JOURNAL OF MORPHOLOGY 166203-216
(1980)
Fetal Tooth Development and Adult Replacement in
Dermophis mexicanus (Amphibia: Gymnophiona):
Fields Versus Clones
MARVALEE H. WAKE
Department of Zoology and Museum of Vertebrate Zoology, University of California,
Berkeley, California 94720
ABSTRACT
Teeth of fetuses of a caecilian, Dermophis mexicanus (Amphibia:
Gymnophiona), show ontogenetic variation in crown structure from small,
multidenticulate, and non-pedicellate to larger, spoon-shaped, pedicellate teeth
with a single apical spike. Number of denticles decreases as enamel-secreting
cells mature. Numbers of teeth and of tooth rows increase ontogenetically. A
fetal vomeropalatine set of teeth is present in D . mexicanus but absent in species
previously examined. Teeth transitional to the adult shape and arrangement
appear shortly before birth. The transition is correlated with birth, not fetal
size. There is relatively little increase in numbers of teeth during the juvenile
period. The pattern of development does not fully agree with either morphogenetic field theory or with clone theory, both as defined by Osborn (‘78). Sequence
of initiation is appropriate to either. Tooth shape changes agree with aspects of
clone theory. Multiple rows of fetal teeth and the transition to adult follow field
theory. Clone theory holds that patterns of development and shape are selfregulated, field theory that they are controlled extrinsically. I suggest that
substances regulating differentiation mediate early development, and hormones
later development, including inception of adult teeth, and are comparable to
“field substances” influencing primordia that originate according to clone theory.
Components of both theories are appropriate to analyzing tooth development
phenomena.
Recently attention has been paid to the
considerable species-specific and ontogenetic
variation in fetal and adult dentitions of caecilians (Amphibia: Gymnophiona). Fetal tooth
crown morphology and tooth position are modified extensively during the abrupt transition
from fetal state to adult; this change, and
some adult variation, are correlated with
mode of feeding. Parker (’56)and Parker and
Dunn (’64)considered the teeth non-adaptive,
non-functional vestiges of a piscine ancestry;
Wake (’76, ’77a,b, ’78) emphasized the functional nature of these teeth in fetal nutrition.
The teeth of oviparous embryos and larvae are
similar to those of the adult, without marked
ontogenetic variation (Sarasin and Sarasin,
1887-90; Marcus, ’20; Reuther, ’31; Lawson,
’65a,b; Wake, ’76), and there is considerable
interspecific variation in adult dentition as
well (Wake and Wurst, ’79), some correlated
with feeding specializations (Wake, ’78). This
system lends itself to 1) analysis of tooth
0362-252518011662-0203$02.60 0 1980 ALAN R. LISS, INC.
replacement phenomena, 2) comparison with
tooth replacement analyses in other amphibians as well as in fish, reptiles, and mammals,
and 3) consideration of current theories explaining tooth replacement and the evolution
of tooth size, shape, and position in vertebrates.
This paper focuses on the ontogeny and
replacement of teeth in Dermophis mexicanus,
a Mexican and Central American terrestrial,
burrowing, viviparous member of the gymnophionan family Caeciliidae. All specimens
are from a single population and were collected a t various times of the year. Since gestation
virtually
in this species is 1year (Wake, ’80),
the entire pre-birth ontogeny as well as adult
variation can be examined.
Parker and Dunn (‘64) described the fetal
teeth of D . mexicanus as spoon-shaped with a
single apical cusp. They commented that some
populations of the species appeared to have
multicusped teeth and reported that fetuses
204
MARVALEE H. WAKE
near birth size had dentitions of mixed fetal
and adult teeth. Wake (‘76) described the teeth
of a 22-mm D. mexicanus embryo and commented that the tooth crowns with long, flexible denticles and no mineralized pedicel may
not be functional. It appeared that D. mexicanus, and other species, had two fetal tooth
morphs. Current material available for study
indicates that, through ontogeny, there is a
transition from a multi-denticled “pref‘unctional” tooth crown through a reduction of
cusps and an increase in enamel deposition
and mineralization to the spoon-shaped,monodenticled “functional fetal” form and then to
the adult size, shape, and position on the jaws.
The tooth ontogeny in this species is typical of
live-bearing caecilians in many ways, but unusual in others, particularly in the ontogenetic
changes in dentition as the fetus grows, as
described below. The completeness of the series and the properties of the morphology
provide an unusual opportunity for analysis of
tooth growth and replacement. Several problems are addressed, utilizing the Dermophis
mexicanus data and comparison with data
from other caecilians, salamanders, frogs, lizards, and mammals. The pattern of change in
fetal tooth crown morphology, the pattern of
fetal tooth addition and replacement, and
changes in the basic structure of the tooth
during development are described and evaluated. The pattern of the transition from fetal
to adult dentitional shape, size, and position,
clues to the mechanism of this shift associated
with developmental phenomena, and whether
these patterns are self-generated (morphogenetic clones) or externally controlled (morphogenetic fields), or more complexly produced,
are considered.
MATERIALS AND METHODS
ually. Other specimens were cleared and
stained with alizarin Red-S or double-stained
with alizarin and alcian blue (Wassersug: ’76).
Tooth replacement series are much more easily analyzed in cleared and stained than in
dried skeletal specimens. Locus maps were
constructed as in Wake (’76). In maps of fetal
teeth, a teeth are the non-pedicellate, multidenticulate “prefunctionaYteeth; b teeth are
the pedicellate, spoon-shaped, monocusped
functional fetal teeth; B teeth are the transitional fetal-adult morph. Loci were counted as
in Wake (’76). Type b teeth include b and c
(labial, sloughing) types described by Wake
(’761, since the significant morphological distinction in D. mexicanus is between a and b
types.
OBSERVATIONS
Tooth crown ontogeny
Wake (‘76) described the tooth crowns of a
22-mm embryo of D. mexicanus as having
several long, flexible denticles, being weakly
mineralized, and lacking pedicels. It is apparent from new material that this is an early
stage in tooth crown ontogeny in the species,
as can be seen from the following sequence.
