BSA of Norops lineatopus

Geometric Morphometric Analysis of the Shoulder of Jamaican Anoles

garmani mating trivers IIxBirds are lovely animals. Our avian friends swoop through the air, defecate on field equipment, and consume lizards. What’s not to like?! Well, their shoulder region, for example. Lost interclavicle, reverted muscle pathways, and so many other anatomical adaptations that appear crucial for the modern avian life style, but that are hard to explain in a gradual-evolutionary context. Reconstructing the structural evolution of the avian shoulder remains a challenging task to students of biomechanics and kinematics. When I left my European homestead to enter the Canadian realm of biological sciences, I was hoping to solve the evolutionary mystery of the avian shoulder, at least in part. Alas, the discovery of anoles sent me on a much more convoluted journey.

Here is the first tale that resulted from that endeavour (Tinius & Russell 2014).

Anole Shoulders

When I started my Ph.D. project at the University of Calgary my supervisor Anthony Russell had a few theoretical objections against my idea of generating a holistic understanding of the evolution of the avian shoulder. It did not take him long to convince me that the shoulder of theropods is too complex to grasp within the constraints of one thesis. Instead, he presented me with a cabinet full of pickled anoles. More than four hundred specimens, all collected in 1978 and 1979 near Mandeville in Jamaica. These anoles represent five of the seven species of “native” Jamaican anoles, excluding the very rare Norops reconditus.

[At this point I should explain why I continue the genus name Norops for the Jamaican radiation, despite Castañeda and de Queiroz (2013) convincingly arguing for an alternative classificatory system within Anolis. Guyer & Savage (1986) advocated the name Norops as genus name for the beta-anoles, and its content has been relatively stable ever since. As long as the Norops taxon has not been refuted in a taxonomic discussion, a return to the name Anolis is not warranted by the current rules of taxonomy. Maybe I am being stereotypically German, but to me it makes sense to retain the name Norops.]

CT-Images and Geometric Morphometrics

Anoles have various advantages over birds. They are terrestrial quadrupeds, featuring a relatively complete pectoral girdle. Since the external morphology and morphometry appear to correlate with their ecological preferences, the anoles of the Greater Caribbean islands presented themselves as ideal study subjects for my study of diapsid shoulders. As has been shown in many studies already, various aspects of the differential morphology of anoles is directly correlated with their lifestyle. The questions for us were: does the morphology of the shoulder girdle differ between Jamaican anoles? And does external similarity indicate skeletal similarity? With any luck we could establish some ground rules for the structure and composition of the diapsid shoulder region, and then move on towards birds.

BSA of Norops lineatopus

CT reconstruction of the breast-shoulder apparatus of Norops lineatopus.

From the cabinet of anoles available to me, I selected all specimens that retained an intact shoulder region, and dragged them through a micro-CT-scanner, thus obtaining a 3D image of the density contrasts in the shoulder region of these anoles. I generated surface renditions of the humeri, scapulocoracoids, clavicles, interclavicle, presternum, ribcage, and vertebral column, representing the complete breast-shoulder apparatus of every anole specimen. I then rotated the virtual scapulocoracoid in the coracosternal groove until its relative position and orientation was comparable between all specimens. Finally, I distributed forty-three 3D-landmarks on the skeletal elements of the pectoral girdle, and subjected them to geometric morphometric analysis.

Interclavicle, presternum and scapulocoracoid of Norops lineatopus.

Interclavicle, presternum and scapulocoracoid of Norops lineatopus.

Obviously, the elements of the pectoral girdle are mobile with respect to each other, which makes a geometric morphometric analysis rather challenging. In order to eliminate any variability in the data that was caused by element mobility, we analysed the scapulocoracoid and the clavicle in isolation of each other and of the interclavicle-presternum-moiety. The statistical analysis via principal component analysis (PCA) and canonical variate analysis (CVA) may appear very traditional, but since this was only an explorative study, we did not want to venture too far from the beaten stats track.

Morphometric Differentiation among Jamaican Anoles

The outcome was positively surprising. Considering that the Jamaican anoles represent only a small monophyletic radiation, the elements of the shoulder girdle show a remarkable degree of morphological variability between species.

The twig-anole N. valencienni, in particular, appears to differ greatly from all other species examined. 1) Contrary to observations in other Jamaican anoles, the third sternal rib of N. valencienni articulates with the mesosternum, instead of the presternum, and the presternum appears anteroposteriorly elongated. 2) The scapulocoracoid of N. valencienni is anteroposteriorly shorter, and dorsoventrally taller than in any other species examined.

Canonical Variate Analysis of the scapulocoracoid

Canonical Variate Analysis of the scapulocoracoid of Jamaican anoles.

Twig-anoles have often been regarded as special. Peterson (1973) reported that twig-anoles are in many aspects of their locomotor style more similar to chameleons than to other anoles. Many of the characteristics that Peterson (1974) associated with twig anoles, like a potentially great excursion of the scapulocoracoid, or a tall scapula and suprascapula, could be corroborated in our current study.

The geometric shape of the scapulocoracoid also differentiates most of the other species of Jamaican Norops, but none of these differences really stands out in the statistical analysis. Furthermore, without other island radiations to compare to, any differences between these five anole species are not really indicative of ecomorphological adaptations. Many of the patterns observed in our data were already reported by Jane Peterson forty years ago, and we are still trying to catch up with her work …

Left clavicle of Jamaican anoles

Left clavicle of Jamaican anoles. Phylogenetic relationship and ecomorph association.

The shape of the clavicle is not well represented in our data, since this element only bears three landmarks. However, even in a qualitative comparison, it became apparent that the form of the clavicle is influenced by phylogenetic relationship as well as by ecological aspects. Authors like Sukhanov (1961) or Lécuru (1968) already attributed special powers of systematic characterization to the clavicle, but, again, this aspect has never been discussed in detail.


We were rather conservative in our statistical approach, and there are obvious methodological problems when landmarking 3D-structures that are a) not completely ossified, and b) mobile with regard to each other. Considering these difficulties, I think these results are rather promising. In both the description and in the geometric morphometric analysis we were able to show that the form of the pectoral girdle differs between the species of Jamaican anoles. In a future analysis we hope to show that morphological trends identified in this paper are consistent with variations between ecomorphs. We are currently expanding the data set with anoles from Hispaniola and Puerto Rico, and a preliminary data analysis has reinforced that hope.


Castañeda MDR, de Queiroz K. 2013. Phylogeny of the Dactyloa clade of Anolis lizards: new insights from combining morphological and mo

About Alex Tinius

As a Ph.D. student at the University of Calgary I am currently working on the morphology of the skeletal elements of the pectoral girdle of island anoles. My greater goal is to develop a holistic understanding of the functional morphology of the breast-shoulder apparatus of diapsids.