3.2)
The autopalatine is rod-shaped and straight, with a flat lateral surface. It bears
a small posterior process, which contacts the lateral ethmoid, a shape which is
different from that in less advanced catfish families like Diplomystidae and
†Hypsidoridae (Grande, 1987; Schaefer, 1990; Arratia, 1992). Anteriorly, the
bone bears a large cartilaginous condyle for articulation with the maxillary bone,
to which it is also ligamentously connected. The posterior process, on the other
hand, bears no cartilaginous tip and is small compared to primitive catfishes
(Arratia, 1992; Reis, 1998), but not as small as in Callichthys (Reis, 1998). The
absence of this cartilaginous tip is variably present among catfishes, but common
to all non-nematogenyid loricarioids (de Pinna, 1993). The process serves as the
insertion site of the extensor tentaculi muscle (Fink & Fink, 1981; Reis, 1998).
Mandibular arch (figs. 3.2A, 3.4)
The metapterygoid bone is nearly triangular and has a narrow, elongated,
anterior process that ends near the autopalatine. This process is single in
Corydoras, as it is in Aspidoras, whereas, in Callichthys, it is bifurcated (Reis,
1998). The metapterygoid itself was first described as a fusion of ecto-, ento-
25
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
and metapterygoid by Howes & Teugels (1989), although other authors reported
the ecto- and entopterygoid to be absent (Regan, 1911; Arratia, 1990; Reis,
1998). In addition, the hypothesis by Howes & Teugels (1989) was, due to the
lack of ontogenetic evidence, contradicted by Arratia (1992), who thus defined
the bone as being the metapterygoid only. Next to this commonly used
hypothesis, Diogo et al. (2001) also hypothesized the catfish hyomandibula as
being a fusion of the hyomandibula and metapterygoid of other teleosts and the
quadrate as being a fused quadrate + symplectic. Next to this, he also proposed
a homology between the metapterygoid and the gymnotiform entopterygoid. In
this case, I chose to follow the hypothesis as presented by Arratia (1990) since
no evidence against it is presented here and it is commonly used in catfish
literature. The metapterygoid further contacts the hyomandibula on its posterior
margin through a serrated suture, as in all Corydoradinae (Reis, 1998). The
metapterygoid
is
also
joint
synchondrally
to
the
quadrate
bone
and
ligamentously attached to the lateral ethmoid. The quadrate bone is a simple,
small, triangular bone, the typical condition found in Diplomystidae, as well as in
most Siluriformes (Arratia, 1992; Reis, 1998). The bone connects synchondrally
to both the metapterygoid and the hyomandibula. On its postero-ventral margin
it articulates with the articular bone complex. This complex is considered to
consist of the fused angular, the articular and the retro-articular bone (Arratia,
2003a). This compound bone is small, not canal-bearing, and connected to the
dentary bone complex. It articulates with the quadrate dorsally and is
ligamentously connected with both the interopercle and posterior ceratohyal
bone. The articular bone complex further bears a laminar coronoid process,
which serves as an insertion site for parts of the adductor mandibulae muscle
complex (Reis, 1998). The last bone of the mandibular arch is another compound
bone called the dentary complex. The bone is a fusion of the mento-meckelium
and the dental bone. It forms the main part of the lower jaw and is toothless in
adult specimens of C. aeneus, a condition that is different from that in the early
ontogenetic stages (Huysseune & Sire, 1997). It bears a small process anteromedially for insertion of the anterior intermandibular muscle. Further it medially
encloses the Meckel’s cartilage. The fact that the Meckel’s cartilage is small and
that no coronomeckelian bone is present are conditions the Callichthyidae share
with Astroblepidae, Loricariidae, Trichomycteridae and several other nonloricarioid catfishes (de Pinna, 1993). As in the articular complex, the dentary
26
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
complex does not bear a part of the preoperculo-mandibular branch of the lateral
line system, a condition shared by all Loricarioidea, except N. inermis (Schaefer,
1990).
Hyoid arch (figs. 3.4A, 3.5)
The hyomandibula articulates with the neurocranium through the sphenotic and
posttemporo-pterotico-supracleithrum. It also bears a large process on its dorsoposterior margin for the articulation with the opercle. The perichondral part of
the hyomandibula is long and bears a bony plate on its ventro-anterior side,
which contacts the metapterygoid and quadrate. On its medial side the bone
articulates with the rest of the hyoid arch through the small interhyal bone. This
interhyal articulates with the posterior ceratohyal, which, in turn, synchondrally
contacts the anterior ceratohyal. The anterior ceratohyal has a twisted surface
with a medial, bony outgrowth and articulates with three branchiostegal rays on
its medial posterior margin and with the larger, fourth ray on its lateral posterior
margin. The anterior part of the hyoid arch consists of both a ventral and a
dorsal hypohyal, both square-shaped and articulating with the ventral, plate-like
parurohyal. The presence of both dorsal and ventral hypohyals in C. aeneus and
in most other Corydoradinae (Reis, 1998) is in contrast to other Loricarioidea.
According to Arratia & Schultze (1990) most catfishes have two pairs of
hypohyals, except for Trichomycteridae, Loricariidae and Callichthyidae, which is
contradicted by my findings. The former study, however, was solely based on
observations on Callichthys, where indeed only the ventral hypohyals are present
(Arratia & Schultze, 1990; Reis, 1998). Schaefer (1987) confirms this and
mentions a loss of the dorsal hypohyal only in Trichomycteridae and Loricariidae,
but contrary to Arratia & Schultze (1990), he also mentions a similar loss in
Astroblepidae.
Branchial arches (figs. 3.5)
In C. aeneus, the branchial basket bears the typical siluriform configuration in
which five branchial arches are present. Only basibranchials II and III are
present as distinct, ossified elements. The posterior copula remains cartilaginous.
Ossified hypobranchials I and II are present, whereas separate hypobranchials
III and IV show no ossification. The fifth hypobranchial is absent. The
ceratobranchials of all five arches are well ossified, bearing cartilaginous tips
27
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
(with exception of the posterior tip of the fifth one). All ceratobranchials support
hemibranchs. The fifth ceratobranchial bears the lower pharyngeal tooth plate
and is the only ossified bone in this arch. The first four epibranchials are very
variable in shape, with the second and fourth bearing an uncinate process. All
four are fully ossified and bear hemibranchia. Furthermore, the two first
epibranchials are synchondrally connected to each other distally. They contact
the ossified third infrapharyngobranchial bone through a fused cartilaginous first
and second infrapharyngobranchial. The latter is synchondrally connected to the
third
epibranchial
and
to
the
fourth
infrapharyngobranchial.
This
fourth
infrapharyngobranchial is connected to the fourth epibranchial bone and supports
the upper pharyngeal tooth plate.
Opercular series (figs. 3.2A, 3.3)
The opercular series consist of the opercular, interopercular, preopercular and
suprapreopercular bones. This condition differs within different groups of
catfishes and even loricarioids. In Loricariidae, for example, the interopercular
bones have been lost entirely. The opercle itself is large, more or less triangular,
and is connected to the interopercle on its ventro-anterior margin. It also bears a
process for the articulation with the hyomandibula on its dorso-anterior margin.
In Astroblepidae and Loricariidae this articulation shifts toward the dorsal side of
the opercular bone (Schaefer, 1987, 1988). The interopercle is a small,
triangular bone, which is ligamentously connected to the lower jaw at the level of
the angulo-retroarticular bone. Dorso-anteriorly from the interopercle and
anterior to the ventral part of the opercle, lies the preopercle. This bone, present
in all loricarioid families, bears part of the preopercular canal with two of its
pores, one centrally and one anteriorly, which are homologous to pores 4 and 5
in Diplomystes (Schaefer, 1988). The part of the preopercular canal running
through the preopercle and suprapreopercle in C. aeneus, as in all Callichthyidae,
does no longer connect to the part of the preopercular canal that is present in
the posttemporo-pterotico-supracleithrum. In N. inermis and in several nonloricarioid catfishes, this canal continues into the mandible and is consequently
referred to as the preoperculo-mandibular branch. In trichomycterids, on the
other hand, the preopercular canal is extremely reduced and does not even enter
the preopercle, but remains limited to an opening in the pterotic bone (Baskin,
1972; Schaefer, 1988). Finally, the presence of a suprapreopercular bone in C.
28
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
aeneus, is a condition that has never been mentioned within the Callichthyidae
but which was present in all specimens examined. Therefore, further ontogenetic
research will focus on the development of this bone, attempting to reveal
whether this bone is truly homologous to the suprapreopercular bone found in
other fish groups.
Weberian apparatus (fig. 3.1B)
In C. aeneus the Weberian apparatus is part of a complex structure, comprising
a fusion between several vertebrae. Normally, the complex vertebral centre of
the Weberian apparatus is a fusion of the second to the fifth vertebral centre in
all Siluriformes, except Diplomystes, where the fifth centre is excluded from the
complex (Arratia, 1987; Fink & Fink, 1996). Additionaly, in Loricarioidea the first
vertebral centre is also fused to the complex (Schaefer, 1990; Reis, 1998).
Coburn & Grubach (1998), however, discovered, after ontogenetic research, that
in Corydoras paleatus only three vertebrae are fused within the complex and that
the first two vertebrae are missing. The gas bladder is divided into two chambers
which are encapsulated in the expansions of the transversal processes of this
complex centre. Laterally to the compound centre, two foramina are situated
through which passes the duct that connects these two chambers. Gas bladder
contact
with
the
external
medium
occurs
through
an
aperture
in
the
posttemporo-pterotico-supracleithrum, covered by a hollow expansion bearing
the latero-sensory canal. This condition is possibly homologous to the condition
found in Astroblepidae and Loricariidae, where the aperture is completely
covered by the posttemporo-pterotico-supracleithral bone (Reis, 1998). The
connection between the gass bladder and inner ear is made up of one compound
bone referred to as the compound tripus (Schaefer, 1990). The compound tripus
found in C. aeneus was suggested to be a fusion between the tripus,
intercalarium, scaphium and interossicular ligament, typically found in all
Siluriformes, but with the loss of the claustrum in Callichthyidae (Schaefer,
1990; Reis, 1998). However, since Coburn & Grubach (1998) mention the loss of
the first two vertebrae, their derivatives (claustrum, scaphium and intercalarium)
are also missing and their results show the tripus to be a myoseptal tripus
(formed in the paravertebral sac and the dorsal myoseptum of vertebra III).
29
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
Vertebral column (figs. 3.1B, 3.6)
In the specimens examined, the total number of vertebrae, including the first
five incorporated in the Weberian apparatus, was 28. This number equals that
found in several other Corydoras and callichthyid species, and is one more than
the number found in Scoloplacidae and some Loricariidae (e.g. Otocinclus,
Hypoptopoma) (Schaefer, 1990). Britto (2000) mentions the presence of 28-31
vertebrae in several Aspidoras-species, which corresponds to the 27-32
vertebrae described by Regan (1911) for the family of the Callichthyidae. The
first articulating vertebra, the sixth vertebra, has two large parapophyses that
articulate
with
the
complex
centre
of
the
Weberian
apparatus.
These
parapophyses further support a pair of large, hollow ribs, which contact the first
ventrolateral bony scutes behind the pectoral girdle. Vertebrae 7-12 each carry a
pair of small, thin ribs. The presence of a pair of such large, hollow ribs on the
parapophysis of the sixth vertebral centre, followed by several pairs of small ribs
is typical for all Callichthyidae (Regan, 1911; Hoedeman, 1960d; Reis, 1998;
Britto, 2000). In contrast to Hoedeman (1960d) mentioning only four to five of
these small ribs in Corydoras-species, six were found here. The number of caudal
vertebrae is 14, of which, in the first three to four, the haemal spines are
expanded and plate-like. These haemal spines are ventroposteriorly oriented,
thus forming a protective, posterior wall for the abdominal cavity. Furthermore,
the last preural vertebra is incorporated within the ural complex together with
the last vertebra, the first ural vertebra (Lundberg & Baskin, 1969).
Dorsal fin (fig. 3.6A)
In C. aeneus the dorsal fin bears a first small fin ray, modified to serve a spinelocking mechanism, followed by a second, large one (Alexander, 1965). After
this, seven branched dorsal fin rays are present, of which the last is split up to its
base. The pterygiophores of these spines plus the first five fin rays are connected
to the 8th to 11th vertebral neural spine. The first pterygiophore bears a large
transverse process, which connects to the lateral body scutes. In callichthyids,
this process is further ligamentously connected to the sixth rib, whereas in
scoloplacids, astroblepids and loricariids, this ligament ossifies into a lateral bone
(Schaefer, 1990). Preceding the first dorsal fin spine, a nuchal plate is present,
which is connected to the
seventh
vertebra
and contacts the parieto-
supraoccipital at its anterior side. The condition where two fin spines with seven
30
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
fin rays are present, as is the case here, fits the plesiomorphic nine fin rays
found in Diplomystes (Alexander, 1965). Although most authors do not count the
first modified fin spine, the number of remaining true fin rays fits my findings. It
must, however, be stated that the low number of specimens studied here does
not allow a detailed description of meristic characters and that such a description
is also beyond the scope of this study. Nonetheless, seven fin rays, with the last
ray branched up to the base, does corresponds with the original description of C.
aeneus by Gill (1858) and of other Corydoras-species (Isbrücker & Nijssen,
1973, 1992a; Nijssen, 1970) and could indicate a low variability. The number of
dorsal-fin rays is a character of great taxonomic value within the Corydoradinae
and a number of 7-9 is determinative for Corydoras-species, whereas a number
of 10 or more is determinative for species formerly belonging to the genus
Brochis (and different from the plesiomorphic siluriform condition). The
distinction between Aspidoras, on the one hand, and Corydoras, on the other
hand, can also be done based on dorsal fin morphology. Here, a lack of contact
between the nuchal plate and the posterior process of the parieto-supraoccipital
is held as being typical for Aspidoras-species (Reis, 1996). Within the
Loricarioidea, all families have the plesiomorphic siluriform number of branched
dorsal fin rays, except for the Scoloplacidae, where a reduction of the number
has occurred and only four are present (Reis, 1998).
Anal fin
The anal fin consists of a single unbranched and seven branched fin rays. The
bases of the first four rays articulate with the haemal spines of vertebrae 20 to
22. The number of fin rays (n=7) found in C. aeneus corresponds with the
number given in the original description by Gill (1858), although the presence of
a possible eighth branched ray is mentioned. The presence of a single,
unbranched, anal fin ray is a derived condition within the Callichthyidae, which
only occurs in some Corydoras-species and all Lepthoplosternum-species. Among
the Loricarioidea, this condition is also found in Scoloplacidae, Astroblepidae and
Loricariidae, in contrast to Trichomycteridae and N. inermis, where two
unbranched rays are present (Reis, 1998).
31
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
Adipose fin
The adipose fin consists of a single spine, derived from a small dorsal bony
plate. The homology of this spine initially was unclear. The spine could be
considered as a transformed bony scute or a true fin ray that is covered by these
dermal plates. Hoedeman (1960d), however, mentions the presence of ‘two
strong muscle bundles’ used for erection of the spine in Callichthyinae, which
would mean that the spine in Corydoras, although not movable in the
Corydoradinae, is homologous to a fin ray.
Caudal skeleton (fig. 3.6B)
As mentioned before, in many Siluriformes and Cypriniformes, the first preural
vertebra is fused to the complex centre of the caudal skeleton (Lundberg &
Baskin, 1969). The caudal skeleton in C. aeneus is of the pleurostyl type and
consists of two bony plates. The dorsal plate is formed by a fusion of the urostyl
and the dorsal hypurals III, IV and V, a fusion which, as well as the development
of a plate-like lamina on the epural (the neural spine of the first preural centre),
could be revealed by preliminary ontogenetic data (see 4.6). The ventral plate
comprises the parhypural and hypurals I and II. Hypurals II and III are variably
fused on their left and/or right side or on neither side. The fact that the dorsal
hypurals are also fused to the compound centre is common to all loricarioids,
except for N. inermis (Schaefer, 1990). The number of principal rays is 7/7,
which is common among Corydoras species (Isbrücker & Nijssen, 1973, 1992a).
Surprisingly, the number found here differs from that given in the original C.
aeneus description by Gill (1858) (n=6/6). Further, both the neural and haemal
spine of the preural vertebral centre II are heavily ossified and branched. This
state is, to a lesser extent, also present in the preural vertebral centre III.
Pectoral girdle (fig. 3.7)
The pectoral girdle consists of the cleithrum, which articulates with the
supracleithrum, part of the posttemporo-pterotico-supracleithrum and embedded
in the skull. The cleithral bones are medially connected by a simple suture. The
ventral part of the pectoral girdle consists of the scapulocoracoid. As in all
Siluriformes the scapulacoracoid bone is a compound bone, comprising the
scapula, the coracoid and the mesocoracoid. In Callichthyidae, the posterior
process of this scapulacoracoid and of the cleithrum are sutured behind the
32
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
articulation of the fin with the girdle, this way forming a bony shield around the
entire base of that fin (Reis, 1998). The scapulocoracoid bones also connect
ventrally, but in contrast to the cleithral bones, here a heavily interdigitating
suture is present. Cleithrum and scapulacoracoid connect medially by means of a
coracoid bridge (see also Diogo et al., 2001). The pectoral spine is pungent as in
all Corydoradinae (this in contrast to the Callichthyinae), bears serrations on
both its anterior and posterior face and a large articulation head, which also suits
a spine locking mechanism (Hoedeman, 1960d; Alexander, 1965). Ten branched
rays together with two proximal radials are present.
Pelvic girdle (fig. 3.8)
The pelvic girdle consists of two basipterygia, which bear both an internal and
an external anterior process. The homology of both these processes was
questioned by Shelden (1937), and both were referred to as ‘projections’. Since
no obvious motivation was given to support this idea, I do not follow his views on
this matter and consider them to be the internal and external process (see 4.6).
The internal process is well developed and bears a small dorsal lamina. The
presence of a dorsal lamina on the internal process is a typical callichthyid
feature (Reis, 1998). Second, the external process also bears a lamina, which, in
C. aeneus is connected to a scute of the lower, lateral series of bony scutes, by
means of connective tissue (Reis, 1998). A third laminar process is present on
the ischiac process, where it connects to the ventral tip of a scute of the lower,
lateral series (Reis, 1998). This ischiac process is further divided into a dorsal
and a ventral process, of which the dorsal part is bent laterally. These
callichthyid features in the pelvic girdle are possibly related to reproductive
strategies (Reis, 1998). Furthermore, the pelvic fin bears six branched rays.
33
3 ADULT MORPHOLOGY – ADULT OSTEOLOGY
34
3 ADULT MORPHOLOGY – ADULT MYOLOGY
3.2. ADULT MYOLOGY
Modified from:
Huysentruyt F, Geerinckx T & Adriaens D
A descriptive myology of Corydoras aeneus (Gill, 1858)
(Siluriformes: Callichthyidae), with a brief discussion
on adductor mandibulae homologies
Animal Biology 2007, 57(4): 433-452.
Abstract
Cranial and postcranial myology of adult Corydoras aeneus is described and results
discussed in comparison to other ostariophysan, siluriform and/or loricarioid fishes.
Further, a brief discussion on m. adductor mandibulae homologies is given providing
arguments for the use of the terminology proposed by Diogo & Chardon (2000) in future
studies dealing with the myology of Siluriformes. Doing this, I here identified an A1OST,
A2, and A3’ section in C. aeneus and recognized the homology of the m. retractor
tentaculi with the A3”. Next to this, the opercular system is discussed, focusing on
similarities in this system in both a callichthyid (C. aeneus) and loricariid (Ancistrus cf.
triradiatus) representative. In both these families, the m. dilatator operculi is enlarged
and the direction of the operculo-hyomandibular articulation has shifted. In addition, in
both lineages, the m. hyohyoidei abductor has also shifted its orientation, acting as an
adductor. These similarities seem to corroborate the ‘decoupling hypothesis’ by Schaefer
& Lauder (1986) in which a decoupling of elements in primitive members of a lineage
leads to a higher morphological diversity within that lineage combined with the possible
acquisition of new functions.
Brief introduction
In chapter 3.1, a descriptive study on the osteology of the species is presented,
but a complete overview of the cranial and postcranial myology is still absent.
This information, however, is quite relevant for ongoing phylogenetic research on
Loricarioidea, to which these callichthyids belong (Reis, 1998; Britto & Castro,
2002; Britto, 2003). In his review, Britto (2003) already stated that the poor
knowledge about the phylogenetic relationships among the genera of the
35
3 ADULT MORPHOLOGY – ADULT MYOLOGY
Corydoradinae is largely due to the scarcity of data on internal anatomy. The
review aims to counter this and provides 71 osteological characters, resolving a
large part of this intergeneric phylogeny. However, no myological characters are
included in his study. Reis (1998) also mentioned the importance and scarcity of
anatomical data on the family level, but he also provided no myological data in
his analysis. The lack of myological characters in these various phylogenetic
analyses of this group is even more puzzling, considering the fact that Schaefer
& Lauder in 1986 already found that, as a consequence of structural innovations
in the jaw adductor complex in Loricarioidea, a lot of confusion exists about the
homology of these different parts, and thus about their evolutionary history. In
addition, Schaefer (1990) has also proposed the presence of a divided musculus
adductor mandibulae as a synapomorphy for the Callichthyidae. Finally, even
recent molecular research on the family of the Callichthyidae pointed out that
“further analyses of additional morphological data may be very helpful in the
understanding of the phylogeny of the Callichthyidae” (Shimabukuro-Dias et al.,
2004).
Consequently, the aim of this study is to provide a full description of the
species’ adult myology, as a basis for further ontogenetic research on this
species, as well as to contribute to future phylogenetic studies.
Brief material and methods
The specimens examined are presented in table 2.1. For clearing and staining
procedures, procedures on serial sectioning and 3D-reconstructions see 2.2. For
details on skeletal elements, I refer to chapter 3.1. Muscle terminology follows
Winterbottom (1974), except where indicated otherwise.
Results
Jaw musculature (figs. 3.9, 3.11, 3.12)
The m. adductor mandibulae complex in Corydoras aeneus consists of two
different muscles, the A2 and A1OST-A3’. The A2 originates laterally on the
36
3 ADULT MORPHOLOGY – ADULT MYOLOGY
hyomandibula, on a ridge near the suture with the metapterygoid and has its
insertion on the ligamentum primordium. The A1OST-A3’ also originates on the
hyomandibula, just below the A2, but the site of origin of this muscle expands
further ventrally, covering the entire caudal margin of the quadrate bone. The
muscle then runs forward, covering the entire suspensorial arch, splitting up at
the level of the caudal margin of the lower jaw, with a lateral bundle (A1OST)
inserting on the dorsolateral side of the lower jaw and a medial bundle (A3’) on
the dorsomedial side. Both bundles attach to the lower jaw between the
suspensorial joint caudally and the coronoid process rostrally.
The m. retractor tentaculi in C. aeneus originates on both the hyomandibula
and metapterygoid. The muscle then runs medially from both the A2 and A1OSTA3’ and, like the A2, also attaches to the lig. primordium.
The m. intermandibularis anterior connects the medial sides of both dentary
bones at their rostral margin.
The m. intermandibularis posterior is subdivided into a pars dorsalis and a pars
ventralis which merge rostrally and insert on the lower jaw. Both parts follow the
same path and are difficult to be discerned through dissection. Serial sections,
however, clearly show a subdivision into a dorsal and ventral part. These serial
sections also revealed that the muscle was innervated by the inferior mandibular
nerve branch of the trigeminal nerve (V) and no myocomma was present. This
indicates that the muscle merely consists out of the m. intermandibularis
posterior and that no m. protractor hyoidei part is present.
The m. extensor tentaculi originates both from the lateral side of the
orbitosphenoid and the caudal side of the lateral ethmoid bone. Rostrally, the
muscle inserts as a single bundle on the caudal margin of the autopalatine,
posterior to its articulation with the lateral ethmoid.
Suspensorial musculature (figs. 3.9, 3.10)
The m. levator arcus palatini originates from the skull at the level of the
sphenotic bone and inserts on the laterodorsal surface of the hyomandibula. The
m. adductor arcus palatini originates from the base of the skull along the lateral
side of the parasphenoid bone and inserts on the medial side of both
hyomandibula and metapterygoid. Posterior to the m. adductor arcus palatini, a
m. adductor hyomandibulae is present, originating on the ventral side of the
prootic and inserting on the medial surface of the hyomandibula.
37
3 ADULT MORPHOLOGY – ADULT MYOLOGY
Opercular musculature (fig. 3.10)
The m. dilatator operculi is a conical muscle originating from a cavity in the
sphenotic and from the hyomandibular bone and running ventroposteriorly to
insert, through a tendon, on the anterodorsal side of the opercle, above the
articulation with the hyomandibula. The anterior fibers of this muscle are
partially continuous with the m. levator arcus palatini.
The m. adductor operculi originates entirely on the posttemporo-pteroticosupracleithrum. The muscle runs ventrally to insert on a ridge on the
dorsomedial side of the opercle.
The m. levator operculi is a broad muscle originating on the posttemporopterotico-supracleithrum and the prootic. This muscle runs ventrally to insert
medially on the operculare on a large ridge that starts just below the rostral side
of the insertion ridge of the dilatator operculi and that runs caudoventrally to the
middle of the opercle.
Hyoid musculature (fig. 3.12)
As mentioned, no m. protractor hyoidei is present. The m. hyohyoideus inferior
connects both hyoid arches and the left and right part of this muscle are medially
fused through an aponeurosis. The m. hyohyoideus abductor in C. aeneus
originates on the most medial branchiostegal rays, runs rostrally and fuses at the
midline just behind the caudal margin of the hyohyoideus inferior. The
aponeurosis through which these two halves are fused connects to the midline
aponeurosis of the m. sternohyoideus (q.v.). However, a second bundle runs
from the branchiostegal rays caudally to insert on the scapulocoracoid. Further,
the mm. hyohyoidei adductores interconnect all branchiostegal rays, the opercle
and the interopercle. Finally, the m. sternohyoideus is a large muscle, originating
from the cleithrum and running rostrally, inserting on the parurohyal bone and
further connecting to the dorsal and ventral hypohyal through a double
ligamentous connection. This muscle is split up over its entire length into a right
and left half, which medially connect through an aponeurosis.
Gill arch musculature (figs. 3.13)
Dorsally, in C. aeneus, gill arches I-IV bear a m. levator externus. These
muscles originate grouped on the ventrolateral side of the prootics and insert on
the respective epibranchial bones. On the third arch, however, the insertion site
38
3 ADULT MORPHOLOGY – ADULT MYOLOGY
of the m. levator externus has also extended onto the cartilage connecting the
epibranchial and infrapharyngobranchial III to the infrapharyngobranchial IV,
with some fibers even inserting on the latter bone. Only arches I, II and IV bear
a m. levator internus. This muscles is a single muscle in the first two arches,
originating on the prootic, on the same site as the mm. levatores externus, and
inserting on the cartilaginous infrapharyngobranchial I+II and epibranchial II.
Since the infrapharyngobranchials of the first two arches are fused, I chose to
consider this muscle as the combined m. levator internus of the first two arches.
On
the
fourth
arch
the
m.
levator
internus
also
inserts
on
the
infrapharyngobranchial IV, with the upper pharyngeal toothplate attached, but
the site of origin is on the ventromedial part of the posttemporo-pteroticosupracleithrum. A m. obliquus dorsalis is only present on arches III and IV.
These muscles interconnect the epibranchials and infrapharyngobranchials of
both arches. Only the third arch bears a m. transversus dorsalis which
interconnects the left and right infrapharyngobranchial. Additionally, the m.
obliquus posterior connects the ceratobranchial V to the epibranchial IV. It is a
short muscle, with its sites of origin and insertion close to the lateral tips of both
bones. No m. retractor dorsalis or m. levator posterior is present. In gill arches I,
II, III and IV a m. adductor arcuum branchialium is present connecting the
ceratobranchials to the epibranchials.
Ventrally, all arches bear a m. rectus ventralis, with the exception of arch III.
This muscle connects the ceratobranchial of the respective arch to the
hypobranchial of the proceeding one, except for the first arch, where this muscle
inserts on the dorsal hypohyal. Arch III bears the m. rectus communis, which
has the same origin as the m. rectus ventralis in the other arches, but with its
insertion on the dorsal hypohyal. A m. obliquus ventralis is only present in arches
II and III, connecting the arches’ hypobranchials to their ceratobranchials. In
arch II, these muscles also insert on the basibranchial, connecting the right and
left m. obliquus ventralis at this point. A m. transversus ventralis was found in
arches IV and V, interconnecting the left and right cartilaginous hypobranchial in
the fourth arch and the left and right ceratobranchial in the fifth. Further, both
the m. pharyngoclavicularis internus and externus connect the ventral part of the
gill basket to the pectoral girdle. Posteriorly, both ceratobranchial V and
epibranchial IV are confluent with the m. sphincter oesophagi.
39
3 ADULT MORPHOLOGY – ADULT MYOLOGY
Eye musculature (fig. 3.14A)
The extrinsic eye musculature in C. aeneus is similar to the pattern commonly
found in teleosts. Two oblique muscles, the m. obliquus superior and m. obliquus
inferior originate on the orbitosphenoid bone and insert on the most dorsal and
most ventral part of the eyeball respectively. At the same sites respectively, both
the m. rectus superior and m. rectus inferior insert, originating on the
parasphenoid. The last two extrinsic eye muscles, the m. rectus externus and the
m. rectus internus both originate in a large posterior myodome between the
prootics and the parasphenoid at the level of the prootic bridge, inserting
respectively on the most caudal and most rostral side of the eyeball.
Pectoral musculature (fig. 3.14B, 3.15)
The pectoral girdle is connected to the splanchnocranium through the m.
sternohyoideus and to the posttemporo-pterotico-supracleithrum through the m.
protractor pectoralis (fig. 3.9, 3.10). Ventrally, the main muscle mass for
movement of the pectoral fin is made up of the m. abductor pectoralis
superficialis pars ventralis. This muscle originates on the ventrolateral face of the
cleithrum and scapulocoracoid and runs backwards where it attaches to the
bases of all fin rays except the pectoral spine. Dorsal to this muscle lies the m.
abductor pectoralis superficialis pars dorsalis, which also inserts on the fin rays,
but originates from the scapulocoracoid. Lateral to these muscles lies the m.
arrector pectoralis ventralis, which originates from the dorsolateral side of the
cleithrum and inserts ventrally on the pectoral fin spine. Dorsal to this the m.
arrector pectoralis dorsalis originates on the dorsomedial side of the cleithrum,
runs through a foramen in the cleithrum and inserts on the anterodorsal margin
of the pectoral fin spine. Also originating on the dorsomedial face of the pectoral
girdle, but on the scapulacoracoid are the m. adductor pectoralis superficialis and
the m. adductor pectoralis profundus, which also pass through a foramen to
insert dorsally on the bases of all fin rays except the pectoral spine. Here, the
lateral muscle fibers finally insert on the more medial fin rays and the medial
fibers insert on the lateral rays, this way resulting in a ‘cross-over’ of muscle
fibers. The pectoral spine is abducted by contraction of the m. abductor
pectoralis
profundus,
which
also
originates
on
the
medial
side
of
the
scapulacoracoid, but ventrally, and inserts on the head of this pectoral fin spine.
40
3 ADULT MORPHOLOGY – ADULT MYOLOGY
Additionally, the pectoral girdle is connected to the pelvic girdle through the m.
infracarinalis anterior.
Pelvic musculature (fig. 3.16)
Ventrally, the pelvic girdle is covered by three separate muscles. The most
ventral muscle is the m. abductor pelvicus superficialis, which originates medially
on the pelvic girdle and inserts on all but the most lateral fin ray. Underneath
this muscle, but with a more anterior origin lies the m. abductor pelvicus
profundus, which also inserts on all fin rays except the first. Also ventrally lie the
m. arrector pelvicus ventralis and the m. arrector pelvicus dorsalis, which both
originate on the ventrorostral margin of the anterior internal process and insert
on the first pelvic fin ray. On the anteromedial margin of the anterior internal
process of the pelvic girdle the m. adductor pelvicus superficialis originates and
posterior to this up to the posterior margin of the girdle, the m. adductor
pelvicus profundus has its origin. Both these muscles insert on all pelvic fin rays.
The pelvic girdle further attaches to the anal fin through the m. infracarinalis
medius and to the pectoral girdle through the m. infracarinalis anterior.
Unpaired fin and body musculature (fig. 3.17, 3.18)
The caudal fin musculature mainly consists of a m. flexor dorsalis and a m.
flexor ventralis, which both originate on the third preural vertebral centre and
insert on the bases of the principal fin rays. In addition, dorsal to this muscle lies
the m. flexor dorsalis superior which originates on the neural spine of the third
preural vertebra and inserts on the dorsal proximal fin rays and the upper
principal fin ray. Also, ventrally a m. flexor ventralis inferior is present, which
originates on the haemal spine of the third preural vertebra and inserts on the
ventral proximal fin rays. Additionally, on the hypurals I, II and the parhypural,
the m. hypochordalis longitudinalis originates, which ligamentously inserts on the
upper principal fin ray. Further, the principal fin rays are interconnected through
the mm. interradiales, which insert on the lepidotrichia shafts, distal to the
insertion of the flexor muscles. Dorsally, the posterior margin of the dorsal fin is
connected to the neural spine of the third preural vertebra by the m.
supracarinalis posterior, which inserts on this spine through a tendon. Ventrally,
the haemal spine of the third preural vertebra is connected to the anal fin by the
m. infracarinalis posterior through a musculous insertion.
41
3 ADULT MORPHOLOGY – ADULT MYOLOGY
Three types of muscles control the movement of the dorsal fin: the mm.
erectores dorsales, the mm. inclinatores dorsales and the mm. depressores
dorsales. The first two types of muscles, of which the mm. inclinatores dorsales
lie more lateral and cover the mm. erectores dorsales, originate from the
anteroventral part of the pterygiophore supporting the proceeding fin ray, while
the m. depressor dorsalis of each fin ray has its origin on the ventral side of the
pterygiophore of the corresponding fin ray. On the first dorsal fin spine only a m.
erector dorsalis inserts which originates on the anterodorsal side of the seventh
vertebra. A m. erector dorsalis also inserts on the second dorsal fin spine, which
also originates on the seventh vertebra, albeit on the posterodorsal side.
Additionally, this second spine bears a m. depressor dorsalis which originates on
the anterodorsal side of the eighth vertebra and runs posterior to the transverse
process to insert on the spine’s posterior margin. Posteriorly, the dorsal fin is
connected to the third preural vertebra through the m. supracarinalis posterior.
Anteriorly, the dorsal fin connects to supraoccipital bone through the m.
supracarinalis anterior.
Lateral to all other muscles of the anal fin lie the mm. inclinatores anales,
which originate on the pterygiophores and insert laterally on the fin rays bases.
Underneath these muscles lie the mm. erectores anales and the mm.
depressores anales. Both these muscles originate on the pterygiophores and
insert on the fin rays, the mm. erectores anales anteriorly and the mm.
depressores anales posteriorly. Posteriorly, the m. infracarinalis posterior
connects the anal fin to the third preural centre and anteriorly, the m.
infracarinalis medius connects the anal fin to the pelvic girdle.
Discussion
Next to structural changes in the morphology of the m. hyohyoideus inferior,
the presence of a cartilage plug between the lower jaw and the hyoid in
Astroblepidae and Loricariidae, and the shift and neoformation of several
ligamentous connections in the jaw region, the Loricarioidea differ structurally
from the plesiomorphic siluriform situation found in Diplomystidae as described
by Diogo & Chardon (2000) at the level of the jaw musculature. Because of this,
42
3 ADULT MORPHOLOGY – ADULT MYOLOGY
the superfamily Loricarioidea has been considered “a clade showing a pattern of
progressive increase in the mechanical complexity of structures associated with
feeding”, with consequent homology ambiguities (Schaefer & Lauder, 1986,
1996).
A first example of this is the nature of the dorsal division of the m. adductor
mandibulae inserting onto the maxillary bone, the so-called A1 (Winterbottom,
1974; Gosline, 1986, 1993). Since such a muscle in ostariophysans is believed to
have evolved independently from the A1 in eurypterygians (Fink & Fink, 1996), a
true A1 part, as present in Acanthomorpha, is absent in Ostariophysi (Adriaens &
Verraes, 1996; Diogo & Chardon, 2000; Wu & Shen, 2004). Alexander (1965),
however, did recognize an A1 part in Callichthyidae and Loricariidae, an error
later adapted by Howes (1983), Schaefer & Lauder (1986) and Schaefer (1997).
Diogo & Chardon (2000) chose to name the different muscle bundles according
to their position, giving the more lateral bundles inferior numbers. This way, they
named this dorsolateral section in Ostariophysi A2 (medial to A0 and A1).
Following the same logic, they suggested the term A1OST for the ventrolateral
cheek muscle inserting on the dorsal face of the lower jaw (the term OST was
added to avoid confusion with the acanthomorph A1). Wu & Shen (2004), on the
other hand, avoided the use of the term A1 in their terminology, using only
derivatives of the A2 and A3 terms. Still, I believe the terminology proposed by
Diogo & Chardon (2000) to be more appropriate, given the thoroughness of their
comparative work on siluriforms, a group dealt with in a rather limited way by
Wu & Shen (2004). However, it must be noted that in the terminology of Diogo &
Chardon (2000) the term A1 in A1OST is somewhat misleading due to the lack of
any relation of this muscle with the maxillary, as already stated by Wu & Shen
(2004); as well as that their A2 is not homologous to the A2 as described by
Vetter (1878), and maybe the term A2OST would have been more appropriate.