3 5 - m m fetus. The dentary tooth crowns
are in four rows antero-medially, reducing in
number posteriorly on the dentary ramus (Fig.
1A). They are small, approximately 8 pm high
(crown base to tip of longest denticle), and
protrude little above the oral mucosa (Figs.
8a, b). There are two series of teeth in the roof
of the mouth: a double row of multi-denticled
teeth occurs on the ossifying maxillae and
premaxillae, and a single row in a vomeropalatine series. These teeth show no alizarin
affinity, so may be unmineralized.
An ontogenetic series of Dermophis mexi39-mm. Crowns are broader; denticles are
canus including embryos, fetuses, juveniles, reduced in height and frequently fused as
and adults was prepared for study of denti- paired units (Fig. 8c,d). The cup- or spoontional and other morphology. Table 1lists the shaped face of later teeth is foreshadowed by
specimens and their manner of preparation. the curvature medially of the outer margins
Preparation for scanning electron microg- of the tooth crowns (Figs. 8c, d, e). Some
raphy involved the technique described by flexibility of denticles and the crown body is
Wake and Wurst (’79).Material was embedded indicated by the variation in curvature of both
in paraffin according to standard techniques components (Fig. 8d). Pedicels are forming.
and sectioned so that frontal, sagittal, and
transverse series stained with hematoxylin
46-mm. Further denticle reduction is eviand eosin and Mallory’s azan were available. dent (Fig. 8e); the body of the crown is broadHeads were embedded in plastic according to ening and thickening. The apical denticles
Lombard (’701, sectioned sagittally a t 10 pm curve medially, and the lingual face of the
on a standard microtome, stained with he- crown is concave. Only the tooth tips are
matoxylin and safranin, and mounted individ- mineralized. There are four rows of dentary
205
FETAL AND ADULT CAECILIAN TEETH
TABLE I , Specimens
Total length
in m m
4
5
6
a
10
10
10
14
15
15
22
23
25
34
35
*35
38
*39
44
* 46
48
*51
58
58
*60
64
68
72
77
80
80
* 101
110
*110
* 120
* 123
140
* 141
* 148
* 150 (2)
* 151
155
162
164
* 185
199
265
290
325
327
* 349
398
of
Dermophis mexicanus prepared for dentitional analvsis
Maturation s t a t e
Preparation
Embryo
Embryo
Embryo
Emhryo
Embryo
Embryo
Embryo
Embryo
Embryo
Embryo
Embryo
Embryo
Embryo
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Fetus
Juvenile
Juvenile
Newborn
Juvenile
Juvenile
Fetus
Juvenile
Juvenile
Juvenile head
Juvenile head
Juvenile head
Adult head
Adult head
Adult head
Adult head
Adult head
Adult head
Pa S
Pa S
Pa S
Pa S
Pa S
Pa F
Pa X
Pa S
Pa S
Pa F
C&S
Pa X
Pa F
Pa X
Pa F
SEM
C C S
SEM
Pa F
SEM
C&S
SEM
Pa X
Pa F
SEM
DC&S
Pa X
Pa F
SEM
Pa X
Pa F
DC&S
PI s
DC&S
DC&S
DC&S
DC&S
DC&S
DC&S
DC&S
DC&S
DC&S
PIF&S
Pa F
DCCS
Pa F
Pa F
Pa F
PlF&S
Pa F
C&S
P1 s
~~
Tooth type
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
PFF
FF
mixed P F F & FF
mixed PFF & FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
+A
+A
A
A
A
A
A
FF
A
A
A
A
A
A
A
A
A
A
A
Abbreviations: A = adult; C & S = cleared and alizarin-stained D C & S = cleared and
alizarin- and alcian blue-stained, F = frontal sections; FF = functional fetal; Pa = paraffin;
PFF = prefunctional fetal; P1 = plastic; s = sagittal sections; SEM = scanning electron
microscopy; x = transverse sections. Asterisks indicate specimens described in detail in the
text; other specimens were prepared to provide more histological detail or other developmental
stages for comparison.
206
MARVALEE H. WAKE
A
4
3
2
I
0)
b'
b2 b'
b'
b2a2 F
b'
l
b'
b'
b'
b'
b'
B
C
101-mm. Tooth crowns are predominately
the monocuspid, spoon-shaped type throughout the jaw. Multi-denticled crowns are present antero-medially and labially, extending
posteriorly along the labial margin, but decreasing in number posteriorly on the dentary.
b' b' b'
b' b' b'
b' b' b' b' b'
B
b
5
Loci
V b'
b'
B b'
b
I0
B
b'
15
b' b'
b
b b
the jaw. The overall dimensions of the crowns
are little changed from the 35-mm stage, but
the shape is markedly changed. Teeth are in
five rows anteriorly on the lower jaw (Figs.
l C , 8e). All of the teeth are well mineralized.
Similar but smaller fetal teeth are disposed in
four rows in a maxillary-premaxillary series
and in four rows in a vomeropalatine series
(Figs. 2A, 3A).
20
25
posterior 4
Fig. 1. Locus maps of fetal teeth on the right dentary
ramus of Derrnophis rnezicanus. A) 35-mm specimen; B)
46-mm specimen; C) 64-mm specimen; D) 110-mm specimen; E) 151-mm specimen. Abbreviations: a, b, and B as
described in the text. A numerical superscript indicates the
cusp number (superscript indicates a multi-cusped tooth),
absence of superscript indicates that all are monocusped,
and e indicates that the crown is just erupting.
teeth (Fig. lB), the labialmost having developed earliest; three rows of maxillary-premaxillary teeth and a single row of vomeropalatine teeth.
51-mm. Many tooth crowns are highly
curved, relatively stout-bodied, and have a
single medial apical cusp. These are well mineralized, as shown by alizarin staining, and
have mineralized pedicels. An anterior labial
view (Fig. 8f) shows several such teeth and
one, closest to the labial margin, that lacks
the medial apical cusp but has virtually no
lateral denticles. On the posterior lingual aspect of the same jaw (Fig. 8g), the largest,
thickest, most concave crowns are those with
the single apical cusp; those more posterior
and more labial are similar to crowns of the
35- to 46-mm specimens. This is of significance to the tooth succession proposed below.
