Powell Et Al. Shape Variation in the Dermatocranium of the Greater Short-Horned Lizard

Powell et al. – Shape Variation in the Dermatocranium of the Greater Short-Horned Lizard Phrynosoma hernandesi– Online Resource

Online Resource - Powell et al., Shape Variation in the Dermatocranium of the Greater Short-Horned Lizard Phrynosoma hernandesi (Reptilia: Squamata: Phrynosomatidae)

Article Title:Shape Variation in the Dermatocranium of the Greater Short-Horned Lizard Phrynosoma hernandesi(Reptilia: Squamata: Phrynosomatidae)

Journal Name: Evolutionary Biology

Author Names: G. L. Powell1, A.P. Russell, H.A. Jamniczky, and B. Hallgrímsson

1Corresponding Author:

Affiliation: Dept. of Biological Sciences, University of Calgary, 2500 University Dr. NW, Calgary, Alberta, Canada T2N 1N4

e-mail:

Telephone: (403) 220 – 7638

Fax: (403) 289 – 9311

S.1. Specimens of Phrynosoma hernandesiExamined in This Study

CMN: 35b;1020; 1147-1; 1147-2; 1147-3; 1284; 1634; 1829; 1921 I-IX; 1921-1; 1921-2; 1921-3; 1922; 3479; 3482; 7333-1; 7333-2; 8190; 13496; 15499; 18301; 18302; 18591; 18592; 35199; 35977

PMA(Z): 672420; 71546; 77471

ROM: 7454; 7770 – 7777; 7876

SMNH: 13281

UAMZ: 13; 15; 41; 42; 94; 97; 98; 101; 102; 122; 123; 125; 126; 128; 130; 156; 276; 278 – 281; 776

UCM(Z): 1980.20 – 23; 1980.26; 1994.02 – 05; 1994.010; one uncatalogued specimen

Institutional Abbreviations

Collections Housing Specimens of Phrynosoma hernandesi: CMN – Canadian Museum of Nature; PMA(Z) – Zoology Collection of the Royal Alberta Museum; ROM – Royal Ontario Museum; SMNH – Royal Saskatchewan Museum of Natural History; UAMZ – University of Alberta Museum of Zoology; UCM(Z) – University of Calgary Museum of Zoology.

Collections Housing Specimens of PhrynosomaListed in Fig. 1: UTACVR – University of Texas at Arlington; CAS – California Academy of Sciences; UCMVZ – University of California Museum of Vertebrate Zoology; TCWC – Texas Cooperative Wildlife Collection; UAZ – University of Arizona; UM –University of Michigan Museum of Zoology; USNM – United States National Museum of Natural History.

S.2. Technical Difficulties Related to Landmarking

Because the material examined includes neonates and juveniles, the degree of ossification varied considerably among specimens, and was particularly poor in very young specimens. Ossification within skulls of younger specimens appeared to proceed posteriorly and dorsally from the premaxilla and maxillae, and anteriorly and posteriorly from the frontal.Setting of thresholds on an isosurface reconstruction of a neonate lizard at a level that would reveal useful detail of the posterior portion of the dermatocranium generally resulted in an isosurface reconstruction for which landmarks associated with the frontal, jugal and maxillary were difficult to locate repeatably, the sutures between these and adjacent bones being rendered as fused. The method used to produce a more useful isosurface rendering in such cases is outlined below.

Dermatocranial sutures examined were frequently unclosed, either partially or completely. This was most commonly so for the following:the fronto-parietal suture as it approached the pineal foramen (Landmarks 38 and 39; Fig. 3 a); the posterior ends of the postorbital-squamosal sutures (Landmarks 42 and 43; Fig. 3 a, b); and the common juncture of nasals and frontal (Landmarks 24 and 25; Fig. 3 a, c). In such cases, the landmark defined by such a juncture was placed on bone as close as possible to where the juncture would be if the suture was completely closed, or where the suture between two bones was closed if there were three elements involved. In the case of the juncture between the nasals and the frontal (Landmarks 24 and 25; Fig. 3 a, c), the anterior tip of the frontal was almost always unossified (Fig. 3 a), and the suture between the nasals unclosed; these landmarks were placed upon the medioposteriorextremities of the nasals.

In very small individuals, the median parietal horn P1 (Landmarks 68, 69, 70, Fig. 3 a, b, d) was not developed. This feature arises later in ontogeny, which implies that it is potentially present at birth and thus that its locus can be inferred. The medial extremities of the paired parietal horns (P2 - Landmarks 66, 67, Fig.3 a, b, d) were used to determine a line corresponding to the posterior margin of the parietal in very young individuals, and the median point on this presumptive margin was used to place the locus of the apex of P1; Landmark 70 was placed here, flanked immediately by Landmarks 68 and 69 (Fig. 3 a, d).

