Read an Excerpt
CHAPTER 1
Intraspecific and Interspecific Variability in Pedal Phalangeal and Digital Dimensions and Proportions in Non-avian Dinosaurs, Birds, and Crocodylians
I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind.
William Thomson (Lord Kelvin), Popular Lectures 1: 73, 1883
REGRETTABLY, NON-AVIAN DINOSAURS ARE EXTINCT. WE can no longer watch such animals in the act of making footprints, and so it is no longer possible to be certain that footprints we think might have been made by dinosaurs of the same species really were. We must take an indirect approach. Knowing what the intact sole of the foot of dinosaurs was like would help here; dinosaur "mummies" and other remarkably preserved specimens sometimes provide such information (e.g., Brown 1916; Cuesta et al. 2015; Wang X. et al. 2017), but there aren't enough of them.
Non-avian dinosaur species are usually named on the basis of skeletal material. We can look at foot skeletons of non-avian dinosaurs and other archosaurs. If pedal intraspecific size and shape variability turned out to be comparable among non-avian dinosaurs, birds, and crocodylians, this would bolster our confidence that the same would have been true for their intact feet and footprints — although we must always keep in mind that ascertaining that skeletons of extinct animals like dinosaurs were members of the same species is itself a challenging problem.
The present chapter examines size and shape variability of foot skeletons of extant crocodylian and bird species that are unambiguously members of the same species, as well as variability of foot skeletons of what are thought to be members of the same species of several kinds of extinct birds and non-avian dinosaurs. The question is, essentially, how different in size and shape of foot can skeletons be and still belong to the same species? And is the amount of variability among species within a group greater than that within a species? We will consider several different measures of variability, themselves of highly variable sophistication, for comparing intraspecific and interspecific variability in pedal proportions.
MATERIALS AND METHODS
I collected measurements of the phalangeal skeletons of numerous specimens of crocodylians, birds, and non-avian dinosaurs from collections in North America, Europe, Australia, and New Zealand (table A1.1; figs. 1.1–1.12); these measurements are used in this and later chapters. As in earlier studies (Farlow and Lockley 1993; Farlow and Chapman 1997; Farlow 2001; Smith and Farlow 2003; Farlow et al. 2012, 2013, 2014), I measured non-ungual phalangeal lengths (fig. 1.1) along the medial and lateral sides of the toe bone, parallel to the long axis of the bone, from about the dorsoventral midpoint (or the most concave point, if different) of the concave proximal cotyla (terminology of Baumel et al. 1979) to the dorsoventral midpoint (or most distal point, if different) of the convex distal articular trochlea of the bone. If it was not possible to measure phalanx lengths on both the medial and lateral sides of the bone, whichever length could be measured was used instead of the mean length. Non-ungual proximal widths were taken as the greatest transverse distance across the proximal face of the phalanx, perpendicular to the long axis of the bone. Non-ungual phalanx distal widths were taken as the greatest transverse distance across the distal trochlea, perpendicular to the long axis of the bone. Ungual lengths were measured in a straight line, again on both medial and lateral sides (where possible) of the bone, from the dorsoventral midpoint (or the most distal point) of the concave proximal cotyla to the distal tip of the ungual, located near the midline of the bone.
There were certain unavoidable problems in the database. Because there are numerous phalanges in an archosaurian foot, it is common (particularly for non-avian dinosaurs) for fossil specimens to have incomplete or poorly preserved feet. This is less of a problem for foot skeletons of extant or subfossil birds. For specimens of extant crocodylians, however, I only measured lengths of phalanges that remained bound together in articulation by dried soft tissues; I was less confident of my ability to correctly identify loose phalanges of crocodylians in specimen boxes than of my ability to make such identifications for bird and dinosaur specimens. Another problem for most crocodylian feet in osteological collections is that the horny claw does not easily separate from the ungual, making accurate measurements of unguals impossible, and so crocodylian unguals were not used in the present study.
