Supplementary information* for
Evolution of cranial shape in caecilians (Amphibia: Gymnophiona)
Emma Sherratt, David J. Gower , Christian Peter Klingenberg, Mark Wilkinson
* except Table S1 which is a separate .xlsx file.
Supplementary Methods: Phylogeny
Evidence for our composite phylogeny (Figure 1) comes from several sources. The most
important source is the mitogenomic phylogenetic study of San Mauro et al. (2014, abbreviated to SM14)
which is the most recent large-scale phylogenetic study of caecilians, and which yielded a tree in which
all internal branches were maximally supported. This study determines the broad outline of our
phylogeny. Other studies (and abbreviations) providing support for particular internal branches within
major clades are: Boulengerula (Loader et al. 2011, L11); Indotyphlidae (Gower et al. 2011, G11);
Ichthyophiidae (Wilkinson et al. 2014, W14); Scolecomorphidae (Loader 2005, L05; Doherty-Bone et al.
2011, DB11); Typhlonectidae (Wilkinson and Nussbaum 1999) and Uraeotyphlus (Gower et al.
unpublished data, Gu). Because our composite tree includes more taxa than are included in the
phylogenetic trees from any of these sources, the positions of the additional taxa are necessarily based
upon assumptions and in some cases internal branches are based entirely upon assumptions of
monophyly.
Here we give, for each split in the composite tree, the supporting phylogenetic study or
assumptions upon which they are based (Table S2). In the former case support is maximal with at least
one numerical phylogenetic method unless otherwise noted in which case maximal support achieved is
given in parentheses. Note however that support values for less inclusive phylogenies do not apply
without caveat to corresponding internal branches in more inclusive phylogenies (see Wilkinson et al.,
2005). Where both assumptions and support are given this reflects that there is some support for
monophyly from explicit phylogenetic analyses but it is not sufficient (because of limited taxon sampling)
to fully justify the node. The assumptions of monophyly together with the support for internal branches
from the relevant phylogenetic studies completely determine our composite phylogeny in the sense that
if all the assumptions and supporting phylogenetic results are true then our composite tree must be true.
References
Doherty-Bone TM, Ndifon RK, San Mauro D, Wilkinson M, LeGrand GN, Gower DJ. 2011. Systematics and
ecology of the caecilian Crotaphatrema lamottei (Nussbaum)(Amphibia: Gymnophiona:
Scolecomorphidae). Journal of Natural History 45(13-14):827-841.
Gower DJ, San Mauro D, Giri V, Bhatta G, Venu G, Ramachandran K, Oommen OV, Fatih FA, MackenzieDodds JA, Nussbaum RA and others. 2011. Molecular systematics of caeciliid caecilians (Amphibia:
Gymnophiona) of the Western Ghats, India. Molecular Phylogenetics and Evolution 59(3):698-707.
Loader S, Wilkinson M, Cotton J, Müller H, Menegon M, Howell KM, Gower DJ. 2011. Molecular
phylogenetics of Boulengerula (Amphibia: Gymnophiona: Caeciliidae) and implications for
taxonomy, biogeography and conservation. Herpetological Journal 21(1):5-16.
Loader SP. 2005. Ph.D. Thesis: Systematics and biogeography of amphibians of the African Eastern Arc
mountains. Glasgow, UK: University of Glasgow.
San Mauro D, Gower DJ, Müller H, Loader SP, Zardoya R, Nussbaum RA, Wilkinson M. 2014. Life-history
evolution and mitogenomic phylogeny of caecilian amphibians. Molecular Phylogenetics and
Evolution.
Wilkinson M, Nussbaum RA. 1999. Evolutionary relationships of the lungless caecilian Atretochoana
eiselti (Amphibia: Gymnophiona: Typhlonectidae). Zoological Journal of the Linnean Society
126(2):191-223.
Wilkinson, M. & Nussbaum, R.A. 1992: Taxonomic status of Pseudosiphonops ptychodermis Taylor
and Mimosiphonops vermiculatus Taylor (Amphibia: Gymnophiona: Caeciliaidae), with description
of a new species. Journal of Natural History 26: 675-688.