The crowns form five to six rows on the lower
jaw.
64-mm. Crowns of the concave, monocusped type predominate along the length of
110-mm. Dentary tooth crowns are predominately monocuspid and are in eight rows
medially (Fig. ZD).However, seven teeth on
the lingual anteromedial margin are larger,
have a reduced apical cusp and a much shallower lingual face. These transitional teeth
are designated type B . Each appears to occupy
the equivalent of two of the spaces held by
smaller teeth. These teeth are momholoeicallv
more like adult than fetal teeth 6ut lGk th;
lateral flanges and slight convexity shown by
Wake and Wurst ('79). In the maxillary-premaxillary series there are four rows of fetal
teeth, with five adult-like teeth interspersed
medio-laterally in the first and second rows a t
loci 2, 7, 8, 10, and 14 (Fig. 2B). The vomeropalatine series is of fetal teeth in four rows,
with a single adult-type tooth at locus 10 in
the first row (see Fig. 3B).
120-mm. The fetus has a full complement
of cusped fetal teeth. They are disposed in six
rows from lingual to labial aspect antero-me-
-
4
3bbbbbb!bbbb;bb
2 - B ,bbb
b
I - , ,b, , , ,B,B: ,B, , ,
6
5B
bbb
b
:-B
B
b
b
b
,,,
I
LOCI
b b
,b , ,
5
bb
b b
B
B
B
-
2
I
b
b
labial
b b b b bbb
b b b
b
,B, , , ,b, , ,b, , , b lingual
,
,
Be
, ,,,
10
t
b
b bb
b b b
bb
b B b
b
b
B
, ,B: ,
15
b
b
B. b
-
, , , , , , , ,b', ,,'b
20
posterior
b
25
Fig. 2. Locus maps of fetal teeth of the left maxillarypremaxillary row. A) 64-mm specimen: B) 110-mmspecimen;
C) 151-mm specimen. Abbreviations as in text and Figure 1.
FETAL AND ADUL'I' CAECILlAN
A
2
I
b
e d
be
C
b
a
~ ~ , b , , ~ , ~ ~ , b , b , : , Bb,;! 8
2
b
I
b
I
Loci
B b
bb'
5
b'
b
B
j!, ,! ,:,I:,
b
b
10
15
b b
b bb'
20
posterior
-
,
b
b
m
1
25
Fig. 3. Locus maps of fetal teeth of the left vomeropalatine row. A) 64-mm specimen; B) 110-mmspecimen; C)
151-mm specimen. Abbreviations as in text and Figure 1.
dially, reducing to three, then one, posteriorly
on the lower jaw and in the maxillary-premaxillary series.
123-mm. This free-living juvenile has a
full complement of adult teeth arranged in
single rows (Fig. 4A, B, C).
141-mm. This 2-day-old specimen has a
full complement of adult teeth as above, but
is also sloughing medially and labially its
remaining fetal teeth.
148-mm. This juvenile has a full complement of adult teeth.
150-mm. A juvenile has a full complement
of adult teeth (Fig. 5A, B, C).
151-mm. This large fetus has a full set of
fetal monocusped teeth in up to eight rows
medially on the lower jaw, six and five in the
maxillary and vomeropalatine series, respectively. It resembles the complement in the
110-mm fetus, but includes greater numbers
of large functional fetal teeth and of the transitional tooth type (Figs. lE, 2C, 3C).
185-mm. This advanced juvenile has normal adult teeth (Fig. 6A, B, C).
349-mm. The complement of adult teeth is
present, numerically greater in both functional teeth and replacement stages than in smaller specimens (see Fig. 7).
Other aspects of tooth crown morphology
are considered based on analysis of 1) anomalous development, and 2) crown structure as
revealed by accidental SEM beam damage as
a result of overexposure. One tooth in the 35mm fetus (Fig. 8a, left middle of field) showed
Loci
posterior 4
Fig. 4. Locus map of teeth in various successional stages
in a 123-mm juvenile D . mexicanus. A) Right dentary
ramus; B) left premaxillary-maxillary row; C) left vomeropalatine row. Stages a-d, erupting to assuming fixed
position; e is an ankylosed functional tooth; f is a tooth in
which the pedicel is eroding; g is the space left by a
sloughed tooth. See Wake ('76) for more detail.
an irregular surface and single medial extension (Fig. 9b). Similar tooth crowns were observed, although infrequently in other specimens. This tooth lacks an enamel layer and
therefore lacks the denticles typical of teeth
in this size fetus. The irregular crown surface
is assumed to be due to irregular pulses of
dentine deposition as the tooth forms. The
crown from a 77-mm specimen that suffered
beam damage is shown t o have an internal
dentine structure similar t o the tooth from the
35-mm specimen mentioned above. The tooth
appears to have enamel that is lain down in
three coats with slightly different orientation
in this specimen. Based on examination of
internal and external lamellar cells of tooth
buds of small specimens, there appears to be
a correlation between number of dentine
pulses (surface blips) and lamellar cell counts.
This suggests that the dentine "core" of all
fetal teeth is similar, and crown cusp variation
is a property of change in enamel deposition
with increased age in this species. Microanatomy of tooth development is the subject of
another manuscript.
208
MARVALEE H. WAKE
A
0
b
e
,
0
:
C
l
f
, ,
,0 , ,
,
, , , ,
C
:.' ,*;:
g.
0
,;
b
0
5
Loci
,
10
posterior
, , ,
,
15
Fig. 5. Locus map of teeth in various stages in various
successional stages in a 150-mm juvenile. A) Right dentary
ramus; B) left premaxillary-maxillary row; C) left vomeropalatine row. Stages as in Figure 4.