Given this, I identified the dorsolateral division of the m. adductor mandibulae
in Corydoras aeneus as the A2 according to the definition by Diogo & Chardon
(2000). In C. aeneus, in contrast to other non-loricarioid catfishes, this muscle
has shifted its insertion onto the lig. primordium, through which it indirectly
inserts on both upper and the lower jaw, altering the entire jaw mobility (Howes,
1983; Schaefer & Lauder, 1986). Next to this, in C. aeneus, the m. retractor
tentaculi also inserts on the lig. primordium, a muscle I believe to be
homologous to, or derived from the A3”, as already suggested by various authors
43
3 ADULT MORPHOLOGY – ADULT MYOLOGY
(Lubosch, 1938; Alexander, 1965; Howes, 1983; Adriaens & Verraes 1996,
1997a; Diogo & Chardon, 2000; Diogo, 2005).
In the ventral region of this cheek muscle complex, the m. adductor
mandibulae is composed of two separate bundles: one attaching to the lower jaw
laterally and a second one attaching to the lower jaw on the medial side. Given
the ventrolateral position of the former bundle I identified it as the A1OST. The
latter bundle was identified as the A3’, based on its medial position. Gosline
(1989, 1993) and Adriaens & Verraes (1996) also confirm the presence of a
compound lower jaw muscle in catfishes, and conclude that this muscle is the
composed A2A3’, as suggested earlier by Takahashi (1925). Given the synonymy
between the A2 as described by Takahashi (1925) to the A1OST as described by
Diogo & Chardon (2000), this conclusion also fits my findings and I put this
compound muscle in homology with the A1OST-A3’. This insertion of the A1OSTA3’ directly onto the medial surface of the dentary is considered a derived state
in Siluriformes (Diogo, 2005). Diogo (2005), also claims that in Corydoras, the
insertion of the A3’ is lateral to that of the A2 and A3”, and that the A3” inserts
partially on the mandible and partially on the lig. primordium. When regarding
the A3” as the m. retractor tentaculi, the latter statement is indeed correct, but
my observations contradict the lateral position of the A3’ as it is situated on the
medial side of the A2. Wu & Shen (2004) do not mention a further subdivision of
the A3 and place the A3 in synonymy with the A3’ as described by Takahashi
(1925). They, however, do not mention the m. retractor tentaculi in their survey,
and since this is believed to be homologous to the A3”, retaining the name A3’ for
the medial bundle in C. aeneus seems justifiable. Finally, in C. aeneus both an A0
and Aω are absent, a condition common in Siluriformes (Alexander, 1965; Diogo
& Chardon, 2000).
Next to this, the subdivision of the m. extensor tentaculi in C. aeneus is a
character already mentioned by Diogo (2005) as a derived siluriform state but
which was not reported by Schaefer & Lauder (1986, 1996). This subdivision of
the m. extensor tentaculi fits the general trend of an increase in morphological
complexity of the functional design through decoupling present throughout the
loricarioid evolution (Schaefer & Lauder, 1986, 1996). The definition of
decoupling in this case can be stated as the repetition of individual elements as
redundant design components, followed by the specialization of one or more of
these
44
elements
as
a
mechanistic
basis
for
the
evolution
of
novel
3 ADULT MORPHOLOGY – ADULT MYOLOGY
structure/function (Schaefer & Lauder, 1996) (see also 1.1). In these loricarioids,
such a decoupling has mainly occurred at the level of the jaws. The upper jaws
have been decoupled from the cranium, while the lower jaw has been decoupled
from the opercular series, increasing mobility and independency of both upper
and lower jaw (Schaefer & Lauder, 1996). Next to this, the lower jaws and hyoid
musculature
have
become
decoupled
from
their
plesiomorphic
bilaterally
constrained midline attachments and a new redundant linkage was acquired
(Schaefer & Lauder, 1996). All these factors, combined with the increase in
myological complexity at the level of the m. adductor mandibulae has ultimately
facilitated the evolution of a suckermouth in astroblepids and loricariids and the
evolution of an algae scraping feeding apparatus in the latter family.
Another example in which an evolutionary pattern of increased morphological
complexity through decoupling is present throughout the loricarioid lineage
involves the opercular system. This way, in the family of the Loricariidae, the
opercle has been decoupled from the lower jaw and has lost his function in
expiration (Geerinckx & Adriaens, 2006). Within this family, in the Ancistrinitribe, this has ultimately led to the formation of an erectile opercle, armoured
with large denticles, acting as a defensive mechanism (Geerinckx & Adriaens,
2006). One of the most important innovations in this process has been the
development of a hypertrophied m. dilatator operculi, which has ultimately led to
the formation of several myodomes in the skull bones in Ancistrus cf. triradiatus,
thus forming a ‘secondary skull roof’ (Geerinckx & Adriaens, 2006).
In C. aeneus, the m. dilatator operculi is not hypertropied nor is it split into
several different bundles. This corresponds to the plesiomorphic condition for
Siluriformes as described by Diogo (2005), who only mentions a subdivided m.
dilatator operculi in members of the Aspredinidae and Trichomycteridae. He,
however, does not mention this for loricariids, though only two species of
Hypoptopoma were included in the study. However, in C. aeneus, the m.
dilatator operculi does extend rostrally, originating in a cavity in the sphenotic
bone, a situation which is not mentioned in the study by Diogo (2005). In his
study, the origin of the m. dilatator operculi on the dorsal surface on the
neurocranium is mentioned as a derived character state in Trichomycteridae and
members of the genus Plotosus but the presence of a sphenotic cavity was not
detected in any of the studied siluriforms. As in A. cf triradiatus, the cavity in C.
aeneus splits up the sphenotic bone in a deeper and outer layer, although both
45
3 ADULT MORPHOLOGY – ADULT MYOLOGY
situations differ structurally. In C. aeneus the deeper layer of this cavity is
formed by perichondral part of this bone (the autosphenotic), where the outer
layer is formed by the dermosphenotic part (as shown from preliminary
ontogenetic data). In the skull bones where a myodome is formed in A. cf
triradiatus, however, the myodome is positioned in between layers of mixed
origin, since canals of the lateral line system are present in the deeper layers
(Geerinckx & Adriaens, 2006). Therefore, both cavities in A. cf. triradiatus and C.
aeneus can not be considered homologous and the term myodome was not
applied to the situation in C. aeneus. In addition, in its most narrow sense the
term ‘myodome’ is restricted to those cavities housing the external eye muscles
in teleosts (Rojo, 1991). In a broader sense, as applied by Geerinckx & Adriaens
(2006) the term applies to a cavity formed in a bone housing a muscle, which is
clearly the case in the A. cf. triradiatus skull, but not in the C. aeneus sphenotic,
where the dermosphenotic merely forms an outgrowth covering the m. dilatator
operculi.
Nonetheless, the rostral expansion of the m. dilatator operculi in C. aeneus,
combined with the more oblique direction of the opercle-hyomandibular
articulation, implies an increase in efficiency in the dilatation of the opercle. Next
to this, the anterior bundle of the m. hyohyoidei abductor has shifted its
orientation to the transverse plane, acting as an adductor of the branchiostegal
membrane. A similar state is found in A. cf triradiatus, in which the m.
hyohyoidei abductor has also shifted its direction medially, acting as an adductor
of the branchiostegal membrane, whereas fibers of the m. hyohyoideus inferior
assist in the abduction (Geerinckx & Adriaens, in press). This way, the closing of
the branchiostegal membrane is at least partially decoupled from the opercle in
both species, a factor which could have facilitated the alteration of the opercle
into a defensive mechanism in A. cf. triradiatus.
This
hypothesis
would
again
fit
the
general
hypothesis
of
increased
morphological complexity through decoupling, which shows the importance of
decoupling combined with shifts in function as a common evolutionary pathway
for accomplishing innovative structural design. It is my belief that further
investigation of not only the adult morphologies, but also extensive ontogenetic
and functional comparative research of several loricarioid lineages will elucidate
the full impact of such an evolutionary pathway and quantify their role in the
evolution of these extremely successful and diverse taxa.
46
3 ADULT MORPHOLOGY – ADULT MYOLOGY
47
3 ADULT MORPHOLOGY – ADULT MYOLOGY
48
Chapter
4
Ontogeny
4 ONTOGENY - EGG
4.1 THE EGG
Modified from:
Huysentruyt F & Adriaens D
Adhesive structures in the eggs of Corydoras
aeneus (Gill, 1858; Callichthyidae)
Journal of Fish Biology 2005, 66: 871-876.
Abstract
The surface structure of the eggs of the catfish Corydoras aeneus was examined and
showed to be a unique pattern among teleosts. The surface was covered with small
protuberances, which resemble attaching-filaments of teleost eggs. Eggs were, however,
found to be very adhesive and since the species is known to inhabit turbid waters, this
rare egg attachment mode could well be related to these environmental settings.
Introduction
Modifications in the egg morphology of various teleosts often reflect the
ecological challenges a species is faced with during its embryonic life stages. In
this context, for instance, the thickness of the zona radiata interna, which has a
protective function, is directly related to the exposure to mechanical strains
(Riehl, 1996). Next to this, the zona radiata externa has been known to mediate
egg adhesion in several teleosts and various modes of egg attachment are
known (Laale, 1980; Patzner & Glechner, 1996; Riehl & Patzner, 1998; Rizzo et
al., 2002). Again, the presence and manner of egg attachment could reflect
environmental constraints placed upon the eggs, or the species in general,
during these developmental stages. Patzner & Glechner (1996) found fishes from
different environments, which exhibited the same variety in attachment
structures. Rizzo et al. (2002), on the contrary, related the absence or presence
of egg adhesion in different species to their migratory behavior and Morin & Able
(1983) related the nature of adhesive structures found in the eggs of Fundulus
heteroclitus to a variance in egg deposit sites. In addition Rizzo et al. (2002)
49
4 ONTOGENY - EGG
point out that adhesive eggs are usually larger and laid in smaller numbers, as
well as they relate egg adhesiveness to both the sedentary nature of species and
possible parental care. In any case, the presence of adhesive eggs certainly
indicates a behavioral strategy in which eggs are confined to a single locus until
hatching, a strategy which could relate to environmental settings, making it
useful to take environmental factors into account when discussing egg
morphology.
Therefore, in the armored catfish Corydoras aeneus, known to inhabit fast
current fresh water systems in Southern America (Gosline, 1940; Nijssen, 1970;
Kramer & Braun, 1983; Burgess & Quinn, 1992; Kohda et al., 1995; Fuller,
2001), adaptations in egg morphology in order to cope with a similar turbulent
environment can be expected. More so, since Kohda et al. (1995) found
adaptations in the species’ insemination strategy, presumably to ensure a high
insemination rate even in a turbulent habitat. This unique mode of insemination
in C. aeneus consists of the female swallowing sperm and quickly transporting it
through the intestine, emitting it at the anal opening into a previously emitted
ventral egg pouch (Kohda et al., 1995). However, since no motivation was
presented by these authors describing how the sperm is protected from the
intestinal environment and since this method has not been observed in any other
species, this hypothesized insemination mode remains doubtfull and needs
further study.
For this study, eggs of C. aeneus were examined in order to obtain information
on their morphology, this way possibly gathering some information on the
environmental demands met by the species during its primary developmental
stages.
Brief materials and methods
Breeding in Corydoras aeneus was induced as described under 2.1. The
fertilized eggs were photographed and measured as described under 2.2. After
this, eggs were fixed in 4% glutaraldehyde in a 0,2 M cacodylate buffer (pH 7,4).
After washing, the eggs were dehydrated in a graded ethanol series and critical
point dried under liquid CO2. The eggs were further mounted and coated with
50
4 ONTOGENY - EGG
gold using a Balzers SCD040 sputtercoater. For examination a Jeol SM840
scanning electron microscope was used at 15 kV.
Results and discussion
The eggs (n = 21) examined have a diameter of 1.47 ± 0.20 mm (± SD) and
were always laid against either a leaf or the aquarium walls. This size matches
the egg size of 1.5 mm described by Fuller (2001) for the species. A large yolk
sac is present and, like in most teleosts (Riehl & Patzner, 1998), eggs are
spherical in shape (fig. 4.1). The egg-surface is covered with little projections,
with the exception of a small circular region surrounding the micropyle (fig. 4.1B;
4.2E). These projections are c. 10 µm long, c. 5 µm wide and continuous with
the outer layer of the zona radiata, i.e. the zona radiata externa (fig. 4.2A),
which is in total c. 13 µm thick. Both the size and position of these projections
would suggest them to be villi-like protuberances as according to the definition
by Riehl & Patzner (1998). However, when comparing them to the villi described
in Leuciscus leuciscus, Alburnoides bipunctatus and Rutilus rutilus by Patzner &
Glechner (1996), the shape of the villi seems to differ from all these species. The
protuberances found here were regularly hexagonally shaped (fig. 4.2C), whilst
the villi found in the former species were irregularly shaped. In addition, after
spawning, these protuberances can lengthen considerably and act as adhesive
structures in contact to both the substrate or other eggs (fig. 4.2B). After
lengthening the projections more resemble attaching-filaments than villi-like
protuberances. Attaching-filaments, however, are only fixed at the vegetal or
animal pole or arranged in a disc or annular bulge (Riehl, 1996; Riehl & Patzner,
1998) excluding these projections from that category. In addition, this part of
the zona radiata is perforated in Corydoras aeneus with several small pores that
lie in a clear hexagonal pattern in between the projections (fig. 4.2C). Similar
hexagonal
structures
are
suggestive
of
the
honey-comb
like
pattern
characteristic of fishes of the family Percidae (Riehl & Bless, 1995; Riehl &
Patzner, 1998). This way, the arrangement and attachment mode found in the
eggs of C. aeneus was not described in other teleosts. Further, under the zona
radiata externa lies the zona radiata interna, which is c. 7 µm thick and exist
51
4 ONTOGENY - EGG
entirely out of pillar-shaped structures separated by pore-canals (fig. 4.2A). The
thickness fits the teleost average, where a zona radiata of 5-15 µm is commonly
found (Riehl, 1999), which wrongfully suggests that the eggs of C. aeneus do not
have to cope with more than average mechanical stress. The deepest zone of the
zona radiata is the zona radiata subinternus, which forms the boundary with the
oocytoplasma and is ± 0.2 µm thick. This part of the zona radiata is also highly
perforated but pores are more randomly organized throughout this inner layer.
The micropylar apparatus in C. aeneus exists of a flat, ellipse-shaped, micropylar
pit which is approximately 50 µm long and 20 µm wide (fig. 4.2D). In this pit lies
the micropylar canal, which is c. 13 µm wide (fig. 2D; F). Although the latter
measurement is somewhat unreliable due to the unclear borders of the canal,
this condition fits the type 2 micropylar configuration as described by Riehl &
Götting (1974) and Riehl (1999). The micropyle itself lies at the end of the pit
and is, after insemination, closed off by a micropylar plug to avoid polyspermy
(Riehl, 1996).
Conclusively, it can be stated that the eggs in C. aeneus exhibit a unique
surface pattern with small villi-like protuberances which resemble attachingfilaments of teleost eggs. The presence of these structures could be related to
the turbid habitat in which this species lives, but further studies on eggs of
phylogenetically related and of non-related sympatric fishes will further have to
clarify this hypothesis and demonstrate the true relation. The results of this
study
could
possibly
also
be
used
in
a
taxonomic
perspective,
since
morphological characters of teleost eggs have already been used this way (Kim &
Park, 1996; Park & Kim, 2001) and phylogenetic relations within both Corydoras
and the Callichthyidae are still unclear (Britto & Castro, 2002). However, Patzner
& Glechner (1996) also question such an application since no association
between egg morphology and taxonomy exists, at least not within the family
Cyprinidae. Morin & Able (1983) seem to confirm this by reporting interspecific
variation in Fundulus heteroclitus, but, however, do state that egg morphology
has
proved
to
help
Cyprinodontiformes.
identify
Therefore,
major
groups
information
on
and
egg
species
within
morphology
of
the
other
callichthyids will have to reveal whether a similar pattern is also present in those
species and to what extent, if any, the use of a similar character in the
phylogenetic reconstruction of this family would be advisable.
52
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
4.2 EARLY DEVELOPMENT AND GROWTH
Modified from:
Huysentruyt F, Moerkerke B, Devaere S & Adriaens D
Early development and allometric growth in the
armoured catfish Corydoras aeneus (Gill, 1858)
Hydrobiologia, submitted.
Abstract
Corydoras aeneus larvae were bred in captivity and collected at different ages in a
closely set ontogenetic series. The development of external morphology and of
allometries of several body parts was studied, attempting to reveal important steps in the
species’ early life history. Based on external morphology, the different stages in early
development of C. aeneus were identified, as described by Balon (1975). After hatching,
at a SL of 3.5 mm, an eleutherembryonic phase was present, followed by the
protopterygiolarval phase (4.4-5.7 mm SL), the pterygiolarval phase (5.7-14.0 mm SL)
and the juvenile period. In addition, an overall growth curve was established and
inflexion points were determined. Hence, it was of interest to determine the growth
coefficient k in SL = b agek with the possibility to allow for changing k over different time
points. To this end, both variables were transformed on the log-scale (log(SL) = log(b)+k
log(age)) and a piecewise linear regression method was applied where I followed the
ideas of regression spline smoothing procedures. This way, the growth curve was divided
into six different stages of growth rate. Initially, the slope was 0.05 until 0.7 dph, then
increasing to 0.18 until 4 dph, and 0.36 until 10 dph. After this, growth rate reached a
maximum of 0.76 until 24 dph, slowed down to 0.47 until 37 dph and then finally again
slowed down to 0.36. A similar analysis was also done on the data of growth in the
different body parts and these results were compared to both morphological and
literature data. This led to the conclusion that the inflexion points found in the early
development of C. aeneus matched the different key-events known in teleost early
development. The transition from endo- to exogenous feeding, when priorities also focus
on respiratory functions, was the first point at which allometries changed together with
functional demands. A second, similar alignment occurred at the transition to the
pterygiolarval phase, when priorities shift toward locomotory needs. Finally, my results
also indicated a transition to a carangiform swimming mode at approximately 8 mm SL.
53
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
Brief introduction
Morphogenesis and differentiation are very intense during early life stages (van
Snik et al., 1997) and teleost larvae commonly functionally optimize growth to
increase
fitness
(Fukuhara,
1992).
Both
of
these
processes
lead
to
a
discontinuous larval growth (in terms of rate) (Gisbert, 1999), of which a
thorough knowledge would help to understand functional trends and ecology of
the species at different developmental stages (Fukuhara, 1992).
Therefore, it was my objective to study the ontogeny of C. aeneus and, as a
first step in this study, I gathered information on growth of the species in order
to obtain insights into the timing of changes in ontogenetic allometries. Further,
knowledge on the ontogeny can also serve aquacultural applications (Fukuhara,
1992).
Materials and methods
Breeding in Corydoras aeneus was induced as described under 2.1. Afterwards,
specimens from various nests were removed post hatching at 1h, 2h, 3h, 6h,
12h, 24h, 2d, 3d up to 14d, every two days up to 24d, at 25d, 27d, 29d, 30d,
33d then every four days up to 45d, followed by one specimen every five days
until 70d, one specimen every ten days until 90d and the last two specimens at
120d (in total 41 specimens were used). All specimens were collected and
photographed as described under 2.1.
Based on standard length (SL) in function of age (in days), an overall growth
curve was established and inflexion points were determined. Hence, it was of
interest to determine the growth coefficient k in SL = b agek (Fuiman, 1983),
with the possibility to allow for changing k over different time points. To this end,
both variables were transformed on the log-scale (log(SL) = log(b)+k log(age))
and a piecewise linear regression method was applied where I followed the ideas
of regression spline smoothing procedures (see e.g. Friedman & Silverman,
1989).
54
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
Suppose we have n data points (xi,yi) (i = 1,…,n). In this case, xi and yi
respectively represent the age and standard length of observation i on the logscale. Assume further that the data points satisfy:
y i = f ( xi ) + ε i
ε i ~ N (0, σ 2 ); i = 1, K, n (*)
with
f ( xi ) = β 0 + β1 xi + ∑ β j +1 ( xi − t j ) I ( xi > t j ) .
m
j =1
m is the number of ‘knots’ or inflexion points and {t j ; j = 1,K, m} is the set of
these knots ( {10 j ; j = 1,K, m} is then the set on the original scale). Note that
t
min( xi ) < t1 < K < tm < max( xi ) and that we restrict {t j ; j = 1,K, m} to be a subset of
{xi ; i = 1,K, n} . I ( xi > t j ) = 1 as soon as xi > t j and 0 otherwise. Model (*) implies that
the slope or growth rate before the first knot equals β1 . After the first l knots
( l ≤ m ), the slope becomes β1 +
∑β
l
j =1
j +1
. The goal is to estimate f ( xi ) based on
the observed data points. This means that a set of knots needs to be chosen and
that the parameters β = ( β 0 , β 1 , β 2 , K , β m +1 ) need to be estimated.
In the absence of knots, f ( xi ) = β 0 + β1 xi and the model to be fitted is a simple
linear regression model. I opted to select a number of knots using a forward
search procedure and to estimate β
using ordinary least squares. More in
particular, I followed the forward addition strategy as described in Lee (2002)
who gives an overview of algorithms for ordinary least squares regression spline
fitting.
The generalized cross-validation (GCV, see Friedman & Silverman, 1989) was
used as the criterion to select the ‘best’ model. The best model is then the model
which minimizes
1 / n∑ ( y i − fˆ ( xi )) 2
n
GCV =
i =1
(1 − (3m + 1) / n) 2
with fˆ ( xi ) the estimate for f ( xi ) .
The sequential search procedure works as follows. The initial model is the
model without knots and its GCV-value is calculated. In each step of the search
procedure, a knot is added such that the largest decrease or smallest increase in
55
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
GCV is seen. This process is repeated until a user-defined maximum number of
knots is obtained. In all my analyses, this maximum was set equal to 10. After
this procedure, the model with the smallest GCV among all candidate models is
chosen as the final model.
In order to investigate allometric growth, a similar analysis was done on growth
of the different body parts, this time using SL as the independent variable. For
this purpose, head length, abdominal length, post-anal length, caudal fin length,
pectoral fin length and eye diameter (measured in a horizontal plane) were
measured (fig. 4.3) and used as dependent variables. Not all measurements
could be made on all specimens (e.g. a damaged caudal fin, anal opening not yet
present, etc.), where this was the case, specimens were omitted from the
analysis.
This
resulted
in
slightly
different
sample
numbers
for
these
measurements.
Afterwards,
the
specimens
were
drawn
and
external
morphology
was
described. The occurrence of important events was described providing the
minimal SL at which these events were observed. Based on morphological
criteria different stages of early development were demarcated and terminology
was used according to Balon (1975, 1999). However, testing whether ontogeny
in C. aeneus occurred saltatorily, was beyond the scope of this study, since this
would require a more extensive study of disruptive morphogenetic events
throughout the species’ ontogeny (see 5.1 for a more thorough approach of this
matter).
Results
Hatching occurs after an incubation period of three to four days. Early
development
after
eleutherembryonic
hatching
phase
was
divided
(between
into
hatching
four
and
different
exogenous
parts:
the
feeding),
protopterygiolarval phase (until start of finfold differentiation), pterygiolarval
phase (until completion of finfold differentiation) and juvenile period (until sexual
maturation).
56
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
Morphological development
At hatching, larvae have a SL of 3.5 ± 0.2 mm (± SD) (n = 6). From this size
up to 4.4 mm SL a yolk sac is present. During this eleutherembryonic phase (fig.
4.4A), specimens are unpigmented and the mouth and anal opening are still
closed. Distinctly large, ovally shaped, pectoral fins are present, which are more
like large skin folds than true fins and thus do not yet bear any fin rays. Two
pairs of oral barbels, the maxillary and external mandibular barbels, are present
and the median finfold is undifferentiated. The eye diameter in these free
swimming embryos is small and the eyes are still almost devoid of pigment.
From three to five dph, at a SL of 4.4 to 5.7 mm, a yolk sac is no longer
present and an extra pair of oral barbels, the internal mandibular barbels, begin
to develop (fig. 4.4B). The anus and mouth open, as does the opercular cavity.
At this protopterygiolarval phase, faint body pigmentation appears on the lateral
side and head of these specimens. The pigmentation on the head is concentrated
in a double line, which runs from the buccal area over the eye, to the back of the
head. Further pigmentation is concentrated in spots on the lateral side of the
body and in the finfold and pectoral fins. In these fins, pigmentation is
concentrated where future fin rays are to develop. The eye in these specimens is
larger and more pigmented. Near the end of this phase the median finfold begins
to differentiate, which starts with an indentation behind the future dorsal fin and
a slight narrowing of the finfold near the caudal peduncle.
In specimens aged 7-23 dph, with a SL between 5.7 and 14.0 mm (the
pterygiolarval phase), the finfold further differentiates (fig. 4.4C; D; E). At 11
dph (7-8 mm SL) the dorsal fin is almost fully detached from the finfold, pelvic
fins start to develop and pigmentation is present at the base of the future anal
and adipose fin, where in both cases the finfold also starts to indent. At 16 dph
(11-12 mm SL) all fins are fully detached, except for the adipose fin. This fin is
still surrounded anteriorly by a small part of the median finfold, which totally
disappears at a SL of 14 mm. The fin rays in the different fins develop in a
partially different order. Fin rays begin to develop in the caudal fin, pectoral and
dorsal fins first and at a SL of 8 mm both dorsal and caudal fin have reached
their definite number of fin rays. At this time the pelvic and anal fin rays start to
develop, both of which become fully developed at a SL of 9-10 mm. A spine
within the adipose fin is present from a SL of 11-12 mm on and in the pectoral
fins the definite number of fin rays is reached at 22-25 mm SL. At 11 mm SL, the
57
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
first bony scutes, typical for callichthyid fishes, start to develop as well. The first
scute that develops is the one anterior to the adipose fin, which partially covers
the adipose fin spine. From there on three more, small, dorsal scutes develop in
an postero-anterior order. At a SL of 15 mm, several small scutes develop
dorsally and ventrally on the caudal peduncle. Finally, at approximately 16 mm
SL, the first lateral scutes start to develop postero-anteriorly along the midline of
the body with a further centripetal development of all scutes.
Allometric growth
In the growth curve of Corydoras aeneus, which ranges from 0 until 120 dph,
five different inflexion points were calculated with subsequent changes in growth
rate (slope) (fig. 4.5). The first inflexion in the growth curve is located at 0.7 dph
(which corresponds to a fitted value of 3.9 mm SL), at which the estimated
growth rate increases from 0.05 to 0.18 (p<0.01). Further, the growth curve in
specimens over 0.7 dph shows a second inflexion at 4 dph (5.4 mm SL), where
the slope further increases to 0.36 (p<0.01). A subsequent third inflexion is
present at 10 dph (7.4 mm SL), where growth rate increases even further up to
0.76 (p<0.001). After this, at 24 dph (14.5 mm SL), growth slows down to a rate
of 0.47 (p<0.01). Finally, the last inflexion in this growth curve is located at 37
dph (17.9 mm SL), where growth rate again decreases to 0.36, but this growth
rate is not significantly different from the rate found before this point (p = 0.20).
Note that I used the GCV-criterion to select the best model and not statistical
significance.
In the context of body part allometries, the model with one inflexion point (R²
= 0.958) has the smallest GCV. This model shows that growth of the head is
positively allometric (slope = 1.44) up to a SL of 13.1 mm, after which the rate
drastically decreases to 0.69 (p<0.001) (Fig. 4.6A). Growth of the abdomen is
also best modelled by a single inflexion point model (R² = 0.947) and in this
case, growth starts nearly isometric (slope = 1.19), and then declines to 0.80
(p<0.05) from 9.8 mm SL on (fig. 4.6B). In the regression which describes
growth in the postanal region, no improvements over the simple linear
regression model were found in GCV by introducing inflexion points. This way,
growth rate in the postanal region in relation to SL was found to remain constant
at 0.81 and significantly different from 0 (p<0.001) throughout the entire
ontogenetic series studied here (R² = 0.909) (fig. 4.6C). The analysis of growth
58
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
in the caudal region showed a two inflexion point model as best suited (R² =
0.978). Growth in this region starts out highly positively allometric (slope =
4.72) up to a SL of 6.2 mm, after which it slows down between 6.2 and 21 mm
SL to a rate of 1.38 (p<0.001). After this it finally decreases to a rate of -0.20
(p<0.001) (fig. 4.6D). Further, a single inflexion point was also found in the
growth curve of the eye (R² = 0.973). This inflexion was situated at a SL of 9.8
mm where the rate drastically decreases (slope = 1.89 to 1.02, p<0.001) (fig.
4.7A). Finally, in the growth of the pectoral fins, no inflexions were detected,
which again leaves the simple linear regression as the best fit (R² = 0.924), with
a constant growth rate of 1.22 (significantly different from 0 with p<0.001) as a
result (fig. 4.7B).
Discussion
Hatching size in Corydoras aeneus is found to be quite constant (3.5 ± 0.2 mm
(n = 6)) at 25°C, which is presumed to be correlated to the low variance in egg
size (1.47 mm ± 0.20 (n = 21)) (Osse & van den Boogaart, 1995) (see also 4.1).
The scope of this paper, however, was to study growth allometries in C. aeneus,
and the timing of shifts in growth rates, both for overall growth as for growth of
specific body regions (head, abdomen, post-anal region, caudal fin, pectoral fin
and eye diameter). I hypothesized that the chronology of these allometric
changes would be related to the chronology of important early life history events,
and would therefore reflect an evolutionary ontogenetic response to functional
demands. It has already been suggested that allometric growth patterns closely
match the expected priorities for executing the necessary biological roles (Osse
et al., 1997), or as Kováč & Copp (1999) put it: “a certain level of development
is necessary,... , to coincide in functional readiness”. Therefore, in this study, the
chronology of relevant shifts in allometric growth rate was statistically assessed
by calculating inflexion points in growth rates of different body regions during
ontogeny. In addition, literature on early life history traits in teleosts in general
provided the framework to verify possible correlations with the allometries found
in this study.
59
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
A first correlation I hypothesized was based on the fact that predation is one of
the main agents that result in larval mortality (Bailey & Houde, 1989). As a
result, a higher efficiency in the use of structures that counter this threat can be
expected in these early life stages. In contrast to some callichthyid congeners
that produce foam nests for their hatchlings (e.g., Megalechis), such hiding
facilities are not provided for larval C. aeneus. Consequently, predator avoidance
would require an increase in swimming efficiency, and thus burst swimming can
be expected to be an important escape response in C. aeneus hatchlings. Weihs
(1980), in his study on Engraulis mordax, found a continuous swimming mode to
be more efficient in larvae under 5 mm, while in larvae with a length between 5
and 10 mm burst-and-coast swimming became more effective. Blaxter (1986),
on the other hand, mentions that a change in swimming mode occurs as the
caudal fin develops and inertial forces start to play a bigger role in locomotion.
My observations indeed show a higher growth rate of the caudal region (slope =
4.72) up to 6.2 mm SL, which provides a ‘functional readiness’ for burst and
coast swimming after this stage. At a SL of about 5 mm, notochord flexion
occurs, which precedes both a drastic increase in overall growth rate (0.18 to
0.36) and the start of the pterygiolarval phase, both of which occur around 5.5
mm SL. These factors indeed indicate a shift in the components of the
locomotory apparatus which may be related to an onset of a change in swimming
mode at 5 mm SL. Such a shift toward burst and coast swimming is believed to
enlarge the dispersal and foraging range and improve predator avoidance
(Gisbert, 1999). Therefore, based on these findings, I expect a similar change in
swimming mode to occur at that point but a kinematic analysis of the ontogeny
of larval swimming in C. aeneus will have to confirm this.
Apart from predation, a second correlation I predict is based on the fact that
starvation becomes a major threat for larval survival once the yolk sac becomes
depleted (Bailey & Houde, 1989). Accordingly, the point at which this depletion
occurs is critical during larval development (Pedersen et al., 1990; Jardine &
Litvak, 2003). After this depletion, exogenous feeding becomes obligatory and
the presence of a functional feeding apparatus is required (van Snik et al., 1997;
Jardine & Litvak, 2003). Given this, I expected that under 4.4 mm SL (= point of
yolk sac depletion) development would, next to predator avoidance, also be
focused on the completion of the differentiation of the feeding apparatus to a
first functional stage. My observations show that, around this transition in C.
60
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
aeneus (4.4 mm SL), growth rate in head length is positively allometric (slope =
1.44). During the early life stages, protrusion of upper jaw bones is not yet
functional so that high suction forces must be generated in order to catch prey.
Higher suction forces can be related to head volume (Osse, 1990), which implies
that a fast head growth (as is the case here) would indeed allow an increase in
suction efficiency.
The fact that, in the earliest stages of development in some teleosts, growth
intensities are not distributed uniformely across the body was already observed
by Fuiman (1983). For the earliest stages, Fuiman (1983) describes a continuous
U-shaped gradient across the body axis in between a high intensity of growth
rates at the terminal growth centres (head and tail). Fuiman (1983) also
described that in juveniles, this gradient approaches isometry, which probably
remains throughout adulthood. As discussed above, in C. aeneus, a higher
growth rate in both head and tail are also present in the earliest phases. Fig. 4.8
shows those different growth rates mapped along the body axis for three
different phases of development in C. aeneus. The graph also shows that growth
turns almost isometric near the start of the juvenile phase, with the exception of
the tail region. However, since standard error in the growth rate of the tail region
is very high, it seems probable that the model of Fuiman (1983) also applies
here. Fuiman (1983) explained this U-shaped profile in early development
discussing the need of propulsive force provided by the tail region (to escape
predators), and the need of a large head for the elaboration of feeding and
sensory mechanisms, as we also already suggested for C. aeneus.
Given the properties of the physical environment, like the high viscosity of the
aqueous medium and of the biological environment such as the relatively large
size of the available prey organisms considered, size acts as an important
constraint on development of form and function during early ontogeny. This is a
view that gains importance in many recent developmental studies on fish larvae
(Strauss, 1984; Buckel et al., 1995; Stern & Elmen, 1999; Adriaens & Verraes,
2002) and can be exemplified here too. Feeding strategies of fish larvae as well
as
locomotor
habits
demonstrably
show
pronounced
changes
at
certain
developmental stages and certain size ranges are rather strictly correlated to
typical morphogenetic events or differentiation of particular organ systems. In
Cyprinus carpio (van Snik et al., 1997), for example, a second change in
swimming mode from anguilliform to subcarangiform was found around 8 mm
61
4 ONTOGENY – EARLY DEVELOPMENT AND GROWTH
SL. Gisbert (1999) found a change in swimming style to be correlated to the first
appearance of caudal fin rays in the chondrostean fish Acipenser baeri. My
observations show that at 8 mm SL caudal fin rays start to develop and anal,
adipose and pelvic fins separate from the respective finfolds and that, just prior
to this, at 7 mm SL, overall growth rate reaches a maximum (0.76). Therefore, a
similar change in swimming mode is expected around 8 mm SL in C. aeneus.
This altered swimming would also imply a decrease of the head yaw (Osse,
1990), which again results in better aiming skills for prey capture. In this context
a decrease of eye and abdomen growth rate (1.19 to 0.8 and 1.89 to 1.02,
respectively) shortly after this shift indeed may indicate changes in prey
preference and/or visual performance in relation to predation.
Conclusively, the analysis of inflexion points in growth curves, in combination
with observed changes in morphological development, holds valuable information
on changes in functional demands throughout ontogeny. Such an analysis can
provide a framework of these shifting functional demands placed on a developing
larvae, useful for both the comparison with the results of additional ontogenetic
research on the species, as well as for aquacultural applications.
62
4 ONTOGENY – CRANIUM
4.3 THE CRANIUM
Modified from:
Huysentruyt F, Brunain M & Adriaens D
Ontogeny of the chondrocranium in Corydoras aeneus
(Gill, 1858) (Callichthyidae, Siluriformes)
Journal of Morphology, accepted.
and:
Huysentruyt F, Geerinckx T, Brunain M & Adriaens D
Development of the osteocranium in Corydoras aeneus
(Gill, 1858) (Callichthyidae, Siluriformes)
Journal of Morphology, submitted.