Wilkinson, M., Pisani, D., Cotton, J. A. & Corfe, I. 2005. Measuring support and finding unsupported groups
in supertrees. Systematic Biology 54: 823-831.
Wilkinson M, Presswell B, Sherratt E, Papadopoulou A, Gower DJ. 2014. A new species of striped
Ichthyophis Fitzinger, 1826 (Amphibia: Gymnophiona: Ichthyophiidae) from Myanmar. Zootaxa
3785(1):45-58.
Table S2 The phylogenetic study supporting the mode, or assumptions upon which they are based, for
each split in the composite tree, Figure 1. Abbreviations in supplementary methods above.
Split Source of Support Assumption of Monophyly
1
SM14
2
SM14
Rhinatrematidae
3
-
Rhinatrema
4
SM14
5
SM14
6
SM14
Ichthyophis minus I. bombayensis
7
W14
Sri Lankan Ichthyophis
8
-
Peninsular Indian striped Ichthyophis
9
W14
South-East Asian Ichthyophis
10
SM14
11
SM14
12
-
13
Gu
14
Gu
15
Gu
16
Gu
17
Gu
18
Gu
19
SM14
20
SM14
21
L05
22
L05
23
L05
24
DB11
25
L05
26
L05 (89)
27
L05
28
DB11
29
SM14
Uraeotyphlus malabaricus group.
Scolecomorphus vittatus + S. kirkii
Crotaphatrema
30
SM14
31
SM14
32
SM14
33
L12
34
L12
35
L12
36
L12
37
L12
38
L12
39
SM14
40
SM14
41
SM14
42
SM14
43
W99
44
SM14
45
W99
46
W99
47
-
48
SM14
49
SM14
50
SM14
51
SM14
52
G12
53
SM14
54
G12
55
G12 (89)
56
G12 (98)
57
G12 (93)
58
-
59
SM14
60
SM14
Caeciliidae
Chthonerpeton
Idiocranium
61
SM14
Microcaecilia
62
SM14
Siphonoforms sensu Wilkinson and Nussbaum (1992)
63
SM14
64
-
65
SM14
66
SM14
67
SM14
Geotrypetes
Neotropical dermophiids
Table S3 Anatomical definitions of the landmarks used in this study. The composition of the caecilian
skull varies across taxa because some elements are lost or fused to others. This variation dictated that
some potential landmarks were ruled out through not being applicable to all taxa. The stapes is absent
and the fenestra ovalis closed in adult species of family Scolecomorphidae, but a cartilaginous rod
(precursor to stapes) is present in early ontogenetic stages, so the landmarks (47-50) were placed in a
central position on the occipital bulb. This approach was not, however, used in the case of the orbit
because of the variable location of this foramen.
Landmarks
1&2
Anteromedial point of nasal bone at nares opening.
*3 & 4
Anterior suture of premaxila/nasopremaxilla (n.pmx), where left and right n.pmx tooth rows meet.
5&6
Anterior corner of maxillopalatine, where left and right max.p. tooth row begin.
7&8
Posteromedial point of nares opening, on the septomaxilae when present, otherwise on nasal bones.
9 & 10
Posterior end of vomerine tooth row.
11 & 12
Posterior end of maxillopalatine, where skull widens for lower temporal fossa.
13 & 14
Posterior tip of vomers projecting over the paraspheniod.
15 & 16
Anteriolateral point on processus ascendens of the quadrate.
17 & 18
Widest point of osbasale, medially situated on otic capsule, on basipterygoid process.
19
Intersection of nasal and frontal bones.
*20 & 21
Parietal ridge, either side of suture between parietals (muscle scar for m. cutaneous dorsalis and M.
depressor mandibulae)
*22 & 23
Dorsomedial point of foramen magnum.
24 & 25
Lateral point of foramen magnum, above occipital condyls.
26 & 27
Lateral point of occipital condyls
28 & 29
Posterolateral corner of frontals where they contact parietals.
30 & 31
Posterolateral corner of parietal, by otic capsule.
32 & 33
Posterior point on processus ascendens of the quadrate.
34 & 35
Anterodorsal point of orbital foramen, on the sphenethmoid (point of insertion for cartilage taenia
marginalis dorsalis)
36 & 37
Anterior point of exterior foramina for jugular nerve.