Fig. 6. Locus map of teeth in various stages in a 185mm juvenile. A) Right dentary ramus; B) left premaxillarymaxillary row; C) left vomeropalatine row. Stages as in
Figure 4.
Fetal and adult tooth replacement
Tooth replacement patterns in D . mexicanus
are similar in many ways to patterns described by Wake ('76) for Gymnopis, Typhlonectes, and Schistometopum. 'Ibvo sets of fetal
teeth are present; numbers of teeth may increase as new loci are added anteriorly before
the jaw is mineralized early in development
and posteriorly afterward; numbers of rows of
teeth increase as age increases, although pedicels do not fuse to form the dentary patch
seen in Typhlonectes. Observations will therefore be confined to those aspects in which D.
mexicanus either differs in pattern or adds
new information about replacement patterns.
A major difference in tooth crown morphology exists between type a prefunctional teeth
and type b functional teeth in D. mexicanus.
The former are multi-denticulate and fairly
flat-faced, with pedicels that are unmineralized t o weakly mineralized. The latter, monocusped and concave lingually, with stout, wellmineralized pedicels. The "functional" designation refers to tooth use in oviducal feeding
(Wake, '76). This is unlike other species reported by Wake, in which the prefunctional
tooth crowns are miniatures of the functional
crowns.
D . mexicanus is unusual in having both
maxillary-premaxillary and vomeropalatine
tooth series present during the fetal period. In
other species the maxillary-premaxillary series occurs late in fetal development, if at all,
and the vomeropalatine series is not represented. As noted above, the latter two series
of teeth increase in numbers of teeth and
numbers of tooth rows in a pattern similar to
that of the lower jaws (compare Figs. 2 and 3).
Prefunctional a teeth occur in the labialmost
tooth rows in all three series, presumably the
result of aggregation of progressive teeth in
loci from the more lingually situated dental
buds. Several functional fetal teeth aggregate
in each tooth family and are slowly sloughed.
This contrasts with the adult state of a single
functional tooth per family. Multiple functional teeth are typical of all fetal dentitions
examined.
The several fetuses near birth size, newborns, and juveniles yield considerable new
information about the fetal-to-adult dentitiona1 transition. Birth occurs between 108 and
155 mm fetal total length, usually between
125 and 140 mm (Wake, '80). The teeth of a
110-mm fetus, a 123-mm juvenile, a 151-mm
fetus, and a 150-mm juvenile are mapped for
209
FETAL AND ADULT CAECILIAN TEETH
DISCUSSION
Loci
posterior +
Fig. 7. Locus map of teeth in various stages in a 349mm adult. Right dentary ramus. Stages as in Figure 4.
comparison (Figs. 1-51, and other specimens,
fetal and free-living, in this range were prepared and examined (see Table 1).A key
feature of both the 110-mm and the 151-mm
fetal dentitions is the presence of teeth that
resemble the adult morph (type B ) more than
the type b morph. These are interspersed
among b teeth in all three tooth series. The
larger fetus has more numerous teeth of both
types in more rows (except for the dentary)
than the smaller (Figs. 1-3). The transitional
teeth are much larger (two times, in both
height and breadth of crown) and at nearly
equally spaced intervals on the jaw. In most
tooth series of both specimens, more anterior
(medial) B teeth are in more lingual tooth
rows. This suggests that the transition begins
anteriorly on the jaw and subsequently develops posteriorly. However, in the vomeropalatine row, it appears that the first B tooth is
not most anterior. The only B tooth in that
row in the 110-mm specimen is in the tenth
locus (first on the palatine element); a B tooth
in locus 10 (last on the vomer) of the 151-mm
specimen is in a near-labial row, whereas the
other two B teeth, one more anterior and one
more posterior, are in lingual positions.
A comparison of the 123-mm, 150-mm, and
185-mm juveniles (Figs. 4-6) shows that number of functional adult teeth of the lower jaw
is about the same in each (9, 12, and 11 on
the right ramus), whereas the number increases with increased size in the other two
tooth series. This suggests that there is a
fundamental tooth complement for adult feeding, and that there is high significance to the
dentary complement. The proportion of earlier
developmental stages of teeth also increases
with age, as noted by Wake ('76) for other
species. The large (349-mm TL) mature specimen (Fig. 7) has only one more functional
dentary tooth, but two to four more loci are
involved and many more replacement stages
are present; maxillary and vomeropalatine
series show greater numbers of functional and
replacement teeth as well.
Developmental shape-transition sequence
The earliest tooth crowns are tiny structures
with a dentine base and a thin enamel layer
extended somewhat irregularly into a number
(three to seven) of apical spikes. These crowns
are not mineralized or are weakly mineralized
and are not ankylosed. They are shed during
the later fetal period. This state is similar t o
the situation that Osborn ('71) noted in prehatching embryos of Lacerta. The developmental sequence involves restriction of the
number of enamel apical spikes so that two
occur in early functional fetal teeth. These
sequential changes are described below and
can be followed in the developmental series
shown in Figure 8. The breadth of the body of
the tooth crown increases. There is also an
increase in curvature of the two sides of a
crown toward each other so that a concave
lingual aspect is achieved. This also brings
the spike regions into closer proximity (Fig.
8d). The increase in crown breadth and thickness and reduction of denticle or spike height
results in the crown configuration shown in
the tooth in the upper righthand corner of
Figure 8e. Crowns produced subsequently
have the lateral apical regions fused lingually
to produce the enclosed spoon shape and a
new, single apical extension. This is suggested
by comparison of labial and lingual aspects of
teeth in Figure 8f-h. Since it appears that the
dentine base component of the crown is only
slightly curved and has either a single median
or paired extension in close proximity (Fig.
9a, b), much of this variation may be due to
the developing nature of enamel deposition.
Size of crown and breadth of dentine and
enamel layers increase as cell number (to a
limited degree; see Meredith-Smith and Miles,
'71) and output increase. The transitional fetal-adult tooth is larger in size, more flattened,
and without the pronounced apical spike of
functional fetal teeth. It may be construed to
be the result of further maturation of the
dental progenitor, suppressing the spike and
strengthening the body of the crown so that
its sides do not curve medially, resulting in a
flatter, more conical tooth. These transitional
teeth are smaller than posterior (new) adult
teeth and lack the lateral flanges typical of
the arrowhead-shaped tooth crowns of the species (Wake and Wurst, '79).
