Morphometrics: Applications in Biology and Paleontology Edited by Ashraf Elewa Springer, Berlin, 2004, 263 p.  ISBN: 3-540-21429-1. $129.00

It is not a good sign when methodologists call your work "interesting." That rarely means that you’ve stumbled across a creative solution to a longstanding problem. More often, it means you’ve stumbled on a new problem, one so novel that it is not clear if your solution (or any other) is valid. A methodologist’s notion of "interesting" conveys something ominous to an empiricist, reminiscent of the curse "may you live in interesting times." Morphometrics was interesting, in that ominous sense, from about 1985-1996, the decade bridged by the publication of the Red Book and White Book. The papers published during that decade were not necessarily interesting (in the more positive sense of the term). Instead, they were frequently just methodological exercises, useful for demonstrating and comparing methods but not for advancing biological theory. That is not to say that they were therefore unimportant; on the contrary, they were vital. We desperately needed sound case-studies, on any subject, however tangential they might be to our theoretical interests. At a time when methodologists are writing to each other instead of to us, and the field is changing so rapidly changes that papers become obsolete before they are published, and last year’s methodological dogma becomes this year’s heresy, we cannot hope to keep up without papers demonstrating the state-of-the-methodological-art.

That need was satisfied by three volumes that compiled proceedings of workshops held by preeminent methodologists . Those highly eclectic volumes comprised a mélange of papers on various methodological and biological topics. Only the first had any sort of unifying biological theme (that of bringing morphometricians and systematists together), but their lack of any thematic coherence was immaterial to their value. Their theme was methodological, and their value lay precisely in explaining and demonstrating methods. Now that we no longer live in such interesting times, we can ask for more; we can ask that the volumes cohere, and that the papers ask interesting biological questions. That is not to say methods have finally become dull; we still lack enough methods for capturing morphological attributes not captured by configurations of landmarks, and we still need methods for more specialized purposes, such as modeling the evolution of shape. But subject matter can finally dominate, and several volumes have focused more sharply on subject matter, even when the theme is methodological. For example, two recent books have focused on applications of morphometrics in systematics . Although both are oriented to methods, neither is about methods of shape analysis. Rather, both present varied viewpoints on the relationship between morphometrics and systematics. Other books concern biological concepts and theory, presenting morphometric case-studies in those contexts. In these books, such as the volume on phenotypic integration , morphometrics is just the means to answering a biological question.

In Morphometrics: Applications in Biology and Paleontology, Elewa assembles an eclectic assemblage of papers covering a diverse array of topics, making this book structurally similar to the workshop proceedings. Five of the chapters are primarily methodological in focus, and rather specialized; the remaining eleven are empirical, and cover a range of subjects (although taxonomy predominates). The empirical chapters generally offer a thoughtful discussion of the biological context for the studies, and do more than just demonstrate methods. In one respect, this book differs strikingly from the workshop proceedings—this book lacks an authoritative summary of the state-of-the-art, and only one chapter breaks new ground in shape analysis. That this book lacks a survey of methods is not necessarily a serious omission because those are readily available elsewhere. However, it would have been useful to include one because the empirical studies tend to be methodologically uniform. As a result, some interesting methods are neither discussed as part of a survey of methods, nor demonstrated in the empirical chapters. Nearly all the empirical studies use the standard ordination methods: canonical variates analysis (CVA), with the associated one-way multivariate analysis of variance (MANOVA), and/or principal components analysis (PCA). Some also regress shape or one or more principal components on size. These are widely used and useful methods. And they might suffice for many studies, especially for the taxonomic ones that figure so prominently in this book. But they do not always suffice, especially for historical or causal investigations of morphology. Evolutionary studies need tree-based methods to reconstruct the evolution of form and tree-based methods are also needed for comparative studies that take phylogeny into account when testing hypotheses statistically. Examples of both kinds of studies would have added to the value of this book for evolutionary morphologists. Additionally, it would have been useful to include studies relating shape to more complex factors than size. One chapter conducts such an analysis, correlating morphology and ecological and behavioral variables (Ch. 13, Korner-Nievergelt). But a case study using partial least squares (PLS) would have made a useful contribution to the book (for an introduction to PLS see Rohlf and Corti 2000; for an interesting application, that also takes phylogeny into account in the statistical analysis, see Rüber and Adams 2001). Unfortunately, there is a fairly long list of methods not demonstrated in this book, among them: methods for quantifying disparity and analyzing its temporal and spatial structure, those required to test hypotheses about tempo and mode of evolution, and those need for analyses of developmental regulation (canalization and fluctuating asymmetry, as well as the relationship between them).

There is one other reason why I would hesitate to recommend this book as a source of information about landmark-based geometric morphometrics. That is because several chapters separate the uniform (=affine) and nonuniform (=nonaffine) components of shape, without also treating them in combination. Readers familiar with the recent history of morphometrics may recognize that this tactic traces back to the days when those two components were seen as mathematically incommensurate. Once the Procrustes distance became generally accepted as the metric for shape analysis, there was no longer a mathematical justification for separating those two components. In the absence of a mathematical justification for that separation, a good biological one is necessary. Without that, it arguably makes no more sense to treat the uniform and nonuniform components separately than to decompose the nonaffine part into its partial warps and analyze all of them separately.