For many dinosaur specimens it was not possible to measure phalangeal lengths on both the medial and lateral sides of the bone, either because the foot was still embedded in matrix or because of poor preservation. This could potentially introduce "bogus" variability within a sample for a particular taxon, if phalanx lengths of some specimens used in an analysis were means of both medial and lateral lengths, while lengths of other specimens were only medial or lateral lengths. This problem would likely be greater for phalanges of digits II and IV (the peripheral toes of functionally tridactyl dinosaurs), and less for the more symmetrical digit III phalanges.
If a given specimen had complete sets of phalanges for both feet, I commonly used measurements for only one foot. If, however, both feet were incomplete, I often combined measurements from both feet to create data for a composite, "synthetic" foot of that individual. Where possible, I took the average of medial and lateral lengths of each phalanx as its phalangeal length, but used just the medial or the lateral length if this was the only measurement of phalangeal length that could be made; I call these averages "blended" phalangeal measurements. As already noted, employing data for feet in which some, but not all, phalangeal lengths were means of medial and lateral lengths could introduce artifacts in calculations of intraspecific or other kinds of within-group variability in phalangeal size and shape, a possibility that will be considered in this chapter for the dinosaur species with the largest sample of feet, Iguanodon bernissartensis. In some fossil specimens the tips of unguals had broken off; if I thought I could accurately estimate ungual length within a few millimeters, or if I thought that the measured ungual length was a reasonable approximation of its true length, I went ahead and used the ungual length in my analyses. Because of this postmortem damage to unguals, however, variability in ungual lengths may be artificially higher in comparison with variability in lengths of non-ungual phalanges.
Most measurements of phalanx or digit lengths and widths were made to the nearest millimeter. For small phalanges of young crocodylians, however, lengths were measured to the nearest 0.5 or even 0.1 mm.
In some analyses, data just for the larger pedal phalanges (II1–3, III1–4, IV1, IV2, IV5) were analyzed, with phalanges II1, II2, III1–III3, IV1, and IV2 (here designated the "big seven" phalanges) given particular attention; phalanges IV3 and IV4 were not so treated, but their measurements were obviously used in analyses of overall digit lengths. Overall, cumulative digit length was calculated as the summed lengths of the phalanges of each digit. Digit lengths were examined both with and without the unguals.
Apart from issues of completeness and quality of preservation of foot skeletons, the biggest issue in the present study was the sample size of feet for each species, which was limited by what was available in museum collections or (in the case of alligators) what I was able to collect myself. Consequently, my most detailed analyses were for alligators (Alligator mississippiensis), emus (Dromaius novaehollandiae), ostriches (Struthio camelus), rheas (Rhea americana and Pterocnemia [assigned to Rhea by Sibley and Monroe (1990), but not Dickinson (2003) or Clements (2007)] pennata), kiwi (Apteryx australis), moa (particularly Anomalopteryx didiformis, Pachyornis elephantopus, and Dinornis robustus), and the non-avian dinosaurs Coelophysis bauri, Tyrannosaurus rex (provisionally including Nanotyrannus lancensis), and Iguanodon bernissartensis, forms represented by a reasonable (but not huge) number of specimens.
Some of these taxa deserve further mention. Early in my work I recognized that moa (Dinornithiformes) could be of particular interest for this project. The nine generally recognized species constitute an adaptive radiation of big flightless birds that lived in prehistoric New Zealand (Cooper et al. 1992, 2001; Worthy and Holdaway 2002; Bunce et al. 2003, 2009; Huynen et al. 2003; Baker et al. 2005; Worthy 2005; Allentoft and Rawlence 2012; Worthy and Scofield 2012; Brassey et al. 2013; Olson and Turvey 2013; Bishop 2015; Angst and Buffetaut 2017; Mayr 2017), an area roughly comparable in size, interestingly enough, to the collected depositional basins of the Newark Supergroup of eastern North America during the Early Jurassic (home to the historically important footprint fauna of Edward Hitchcock). As subfossil birds, moa have only recently become extinct, and so were sure to be represented by larger sample sizes of foot skeletons than non-avian dinosaurs. In addition, the moa would have been as close in size to non-avian dinosaurs as any other avian clade. For all these reasons they have been given particular emphasis in this study.