Properties of the tooth recruitment sequence
are also evident in Figure 8. In Figures 8c, e,
f, and g, it can be seen that the multi-cusped,
210
W V A L E E H. WAKE
Figure 8
211
Fig. 9. Scanning electron micrographs of anomalous fetal tooth conditions. a) Crown in a 77-mm specimen
in which beam damage has caused the enamel layers to shear away from the dentine. Note the single apical
spike and the irregular surface of the dentine. b) Crown in a 35-mm specimen (also shown in Fig. 8a) in which
enamel was not deposited over dentine. Note the curvature of the crown and the medial and apical dentine
elongations. Bar = 10 pm. See text for discussion.
Fig. 8. Scanning electron micrographs of fetal teeth of the right dentary a t various sizes and stages of development.
a) 35-mm specimen, anterior lingual view; b) 35-mm, posterior lingual view; c) 39-mm, posterior lingual view; d) 39-mm,
mid-ramus lingual view; e) 46-mm, anterior labial view; f) 51-mm, anterior labial view; g) 51-mm, posterior lingual view;
h) 60-mm, posterior lingual view. Bar = 10 p m in all micrographs. Compare the cusp numbers and curvature of the crown
faces as size of fetus increases. Note in c and g that multi-cusped crowns, which erupt earlier, are on the labial margin of
the jaw relative to later-erupting, larger crowns that have fewer cusps and are in more lingual series. Arrow in 8a
indicates tooth enlarged in Figure 9b.
212
MRVALEE H. WAKE
of the B tooth in the middle of the vomeropalatine row may be similar to the mid-jaw stem
progenitor position Osborn ('71) found in Lacerta. Further, D. mexicanus is not typical of
viviparous caecilian species in having a maturational sequence of fetal and adult tooth
types. As noted above and in Wake ('761, in
most other species, the nonfunctional fetal
tooth crowns are smaller but of the same
shape as the functional fetal teeth. In addition,
there are several reports (see Parker and
Dunn, '64) of the presence of fetal teeth in
free-living specimens, without indications of
teeth of transitional form to the adult. It is
therefore possible that fetal tooth morphology
is often more genetically "fixed" than in D.
mexicanus, and that the transition from fetal
to adult shape and position is a more abrupt
and radical change. However, prebirth ontogenetic series are not yet available for other
species, and the virtually complete D. mexicanus series provides substantial insight into
fetal tooth development, although there may
be some variation in timing and pattern
among viviparous caecilian species. Several
problems of interpretation of tooth maturation
and replacement arise: First, that of explaining the different patterns of replacement of
fetal and of adult teeth; second, explaining the
difference in pattern of organization (position)
of adult vs. fetal teeth; third, explaining the
shape transitions and changes in function,
'80).
The presence of the vomeropalatine series correlated with other developmental properin fetuses of this species suggests that this is ties. If adult teeth (since they are a single
a function of dental primordium organization, tooth class as defined by Osborn "781) have a
not of functional involvement. If Wake ('76,
single primordium, as suggested above, does
'77a,b) and Salthe and Mecham ('74) are cor- this mean that the primordium is different
rect, the protruding teeth of the maxillary from that of the fetal teeth with which it is
series and especially the dentary are used to surrounded, or is it a maturation phase of a
stimulate secretion from oviducal epithelial pre-existing primordium? The above quescells. The vomeropalatine teeth do not partic- tions, and the data presented herein, will be
ipate in this function. That the marginal rows interpreted according to current theories of
of teeth are so involved is corroborated by the tooth development and replacement, primarily
presence of shredded epithelium and especial- the morphogenetic field theory (including Edly of red blood cells, presumed of maternal mund's "601 Zahnriehe theory) and the mororigin, as seen in Figure 8f.
phogenetic clone theory, both as defined by
As noted above, a number of aspects of tooth Osborn ('78).
development and replacement are typical of
other caecilian species and of lower verteThe field theory and the clone theory
brates with polyphyodonty and indeterminate
Both theories have been proposed to account
growth in general. Tooth loci are added anteriorly until the jaws mineralize, and poste- for ontogenetic differentiation of regions,
riorly on the jaws subsequently. In fetuses, units, shapes, sizes, and sequences of teeth.
tooth rows aggregate on the jaws, showing Major distinctive properties of the two ideas,
maturation sequences, and sloughing is slow. as summarized by Osborn ('78), are as follows:
New tooth types or earliest maturation occur 1) patterns controlled from extrinsic sources
anteriorly (medially). However, the inception (field) vs. self-generated (clone); 2) all primor-
flatter-crowned tooth characteristic of early
stages of development is labial to the later,
functional, mono-cusped, spoon-shaped morph.
The multi-cusped tooth therefore is assumed
to be crowded labially as successive rows of
more mature fetal teeth accrue. Yet, the posteriormost new loci, added in more advanced
fetuses, do not replicate the entire sequence of
tooth development. Frequently the first tooth
of such a new locus is a type b tooth (as in
Fig. 8h), suggesting that there is a maturational phenomenon affecting the entire dental
primordium, organized or not.
The transitions from a to b tooth crown
types, and to B and adult, are cued to other
transitions in the development of the organism. The a to b transition (prefunctional
fetal tooth to functional) occurs as yolk is
exhausted by the embryo (at approximately
25 mm total length). The fetus shifts to ingesting the nutrient secretion of the oviducal
epithelium, and the fetal teeth are indeed
functional (Wake, '77a,b). Type B teeth accrue
in fetuses 110-160 mm, but just before birth.