The book is also fairly uniform in its taxonomic scope. All but two chapters, both primarily methodological, are about vertebrates or arthropods. That might not seem particularly important because methods are methods and it should not matter whether all are applied to a rat skull, or each to different phyla. But the particular methods most suited to capturing important aspects of form do depend, at least in part, on the forms being captured. This becomes obvious in the chapter on the measurement of complex three-dimensional branching forms (Ch. 7, Kanndorp and Leiva). That chapter (on corals and sponges) is illuminating, and gives a sense of the diversity of measurement theories that might appear in a book that plants, and a variety of modular organisms (both plant and animal) as well as spiral-shaped animals. Nearly all the chapters in this book use landmark-based morphometrics (even the two that analyze traditional morphometric variables measure them between landmarks: Ch. 3 García-Rodriguez, Agüero, Pérez-Enriquez and MacLeod; Ch. 9, Agüero and García-Rodriguez). The one other exception is a Fourier analysis of passerine (bird) foot soles (Ch. 13, Korner-Neivergelt). Although landmark-based methods have some notable benefits, they are not universally suitable.

Despite being methodologically and taxonomically too uniform, the book is topically diverse. For that reason, no dialogue emerges by comparing views on a common biological theme. Each chapter stands on its own and the book is the sum of its parts. However, topical diversity does have an advantage in that it increases the chance of finding an interesting paper, especially when most of the papers are on highly specialized topics.

The five methodological chapters are typically interesting even if most of them are also quite specialized. As mentioned above, one offers a method for quantifying features of complex, three-dimensional branching structures, using an extension of the medial axis (Ch. 7, Kanndorp and Leiva). Another provides a new method for identifying unknown specimens using maximum-likelihood (Ch. 14, Polly and Head). A third, which is more didactic than methodological, discusses semi-landmarks (i.e., anatomical points that are not homologous as individual loci, such as points along an outline), explaining how to acquire and analyze them using freely available software (Ch. 6, Sheets, Kim and Mitchell). Another deals with preservational rather than morphometric methods, comparing crab carapaces preserved by alcohol and freezing relative to fresh specimens (Ch. 4, Rufino, Abelló and Yule). The results are reassuring in that the difference is slight, albeit statistically significant. The final methodological chapter argues for measurement of process rather than outcome, that is, measuring morphogenesis and growth rather than shape (Ch. 5, Hammer). The method for doing so is only sketched out, and its starting point is a phenomenological model (Huxley’s Power Law) but the approach may prove valuable when further developed. These methodological chapters are worth reading if you have any occasion to need the method, or a general interest in methods of shape analysis.

Four of the empirical chapters concern morphology, analyzed for its own sake rather than as an indicator of something else (e.g., taxonomic status of the sampled populations). Of these four, two are descriptive. One, a structuralist analysis of the diversity of avian skull architectures, samples 25 orders of modern birds, and also analyzes these in conjunction with extinct therapods (Ch. 12, Marugán-Lobon and Buscalioni). As well as describing the major dimensions of skull architecture, the analysis finds that region of morphospace occupied by extinct therapods is emptied by their extinction. It is unfortunate that the graphics are too cramped to be fully readable because this is an interesting study. The other descriptive study depicts dimensions of variation in cephalofoil shape across four species of hammerhead sharks (Ch. 8, Cavalcanti). Adaptive explanations for the cephalofoil are discussed as well as the hypothesis of allometric scaling. There are statistical analyses of allometry, but their designs are sometimes problematic or ambiguous so the results are as well. In one case, the significance of the correlation between size and shape is tested for the pooled data (i.e., pooled across ontogenetic stages and species), so it is not clear whether the allometry thereby detected is ontogenetic, static or evolution (or a combination of them all). And the analysis is based on a regression of the first uniform and first nonuniform components, rather than on a regression of shape on size. The ontogenetic analyses are ambiguous because it is not clear whether the dependent variable, a relative warp (= a PC) is extracted from the pooled data, or from each ontogenetic series analyzed separately; regardless, a multivariate regression of shape on size would be preferable to either. One fascinating chapter takes a functionalist approach to morphology, looking at how well passerine foot sole morphology predicts substrate and locomotory behavior (Ch. 13, Korner-Neivergelt). The results are intuitively appealing: foot sole morphology predicts the substrate variables "needles" "trunk" and "vertical," and the behaviors related to bipedal locomotion (other than "upside down" and "below"). The final morphological chapter focuses on the static allometry of mandibles of adult hominids Ch. 16, Bastir and Rosas). Three intriguing hypotheses are tested (1) that the three samples (humans, Neanderthals, and a sample from Atapuerca, Sima de los Huesos) share a common spatial pattern of static allometry; (2) that smaller individuals from Atapuerca-SH resemble the primitive condition whereas larger ones resemble the derived Neanderthal shape; and (3) that size has a larger impact on the shape of the supra-alveolar than infra-alveolar region of the mandible (which would make sense in functional terms). The first hypothesis is not supported but the other two are. Unfortunately, the table presenting the results for the first hypothesis errs in labeling the difference between humans and the other samples as not significant; the difference between the Atapeurca-SH and Neandethal samples, which is estimated at 92.5° would also likely become significant when samples sizes increase above eight for the Atapuercan, and 17 for the Neanderthals. Being a morphologist with a taste for both structuralist and functionalist analyses, and also an interest in allometry and mammalian mandibles, I would regret having not read the chapters on avian skull architecture, ecomorphology of passerine foot soles, and hominid mandibular allometry.