38 & 39
Posteroventral point of orbital foramen, on the pleurosphenoid portion of the os basale (point of insertion
for cartilage taenia marginalis ventralis)
40 & 41
Dorsal point of squamosal ridge (muscle scar for M. depressor mandibulae insertion)
42
Posterior projection of infraorbital extension of the sphenethmoid
43
Medial of sphenethmoid on posterior side, between the olfactory nerve (Id and Iv) foramina
44 & 45
Anterior point of external foramina for carotid artery.
46
Ventromedial point of foramen magnum
47 & 48
Posterior point of foramen ovale for the stapes
49 & 50
Anterolateral point on stapedial process, by contact with quadrate.
51 & 52
Anterior point of orbital foramina, on the sphenethmoid, where foramen is widest.
53 & 54
Posterior inflexion point where maxillopalatine splits to surround the choanae, where when present,
pterygoid contacts max.p.
55 & 56
Posterior end of maxillopalatine inner tooth row.
57 & 58
Lateral point of left and right olfactory (Iv) nerve foramina on posterior side of sphenethmoid.
*59 & 60
Anteromedial suture of vomers, where vomerine tooth rows meet.
Supplementary Figures
Figure S1 Principal components (PC) analysis of the raw species means prior to the phylogenetic
allometry correction. The dots are sized proportionally to log centroid size of the cranium. It is evident
that no single PC axis defines the covariation of shape with size; rather size variation is dispersed among
species. The dots are coloured by clade as in Figure 1. Other than the position of Atretochoana, the large
purple dot, away from the others (divergent in a direction oblique to PC1 and PC3), all other species
remain in the distinct clade clusters as seen in Figure 6.
Figure S2 The fourth and fifth principal axes of cranial shape variation, visualised as warped
crania surfaces, as a complement to Figure 5. PC axes are from a PCA of species means, corrected for
evolutionary allometry Shape changes associated with the PCs are shown as extreme cranial shapes
representing the positive and negative end of each axis. In each case, the magnitude of shape change
from the mean corresponds to PC scores in Figure 6.
Shape changes in detail:
Depth of the front of the cranium dominates the fourth axis of shape variation. PC4 (7.9%) is associated
with relative shifts of landmarks in the boundary between the braincase and snout, marking changes in
the length of the sphenethmoid and the size of the internal orbital foramen, and change in the depth of
the front of the cranium relative to the braincase. In the negative direction, the front of the cranium is
almost as deep as the braincase, the internal orbital foramen is wide and the sphenethmoid is long. In the
positive direction the front of the cranium is dorsoventrally compressed and the internal orbital foramen
and sphenethmoid become relatively shorter. Although this axis contributes little to the overall variation
among species, it is important because it appears to describe morphological changes associated with the
niche-shift from terrestrial to more aquatic habitats, with members of the Typhlonectidae (even those
that are semi-aquatic) lying far away from all other species along this axis (Figure 6).
The final axis of major variation in caecilian cranial shape (PC5) is associated with mouth shape and the
position of the jaw articulation relative to the back of the skull, but overall contributes relatively little to
the variation among species (6.7%). This axis is associated with relative shifts of landmarks in the cheek
region and changes in the relative size of the mouth. In the negative direction, the cheek region lies
further posteriorly in relation to the braincase, and as such is associated with longer tooth rows and a
more extensive mouth. In the negative direction, the cheek region lies further forwards, with a less
extensive mouth. Species at the extreme limits along this axis belong to the Caeciliidae (negative) and
Typhlonectidae (positive; with the exception of Atretochoana, which lies closer to species of Caeciliidae).
Figure S3 Clade disparity is not correlated with clade age (A) nor with species richness (B).
Morphological diversity (disparity), calculated from shape data corrected for evolutionary allometry, as
determined by Procrustes variance. Abbreviations as in Figure 2, except Herpelidae + Chikilidae represented as
H for simplicity. Clade ages were estimated from published sources, details in text. Scolecomorphidae is
excluded from clade age analysis because no published dating estimate exists including the genus
Crotaphatrema.