The size and placement, including spacing, of
these teeth foreshadow the state of the adult
tooth. Birth in this species is correlated with
the onset of the rainy season. Fat bodies of
large fetuses are very small, so it is inferred
that feeding upon invertebrate prey should be
established shortly after birth (Wake, '77a,b,
FETAL AND ADULT CAECILIAN TEETH
dia equivalent, but influenced by field substances (field) vs. all primordia different, with
final shape largely predetermined (clone); 3)
primary gradient (field substances) inducing
development of a secondary gradient (shape)
vs. gradients being concomitants of growth.
With regard to tooth development, the field
theory would have a specialized field generator anteriorly on the jaw, or presumptive jaw,
which produces a field substance that diffuses
through the growing region or dental lamina.
The further a cell is from the field generator,
the less field substance it is exposed to. Tooth
primordia differentiate, probably due to a different field effect, and the identical primordia
are subject to different concentrations of field
substances due to differences in distance from
the field generator, so that a gradient of structure is produced.
In the clone model, tissue equivalent to the
field generator contains three regions, the
middle one of which gives rise to primordia
and the tissue surrounding them. The middle
region is the clone, and it grows posteriorly.
Cell division ensues, and the tissues become
competent t o initiate primordia (suggesting a
maturation effect). The clone expands posteriorly, and new primordia arise as space becomes available within the cell region called
the clone. Osborn ('71, '74b, '78) suggests that
a zone of inhibition develops around a primordium, restricting development of new primordia nearby. Later primordia (interspersed anteriorly in growth space and added posteriorly)
are derived from cells that have divided more
times than those of earlier, more anterior
primordia, so that gradients in regional structures (such as anterior vs. posterior teeth) are
the result of a gradient in cell ancestry.
Tooth shape, according to the field theory as
discussed by Butler ('39, '56) and Van Valen
('701, is thought by those authors to be the
result of the ectodermal enamel organ and the
mesodermal dental papilla folding into a
shape that approximates the shape of the later
tooth. Sites that will develop into tooth cusps
mature earlier than surrounding tissue, and
so have different sensitivities to field substances. The clone theory does not require
sensitive, or target, sites. Osborn ('74b) proposed that the enamel epithelium grows over
the surface of the dental papilla a t a constant
rate in all directions. The dental papilla grows
a t constant, but different, rates in various
directions, so that the shape into which the
enamel organ folds is dependent on the growth
of the dental papilla. The clone theory, then,
213
says that the shape of a tooth is determined
by the growth rate of the primordium and,
ultimately, the clone. Shapes therefore may
be different but graded. "Shape potential" of
cells in a clone therefore changes as the clone
matures and extends. Some clone cells do not
become incorporated into primordia. Osborn
('78) presents experimental evidence to show
that all cells of the dental papilla have the
same developmental information, so that
shape is intrinsic.
Considerations of tooth number by Osborn
('72, '74a,b, '77) are aimed a t interpretation of
the evolution of diphyodont, heterodont mammalian dentitions from polyphyodont, homodont reptilian precursors, but contrasts between field and clone theories can be made.
According to field theory, response t o field
substances peaks at various places along the
jaw, resulting in a number of teeth, or, as
Osborn ('78) interprets Wolpert ('69), a gradient of field substance gives "positional information" and cells respond to specific concentrations of field substance, thereby giving rise
to evenly spaced teeth. The clone theory proposes that primordia can be initiated anywhere in the clone, but once initiated, generate a zone of inhibition, giving a spacing of
developing teeth. Few primordia can develop
in small clones or those with large inhibitory
zones, so the number is self-generated.
Sequence of tooth initiation is controlled by
an external substance, such as diminished
field substance producing later or smaller
teeth (field theory). The clone theory holds
that new primordia arise at the margins of
the extending clone, and shape potential of
new primordia is the result of cell ancestry.
Tooth loss during evolution is attributed to
weakening of the field effect (field theory) or
a reduction of shape potential of clone cells
below a threshold before the clone finishes
growing, so that primordia are generated but
do not develop (clone theory). According to the
latter, teeth are lost in sequence from the
margins of tooth classes (for example, incisors
and canines) generated by the clones.
There are many similarities between the D.
mexicanus data and those for Lacerta uiuipara,
the viviparous lizard carefully analyzed by
Osborn ('71) which provided the basis for
much of the presentation of the clone model.
Many of the developmental phenomena shown
in D mexicanus can be interpreted according
to the clone model, but, a t the same time, the
data invoke questions not answerable by clone
theory, but by a field model.
214
MARVALEE H. WAKE
The sequence of initiation of fetal teeth is
anterior to posterior, appropriate to either
theory. Posteriorly on the jaw, primordia generate type b teeth without going through a n
a stage. This is particularly logical according
to clone theory, for these primordia are generated at the margin of the growth unit and
may be the result of greater numbers of cell
division, thus an alteration of the cell ancestry, so that the more advanced tooth type is
developed. Yet, a major question arises from
this idea. It is assumed that cells giving rise
to later primordia have divided more frequently than those of earlier primordia (Osborn,
‘78, p. 173). Why is this necessarily so? The
jaw itself organizes anteriorly in caecilians,
with extensive differentiation and cell division
(Marcus, ’ 2 0 Wake, unpublished), so that posteriorly, there have been fewer cell divisions
at the same point in time. Posterior dental
buds form before the jaw is well-organized. It
is a t best possible that these are the results of
as many divisions as more anterior buds, but
more likely fewer, as they appear later. This
is not just a growth phenomenon, but one of
differentiation as well. Osborn (’78, p. 175)
also suggests that the clone may grow progressively more slowly, so that the growth
rates of the tooth primordia are equivalently
reduced in initial growth rates, resulting in
graded shapes of teeth. This suggests that
there would result smaller or more embryonic
teeth, not the more advanced type found in
caecilians.