Abstract
Callichthyids take a basal position in the loricarioid evolutionary lineage leading up to
an algae scraping feeding mechanism in the loricariid family. Therefore, the study of the
morphology and development of a callichthyid representative would contribute to a better
knowledge on the differences in cranial morphology and their impact on feeding ecology
within this superfamily. Therefore, development in the cranium of Corydoras aeneus was
studied based on 48 cleared and stained specimens and ten series of serial sections. The
latter sections were also digitized and used for 3D-reconstructions. Development overall
follows the typical siluriform trends in cranial development. Even the low complexity of
the chondrocranium at hatching fits the trend observed in other siluriforms, although
other
studies
showed
loricarioid
hatchlings
to
generally
show
more
complex
chondrocrania. In contrast to other catfish, in C. aeneus, the notochord was never found
to protrude into the hypophyseal fenestra. In addition, also differing from other
siluriforms, a commissura lateralis is present, a state also reported for Ancistrus cf.
triradiatus (Geerinckx et al., 2005). The splanchnocranium again has the typical
siluriform shape during its ontogeny, with the presence of a compound hyosymplecticpterygoquadrate plate, although not fused to the neurocranium or interhyal at any time
during ontogeny, a state described earlier for Callichthys callichthys (Hoedeman, 1960b;
Howes & Teugels, 1989). The most striking differences found in comparison to other
catfishes, however, involves the branchial basket, which arises as a single element with a
further differentiation from the middle arches on in both a rostral and caudal direction. As
in other studies on siluriform osteocranial formation, ossifications generally appeared as
a response to functional demands (Tilney & Hecht, 1993; Vandewalle et al., 1995;
63
4 ONTOGENY – CRANIUM
Vandewalle et al., 1997; Adriaens & Verraes, 1998). This way, early development of the
skull was found to occur in two distinct phases. In a first phase, several new bony
elements, all of dermal origin and related to feeding, appear shortly after yolk depletion
(4.4 mm SL). After this, in between 5 and 8 mm SL, developmental priorities seem to
shift to size increase of the cartilaginous skull and no new bony elements appear. Finally,
a second phase of osteogenesis occurs from 8-18 mm SL, in which all remaining dermal
and perichondral bones appear.
Brief introduction
A thorough knowledge of anatomical ontogeny is of critical importance in
understanding the functional trends during early development, since organisms
must be functional at each moment, including the early ‘temporary’ stages
(Fukuhara, 1992; Koumoundouros et al., 2001a, b; Geerinckx et al., 2005).
In this context, the study of the morphology and development of C. aeneus can
contribute to a better knowledge on the differences in cranial morphology and
their impact on feeding ecology within this superfamily. This is especially the
case for the earliest stages in ontogeny, at which point the yolk sac often
becomes depleted and feeding becomes obligatory.
Therefore, the aim of this study is to provide a complete description of the
ontogeny of the cranium in C. aeneus, adding to the wide variety of data on
catfish cranial development (Kindred, 1919; Ballantyne, 1930; Bamford, 1948;
Srinivasachar, 1957a, b, 1958a, 1959; Hoedeman, 1960b; Vandewalle et al.,
1985; Surlemont et al., 1989; Surlemont & Vandewalle, 1991; Adriaens &
Verraes, 1994, 1997b; Geerinckx et al., 2005) and providing a basis for further
ongoing ontogenetic research on this species. The data generated this way were
compared to the general trends in siluriforms and to similar data published
earlier by Geerinckx et al. (2005) on Ancistrus cf. triradiatus, a loricariid
representative. The comparison of the differences and similarities of both
ontogenetic sequences provides additional information on the evolutionary
processes leading up to the remarkable niche diversification in the Loricariidae
family. This way, I hypothesize that, given the close phylogenetic relationship of
the species compared, both ontogenies would exhibit a great amount of
64
4 ONTOGENY – CRANIUM
similarity, only showing significant differences in the development of those
elements associated with the specialized feeding mechanism in A. cf. triradiatus.
Given the fact that the study of the developing cranium in C. aeneus is hereto
undescribed, an additional aim of this study is to link the morphological aspects
of the developing skull to the species’ most important early life history traits.
Therefore, results are compared to data on the species’ early life history (see
also 4.2), providing knowledge on morphological trends guided by possible
functional constraints and environmental preferences at different developmental
stages (Fukuhara, 1992). This way, it is also expected that differences in
functionality during the early ontogeny of the skull between the different species
studied would be reflected in the timing and sequence of the developing
structures associated with these specific functions.
Brief materials and methods
The specimens examined are presented in table 2.1. For clearing and staining
procedures, procedures on serial sectioning and 3D-reconstructions see 2.2.
Bone terminology used throughout this paper follows Harrington (1955),
Adriaens & Verraes (1998) and Arratia (2003a). In the earliest stages, 3Dreconstructions sometimes yielded some left-right asymmetry, which, however,
could very well be an artifact of the digitization, since very low amounts of
cartilage are present in these stages. Therefore, and given the fact that, in no
other study done on early catfish chondrocrania, as well as in my cleared and
stained specimens a left-right asymmetry was found, this asymmetry was
treated as an artefact and not further discussed.
Results
Stage 1: 3.3 mm SL (hatchling) (fig. 4.9A)
Neurocranium
65
4 ONTOGENY – CRANIUM
At
hatching,
the
neurocranium
is
rudimentary
and
chondrification
is
concentrated around the notochord tip. Bordering this notochord tip laterally are
two parachordal cartilages, which interconnect anterior to the notochord through
the acrochordal cartilage. On their lateral margins, both parachordal cartilages
connect to the basiotic laminae, making up the primordium of the basal plate.
Further laterally this plate contacts the dorsocaudally oriented anterior part of
the otic cartilages through the anterior basicapsular commissure. This way, two
large
posterior
metotic
fissures
appear,
through
which
the
nervus
glossopharyngeus (IX) and vagus (X) pass. At the rostral margin of this basal
cartilaginous skull, two polar cartilages are present.
Splanchnocranium
At about the time of hatching, the splanchnocranium is confined to the
suspensorium, which arises as a single chondrification or hyosymplecticpterygoquadrate plate (sensu Arratia, 1992) and articulates dorsally with the
neurocranium
at
the
level
of
the
otic
cartilage.
This
hyosymplectic-
pterygoquadrate plate does not yet bear a process for articulation with the
operculare, nor is the foramen truncus hyomandibularis formed at this stage.
Stage 2: 4.0 mm SL (figs. 4.9B; 4.13A; B)
Neurocranium
At 4.0 mm SL, the trabecular bars are formed as rostral expansions of the polar
cartilages. As is to be expected in a platybasic teleost skull, these trabecular bars
lie well separated from each other and are slightly curved. Not yet touching
rostrally, they leave a wide hypophyseal fissure at the anterior margin of the
skull. Since the internal carotid artery still passes through this large fissure (and
not through a smaller fissure in the cartilaginous trabecular bars themselves) it is
impossible to distinguish the trabecular bars from the polar cartilages at this
point. Observations on later stages and other siluriform fish, however, lead to
the assumption that the posterior part at this point would correspond to the polar
cartilage (Adriaens & Verraes, 1997b; Geerinckx et al., 2005).
Splanchnocranium
At 4.0 mm SL, the entire hyoid arch is also formed as a single element,
incorporating both left and right hypohyals and ceratohyals. In addition, the
hyosymplectic-pterygoquadrate plate has broadened and shifted its orientation
66
4 ONTOGENY – CRANIUM
from oblique to vertical. At this level, the dorsal and ventral parts of the
suspensorium can be distinguished as the hyosymplectic and quadrate part
respectively. In the branchial basket, which is still very much compressed
dorsoventrally, the presence of chondrocytes, as observed on serial sections,
already indicates the onset of branchial arch formation, in which all five arches
are present and fused together ventrally. In addition, the first four arches are
also fused at the dorsal side of the basket, forming a large dorsal plate.
Stage 3: 4.4 mm SL (figs. 4.10; 4.13C; 4.14A; B)
Neurocranium
At this stage, both trabecular bars contact each other rostrally, forming the
ethmoid cartilage, and expanding laterally into the solum nasi. Thus, the fissura
hypophysea is closed off, forming the fenestra hypophysea. At the caudal side of
this fenestra, the otic cartilages have expanded ventrolaterally, contacting the
polar cartilages through the lateral commissures, splitting up the fenestra
sphenoidea into anterior and posterior parts. At this stage, serial sections already
show the passage of the nervus oculomotorius (III), trochlearis (IV), trigeminus
(V), and the mandibular branch of the nervus facialis (VII) through the anterior
part and the passage of the ramus hyomandibularis of the nervus facialis (VII)
through the posterior part. At the level of the basal plate an additional
connection with the otic cartilage is formed through the posterior basicapsular
commissure, splitting up the metotic fenestra. Serial sections further indicate
that
the
aperture
present
between
anterior
and
posterior
basicapsular
commissure is penetrated by the nervus glossopharyngeus (IX), but not by the
nervus vagus (X), which makes this the basicapsular fenestra. The posterior
subdivision of the metotic fenestra is not penetrated by any nerves and will here
be called foramen “A.” Caudally, two occipital pilae have emerged, which fuse at
the dorsal side of the neural tube, forming the tectum posterius. At the caudal
margin of this chondrocranium, the lamina basiotica has further expanded,
rigidifying the neurocranial floor. At the dorsal side of the chondrocranium, both
otic capsules have expanded rostrally forming the taeniae marginales, which
already interconnect through the epiphyseal bridge but do not yet protrude
further rostrally.
67
4 ONTOGENY – CRANIUM
Splanchnocranium
At 4.4 mm SL, both the interhyal, Meckel’s cartilage and autopalatine first
appear, all as separate elements. The hyosymplectic-pterygoquadrate plate has
developed a processus opercularis for the later articulation with the opercular
bone. In the branchial basket, the central copula has split up into an anterior and
a posterior part, with the former part situated near the hypobranchial parts of
the first three branchial arches and the latter part with those of arches IV and V.
The medial tip of the hypobranchial part of the third arch is situated in between
these two copulae. Furthermore, hypobranchials II and III are already present as
separate elements. All arches already have ceratobranchial cartilages and arches
I-IV bear epibranchials, none of which has detached from the ceratobranchial
parts at this point. One large infrapharyngobranchial element is present as a
separate element, with the anterior tip articulating with epibranchials one and
two, and the posterior part articulating with epibranchials III and IV.
Teeth are already present in both the dorsal and ventral part of the basket, but
no ossifications supporting these teeth have been formed at this point. This is
also the case for both both lower and upper oral jaw, where teeth primordia are
also present without the presence of supportive ossifications. The autopalatine
still articulates with the maxillary barbel through the submaxillary cartilage and
the
cartilaginous
hyosymplectic-pterygoquadrate
plate
has
developed
an
articulation with the opercular bone, the first ossified element to appear.
Stage 4: 4.9 mm SL (figs. 4.11; 4.13D; 4.14C; D))
Neurocranium
At this stage, the taenia marginalis makes contact with the solum nasi, through
the commissura spheno-ethmoidalis, forming the lamina orbitonasalis. Rostrally,
the taenia marginalis further expands, forming the onset of the commissura
sphenoseptalis. In the skull base, the fenestra basicapsularis is split up by the
commissura basivestibularis into an anterior and a posterior part, with the latter
accommodating the nervus glossopharyngeus (IX).
Splanchnocranium
The hyosymplectic-pterygoquadrate plate further differentiates into a more
distinct pars quadratum, which articulates with Meckel’s cartilage and a pars
hyosymplecticum, articulating with the neurocranium. In the latter, the foramen
68
4 ONTOGENY – CRANIUM
truncus hyomandibularis is present from this stage on (not visible in fig. 4.11).
Additionally, the onset of a pterygoid process is present. The hyoid arch has
started to split up medially and the two lateral halves have begun to differentiate
into a more distinct ceratohyal and hypohyal part, with the latter articulating with
the interhyal. In the branchial basket, epibranchials II and III have detached
from the ceratobranchials but in this specimen, the hypobranchial of the second
arch was still loosely connected to the ceratobranchial (even though they were
already separated in the specimen used for the previous stage). Cartilage had
also disappeared at the level of the articulation between cerato- en epibranchial
IV, indicating that both elements probably detach soon after this stage. In
addition, the infrapharyngobranchial cartilage has expanded anteriorly and has
started to differentiate into a more clear anterior and posterior part, again clearly
articulating with, respectively, the first and last two epibranchials.
Both a thin layer of dentary bone and the maxillary and premaxillary bones are
present, all of which are dermal ossifications supporting teeth. The maxillary also
supports the maxillary barbel at this point and articulates through a double
dorsal process and two submaxillary cartilages with the autopalatine. The bone
further has a medial articulation with the premaxillary, which in turn bears a
dorsal process for articulation with the ethmoid cartilage. In addition, the
parurohyal
bone
has
started
to
ossify
in
the
tendon
connecting
the
sternohyoideus muscle to the hyoid.
Stage 5: 5.3 mm SL (figs. 4.12; 4.13E; 4.14E; F))
Neurocranium
At this stage, the solum nasi has expanded laterally and the commissura
sphenoseptalis now is fused to the ethmoid cartilage at the level of the lamina
precerebralis. The preorbital base forms another vertical connection between
taeniae marginales and trabeculae cranii slightly behind the lamina orbitonasalis,
leaving a small orbitonasal foramen, which disappears later during ontogeny.
Therefore a large foramen fila olfactoria appears. The cartilage of the skull floor
has expanded further, comprising a caudal expansion of the acrochordal cartilage
and closing off both the foramen “A” as well as narrowing the foramen for the
passage of the nervus vagus (X).
69
4 ONTOGENY – CRANIUM
Splanchnocranium
The pterygoid process of the hyosymplectic-pterygoquadrate has expanded
significantly. In the branchial basket the first epibranchial has detached from the
rest of the arch but in the fourth arch, this element is still slightly connected to
the ceratobranchial. In middle of the fifth ceratobranchial, along its medial side,
a thin bony plate supporting the ventral pharyngeal teeth has started to develop.
Along the ventromedial side of infrapharyngobranchials III and IV, a similar
ossification was observed, supporting the dorsal pharyngeal teeth.
Stage 6: 6.3 mm SL (fig. 4.13F)
Neurocranium
No additional changes have occurred in the chondrocranium at this point.
Splanchnocranium
At this point, in the gill arches, all epibranchials have detached from the
ceratobranchials and the infrapharyngobranchial part has split up into a posterior
and anterior part articulating with the last and first two arches, respectively. In
both the fourth and first arch, the hypobranchial part is still fused to the
ceratobranchial part. In the first arch, these elements will separate shortly after
this stage, while, in the fourth arch, both elements never detach.
8.2 mm SL (figs. 4.15; 4.20A; 4.21A)
Neurocranium
In the skull roof, three ossification centres are present. Rostrally, the dorsal
perichondral supraethmoid bone expands over the ethmoid cartilage. Next to
this, on the dorsomedial side of the taeniae marginales, just behind the
epiphyseal bridge, the frontal bones appear. On the caudal margin of the
neurocranium, the parieto-supraoccipital bone has started to ossify at the level of
the tectum synoticum. The skull floor is made up of the ventral hypoethmoid
bone (not illustrated), caudally contacting the prevomeral bone, which is flanked
by two palatal splints at is posterior margin. Caudal to this ossification, the
borders of the hypophyseal fenestra have started to ossify, forming the early
parasphenoid
bone.
Caudolaterally,
the
early
posttemporo-pterotico-
supracleithra are present, in which the separate elements composing this
complex bone can not be distinguished.
70
4 ONTOGENY – CRANIUM
Splanchnocranium
Ossification of the palatine has started at its posterior margin and, in the lower
jaw, Meckel’s cartilage has also started to ossify. This way, at its anterior margin,
the mentomeckelian appears, while posteriorly, the articular and retroarticular
bone become apparent, still separated at this point. In the suspensorium, the
hyomandibular
and
quadrate
parts
show
ossification
centres,
enforcing
articulations with the opercular and articular bone respectively. In the hyoid
arch, the ventral hypohyal and anterior and posterior ceratohyals ossify, as do
the branchiostegal rays. Finally, in the opercular series, the interopercle and
suprapreopercle have formed. The opercle has now developed its typical oval
shape and oblique dorsoventral orientation.
9.7 mm SL (figs. 4.15B; 4.18A; 4.20B; 4.21B)
Neurocranium
The supraethmoid bone has expanded caudolaterally and ventrally, it has fused
to the hypoethmoid bone, forming the mesethmoid bone. Laterally, this complex
shows two expansions of dermal origin, which represent the laterodermethmoid
parts. The frontals have expanded in both the anterior and posterior direction,
next to a small expansion which covers the epiphyseal bridge and connects both
at the midline. At the caudal margin of the neurocranium, anterior expansions of
the parieto-supraoccipital bone almost reach the frontal bones. In the skull floor,
the hypophyseal fenestra has started to close, due to further ossification of the
parasphenoid. Next to this, the orbitosphenoids have started ossifying at the
anterior border of the sphenotic fenestra. In the otic region, both the prootics in
the skull floor as well as the sphenotics in the skull roof are present. In the latter
bones, the dermosphenotic and autosphenotic parts are already fused. At the
caudal margin of the otic region, the posttemporo-pterotico-supracleithrum has
expanded dorsally, housing the posttemporal branch of the cranial lateral line
system. Ventrally, the basioccipitals and exoccipitals are present.
Splanchnocranium
The
pterygoid
process
has
started
to
ossify
at
its
tip,
forming
the
metapterygoid and, next to this, the preopercular bone is present.
71
4 ONTOGENY – CRANIUM
Gill arches
Ceratobranchials III-IV, epibranchials I-IV, infrapharyngobranchial III (anterior
infrapharyngobranchial) and infrapharyngobranchial IV (posterior infrapharyngobranchial) have all started to ossify in the centre of their respective cartilaginous
elements.
11.2 mm SL (figs. 4.16A; 4.20C; 4.21C)
Neurocranium
All skull roof bones have expanded, rigidifying the entire skull. The lateral
ethmoid bone is present, with the prefrontal and parethmoid parts directly
appearing as fused elements. Further, the pterosphenoid ossifies at the
posterodorsal margin of the fenestra sphenoidea and, in the skull floor, the
hypophyseal fenestra has closed entirely.
Splanchnocranium
In the splanchnocranium, ossification of the interhyal bone has started.
Gill arches
Cerato- and hypobranchials I-II have began to ossify, although, in contrast to
the gill arch bones formed earlier, ossification starts at the rostrolateral side of
the cartilaginous hypobranchials.
13.9 mm SL (figs. 4.16B; 4.18B; 4.20D; 4.21D)
Neurocranium
Bone expansions have further rigidified the skull structure.
Splanchnocranium
All ossified parts have expanded and in the hyoid arch, with the appearance of
the dorsal hypohyal, all bones are present.
Gill arches
Basibranchials II and III ossify at this stage, completing ossification in the
branchial arches.
72
4 ONTOGENY – CRANIUM
16.0 mm SL (figs. 4.17A)
Neurocranium
At this stage, only the nasal and first infraorbital appear as new ossifications in
the neurocranium.
18.0 mm SL (figs. 4.17B; 4.19)
With
the
second
infraorbital
present,
the
skull
resembles
the
adult
configuration. Next to this, the palatal splints have merged to the prevomer,
although additional observations on serial sections, show them to still be present
in some adult specimens, albeit highly diminished in size. This suggests that their
reported absence in other callichthyid genera like Callichthys, Dianema,
Lepthoplosternum, Hoplosternum and Megalechis is probably the result of a
fusion or even secondary loss during ontogeny.
Discussion of chondrocranial development
At hatching the chondrocranium is rudimentary and concentrated around the
notochord tip, a situation common in catfishes (Adriaens & Verraes, 1997b). The
fact
that
chondrification
has
started
in
both
the
neurocranium
and
splanchnocranium at this stage is a state also described in most studies dealing
with teleost ontogeny (e.g., de Beer, 1937; Adriaens & Verraes, 1997b;
Geerinckx
et al.,
2005).
In
contrast,
however,
the
complexity
of
the
chondrocranium at hatching is highly variable in teleosts. Species like Salmo
letnica, Salmo trutta fario (Salmoniformes) and Oryzias latipes (Beloniformes)
already exhibit a fairly high level of neurocranial chondrification at hatching,
while
in
other
teleost
(Synbranchiformes),
Solea
species
solea
such
as
Mastacembelus
(Pleuronectiformes)
and
armatus
Hepsetus
odoe
(Characiformes) no chondrification is present at that time (de Beer, 1927;
Bhargava,
1958;
Bertmar,
1959;
Langille
&
Hall,
1987;
Wagemans
&
Vandewalle, 1999; Ristovska et al., 2006). However, several studies indicate that
siluriform chondrocrania are generally limited in complexity at the time of
hatching. Species like Chrysichthys auratus (Claroteidae) and Heterobranchus
longifilis and Clarias gariepinus (Clariidae) show little to no chondrification at
73
4 ONTOGENY – CRANIUM
hatching while the chondrocranium in hatchlings from other species like
Heteropneustes
fossilis
(Heteropneustidae)
is
only
slightly
developed
(Srinivasachar, 1959; Adriaens & Verraes, 1997b; Vandewalle et al., 1997,
1999). Studies on species of the Loricarioidea superfamily, like Ancistrus cf.
triradiatus (Loricariidae) and Callichthys callichthys (Callichthyidae), however,
show hatchlings of this group to have a more developed chondrocranium
(Hoedeman, 1960b; Geerinckx et al., 2005). As an exception, Hoplosternum
littorale (Callichthyidae), only showed localized patches of connective tissue
where chondrification was taking place, but without actual cartilage formed at
the time of hatching (Ballantyne, 1930). This way, the complexity of the
chondrocranium at hatching in Corydoras aeneus seems to fit the general trend
found in Siluriformes, rather than that found in the Loricarioidea. In this context,
egg size has often been suggested as a key factor influencing larval development
and chondrocranium complexity at hatching since this implies an increase in yolk
material, frequently related with the duration of the prehatching development
(Araujo-Lima, 1994; Adriaens & Vandewalle, 2003). For example, in Galeichthys
feliceps (egg size: 15.6 mm) (Ariidae) hatchlings already possess complete bony
skulls, whereas in the loricariid A. cf. triradiatus (egg size: 3.1 mm) a highly
developed chondrocranium is present at that point (Tilney & Hecht, 1993;
Geerinckx et al., 2005). In comparison, the egg size of 1.5 mm in C. aeneus is
rather small, indeed resulting in a less developed chondrocranium at hatching
(see 4.1). Next to body size and clutch size, egg size is often suggested to be
related to parental care (Kolm & Ahnesjö, 2005). Such parental care, observed in
many loricariid species (Suzuki et al., 2000), is unreported in Callichthyidae,
which could explain the smaller egg size in members of this group. As a result,
yolk sac amount is also limited, which speeds up the time of hatching during
development. Indeed, C. aeneus hatches three days after fertilization at a SL of
3.5 mm, at which stage the chondrocranium is still primordial, shortening the
period between onset of chondrification (hatching) and yolk sac depletion (4.4
mm) to three days. For comparison, the larger egg size in A. cf. triradiatus
indeed leads to a prolonged prehatching period of five days, yielding a much
more complex neurocranium at hatching. After this, development takes another
five days until the yolk sac becomes depleted, this way prolonging the entire
period of chondrocranial formation up to the level of yolk sac depletion (and thus
functional complexity) up to 10 days. Therefore, although the pathways of
74
4 ONTOGENY – CRANIUM
chondrocranial development seem to be more or less rigid within large taxa, the
time available for this development could very well play an important role in
creating the necessary opportunities for the evolution of innovative structures
from the earliest stages on.
Neurocranium
Skull floor
As in other siluriform fish, the C. aeneus skull is of the platybasic type, often
related to the reduction in eye size, as is typical for this group (Daget, 1964;
Adriaens & Verraes, 1997b).
At hatching, the rostral cartilages of the
chondrocranial floor do not yet expand beyond the polar cartilages, as could be
observed by examination of serial sections revealing that the path of the arteria
carotis interna (commonly accepted to mark the border between the trabeculae
cranii s.s. and the polar cartilages (Goodrich, 1958; Bertmar, 1959; Adriaens &
Verraes, 1997b)) passes in front of them. Examination of sections of later
ontogenetic stages showed the trabeculae cranii s.s. to be present only from 4.0
mm SL on, with the arteria carotis interna running medial to this cartilage
through the hypophyseal fenestra. Finally, from 4.9 mm SL on, these cartilages
expand medially, forming a fissura through which this artery runs. At 5.3 mm SL
this fissura closes medially, incorporating the artery in the trabecular bars. Since,
however, like in A. cf. triradiatus, no evidence of cartilage resorption is seen, the
mechanism suggested for the appearance of this fissure is that of allometric
growth of the trabecular bars (Geerinckx et al., 2005). The notochord never
protrudes into the hypophyseal fenestra in C. aeneus, a situation different from
other catfish like Ariopsis felis and Arius jella (Ariidae), other loricarioids like A.
cf. triradiatus and even other callichthyids like C. callichthys (Bamford, 1948;
Srinivasachar, 1958a; Hoedeman, 1960b; Geerinckx et al., 2005). Further, at
4.4 mm SL, the trabecular bars fuse with the ethmoid cartilage closing the
hypophyseal fenestra. This fusion occurs at the caudal border of the ethmoid
plate, as is generally the case, although this fusion has been reported to occur at
the dorsal face of the ethmoid plate in C. callichthys and H. littorale (Ballantyne,
1930; Hoedeman, 1960b). The ethmoid cartilage then further expands in both a
dorsal and lateral direction, forming the precerebral lamina en solum nasi
respectively. Also at 4.4 mm SL the metotic fenestra is split up into an anterior
75
4 ONTOGENY – CRANIUM
and posterior part, with the first accommodating the passage of the nervus
glossopharyngeus (IX). Since in teleosts no nerves pass through the anterior
basicapsular fenestra, the anterior fenestra found here is homologous to the
basicapsular fenestra and the commissure dividing both the anterior and
posterior fenestra can be designated the commissura basicapsularis posterior
(Daget, 1964). As a consequence, the posterior fenestra can not be homologized
to any of the described fenestrae in teleost chondrocrania and is therefore here
called foramen “A”. At 4.8 mm SL the basicapsular fenestra is again split up, this
time providing passage for the glossopharyngeus (IX) nerve in the posterior part,
making these openings the fenestra basicapsularis anterior and posterior, split up
by the commissura basivestibularis. From a SL of 5.3 mm SL on the foramen “A”
disappears and caudally, the lamina basiotica has further expanded filling up the
neurocranium floor, leaving the fenestra of the vagus (X) nerve.
Skull roof
An actual skull roof in C. aeneus is formed from a SL of 4.4 mm SL on. At this
point, at the rostral skull roof margin, the otic capsules have formed taeniae
marginales which interconnect through the epiphyseal bridge. At its caudal side
both halves of the neurocranium also interconnect, forming the tectum posterius
as a skull roof element. In other catfish, loricarioids and even callichthyids, the
tectum posterius generally closes at a much higher SL. This occurs at a SL of
about 7 mm in C. gariepinus (7.1 mm), A. cf. triradiatus (6.8 mm) and C.
callichthys (6-7 mm), while in H. fossilis (12 mm) and A. jella (29 mm), this
state is first reported at even higher body lengths (Srinivasachar, 1958a, 1959;
Hoedeman, 1960b; Adriaens & Verraes, 1997b; Geerinckx et al., 2005).
Rostrally, the taeniae marginalis anteriores (in front of the epiphyseal bridge) are
significantly reduced in C. aeneus and the bifurcation into an anteriorly directed
commissura
sphenoseptalis
and
a
laterally
directed
commissura
spheno-
ethmoidalis is situated almost at the level of the epiphyseal bridge itself. A
reduction of these anterior parts is indeed common in siluriforms (Adriaens &
Verraes, 1997b; Geerinckx et al., 2005), but such a significant reduction or a
complete absence of these structures is a condition commonly found in
loricarioids like the loricariid A. cf. triradiatus and the callichthyids C. callichthys
and H. littorale (Ballantyne, 1930; Hoedeman, 1960b; Geerinckx et al., 2005).
76
4 ONTOGENY – CRANIUM
Skull wall
As in other catfish, the lamina orbitonasalis is the first preotic vertical
commissure to develop in C. aeneus, separating the fenestra sphenoidea from
what is later to be the foramen fila olfactoria (Geerinckx et al., 2005). Shortly
after this, the preorbital base and commissura sphenoseptalis are formed as
additional
vertical
connections
between
the
taeniae
marginales
and
the
trabeculae cranii. At the caudal margin of the orbital region, a commissure splits
up the fenestra sphenoidea into an anterior and posterior part. Formerly the
absence of a lateral commissure was believed to be a typical siluriform feature,
but, recently, Geerinckx et al. (2005) showed a true lateral commissure to be
present in A. cf. triradiatus. Given this, and given the fact that, as in A. cf.
triradiatus, the nervus oculomotorius (III), trochlearis (IV), trigeminus (V) and
the mandibular branch of the nervus facialis (VII) pass through the anterior part
of the sphenoid fenestra and that the ramus hyomandibularis of the nervus
facialis (VII) passes through the posterior part, I also designated this
commissure to be a true lateral commissure.
Splanchnocranium
In the splanchnocranium, no fusion of the “hyosymplectic-pterygoquadrate
plate” with the interhyal or neurocranium is observed. Such a fusion has been
suggested by Hoedeman (1960b) for C. callichthys and has since been confirmed
by Howes & Teugels (1989) for Corydoras paleatus, at least for the earliest
stages. Observations on cleared and stained specimens and serial sections,
however, clearly indicate that in C. aeneus, the interhyal is formed as a separate
element
from
the
earliest
stages
on
and
that
the
“hyosymplectic-
pterygoquadrate plate” never fuses to the neurocranium. The palatine also arises
as a single element, a state which is indeed typical for siluriform fish (Adriaens &
Verraes, 1997b). Hoedeman (1960b) also mentions the fact that both halves of
Meckel’s cartilage in C. callichthys are “anteriorly connected by ligamentous
tissue, but do not fuse”. In all C. aeneus specimens examined here, both halves
of Meckel’s cartilage remain fused during almost the entire chondrocranial
development, only separating near the start of ossification. In A. cf. triradiatus,
this situation is remarkably different with both halves of Meckel’s cartilage arising
separately, after which they fuse and later separate again at the onset of
ossification in the lower jaw. This increase in lower jaw mobility has been argued
77
4 ONTOGENY – CRANIUM
to be an important structural innovation in the families Astroblepidae and
Loricariidae and thus seems to have its basis early in ontogeny. The situation of
the early Meckel’s cartilage in A. triradiatus also differs structurally from that
found in C. aeneus, since both these cartilages are directed medially in the
former species and rostrally in the latter and in siluriform species in general.
All elements of the splanchnocranium in C. aeneus are present from a SL of 4.4
mm on, which corresponds to the point of yolk sac depletion. Since, at that point
starvation becomes a major threat for larval survival (Bailey & Houde, 1989),
this point is critical during larval development (Pedersen et al., 1990; Jardine &
Litvak, 2003). From this stage on exogenous feeding becomes obligatory and the
presence of a functional feeding apparatus is required (van Snik et al., 1997;
Jardine & Litvak, 2003). It has also been suggested that a fusion between lower
jaw and hyosymplectic-pterygoquadrate plate is crucial in the passive mouth
opening mechanisms acting during early ontogeny of fishes, but again, such a
fusion was not present in any of the stages examined here (Surlemont et al.,
1989; Adriaens et al., 2001).
Branchial arches
The development of the branchial arches in C. aeneus differs from the situation
described in siluriform fish up to this point. Generally, in Siluriformes,
chondrification in each arch starts with the ceratobranchials, followed by the
hypobranchials
and
pharyngobranchials
basibranchials,
(Srinivasachar,
and
1959;
eventually
Adriaens
&
the
epi-
Verraes,
and
1997b;
Vandewalle et al., 1997). In C. aeneus, however, I found that all different
elements arose simultaneously and continuously with each other. Since I did not
notice this initial single formation of the branchial basket in the cleared and
stained specimens, these observations are probably the result of the detailed
observation method used in this study. In this study, serial sections for 3-D
reconstructions of these early branchial baskets were also used, in contrast to
only the cleared in stained specimens used in former studies on branchial arch
ontogeny of fish. However, Adriaens & Verraes (1997b), also mention that the
general
ontogenetic
sequence
in
siluriform
branchial
arches
involves
a
differentiation in an antero-posterior direction. In C. aeneus, differentiation
started with the separation of an infrapharyngobranchial plate, which, based on
its position and articulation with the epibranchials, probably resembles a
78
4 ONTOGENY – CRANIUM
continuous
infrapharyngobranchial
I-IV.
At
the
same
moment
of
this
differentiation, the central copula detaches from all arches and splits up into the
typical anterior and posterior copula. Several authors have argued that the
anterior part is the fusion of the first three basibranchials, with the posterior
copula being a fused basibranchial IV and V (Srinivasachar, 1959; Adriaens &
Verraes, 1997b; Vandewalle et al., 1997). At the same time of the formation of
both copulae, the hypobranchials of arches II and III separate as well. However,
in the specimen of stage 3 (4.4 mm SL) examined here, I found the
basibranchial of arches II an III to be separate, while in the stage 4 specimen
(4.9 mm SL), the basibranchial was still loosely connected to the ceratobranchial
part. At this point, however, the epibranchials of arches II and III have also
separated,
indicating
that
both
arches
differentiate
rapidly
and
almost
simultaneously, prior to arches I and IV. After this, the epibranchial of arch I first
detaches at stage 5 (5.3 mm SL), shortly after which the epibranchial of arch IV
also
separates
from
the
ceratobranchial
part.
At
this
point,
the
infrapharyngobranchial plate has also split up into an anterior and posterior part.
The last element to detach is the first hypobranchial, which separates at around
8 mm SL. The configuration of the gill arch basket at the end of chondrocranial
differentiation does resemble the typical siluriform state. The basket has five
arches with an anterior and posterior copula, separate hypobranchials in the first
three arches and separate epibranchials in the first four (Adriaens & Verraes,
1997b). Adriaens & Verraes (1997b), however, also suggested the presence of a
remnant cartilaginous infrapharyngobranchial I and II, articulating with the
epibranchials of these two arches in adult C. gariepinus. In adult C. aeneus such
a small cartilaginous mass is also observed, and was placed in homology with a
fused infrapharyngobranchial I and II (see 3.1). The ontogenetic sequence
observed here, however, suggests otherwise. In stage 4, an anterior projection
of the cartilaginous infrapharyngobranchial plate is indeed observed, articulating
with the epibranchial parts of these first two arches. In the later stages,
however, this projection degenerates, with even a loss of contact between the
epibranchial of the first arch and the infrapharyngobranchial plate in stage 6.
Therefore, the cartilaginous mass seen in adult C. aeneus can not be homologous
to these infrapharyngobranchials and probably resembles the cartilaginous tips of
epibranchials I and II, which have come in close contact to each other or have
fused.
79
4 ONTOGENY – CRANIUM
Conclusions
In general, the chondrocranium in C. aeneus follows the typical siluriform trends
in chondrocranial development as described by Adriaens & Verraes (1997b). The
skull is of the platybasic type, as described for all Siluriformes and the notochord
does not protrude into the hypophyseal fenestra due to the presence of the
acrochordal cartilage (Adriaens & Verraes, 1997b; Vandewalle et al., 1999). In
contrast to the general situation found in catfishes, but similar to the situation
described for A. cf. triradiatus by Geerinckx et al. (2005) a commissura lateralis
was found. The splanchnocranium again has the typical siluriform shape during
its ontogeny, with the presence of a compound hyosymplectic-pterygoquadrate
plate present, although not fused to the neurocranium or interhyal as described
for C. callichthys by Hoedeman (1960b) and Howes & Teugels (1989). In
general, as hypothesized, development of the chondrocranium was very similar
to that of A. cf. triradiatus, with one of the main differences found in the
development of the lower jaw. This resulted in an increase in jaw mobility in the
loricariid species, used for algae scraping. In comparison to all siluriforms,
including A. cf. triradiatus, however, the most striking differences found involved
the ontogeny of the branchial basket. Whereas, generally, chondrification occurs
in a ventrodorsal and antero-posterior direction, in C. aeneus, the branchial
basket arises as a single element and differentiation first starts dorsally, followed
by arches II and III and then proceeds in both an anterior and posterior
direction. Since I did not notice this initial single formation of the branchial
basket in the cleared and stained specimens, these observations are probably the
result of the detailed observation method used in this study. To find out whether
or not this pattern is indeed more common in siluriforms, further histological
comparative research on branchial basket development is needed.
Discussion of osteocranial development
Ossification sequence in teleosts is related to functional demands arising in
developing organisms (Adriaens & Verraes, 1998). In this context, the moment
of yolk sac depletion is critical, since, from that point on, exogenous feeding
becomes obligatory and starvation becomes a major threat for larval survival
80
4 ONTOGENY – CRANIUM
(Bailey & Houde, 1989; Pedersen et al., 1990; Jardine & Litvak, 2003). After this
transition, larvae are faced with high functional demands on the structures
associated with feeding (van Snik et al., 1997; Jardine & Litvak, 2003). In
addition, in most teleost larvae the common feeding method is suction feeding,
resulting in additional functional demands on the structures necessary to create a
negative buccal cavity pressure (Wagemans & Vandewalle, 2001). Therefore, it is
expected that development in both the chondro- and osteocranium reflects these
demands and that priorities during early cranial ontogeny would be focused on
the
completion
of
a
functional
feeding
apparatus,
prior
to neurocranial
fortification. It has also been noted that the first bones that appear during
teleostean development always seem to be of dermal origin (Wagemans et al.,
1998). Indeed, throughout studies dealing with osteological development in
teleosts, onset of ossification generally involves the maxilla, premaxilla, dentary,
upper and lower pharyngeal jaws, branchiostegal rays and opercular elements,
most of which are of dermal origin and all of which are associated with jaw
movement and the allied act of respiration (Weisel, 1967; Kobayakawa, 1992;
Tilney & Hecht, 1993; Mabee & Trendler, 1996; Adriaens & Verraes, 1998;
Faustino & Power, 2001; Vandewalle et al., 2005; Geerinckx et al., 2007).