S.3. Measurement Error Analysis

The Procrustes fit of the pooled data set, and all subsequent analysis of measurement error,were carried out using MorphoJ 1.05a (Klingenberg 2011). The cumulative squared Procrustes distance distribution of the entire sample, plotted against the expected cumulative distribution under the assumption of multivariate normality (Fig. S-1), was examined for indication of outliers. Although the observed cumulative squared Procrustes distance distribution does not conform closely to the expectations of normality (Fig. S-1), it does not display long tails, indicating that there are no replicates in the sample that display strong idiosyncratic variance (that is, possibly large measurement errors). The entire sample was thus retained for further analysis.

In order to quantify measurement error of the entire replicated sample, we subjected the Procrustes-fitted data to a repeated-measures Procrustes ANOVA, using replicate as the error effect, incorporating individual specimen identity and side (right or left) as main effects and an interaction term for individual identity and side (Klingenberg and McIntyre 1998). This model assumes that variance about each landmark is isotropic (Goodall 1991), an assumption which is unlikely to be sustained for data from real organisms (Klingenberg and Monteiro 2005). However, we areconcerned with the relative magnitudes of effects rather than in drawing statistical inferences from such an analysis, and a Procrustes ANOVA is sufficient for this purpose. Relative magnitudes of effects are estimated by calculating the ratios of the Procrustes mean squares of, individually, the individual identity, side, and the individual identity*side interaction effects, to the Procrustes mean square of the replicate effect, which is used as the error effect in this experimental design.

Individual identity, side, and the interaction effect between individual and side are all highly significant (Table S-1). The Procrustes mean square error of the replicate effect was 26.53% of the Procrustes mean square error of the individual identity*side effect, sufficiently small that the use of this estimate of asymmetry can be justified in the evaluation of hypotheses concerning modularity in the dermatocranium of Phrynosoma hernandesi, although a smaller percentage would be desirable. The significance of the interaction effect (p < 0.0001; Table S-1) mitigates this concern. The Procrustes mean square error of the replicate effect was 8.04% of the Procrustes mean square error for the side effect, indicating that measurement error does not contribute a great deal to the significant effect of side (p < 0.0001; Table s-1). The Procrustes mean square error of the replicate effect was 1.56% of the Procrustes mean square error for the individual effect. This strongly suggests that measurement error does not contribute sufficiently to the total variance of the sample to be a concern for further analysis of shape in the sample of P. hernandesi.

Table S-1. Statistics for main effects of repeated-measures ANOVA of replicated Phrynosoma hernandesi landmark data.

Literature Cited

Goodall, C. (1991)Procrustes methods in the statistical analysis of shape. Journal of the Royal Statistical Society. Series B,53,285-339.

Klingenberg, C.P. (2011) MorphoJ: an integrated software package for geometric morphometrics. Molecular Ecology Resources, 11, 353-357. doi: 10.1111/j.1755-0998.2010.02924.x

Klingenberg,C.P., McIntyre, G.S. (1998) Geometric morphometrics of developmental instability: Analyzing patterns of fluctuating asymmetry with Procrustes methods. Evolution, 52, 1363-1375.

Klingenberg, C.P, Barluenga, M., Meyer, A. (2002) Shape analysis of symmetric structures: Quantifying variation among individuals and asymmetry. Evolution,56, 1909-1920.

Klingenberg, C.P.,Monteiro, L.R. (2005)Distances and directions in multidimensional shape spaces: implications for morphometricapplications. Systematic Biology, 54,678-688.

<Bookmark(9)>

Fig. S-1. Cumulative squared Procrustes distance distribution for the pooled Phrynosoma hernandesi sample (red line) plotted over distribution expected under a multivariate normal distribution fitted to the data (blue line).

Fig. S-2. Plot of female (red) and male (blue) ln(centroid size) against ln(body length) for entire sample of Phrynosoma hernandesi used in the geometric morphometric analysis.

Table S-2. Analysis of covariance for sexual dimorphism in allometric relationship between ln(centroid size) and ln(body length) in the sample of Phrynosoma hernandesi used in the geometric morphometric analysis. Adjusted R2 = 0.9784.

S.4. Retention of Principal Components

A broken-stick distribution for fifty-two segments was generated and used to determine which principal components were to be retained for further analysis (Legendre and Legendre 1998). In addition, a histogram of the percent of the total variance explained for each principal component was inspected for discontinuities in the rate of decrease from greatest to least. Any discontinuity found in the histogram were taken to indicate that the principal components plotted to the left were to be disregarded.

Literature Cited

Legendre, P., Legendre, L. (1998)Numerical Ecology. Second English Edition. Amsterdam:Elsevier Science B.V..

1