The phenomena of multiple rows of fetal
teeth in caecilians (see Wake, ’76) and tooth
patches in other forms (for example, in the
plethodontid salamander Eurycea described by
Lawson et al., ’711, and the transition to the
adult single rows of teeth, is poorly reconcilable to the clone model. For fetal teeth to
aggregate in the manner that they do, there
must be very weak “zones of inhibition” generated to provide spacing. In fact, spacing,
especially in terms of numbers of rows (teeth
produced per locus), seems unregulated, and is
mediated only by birth. Spacing of B teeth
seems to take up the equivalent of three to
four loci of fetal teeth; hence, a “zone of inhibition” might be postulated. However, the
sudden increase in this phenomenon a t birth
is not explained by the clone theory. The
reduction in numbers of teeth per row and in
numbers of rows also cannot be reconciled to
this theory. The clone model predicts that
teeth are lost sequentially a t the margins of
tooth classes generated by clones. It appears
that teeth of the lower jaw of D. mexicanus
represent a single class, and therefore a single
clone, by analogy to Osborn’s L. viuipara, so
tooth loss should be posterior, not interspersed
throughout the jaw. The “zone of inhibition”
weakens as a tooth ages, so that alternate
replacement may occur. It is not known to
totally suppress loci so that the teeth are lost.
Tooth shape itself is poorly explained by the
clone theory. The earliest fetal teeth in D.
mexicanus are nonfunctional, as are those of
L. vivipara (Osborn, ’71).Osborn suggests that
shape potential in Lacerta decreases early in
life to a minimal level represented by simple
conical teeth, accounting for the loss of tiny
accessory cusps on the teeth of hatchlings.
Such an explanation might be invoked for the
overall shape transition in Dermophis, but
does not explain the differences in tooth shape
correlated with feeding function in fetus vs.
adult. Osborn (’78,p. 81) comments that while
the idea that sequences evolved in response to
the requirements of the animal suggests selective advantage for the sequence, it does not
indicate the control mechanism. Yet, Osborn’s
control mechanism, I infer, would involve regulation of change on the margins of clones.
This may apply to a series of heterodont teeth
in a mammal, with each type derived from a
different clone, but is weakly applicable to the
caecilian ontogenetic sequence. A shape transition has been hypothesized for these teeth,
suggesting that they represent a single clone
and a single tooth class, but change in tooth
types does not occur at the margins of clones.
This does not explain the timing of the fetalto-adult change, or the change from the nonfunctional fetal to the functional form, both of
which are correlated with major changes in
nutrition. Surely, “shape potentials” may
change as cell “competency” increases. However, this does not control or mediate change,
but is another product of change.
Osborn (‘78, p. 185) indicates that tooth
shape potential and size are affected by hormones, citing evidence from the literature on
salamander sexual dimorphism and seasonal
change (although he inappropriately suggests
that such effects might not be a valid conclusion from the data). He does not explain the
nature of hormonal influence. It is surely a n
“external” difference, so it calls into question
the degree to which shape is intrinsic. The
situation in caecilians is clearly one in which
hormones play a part. The change from yolk
nutrition to that of the maternal oviducal
secretion is at least indirectly hormonally me-
FETAL AND ADULT CAECILIAN TEETH
diated. The female’s corpora lutea produce
considerable progesterone during the time
that oviducal secretory activity is maintained
(McCreery, personal communication). Secretory activity probably diminishes shortly before birth (Wake, unpublished), so fetuses may
be subject to a reduced progesterone level as
well. A metamorphosis occurs in the oviduct
in fetal caecilians, involving loss of gills
(Wake, ’67), changes in the structure of the
eye and the skin, inception of the tentacle,
and change in hemoglobin (Toews and Macintyre, ’77). I have no reason to doubt that this
is under the control of thyroxin and prolactin,
as is typical of all amphibians investigated
thus far. The transition from fetal to adult
dentition, which begins in the oviduct shortly
before birth, is likely under similar hormonal
control. This aspect is similar to the field
model, in which tooth shape, etc., are induced.
An anteriorly placed “field generator” of unknown origin need not be postulated, since
maturation of teeth is consistently anterior in
most tooth series, so that more mature loci
probably respond earliest to hormonal presence. In fact, the most mature locus anywhere
on the jaw would probably respond first. Another aspect of tooth initiation reconcilable to
the field model has to do with the nature of
differentiation and organization on the jaws
and dental laminae. Since differentiation is
anterior first, proceeding posteriorly, it is likely that either “inducer” substances or the
byproducts of newly differentiating cells are
the equivalent of Van Valen’s (’70) several
different field substances, since they would
have a directionality along the jaw as a result
of developmental timing.
Finally, many of the arguments are based
on interpretations of the genome and its responses, without any substantive knowledge
of the nature of the genome. Some recent
experimental embryology sheds some light on
this problem (for example, Kollar, ’78; Miller,
’78; Flynn-Miller and Miller, ’78; and the work
of their collaborators), but not to the point
that it deals with the questions raised by
Osborn and herein.
Aspects of both the clone model and the
field model are useful in interpreting the Dermophis data. Osborn (’78) believes that the
field model is often applied to any aspect of
variation and does “little more than disguise
the statement of an observation as an explanation,” and lacks predictive power “because
fields seem to be infinitely variable.” I suggest
that the clone model, while explaining certain
215
developmental phenomena of dentitions, has
serious weaknesses as well. I suspect that a
biological explanation of these phenomena
will involve reconciliation of some aspects of
both extant theories in a model that also
includes information on the genome and hormonal and other extrinsic mediation.
ACKNOWLEDGMENTS
I thank Gloria Wurst for taking the scanning electronmicrographs and sectioning many
heads, Willy Bemis for preparing heads according to several techniques, and other assistance. I appreciate the use of facilities,
equipment, and advice of the Electron Microscope Laboratory at the University of California, Berkeley. I am grateful to John Cadle,
Theodore Papenfuss, Robert Seib, David
Wake, and Thomas Wake for collection of
specimens, and to David Wake for criticizing
the manuscript. All material will be deposited
in the Museum of Vertebrate Zoology, University of California, Berkeley. I also thank
Diane Nakamura for typing the manuscript
and Gene Christman for preparing the graphs.
This work was supported by National Science Foundation grant DEB 77-22642.
LITERATURE CITED
Butler, P.M. (1939) Studies of the mammalian dentition.