As
expected,
in
Corydoras
aeneus
all
elements
of
the
cartilaginous
splanchnocranium are present at the time of yolk sac depletion (4.4 mm SL). At
this point also, the opercle is the first ossified element to appear and teeth are
already present on both lower and upper oral and pharyngeal jaws, although not
yet erupted and not yet supported by ossified elements. Buccal teeth appearing
prior to ossification of their supporting elements is a situation uncommon in
teleost fishes, where teeth usually appear at the same time or after premaxillary
and dentary bones (Vandewalle et al., 2005). In contrast to this, pharyngeal
teeth do generally appear separated from their respective supportive skeletal
elements (Vandewalle et al., 2005), as is also the case here. In general, teeth
appearing prior to the formation of their supporting ossified elements seems
plausible, since it has been argued that bone morphogenesis is possibly
influenced by the presence or absence of teeth (Huysseune, 1989).
Shortly after yolk depletion, at 4.9 mm SL, the appearance of the parurohyal
may indicate a response to mechanical stress resistance due to activity of the
sternohyoideus muscle (Adriaens & Verraes, 1998), which implies an increase in
lower jaw mobility shortly after this dietary shift. After such a functional and
81
4 ONTOGENY – CRANIUM
ossified feeding and breathing apparatus has developed, in between 5 and 8 mm
SL, developmental priorities seem to shift to size increase of the cartilaginous
skull.
After
this,
cartilage
resorption
and
further
ossification
begins,
accomplishing a rigidification of the skull. Such a fortification becomes necessary
since, from the moment of exogenous feeding on, the overlying brain has to be
protected from the particles passing below (Adriaens & Verraes, 1998). Next to
this, it has also been hypothesized that respiratory movements and buccal
pressure would also generate mechanical loadings inducing skull floor ossification
(Mabee
&
Trendler,
1996;
Geerinckx et
al., 2007). This explains why
ossificiations in the skull floor also generally ossify early in the ontogeny of
teleost fish (Vandewalle et al., 1995, 1997; Adriaens & Verraes, 1998). In C.
aeneus, these ossification indeed occur early, however not directly after the shift
from endo- to exogenous feeding, as would be expected. In this case, skull floor
bones like the ventral mesethmoid part, the prevomer and parasphenoid only
start to ossify from 8 mm SL on, well after this shift in feeding style (which
occurs at 4.4 mm SL). At this stage, the prevomer also clearly shows two lateral
cornua or palatal splints. The presence of such structures in adult callichthyids
has been documented in the past for species of Corydoras and Aspidoras (Reis,
1998), but was not observed in my earlier study on the osteology of adult C.
aeneus (see 3.1). Observations now show these splints to fuse to the prevomeral
bone, although additional observations on serial sections show them to be still
present in some adult C. aeneus-specimens, but, in that case, highly diminished
in size. This suggests that the reported absence of such splints in other
callichthyid genera like Callichthys, Dianema, Lepthoplosternum, Hoplosternum
and Megalechis is probably the result of a secondary loss (through fusion or
resorption) during ontogeny. The prevomeral bone in C. aeneus is toothless, in
contrast to the general siluriform state. Also unlike other catfish, the bone arises
as un unpaired element, but since it has been argued that the prevomeral bone
generally consists out of an unpaired dermal toothless bone fused to paired
autogenous toothplates, its seems that only the latter are absent in C. aeneus
development (Adriaens & Verraes, 1998). At 8 mm SL, the neurocranium also
has started to ossify at its dorsal side, since such reinforcements become
necessary due to skull growth (Adriaens & Verraes, 1998). Ossification
simultaneously starts at both the level of the frontal bones as well as the level of
the parieto-supraoccipital bone. The latter bone arises as a single ossification and
82
4 ONTOGENY – CRANIUM
a separate parietal and supraoccipital part could not be discerned at any stages.
Hoedeman
(1960c),
in
his
studies
on
the
callichthyids
Callichthys and
Hoplosternum also did not find ontogenetic evidence for the compound nature of
this bone. However, various authors have observed such a fusion between both
the dermal parietal and perichondral supraoccipital in various Siluriphysi and
have argued this state to be typical for the group (Bamford, 1948; Lundberg,
1975; Fink & Fink, 1996). Therefore, the bone found here was homologized with
this fused parieto-supraoccipital. Another example of such a compound bone
arising at this stage is the posttemporo-pterotico-supracleithrum. Again, during
ontogeny, no signs of fusion of the separate elements constructing this
compound bone were observed, but, nonetheless, the compound nature of this
bone was accepted based on the arguments provided under 3.1.
From 8 mm SL on, ossification in the splanchnocranium drastically increases,
with various centers of ossification in the autopalatine, opercular series, lower
jaw, suspensorial and hyoid arch. Also at this point all articulatory facets have
started ossifying, rigidifying the entire splanchnocranium structure. In the lower
jaw, the mentomeckelian, dentary, articular and retroarticular bones, which will
later make up the compound dentary bone (s.l.), are still present as separate
elements. As in all callichthyids and various other loricarioids like loricariids,
astroblepids and most trichomycterids, no coronomeckelian bone is present (Mo,
1991; de Pinna, 1993; Geerinckx et al., 2007). At this point in the opercular
series, a suprapreopercular bone has also developed. The presence of this bone
was already mentioned in adult C. aeneus specimens as a condition uncommon
for Callichthyidae and it was suggested that further ontogenetic research on the
ontogeny of this bone would have to confirm its hypothesized homology (see also
3.1). Given the fact that the data here shows the bone to develop in close
contact to the preopercular canal and given its position, the small bone shown in
the figures would indeed appear to be a suprapreopercular bone. However, also
as shown in the figures, the bone described as the suprapreoperculare in the
ontogenetic series fuses to the hyomandibula and the preopercular canal does
not protrude into it. Next to this, the bone develops early in ontogeny, while in a
related species like A. triradiatus, it arises much later in ontogeny (Geerinckx et
al., 2007) and does not resemble the bone described as suprapreopercle in the
adult specimens of C. aeneus. In addition, study of additional adult specimens
shows the bone to be variably present in these adults. Therefore it is
83
4 ONTOGENY – CRANIUM
questionable whether or not the bone described in early ontogeny as the
suprapreopercle is homologous to the bone seen in some adult specimens.
Nonetheless, since both bones, when present, are in close contact with the
preopercular canal and are situated above the preopercular bone, the use of
name suprapreopercle seem justifiable in both cases.
In contrast to the 8 mm SL-stage, at around 10 mm SL, ossification appears
more focused on the neurocranium, with additional ossifications appearing in the
skull roof, wall and floor. In the splanchnocranium, however, only the
preopercular and metapterygoid bone appear at this stage. The catfish
metapterygoid has been described as a fusion of ecto-, ento- and metapterygoid
by Howes & Teugels (1989). Other authors, however, reported the ecto- and
entopterygoid to be absent, which I believe to be the case in this species as well
(Regan, 1911; Arratia, 1990, 1992; Adriaens & Verraes, 1998; Reis, 1998). Also
at 10 mm SL, most elements of the branchial basket start to ossify, all of which
start in the center of the respective cartilaginous elements. Shortly after this, at
11 mm SL, ceratobranchials and hypobranchials I-II also start to ossify,
although, in this case, ossification starts at the rostrolateral side of the
cartilaginous hypobranchials. Finally, at 14 mm SL, the basibranchials are the
last bony elements to develop. This way, the entire branchial basket, apart from
the tooth bearing elements, has completely ossified within the short range of 1014 mm SL and together with the hyoid arch, it is the last large functional unit to
complete ossification. Next to this, the interhyal first ossifies at this point. The
late emergene of a bony interhyal indicates that the fact that it is lost in several
siluriform catfishes like Clarias gariepinus, Loricariidae and Scoloplacidae
(Adriaens & Verraes, 1998; Bailey & Baskin, 1976) could be the result of a
truncation in development. This could also explain the variability in the number
of bones in the infraorbital series, found throughout siluriform phylogeny. Only
three additional ossifications occur after this point. At 16 mm SL, the nasal and
first infraorbital bone are the last bones to appear in the development of the C.
aeneus-cranium.
Conclusively, overall ossification sequence in C. aeneus follows the general
trends observed in siluriform development, with ossifications appearing as a
response to functional demands (Tilney & Hecht 1993; Vandewalle et al., 1995,
1997; Adriaens & Verraes 1998). This way, early development of the skull is
84
4 ONTOGENY – CRANIUM
focused on the completion of a functional feeding and respiratory apparatus,
prior to rigidification and growth.
85
4 ONTOGENY – CRANIUM
86
4 ONTOGENY – CRANIAL MYOLOGY
4.4 THE CRANIAL MYOLOGY
Modified from:
Huysentruyt F, Brunain M & Adriaens D
Ontogeny of the cranial musculature in Corydoras aeneus
(Gill, 1858) (Callichthyidae, Siluriformes)
Journal of Fish Biology, submitted.
Abstract
Very few studies have dealt with the ontogeny of the cranial musculature in teleosts in
general and Siluriformes in particular and the ones that have often fail to describe the
earliest stages of ontogeny. Nevertheless, the study of these earliest stages could yield
valuable information on the evolutionary-anatomical response to functional needs placed
on the developing organism; information on the presence of muscular tissue and sites of
origin and insertion would be indicative of functionality. To fill this gap, a complete study
of the early ontogeny of the cranial muscles of Corydoras aeneus (Callichthyidae) was
done. This species belongs to the Loricarioidea, a superfamily in which an evolutionary
trend has been observed that has led to the formation of a suckermouth in two lineages
(Astroblepidae and Loricariidae) and an additional highly specialized feeding mechanism
(i.e. algae-scraping) in the Loricariidae. Results found here were compared to those
found by Geerinckx et al. (in press) for Ancistrus cf. triradiatus, a loricariid
representative. As expected, this comparison revealed a high degree of similarity in the
ontogeny of both species’ cranial muscles. This way, both species lack a m. protractor
hyoidei, and the m. intermandibularis posterior is divided into two different parts which
have partly obtained a novel function in A. cf. triradiatus. A similar increase in complexity
in this species is found in the dorsal constrictor of the hyoid muscle plate. This constrictor
gives rise to the same muscles as in C. aeneus, but, in A. cf. triradiatus, the m. levator
operculi later hypertrophies. In addition, in A. cf. triradiatus, the m. extensor tentaculi
further differentiates into two separate bundles (as opposed to a single muscle diverging
posteriorly in C. aeneus) and a loricariid neoformation is present called the m. levator
tentaculi (Geerinckx et al., in press).
87
4 ONTOGENY – CRANIAL MYOLOGY
Brief introduction
In order to detect functional changes as they occur throughout ontogeny, a
good starting point of study are the higher levels of organization, linked to these
changes (Simonovic et al., 1999). The development of the cartilaginous as well
as the bony skull has been the subject of various studies on siluriforms in general
(Kindred, 1919; Bamford, 1948; Srinivasachar, 1957a, b, 1958a, b, 1959;
Vandewalle et al., 1985, 1997; Tilney & Hecht, 1993; Adriaens & Verraes, 1994,
1997b, 1998; Geerinckx et al., 2005) and callichthyids in particular (Ballantyne,
1930; Hoedeman, 1960a, b, c). However, only in the cases of Clarias gariepinus
and Ancistrus cf. triradiatus has the focus of ontogenetic research been extended
to the level of the development of the muscular system (Surlemont et al., 1989;
Surlemont & Vandewalle, 1991; Adriaens & Verraes, 1996, 1997b, c, d;
Geerinckx & Adriaens, 2007, in press; Geerinckx et al., in press). As far as
teleosts in general are concerned, a similar lack of studies dealing with the issue
in general is found and the ones that have, often fail to describe the very earliest
stages of ontogeny. Nevertheless, the study of this system up to the earliest
stages of development could yield valuable information on the evolutionaryanatomical response to functional needs placed on the developing organism
since, although morphogenesis and differentiation are very intense during early
life stages (van Snik et al., 1997), functional constraints of vital importance act
on each of those stages. Information on the presence of muscular tissue and on
sites of origin and insertion, necessary for performing movements in various
functional units, would therefore be indicative of the functionality of these units
throughout ontogeny.
To fill this gap, the aim of this study is to provide a complete description of the
ontogeny of the cranial muscles of Corydoras aeneus, in continuation of previous
studies published on the species’ overall development and the ontogeny of the
chondrocranium, osteocranium and postcranial skeleton (see 4.3, 4.4, 4.5, 4.7).
Additionally, results found here will also be compared to the results found in A.
cf. triradiatus by Geerinckx et al. (2005).
88
4 ONTOGENY – CRANIAL MYOLOGY
Brief materials and methods
The specimens examined are presented in table 2.1. For clearing and staining
procedures, procedures on serial sectioning and 3D-reconstructions see 2.2.
Terminology on muscles and their development follows Winterbottom (1974) and
Jarvik (1980). For a discussion on homology issues and some aspects of the
innervation pattern, I refer to 3.2.
Results
Stage 1: 3.3 mm SL (hatchling) (fig. 4.22)
At hatching, the chondrocranium of Corydoras aeneus is still very rudimentary
with
the
neurocranium
concentrated
around
the
notochord
tip
and
the
splanchnocranium still confined to the hyosymplectic-pterygoquadrate plate
(sensu Arratia, 1992). At this point, a small sheet of muscular tissue, the
mandibular muscle plate, is present anterior to this early suspensorium. The
sheet is broad dorsally and orientated in a dorsoventral direction with a slight
antero-ventral inclination in the lower half, narrowing toward its ventral tip.
Stage 2: 4.0 mm SL (fig. 4.23)
Shortly after hatching, the neurocranium has expanded rostrally and, at the
level of the splanchnocranium, the suspensorium has differentiated into a
quadrate and hyosymplectic part and the hyoid arch has been formed. The dorsal
part of the mandibular muscle plate has differentiated into an anterior musculus
levator
arcus
palatini,
which
connects
the
anterodorsal
margin
of
the
hyosymplectic to the anterior margin of the otic capsule, and a posterior m.
dilatator operculi. Both of these muscles remain in close contact with each other
and even share some fibers in adult specimens (see 3.2). The m. dilatator
operculi also originates on the otic capsule but its insertion site, the opercle, is
still absent at this point in development. Posterior to the suspensorium, the
constrictor dorsalis and constrictor ventralis of the hyoid muscle plate are
present with the former associated with the otic capsule and lying posteromedial
to the hyosymplectic. The ventral constrictor (or m. interhyoideus posterior, see
89
4 ONTOGENY – CRANIAL MYOLOGY
discussion), on the other hand, is situated just posterior to the early hyoid. At
the posterior margin of the neurocranium the m. protractor pectoralis is present,
originating at the ventroposterior side of the otic capsule.
Stage 3: 4.4 mm SL (fig. 4.24)
At this stage, the neurocranium has further differentiated and its roof has
closed anteriorly at the level of the ethmoid cartilage and posteriorly where the
otic capsules and neurocranial floor have expanded significantly. In the
splanchnocranium, a lower jaw is present and ventral to it, the anterior part of
the mandibular muscle plate has differentiated into a m. intermandibularis
anterior and posterior. The anterior part is situated and inserts on the anterior
margin of the still unpaired cartilaginous lower jaw, while the posterior part
originates on both ceratohyals, but does not yet insert on the lower jaw. In the
larger middle section of the mandibular muscle plate, the dorsal fibers have
differentiated into a distinct m. retractor tentaculi, originating on the anterior
side of the suspensorium and inserting on the primordial ligament, near the
autopalatine (not shown in fig. 4.24). The remaining part of the mandibular
muscle plate forms the still undifferentiated m. adductor mandibulae complex
which also originates on the suspensorium but which, next to an insertion on the
primordial ligament, also inserts on the lower jaw. The constrictor dorsalis of the
hyoid muscle plate has formed a m. adductor arcus palatini, originating on the
trabeculae cranii and inserting on the medial side of the supensorium. More
anteriorly, fibers of this hyoid muscle plate have differentiated further into a m.
extensor tentaculi, which also originates on the trabeculae cranii and inserts on
the posterior margin of the autopalatine. In addition, the constrictor dorsalis of
the hyoid muscle plate has differentiated into two parts, which both originate on
the ventral side of the otic capsule. The anterior part can be discerned as the m.
adductor hyomandibulae, and the posterior part as the remaining constrictor
dorsalis of the hyoid muscle plate. This part, although it already inserts on the
dorsomedial margin of the operculum, is not yet differentiated into a m. adductor
and m. levator operculi. Ventrally, the m. sternohyoideus is now present,
connecting the hyoid arch with the pectoral girdle.
90
4 ONTOGENY – CRANIAL MYOLOGY
Stage 4: 4.9 mm SL (fig. 4.25; 4.26A)
At 4.9 mm SL the chondrocranium is further rigidified anterodorsally through
the taeniae marginales and all parts of the splanchnocranium have become more
differentiated. At the level of the cranial muscles, the m. extensor tentaculi has
enlarged and inserts on the solum nasi. The m. retractor tentaculi connects to
the primordial ligament (not shown in fig. 4.25) and the m. adductor mandibulae
has not yet differentiated into an A2 and A1OST-A3’ part (fig. 4.26A). Further, at
the dorsal margin of the opercle the muscle sheet has become fully differentiated
and a separate m. levator operculi and m. adductor operculi are now present.
Stage 5: 5.3 mm SL (fig. 4.26B; C)
At the level of the jaw musculature, the m. adductor mandibulae has now fully
differentiated into an A2 and A1OST-A3’ part (fig. 4.26B) and, at this stage, the
last cranial muscles to develop are the m. hyohyoideus abductor and adductores
(fig. 4.26C).
Discussion
As mentioned, very few studies have dealt with the ontogeny of the cranial
musculature in teleosts in general and Siluriformes in particular and the ones
that have, often fail to describe the very earliest stages of ontogeny. However, a
detailed overview on the developmental pattern of cranial muscle development in
Amia calva was given by Jarvik (1980), based on work done by Allis (1897) and
Edgeworth (1928). Given the phylogenetic position of the Amiiformes as a sister
group to the Teleostei (Hurley et al., 2007), and given the lack of other studies
dealing with the subject, the pattern described by Jarvik (1980) has often been
regarded as the plesiomorph teleostean pattern. In this pattern, all of the cranial
muscles originate from an ontogenetic primordium of the visceral musculature,
with the exception of the eye musculature and the m. sternohyoideus, which
originate from somatic musculature (Jarvik, 1980; Adriaens & Verraes, 1996). Of
the three parts of visceral muscle primordium (mandibular, hyoid and branchial
muscle plates) only the first two are dealt with in this paper on Corydoras
91
4 ONTOGENY – CRANIAL MYOLOGY
aeneus. In addition, results are compared to those on Ancistrus cf. triradiatus
(Geerinckx & Adriaens, in press; Geerinckx et al., in press).
Mandibular muscle plate
The fact that, at hatching in C. aeneus, the small sheet of muscular tissue
present is situated anterior to the suspensorium, suggests this muscle sheet to
be the early mandibular muscle plate. This is confirmed by the fact that, at 4.9
mm SL, the different muscular units developing from this plate are all innervated
by parts of the mandibular trunk of the trigeminal nerve (V), a discriminative
character of the mandibular muscle plate (Edgeworth, 1935; Jarvik, 1980).
Shortly after hatching, at 4.0 mm SL, a dorsal section splits off, forming the m.
dilatator operculi and m. levator arcus palatini. This is in accordance with the
constrictor dorsalis of the mandibular muscle plate as described by Winterbottom
(1974) and Jarvik (1980). In A. calva, the remaining part of the mandibular plate
further gives rise to the m. palatomandibularis, a compound m. adductor
mandibulae and a ventral m. intermandibularis (Jarvik, 1980). Afterwards, in A.
calva, the m. palatomandibularis forms the m. nasalis, preorbitalis and
parabasalis, all of which are absent in teleosts, with the exception of a preorbital
muscle described in two species of cobitids (Takahasi, 1925; Winterbottom,
1974; Jarvik, 1980). Takahasi (1925) has further proposed that at least part of
this m. palatomandibularis has been incorporated into the A1 part of the adductor
mandibulae
complex
in
teleosts.
Given
this,
I
believe
a
primitive
m.
palatomandibularis to be absent in teleosts and have designated the ventral part
of the muscle plate at this point as constrictor ventralis. At 4.4 mm SL, this
constrictor ventralis has further differentiated into a m. intermandibularis
anterior and posterior (Winterbottom, 1974; Edgeworth, 1935). The latter
muscle is generally believed to have fused to the m. interhyoideus anterior
forming the compound m. protractor hyoidei in teleosts, based on an observed
double innervation of the muscle by both the mandibular trunk of the trigeminal
nerve (V) and the hyoid trunk of the facial nerve (VII) (Winterbottom, 1974).
However, since serial sections revealed this muscle in C. aeneus to be innervated
only by the inferior mandibular nerve branch of the trigeminal nerve (V) and
since no myocomma was present, it, in this case, merely consists of the m.
intermandibularis posterior and thus no m. protractor hyoidei part is present (see
3.2). This is also the case in A. cf. triradiatus and possibly in all Siluriformes
92
4 ONTOGENY – CRANIAL MYOLOGY
(Geerinckx & Adriaens, 2007). The m. intermandibularis posterior of other adult
non-loricarioid siluriforms generally has a secondary subdivision, with some
bundles serving the mandibular barbel base (Diogo & Chardon, 2000). In adult C.
aeneus, such a subdivision is also present, although no insertion on the
mandibular barbels is found (see 3.2). In A. cf. triradiatus, these subdivisions are
also found, but, given the rotation of the lower jaw in A. cf. triradiatus and the
subsequent orientation of these bundles in combination with the absence of
mandibular barbels in loricariids, Geerinckx & Adriaens (2007) have termed them
the labial and dental parts. Nonetheless, is seems plausible that they are
homologous to the subdivisions found in C. aeneus, which have partly obtained a
novel function in the Loricariidae.
At 4.4 mm SL, the remaining part of the adductor division of the mandibular
muscle plate differentiates into a dorsal m. retractor tentaculi, inserting only on
the primordial ligament, and a more ventral compound m. adductor mandibulae.
At 4.9 mm SL, the latter splits up further into a dorsomedial A2 part, which also
inserts on the primordial ligament and a compound A1OST-A3’, diverging
anteriorly into an A1OST bundle inserting on the dorsolateral side of the lower
jaw and an A3’ bundle inserting on the dorsomedial side. The fact that the m.
retractor tentaculi is also derived from the adductor division of the mandibular
plate, adds evidence to the hypothesized homology of this muscle to part of the
A3”, as already suggested by various authors (Lubosch, 1938; Alexander, 1965;
Howes, 1983; Adriaens & Verraes 1996, 1997a; Diogo & Chardon, 2000; Diogo,
2005).
In A. cf. triradiatus, the adductor mandibulae complex consists of the same
bundles, with the addition of a m. retractor veli (Geerinckx et al., subm.). The
bundle homologous to the m. retractor tentaculi, on the other hand, is present
but its insertion point has shifted from the primordial ligament directly onto the
premaxilla. Given this, and since in various studies on loricariids the term m.
retractor tentaculi was already mistakenly used for what Geerinckx et al. (subm.)
called the m. levator tentaculi (Howes, 1983; Schaefer & Lauder, 1986, 1996;
Diogo & Vandewalle, 2003), Geerinckx et al. (subm.) proposed the term m.
retractor premaxillae for this muscle.
93
4 ONTOGENY – CRANIAL MYOLOGY
Hyoid muscle plate
In C. aeneus, the hyoid muscle plate starts to develop shortly after the
appearance of the mandibular muscle plate. In the specimens examined, at its
time of first appearance, the plate is already split up into a constrictor dorsalis
and constrictor ventralis. In A. calva, the constrictor ventralis is described to
differentiate into a m. interhyoideus anterior and m. interhyoideus posterior in
later development (Jarvik, 1980). Since the m. interhyoideus anterior is then
believed
to
further
differentiate
into
part
of the
m.
protractor
hyoidei
(Edgeworth, 1935; Winterbottom, 1974), which is absent in C. aeneus (see 3.2),
the ventral part present here can be homologized with the m. interhyoideus
posterior of A. calva. This is also confirmed by the fact that this muscle plate
further differentiates into the m. hyohyoideus inferior, the m. hyohyoideus
abductor and the mm. hyhyoidei adductores, all of which are believed to
differentiate from the m. interhyoideus posterior (Winterbottom, 1974, Jarvik,
1980). Jarvik (1980) further describes the constrictor dorsalis to differentiate
into a single anterior muscle, the m. adductor hyomandibulae and two posterior
muscles, the m. adductor operculi and m. levator operculi. At 4.4 mm SL,
however, only two muscle bundles were found. Given its position and given the
fact that the anterior part inserts on the medial side of the hyosymplectic, it was
homologized with the m. adductor hyomandibulae. This way, the posterior part
was still homologous to the posterior portion of the constrictor dorsalis, which
had not yet differentiated into a m. adductor operculi and m. levator operculi. In
addition, the anterior part of this muscle plate is described to further
differentiate, giving rise to the m. adductor arcus palatini, which, in turn,
anteriorly gives rise to m. extensor tentaculi (Winterbottom, 1974; Diogo &
Vandewalle, 2003). It has also been argued that the m. adductor arcus palatini
has evolved from a shifted position of the m. adductor hyomandibulae
(Winterbottom, 1974). In several teleosts, however, both a m. adductor arcus
palatini as well as a m. adductor hyomandibulae are frequently found
(Winterbottom, 1974), as is the case here in C. aeneus. Given this, Winterbottom
(1974) has argued that in these cases, the m. adductor hyomandibulae is a
secondary derivative of the anterior fibers of the m. adductor operculi or of the
posterior fibers of the m. adductor arcus palatini. Diogo & Vandewalle (2003), in
their overview on siluriform cranial muscles, accept the former possibility, but
provide no arguments for this choice. The data on C. aeneus now shows that at
94
4 ONTOGENY – CRANIAL MYOLOGY
4.4 mm SL, the dorsal constrictor is divided into a m. extensor tentaculi, a m.
adductor arcus palatini, a m. adductor hyomandibulae and a dorsal muscle
associated with the opercle, which later differentiates into a m. levator operculi
and m. adductor operculi part. Therefore, I reject the possibility of the m.
adductor hyomandibulae being a secondary derivate of the m. adductor operculi
in this case, since the former differentiates earlier in ontogeny than the latter.
The other option, where the m. adductor hyomandibulae is derived from the
posterior part of the m. adductor arcus palatini would therefore seem more
plausible here.
In comparison to A. cf. triradiatus, a similar pattern of development in the
hyoid muscle plate is found, again with an increase in complexity, as is the case
in the mandibular muscle plate. This way, the dorsal constrictor gives rise to the
same muscles as in C. aeneus. The m. levator operculi, in A. cf. triradiatus,
however, later hypertrophies, resulting in the formation of large myodomes in
the neurcranial roof (Geerinckx & Adriaens, in press). Additionally, in A. cf.
triradiatus and some other loricariids, the m. extensor tentaculi is differentiated
into two separate bundles (as opposed to a single muscle diverging posteriorly in
C. aeneus) and a loricariid neoformation is present called the m. levator tentaculi
(Diogo, 2005; Geerinckx et al., in press). This fits the evolutionary trend of an
increased differentiation of the constrictor dorsalis of the hyoid muscle plate in
teleosts, in which anterior fibers of the m. adductor arcus palatini give rise to
new muscles in various groups. While, in this case, the m. extensor tentaculi
arises, later splitting up into two separate bundles with a subsequent
differentiation of the m. levator tentaculi in Loricariidae, other examples exist in
teleosts. These examples comprise the formation of a m. retractor arcus palatini
in Acanthuridae, Balistidae and Ostraciidae and of a m. retractor palatini in
Balistes, both of which are also anterior derivations of the m. adductor arcus
palatini.
The ontogeny of the m. adductor hyomandibulae, however, differs substantially
from the situation found in C. aeneus. In A. cf. triradiatus, this muscle appears in
a later stage of ontogeny than the m. adductor arcus palatini and m. adductor
operculi. In addition, it appears in close contact with the latter muscle (Geerinckx
et al., in press), which, in this case, would confirm the hypothesis of
Winterbottom (1974) and Diogo & Vandewalle (2003), where it is a secondary
derivative of the m. adductor operculi. This would, however, imply a non-
95
4 ONTOGENY – CRANIAL MYOLOGY
homology with the m. adductor hyomandibulae, as described in more primitive
species like A. calva. Given the results found in both C. aeneus and A. cf.
triradiatus, and given the fact that, in more primitive lineages like A. calva, a m.
adductor hyomandibulae and no m. adductor operculi are present, it would seem
that indeed multiple pathways toward the formation of a m. adductor
hyomandibulae exist in teleosts. However, given the function of this muscle and
its consistent origin on the otic region of the neurocranium and insertion on the
dorsomedial side of the suspensorium, the use of the name m. adductor
hyomandibulae seems highly justifiable throughout the different actinopterygian
groups, as long as a true homology is unknown.
96
4 ONTOGENY – POSTCRANIAL SKELETON
4.5 THE POSTCRANIAL SKELETON
Abstract
Most ontogenetic research on teleosts focuses on the cranium. A detailed study of the
ontogeny of the postcranial skeleton would, however, help understanding function in the
developing larva since development of the body axis provides firmness, influenced by
increasing forces acting on it during ontogeny, and since fin development also relates to
changes in swimming mode, swimming velocity and feeding techniques and preferences
(Fukuhara, 1992; Koumoundouros et al., 1999; Koumoundouros et al., 2001b).
Therefore, such a study was performed on Corydoras aeneus, and results were compared
to some of the data presented in 4.2. Indeed results confirmed that swimming mode in
this species changed at two points during ontogeny. Observations on the ontogeny of the
caudal fin skeleton showed that all cartilaginous elements in this fin had developed
between a SL of ca. 5–6 mm, at which point also notochord flexion occurred. This
corresponds to the suggested change to burst swimming (rapid generation of trust
propelling the animal away) at 5 mm SL (Weihs, 1980; Verhagen, 2004)(see also 4.2).
In the analysis of early development a second shift in swimming mode from anguilliform
(with trust generated by the entire body resulting in high undulatory amplitudes along
the body surface) to subcarangiform swimming (with trust being generated mostly
posteriorly in the body and low amplitudes of anterior body undulations) was believed to
be present at around 8 mm SL (see 4.2). Indeed, at 8 mm SL, the anal, adipose and
pelvic fins were found to separate from the larval finfold and ossification in the vertebral
column and caudal skeleton had started. Also, development of the anal and pelvic fin had
started and ossification of the scapulocoracoid plate and development of the proximal
radials and fin spine in the pectoral fin also took place.
Brief introduction
In C. aeneus, general early development as well as the ontogeny of the
chondrocranium and osteocranium (see 4.2; 4.3; 4.4) has been studied, but no
data
exist
on
postcranial
ontogeny.
Such
data
would,
however,
help
understanding function in the developing larva since development of the body
axis provides firmness, influenced by increasing forces acting on it during
97
4 ONTOGENY – POSTCRANIAL SKELETON
ontogeny, and since fin development also relates to changes in swimming mode,
swimming velocity and feeding techniques and preferences (Fukuhara, 1992;
Koumoundouros et al., 1999; Koumoundouros et al., 2001b). In addition,
ontogenetic data could also help solve some homology issues. This way, e.g. in
the caudal skeleton of adult specimens, the dorsal plate is believed to be a fusion
of the urostyl and the dorsal hypurals III, IV and V while the ventral plate would
comprise the parhypural and hypurals I and II (Lundberg & Baskin, 1969) (see
also 3.1). In the case of the Callichthyidae, however, little ontogenetic data
support this. Therefore, the objective of this study is to describe the normal
pattern of postcranial development in C. aeneus and link it to possible changes in
function of the locomotory and feeding apparatus as well as to determine the
homology of the different skeletal elements involved. In this study, the
development of the Weberian apparatus was not included, since a thorough
description of this was already given by Coburn & Grubach (1998) for C.
paleatus.
Brief materials and methods
The specimens examined are presented in table 2.1. For clearing and staining
procedures, see 2.2. Bone terminology used throughout this paper follows Arratia
(2003b), Lundberg & Baskin (1969) and Grandel & Schulte-Merker (1998).
Results
Axial skeleton
The vertebral column (fig. 4.27)
Development in the vertebral column of Corydoras aeneus takes place over a
narrow range in standard length. The first signs of vertebral development were
present from a SL of 5.2 mm on, but until 6.2 mm SL, the vertebral column
shows no signs of ossification. Shortly after, at 7.2-7.7 mm SL (see also fig.
4.28F), the first ossifications appear and by 8.2 mm SL, all vertebrae have fully
98
4 ONTOGENY – POSTCRANIAL SKELETON
developed and all neural and haemal processes have ossified. At this point, the
parapophyses of the sixth vertebral centre already support a large hollow rib,
which will later contact the first ventrolateral bony scutes behind the pectoral
girdle. Further, the remaining precaudal vertebrae (7-12) each bear a thin rib at
this point (not shown in figure 4.27). Behind these, the haemal spines of
vertebrae 13-17 are gradually longer in a posterior direction, making space for
the abdominal cavity. In addition to these five caudal vertebrae, nine more are
present and an additional one is incorporated in the caudal skeleton, which totals
the number of vertebrae to 28 (see also 3.1).
The caudal skeleton (figs. 4.27; 4.28)
The caudal skeleton is of the pleurostyl type and consists of two bony plates.
The ontogenetic series studied here revealed the fusion of the urostyl to the
dorsal hypurals, as well as the development of a plate-like lamina on the epural
(the neural spine of the first preural centre). An actual fusion of the hypurals,
however, was not detected. At 4.8 mm SL a ventral hypural cartilage is present,
soon after which, at 5.2 mm SL, the dorsal hypural cartilage appears. At 5.4 mm
SL, notochord flexion starts and the first five principal fin rays appear. Of these,
four articulate with the dorsal and one with the ventral hypural plate. At 5.7 mm
SL, the cartilaginous haemal arch of the second preural centre is present and
three additional principal fin rays appear that articulate with the ventral hypural
plate. In addition, ventrally, the parhypural cartilage also appears and is already
fused to the ventral hypural plate. At 6.2 mm SL, six principal rays and one
procurrent ray are present ventrally and the epural cartilage has appeared
dorsally. Shortly after this, at 7.2 mm SL, all principal rays are present together
with two procurrent rays on both the dorsal and ventral side. At this stage,
ossification of the vertebrae has started, clearly showing ossification centres on
the anterior and posterior margin of the second preural vertebra and in its neural
and haemal spines. The first preural vertebral centre, which will later fuse to the
tail skeleton, only shows ossification at its anterior margin. After this, at 7.7 mm
SL, the haemal spine of the second preural centre also ossifies and at 8.2 mm SL
almost all parts of the caudal skeleton become completely ossified. After this
stage, the epural further ossifies and both hypural plates partially fuse followed
by the second preural haemal arch, which is the last part to ossify.
99
4 ONTOGENY – POSTCRANIAL SKELETON
The dorsal fin (figs. 4.27; 4.29)
Ossification in the dorsal fin starts around 7.7 mm SL in C. aeneus, with the
ossification of the first two fin rays. Before this, from a SL of 5.7 mm on, eight
cartilaginous pterygiophores are present supporting seven actinotrichia. At 8.2
mm SL, an additional double actinotrichium appears at the level of the 8th
pterygiophore. During ontogeny, it is clear that both the second fin spine and the
first fin ray appear in relation to the second pterygiophore. Finally, in adult
specimens of C. aeneus, the second dorsal fin spine will also develop a lateral
process through which it articulates with the transverse process on the second
pterygiophore. Further in ontogeny, pterygiophores II-VI all develop a proximal
and distal part. In the second pterygiophore, articulation with the second fin
spine occurs at the level of the proximal part, while the first fin ray articulates
with the cartilaginous distal part. In all other fin rays, articulation occurs only
through the distal part. In addition, the first pterygiophore also fuses to the
nuchal plate forming a large rigid bone proceeding the dorsal fin. Therefore, this
bone was not recognized as a pterygiophore and the transverse process was
mistakenly reported to be present on the first pterygiophore in my description of
the species’ adult osteology (see 3.1).
The anal fin (figs. 4.27; 4.30)
At 7.7 mm SL, the first three proximal pterygiophores of the anal fin become
apparent, shortly after which, at 8.2 mm SL, all six pterygiophores are present.
Ossification in the fin rays starts around 11 mm SL, at which time the onset of
ossification in the first pterygiophore can also be observed and the last, double
fin ray first appears. The first two fin rays both develop in relation to the first
pterygiophore and the distal parts of all pterygiophores remain unossified.
Appendicular skeleton
The pectoral girdle (fig. 4.31)
At 6.2 mm SL, a thin cleithral bone is present, caudally bordered by an
endoskeletal cartilaginous disc which articulates with five actinotrichia. At this
point, the anterior dorsoventrally orientated cartilaginous rod, corresponding to
the scapulocoracoid cartilage, is still connected to this rostro-caudally orientated
endoskeletal disc. At 6.6 mm SL, the fin plate splits into three lobes, and three
cartilaginous distal radials individualize at this point. The upper lobe disappears
100
4 ONTOGENY – POSTCRANIAL SKELETON
during further ontogeny (or fuses with the scapulo-coracoid cartilage), while the
lower two ossify as proximal radials. From 8 mm SL on, the first fin ray, later
forming the pectoral fin spine, starts to ossify, while the remaining fin rays ossify
much later during ontogeny (after 12.5 mm SL). Also around 8 mm SL, the
dorsal part of the pectoral girdle, the posttemporo-pterotico-supracleithrum, is
incorporated into the neurocranium and also starts to ossify (see 4.4; fig.