The remaining seven chapters use morphology to draw inferences about biogeography or taxonomy. The biogeographic analysis, of sauropod femurs, concludes that the morphological differences between Laurasian and Gondwanan Titanosauria indicate that they differentiated prior to the break up of Pangea, making dispersal (and intercontinental bridges) unnecessary to explain the presence of Laurasian Titanosauria (Ch. 11, Canudo and Cuenca-Bescós). The six taxonomic chapters include three pilot studies based on very small samples that explore the value of the shape data for taxonomic studies (Ch. 2, Elewa; on polymorphism of ostracodes from Egypt and Spain; Ch. 15, Pavlinov, on the anteomolar row configuration of brown-toothed shrews of the genus Sorex, and Ch. 10, Rodrigues and dos Santos on sauropod footprints). Aside from small sample sizes, a common feature of those chapters is the difficulty of quantifying the morphologies. The three based on larger samples include a comparison between populations of the blue spiny lobster Panulirus inflatus from the Pacific coast of Mexico, using traditional morphometric data (Ch. 3 García-Rodriguez, Agüero, Pérez-Enriquez and MacLeod). The low level of differentiation found suggests plastic responses to environmental conditions. Agüero and García-Rodriguez (Ch. 9) also use traditional morphometrics data, and compare two populations of sardines (Sardinops sagax) that are also from the Pacific coast of northern Mexico. When the two sexes are analyzed separately, significant differences were found between localities, and the misclassification rates are moderate to low (ranging from 7 to 19% for females, and 18 to 22% for males). Havarti (Ch. 17) uses 3-D morphometrics of the temporal bone to compare nine populations of modern humans, Neatherthals, and a Middle Pleiostocene specimen (Kabwe). The major results are the confirmation that Neantherthals are separated from all modern human populations, that Upper Paleolithic Europeans are very similar to modern humans. Not being a taxonomist, I cannot comment on the sampling designs, although the three that examine reasonably large samples seem to draw well-supported conclusions. Although I did not find these interesting, that is more a reflection of my own interests than the merits of the chapters.

Both the strength and weakness of Morphometrics: Applications in Biology and Paleontology lie is its emphasis on the biological significance of the chapters. The papers are not just methodological exercises, which is certainly a strength of the book. But it becomes a weakness when the topics of the chapters to not match the interests of the reader. And it is also a weakness when the strength of a book that is really about methods lies in the topics of its chapters. As a survey of morphometric methods, this book falls short for three major reasons. First, not every morphology can be described by landmark-based methods; while undeniably useful, and perhaps preferable to other approaches to measurement when applicable, there are morphologies that cannot be described by landmarks, and these are severely underrepresented in the book. Second, the same few methods are repeatedly applied and this underrepresents the variety of available methods. Having an unnecessarily limited morphometric tool-box creates the same kind of problem for biologists that having a nearly empty tool-box does for a carpenter—if you only have a hammer, everything looks like a nail. The tool-kit surveyed in this book is too sparse to answer many of the questions motivating paleontological studies (including those that require tree-based methods, or quantifying disparity, or testing hypotheses too complex for a one-way MANOVA). Sophisticated readers may recognize that many other methods exist and could adapt them to the distinctive features of geometric shape data. However, morphometrically naïve readers might not realize that the limitations are those of the book, not the methods. Third, not every paper in this book is an excellent model of a well-crafted study; some are slightly problematic, others more so. In some cases, papers are problematic only because they make fairly obvious errors (such as misreading significant differences for non-significant ones). In other cases, the flaws lie in either the sampling design or data analysis, and these can be confusing to interpret. Readers without the expertise to recognize those errors and flaws may be led astray, whereas those having more expertise may be too sophisticated to find more than two or three chapters interesting.

I should note that both the Preface (by Rohlf) and the Introduction indicate that the audience of the intended audience for this book is systematists, not the much broader one implied by the title. A major aim of the book, as articulated by Elewa, is to provide answers to questions asked by systematists. A more specific one is to bring paleontologists and biologists together, to start solving taxonomic problems in a compatible manner. In light of these aims, it is not surprising that the book focuses so much on taxonomy, and taxonomists may well find it both useful and interesting. But I cannot recommend it to morphologists or evolutionary biologists or to the morphometrically naïve.