Differentiation of the postcanine dentition. Proc. Zool.
Soc. Lond., 109:l-36.
Butler, P.M. (1956) The ontogeny of molar pattern. Biol.
Rev., 31:30-70.
Edmund, A.G. (1960) Tooth replacement phenomena in the
lower vertebrates. Contrib. Life Sci. Div. R. Ontario
Museum, 52: 1-190.
Flynn-Miller, K.L., and W.A. Miller (1978) Dental morphological variations associated with murine chondrodystrophies (with a comment on the histology of the
cartilage disturbances). In: Development Function and
Evolution of Teeth. P.M. Butler and K.A. Joysey, eds.
Academic Press, New York, pp. 111-126.
Kollar, E.J. (1978) The role of collagen during tooth morphogenesis: some genetic implications. In: Development
Function and Evolution of Teeth. P.M. Butler and K.A.
Joysey, eds. Academic Press, New York, pp. 1-12.
Lawson, R. (1965a) The teeth of Hypogeophis rostrutus
(Amphibia, Apoda) and tooth structure in the Amphibia.
Proc. Zool. Soc. Lond., 145:321-326.
Lawson, R. (1965133 The development and replacement of
teeth in Hypogeophis rostrutus (Amphibia, Apoda). J.
Zool., 147:352-362.
Lawson, R.. D.B. Wake, and N.T. Beck (1971)Tooth replacement in the red-backed salamander, Plethodon cinereus.
J. Morphol., 134:259-270.
Lombard, R.E. (1970) A comparative morphological analysis of the salamander inner ear. Ph.D. Thesis, University
of Chicago.
Marcus, H. (1920) Uber die Zahne und die Korrelation
ihrer Zahl mit dem Alter, untersucht an Blindwiihlen,
Krokodilen, und Seekiihen. Deutsche Zahn. Forsch.
Prax., George Thieme, Leipzig, pp. 145-157.
216
MARVALEE H. WAKE
Meredith-Smith, M., and A.E.W. Miles (1971) The ultrastructure of odontogenesis in larval and adult urodeles;
differentiation of the dental epithelial cells. Z. Zellforsch.,
121: 470-498.
Miller, W.A. (1978) The dentitions of tabby and crinkled
mice (an upset in mesodermal: ectodermal interaction).
In: Development Function and Evolution of Teeth. P.M.
Butler and K.A. Joysey, eds. Academic Press, New York,
pp. 99-108.
Osborn, J.W. (1971) The ontogeny of tooth succession in
Lacertu uiuipuru Jacquin (1787). Proc. R. SOC.Lond.,
B:261-289.
Osborn, J.W. (1972) O n the biological improbability of
Zahnreihen as embryological units. Evol., 26r601-607.
Osborn, J.W. (1974a) Tooth succession in Diudemodon.
Evol.. 28:141-157.
Osborn, J.W. (1974b) On the control of tooth replacement
in reptiles and its relationship to growth. J. Theor. Biol.,
46:509-527.
Osborn, J.W. (1977) The interpretation of patterns in dentitions. Biol. J. Linn. Soc., 9:217-229.
Osborn, J.W. (1978) Morphogenetic gradients: fields versus
clones. In: Development Function and Evolution of Teeth.
P.M. Butler and K.A. Joysey, eds. Academic Press, New
York, pp. 171-202.
Parker, H.W. (1956) Viviparous caecilians and amphibian
phylogeny. Nature, 178:250-252.
Parker, H.W., and E.R. Dunn (1964) Dentitional metamorphosis in the Amphibia. Copeia, 1964t75-86.
Reuther, G. (1931) Die Zahnleiste vonHypogeophis. Beitrag
zur Kenntnis der Gymnophionen. XIV. Morphol. Jahrb.,
68: 105-112.
Salthe, S.N., and J.S. Mecham (1974) Reproductive and
courtship patterns. In: Physiology of the Amphibia. B.
Lo&, ed. Vol. 2. Academic Press, New York, pp. 309-521.
Sarasin, P., and F. Sarasin (1887-90) Ergebnisse naturwissenschaftlichen Forschungen auf Ceylon in den Jahren
1884- 1886. Zur Entwicklungsgeschichte und Anatomie
der ceylonischen Blindwiihle Ichthyophis glutinosus. C.
W. Kreidel’s Verlag, Wiesbaden.
Toews, D., and D. Macintyre (1977) Blood respiratory properties of a viviparous amphibian. Nature, 266:464-465.
Van Valen, L. (1970) An analysis of developmental fields.
Dev. Biol., 23:456-477.
Wake, M.H. (1967) Gill structure in the caecilian genus
Gymnopis. Bull. So. Calif. Acad. Sci., 66:109-116.
Wake, M.H. (1976) The development and replacement of
teeth in viviparous caecilians. J. Morphol., 248t33-64.
Wake, M.H. (1977a) The reproductive biology of caecilians:
a n evolutionary perspective. In: Reproductive Biology of
the Amphibia. S. Guttman and D. Taylor, eds. Plenum
Press Publ., New York, pp. 73-102.
Wake, M.H. (1977b) Fetal maintenance and its evolutionary
significance in the Amphibia: Gymnophiona. J. Herpetol.
11:379-386.
Wake, M.H. (1978) Ontogeny of Typhlonectes obesus, with
emphasis on dentition and feeding. Papeis Avulsos, Zool.,
22: 1-13.
Wake, M.H. (1980) Reproduction, growth, and population
structure of Dermophis mexicanus (Amphibia: Gymnophiona). Herpetologica (in press).
Wake, M.H., and G.Z. Wurst (1979) Tooth crown morphology in caecilians (Amphibia: Gymnophiona). J. Morphol.,
159t331-342.
Wassersug, R.J. (1976) A procedure for differential staining
of cartilage and bone in-whole formalin-fixed vertebrates.
Stain Technol., 5lt131-134.
Wolpert, L. (1969) Positional information and the spatial
pattern of cellular differentiation. J. Theor. Biol.,
24: 1-47.