4.15A). At 9.7 mm SL, the scapulocoracoid cartilage ossifies, with both left and
right bones extending much further ventrally than the cartilage itself and
contacting each other at the midline. After this stage, all bones further ossify,
and at 12.5 mm SL, both scapulocoracoid and cleithrum have developed a
median process, contacting each other. At this stage, the first serrations on the
pectoral fin spine appear and eight additional ossified fin rays are present.
Pelvic girdle (fig. 4.32)
The first signs of a pelvic girdle are observed in specimens with a SL of around
8.2 mm, in which two cartilaginous basipterygia articulate with six actinotrichia,
already approaching the definite number of pelvic fin rays. These rays become
ossified at 9.7 mm SL. The cartilaginous basipterygium further develops in an
anterior direction, forming the cartilaginous precursor anterior external and
internal processes, which later fuse. Around 16 mm SL, ossification starts,
through which a lateral process and dorsal and ventral ischiac processes develop.
Discussion
Data on postcranial siluriform morphology is mostly dedicated to the Weberian
apparatus (for a review, see Chardon et al., 2002), although some recent data is
present on the pectoral girdle (Diogo et al., 2001) and caudal skeleton (Lundberg
& Baskin, 1969; Arratia, 1983; Lakshmi & Srinivasa Rao, 1989; Fujita, 1992;
Arratia, 2003b; de Pinna & Ng, 2004; De Schepper et al., 2004). In the specific
case of loricarioids, a detailed description on the pectoral girdle of Nematogenys
inermis has been provided by Diogo et al. (2006) and the postcranial osteology
of Corydoras aeneus is dealt with under 3.1. In addition, some literature on the
ontogeny and ossification of loricarioids and even callichthyids exists, with a
study on the Weberian apparatus in Corydoras paleatus by Coburn & Grubach
101
4 ONTOGENY – POSTCRANIAL SKELETON
(1998), and a description of the postcranial ontogeny in Trichomycterus
areolatus and Diplomystes nahuelbutaensis (Arratia, 2003b) and Callichthys
callichthys and Megalechis thoracata (Hoedeman, 1960d).
Ossification of the vertebral column in C. aeneus occurs much faster than has
been described for some marine teleosts like Hippoglossus hippoglossus (Lewis &
Lall, 2006), Diplodus sargus (Koumoundouros et al., 2001b), Anarhichas lupus
(Pavlov & Moksness, 1997), Dentex dentex (Koumoundouros et al., 1999) and
Pagellus erythrinus (Sfakianakis et al., 2004). In the freshwater siluriform
Ictalurus punctatus (Arratia, 2003b), this development also occurs slower than in
C. aeneus but Arratia (2003b) does mention a higher speed of development in
the more closely related Trichomycterus areolatus.
While the primitive teleost condition is to have two separate ural centra
posterior to the centrum that bears the last haemal arch (i.e. first preural
centrum), in catfishes and many other advanced groups of fishes it is generally
believed that both the first and second ural centra are co-ossified with the first
preural centrum (Lundberg & Baskin, 1969). Next to the evidence found in
primitive teleostean lineages, this hypothesis is partially confirmed for siluriforms
by the discovery of a separate second ural centrum in certain siluriform groups
(Lundberg & Baskin, 1969; De Pinna & Ng, 2004). However, throughout the
ontogeny of C. aeneus, in none of the stages examined, separate ural centra
were found. This may indeed imply that fusion of these elements already occurs
at the very early stages of development, as already suggested by Lundberg &
Baskin (1969). In the same study, Lundberg & Baskin (1969) described the
dorsal plate of C. aeneus as a fusion of the urostyl and hypurals III, IV and V and
the ventral plate as the fusion of parhypural and hypurals I and II. Although the
data show no separate hypurals at any point during ontogeny, the fusions
between the cartilaginous parhypural and the ventral plate, and between the
ossifying urostyl and dorsal plate were indeed clearly observed. The fact that not
all hypurals develop as individual elements is a state which has only occasionaly
been reported in teleosts (e.g. Pavlov & Moksness, 1997). In catfishes, however,
not much data is present on the ontogeny of the caudal fin skeleton. Fujita
(1992), however, in his paper on caudal fin ontogeny in Clarias batrachus also
reports a first hypural that is fused to the second hypural from the moment it
appears. Given the high prevalence of hypural fusions in catfish (Lundberg &
102
4 ONTOGENY – POSTCRANIAL SKELETON
Baskin, 1969), this could indicate a different ontogenetic pattern in Siluriformes,
but additional data is required to confirm this.
Literature on the ontogeny of the pectoral girdle in Siluriformes is limited to a
single publication by Arratia (2003b) based on an ontogenetic series of Ictalurus
punctatus; the results found are highly comparable to those of C. aeneus. Both
species exhibit a chondral endoskeletal disc that splits up distally, with two
proximal radials ossifying from this during later ontogeny.
Again, in the case of the pelvic girdle, literature is limited to the studies done
by Arratia (2003b) on Ictalurus punctatus, Diplomystes nahuelbutaensis and
Trichomycterus areolatus. These data clearly show that the presence of the
external and internal anterior process is determined early in ontogeny and visible
on the cartilaginous basipterygium (Arratia, 2003b). This implies that the two
cartilaginous processes visible in the early ontogeny of C. aeneus represent both
the internal and external anterior processes, which fuse during later ontogeny.
As a result, the external anterior process, as described under 3.1, actually
represents a lateral process. The homology of these anterior processes has
already been questioned by Shelden (1937), and as a result both were referred
to as ‘projections’. However, no obvious motivation was given to support this
idea, and therefore both were considered as the internal and external process
under 3.1. Evidence now thus shows the external process to be a true lateral
process, which might very well also be the case for Callichthys callichthys as
described by Reis (1998). Posteriorly, in callichthyids, two bony ischiac processes
are present which also show no chondral precursor in early ontogenetic stages.
According to Arratia (2003b), these processes, together with the lateral
processes, are therefore not homologous to the anterior and ischiac processes,
and should be considered as mere projections of the bony surface of the
basipterygial plate. The fact that the situation found in the callichthyid pelvic
girdle differs from that found in other siluriforms has already been argued to
relate to reproductive strategies (Reis, 1998), since during spawning female
callichthyids carry their eggs in a ‘shell-shaped’ pouch made by the pelvic fins
(Kohda et al., 1995).
The study of the changes in growth and allometries during ontogeny (see 4.2)
seemed to confirm that around ca. 5 mm SL a shift in swimming mode toward a
burst and coast swimming technique was present. Both Weihs (1980) and
Verhagen (2004) indeed found that from this SL on, burst swimming became the
103
4 ONTOGENY – POSTCRANIAL SKELETON
most effective swimming mode in fish larvae. Given that propulsive forces in this
swimming mode are obtained by tail beats, development in the caudal fin was
expected to be at a peak at this time. At this point in development (up to 6.2
mm SL), growth in the caudal fin indeed was at a maximum reaching a growth
coefficient of 4.72 (see 4.2; fig. 4.6D). Observations on the ontogeny of the
caudal fin skeleton now also show that all cartilaginous elements in this fin
develop between a SL of ca. 5–6 mm, at which point also notochord flexion
occurs. After this, between 6.2 and 7.2 mm SL, no major changes in the
development of the caudal fin skeleton occurred. In the analysis of early
development
a
second
shift
in
swimming
mode
from
anguilliform
to
subcarangiform swimming was believed to be present at around 8 mm SL.
Indeed, at this point, anal, adipose and pelvic fins were found to separate from
the larval finfold and abdominal and postanal growth increased at around this SL
(see 4.2). The results here again confirm this, with ossification starting in the
vertebral column and caudal skeleton in between 7 and 8 mm SL. In addition,
development of the anal and pelvic fin starts around 8 mm SL and ossification of
the scapulocoracoid plate and development of the proximal radials and fin spine
in the pectoral fin also concentrate around 8–10 mm SL. In conclusion, all
elements presented here seem to confirm that during ontogeny in C. aeneus, two
separate shifts in swimming style occur, making the presence of a functional
locomotory apparatus crucial at those points in development. As suggested by
Weihs (1980) and Verhagen (2004), these shifts are caused by the differences in
the physical properties of a longer larval body in an aqueous environment and
therefore
relate
to
developmental state.
104
the
absolute
length
of
the
larvae
rather
than
its
4 ONTOGENY – MINIATURIZATION
4.6 ONTOGENY OF CORYDORAS PYGMAEUS,
A MINIATURIZED CONGENERIC
Abstract
The ontogenetic pattern of a miniaturized Corydoras species, Corydoras pygmaeus, was
studied
and
results
on
growth,
external
morphology
and
development
of
the
chondrocranium, osteocranium and postcranial skeleton were compared to comparable
data on C. aeneus, a non-miniaturized Corydoras species. Given the low number of
miniaturized Corydoras species, miniaturization was believed to have been secondary
derived in this genus through the occurrence of paedomorphic events. Results showed
high similarities in the overall pattern of development in both species with some
difference in the timing of onset of ossification of different elements. In conclusion,
miniaturization in the body of C. pygmaeus was indeed found to be the result of
paedomorphic events, including a lower growth rate and subsequent postdisplacement of
several traits. These paedomorphic events have, however, besides the smaller size, not
led to the retention of a large number of paedomorphic traits.
Introduction
The highly diverse catfish genus Corydoras comprises 152 species with species
ranging in maximal standard length (SL) between 20 mm (Corydoras habrosus)
and 88 mm (C. britskii) (Reis, 2003; Froese & Pauly, 2006; Ferraris, 2007). The
smallest species, C. habrosus, C. boehlkei, C. gracilis, C. hastatus and C.
pygmaeus even qualify as miniaturized, since in freshwater teleostean fishes, a
maximal SL of 25-26 mm is commonly set as the threshold for miniature species
(Hanken & Wake, 1993; Weitzman & Vari, 1988). In studies by Reis (1998) and
Britto (2003), morphological data provided a phylogenetic framework for both
the family (Callichthyidae), and subfamily (Corydoradinae) to which this genus
belongs and in a study by Shimabukuro-Dias et al. (2004) molecular data
provided this for the Callichthyidae. The phylogeny of the genus itself, however,
remains largely unresolved. This way, the dwarf species C. hastatus and C.
105
4 ONTOGENY – MINIATURIZATION
pygmaeus
were
first
described
as
belonging
to
the
separate
genus
Microcorydoras by Myers (1953), which was later put into synonymy with
Corydoras. In 1970, Nijssen also suggested a close relationship between C.
hastatus and C. pygmaeus based on their minute size and suggested the
resurrection of the Microcorydoras genus. In their revision of the genus in 1980
however, Nijssen & Isbrücker rejected this hypothesis and put both these dwarf
species together with C. gracilis and other non-pygmy species into the “elegans”group of the Corydoras-genus. In this revision, the other dwarf species, C.
habrosus, was even placed into an entirely different group, the “aeneus”-group.
In 1985, Strauss performed a morphometric analysis of some representatives of
the genus and he also found that there seemed to be: “no justification on
morphometric grounds for isolating the dwarf taxa (C. hastatus, C. australe and
C. pygmaeus) into a separate genus”. He found that C. hastatus and C.
pygmaeus lie well within the normal patterns of variation for the genus but did
mention that C. australe differed considerably from all other taxa in sizeindependent form (Strauss, 1985). The latter statement thus questioned the
earlier hypothesis of Nijssen & Isbrücker (1980), in which they synonymized C.
australe with C. hastatus, a synonymy which since has been generally validated
(Eschmeyer, 1998). However, Schaefer et al. (1989) again proposed to put the
two dwarf species C. hastatus and C. pygmaeus into one monophyletic group
based on their minute size, color pattern and reduced dorsal fin ray number
(six), but they were unable to retain any additional synapomorphies (besides
small size) uniting all miniature Corydoras species. This led them to conclude
that they were: “unable to reject the hypothesis that miniaturization arose more
than once in the genus” (Schaefer et al., 1989). Finally, in the most recent study
done on this group which included dwarf species, Britto (2003) placed dwarf
species like C. habrosus and C. gracilis throughout the family’s phylogeny. They
did, however, find four synapomorphies to place both C. pygmaeus and C.
hastatus into one monophyletic group. Of these four characters, however, two
characters are based on pigmentation and one on ethology. This leaves the
reduction of the suture between metapterygoid and hyomandibula as a single
morphological synapomorphy to group both dwarf species. This confirms the
hypothesis of Schaefer et al. (1989) that C. pygmaeus and C. hastatus are
indeed sister species, but it also confirms their assumption that miniaturization
arose more than once in the genus. In this context, heterochronic events are
106
4 ONTOGENY – MINIATURIZATION
often suggested as key factors in the evolution of miniaturized species. Also, of
the few synapomorphies indicating a sister-species relationship between C.
pygmaeus and C. hastatus (next to body size itself) the two main characters
(reduced dorsal fin ray number (Schaefer et al., 1989) and reduction of the
suture between metapterygoid and hyomandibula (Britto, 2003)) are both
reductive characters, which also indicates the importance of heterochronic events
in this miniaturization process.
Given this and given the fact that miniaturization has often been postulated as
a key feature in the evolution of high level taxa (Hanken, 1993; Hanken & Wake,
1993), the question whether miniaturization within the Callichthyidae is also the
result of heterochronic events and the extent of their role in callichthyid evolution
becomes apparent.
The adult morphology of the family has been studied thoroughly in a
phylogenetic context (Reis, 1998) and an extensive study on the adult osteology
and myology of C. aeneus has been done recently (see chapters 3.1, 3.2), but
very little is known about the ontogeny of the genus’ representatives. In
chapters 4.2, 4.3, 4.4, 4.5 and 4.6, the ontogeny of the cranial muscles and
skeleton of C. aeneus is dealt with in detail, but ontogenetic data on a
miniaturized congeneric is still lacking. Therefore, the aim of this study is to
describe the early ontogenetic pattern of a miniaturized Corydoras species, C.
pygmaeus, and compare this pattern to the pattern found in C. aeneus. Doing
this, I hypothesize that the miniaturized body in C. pygmaeus is the result of
paedomorphic processes. In addition, a comparison of developmental sequences
of related taxa within a phylogenetic framework, would serve the purpose of
identifying such possible heterochronic events (Mabee & Trendler, 1996).
Although the phylogenetic framework of the Corydoras genus, as mentioned, is
lacking, it seems plausible to assume that the non-miniaturized form is the
plesiomorphic condition of the genus. This assumption is based on the fact that
only five of the 152 species of Corydoras and only one of the 24 species in the
closely related genera Aspidoras and Scleromystax are miniaturized (Reis, 2003;
Ferraris, 2007). In a first stage of this research, the focus was limited to an
analysis of overall growth, external morphology and of the skeletal cranial and
postcranial elements. This was done through the study of an ontogenetic series
of cleared and stained specimens of C. pygmaeus.
107
4 ONTOGENY – MINIATURIZATION
Brief materials and methods
The specimens examined are presented in table 2.2. For clearing and staining
procedures, see 2.2. Bone terminology used throughout this paper follows Arratia
2003b), Lundberg & Baskin (1969) and Grandel & Schulte-Merker (1998). For
details on the analysis of growth, I refer to 4.2.
Results
Growth
In the growth curve of Corydoras pygmaeus, ranging from 0 until 50 dph, only
a single inflexion point was calculated where growth rate (slope) changed (fig.
4.33A). This inflexion is located at 5 dph (which corresponds to a fitted value of
5.17 mm SL), at which the estimated slope increases from 0.12 to 0.37
(p<0.001).
External morphology
At hatching, C. pygmaeus larvae have an average SL of 3.7 mm and a yolk sac
remained present up to 4.4 mm SL. During this eleutherembryonic phase (fig.
4.34A) (for the terminology of the different phases I refer to Balon (1975); see
also 4.2), pigmentation was limited to small patches in the dorsal finfold and on
the dorsal and lateral sides of the head. At 4.4 mm SL, the anus, mouth and
opercular cavity open and the protopterygiolarval phase starts (fig. 4.34B). At
this point, body pigmentation in the dorsal finfold has expanded, while
pigmentation on the head is present as a double line running from the buccal
area, over the eye, to the back of the head. The median finfold further remains
undifferentiated until 6.2 mm SL, at which time the pterygiolarval phase starts
(fig. 4.34C). This pterygiolarval phase is characterized by a substantial expansion
of the body pigmentation and the onset of fin ray formation in all fins. At 8.9 mm
SL all fins have fully detached from the finfold and the juvenile period starts (fig.
4.34D). At this point, pigmentation on the body has already began to
concentrate into a median lateral line expanding into the basis of the caudal fin,
as is the case in adult C. pygmaeus.
108
4 ONTOGENY – MINIATURIZATION
Skeleton
Stage 1: 3.7 mm SL (hatching) (Fig. 4.35A; B; 4.40A)
At hatching in C. pygmaeus, the polar and otic cartilages are already
interconnected through the lateral commissure and the basiotic laminae contact
the otic cartilages through the anterior basicapsular commissure. At the caudal
edge of the neurocranium floor, the parachordal cartilages connect to the otic
cartilage through the occipital pilae, leaving two large fenestrae metotica. Both
trabecular bars are present, running forward, merging at the level of the ethmoid
cartilage, bordering the fenestra hypophysea. Dorsally, both taeniae marginales
posteriores are present and interconnect through the epiphyseal bridge, but do
not yet contact the ethmoid cartilage. Dorsocaudally, both sides of the
neurocranium are extended mediocaudally, forming the onset of the tectum
posterius.
Most parts of the cartilaginous splanchnocranium have been formed by the time
of hatching. In the hyoid arch, the hypohyal, ceratohyal and interhyal parts are
present
and
in
the
hyosymplectic-pterygoquadrate
plate,
distinguishable
hyosymplectic and quadrate parts are present. In the lower jaw, Meckel’s
cartilage is present, rostrally bearing a small layer of teeth-bearing dentary
bone. In the premandibular arch, the autopalatines and both teeth-bearing
premaxillary bones are visible.
A ventral and dorsal hypural are present in the caudal skeleton, but no
notochord flexion was observed yet.
Stage 2: 4.3 mm SL (Fig. 4.35C)
The neurocranium floor has further expanded splitting up the metotic fenestra.
Given the situation described under 4.2, in C. pygmaeus, the anterior fenestra
was also designated as the fenestra basicapsularis anterior, while the posterior
aperture was called foramen “A”. This is corroborated by the fact that, at the
caudal edge of the neurocranium, as in C. aeneus, two foramina for the vagus
(X) nerve are present.
Stage 3: 4.8 mm SL (Fig. 4.36)
On the dorsocaudal side of the neurocranium, the otic capsules have expanded
medially and the tectum posterius has closed dorsocaudally. Rostrally, the taenia
marginalis contact the solum nasi and ethmoid plate through the commissura
spheno-ethmoidalis and commissura sphenoseptalis respectively.
109
4 ONTOGENY – MINIATURIZATION
Stage 4: 5.6 mm SL (fig. 4.40B)
In the skeleton of the caudal fin, both the dorsal and ventral hypural plates
have expanded substantially.
Stage 5: 6.6 mm SL (Fig. 4.37; 4.40C; 4.42A)
In specimens with a SL of 6 to 7 mm, the otic capsules and neurocranium floor
have both expanded. At the rostral end of the neurocranium the ethmoid region
has broadened by lateral expansions of both solum nasi.
The suspensorium has become more rigid and the hyosymplectic has developed
a distinct articulatory facet for future articulation with the opercle. A pterygoid
process is present at the ventrorostral margin of the suspensorium and the
parurohyal bone has appeared in the hyoid arch.
In the postcranial skeleton, notochord flexion has now occurred and vertebral
centra start to develop. In the caudal fin, a cartilaginous haemal spine has been
formed on the second preural vertebral centre. In the cartilaginous pectoral
girdle, a dorsal scapular and ventral coracoid process are present and matrix
decomposition has started in the endoskeletal disc forming the first distal radials.
In addition, in the pelvic girdle, two small cartilaginous basipterygia have been
formed.
Stage 6: 7.5 mm SL (fig. 4.40D; 4.41A)
From this stage on, the epural cartilage is present in the caudal skeleton. In the
anal fin, fin rays have appeared, supported by pterygiophores that are not yet
split up into proximal and distal radials. Cartilaginous pterygiophores have also
appeared in the dorsal fin and are already split up into proximal and distal
radials.
Stage 7: 7.9 mm SL (fig. 4.38A)
In the splanchnocranium, additional to the maxillary, premaxillary, opercular
and parurohyal bone, ossifications have appeared in the autopalatine and
anterior ceratohyal.
Stage 8: 8.5 mm SL (fig. 4.40E)
No additional ossifications are present in the neuro- and splanchnocranium and
in the caudal fin, the first signs of early ossification (not yet stained) are seen.
110
4 ONTOGENY – MINIATURIZATION
Stage 9: 9-9.5 mm SL (fig. 4.38B; 4.40F; 4.41B; 4.42B)
At this stage, ossification has started throughout the entire neurocranium. In
the skull roof, the mesethmoid, frontal, sphenotic, parieto-supraoccipital and
posttemporo-pterotico-supracleithrum have formed, while in the skull floor the
prevomer, parasphenoid, prootics and occipital complex are present. At the level
of the skull wall, the pterosphenoid and orbitosphenoid are still confined to small
layers of bone surrounding the fenestra sphenoidea.
With the appearance of the preopercle and interopercle and the ossification of
the quadrate, interhyal, hyomandibula, metapterygoid, articular complex and
posterior ceratohyal, almost all elements of the bony splanchnocranium are also
present from this stage on.
All vertebrae and their neural and haemal spines have started ossifying at this
stage, as is also the case for the elements of the caudal, dorsal and pectoral fin
skeleton, although in the latter, no ossification was observed at the level of the
fin rays yet (only at the level of the fin spine). In the pelvic girdle and anal fin,
on the other hand, ossification of the fin rays had started but no ossification was
present in the supportive skeletal structures yet.
Stage 10: 10.0-11 mm SL (4.39A; 4.40G; 4.41C; 4.42C)
At the level of the skull roof, the lateral ethmoid has appeared while all other
bones now contact each other. In the skull wall, orbito- and pterosphenoid have
enlarged substantially.
In the splanchnocranium, all bones have ossified further, contacting each other
and rigidifying the entire structure.
Ossification in the postcranial skeleton is completed in all fins.
Stage 11: 16 mm SL (4.39B)
At this point the two tubular infraorbital bones are the last bones to appear.
Discussion
Growth
In comparison to the five different inflexions found in the growth curve of
Corydoras aeneus (see 4.2), only a single inflexion was found in the analysis of
growth in C. pygmaeus (fig. 4.33B). This can partially be related to the lower
111
4 ONTOGENY – MINIATURIZATION
number of specimens used in this analysis (26 compared to 39 used for C.
aeneus). In addition, differences in growth could also be an effect of differences
in dietary requirements in both species. Since the nutritional requirements of
both species are unknown, they were fed a similar diet throughout their
ontogeny (see 2.1), this way attempting to minimize the impact of nutrition. A
thorough analysis of the optimal dietary conditions could however, prove usefull
in eliminating this possible effect on differences in growth.
Nevertheless, the SL at which this inflexion point is situated (5.2 mm) does
correspond to the second inflexion found in the growth of C. aeneus (5.4 mm).
In C. aeneus the occurrence of this inflexion was linked to a change in swimming
mode around this SL, combined with the onset of finfold differentiation (start of
pterygiolarval phase) and notochord flexion. Weihs (1980), in his study on
Engraulis mordax, found a continuous swimming mode to be more efficient in
larvae under 5 mm, while in larvae with a length between 5 and 10 mm burstand-coast swimming became more effective. Blaxter (1986), on the other hand,
mentions that a change in swimming mode occurs as the caudal fin develops and
inertial forces start to play a bigger role in locomotion. Although in C. pygmaeus,
no notochord flexion is observed, development in the caudal fin skeleton is
ongoing at that point and both finfold differentiation as well as notochord flexion
occur only slightly later in ontogeny (at 6.2 and 6.6 mm SL respectively). This all
indicates that the single inflexion point observed in the growth of C. pygmaeus
may also be related to a change in swimming mode similar to the one suggested
for C. aeneus.
External morphology
At hatching, both species are of comparable size (3.5 mm SL in C. aeneus; 3.7
mm SL in C. pygmaeus), although duration of prehatching development in C.
aeneus (3.3 days ± 0.5; n=120 (MEAN ± S.D.)) was significantly shorter (p=
4.3E-06) than in C. pygmaeus (4.1 days ± 0.3; n=12). After this, the yolk sac
becomes depleted in both species at a SL of 4.4 mm, respectively at 1 dph (5
dpf) in C. pygmaeus and 3 dph (6 dpf) in C. aeneus. Although no information on
egg size in C. pygmaeus was recorded, the delay in time seems to be result of a
decreased rate in the development of C. pygmaeus since yolk sac size in
hatchlings was highly comparable (fig. 4.4A; 4.34A). After this point, the SL at
which transitions between the different developmental phases occur starts to
112
4 ONTOGENY – MINIATURIZATION
differ between both species. While the shift from protopterygiolarval to
pterygiolarval phase is somewhat comparable between both (6.2 mm SL in C.
pygmaeus and 5.7 in C. aeneus), the start of the juvenile period in C. pygmaeus
occurs at 8.9 mm SL, which is substantially lower than the 14 mm SL at which
this shift takes place in C. aeneus. These shifts, however, do correspond to a
comparable age of 22 days in C. pygmaeus and an age of 23 days in C. aeneus
(fig. 4.43A; B).
Chondrocranium
Generally, during development of the chondrocranium, the cranial morphology
of C. aeneus and C. pygmaeus is comparable and no major differences are
noted. The fact that both halves of the Meckel’s cartilage are medially fused in
the earliest stage described for C. pygmaeus is one of such few differences but
this could also be the result of staining of the ligamentous tissue between both
halves at this point, subsequently masking the separation of both halves (no
histological sections were available to verify this). The fact that, from that point
on, in no other stages a similar fusion has been found, seems to confirm this.
However, the level of development in the chondrocranium of both investigated
species at the time of hatching, is remarkably different, which could relate to the
longer prehatching stage in C. pygmaeus. The level of chondrification in C.
pygmaeus is very high, whereas in C. aeneus only the onset of chondrification is
present in hatchlings. In general, the complexity of the chondrocranium at
hatching, although highly variable in teleosts, is generally limited in Siluriformes,
while studies on species of the Loricarioidea group, however, tend to show a
more developed chondrocranium at that point (see 4.3). The fact that, in both
congeneric species investigated here, the degree of chondrification of the
neurocranium also differed remarkably, confirms the high variability of the
chondrification state at hatching in general. Again, the difference observed
probably relates to the longer prehatching stage in C. pygmaeus. Fig. 4.43
indeed shows that development of the chondrocranium takes place over a similar
age-range in both species, with a delay of hatching in C. pygmaeus. At the time
of yolk depletion, the level of development of the chondrocranium in both species
has become highly comparable (figs. 4.9; 4.36). This condition evidently may
relate to the fact that, from this stage on, exogenous feeding becomes obligatory
113
4 ONTOGENY – MINIATURIZATION
and the presence of a functional feeding apparatus is required (van Snik et al.,
1997; Jardine & Litvak, 2003).
Osteocranium
In both species the onset of ossification in the neurocranium was situated
around 8-9 mm SL. This size, however, corresponded to an age of around 10 dph
in C. aeneus, but an age of around 20 dph in C. pygmaeus. In both species
ossification of most elements continues up to the age of 35-40 days post
hatching, but in C. pygmaeus, as mentioned earlier, the onset of ossification is at
20 days, while this occurs at 10 days in C. aeneus. During this period, growth in
the skull of C. aeneus continues, while in C. pygmaeus it almost stops. In C.
pygmaeus,
however,
the
last
elements,
the
infraorbital
bones,
appear
substantially later, at 105 dph. The result of this is that, at the end of cranial
ossification, next to the skull being smaller in C. pygmaeus, tubular infraorbital
bones, typically regarded as typical paedomorphic features (Weitzman & Vari,
1988) are also retained.
Postcranial skeleton
While initial formation of the cartilaginous postcranial skeleton starts at a lower
SL (and age) in C. pygmaeus, ossification in most parts, as is the case in the
cranium, starts at 8-9 mm SL in both species. This corresponds to a substantial
time delay in C. pygmaeus. Again, this could help to explain the smaller size of
the adult, as well as the lower number of dorsal fin rays, both considered typical
paedomorphic, or reductive, traits (Weitzman & Vari, 1988).
Heterochronies
Because heterochrony deals with changes in the rates and timing of
developmental processes, the most straightforward way to study it is to compare
the actual curves depicting measures of size as a function of developmental time
(Klingenberg, 1998). In this case, the analysis of both growth curves points
toward neoteny as the main paedomorphic mechanism in the miniaturization of
C. pygmaeus, since, in the formalism of Alberch et al. (1979), neoteny is defined
as a decrease in the rate of development. In both C. pygmaeus and C. aeneus,
growth rate is comparable in specimens under 10 dph (SL of around 6.5 mm),
after which, growth speeds up in C. aeneus, but not in C. pygmaeus (fig. 4.33B).
114
4 ONTOGENY – MINIATURIZATION
In the light of specific traits, next to rate, shifts in the point of onset and offset
can also produce paedomorphisms. In this case, the clearest example is the
postdisplacement of onset of ossification in the cranium in C. pygmaeus. The
results also show this onset in ossification to be size-related (8 mm SL), as is
also the case for offset of development in the chondrocranium (5-6 mm SL). On
the other hand, offset of ossification appears to be age-related (35-40 dph).
Overall, the ontogenetic pattern in these species appears to be very conservative
with both age- and size related aspects. This way, growth rate and the
subsequent postdisplacement of size-related traits (like the onset of ossification
at 8 mm SL) appear to be the main cause of a miniaturized body in C.
pygmaeus.
In the context of the identification of heterochronic events through interspecific
comparisons, Fuiman et al. (1998) suggested the use of an ontogenetic index.
This index OL gives a measure of the relative timing (in size) of an ontogenetic
event in relation to the size at metamorphosis. These measures are put on the
log scale and the relation is mutiplied by 100, placing these events on a scale
between 0 and 100. In this study, calculations are based on SL, and this way, for
example, the ontogenetic index of the moment of hatching is calculated as: log
(SLhatch)/log(SLmetamorph.)*100. These indexes were calculated for hatching, yolk
depletion and finfold differentiation and results are shown in fig. 4.44. This shows
a clear neotenic mechanism behind the miniaturization in C. pygmaeus since
similar ontogenetic events have been delayed in time in this species.
In conclusion, miniaturization in the body of C. pygmaeus is indeed the result
of paedomorphic events, which include a lower growth rate and subsequent
postdisplacement of several traits. These paedomorphic events have, however,
besides the smaller size, not led to the retention of a large number of
paedomorphic traits.
115
4 ONTOGENY – MINIATURIZATION
116
Chapter
5
General Discussion
5 GENERAL DISCUSSION
5.1 THE ONTOGENY OFCORYDORAS AENEUS
The theory of saltatory ontogeny
Studies on the early development of fish have viewed ontogeny either as a
continuous, inconspicuous accumulation of small changes (gradual ontogeny), or
as a series of steps and thresholds (saltatory ontogeny) (Vilizzi & Walker, 1999).
The concept of saltation in ontogeny was first proposed by Vasnetsov (1953) and
Kryzhanovsky et al. (1953) who proposed a ‘step theory’ in which they
recognized ontogeny to be a sequence of ‘etapes’ of quantative morphogenesis
and growth, separated by a combination of brief but distinct qualitative changes
(‘leaps’) in a developmental process. Balon (1959, 1975, 1979, 1990) later
adapted this theory and formulated saltatory ontogeny as a sequence of longer
stabilized states, or steps, punctuated by rapid changes in ‘integrative actions’,
which form ‘thresholds to survival’. Such thresholds are considered abrupt
functional changes in ontogeny that produce relationships to the environment
(McElman & Balon, 1980). Balon (1959, 1971) also stated that: “these
thresholds could very well be the most important intervals of ontogeny, during
which adaptations to the environment express themselves and the future course
of ontogeny is determined”. A clear recognition of these thresholds would
however, prove to be extremely difficult based on the mere study of adult
specimens (Balon, 1979). Therefore, ontogenetic studies aiming to recognize
these important thresholds should aim to describe as many features of the
developing larvae as possible (Balon, 1979). In my study of Corydoras aeneus, I
examined the ontogeny of skeletal and muscular elements during cranial and
postcranial development. Although no additional internal organ systems were
studied, the data presented here should suffice in recognizing the position of
some of these important thresholds during ontogeny. In addition, under 4.2, an
analysis of several mensural characters was performed. Mensural characters with
inflexion points occurring at similar size are also suggested to be associated with
some common ecological, physiological and/or behavioural function and could
therefore also prove useful in recognizing thresholds during development (Kováč
& Copp, 1999).
117
5 GENERAL DISCUSSION
Finally, similar data on Corydoras pygmaeus (see 4.7) and equally extensive
comparable data on Ancistrus cf. triradiatus (Geerinckx, 2006) allow for a
comparison of these thresholds in all three species thus adding an evolutionary
dimension. Given the extensive differences in the development, morphology and
early life history between both A. cf. triradiatus and C. aeneus, I expect the
position of these thresholds to differ substantially between both ontogenies.
Since the objective of this part is to compare the differences in the number and
relative position of thresholds related to the differences in the ontogenetic
trajectories of all three species, it is sufficient that only a single axis of
comparison was used. For this purpose, I chose SL, since most data on all three
species describe the occurrence of the different developmental events in a SLbase staging system.
The ontogeny of Corydoras aeneus
In this part, the results presented under chapter 4 are briefly recapitulated,
with the exception of the egg morphology. The beginning and end of the
development of different morphological units and the position of other important
events in the development are subsequently mapped onto a SL axis depicting
development in Corydoras aeneus as was done under 4.7 (Fig. 4.43). This way,
those positions in development in which several important events (inflexions in
growth, onset and offset of development in certain structures and the transitions
between the different developmental phases) co-aligned, were designated as
thresholds. As an arbitrary marker, I chose only to describe those points at which
at least three events clearly co-aligned as thresholds. This allowed for a more
objective comparison with the ontogenies described for C. pymaeus and
Ancistrus cf. triradiatus, as presented below.
Ontogeny
Early development and growth
Based on the study of ontogeny of external morphology, four different phases
were recognized in the ontogeny of C. aeneus. First an eleutherembryonic phase
is present from hatching until 4.4 mm SL, during which specimens are
118
5 GENERAL DISCUSSION
unpigmented and the mouth and anal opening are still closed. During the
protopterygiolarval phase, at a SL of 4.4 to 5.7 mm, finfold differentiation starts,
the yolk sac disappears and the anus, mouth and opercular cavity open. After
this, at a SL between 5.7 and 14.0 mm (the pterygiolarval phase), the finfold
further differentiates until all fins have fully detached from the larval finfold. In
addition to an analysis of external morphology, a separate analysis of change in
overall growth yielded five different inflexion points at which growth rate
changed significantly. These points existed at 3.9, 5.4, 7.4, 14.5 and 17.9 mm
SL.
Chondrocranium
The development of the chondrocranium starts around the time of hatching at a
SL of 3.3 mm. At 5.3 mm SL, almost all elements of the cartilaginous skull are
present with the exception of some structures in the branchial basket which are
still fused at that point.
Osteocranium
The first ossifications in the skull of C. aeneus were found at the level of the
splanchnocranium in which the opercle and tooth-bearing elements in the oral
and pharyngeal jaws develop a thin layer of bone in between 4.9 and 5.3 mm SL.
After this, no new centers of ossification are found in the skull until 8.2 mm SL.
From that point on until 13.9 mm SL the remaining elements of the
splanchnocranium ossify, while ossification of the neurocranium continues until
18 mm SL.
Cranial myology
The first signs of cranial muscles become apparent at the time of hatching, at
3.3 mm SL. After this, development of the cranial musculatures occurs rather
fast, with all different muscles present from a SL of 5.3 mm on.
Postcranial skeleton
In C. aeneus the development of the postcranial skeleton takes place between
4.8 and 15-16 mm SL.
Thresholds
Fig. 4.45A shows all elements described above mapped onto a SL axis. As the
two most important events during early development, hatching and yolk
depletion are also indicated on this axis. Four thresholds were found at those SL
119
5 GENERAL DISCUSSION
at which several events co-aligned. The overview clearly shows that development
in C. aeneus is discontinuous with hatching co-occurring with a first threshold, at
which point development of the chondrocranium and the cranial musculature
starts and shortly after which a first inflexion in growth occurs. This is followed
by a first stabilized step occurring between 3.3 and 5.3 mm SL in which yolk is
depleted, the chondrocranium further develops and all cranial muscles and the
first elements of the bony splanchnocranium (the tooth bearing elements)
appear. A second threshold occurs at the time at which all cranial muscles, the
tooth
bearing
elements
of
the
splanchnocranium
and
all
parts
of
the
chondrocranium have been formed. At this point, a second inflexion in growth
was also found. Just prior to this point, development of the cartilaginous
postcranial skeleton has started and shortly after this point finfold differentiation
starts. Between 5.3 and around 8 mm SL, a second stabilized state is present in
which only in the postcranial cartilaginous skeleton, new elements arise. The
third threshold is located at 8 mm SL, shortly after a third inflexion in growth and
at the same time of the onset of ossification in the rest of the splanchnocranium,
the postcranial skeleton and the osteocranium. After another stabilized period
between c. 8 and 14 mm SL, the last threshold is found. At this point, ossification
in the splanchnocranium has completed, a fourth inflexion in growth is found and
the juvenile period starts.