In collecting measurements of moa foot skeletons in museum collections in New Zealand, Britain, and the United States, I generally used the museum specimen identifications associated with those specimens, supplemented by discussions with moa specialists (particularly Trevor H. Worthy). Allentoft et al. (2010) compared species and specimen identifications based on morphology with identifications based on ancient mitochondrial DNA, and found fairly good agreement between molecular and morphological species identifications of moa in their sample; discrepancies involved Euryapteryx curtus vs. Pachyornis elephantopus, Emeus crassus vs. P. elephantopus, E. curtus vs. E. crassus, or juvenile birds identified as emeids vs. Dinornis robustus. One specimen in my sample (Canterbury Museum CM AV 8622) that was identified on the basis of morphology as being E. curtus was assigned by Allentoft et al. (2010) on the basis of DNA to P. elephantopus. In analyses of intraspecific variability in pedal proportions of P. elephantopus reported in this chapter, results will be reported both excluding and including CM AV 8622.
As the potential maker of tridactyl dinosaur footprints represented by the largest sample size, Iguanodon bernissartensis figures prominently in this chapter. Consequently some comment about the nature and taxonomy of these specimens is in order.
My sample of I. bernissartensis and related forms comprises skeletons (table 1.1) in the collections of the Institut Royal des Sciences Naturelles de Belgique (IRSNB) and London's Natural History Museum (NHMUK; formerly BMNH) (Norman 1980, 1986, 2010, 2012, 2014; Paul 2007, 2008b, 2010; Verdú et al. 2017). Norman (1987a: 247) speculated that the Iguanodon assemblage at Bernissart accumulated over "an appreciable period of time (?10-100 years)." This is a short enough time that it is unlikely that there would have been any evolutionary change in pedal shape in the species present, and so I will treat the Bernissart Iguanodon specimens as an essentially contemporaneous assemblage.
The Bernissart Iguanodon specimens fall into three categories, as recognized by Norman (1980, 1986): three "sub-adult" individuals of I. bernissartensis, several adult individuals of the same species (but obviously we cannot be certain about the sexual maturity of these animals in the way that is possible for, say, extant American alligators — another species that figures prominently in this book), and at least two specimens of a second species. Although the two putative species have been interpreted as sexual dimorphs of a single species (Carpenter 1999), Norman (1986) provisionally concluded that they were probably distinct species.
At the time I examined the specimen, NHMUK R1829 was labeled in the Natural History Museum (London)'s collection as I. mantelli. Norman (1986) regarded this taxon as an objective junior synonym of I. anglicus, which in turn is a dubious taxon based on teeth (Norman 1986; Charig and Chapman 1998, 2000). I. mantelli was therefore not an appropriate name for the small, gracile form (IRSNB R 57 [old 1551]) of Iguanodon at Bernissart, even though it had previously been identified under this name. Norman (1986) concluded that IRSNB R 57 was appropriately assigned to I. atherfieldensis. Although he did not assign NHMUK R1829 to I. atherfieldensis in his 1986 monograph, Norman (2012, 2014) indicated that atherfieldensis would be the appropriate species name for the specimen.
Paul (2007) went even further than that, first assigning IRSNB R 57, the Bernissart specimen of I. atherfieldensis, to a new genus, such that the full name of the species would be Mantellisaurus atherfieldensis, and subsequently (Paul 2008b, 2010) assigning this specimen to a new genus and species, Dollodon bampingi. Norman (2010, 2012, 2014) accepted the validity of Mantellisaurus as the generic name for specimens previously assigned to Iguanodon atherfieldensis, an interpretation that will be followed here (see Lomax and Tamura 2014 for a summary of current thinking about the status of these and other Early Cretaceous iguanodonts from Europe and Britain).
(Continues…)
Excerpted from "Noah's Ravens"
by .
Copyright © 2018 James O. Farlow.
Excerpted by permission of Indiana University Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