Conclusively, the analysis of the development of both the musculoskeletal
system and external development indeed show strong indications of saltatory
ontogeny in C. aeneus. Since this can not be determined definitively without a
further analysis of additional organ systems beyond the musculoskeletal
apparatus, this conclusion appears somewhat preliminary. However, the different
morphological traits used for the demarcation of the different developmental
phase also include several aspects of the development of the internal organ
systems. This way, depletion of yolk and opening of anal and mouth opening
indicates a functional digestive tract, whereas fin development is indicative of
developing locomotory functions. Therefore, although additional data is needed
to confirm the exact points at which thresholds occur in the ontogeny of C.
aeneus, it is clear that ontogeny in this species follows a saltatory pathway.
During this ontogeny, at least four thresholds are present in which important
events co-align.
120
5 GENERAL DISCUSSION
Corydoras pygmaeus and Ancistrus cf. triradiatus
Corydoras pygmaeus (fig. 4.45B)
In the early development of C. pygmaeus also four different phases are found.
The eleutherembryonic phase starts at 3.7 mm SL until, at 4.4 mm SL, yolk
becomes depleted (fig. 4.34A;B). After this, from 4.4 mm SL until 6.2 mm SL,
the protopterygiolarval phase is present. At the end of this phase, finfold
differentiation starts and specimens reach the pterygiolarval phase (fig. 4.34C).
At 8.9 mm SL all fins are fully detached from the larval finfold and the juvenile
period starts (fig. 4.34D). The growth curve of C. pygmaeus only shows a single
inflexion point at 5.2 mm SL. Development of the chondrocranium starts during
the embryonic phase which leads to a high level of development in the
chondrocranium at hatching. By around 4.8 mm SL, all elements of the
cartilaginous cranium have been formed. In C. pygmaeus, the first ossifications
found in the developing cranium are located at the level of the splanchnocranium
(opercle, dentary and premaxilla). These early ossifications arise between
hatching and 4.3 mm SL. After this, ossification of the other cranial elements
does not start until 8-9 mm SL. At 16 mm SL, the infraorbital bones are the last
elements to ossify. Development in the postcranial skeleton of C. pygmaeus
starts at hatching and the different elements start to ossify at 8-9 mm SL.
Ossification in the postcranial skeleton is completed at around 10-11 mm SL.
As in C. aeneus, it is clear that the ontogenetic trajectory fits the definition of
saltatory ontogeny as given by Balon (1959, 1975, 1979, 1990). Also similar to
C. aeneus, the first threshold is located at hatching, while the second threshold
occurs around the end of development of the chondrocranium at a point at which
growth rate changed. In contrast to the situation in C. aeneus, however, this
point also corresponds to the time of yolk depletion in C. pygmaeus. A third and
last inflexion, again similar to C. aeneus, is found at the time at which
ossification in the neurocranium and postcranial skeleton starts. In contrast to
the situation in C. aeneus, this point does not correspond to the onset of
ossification in the non-tooth-bearing elements of the splanchnocranium, which is
already ongoing at that point. Another difference with C. aeneus is that this point
corresponds to the start of the juvenile phase.
121
5 GENERAL DISCUSSION
Ancistrus cf. triradiatus (fig. 4.45C)
In the early development of A. cf. triradiatus, only three distinct periods were
recognized. Embryos hatch at 6 mm SL, after which the yolk sac remains present
until 10 mm SL, which is a distinctly longer period than in C. aeneus. At the
moment of yolk sac depletion, specimens of A. cf. triradiatus already have an
adult-like appearance and therefore, a true larval stage is absent (cfr. Balon,
1975). Data by Geerinckx (2006) also showed that the development of the
chondrocranium in A. cf. triradiatus starts around 5.2 mm SL, well before
hatching and that most structures of the cartilaginous skull are formed by 8.9
mm SL. Also in this species, the first elements of the skull to ossify were the
tooth-bearing elements and opercle (at 5.6 mm SL), but after this, ossification of
the splanchnocranium continues in each of the stages examined. During the
ontogeny of the cranial muscles in A. cf. triradiatus, the first elements appear at
a SL of 6.1 mm, while the last elements are formed by a SL of 10.2 mm. The
method described under 4.2 for the calculation of inflexion points in growth was
not applied to the growth data of A. cf. triradiatus and therefore no comparable
data in this segment is present, nor is this the case for data on postcranial
development. However, it seems clear that the ontogeny of A. cf. triradiatus
follows a totally different path than those found in both Corydoras species. The
only point at which a possible threshold is found is at the point of hatching, just
after the onset of splanchnocranium formation, and co-aligned with the onset of
development of the cranial muscles. Due to this limited number of thresholds and
the absence of stages in which very little new structures are developed (compare
to both Corydoras species), the term ‘saltatory ontogeny’ does not apply to the
ontogeny of A. cf. triradiatus.
Thresholds in the development of Loricarioidea
The results clearly show that the main difference found between the loricariid
and callichthyid ontogenies studied here is the fact that the number of thresholds
is substantially lower (or even totally zero) in the ontogeny of the loricariid. The
only threshold found in all three species was the point of hatching itself, although
the importance of hatching as a threshold of ultimate survival value has already
122
5 GENERAL DISCUSSION
been questioned (Balon, 1984). Additional thresholds found in the ontogenies of
both callichthyids correspond to other important life history events such as the
transition to exogenous feeding and the point at which the larvae become
juveniles. Since the point at which the yolk sac is depleted in A. cf. triradiatus
has been postponed in comparison to both callichthyids, most cranial structures
have been formed by that point and the point probably forms a smaller threshold
to survival than in both Corydoras species. In addition, it is known that in
loricariid fishes a true larval stage is absent (Geerinckx, 2006), again eliminating
that point as a threshold in early development. In general, such a reduction of
the larval period in fishes is believed to be associated with the development of
reproductive hiding strategies (or other forms of parental care), which itself is
directly related to an increase in egg size and yolk density (Balon, 1979; Kolm &
Ahnesjö, 2005). In general Balon (1984) formulated the relation as: “In fishes
with an increased endogenous food supply and parental care (increased
reproductive cost per embryo), the embryo develops permanent organs directly,
bypassing the larva period with its remodelling of temporary structures”. Indeed,
in this case, the differences in egg size seem to have clearly influenced the
differences found in the early ontogenies of Corydoras and Ancistrus species.
This way, it was already shown that the appearance of a highly developed
chondrocranium at hatching in A. cf. triradiatus was probably the direct result of
this increase in egg size. The analysis presented here now additionally shows
that the prolongation of the embryonic period also seems to affect the
occurrence of thresholds and a subsequent saltatory developmental pattern in
early ontogeny, with the elimination of a metamorphosis from a larva into a
juvenile as the most striking result. The elimination of this larval period is
generally accepted as an important ecological and evolutionary phenomenon
(Balon, 1984) and may therefore have played a crucial part in the evolution of a
specialized feeding apparatus in Loricariidae. The prolonging of the embryonic
period would, after all, have provided the time necessary for the development of
these structures since active feeding only becomes necessary at a point much
later in development.
Given the well documented relation between the elimination of the larval
period, egg size and parental care, it seems very likely that selection at the level
of the reproductive strategies in loricariids has also played an important part in
123
5 GENERAL DISCUSSION
the radiation and evolutionary success of this family, but this will be further
discussed under 5.2.
124
5 GENERAL DISCUSSION
5.2 THE EVOLUTION OF ALGAE SCRAPING
IN
LORICARIOIDEA
Scaridae, another example of adaptive radiation and algivory in
teleosts
It has been argued that seemingly different vertebrate radiations follow similar
evolutionary trajectories in which groups diverge along axes of habitat, trophic
morphology and communication, often in that order. Speciation events in stages
one (habitat) and two (trophic morphology) probably follow the rules of
ecological
selection
models,
while
lineage
diversification
in
stage
three
(communication) would occur according to sexual selection models (Streelman &
Danley, 2003). Seehausen (2004) has also suggested a similar model of
radiation, although he did not suggest an order of the different stages, but does
state that three main factors influence radiation. These factors include release
from competition in an underutilized environment, key evolutionary innovations
that enable utilization of resources that were there but could not be utilized and
mating systems that are conducive to rapid divergence (Seehausen, 2004). In
general, both hypotheses merge the traditional ecological radiation theory with
the sexual model alternative, with or without presenting an order in which both
effects occur. Streelman & Danley (2003) exemplify their statement, including
the order of events, using various examples including the example of the family
of the parrotfishes (Scaridae), another teleost group in which algivory has
evolved. Although it must be stated that not all loricariids are algivorous, the
evolution of a highly specialized feeding apparatus with associated algivory in
most species does seem to form the cornerstone of their radiation. Therefore,
given the relevance of the similarity in trophic specialization between Scaridae
and Loricariidae and the fact that the example is situated within the teleosts, it
seems appropriate to discuss it here. In Scaridae, divergence has clearly started
along the habitat axis when 40 million years ago, scarids diverged into a ‘reef’
and ‘seagrass’ clade. Second, members of the Scaridae family in the reef habitat
diverged into lineages that either scrape or excavate algae from coral and rock
125
5 GENERAL DISCUSSION
surfaces (Streelman et al., 2002). These differences in trophic utilization were
associated with changes in the craniofacial skeleton and musculature, feeding
rate and kinematics (Bellwood & Choat, 1990; Bellwood, 1994; Alfaro &
Westneat, 1999). By contrast, the seagrass lineage of parrotfishes does not
exhibit secondary divergence in morphological features; nearly all species are
browsers (Streelman & Danley, 2003). Finally, within the reef clade, a tertiary
radiation is observed which involves a sexual dimorphism at the level of male
mating coloration, correlated with behavioral differences (e.g. territoriality and a
haremic mating system) (Streelman et al., 2002; Streelman & Danley, 2003).
Given this, it is beyond doubt that the radiation enclosing the evolution of algae
scraping in Scaridae has evolved along the three axes mentioned before and in
that specific order.
The question dealt with in the following chapter, however, is whether or not the
same model applies for the Loricarioidea. Given the fact that Streelman & Danley
(2003) argued that this model applied to different vertebrate radiations, I
hypothesize that the model is also applicable here. One can test for evidence
supporting these different evolutionary steps in morphology, ontogeny, ethology
and biogeography of the extant loricarioids, as well as in the known fossil record
of the superfamily. Given the limited record existing on Scoloplacids, the
following discussion will be limited to evidence found in Callichthyidae,
Astroblepidae and Loricariidae.
Three phases of evolution
Habitat divergence
As habitat divergence and subsequent lineage divergence is believed to follow
ecological speciation models, both allopatric and non-allopatric models of
divergence can apply to such divergence and examples from both are known
(Streelman & Danley, 2003). In this context, allopatric speciation is defined as a
process in which a reduction to zero of the number of breeding individuals in a
population that are immigrants from other populations, is accomplished by a
physical barrier extrinsic to the organism (Futuyama & Mayer, 1980). Nonallopatric (albeit parapatric or sympatric) speciation is then accomplished by
126
5 GENERAL DISCUSSION
isolation through biological characters of the organisms themselves (Futuyama &
Mayer, 1980). The idea of sympatric and/or parapatric speciation has long been
considered highly unlikely (Coyne & Orr, 2004; Seehausen, 2004) but recently,
several studies have showed evidence for such sympatric/parapatric speciation
models. Since most adaptive radiations occur in geographically narrowly confined
regions, sympatry and parapatry during part of the speciation process is invoked
in many (Seehausen, 2004). The divergence of ‘reef’ and ‘seagrass’ clades in
Scaridae provides an example of such a sympatric process in which the split in
habitat might have occurred in the absence of physical barriers that would
prevent gene flow (Streelman & Danley, 2003). Also, the cichlid fishes of Lake
Barombi Mbo (Cameroon) provide one of the most convincing examples of
sympatric speciation in animals (Turner, 2007). Nonetheless, although sympatric
speciation is theoretically plausible and supported, there is little evidence that it
is common (Coyne & Orr, 2004). More common are examples in which lineages
diverge in allopatry, like the Hawaiian honeycreepers and Galápagos finches
(Lack, 1947; Grant, 2000). In these cases, lineage splitting is believed to have
begun in allopatry, but to have been accelerated greatly following secondary
contact (Grant, 2000). Given the more common nature of the allopatric
speciation model and the broad geographical range in which the adaptive
radiation in Loricarioidea has occurred (in contrast to other known radiations), it
would seem plausible that habitat divergence in Loricarioidea has also occurred
in allopatry.
The geographic separation needed for allopatric speciation can occur through
diverse climatic or geological events or the extinction of intermediate populations
(Coyne & Orr, 2004). In order to elucidate this, one needs to look at the
biogeographical context and fossil evidence of the Loricarioidea. Since loricarioids
are generally believed to form the sister group of African amphiliids (de Pinna,
1993; 1998), the superfamily itself could have originated >112 Ma, the point at
which both the separation of Africa and South America is completed through the
opening of the South Atlantic (Lundberg, 1998; Maisey, 2000). Recently, Sullivan
et al. (2006) placed the loricarioid superfamily as a sistergroup to all other
catfish based on molecular data. This subsequently led to a estimate of
divergence between Loricarioidea and all other catfish to 125 Ma, predating the
Gondwana break up (Lundberg et al., 2007). In both cases, these estimates by
far predate the hypothesis formulated in the refuge theory by Haffer (1982),
127
5 GENERAL DISCUSSION
which has long been one of the most popular theories explaining Neotropic
diversity. This theory states that most of the Neotropic divergence occurred
recently and that the primary forces were the climatological fluctuations of the
Pleistocene (1.8 million to 11.000 years ago), which fragmented and reconnected
habitats several times, promoting allopatric speciation (Montoya-Burgos, 2003).
Since new morphological as well as molecular data predates this, diversification
has been linked to the particularly eventful earlier (< 1.8 Ma) history of
Neotropical river systems. Based on this, the so-called ‘hydrogeological’
hypothesis was formulated (Montoya-Burgos, 2003). This states that the
formation of drainage divides, shifting courses of rivers and repeated incursions
and regressions of marine waters must have produced many vicariant events
promoting biotic enrichment (Lundberg, 1998). The fossil evidence found in the
loricarioid group also corroborates this hydrogeological hypothesis. The earliest
fossil known in this superfamily is a callichthyid described as Corydoras
revelatus, from the late Paleocene, ca. 58.5 Ma, found in the Argentinian Maíz
Gordo formation (Lundberg, 1998; Lundberg et al., 1998). This indicates that, at
least up to the level of Callichthyidae, divergence in loricarioids also long
predates the Pleistocene. Based on the remaining loricarioid fossil record,
Lundberg (1998) infers that a considerable diversity of modern loricarioids had
evolved by at least 13.5 Ma. This is recently confirmed by a molecular analysis of
some
members
of
the
Hypostomus genus, which dates back loricariid
diversification at the genus level to 12 – 4 Ma (Montoya-Burgos, 2003). In
addition, recent fossil calibrations estimate the Callichthyidae – (Astroblepidae +
Loricariidae) divergence at 110 Ma (Lundberg et al., 2007). This way, it is clear
that various vicariance events may have played a crucial role in the early
divergence of loricarioid clades, and have given the onset of habitat divergence.
Regarding habitat divergence in the case of vicariance, much depends on the
habitat both groups are isolated into. The most evident manner in which
allopatry can lead to a divergence in habitat use is the case in which both groups
get isolated into different habitat types. On the other hand, both groups could
get isolated in similar habitat types but other selective forces (like competition)
or genetic drift could drive both groups into different directions of habitat use.
The former option seems the most plausible in this specific case, since the Maíz
Gordo formation, in which the Corydoras revelatus fossil was found, was believed
to be a wetland to lacustrine environment (Lundberg et al., 1998), while the
128
5 GENERAL DISCUSSION
distribution of extant Astroblepidae (no fossil taxa within this family are known
(Ferraris, 2007)) is limited to high elevation streams of the Andes (Schaefer,
2003). This limited distribution could indicate that the distribution of the common
ancestor of Astroblepidae + Loricariidae is also located in this narrow
geographical range, with a secondary recolonization of the Brazil and Guyana
shield by loricariids. Another hypothesis could be a broad distribution in the
common ancestor to both families with a secondary limitation in the distribution
of Astroblepidae. This would imply that Astroblepidae have evolved from an
Andes clade of this common ancestor into a separate family. The divergence
between Astroblepidae and Loricariidae indeed dates back to 90 Ma (Lundberg et
al., 2007), at which point formation of the Andes started, altering the entire
hydrogeography of the South American continent (Lundberg et al., 1998). The
fact that the distribution of extant Loricariidae also covers the high Andes would,
however, imply a secondary recolonization of the Andes by this group. In
addition, Astroblepidae possess several derived features like a suctorial mouth
and a specialized pectoral to pelvic fin muscle system, two structures that can be
used in an alternating manner, giving astroblepids the capability of climbing
vertical surfaces and colonize fast water currents (Shelden, 1937). This matter
also partially answers the question posed by Adriaens (2003) and Geerinckx
(2006) on the evolutionary order of suckermouth attachment versus scraping
feeding mode in loricarioids. The evolution of attachment modes in species
colonizing turbid environments is indeed common, albeit often developed from
other parts than the mouth (Geerinckx, 2006). Also, the fact that a sister family
like the Astroblepidae only possesses a suckermouth without the specialized
feeding mechanism indicates an appearance of this suckermouth earlier in
evolution. In addition, Lundberg et al. (2007) place the common ancestor to
(Scoloplacidae + Astroblepidae + Loricariidae) only at 95 Ma, which places the
evolution of a suckermouth in Loricarioids between 95 and 90 Ma, at the same
time of the start of Andes formation (Lundberg et al., 1998).
All this indicates that, at the basis of loricarioid evolution, several vicariant
events took place splitting up the Callichthyidae and the common ancestor to
Astroblepidae and Loricariidae into habitats with highly different hydrological
characteristics.
129
5 GENERAL DISCUSSION
Trophic morphology
The fact that a great extent of the divergence of the different families of
loricarioids has occurred at the level of their trophic morphology forms the main
reason this project was started (see 1.1). This is related to the fact that radiation
in Loricarioidea has peaked in the Loricariidae, the family in which, in addition to
morphological changes, dietary specializations have also evolved, indicating the
importance of a divergence in trophic morphology. In addition, loricariids also
greatly exceed all other loricarioids in terms of diversity of jaw and teeth size and
shape (Schaefer & Lauder, 1986; Geerinckx & Adriaens, 2006). Therefore, it is
generally believed that the specialized trophic morphology of the Loricariidae
forms the key-stone of their evolutionary success.
In 1981, Lauder proposed a general hypothesis concerning evolution of
structural patterns as:
“primitive
members of a
morphologically diverse
monophyletic lineage possess functional, structural, or morphogenetic networks
which have a greater number of independent (decoupled) elements as compared
to
similar
networks
in
closely
related
but
less
morphologically
diverse
monophyletic lineages”. Schaefer & Lauder (1986) later referred to this as the
‘decoupling hypothesis’. The definition of decoupling in this case can be stated as
the repetition of individual elements as redundant design components, followed
by the specialization of one or more of these elements as a mechanistic basis for
the evolution of a novel structure/function (Schaefer & Lauder, 1996) (also see
1.1; 3.2). Schaefer & Lauder (1986, 1996) showed that the Loricarioidea are a
clear example of a clade showing a pattern of progressive increase in mechanical
complexity, in this case, in those structures associated with feeding (Schaefer &
Lauder, 1986). In the loricarioids, decoupling has mainly occurred at the level of
the jaws. In Callichthyidae, Astroblepidae and Loricariidae (Scoloplacidae were
not dealt with in the study by Schaefer & Lauder (1986, 1996)), the
premaxillaries have been decoupled from the cranium, increasing their mobility
through modifications in the ethmoid hinge joint (see 3.1). In addition, Schaefer
& Lauder (1986) mention a shift of a dorsal extension of the m. adductor
mandibulae onto a sheet of connective tissue between lower and upper jaw. This
muscle has been called m. retractor tentaculi by Howes (1983) and under 3.2
and 4.5 in this thesis. In addition, a subdivision of the m. extensor tentaculi in C.
aeneus was not reported by Schaefer & Lauder (1986, 1996), although it also fits
the general trend of an increase in morphological complexity through decoupling:
130
5 GENERAL DISCUSSION
in
loricariids,
the
muscle
becomes
completely
subdivided,
allowing
an
independent mobility of the maxilla and premaxilla (Geerinckx, 2006) (see 3.2).
In Astroblepidae and Loricariidae, upper jaw mobility is further improved
through a direct insertion of the m. retractor tentaculi on the premaxillaries and
a loss of the connection between the lig. primordium and the maxilla. In addition,
the lower jaw has been decoupled from the opercular series, allowing an
independent mobility of both upper and lower jaw (Schaefer & Lauder, 1996).
Decoupling events have also played a major role at the level of the lower jaws
and hyoid musculature. These have become decoupled from their plesiomorphic
bilaterally constrained midline attachments and a new linkage between hyoid and
mandible was acquired (Schaefer & Lauder, 1996).
It is clear that all these decoupling events have facilitated the formation of a
functional suckermouth in a first phase with a subsequent algae scraping feeding
apparatus in a second phase. This divergence in trophic morphology has this way
played a major role in the radiation process of the loricariids.
Communication
In the model by Streelman & Danley (2003) this stage in vertebrate radiation
occurs according to sexual selection models, which explain divergence through
competition for mating opportunities. This often involves the development of
male secondary sexual characteristics (Streelman & Danley, 2003). Indeed, also
in the specific case of loricarioid evolution, a high degree of sexual selection
seems to have played a significant role. In Callichthyidae, sexual dimorphism is
confined to differences in size and length of the dorsal and pectoral fin spine
(Kohda et al., 2002; Pruzsinszky & Ladich, 1998), while many Loricariidae are
known for their elaborate sexually dimorphic modifications present in mature
males. Some of these sex dimorphisms are pervasive throughout loricariids and
are considered generalized traits for the family (Rapp Py-Daniel, 2000). These
traits include the presence of large odontodes on head, fins and body on mature
males (Rapp Py-Daniel, 2000). Examples of sexual dimorphism in Loricariidae are
well
documented
in
literature
and
include
studies
on
Rineloricaria,
Pseudancistrus, Otocinclus, Hypostomus, Farlowella and many more (Aquino,
1994; Mazzoni & Caramaschi, 1995; Retzer & Page, 1997; Rodríguez &
Miquelarena, 2005; Lujan et al., 2007). In species of the hypostomine Ancistrinitribe, which is considered the most derived clade in loricariid phylogeny
131
5 GENERAL DISCUSSION
(Armbruster, 2004), this sexual dimorphism has even evolved beyond these
general traits. In these species an erectile cheek-spine apparatus has evolved as
one of the synapomorphies of this tribe (Schaefer, 1987). An example of such
sexual dimorphism has been described in Ancistrus cf. triradiatus, in which a tuft
of enlarged odontodes anterior to the opercle can be erected with great velocity,
acting as a defensive/offensive mechanism and which is, on average, more
developed in male fishes (Geerinckx & Adriaens, 2006). The evolution of such a
cheek-spine apparatus has again been made possible by a decoupling event. The
evolutionary decoupling of the opercle from the lower jaw movements has
evidently not only increased lower jaw mobility; it has also altered the function of
the opercular apparatus, which has lost its function in expiration (redundancy)
and became ‘available’ for the acquisition of a novel function. In combination
with the presence of odontodes on bony platelets in the skin, the possibility to
develop a series of articulations between the opercle, the cheek spines, the
cheek plates anterior to them and the quadrate, and, the possibility of the
opercular musculature to expand substantially inside newly evolved skull cavities
has led to the evolution of this cheek-spine apparatus (Geerinckx & Adriaens,
2006).
The appearance of sexual dimorphism in Loricariidae also appears to coincide
with the evolution of parental care and several loricariid species are known to
exhibit parental care with great diversity in reproductive strategies. In several
Otocinclus species, for example, no parental care is present (Schmidt, 2001).
This fits the hypothesis of Isbrücker & Nijssen (1992b) that, in the primitive
loricariid
subfamilies
Neoplecostominae
and
Hypoptopomatinae,
sexual
dimorphism is totally absent. Reis & Schaefer (1992) and Aquino (1994),
however, did find some extent of sexual dimorphism in Eurycheilichthys
pantherinus and Otocinclus flexilis and O. vittatus, but this was very limited.
Males of some Hypostominae and Ancistrinae are known to guard their eggs in
holes constructed in stream banks (Schmidt, 2001) and in both subfamilies
examples of sexual dimorphism have been described (Armbruster, 2004;
Geerinckx & Adriaens, 2006). In Loricariini, Isbrücker (1981) described several
types of secondary sexual dimorphism and various examples of parental care
have been documented in members of this tribe. This way, in Loricaria simillima
and L. piracicabae, adhesive eggs are carried on the ventral surface of the body
by the males and even by both sexes in L. cataphracta (Schmidt, 2001). In
132
5 GENERAL DISCUSSION
Loricariichthys anus, L. maculatus, L. platymetopon and Paraloricaria vetula this
attachment is located in an expansion of the lower lip (Taylor, 1983; Schmidt,
2001) and has therefore even been described as a form of mouth brooding in L.
platymetopon (Lassala & Renesto, 2007). Males of species of Rineloricaria,
Ricola, Sturisoma and Sturisomatichthys are even known to, not only guard
(Covain & Fisch-Muller, 2007), but also fan the eggs and remove fungused
individuals
(Schmidt,
2001).
In
all
these
genera,
except Ricola,
sexual
dimorphism included the presence of prominent odontode development in males
(Isbrücker, 1981; Covain & Fisch-Muller, 2007). Covain & Fisch-Muller (2007)
further described sexual dimorphism in Rineloricaria to include hypertrophied
development of the odontodes along the sides of the head, on the pectoral spines
and rays, and the predorsal area of mature males. In male specimens of
Loricaria they describe hypertrophied development of the pectoral spines, blunt
odontodes on the pelvic and anal fin spines, and tooth crowns becoming
shortened and rounded, while in Loricariichthys hypertrophied development of
the lips was reported (Covain & Fisch-Muller, 2007). In general, in their overview
of the Loricariinae, Covain & Fisch-Muller (2007), provide data on the
reproductive biology of members of 21 genera, 12 of which exhibit both sexual
dimorphism and parental care and two of which exhibit neither. In contrast to
the hypothesized relation between parental care and sexual dimorphism three
genera exhibit parental care only and in four genera sexual dimorphism is
described without parental care (Covain & Fisch-Muller, 2007). Nonetheless,
since 14 of the 21 genera described fit this hypothesis, there does seem to be a
relation between parental care and sexual dimorphism.
In Callichthyidae, the extent of parental care found is substantially lower, as is
the case for sexual dimorphism. Only in the species Hoplosternum littorale,
Megalechis thoracata and Callichthys callichthys some form of parental care has
been observed. These species are known to build and guard foam nests in which
eggs are deposited (Pascal et al., 1994; Ramnarine, 1994; 1995; Mol, 1996;
Andrade & Abe, 1997; Hostache & Mol, 1998).
Next to sexual dimorphism, in those groups of loricariids that have developed
parental care, a lower fecundity and higher egg size are also noted (Suzuki et al.,
2000; Geerinckx, 2006). Both the inverse relationship between parental care and
fecundity as well as the positive correlation with egg size have been reported
repeatedly throughout teleosts (Suzuki et al., 2000; Kolm & Ahnesjö, 2005). In
133
5 GENERAL DISCUSSION
the context of the relation between egg size and parental care, Kolm & Ahnesjö
(2005) have recently summarized four, not mutually exclusive, explanations for
this general and positive correlation. A first hypothesis, the ‘safe harbour’
hypothesis states that selection will favor a prolongation of the time spent in the
more protected embryo stage, while the ‘eggs require care’ hypothesis presumes
that the larger eggs would require more care and that selection would therefore
act upon the amount of care provided (Kolm & Ahnesjö, 2005). As a third and
fourth hypothesis a mutual relation between both former hypotheses as well as
the influence of other independent factors is suggested to select simultaneously
for more care and larger eggs (coevolution) (Kolm & Ahnesjö, 2005). Although
Kolm & Ahnesjö (2005) were not able to elucidate this matter and suggested
more research (they proposed Cichlidae, Syngnathidae and Gasterosteiformes as
suitable case studies, but the examples clearly indicate the suitability of the
Loricariidae), the fact whether or not larger eggs preceded care of vice versa, is
beyond the scope of this analysis. The fact remains that, in general, a relation
between egg size and parental care exists and that this has influenced not only
the matter of selection placed on the radiation process (from ecological to sexual
selection), but also the ontogenetic trajectories that generate the morphological
diversity that co-occurs with it. It is known that species with the least developed
parental care retain a longer vestigial larval period during and immediately after
the transitory interval of mixed feeding, which separates the embryonic period of
endogenous nutrition and the juvenile period in which all final adult structures
have appeared (Balon, 1979). In fishes with more advanced hiding strategies
this transitory vestige of the larval period does not exist and, in these, all
juvenile characters develop while feeding is still endogenous (Balon, 1979). With
the advancement of hiding strategies, the transition from endo- to exogenous
feeding is more sudden, and the interval of mixed feeding is very brief or
nonexistent (Balon, 1979). It seems plausible that all these differences occur as
a direct result of the increase in egg size, since inadequate provision for
embryonic nutrition is the primary factor governing the occurrence of a larval
stage (Orton, 1953), or as Balon (1986) puts it: ‘the endogenous food supply is
the most important, for it determines and changes the entire remaining life
history’. This is clearly exemplified in my comparison of ontogeny in Corydoras
aeneus and Ancistrus cf. triradiatus under 5.1.
134
5 GENERAL DISCUSSION
In conclusion, in comparison to other loricarioid families, parental care and
sexual dimorphism are clearly more common in the Loricariidae, which confirms
that a substantial degree of ‘communicative divergence’ with subsequent sexual
selection has occurred in loricariid radiation. In addition to this, the increase in
egg size has led to a prolongation of the embryonic period and a loss of the
larval stage in, at least some loricariiids, altering the entire ontogenetic pathway
and lowering the number of crucial thresholds throughout ontogeny (see 5.1).
Three stages
All elements presented above clearly show that, as suggested by Streelman &
Danley (2003) for all vertebrate radiations, radiation in this example has also
followed an evolutionary trajectory along axes of habitat divergence, trophic
morphology and communication. It has also become apparent that the part of
this divergence which follows models of sexual selection through sexual
dimorphism, parental care and the probable co-evolution of egg size has played a
much more crucial role in the radiation of the Loricariidae than expected. The
order in which all three phase occur, however, remains unclear. In the cases of
trophic morphology and communicative divergence, both paths seem to have coevolved. On the one hand parental care and the suggested co-evolution of egg
size would have delayed yolk depletion and eliminated the larval stage, allowing
the necessary time for the development of new structures related to feeding
morphology. On the other hand, the decoupling of various structures related to
feeding has also led to a decoupling of the opercular apparatus, resulting in a
sexually dimorphic mechanism (used for nest guarding), which, in turn would
have led to the evolution of extensive parental care and larger egg size.
Therefore, I believe that the evolutionary basis of radiation in the Loricariidae
has been established by various events of vicariance throughout the South
American history but that a co-evolution of decoupling events in trophic
morphology and sexual dimorphism with related parental care and increase in
egg size have ultimately led to the extreme diversity found in only the loricariid
family.
135
5 GENERAL DISCUSSION
136
Chapter
6
Summary & Samenvatting
6 SUMMARY & SAMENVATTING
6.1 SUMMARY
This doctoral thesis provides a detailed description of the adult morphology of
Corydoras aeneus combined with several aspects of the species’ ontogeny. For
comparison, similar results were gathered from a miniaturized congeneric,
Corydoras pygmaeus and results were further compared to a description of the
ontogeny of the loricariid Ancistrus cf. triradiatus (Geerinckx, 2006). All results
are subsequently discussed in the context of the evolution of algae scraping in
the loricarioid superfamily and adaptive radiation in the Loricariidae. In this
summary I list a brief recapitulation of the main results and conclusions per
chapter.
1.
In the introduction to this dissertation, the example of adaptive
radiation in the family of the Loricariidae is presented. These Loricariidae belong
to the superfamily of the Loricarioidea, a superfamily in which the various
families that belong to it exhibit a trend of increasing morphological complexity
in which various elements have been decoupled, creating opportunities for the
evolution of new structures and functions (Schaefer & Lauder, 1986, 1996). This
way, in members of the more basal loricarioid family of the Callichthyidae a
suckermouth or algae scraping apparatus is still lacking; but the mouth already
has a ventral position. In the family of the Astroblepidae, such a suckermouth is
present, but no algivory is known and finally, in the family of the Loricariidae,
both a suckermouth and algae scraping feeding apparatus are often found. In
this context, the aims of this research were to study the morphology and early
ontogeny of a basal representative of this evolutionary lineage and compare the
results found to those of a highly specialized species.
2.
In the second chapter, the materials and methods used throughout
this dissertation are listed. First, the choice of C. aeneus as the main research
object is discussed, giving arguments that the family the species belongs to
takes a basal position within the lineage leading up to the highly specialized
morphology in the Loricariidae. Within the Callichthyidae, the choice of genus
and species is explained in relation to the amount of data already known from
137
6 SUMMARY & SAMENVATTING
existing literature. Further, breeding and collection protocols and lists of all
specimens collected this way are presented.
In a second part of this chapter, the different methods use are given and, when
appropriate, detailed protocols are included. The different methods explained
include: in toto clearing and staining, dissecting, serial sectioning and the
generation of 3D-reconstructions.
3.
The third chapter describes the adult morphology of C. aeneus and
is divided into two main parts.
A first part deals with the osteology of the species in which results are
compared to the condition described for other loricarioid fishes. Most results are
found to correspond with the expected pattern, although some differences are
observed. One of these differences is the presence of a suprapreopercular bone,
a bone that was previously unmentioned in any callichthyid species. Further,
several homologies are resolved and mainly confirm existing hypotheses. This
way, for example, the compound dorsolateral bone of the otic region is identified
as the posttemporo-pterotico-supracleithrum while the compound bone forming
the neurocranial floor in the occipital region is recognized as a fusion of both
basi- and exoccipital bones. This part of the study, however, still fails to resolve
the homology of the lacrymo-antorbital and suprapreopercular bones.
In a second part, the musculature of the species is described and discussed.
Given the importance of the jaw and opercular muscles in the context of
decoupling and the acquisition of new functions in relation to the specialized
feeding apparatus of Loricariidae, homologies of these bundles are discussed in
detail. In the case of the jaw musculature of adult C. aeneus, an A1OST, A2, and
A3’ section is identified and arguments supporting the homology of the m.
retractor tentaculi with the A3” are presented. In the opercular region, as is the
case in A. cf. triradiatus, the m. dilatator operculi is enlarged and the direction of
the operculo-hyomandibular articulation has shifted. In addition, in both species,
the m. hyohyoidei abductor has also shifted its orientation, acting as an
adductor.
4.
The third chapter deals with several aspects of the ontogeny of both
C. aeneus and C. pygmaeus.
138
6 SUMMARY & SAMENVATTING
In the first part, the surface structure of the eggs of C. aeneus is described.
This showed to be a unique pattern among teleosts. The surface is covered with
small protuberances, which resemble attaching-filaments of teleost eggs. In
addition, eggs were also found to be very adhesive, which probably relates to the
fact that the species is known to inhabit turbid waters.
The second part of this chapter presents an overview of the early development
and the growth pattern of C. aeneus. Based on external morphology, the
different stages in early development, as described by Balon (1975), are
identified. After hatching, from a SL of 3.5 mm, an eleutherembryonic phase is
present, followed by the protopterygiolarval phase (4.4-5.7 mm SL), the
pterygiolarval phase (5.7-14.0 mm SL) and the juvenile period. In addition, an
overall growth curve is presented in which inflexion points were determined,
applying a piecewise linear regression method, which follows the ideas of
regression spline smoothing procedures. This way, the growth curve was divided
into six different stages of growth rate. Initially, the slope is 0.05 until 0.7 dph,
then increasing to 0.18 until 4 dph, and 0.36 until 10 dph. After this, growth rate
reaches a maximum of 0.76 until 24 dph, slows down to 0.47 until 37 dph and
then finally again slows down to 0.36. A similar analysis was also done on the
data of growth in the different body parts. The inflexion points found this way
match the different key-events known in teleost early development. Such events
are: the transition from endo- to exogenous feeding, the transition to the
pterygiolarval phase, when priorities shift toward locomotory needs and the
transition to a carangiform swimming mode at approximately 8 mm SL.
In a third part of this chapter on ontogeny, the development of the cranium is
described. For the study of the chondrocranium, serial sections were digitized
and used for 3D-reconstructions. Development overall follows the typical
siluriform trends in chondrocranial development. Even the low complexity of the
chondrocranium at hatching fits the trend observed in other siluriforms, although
other studies show loricarioid hatchlings to generally have more complex
chondrocrania. As in A. cf. triradiatus, a true commissura lateralis is present,
which differs from the situation described for other siluriforms. The most striking
differences found in comparison to other catfishes, however, involve the
branchial basket, which arises as a single element in C. aeneus, with a further
differentiation from the middle arches on in both a rostral and caudal direction.
With the description of development of the cranial bones, this part continues to
139
6 SUMMARY & SAMENVATTING
describe the development of the cranium. Development of this osteocranium is
found to occur in two distinct phases. In a first phase, several new bony
elements, all of dermal origin and related to feeding, appear shortly after yolk
depletion (4.4 mm SL). After this, in between 5 and 8 mm SL, developmental
priorities shift to size increase of the cartilaginous skull and no new bony
elements appear. Finally, a second phase of osteogenesis is present from 8-18
mm SL, in which all remaining dermal and perichondral bones appear.
In the fourth part, ontogeny of the cranial muscle system is dealt with.
Comparison of these results to those of A. cf. triradiatus reveals a high degree of
similarity. This way, both species lack a m. protractor hyoidei, and the m.
intermandibularis posterior is divided into two different parts which have partly
obtained a novel function in A. cf. triradiatus. A similar increase in complexity in
this species is found in the dorsal constrictor of the hyoid muscle plate. This
constrictor gives rise to the same muscles as in C. aeneus, but, in A. cf.
triradiatus, the m. levator and dilatator operculi later become hypertrophied. In
addition, in A. cf. triradiatus, the m. extensor tentaculi further differentiates into
two separate bundles (as opposed to a single muscle diverging posteriorly in C.
aeneus) and a loricariid neoformation is present called the m. levator tentaculi
(Geerinckx et al., subm.).
In contrast to the former three parts in this chapter, the fifth part does not
describe cranial ontogeny in C. aeneus, but focuses on the development of the
postcranial skeleton. Results indicate a possible shift in swimming mode at two
points during ontogeny. Observations on the ontogeny of the caudal fin skeleton
show that all of the cartilaginous elements in this fin develop between a SL of c.
5–6 mm, at which point also notochord flexion occurs, which corresponds to a
suggested change to burst swimming at 5 mm SL (Weihs, 1980; Verhagen,
2004). At 8 mm SL, the anal, adipose and pelvic fins separate from the larval
finfold and ossification in the vertebral column and caudal skeleton starts. Also,
development
of
the
anal
and
pelvic
fin
starts
and
ossification
of
the
scapulocoracoid plate and development of the proximal radials and fin spine in
the pectoral fin takes place. This all corresponds to a second shift in swimming
mode suggested to occur at that point, based on the analysis of early
development.
The sixth and final part of this chapter gives a brief overview of similar results
in a miniaturized congeneric species, Corydoras pygmaeus. Results show high
140
6 SUMMARY & SAMENVATTING
similarities in the overall pattern of development in both species with some
difference in the timing of appearance of ossification of different elements. In
conclusion, miniaturization in the body of C. pygmaeus was found to be the
result of paedomorphic events, including a lower growth rate and subsequent
postdisplacement of several traits. These paedomorphic events have, however,
besides the smaller size, not led to the retention of a large number of
paedomorphic traits.
5.
In the fifth chapter, the results of this dissertation are discussed in
relation to the evolution within the Loricarioidea.
In the first part, the ontogenies of C. aeneus, C. pygmaeus and A. cf.
triradiatus are briefly recapitulated and the theory of saltatory ontogeny, as
presented by Balon (1959, 1975, 1979, 1990) is applied to these results. This
shows that the ontogenies of both Corydoras species follow a saltatory path,
altering stabilized periods of quantitative morphogenesis and growth with brief
but distinct qualitative changes. The situation in A. cf. triradiatus, however,
substantially differs from this, showing no saltatory development. This was
related to the larger egg size (and the evolution of parental care) and a
subsequently longer prehatching period, which also probably led to the skipping
of a true larval stage during the species’ ontogeny. All this is hypothesized to
have ultimately facilitated the evolution of a highly specialized feeding apparatus,
given the increase in the time available for development of this specialized
morphology through a delay in yolk depletion.
In the second part, results presented in this dissertation and additional results
from literature are combined to apply the theory of Streelman & Danley (2003)
to the evolution of algae scraping and subsequent (?) adaptive radiation in
Loricarioidea in general and Loricariidae specifically. This leads to the conclusion
that the evolutionary basis of radiation in the Loricariidae was established by
various events of vicariance throughout the South American geological history. In
addition, a co-evolution of decoupling events in trophic morphology and sexual
dimorphism (with related parental care and increase in egg size) have ultimately
led to the extreme diversity found in the loricariid family only.
141
6 SUMMARY & SAMENVATTING
142
6 SUMMARY & SAMENVATTING
6.2 SAMENVATTING
Deze doctoraatsthesis geeft een gedetailleerde beschrijving van zowel de adulte
morfologie als een aantal aspecten van de ontogenie van Corydoras aeneus. Ter
vergelijking werd een summiere vergelijkbare studie van een geminiaturiseerde
soort uit hetzelfde genus, Corydoras pygmaeus, uitgevoerd. De gevonden
resultaten werden verder vergeleken met resultaten uit een gelijkaardige studie
uitgevoerd op de loricariide Ancistrus cf. triradiatus door Geerinckx (2006).
Vervolgens werden alle resultaten bediscussieerd in de context van de evolutie
van het algenschrapen in de superfamilie van de Loricarioidea en van de
adaptieve radiatie binnen de Loricariidae. In deze samenvatting wordt een korte
recapitulatie van de belangrijkste resultaten en conclusies per hoofdstuk
gepresenteerd.
1.
In de inleiding van deze verhandeling wordt het voorbeeld van de
adaptieve radiatie binnen de familie van de Loricariidae voorgesteld. Deze
Loricariidae behoren tot de superfamilie van de Loricarioidea, een superfamilie
waarin de verschillende families die er deel van uitmaken een trend naar een
stijgende morfologische complexiteit vertonen, waarbij verschillende elementen
werden ontkoppeld, daarbij de gelegenheid tot de evolutie van nieuwe structuren
en functies creërend (Schaefer & Lauder, 1986, 1996). In vertegenwoordigers
van de meer basale loricarioide familie Callichthyidae is een zuigmond of
algenschrapend voedingsmechanisme afwezig; maar de mond bevindt zich wel
reeds ventraal. Bij de Astroblepidae is zo’n zuigmond wel aanwezig is, hoewel
geen algivorie gekend is en uiteindelijk zijn bij vertegenwoordigers van de
Loricariidae zowel een zuigmond als vaak ook algivorie aanwezig. In deze context
was het doel van deze studie het bestuderen van de morfologie en vroege
ontogenie van een basale vertegenwoordiger van deze evolutieve lijn en om de
resultaten te vergelijken met deze van een zeer gespecialiseerde soort.
2.
In het tweede hoofdstuk worden het materiaal en de gebruikte
methoden opgesomd. In eerste instantie wordt de keuze van C. aeneus als
studieobject beargumenteerd, waarbij wordt aangehaald dat de familie waartoe
deze soort behoort een basale positie inneemt binnen de lijn naar een
143
6 SUMMARY & SAMENVATTING
gespecialiseerde morfologie binnen de Loricarioidea. Binnen deze familie wordt
de keuze van het genus en de soort geplaatst binnen de context van de mate
waarin literatuurgegevens voorhanden zijn. Vervolgens worden de verschillende
protocols m.b.t. het kweken en verzamelen van specimens weergegeven,
aangevuld met een lijst van de gebruikte specimens. In een tweede deel van dit
hoofdstuk worden de verschillende methoden uiteengezet, en, waar aangewezen,
worden gedetailleerd protocols meegegeven. De verschillende methoden gebruikt
zijn: opheldering en kleuring, dissecties en het maken van seriële coupereeksen
en 3D-reconstructies.
Het derde hoofdstuk beschrijft de adulte morfologie van C. aeneus
3.
en is onderverdeeld in twee delen.
Een eerste deel behandelt de osteologie van de soort en resultaten hiervan
worden vergeleken met deze van andere loricarioide vissen. Het merendeel van
deze resultaten beantwoordt aan het verwachte patroon, hoewel enkele
verschillen werden gevonden. Een van deze verschillen was de aanwezigheid van
een suprapreoperculair been, voorheen onbeschreven in Callichthyidae. Verder
worden
verschillende
bestaande
hypothesen
homologieproblemen
worden
bevestigd.
opgehelderd,
Zo
wordt,
waarbij
vaak
bijvoorbeeld,
de
het
samengesteld dorsolateraal been van de otische regio geïdentificeerd als het
posttemporo-pterotico-supracleithrum, terwijl het samengesteld been dat de
bodem van het neurocranium in de occipitale regio vormt wordt herkend als een
fusie van zowel de basi- als exoccipitale beenderen. De homologie van het
veronderstelde lacrimo-antorbitale en suprapreoperculare konden daarentegen in
dit deel van de studie niet worden bevestigd.
In een tweede deel wordt de musculatuur van de soort beschreven en
bediscussieerd. Gezien het belang van kaak- en operculaire spieren in de context
van ontkoppeling en het verwerven van nieuwe functies in relatie tot het
gespecialiseerde voedingsapparaat bij Loricariidae, werden homologieën van
deze spieren in detail bestudeerd. Bij C. aeneus worden zo binnen het
kaakspiercomplex A1OST, A2 en A3’ secties herkend en worden argumenten
aangereikt die de homologie van de m. retractor tentaculi met de A3’’
ondersteunen. Binnen het operculaire spiercomplex wordt, net zoals bij A. cf.
triradiatus, een sterk vergrote m. dilatator operculi aangetroffen en was de
oriëntatie van het operculo-hyomandibulair ligament verschoven. Daarnaast
144
6 SUMMARY & SAMENVATTING
werd bij beide soorten ook een verschuiving van de m. hyohyoidei abductor
vastgesteld, die daardoor als een adductor fungeerde.
4.
Het vierde hoofdstuk behandelt verschillende aspecten van de
ontogenie van zowel C. aeneus als C. pygmaeus.
In het eerste deel wordt de oppervlaktestructuur van de eitjes van C. aeneus
beschreven. Daarbij wordt een patroon aangetroffen dat uniek blijkt bij
teleosten. Het oppervlak was bezet met kleine uitsteeksels, sterk gelijkend op
filamenten die voor de vasthechting zorgen bij de eitjes van andere teleosten.
Daarnaast vertonen de eitjes ook een sterk vermogen tot vasthechting,
waarschijnlijk gerelateerd aan het feit dat van de soort gekend is dat ze in
snelstromend water leeft.
Het tweede deel van dit hoofdstuk geeft een overzicht van de vroege
ontwikkeling en de groeipatronen bij C. aeneus. Gebaseerd op externe
morfologie worden de verschillende stadia, zoals beschreven door Balon (1975),
in de vroege ontwikkeling geïdentificeerd. Na kippen, vanaf een SL van 3.5 mm,
is
een
vrijlevende
embryonale
fase
aanwezig,
gevolgd
door
de
proto-
pterygiolarvale fase (4.4-5.7 mm SL), de pterygiolarvale fase (5.7-14 mm SL) en
de juveniele periode. Daarnaast wordt een groeicurve gepresenteerd waarin
verschillende inflexiepunten werden gezocht, daarbij gebruik makend van een
stapsgewijze lineaire regressiemethode volgens de procedures van ‘regression
spline smoothing’. Op deze manier kon de groeicurve in zes fases met
verschillende groeisnelheid worden onderverdeeld. Bij aanvang is de hellingshoek
in de curve 0.05 tot aan 0.7 dph, waarna deze toeneemt tot 0.18 tot aan 4 dph,
en 0.36 tot 10 dph. Hierna wordt een maximum in groeisnelheid van 0.76 bereikt
tot aan de leeftijd van 24 dph, die daarna afneemt tot 0.47 tot 37 dph en
uiteindelijk nogmaals tot 0.36 daarna. Een gelijkaardige analyse werd ook
uitgevoerd op de groei van verschillende lichaamsdelen. De inflexiepunten die op
deze manier werden gevonden komen overeen met de verschillende gekende
‘key-events’ in de vroege ontwikkeling van teleosten. Voorbeelden daarvan zijn:
de transitie van endo- naar exogene voeding, de overgang naar de pterygiolarvale fase, waarbij prioriteiten verschuiven naar locomotie toe en de
transitie naar een carangiforme zwemstijl rond c. 8 mm SL.
In een derde deel van dit hoofdstuk over ontogenie wordt de ontwikkeling van
het cranium beschreven. Voor de studie van het chondrocranium werden seriële
145
6 SUMMARY & SAMENVATTING
coupereeksen
gedigitaliseerd
en
voor
3D-reconstructies
gebruikt.
De
ontwikkeling volgt de typisch siluriforme trends in chondrocranium ontwikkeling.
Zelfs de geringe complexiteit van het chondrocranium op het moment van
hatching past in de algemene trend die bij katvissen wordt waargenomen,
alhoewel andere studies bij loricarioide hatchlings in de regel een meer complex
chondrocranium aantonen. Zoals bij A. cf. triradiatus wordt ook hier een echte
commissura lateralis gevonden, wat verschilt van de situatie zoals die bij andere
Siluriformes wordt beschreven. Het meest markante verschil in vergelijking met
andere katvissen, daarentegen, heeft betrekking op de kieuwkorf, die als een
enkelvoudig element ontstaat bij C. aeneus, met een verdere differentiatie vanuit
de middelste bogen in zowel een rostrale als caudale richting. Met de
beschrijving van de ontwikkeling van de craniale beenderen, vervolgt dit
hoofdstuk de beschrijving van de ontwikkeling van het cranium. De ontwikkeling
van dit osteocranium blijkt in twee verschillende fasen te verlopen. In een eerste
fase ontwikkelden zich, kort na het verdwijnen van de dooier (4.4 mm SL),
verschillende
beenderen
die
allen
van
dermale
origine
en
aan
voeding
gerelateerd zijn. Na dit, tussen 5 en 8 mm SL, vershuiven de prioriteiten in de
ontwikkeling in de richting van groei van de kraakbeenschedel en worden geen
nieuwe benige elementen gevormd. Uiteindelijk is een tweede fase van
osteogenese aanwezig in specimens tussen 8 en 18 mm SL, waarbij alle overige
dermale en perichondrale beenderen worden gevormd.
In het vierde deel wordt de ontogenie van de craniale spieren behandeld. Een
vergelijking van deze resultaten met die van A. cf. triradiatus toont een grote
mate van gelijkenis. Zo ontbreekt bij beide soorten de m. protractor hyoidei en is
de m. intermandibularis posterior onderverdeeld in twee verschillende delen die
in A. cf. triradiatus deels een nieuwe functie hebben aangenomen. Een
vergelijkbare
verhoging
van
de
complexiteit
in
deze
soort
wordt
ook
teruggevonden in de dorsale constrictor van de hyoid spierplaat. Uit deze
constrictor worden dezelfde spieren gevormd als bij C. aeneus, waarbij, in een
latere fase bij A. cf. triradiatus, de m. levator en dilatator operculi hypertrofieert.
Daarnaast differentieert de m. extensor tentaculi bij A. cf. triradiatus verder in
twee aparte bundels (in tegenstelling tot een enkelvoudige spier die achteraan
vertakt bij C. aeneus) en is ook een loricariide neoformatie aanwezig die m.
levator tentaculi wordt genoemd (Geerinckx et al., subm.).
146
6 SUMMARY & SAMENVATTING
In tegenstelling tot de vorige drie delen uit dit hoofdstuk worden in het vijfde
deel geen elementen van de craniale ontogenie van C. aeneus behandeld, maar
wordt er gefocused op de ontwikkeling van het postcraniale skelet. Resultaten
hiervan wijzen in de richting van een mogelijke verschuiving in zwemstijl op twee
punten gedurende de ontogenie. Observaties met betrekking tot de ontogenie
van het staartskelet tonen aan dat alle kraakbenige elementen in deze fin zich
ontwikkelen bij een SL van c. 5-6 mm, hetzelfde punt als waarbij flexie in het
notochord optreedt, wat overeenkomt met een gesuggereerde verandering naar
‘burst swimming’ bij 5 mm SL (Weihs, 1980; Verhagen, 2004). Bij 8 mm SL,
scheiden zowel de anale, vet- en pelvische vinnen zich af van de larvale vin en
start de verbening in de wervelkolom. Daarnaast worden de eerste tekenen van
verbening gevonden in de scapulocoracoidplaat en beginnen de proximale
radialen en vinstekel in de borstvin zich te ontwikkelen. Dit alles komt overeen
met een tweede verschuiving in zwemstijl rond dit punt, die wordt gesuggereerd
op basis van de analyse van de vroege ontwikkeling.
Het zesde en laatste deel van dit hoofdstuk behandelt een kort overzicht van
een vergelijkbare studie bij een miniature soort uit hetzelfde genus, C.
pygmaeus. Resultaten hiervan tonen een hoge mate van overeenkomst in het
ontwikkelingspatroon van beide soorten, met enkele verschillen in de timing van
het verschijnen van verschillende elementen. Als besluit kan worden gesteld dat
miniaturisatie in het lichaam van C. pygmaeus het resultaat is van paedomorfe
processen, waaronder een
displacement
van
lagere groeisnelheid
verschillende
kenmerken.
en daaruitvolgende post-
Deze
paedomorfiën
hebben
weliswaar, behalve de kleinere afmetingen, niet geleid tot de retentie van een
hoog aantal paedomorfe kenmerken.
5.
In het vijfde hoofdstuk worden de resultaten van deze verhandeling
bediscussieerd in relatie tot de evolutie binnen Loricarioidea.
In een eerste deel worden de ontogenetische patronen van zowel C. aeneus, C.
pygmaeus en A. cf. triradiatus kort hernomen en wordt de theorie van de
‘saltatory ontogeny’ van Balon (1959, 1975, 1979, 1990) toegepast op deze
resultaten. Dit toont aan dat een dergelijke ‘saltatory ontogeny’ inderdaad
aanwezig was bij beide Corydoras-soorten, waarbij stabiele periodes van
kwantitatieve
morfogenese
worden
afgewisseld
met
korte
maar
distincte
kwalitatieve veranderingen. De situatie in A. cf. triradiatus, daarentegen,
147
6 SUMMARY & SAMENVATTING
verschilt hier in grote mate van, gezien hier geen ‘saltatory’ ontogenie wordt
waargenomen. Dit wordt gerelateerd aan een groter ei (en de evolutie van
broedzorg) en een daaruitvolgende langere periode voorafgaand aan kippen, wat
daarnaast ook leidde tot het overslaan van een echt larvaal stadium bij de soort.
Hieruit volgt de hypothese dat al deze verschillen uiteindelijk de evolutie van een
zeer gespecialiseerd voedingsapparaat hebben gefaciliteerd, gezien de verhoging
van de tijd die beschikbaar is voor de ontwikkeling van deze structuren door het
uitstellen van de dooierabsorptie.
In een tweede deel worden de resultaten van deze verhandeling en bijkomende
resultaten uit de literatuur gebundeld in een toepassing van de theorie van
Streelman
&
Danley
(2003)
op
de
evolutie
van
het
algenschrapen
en
daaruitvolgende (?) adaptieve radiatie bij Loricarioidea in het algemeen en meer
specifiek bij de Loricariidae. De algemene conclusie hierbij is dat de evolutieve
basis van de radiatie bij de Loricariidae moet worden gezocht in de verschillende
gebeurtenissen die tot periodes van vicariantie hebben geleid gedurende de
Zuid-Amerikaanse geologische geschiedenis. Daarbij komt dat een co-evolutie
van ontkoppeling van structuren gerelateerd aan de voedingsmorfologie en het
seksueel dimorfisme (met de daaraan gerelateerde broedzorg en grotere
dooieromvang) uiteindelijk tot de extreme diversiteit, die enkel bij Loricariidae
wordt gevonden, hebben geleid.
148
7
References
7 REFERENCES
7. REFERENCES
Adriaens D. 2003. Feeding mechanisms. In: Arratia G, Kapoor AS, Chardon M, Diogo R,
editors. Catfishes, vol. 1. Enfield, NH: Science Publishers. p 221-248.
Adriaens D, Verraes W. 1994. On the functional significance of the loss of the interhyal
during ontogeny in Clarias gariepinus Burchell, 1822 (Teleostei, Siluroidei). Belg J
Zool 124:139–155.
Adriaens D, Verraes W. 1996. Ontogeny of cranial musculature in Clarias gariepinus
(Siluroidei: Clariidae): The adductor mandibulae complex. J Morph 229:255-269.
Adriaens D, Verraes W. 1997a. Ontogeny of the maxillary barbel muscles in Clarias
gariepinus (Siluroidei: Clariidae), with some notes on the palatine-maxillary
mechanism. J Zool 241:117-133.
Adriaens D, Verraes W. 1997b. The ontogeny of the chondrocranium in Clarias
gariepinus: trends in siluroids. J Fish Biol 50:1221–1257.
Adriaens D, Verraes W. 1997c. Ontogeny of the hyoid musculature in the African catfish,
Clarias gariepinus (Burchell, 1822) (Siluroidei: Clariidae). Zool J Linn Soc 121:105–
128.
Adriaens D, Verraes W. 1997d. Ontogeny of the suspensorial and opercular muscles in
Clarias gariepinus (Siluroidei: Clariidae), and the consequences for respiratory
movements. Neth J Zool 47:61-89.
Adriaens D, Verraes W. 1998. Ontogeny of the Osteocranium in the African Catfish,
Clarias gariepinus Burchell (1822) (Siluriformes : Clariidae): Ossification sequence as
a response to functional demands. J Morph 235:183–237.
Adriaens D, Verraes W. 2002. An empirical approach to study the relation between
ontogeny, size and age using geometric morphometrics. In: Aerts P, D’Août K, Herrel
A, Van Damme R, editors. Topics in functional and ecological vertebrate morphology.
Maastricht: Shaker Publishing. p 293-324.
Adriaens D, Vandewalle P. 2003. Embryonic and larval development in catfishes. In:
Arratia G, Kapoor AS, Chardon M, Diogo R, editors. Catfishes, vol. 1. Enfield, NH:
Science Publishers. p 639–666.
Adriaens D, Verraes W, Taverne L. 1997. The cranial lateral-line system in Clarias
gariepinus (Burchell, 1822) (Siluroidei: Clariidae): morphology and development of
canal related bones. Eur J Morphol 35:181-208.
Adriaens D, Aerts P, Verraes W. 2001. Ontogenetic shift in mouth opening mechanisms in
a catfish (Clariidae, Siluriformes): a response to increasing functional demands. J
Morphol 247:197-216.
Alberch P, Gouls SJ, Oster GF, Wake DB. 1979. Size and shape in ontogeny and
phylogeny. Paleobiology 5:296-317.
Albertson RC, Streelman JT, Kocher TD. 2003. Directional selection has shaped the oral
jaws of lake malawi cichlid fishes. Proc Natl Acad Sci USA 100:5252-5257.
Albertson RC, Streelman JT, Kocher TD, Yelick PC. 2005. Integration and evolution of the
cichlid mandible: the molecular basis of alternate feeding strategies. Proc Natl Acad
Sci USA 102:16287-16292.
Alexander RM. 1965. Structure and function in the catfish. J Zool 148:88-152.
149
7 REFERENCES
Alfaro M, Westneat MW. 1999. Motor patterns of herbivorous feeding: electromyographic
analysis of biting in the parrotfishes Cetoscarus bicolor and Scarus iserti. Brain Behav
Evol 54:205–222
Allis E. 1897. The cranial muscles and cranial and first spinal nerves in Amia calva. J
Morph 12:487-809.
Andrade DV, Abe AS. 1997. Foam nest production in the armoured catfish. J Fish Biol
50:665-667.
Aquino AE. 1994. Secondary sexual dimorphism of the dermal skeleton in two species of
the hypoptopomatine genus Otocinclus (Siluriformes: Loricariidae). Ichthyol Explor
Freshw 5:217-222.
Aquino AE, Schaefer SA. 2002. Revision of Oxyropsis Eigenmann and Eigenmann, 1889
(Siluriformes, Loricariidae). Copeia 2002:374-390.
Aranha JMR, Takeuti DF, Yoshimura TM. 1998. Habitat use and food partitioning of the
fishes in a coastal stream of Atlantic forest, Brazil. Rev Biol Trop 46:951-959.
Araujo-Lima CARM. 1994. Egg size and larval development in Central Amazonian fish. J
Fish Biol 44:371-389.
Armbruster JW. 2004. Phylogenetic relationships of the suckermouth armoured catfishes
(Loricariidae) with emphasis on the Hypostominae and the Ancistrinae. Zool J Linn
Soc 141:1-80.
Arratia G. 1983. The caudal skeleton of ostariophysan fishes (Teleostei): intraspecific
variation in Trichomycteridae (Siluriformes). J Morph 177:213-229.
Arratia G. 1987. Description of the primitive family Diplomystidae (Siluriformes,
Teleostei, Pisces): morphology, taxonomy and phylogenetic implications. Bonn Zool
Monogr 24:11-20.
Arratia G. 1990. Development and diversity of the suspensorium of trichomycterids and
comparison with loricarioids (Teleostei: Siluriformes). J Morphol 205:193-218.
Arratia G. 1992. Development and variation of the suspensorium of the primitive
catfishes (Teleostei: Ostariophysi) and their phylogenetic relationships. Bonn Zool
Monogr 32:1-148.
Arratia G. 2003a. Catfish head skeleton: an overview. In: Kapoor AS, Arratia G, Chardon
M, Diogo R, editors. Catfishes. Enfield, NH: Science Publishers. p 3-46.
Arratia G. 2003b. The Siluriform postcranial skeleton: an overview. In: Kapoor AS,
Arratia G, Chardon M, Diogo R, editors. Catfishes. Enfield, NH: Science Publishers. p
121-157.
Arratia G, Schultze H-P. 1990. The urohyal: development and homology within
osteichthyans. J Morphol 203:247-282.
Arratia G, Huaquín L. 1995. Morphology of the lateral line system and of the skin of
diplomystid and certain primitive loricarioid catfishes and systematic and ecological
considerations. Bonn Zool Monogr 36:5-110.
Bailey RM, Baskin JN. 1976. Scoloplax dicra, a new armored catfish from the Bolivian
Amazon. Occ Pap Mus Zool, University of Michigan 674.
Bailey K, Houde ED. 1989. Predation on eggs and larvae of marine fish and the
recruitment problem. Adv Mar Biol 25:1–83.
Ballantyne FM. 1930. Notes on the development of Callichthys littoralis. Trans R Soc
Edinb 56:437–466.
Balon EK. 1959. Die embryonale und larvale Entwicklung der Donauzope (Abramis
ballerus subsp.). Biologické práce 5:1-87.
150
7 REFERENCES
Balon EK. 1971. The intervals of fish development and their terminology (a review and
proposals). Věst Čs Spol Zool 35:1-8.
Balon EK. 1975. Terminology of intervals in fish development. J Fish Res Board Can
32:1663-1670.
Balon EK. 1979. The theory of saltation and its application in the ontogeny of fishes:
steps and thresholds. Environ Biol Fish 4:97-101.
Balon EK. 1984. Reflections of some decisive events in the early life of fishes. T Am Fish
Soc 113:178-185.
Balon EK. 1986. Types of feeding in the ontogeny of fishes and the life-history model.
Environ Biol Fish 16:11-24.
Balon EK. 1990. Epigenesis of an epigeneticist: the development of some alternative
concepts on the early ontogeny and evolution of fishes. Guelph Ichthyol Rev 1:1-42.
Balon EK. 1999. Alternative ways to become a juvenile or a definitive phenotype (and on
some persisting linguistic offences). Environ Biol Fish 56:17-38.
Bamford TW. 1948. Cranial development of Galeichthys felis. Proc Zool Soc Lond
118:364-391.
Baskin JN. 1972. Structure and relationships of the Trichomycteridae. New York:
Unpublished PhD thesis.
Bellwood DR. 1994. A phylogenetic study of the parrotfishes, family Scaridae (Pisces:
Labroidei), with a revision of genera. Rec Aust Mus 20:1–86.
Bellwood DR, Choat JH. 1990. A functional analysis of grazing in parrotfishes (family
Scaridae): the ecological implications. Environ Biol Fish 28:189-214.
Berra TM. 2001. Freshwater fish distribution. London: Academic Press.
Bertmar G. 1959. On the ontogeny of the chondral skull in Characidae, with a discussion
on the chondrocranial base and the visceral chondrocranium in fishes. Acta Zool
40:203–364.
Bhargava HN. 1958. The development of the chondrocranium of Mastacembelus armatus
(Cuv. and Val.). J Morph 102:401–426.
Blaxter JHS. 1986. The development of sense organs and behavior in teleost larvae with
special reference to feeding and predator avoidance. T Am Fish Soc 115:98-114.
Bock WJ, Shear CR. 1972. A staining method for gross dissection of vertebrate muscles.
Anat Anz 130:222-227.
Britto MR. 1998. Two new species of the genus Aspidoras (Siluriformes: Callichthyidae)
from Central Brazil. Ichthyol Explor Freshw 8:359-368.
Britto MR. 2000. Aspidoras depinnai (Siluriformes: Callichthyidae): a new species from
northeastern Brazil. Copeia 2000:1048-1055.
Britto MR. 2003. Phylogeny of the subfamily Corydoradinae Hoedeman, 1952
(Siluriformes: Callichthyidae), with a definition of its genera. Proc Acad Nat Sci Phil
153:119-154.
Britto MR, Castro RMC. 2002. New corydoradine catfish (Siluriformes: Callichthyidae)
from the Upper Paraná and Săo Francisco: the sister group of Brochis and most of
Corydoras species. Copeia 2002:1006-1015.
Buckel JA, Steinberg ND, Conover DO. 1995. Effects of temperature, salinity, and fish
size on growth and consumption of juvenile bluefish. J Fish Biol 47:697-706.
Burgess WE, Quinn J. 1992. Colored atlas of miniature catfish. New Yersey: T.F.H.
Publications.
151
7 REFERENCES
Chardon M, Parmentier E, Vandewalle P. 2002. Morphology, development and evolution
of the Weberian apparatus in catfish. In: Kapoor AS, Arratia G, Chardon M, Diogo R,
editors. Catfishes. Enfield, NH: Science Publishers. p 71–119.
Coburn MM, Grubach PG. 1998. Ontogeny of the Weberian apparatus in the armored
catfish (Siluriformes: Callichthyidae). Copeia 1998:301-311.
Covain R, Fisch-Muller S. 2007. The genera of the Neotropical armored catfish subfamily
Loricariinae (Siluriformes: Loricariidae): a practical key and synopsis. Zootaxa
1462:1-40.
Coyne JA, Orr HA. 2004. Speciation. Sunderland, MA: Sinauer Associates.
Daget J. 1964. Le crâne des téléostéens, I. Origine et mise en place des tissus
squelettogènes chez l’embryon. Mésomésenchyme et ectomésenchyme. Mém Mus
Natl Hist Nat Sér A Zool 31:164-340.
Darwin C. 1859. On the origin of species by means of natural selection, or preservation
fo favored races in the struggle for life. London: John Murray.
Darwin C. 1871. The descent of man and selection in relation to sex. New York: Random
House, The Modern Library.
de Beer GR. 1927. The early development of the chondrocranium of Salmo fario. Q J
Microsc Sci 71:259–312.
de Beer GR. 1937. Studies on the vertebrate head. Oxford: Clarendon Press.
Delariva RL, Agostinho AA. 2001. Relationship between morphology and diets of six
neotropical loricariids. J Fish Biol 58:832-847.
de Pinna MCC. 1993. Higher-level phylogeny of Siluriformes, with a new classification of
the Order (Teleostei, Ostariophysi). New York: Unpublished PhD thesis.
de Pinna MCC. 1998. Phylogenetic relationships of Neotropical Siluriformes (Teleostei:
Ostariophysi): historical overview and synthesis of hypotheses. In: Malabarba LR,
Reis RE, Vari RP, Lucena ZM, Lucena CAS, editors. Phylogeny and classification of
Neotropical fishes. Porto Alegre: Edipucrs. p 279-330.
de Pinna MCC. 2003. Family Nematogenyidae. In: Reis RE, Kullander SO, Ferraris CJ,
editors. Check list of the freshwater fishes of South and Central America. Porto
Alegre: Edipucrs. p 268-269.
de Pinna MCC, Wosiacki W. 2003. Family Trichomycteridae. In: Reis RE, Kullander SO,
Ferraris CJ, editors. Check list of the freshwater fishes of South and Central America.
Porto Alegre: Edipucrs. p 270-290.
de Pinna MCC, Ng HH. 2004. The second ural centrum in siluriformes and its implication
for the monophyly of superfamily sisoroidea (Teleostei, Ostariophysi). Am Mus Novit
3437:1–23.
De Schepper N, Adriaens D, Teugels GG, Devaere S, Verraes W. 2004. Intraspecific
variation in the postcranial skeleton morphology in African clariids: a case study of
extreme phenotypic plasticity. Zool J Linn Soc 140:437–446.
Diogo R. 2005. Morphological evolution, aptations, homoplasies, constraints and
evolutionary trends. Catfishes as a case study on general phylogeny and
macroevolution. Plymouth: Science Publishers.
Diogo R, Chardon M. 2000. Homologies among different adductor mandibulae sections of
teleostean fishes, with special regard to catfishes (Teleostei: Siluriformes). J Morph
243:193-208.
Diogo R, Vandewalle P. 2003. Catfishes as a case study for discussions on general
evolution: the importance of functional uncouplings in morphological macroevolution.
Eur J Morph 41:139-148.
152
7 REFERENCES
Diogo R, Oliveira C, Chardon M. 2001. On the osteology and myology of catfish pectoral
girdle, with a reflection on catfish (Teleostei: Siluriformes) plesiomorphies. J Morphol
249:100-125.
Diogo R, Chardon M, Vandewalle P. 2006. On the osteology and myology of the cephalic
region and pectoral girdle of Nematogenys inermis (Ghichenot, 1848), with comments
on the autapomorphies and phylogenetic relationships of the Nematogenyidae
(Teleostei : Siluriformes). Belg J Zool 136:15-24.
Edgeworth FH. 1928. The development of some of the cranial muscles of ganoid fishes.
Phil Trans R Soc (B) 217:39-89.
Edgeworth FH. 1935. The cranial muscles of vertebrates. Cambridge: University Press.
Eschmeyer WN. 1998. Catalog of fishes. San Francisco: California Academy of Sciences.
Faustino M, Power DM. 2001. Osteologic development of the viscerocranial skeleton in
sea bream: alternative ossification strategies in teleost fish. J Fish Biol 58:537–572.
Ferraris CJ. 2007. Checklist of catfishes, recent and fossil (Osteichthyes: Siluriformes),
and catalogue of siluriform primary types. Auckland: Magnolia Press.
Fink SV, Fink WL. 1981. Interrelationships of the ostariophysan fishes (Teleostei). Zool J
Linn Soc 72:297-353.
Fink SV, Fink WL. 1996. Interrelationships of ostariophysan fishes (Teleostei). In:
Stiassny MLJ, Parenti LR, Johnson GD, editors. Interrelationships of fishes. London:
Academic Press. p 209-249.
Friedman JH, Silverman BW. 1989. Flexible parsimonious smoothing and additive
modeling. Technometrics 31:3-21.
Friel JP, Vigliotta TR. 2006. Synodontis acanthoperca, a new species from the Ogôoué
river system, Gabon with comments on spiny ornamentation and sexual dimorphism
in mochokid catfishes (Siluriformes: Mochokidae). Zootaxa 1125:45-56.
Froese R, Pauly D. 2006. Fishbase. www.fishbase.org.
Fuiman LA. 1983. Growth gradients in fish larvae. J Fish Biol 23:117-123.
Fuiman LA, Poling KR, Higgs DM. 1998. Quantifying developmental progress for
comparative studies of larval fishes. Copeia 1998:602-611.
Fujita K. 1992. Ontogeny of the caudal skeleton in the clariid catfish Clarias batrachus.
Jpn J Ichthyol 38:430–432.
Fukuhara O. 1992. Study on the development of functional morphology and behaviour of
the larvae of eight commercially valuable teleost fishes. Contrib Fish Res Japan Sea
Block 25:1–122.
Fuller IAM. 2001. Breeding corydoradine catfishes. Kidderminster: Ian Fuller Enterprises.
Futuyma DJ, Mayer GC. 1980. Non-allopatric speciation in animals. Syst Zool 29:254271.
Gavaia PJ, Sarasquete C, Cancele ML. 2000. Detection of mineralized structures in early
stages of development of marine Teleostei using a modified alcian blue-alizarin red
double staining technique for bone and cartilage. Biotech Histochem 75:79-84.
Geerinckx T. 2006. Ontogeny and functional morphology of a highly specialized trophic
apparatus: a case study of neotropical suckermouth armoured catfishes (Loricariidae,
Siluriformes). Unpublished PhD-thesis, Ghent University.
Geerinckx T, Adriaens D. 2006. The erectile cheek-spine apparatus in the bristlenose
catfish Ancistrus (Loricariidae, Siluriformes), and its relation to the formation of a
secondary skull roof. Zoology 109:287-299.
Geerinckx T, Adriaens D. 2007. Ontogeny of the intermandibular and hyoid musculature
153
7 REFERENCES
in the suckermouth armoured catfish
Siluriformes). Anim Biol 57:339-357.
Ancistrus
cf.
triradiatus
(Loricariidae,
Geerinckx T, Adriaens D. Ontogeny of the suspensorial and opercular musculature in the
suckermouth armoured catfish Ancistrus cf. triradiatus (Loricariidae, Siluriformes).
Zoomorphology in press.
Geerinckx T, Brunain M, Adriaens D. 2005. Development of the chondrocranium in the
suckermouth armored catfish Ancistrus cf. triradiatus (Loricariidae, Siluriformes). J
Morphol 266:331–355.
Geerinckx T, Brunain M, Adriaens D. 2007. Development of the osteocranium in the
suckermouth armored catfish Ancistrus cf. triradiatus (Loricariidae, Siluriformes). J
Morph 268:254-274.
Geerinckx T, Huysentruyt F, Adriaens D. Ontogeny of the jaw and maxillary barbel
musculature in the armoured catfish families Loricariidae and Callichthyidae
(Loricarioidea, Siluriformes), with a discussion on muscle homologies. Zool J Linn Soc
in press.
Gill TN. 1858. Synopsis of the fresh water fishes of the western portion of the island of
Trinidad, W I. Ann Lyc Nat Hist N Y 363-430
Gisbert E. 1999. Early development and allometric growth patterns in Siberian sturgeon
and their ecological significance. J Fish Biol 54:852-862.
Goodrich ES. 1958. The skull. In: Goodrich ES, editor. Studies on the structure and
development of vertebrates. New York: Dover Publications. p 230–395.
Gosline WA. 1940. A revision of the neotropical catfishes of the family Callichthyidae.
Stanf Ichthyol Bull 2:1-29.
Gosline WA. 1975. The cyprinid dermosphenotic and the subfamily Rasborinae. Occas
Pap Mus Zool, Univ Michig 673:1-13.
Gosline WA. 1986. Jaw muscle configuration in some higher teleostean fishes. Copeia
1986:705-713.
Gosline WA. 1989. Two patterns of differentiation in the jaw musculature of teleostean
fishes. J Zool 218:649-661.
Gosline WA. 1993. A survey of upper jaw musculature in higher teleostean fishes. Occas
Pap Mus Zool, Univ Michig 724:1-26.
Grande L. 1987. Redescription of †Hypsidoris farsonensis (Teleostei: Siluriformes), with a
reassessment of its phylogenetic relationships. J Vert Paleontol 7:24-54.
Grande L, de Pinna MCC. 1998. Description of a second species of the catfish †Hypsidoris
and a reevaluation of the genus and the family †Hypsidoridae. J Vert Paleontol
18:451-474.
Grandel H, Schulte-Merker S. 1998. The development of the paired fins in the Zebrafish
(Danio rerio). Mech Develop 79:99-120.
Grant PR. 2000. R.C.L. Perkins and evolutionary radiations on islands. Oikos 89:195–
201.
Grant PR, Grant BR. 2002. Unpredictable evolution in a 30-year study of Darwin's
finches. Science 296:707-711.
Haffer J. 1982. General aspects of the refuge theory. In: Prance GT, editor. Biological
Diversification in the Tropics. New York, NY: Colombia University Press. p 6-24.
Hanken J. 1993. Adaptation of bone growth to miniaturization of body size. In: Hall BK,
editor. Bone, Vol. 7: Bone growth – B. Boca Raton, FL: CRC Press. p 74-104.
154
7 REFERENCES
Hanken J, Wassersug R. 1981. The visible skeleton. A new double-stain technique reveals
the native of the "hard" tissues. Funct Photogr 16:22-26.
Hanken J, Wake DB. 1993. Miniaturization of body size: organismal consequences and
evolutionary significance. Annu Rev Ecol Syst 24:501-519.
Harrington RWJ. 1955. The osteocranium of the American cyprinid fish, Notropis
bifrenatus, with an annotated synonymy of teleost skull bones. Copeia 4:267–291.
Hoedeman JJ. 1960a. Studies on callichthyid fishes (3): Notes on the development of
Callichthys (1) (Pisces: Siluriformes). Bull Aquat Biol 1:53-68.
Hoedeman JJ. 1960b. Studies on callichthyid fishes: (4) Development of the skull of
Callichthys and Hoplosternum (1) (Pisces - Siluriformes). Bull Aquat Biol 1:73–84.
Hoedeman JJ. 1960c. Studies on callichthyid fishes: (5) Development of the skull of
Callichthys and Hoplosternum (2) (Pisces - Siluriformes). Bull Aquat Biol 2:21-36.
Hoedeman JJ. 1960d. Studies on callichthyid fishes (6): The axial skeleton of Callichthys
and Hoplosternum (Pisces: Siluriformes). Bull Aquat Biol 2:37-44.
Hostache G, Mol JH. 1998. Reproductive biology of the neotropical armoured catfish
Hoplosternum littorale (Siluriformes – Callichthyidae): a synthesis stressing the role
of the floating bubble nest. Aquat Living Resour 11:173-185.
Howes GJ. 1983. The cranial muscles of loricarioid catfishes, their homologies and value
as taxonomic characters. Bull Bri Mus Nat Hist 45:309-345.
Howes GJ, Teugels GG. 1989. Observations on the ontogeny and homology of the
pterygoid bones in Corydoras paleatus and some other catfishes. J Zool 219:441-456.
Hurley IA, Mueller RL, Dunn KA, Schmidt EJ, Friedman M, Ho RK, Prince VE, Yang ZH,
Thomas MG, Coates MI. 2007. A new time-scale for ray-finned fish evolution. Proc
Roy Soc B-Biol Sci 1609:489-498.
Huysseune A. 1989. Morphogenetic aspects of the pharyngeal jaws and neurocranial
apophysis in postembryonic Astatotilapia elegans (Trewavas, 1933) (Teleostei:
Cichlidae). Mededelingen van de Koninklijke Academie voor Wetenschappen, Letteren
en Schone Kunsten van België 51:11–35.
Huysseune A, Sire J-Y. 1997. Structure and development of teeth in three armoured
catfish, Corydoras aeneus, C. arcuatus and Hoplosternum littorale (Siluriformes,
Callichthyidae). Acta Zool 78:69-84.
Isbrücker IJH. 1981. Revision of Loricaria Linnaeus, 1758 (Pisces, Siluriformes,
Loricariidae). Beaufortia 31:51-96.
Isbrücker IJH, Nijssen H. 1973. Two new species of the callichthyid catfish genus
Corydoras from Brazil (Pisces, Siluriformes, Callichthyidae). Beaufortia 21:1-7.
Isbrücker IJH, Nijssen H. 1992a. Corydoras breei, a new species of callichthyid catfish
from the Corantijn River basin in Surinam (Pisces, Siluriformes, Callichthyidae).
Beaufortia 43:9-14.
Isbrücker IJH, Nijssen H. 1992b. Sexualdimorphismus bei Harnischwelsen (Loricariidae).
Odontoden, Zähne, Lippen, Tentakel, Genitalpapillen und Flossen. DATZ-Sonderheft
(Harnischwelse) 19-33.
Jardine D, Litvak MK. 2003. Direct yolk sac volume manipulation of zebrafish embryos
and the relationship between offspring size and yolk sac volume. J Fish Biol 63:388–
397.
Jarvik E. 1980. Basic structure and evolution of Vertebrates. vol. 1. London: Academic
Press.
155
7 REFERENCES
Kavanagh KD, Alford RA. 2003. Sensory and skeletal development and growth in relation
to the duration of the embryonic and larval stages in damselfishes (Pomacentridae).
Biol J Linn Soc 80:187-206.
Kim I, Park J. 1996. Adhesive membranes of oocyte in four loaches (Pisces: Cobitidae) of
Korea. Korean J Zool 39:198-206.
Kindred JE. 1919. The skull of Amiurus. Ill Biol Monogr 5:1–120.
Klingenberg CP. 1998. Heterochrony and allometry: the analysis of evolutionary change
in ontogeny. Biol Rev 73:79-123.
Kobayakawa M. 1992. Comparative morphology and development of bony elements in
the head region in three species of Japanese catfishes (Silurus: Siluridae;
Siluriformes). Japan J Ichthyol 39:25–36.
Kohda M, Tanimura M, Kikue-Nakamura M, Yamagishi S. 1995. Sperm drinking by female
catfishes: a novel mode of insemination. Environ Biol Fish 42:1-6.
Kohda M, Yonebayashi K, Nakamura M, Ohnishi N, Seki S, Takahashi D, Takeyama T.
2002. Male reproductive success in a promiscuous armoured catfish Corydoras
aeneus (Callichthyidae). Environ Biol Fish 63:281-287.
Kolm N, Ahnesjö I. 2005. Do egg size and parental care coevolve in fishes? J Fish Biol
66:1499-1515.
Koumoundouros G, Divanach P, Kentouri M. 1999. Ontogeny and allometric plasticity of
Dentex dentex (Osteichthyes: Sparidae) in rearing conditions. Mar Biol 135:561–572.
Koumoundouros G, Divanach P, Kentouri M. 2001a. Osteological development of Dentex
dentex (Osteichthyes: Sparidae): dorsal, anal, paired fins and squamation. Mar Biol
138:399-406.
Koumoundouros G, Sfakianakis DG, Maingot E, Divanach P, Kentouri M. 2001b.
Osteological development of the vertebral column and of the fins in Diplodus sargus
(Teleostei: Perciformes: Sparidae). Mar Biol 139:853-862.
Kováč V, Copp GH. 1999. Prelude: looking at early development in fishes. Environ Biol
Fish 56:7-14.
Kramer DL, McClure M. 1980. Aerial respiration in the catfish, Corydoras aeneus
(Callichthyidae). Can J Zool 58:1984-1991.
Kramer DL, McClure M. 1981. The transit cost of aerial respiration in the catfish
Corydoras aeneus (Callichthyidae). Physiol Zool 54:189-194.
Kramer DL, Braun EA. 1983. Short-term effects of food availability on air-breathing
frequency in the fish Corydoras aeneus (Callichthyidae). Can J Zool 61:1964-1967.
Kryzhanovsky SG, Disler NN, Smirnova EN. 1953. Ecomorphological principles of
development in percids.Trudy Instituta Morphologii Zhivotnych Severtsova 10:3-138.
Laale HW. 1980. The perivitelline space and egg envelopes of bony fishes: a review.
Copeia 1980:210-226.
Lack D. 1947. Darwin’s Finches. Cambridge: University Press.
Lakshmi K, Srinivasa Rao K. 1989. Postcranial skeleton of the marine catfish Arius
tenuispinis (Osteichthyes: Siluriformes, Ariidae). J Morph 202:361-377.
Langille RM, Hall BK. 1987. Development of the head skeleton of the Japanese Medaka,
Oryzas latipes (Teleostei). J Morphol 193:135–187.
Lassala MDP, Renesto E. 2007. Reproductive strategies and genetic variability in tropical
freshwater fish. Genet Mol Biol 30:690-697.
Lauder GV. 1981. Form and function: structural analysis in evolutionary morphology.
Paleobiology 7:430-442.
156
7 REFERENCES
Lauder GV, Crompton AW, Gans C, Hanken J, Liem KF, Maier WP, Meyer A, Presley R,
Rieppel OC, Roth G, Schluter D, Zweers GA. 1989. Group report: Howe are feeding
systems integrated and how have evolutionary innovations been introduced? In:
Wake DB, Roth G, editors. Complex and organismal functions: integration and
evolution in vertebrates. New York: Wiley & Sons. p 97-115.
Lee TCM. 2002. On algorithms for ordinary least squares regression spline fitting: a
comparative study. J Stat Comput Sim 72:647-663.
Lewis LM, Lall SP. 2006. Development of the axial skeleton and skeletal abnormalities of
Atlantic halibut (Hippoglossus hippoglossus) from first feeding through
metamorphosis. Aquaculture 257:124–135.
Liem KF. 1973. Evolutionary strategies and morphological innovations: cichlid pharyngeal
jaws. Syst Zool 22:425-441.
Losos JB, Leal M, Glor RE, de Queiroz K, Hertz PE, Rodriguez Schettino L, Chamizo Lara
A, Jackman TR, Larson A. 2003. Niche lability in the evolution of a Caribbean lizard
community. Nature 424:542-545.
Lubosch W. 1938. Muskeln des Kopfes: Viscerale Muskulatur (Fortsetzung). B. Teleostier.
In: Bolk L, Göppert E, Kallius E, Lubosch W, editors. Handbuch der vergleichenden
Anatomie der Wirbeltiere. Berlin & Wien: Urban & Schwarzenberg. p 1011-1024.
Lujan NK, Armbruster JW, Sabaj M. 2007. Two new species of Pseudancistrus from
southern Venezuela (Siluriformes: Loricariidae). Ichthyol Explor Freshwaters 18:163174.
Lundberg JG. 1975. Homologies of the upper shoulder girdle and temporal region bones
in catfishes (Order Siluriformes), with comments on the skull of the Helogeneidae.
Copeia 1975:66-74.
Lundberg JG. 1982. The comparative anatomy of the toothless blindcat, Trogloglanis
pattersoni Eigenmann, with a phylogenetic analysis of the ictalurid catfishes. Misc
Publ Mus Zool, Univ Michig 163:1-85.
Lundberg JG. 1998. The temporal context for the diversification of Neotropical fishes. In:
Malabarba LR, Reis RE, Vari RP, Lucena ZM, Lucena CAS, editors. Phylogeny and
classification of Neotropical fishes. Porto Alegre: Edipucrs. p 49-68.
Lundberg JG, Baskin JN. 1969. The caudal skeleton of the catfishes, order Siluriformes.
Am Mus Nov 2398:1-19.
Lundberg JG, Marshall LG, Guerrero J, Horton B, Malabarba MCSL, Wesselingh F. 1998.
The stage for Neotropical fish diversification: a history of tropical South American
rivers. In: Malabarba LR, Reis RE, Vari RP, Lucena ZM, Lucena CAS, editors.
Phylogeny and classification of Neotropical fishes. Porto Alegre: Edipucrs. p 13-48.
Lundberg JG, Sullivan JP, Rodiles-Hernández R, Hendrickson DA. 2007. Discovery of
African roots for the Mesoamerican Chiapas catfish, Lacantunia enigmatica, requires
an ancient intercontinental passage. Proc Acad Nat Sci Phila 156:39-53.
Mabee PM, Trendler TA. 1996. Development of the cranium and paired fins in Betta
splendens (Teleostei: Percomorpha): Intraspecific variation and interspecific
comparisons. J Morph 227:249-287.
Machado-Allison A, Garcia C. 1986. Food habits and morphological changes during
ontogeny in three serrasalmin fish species of the Venezuelan floodplains. Copeia
1986:193-196.
Maisey JG. 2000. Continental break up and the distribution of fishes of Western
Gondwana during the Early Creataceous. Creataceous Res 21:281-314.
Mayr E. 1996. What is a species and what is not? Philos Sci 63:626-277.
Mayr E. 2001. What evolution is. New York: Basic Books.
157
7 REFERENCES
Mazzoni R, Caramaschi EP. 1995. Length and weight relationship and condition factor of
two species of Hypostomus Lacépède, 1803 (Siluroidei, Loricariidae) in the lower
Paraíba de Sul River, Rio de Janeiro, Brazil. Arq Biol Tecnol 38:135-140.
McElman JF, Balon EK. 1980. Early ontogeny of white sucker, Catostomus commersoni,
with steps of saltatory development. Env Biol Fish 5:191-224.
Mo T. 1991. Anatomy and systematics of Bagridae (Teleostei), and siluroid phylogeny.
Koenigstein: Koeltz Scientific Books.
Mol JH. 1996. Reproductive seasonality and nest-stie differentiation in three closely
related armoured catfishes (Siluriformes: Callichthyidae). Environ Biol Fish 45:363381.
Montoya-Burgos JI. 2003. Historical biogeography of the catfish genus Hypostomus
(Siluriformes: Loricariidae), with implications on the diversification of Neotropical
ichthyofauna. Mol Ecol 12:1855-1867.
Morin RP, Able KW. 1983. Patterns of geographic variation in the egg morphology of the
fundulid fish, Fundulus heteroclitus. Copeia 1983:726-740.
Myers GS. 1953. Habits and classification of Corydoras hastatus. The Aquarium Journal
24:268-270.
Nelson JA. 2002. Metabolism of three species of herbivorous loricariid catfishes: influence
of size and diet. J Fish Biol 61:1586-1599.
Nelson JS. 2006. Fishes of the world. 4th ed. Hoboken, New Jersey: John Wiley & Sons.
Ng HH, Bailey RM. 2006. Chiloglanis productus, a new species of suckermouth catfish
(Siluriformes: Mochokidae) from Zambia. Occas Pap Univ Mich Mus Zool 738:1-13.
Nijssen H. 1970. Revision of the Surinam catfishes of the genus Corydoras Lacepède,
1803 (Pisces, Siluriformes, Callichthyidae). Beaufortia 18:1-75.
Nijssen H, Issbrücker IJH. 1980. A review of the genus Corydoras Lacépède, 1803
(Pisces, Siluriformes, Callichthyidae). Bijdr Dierk 50:190-220.
Oliveira C, Almeida-Toledo LF, Mori L, Toledo-Filho SA. 1992. Extensive chromosomal
rearrangements and nuclear DNA content changes in the evolution of the armoured
catfishes genus Corydoras (Pisces, Siluriformes, Callichthyidae). J Fish Biol 40:419–
431.
Oliveira C, Almeida-Toledo LF, Mori L, Toledo-Filho SA. 1993. Cytogenetic and DNA
content studies of armoured catfishes of the genus Corydoras (Pisces, Siluriformes,
Callichthyidae) from the southeast coast of Brazil. Rev Bras Genét 16:617–629.
Orton GL. 1953. The systematics of vertebrate larva. Syst Zool 2:63-75.
Osse JWM. 1990. Form changes in fish larvae in relation to changing demands of
function. Neth J Zool 40:362-385.
Osse JWM, van den Boogaart JGM. 1995. Fish larvae, development, allometric growth,
and the aquatic environment. ICES Marine Science Symposia 201:21-34.
Osse JWM, van den Boogaart JGM, van Snik GMJ, van der Sluys L. 1997. Priorities during
early growth of fish larvae. Aquaculture 155:249-258.
Park J, Kim I. 2001. Fine structure of the oocyte envelopes of three related cobitid
species in the genus Ikookimia (Cobitidae). Ichthyol Res 48:71-75.
Paugy D, Roberts TR. 1992. Mochokidae. In: Lévêque C, Paugy D, Teugels GG, editors.
Faune des poissons d’eaux douces et saumâtres de l’Afrique de l’Ouest. Tervuren:
MRAC. Paris: ORSTOM. p 500-563.
Pascal M, Hostache G, Tessier C, Vallat P. 1994. Cycle de reproduction et fécondité de
l’atipa, Hoplosternum littorale (Siluriforme) en Guyane Française. Aquat Living Resour
158
7 REFERENCES
7:25-37.
Patzner RA, Glechner R. 1996. Attaching structures in eggs of native fishes. Limnologica
26:179-182.
Pavlov DA, Moksness E. 1997. Development of the axial skeleton in wolffish, Anarhichas
lupus (Pisces, Anarhichadidae), at different temperatures. Environ Biol Fish 49:401–
416.
Pedersen BH, Ugelstad I, Hjelmeland K. 1990. Effects of a transistory, low food supply in
the early life of larval herring (Clupea harengus) on mortality, growth and digestive
capacity. Mar Biol 107:61–66.
Pruzsinszky I, Ladich F. 1998. Sound production and reproductive behaviour of the
armoured catfish Corydoras paleatus (Callichthyidae). Environ Biol Fish 53:183-191.
Ramnarine IW. 1994. Larval culture, development and growth of the cascadu,
Hoplosternum littorale (Hancock 1828; Callichthyidae). Aquaculture 126:291-298.
Ramnarine IW. 1995. Induction of nest building and spawning in Hoplosternum littorale. J
Fish Biol 47:555-557.
Rapp Py-Daniel L. 2000. Tracking evolutionary trends in Siluriformes (Teleostei;
Ostariophysi). In: International congress on the biology of fish, Aberdeen. Evolution
Symposium Proceedings. University of Aberdeen: Aberdeen. p 57-72.
Regan CT. 1911. The classification of the teleostean fishes of the order Ostariophysi: 2.
Siluroidea. Ann Mag Nat Hist 8:553-577.
Reis RE. 1996. Callichthyidae. The tree of life web project. www.tolweb.org.
Reis RE. 1997. Revision of the neotropical catfish genus Hoplosternum (Ostariophysi:
Siluriformes: Callichthyidae), with the description of two new genera and three new
species. Ichthyol Explor Freshw 7:299-326.
Reis RE. 1998. Anatomy and phylogenetic analysis of the neotropical callichthyid
catfishes (Ostariophysi, Siluriformes). Zool J Linn Soc 124:105-168.
Reis RE. 2003. Family Callichthyidae. In: Reis RE, Kullander SO, Ferraris CJ, editors.
Check list of the freshwater fishes of South and Central America. Edipucrs: Porto
Alegre. p 291-309.
Reis RE, Schaefer SA. 1992. Eurycheilus pantherinus (Siluroidei: Loricariidae), a new
genus and species of Hypoptopomatinae from Southern Brazil. Copeia 1:215-223.
Retzer ME, Page LM. 1997. Systematics of the stick catfishes, Farlowella Eigenmann &
Eigenmann (Pisces, Loricariidae). Proc Acad Nat Sci Phila 147:33-88.
Riehl R. 1996. The ecological significance of the egg envelope in teleosts with special
reference to limnic species. Limnologica 26:183-189.
Riehl R. 1999. The micropyle of teleost fish eggs: morphological and functional aspects.
Proceedings of the 5th Indo-Pacific Fish Conference, Nouméa 589-599.
Riehl R, Götting KJ. 1974. Zu Struktur und Vorkommen der Mikropyle an Eizellen und
Eiern von Knochenfischen (Teleostei). Arch Hydrobiol 74:393-402.
Riehl R, Bless R. 1995. First report on egg deposition and egg morphology of the
endangered endemic Romanian perch. J Fish Biol 46:1086-1090.
Riehl R, Patzner RA. 1998. The modes of egg attachment in teleost fishes. Ital J Zool
65:415-420.
Ristovska M, Karaman B, Verraes W, Adriaens D. 2006. Early development of the
chondrocranium in Salmo letnica (Karaman, 1924) (Teleostei: Salmonidae). J Fish
Biol 68:458-480.
Rizzo E, Sato Y, Barreto BP, Godinho HP. 2002. Adhesiveness and surface patterns of
159
7 REFERENCES
eggs in neotropical freshwater teleosts. J Fish Biol 61:615-632.
Rodríguez MS, Miquelarena A. 2005. A new species of Rineloricaria (Siluriformes:
Loricariidae) from the Paraná and Uruguay River basins, Misiones, Argentina. Zootaxa
945:1-15.
Rojo AL. 1991. Dictionary of evolutionary fish osteology. Florida: CRC Press.
Schaefer SA. 1987. Osteology of Hypostomus plecostomus (Linnaeus) with a
phylogenetic analysis of the loricariid subfamilies (Pisces: Siluroidei). Contrib Sci
394:1-31.
Schaefer SA. 1988. Homology and evolution of the opercular series in the loricarioid
catfishes (Pisces: Siluroidei). J Zool 214:81-93.
Schaefer SA. 1990. Anatomy and relationships of the scoloplacid catfishes. Proc Acad Nat
Sci Phila 142:167-210.
Schaefer SA. 1997. The neotropical cascudinhos: systematics and biogeography of the
Otocinclus catfishes (Siluriformes: Loricariidae). Proc Acad Nat Sci Phila 148:1-120.
Schaefer SA. 2003. Family Astroblepidae. In: Reis RE, Kullander SO, Ferraris CJ, editors.
Check list of the freshwater fishes of South and Central America. Porto Alegre:
Edipucrs. p 312-317.
Schaefer SA, Lauder GV. 1986. Historical transformation of functional design:
evolutionary morphology of feeding mechanisms in loricarioid catfishes. Syst Zool
35:489-508.
Schaefer SA, Lauder GV. 1996. Testing historical hypotheses of morphological change:
biomechanical decoupling in loricarioid catfishes. Evolution 50:1661-1675.
Schaefer SA, Aquino AE. 2000. Postotic laterosensory canal and pterotic branch
homology in catfishes. J Morphol 246:212-227.
Schaefer SA, Weitzman SH, Britski HA. 1989. Review of the Neotropical catfish genus
Scoloplax (Pisces: Loricarioidea: Scoloplacidae) with comments on reductive
characters in phylogenetic analysis. Proc Acad Nat Sci Phila 141:181-211.
Schmidt RE. 2001. Loricaria cataphracta: parental care and description of early larvae.
Ichthyol Explor Freshwaters 12:235-240.
Seehausen O. 2004. Hybridization and adaptive radiation. Trends Ecol Evol 19:198-207.
Sfakianakis DG, Koumoundouros G, Divanach P, Kentouri M. 2004. Osteological
development of the vertebral column and of the fins in Pagellus erythrinus (L. 1758).
Temperature effect on the developmental plasticity and morpho-anatomical
abnormalities. Aquaculture 232:407–424.
Shelden FF. 1937. Osteology, myology, and probable evolution of the nematognath pevic
girdle. Ann NY Acad Sci 37:1-96.
Shiba Y, Sasaki Y, Enomoto T, Kanno Y. 1982. Microtubule formation in regenerating
terminal buds of the silurid fish, Corydoras aeneus. Dev Growth Differ 24:199-203.
Shimabukuro-Dias CK, Oliveira C, Reis RE, Foresti F. 2004. Molecular phylogeny of the
armored catfish family Callichthyidae (Ostariophysi, Siluriformes). Mol Phylogenet
Evol 32:152-163.
Simonović PD, Garner P, Eastwood EA, Kováč V, Copp G. 1999. Correspondence between
ontogenetic shifts in morphology and habitat use in minnow Phoxinus phoxinus.
Environ Biol Fish 56:117-128.
Simpson GG. 1953. The major features of evolution. New York, Columbia University
Press.
Sire JY, Huysseune A. 1996. Structure and development of the odontodes in an armoured
160
7 REFERENCES
catfish, Corydoras aeneus (Siluriformes, Callichthyidae). Acta Zool 77:51-72.
Srinivasachar HR. 1957a. Development of the skull in catfishes: Part I - Development of
chondrocranium in Silonia, Pangasius and Ailia (Schilbeidae). P Natl Inst Sci India
22B:335–356.
Srinivasachar HR. 1957b. Development of the skull in catfishes: Part II - Development of
chondrocranium in Mystus and Rita (Bagridae). Gegenbauers Morphol Jahrb 98:244–
261.
Srinivasachar HR. 1958a. Development of the skull in catfishes: Part IV - The
development of the chondrocranium in Arius jella Day (Ariidae) and Plotosus canius
Ham. (Plotosidae) with an account for their interrelationships. Gegenbauers Morphol
Jahrb 99:986-1016.
Srinivasachar HR. 1958b. Development of the skull in catfishes: Part V - Development of
the skull in Heteropneustes fossilis (Bloch). Proc Natl Inst Sci, India 24B:165-190.
Srinivasachar HR. 1959. Development of the skull in catfishes: Part III - The
Heteropneustes fossilis (Bloch)
development
of
the
chondrocranium
in
(Heteropneustidae) and Clarias batrachus (Linn) (Clariidae). Gegenbauers Morphol
Jahrb 101:373–405.
Stern DL, Elmen DJ. 1999. The developmental basis for allometry in insects.
Development 126:1091-1101.
Strauss RE. 1984. Allometry and functional feeding morphology in haplochromine
cichlids. In: Echelle AA, Kornfield I, editors. Evolution of Fish Species Flocks. Maine:
Orono Press. p 217-229.
Strauss RE. 1985. Evolutionary allometry and variation in body form the South American
catfish genus Corydoras (Callichthyidae). Syst Zool 34:381-396.
Streelman JT, Danley PD. 2003. The stages of vertebrate evolutionary radiation. Trends
Ecol Evol 18:126-131.
Streelman JT, Alfaro M, Westneat MW, Bellwood DR, Karl SA. 2002. Evolutionary history
of the parrotfishes: biogeography, ecomorphology, and comparative diversity.
Evolution 56:961–971.
Sullivan JP, Lundberg JG, Hardman M. 2006. A phylogenetic analysis of the major groups
of catfishes (Teleostei: Siluriformes) using rag1 and rag2 nuclear gene sequences.
Mol Phylogenet Evol 41:636-662.
Surlemont C, Vandewalle P. 1991. Développement postembryonnaire du squelette et de
la musculature de la tête de Clarias gariepinus (Pisces, Siluriformes) depuis l'eclosion
jusqu'à 6,8 mm. Can J Zool 69:1094–1103.
Surlemont C, Chardon M, Vandewalle P. 1989. Skeleton, muscles and movements of the
head of a 5,2 mm fry of Clarias gariepinus (Burchell) (Pisces Siluriformes). In:
Splechtna H, Hilgers H, editors. Trends in vertebrate zoology. Stuttgart: Gustav
Fischer Verlag. p 459-462.
Suzuki HI, Agostinho AA, Winemiller KO. 2000. Relationship between oocyte morphology
and reproductive strategy in loricariid catfishes of the Parana River, Brazil. J Fish Biol
57:791-807.
Takahasi, N. 1925. On the homology of the cranial muscles of the cypriniform fishes. J
Morphol Physiol 40:1-109.
Tamaru CS, Cole B, Bailey R, Brown C. 1997. A manual for commercial production of the
tiger barb, Capoeta tertrazona, a temporary paired tank spawner. Hawaï: Center for
Tropical and Subtropical Aquaculture.
Taylor JN. 1983. Field observations on the reproductive ecology of three species of
armored catfishes (Loricariidae: Loricariinae) in Paraguay. Copeia 1:254-259.
161
7 REFERENCES
Taylor WR, Van Dyke GC. 1985. Revised procedures for staining and clearing small fishes
and other vertebrates for bone and cartilage study. Cybium 9:107–119.
Turner GF. 2007. Adaptive radiation of cichlid fish. Curr Biol 17:R827-R831.
Tilney RL, Hecht T. 1993. Early ontogeny of Galeichthys feliceps from the south east
coast of South Africa. J of Fish Biol 43:183-212.
Vandewalle P, Surlemont C, Sanna P, Chardon M. 1985. Interprétation fonctionnelle de
modifications du splanchnocrâne pendant le développement post-embryonnaire de
Clarias gariepinus (Téléostéens, Siluriformes). Zool Jahrb Anat 113:91–100.
Vandewalle P, Lalèyé P, Focant B. 1995. Early development of cephalic bony elements in
Chrysichthys auratus (Geoffroy Sint-Hilaire, 1808) (Pisces, Siluriformes, Claroteidae).
Belg J Zool 125:329–347.
Vandewalle P, Gluckman I, Baras E, Huriaux F, Focant B. 1997. Postembryonic
development of the cephalic region in Heterobranchus longifilis. J Fish Biol 50:227–
253.
Vandewalle P, Chikou A, Lalèyé P. 1999. Early development of the chondrocranium in
Chrysichthys auratus. J Fish Biol 55:795–808.
Vandewalle P, Germeau G, Besancenet P, Parmentier E, Baras E. 2005. Early
development of the head skeleton in Brycon moorei (Pisces, Ostariophysi,
Characidae). J Fish Biol 66:996–1024.
Van Snik GMJ, Van Den Boogaart JGM, Osse, JWM. 1997. Larval growth patterns in
Cyprinus carpio and Clarias gariepinus with attention to the finfold. J Fish Biol
50:1339–1352.
Vasnetsov VV. 1953. Etaps in the development of bony fishes. In: Pavlovsky EN, editor.
Otcherky po obshtchim voprosam ichtiologii. Moskva-Leningrad: AN SSSR Press. p
207-217.
Verhagen JHG. 2004. Hydrodynamics of burst swimming fish larvae; a conceptual model
approach. J Theor Biol 229:235–248.
Vetter B. 1878. Untersuchungen zur vergleichenden Anatomie der Kiemen und
Kiefermusculatur der Fische. Jena Zeits J Naturwiss 12:431-550.
Vigliotta TR. 2007. A phylogenetic study of the mochokidae (Teleostei, Siluriformes).
Unpublished Master Thesis, Cornell University.
Vilizzi L, Walker KF. 1999. The onset of juvenile period in carp, Cyprinus carpio: a
literature survey. Environ Biol Fish 56:93-102.
Wagemans F, Vandewalle P. 1999. Development of the cartilaginous skull in Solea solea:
trends in pleuronectiforms. Ann Sci Nat 1:39–52.
Wagemans F, Vandewalle P. 2001. Development of the bony skull in common sole: brief
survey of morpho-functional aspects of ossification sequence. J Fish Biol 59:1350–
1369.
Wagemans F, Focant B, Vandewalle P. 1998. Early development of the cephalic skeleton
in the turbot. J Fish Biol. 52:166–204.
Weihs D. 1980. Energetic significance of changes in swimming modes during growth of
larval anchovy, Engraulis mordax. Fish Bull 77:597-604.
Weisel GF. 1967. Early ossification in the skeleton of the sucker (Catostomus
macrocheilus) and the guppy (Poecilia reticulata). J Morph 121:1–18.
Weitzman SH, Vari RP. 1988. Miniaturization in South American freshwater fishes; an
overview and discussion. Proc Biol Soc Wash 101:444-465.
162
7 REFERENCES
West-Eberhard MJ. 2003. Adaptive radiation. In: West-Eberhard MJ,
Developmental plasticity and evolution. Oxford: University Press. p 564-597.
editor.
Winterbottom R. 1974. A descriptive synonymy of the striated muscles of the Teleostei.
Proc Acad Nat Sci Phila 125:225-317.
Wright JJ, Page LM. 2006. Taxonomic revision of the lake Tanganyikan Synodontis
(Siluriformes: Mochokidae). Bull Fla Mus Nat Hist 46:99-154.
Wu K-Y, Shen S-C. 2004. Review of the teleostean adductor mandibulae and its
significance to the systematic positions of the Polymyxiiformes, Lampridiformes, and
Triacanthoidei. Zool Stud 43:712-736.
163
7 REFERENCES
164
8
Publication List
8 PUBLICATION LIST
PUBLICATION LIST
Huysentruyt F, Adriaens D, Teugels GG, Devaere S, Herrel A, Verraes W, Aerts P.
2004. Diet composition in relation to morphology in some African anguilliform
clariid catfishes. Belg J Zool 134(1):25-30.
Devaere S, Teugels GG, Adriaens D, Huysentruyt F, Verraes W. 2004.
Redescription of Dolichallabes microphthalmus (Poll, 1942) (Siluriformes,
Clariidae). Copeia 1:108-115.
Huysentruyt F, Adriaens D. 2005. Adhesive structures in the eggs of Corydoras
aeneus (Gill, 1858; Callichthyidae). J Fish Biol 66(3):871-876.
Huysentruyt F, Adriaens D. 2005. Descriptive osteology of Corydoras aeneus
(Gill, 1858) (Siluriformes: Callichthyidae). Cybium 29(3):261-273.
Adriaens D, Geerinckx T, Huysentruyt F, Schaefer SA, Herrel A. 2005. Evolution
of trophic specialisations in neotropical catfishes: more than a mouthful. Integr
Comp Biol 45(6):955-955.
Van Wassenbergh S, Herrel A, Adriaens D, Huysentruyt F, Devaere S, Aerts P.
2006. A catfish that can strike its prey on land. Nature 440:881.
Huysentruyt F, Geerinckx T, Adriaens D. 2007. A descriptive myology of
Corydoras aeneus (Gill, 1858) (Siluriformes: Callichthyidae), with a brief
discussion on adductor mandibulae homologies. Anim Biol 57:433-452.
Huysentruyt F, Brunain M, Adriaens D. ———. Ontogeny of the chondrocranium
in Corydoras aeneus (Gill, 1858) (Callichthyidae, Siluriformes). J Morph DOI:
10.1002/jmor.10603.
Geerinckx T, Huysentruyt F, Adriaens D. ———. Ontogeny of the jaw and
maxillary barbel musculature in the armoured catfish families Loricariidae and
Callichthyidae (Loricarioidea, Siluriformes), with a discussion on muscle
homologies. Zool J Linn Soc in press.
Huysentruyt F, Geerinckx T, Brunain M, Adriaens D. ———. Development of the
osteocranium in Corydoras aeneus (Gill, 1858) (Callichthyidae, Siluriformes). J
Morph submitted.
Huysentruyt F, Brunain M, Adriaens D. ———. Ontogeny of the cranial
musculature in Corydoras aeneus (Gill, 1858) (Callichthyidae, Siluriformes). J
Fish Biol submitted.
Huysentruyt F, Moerkerke B, Devaere S, Adriaens D. ———. Early development
and allometric growth in the armoured catfish Corydoras aeneus. In prep.
Huysentruyt F, Adriaens D. ———. Ontogeny of the postcranial skeleton in
Corydoras aeneus (Gill, 1858) (Callichthyidae, Siluriformes). In prep.
165
8 PUBLICATION LIST
Ide C, De Puysseleir J, Huysentruyt F, Geerinckx T, Adriaens D. ———. Nonalgae scraping vs algae scraping, comparison between Synodontis and
Atopochilus (Siluriformes: Mochokidae). In prep.
166